16683522 alcohol textbook 4th ed

THE ALCOHOLTEXTBOOK

4TH EDITION

THE ALCOHOLTEXTBOOK

4TH EDITIONA reference for the beverage, fuel and industrial alcohol industriesA reference for the beverage, fuel and industrial alcohol industries

Edited by KA Jacques, TP Lyons and DR Kelsall

Foreword iii

The Alcohol Textbook4th Edition

A reference for the beverage, fuel and industrial alcohol industries

K.A. Jacques, PhDT.P. Lyons, PhDD.R. Kelsall

iv T.P. Lyons

Nottingham University PressManor Farm, Main Street, ThrumptonNottingham, NG11 0AX, United Kingdom

NOTTINGHAM

Published by Nottingham University Press (2nd Edition) 1995Third edition published 1999Fourth edition published 2003© Alltech Inc 2003

All rights reserved. No part of this publicationmay be reproduced in any material form(including photocopying or storing in anymedium by electronic means and whether or nottransiently or incidentally to some other use ofthis publication) without the written permissionof the copyright holder except in accordance withthe provisions of the Copyright, Designs andPatents Act 1988. Applications for the copyrightholder’s written permission to reproduce any partof this publication should be addressed to the publishers.

ISBN 1-897676-13-1

Page layout and design by Nottingham University Press, NottinghamPrinted and bound by Bath Press, Bath, England

Foreword v

Contents

Foreword ixT. Pearse LyonsPresient, Alltech Inc., Nicholasville, Kentucky, USA

Ethanol industry today

1 Ethanol around the world: rapid growth in policies, technology and production 1T. Pearse LyonsAlltech Inc., Nicholasville, Kentucky, USA

Raw material handling and processing

2 Grain dry milling and cooking procedures: extracting sugars in preparation forfermentation 9Dave R. Kelsall and T. Pearse LyonsAlltech Inc., Nicholasville, Kentucky, USA

3 Enzymatic conversion of starch to fermentable sugars 23Ronan F. PowerNorth American Biosciences Center, Alltech Inc., Nicholasville, Kentucky, USA

4 Grain handling: a critical aspect of distillery operation 33David J. RadzanowskiAlltech Inc., Nicholasville, Kentucky, USA

Substrates for ethanol production

5 Lignocellulosics to ethanol: meeting ethanol demand in the future 41Charles A. AbbasArcher Daniels Midland, Decatur, Illinois, USA

6 Ethanol production from cassava 59Nguyen T.T. VinhThe Research Institute of Brewing, Hanoi, Vietnam

7 Whey alcohol - a viable outlet for whey? 65Jack O’SheaAlltech Inc., Dunboyne, County Meath, Ireland

8 Treatment and fermentation of molasses when making rum-type spirits 75Robert PiggotAlltech Inc., Nicholasville, Kentucky, USA

vi T.P. Lyons

Yeast and management of fermentation

9 Understanding yeast fundamentals 85Inge RussellInternational Centre for Brewing and Distilling, School of Life Sciences,Heriot-Watt University, Edinburgh, UK

10 Practical management of yeast: conversion of sugars to ethanol 121Dave R. Kelsall and T. Pearse LyonsAlltech Inc., Nicholasville, Kentucky, USA

11 Continuous fermentation in the fuel alcohol industry:How does the technology affect yeast? 135W. M. IngledewUniversity of Saskatchewan, Saskatoon, Saskatchewan, Canada

12 Understanding near infrared spectroscopy and its applications in the distillery 145Don Livermore1, Qian Wang2 and Richard S. Jackson2

1Hiram Walker & Sons Ltd., Walkerville, Ontario, Canada2Bruker Optics Inc., Billerica, Massachusetts, USA

13 Emerging biorefineries and biotechnological applications of nonconventionalyeast: now and in the future 171Charles A. AbbasArcher Daniels Midland, Decatur, Illinois, USA

Beverage alcohol production

14 Production of Scotch and Irish whiskies: their history and evolution 193T. Pearse LyonsAlltech Inc., Nicholasville, Kentucky, USA

15 Tequila production from agave: historical influences and contemporary processes 223Miguel Cedeño CruzTequila Herradura, S.A. de C.V. Ex-Hda San Jose del Refugio Amatitán, Jalisco, México

16 Production of heavy and light rums: fermentation and maturation 247Robert PiggotAlltech Inc., Nicholasville, Kentucky, USA

17 From pot stills to continuous stills: flavor modification by distillation 255Robert PiggotAlltech Inc., Nicholasville, Kentucky, USA

18 From liqueurs to ‘malternatives’: the art of flavoring and compounding alcohol 265Andy Head and Becky TimmonsNorth American Biosciences Center, Alltech Inc., Nicholasville, Kentucky, USA

Foreword vii

19 Production of American whiskies: bourbon, corn, rye and Tennessee 275Ron RalphRon Ralph & Associates Inc., Louisville, Kentucky, USA

Contamination and hygiene

20 Bacterial contamination and control in ethanol production 287N.V. NarendranathNorth American Biosciences Center, Alltech Inc., Nicholasville, Kentucky, USA

21 Managing the four Ts of cleaning and sanitizing: time, temperature, titration andturbulence 299Jim Larson and Joe PowerNorth American Biosciences Center, Alltech Inc., Nicholasville, Kentucky, USA

Recovery

22 Ethanol distillation: the fundamentals 319P. W. MadsonKATZEN International, Inc., Cincinnati, Ohio, USA

23 Development and operation of the molecular sieve: an industry standard 337R. L. Bibb SwainDelta-T Corporation, Williamsburg, Virginia, USA

Engineering ethanol fermentations

24 Water reuse in fuel alcohol plants: effect on fermentationIs a ‘zero discharge’ concept attainable? 343W. M. IngledewUniversity of Saskatchewan, Saskatoon, Saskatchewan, Canada

25 Understanding energy use and energy users in contemporary ethanol plants 355John MeredithRo-Tech, Inc., Louisville, Kentucky, USA

The dryhouse, co-products and the future

26 Dryhouse design: focusing on reliability and return on investment 363John MeredithRo-Tech, Inc., Louisville, Kentucky, USA

27 Ethanol production and the modern livestock feed industry: a relationshipcontinuing to grow 377Kate A. JacquesNorth American Biosciences Center, Alltech Inc., Nicholasville, Kentucky, USA

viii T.P. Lyons

28 Biorefineries: the versatile fermentation plants of the future 387Karl A. DawsonNorth American Biosciences Center, Alltech Inc., Nicholasville, Kentucky, USA

The Alcohol Alphabet 399A glossary of terms used in the ethanol-producing industries

Index 441

Foreword ix

Foreword

Alcohol production: a traditional process changing rapidly

T. PEARSE LYONS

President, Alltech Inc., Nicholasville, Kentucky, USA

In its simplest form, alcohol production is theprocess of preparing starch- or sugar-containingraw materials for fermentation by yeast, whichis currently the only microorganism used forconverting sugar into alcohol. The ethanol isthen concentrated and recovered in a processcalled distillation. Though essentially a simpleprocess, making it happen with maximumefficiency on a very large scale is a remarkablecombination of microbiology and engineering.

Over the past 25 years progress has been madein virtually all aspects of the process, but whatare the challenges facing alcohol in the future?Beginning with raw material processing, hereare the areas that we feel need to be addressed:

Raw material reception and processing

Despite many declarations to the contrary, it isvirtually impossible for a distillery to accuratelycalculate true yields. A sampling of grains iseither done in very cursory fashion or not at all.The distiller must see himself first as a processorof grain rather than a producer of alcohol. Hemust know accurately how much starch ispresent in his raw material and he must be ableto measure that immediately. NIR spectroscopyhas much to offer in terms of raw materialapplications; and this topic is detailed in thisvolume. However, if we control the process anduse the appropriate enzymes from solid statefermentation, such as Rhizozyme™ andAllzyme™ SSF, which are capable of releasingnot just bound starch but also hemicellulose andcellulose, we can and indeed some plants do,get 2.9+ gallons per bushel (116 gallons or 420

liters/tonne. In fact, we predict yields of 3.2gallons in the near future. For molasses, theadvent of detranases can ensure maximumyields. Every aspect of cooking and fermentationmust be controlled, however.

The cooking process

Much debate exists at the moment regarding thecooking process. It is a sad reflection perhapson our industry that since the pioneering PhDwork by Dr. John Murtagh in 1972 no furtherscientific thesis has been published on cooking.Should grain be finely ground? Should cookinginvolve high temperature α-amylase or shouldan entirely different type of enzyme be used?Often the enzyme selected is one developedbased on experience in the corn wet millingsector, which may be far from ideal for dry millalcohol production.

Our industry needs a new enzyme, one tailoredto our specific needs: converting grain tofermentable sugar. Raw materials contain notonly starch, but also hemicellulose (a polymerof 5-carbon sugars) and cellulose. The cookingprocess must be designed to loosen thesematerials such that they release bound starch andalso become available in turn for fermentation.

Yeast and fermentation

Despite all we know about it, yeast remainsunderrated and misunderstood. It is theworkhorse of the distillery, yet ‘economies’ onquality and quantity of yeast are made; and we

x T.P. Lyons

often force it to grow and operate underunsuitable conditions. As Professor MikeIngledew has pointed out many times, thefermentor is not a garbage can! We must makesure that conditions in fermentation are optimumfor yeast.

Like our need for appropriate enzymes, ethanolproduction also requires yeast with higherethanol and temperature tolerance if we are tokeep progressing to higher levels of alcohol inthe fermentor. Today, 17-20% abv is becomingstandard in our industry as we move to hightemperature ThermosaccTM-like yeast. Whenused in combination with enzymes capable ofextracting increasing amounts of sugar, such asthose produced by surface culture fermentation,yeast can bring us to a whole new level.

While 23% ethanol by volume in the fermentorwithout loss of performance is possible giventhe right yeast and enzyme combination, it canonly be done by careful management of the so-called stress factors. Factors which, if left

uncontrolled, will prevent the yeast fromperforming. Operating a fermentation at 23%ethanol is totally different than operating adistillery at 8% ethanol. Any minor change,whether in mycotoxin level, salt level or nutrientlevel can have a major impact on performance.

Microbial contamination

A distillery is never going to be a sterilefermentation and even those of us who have hadthe good fortune to operate in sterile fermentationconditions know how easily infection can takehold. Since the yeast is a relatively slow-growingmicroorganism compared to most of theinfectious microorganisms (Lactobacillus, etc.),it is critical that we have an arsenal ofantimicrobial agents and effective cleaningproducts and programs ready to protect thefermentor environment.

PRIMARY ALCOHOL

SUGAR

POT STILL CONTINUOUS STILL

Malt

Whisky

Scotch Whisky

Vodka

Tequila

Grain

Whisky

RumIrish

Whiskey

Grain

Alcohol

GinBourbon

Blended Whisky

MATURATION SPIRITS

WHITE SPIRITS

Fuel

Alcohol

Molasses

Sugarcane

Whey

Yeast

Enzymes

DEXTRINS

STARCHLIQUIFACTION

Agave

Sugarbeets

Corn

Wheat

Barley

Rye

Potatoes

Milo (sorghum)

Cellulose

CO-PRODUCTSCO2

DDGS

Wet cake

Solubles

BIOREFINERYValue-added

co-products

Foreword xi

Recovery and utilization of co-products

As we review our overall process, it is obviouslycritical that we maximize utilization of co-products. Assuming the feedstock is corn, themajor cost factors are the raw material, energy,enzymes, processing chemical cost (watertreatment, etc.) and labor. Much of these costscan be offset with 17 lbs of distillers grainscoming from each bushel of corn; but we mustmaximize the return on this essential product.Distillers grains with solubles (DDGS) arecurrently (July 2003) selling for around $80/tonand corn for $2.20 per bushel. At $0.04 per lb($80/ton), 17 lbs of DDGS gives us a credit of$0.68. This is set against the $2.20 bushel ofcorn or the $0.03-0.04 for enzymes. If on theother hand, new or novel products could bemade from distillers grains, (i.e. improved by-pass proteins for cattle, pre-hydrolyzed DDGSfor human foods), the financial impact could besubstantial. The future will see many newproducts built around DDGS. This is the subjectof a separate chapter in this volume.

As dry mills shift toward biorefining, insteadof simply grinding, corn, yield of ethanolbecomes only one part of the economicequation. Perhaps in the future, furtherprocessing of distillers grains will create productsthat make spent grains the single most importantraw material from a distillery. After all, they arerich and valuable protein sources, a commodityin very short supply across the world.

Education

Perhaps the biggest surprise, as one surveys theprogress that has been made in the last 20 yearsis how little progress has been made in the areaof education. Unlike the brewing industry, whichseems to have an abundance of professionals

with PhD and masters degrees, our industry hasvery few. Unlike the brewing industry, whichhas no less than five brewing schools aroundthe world, the annual Alltech Alcohol Schoolremains the only venue for industry-widetraining.

The Alltech Institute of Brewing and Distilling,established in 2000, set out to address thisshortfall by working in conjunction with theHeriot-Watt University in Scotland. Perhaps ourindustry should use the brewing industry as amodel. Scholarships should be funded andperhaps a bioscience center concept established.Bioscience centers bridge the gap betweenuniversity and industry by creating anenvironment where students are encouraged tocomplete higher degrees while doing industry-focused research.

Conclusions

The alcohol industry, therefore, is alive and well,and we have been given an opportunity toachieve something that many of us thoughtwould have been achieved by the mid-1980s.We have a superb oxygenate that will helpreduce global warming and at the same timeenable us to add value to our grains. The worldis not short of starch and sugar, the world is shortof protein. The fuel alcohol industry andbeverage alcohol industries therefore are in factgenerators of protein; it only remains for us tomake sure the proteins we generate are the mostadvantageous possible to man and beast.

The chapters in this book we hope will helpthe reader realize the complexity and at the sametime the simplicity of the process of convertingsugars to ethanol. If we make your search forinformation just a little easier, we will haveachieved our objective. If we can encouragemore research, then we will be well-rewarded.

Ethanol industry today

Ethanol around the world: rapid growth in policies, technology and production 1

Chapter 1

Ethanol around the world: rapid growth in policies, technology andproduction

T. Pearse Lyons


Fuel ethanol industry history: how didtoday’s ethanol ‘boom’ happen?

The amazing expansion of the past few years ishappening in a wholly different industry thanwhen the first ‘boom’ occurred. In the late 1970sand early 1980s an embryonic ethanol industrywas heralded as the renewable answer toworldwide shortages of fossil fuels. Farmers inAmerica’s heartland had abundant raw materialsin the form of cereal grain; and it seemed at onestage as if every farm would have its own‘gasohol’ plant. Small-scale operations with amyriad of new technologies emerged withunique names like ‘OPEC Killer’ or ‘A Step toIndependence’.

Few of the early fuel ethanol plants were reallybig, as it seemed the beverage plant size wastaken as an example. Many were only capableof producing hundreds of gallons per day, butothers were well-engineered plants withcapacities up to 15-20 million liters per annum.Competitions were held at state fairs in order toenable farmers to determine which design wasbest. Enthusiasm, regardless of expertise, wasthe order of the day. The industry was cryingout for training and technology; and during thistime Alltech Inc. emerged, with its philosophyof marketing through education. Here alsoemerged the concept of the ‘Alcohol Textbook’,and the annual Alcohol School, which is now inits 23rd year.

Figure 1. The alcohol industry reference books byAlltech Inc.

By the mid-1980s, the smaller plants werebeginning to run into typical start-up problems,many associated with poor design, little capital,and relatively high labor costs. Perhaps theywere doomed to fail from the beginning, withor without government support. Production was

2 T.P. Lyons

nowhere near the forecasted 7 billion liters (1.85billion gallons), and instead languished around1.5 billion liters (0.4 billion gallons). Beverageethanol remained the dominant ethanol product.To most, it looked like the gasohol boom wasover.

While smaller operations disappeared, otherlarger companies, well funded and with well-designed plants, continued to grow. Corporationssuch as Cargill, Staleys and ADM becameindustry leaders. Today, one of these (ADM) isresponsible for processing 45% of the corn thatis converted into ethanol in the US. Thiscompany and others of this size knew the ethanolindustry had long term prospects and theyplanned survival strategies accordingly.

Meanwhile in Brazil during the 1970s and1980s, a similar revolution was occurring. In1975 the so-called ‘Proalcohol’ program waslaunched, with massive support by thegovernment. Initially, the program focused onthe production of anhydrous ethanol, which wasblended with gasoline to produce the Braziliangasohol. The emphasis quickly changed tohydrous ethanol (95% ethanol/5% water), whichcould be used in its pure form in cars withspecially designed engines. By the mid-1980s,such cars represented the vast majority of themarketplace and it seemed the Proalcoholproduct was going to replace gasoline on a 1:1basis.

The system, however, was very inflexible.When ethanol became difficult to get in the late1980s, consumers began to switch back toconventional cars in which gasoline (stillblended with a minimum level of ethanol) couldbe used. Was the fuel ethanol program alsogoing to falter in Brazil?

Growing interest in fuel ethanol aroundthe world

Against such a background of declining interestin ethanol in the early 1990s, what brought aboutthe resurrection of the ethanol boom in the USand elsewhere? The first driving factor was aninternational agreement called the KyotoProtocol, in which most industrial nations agreedto improve the environment and lessen thegreenhouse effect by burning less fossil fuel ordoing so more efficiently. It was agreed to rollback carbon dioxide emissions to a given year.The impact was to be enormous.

UNPRECEDENTED GROWTH IN THE US

In the United States, domestic events have giventhe industry a big push. The US Clean Air Actmandated an average of 2% oxygen be presentin what was called ‘reformulated gasoline’(RFG). Representing some 30% of the vast USgasoline market (100 billion gallons), it createda clear demand for an oxygenate. MTBE (methyltertiary butyl ether), with approximately 18%oxygen, and ethanol with 30% oxygen werechosen. Both were excellent oxygen boosters.MTBE, a petroleum derivative, was the earlychoice, but a series of scares regarding itspossible carcinogenic nature and its detectionin groundwater led to its planned phaseout. Thealcohol industry, meanwhile, was beginning toemerge, particularly in the Corn Belt (Nebraska,Iowa, Minnesota and Illinois). Its capacity hadreached 10 billion liters (2.7 billion gallons) by2003 (Figure 2) and in anticipation of thephaseout of MTBE (not just in California, butaround the country) some 73 plants had beenbuilt. At the time we are going to press, these 73plants, located in 20 states, have the capacity of11 billion liters (Figure 3). A further 13 plantsare under construction, which will add some 2billion liters (500 million gallons) of capacity.

The ethanol industry also continues its streakof monthly production records. An all-timemonthly production record of 181,000 barrelsper day (b/d) was set in June of 2003, accordingto the US Energy Information Administration(Figure 4). The previous all-time record was179,000 b/d, set in April of this year.

Driving industry growth: Farmer-owned ethanolplants

Unlike the earlier plants, which were dominatedby the large corporations, most of the new plantsare farmer-owned projects, as farmers seek toadd value to the commodities they grow andparticipate in profit sharing when the plantprospers. These plants have a typical size of 150million liters and a construction cost per liter of$0.40-$0.50. The predominant raw material islocally-grown corn (maize); and ethanolproduction is the third largest and fastest growingmarket for US corn (Figure 5). In 2002, over800 million bushels of corn were processed intoethanol and valuable feed co-products. The USDepartment of Agriculture estimates that morethan one billion bushels of corn will be


0

Mill

ion

so

f g

allo

ns

2500

2000

1500

1000

500

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

Figure 2. Historic US fuel ethanol production.

Figure 3. Fuel ethanol plants in the US (source: RFA).

n facilityEthanol productio

n in 2002Began production

onUnder constructio

els AssociationSource: Renewable Fue

4 T.P. Lyons

processed into ethanol in 2003, or about 10%of the total US harvest. Since 1999, farmer-owned ethanol plants have increased theirpercentage of total production capacity from20% to nearly 35%. As a whole, farmer-ownedethanol plants are the largest ethanol producerin the US.

New marketplace opportunities

Following concerns regarding watercontamination with MTBE, the state of Californiapassed landmark legislation phasing out the useof MTBE in the state’s motor fuel supply. Whilethe original deadline was delayed by one year

to December 31, 2003, all major refinersvoluntarily switched to ethanol in early 2003,resulting in a new market for more than 600million gallons. In 2004, with the MTBE ban inplace, California’s ethanol consumption willincrease to 750-800 million gallons annually.

MTBE bans in New York and Connecticut arealso scheduled to take effect on January 1, 2004.These two states combined represent a marketfor more than 500 million gallons of ethanol. Ifall of the northeast were to replace current MTBEuse with ethanol, a market for more than a billiongallons of ethanol would be created. In total, 17states have passed legislation to phase out MTBEuse; and the future for ethanol looks bright.

Figure 4. Monthly US ethanol production, 2002-2003 (1 barrel = 42 US gallons or 159 liters).

Th

ou

san

ds

of

bar

rels

/day

190

180

170

160

150

140

130

120

110

Jun Jul Aug Sep Oct Nov Dec Jan Feb AprMar May Jun

0

Mill

ion

so

f b

ush

els

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

1000

800

600

400

200

Figure 5. Corn utilized in US ethanol production.


Comprehensive energy legislation offersopportunities for ethanol

The nation’s increasing reliance on importedenergy, particularly from unstable regions of theworld, combined with erratic natural gas andworld oil markets has prompted theadministration and US Congress to debatecomprehensive energy policy legislation. As partof this legislation, an historic fuels agreementsupported by an unprecedented coalition ofagriculture, oil and environmental interestswould provide a growing market for renewablefuels while addressing concerns regardingMTBE water contamination and providingneeded flexibility in the fuels marketplace. Thisagreement would establish a Renewable FuelStandard (RFS), which would set an annual andincreasing minimum level of renewable fuelssuch as ethanol and biodiesel permitted in themarketplace, growing to 5 billion gallons by2012.

New uses for ethanol

While ethanol today is largely used as a blendingcomponent with gasoline to add octane andoxygen, opportunities abound for ethanol toplay a role in advanced vehicle technology andalternative fuels markets. For example, as arenewable fuel with an established fuelinginfrastructure in the US, ethanol is an attractivefuel option for the fuel cell market in both powergeneration and transportation. Research is alsounderway on E dieselTM, a mixture of ethanol,diesel and a blending agent. And there is alargely untapped market for E85, a blend of 85%ethanol and 15% gasoline, to power the growingnumber of available flexible fuel vehicles.

Valuable feed co-product production increasing

Record ethanol production means recordproduction of valuable feed co-products forlivestock. In 2002, US dry mill ethanol facilities,which account for approximately 65% of ethanolproduction, produced about four million shorttons of distillers dried grains with solubles(DDGS). Ethanol wet mills producedapproximately 450,000 US tons of corn glutenmeal, 2.5 million tons of corn gluten feed andgerm meal, and 530 million pounds of corn oil.With estimates for 2003 production nearly 30%

higher than 2002, we can anticipate a parallelincrease in production of feed co-products. Awaiting market exists in the world’s animal feedindustry (>600 million tonnes) wheretraditionally used sources of protein such asanimal by-products and fish meal have beeneither eliminated due to concern surroundingMad Cow Disease (BSE) or reduced availabilityand higher prices. The combined protein andenergy value of ethanol by-products gives themtremendous potential in animal feeds across theworld.

BRAZIL AND SOUTH AMERICA

The US, however, is not the only countryexperiencing an increase in alcohol production.In Brazil, new initiatives are underway and theinflexible Proalcohol program has been changed.Flex-fuel engines, which run equally well oneither pure hydrous alcohol or gasoline, havebeen launched. This technology can cope withanything from 100% hydrous ethanol to agasohol mix of 80% gasoline and 20%anhydrous ethanol, adjusting to any changewithin a fraction of a second. This compares veryfavorably to the engines in the US, which wouldnormally handle up to 85% anhydrous in a blendand cannot use pure hydrous ethanol.Furthermore, since ethanol sells in Brazil for lessthan gasoline at the pumps (50-60% of thegasoline price) even taking into account that 30%more ethanol is required to travel the samedistance, the new cars are bound to be a successand will drive up ethanol demand.

It is expected that a million cars with Flex-fuelengines could be sold each year. With an average240 liters (63 gallons) of ethanol per monthconsumption, this could mean an additional 3billion liters would be needed over Brazil’scurrent 12-13 billion liters of production.

Table 1. Projected production of ethanol (billion liters)1.

2003 2005

USA 10 18Brazil 13 16European Union Unknown 8India 1.5 1.8Thailand Unknown 0.7Australia .04 .4China Unknown 21F.O. Licht, July 2003

6 T.P. Lyons

While few people will say exactly how much itcosts to produce a liter of fuel ethanol in Brazil,it is believed to be between $0.10 and $0.15.This puts the industry in a very favorableposition to export ethanol, since a similar costin the US from corn would be more like $0.20.

Fuel ethanol emergence in South America isnot limited to Brazil. In Peru and Colombiainterest is also growing. Colombia estimates anannual need for almost one billion liters and 12production plants.

THE EUROPEAN UNION

Driven by the Kyoto Agreement, the EuropeanCommission proposed in 2001 to increase theuse of crop-based, non-polluting renewablefuels. The target was for biofuels to account for2% of all fuels in EU transport by 2005. It wasmandated this would increase to 5.75% by 2010(from 0.3% in 2001).

Both ethanol and biodiesel (from canola) areconsidered biofuels. In 2003 the targets weremade voluntary, but a direction on tax relief forrenewable energy was expected to be put intoplace. These would augment the existing taxrelief schemes, which vary from country tocountry. Spain, Germany, and Sweden, forexample, have 100% tax relief on renewablefuels, while France has 60% and Britain 40%.

Across Europe, from Poland to Ireland, interestis rising as the new legislation begins to takeeffect.

ASIA

India’s transport sector accounts for over halfof the country’s oil consumption. It is notsurprising, therefore, that interest is growing inreplacing dependence on fossil fuels. By January2003, nine states had mandated at least 5%ethanol (E5) in transport fleets. Local distillerieswith a capacity of some 3.2 billion liters (900million gallons) are currently only operating at50% capacity. The demand for an additional 400million liters under this new mandate is thereforeimmediately being met. Although some politicalmaneuvering is still needed, the fuel ethanolprogram looks sound.

In Southeast Asia, Thailand’s government, likemany in the region, had an early interest in

bioethanol as a way to reduce oil imports. It wasestimated that 0.7 billion liters of ethanol wouldbe required should they decide to introduce theoxygenate at 10% of gasoline. Both cassava(tapioca) and molasses were suggested as rawmaterials and eight projects were approved. Atthis time only one project is under construction.

Two distinct areas for ethanol production existin China: the maize-rich north and the sugar-cane rich south. Both will feature in the country’sethanol future. The government sees ethanol aspart of its program to reduce pollution –important with the looming 2004 Olympics –and at the same time encourage ruralemployment. China’s largest fuel ethanolfacility, costing almost $300 million to construct,is in Jilin. With a capacity of some 600 millionliters (165 million gallons), it will use 2 milliontons of maize a year (13% of production in thatprovince). By October 2003, all vehicles in theProvince of Jilin will use fuel containing at least10% ethanol.

AUSTRALIA

Uncertainty also exists in Australia, where atarget of 350 million liters was encouraged withtax exemptions and grants for up to six years.Local governments have tried to boost ethanolconsumption. For example, in Cairns,Queensland’s largest city, government-ownedvehicles were mandated to use E10 (90%gasoline, 10% ethanol). Using some 20,000 litersper day, the government hoped in this way toavoid some of the negative publicity beinggenerated by the anti-ethanol lobby.

Conclusions

Ethanol, as a global product, is experiencing asignificant jump in production and productioncapacity. Whether it be used domestically orexported to areas around the world (such asJapan, Singapore and even California) whereproduction is not sufficient or not possible, it isclear it will become a global commodity. Thefuture, therefore, is very positive; and this futurewill not be one dependent on subsidies. A 1997study by the Kellogg School of Management atNorthwestern University concluded the net effect


of ethanol production was a reduction in the USfederal budget deficit. This was brought aboutby increased tax revenues due to the increasesin personal income from wages and salaries,higher farm income, more corporate tax revenue,and declining unemployment compensationpayments. Alcohol as a fuel is clearly with us tostay.

Raw

material handling and processing

Grain dry milling and cooking procedures 9

Chapter 2

Grain dry milling and cooking procedures: extracting sugars in preparationfor fermentation

DAVE R. KELSALL AND T. PEARSE LYONS

Alltech Inc., Nicholasville, Kentucky, USA

This chapter deals with the milling and cookingstages of alcohol production from whole cereals.In brief, in this dry milling process the wholecereal, normally corn (maize), is ground in a millto a fine particle size and mixed with liquid,usually a mixture of water and backset stillage.This slurry is then treated with a liquefyingenzyme to hydrolyze the cereal to dextrins, whichare a mix of oligosaccharides. The hydrolysisof starch with the liquefying enzyme, called α-amylase, is done above the temperature ofgelatinization of the cereal by cooking the mashat an appropriate temperature to break down thegranular structure of the starch and cause it togelatinize. Finally the dextrins produced in thecooking process are further hydrolyzed toglucose in a saccharification process using theexoenzyme glucoamylase, and possibly anotherenzyme (Rhizozyme™). These enzymes mayalso be added to the yeast propagation tank orthe fermentor. These separate stages of milling,cooking and saccharification will be explainedin more detail. This is called the dry millingprocess. Most of the new distilleries use thisprocess or a minor variation of it.

Milling

The purpose of milling is to break up the cerealgrains to appropriate particle size in order to

facilitate subsequent penetration of water in thecooking process. The optimum size of theground particle is the subject of disagreement.Some engineers believe the particle should beas fine as possible to allow the water maximumaccess for hydrolysis of starch while othersbelieve a better yield is obtained if the particlesare larger and the jet cooker can act on theparticles. The key is simply to expose the starchwithout grinding so fine as to cause problems inco-product recovery. A wide variety of millingequipment is available to grind the whole cerealto a meal. Normally, most distilleries use hammermills, although some may use roller mills,particularly for small cereal grains.

HAMMER MILLS

In a hammer mill, the cereal grain is fed into agrinding chamber in which a number ofhammers rotate at high speed. The mill outletcontains a retention screen that holds back largerparticles until they are broken down further sothat there will be a known maximum particlesize in the meal. The screens are normally in thesize range of 1/8-3/16 in. (2-4 mm).

Sieve analysis (particle size distribution) showswhether the hammer mill screens are in goodorder and whether the mill is correctly adjusted,and should be conducted regularly. Table 1

10 D.R. Kelsall and T.P. Lyons

shows a typical sieve analysis for corn. The twolargest screens retain only 11% of the particleswhile the quantity passing through the 60-meshscreen is also fairly low at 7%. For efficientprocessing of the cereal starch into alcohol, theparticles should be as fine as possible. Howevera compromise must be made such that theparticles are not so fine that they cause ballingin the slurry tank or problems in the by-productrecovery process. Sometimes the sieve analysiswill vary depending on the severity of the shakerand a consistent shaking should be done eachtime the test is done.

Table 1. Typical results of a sieve analysis of corn meal.

Screen size Hole size (in) Corn on screen (%)

12 0.0661 3.016 0.0469 8.020 0.0331 36.030 0.0234 20.040 0.0165 14.060 0.0098 12.0Through 60 7.0

Fineness of the grind can be significant factorin the final alcohol yield. It is possible to obtaina 5-10% difference in yield between a fine anda coarse meal. Table 2 shows the typical alcoholyield from various cereals. It can be seen thatthe normal yield from corn is 2.85 gallons ofanhydrous ethanol per bushel (56 lbs). However,the yield with coarsely ground corn may dropto 2.65 gallons per bushel, a reduction in yieldof 7.5%. This is a highly significant reductionand would have serious economic consequencesfor any distiller. Grinding is about expandingthe amount of surface area. When more surfacearea is exposed, water and enzymes are betterable to penetrate the kernel. A particle 1 in. indiameter has a total surface area of 6 in2. If yougrind the 1 in. particle into 1,000 particles,surface area becomes 6,000 in2. No amount ofheating or ‘exploding’ will overcome poor grindas measured by yield. The key is to expose moresurface area, but to do so in a controlled manner.

Since grind is so important, it is recommendedthat a sieve analysis of the meal be done at leastonce per shift. The distiller should setspecifications for the percentage of particles oneach sieve; and when the measured quantity falls

outside these specifications the mill should beadjusted. Normally the hammers in a hammermill are turned every 15 days, depending onusage; and every 60 days a decision should bemade as to whether or not to replace the hammersand screens.

Table 2. Typical alcohol yields from various cereals.

Cereal Yielda

(US gallons of anhydrousalcohol/bushel)

Fine grind corn1, 3/16 in. 2.85Coarse grind corn1, 5/16 in. 2.65Milo 2.60Barley 2.50Rye 2.40

aNote that a distiller’s bushel is always a measure of weight.It is always 56 lb, regardless of the type of grain.1Laboratory data using Rhizozyme™ as glucoamylasesource.

Since sieve analysis is critical, a case can alsobe made for recycling to the hammer mill anygrain not ground sufficiently fine. Fineness ofthe grind also has an important bearing oncentrifugation of the stillage post-distillation. Afiner grind may yield more solubles and henceplace a greater load on the evaporator. Howeversince the key is to maximize yield, dry houseconsiderations, while important, cannot overrideyield considerations.

ROLLER MILLS

Some distillers use roller mills (e.g. malt whiskyproducers), particularly where cereals containingsubstantial quantities of husk material are used.In a roller mill, the cereal is nipped as it passesthrough the rollers, thus exerting a compressiveforce. In certain cases, the rollers operate atdifferent speeds so that a shearing force can beapplied. The roller surfaces are usually groovedto aid in shearing and disintegration. Figure 1shows the general configuration of a roller mill.

In Scotland the solids in whisky mash, madeentirely from malted barley, are usually removedby using a brewery-type lauter tun, which is avessel with a perforated bottom like a largecolander. In this case, a roller mill should beused as the shear force allows the husk to be


separated with minimal damage. The husk thenacts as the filter bed in the lauter tun for theefficient separation of solids and liquid.

Cooking

The purpose of cooking is 5-fold:

Sterilization. The mash must be sterilized so thatharmful bacteria are minimized. A brewerachieves this by boiling mash (wort) while thedistiller achieves it by cooking.

Solubilization of sugars. Sugars are solubilizedto a certain point, typically 2-3% free sugars. Inthis way yeast growth can occur rapidly but notexcessively as would happen if too much sugarwas released at once.

Release of all bound sugars and dextrins (chainsof sugar). Extraction must occur such thatsubsequent enzymatic hydrolysis can ensure thatall sugars are utilized. Starch (precursor ofsugars) is in large bound to protein and fiber,which is are freed during cooking.

Protein breakdown to amino acids should beminimized. Amino acids and peptides canbecome bound to sugars in Maillard reactions,which leaves sugar unavailable.

Mash in

Feed roll

1st pair

of rolls

Anti-explosion

device

Cylindrical

screen

Revolving

beater

2nd pair

of rolls

Husks

To meal bin

Coarse gritsFine grits

and flour

Figure 1. Schematic of a roller mill.

Reduction in viscosity. Following gelatinization,viscosity is reduced allowing the slurry to bemoved through lines for subsequent processing.

Cooking is the entire process beginning withmixing the grain meal with water (and possiblybackset stillage) through to delivery of a mashready for fermentation. Figure 2 shows thecomponents that make up a typical milling andcooking system. This schematic diagram couldrepresent the processes involved in beverage,industrial or fuel alcohol production, except thatnowadays only the whisky distillers use malt asa source of liquefying and saccharifyingenzymes. All other alcohol producers usemicrobial enzyme preparations. The fastestenzyme systems have low calcium requirements(2-5 ppm), lower pH optima (less bufferingrequired) and high temperature stability. Theyare designed to reduce viscosity, with lessemphasis on DE. The key to cooking is to liquefythe starch so it can be pumped to the fermentor,hence viscosity is critical.

HYDROLYSIS OF STARCH TO FERMENTABLESUGARS

Step 1: gelatinization

The purpose of cooking and saccharification isto achieve hydrolysis of starch into fermentablesugars, which is accomplished by theendoenzyme α−amylase, followed by theexoenzyme glucoamylase (amyloglucosidase) torelease glucose. However in order for the α-amylase to gain access to the starch molecules,the granular structure of the starch must first bebroken down in the process known asgelatinization. When the slurry of meal andwater are cooked, the starch granules start toadsorb water and swell. They gradually lose theircrystalline structure until they become large, gel-filled sacs that tend to fill all of the availablespace and break with agitation and abrasion.

The peak of gelatinization is also the point ofmaximum viscosity of a mash. Figures 3, 4 and5 show the progressive gelatinization ofcornstarch, as viewed on a microscopic hot stage.In Figure 3 the granules are quite distinct andseparate from the surrounding liquid. In Figure


4 these same granules have swollen in size andsome of the liquid has entered the granules.Figure 5 shows the granules as indistinct entitiesin which the liquid has entered to expand themconsiderably.

Gelatinization temperatures vary for thedifferent cereals (Table 3). Some distillersconsider it important for the slurryingtemperature of the meal to be below thetemperature of gelatinization. This avoidscoating of grain particles with an imperviouslayer of gelatinized starch that prevents theenzymes from penetrating to the starch granulesand leads to incomplete conversion. Manydistillers, however, go to the other extreme andslurry at temperatures as high as 90°C (190°F).At these temperatures starch gelatinizes almostimmediately and with adequate agitation thereis no increase in viscosity and no loss of yield.

Table 3. Gelatinization temperature ranges of variousfeedstocks.

Gelatinization range (°C)

Corn Standard 62-72 High amylose 1 67->80 High amylose 2 67->80Barley 52-59Rye 57-70Rice (polished) 68-77Sorghum (milo) 68-77Wheat 58-64

Step 2. Liquefaction: hydrolysis of starch todextrins

Liquefaction is accomplished by the action ofα−amylase enzyme on the exposed starchmolecules. Starch exists in two forms. One form

Liquefaction

tank 90°C

Mash cooler

Yeast

starter tank

Saccharification tank

60°C

Fermentors

32°C

Cooker

90 - 120°C

Pre-mixing tank

40 - 60°CGlucoamylase

(0.06 - 0.12%)

(Allcoholase II L400�)

or

malt (up to 10%) (beverage alcohol only)

or

Rhizozyme� (0.05%)

CoolHeatWater

Still

Hammer mill

α-amylase

(0.04 - 0.06%)

(High T L120�)

α-amylase (0.02%)

(High T L120�)

Raw material

Yeast food

Rhizozyme�

(0.01% supplement or

0.05% total replacement)

(Rarely used)

Figure 2. Components of a milling and cooking system.


is the straight-chained amylose, where theglucose units are linked by α−1,4 glucosidiclinkages (Figure 6). The amylose content of cornis about 10% of the total starch; and the amylosechain length can be up to 1,000 glucose units.The other form of starch is called amylopectin,which represents about 90% of the starch in corn.

Amylopectin has a branched structure (Figure7). It has the same α−1,4 glucosidic linkages asin amylose, but also has branches connected byα−1,6 linkages. The number of glucose units inamylopectin can be as high as 10,000. Corn,wheat and sorghum (milo), the three mostcommon feedstocks for ethanol production, have

Figure 3. Gelatinization of cornstarch. Starch granules viewed on microscope hot stage at 67°C (under normal illumination).

Figure 4. Gelatinization of cornstarch. Same granules as in Figure 3 at 75°C (under normal illumination).


CH2OH

O

O

HOH

H

H

OH

α-1,4 linkage

H4 1 1

5

6

3 2

CH2OH

O

HOH

H

H

OH

4 1

n

5

6

3 2

H

O

HH

O

Amylose

Figure 6. Structure of amylose.

Figure 5. Gelatinization of cornstarch. Same granules as in Figures 3 and 4 at 85°C (under normal illumination).

similar levels of starch, but percentages ofamylose and amylopectin differ with grain andwith variety. Starch structures are reviewed ingreater detail in the chapter by R. Power in thisvolume.

The α−amylase enzyme acts randomly on theα−1,4 glucosidic linkages in amylose andamylopectin but will not break the α−1,6linkages of amylopectin. The resulting shorterstraight chains (oligosaccharides) are calleddextrins, while the shorter branched chains arecalled α-limit dextrins. The mixture of dextrinsis much less viscous.

Step 3. Saccharification: release of glucose fromdextrins

Saccharification is the release of the individualglucose molecules from the liquefied mixtureof dextrins. The dextrins will be of varying chainlengths. However, the shorter the chain lengththe less work remaining for the exoenzymeglucoamylase, which releases single glucosemolecules by hydrolyzing successive α−1,4linkages beginning at the non-reducing end ofthe dextrin chain. Glucoamylase also hydrolyzesα−1,6 branch linkages, but at a much slower


rate. One real problem with HPLC analysis isthat dextrins are given as total oligosaccharideswith no differentiation between four glucoseunits and 20 glucose units. The work of theglucoamylase is directly proportional to thelength of the dextrin chain; and there is currentlyno way of knowing how well the α-amylase hasworked to produce small oligosaccharides.

In Scotch whisky production, malted barley isused as a source of both α−amylase and theexoenzyme ß-amylase. The fermentable sugarproduced is maltose, a dimer made up of twoglucose units.

Cooking systems

BATCH COOKING SYSTEMS

In considering all the different processes thatmake up cooking, it should first be explained

that there are a variety of batch and continuouscooking systems. For a batch system there isusually only one tank, which serves as slurrying,cooking and liquefaction vessel. Live steam jetsare typically installed in the vessel to bring themash to boiling temperature along with coolingcoils to cool the mash for liquefaction. Figure 8shows a typical batch cooking system.

In the batch cooking system, a weighedquantity of meal is mixed into the vessel with aknown quantity of water and backset stillage.These constituents of the mash are mixed insimultaneously to ensure thorough mixing. Thequantity of liquid mixed with the meal willdetermine the eventual alcohol content of thefermented mash. When a distiller refers to a ’25gallon beer’, it means 25 gallons of liquid perbushel of cereal. For example, for a corndistillery with an alcohol yield of 2.5 gallons ofabsolute alcohol per bushel, the 25 gallons of

CH2OH

O

O

HOH

H

H

OH

α-1,6 linkage

α-1,6 glucosidic linkages

α-1,4 glucosidic linkages = Glucosidic units

α-1,4 linkageH4 1

5

6

3 2CH2

O

HOH

H

H

OH

4 1

5

6CH2OH6

3

HOH

H

4

5

32

H

O

H H

O

H

O

Amylopectin

Figure 7. Structure of amylopectin.


liquid would contain 2.5 gallons of alcohol.Therefore it would contain 10% alcohol byvolume (abv). Using the distillery alcohol yield,the distiller can determine the quantity of cerealsand liquid to use. Most distilleries traditionallyoperated with beers in the 10-15% abv range,although some beverage plants run at alcohollevels as low as 8%. More recently fuel alcoholplants run at alcohol levels as high as 19-20%by volume, the average being nearer to16-17%by volume.

In the batch cooking system, a small quantityof α−amylase is added at the beginning (0.02%w/w of cereal) to facilitate agitation in the highviscosity stage at gelatinization. After boiling,usually for 30-60 minutes, (sometimes under aslight pressure), the mash is cooled to 75-90°Cand the second addition of α−amylase made(0.04-0.06% w/w cereal). Liquefaction thentakes place, usually over a holding period of 45-90 minutes. The mash should always bechecked at this stage to make certain that nostarch remains. Starch produces a blue or purplecolor with iodine. Mash should not be transferredfrom the liquefaction hold until it is ‘starch-negative’.

The pH range for efficient α−amylase usageis 6.0-6.5, although enzymes with good activityat pH 5.5-5.7 are now available. Therefore, mashpH should be controlled in this range from thefirst enzyme addition until the end ofliquefaction. The glucoamylase enzyme has alower pH range (4.0-5.5), so after liquefactionthe pH of the mash should be adjusted with eithersulfuric acid or backset stillage, or a combinationof the two.

The quantity of backset stillage as a percentageof the total liquid varies from 10 to 50%. Onone hand, the backset stillage supplies nutrientsessential for yeast growth. However too muchbackset stillage can result in the oversupply ofcertain minerals and ions that suppress goodfermentation. Especially noteworthy are thesodium, lactate and acetate ions. Sodiumconcentrations above 0.03% or lactate above0.8% or acetate above 0.03% can inhibit yeastgrowth and can slow or possibly stop thefermentation prematurely. Overuse of backset(or even process condensate water) must beavoided to prevent serious fermentationproblems.

Grain

Water

Backset

Steam

α-amylase

(High T L120�)

Cooling coils

Heat exchanger

Glucoamylase + distillers yeast

(Allcoholase II L400� + Rhizozyme� + Thermosacc�)

To fermentors

30°- 32°C

Figure 8. Batch cooking system.


CONTINUOUS COOKING SYSTEMS

Few distilleries outside of beverage plants usebatch cooking. Most fuel ethanol distilleries usea continuous cooking system. In the continuouscooking process (Figure 9) meal, water andbackset stillage are continuously fed into apremix tank. The mash is pumped continuouslythrough a jet cooker, where the temperature isinstantly raised to 120°C. It then passes into thetop of a vertical column. With plug flow, themash moves down the column in about 20minutes and passes into the flash chamber forliquefaction at 80-90°C. High temperature-tolerant α−amylase is added at 0.04-0.08% w/wcereal to bring about liquefaction. The retentiontime in the liquefaction/flash chamber is aminimum of 30 minutes, but should be at least60 minutes. The pH from slurrying through tothe liquefaction vessel must be controlled withinthe 6.0-6.5 range. The greatest advantage of thissystem is that no enzyme is needed in theslurrying stage, leading to significant savings inenzyme usage. It is critical to have plug flowthrough the chamber along with good enzymedispersion. The mash should have a relativelylow viscosity and dextrose level of 2-3%.

Modern systems have 29-33% solids. From theliquefaction chamber, the mash is pumpedthrough a heat exchanger to fermentation.

CONTINUOUS U-TUBE COOKING SYSTEM

The continuous U-tube cooking system (Figure10) differs from the columnar cooking systemin that the jet cooker heats the mash to 120-140°C prior to being transferred through acontinuous U-tube. The retention time in the U-tube is only three minutes, after which it isflashed into the liquefaction vessel at 80-90°Cand the enzyme is added (high temperature-tolerant α-amylase 0.05-0.08% w/w cereal). Theresidence time in the liquefaction vessel is aminimum of 30 minutes.

The main advantage of this system is therelatively short residence period in the U-tube.If properly designed there is no need to add anyα−amylase enzyme in the slurrying stage.However, because of the relatively narrowdiameter of the tubes, some distillers add a smallamount of enzyme to the slurry tank to guaranteea free flow.

Slurry tank

60°C max

Column

20 min

120°C

α-amylase

(High T L120�)

0.05 - 0.8 w/w

Pressure relief

Vacuum

ejector

Flash chamber 80-90°C

30 min retention time

Glucoamylase + yeast

(Allcoholase II L400� +

Rhizozyme�+ Thermosacc�)

To fermentor

30-32°C

Grain Steam

Water

Backset

(up to 40%)

Heat exchanger

Liquefaction

Figure 9. Continuous columnar cooking system.


COMPARING COOKING SYSTEMS

The relative heat requirements of the threecooking systems can be seen in Table 4.Surprisingly, the batch system is the mostenergy-efficient. Batch systems also generallyuse less enzyme than the other systems, possiblydue to the difficulty of accurate dosing and goodmixing with the continuous systems. The maindisadvantage of the batch system compared tothe continuous system is the poor utilization orproductivity per unit of time. The temperature-time sequences for the three systems shown inFigure 11 demonstrate how much moreefficiently time is used in continuous systemscompared to the batch system.

Table 4. Relative heat requirements of cooking systems.

Batch 1Continuous columnar 1.18Continuous U-tube 1.37

In the continuous systems, the flow diagramsshow steam addition to raise the mashtemperature. This temperature increase is

brought about instantaneously by a jet cookeror ‘hydroheater’.

One purpose of the cooking process is tocleave the hydrogen bonds that link the starchmolecules, thus breaking the granular structureand converting it to a colloidal suspension.Another factor in the breakdown of starch is themechanical energy put into the mash viaagitation of the different vessels in which thecooking process takes place. Well-designedagitation is very important in a cooking system;and the problem is intensified when plug flowis also desired.

Mash viscosities give an indication of therelative ease or difficulty with which somecereals are liquefied. Figure 12 comparesviscosity against temperature for corn and waxymaize (amioca) and illustrates the difference inviscosity profiles.

All of the cooking systems described requirethe addition of enzymes at least for theliquefaction stage where most of the hydrolysistakes place. Many distilleries now use a hightemperature-tolerant α−amylase. The optimumpH range for this enzyme is between 5.8 and6.5, although it shows good stability up to pH

Flash chamber 80-90°C

30 min retention time

Glucoamylase + yeast

(Allcoholase II L400� +

Rhizozyme� + Thermosacc�)

Steam

Retention 'U' tube

3 min retention

140°C

α-amylase (0.05%)

(High T L120�)

Pressure

control

Vacuum

ejector

To fermentor

30-32°C

Heat exchanger

Slurry tank

60°C max

Grain

Water

Backset

(up to 40%)

Figure 10. High temperature, short time, continuous U-tube cooking system.


8.5 (Figure 13) while the optimum temperaturerange is 88°C-93°C (Figure 14). Typically, thistype of enzyme would be used at between 0.04%and 0.08% by weight of cereal. Where it isnecessary to add some α−amylase enzyme tothe slurrying vessel, the dosage rate may beslightly higher.

18

14

10

6

2

65 75 85 85 75 65 55 45 35

Vis

cosit

y in

pois

es

Held at

95°C

Corn

Waxy maize

(Amioca)

Temperature (°C)

Figure 12. Increasing viscosity with cooking temperature.

160

140

120

100

80

60

40

20

0

Continuous U-tube cooker

Continuous columnar cooker

Holding time

Batch cooker

Tem

perature °C

0 25 50 75 100 125

Time (minutes)

Figure 11. Temperature/time sequences in various types of cooking systems.

100

75

50

25

0

4 5 6 7 8 9 10 11 12

% activ

ity rem

ain

ing

pH

Treatment time: 20 minutes

Calcium: 15 ppm

Figure 13. Effect of pH on activity of α-amylase.


The reaction time for enzyme-catalyzedreactions is directly proportional to theconcentration of enzyme. Consequently, distillerswishing to minimize the quantity of enzymeused should design equipment to have longresidence times to allow the reactions to becompleted with minimal dosage of enzyme.

Saccharification: should we have highlevels of sugar going into the fermentor?

Saccharification of distillery mashes is asomewhat controversial subject. Over the last10 years many distillers have changed fromsaccharifying mash in a dedicated sacchari-fication vessel (or ‘sacc’ tank) to adding thesaccharifying enzyme directly to the fermentorin a process referred to as SimultaneousSaccharification and Fermentation (SSF).Saccharification in a separate vessel is still thepractice in some distilleries, particularly forbeverage production and continuousfermentation.

Another factor is use of Rhizozyme™, anenzyme produced by surface culture (solid statefermentation). Rhizozyme™ has an optimum pHof 3.5-5.0 and an optimum temperature of 30-35°C. As such, it is a glucoamylase more suited

100

75

50

25

0

49 54 60 65 71 77 82 88 93

% activ

ity rem

ain

ing

Temperature (°C)

Substrate: 4% w/v soluble starch

+ 10 ppm calcium

Treatment time: 15 minutes

100

Figure 14. Activity of high temperature-tolerant α-amylasein relation to temperature.

to a distillery fermentor (Table 5). Rhizozyme™has particularly found favor in SSF situationsbecause the enzyme can work at near itsoptimum in the fermentor. Because the organismused to produce RhizozymeTM is grown on asolid substrate (eg wheat bran), this enzymesource also contains side activities that assist inreleasing more carbohydrate and protein (Table5). These side activities work to release sugarsfrom other structural carbohydrates besidesstarch, which also releases bound starch that isthen available for conversion. This approach toenzyme addition is able to increase yield tohigher levels than normally seen with singleactivity liquid glucoamylase. Levels in excessof 2.9 gallons/bushel have been reported; andthe target is 3.1 gallons/bushel (122 gallons or456 liters per tonne).

Table 5. Advantages of Rhizozyme™ over conventionalglucoamylase.

Conventional Rhizozyme™glucoamylase

Temperature optimum, °C 60+ 30+pH optimum 4.0-5.5 3.5-5.0Amylase activity, SKB units None 50,000Cellulase activity CMC-ase units None 2500Amylopectinase, AP units None 5000

IF A SACCHARIFICATION STEP IS USED

Mash from the liquefaction vessel is cooled,usually to 60-65°C, and transferred to asaccharification vessel where the glucoamylase(amyloglucosidase) enzyme is added. Thisexoenzyme starts hydrolyzing the dextrins fromthe non-reducing end of the molecule andprogressively, though slowly compared toendoenzymes, releases glucose. The saccharifyingprocess is usually carried out with a residencetime of between 45 and 90 minutes, but can beas long as 6 hrs, and the glucoamylase is addedat 0.06-0.08% by weight of cereal used. Somedistillers measure the quantity of glucoseproduced by measuring the dextrose equivalent(DE) of the mash. A DE of 100 represents pureglucose, while zero represents the absence ofglucose. This test is rarely used nowadays, asmany distillers have high performance liquid


chromatography systems (HPLC) that canmeasure sugars directly. Recent experience,however, shows that DE, provided it is above10, is of no concern. In focusing on 23% ethanol,the key in any cooking and liquefaction processis to liquefy, i.e., lower viscosity, so that mashcan be pumped through the heat exchanger tothe fermentor (for SSF) or to the saccharificationtank.

The functional characteristics of liquidglucoamylase prepared from the microorganismAspergillus niger, can be seen in Figures 15 and16. Two parameters, temperature and pH, dictatehow enzymes can be used. While liquefactionis carried out at a pH of 6.0-6.5 and atemperature of 90°C, this is not at all acceptablefor saccharification. The pH must be in the 4-5range for saccharification; and the optimumtemperature for the glucoamylase activity is60°C. The mash, therefore, must be acidifiedwith either sulfuric acid or backset stillage orboth before addition of the glucoamylase.Temperature must also be adjusted. As mentionedpreviously, normal mash saccharificationtemperature is 60-65°C; although formicrobiological reasons 70-75°C would bepreferable. Lactobacillus can survive at 60°C; andfrequent infection of saccharification systems hascaused many distillers to change tosaccharifying in the fermentor.

IF NO SACCHARIFICATION STEP IS USED

If no saccharification step is planned, theliquefied mash is simply cooled from 90°Cthrough a heat exchanger and transferred to thefermentor. A portion of the liquefied mash isdiverted to a yeast starter tank where yeast,glucoamylase and the RhizozymeTM is added.Conventional glucoamylase (L400) is added at0.06% with 0.01% Rhizozyme™ recommendedas a supplement. Rhizozyme™ alone can beadded at 0.05%, in which case no glucoamylaseis required.

Recommendations

Given the goals and the variables involved, whichcooking system should be chosen? Acomparison of systems used in four distilleries

demonstrates the diversity of approach possible,yet points out many similarities (Table 6). Itwould appear that there are three schools ofthought:

a) A long liquefaction period (1-2 hrs at 90+ °C)after the high temperature jet cooker and lowtemperature slurry (no enzymes).

b) A short high temperature slurry prior to along liquefaction period (1-2 hrs at 90+ °C)after the high temperature jet cooker.Enzymes used in slurry.

c) Choice of saccharification stage or SSF.

From the experience of the authors,recommendations are for a short slurry at 140-150°F followed by jet cooking (10-15 min.) anda long (1.5 hrs) liquefaction step. The mashshould have a sugar profile similar to Table 7. Itshould then be cooled to fermentationtemperature on the way to the fermentor. Thisprofile will ensure a rapid start by yeast with nosugar overload. In the presence of Rhizozyme™or glucoamylase, dextrins will continue to be‘spoon-fed’ to the yeast during fermentation.Enzyme additions should include all the amylaseduring liquefaction and Rhizozyme™ duringfermentation.

Table 7. The target sugar profile following liquefaction.

pH 4.2Solids 29-35Glucose, % 1.8-2.5Maltose, % 0.7-0.9Maltotriose, % 0.2-0.3Lactic, % 0.07-0.08Glycerol, % 0.2-0.3

The future

Many new enzyme systems including xylanases,hemicellulases, ligninases (esterases) and othersare being developed and the cooking systemmust be flexible enough to handle them. Theobjective is to maximize the biochemical activityof the enzyme so as to maximize alcohol yield,not to fit an enzyme to an existing engineeringdesign.

Enzymatic conversion of starch to fermentable sugars 23

Chapter 3

Enzymatic conversion of starch to fermentable sugars

RONAN F. POWER

North American Biosciences Center, Alltech Inc., Nicholasville, Kentucky, USA

Introduction

Starch is the second most abundant carbohydratein the plant world after cellulose. It is theprincipal plant storage polysaccharide and issimilar in structure and function to glycogen,which is the main storage polysaccharide of theanimal kingdom. Both starch and cellulose areglucose-based polymers that differ in theorientation of the linkages between the glucosesubunits in the respective macromolecules.Saccharomyces cerevisiae is the primarymicroorganism involved in the transformationof starch to ethyl alcohol, and corn starchremains the major raw material for industrialalcohol fermentation although potato starch alsohas limited use, particularly in Europe. Forexample, in the 1999 US crop year, 526 millionbushels of corn (14.7 million tons at 15%moisture) were used in the fuel ethanol industry(McAloon et al., 2000).

The starting compound for alcohol synthesisby yeast is glucose (dextrose), which is a six-carbon sugar. Once fermented, each moleculeof glucose yields two molecules of ethanol andcarbon dioxide, respectively.

C6H12O6 → 2C2H5OH + 2CO2

The sequence of biosynthetic reactions by whichyeast converts glucose to ethanol is termed the

Embden-Meyerhof-Parnas (glycolytic) pathway(Figure 1). In this pathway, glucose isphosphorylated, then split into twophosphorylated 3-carbon derivatives,glyceraldehyde-3-phosphate and dihydroxy-acetone phosphate. Both products aresubsequently converted into pyruvic acid, whichunder anaerobic conditions yields ethanol andcarbon dioxide. It is obvious from Figure 1 thatthe yeast, S. cerevisiae, is very well equipped toferment glucose and fructose. This is why winefermentations are technically very simple. Grapepressings deliver pure glucose and, of course,the required yeast are already present on thesurface of the grapes. Likewise, molassesrepresents a readily assimilated feedstock foryeast since the predominant sugar in molassesis the disaccharide, sucrose. Yeast contains theenzyme sucrase, which cleaves sucrose into itsconstituent monosaccharides, glucose andfructose.

Starch, however, cannot be fermented directlyby S. cerevisiae, because the organism lacks therequisite starch-degrading or amylolyticenzymes to liberate glucose from this storagepolysaccharide. Whether the starch is derivedfrom corn, potatoes, sorghum (milo), barley orother cereal grains, the bonds between theglucose subunits in the starch chain must first

24 R.F. Power

GLUCOSE

Glycolysis

Fructose-6-P

Glucose-6-phosphate

Fructose 1,6bisphosphate

Dihydroxyacetone phosphate

3-phosphoglycerate

Glyceraldehyde3-phosphate

Phosphoenolpyruvate

Pyruvate

ATP

ADP

ATP

ADP

1,3 bisphosphoglycerate

2-phosphoglycerate

H2O

Glucokinase

Phosphohexose

isomerase

Phosphofructokinase

Aldolase

Triosephosphate

isomerase

NAD+

NADH

Glyceraldehyde-3-P

dehydrogenase

ATP

ADPPhosphoglycerate

kinase

Phosphoglycerate

mutase

Enolase

ATP

ADPPyruvate

kinase

Anaerobichomolactic

fermentation

Anaerobicalcoholic

fermentation

Aerobicoxidation

CO2+ Water

CO2+ Ethanol

Lacticacid

YeastLactic bacteria TCA cycle

Figure 1. The Embden-Meyerhof-Parnas (glycolytic) pathway for ethanol production by yeast.


be hydrolyzed to liberate free glucose moleculeswhich yeast can utilize. Brewers achieve this byexploiting the endogenous starch-digestingenzymes produced by seeds when theygerminate. Grains, such as barley, destined forfermentation, are first sprouted to initiate enzymesynthesis within the seeds. This process istermed malting. The malted grains are thensteeped in hot water to allow the enzymes toconvert starch to fermentable sugar. Thismashing phase is then followed by a lauteringor mash filtration stage to separate the solidsfrom the glucose-rich liquid ‘wort’ to which yeastis added.

The first step in the industrial production ofalcohol from starch also requires that the starchis saccharified; and this is generally recognizedto consist of two phases: a) the dextrinization orliquefaction phase in which starch is partiallydegraded to soluble dextrins, and b) the truesaccharification or ‘conversion’ phase in whichthe dextrins are hydrolyzed to fermentablesugars. The two phases of saccharification maybe carried out by either acid or enzymatichydrolysis. In the past, acid combined with hightemperatures and pressures was used toaccomplish the same reaction that enzymesperform today under relatively mild conditions.Acid hydrolysis was extremely hazardous,required corrosion-resistant vessels and resultedin a product with a very high salt content, dueto neutralization. In addition, dextrose yield waslimited to approximately 85% and off-colors andflavors were often a result of the acid hydrolysisprocess. The change to enzymatic hydrolysis inthe 1960s eliminated all the drawbacks of theacid hydrolysis process and increased the yieldof dextrose to between 95 and 97%. This chapterwill focus on the physical and enzymatic stepsrequired to convert starch to sugars that areutilizable by yeast.

Carbohydrate chemistry: the basics

When most people think of carbohydrates, theythink of sugar or starch. Carbohydrates are thuscalled because they are essentially hydrates ofcarbon. They are composed of carbon and waterand most may be represented by the generalformula (CH2O)n. For example, the empiricalformula for D-glucose is C6H12O6, which can

also be written (CH2O)6. There are four majorclasses of carbohydrate that are of generalinterest: monosaccharides, disaccharides,oligosaccharides and polysaccharides.Monosaccharides, or simple sugars, consist of asingle polyhydroxy aldehyde or ketone unit. Themost abundant monosaccharide in nature is, ofcourse, the 6-carbon sugar D-glucose.

Disaccharides encompass sugars such assucrose, which is an important transportcarbohydrate in plants. Lactose, a disaccharideof glucose and galactose, is commonly calledmilk sugar. Oligosaccharides (Greek oligos,‘few’) are made up of short chains ofmonosaccharide units joined together bycovalent bonds. Most oligosaccharides in naturedo not occur as free entities but are joined asside chains to polypeptides in glycoproteins andproteoglycans. Polysaccharides consist of longchains of hundreds or even thousands of linkedmonosaccharide units and, as will be discussed,exist in either linear or branched chain form.Knowledge of the terminology associated withthe organizational structure of monosaccharides,such as D-glucose, is fundamental to anunderstanding of how polysaccharides differ andhow polysaccharide-degrading enzymes, suchas amylases, function.

Monosaccharide structures are often drawn asopen Fischer projection formulae in which thesugar is depicted in either its D- or L-configuration. Most naturally occurring sugarsare D-isomers, but D- and L-sugars are mirrorimages of each other. The D or L designationrefers to whether the asymmetric carbon farthestfrom the aldehyde or keto group points to theright or the left as shown for D-glucose and L-glucose (Figure 2).

The ring or cyclic forms of monosaccharidesare most commonly drawn as Haworth projectionformulae, as shown for glucose in Figure 3.Although the edge of the ring nearest the readeris usually represented by bold lines, the six-membered pyranose ring is not planar, asHaworth projections suggest. More importantly,however, the numbering system for the carbonatoms in the glucose structure should becarefully noted. Carbon atom number one, alsoknown as the anomeric carbon atom, is the firstcarbon encountered when going in a clockwisedirection from the oxygen atom in the ringstructure, and the other carbon atoms, C2 to C6,

26 R.F. Power

are numbered as indicated in Figure 3. Whenthe hydroxyl or OH group at the anomericcarbon atom extends below the ring structure ofD-glucose, the sugar is stated to be in the α(alpha) configuration. When the hydroxyl groupat C1 extends above the ring, the sugar is in theß (beta)-configuration. These seemingly minorstructural changes represent the major differencebetween key natural polysaccharides such asstarch and cellulose.

C

C

C

C

C

CH2OH

OH

H

OH

OH

H

HO

H

H

H O1

2

3

4

5

6

D-glucose

C

C

C

C

C

CH2OH

H

OH

H

H

HO

H

HO

HO

H O1

2

3

4

5

6

L-glucose

Figure 2. Fischer projection formulae for D- andL- glucose.

STARCH

Native starch is a large polymer made up of D-glucose molecules linked together to form alinear polymer called amylose or a branchedpolymer called amylopectin (Figure 4).

In amylose, the glucose subunits are linkedtogether by bonds between the anomeric carbonof one subunit and carbon number 4 of theneighboring subunit. Because the OH group at

the anomeric carbon is below the apparent planeof the glucose subunit, the glucosidic linkageformed is stated to be in the α-configuration andthe bond is termed α(1→4). Amylose chainstypically contain between 500 and 2,000 glucosesubunits (Lehninger, 1982).

In amylopectin, glucose subunits are alsolinked in α(1→4) configuration in the linearparts of the polysaccharide but branch pointsoccur approximately once every 25 glucose unitswhere the terminal glucose in a parallel amylosechain is involved in an α(1→6) linkage. Thesebranch points are resistant to hydrolysis by manyamylases. Some amylases are totally incapableof hydrolyzing the α(1→6) bonds while othershydrolyze the 1→6 bonds more slowly than the1→4 bonds.

Glycogen is a very closely related structure toamylopectin and serves as the glucose storagepolysaccharide in animals. The main differenceis that the branch points in glycogen occur morefrequently, approximately once every 12 glucoseunits. Cellulose is a linear, unbranchedhomopolysaccharide of 10,000 or more D-glucose units linked by 1→4 glucosidic bonds.It might be expected, therefore, to closelyresemble amylose and the linear chains ofglycogen, but there is a critical difference: incellulose the 1→4 linkages are in the ß-configuration (Figure 5). This apparently trivialdifference in the structures of starch and celluloseresults in polysaccharides having widelydifferent properties. Because of their ß-linkages,the D-glucose chains in cellulose adopt anextended conformation and undergo extensiveside by side aggregation into extremelyinsoluble fibrils. Since no enzyme capable ofhydrolyzing cellulose is secreted by the intestinaltract of vertebrates, cellulose cannot easily be

CH2OH

O

OH

H

H

OH

H H

OHOH

β-D-glucose

CH2OH

O

OH

H

H

OH

H OH

HOH

α-D-glucose

1

23

4

5

6

1

23

4

5

6

Figure 3. Haworth projections of α-D-glucose and ß-D-glucose.


Figure 4. Structure of amylose and amylopectin; the polysaccharides that comprise starch.

Figure 5. Structure of cellulose showing ß(1→4) linkages between D-glucose units.

OH

OOH

OH

CH2OH

H

H

HH O

Amylose

OH

OH

CH2OH

H

H

HH O

O

OH

OH

CH2OH

H

H

HH O

O

OH

OOH

CH2OH

H

H

HH O

α-(1-4) linkage

Amylopectin

O

H

O

OH

OH

CH2OH

H

H

H

H

O

O

OH

OH

CH2OH

H

H

H

O

OH

OOH

CH2OH

H

H

HH O

OH

OH

CH2OH

H

H

HH O

O

OH

OH

CH2OH

H

H

HH O

O

OH

OOH

CH2

H

H

HH O

O

α-(1-6) linkage

Cellulose (β 1-4 linkages)

OH

O

OHH

H

H

OH

OHO

CH2OH

H

H

H

HO

O

H O

CH2OH OH

OH

CH2OH

H

H

H

HO

O

H

OH

OH

CH2OH

H

H

OO

OH

OH

CH2OH

H

H

H

OOH

H

H

H

H

digested and its D-glucose units remainunavailable as food to most non-ruminantanimals.

In contrast, the α-linkage of starch allows themolecule to be more flexible, taking on a helicalconfiguration in solution which is stabilized byhydrogen bonding. Furthermore, thebottlebrush-like structure of amylopectin lendsitself to less densely packed associationsbetween adjacent starch molecules and furtherenhances water solubility and access by starch-degrading enzymes. The ratio of amylose toamylopectin is a characteristic of each starch

source and has an impact on its gelatinizationproperties (see below). For example, four classesof corn starch exist. Common corn starchcontains approximately 25% amylose, whilewaxy maize is almost totally comprised ofamylopectin. The two remaining corn starchesare high-amylose varieties with amylose contentsranging from 50 to 75%. Other typical amylosecontents are 20% for potato, 17% for tapioca and25% for wheat. Common rice starch has anamylose:amylopectin ratio of approximately20:80 while waxy rice starch contains only about2% amylose (Table 1).

28 R.F. Power

Table 1. Ratio of amylose to amylopectin (percent oftotal starch) in various starch sources.

Starch source Amylose (%) Amylopectin (%)

Common corn 25 75Waxy maize 1 99High amylose corn 50-75 25-50Potato 20 80Rice 20 80Waxy rice 2 98Tapioca/cassava/manioc 17 83Wheat 25 75Sorghum 25 75Waxy sorghum <1 >99Heterowaxy sorghum <20 >80

In plants, starch occurs in the form of minute,discrete granules which are visible micro-scopically. Their size varies depending on starchsource. At room temperature, starch granules arequite resistant to penetration by both water andhydrolytic enzymes, due to the formation ofhydrogen bonds within the same molecule andbetween neighboring molecules. However, theseintra- and inter-molecular hydrogen bonds canbe weakened as the temperature is increased. Infact, when an aqueous suspension of starch isheated, the hydrogen bonds rupture, water isabsorbed and the starch granules swell. Thisprocess is termed gelatinization because theresulting solution has a highly viscous,gelatinous consistency. This is the same processlong employed to thicken broth in foodpreparation.

The exact temperature required forgelatinization to occur depends on a variety offactors, including granule size and structure andthe aforementioned amylose to amylopectinratio. For example, tapioca starch has a lowergelatinization temperature than rice starch,although both have similar amylose contents.This is because tapioca starch granules are muchlarger than rice starch granules and, as a result,swell more easily. Simply put, the larger thegranule, the less molecular bonding is presentand thus the granules swell faster in the presenceof water and elevated temperatures. Waxy cornand common corn starch both have the samegranule size but waxy corn will swell to a greaterdegree and each will gelatinize at differenttemperatures. This is due to their differentamylose:amylopectin ratios. One can visualize

that amylose molecules, because of their linearstructure, will lie in closer proximity to eachother than the brush-like amylopectin molecules.More extensive inter-molecular hydrogenbonding will occur between the closely bunchedamylose polymers in high-amylose contentstarch sources than will occur in high-amylopectin sources. Consequently, it requiresmore energy to break the hydrogen bonds inhigh-amylose starch sources, which results inhigher gelatinization temperatures.

Table 2. Typical temperature ranges for thegelatinization of various cereal starches.

Cereal starch source Gelatinization range (oC)

Barley 52-59Wheat 58-64Rye 57-70Corn (maize) 62-72High amylose corn 67->80Rice 68-77Sorghum 68-77

Kelsall and Lyons, 1999

Gelatinization must occur in order for the starch-degrading amylase enzymes to begin the processof converting this plant storage polysaccharideto fermentable sugars.

Enzymatic hydrolysis of starch

Although the use of industrial enzymes for starchhydrolysis did not become widespread until the1960s, Payen and Persoz reported the reactionsinvolved some 130 years earlier (Payen andPersoz, 1833). The enzyme used for the initialdextrinization or liquefaction of starch is α-amylase. The more scientifically correct orsystematic name for this type of enzyme is 1,4-α-D-glucan glucanohydrolase and it is alsodescribed by the following Enzyme Commission(EC) number, as laid down by the InternationalUnion of Biochemistry (IUB): 3.2.1.1. Thisnumber denotes that it is a hydrolase (3) of theglycosidase subcategory (3.2), which hydrolyzesO- glycosyl linkages (3.2.1). Because it is thefirst enzyme appearing in this category, itscomplete IUB number is 3.2.1.1. When faced


is effected by a class of amylases calledglucoamylases. Also included in Figure 6 is adepiction of the mode of action of pullulanases(α-dextrin 6-glucanohydrolase; EC No.3.2.1.41), which belong to a category of starch-degrading enzymes known as debranchingenzymes. The action of pullulanases concernsthe hydrolysis of 1,6-α-D-glucosidic linkagesin amylopectin, glycogen and their nascent limitdextrins, generated by α-amylase activity. Theapplication of dedicated pullulanases in theethanol industry is not widely encountered,although commercial preparations are available.This is due both to economic considerations andthe fact that, in addition to a predominantly exo-acting α(1→4) hydrolytic activity, glucoamylases(see below) can also hydrolyze 1→6 bonds to alimited degree. Given that the use of aglucoamylase component is absolutely requiredfor efficient saccharification to occur,glucoamylase inclusion is normally relied uponto perform the dual functions of anamyloglucosidase and a debranching enzyme.Nevertheless, pullulanases deserve to bementioned as a category of enzymes in their ownright because they are likely to become muchmore affordable in the coming years.

In the beverage alcohol industry, ß-amylase(1,4-α-D-glucan maltohydrolase; 3.2.1.2) isanother enzyme encountered in the starchconversion process. Like α-amylase, thisenzyme cleaves α(1→4) linkages but attacksstarch in an ‘exo’ rather than an ‘endo’ fashion.The enzyme cleaves maltose (a disaccharide ofglucose) in a stepwise manner from the non-reducing end of the starch polymer. The enzymecannot bypass α(1→6) branch points to attacklinear 1→4 bonds on the other side andgenerates ß-limit dextrins as a result. Hence, theenzyme is most effective when used inconjunction with a debranching enzyme. ß-amylases are also utilized in the syrup industryfor the production of high-maltose syrups fromstarch (McCleary, 1986). This enzymerepresents a good example of the confusion thatsometimes arises in enzyme terminology. Onecould easily be forgiven for questioning how aß-amylase can act upon α(1→4) linkages. Themain reason for this confusion is due to the factthat enzyme nomenclature is based upon theconfiguration of the released product rather thanthat of the bond being hydrolyzed. As ß-amylase

with the IUB-designated nomenclature forenzymes, the value of even a modicum ofknowledge in the area of carbohydrate linkageterminology becomes apparent. For example, inthe case of α-amylase, one can tell from itssystematic name that it hydrolyzes 1→4 bondslinking α-D-glucose residues. A furthercharacteristic of α-amylases is that they areendo-acting enzymes, meaning that they attackthe starch polymer from within the chain of linkedglucose residues rather than from the ends. Theyrandomly cleave internal α(1→4) bonds to yieldshorter, water-soluble, oligosaccharide chainscalled dextrins, which are also liberated in theα-configuration. α-Amylases are metallo-enzymes and require calcium ions to be presentfor maximum activity and stability. Finally, α-amylases cannot cleave α(1→6) bonds andbypass the branch points in amylopectin. Whenthis occurs, the residual products are called α-limit dextrins.

Commercial sources of α-amylase areproduced mainly by Bacillus species, forexample, Bacillus amyloliquefaciens and B.licheniformis. Because cooking conditions forstarch will be dealt with in detail elsewhere, it issufficient to point out here that the choice of α-amylase is based principally on tolerance to hightemperatures and that this varies quite widelyamong enzyme sources. One common approachis to cook the starch at approximately 105oC inthe presence of a thermostable α-amylase,followed by a continued liquefaction stage at90-95oC. The maximum extent of hydrolysis,or dextrose equivalence (DE), obtainable usingbacterial α-amylases is around 40, but careshould be exercised not to overdose or prolongtreatment since this can lead to the formation ofmaltulose (4-α-D-glucopyranosyl-D-fructose),which is resistant to hydrolysis by α-amylasesand glucoamylases. It must be remembered thatthe principal function of liquefaction is to reducethe viscosity of the gelatinized starch, therebyrendering it more manageable for subsequentprocessing.

The general action of amylases on starch isshown in Figure 6, where the first phase ofenzymatic starch degradation, i.e. the productionof both dextrins and α-limit dextrins from starch,is schematically depicted. As stated earlier, thenext stage is termed the true saccharificationphase, which under most practical circumstances

30 R.F. Power

catalyzes the removal of maltose units fromstarch, it produces what is known as a Waldeninversion of the OH group at C-1, giving rise toß-maltose (Belitz and Grosch, 1999). Thus, theproducts of bacterial and fungal α-amylases areall in the α-configuration, while the products ofß-amylases exist in the ß-configuration.

As mentioned, the true saccharification orconversion of dextrins to glucose is conducted,in the vast majority of cases, by glucoamylase,although other enzymes, including some α-amylases, have significant saccharifying activityin their own right. Glucoamylase is also knownas amyloglucosidase (glucan 1,4 –α-glucosidase;EC No. 3.2.1.3) and its main activity is thehydrolysis of terminal 1,4-linked α-D-glucoseresidues successively from the non-reducing

ends of dextrins causing the release of ß-D-glucose. In other words, it is an exo-actingenzyme that cleaves single molecules of glucosein a stepwise manner from one end of the starchmolecule or dextrin. Reference has also beenmade to the fact that glucoamylases canhydrolyze α(1→6) bonds to a certain extent.However, these branches are cleaved at a rateapproximately 20 to 30 times slower than thecleavage of α(1→4) bonds by the enzyme. Extraglucoamylase can be added to compensate forthis slower rate but this can cause undesirableside reactions to take place, whereby glucosemolecules repolymerise in a reaction termedreversion, forming isomaltose and causing adecrease in glucose yield.

Glucoamylases are isolated from fungal

Figure 6. Schematic representation of the action of amylases on starch.

LIQUEFACTION

Production of dextrins

and α-limit dextrins

by α-amylase

AMYLOSE

AMYLOPECTIN

DEXTRINS

α-LIMIT DEXTRINS

SACCHARIFICATION

Hydrolysis of dextrins to

fermentable sugars

GELATINIZATION

Heat and moisture solubilize starch

α-amylase

endoenzyme

hydrolyzes random α(1-4) bonds

α-amylase

α-amylase

α-amylase

Pullulanase

(debranching)

Glucoamylase

α(1-4), α(1-6 ) bonds

Glucoamylase

STARCHamylose + amylopectin

Slurry

Cooking

FERMENTABLE SUGARS Glucoamylase

exoenzyme

hydrolyzes α(1-4) bonds from non-reducing ends

hydrolyzes α(1-6) bonds more slowly


sources such as Aspergillus niger and Rhizopussp . Fungal enzymes, by nature, are lessthermotolerant than their bacterial counterpartsand therefore the temperature maxima forglucoamylases tend to be in the region of 60oCwhile their pH optima normally lie between pH4.0 and 4.5. These conditions of temperature andpH can cause significant problems in plantsusing dedicated saccharification tanks becausethey are quite conducive to the growth of certainbacterial contaminants. The claimed benefit ofpre-saccharification of mash prior tofermentation is that a high level of glucose ismade available at the start of fermentation.However, in addition to infection problems, suchan approach can lead to the generation of non-fermentable reversion products, such asisomaltose.

Because of the drawbacks of mash pre-saccharification, a popular approach is tosaccharify simultaneously with fermentation(SSF or Simultaneous Saccharification andFermentation). The requirement for high glucoseconcentrations at the start of fermentation is metby the addition of glucoamylase during thefilling of the fermentor. Although the concernwith such a system is that glucose may becomelimiting during the fermentation, this is not aproblem in practice and is merely a case ofmetering in the correct amount of enzyme.Glucose is used up as it is produced and thusthere is little risk of reversion reactions occurring.An additional attraction to this approach is thatit lends itself to the use of somewhat non-conventional enzymes, which can boost ethanol

yield by liberating fermentable sugars from non-starch polysaccharide sources. For example,multi-enzyme complexes isolated from fungalcultures grown on fiber-rich, semi-solid orsurface culture systems, contain a mixture ofamylolytic, cellulolytic and otherpolysaccharidase activities which havetemperature and pH optima that quite closelymatch those found in distillery fermentors. Onecommercial preparation, Rhizozyme™, is asurface culture enzyme gaining increasing usein the fuel ethanol industry. These enzymessatisfy the requirement to saccharify dextrins toglucose but have the additional activitiesrequired to attack, for example, ß-linked glucosepolymers, if present.

Summary

The objective of industrial alcohol productionis to produce the highest ethanol yield possible.This requires the greatest attainable conversionof starch to fermentable sugars because yeastcannot use starch in its native state. The use ofstarch-degrading enzymes represented one of thefirst large-scale industrial applications ofmicrobial enzymes. Traditionally, two mainenzymes catalyze the conversion of starch toglucose. In the first stage of starch conversion,α-amylase randomly cleaves the large α-1,4-linked glucose polymers into shorteroligosaccharides at a high reaction temperature.This phase is called liquefaction and is carriedout by bacterial enzymes. In the next stage,

Table 3. Enzymes used in the starch hydrolysis process.

Enzyme EC number Source Action

α-amylase 3.2.1.1 Bacillus sp. Only α-1,4-linkages are cleaved to produce α-dextrins, maltoseand oligosaccharides G(3) or higher

Aspergillus oryzae, A. niger Only α-1,4-linkages are cleaved producing α-dextrins, maltose andG(3) oligosaccharides

α-amylase 3.2.1.1 Bacillus amylosacchariticus Cleaves α-1,4-linkages yielding α-dextrins, maltose, G(3), G(4)(saccharifying) and up to 50% (w/w) glucose.

ß-amylase 3.2.1.2 Malted barley α-1,4-linkages cleaved, from non-reducing ends, to yield limitdextrins and ß-maltose.

Glucoamylase 3.2.1.3 Aspergillus sp. Rhizopus sp. α-1,4 and α-1,6-linkages are cleaved, from the non-reducing ends,to yield ß-glucose.

Pullulanase 3.2.1.41 B. acidopullulyticus Only α-1,6-linkages are cleaved to yield straight chain dextrins.

32 R.F. Power

called saccharification, glucoamylasehydrolyzes the oligosaccharides or dextrins intofree glucose subunits. Fungal enzymes arenormally employed to effect saccharification andthese operate at a lower pH and temperaturerange than α-amylase. Depending on theindustry and process involved, one sometimesencounters additional enzymes, such aspullulanase, which is added as a debranchingenzyme to improve the glucose yield.

Despite the intricate nature of enzymenomenclature and the rather strict reactionconditions which must be adhered to in order toachieve maximal enzyme performance, theirpurpose is very simple: to provide fuel to yeastfor the biosynthesis of ethanol. This goal mustnot be lost sight of and new approaches,enzymatic or otherwise, need to be explored inan effort to liberate additional fermentablesugars from common cereal sources. Theapplication of surface culture and fungi thatproduce multi-enzyme complexes, encompassingseveral polysaccharidase activities in additionto the requisite, traditional, amylolytic activities,represents one promising approach towardsmaximizing the release of utilizablecarbohydrate from both starch- and non-starch-based cereal polysaccharides.

References

Belitz, H.-D. and W. Grosch. 1999. FoodChemistry. (M.M. Burghagen, D. Hadziyev, P.

Hessel, S. Jordan and C. Sprinz, eds). Springer-Verlag, Berlin Heidelberg. pp. 237-318.

Kelsall, D.R. and T.P. Lyons. 1999. Grain drymilling and cooking for alcohol production:designing for 23% ethanol and maximumyield. In: The Alcohol Textbook, 3rd ed.(K.A.Jacques, T.P. Lyons and D.R. Kelsall, eds).Nottingham University Press, UK. pp. 7-24.

Lehninger, A.L. 1982. Principles ofBiochemistry. (S. Anderson and J. Fox, eds).Worth Publishers Inc. New York. pp. 277-298.

McAloon, A., Taylor, F., Yee, W. Ibsen, K. andR. Wooley. 2000. Determining the cost ofproducing ethanol from corn starch andlignocellulosic feedstocks: A Joint StudySponsored by U.S. Department of Agricultureand U.S. Department of Energy. NationalRenewable Energy Laboratory TechnicalReport/ TP-580-28893.

McCleary, B.V. 1986. Enzymatic modificationof plant polysaccharides. Int. J. Biol.Macromol. 8:349-354.

Payen, A. and J.F. Persoz. 1833. Mémoire sur ladiastase, les principaux produits de sesréactions, et leurs applications aux artsindustriels. Ann. Chim. Phys. 53: 73-92.

Grain handling: a critical aspect of distillery operation 33

Chapter 4

Grain handling: a critical aspect of distillery operation

DAVID J. RADZANOWSKI


Introduction

Until recently, sanitation in grain handling hasbeen basically ignored by the distilling industry,except for those engaged in beverage alcoholproduction. Now that we have begun to discussand accept the ‘Biorefinery’ concept, we areforced to realize that a good sanitation programand procedures are essential to the bottom line.Outbreaks of food animal diseases such as BSEand food-borne diseases such salmonella, as wellas prohibitions against mycotoxins andinsecticides in grains going into food animalsand ultimately the human food chain, havegrabbed our attention. We suddenly realize thatdistillers dried grains with solubles (DDGS) is avaluable co-product and not just an unavoidablenuisance. Also, if we cannot sell it because ofcontamination, we are out of business.

We first became aware of the importance of asanitation program in grain handling systemssome years ago in the brewing industry.Breweries by definition prepare a food product,beer, and are subject to all possible scrutiny bythe Food and Drug Administration (FDA) andrelated state and local food safety bureaus.Breweries routinely underwent very involvedinspections for food contamination and lookedforward to them with much dread. In the 1970s,a group of federal inspectors appeared at oneplant and after providing the proper credentials

and the warrants necessary for the inspection,surprised the staff by asking only for a sampleof brewer’s spent grains. The staff relaxed andgladly obtained a sample of spent grains underthe inspectors’ directions.

Two weeks later the plant was upset with anotice of citation for improper sanitationprocedures since insect fragments, mold,mycotoxins and insecticide residuals were foundin those spent grains. Since animal feed lawswere less stringent in the 1970s than today, theplant was allowed to continue selling the spentgrains. The agency just used the contaminationdetected as a means of bringing attention to aninadequate sanitation program as established forbreweries under the Federal Food, Drug andCosmetic Act.

Good manufacturing practices

Since 1969, the Act has contained guidelinescalled the Good Manufacturing Practices or theGMPs (FDA, 2000). This Amendment to the Actis titled ‘Human Foods, Current GoodManufacturing Practice (Sanitation) inManufacture, Processing, Packing or Holding’.Sections 3 through 8 of the amendment applyin determining whether the facilities, methods,

34 D.J. Radzanowski

practices and controls used in the manufacture,processing, packing or holding of food are inconformance with and are operated oradministered in conformity with goodmanufacturing practices to produce undersanitary conditions, food for humanconsumption.

Knowledge of the GMPs, with an eye towardmeeting the outlined standards, is a valuableguide toward setting up an operation thatfacilitates the proper handling of grain. Section3 of the amendment provides guidelines for thecorrect positioning of the plant and the care ofthe grounds. It goes further to detail theappropriate construction of the facility utilizingcleanable surfaces, adequate lighting, ventilation,employee facilities and screening and protectionfrom pests. Section 4 contains the requirementsfor the equipment and utensils. All must besuitable for the intended use, durable, cleanableand kept in good repair.

Sections 5 and 6 are closely related. They dealwith the needs for adequate water, plumbing andsewage disposal, toilet and hand washingfacilities for employees, and the proper disposalof offal and rubbish. The requirements forgeneral maintenance, pest control and sanitizingof equipment and utensils are described in detail.

To ensure that food will not be contaminated,Section 7 requires the inspection of all rawmaterials and ingredients as well as the inspectionof the containers and the common carriersdelivering the goods. This section also requiresfrequent cleaning and sanitization, processingunder conditions to minimize bacterial andmicro-organic growth, chemical and biologicaltesting and the maintenance and retention ofrecords.

Under Section 8, plant management is requiredto assure that employees are free from disease,observe good cleanliness practices in clothingas well as personal habits and that they areproperly trained and educated to perform theirduties in a safe and sanitary manner.

STRICTER REQUIREMENTS LIKELY

Distillers in the United States, as well as in othercountries, do not normally come under thejurisdiction of the FDA or related organizations.They are normally governed by bureaus like theUS Bureau of Alcohol, Tobacco and Firearms

(BATF), a regulatory agency that imposes andcollects tax monies on the ethanol generated bya distiller. Recent changes in the US with thecreation of the Homeland Security Act have splitup the BATF and reduced the strength of theorganization pertaining to alcohol producerenforcement. This leaves a gap that the FDAseems to be filling, slowly but forcefully. Whatdoes this mean to the distiller? In all probability,stricter requirements demanding reducedcontamination in co-products in animal feedssuch as DDGS will be established. We havealready seen some activity from other countriesrejecting DDGS generated in the US due to thecontent of mycotoxins and other unwantedcontamination. How can we help preventrejection or seizure of this valuable by-product?Adopting strict sanitation programs is an aid todesigning or remodeling plants to conform tothe guidelines needed.

Where do we start?

William Schoenherr, formerly of the LauhoffGrain Company, which is now part of BungeMilling is considered the father of sanitation ingrain handling and storage facilities as well asin breweries. Schoenherr started as an FDAinspector and brought his enormous expertiseto the grain industry. Lauhoff recognized that ithad obtained a great asset and allowed Bill toshare his expertise with the grain handling worldand particularly their customers. Schoenherralways talked about a number of key factors thatmust be considered in a good sanitation program.We will use these as guidelines for the followingdiscussion.

FORMS OF CONTAMINANTS, DETECTION ANDREMOVAL

First, it is important to be aware of the differentforms of contamination that may be present inthe raw materials brought in for processing.Second, everything possible must be done toensure detection of such contamination beforeit enters the facility. And, third, everythingpossible must be done to avoid, remove ordestroy the contaminants before using these rawmaterials. These three are very closely related.

What kind of contaminants can we expect?


Among others, insects, rodents and birds andtheir droppings, stone, cobs, weed seeds, glass,wood, rags, tramp metal, residual insecticidesand fumigants, water, bacteria, mold andmycotoxins. Some of these are from uncleangrain handling and transport. Others have beenadded as fillers, adding weight withoutproducing usable substrate. Some, like stone andtramp metal, are dangerous and may cause firesand explosions in the grain handling system. Themore difficult contaminants to detect such asresidual insecticides and fumigants as well asmycotoxins, can be present at levels that canresult in unmarketable DDGS.

Most operations are located close to the sourceof the substrate used for fermentation, so mostof the material is received in trucks with only alimited amount of rail transit. Samples aregenerally obtained by probing each truck in twoplaces, front and back, and then composited.Statistically, this is not sufficient. Two poundsof sample from a 50,000-pound load is not veryindicative of the condition of the contents of thetruck.

Bushel weight, moisture, damaged kernel andbroken kernel/foreign material tests are runimmediately. High insect infestation may bedetermined by visual inspection, but lowinfestations may be more difficult to find. Also,the true grain weevils, because they bore intokernels, may not be visible on the surface of thegrain. Only a few plants use a black light todetect aflatoxin presence. Black light is alsousable for the detection of rodent urinecontamination. This should be a routine test forall shipments in all plants. NIR technology offersa solution to many of the grain testing needs forquick decision making by distillers. That subjectis covered in a separate chapter in this volume.

Do not neglect the senses. Examine the grain,checking for insect presence, rodent droppings,bird feathers as well as foreign seeds andextraneous matter such as cobs, straw, wood andmetal. Smell the material for any hint of moldpresence or chemical contamination. Feel thegrain for moisture and slime presence.

Most plants use the specifications for No. 2corn as a guideline, which call for a bushelweight of 56 pounds, a limit of 0.2% heatdamaged kernels, 5% total damaged kernels and3% total broken kernel/foreign materialpresence. A recent discussion with a group of

farmers supplying corn to some of the newestfuel ethanol plants coming on stream was veryrevealing concerning the state of the corn marketin some areas of the country. The farmers werequite happy that the new distillers were payingthe same price for No. 2, 3, 4 and 5 grain asthey are for No. 1. This seems not only wastefulon the part of the distiller, but potentiallydangerous. The more allowable extraneouscontent in the lower grade incoming grain, thegreater the contribution to lower yields and thehigher likelihood of gross contamination andpossible ignition of fires and explosions.

High moisture is the most common reason forrejection of a shipment. This is a good start. Highmoisture contributes to difficulties in milling andhandling and to lower yields. High moisture alsocontributes to elevated mold growth, whichtypically leads to contamination with mycotoxinsthat become concentrated in the DDGS.

First, decide on specifications for acceptanceof the incoming substrate. Next, provide thepeople and the testing equipment to determineif the substrate meets those specifications. Stickto the established specifications. Only whensatisfied should you begin to unload the materialinto the plant’s infrastructure.

RECEIPT AND STORAGE

Specifications should call for all shipments tobe received in covered hopper railroad cars andhopper trucks fully covered by tarpaulin or someother means. This will help prevent appearanceof rodents and other pests while the shipment isin transit.

Once the shipment is accepted, it can thenproceed to the unloading area. Most people dropthe contents of the railcar or truck into a pit fromwhich the material is conveyed in some mannerto the storage bin. It is important that this stationbe constructed in a manner that allows it to bekept in good sanitary condition without mucheffort. Pay attention to the grid covering the pit.Some grids have wide-spaced bars that couldallow whole rats to pass through to the pit andonward to the bin. This is unacceptable. The gridsystem should be designed for the rapid flow ofthe grain and nothing else.

There are many different styles of grainconveying systems, including bucket elevators,

36 D.J. Radzanowski

positive and negative pneumatic systems, beltconveyors and the newer pulse pneumaticsystems. The most user unfriendly anddangerous system seems to be the bucketelevator. This is due to the high number of partswith a greater maintenance requirement and thehigher number of possible ignition points.Manufacturers have attempted to alleviate someof this by the introduction of plastic buckets,but the hangers and chains are still, for the mostpart, made of metal. Bucket elevators alsoprovide more harborages for insects and areharder to clean and fumigate. Pneumatic systemstend to be abraded by the impingement of thegrain on the tubing system, but a well-designedand balanced system can extend the lifeexpectancy of the tubes indefinitely.

Grain should ideally be cleaned on the way tothe bin. Why store dust, straw, cobs, wood, metal,dead rats and other extraneous material?Unfortunately, most of the systems we see aresimilar to Figure 1.

The facility illustrated in Figure 1 is typical in

breweries since the brewers are not reallyreceiving raw materials, but barley that has beencleaned and processed by malting. For brewerssuch a system works well. The earlier fuelethanol plants had systems like this installedbecause the engineers had been designing forbreweries for a number of years and did notrecognize that distillers are involved in handlinga different type of material. Distillers havebecome aware of this and are requesting thatcleaning systems be placed ahead of the bins.

Keep records of the amount of debris removed,as you do not want to pay for trash. Keep a recordof the net amount of clean substrate entering thestorage bin. This figure will be helpful later indetermining true yield.

Bins should be constructed of concrete or metaland be well sealed. We have seen binsconstructed of corrugated metal sheets thatleaked so badly that the grain was beginning toferment in the storage bin after a very lightrainfall. One of the most important design criteriafor grain bins is their ability to be fully emptied.

Receiver

Airlock

Venthopper

Rotary feeder

Exhaust

Bin selector valve

Receiver

Airlock

Exhaust

Blower

Cleaner

Mill

Surgehopper

Rail car

Figure 1. Typical configuration of a grain storage receiving system.


We have seen too many storage bins atdistilleries that hold 3-5 days of material thatcannot be totally emptied. This can be verydangerous and conducive to infestation. Moldgrowing in stationary grain simply contaminatessuccessive loads.

One of the great tools for protection againstinsect infestation is the understanding of the lifecycles of the various insect pests. Interruptingthat life cycle prevents an infestation frombecoming established. A 3-5 day storage limitof grain in a bin will interrupt the life cycle ofthe common grain insects, but the bin must betotally emptied each time it is filled. If the designdoes not allow total emptying, there will be acontinuous infestation.

STORAGE CAPACITY

How large should storage capacity be? Financialand logistic experts must contribute to thisdecision. Storage capacity must accommodateboth the processing schedule and cash flowconsideration. If running 24-hrs a day, sevendays a week and 50 weeks of the year, theneither storage capacity or local sources ofsubstrate must keep the plant supplied. Localsupplies reduce storage needs, so that capacitymay be needed only to maintain supplies throughweekends and holidays. If shipments must comefrom afar, more capacity is needed. This is wherethe logistics advice is needed.

The sanitation program

Sanitation programs must be adjusted to the typeof facility. Very few large operations today arecontained within a structure. Most are of an openair design. The enclosed plant has the sanitationadvantage of controlled entry and egress,making it easier to exclude rodents and birds.Regardless of design, full inspection andcleaning techniques must not be relaxed. Spillsmust be cleaned up quickly to prevent attractingpests. Scrape, shovel and vacuum the debris.Never use an air hose since the subsequentdispersal just spreads infestation and infection.

Any inspection and cleaning program shouldbe familiar with storage insect life cycles andtake advantage of this knowledge. Fullinspections of the total operation should be made

every two weeks and inspections of the morecritical zones should take place on a weeklybasis. Any evidence of an unwanted invasionshould be dealt with immediately. The cleaningschedule should also be devised with the insectlife cycle in mind. Certain areas of the plant, likethe receiving station and subsequent conveyors,dust collectors and air conveyor systems requiremore frequent inspection, cleaning and possiblyfumigation on a very regular basis. The programshould determine the risk and set the appropriateschedule.

A pest control program can be dangerous inthe hands of unqualified people. Bait stations,insecticides, and fumigants are poisons andharmful to humans, animals and pets as well asthe pests they are intended to control. Thesecompounds are handled and applied by licensedindividuals, however it is necessary for theperson in the company in charge of the sanitationprogram to have training and knowledge of pestcontrol operations in order to evaluate the workof the hired contractors.

With the direction and agreement ofmanagement, establish the program. Determinethe inspection schedule. Hire the pest controlfirm. Determine the schedule for application ofthe tools and chemicals of the pest controloperation. Determine the cleaning schedule.Keep complete records. Record keeping cannotbe stressed too strongly. Obtain all reports fromthe contract pest control operation including thedates of application, the places of application,the chemicals used in the application, the strengthof the materials used, the technician performingthe service and the records and results of theirinspections as well as the results of the service.Review the work of the hired contractor foreffectiveness. Review the program and makeappropriate changes as they are necessary.

MILLING

When grain leaves the bin on the way to the mill,it should pass through de-stoners to protect themills from damage and magnets to preventsparks and explosions. Dry milling operationsshould also incorporate explosion dampers in thesystem. This brings us to the question of whattype of a mill to use.

Brewers and Scotch or Irish whisky distillerscan work very well with a roller mill, preferably

38 D.J. Radzanowski

with five or six rolls. The roller mills workextremely well with the soft, already modifiedand malted barley, which is the principalsubstrate of these operations. Simple adjustmentscan be made in the grind according to the needsof the wort separation systems in use, whetherthey incorporate mash filters or lauter tuns.

Table 1 contains the recommended malt sieveanalysis for the different wort separationsystems. This is based on the American Societyof Brewing Chemists (ASBC) testing sievesystem with a 100 g (0.22 lb) sample.

For most grains such as corn and othersubstrates such as cassava chips, hammer millswill probably perform the most efficiently. Theproper screen size to use in the US would benumber 6 through 8. This is the equivalent ofopenings in the range of 1/8 to 3/16 inches. InEurope, the screens would be 2 to 4 mm slotsize.

A sieve analysis should be accomplished daily(Table 2). The importance of a fine grind cannotbe overemphasized, however a very fine grindmust be accompanied by good mixing agitationto avoid formation of pills and lumps, creatinginsufficient exposure of the solids to the liquid.

A fine grind exposes more surface area,facilitating better slurry, cook and liquefyingoperations. The more surface area exposed, thegreater the effectiveness of the enzymesallowing for greater activity and more completereactions. An idea of the importance of particlesize in determining surface area exposure canbe obtained by reviewing Table 3, which usesthe explanation of the need for small bubblesfor better aeration as an example.

Table 2. Typical sieve analysis of corn meal.

Screen size Hole size (inches) Corn on screen (%)

12 0.0661 3.016 0.0469 8.020 0.0331 36.030 0.0234 20.040 0.0165 14.060 0.0098 12.0Through 60 7.0

Most distillers weigh the grist on the way to thecooking operation by the use of weigh hoppersand weigh belts. They use these figures forproportioning solids and liquid in the slurryingoperation.

Table 3. The free surface area of one inch of air when split up into bubbles of varying size1.

Bubble diameter Volume of bubble Surface area of bubble Surface area of 1 cubic inch of air = A/V(In.) V=0.5236 d3 (cubic inches) A=3.1416 d2 (square inches) (square inches)

1.0 0.5236 3.1416 60.5 0.0645 0.7854 120.1 0.0005236 0.031416 600.01 0.0000005236 0.00031416 6000.001 - - 6,0001Note: A = 3.1416 d2 = 6 = surface area

V 0.5236 d3 d

Table 1. Recommended malt sieve analysis for lauter tun and mash filter wort separation systems.

Fraction US sieve number (inches) For lauter tun (% retained) For mash filter (% retained)

Husks I 10 (0.0787) 14 6Husks II 14 (0.0555) 20 10Coarse grits I 18 (0.0394) 29 6Coarse grits II 30 (0.0234) 18 18Fine grits 60 (0.0098) 9 43Coarse flour 100 (0.0059) 4 10Powder Bottom 6 7

Adapted from ASBC, 1996


Other considerations

Full attention must be paid to sanitation in allother phases of the operation. Sanitary designof equipment and piping systems is essential.The positioning of sample ports can be conduciveto keeping a clean workspace. Containing spillsfrom sample ports by placing buckets to catchthe spills is a very commendable practice, butbuckets must be emptied and cleaned regularly.Spills from sample ports or repair work shouldbe cleaned up immediately. Besides attractingpests, they also provide good growth media forthe bacteria and wild yeast species that causecommon distillery infections. These can beeasily transferred to the vessels.

As mentioned above, open air plants have themost difficulty in the control of pests since thereis no barrier to help prevent the intrusion of theinsects, birds and rodents including bats. Besidesa good sanitation program, it is also importantthat any evidence of droppings, feathers andfragments be cleaned up immediately. The Food,Drug and Cosmetic Act standards will requirepreventing the possibility of productcontamination.

Storage areas for DDGS must also beevaluated from a sanitation perspective. Manydistillers store DDGS on the ground in covered

and sometimes uncovered areas. These areas aretotally open to attack by insects, birds, bats,rodents, dogs, cats, raccoons, possum and otheranimals. The possibility of mold and mycotoxinformation in these areas can also be devastatingto the acceptance of the DDGS in the world feedmarket, which is increasingly recognized as apart of the human food chain.

References

A Guide to Good Manufacturing Practices forthe Food Industry. 1972. Lauhoff GrainCompany, Danville, IL, USA.

ASBC. 1996. Methods of Analysis (8th Rev. Ed).American Society of Brewing Chemists. STPaul, MN, USA.

FDA 2000. Good Manufacturing Practices. 21CFR 110.

The Alcohol Textbook: A reference for thebeverage, fuel and industrial alcohol industries(3rd Ed). 1999. (K.A. Jacques, T.P. Lyons andD.R. Kelsall, eds.) Nottingham UniversityPress, UK.

The Practical Brewer: A Master BrewersAssociation of the Americas. 1999. MastersBrewers Assoc. Amer., Wauwatosa, WI, USA.

Substrates for ethanol production

Lignocellulosics to ethanol 41

Chapter 5

Lignocellulosics to ethanol: meeting ethanol demand in the future

CHARLES A. ABBAS

Director of Yeast and Renewables Research, Archer Daniels Midland, Decatur, Illinois, USA

Introduction

The US alcohol industry has the capacity toprocess about 1 billion bushels of corn perannum; which accounts for most of the fuelethanol produced domestically. This representsapproximately 10% of the expected 2003 USharvest. As production increases to the expectedfive billion gallons, the amount of corn used forethanol will also double. Can we continue to useincreasing amounts of grain? What wouldhappen if all of US gasoline demand (~150billion gallons) was replaced with ethanol? (Thishappened in Brazil in the 1970s and 1980s.)Assuming a harvest of 10 billion bushels (250million tonnes) and excellent conversionefficiency, the amount of ethanol producedwould still fall far short of the target.

Demand for ethanol is increasing at a rate thatwill require serious consideration of alternativesto the primarily starch-based feedstocks over thenext decade. Various lignocellulosic biomasssources such as agricultural residues, wood andforest wastes, municipal solid wastes and wastesfrom the pulp and paper industry have thepotential to serve as low cost and abundantfeedstocks for ethanol production (Bothast andSaha, 1997; Galbe and Zacchi, 2002; Katzen etal., 1999; Monceaux and Madson, 1999; Rogers,1997; Rogers et al., 1997; Singh and Mishra,1995; Tolan, 1999; von Sivers and Zacchi,

1996). According to a recent US Department ofEnergy (DOE) draft report, the totallignocellulosic feedstock potentially available inthe US is estimated at 623 million dry tons(Figure 1). Excluding the biomass contributionfrom energy crops and from sludge and biogas,the remaining lignocellulosics that are primarilyderived from agricultural crop residues, forestresidues and municipal-source wood/paper/organics can be estimated at a total of about 400million dry tons. Assuming that these availablelignocellulosics are converted to ethanol withyields approaching 80 gallons per ton, domesticethanol production can be expanded further byan additional estimated 32 billion gallons. Thisfigure represents up to 20% of the current USgasoline consumption, which is estimated in2003 at 150 billion gallons. This chapter servesto provide an overview of progress of researchon lignocellulosics and the opportunities andtechnical challenges to the production of ethanolfrom these renewable biomass feedstocks.

Major components of lignocellulosics

The composition of lignocellulosics variesgreatly among the major categories of plantsources (i.e. softwoods, hardwoods, straws) and

42 C.A. Abbas

also varies within each category (Tables 1 and2). In spite of these differences, the majorcomponents of all lignocellulosics are cellulose,hemicellulose, lignin and extraneous materials(Singh and Mishra, 1995). Of these, cellulose isby far the most abundant organic substrate.Cellulose is a linear homopolymer ofanhydroglucose residues that are linked via ß-O-1, 4-glycosidic bonds. Different forms ofcellulose exist with varying degrees ofpolymerization and molecular weight (Singh andMishra, 1995). These forms include a highlycrystalline form, which results from the packingof cellulose microfibrils into highly orderedfibers, as well as less ordered and amorphousforms (Lynd et al., 2002). These variations inthe crystal structure of cellulose affect itshydrolysis by the action of enzymes and/orchemicals. It is generally recognized that the

more ordered or crystalline the cellulose, the lesssoluble and less degradable it is. Unlikecellulose, hemicellulose consists of shorterheteropoly saccharide chains composed ofmixed pentosans and hexosans, which are morereadily soluble and consequently amenable toenzymatic degradation. D-xylose and L-arabinose are the major constituents of the xylanor arabinoxylan backbone of the hemicelluloseof straws with D-glucose, D-glucuronic, D-mannose and D-galactose as the other primaryconstituents of the side chains. The principalcomponent of hardwood hemicellulose isglucuroxylan; that of softwoods, glucomannan(Jeffries et al., 1994). Minor constituents, whichare ether or ester-linked methyl, acetyl, or acylgroups are also commonly encountered. The D-xylose in the xylan backbone is linked via 1,4-ß-O-glycosidic linkages. The hexosan, araban

Total feedstocks available: 623 million dry tonnes (MdT) per yearEnergy equivalent: 16 quadrillion BTU

Potential energy crops159 MdT

(primarily switchgrass, also hybrid poplar and willow)

Agricultural crop residues156 MdT

(corn stover, wheat andrice straw, cotton stalks)

Forest residues84 MdT

Primary mill residues2 MdT

(excludes portion currentlyused for fuel wood, fiberand various by-products)

Other wastes161 MdT

(unused organic fraction ofmunicipal solid waste,

construction and demolitionwaste wood, urban tree residues

Biogas11 MdT

(landfill, digesterand sewage gas)

Sludge50 MdT

(manure and biosolids)

Figure 1. Feedstocks available in the US by biomass type(from: Biobased Products and Bioenergy Roadmap, July 2001 draft).


and other minor constituents of hemicelluloseare readily hydrolyzed from the backbone bychemical means, enzymatic means, or acombination of the two. By comparison, theenzymatic hydrolysis of the xylan backboneproceeds at a much slower rate in the presenceof these constituents. The considerable variationthat is present in the hemicellulose among plantsources is responsible for the great differencesin the activity of commercial xylanase orhemicellulases on lignocellulosics.

Table 1. Distribution of cellulose and hemicellulose inwood and straw.

Substrate Xylan Araban Galactan Mannan Glucan%

Hardwood 17.4 0.5 0.8 2.5 50.1Softwood 5.7 1.0 1.4 11.2 46.3Straw 16.2 2.5 1.2 1.1 36.5

Table 2. Composition (%) of common lignocellulosicsubstrates.

Substrate Hexosans Pentosans Lignin

Bagasse 33 30 29Birch 40 33 21Corn stover 42 39 14Groundnut shells 38 36 16Oat straw 41 16 11Pine 41 10 27Rice straw 32 24 13Sawdust 55 14 21Wheat straw 30 24 18

The composition of lignin and extraneousmaterials in plant sources varies considerably(Singh and Mishra, 1995). Lignins are generallydistributed with hemicelluloses in the spaces ofintracellular microfibrils in primary andsecondary cell walls and in the middle lamellaeas cementing components to connect cells andharden the cell walls of the xylem tissues(Higuchi, 1990). Lignins have been generallyclassified into three major groups based on thechemical structure of the monomer units. Thegroups are softwood lignin, hardwood lignin andgrass lignin (Higuchi, 1990). Because the ligninand hemicellulose constituents differ, the cross-links between these polymers vary from plantto plant and tissue to tissue (Jeffries et al., 1994).

The primary function of lignin in lignocellulosicsis to provide plants with structural rigidity and italso serves as a waterproofing agent in woodytissue (Singh and Mishra, 1995). Lignin is anamorphous noncrystalline polymer composed ofhighly branched polymeric molecules ofphenylpropane-based monomeric units that arelinked by different types of bonds, which includealkyl-aryl, alkyl-alkyl and aryl-aryl ether bondssome of which are not readily hydrolyzable(Singh and Mishra, 1995). Thus, lignin isconsiderably resistant to microbial degradationin comparison to polysaccharides and othernatural polymers (Ericksson, 1981; Jeffries etal., 1994; Odier and Artaud, 1992). Lignin isacid stable and insoluble in water but can besolubilized under alkaline conditions. Inlignocellulosics the lignin is found closely boundto cellulose and hemicellulose and as a resultcannot be readily separated from thesecarbohydrate polymers using conventionalmethods (Jeffries et al., 1994; Singh and Mishra,1995). In lignocellulosics the association oflignin with cellulose and hemicellulose impedesenzymatic degradation of these carbohydratepolymers. For this reason, physical or chemicalpretreatments have been designed to disrupt thelignin-carbohydrate matrix to obtain goodenzymatic accessibility.

Extraneous materials in lignocellulosics consistof extractable and nonextractable organics suchas terpenes, phenols and resins, which includeoils, alcohols, tannins, alkaloids, gums, silicate,oxalates and proteins. In some straws and hullsnonextractables such as silica can account forover 10% of the dry weight (Singh and Mishra,1995).

Processing of lignocellulosics

Processing of lignocellulosics is an essential stepin releasing the carbohydrate components inorder to derive fermentable feedstocks for theproduction of ethanol. This goal can beaccomplished by combining pretreatment andhydrolysis steps that involve physical, chemical,thermal and/or enzymatic treatment (Bothast andSaha, 1997; Galbe and Zacchi, 2002; Katzen etal., 1999; Tolan, 1999; McMillan, 1994a, 1994b;Ramos and Saddler, 1994; Singh and Mishra,1995; Stenberg et al., 1998; von Sivers and

44 C.A. Abbas

Figure 2. Enzymatic degradation of cellulose.

Figure 3. Enzymatic degradation of hemicellulose.

CH2OH

CH2OH

OO

HOH

H

H

OH

H

OH H

H

H OH

O

OHH

H

CH2OH

CH2OH

OO

HOH

H

H

OH

H

OH

CH2OH

OO

HOH

H

H

OH

OHH

H OH

O

OHH

H

β 1-4 cellobiohydrolase

CH2OH

CH2OH

OO

HOH

H

H

OH

H

OH H

H

H OH

O

OHH

H

CH2OH

CH2OH

OO

HOH

H

H

OH

H

OH

CH2OH

OO

HOH

H

H

OH

OHH

H OH

O

OHH

H

CH2OH

CH2OH

OO

HOH

H

H

OH

H

H H H

OH

H OH

OHO

OHH

H

β−glucosidase

CH2OH

CH2OH

OO

HOH

H

H

OH

H

OH H

H

H OH

O

OHH

H

CH2OH

CH2OH

OO

H

H

H

OH

H

O

H

CH2OH

OO

HOH

H

H

OH

OHH

H OH

O

OHH

H

HO

H H

H H

β 1-4 endoglucanase

H

H

OO

HOH

H

H

O

H

OH

COOH

O

HOH

H

H

OH

H

CH3O

H

H

H OH

O

OHH

H

H

H

OO

HO

H

H

OAc

H

OH

H

OO

HOH

H

H

OH

OHH

H OH

O

OHH

H

O

OH

H

H

OH

H

HOH2C α−arabinofuranosidase

β−endoxylanase

Acetylxylan esterase

α−glucuronidase

H

OH

H

OHO

HOH

H

H

OH

OHH

H OH

O

OHH

H

β−xylosidase

Cellulose

Cellobiose

Hemicellulose

Xylobiose


Zacchi, 1996). A list of pretreatment methods isprovided in Table 3 and the reader is urged toconsult the references cited for greater detail.The enzymes involved in the hydrolysis of thepolysaccharide components of lignocellulosicsare covered in the next section of this chapterand some of their substrates and catalytic actionsites are illustrated in Figures 2 and 3. Thesuccessful integration of the pretreatment andhydrolysis steps when combined with multistepseparation and fractionation, along withsimultaneous saccharification and fermentation(SSF), can result in an environmentally friendlyprocess that yields a fermentable sugar slurrythat is nontoxic or minimally toxic to ethanolproducing microorganisms. The requirementsfor an economically viable process to meet theabove criteria are as follows:

1) a well characterized lignocellulosic feedstock

2) high solids loading by employment ofphysical means to increase surface area

3) a treatment method to hydrolyze thecarbohydrate polymers to monosaccharideswhile minimizing the formation of sugardegradation products (HMF, furfural,levulinic acid)

4) little or no by-product waste stream thatrequires waste treatment and/or disposal

5) reduced enzyme cost and usage through

proper use of enzyme loading with the useof recycling and/or the use of recombinantor naturally-occurring microorganisms thatproduce the required hydrolytic enzymes

6) good sugar conversion and fermentationyields

7) utilization of a well designed fermentationprocess that takes advantage of the currentlyavailable genetically engineered micro-organisms

8) a high ethanol concentration in the finalfermentation broth which would beeconomical to recover by conventionaldistillation or through the use of newtechnologies such as solvent extraction ormembrane pervaporation.

Based on these criteria it becomes apparent thatthe use of lignocellulosics for the production ofethanol requires significant process integration.This is best described by the employment of aholistic approach to processing referred to asintegrative bioprocessing (Abbas, 1996; Abbas,2000). Integrative bioprocessing has numerousadvantages over conventional processing in thatit combines advances in chemical engineering,bioprocessing and molecular biology to developan optimized integrated system as compared tooptimizing the individual components of aprocess separately (Figure 4).

Table 3. Physiochemical methods for pretreatment of lignocellulosics.

Chemicals Alkali Sodium hydroxide, ammonium hydroxide

Acids Dilute or concentrated sulfuric acid, dilute or concentratedhydrochloric acid, nitric, phosphoric, acetic

Oxidizing agents Peracetic acid, sodium hypochlorite, sodium chlorite, hydrogen peroxide

Solvents (Organosolv) Methanol, ethanol, butanol, phenol, ethylamine, hexamethylene-diamine, ethylene glycol

Gases Ammonia, chlorine, nitrous oxide, ozone, sulfur dioxide.

Physical/thermal Mechanical Milling, grinding, extrusion, pressing

Autohydrolysis Steam pressure, steam explosion, hydrothermolysis, steam andmechanical sheer, pyrolysis, dry heat expansion, moist heat expansion

Irradiation Gamma, electron beam, photooxidation

46 C.A. Abbas

One of the most promising feedstocks that canmeet many of the criteria listed is corn fiber hulls,a lignocellulosic by-product of the corn wetmilling process (Abbas, 1996; Abbas et al.,2002; Beery et al., 2003; Gulati et al., 1996;Moniruzzaman et al., 1996; Saha et al., 1998;Saha and Bothast, 1999; Werpy et al., 2001).Other promising feedstocks that are currentlyavailable and meet many of the criteria listedabove are agricultural residues and bioprocessingplant by-products, which include soybean hulls,wheat straw, pentosan-containing fibrous starchstreams from wheat processing, corn mesaprocessing, fibrous by-product from dry grindtortilla flour operations, sugar beet pulp fibers,rice straw, sugar cane bagasse, spent brewersgrains and dried distillers solubles from dry grindethanol plants. Beyond the above readilyavailable feedstocks, municipal wastes, paperand pulp wastes, corn stover and corncobsrepresent additional attractive feedstocks. Theconversion of these feedstocks can beeconomically feasible provided that a cost-effective process is delineated for theircollection, delivery, analysis, pretreatment,hydrolysis of the carbohydrate fractions and/orby-product disposal. The development of bothcost effective hydrolytic enzymes and highsolids pretreatment with good yields will be themajor drivers for greater utilization of this lastgroup of lignocellulosics.

Enzymes for the processing oflignocellulosics

The enzymatic depolymerization oflignocellulosics requires an extensive battery ofmicrobial hydrolytic enzymes from bacterial andfungal sources that are capable of hydrolyzingcellulose, hemicellulose and lignin (Baker et al.,1995; Ericksson, 1981; Galbe and Zacchi, 2002;Ghosh and Ghosh, 1992; Jeffries, 1994;Joseleau et al., 1994; Lynd et al., 2002; Malburget al., 1992; Nieves et al., 1997; Pelaez et al.,1995). These enzymes, which includeglycohydrolyases, esterases and ligninases, mustact on the polysaccharide backbones as well asthe branched chains and substituents attachedto lignin. Enzymatic digestion can be facilitatedby pretreatment and hydrolysis, which enhancesenzyme reaction rate by improving binding tothe solubilized polymers and consequentlyimproves process economics by reducingenzyme loading. Enzyme preparations that areknown to be effective on lignocellulosics containseveral exo- and endo-glycohydrolytic activitiesand include cellulase enzyme complexes thatdegrade cellulose and hemicellulases such as D-xylanases, L-arabinases, D-galactanases,D-mannases and esterases that release acetylgroups from branched sugar polymers or thatrelease ferulic acid or/other lignin precursors(Figures 2 and 3). While an understanding ofthe enzymatic arsenal involved in the primaryattack on lignin is still incomplete, several

1. Pretreatment1. Pretreatment 2. Hydrolysis2. Hydrolysis 3. Biocatalysts3. Biocatalysts 4. Fermentation4. Fermentation 5. Recovery5. Recovery

Conventional Processing

Integrative Processing

Figure 4. Illustration of integrative bioprocessing.


important enzymes are known to be involved(Ericksson, 1981; Ericksson, 1993; Hatakka,1994; Singh and Mishra, 1995). These enzymesinclude the extracellular lignolytic enzymesderived from white rot fungi, which consist oflignin peroxidase (LiP), manganese peroxidase(MnP) and laccases (Ericksson, 1981; Ericksson,1993; Fiechter, 1993; Hatakka, 1994; Odier andArtaud, 1992; Pelaez et al., 1995; Youn et al.,1995). Unlike the polysaccharide degradingenzymes, which are substrate specific, theenzymes that initiate degradation of ligninappear to have a broad range of substrates(Joseleau et al., 1994).

Most of the current commercially availablecarbohydrate and lignin degrading enzymes arederived primarily from fungal sources usingsubmerged liquid or in some cases solid statesurface fermentation (Hatakka, 1994; Duarte andCosta-Ferreira, 1994; Nieves et al., 1997). Whentested on a lignocellulosic feedstock of interest,considerable variation has been observed incommercial enzyme preparations in thedistribution of key enzyme activities and in therelative activities of cellulases and hemicellulases(Abbas, unpublished observations). For thisreason careful attention must be paid to themethod, substrate and conditions of the assayused by suppliers to measure enzyme activity.Whenever possible, additional information onother enzyme activities assayed should berequested. Knowledge of the microbial sourceand feedstock used by the manufacturer can beuseful in selecting enzyme preparations forfurther evaluation of a targeted feedstock.

In addition to selection and use of the properenzyme activities, care should be taken to designa process that integrates enzyme additionwhenever possible into pretreatment, hydrolysisand fermentation. In view of the currentlimitations of commercial enzymes and theirprohibitive cost, there exists the need to integrateand locate enzyme production facilities in thevicinity of, or as a part of, lignocellulosicprocessing plants (Tolan, 1999). The enzymeproduction facility should utilize a pretreatedfeedstock derived from the lignocellulosic ofinterest and the end user should employ aminimally treated or a crude enzymepreparation. This ensures a cost-effectiveenzyme with the necessary mixture and ratiosof activities needed to process a given

lignocellulosic (Adney et al., 1994). Additionalcost reduction in processing lignocellulosics canbe realized from reduced enzyme loading bytiming enzyme addition with pretreatment andhydrolysis and the use of simultaneoussaccharification and fermentation (SSF). Use ofSSF aids in improved enzyme catalysis byreducing feedstock inhibition from the endproducts, as is the case with glucose andcellobiose inhibition of cellulases.

Microbial fermentation of lignocellulosics

For several decades microbial utilization ofsugars obtained from the hydrolysis oflignocellulosics for the production of fuel ethanolhas been an active area of research (Corringtonand Abbas, 2003; DeCarvalho et al., 2002;Dennison and Abbas, 2000; Dennison andAbbas, 2003; Dien et al., 1997; Green et al.,2001; Ho et al., 1999; Ho et al., 2000a; Ingramand Doran, 1995; Ingram, 2000; Jeffries, 1985;Lai et al., 1996; Wood and Ingram, 1996;Sreenath and Jeffries, 1996; Sreenath andJeffries, 2000; Jeffries et al., 1994; Jeffries etal., 1995 Martin et al., 2002; Palmqvist andHahn-Hagerdal, 2000; Rogers et al., 1997;Supple et al., 2000; Joachimsthal and Rogers,2000; Lawford and Rousseau, 1993a, 1993b;Lawford et al., 2001; Lawford and Rousseau,2002; Lawford and Rousseau, 2003; Yamada etal., 2002; Tengborg et al., 2001; Stengberg etal., 2000 ; van Zyl et al., 1999 ). This has beenlargely due to the absence of suitableethanolgens that can utilize the mixture of thevarious pentose, hexose and higher sugarspresent in hydrolysates (Singh and Mishra,1995). Research has confirmed considerabledifferences in the uptake and utilization of thevarious sugars by bacteria, yeast and molds (Hoet al., 2000b; Ingram, 2000; Jeffries et al., 1994;Jeffries et al., 1995; Green et al., 2001; Zhanget al., 1998; Zhang, 2002; Picataggio et al.,1994). Several strategies have been employedto remedy these limitations ranging from hostselection, host modification by classical straindevelopment approaches and geneticengineering of new strains (Ho et al., 1999; Ho,2000b; Jeffries et al., 1995; Ingram, 2000;Picataggio et al., 1994). While an extensive listof microbial systems that have resulted from

48 C.A. Abbas

Table 5. Fermentations of lignocellulosics.

Single or pure culture Pentose fermenting ethanolgenic yeast orRecombinant bacteria or yeast

Co-culture Utilize an ethanolgenic pentose fermenting yeast in combination with abrewing yeast

Isomerization followed by fermentation Carry out isomerization at neutral to slightly alkaline pH followed bylowering pH and temperature and fermentation using brewing yeast

Simultaneous saccharification and Commonly employed by combining the use of free cellulases/fermentation (SSF) hemicellulases/ligninases with an ethanolgenic organism or an

ethanolgen genetically engineered to utilize lignocellulosics

Table 4. Ethanolgens of interest.

Candida shehatae and Pichia D-xylose fermenting yeast mutants; T. Jeffries, University of Wisconsinstipitis recombinants

Saccharomyces cerevisiae Recombinant D-xylose fermenting yeast N. Ho, Purdue University; B. HahnHagerdal, Lund University (Sweden)

Recombinant L-arabinose fermenting yeast E. Boles, Heinrich-Heine University(Germany); VTT (Finland)

Zymomonas mobilis Recombinant D-xylose/L- arabinose NREL, US Department of Energyfermenting

E. coli/Klebsiella oxytoca Recombinant D-xylose/L-arabinose L.O. Ingram, University of Floridafermenting

Bacillus stearothermophilus Recombinant D-xylose/L-arabinose University of London, AGROLfermenting

research in this area is outside the scope of thischapter, a short list of ethanolgens of interest issummarized in Table 4. A number of excellentreferences on the growth of these organisms onlignocellulosic hydrolysates and/or sugars isprovided at the end of this chapter. In someinstances, modifications in the fermentationprocess have been employed to enhance sugaruptake and utilization as well as to circumventhydrolysate toxicity and end product toxicity.The list provided in Table 5 illustrates some ofthe available options that can be used in thefermentation of sugars in lignocellulosichydrolysates by combining an appropriatehydrolysate and process.

No discussion of lignocellulosic hydrolysatefermentation is complete without addressing themajor differences in the metabolism of thepentoses D-xylose and L-arabinose by fungi andbacteria. These differences are illustrated inFigures 5 and 6. In bacteria, D-xylose utilizationinvolves the action of D-xylose isomerasefollowed by phosphorylation of D-xylulose by

D-xylulose kinase (Amore et al., 1989; Gulatiet al., 1996; Ho et al., 1999; Singh and Mishra,1995). By comparison the utilization of D-xylosein fungi proceeds with the action of D-xylosereductase with the formation of xylitol as theproduct (Bruinenberg et al., 1984; Jeffries,1985; Jeffries et al., 1994; Jeffries et al., 1995;Singh and Mishra, 1995; Jeppsson et al., 1999).This is followed by dehydrogenation of xylitolby the action of xylitol dehydrogenase to D-xylulose, which in turn is acted on by D-xylulosekinase (Singh and Mishra, 1995; Amore et al.,1989). The product of the phosphorylation, D-xylulose-5-phosphate, is assimilated into thepentose phosphate pathway, which feeds intothe glycolytic pathway leading to the productionof ethanol.

A significant difference in cofactor use existsbetween the naturally pentose-fermenting yeastsuch as Pichia stipitis and the brewing yeast,Saccharomyces cerevisiae, as illustrated inFigure 7. In Pichia stipitis and in other generaof pentose fermenting yeast, the reductases that


yeast is unable to produce significant levels ofethanol from D-xylose under anaerobicconditions (Amore et al., 1989; Bruinenberg etal., 1984; Jeffries, 1985). Cloning of the Pichiastipitis D-xylose reductase that utilizes NADHinto S. cerevisiae improved the ability of this

Figure 5. Microbial D-xylose metabolism.

D-XYLOSE

Xylitol

In Bacteria In Fungi

Xyloseisomerase (XI)

XyluloseATP

ADPXylulokinase (XK)

Xylulose-5-P

Pentose phosphatepathway

Embden-Meyerhoff pathway(Glycolysis)

NADPH

NADP+

NADH

NAD+

Xylose reductase (XR)

Xylitol dehydrogenase (XDH)

Figure 6. Microbial metabolism of L-arabinose.

Penitol-NAD- dehydrogenase

L-ribulose

L-ribulose-5-phosphate

D-xylulose-5-phosphate

Pentosephosphatepathway

L-arabinoseisomerase

NAD(P)H

NAD(P)+

NADPH-dependent aldopentose reductase

L-arabitol

ATP

ADP

L-ARABINOSE

D-xylulose kinase

L-xylulose

D-xylulose

L-xylulose-NADPH reductase

Xylitol

Kinase

Epimerase

Penitol-NAD- dehydrogenase

NAD+

NADH

In Bacteria In Fungi

reduce D-xylose to xylitol can utilize eitherNADH or NADPH. In contrast, in S. cerevisiaethe host reductase that acts on D-xylose is limitedto NADPH. This cofactor use in S. cerevisiae isresponsible for the cofactor regenerationimbalance and is one of the reasons why this

50 C.A. Abbas

yeast to utilize D-xylose and resulted inimproved ethanol production by the geneticallyengineered Saccharomyces strain (Table 6).Further genetic engineering of S. cerevisiae toincrease D-xylulose kinase activity yielded newrecombinant yeast strains with significantethanol production from D-xylose (Ho et al.,1999; Ho et al., 2000a; Eliasson et al., 2000).Most recently, employment of metabolic fluxanalysis to recombinant S. cerevisiae grown ona mixture of D-glucose and D-xylose has beenused to further delineate additional genes thatcan be targeted to improve ethanol productionfrom D-xylose (Figure 8). This latest approachhighlights the great potential offered bymetabolic engineering to achieve furtherimprovements in ethanol yield and productivityin recombinant Saccharomyces cerevisiae(Nielsen, 2001; Ostergaard, 2000).

Table 6. Effect of XK overexpression by S. cerevisiae.

+XR +XDH Brewing yeast expressing these twogenes are poor producers of ethanol.

+XR +XDH +XK Brewing yeast expressing thesethree genes show greater ethanolproduction.Recombinants expressing atruncated XK are good D-xylosefermenters

The metabolism of L-arabinose in bacteria(outlined in Figure 6) involves the enzyme L-

arabinose isomerase, which yields L-ribulose asthe product (Deanda et al., 1996; Saha andBothast, 1995). Further metabolism of L-ribuloseproceeds with its phosphorlyation by the actionof a L-ribulose kinase to form L-ribulose-5-phosphate which is converted to D-xylulose -5-phosphate. As indicated earlier, D-xylulose-5-phosphate is assimilated further via the pentosephosphate pathway into glycolysis, which inethanolgenic recombinant bacteria culminates inthe production of ethanol. In fungi the utilizationof L-arabinose involves the initial action of analdopentose reductase with the production of L-arabitol. The metabolism of L-arabitol is carriedon further through the action of a penitoldehydrogenase, an L-xylulose reductase, asecond penitol dehydrogenase and finally D-xylulose kinase. The product of the last step,D-xylulose-5-phosphate, is assimilated into thepentose phosphate pathway and from therefurther into the glycolytic pathway.

Recently two approaches that utilizedknowledge gained from L-arabinose metabolismin bacteria and fungi were successfully employedto engineer strains of S. cerevisaie to fermentthis sugar to the ethanol. In the first approach,the bacterial genes from Escherichia coli (Ara Band Ara D) and Bacillus subitilis (Ara A) werecloned into a S. cerevisaie host strain(Becker and Boles, 2003). This strain was furtherengineered to over express Gal 2 (galactosepermease) and to enhance the enzymetransaldolase activity while reducing L-

Figure 7. Xylose pathways of Pichia stipitis and Saccharomyces cerevisiae.

P. stipitis

Xylose

XR

NADPHNADH

NADPNAD NAD NADH

XDH

Xylitol Xylulose

Xylose

XR XDH

Xylitol Xylulose

S. cerevisiaeNADPH NADP NAD NADH


Fig

ure

8. M

etab

olic

flux

ana

lysi

s of

D-g

luco

se a

nd D

-xyl

ose

in S

acch

arom

yces

cer

evis

iae.

XY

LOS

E

Xyl

itol

Xyl

itol

NA

DH

NA

DP

H

Xylo

se

reductase

Xylo

se

dehydrogenase

Xyl

ulos

e

GLU

CO

SE

Xyl

ulos

e-5-

P

Rib

ulos

e-5-

P

Rib

ose-

5-P

Glu

cose-6

-P

dehydrogenase

Phosphofructokin

ase

Fruc

tose

1,6

bisp

hosp

hate

Dih

ydro

xyac

eton

e p

hosp

hate

Phosphofructoald

ola

se

Trio

sephosphate

isom

erase

Ace

tald

ehyd

eE

than

ol

3-ph

osph

ogly

cera

te

Gly

cera

ldeh

yde

3-ph

osph

ate

Gly

cerald

ehyde-3

-P

dehydrogenase

Pho

spho

enol

pyru

vate

Pyr

uvat

e

Enola

se

Pyruvate k

inase

Glycolysis

Pyruvate

decarboxyla

se

Ace

tate

NA

DH

CO

2

CO

2

CO

2

CO

2

CO

2

Pyruvate c

arboxyla

se

NA

DP

H

Acetald

ehyde

dehydrogenase

Ace

tyl C

oA

Oxa

loac

etat

e

Alc

ohol

dehydrogenase

NA

DH

2N

AD

PH

Transla

dola

se

Sed

ohep

tulo

se-7

-P

Ery

thro

se-4

-P

Transketola

se

Xylo

se

kin

ase

Oxa

loac

etat

e

Ace

tyl C

oA

Suc

cina

te

Mal

ate

Isoc

itra

te

α-ke

togl

utar

ate

NA

DP

H

NA

DH

O2

H2O

Respir

atio

n

NA

DH

Isocitrate

dehydrogenase

NA

DH

NA

DP

H

Mala

te

dehydrogenase

Succin

ate

dehydrogenase

Fum

erase

Pyruvate

dehydrogenase

Fruc

tose

-6-P

Glu

cose-6

-P

isom

erase

Glu

cose

-6-P

Gly

cero

l

NA

DH

Gly

cerol-

3-phosphate d

ehydrogenase

Gly

cerol-

3-phosphatase

CO

2

Citric

acid

cycle

*

*

* *

* *

*

*

*

*

* Bio

mas

s pr

oduc

tion

52 C.A. Abbas

ribulokinase activity. The engineered strainfermented L- arabinose to ethanol under reducedoxygen conditions with a doubling time of 7.9hrs in a medium containing L-arabinose as thesole carbon source. In the second approach, thefive genes involved in L-arabinose metabolismin fungi were cloned into a S. cerevisiae hoststrain (Richard et al., 2002; Richard et al.,2003). These genes were over expressed in thehost strain leading to growth on L-arabinose andthe production of ethanol under anaerobicconditions.

Current status of lignocellulosic researchin the US

Over the past decade several initiatives that aimto improve the economics of the utilization oflignocellulosics for fuel ethanol production havebeen proposed and funded by the US DOE.These initiatives focused primarily on threeareas:

1) sugar or feedstock development platform withmost recent focus on lignocellulosics presentin corn fiber hulls from corn wet mills andcorn stover

2) enzyme development platform, whichprimarily targets improvements in cellulaseenzyme technology with the goal of reducingthe cost of these enzymes to bring theircommercial costs in line with other industrialenzymes

3) ethanolgens development platform with afocus on yeast and bacterial recombinantswith broad and improved feedstock utilization

The above DOE-funded initiatives recognize theneed for further research to improve lignocellulosicfeedstock pretreatment and hydrolysis as wellas to further improve biocatalysts to facilitateeconomic scale-up for ethanol production fromrenewable lignocellulosics. Recently, in anattempt to realign research programs andinitiatives with that of emerging biomass to fuelsand chemicals industrial biorefineries, DOE hasundergone an assessment of its role and is inthe process of establishing new priorities. Thisassessment would serve to ensure that all futuredevelopment in the area of lignocellulosic

research would be industry driven. Over the nextdecade the successful integration of the researchresults on lignocellulosics from the DOEplatforms and from DOE co-funded industry-led pilot plant biorefineries will be key to furtherUS efforts in improving the economics ofbiomass conversion to ethanol and chemicals.

Concluding remarks

In the US, lignocellulosics are attractivefeedstocks for further expansion of fuel ethanolproduction. The recent advances in pretreatmentand hydrolysis when combined with theavailability of new ethanolgens have increasedthe feasibility of hemicellulose conversion toethanol. The conversion of cellulose continuesto be economically and technically challengingdue to the current high cost of commerciallyavailable enzymes as well as the energy cost forpretreatment and hydrolysis and cost ofsubsequent detoxification. Processing facilitiesthat can utilize corn crop residues such as cornstover or corn wet mill biorefineries thatcurrently produce several hundred millionpounds of corn fiber hulls annually will representthe first generation of lignocellulosic-basedethanol production plants built. These facilitieswill be adjacent and fully integrated into existingplant operations and thereby benefit from theavailable infrastructure. Based on currentresearch trends it is predicted that these plantswill be in operation within the next decade andtherefore will represent in the near term thesuccessful commercialization of lignocellulosic-based biorefineries.

References

Abbas, C.A. 1996. Agribusiness by-productspotential for higher value added products. 18th

Symposium on Biotechnology for Fuels andChemicals. Gatlinburg, TN (abstract).

Abbas, C.A. 2000. Integrative bioprocessing ofcorn fiber hulls (CFH) to ethanol. AnnualRenewable Fuels Association Board ofDirectors Meeting. Minneapolis, MN (abstract).

Abbas, C.A., K.E. Beery, A. Rammelsberg, M.Alnajjar, J. Franz, R. Orth, A. Schmidt, T.Werpy and R. Shunk. 2002. Thermochemical


hydrolysis of corn fiber. 24th Symposium onBiotechnology for Fuels and Chemicals.Gatlinburg, TN (abstract).

Adney, W.S., C.I. Ehrman, J.O. Baker, S.R.Thomas and M.E. Himmel. 1994. Cellulaseassays. Method for empirical mathematicalmodels. ACS Symposium Series 566:218-235.

Amore, R., M. Wilhelm and C.P. Hollenberg.1989. The fermentation of xylose-an analysisof the expression of Bacillus and Actinoplanesxylose isomerase genes in yeast. Appl.Microbiol. Biotechnol. 30:351-357.

Baker, J.O., W.S. Adney, S.R. Thomas, R.A.Nieves, Y-C. Chou, T.B. Vinzant, M.P. Tucker,R.A. Laymon and M.E. Himmel. 1995.Synergism between purified bacterial andfungal cellulases. ACS Symposium Series618:113-141.

Becker, J. and E. Boles. 2003. A modifiedSaccharomyces cerevisiae strain that consumesL-arabinose and produces ethanol. Appl. Env.Microbiol. 69(7):4144-4150.

Beery, K.E., C.A. Abbas, T. A. Werpy, A.J.Schmidt and R.J. Orth. 2003. Separation andconversion of corn fiber. 225th AmericanChemical Society National Meeting. NewOrleans, LA (abstract).

Bothast, R.J. and B.C. Saha. 1997. Ethanol fromagricultural biomass substrates: a review. Adv.Appl. Microbiol. 44:261-286.

Bruinenberg, P.M., P. H.M. de Bot, P. van Dijkenand W. A. Scheffers. 1984. NADH-linkedaldose reductase: the key to anaerobicalcoholic fermentation of xylose by yeasts.Appl. Microbiol. Biotechnol. 19:256-260.

Corrington, P. and C. Abbas. 2003. Screeningof ethanolgens on corn fiber hydrolysates. 25th

Symposium on Biotechnology for Fuels andChemicals. Breckenridge, CO (abstract).

De Carvalho Lima, K.G., C.M. Takahashi andF. Alterthum. 2002. Ethanol production fromcorn cob hydrolysates by Escherichia coliKO11. J. Indust. Microbiol. and Biotech.29:124-128.

Deanda, K., M. Zhang, C. Eddy and S.Picataggio. 1996. Development of anarabinose-fermenting Zymomonas mobilisstrain by metabolic pathway engineering;metabolic engineering for ethanol productionfrom agriculture residue waste. Appl. Environ.Microbiol. 62(12): 4465-4470.

Dennison, E.K. and C.A. Abbas. 2000.Evaluation of recombinant Zymomonas mobilisstrains on corn fiber hydrolysate. 22nd

Symposium on Biotechnology for Fuels andChemicals. Gatlinburg, TN (abstract).

Dennison, E.K. and C.A. Abbas. 2003.Evaluation of recombinant microorganismethanol fermentation of corn fiber hydrolysate.25th Symposium on Biotechnology for Fuelsand Chemicals. Breckenridge, CO (abstract).

Dien, B.S., R.B. Hespell, L.O. Ingram and R.J.Bothast. 1997. Conversion of corn millingfibrous co-products into ethanol byrecombinant Escherichia coli strains K011 andSL40. World J. Microbiol. and Biotech.13:619-625.

Duarte, J.C. and M. Costa-Ferreira. 1994.Aspergilli and lignocellulosics: enzymologyand biotechnological applications. FEMSMicrobiology Reviews 13:377-386.

Eliasson, A., C. Christensson, C.F. Wahlbom andB. Hahn-Hagerdal. 2000. Anaerobic xylosefermentation by recombinant Saccharomycescerevisiae carrying XYL1, XYL2 and XKS1 inmineral medium chemostat cultures; plasmidYIpXR/XDH/XK-mediated recombinant yeastconstruction for improved ethanol production.Appl. Environ. Microbiol. 66(8):3381-3386.

Eriksson, K-E. 1981. Microbial degradation ofcellulose and lignin. Ekman-Days 1981, Int.Symp. Wood Pulping Chem.

Eriksson, K-E.L. 1993. Lignin biodegradationand practical utilization: concluding remarks-where do we stand and where are we going?J. Biotech. 30:149-158.

Fiechter, A. 1993. Function and synthesis ofenzymes involved in lignin degradation. J.Biotech. 30:49-55.

Galbe, M. and G. Zacchi. 2002. A review of theproduction of ethanol from softwood; alcoholproduction via hemicellulose, cellulose andlignin using cellulase hydrolysis process,fungus fermentation and distillation: a review.Applied Microbiol. and Biotech. 59(6):618-628.

Ghosh, B.K. and A. Ghosh. 1992. Degradationof cellulose by fungal cellulase. In: MicrobialDegradation of Natural Products (G.Winkelmann, ed). VCH Publishers, New York,NY, U.S.A.

54 C.A. Abbas

Green, E., N. Baghael-Yazdi and M. Javed.2001. Fermentative production of ethanol byheterologous expression of a pyruvatedecarboxylase gene in Bacillus sp.WO0149865A1 International patentapplication, 12 July, 2001

Gulati, M., K. Kohlmann, M.R. Ladisch, R.Hespell and R. Bothast. 1996. Assessment ofethanol production options for corn products.Bioresource Tech. 58:253-264.

Hatakka, A. 1994. Lignin-modifying enzymesfrom selected white-rot fungi: production androle in lignin degradation. FEMSMicrobiology Reviews 13:125-135.

Higuchi, K. 1990. Lignin biochemistry:biosynthesis and biodegradation. Wood Sci.Technology. 24:23-63.

Ho, N.W., Z. Chen, A.P. Brainard and M. Sedlak.1999. Successful design and development ofgenetically engineered Saccharomyces yeastsfor effective cofermentation of glucose andxylose from cellulosic biomass to fuel ethanol.In: Advances in Biochemical Engineering/Biotechnology Vol. 65:163-192 (T. Scheper,ed). Springer-Verlag and Berlin Heidelberg,Germany.

Ho, N.W., Z. Chen, A.P. Brainard and M. Sedlak.2000a. Genetically engineered Saccharomycesyeasts for conversion of cellulosic biomass toenvironmentally friendly transportation fuelethanol. ACS Symposium Series, pp. 143-159.

Ho, N.W., Z. Chen and M. Sedlak. 2000b.Strategies for successful metabolic engineeringof Saccharomyces yeasts to effectively co-utilize glucose and xylose from renewablecellulosic biomass for the production ofethanol and other industrial products. Am.Chem. Soc. 219th ACS National Meeting(abstract).

Ingram, L.O. 2000. Engineering bacteria forethanol from lignocellulosics. Am. Chem. Soc.219th ACS National Meeting (abstract).

Ingram, L.O. and J.B. Doran. 1995. Conversionof cellulosic materials to ethanol. FEMSMicrobiology Reviews 16: 235-241.

Jeffries, T.W. 1985. Emerging technology forfermenting D-xylose. Trends in BiotechnologyVol.3(8):208-212.

Jeffries, T.W. 1994. Biodegradation of lignin andhemicelluloses. In: Biochemistry of Microbial

Degradation (C. Ratledge, ed). KluwerAcademic Publishers, The Netherlands, pp.233-277.

Jeffries, T.W., K. Dahn, W.R. Kenealy, P. Pittman,H.K. Sreenath and B.P. Davis. 1994. Geneticengineering of the xylose-fermenting yeastPichia stipitis for improved ethanol production.Am. Chem.Soc. 208th ACS National Meeting(abstract).

Jeffries, T.W., B.P. Davis, H.K. Sreenath, J.Y.Cho and K. Dahn. 1995. Genetic developmentof the xylose fermenting yeast Pichia stipitis.Am. Chem. Soc. 209th ACS National Meeting(abstract).

Jeppsson, H., K. Holmgren and B. Hahn-Hagerdal. 1999. Oxygen-dependent xylitolmetabolism in Pichia stipitis; influence ofoxygen on utilization of xylitol alone, or as amixture with xylose or glucose, for ethanolproduction. Appl. Microbiol. Biotechnol.53(1):92-97.

Joachimsthal, E.L. and P.L. Rogers. 2000.Characterization of a high-productivityrecombinant strain of Zymomonas mobilis forethanol production from glucose/xylosemixtures. Applied Biochemistry andBiotechnology - Part A Enzyme Eng. andBiotech. 84-86:343-356.

Joseleau, J-P. , S. Gharibian, J. Comtat, A.Lefebvre and K. Ruel. 1994. Indirectinvolvement of ligninolytic enzyme systemsin cell wall degradation. FEMS MicrobiologyReviews 13:255-264.

Katzen, R., P.W. Madson and D.A. Monceaux.1999. Lignocellulosic feedstocks for ethanolproduction: the ultimate renewable energysource. In: The Alcohol Textbook. 3rd Edition(K.A. Jacques, T.P. Lyons and D.R. Kelsall,eds). Redwood Books, Trowbridge, Wiltshire,UK, pp. 117-127.

Lai, X., L.O. Ingram and R.B. Hespell. 1996.Cloning of celliobiose utilization genes fromG- and G+ bacteria: functional expression inEscherichia coli. Gen. Meet. Am. Soc.Microbiol. 96th Meeting (abstract).

Lawford, H.G. and J.D. Rousseau. 1993a.Production of ethanol from pulp mill hardwoodand softwood spent sulfite liquors bygenetically engineered E. coli; Escherichiacoli application in wood pulp waste-disposal


from spent sulfite liquor. Appl. Biochem.Biotechnol. 39-40:667-685.

Lawford, H.G. and J.D. Rousseau.1993b.Developments in hemicellulose bioconversiontechnology; ethanol production by xylose,newsprint or spent sulfite liquor fermentationby recombinant Escherichia coli carryingplasmid pLOI297: a review. Energy BiomassWastes XVI: 559-97.

Lawford, H.G. and J.D. Rousseau. 2002.Performance testing of Zymomonas mobilismetabolically engineered for cofermentationof glucose, xylose and arabinose. AppliedBiochemistry and Biotechnology - Part AEnzyme Engineering and Biotechnology 98-100:429-448.

Lawford, H.G. and J.D. Rousseau. 2003.Cellulosic fuel ethanol: Alternative processdesigns with wild-type and recombinantZymomonas mobilis. Applied Biochemistryand Biotechnology – Part A EnzymeEngineering and Biotechnology 105-108:457-469.

Lawford, H.G., J.D. Rousseau and J.S. Tolan.2001. Comparative ethanol productivities ofdifferent Zymomonas recombinants fermentingoat hull hydrolysate. Applied Biochemistry andBiotechnology- Part A Enzyme Engineeringand Biotechnology 91-93:133-146.

Lynd, L.R., P.J. Weimar, W.H. van Zyl and I.S.Pretorius. 2002. Microbial cellulose utilization:fundamentals and biotechnology. Microbiol.and Molecular Biol. Rev. 66(3):506-577.

Malburg, L.M. Jr., J.M. Tamblyn Lee and C.W.Forsberg. 1992. Degradation of cellulose andhemicelluloses by rumen microorganisms. In:Microbial Degradation of Natural Products (G.Winkelmann, ed). VCH Publishers, New York,NY, U.S.A.

Martin, C., M. Galbe, C.F. Wahlbom, B. Hahn-Hagerdal and L.J. Jonsson. 2002. Ethanolproduction from enzymatic hydrolysates ofsugarcane bagasse using recombinant xylose-utilizing Saccharomyces cerevisiae; involvingsugarcane bagasse hydrolysate, detoxification,phenoloxidase laccase, recombinant xylose-utilising Saccharomyces cerevisiae andfermentation. Enzyme and Microbial. Tech.31(3):274-282.

McMillan, J.D. 1994a. Conversion of

hemicellulose hydrolyzates to ethanol. ACSSymposium 566:411-437.

McMillan, J.D. 1994b. Pretreatment oflignocellulosic biomass. ACS SymposiumSeries 566:292-324.

Monceaux, D.A and P.W. Madson. 1999.Alternative feedstocks: a case study of wasteconversion to ethanol. In: The AlcoholTextbook. 3rd Edition (K.A. Jacques, T.P. Lyonsand D.R. Kelsall, eds). Redwood Books,Trowbridge, Wiltshire, UK, pp.129-135.

Moniruzzaman, M., B.S. Dien, B. Ferrer, R.B.Hespell, B.E. Dale, L.O. Ingram and R.J.Bothast. 1996. Ethanol production from AFEXpretreated corn fiber by recombinant bacteria.Biotech. Letters 18(8):985-990.

Nielson, J. 2001. Metabolic engineering. Appl.Microbiol. Biotechnol. 55:263-283.

Nieves, R.A., C.I. Ehrman,W.S. Adney, R.T.Elander and M.E. Himmel. 1997. Technicalcommunication: survey and analysis ofcommercial cellulase preparations suitable forbiomass conversion to ethanol; cellulasecomplex characterization a view to its use inethanol production from lignocellulose. WorldJ. Microbiol. Biotechnol. 14(2):301-304.

Odier, E. and I. Artaud. 1992. Degradation oflignin. In: Microbial Degradation of NaturalProducts (G. Winkelmann, ed). VCHPublishers, New York, NY, U.S.A.

Ostergaard, S., L. Olsson and J. Nielsen. 2000.Metabolic engineering of Saccharomycescerevisiae. Microbiol. Mol. Biol. Rev. 64:34-50.

Palmqvist, E. and B. Hahn-Hagerdal. 2000.Fermentation of lignocellulosic hydrolysates.II: Inhibitors and mechanisms of inhibition; theinhibitory effect of weak acid, furan derivativeand phenolic compound on ethanol productionby Saccharomyces cerevisiae from ahydrolysate substrate; a review. BioresourceTechnol. 74(1):25-33.

Pelaez, F., M.J. Martinez and A.T. Martinez.1995. Screening of 68 species ofbasidiomycetes for enzymes involved in lignindegradation. Mycol. Res. 99(1):37-42.

Picataggio, S.K., M. Zhang; and MarkFinkelstein.1994. Development of geneticallyengineered microorganisms for ethanolproduction. ACS Symposium Series 566:342-362.

56 C.A. Abbas

Ramos, L.P. and J.N. Saddler. 1994.Bioconversion of wood residues: mechanismsinvolved in pretreating and hydrolyzinglignocellulosic materials. ACS SymposiumSeries 566:325-41.

Richard, P., M. Putkonen, R. Vaananen, J.Londesborough and M. Penttila. 2002. Themissing link in the fungal L-arabinosecatabolic pathway, identification of the L-xylulose reductase gene. Biochemistry. 2002May 21;41(20):6432-6437.

Richard, P., R. Verho, M. Putkonen, J.Londesborough and M. Penttila. 2003. Thefungal L-arabinose catabolic pathway. In: YeastVol 20, Number S1. XXIst InternationalConference on Yeast Genetics and MolecularBiology, Gothenburg, Sweden.

Richard, P., R. Verho, M. Putkonen, J.Londesborough and M. Penttila. 2003.Production of ethanol from L-arabinose bySaccharomyces cerevisiae containing a fungalL-arabinose pathway. FEMS Yeast Research3:185-189.

Rogers, P.L. 1997. Bioenergy: Ethanol frombiomass. Australas. Biotechnology. 7(5): 283-284.

Rogers, P.L., E.L. Joachimsthal and K.D.Haggett. 1997. Ethanol from lignocellulosics:potential for a Zymomonas-based process;recombinant bacterium fermentation.Australas. Biotechnol. 7(5):304-309.

Saha, B.C. and R.J. Bothast. 1995. Fuel ethanolproduction from L-arabinose: constraints,challenges, current status and future trendsusing recombinant yeast. 209th Am. Chem.Soc. Meeting (abstract).

Saha, B.C. and R. J. Bothast. 1999. Pretreatmentand enzymatic saccharification of corn fiber;acid pretreatment and cellulase and betaglucosidase-mediated saccharification ofmaize starch to yield monomeric sugar, usefulfor the economic production of ethanol. Appl.Biochem. Biotechnol. 76(2):65-78.

Saha, B.C., B.S. Dien and R.J. Bothast. 1998.Fuel ethanol production from corn fiber:current status and technical prospects. AppliedBiochemistry and Biotechnology - Part AEnzyme Engineering and Biotechnology. 70-72:115-125.

Singh, A. and P. Mishra. 1995. Microbial pentose

utilization. Current applications inbiotechnology. Vol. 33. Elsevier Science B.V.,Amsterdam, The Netherlands.

Sreenath, H.K. and T.W. Jeffries. 1996. Effectof corn steep liquor on fermentation of mixedsugars by Candida shehatae FPL-702. AppliedBiochemistry and Biotechnology Papers from17th Symposium on Biotechnology for Fuelsand Chemicals. Vol. 57/58:551-561.

Sreenath, H.K. and T.W. Jeffries. 2000.Production of ethanol from wood hydrolyzateby yeasts. Bioresource Technology 72:253-260.

Stenberg, K., C. Tenborg, M. Galbe and G.Zacchi. 1998. Optimization of steampretreatment of SO2-impregnated mixedsoftwoods for ethanol production; softwoodpretreatment by SO2 impregnation and steam.J. Chem. Technol. Biotechnol. 71(4): 299-308.

Stenberg, K., M. Bollok, K. Reczey, M. Galbeand G. Zacchi. 2000. Effect of substrate andcellulase concentration on simultaneoussaccharification and fermentation of steam-pretreated softwood for ethanol production;alcohol production by Saccharomycescerevisiae. Biotechnol. Bioenrg. 68(2):204-210.

Supple, S.G., E.L. Joachimsthal, N.W. Dunn andP.L. Rogers. 2000. Isolation and preliminarycharacterization of a Zymomonas mobilismutant with an altered preference for xyloseand glucose utilization; the mutant, CP4(plasmid pZb5) M1-2, metabolizing xylosemore rapidly than glucose in mixed glucose/xylose media. Biotech. Letters 22(2):157-164.

Tengborg, C., M. Galbe and G. Zacchi. 2001.Reduced inhibition of enzymatic hydrolysis ofsteam-pretreated softwood; for economicallyfeasible ethanol production usingSaccharomyces cerevisiae and enzyme.Enzyme Microb. Technol. 28(9-10):835-844.

Tolan, J.S. 1999. Alcohol production fromcellulosic biomass: the Iogen process, a modelsystem in operation. In: The Alcohol Textbook.3rd Edition (K.A. Jacques, T.P. Lyons and D.R.Kelsall, eds). Nottingham University Press,Nottingham, pp. 107-116.

Van Zyl, W. H., A. Eliasson, T. Hobley and B.Hahn-Hagerdal. 1999. Xylose utilization byrecombinant strains of Saccharomyces


cerevisiae on different carbon sources; xyloseconversion into ethanol by fermentation ofrecombinant Saccharomyces expressingPichia stipitis xylose-reductase and xylitoldehydrogenase. Appl. Microbiol. Biotechnol.52(6):829-833.

Von Sivers, M. and G. Zacchi. 1996. Ethanolfrom lignocellulosics: a review of theeconomy; alcohol production andlignocellulose degradation and saccharifi-cation. Bioresource Technol. 56(2-3):131-140.

Werpy, T.A., J. Franz, M. Alnajjar, A. Schmidt,R.J. Orth, J. Magnuson, T. P. Binder, C. A.Abbas, T. A. Stoa and B. Sedlacek. 2001. Cornfiber separation and subsequent conversion tofuels and chemicals. 23rd Symposium onBiotechnology for Fuels and Chemicals.Breckenridge, CO.

Wood, B.E. and L.O. Ingram.1996. Productionof recombinant bacterial cellulases byethanolgenic bacteria: evaluation for cellulosefermentation. Gen. Meet. Am. Soc. Microbiol.96th Meet (abstract).

Yamada, T., M.A. Fatigati and M. Zhang. 2002.Performance of immobilized Zymomonasmobilis 31821 (pZB5) on actual hydrolysatesproduced by Arkenol technology. AppliedBiochemistry and Biotechnology 98-100:899-907.

Youn, K-D, Y.C. Hah and S-O. Kang. 1995. Roleof laccase in lignin degradation by white-rotfungi. FEMS Microbiol. Letters 132:183-188.

Zhang, M. 2002. Novel recombinantZymomonas mobilis 31821/pZB5 strainderived from Z. mobilis ATCC 31821 usefulfor producing ethanol by fermenting xylose;xylose utilization in ethanol production byvector-mediated gene transfer, expression inhost cell. WO 2002038740.

Zhang, M., C.Y. Chou, S.K. Picataggio and M.Finkelstein. 1998. New Zymomonas mobilisstrain containing exogenous genes; for growthon xylose and arabinose for ethanolproduction. WO 9850524.

Ethanol production from cassava 59

Chapter 6

Ethanol production from cassava

NGUYEN T. T. VINH

The Research Institute of Brewing, Hanoi, Vietnam

Introduction to cassava

Cassava (Manihot esculenta), which is believedto have originated in Latin America, is animportant food crop in many tropical countriesof Africa, South America and Asia (Dougrell,1982). A number of attributes have made it anattractive crop for small farmers with limitedresources in marginal agricultural areas:

• it is one of the most efficient carbohydrate-producing crops;

• it is tolerant of poor soil fertility and drought;

• it has the ability to recover from the damagecaused by most pests and diseases;

• the roots can be left in the ground for longperiods as a food reserve and thus provideinsurance against famine;

• the crop is well adapted to traditional mixedcropping agricultural systems andsubsistence cultivation in which farmers seekto minimize the risk of total crop failure.

Yields of cassava range from 15-150 tonnes perha. In Brazil it can reach 80-90 tonnes; and inColombia yields of >50 tonnes/ha areconsistently obtained.

The roots are the principle edible portion ofthe plant and typical ranges of composition arewater, 62-65%; total carbohydrate, 32-35%;

crude protein, 0.7-2.6%; fat, 0.2-0.5%; fiber, 0.8-1.3%; ash, 0.3-1.3%; cellulose, 1-3.1%;polyphenol, 0.1-0.3%. Cassava chips typicallycontain 65-70% of starch, which makes it oneof the richest sources of fermentable substratefor the production of alcohol. In nutritional terms,cassava is primarily a source of carbohydrate-derived energy, most of which is derived fromstarch. In addition, cassava contains a smallamount of B vitamins and toxic substances.Vitamins include B1, B2 and B6. The toxiccompound in cassava is phazeounatine,consisting of two cyanogenic glucosides,linamarin and lotaustralin. Linamarin accountsfor more than 80% of the cassava cyanogenicglucosides. Phazeounatine content ranges from0.001-0.04 mg/100g; and is present mainly inthe outer layer of bitter cultivars. Phazeounatineitself is not toxic, but it is easily degraded torelease hydrocyanic acid (HCN). Typicalcomposition of cassava chips is shown in Table1.

Table 1. Composition of cassava chips.

Composition %

Moisture 14Carbohydrate 67.6Protein 1.75Lipid 0.87Cellulose 3.38Minerals 1.79

60 N.T.T. Vinh

WORLD CASSAVA PRODUCTION

Total world production has increased from 70million tonnes in 1960 to an estimated 174million tonnes in 2001 (Table 2). Of this total,52% is produced in Africa, 28.5% in Asia and19.5% in Latin America. Cassava production inAsia has risen from 48.5 million tonnes in 1985to 49.4 million tonnes in 2001 (FAO, 2001). Thetwo major Asian cassava-growing countries,Thailand and Indonesia, have experienced thelargest increases in production. Plants forproducing modified cassava starch and otherstarch derived products were established duringthe period 1985 to 1990; and governmentsupport of biofuels programs in Asia willcontinue to prompt higher production (Sriroth,2002).

CASSAVA UTILIZATION AS A FOODSTUFF

About 40% of cassava is used for direct humanconsumption, especially in processed form.Most of the remainder is destined for animal feedor processed for starch.

Global cassava utilization is estimated to haverisen by about 2% in 2000, very much in line

with production, given the fact that cassava foodreserves are mostly kept in the ground in theform of roots until needed for harvest. Propercassava stocks are held only in relatively modestquantities and in dried form.

Cassava global utilization as food is put at 102million tonnes in 2000, the bulk of which isconsumed in Africa in the form of fresh rootsand processed products such as gari, foufou,attiéké, etc. Global cassava utilization as feed isestimated to have remained around 34 milliontonnes, most of which concentrated in LatinAmerica, the Caribbean and in the EC(Balagopalan, 2002).

Ethanol from cassava: process overview

Dried cassava chips are used in most distilleries.The cassava chips are ground to a meal - drymeal for dry milling processes and wet mashfor wet mill. The fresh roots are washed, crushedinto a thin pulp and then screened. In Malaysiaand other countries many factories are equippedto use cassava roots.

Cassava starch/flour is gelatinized first bycooking and further converted to simpler sugarswith the help of mild acids or amylase enzymes.Cassava starch, having a low gelatinization

Table 2. World cassava production (in million tonnes).

1985 1986 1987 1988 1989 1990 1999 2000 2001

World 136.6 133.6 136.8 141.3 148.6 150.0 172.6 175.5 174.0Africa 58.2 58.6 58.4 59.6 62.9 64.1 92.4 92.7 90.9

Ghana 3.1 2.9 2.7 2.8 3.3 3.0 7.8 7.5 7.8Madagascar 2.1 2.4 2.2 2.2 2.3 2.3 2.5 2.2 2.4Mozambique 3.2 3.3 3.4 3.2 3.5 3.6 5.4 4.6 4.5Nigeria 13.5 14.7 14.0 15.0 16.5 17.6 32.7 33.9 34.0Tanzania 6.8 6.2 6.0 6.1 6.2 5.5 7.2 5.8 5.0Uganda 2.7 1.9 2.8 2.5 3.1 3.2 3.3 5.0 5.5

Asia 48.5 42.7 47.6 52.3 54.1 52.0 50.9 50.5 49.4China 3.6 3.5 3.3 3.3 3.2 3.2 3.6 3.6 3.8India 5.7 4.9 4.8 5.4 4.5 4.6 6.1 6.2 6.2Indonesia 14.0 13.3 14.3 15.5 17.1 16.3 16.5 15.7 15.5Phillippines 1.7 1.7 1.8 1.8 1.8 1.9 1.8 1.8 1.8Thailand 19.3 15.2 19.5 22.3 23.5 21.9 20.3 20.2 19.2Vietnam 2.9 2.8 2.7 2.8 2.9 3.0 1.8 2.0 2.0

Latin America 29.6 32.1 30.6 29.2 31.4 33.7 29.2 32.1 33.5Brazil 23.1 25.6 23.5 21.7 23.4 25.4 20.9 23.4 24.6Colombia 1.4 1.3 1.3 1.3 1.5 1.7 1.8 1.9 2.0Paraguay 2.9 2.9 3.5 3.9 4.0 4.0 3.5 3.5 3.7

Note: Figures are approximations. Source: FAO Production Yearbooks


temperature, is easily liquified and saccharified.The main advantage of cassava over any othercrop for this purpose is the presence of highlyfermentable sugars after saccharification. Largevolumes of the saccharified starch are fed intofermentation vessels and inoculated with activelygrowing yeast (Saccharomyces cerevisiae). Acontinuous saccharification system is shown inFigure 1. The pH of the mash for fermentation isoptimally 4–4.5 and the temperature range is 28-32°C. Alcohol is recovered from fermented mashafter 48-72 hrs by distillation.

Ethanol yields are 70-110 liters of absolutealcohol per ton of cassava roots depending onthe variety and method of manufacture, whichcompares to an expected yield of 350-400 liters/tonne from maize. The crude alcohol of cassavais described as average in quality. It has adisagreeable odor, but can be improved if thefirst and last fractions in the distillation processare discarded. It is utilized for industrialpurposes, including cosmetics, solvents andpharmaceutical products. If production isrequired for human consumption, special careshould be taken in handling the roots to rid themof hydrocyanic acid.

Processing cassava for alcohol production

MILLING

The purpose of milling is to break up the cassavaroots to as small a particle size as possible inorder to facilitate subsequent penetration of waterin the cooking process. Most distilleries usehammer mills for the cassava dried chips. Incontrast to corn and rice, cassava chips are softand easy to mill. With a hammer mill, the rawmaterial is introduced into a chamber in which anumber of hammers rotate at high speed from75-80 m/s. The collision of the hammers withthe raw material causes breakdown to a meal.The mill outlet contains a retention screen thatholds back larger particles until they are brokendown further so that there will be a knownmaximum particle size in the meal. The screensare normally in the size range of 2-3 mm for afine grind and 6-8 mm for a coarse grind.Depending on the cooking method, the targetsize of particles may differ. Liquefaction inresponse to cooking and amylase will improvewith fineness of the grind, which is a significantfactor in the final alcohol yield. It is possible toobtain a 5-10% difference in yield when movingfrom coarse to finely ground meal, as is the casewith cereal grains.

Figure 1. Typical continous saccharification system.

Saccharification

system 2

10 - 15 min

Cooler

Pump

Mash

30%

70%

Saccharification

tank 1

55 - 60°C

15 - 20 min

Fermentors

Enzyme

tank

62 N.T.T. Vinh

COOKING

Cooking is the entire process beginning withmixing the cassava meal with water through todelivery of a mash ready for fermentation. Thekey to cooking is to liquefy the starch so it canbe pumped. There is always a protectivemembrane around starch granules, whether thefeedstock is corn, rice, wheat or tubers such asmanioc, potato, and sweet potato. Therefore theessential purpose of cooking is to break themembrane, releasing the starch.

Gelatinization is the phenomena by which agranule of starch swells in presence of water andheat so that surface area exposed to enzymes isincreased. During gelatinization water penetratesthe granule making the breakdown of the starcheasier for the α-amylase. Starch is composed ofamylose and amylopectin. While starch in corncontains 85-90% amylopectin, cassava starchcontains 70% amylopectin. For each type of rawmaterial there is a typical gelatinizationtemperature range, which depends on starchgranule size. When granule size is smallgelatinization temperatures are lower.Gelatinization temperature range of corn starch is67-80°C, rice starch 65-85°C, and 61-70°C forcassava starch. In addition, gelatinizationtemperature depends on the electrolytes in solution.Alkaline substances decrease gelatinizationtemperature, while sugars increase gelatinizationtemperature.

When cooking in weak acid conditions,cellulose normally cannot be hydrolyzed.Hemicellulose is hydrolyzed by xylanase in thepresence of H+ and high temperatures formingdextrins and lower molecular weight substancesincluding arabinose and xylose.

During cooking, a small amount of starch willbe converted to sugars by endogenous amylase.Sucrose, glucose, fructose and a little maltoseare formed in cooking. At high temperaturesthese sugars may be converted into caramel,furfural, oxymethyl furfural and othercompounds. The rate of conversion depends ontemperature and pH. Experience has shown thatadjusting mash pH to 3.5 may reduce the rate ofsugar conversion (5% for glucose and 26% forfructose). If pH is raised to 6.5, then conversionrate will be 80% for glucose and 90% forfructose.

Cooking procedures for cassava

There are three methods for cooking cassava:batch, semi-continuous and continuous. Theratio of water to material is normally 3.5:1-3:1with amylase. The batch method is still used inmany countries, but has the disadvantages oflong cooking times at high temperature withaccompanying sugar loss. The continuousmethod requires strict control. The heat-stableamylase is popular for alcohol plants, and is usedat a rate of 0.02–0.03%. Around 80% of theamylase is used for the first liquefaction and 20%is added for additional liquefaction after boilingand cooling to 90°C.

Liquefaction and saccharification

Hydrolysis by the endoenzyme α-amylase israndom and rapid, reducing starch to detrins ofvarying chain length. The shorter the chainlength, the less work remains for the exoenzymeglucoamylase, which releases single glucosemolecules by hydrolyzing successives α-1,4linkages beginning at the non-reducing end ofthe dextrin chain. Glucoamylase also hydrolyzesα-1,6 branch linkages, but at a much slower rate.

Recently simultaneous saccharification andfermentation (SSF) of cassava has been of interestas a means of reducing the cost and timeinvolved in ethanol production. In an experimentby Keawsompong and co-workers (2003) inThailand, conventional liquefaction/sacchari-fication was compared with an SSF process thatemployed RhizozymeTM, a suface cultureglucoamylase targeted for the temperature and pHconditions of the fermentor. After 48 hrs offermentation, similar amounts and conversionefficiencies were obtained, however by leavingout the saccharification step, the SSF process wascompleted 25% faster than the conventionalmethod of sequential liquefaction andsaccharification (Figure 2). Yield increases werealso observed, presumably because othermaterials became available for fermentation. At36 hrs the conventional system had 8-9% ethanolcompared to 10%+ in the RhizozymeTM

fermentor. Cell counts were also significantlyhigher.

FERMENTATION

After saccharification, mash is cooled down to


28-32°C and is pumped into the fermentor.Fermentation is the critical step in a distillery. Itis here that yeast converts sugar to ethanol,carbon dioxide and other by-products.

Saccharomyces cerevisiae is typically used forethanol fermentation from cassava. Duringfermentation, foaming can be a problem. Thefermentation media from cassava is usually lowin nitrogen, so yeast food or other nitrogensources should be used. Fermentation time isnormally of 56-72 hrs.

As with other fermentations, infections are ofconcern. LactosideTM, a broad-spectrumantimicrobial, is added at 2-3 ppm in fermentorsin some distilleries.

DISTILLATION

Distillation procedures are those standardthroughout the industry. Cassava poses nospecial problems in the still. Yield is around 94-95% of calculated values.

References

Akpan, I., M.J. Ikenebomeh, N. Uraih, and C.O.Obuekwe. 1988. Production of ethanol fromcassava whey. Acta Biotechnol. 8(1):39-45.

Balagopalan, C. 2002. Cassava Utilization inFood, Feed and Industry. CCAB International.

Figure 2. Changes in total soluble solids (TSS), brix, cell count and ethanol during cassava fermentation by yeast in a conven-tional (top) liquefaction and saccharification process and a simultaneous saccharification and fermentation (bottom) process using

RhizozymeTM. The SSF process required 25% less time (from Keawsompong et al., 2003).

Time (hr)

0 12 24 36 48

Totalsoluble solids

(Brix)

10

12

14

16

18

20

22

24

26

Cel

l (10

8 ce

lls/m

l)

Totalsoluble solids

(Brix)

Cel

l (10

8 ce

lls/m

l)

Eth

ano

l (%

, v/v

)E

than

ol (

%, v

/v)

0.0

.5

1.0

1.5

2.0

0

2

4

6

8

10

12

Ethanol

Cell

TSS

Time (hr)

0 12 24 36 48

10

12

14

16

18

20

22

24

26

0.0

.5

1.0

1.5

2.0

0

2

4

6

8

10

12

Ethanol

Cell

TSS

Conventional

RhizozymeTM

64 N.T.T. Vinh

Dougrell, L.M. 1982. Cassava - modernagricultural system development towardsethanol production in Australia. InternationalAlcohol Fuel Technology Symp. NZL, p.63-72.

FAO/GIEWS - Food Outlook No 3, June 2001,p.9.

Keawsompong, S., K. Piyachomkwan, S.Walapatit, C. Rodjanaridpiched and K. Sriroth.

2003. Ethanol production from cassava chips:simultaneous saccharification and fermentationprocess. BioThailand Conference, Pattaya,Thailand.

Sriroth, K., K. Piyachomkwan, K. Sangseethongand C. Oates. 2002. Modification of cassavastarch. X International Starch Convention, 11-14 June 2002, Kracow, Poland.

Whey alcohol - a viable outlet for whey? 65

Chapter 7

Whey alcohol - a viable outlet for whey?

JACK O’SHEA

Alltech Inc., Dunboyne, County Meath, Ireland

Introduction

Whey is the main by-product from casein andcheese manufacture. In the past, cheese makingwas a traditional cottage industry using the milkfrom various animals such as goats, camels andreindeer. Today cheese is usually made in largeproduction facilities from cow’s milk by theapplication of microbiology and chemistry.Casein, the chief milk protein, is coagulated bythe enzymatic action of rennet or pepsin, by lacticacid producing bacteria, or by a combination ofthe two. The next step involves separating thecurds from the whey, a process facilitated byheating, cutting, and pressing. With theseparation of the curds most of the butterfat anda number of the other constituents are removed.The butterfat usually comprises only 3-4% byvolume of whole milk. Whey makes up to 80-90% of the total volume of milk that enters thecheese making process and contains about 50%of the nutrients. These nutrients are solublewhey protein, lactose, vitamins and minerals.Whey from cheese and rennet casein is knownas sweet whey; and the manufacture of acidcasein yields acid whey. Acid whey is producedby removal of the fat by centrifugation to makebutter. The milk is then acidified to coagulatethe protein. This is then centrifuged off as casein,leaving acid whey.

Table 1. Approximate composition of whey.*

Component %

Water 94.50Dry matter 5.75Protein 0.80Lactose 4.40Ash 0.50Fat 0.05

*This is an average analysis, the type of cheese producedwill influence composition of the whey.

Approximately 100 kg of milk are required toproduce 10 kg of cheese. Quantities of milkproduced in the top 40 milk producing countriesin the world are outlined in Table 2. The top 40cheese producing countries are listed in Table3. The whey production from all this milk isconsiderable and can present a disposal problem.The challenge to dairy production plants is todispose of whey in an environmentally friendlyand economic manner. The problems facingwhey producers and the uses for whey areexplored below.

Problems with whey

Whey produced by dairies has historically beenand remains today a potential pollution problem.

66 J. O’Shea

Table 2. International production of milk (thousand tonnes).1

Country 1995 1996 1997

1 USA 70500 70003 710722 India 32000 35500 345003 Russian Federation 39098 35590 340004 Germany 28621 28799 287005 France 25413 25083 249576 Brazil 16985 18300 191007 Ukraine 17060 15592 163008 United Kingdom 14683 14680 148499 Poland 11642 11696 1209910 New Zealand 9285 9999 1113111 Netherlands 11294 11013 1092212 Italy 10497 10799 1053113 Argentina 8792 8947 979514 Turkey 9275 9466 946615 Australia 8460 8986 930316 Japan 8382 8657 864217 Mexico 7628 7822 809118 Canada 7920 7890 780019 China 6082 6610 694620 Spain 6150 6084 599721 Columbia 5078 5000 540822 Irish Republic 5415 5420 538023 Romania 4646 4676 512624 Belarus 5070 4908 485025 Denmark 4676 4695 463226 Pakistan 4223 4379 454027 Switzerland 3913 3862 391328 Iran 3450 3809 389729 Kazakhstan 4576 3584 358430 Belgium/Luxembourg 3644 3682 347731 Sweden 3304 3316 333432 Uzbekistan 3665 3183 320033 Czech Republic 3155 3164 316434 Austria 3148 3034 309035 Sudan 2760 2880 288036 South Africa 2794 2592 272037 Finland 2468 2431 246338 Kenya 2170 2210 221039 Korea 1998 2034 207240 Chile 1890 1924 2040

Total world milk production 421,810 422,299 426,181

1National statistics, Eurostat, 1999.


Table 3. International cheese production (thousand tonnes).1

Country 1995 1996 1997

1 USA 3493 3627 36272 France 1617 1635 16603 Germany 1453 1531 15914 Italy 982 985 9405 Netherlands 700 709 7136 Russian Federation 477 477 4777 Argentina 369 376 4058 Poland 354 397 3979 United Kingdom 362 377 37810 Egypt 344 346 34911 Canada 301 312 32312 Denmark 311 299 29113 New Zealand 197 239 27414 Australia 234 261 27015 Spain 240 247 24616 China 202 202 20617 Iran 196 196 19718 Greece 202 192 19219 Czech Republic 137 139 13920 Turkey 136 139 13921 Switzerland 132 133 13322 Mexico 126 123 12323 Sweden 129 127 11824 Japan 105 109 11425 Austria 101 100 10726 Irish Republic 84 98 9727 Israel 92 92 9228 Norway 85 84 8929 Finland 96 95 8830 Hungary 83 88 8831 Syria 66 75 8632 Sudan 75 76 7633 Belgium/Luxembourg 72 73 7634 Bulgaria 82 71 7235 Ukraine 88 71 7136 Venezuela 103 70 7037 Portugal 67 67 6538 Brazil 60 60 6039 Columbia 51 51 5140 Chile 48 49 51

Total world 14,052 14,398 14,541

1National statistics, Eurostat, 1999.

68 J. O’Shea

As a consequence, the majority of cheeseproducing countries have strict laws defining thepermitted biological and chemical oxygendemand (BOD and COD) limits for the effluentfrom dairy plants. A cheese manufacturer’sability to dispose of the whey produced may bethe limiting step in the volume of cheese thatcan be produced.

A number of problems are common to all wheyusers. The first complication is that the whey isextremely perishable. This is especially true forcheese whey, which contains bacteria addedduring cheese production. The whey musttherefore be protected from biologicaldegradation; and the most effective method ofprotection is pasteurization. Alternatively, thewhey can be cooled to 4°C or protected by theuse of chemical additives, e.g., hydrogenperoxide. A second problem with whey is thevery high water content, over 93%. This meansthat transport and concentration costs are high,and form a large proportion of the totalprocessing cost. Another factor influencing theeconomics of whey processing in many

countries is one of seasonality. Most cheeseproduction occurs between March andSeptember. This means that the processing plantis used for only half of the year. Finally, cheesewhey contains casein fines and fat that must beremoved prior to most processing methods. Thisis not particularly a disadvantage, since therecovered fines can be added back to the cheeseor used to make processed cheese while theseparated fat can be used to make whey butter(Figure 1). The fines are usually separated onvibrating screens and the fat by centrifugation.

Uses for whey

WHEY AS AN ANIMAL FOOD

One solution to the whey problem is to sell it topig farmers, however there are many limitationsto the use of whey as a pig feed. Due to its lowdry matter content, whey-fed pigs typicallyproduce very large volumes of urine. Manypiggeries are not in a position to deal with the

Figure 1. Uses for processed whey (adapted from Nielson, 1997).

Whey

Pre-treatment separation

Pre-treated wheyWhey cream

Whey butter

Concentrationof solids

Whey protein concentrate

Whey protein isolate

Brown cheese

Whey powder

Concentrated wheyfor animal feed

Fractionationof solids

Lactoseconversion

Demineralized wheyLactose

Hydrolysis Fermentation

Glucose/galactosesyrup

Whey drinks

Whey syrup

Alcohol

Lactic acid

Gas

Penicillin

Cheese fines


high humidity in barns housing whey-fed pigs,and respiratory problems are a very real deterrentto maximizing profits from intensive piggeries.In addition, the whey supply in many regions isseasonal; it dries up between October and April.The pig farmer is therefore vulnerable to thesefluctuations in feed supply, and is more likelyto opt for feed sources available all year such asbarley plus soya. At the same time, the summersurplus of whey cannot be utilized by the pigfarmer with a fixed number of pig spaces. Inaddition, whey is deficient in protein and fat-soluble vitamins and requires balancing withappropriate additives. The difficulties recentlyexperienced by the European pig industryindicate how volatile this industry is and that itis not a guaranteed ‘sink’ for whey. Many wheyproducers would prefer to have several relativelysecure outlets for their whey.

WHEY POWDER

Whey is processed by several methods. One ofthe most common methods is the production ofwhey powder by drying. Table 4 outlines wheypowder production in Europe from 1995 to1998. Drying sounds like a simple process, buta number of stages are necessary to produce a

Table 4. Whey powder production in Europe (thousand tonnes).1

Europe 1995 1996 1997 1998(15 countries) (estimated)

Production 1280 1310 1280 1310Consumption 1235 1235 1172Import 539 534 485 of which, from outside EU 9 2 2Export 654 703 703 of which, from outside EU 54 77 110 100

1Eurostat, 1999.

Table 5. Production of lactose in Europe (1995-1998) (thousand tonnes).1

Europe 1995 1996 1997 1998(15 countries) (estimated)

Production 280 310 330 330Consumption 195 225 230 240Export to outside EU 85 85 100 90

1Eurostat, 1999.

high quality powder. The free-flowing wheypowder resulting from drying has a number ofapplications in the food industry, includingincorporation into dry mixes, baked foods,confections, ice cream and processed cheese.Whey may also be made into cheese such as theScandinavian primost and mysost (Edelman andGrodnick, 1986).

Reverse osmosis may also be used toconcentrate whey. It works on the same principleas ultrafiltration except that membranes are suchthat only water is allowed to pass through them.Reverse osmosis is cheaper than thermalevaporation for removing up to 70% of the waterfrom whey. Savings are therefore possible byusing reverse osmosis for pre-concentrationbefore evaporation or membrane processing(Burgess, 1980).

LACTOSE MANUFACTURE

One of the oldest methods of whey utilization isthe manufacture of lactose. The first stage isconcentration of the whey by low temperatureevaporation, which ensures that the whey proteinis not damaged. The concentrate, containing 60-65% solids, is fed to a crystallization tank andseeded with small crystals of lactose. Edible

70 J. O’Shea

lactose is able to absorb flavor; and it is thereforeused in certain applications in the food industry.Lactose may also be used as a wine sweetener.For pharmaceutical applications, e.g., tabletformulation, the lactose crystals must be furtherrefined to remove traces of protein and minerals.Table 5 outlines lactose manufacture in Europefrom 1995 to 1998.

LACTOSE HYDROLYSIS

Lactose is not very sweet compared to canesugar or even its constituent monosaccharidesglucose and galactose. The sweetness of lactosecan be considerably increased by hydrolysis. Anumber of technologies are available for lactosehydrolysis. These can be classified as eitherenzymatic methods using ß-galactosidase or ionexchange methods (Burgess, 1980).

DEMINERALISED WHEY

One of the main disadvantages of whey as afood product is its high mineral content. Wheydemineralisation has opened up the baby foodmarket to whey producers. Two differenttechnologies may be used for thedemineralisation of whey, electrodialysis and ionexchange. Of these techniques mostmanufacturers use electrodialysis. In the ionexchange process whey is passed through acolumn containing a strong cation resin and thenthrough a column containing a weak anion resin.The former absorbs the cations in the whey andreleases hydrogen ions, while the latterneutralizes the hydrogen ions and absorbs theanions. The advantage of ion exchange lies inits relative simplicity and low capital cost.However, the regeneration costs are high and alarge volume of waste is generated. Ionexchange is probably cheaper thanelectrodialysis for high levels ofdemineralization. In contrast, the capital costsof electrodialysis are high while running costsare relatively low.

PROTEIN RECOVERY FROM WHEY

Whey protein is a relatively high value product.Ultrafiltration is one of the processes used toextract whey proteins. Whey proteins with a

molecular weight of 12,000-13,000 areextracted, leaving the permeate suitable forfermentation. The main feature of ultrafiltrationis the semi-permeable membrane, usually madeof cellulose acetate, polyamide or othercopolymers. The membrane acts as a filter thatis permeable to the low molecular weight saltsand lactose, but impermeable to the highermolecular weight proteins. Pressure is appliedto the whey side of the membrane and water,salts and lactose are forced through it whileprotein is retained. The protein concentrationin the whey therefore increases in proportion toits reduction in volume and a whey proteinconcentrate containing up to 60% protein on adry weight basis can be produced. Afterultrafiltration, the whey protein concentrate isevaporated to 40-50% total solids by lowtemperature vacuum evaporation and spray-dried. The whey proteins have a high nutritionalvalue.

ALCOHOL PRODUCTION

After ultrafiltration there remains a large volumeof permeate that needs to be processed. Theprotein-free permeate is a yellow liquidcontaining 90% of the whey solids, includingmost of the lactose. The composition of wheypermeate (Table 6) and that of skim milk areessentially similar.

Table 6. Average composition of whey permeate.

Component %

Lactose 4.8Protein 0.6Non-protein nitrogen 0.1Ash 0.5Fat 0.5Lactic acid 0.15Total solids 5.7

It has been known for some time that whey canbe fermented to produce alcohol; but in the pastsynthetic alcohol produced from crude oil hasbeen cheap. Following the energy crisis in the1970s, the price of synthetic alcohol increased,making fermentation alcohol productionprocesses more economically attractive. Wheypreviously regarded as a waste product came to


be viewed as a potential substrate for alcoholproduction. The technology was readilyavailable; and today a number of industrialplants successfully produce alcohol from wheypermeate.

Lactose is the major constituent of wheypermeate and accounts for approximately 70%of the total solids. It must be stressed that this isan average analysis for sweet whey. The typeof cheese produced will obviously influence thewhey composition. Small amounts of othernutritionally desirable materials includingriboflavin, amino acids and relatively high levelsof calcium and phosphorus are also present inwhey. To obtain the maximum reduction inBOD, it is necessary to develop a processwhereby practically all the lactose is convertedto alcohol.

In practice, a lactose content of 4.3-5.0% isobtained in the permeate. Using the yeast strainKluyveromyces marxianus a yield of 86% isattainable from an initial lactose concentrationof 4.3%. This yeast has the ability to fermentlactose, a disaccharide composed of glucose andgalactose (Figure 2).

The alcohol concentration followingfermentation of permeate is low thus requiringa very high energy input in the subsequentdistillation. On the other hand, the pollutionproblem is greatly decreased with the total solids(lactose and protein) being reduced to 10% ofthat initially present.

Biomass produced during the fermentation willadd to the effluent load. Whey pH ranges from4.6 to 5.6. The temperature range forfermentation can be 25-35°C. Using a batch

process the fermentation time will be between18 and 24 hrs. At the end of fermentation theyeast cells may be separated for re-use. If a yieldof 80% is attained from a lactose concentrationof 4.4%, then production of 1 liter of 100%ethanol requires 42 liters of whey. The qualityof the alcohol produced is very good with onlyvery low levels of acetaldehyde and fusel oils.The permeate can be further concentrated; andthe higher lactose content will increase alcoholcontent and thus lower distillation costs. This isvery useful where permeate must be transportedto alcohol plants.

One of the greatest difficulties facing a wheyalcohol plant manager is infection. In certaininstances there will be a carryover of lactic acidbacteria from the cheese plant. Thesemicroorganisms grow very easily on wheypermeate thus reducing the amount of substrateavailable for alcohol production. Infection insuch fermentations is usually controlled by usingantibiotics.

Not withstanding such difficulties, utilizationof whey as a substrate for alcohol productionon a worldwide basis has become increasinglypopular. This is evidenced by the diversegeographical locations of whey alcohol plants,examples of which are given in Table 7.

Considerations relevant to constructionof a whey alcohol plant

EFFLUENT TREATMENT

Although ultrafiltration removes the protein from

C12H22O11 + H2O →→→→→ 4C2H5OH + 4CO2

Lactose + water Alcohol + carbon dioxide

Figure 2. Structure of lactose and its conversion to alcohol by Kluyveromyces marxianus.

α-Lactose

HOCH2

O

HOH

H

H

OH

HO

O

H

H

OH

HOCH2

O

HOH

H

H

OH

H

72 J. O’Shea

whey and the fermentation process converts themain ingredient (lactose) into alcohol, biomassis produced. The other ingredients of whey willcontribute to the BOD/COD of the final effluentfrom the alcohol plant; and adequate effluenttreatment facilities must be provided. One optionis to use an anaerobic digestion system. Thiswill convert a portion of the BOD into gas thatcan be used to fuel the boilers in the distillationplant.

MARKETS FOR ALCOHOL

The alcohol markets in the US and Europe arevery different. Because of governmentintervention in the US, ethanol production hasgrown from an insignificant amount in 1978 toa record 6.4 billion liters in 1998. In Europe thesituation is very different. There has been nosuch government intervention and anoversupply of alcohol has existed for some time.Until recently a large portion of this alcohol wassold to Russia, but this market has diminished.The low price of oil has helped the majority ofalcohol producers to keep production costsdown; however as oil prices rise these productioncosts will also increase.

In the US, ethanol was promoted as a solutionfor a variety of complex problems. Among them:US dependence on foreign oil supplies, which

was made apparent by two oil crises of the 1970s,and the low gasoline octane ratings caused byreduced use of lead after the approval of theClean Air Act in 1977. Ethanol production fromcorn was seen as a way to boost farm incomescaused by the grain surplus in the wake of theSoviet embargo. Also, addition of ethanol togasoline was seen as a means to reduce airpollution. These provide a ready market for anyalcohol produced. Within the EU there has beenmuch discussion about bio-ethanol, but onlyrecently has there been any commitment to bio-fuels. Tax concessions for pilot plants producingbio-fuels have been allowed; and as aconsequence new bio-fuel projects have beenannounced in the Netherlands, Sweden andSpain. France has made the most progress. In1996 the French approved a draft law that madethe use of oxygenated components in fuelmandatory by 2000.

Conclusions

The attachment of a whey alcohol plant to a dairyplant producing the appropriate substrate solvesa number of major problems confronting wheyproducers. Alcohol production provides aguaranteed outlet for the whey that is not subjectto the fortunes of pig markets or other elementsoutside their control. Whey production (or

Table 7. Alcohol plants using whey as a substrate.

Ireland Carbery Milk products, of Ballineen, County Cork, operate an alcohol plant using cheese whey assubstrate. This alcohol is used for vodka and cream liqueurs (Murtagh, 1995).

United States A plant to produce industrial alcohol from whey has been installed in California in one of the largestcheddar cheese plants in the world. This plant processes 1,000,000 kg of milk per day into cheese. Thewhey from the process is converted into permeate and protein by ultrafiltration. The protein is dried andthe lactose in the permeate is converted into alcohol. The alcohol is then distilled in a distillation plantcapable of processing 200,000 liters of whey per day, producing 1.5 million liters of pure alcohol per year.

New Zealand Whey is the main by-product of the New Zealand dairy industry and is produced in the manufacture ofcheese and casein. Anchor Ethanol was the first company in the southern hemisphere to manufacturealcohol from casein whey on a commercial basis. Anchor Ethanol began production at Anchor ProductsRepora in 1980 and a second plant was commissioned at Anchor Products Tirau the following year. Theseplants produce a combined total of 10 to 11 million liters of ethanol per year, over half of which isexported. This alcohol is the basis for a range of alcohol beverages. It is also used in the manufactureof solvents, methylated spirits, white vinegar, surgical spirit, food colouring, deodorants, perfumes,aerosols and pharmaceutical products (New Zealand Dairy,1999).

Russia In Russia the range of feedstocks used for alcohol production is very diverse, including grains, fruit, whey,wine and molasses (Berg, 1999).


disposal) does not form a limiting step in cheeseor casein production. The whey proteinrecovered by ultrafiltration is a valuable item,the sale of which contributes to processeconomics. There is also a universal movetowards bio-fuels, therefore the markets foralcohol should be increasingly secure in thefuture.

References

Berg, C. 1999. World Ethanol Production andTrade to 2000 and Beyond. http://www.distill.com/berg.

Edelman, E. and S. Grodnick. 1986. The IdealCheese Book. http://cbs.infoplease.com/ce5/CE010398.html.

Eurostat. 1999. Institut National de Statistique,Brussells, Belgium, http://europe.eu.int/en/comm/eurostat/eurostat.html.

Burgess, K.J. 1980. Uses of waste dairy products.Technology Ireland, June, pp. 43-44.

Murtagh, J.E. 1995. Molasses as a feedstock foralcohol production. In: The Alcohol Textbook(T.P. Lyons, D.R. Kelsall and J.E. Murtagh,eds). Nottingham University Press,Nottingham, UK. pp. 27-34.

New Zealand Dairy.1999. http://www.nzdairy.co.nz/public/educational/whey/ethanol/Ethanol~t.htm.

Nielson, W.K. 1997. Whey processing. TechnicalBulletin. APV Ireland Ltd, Dublin.

Treatment and fermentation of molasses when making rum-type spirits 75

Chapter 8

Treatment and fermentation of molasses when making rum-type spirits

ROBERT PIGGOT


Global molasses production increasing

World molasses production has increasedmarkedly over the past 40 years from around15 million mT in the early 1960s to around 45million mT by 2000 (Anon., 2002, Table 1).While intricately tied to the production of rumin the Caribbean region, sugar crops are alsothe feedstock for over 60% of world fuel ethanolproduction (Berg, 1999). Around 75% of theworld’s molasses comes from sugar cane(Saccharum officinarum) grown in tropicalclimates (Asia, South America), while theremainder comes primarily from sugar beet (Betavulgaris) grown in the more temperate climatesof Europe and North America.

Table 1. World molasses production by region.

2002/03* 2001/02 2000/011000 mT

EU 3,950 3,650 3,966Western Europe 4,800 4,359 4,856Eastern Europe 2,420 2,391 2,224Africa 3,340 3,307 3,260North/Central America 6,830 6,677 6,874South America 12,900 11,626 10,182Asia 18,120 18,019 16,511Oceania 1,080 1,023 1,030World 49,490 47,402 44,957

Anon., 2002*Predicted

The perfect feedstock?

At first sight molasses appears to be the perfectfeedstock for production of alcohol. All thesugars are present in a readily fermentable form.Unlike grains, which contain starch instead ofsugar, molasses need not be cooked orprocessed. It is also in liquid form, which makesit easily handled. However the reality is thatmolasses has a range of problems and quirksthat make it far from a perfect feedstock, andknowledge of its composition and characteristicsare very important if it is intended as a substratefor alcohol production.

Molasses production

There are several types of molasses (Table 2)defined by sources and process, however theinformation in this chapter will focus on themolasses produced as a by-product of sugarproduction from cane and beets. Cane or beetmolasses is the end product after the sugar hasbeen removed by crystallization. Cane sugar isproduced by first heating the juice crushed fromcane. It is then clarified by filtration with limeadded to precipitate fiber and sludge. The slurryis then evaporated to concentrate the sugar andallow it to crystallize. The syrup containing the

76 R. Piggot

crystals is then centrifuged to separate the crystalsand the syrup residue. This first residue, whichstill has a high content of sugar, is referred to as‘A molasses’. The process is repeated to yieldmore sugar and subsequent B, C, or even Dmolasses grades. Sugar mills normally evaporateand centrifuge a maximum of three times, butthe number of treatments will depend onmarketplace economics. Obviously successivecrystallizations reduce the amount of sugarpresent and lessen its value as a fermentablesubstrate for ethanol production. Repeatedcrystallization also tends to degrade molassessugars to unfermentable compounds (Murtagh,1999).

Table 2. Types of molasses.

Blackstrap molasses By-product of sugar productionfrom sugar cane

High test (cane) molasses Primary product: extracted fromsugar cane

Refiners cane or beet By-product of refining rawmolasses brown sugar to produce white

sugar

Beet molasses By-product of sugar productionfrom sugar beets

Citrus molasses Juices extracted from themanufacture of dried citrus pulp

BLACKSTRAP MOLASSES

Blackstrap (cane) molasses is used in animalfeeds and is also the most common feedstockwhen using molasses to make alcohol. Since itis a waste product and a commodity typicallygathered from many plants, unless the source isknown it can be expected to vary considerably.Even from the same plant, different operatingconditions can produce a product that fermentsvery differently from batch to batch.

HIGH-TEST MOLASSES

In contrast to other types of molasses, high-testmolasses is not a by-product. It is producedinstead of sugar as opposed to being a by-product of sugar production. It is clear, lightbrown and of controlled composition. The

process for making high test molasses is exactlythe same as the production of the sugar, however,the sugar cane syrup is instead just concentrated,partially inverted using either invertase fromSaccharomyces cerevisiae or acid (not favoredas it destroys about 5% of sugars) to preventcrystallization in the final product. Apart fromthe fact that high test molasses is more expensive,this product has several advantages over typicalmolasses as a feedstock for the production ofethanol. It is lower in ash, has fewer dissolvedimpurities and has a more accurate and controlledsugar profile and sugar content. Typicalcomposition is summarized in Table 3.

Table 3. Typical composition of high test molasses.

°Brix 85.0Sucrose, % 27.0Reducing sugars, % 50.0Calcium, % 0.2Ash, % 2.25Water, % 15.50

Chen and Chou, 2003

CANE VS. BEET MOLASSES

The molasses obtained from cane and beet sugarproduction are similar but not identical. Molassesproduced from beets cannot legally be used forrum and beverage produced from beet molasseshas an off flavor and character. Beet molasses isalso slightly lower in sugars than cane molasses;and its sugar profile consists almost entirely ofsucrose (Table 4). Organic acid types andamounts in beet molasses vary considerably, asdoes the ash content. The content of otherminerals is slightly lower in beet than in canemolasses. The pH of beet molasses is slightlymore neutral (around pH 7) than cane molasses,which is typically pH 5 to 5.5.

Since cane and beet molasses are agriculturalproducts, chemical composition will be affectedto varying extents by climatic conditions, soil,fertilization practices, moisture, etc., as well asthe method of production and sugar removalfrom the product. This also means that molassescomposition can vary greatly among refineriesor even from the same refinery during a singlecrushing season.


Table 4. Comparative chemical composition of beet andcane molasses.

Cane Beet

Range or typical value (%)

Total sugars 45-55 48-52Sucrose 25 to 35 46-52Reducing sugars 20 to 35 0.5-4.0pH 5-5.5 ~7.0Ash 10-16 10-12Non-sugar organic matter 10-12 12-17Starch/polysaccharides 0.5%

Calcium 0.4-0.8 0.1-0.5Sodium 0.1-0.4 0.5-1.0Potassium 1.5-5.0 2.0-7.0Magnesium 0.05-0.98 0.1-0.15Phosphorous 0.03-0.1 0.02-0.07Sulfur 0.3-0.8 0.15-0.5

Range or typical value (ppm)

Biotin 1.2-3.2 0.04-0.13Folic acid 0.04 0.2Riboflavin 2.5 0.4Copper 2.2-38 2.5Manganese 4-300 3.0Zinc 4-48 0.15

United Molasses

Basic sugar chemistry helpful inunderstanding molasses as a substrate

SUCROSE STRUCTURE AND HYDROLYSIS

Sucrose, common table sugar, is a disaccharidecomposed of one glucose and one fructose

molecule. Sucrose is the principal sugar inmolasses. On hydrolysis (reaction with water),sucrose yields the two monosaccharides (Figure1). It should be noted that the molecular weightof sucrose is less than the total of its componentmonosaccharides because a molecule of wateris displaced from sucrose and transferred to thesimple sugars.

INVERT AND REDUCING SUGARS, SUGARMEASUREMENTS IN MOLASSES

In aqueous solutions, monosaccharidescontaining five or more carbons occur in cyclicforms. These pyranose (6-carbon) or furanose(5-carbon) rings form when the aldehyde (orketone) group reacts with a hydroxyl group atthe other end of the molecule to form ahemiacetal (or hemiketal) linkage. Another chiralcenter is produced, known as the anomericcarbon (Figure 2). The hydroxyl group (OH) canbe either above the plane of the ring (ß-glucose),or below the plane (α-glucose). α- and ß-glucoseare known as anomers, because they differ onlyin the position of the group at the anomericcarbon.

It is this free anomeric carbon that makesglucose and fructose reducing sugars, whichrefers to their ability to reduce a standard coppersulfate solution such as Fehling’s solution.Sucrose does not have this property. The twomonosaccharides are also referred to as invertsugar. This is derived from the fact that sucroserotates polarized light to the right (dextrorotary),while the two hydrolysis products rotatepolarized light to the left (levorotatory). During

Sucrose(MW=342)

+ water(MW=18)

Glucose + Fructose(MW=180 + MW=180)

C12H22O11 C6H12O6 C6H12O6H2O

+

SucroseCH2OH

O

O

HOH

H

H

OH

H H

OH

CH2OH OHO

H

OH

HO

H

CH2OH

FruGlu+ water

H2OCH2OH

OH

O

HOH

H

H

OH

H H

OH

CH2OH OHO

H

OH

HO

H

CH2OH

FruGluH

+

Figure 1. Hydrolysis of sucrose.

78 R. Piggot

fermentation yeast produces an enzyme(invertase) that breaks sucrose down into invertsugars.

Total sugars as invert (TSAI) is a measurementcommonly used to gauge molasses sugarcontent. This is either a calculated value or basedon reducing power of a sample, which is at besta crude measurement. Measuring reducingpower assumes that all the reducing compoundsin molasses are sugars. Thus any sodium sulfiteadded would appear as sugar in such a test. Thereis often up to 5% difference in this test methodover the actual amount of reducing sugarspresent. Sugar measurements for molasses aresummarized in Table 5.

DEXTRANS

Dextrans are sucrose molecules that havepolymerized into long chains. This polymerization

can occur when molasses has been stored for along period of time or if it has been contaminatedwith Leuconostoc mesenteroides. While dextransappear as reducing sugars, they are notfermentable. A bad contamination will result inmolasses having long ‘strings’ or a ropeyappearance caused by dextran formation. Arecent development by Alltech has been use ofdextranase (used at a rate of 0.01% by weightof molasses) to hydrolyze the dextran chains.

Both sucrose and its component mono-saccharides are readily fermented by yeast.

BRIX VALUES

Repeated evaporation and centrifugationincrease molasses viscosity and concentratessalts and other impurities to form a thick, viscous,brown liquid. The concentration of molasses isnormally measured in degrees Brix. The Brix

β-D-glucose

C

C

C

C

C

CH2OH

OH

H

OH

OH

H

HO

H

H

H O1

2

3

4

5

6

D-glucose (linear)

α-D-glucose

Free anomeric carbon(the reducing end of the sugar)

CH2OH

O

OH

H

H

OH

OH

H4 1

5

6

3 2

OH

H

CH2OH

O

OH

H

H

OH

OH

H4 1

5

6

3 2 OH

H

Figure 2. Linear and ring structures of glucose.

Table 5. Measurements of molasses sugar content used in trade.

Term Description/calculation

TSAI (Total sugars as invert) (Sucrose (%)/0.95) + invert sugars (%)TSM (Total sugars) Sucrose (%) + invert sugars (%)TSAS (Total sugars and sucrose) Sucrose plus the amount of invert sugar converted to sucrose: sucrose (%) + (invert (%) x 0.95)Brix The sugar content of a solution with the same specific gravity of a molasses sample (same

as the Balling scale). Measured using either a refractometer or hydrometer.


scale is the same as the Balling scale, which is ameasure of what the sugar content of a liquidwould be if all the dissolved and suspendedsolids were sugar. Expressed another way, it isthe sugar content of a sugar solution with thesame specific gravity as the sample. Thus, an80°Brix molasses has a specific gravity of1.416, which is the same as a sugar solutioncontaining 80% sugar by weight. Brix ismeasured using a hydrometer originallyintended only for sugar solutions. Unfortunatelywith molasses, the presence of other dissolvedsolids means that Brix readings often bear littlerelationship to the amount of sugar actuallypresent. Despite this, Brix values are quitecommonly used in the sugar industry; andmolasses is often defined as around 80°Brix.Brix values determined by refractometer andhydrometer rarely agree, so the method bywhich the reading was obtained should beclearly stated.

Molasses handling and storage

Molasses is generally quite stable and can bestored for long periods of time if properlymanaged. That said, a 2.3 million gallon tankof molasses measuring 90 ft in diameter and 52ft high exploded in Boston in 1919 killing 21people. The ‘wall of goo’ smashed buildings andknocked out stone supports for an elevated train.Needless to say it took months to clean up, andthe distillery was fined extensively for damages.Paradoxically, this odd, but true urban legendoccurred in January when temperatures warmedto an unseasonable 46°F. Clearly proper designand temperature control are important formolasses storage tanks. It can be stored in eithersteel tanks or underground concrete tanks. Bothare very satisfactory, however, it is criticallyimportant that the molasses is kept above70°Brix. This means it must be kept out of therain and free from any condensation inside thestorage vessel. The storage should be wellventilated. If moisture collects even in smallpools on the surface, then mold and otherorganisms held in check by the high sugarcontent in the molasses will flourish.

During processing in the sugar refinery andduring handling, air can become entrained in

the molasses. In addition, carbon dioxide isevolved during molasses degradation. Carbondioxide bubbles rise very slowly in the molasses,meaning that density at various levels in a storagetank will be quite different. This makes itimpossible to calculate with precision from amolasses sample taken from a single point in thetank and a measure of the liquid height to tellexactly how much molasses is present in the tank.The way to more accurately measure the amountof molasses present is to take a reading with adevice such as a pressure cell that will measurethe actual weight of the molasses in the tank.This will not significantly vary with the amountof entrained air. It also means that productionprocedures are more consistent in that it ispossible to know exactly how much molasseshas been processed.

VISCOSITY: DILUTING MOLASSES

Molasses is well known for being viscousespecially at lower temperatures. Molassesviscosity varies widely, depending on severalfactors including dry solids content. Someequipment is designed to take advantage of thisviscosity to help move the molasses. As a generalrule, pumps should be positive displacement withinlets as large (at least 8 to 10 cm) and as directas possible. The pump should have large inletcavities and turn as slowly as possible to allowtime for the molasses to properly fill the chamber.A pump turning too quickly will cavitate.

Dilution will assist in movement; and dilutionto 40°Brix is typical. Dilution should occur justprior to the fermentation or clarification/pasteurization process, however, as microbes canstart to grow at the reduced osmotic pressurecaused by dilution.

When diluting molasses, it must be rememberedthat the Brix scale measures on a weight % basisand all calculations must be based on weight andnot volume. 80oBrix molasses has a specificgravity of 1.416, therefore a gallon weighs about11.8 lbs and a ton contains about 169.5 gallons.Example calculations for diluting 80°Brixmolasses to 40° and 25°Brix are given in Table6. Temperature of the water and molasses willhave a large impact on these calculations (Figure3).

80 R. Piggot

DEGRADATION DURING STORAGE ANDPROCESSING

As illustrated in Figure 4, heating will lowerviscosity; however little is gained bytemperatures above 35°C and it is advised tokeep temperature below 40°C to reduce thermaldegradation (Figure 5). If the molasses is to beheated in a tank it is recommended to use a coilhot water system rather than steam. Heatingshould be done gently to avoid spot heating.Three hours at 120°C can reduce sugar contentby as much as 10 to 15%. At 130°C sugar lossincreases to 25-35% (United Molasses).

Fermentation

Molasses makes a good base for fermentation.In fact, most of the yeast produced in the world

is made using a molasses substrate. Howeverthere are several potential problems withmolasses that must be considered if fermentationsare to be efficient and cost-effective.

The amount of fermentable sugar varies greatlyin molasses. In order to standardize, yieldcalculations are usually based on 55% sugars.A good fermentation with high quality molassesshould yield about 290 liters of ethanol per tonof molasses. However recent molassesshipments have yielded more on the order of260 liters/tonne. Molasses prices are setinternationally; and typically run $USD35-40 perton FOB with freight added raw material costper liter of alcohol can be $0.20-0.25.

With the increasing efficiency of the sugarindustry in extracting sugar, the molassesremaining for use in fermentation/distillationoperations is becoming poorer in quality. Someplants are finding it necessary to sterilize and

Table 6. Molasses dilution: dilute 1 ton (or tonne) of 80oBrix molasses to 25oBrix.

Gallons Liters

Total tons: 80/25 = 3.2 tons (or tonnes) at 25oBrixVolume required at 25oBrix Molasses (1 ton or tonne at 80oBrix*) 169.5 706 Water (2.2 tons or tonnes) 528 2,200

Total volume produced at 25oBrix 697.5 2906

*assuming 80oBrix molasses has a specific gravity of 1.416 at 20°C.

Temperature = 20O

C

Brix = 80O

Purity = 80%

Density = 1414.1 kg/cm3

20 30 40 50 70 80

1375

1380

1385

1390

1395

1400

1405

1410

1415

Temperature (OC)

Densit

y (kg/m

3)

60

Figure 3. Effect of temperature on density of molasses.


clarify before fermentation, while others employdextranase to ensure maximum extraction ofwhatever sugar is present. This is accomplishedby diluting the molasses to approximately40°Brix, then heating it. Gums and othersuspended solids drop out of the diluted solution.Following heating to pasteurization temperaturefor an appropriate time the molasses is cooledand sent to fermentation. The disadvantage ofthis process is the energy input required as wellas some loss of sugar.

Molasses at 80oBrix will not ferment withoutdilution as the sugars and salts exert a very highosmotic pressure. It is therefore necessary to

dilute the molasses to below 25oBrix. Yeast willnot start fermenting rapidly above this point; andcontamination may develop before the yeastbecome established since molasses is laden withcontaminating bacteria. Some distilleries acidifythe wash to help reduce infection. Sulfate addedat this point can come out later as calcium sulfatescale in the wash column.

Brix values of about 25o are the highestosmotic pressure that yeast will tolerate to beginfermentation, however this means a low sugarcontent and potential low fermentor alcohollevels. If sugar content is initially 46%, thendilution to 25°Brix results in only ~14.3% sugar

0

10 15 20 25 30 35 40

Temperature (°C)

Vis

cosi

ty c

enti

po

ise

40000

30000

20000

10000

Figure 4. Effects of temperature on the viscosity of cane molasses (R&H Hall, 1999).

-4

-3

-2

-1

03 6 9 12 15

Days

%

60°C

40°C

Figure 5. Sugar losses in molasses stored at 60 or 40°C over 15 days (United Molasses).

82 R. Piggot

(Calculated: 25 x 46/80 = 14.3). This results inabout 8% alcohol. This is low by today’sstandards and results in high hydraulic loadingand large volumes of effluent. One way toincrease the alcohol content of the fermentationis by incremental feeding. After the fermentationhas been going several hours and has droppedto 15 or 16°Brix, additional molasses (at40°Brix) is added to bring concentration to20°Brix and this mixture is then fermented. Thisallows yeast to become rapidly established.

YEAST AND YEAST CONDITIONING

It is important that very high quality yeast areadded at high cell counts to the fermentation.Unlike grain fermentations where the sugar orglucose is produced throughout fermentation,all the fermentable substrate in molasses ispresent at the beginning of fermentation and itis equally available to yeast and any undesirablecontaminants that might be present. In the battledetermining what species is to dominatefermentation, healthy yeast are critical. Manydistilleries condition the yeast before adding itto fermentation. This is done by using ~10% ofthe final fermentor volume diluted to about16°Brix. Yeast is added, as is a small amount ofair, yeast food (eg, AYF™ at 10-15 ppm) andpossibly antibiotics to ensure that otherorganisms are not being grown. A low pH and atemperature around 28°C also help keepcontamination in check. It is important thatconditioning tanks are kept clean and that yeastis pumped to the fermentor while still in theexponential growth phase.

Molasses fermentations are very rapid, oftencompleted in 24 hrs, and so much heat isgenerated that a good cooling system is veryimportant. Since many of these fermentationsystems are located in tropical regions wherecooling is limited, a temperature tolerant yeaststrain can be of great value. Thermosacc™, ahigh temperature yeast (97-100°F) designedoriginally for grain distilleries, is now findingapplication in this industry.

Yeast nutrition

The free amino nitrogen (FAN) available inmolasses varies widely for a number of reasons,

including variable amounts in the original syrupand/or losses during during processing andstorage. Unless FAN can be tested on a veryregular basis it is recommended that a sufficientsource of FAN is added to ensure goodfermentations. Nitrogen should not be in theform of ammonium sulfate as it will add to thescaling problem by forming calcium sulfate.Likewise, liquid ammonia is undesirable as ittends to raise the pH and encourage bacterialcontamination unless it is counteracted with acidadditions. (Some plants using liquid ammoniaeither use sulfuric acid to balance the pH, whichintroduces the undesirable sulfate anion, or usephosphoric acid, which generally introducesmore phosphate than necessary).

Urea may be used to supply nitrogen inmolasses fermentations for fuel ethanolproduction, but its use for beverage alcoholproduction should be approached with caution.Urea usage may lead to the production ofcarcinogenic ethyl carbamate, which is un-acceptable in alcoholic beverages. If phosphorusis deficient in the molasses, diammoniumphosphate may be added with a correspondingreduction in urea or other nitrogenous nutrient.Generally, blackstrap molasses requires no otheradded nutrients for fermentation.

The three B vitamins most likely to be essentialfor the satisfactory growth of any particularstrain of Saccharomyces cerevisiae yeast arebiotin, pantothenic acid (pantothenate) andinositol. Biotin is normally present in greatexcess in cane molasses, but is normally deficientin beet molasses. If a yeast strain that does notrequire biotin cannot be obtained (they are rare),it may be advantageous to mix in 20% canemolasses when trying to ferment beet molasses,or at least to use a 50:50 cane:beet molasses mixfor initial propagation of the yeast.

Pantothenate is normally borderline-to-adequate in cane molasses, but is generally inexcess in beet molasses. Mixing some beetmolasses with blackstrap cane molasses may beadvantageous if it is available. Pantothenateconcentrations may also be low in HTM and inrefiners cane molasses (the by-product ofrefining raw brown sugar to produce whitesugar). In these instances, it may help to addsome beet molasses to the mash although onemay obtain yeasts that do not require externalsources of pantothenate for growth.


Inositol may be deficient in HTM, but thisdeficiency is less critical as it is relatively easyto find yeasts that do not require external sourcesof inositol.

CONTAMINATION OF MOLASSESFERMENTATIONS

The principal contaminant of molassesfermentations is Leuconostoc mesenteroides,which causes sucrose molecules to polymerizeinto unfermentable dextran chains. If thecontamination is very extensive, the molassesmay appear to be ‘ropey’. Dextran may also befound in refiners molasses produced from rawsugar that has been stockpiled for a number ofyears. Apart from the loss of yield, the presenceof dextran increases foaming in fermentors.

Another important occasional bacterialcontaminant is Zymomonas mobilis. Thisbacteria can ferment sugars to ethanol, but hasthe side effect of reducing sulfur compounds toproduce a hydrogen sulfide smell, which is veryundesirable for production of good quality rum.

Distillation

There are several issues that are somewhatunique to the distillation of molasses. Scaling inthe wash column is a serious problem. It is largelycaused by lime used in the clarification processat the refinery. The most common ways tocombat this are to use a non-fouling tray designand/or run the column under vacuum to keepthe temperature down and reduce scaleformation. Additives can be added to reducescaling and special care should be taken to keepthe alcohol percentage on the feed tray as lowas possible. High strength spirit on this tray willincrease scale accumulation.

Mud, or sludge, often settles at the bottom ofthe fermentor. If this occurs, then wash can bedrawn off at a point above the fermentor base.Sludge that has settled in the bottom of thefermentor can be discarded separately. A non-fouling tray should be able to handle these solidsand allow them to pass out in the dunder. If not,then evaporator fouling can occur whenevaporating the stillage to produce condensedmolasses solubles (CMS).

Molasses fermentations have a tendency to behigh in sulfur. This is especially true if the

fermentation is infected with Zymomonasmobilis.

Methanol is not common in molassesfermentation, because molasses contains verylittle pectin which is broken down in somefeedstocks during fermentation to formmethanol. Hence the amount of methanol inmolasses spirit is low.

Effluent treatment

The liquid effluent from the rum productionprocess (known as spent wash, stillage, vinasseor dunder) is a significant environmentalproblem. Biological Oxygen Demand (BOD) isin the range of 50,000 mg/l and the effluent isvery dark in color. There is no 100% effectivemethod of treatment available; and no singletreatment would suit all distilleries.

At present there are six primary ways thatmolasses effluent is treated.

1. Dilution and discharge into the sea

2. Land application

3. Deep well injection

4. Anaerobic digestion prior to one of the above

5. Lagoon treatment prior to one of the above

6. Condensation to CMS then used as a cattlefood supplement or incinerated

DILUTION AND DISCHARGE INTO THE SEA

This is a controversial method of disposal, butis used by several distilleries. It has been studiedin at least two cases and found not to be harmfulto the environment if done properly. The BODdissipates rapidly, but the color change can bevisible for miles from the outlet. This is bothvisibly unappealing and it also blocks light frompenetrating the water. In areas of the Caribbeanwhere tourism is a major industry, foul smellsand brown seawater are not acceptable.

LAND APPLICATION

Stillage can be applied to the soil surface andincorporated into the top 15 to 30 cm in threedifferent ways:

84 R. Piggot

Soil conditioner. This approach takesadvantage of the available organic and inorganicnutrients in effluent to enhance the growth ofcertain crops. This is a good option when adistillery is located in an area with the correctsoil type and crops. Unfortunately this is oftennot the case.

Compost/soil improvement. In land reclamation,or soil improvement to add organic andinorganic nutrients to marginal lands.

Disposal site. On a plot of land set aside for thepurpose of disposal, whereby the organicconstituents are degraded by natural processes(oxidation and reduction via sunlight) and theinorganic components are trapped within the soilmatrix.

DEEP WELL INJECTION

Injection of stillage in a deep well is one methodof land application employed, but has causedground water contamination in some areas.

ANAEROBIC BIODIGESTERS

Bacardi have done a considerable amount ofwork on anaerobic biodigesters. This processreduces BOD by about 80% but has little effecton effluent color. In order to reduce the BOD toacceptable levels and reduce color, an aerobicfinishing step may be required. This leaves asludge for disposal. This sludge may have someuse as an animal feed supplement. The methanegas from the anaerobic digester can be used torun a boiler in the distillery. The author’sexperience with this process is that it works well,but the value of the captured gas is not muchmore than the cost of operating the process,which results in a long capital payback.

LAGOON

Effluent held in a lagoon is sometimes treatedwith lime to adjust pH. Other types of treatmentsuch as aeration, enzyme and/or bacterialadditives, have been used in a few distillerieswith apparent success.

EVAPORATION TO CONDENSED MOLASSESSOLUBLES

Cane molasses stillage can be evaporated toabout 50°Brix and sold as an animal feedsupplement. Its comparatively low value makesthis option more of a ‘break even’ prospectdepending on the cost of energy to run theevaporators. While CMS can be incinerated,additional cost is incurred. Use as a fertilizer isanother option, however demand is limited.

References

Anon. 2002. Third estimate of world molassesproduction 2001/02. F.O. Licht’s WorldEthanol & Biofuels Report. July 15.

Berg, C. 1999. World Ethanol Production andTrade to 2000 and Beyond. F.O. Licht. 80Calverley Road, Tunbridge Wells, Kent, TN12UN, England.

Chen, J.C. and C.C. Chou. 2003. Cane SugarHandbook: A Manual for Cane SugarManufacturers and Their Chemists. John Wiley& Sons, New Jersey.

Murtaugh, J.E. 1999. Molasses as a feedstockfor alcohol production. In: The AlcoholTextbook, 3rd Ed. (K.A. Jacques, T.P. Lyonsand D.R. Kelsall, eds). Nottingham UniversityPress, UK.

R&H Hall. 1999. Molasses in livestock feeds.Technical Bulletin No. 3. http://www.rhhall.ie/print/issue3_1999.html.

United Molasses.

Yeast and managem

ent of fermentation

Understanding yeast fundamentals 85

Chapter 9

Understanding yeast fundamentals

INGE RUSSELL

International Centre for Brewing and Distilling, School of Life Sciences, Heriot-Watt University,Edinburgh, UK

Introduction

The heart of the fermentation process is the yeastcell. Yeast is the least expensive raw materialinput into the fermentation process whether it isfor beer, distilled spirits or fuel ethanol, and yettoo often the importance of the yeast isundervalued. The yeast is ignored, abused andeconomized. A healthy and well-selected yeastis needed for an efficient fermentation. Whenselecting the strain, attention must be paid towhether a yeast is needed that can function atlow temperatures such as for lager beerproduction or at high temperatures for fuelethanol production. Whether the environmentrequires a yeast with the ability to survive highethanol concentrations, or whether the flavorcompounds that the yeast produces are of interestor concern. The yeast is an amazingly hardyorganism that can survive under many stressfulconditions, but under stress it will not be workingat its optimum. A careful choice as to the particularstrain and meticulous attention to growthconditions will lead to successful fermentations.

Traditionally yeast have been characterizedaccording to morphological and physiologicalcriteria. Taxonomists have grouped togetherbrewing (both ale and lager), distilling, wine,baking, and numerous wild yeasts, under thename Saccharomyces cerevisiae. This chapterwill only deal with Saccharomyces yeast,

specifically S. cerevisiae unless otherwise noted.This is the microorganism used globally forpotable and industrial ethanol production.Ethanol is the largest commodity produced byany microorganism; and world ethanolproduction in calendar year 2003 is expected toincrease to 38 billion liters/year from 34.3 billionin 2002.

What is a yeast?

A common definition is ‘yeast is a fungus wherethe unicellular form is the most predominant andwhich reproduces by budding (or fission)’. Theterm yeast is often used synonymously withSaccharomyces cerevisiae since this is thespecies that is produced in the largest amountsfor both baking and fermentation purposes.Saccharomyces yeast reproduces only bybudding. Yeast are considerably larger in sizethan bacteria and in terms of quantity andeconomics are the most important group ofmicroorganisms exploited for industrialpurposes. Over 700 different species of yeasthave been identified by taxonomists. Many arepoorly characterized, but the ones of industrialimportance have been well characterized. IndeedSaccharomyces, of which there are thousands

86 I. Russell

of unique strains, was one of the first organismsto have its entire genome sequenced andavailable on the internet.

Why is a yeast so effective atfermentation?

Yeast is unique in that it is the only livingmicroorganism that can switch from respirationto fermentation. An important fact to rememberabout yeast is that despite the presence ofoxygen, yeast will always take the fermentativeroute to utilize glucose if the glucose is there inhigh concentration (called the Crabtree effect).

The world of yeast is very small and wemeasure this microscopic universe in terms ofthe Greek letter µ for micron (a micron being10-6 meter). A single yeast cell measuring 7 µmx 7 µm would have a surface area of 153 µm2.Ten grams of pressed yeast would have a contactsurface area of ~10 m2. It is this large contactsurface area that allows yeast to so effectivelytake up sugars from the surroundingfermentation medium. A useful rule of thumb isthat one gram dryweight of yeast equates toapproximately 4.87 x 1010 cells.

Strictly speaking, Saccharomyces cannot beconsidered a true facultative anaerobe since theCrabtree effect takes precedence, i.e. above 1%fermentable sugar in the culture medium (theexact percentage varying between strains) theyeast ferments the sugar by anaerobicmetabolism, despite how well the culturemedium is aerated.

The main parts of a yeast cell

Yeast cells vary from oval to round and in sizefrom roughly 5 to 10 µm length to a breadth of5 to 7 µm, but there are exceptions ranging from2.5 to 21 µm. Yeast are roughly the size of a redblood cell and many times larger than a bacterialcell (Figures 1 and 2). The mean cell size varieswith the growth cycle stage, the growthconditions and the age of the cell (older cellsbecoming larger in size). The main componentsof the yeast cell are illustrated in Figure 3.

Under the microscope one can see the yeastcell wall along with multiple bud scars and abirth scar (Figure 4). The cell wall gives shapeand stability as well as cell-to-cell recognition.It acts as a permeability barrier to large solutesand plays a role in controlling the entry of waterinto the cell. The cell wall also protects the yeastcell membrane. The yeast cell wall is composed

Figure 1. Yeast and bacteria.

0.2 mm(200 µm)

20 µm

CE

LL

S

OR

GA

NE

LL

ES

MO

LE

CU

LE

S

x10

Minimum resolvable

by human eye

Minimum resolvable

by light microscope

Minimum resolvable

by electron microscope

YEAST: 5-10 µm

BACTERIA: 2-5 µm

ATO

MS

x10

2 µm

x10

200 nm

x10

20 nm

x10

2 nm

x10

0.2 nm

Figure 2. Relative sizes of yeast and bacteria.


primarily of mannan and glucan (Figure 5). Theyeast cell plasma membrane is primarily lipidsand proteins with a small amount ofcarbohydrate and acts as the barrier between thecytoplasm and the outside of the cell. It is veryimportant as it regulates what goes in and out ofthe cell (Figure 6).

The cell nucleus consists primarily of DNAand protein and is surrounded by a nuclear

membrane. The nucleus can be seen using phasecontrast microscopy. The periplasmic space isthe area between the cell wall and the cellmembrane. It contains secreted proteins that areunable to cross the cell wall. For example, someenzymes located here help in the catalysis ofsugars such as sucrose. Sucrose is broken downin the periplasmic space by the enzyme invertase,to fructose and glucose.

Figure 3. The main features of a typical yeast cell.

Mitochondrion

Bud vacuole

Nucleus

Bud

Secretoryvesicles

1µm

Golgi complex

Vacuole

Lipid granule

Bud scar

Cell wall

Vacuolar granules

Storage granule

Mitochondrion

Plasma membrane

Pore in nuclear membrane

Vacuole membrane

Ribosomes

Periplasm

Endoplasmic reticulum

Glycogen granule

Figure 4. Electron micrograph of a yeast cell with multiple bud scars (and birth scar).

88 I. Russell

The mitochondria can only be seen usingelectron microscopy and consist of structureswith two distinct membranes, the outer and theinner, and cristae within the mitochondria areformed by the folding of the inner membrane.Mitochondria are of importance since it is herethat the enzymes involved in the tricarboxylicacid (TCA or Krebs) cycle and in electrontransport and oxidative phosphorylation arelocated. Mitochondria contain self-replicatingDNA and protein synthesis systems. Ribosomes,the site of protein synthesis, are located in thecytoplasm. Vacuoles are part of theintramembranous system, which includes theendoplasmic reticulum. The size and shape ofthe vacuoles change during the cell cycle andvacuoles are the most obvious component of thecell when viewed under the light microscope.The vacuoles store nutrients and provide alocation for the breakdown of macromoleculessuch as protein. Lipid granules are found in thecytoplasm, as are high concentrations ofglycogen. Glycogen can be visualized under thelight microscope by staining the cells with iodine.

The lifespan of yeast

Individual yeast cells are mortal. Yeast aging isa function of the number of divisions undertakenby an individual cell and not a function of thecell’s chronological age. Lifespan is measuredby counting the bud scars on the surface of thecell. All yeast have a set lifespan determinedboth by genetics and environment. Themaximum division capacity of a cell is calledthe Hayflick Limit. Once cells reach this limitthey cannot replicate and enter a stage called

senescence, which leads to cell death. This isparticularly important with respect to recyclingyeast, which cannot be done without loss ofsome metabolic function. A new yeast cell(daughter or bud) can divide to produce 10 to33 daughters of its own depending on the strain.Industrial polyploid ale yeast strains have beenreported to have a lifespan varying from a highof 21.7 +/- 7.5 divisions to a low of 10.3 +/- 4.7divisions. Respiratory deficient (petite) mutantsof these strains were shown to have a reducedlifespan. When examining yeast under themicroscope the following are signs of aging: anincrease in bud scar number, cell size, surfacewrinkles, granularity, and daughter cells beingretained. Aging is accompanied by a progressiveimpairment of cellular mechanisms thateventually results in a yeast with a reducedcapacity to adapt to stress and to ferment (Smart,2000). Figure 7 shows the progression throughthe lifespan cycle of a yeast cell.

The chemical composition of yeast

A yeast cell consists of 75% water and 25% drymatter. The dry matter consists of:

Carbohydrate 18-44%Protein 36-60%Nucleic acids 4-8%Lipids 4-7%Total inorganics 6-10%

Phosphorus 1-3%Potassium 1-3%Sulfur 0.4%

Vitamins trace amounts

Figure 6. Structure of the plasma membrane.Figure 5. Structure of the yeast cell wall.

Cell wall

is a matrix

Plasma membrane

Cytoplasm

β-1,6-glucans

~Chitin

Glucan

layer

Manno

protein

layer

Fibrillar

layer

manno-

proteins

manno-

proteins

manno-

protein

protein

β-1,3-glucans

β-1,6-glucans

β-1,3-glucans


When a sample of yeast is analyzed, itscomposition will depend greatly on the genotypeof the yeast as well as the growth parametersand hence the differences in compositionbetween a yeast propagated for the bakingindustry vs a yeast propagated for use in thefermentation industry. Table 1, compiled froma number of sources, gives a good indicationas to the constituents of a typical yeast cell. Theexact values will vary with the particular strainsexamined, the composition of the media, as wellas many environmental factors. The generalcomposition of yeast guides us to the levels ofnutrients that the yeast may require in order to grow.

Yeast growth requirements

Specific requirements for yeast growth include:

i) water

ii) a carbon source - fermentable carbohydratesas an energy source

iii) oxygen/lipids - lipids for membranebiosynthesis, which can be made by the yeastif oxygen is present

iv) a nitrogen source - amino acids and peptidesare needed for growth and enzyme synthesis.Saccharomyces yeast can also use ammonia.

v) growth factors -vitamins

vi) inorganic ions - essential for yeastmetabolism

WATER

Water is a main component of all living materialand most microorganisms need at least 15% water

· Formation of new bud or

daughter cell

· Divisional age of 0

· Known as a virgin cell

· Mother cell continues

to divide until the Hayflick

Limit is reached

· Once reached, it may not

re-enter the cell cycle

· Note increase in cell size

· Virgin cell reaches critical

size for division

· Mother cell divides to

produce first daughter

· Acquires bud scar

· Mother and daughter cell

separate asymmetrically

· Mother cell is larger and will divide

more rapidly than the daughter

· Mother cell enters the

senescence pathway

· No further divisions are permitted

· Mother cell exhibits 'aged' phenotype

· Senescent mother cell

enters death metabolism

· Culminates in cell lysis

Figure 7. The lifespan of yeast (adapted from Smart, 2000).

90 I. Russell

content to grow. Yeasts prefer to grow at acidicpH. Water is discussed in greater detail in thechapter by Ingledew in this volume.

A CARBON SOURCE (SUGARS AND STARCHES)

Low molecular weight sugars are the preferredcarbohydrate source of Saccharomyces.Fermentable carbohydrates include themonosaccharides glucose, fructose, mannoseand galactose as well as the disaccharidesmaltose and sucrose, and the trisaccharides,raffinose and maltotriose (uptake is straindependent). Large polysaccharides such asstarch and cellulose cannot be taken up by yeast

in that form. Most brewing and distilling strainsutilize sucrose, glucose, fructose, maltose andmaltotriose in this approximate sequence,although some degree of overlap does occur.The initial step in the utilization of any sugar byyeast is either its hydrolysis outside themembrane (e.g. sucrose), followed by entry intothe cell by some or all of the hydrolysis productsor its intact passage across the cell membrane(e.g. maltose). Sucrose is the first sugar todisappear from the wort. The hydrolysisproducts, glucose and fructose, are then takeninto the cell by facilitated diffusion (Figure 8).This is a passive uptake and both sugars aredirectly incorporated into the glycolytic pathway

Table 1. Approximate composition of industrially produced aerobically grown baker’s and anaerobically grownbrewer’s yeast (expressed as g/kg dry yeast) (from Ingledew, 1999).

Cell composition (g/kg dry yeast) Baker’s yeast (aerobically grown) Brewer’s yeast (anaerobically grown)

Carbohydrate 180-440 390-600Protein 380-590 370-420Ash 45-75 73-81Nucleic acid (DNA and RNA) 52-95 39-43Lipids 40-50P 10-19 14-20S 3.9K 20-21 17Na 0.12-0.3 0.7Ca 0.6-0.75 1.3Fe 0.02-0.03 0.1Mg 1.3-1.65 2.3Co 0.008 0.0002Cu 0.008 0.033Mn 0.0059 0.0057Zn 0.170-0.197 0.0387Cr 0.0005-0.0025Mn 0.008Ni 0.003Sn 0.003Mo 0.00004Li 0.000017V 0.00004Se 0.005Pantothenate (Coenzyme A) 0.065-0.10 0.110-0.120Choline (membranes) 2.71-4.00 3.80-4.55Thiamin (Vit B1) 0.090 -0.165 0.092-0.150Riboflavin (Vit B2) 0.040-0.100 0.035-0.045Nicotinic acid/niacin (NAD) 0.30-0.585 0.450Pyridoxine (Vit B6) 0.020-0.040 0.043-0.050Biotin (biocytin) 0.0006-0.0013 0.001p-aminobenzoate (folic) 0.160 0.005Inositol (phospholipids) 3.0 3.9-5.0Folic acid (1-C transfer) 0.013-0.015 0.010

Malony, 1998; Reed and Nagodawithana, 1991; Ingledew et al., 1977; Patel and Ingledew, 1973; Peppler, 1970.


immediately after entry into the cell (Figure 9).For the sugar sucrose (a disaccharide consistingof a glucose molecule and a fructose molecule)the enzyme invertase, which is located betweenthe cell membrane and the cell wall, hydrolysesthe sucrose to its two constituents, glucose andfructose, which then enter the cell by passiveuptake. The hydrolysis of sucrose by the yeast’sinvertase is very rapid and so on a sugar profilethe sucrose disappears very quickly from thesubstrate.

Maltose (two linked glucose molecules) andmaltotriose (three linked glucose molecules) passintact across the cell membrane by an activeprocess, and once inside are hydrolyzed to glucoseunits by the α-glucosidase system. Maltotriosefermentation tends to begin later than that of maltosein a batch fermentation. Maltotetraose (four linkedglucose molecules), higher polysaccharides(dextrins) and pentoses are not fermented byregular ethanol strains. If a very high glucoseconcentration relative to the other sugars ispresent, one can encounter ‘hung’ fermentations,due to a condition called glucose repression. Herethe high concentrations of glucose prevent theyeast from synthesizing the enzymes needed toassimilate the other sugars present.

Figure 10 illustrates the typical sugar utilizationpattern in a laboratory whisky mash fermentationwhere only the enzymes from the malt areavailable to break down the sugars and starches.A mash from a fuel ethanol fermentation wouldhave commercial enzymes added to break thestarch into smaller sugars that can be taken upby the yeast cell.

NITROGEN SOURCE

Yeast growth requires the uptake of nitrogen forthe synthesis of protein and other nitrogenouscomponents of the cell, but yeasts can only utilizelow molecular weight nitrogenous materials suchas inorganic ammonium ion, urea, amino acidsand small peptides. The yeast cannot take upproteins or break down peptides larger thantripeptides. The extracellular proteolytic activityof yeast is negligible. Proteolysis does not occurin the fermentation, unless the yeast hasautolysed. Nitrogen uptake slows or ceases laterin the fermentation as yeast multiplication stops.Nitrogen is discussed in more detail later in thischapter.

MALTOSEMALTOTRIOSE

SUCROSESTARCH / DEXTRIN

Maltose

Glucose

Maltotriose

Glucose + FructoseGlucose

FructoseFructose

permease

permease

permease

permease

invertase

α-glucosidaseα-glucosidase

glucoamylase

Figure 8. Uptake of sugars into the yeast cell.

92 I. Russell

Conversion of sugars to ethanol

Yeast is able to utilize sugars in the presence ofoxygen (aerobically) and when oxygen isexcluded (anaerobically). Aerobic breakdown(catabolism) produces more energy and is calledrespiration. Anaerobic catabolism produces lessenergy and is called fermentation.

Once sugars are inside the cell, they areconverted via the glycolytic pathway (also calledEmbden-Myerhof-Parnas or EMP pathway) intopyruvate. The sequence of enzyme catalyzedreactions that oxidatively convert glucose topyruvic acid in the yeast cytoplasm is known asglycolysis (Figure 9).

The very simplest expression for fermentationis the following equation:

Figure 9. The glycolytic pathway (Embden-Meyerhof-Parnas) for sugar utilization.

ATP

ATP

ATP

ATP

HXK

GlucoseGlucoseHexose

transporter

Fructose

Fructose PGI

Glucose-6-phosphate

PFKHXK

Fructose-6-phosphate

FBA

TPI TDH

Fructose 1,6-diphosphate

Glyceraldehyde 3-phosphate

1,3 diphosphoglycerate

3-phosphoglycerate

Dihydroxyacetonephosphate

EthanolEthanol

Acetaldehyde

NAD+

CO2CO2

NADH,H+

PGM

PGK

2-phosphoglycerate

ENO

PDC Phosphoenol pyruvate

Pyruvate

PYK

Plasma membrane

ADH


This equation, named after the French scientistGay-Lussac, shows that glucose yields almostequal parts of carbon dioxide and ethanol as wellas the production of energy. Part of the energyis used for cell metabolism and some of it willbe lost as heat. What this equation does notaddress is that some of the sugar will be usedfor yeast growth and that there are othermetabolites produced (glycerol, lactic andsuccinic acid, etc.). These other metabolites aresmall in quantity compared to the amount ofethanol produced.

The glycolytic (EMP) pathway operates in thepresence or the absence of oxygen to convertglucose to pyruvic acid, energy and reducednicotinamide adenine dinucleotide (NADH +H+). When oxygen is abundant and the sugarlevel is kept low, little or no ethanol is made bythe growing yeast. The yeast uses all of the sugaras energy for new yeast cell production (i.e. thebaker’s yeast propagation process). Howeverwhen oxygen is reduced and/or the glucoselevels exceed 0.1% (w/v), then ethanol is made.Under low oxygen conditions less than 5% ofthe sugar is metabolized for growth and the total

cell yield is approximately 10% of what ispossible under aeration and non-repressingglucose concentrations.

Not all glucose is metabolized by the glycolytic(EMP) pathway described above. Some of thesugar will be metabolized by a route called thehexose monophosphate pathway. This pathwayalso yields energy, but more importantly it yieldspentose sugars, which are important in thesynthesis of yeast nucleotides and yeast nucleicacids.

From glucose to pyruvate: glycolysis inyeast

Glucose is phosphorylated in two stages. TwoATPs are used to produce fructose 1,6disphosphate, which is then split by the enzymealdolase to form two 3-carbon triose phosphates.Inorganic phosphate is assimilated to form twotriose diphosphates from which four H atomsare accepted by two molecules of oxidized NAD.Lastly, four ATPs are formed by transfer ofphosphate from the triose diphosphates to ADP

Figure 10. Uptake of sugars from a laboratory 19.3ºPlato/Brix whisky mash.

0

2

4

6

8

10

12

14

0 10 20 30 40 50

Fermentation time (h)

Maltotriose Maltose

Glucose Glycerol

Ethanol

Car

bo

hyd

rate

(g

/100

ml)

C H O + 2 P + 2 ADP 2 C H OH + 2 CO + 2ATP + 2H O6 12 6 i 2 5 2 2

Glucose 2 carbon dioxide + 2 ethanol + energy

Mol wt 180 2 x 44 2 x 46

94 I. Russell

resulting in the formation of two molecules ofpyruvic acid. Some of the pyruvic acid and otherintermediates are used by the yeast cell througha variety of metabolic pathways as ‘buildingblocks’ for new yeast cells and some glycerol ismade from the intermediate, dihydroxyacetone.However most of the pyruvic acid isimmediately converted to ethanol and carbondioxide (CO2) (Figures 11 and 12).

Energy and the yeast cell

The yeast cell requires energy for three mainactivities.

i) Chemical energy: to synthesize complexbiological molecules (i.e. make more yeastcell components)

ii) Transport: energy must be expended totransport nutrients in and out of the cell.

iii) Mechanical energy: cells need to movestructures internally and this requires energy(i.e. budding).

There are two main carriers of energy in livingcells, Adenosine TriPhosphate (ATP) andNicotinamide Adenine Dinucleotide (NAD+).ATP is a carrier of chemical energy in the formof high energy phosphate bonds. NAD+ is acarrier of hydrogen and electrons and is involvedin many oxidation-reduction reactions in thecell. It can pick up and transport 2e- and 2H+.The cell takes its energy source, converts it intoNADH and ATP, and then uses these carriers toperform needed tasks in the cell. NAD+ and ATPare not the only carriers of energy, but they arethe major carriers.

Batch growth of yeast

The traditional concept of ethanol productionwas that growth occurred prior to thefermentation of most wort/mash sugars and thatfermentation was carried out by non-growingstationary phase cells. It is now known that yeastgrowth, sugar utilization and ethanol productionare coupled phenomena. The rate offermentation by growing, exponential phasecells of an ale yeast strain is 33-fold higher than

that of non-growing cells. Hence for maximumethanol production, best practice is to keep thecells growing as long as possible at the highestpossible cell concentration. When a yeast istransferred into a suitable fermentation mediumunder optimal growth conditions, the growthpattern illustrated in Figure 13 is seen.

Lag

phase

Exponential

growth phase

Stationary

phase200

160

120

80

40

10

Adaptation

to mash

Decline

phase

Viable

cell count

Total

cell count

12 24 36 48 60 72

Time (hours)

Cel

l po

pu

lati

on

(m

illio

ns

per

ml)

Maximum

growth

Figure 13. Typical batch growth curve for yeast.

CELL GROWTH PHASES

Lag phase. During this phase the yeast adaptsto its new environment and activates itsmetabolism (i.e., synthesizing enzymes). It is aperiod of zero growth but a period of intensebiochemical activity as the yeast cell adjusts fromthe previous culture conditions to the conditionsin the fresh media. This phase ends with the firstcell division.

Accelerating phase. Once cells transit from lagphase and commence active cell division theyenter the acceleration phase before exponentialgrowth. Here the rate of division continuouslyincreases.

In a brewery, only an eight to 10-foldmultiplication is possible, i.e. 3 cell generations,and possibly a fourth generation by a smallproportion of cells. Further growth is not possibledue to the lack of the essential membranecomponents (i.e. unsaturated fatty acids andsterols), which cannot be synthesized under the


NADH

Aerobic (respiration)

(with oxygen)

Anaerobic (fermentation)

(without oxygen)

Krebs (TCA)Cycle &

oxidative phosphorylation

Pyruvate

GLUCOSE Glucose-6-phosphate

Triose phosphates

Glucose-6-phosphate

Oxygen

CO2

Yeast biomass

Pyruvate

Ethanol+

CO2

Oxaloacetate

Succinate

CO2

Succinate

Yeast fermentation products

Ethanol

CO2

GLUCOSE

Figure 11. Aerobic and anaerobic catabolic pathways for sugar in yeast.

Figure 12. Energy and the cell.

H2O NAD+

NAD+

O2

NADH

NADH

Lactate NAD+

NADH

Aerobic (respiration)

(with oxygen)

Anaerobic (fermentation)

(without oxygen)

CO2

Ethanol

Acetyl-CoA

Krebs (TCA)Cycle

Acetaldehyde

Pyruvate

Yeast YeastBacteria

36 ATP

generated

2 ATP

generated

2 ATP

generated

GLUCOSEglycolysis

2 ATP

generated

96 I. Russell

anaerobic conditions of fermentation(remembering that brewery yeast are recoveredfrom previous anaerobic fermentations).Distillery yeast however, are pre-grown by theyeast manufacturer under vigorous aerobicconditions, generating an excess of the essentiallipids, and when added to a well-aerated wort, a20-fold multiplication is possible, i.e. slightlyover four generations.

Log phase (exponential growth). In log phasegrowth rate is constant and maximal. It is aperiod of logarithmic cell doubling. The‘generation time’ is the time during which thecell number doubles. Under optimal conditionsgeneration time is between 90 and 120 minutes.

The yeast can be kept in this log phase byemploying ‘fed batch’ culture conditions (i.e.incremental feeding of nutrients in step withyeast growth). This is used in the production ofbaker’s yeast where the aim is to maximize cellmass and minimise ethanol production.

Deceleration phase. Growth rate decreases inthis phase with the reduction in nutrients (i.e.carbohydrates) and/or the increase in growthinhibitors (i.e. ethanol).

Stationary phase. In the stationary phase thenumber of yeast cells stays constant (i.e., it isstationary with respect to the number of viableyeast cells). The number of new cells formed isbalanced by the number that die. In thestationary phase, yeasts can survive forprolonged periods without added nutrients.

Declining phase. During this phase the deathrate of the yeast cells exceeds the birth rate andthe total cell number decreases.

Under well-controlled laboratory conditions thegrowth of yeast at constant temperature can beexpressed as a straight line by plotting cellnumbers logarithmically against actual time(hence the term log or exponential phase). Itshould be remembered that the growth curveshown is for a constant temperature. For plantfermentations, the large amounts of heatgenerated by yeast metabolism cause thetemperature to rise throughout most of the logphase, and growth rate increases accordingly.This is an important difference from the textbookversion of the microbial growth curve illustrated.

Propagation of yeast

Yeast propagated for the baking industryrepresents the largest amount of a single type ofmicrobial biomass produced, with an annualproduction of baker’s yeast netting $1.5 billionglobally. The large-scale production of distillingand fuel ethanol yeasts follows the sameprinciples as those employed for the manufactureof baker’s yeast. A multi-stage propagation in afermentation media usually consisting of sugarcane or beet molasses (as an inexpensive sourceof sucrose) with additional sources of nutrientsadded. The cell density progressively increases.The later stages are highly aerobic and themolasses medium is delivered slowly to avoidthe Crabtree effect (i.e. suppression ofrespiration by high glucose levels where the cellscontinue to ferment despite the availability ofoxygen) and to maximize respiratory growth.Yeast can be supplied to the ethanol industry asa slurry (yeast cream), compressed yeast or anactive dry yeast.

Most of the yeast used in distilleries and fuelethanol plants is purchased from manufacturersof specialty yeast appropriate for the purpose.Breweries in contrast tend to have their ownproprietary strains and scale-up systems sincethey have the ability to crop and re-use theiryeast.

Manufacturers propagate from proprietarylaboratory stock cultures (best practice storesthese cultures at ultra low temperatures) througha succession of fermentation vessels ofincreasing size. A culture medium of molasses(beet or cane) is supplemented with ammoniumsalts and any required trace nutrients. All culturesare grown with accurate temperature control at30oC, with vigorous aeration (Campbell,2003b). For efficient cultivation it is essential tomaintain a low concentration of fermentablesugar in the culture medium. This avoids theCrabtree effect. The sugar is added as sterilizedconcentrated molasses, slowly at first and thenincreasing rapidly as the yeast biomass increases,so that there is a consistent level of about 0.5%sugar present. The yeast concentration willeventually surpass the ability of the aerationsystem of the culture vessel to maintain thenecessary aerobic conditions for efficientgrowth. When the dissolved O2 concentrationapproaches zero (after about 30 hrs growth)propagation is halted. The yeast culture can be


harvested using a rotary vacuum filter. Thecircumference of the filter is a continuous stripof fabric coated with food-grade starch. As thedrum rotates slowly through the culture in thetrough, the spent culture medium is drawn undervacuum through the hollow spokes. The yeastis retained on the filter, scraped off andpackaged.

YEAST SLURRY

Yeast in the slurry form is used primarily bybrewers. They can recycle their yeast at a highviability (>90%). The slurry must be kept cool(4°C) and has a 3-7 day shelf life. If the yeast isfresh it can be acid-washed to lower the bacterialcount. The slurry usually has a cell count in therange of ~3 x 109 cells/mL at 14-18% total solids.

COMPRESSED YEAST

Compressed yeast is more stable than slurry andrequires cooling. The yeast is stored at 3–4oCand has a 30 day shelf life. It loses viability at~5-10 % a week. It can be rehydrated easily andhas an indigenous bacterial count. Thecompressed yeast usually has a yeast count inthe range of ~6-13 x 109 cells/mL at 33% totalsolids.

Compressed yeast is prepared at 4°C by addingsalt to the yeast slurry and then passing the yeastthrough a cloth press or through a rotary drumvacuum filter. The resultant yeast has theconsistency of butter and is then extruded intoblocks, packaged, and both stored and shippedat 4°C. The yeast is sold as 25 kg bags ofcompressed moist yeast (24–33% dry weight)or in bulk as cream yeast slurry of about 18%dry weight. Before using compressed yeast itmust first be slurried aseptically in a sterilizedyeast mixing vessel to provide the requiredconsistency for pitching. Cream yeast, which isdelivered, stored and pitched in bulk, is oftenused for automated large-scale operations.

ACTIVE DRY YEAST (ADY)

Active dry yeast is used by most batch ethanolplants. It is also used to start continuous plantsand to supplement fermentors. This yeast is driedunder partial vacuum at 45–55oC in an

atmosphere of inert gas, usually nitrogen. Driedyeast (92–96% dry weight) has a long shelf lifeand does not require cold storage (but can bechilled for added security). It is very stable atroom temperature and even more stable (over ayear) if stored cool and under nitrogen gas. Itloses ~7% viability/month at room temperatureand <10% year at 4°C under nitrogen gas. Theactive dry yeast usually has a viable cell countin the range of ~2.2 x 1010 cells/g. It has anindigenous bacterial count. The process to makeactive dry yeast involves extruding thecompressed yeast into spaghetti-like strands thatare then dried in air lift dryers (fluidized bed).Reactivation (conditioning) of dried yeast atpitching is carried out to prevent loss of viability.It is important that rehydration is carried outunder the prescribed conditions and oncerehydrated the yeast must be used immediately.

The availability of a shelf-stable dried yeastinoculum where each gram contains ~2.2 x 1010

viable yeasts, is a major biotechnologicaladvance. The yeast, which has been grownunder extreme aeration for yield, is easier to store,handle and transport. Selected strains forindividual processes are available (i.e. beverageversus fuel). The use of active dry yeast hasreduced the need for yeast propagation expertisein the plants and increased the predictability offermentations. Active dry yeasts are used indistilling and fuel alcohol plants to inoculatefermentors to recommended values of 1.0-2.0million viable cells/mL per degree Plato/Brix ofwort/mash.

CONDITIONING (ACTIVATING) THE YEAST

It is usually recommended that the active dryyeast is conditioned (according to the differentmanufacturer’s instructions) to get optimalgrowth. (This is different from propagating astarter culture with the aim of growing up largequantities of yeast from a small lab culture.) Forexample, one manufacturer suggests aconditioning regime for active dry yeast for fuelethanol plants where a yeast conditioning tankwith a 16–18° Plato/Brix mash and a 1-2%glucose level is used. The recommendedconditioning tank is 7–10% the size of thefermentor and the goal is 250-300 million cells/g.Temperature and nutrition at this stage are criticalto ensure optimal yeast growth.

98 I. Russell

For distilled beverages, dried yeast can beadded directly to the fermentation vessel or pre-mixed to a slurry with water. For directinoculation, the wort cooler is adjusted toproduce wort at 38oC. The warm wort iscollected to a depth of 10 cm in the washbackand the required amount of dried yeast issprinkled onto the wort, taking care to avoidclumping. After 5 min. the heat exchanger isadjusted to the normal set temperature, (~20oC)and wort collection is completed.

For the pre-mixing method, sterile water orweak wort, 10 times the volume of dried yeast,is brought to 38oC in a rehydration tank and yeastis added with stirring, to prevent clumping. Aftermixing for 5 min. the yeast slurry is pumpedinto the washback at normal set temperature. Thepre-mixing method uses the same equipment asthe preparation of a slurry from bags of pressedyeast, but for dried yeast the higher temperatureof mixing is critical (Campbell, 2003b).

It should be noted that there are many differentmethods of conditioning the yeast and it willdepend on the particular plant, the instructionswith the particular type of yeast purchased andthe expertise in the plant. The objectives of yeastpropagation/conditioning are to deliver to afermentor a large volume of yeast of highviability, high budding in the log phase of growthand with a very low level of infection.

There are many methods of propagating yeast(batch, continuous) all with pros and consdepending on the final use of the ethanol i.e.beverage or fuel, the unique characteristics ofthe plant, the technical skill in the plant, etc. Theyeast propagator is the heart of a distillery; andit is important to remember the key objectivesthat are critical to the performance of afermentation. Hence cost savings in this area arenot advised since they are usually minimal whencompared to the cost of production per gallonof ethanol. The loss of ethanol production thatis possible when an aged yeast is used can be inthe range of 1% alcohol, which can lead to aloss of $0.10 per gallon produced. Infectionresident in a plant can cause a drop of 3% alcoholand a potential loss of $0.30 per gallon.

Wild yeast contamination

In the beverage alcohol industry wild yeast

contamination is feared primarily due to the offflavors and odors that can be produced inaddition to fermentation problems for the cultureyeast. For non-beverage alcohol the concern isloss of alcohol due to non-utilization of thesugars for ethanol production. The productionof excretion products such as acetic acid by wildyeast such as Dekkera (also called Brettanomyces)can inhibit the production yeast culture andcause serious fermentation problems in a plant.It is important to be familiar with the normalappearance of the production yeast under themicroscope and to be constantly on the watchfor deviations from the standard. Wild yeast willoften (but not always) have a differentappearance and exhibit very differently sizedcells from the normal population. The cells canvary from elongated cells, to lemon shaped cellsto oval cells. There may be buds with elongatednecks or buds at odd angles. There are a numberof specialized media that can be utilized toisolate and identify contaminating yeast cells(see Campbell, 2003a). In order to protectagainst wild yeast infections it is important tobring in a fresh culture yeast on a regular basisand ensure that there is a good cleaning andsanitation program in place. Once a wild yeastinfection takes hold in a plant, that samecontaminant often comes back numerous timesand causes more infections. An extra-thoroughcleaning and evaluation of problem spots in theplant is often necessary to rid the plant of theproblem. Acid-washing of cropped yeast froma fermentation is ineffective against wild yeastsand many of the wild yeast are more acid-tolerantthan the culture yeast. Washing will howeverreduce the number of bacteria present.

KILLER YEAST

Some strains of Saccharomyces secrete aproteinaceous toxin called a zymocide or killertoxin that is lethal to certain other strains ofSaccharomyces. Toxin-producing strains aretermed killer yeasts and susceptible strains aretermed sensitive yeasts. There are strains that donot kill and are not themselves killed, and theseare called resistant. The killer character ofSaccharomyces is determined by the presenceof two cytoplasmically located dsRNA plasmids(termed M and L). A wild yeast containing ‘killer


factor’ can have a negative impact in a beverageor fuel ethanol plant by annihilating theproduction strain and reducing the yield.

Temperature

Thermal damage to a yeast cell leads to a generaldenaturation of the yeast’s proteins and nucleicacids. A yeast cell cannot regulate its temperatureand all fermentation processes produce heat. Thehigher the temperature above the cell’s optimaltemperature for growth, the greater the loss ofcell viability. Yeast thermotolerance is at itsmaximum when the external pH declines to 4.0.There is a general correlation between growthrate and stress sensitivity. Cells growing quicklyin a glucose rich medium (the desired state forrapid ethanol production) are more sensitive toheat and other stresses than cells that are instationary phase.

Many yeasts are capable of growth over a widerange from 5°C to 35°C . Commercial distillingyeast strains can ferment well at 32-35°C, but athigher temperatures metabolic activity rapidlydeclines. The optimal temperature forreproduction is usually lower and around 28°C.

Yeasts purchased from ale breweries have amaximum growth temperature of 37.5-39.8°Cwhile lager yeasts have a maximum growthtemperature of 31.6-34.0°C. For example, iflager yeast is used as part of the inoculum in awhisky fermentation, they assist the distillingyeast only in the early part of the fermentation.Some Saccharomyces strains can survive 35-43°C under ideal conditions, but not understressed conditions. It is important to rememberthat the growth optimum is species dependentas well as growth condition dependent. Figure14 illustrates the effect of temperature on threedifferent commercial yeast strains fermenting a19.3º Plato/Brix whisky mash at three controlledtemperatures. This figure illustrates theimportance of selecting a strain that can functionoptimally at the temperature of the fermentation.All fermentation processes produce heat and theamount of cooling available and cost of coolingneed to be taken into consideration whenselecting a strain. Certain strains soldcommercially have been constructed for theirability to withstand the stress of high ethanoland high temperature and are known to functionwell under these difficult conditions (Figure 15).

0

2

4

6

8

10


Eth

anol

(%v/

v)

30°C 35°C 38°C

Strain A Strain B Strain C

86°F 95°F 100.4°F

Figure 14. Effect of temperature on three different commercial yeast strains fermenting a 19.3º Plato/Brix whisky mash. Barsindicate final ethanol concentration at 48 hrs of fermentation. (Mash is 85% corn and 15% malted barley).

100 I. Russell

HEAT SHOCK AND COLD SHOCK

Yeast is sensitive to sudden temperature changesand shows signs of shock on sudden coolingwith adverse effects on fermentation and cellmultiplication. Cold shocking a yeast strain willinhibit budding and induce vacuolarrearrangements. Yeasts with elevated levels oftrehalose are better able to survive cold storage.Yeast cells exposed to an elevated temperaturewill exhibit a molecular response called “the heatshock response” where certain proteins aresynthesized which aid the yeast in respondingto stress.

Pitching and viability

To follow the progress of a fermentation it iscommon practice to measure the quantity ofyeast biomass or to count the number of cells.For pitching (inoculating) on a production scale,yeast is measured by mass: either directly as thenumber of 25 kg bags of yeast required, orindirectly as the volume of yeast slurry ofstandard concentration by weight. On alaboratory scale, counting the number of yeastcells per ml is usually employed. This requiresa microscope and an engraved countingchamber slide called a haemocytometer(originally designed to count blood cells). Yeastis normally inoculated at a final concentrationof ~1-2 lb active dry yeast/1000 US gallons (1

kg /4000 L) or 5 x 106 to 2 x 107 viable cells. Abrewing rule of thumb is one million cells perdegree Plato/Brix. So a 12o Plato/Brix wort ormash would be pitched at 12 x 106 cells and a20o Plato/Brix wort/mash at 2 x 107 cells.Pitching the yeast at the correct level ensuresreduced infection and improves the processefficiency thus reducing overall costs.

Viability involves the cell’s ability to bud andgrow, however slowly. Vitality is defined as acontinuum of activity of the cell from very activeto not active at all. Vitality tests are related toyeast growth, yeast physiological changes andfermentation performance.

Measurement of percentage viability alone isinsufficient to guarantee the quality of thepitching yeast and hence the concept of yeastvitality (i.e. the ability of the yeast to fermentquickly and efficiently rather than simply beingalive) is of interest. Various methods have beensuggested as indicators of viability and vitality.Heggart et al. (1999) have extensively reviewedthe factors that affect yeast vitality and viability.

In terms of measuring yeast health, to date noone test exists that is accurate, simple,inexpensive and rapid. It is often not possible torely on a single method and a combination ofmethods is often employed. The evaluation ofyeast viability includes methods based on cellreplication (standard plate count and slide counttechnique), brightfield and fluorescent staining,measurement of intracellular ATP and NADH,

0

2

4

6

8

10

12

14

16

32.2°C 38°C 40°C

Alc

oh

ol(%

)

Superstart™ Thermosacc™

90°F 100.4°F 104°F

Temperature

Figure 15. Ethanol production by two commercial yeast strains at three temperatures.


and other methods such as electrokineticpotential and capacitance. Vitality assessmenttechniques include methods that measuremetabolic activity (staining, microcalorimetry,vicinal diketone reduction, yeast proteaseactivity, magnesium ion release, specific oxygenuptake rate, acidification power and intracellularpH), cellular components (ATP, adenylate energycharge, NADH, glycogen and trehalose, sterolsand unsaturated fatty acids) and fermentativecapacity (glycolytic flux rates, CO2 measurementand short fermentation tests). A combination ofvarious methods is often used to obtain anaccurate picture of a yeast’s fermentativecapability.

One newer method that has been widelyadopted in the brewing industry to assess yeastviability is capacitance measurement. When anappropriate radio frequency signal is applied toa yeast suspension, the dielectrical nature of thecell membrane permits a buildup of charge.Depending on the number of yeast presentwithin the field, the resulting capacitance canbe measured and is proportional to the radiusand volume fraction of cells present. Since yeastcells are relatively constant in size, capacitancecan be related to cell number and biomass. Deadcells with disrupted membranes do not generatea capacitance when a radio frequency is applied.Capacitance probes (Aber Instruments Ltd.Aberystwyth, UK) capable of providing anaccurate measurement of viable yeast in real timehave been successfully adopted by a number ofbreweries worldwide (Austin et al., 1994).

Heggart et al., (2000) have extensivelyreviewed the various techniques available formeasuring yeast viability and vitality and thereader is directed to this review paper for furtherinformation on the many techniques available.

METHYLENE BLUE: THE MOST POPULARVIABILITY TEST

Methylene blue has been in use since the 1920sand is recognized as a simple and standardmethod in the fermentation industry formeasuring yeast viability. The buffered dye isadded to a suspension of cells and examinedunder the microscope (ASBC, 1980). Methyleneblue is taken up only weakly, if at all, by livingcells, and any taken up is rapidly converted to

the colorless reduced form by the metabolicactivity of the cells. However, on death, the cellsbecome readily permeable to methylene blue,and, since dead cells do not metabolize, the cellsremain blue. There are a number of other redoxdyes and variations on the makeup of themethylene blue dye buffers that have beensuggested, but unfortunately these dyes are alsounreliable with yeasts of lower viability thanabout 90% viable (Smart, 1999). Although thedye method has many shortcomings it does givea very fast evaluation of the status of a cultureas long as one remembers that if more than 10%of the cells stain blue, the viability of the cultureis in fact much lower. In other words the dyecan be used to confidently confirm that a cultureis very healthy (>90% viable) but is veryinaccurate when a culture has a high number ofdead cells in it.

To see if a culture is capable of reproducing(i.e. can the cell bud and eventually form acolony) a plate count can be carried out. Whencompared to a plate count, at viabilities above90% the staining method consistently tends tooverestimate cell viability by not more than 10–15%. When viability drops below 90–95% (asmeasured by a plate count), methylene bluecounts can give values lower than 50%. At aplate viability of 0% (i.e. no growth on a plate),methylene blue may indicate apparent viabilitiesas high as 30–40% (Heggart et al., 2000).

A well propagated yeast should be close to100% viability, and should retain that value forseveral weeks if stored at 4oC. Storage at ambienttemperature is not recommended as it leads to aloss of viability. It also allows the rapid growthof bacteria from the inevitable initial low levelof resident contamination. This low level ofcontamination can only be avoided if the yeastculture is propagated under pharmaceuticalconditions. The extra expense for pharmaceuticalconditions would make the product priceprohibitive. Chilled yeast from an anaerobicfermentation does not have the storage stabilityof an aerobically grown yeast.

BUDDING INDEX

Cell counts that measure the number of buddingyeast cells are often employed to determine ifthe yeasts are in the active desired growth phase

102 I. Russell

since this is when ethanol is produced mostrapidly. The yeasts are examined under themicroscope with the use of a hemacytometer,and the number of budding cells counted toobtain a budding index. A budding index of~30% is desirable as this indicates that the cellsare in the rapid phase of growth.

pH

Yeast prefers an acid pH and its optimum pH is5.0–5.2, but brewing and distilling strains arecapable of good growth over the pH range ofapproximately 3.5 to 6.0.

During any fermentation, H+ ions are excretedby the yeast, and this results in a pH decline inthe media. For example, an all malt wort pitchedwith a pure culture of distiller’s yeast will havean initial pH of ~5.2-5.5, which will fall to pH4.0–4.5 (depending on the solids), and thenincrease slightly during the stationary phase dueto release of amino acids from autolysing yeastcells. In fermentations showing late growth oflactic acid bacteria, these bacteria can use theyeast autolysate as a nutrient and further lowerthe pH to ~3.8. Contamination by lactic acidbacteria early in the fermentation is undesirableand reduced spirit yield results since the sugarsused by the bacteria are then not available tothe yeast for the production of ethanol. The lossof spirit yield in malt distilleries due to growthof Lactobacillus from the start of fermentationis reported as follows:

Up to 106 bacteria /ml <1% loss1–10 x 106 bacteria/ml 1–3% loss1–10 x 107 bacteria/ml 3–5% lossAbove 108 bacteria/ml >5% loss

No significant loss of spirit yield was observedwhen the lactic acid bacteria grew only at theend of the fermentation on the nutrients fromthe autolysed yeast (Dolan, 1976; Barbour andPriest, 1988). Changes in pH also affect theactivity of α- and β-amylases and limitdextrinase, and consequently the hydrolysis ofpolysaccharides during the fermentation. Thelow pH associated with heavy early contaminationby lactic acid bacteria reduces spirit yield if theenzymes cannot function well at that pH.

Yeast metabolism

Yeast produces a myriad of products. Table 2summarizes some of the main products.

GLYCEROL

Glycerol is quantitatively the next most importantproduct of yeast alcoholic fermentation afterethanol and carbon dioxide. It is an importantcompound in that it helps to maintain the cell’sredox balance. When a yeast cell is growing,removal of pyruvate for cell biosynthesis canlead to a buildup of NADH, which can haltcatabolism. To avoid this the cell reducesdihydroxyacetone phosphate to glycerolphosphate, which is then dephosphorylated toglycerol. The cell then excretes the glycerol intothe growth medium (Figure 16). Redox balanceand energy metabolism are highly integrated andthe biosynthesis of new cellular material resultsin the production of a surplus of reducingequivalents of NADH. This is especially aproblem under anaerobic conditions andSaccharomyces relies on glycerol production to

Table 2. Main products of yeast fermentation.

Alcohols Acids Esters Others

Ethanol Acetic Ethyl acetate and any other CO2

Propanol Caproic combination of acids and AcetaldehydeButanols Caprylic alcohols on the left DiacetylAmyl alcohol Lactic H2SGlycerol PyruvicPhenethyl alcohol Succinic

(Campbell, 2003b)


re-oxidize the surplus of NADH formed duringanaerobic conditions. However, during aerobicconditions, the surplus of reducing equivalents(as NADH) in the cytoplasm is delivered to therespiratory chain in the mitochondria.

The yeast must maintain a redox balance andglycerol carries out a redox balancing role, butit also plays a role as an osmoprotectant. Thestress of high osmotic pressure and heat shockboth enhance glycerol production. Glycerolaccumulation inside the cell is very importantfor the survival of strains during osmotic stress.An increase in the fermentation temperaturecauses the yeast cell to produce larger quantitiesof glycerol. In a fuel ethanol fermentation,glycerol levels range from ~1.2–1.5%, againdepending on a plethora of environmentalfactors.

In Germany in 1914, in order to produce largeamounts of glycerol for nitroglycerineproduction, this yeast pathway was exploited.Using ‘steered fermentation’ acetaldehydebreakdown to ethanol was blocked by theaddition of sodium bisulphite, and the glycolyticintermediates were reduced to give glycerol asthe major fermentation product.

ACROLEIN

Acrolein has a very peppery, lacrimating andirritating effect on workers operating a still.

Acrolein is produced from glycerol (Figure 17).It produces an intensely bitter taste when itcombines with phenolic compounds. Certaincontaminating lactic acid bacteria in afermentation can degrade the glycerol presentvia a dehydratase enzyme to form 3-hydroxy-propionaldehyde, which under acidic conditionsspontaneously degrades slowly to acrolein instorage. The process occurs more rapidly athigher temperatures.

SUCCINIC AND OTHER ORGANIC ACIDS

Organic acids, depending on the carbon chainlength, can impart different flavors and aromasto beverages ranging from acidic to rancid tocheesy. Succinic acid is a main secondary endproduct of alcoholic fermentation. It is believedto be synthesized and secreted by yeasts eitherfollowing limited operation of the citric acid(Kreb’s or TCA) cycle or by reductive pathwaysinvolving some citric acid cycle enzymes. Lowconcentrations of pyruvic, malic, fumaric,oxaloacetic, citric, α-ketoglutaric, glutamic,propionic, lactic and acetic acids are also presentas fermentation products. These are producedas intermediates of the citric acid (TCA) cycle(Figure 18).

The production of both glycerol and organicacids is greatly affected by the pH as illustratedin Table 3.

NADH+H+

Dihydroxyacetone

phosphate

O

CH2OH

C

CH2OP

NAD+ CH2OH

CHOH

CH2OP

Glycerol-3-

phosphate

Glycerol-3

phosphate

dehydrogenase

Pi

Phosphatase

CH2OH

CHOH

CH2OH

Glycerol

Figure 16. Conversion of dihydroxyacetone to glycerol phosphate to glycerol.

Figure 17. Degradation of glycerol to acrolein.

-H2O

Glycerol

O

CH2OH

C

CH2OH

CH2OH

CH2

CHO

3-hydroxy

propionaldehyde

Lactic

bacteria

spontaneous

CH2

CH

CHO

Acrolein

104 I. Russell

Table 3. The influence of pH on production of glyceroland organic acids1.

Producta pH 3.0 pH 4.0 pH 5.0 pH 6.0 pH 7.0

Ethanol 171 177 173 161 150Carbon dioxide 181 190 188 177 161Glycerol 6.2 6.6 7.8 16.2 22.2Acetic acid 0.5 0.7 0.8 4.0 8.7Lactic acid 0.8 0.4 0.5 1.6 1.91Adapted from Neish and Blackwood, 1951 by permission.aMillimoles of product per millimole glucose fermented.

SULFUR METABOLISM

The biosynthesis of S-containing amino acidsand reduction of sulfate salts in the wort/mashare the key sulfur metabolism areas. In whiskydistilling there are benefits from the remedialeffects of copper in the stills, but with the normalratio of copper surface area to pot still volume itis not possible (or desirable) to remove all sulfur

compounds completely. Another source ofsulfury off-flavors is the reduction of SO4

2- toS2- in the course of fermentation. Most of theresulting H2S is stripped off by the evolution ofCO2 during fermentation. This can pose aproblem in the collection of food-grade CO2.High levels of H2S and organic S compoundscan also remain in the wash and affect qualityof the distilled spirit if care is not taken.

OTHER COMPOUNDS

During fermentation the carbonyl compoundfound in highest concentration is acetaldehyde.It is formed during fermentation and is ametabolic branch point in the pathway leadingfrom carbohydrate to ethanol. The acetaldehydeformed may either be reduced to ethanol, oroxidized to acetic acid, and in the final step ofalcoholic fermentation, acetaldehyde is reduced

Figure 18. The citric acid cycle (also called Kreb’s or TCA) for the synthesis of intermediates and energyunder aerobic conditions.

Pyruvate

CO2

CO2

NADH CO2

Acetyl CoA Lipids

Citrate

Oxaloacetate

Malate

Fumarate

Amino

acids

Amino

acids

Succinate

Cytochromes Amino acids

Succinyl CoA

α-ketoglutarate

cis-aconitate

Isocitrate

NADH

NADHGTP

FADH2

NADH

Oxalosuccinate


to ethanol by an enzymatic reaction. Theconcentration of acetaldehyde varies duringfermentation and reaches a maximum during themain fermentation and then decreases.

Diacetyl and pentane-2,3-dione both impart acharacteristic aroma and taste described as‘buttery’, ‘honey- or toffee-like’ or as‘butterscotch’. It is primarily a problem for brewersbut in addition to being a yeast by-product,excessive levels can also be produced by certainstrains of bacteria such as Pediococcus andLactobacillus.

FLOCCULATION

Although brewers traditionally relied on theflocculence of the yeast to achieve separationof the yeast and beer at the end of fermentation(more common practice now is to separate bycentrifugation), yeast for use in distilling isgenerally non-flocculent as flocculent yeast cellssettle on heating surfaces of the still therebyrestricting heat transfer. For beverage alcohol,if the yeast is charred, this will give rise to flavordefects in the spirit. The goal is to maintain auniform yeast suspension throughout fermentationand distillation.

EFFECT OF CO2 CONCENTRATION

Saccharomyces cerevisiae can tolerate CO2

concentrations up to 2.23 volumes of CO2 (v/v)in solution. Excessive CO2 concentrations caninhibit total yeast growth, cause production of arange of volatiles and can affect the uptake ofamino acids.

VITAMINS

Vitamins are important regulators and cofactorsof numerous metabolic processes (Table 4).Their principal function is enzymatic and theygenerally act either as co-enzymes or precursorsfor fully active enzymes. Essential vitaminrequirements for maximum fermentation ratesare strain dependent. Yeast vary widely in theirneed for vitamins for their metabolism and, in agiven strain, this need may also vary betweenactive respiration and growth on the one hand,and alcoholic fermentation on the other. Almostall vitamins (except mesoinositol) are requiredby yeast to function as a part of a coenzyme,serving a catalytic function in yeast metabolism.Most strains have an absolute requirement for

Table 4. Vitamins and their functional role in yeast metabolism1.

Vitamin Active form Metabolic function(s) Optimum concentration/100 ml 2

Biotin Biotin All carboxylation and decarboxylation 0.56 µgreactions

Calcium pantothenate Coenzyme A Keto acid oxidation reactions 45-65 µgFatty acid, amino acid, carbohydrate andcholine metabolism. Acetylation reactions

Thiamin (B1) Thiamin-pyrophosphate Fermentative decarboxylation of pyruvate 60 µgRearrangements in pentose cycle,transketolase reactions, and isoleucine andvaline biosynthesis

Inositol Inositol Membrane phospholipids (structural) 9.3 mg (free) and 18.9mg (total)

Niacin (B4) (nicotinic acid) NAD+, NADP+ Coenzymes in oxidation/reduction reactions 1000 - 1200 µg(under anaerobiosis)

Pyrdoxine (B6), pyridoxal Pyridoxal-phosphate Amino acid metabolism, deamination, 85 µg and pyridoxamine decarboxylation

Riboflavin (B2) FMN, FAD (Under anaerobiosis) as coenzymes in 20 – 50 µgoxidation/reduction reactions

1Adapted from Heggart et al. (1999).2Optimal concentrations in 100 ml of a standard brewing wort.

106 I. Russell

biotin and many require pantothenate. A cornor grain mash should contain sufficient amountsof vitamins and they are not normally addeddirectly to address a deficiency other thanthrough the use of a yeast food. If deficienciesoccur there can be resultant fermentationproblems. When active dry yeast is used, sincethe yeast only multiplies a few generations,vitamin deficiencies are not usually encounteredunless yeast recycling is practiced. Again, it mustbe stressed that each mash has uniquecharacteristics that must be considered whenyeast nutrition is an issue.

INORGANIC IONS

Yeast requires a number of inorganic ions foroptimum growth and fermentation and thisrequirement varies with i) the yeast strainemployed, ii) growth media differences, and iii)interactions with other constituents especiallyinteractions among the trace metals. Importantcations are zinc, manganese, magnesium,calcium, copper, potassium and iron (Table 5).

With the exception of zinc, grain and corn basedfermentation media are not usually deficient inthese ions, but it is important to remember thatdifferences in the fermentation substrate cancause large differences in metal content. A properratio between calcium and magnesium positivelyinfluences fermentation rates. An imbalance ininorganic nutrition is often reflected in complexand often subtle alterations of metabolic patternsand growth characteristics. The role played bymany of these ionic species is both enzymaticand structural. The macro elements potassium,magnesium, calcium, iron, zinc, and manganeseare required at concentrations between 0.1 and1.0 mM. The micro elements cobalt, boron,cadmium, chromium, copper, iodine,molybdenum, nickel and vanadium are requiredbetween 0.1 and 100 mM. Minerals such assilver, arsenic, barium, mercury, lithium, nickel,osmium, lead, selenium and tellurium can beinhibitory when concentrations exceed 100 µM.Table 6 lists the effects of metal ions on yeastviability and vitality. For additional informationon the effect of metal ions such as magnesium

Table 5. Metal ion requirements and function in the yeast cell1.

Ion Function in yeast cells Concentration

Magnesium (Mg2+) Stimulation of synthesis of essential fatty acids 2 – 4 mMAlleviation of inhibitory heavy metal effectsRegulation of cellular ionic levelsActivation of membrane ATPaseMaintain membrane integrity and permeability

Potassium (K+) Central component in the regulation of divalent cation transport 2 – 4 mMEssential for the uptake of H2PO4

-. Potent stimulator of glycolytic flux: significantly increases the levels of ATP, NAD(P)H, ADP and Pi

Zinc (Zn2+) Essential cofactor in a number of important metabolic enzymes, e.g. alcohol 4 – 8 µM dehydrogenaseEnhances riboflavin synthesis, activates acid and alkaline phosphatasesHelps increase protein content of fermenting cells

Calcium (Ca2+) Not a requirement but may stimulate cell growth 4.5 mM optimum forProtects membrane structure and helps maintain membrane permeability growth under adverse conditionsCounteracts magnesium inhibitionPlays an important role in flocculation

Manganese (Mn2+) Essential in trace amounts for yeast growth and metabolism 2 – 4 µMHelps to regulate the effects of Zn2+

Stimulates the synthesis of proteins and biosynthesis of thiamin

Copper (Cu2+) Essential as an enzyme cofactor 1.5 µM

Iron (Fe2+) Small amounts required for the function of haeme-enzymes 1 – 3 µM

1Adapted from Jones et al. (1981); Jones and Greenfield (1984); Heggart et al. (1999).


Table 6. Effects of metal ions on yeast viability and vitality1.

Metal ion Effect on yeast

Calcium (Ca2+) Amino acid uptake inhibition above 1 mM and growth inhibition above 25 mM. Reduced rate of ethanol production and decrease in Yx/s and Yp/s.Mg:Ca ratios that grossly favor Ca are detrimental to yeast growth and ethanol productivity through Ca antagonism of essential biochemical functions of Mg.

Copper (Cu2+) Growth inhibition occurs at ~10 µM, with increasing potency up to 0.15 M (cell growth ceases).Causes changes in the yeast plasma membrane which can lead to the leakage of low molecular weight compounds and disturb the uptake of nutrients.

Iron (Fe2+) Inhibitory only above 10–15 mM; excess Fe lowers activity of malate, pyruvate and succinate dehydrogenases.

Potassium (K+) Growth inhibition observed above 10 mM with total inhibition at about 2 M.Fermentation rate may be decreased above 4 – 10 mM.

Magnesium (Mg2+) Total inhibition of growth reported at approximately 1 M.

Manganese (Mn2+) Growth is progressively inhibited above 10 mM.

Sodium (Na+) No apparent metabolic role; at high concentrations, specific growth rate is decreased due to the diversion of energy to maintain an electrochemical gradient of Na+ against passive diffusion or facilitated diffusion into the cell.

Zinc (Zn2+) Deprivation in S. cerevisiae prevents budding and arrests cells in G1 phase of cell cycle.Excess Zn added to wort inhibits fermentation and yeast growth.Inhibition of growth appears to occur above ~30 µM when Mn is below 7 µM; when Mn is above 7 µM,Zn can be as high as 1 mM before inhibition of growth occurs.

1Adapted from Heggart et al. (1999)

on ethanol tolerance in yeast, the reader isreferred to the research work of Walker (1999).

Phosphorus

Phosphorus is essential to yeast cells forincorporation into structural molecules [e.g.phosphomannan and phospholipids, nucleicacids (DNA and RNA)] and phosphorylatedmetabolites (e.g. ATP and glucose-6-phosphate).Yeast stores polyphosphates as an energyreserve and uses this material during times ofstarvation. Phosphorus is commonly availablein the form of inorganic orthophosphate(H2PO4

-) and this is rapidly metabolized tonucleoside triphosphate (ATP) on entry into theyeast cell. It can be added in the monobasic formor in the form of ammonium, sodium orpotassium salts. A rule of thumb is that availablephosphorus should be present at ~1-2% of thecell dry weight per unit volume of the medium.

Sulfur

Inorganic sulfur, in the form of sulfate anions,is transported via an active process by yeast into

the cell for assimilation into sulfur containingamino acids such as methionine and tripeptidessuch as glutathione (glutamic acid-cysteine-glycine). In the presence of excess sulfate, yeastcan store sulfur intracellularly in the form ofglutathione, which can account for as much as1% of the cellular dry weight. Under conditionsof sulfate limitation or starvation, the glutathionein the yeast can act as a sulfur source for thebiosynthesis of amino acids. Sulfur is requiredat ~0.3–0.5% of the weight of expected yeastmass per volume of medium.

Magnesium

Magnesium is the most abundant intracellulardivalent cation in yeast and it acts primarily asan enzyme cofactor. It exerts a protective effecton yeast cultures subjected to stress conditionssuch as temperature and osmotic pressure, aswell as playing a role in alcohol tolerance.Magnesium stimulates fermentation during themetabolism of high gravity worts/mashes. Therequirement for magnesium is small and isusually amply provided by the raw materials andwater. Exogenous magnesium exerts a protective

108 I. Russell

effect on yeast cells subjected to physical andchemical stresses (Walker, 1999).

Zinc

Trace levels of Zn2+ are essential for yeast growthand other metal ions cannot fill this requirement.Zinc is a cofactor in a number of enzymaticreactions within the cell and the importantfermentation enzyme, alcohol dehydrogenase,is a zinc metalloenzyme. Certain yeast strainsmay require more zinc than occurs naturally inwort/mash, and supplementation with a yeastfood is used. Zinc deficiency results in low yieldsof yeast (can inhibit budding) and slowfermentations. A concentration of 0.3 ppm Znin the medium is considered adequate and higherconcentrations of zinc (1-5 ppm) may inhibityeast activity, but this depends, at least in part,on the level of manganese present.

Potassium

Potassium is the most prevalent cation in theyeast cytoplasm (~150 mM). Uptake of K+ inyeast requires an active transport mechanism andthe presence of glucose or a fermentable orrespirable substrate. Transport of K+ is associatedwith the transfer of H+ ions from within the cell.

Sodium

Yeast cells do not accumulate Na+ ions undernormal growth conditions and continuouslyexcrete Na+ in order to maintain very lowcytoplasmic concentrations of this cation. In thepresence of high salt concentrations, the yeastcells osmoregulate by producing intracellularcompatible solutes such as glycerol and arabitol.

Manganese

Manganese (Mn) is essential for yeast growthand metabolism and in trace levels acts as anintracellular regulator of key enzymes.Manganese ions have been reported to stimulatefermentation.

Calcium

Calcium (Ca) stimulates yeast growth but it isnot a growth requirement. It is involved inmembrane structure and function. Yeast cellsmaintain cytosolic Ca2+ at very low levels. The

significance of Ca2+ uptake in yeast lies in themultifunctional role of this cation as a modulatorof growth and metabolic responses. Calcium alsoplays an important role in flocculation.

Copper and iron

Copper (Cu) and iron (Fe) ions are essentialnutrients for yeast and act as cofactors in severalenzymes including the redox pigments of therespiratory chain. Copper is an essentialmicronutrient at low concentrations, but is toxicat higher concentrations. Strains of yeast differin their sensitivity to copper and negative effectson fermentation can be seen starting atconcentrations of >10 ppm. Iron is usuallyabundantly provided for in brewer’s wort wherethe average content is 0.1 ppm.

HIGHER ALCOHOL PRODUCTION

From a quantitative point of view, the mostimportant higher alcohols (fusel oils) producedare n-propanol, amyl alcohol, isoamyl alcoholand phenethyl alcohol. In addition, over 40 otheralcohols have been identified. Regulation of thebiosynthesis of higher alcohols is complex sincethey may be produced as by-products of aminoacid catabolism or via pyruvate derived fromcarbohydrate metabolism (Figure 19).

The catabolic route (i.e. Ehrlich pathway)(catabolic defined as a biochemical process inwhich organic compounds are digested, usuallyan energy-liberating process) involves a pathwayin which the keto acid produced from an aminoacid transamination is decarboxylated to thecorresponding aldehyde, then reduced to thealcohol via an NAD-linked dehydrogenase(Figure 20). In this way, for example, isobutanolmay be produced from valine, 3-methyl-1-butanol from leucine and 2-methyl-1-butanolfrom isoleucine. They are by-products of theamino acid assimilation pathway.

The anabolic route of de novo synthesis (i.e.biosynthesis pathway) (anabolic defined as abiochemical process involving the synthesis oforganic compounds, usually an energy-utilizingprocess) utilizes the same pathways as thoseinvolved in the biosynthesis of amino acids. Asin the catabolic route, the keto acid intermediateis decarboxylated and the resultant aldehydereduced to the alcohol.


The relative contribution made by the tworoutes varies with individual higher alcohols.Since there is no corresponding amino acid, theanabolic route appears to be the sole mechanismfor the formation of n-propanol.

The concentration of amino acids in themedium is important. If there are only low levelsof assimilable nitrogen then the anabolic routeis predominant. High concentrations of wortamino nitrogen favor the catabolic pathway dueto feedback inhibition of the amino acidsynthetic pathways by their respective products.

Both the level and the composition of the yeast-assimilable nitrogen in the mash influence theformation of higher alcohols. Low levels ofusable nitrogen result in less yeast growth and

increased yields of higher alcohols. Whenassimilable free amino nitrogen is increased inthe mash, the anabolic (biosynthesis) pathwaybecomes less dominant and the Ehrlich pathwaytakes over. If ammonium ion or urea is the mainnitrogen source, higher alcohols are made viathe anabolic (synthetic) pathway.

ESTERS

Esters are a product of yeast metabolism andthere are over 100 distinct esters identified infermentation beverages. The most abundantester is ethyl acetate (fruity/solvent). Other estersproduced include isoamyl acetate (banana/apple), isobutyl acetate (banana/fruity), ethyl

α-ketoglutarate

Leucine

CH3

CH

CH2

glutamic acid

Transaminase

NH2-CH-COOH

α-keto isocaproate

CH3

CH

CH2

O=CH-COOH

CO2

Decarboxylase

CH3

CH

CH2Alcohol

deydrogenase

CHO

CH3

CH

CH2

CH3

CH2OH

Isoamyl alcohol

CH3CH3CH3

Figure 19. Higher alcohols (fusel oils) are produced by yeast metabolism via two routes both of which utilize α-keto acid(adapted from Walker, 1999).

-2H

Amino acid

NH2

COOH

C

R

Glucose

COOH

R

NHCH+H

2O

-NH3

COOH

R

C = OH-CO

2

R CHO+2H

R CH2OH

Higher alcohol

Ehrlich Pathway

(amino acid sufficiency)

Biosynthesis Pathway

(amino acid deficiency)

Example: catabolic pathway Example: anabolic pathway

α-keto

acid

α-keto

acid

Isobutanol

Valine

Isobutanol

Glucose

H

Figure 20. Higher alcohol production. Conversion of the amino acid leucine to isoamyl alcohol.

110 I. Russell

caproate (apple/aniseed) and 2-phenylethylacetate.Esters are minor components with ethyl acetatebeing present at over 10 mg/L and the otherstypically less than 1 mg/L.

Biosynthesis of the esters involves twoenzymes, acyl-CoA synthetase and alcoholacetyl transferase. There appears to be more thanone alcohol acetyl transferase. The metabolic rolefor ester formation is still not well understoodbut it may provide a route for reducing the toxiceffects of fatty acids by their esterification andremoval.

Fermentation components that promote yeastgrowth tend to decrease ester levels; andincreased oxygenation tends to reduce esterlevels. Higher temperatures lead to increasedester formation and increased levels of higheralcohols.

ORGANIC ACIDS FROM CONTAMINATINGBACTERIA

Contaminating bacteria can alter the normalprofile of organic acids. Lactobacilli are veryethanol tolerant and are capable of very rapidgrowth in distillery mashes. Increases in lacticand acetic acids are observed when there iscontamination and these acids can inhibit thegrowth of yeast if produced in excess, or ifrecycled and allowed to build up in backset.Concentrations of lactic acid over 0.8% (w/v)and of acetic acid over 0.05% (w/v) affect thegrowth of yeast.

In addition to affecting the growth of yeast,the low pH can affect continuing glucoamylaseactivity on residual dextrins in the fermentor.Depending on the nitrogen source in the mash,levels of acid can vary from 0.5-1.4 g/L and canquickly rise to over 15 g/L when there is heavybacterial or wild yeast contamination.

Infection always means loss of ethanol. Forevery molecule of lactic acid produced there isa loss of one molecule of ethanol. Thus for every90 g of lactic acid made, 46 g of ethanol couldhave been made. Likewise, for acetic acid wherefor every 60 g of acetic acid made, 46 g ofethanol could have been made.

OXYGEN AND YEAST GROWTH

Saccharomyces cannot grow indefinitely underanaerobic conditions unless it is supplied with

certain nutrient supplements, specifically sterolsand unsaturated fatty acids. These twocompounds are essential components of theyeast cell membrane and yeast cannotsynthesize them anaerobically. As the yeastgrows and buds, it shares its initial fatty acidsand sterols with the daughter cell. This sharing,as the yeast cell continues to grow and bud,eventually ‘dilutes out’ the required fatty acidsand sterols in the cell membrane. After anextended number of generations without a sourceof oxygen or pre-formed lipid precursors of thenecessary cell membrane constituents, the yeastwill be unable to ferment sugars. This is not aproblem under aerobic conditions since therequired sterols and unsaturated fatty acids aresynthesized de novo from carbohydrates. For thefermentation to proceed rapidly it is critical thatthere is adequate yeast growth.

There is a wide range of oxygen requirementsamong yeast strains and several authors havesuggested that the strains be divided into fourclasses in regard to their oxygen requirementfor fermentation, with Group 1 yeast being fairlyinsensitive to low oxygen content and Group 4yeast having a high oxygen requirement. Therange varies from 4 ppm dissolved oxygen toover 40 ppm dissolved oxygen.

It is important to remember that the amount ofair or oxygen injected into the fermentationmedium may bear little relationship to theamount of oxygen that actually dissolves. Thusit is necessary to measure the oxygen in solutionin order to know what is available to the yeast.The amount of oxygen required for sufficientsterol synthesis, yeast growth and satisfactoryfermentation differs widely among differentstrains of yeast. There is a fine balance that mustbe achieved. If the oxygen concentration is toolow then the lack of sterol synthesis can restrictyeast growth resulting in fermentation problems.If the concentration of oxygen is excessive, thenalthough there is a rapid fermentation, there isalso high yeast growth with a consequentreduction in ethanol yield since the substrate hasbeen used for yeast growth rather than ethanolproduction.

A yeast with a high oxygen requirement, oneapproaching air saturation in a 10° Plato/Brixbrewing wort (specific gravity 1.040), will beunable to ferment an air-saturated high gravitywort since not enough oxygen is available. The


concentration of oxygen at 100% air saturationis inversely proportional to the specific gravity(and the temperature), thus an air saturated wortof 10° Plato/Brix would contain 8.5 ppm O2,whereas a wort of 17° Plato/Brix (specificgravity 1.070) would only contain 7.9 ppm O2.Oxygen solubility is also lower in mashes poisedat >35°C that contain increased amounts ofsolids.

Mash is fermented under anaerobic conditionsand the very small amount of air (8-20 ppm O2)that should be supplied around the time of initialinoculation can be added by ‘rousing’ the cooledmash using air or bottled oxygen, or by ‘splashfilling’ the fermentor just prior to adding theyeast. This small amount of oxygen, if usedaerobically, would only be enough to allowrespiration of 0.008% (w/v) glucose so there isno large negative effect on sugar utilization.Since there is excessive glucose present in themash, the oxygen is probably used anaerobically.Oxygenating at low levels at the initiation offermentation has been carried out for many yearsby the brewing industry and the most effectivetime to ensure oxygen availability is when theyeast begins to actively grow. The amount ofoxygen is important, but the timing of the supplyis also key. The oxygen must be supplied duringthe growth and division phase of the yeast. Indistillery practice, it may be sufficient to splashfill tankage just prior to fermentation, especiallyif active dry (aerobically grown) yeasts are used.

Clear substrates without insoluble lipidmaterials present may require more dissolvedoxygen. For example, with high gravity brewer’sworts that have been filtered and are almostdevoid of insolubles, bottled oxygen is used tooxygenate. The amount of oxygen needed byyeast growing anaerobically is small, butimportant, and deprivation leads to fermentationdifficulties.

Reproductive growth eventually ceases whenthe limiting value of sterols is reached. Whencell multiplication stops, the rapid uptake ofnitrogen and carbohydrates is affected and cellshave a reduced ability to complete thefermentation. In addition they are moresusceptible to external stresses such astemperature and pH. This has an impact on therate of ethanol production. The limiting level ofunsaturated fatty acids for brewing yeast is~0.5%.

Active dry yeasts (ADY), which are grownunder conditions of high aeration, have a muchhigher unsaturated fatty acid and sterol contentthan yeasts propagated or conditioned ‘in house’in an ethanol fermentation plant. This high‘credit’ of sterols and unsaturated fatty acidsallows the active dry yeasts to complete afermentation without the same oxygen (or sterol/fatty acid) requirement in the fermentation broth.It is important to note that these yeasts are neverreused. Distillery practices do not allow forharvesting.

Many distilleries and fuel alcohol plants useactive dry yeasts. Aeration through ‘splashfilling’ of batch fermenters or circulation of mashthrough heat exchangers for cooling purposesis probably adequate if active dry yeast is used.Aeration at the heat exchanger should be doneon the cold side of the exchanger to ensuremaximum solubility of oxygen on injection(remembering that oxygen is more soluble atlower temperatures).

Yeast sterols and oxygen

Sterols in aerobically-grown yeast may reach 1%total dry weight, but with anaerobic growth andno usable sterols in the medium, they dilute outto a value of less than 0.1%. There are severalsterols found in yeast of which ergosterol is themost prominent. Sterols play a key role inregulating membrane fluidity. When insufficientoxygen is available for membrane synthesis,yeast cells fail to grow and loss of membraneintegrity results in cell death. The regulation ofsterol synthesis is extremely complex. Yeast cellsaccumulate sterols into the membrane until theyhave met bulk requirement needs. After this,additional sterols are esterified to long chain fattyacids and stored in intracellular lipid vesicles.The most abundant sterol of the sterol pool iszymosterol. In an aerobic culture these sterolesters accumulate during the stationary phaseof growth and when these cells are re-inoculatedinto fresh medium these steryl esters are thenhydrolyzed and incorporated into the growingcell membranes.

Medium length fatty acids (C6–C10) arise viathe activity of fatty acid synthetase asintermediates in the formation of longer chainlength fatty acids, which are incorporated intothe various classes of yeast lipids. The fatty acid

112 I. Russell

biosynthetic pathways lead to the saturated andunsaturated fatty acids used in membrane lipids.Unless oxygen is present in small quantitieswhen growth begins, the desaturation steps donot proceed and unsaturated fatty acids andsterols cannot be synthesized by the yeast cell.

GLYCOGEN

Glycogen is the yeast’s major reserve carbohydrateand it is similar in formation and structure toplant amylopectin. It is a polymer of α–D-glucose with a molecular weight of ~108. It hasa branched structure containing chains of 10–14 residues of α–D-glucose joined by 1→4linkages. Individual chains are connected by(1→6) α–D-glucosidic linkages (Figure 21).

Glycogen serves as a store of biochemicalenergy during the lag phase of fermentationwhen the energy demand is intense for thesynthesis of such compounds as sterols and fattyacids. This intracellular source of glucose fuelslipid synthesis at the same time as oxygen isavailable to the cell. Dissimilation of glycogenand the synthesis of lipid are both rapid. Thehydrolysis of glycogen from approximately 20-30% to 5% and the corresponding productionof lipid from 5% to 11.5% of the cell dry weightoccurs within the first few hours after pitching.

Toward the later stages of fermentation, the yeastrestores its reserve of glycogen.

TREHALOSE

Trehalose is a non-reducing disaccharideconsisting of two glucose units linked togetherby an α-1, 1-glycosidic bond (Figure 22).Trehalose is present in particularly highconcentrations in resting and stressed yeast cells.During sporulation trehalose is virtually the onlysugar present in the yeast spores. Trehalose playsa protective role in osmoregulation. It protectscells during conditions of nutrient depletion,against toxic chemicals (ethanol, oxygenradicals, heavy metals) and starvation. It alsoplays a role in improving cell resistance to highand low temperatures as well as resistance todessication. This protective role may be due tothe stabilising effect of trehalose on cellmembranes. It is believed that the induction ofheat shock proteins and the accumulation oftrehalose are related. Trehalose accumulationoccurs during late stationary phase or in responseto environmental changes. Trehalose levels areimportant in the preparation of active dried yeastsince trehalose acts as a protectant during thedrying phase.

OH

OOH

OH

CH2OH

H

H

HH O

Glycogen

OH

OH

CH2OH

H

H

HH O

O

OH

OH

CH2OH

H

H

HH O

O

OH

OOH

CH2OH

H

H

HH O

α-(1-4) linkage

OH

OOH

CH2OH

H

H

HH O

OH

OH

CH2OH

H

H

HH O

O

OH

OH

CH2OH

H

H

HH O

O

OH

OOH

CH2

H

H

HH O

O

α-(1-6) linkage

Figure 21. Structure of glycogen: a high molecular weight polymer with branched-chain structure composedof D-glucopyranose residues.


OH

OOH

OH

CH2OH

H

H

HHO

H

OHHOH2C

H OH

HH

O

OH H

Figure 22. Structure of trehalose (C12

H22

O11

). An α-D-glucopyranosyl- α-D-glucopyranosyl.

YEAST MUTATIONS

Yeast mutations are a common occurrencethroughout growth and fermentation, but theyare usually recessive mutations, due to functionalloss of a single gene. Since industrial strains areusually aneuploid or polyploid, the strains willbe phenotypically normal and only if themutation takes place in both complementarygenes will the recessive character be expressed.

If the mutation weakens the yeast, the mutatedstrain will be unable to compete and soon beoutgrown by the non-mutated yeast population.The accepted view was that industrial yeaststrains were genetically very stable, however,with techniques such as DNA fingerprinting ithas been found that instability is not uncommonin many production brewer’s yeast strains. Thisfinding suggests that great care must be takenin regard to yeast generation productionspecifications.

Some commonly observed characteristicsresulting from yeast mutation that can negativelyaffect fermentation are: i) The tendency of yeaststrains to mutate from flocculent to non-flocculent ii) The loss of ability to fermentmaltotriose and iii) the presence of respiratorydeficient mutants.

Serial repitching exposes the yeast to a numberof stresses that lead to genetic drift and mutation.The most common mutation encountered is therespiratory deficient (RD) or ‘petite mutation’arising as a result of the accumulation ofsublethal DNA damage. The mutation arisesspontaneously when a segment of the DNA inthe mitochondrion becomes defective to form aflawed mitochondrial genome. Themitochondria are then unable to synthesizecertain proteins.

This type of mutation is also called the ‘petite’mutation because colonies with such a mutationare usually much smaller than the normalrespiratory sufficient culture. The respiratorydeficient mutation occurs at frequencies ofbetween 0.5% and 5% of the population but insome strains much higher figures have beenreported. Deficiencies in mitochondrial functionresult in a diminished ability to functionaerobically (hence the name respiratorydeficient), and as a result these yeasts are unableto metabolize the non-fermentable carbonsources lactate, glycerol or ethanol.

Flocculation, cell wall and plasma membranestructure, and cellular morphology are all affectedby this mutation. As a result the following effectshave been observed: slower fermentation rates,higher dead cell counts, reduced biomassproduction and flocculation ability, alterations insugar uptake and metabolic by-productformation, and reduced tolerance to stress factorssuch as ethanol and temperature.

Nitrogen and yeast growth

Stuck or sluggish fermentations occur when therate of sugar utilization becomes very slow orprotracted, especially toward the end of afermentation. When the sugar content is high andvital nutrients are lacking, a stuck fermentationcan occur. An inadequate level of assimilablenitrogen is often the culprit.

How much nitrogen does a yeast cell requireto grow and in what form? In traditional batchfermentations there is a 3 to 10-fold increase inyeast cell number. Yeast is normally inoculatedto a final concentration of 5 x 106 to 2 x 107

viable cells per mL. In terms of pitching weight,this is approximately 2 lb of active dry yeast per1000 US gallons (or 1 kg/4000 L). After growthin the fermentation medium the yeast producedweighs from 6 to 20 lb in 1000 gallons (or 3–10kg/4000L).

The mash is a microbiological growth mediumand if the right nutrients are not supplied in themash then the yeast will not grow well (or fermentrapidly) and lower levels of ethanol will result.Table 7 illustrates the total and useable levels offree amino nitrogen (FAN) from a number ofsources.

114 I. Russell

Table 7. FAN levels of typical mashes (mg/L mash*).

Total FAN Useable FAN

Wheat 82 64Barley 84 62Hulless barley 124 100Oats 193 159Hulless oats 184 130Rye 103 83Molasses 267 141Corn 70 58Starch slurry ~0 ~0

*all mashes normalized to 22% solids

Table 8 illustrates the FAN levels in a mash withbackset from a corn dry milling plant and Table9 is a corn wet milling example illustrating FANlevels with backset and steep water.

Table 8. FAN in mash and backset: corn dry millingexample1.

Usage level FAN FAN contribution(L/1000 L) (mg/L) (mg/L)

Backset 200 75 15*Mash ~800 ~88.4 70.7

total 85.7

*quality questioned1Ingledew, 1999

Table 9. FAN in backset and steepwater: corn wetmilling example1.

Usage level FAN FAN contribution(L/1000 L) (mg/L) (mg/L)

Backset 436 75 32.7*Light 982 756 74 steepwater

total 107

*quality questioned1Ingledew, 19992more is hard to add

NITROGEN SUPPLEMENTATION

The nitrogen content of a yeast cell variesbetween 3-9% (w/w) and one can thus make arough calculation as to how much suitable

nitrogen nutrient would be needed in afermentation. If yeasts are ~6% N and we wishto grow 6 mg/mL of new yeast (~300 x 106 cells),then 0.36 mg/mL of usable N is needed. Thus ifthe source of N is urea, 770 mg/L is needed. Ifthe source is NH3, 440 mg/L is needed and ifthe source is diammonium phosphate (DAP),1700 mg/L is needed. Tables 8 and 9 aboveillustrate the differences in useable FANdepending on the carbohydrate source and theuse of backset and steepwater.

FAN AND YEAST GROWTH

Yeast growth increases almost linearly with FANlevels up to 100 mg/L. Very high amounts ofFAN (up to 878 mg/L) have been shown to haveno additional effect on yeast growth, howeverthey did stimulate the fermentation rate.

C3.72 H6.11 O195 N0.61 is the formula proposedfor the major elements in yeast cell mass byHarrison (1967) and from this one can calculatethe nitrogen content of a cell. This is of the orderof 9% (w/w). The number is generally accuratefor aerobically-grown yeast, but is anoverestimation for yeast from anaerobicfermentations.

The level of protein inside a yeast is a functionof the concentration of usable nitrogen(ammonium salts, urea, amino acids, di- ortripeptides and nitrogen derived from nucleicacid breakdown) (Figure 23).

SOURCES OF NITROGEN

Ammonium ion, supplied as ammonia or thesulfate or the phosphate salt, is a preferrednitrogen source for yeast cultivation inlaboratory studies. When urea is utilized as thenitrogen source, biotin or other growth factorsmay be required. Urea, which is not a normalconstituent of mashes, is easily broken downinto two molecules of ammonium ion and onemolecule of carbon dioxide. Many countrieshave now banned the use of urea as a yeast foodingredient for potable alcohol manufacturingbecause it leads to the production of smallamounts of urethane (ethyl carbamate). Thiscompound is a suspected carcinogen in foods(Ingledew et al., 1987).


For beverage alcohol the main source ofnitrogen for yeast growth is amino acids formedfrom the proteolysis of grain proteins. The wort/mash contains 19 amino acids. The yeast takesup the amino acids in an orderly manner withdifferent amino acids removed at various pointsin the fermentation cycle (Table 10). Proline, themost abundant amino acid in wort/mash, hasscarcely been assimilated by the end offermentation, whereas over 95% of the otheramino acids have been used by the yeast.

There are a number of permeases for the uptakeof wort amino acids, some specific for individualamino acids and one that is a general amino acidpermease (GAP) with broad substrate specificity.The uptake and regulatory mechanisms arecomplex.

Although ammonia is a preferred nitrogensource for catabolic reactions, the cell will takeup certain amino acids first for directincorporation into proteins. The absorption ofthe amino acids by the yeast and their rate ofassimilation will vary both with the strain ofyeast and the amino acid composition of the wort/mash. Some amino acids such as glycine andlysine are readily taken up by the yeast.However, under nitrogen limiting conditionsthese two amino acids are inhibitory to growthand to fermentation. When usable nitrogen (suchas other amino acids) is in excess, the glycineand lysine are not inhibitory. A balanced nitrogensource (i.e. a mixture of amino acids) is moreefficient in providing nitrogen to the yeast foruse in the various biochemical pathways than asingle nitrogen source.

Figure 23. Nitrogen uptake by the yeast cell (adapted from Slaughter, 2003).

Carbamoyl

phosphate

CO2

CO2

UreaUrea

NH4

+

or

Glutamine

NH4

+

ADPATP

ATP

CO2

ADP

Pi

Allophanate

Amino

acids

Amino

acids

Cell membrane

Amino acids

Small peptidesSmall

peptides

NH4

+

Vacuole

116 I. Russell

NITROGEN DEFICIENCY

In regular gravity malt wort there is usuallyadequate amino nitrogen available for yeastgrowth but not when using adjuncts at highlevels since many of these adjuncts are relativelydeficient in amino nitrogen. When a starch slurryis used for fuel alcohol production, fermentationproblems will result unless a nutritionalsupplement is added. The medium cannotsupport the growth required to ensure enoughyeast biomass to catalyze the completeconversion of sugar to alcohol in a reasonabletime frame. The addition of corn steep liquor,ammonia or other yeast foods is needed to ensureyeast growth. Molasses-based media can bedeficient in ‘yeast-usable’ nitrogen. It can besupplemented with inorganic ammonium ion,urea or free amino acids. There may also be adeficiency in biotin, sulfur and phosphorus.

The addition of sugar syrups (compounds withno nutritive value other than the sugar content)dilute out the non-carbohydrate componentswhile increasing the mash specific gravity. Thiscan lead to a nitrogen deficiency that manifestsitself in a sluggish or stuck fermentation(O’Connor-Cox and Ingledew, 1989).

Attention to the nutritional status of mash isimportant in order to prevent prematuretermination of fermentation. It is too late toprovide nutrients or new yeast afterfermentations have become stuck or sluggish.

YEAST FOODS

Yeast foods and fermentation supplements aredesigned to provide the lacking nutrients of

fermentation media and have been successfullyused for many years to enhance and ensurepredictable fermentation times and productyields. Not only can they provide nitrogen, butdepending on the formulation they can alsoalleviate vitamin and mineral deficiencies andprovide other growth factors.

PROTEASE ADDITION

Yeasts used for ethanol production are notproteolytic and hence cannot break downproteins or utilize peptides much larger thantripeptides to obtain assimilable nitrogen forgrowth. Proteolysis, therefore, is intracellularunless the cells have lysed. Yeasts are only ableto utilize low molecular weight nitrogenousmaterials such as inorganic ammonium ions,urea, amino acids and small peptides. When theyeast is starved for nitrogen, this will triggercatabolism of amino acids.

The basic amino acids are converted on entryinto the cell to their N-free carbon skeletons. Thedeaminated carbon skeletons of hydrophobicamino acids are then excreted as α-keto acidsand the corresponding fusel oils. Nitrogenstarvation also leads to a small increase in proteindegradation inside the cell.

Commercial proteolytic enzymes such asAllprotease™ can be used to create usable FANfrom proteins and peptides in the mash. Figure24 shows the benefit of protease addition. Wherea plant is experiencing bacteriological problems,protease addition is recommended after yeastaddition rather than before so that liberatedamino acids are consumed by the yeasts ratherthan by the bacteria (Jones and Ingledew, 1994).

Table 10. Classes of wort amino acids in order of assimilation during fermentation1.

Group A Group B Group C Group DRapidly absorbed from start Slowly absorbed from start Slowly absorbed, late Absorbed slowlyof fermentation of fermentation in fermentation over fermentation

Aspartic acid/amine Histidine Alanine ProlineGlutamic acid/amine Isoleucine ‘Ammonia’Lysine Leucine PhenylalanineArginine Methionine TryptophanSerine Valine TyrosineThreonine Glycine

1Pierce, 1987.


It is worth noting that surface culture enzymessuch as Rhizozyme™ contain enough naturalprotease activity to make further additionsunecessary.

Endogenous enzymes

In malt and grain distilleries, the mash, as distinctfrom the mash in a brewing operation, still hasresidual enzyme activity and a high proportionof the cereal starch is converted to fermentablesugars throughout the fermentation bycontinuing starch hydrolysis. In fuel ethanolfermentations, the addition of exogenousenzymes to break down starches performs thesame function throughout the fermentation.

High gravity fermentation

High gravity fermentation employs worts/mashes at higher than normal concentration.High gravity fermentation is progressively beingintroduced around the world and yeast nutritionmust be addressed when the change is madefrom regular gravity to high gravity. Sufficientoxygenation, higher nitrogen levels and higheryeast addition are the key success factors.Increased osmotic pressure and ethanol levelscan affect the yeast and the stress tolerance ofthe yeast becomes a much more important factor.

Stress tolerance during the fermentation of highgravity media has been shown to be very straindependent. Although the issues of yeast nutritionin high gravity wort can be addressed by carefulselection of strains and fermentation conditions,failure to address them can result in seriousfermentation problems. Ingledew and coworkershave published extensively on the opportunitiesfor very high gravity fermentation (VHG) andhave extended the upper limits by successfullyproducing 23.8% (v/v) ethanol in a batchfermentation (Ingledew, 1999; Wang et al.,1999a,b).

Stress factors and synergism

Stress factors such as ethanol, temperature, andacids can significantly impact the vitality andviability of the yeast. These factors can actsynergistically and overpower the ability of theyeast to recover from stress. The list of stressfactors is long and yeast has an amazing abilityto recover from them – but there is a limit! It isimportant to remember that yeast is the heart ofthe fermentation process and if reasonable careand attention are given to ensure its health thepayback will be huge.

‘Select and treat your yeast well and it will workwell for you!’

Figure 24. Benefits of protease addition to an alcohol fermentation (Thomas and Ingledew, 1990, by permission).

0

5

10

15

20

0 50 100 150

Time (hours)

Dis

solv

ed s

olid

s (g

/100

ml)

0

100

200

300

400

500

0 50 100 150

Time (hours)

Yea

st c

ells

(m

illio

n/g

mas

h)

Protease

Control

Control

Protease

118 I. Russell

Acknowledgments

The author would like to acknowledge Dr. M.Ingledew and his coworkers (University ofSaskatchewan) whose research work andprevious publications are the foundation for thischapter, in particular their pioneering work inthe fuel ethanol industry in the area of very highgravity fermentation and nitrogen requirementsof yeasts. Acknowledgments are also due to Dr.G.G. Stewart (Heriot Watt University-International Centre for Brewing and Distilling.

References

American Society of Brewing Chemists. 1980.Report of the subcommittee on microbiology.J. Am. Soc. Brew. Chem. 38:109-110.

Austin, G.D., R.W.J. Watson, P.A. Nordstorm andT. D’amore. 1994. An online capacitancebiomass monitor and its correlation with viablebiomass. MBAA Technical Quarterly 31:85-89.

Barbour, E.A. and F.G. Priest. 1988. Someeffects of Lactobacillus contamination inScotch whisky fermentation. J. Instit. Brewing94:89-92.

Campbell, I. 2003a. Wild yeasts in brewing anddistilling. In: Brewing Microbiology, 3rd

Edition. Kluwer Plenum, New York, pp. 247–266.

Campbell, I. 2003b. Yeast and Fermentation. In:Whisky: Technology, Production andMarketing (I. Russell, ed). Academic Press,London, pp. 117-154.

Dolan , T.C.S. 1976. Some aspects of the impactof brewing science on Scotch malt whiskyproduction. J. Instit. Brewing 82:177-181.

Harrison, J.S. 1967. Aspects of commercial yeastproduction. Process Biochemistry. 2(3): 41-46.

Heggart, H.M., A. Margaritis, H. Pilkington, R.J.Stewart, T.M. Dowhanick and I. Russell. 1999.Factors affecting yeast viability and vitality.MBAA Technical Quarterly 36(4):383-406.

Heggart, H., A. Margaritis, R. Stewart, H.Pilkington, J. Sobczak and I. Russell. 2000.Measurement of yeast viability and vitality.MBAA Technical Quarterly 37(4):409-430.

Ingledew, W.M., L.A. Langille, G.S. Menegazziand C.H. Mok. 1977. Spent brewers yeast –

analysis, improvement and heat processingconsiderations. MBAA Technical Quarterly14:231-237.

Ingledew, W.M., C.A. Magnus and J.R.Patterson. 1987. Yeast foods and ethylcarbamate formation in wine. AmericanJournal of Enology and Viticulture. 38: 332-335.

Ingledew, W.M. 1999. Alcohol production bySaccharomyces cerevisiae: a yeast primer. In:The Alcohol Textbook: A Reference for theBeverage, Fuel and Industrial AlcoholIndustries (K.A. Jacques T.P. Lyons and D.R.Kelsall, eds). Nottingham University Press,Nottingham, UK, pp. 49-87.

Jones, A.M. and W.M. Ingledew. 1994. Fuelalcohol production assessment of selectedcommercial proteases for very high gravitywheat mash fermentation. Enzyme Microb.Tech. 16:683-687.

Jones, R.P., N. Pamment and P.F. Greenfield.1981. Alcohol fermentation by yeast-The effectof environmental and other variables. ProcessBiochem. 16(3):42-49.

Jones, R.P. and P.F. Greenfield. 1984. A reviewof yeast ionic nutrition - Part 1: Growth andfermentation requirements. Process Biochem.19(2):48-60.

Maloney, D. 1998. Yeasts. In: Kirk-OthmerEncyclopedia of Chemical Technology, 4th

Edition (J.I. Kroschwitz and M. Howe-Grant,eds). John Wiley & Sons, New York, pp. 761-788.

Neish, A.C. and A.C. Blackwood. 1951.Dissimilation of glucose by yeast at poisedhydrogen ion concentrations. Canadian J.Tech. 29:123-129.

O’Connor-Cox, E.S.C. and W.M. Ingledew.1989. Wort nitrogenous sources – Their useby brewing yeasts: a review. J. Amer. Soc.Brewing Chem. 47:102-108.

Patel, G.B. and W.M. Ingledew. 1973. Internalcarbohydrates of Saccharomyces carlsbergensisduring commercial lager brewing. J. Inst.Brewing 79:392-396.

Peppler, H.J. 1970. Food yeasts. In: The Yeasts,Vol 3, Yeast Technology. Academic Press, NewYork, pp. 421-461.

Pierce, J.S. 1987. The role of nitrogen inbrewing. J. Instit. Brewing 93:378-381.


Reed, G. and T.W. Nagodawithana. 1991. In:Yeast Technology, 2nd Edition. AVI. VanNostrand Reinhold, New York.

Slaughter, J.C. 2003. The biochemistry andphysiology of yeast growth. In: BrewingMicrobiology, 3rd edition. Kluwer Plenum,New York. pp. 19-62.

Smart, K.A., K.M. Chambers, I. Lambert and C.Jenkins.1999. Use of methylene violet stainingprocedures to determine yeast viability andvitality. J. Am. Soc. Brew. Chem. 57(1):18-23.

Smart, K. 2000. The death of the yeast cell. In:Brewing Yeast Fermentation Performance.Blackwell Science, United Kingdom, pp. 105-113.

Thomas, K.C. and W.M. Ingledew. 1990. Fuelalcohol production: effects of free aminonitrogen on fermentation of very high gravitywheat mashes. Appl. Environ. Microb.56:2046-2050.

Walker, G. 1999. Yeast Physiology andBiotechnology. John Wiley and Sons Ltd.United Kingdom.

Wang, S., W.M. Ingledew, K.C. Thomas, K.Sosulski and F.W. Sosulski. 1999a.Optimization of fermentation temperature andmash specific gravity for fuel alcoholproduction. Cereal Chem. 76:82-86.

Wang, S., K.C. Thomas, K. Sosulski, W.M.Ingledew and F.W. Sosulski. 1999b.Grainpearling and the very high gravity (VHG)fermentation technologies for fuel alcoholproduction from rye and triticale. ProcessBiochemistry. 34(5): 421-428.

Recommended reference books

The following books were invaluable in thepreparation of this chapter and are highlyrecommended to anyone wishing to gain morein depth knowledge on yeast.

An Introduction to Brewing Science andTechnology, Series III, Brewers Yeast.1998.G.G. Stewart and I. Russell. Published by theInstitute and Guild of Brewing. Available fromwww.igb.org.uk.

Biotechnology of Malting and Brewing. 1991.J.S. Hough, Cambridge University Press.

Brewing Microbiology. 2002. F.G. Priest and I.Campbell. (eds). Kluwer Academic.

Beers and Coolers. 1994. Manfred Moll.Springer-Verlag, New York.

Brewing Yeast and Fermentation. 2001. C.Boulton and D. Quain. Blackwell Science.

Brewing Yeast Fermentation Performance. 2000.1st Edition. K. Smart (ed). Blackwell Science.

Brewing Yeast Fermentation Performance. 2003.2nd Edition. K. Smart (ed). Blackwell Science.

Handbook of Brewing. 1994. W. A. Hardwick.Marcel Dekker.

Standards of Brewing. 2002. C.W. Bamforth.Brewers Publications.

Technology - Brewing and Malting, 2nd Edition.1999. W. Kunze. Verlag der VLB Berlin.Available from the VLB at www.vlb-berlin.org/english/books/kunze/.

The Alcohol Textbook: A Reference for theBeverage, Fuel and Industrial AlcoholIndustries. 1999. K.A. Jacques, T.P. Lyons andD.R. Kelsall (eds). Nottingham UniversityPress, Nottingham, UK.

The Practical Brewer [MBAA], 3rd Edition. 1999.Available from the Master Brewers Associationof the Americas at www.mbaa.com.

Whisky: Technology, Production and Marketing.2003. I. Russell (ed). Academic Press, NewYork.

Yeast Physiology and Biotechnology. 1998. G.Walker. John Wiley and Sons Ltd., New York.

Practical management of yeast: conversion of sugars to ethanol 121

Chapter 10

Practical management of yeast: conversion of sugars to ethanol

DAVE R. KELSALL AND T. PEARSE LYONS


Introduction

Fermentation is the critical step in a distillery. Itis here that yeast converts sugar to ethanol; andit is here that contaminating microorganisms findan opportunity to divert the process to lactic acid,glycerol, acetic acid, etc. The fermentation step(and to a much lesser degree the cookingprocess) is also where yeast must be ‘coached’to produce maximum levels of ethanol,sometimes as high as 23%. Fermentation willalso have a direct bearing on downstreamprocessing. If sugar levels remaining in beer aretoo high, evaporative capacity may be impairedor syrup contents increased. Both can lead todistillers grains being more difficult to dry andaltered in color as Maillard reactions occurbetween proteins and sugars. Truly thefermentation step is the heart of the distillery.

When reflecting on the overall flow diagramof the distillery (Figure 1) the central role of yeastis obvious. The process of converting sugars toethanol takes between 10 and 60 hrs duringwhich heat is generated. The various factorsaffecting the efficiency of that process need tobe considered. In the chapters by Russell andIngledew, the biochemistry of ethanol productionby yeast is covered in detail. In this chapter wewill discuss the factors involved in taking

liquefied mash from cooking (1-2% glucose) toensure that pre-formed sugar and other boundsugars (starch and other polysaccharides) arereleased in a timely fashion for maximumconversion to ethanol by yeast.

Choosing a yeast strain

One of the most important decisions a distillermust make is the selection of yeast strain. Thebest characterized yeast is Saccharomycescerevisiae, but within this species there arethousands of unique strains. Which one to use?Several decisions must be made (Table 1). Thefirst decision is whether or not to grow the strainin-house. In fact, yeast strain maintenance andgrowth is a job best left to the experts. For thevery small cost per gallon it represents (less than0.5 cents), there is just too much at stake.

Table 1. Choices among yeast strains.

• Grow in-house?• Active dried or pressed?• High alcohol tolerance needed? Thermosacc™• Lower alcohol in fermentor: Superstart™


ACTIVE DRY OR PRESSED YEAST?

The choice between active dried (95% solids)or pressed yeast (35% solids) very often comesdown to two points: adequate cold storage(pressed yeast must be stored at 5°C and has ashelf life of 1-2 months) and whether the yeastmust ‘start immediately’. Yeast requires time toadapt to any new environment and activate itsmetabolism, but first must be rehydrated. Thisis a period of zero growth but intensebiochemical activity. If we are conditioning theyeast, the advantage of a pressed yeast is thatthe lag phase is minimized because the yeast isalready hydrated. Cell numbers are typically 7-10 billion cells/g. Active or instant dry yeast(ADY) are vacuum packed, often undernitrogen, and are easy to store with minimallosses in activity (5% over 6 months at 5°C).ADY requires wetting, and the lag phase tendsto be 1-2 hrs. Cell numbers 15-20 billion cells/g.

IS HIGH TEMPERATURE AND/OR HIGHALCOHOL TOLERANCE NEEDED?

Yeast is sensitive to temperature, high or low,and responds by activating stress mechanisms

1 bushel ofcorn

(56 lbs)

Fermentablesugar

(36 lbs)

Ethanol(17.6 lbs)

CO2(18.4 lbs)

Heat(7450 BTU)

DDGS(17 lbs)

The distillery flow diagram:from corn to ethanol plus

co-products

+

+

+

Starch(32 lbs)

Milling

LiquefactionSaccharification

FermentationRecoveryDryhouse

Figure 1. Distillery product flow.

including elevation of intracellular trehalose.Additionally, certain proteins, called heat shockproteins, are synthesized. Some strains are betterable to withstand temperature stress than others.Such yeast are also usually able to tolerate higheralcohol levels. In the 1980s it was customary torun fermenters at 8-10% ethanol but now 16-17% is the norm. One yeast, Thermosacc™ , isa good example of a stress resistant strain.Smaller in size (5 µm instead of 8-10 µm, Figure2), it has a temperature tolerance of >40°C andan ethanol tolerance in excess of 20%. Anunusual feature is its ability to tolerate high aceticand lactic acid levels. This may explainobservations of ‘cleaner’ fermentations asinfecting microorganisms such as Lactobacillushave less impact. In beverage alcoholproduction, little, if any, change in congenershas been reported following its use.

Controlling yeast stress factors that affectalcohol yield

TEMPERATURE: THE FIRST STRESS FACTOR

Inability to precisely control the temperature of


fermentation is possibly the biggest and mostcommonly encountered factor or problemaffecting alcohol yield. Almost all distilleriessuffer from underestimation of the amount ofenergy released during fermentation. Considerthe general conversion of glucose to alcohol asillustrated in Figure 1. Generation of most ofthe energy of fermentation of ethanol takes placebetween hours 10 and 30 of fermentation. Inother words, the fermentor cooling systemshould be designed to cool 44,000 BTU per 100lbs of ethanol produced over a 20 hr period.Often this amount of cooling is not designedinto the system and the fermentors becomeoverheated. Engineers did not anticipate theenormous strides that would be made in yeastfermentation (higher alcohol tolerance, fasterconversion) and designed for shorter residencetimes and less cooling capacity.

The theoretical temperature rise for a 25 gallonbeer (25 gallons of water per bushel of grain) is22oC per bushel (Figure 3). The optimumtemperature of fermentation for yeast(Saccharomyces cerevisiae) is 35°C and theoptimum temperature of reproduction for theyeast is 28°C (yeast capable of reproducing at35°C is now available). It can also be seen (in

Figure 3) that without adequate cooling, afermentor set at 35oC would inevitably increasein temperature. Every degree in temperatureabove 35°C will have a tendency to depressfermentation, as the yeast cannot handle thesehigher temperatures. Furthermore, the increasein temperature favors the growth of lactobacilli,

Figure 2. Thermosacc™ vs regular yeast.

25 30 35 40 45 50

Mash bill gallons (gallons of water per bushel)

5

10

15

20

25

30

Tem

per

atu

re r

ise

(°C

)

Figure 3. Theoretical temperature increase duringfermentation.


the microorganism that competes with yeast forglucose. Heat also increases the impact of otherstress factors such as mycotoxins and salt levels.

Combining enzyme activities to control sugardelivery to yeast

How can we best control fermentationtemperature? The easiest way to reducefermentor temperature would be to reduce thesugar level going in and therefore reduce yeastgrowth; however, less alcohol would beproduced. Since the objective is more, not lessalcohol, this is not an option. However, sugarlevel, or more importantly sugar type, isimportant.

Yeast must be maintained in a growth(budding) phase, since their ability to producealcohol is more than 30 times greater duringgrowth than in non-growth mode. Yeast howevercan respond to high glucose levels in two ways.Cell growth is either inhibited by high glucoselevels or yeast may grow rapidly and then stop.In either case there is an over-dosing effect onfermentation efficiency and possibly a spike intemperature. Glucose must be ‘spoon-fed’ to theyeast; and for this reason it is recommended thatglucoamylases be used both sparingly and inconjunction with an enzyme produced in surfaceculture called Rhizozyme™. The combinationprovides ideally balanced sugar delivery to

maintain good yeast growth and maximumalcohol productivity while avoiding hightemperature peaks. Glucoamylase (AllcoholaseII L400) provides glucose at the start offermentation while the Rhizozyme™, with pHand temperature optimums more closely alignedto fermentation conditions, continues to workduring fermentation. The net result is a steady,slow release of glucose and a steady formationof alcohol by yeast.

RhizozymeTM also affects ethanol yield as it iscapable of hydrolyzing starch that isunconverted in the cook process and alsoconverts some of the cellulose in the corn toglucose.

The key is to consider cooking andfermentation together. Cooking must bedesigned to release as much starch, both boundand unbound, as possible without overloadingthe yeast with glucose. It must also be recognizedthat corn has, in addition to starch, otherpotentially useful carbohydrates includingcellulose and hemicellulose. Conventionalglucoamylases do not degrade either, whereasglucoamylase produced by surface culturefermentation (Koji) also contains a range ofcarbohydrase enzymes as ‘side activities’ thatdegrade some of the cellulose thereby releasingmore sugar, which boosts yield (Table 2,Figure 4).

Glucoamylase

Eth

ano

l (%

v/v

)

Rhizozyme

0

5

10

15

20

25 0.04% 0.06% 0.10%

Figure 4. Comparison of alcohol yields from high solids corn mash with traditional and surface culture glucoamylases.


Table 2. Comparison of surface culture andconventional glucoamylase.

Production Surface culture Conventional(Koji) fermentation deep tank

liquid

ActivitiesGlucoamylase 40 400Cellulase 400 —Hemicellulase 350 —α-amylase 5000 —Phytase 300

pH profile Broad NarrowTemperature

optimum, °C 30-40 0-60Activity on

gelatinized starch Rapid SlowActivity on raw

starch Reasonable Slow Yeast stimulation Significant Poor

Modeling provides a prediction of the increasein yield to be expected. Predicted yields of 3.1gallons per bushel, 17-18% higher than thestandard 2.65, have been calculated based onincreased sugar release. This is 18 gallons (>65liters) more per tonne of grain. Therecommendation is to use a 50/50 blend ofRhizozyme™ Koji and conventionalglucoamylase to maximize yield. Where distillerscan reliably measure yield, the results in termsof costs per gallon reveal optimum returns.

Equipment for heat control

Despite the sequential feeding of glucose, heatis generated during fermentation and must beremoved. Four types of heat exchangers aretypically used; and these are illustrated later inthis chapter when fermentation design isdiscussed.a) Internal cooling coils. These are the least

desirable option because they are practicallyimpossible to clean and sterilize.

b) Internal cooling panels. Like cooling coils,panels are suspended in the fermentationvessel. Also like cooling coils, cooling isadequate but sterilization may be difficult.

c) External cooling jacket. Cooling jackets,often dimpled for added contact with thefermentation, can cover either part of the

vessel or the entire vessel. Recirculatingcoolant through the jacket allowsfermentation temperatures to be monitored.If jackets are spaced out over the fermentorsides, then cooling can commence as thefermentor is filled.

d) Heat exchanger with recirculation. Contentsof the fermentor are pumped through a heatexchanger (typically spiral or plate-and-frame) and returned to the fermentor. Thishas the added advantage of acting as a meansof keeping the fermentor agitated; andnutrients (sterol donors, peptides, oxygen,etc.) can be introduced at critical times. It issuggested that optimum control oftemperature can be obtained by placing theheat exchangers in such a position thatrecirculation can start at the beginning of thefill when the yeast starts producing heat.

Of the four types, the cooling jacket is best forminimizing infection, but the heat exchanger isthe most effective at heat control. As the distillermoves towards very high levels of ethanol (over20% by volume), heat recovery and maintenanceof low temperatures will become even moreimportant as temperature rises of 6-8°C willbecome the norm. A word of caution: somedistilleries, particularly in continuous operations,share external heat exchangers betweenfermentors in order to economize. Based on theexperience of many existing distilleries using thissharing concept, infection can easily betransferred among fermentors as well as havinginadequate cooling.

INFECTION: THE SECOND STRESS FACTOR

The various types of infection are described indetail elsewhere in this book, however severalpoints are relevant in this context. Lactobacilliconsume glucose to produce lactic acid andacetic acid, which is the second major factoraffecting the yield of alcohol in fermentation,which in turn has a major impact on distilleryeconomics (Table 3). Yeast do produce someorganic acids during fermentation, butconcentrations are relatively low compared tothose produced by lactobacilli and othercontaminating bacteria. As a general rule of


thumb, if the titratable acidity of an uninoculatedmash is X, then the titratable acidity of thefermented beer is usually about 1.5X. This is ageneral rule and seems to work quite well inpractice, although it is not based on knownscientific principles. The 1.5X represents thetitratable acidity of the finished beer when thereis no significant contamination from lactobacilli.When lactobacilli are active, the generation oflactic and acetic acids substantially increases thetitratable acidity and often the high acid contentwill cause the yeast fermentation to stop ordramatically slow down. Very early into aninfection the titratable acidity reaches 2X.

Practical means of controlling infection

The most fundamental way to control infectionis to control the fermentation environment (Table4). Fermentors must be cleaned on a regularbasis; and designs must avoid sharing of heatexchangers and dead legs in piping. Fermentorsshould be designed to facilitate completeemptying. Scaling will occur regardless of thefermentor design, and should be removed usingeither a scale inhibitor (such as Scale-BanTM) ora de-scaling chemical (an acid wash).Fermentations must be fast; and yeast should beadded either in a form that will quicklyrehydrate or preferrably one that has been‘started’ in a conditioning tank. Saccharificationtanks can be a source of infection aslactobacillus strains can adapt and grow at thetemperatures used in this process. The goal is tomake conditions as suitable as possible for yeastgrowth (without jeopardizing alcohol yield) andas unsuitable as possible for infectiousmicroorganisms. By restricting glucose releasein fermentation by using Rhizozyme™ andlower levels of glucose-releasing glucoamylase

(typically 50-70%), the yeast uses the sugar asit becomes available and competitively excludeslactic and acetic acid-producing bacteria byreducing available sugar. Equally, when yeastgrowth rate begins to decline (typically at 20-24 hrs) sterol donors and oxygen, which at thatstage have become limiting, will revive activityby ensuring that fatty acids are available foryeast reproduction.

Table 4. Practical means of controlling infection.

Avoid deadlegs in linesAvoid sharing fermentor heat exchangersProper cleaning programs

Periodic use of de-scaling chemicalsControl fermentation temperatures Avoid the saccharification step in cooking

Maximize yeast growth (budding)Use yeast yeast conditioning

Use sterol donors and peptides as yeast foods in yeast conditioningAntimicrobials

AllpenTM (against Gram-(+) organisms) LactosideTM (against Gram-(+) and Gram-(-) organisms)

Regardless of precautions, a distillery is not asterile environment and opportunities for growthof both Gram (+) and Gram (-) bacteria exist.Because their growth rate is so much faster thanyeast, strict controls to prevent infection mustbe in place. As early as 1954 penicillin was usedin ethanol fermentation; and owing to its quickinactivation caused no problems with regulatoryagencies. Other antibiotics such asvirginiamycin are effective antimicrobials, butsome are now known to suppress yeastfermentation and in addition can leave residues(2.6%) in spent grains. With the ban in Europeagainst use of virginiamycin in products destinedfor animal feeds, caution is urged. Furthermore,

Table 3. Relationship between bacterial numbers and ethanol production losses.

Viable bacteria Ethanol lost1 Ethanol lost2 Financial loss3

in mash at set (CFU/ml) (% v/v) (gallons) (in US dollars)

105 0.1 - 0.2 80,000 112,000106 0.2 - 0.4 160,000 224,000107 0.6 - 1.0 400,000 560,000108 0.9 - 1.2 480,000 672,000109 1 - 1.5 600,000 840,000

1Depending on the strain of Lactobacillus.2Maximum loss calculated based on a 40,000,000 gallon/yr.3Calculated based on ethanol price of $1.40/gal.


resistance to both penicillin and virginiamycinis increasingly common.

One of the strategies to avoid this risk is totake advantage of the synergy that occurs whenantimicrobial substances are combined, whichlowers the amounts used. Lactoside™, acombination of antimicrobials withoutvirginiamycin, has proven very effective for thisreason. A routine maintenance dose of 1-2 ppmis practical as a precautionary measure. Whenthere is infection the dosage is increased to 2-5ppm. There is no evidence of carryover of theseantibiotics into spent grains and solubles.

ALCOHOL LEVELS: A THIRD STRESS FACTOR

High alcohol beers have a tendency to stopfermenting. Understanding how alcohol levelsact to stress yeast requires a working knowledgeof yeast metabolism and growth. It has beenshown that when yeast are in the reproductivephase they produce alcohol over 30 times fasterthan when not reproducing. Figure 5 showstypical yeast growth patterns or phases duringfermentation. The first phase is the so-called lagphase, during which the yeast adapt to thefermentor environment. During this period thereis little or no yeast growth and consequently littleor no alcohol production. This period can lastfrom 4 to 12 hrs. Ways of managing andreducing the length of the lag phase are shownin Figure 6 in reference to yeast conditioning.

Lag

phase

Exponential

growth phase

Stationary

phase200

160

120

80

40

10

Adaptation

to mash

Decline

phase

Viable

cell count

Total

cell count

12 24 36 48 60 72

Time (hours)

Cel

l po

pu

lati

on

(m

illio

ns

per

ml)

Maximum

growth

Figure 5. Typical yeast growth curve in distillery mash.

10,000 - 20,000 gallonsHold for 8 hours

• Good agitation

• Good cooling

• Can be sterilized

Figure 6. Yeast conditioning tanks: A key factor inmaximizing yeast cell numbers and improving alcohol levels.

The second phase is the exponential growthphase. This is the most important phase wherenearly all of the alcohol is produced. There is alimited time in which the yeast stay in this phase;and this is the factor limiting the quantity ofalcohol produced during a given fermentation.The length of time during which the yeast canremain reproducing depends on the nutritionavailable in the fermentor.

The three major factors that significantly affectalcohol yield are therefore temperature, acidityof the mash and the alcohol level producedduring fermentation. Yeast is tremendouslyresilient and usually produces efficiently despiteall the negative conditions imposed. Yeastfermentation will yield well if the temperature isa few degrees high or if the acidity is a littlehigh indicating a mild infection. Beerscontaining up to 20% alcohol v/v can beproduced without slowing fermentation.However, when changes in two or more of thesefactors happen simultaneously, serious losses inyield usually occur. For example, in a mash at37°C contaminated with lactobacilli at 10 millionCFU/ml, it is likely that fermentation would stopat 8% alcohol, a 33% loss in yield.

MAXIMIZE YEAST PERFORMANCE BYCONDITIONING

Yeast is the powerhouse of any distillery; andwithout healthy yeast, alcohol percentages in thefermentor and alcohol yield will drop.Conditioning tanks are a way to ensure goodcell numbers in the fermentor (Table 5). At thestart of batch fermentation there should be a


minimum of 50 million cells/ml, which shouldincrease to 150-200 million cells/ml at the heightof fermentation and drop to 100 million cells/ml at the end. Continuous fermentors shouldmaintain cell numbers of 150 million from a highof 250-300 million cells/ml. Yeast viability asmeasured by methylene blue stain should be 98-99% in the conditioning tank and 90% at thestart of fermentation. There will be a drop toaround 50% viability at the end of fermentation.In a cascade continuous fermentation system,viability was seen to drop to 30% in the lastfermentor.

Table 5. Keys to a good yeast conditioning tank (pre-fermentor).

• Keep Brix to 10-20°, preferably ~14°Brix• Provide good agitation• Use peptide based yeast food• Achieve minimum of 200 ppm FAN (free amino nitrogen)• Add Rhizozyme™ to ‘spoon-feed’ yeast glucose

(0.05% of mash)• Hold for 4-8 hrs with agitation• Achieve 300-500 million cells/ml before transfer to main

fermentor• Pump to fermentor• Clean and sterilize• Repeat

A typical charge for a continuous yeastconditioning tank is shown in Table 6. The mashwas sterilized and cooled to 90°F. Yeast wasadded with good agitation along with yeast foodand saccharifying enzyme (Rhizozyme™). Themash was held for 8 hrs at which stage a cellcount in excess of 500 million was achieved.The mash was transferred to the fermentor andthe tank cleaned and sterilized. In this waydistillers ensure good yeast growth withminimum ‘lag phase’ periods.

Table 6. Typical charge for yeast conditioning tank for a500,000 gallon fermentor.

Prefermentor size 20,000 gallonsMash Direct from jet cooker: dilute to

14°BrixActive yeast 200 lb ThermosaccTM

Peptide yeast food 20 lbs AYFRhizozyme™ l lbYeast cell numbers expected 400-500 millionTime 8 hrs

MYCOTOXINS: A STRESS FACTOR

In 1985 the Food and Agricultural Organization(FAO) of the United Nations estimated that at least25% of the world’s grain supply is contaminatedwith mycotoxins. The type and degree ofcontamination depend on many factors includingconditions during growth and harvest andgeographic area. Toxins such as zearalenone andvomitoxin (deoxynivalenol) produced by growthof Fusarium molds in stored grain tend to be themycotoxins of concern in the more temperateclimates of North America and Europe whileaflatoxin, produced by various species ofAspergillus, predominates in tropical climates andcan be a serious factor in parts of the US. Howevera wide range of toxins and mixtures of toxins areusually present and all are of concern in the animalfeed and human food chain. To a distiller, the mainconcern with mycotoxins is whether they passthrough into the DDGS. The FDA have setdefinitive limits on aflatoxin levels for interstategrain shipments; and the increasing knowledge ofthe variety of mycotoxins and extent ofcontamination possible have heightened concernin all of the animal feed industry.

Of more recent concern to the distiller has beenthe finding that mycotoxins can also affect yeastgrowth (Table 7). Perhaps presence of these toxinsmay provide at least part of the reason behindunexplained unfinished fermentations. Since notall mycotoxins are destroyed by heating, it isunlikely that the cooking step would remove orinactivate them. In animals many of themycotoxins damage protein metabolism causingreduced growth and increased diseasesusceptibility. Research into ways of nutritionallycompensating for mycotoxin presence in animalsfeeds led to discovery that yeast cell wallglucomannans have characteristics that make themvery specific adsorbents for certain mycotoxins.For example, about 80% of the zearalenone insolution can be bound and inactivated by addingyeast glucomannans. These products are alsobeginning to be used successfully in thefermentation industry.

Table 7. Effects of mycotoxins on yeast growth.

Mycotoxin Level required to inhibityeast growth (ppm)

Zearalenone 50Vomitoxin 100Fumonisin 10


PHYTIC ACID: AN ANTINUTRITIONAL FACTOR

Phytic acid, the form in which 60-80% of thephosphorus is stored in cereal grains, is a stressfactor for yeast that may be converted into anutritional benefit. The phytic acid molecule hasa high phosphorus content (28.2%) and largechelating potential. In addition to holdingphosphorus in tightly bound form, phytic acidcan form a wide variety of insoluble salts withdi- and trivalent cations including calcium, zinc,copper, cobalt, manganese, iron andmagnesium. This binding potentially rendersthese minerals unavailable in biological systems;and makes phytin content a familiar‘antinutritional factor’ when cereal grains are fedto livestock. Animal nutritionists eithercompensate for the presence of phytin withextra-nutritional levels of zinc, calcium andphosphorus or add microbial phytases to dietsfor pigs and poultry.

Phytin may also have an antinutritional effecton yeast. Calcium ion concentration affectsenzyme activity; and minerals such as zinc areneeded for yeast growth. Phytic acid is alsoknown to inhibit proteolytic enzymes; and

phytate-protein or phytate-mineral-proteincomplexes may reduce the breakdown of proteinto amino acids required by yeasts for growth.Starch is also known to be complexed by phytate.Finally, inositol, a limiting nutrient for yeastgrowth, forms a part of the phytin molecule. Theaddition of a phytase enzyme (often presentnaturally in Koji surface culture enzyme sources)can overcome this problem.

SUMMARY: STRESS FACTORS AFFECTINGYEAST ACTIVITY

As we ‘coach’ yeast to produce very high levelsof ethanol, it becomes of paramount importancethat we both understand the stress factorsinvolved and design equipment to allow us tocontrol them (Figure 7). Critical stress factorssuch as temperature, alcohol levels and infection,while understood individually, combine toproduce unique conditions in everyfermentation. Such conditions will vary withfeedstocks used, but also with changes in thecomposition of individual feedstocks due toseasonal or climatic variables. An understanding

Figure 7. Stress factors affecting yeast.


of other feedstock-related stress factors such asmycotoxins or phytin content will also becomemore important as we move toward higherethanol production.

Fermentation systems used in distilleries

Before considering the specific managementtools available to the distiller, it is necessary toconsider the differences among the fermentationsystems in use. Over the last 30 years thebrewing and distilling industries have developednew fermenting systems; and the distilling

industry has largely converted to rapid batchfermentation with cylindro-conical or slopingbottom fermentors, or to cascade continuousfermentation using cylindro-conical fermentors.

BATCH FERMENTATION

A typical cylindro-conical fermentor with skirtsupport is illustrated in Figure 8. Thesefermentors are usually designed to ferment from30,000 gallons to 750,000 gallons. Fermentorsneed to be fabricated on site if the diameter andlength exceed a size that can be safelytransported by road.

Figure 8. Alcohol fermentor with sloping bottom (Alltech/Bishopric, 1981).

Inspection light

Vent

From CIP moduleMash inYeast/enzymes in

20 in. Manway/sampling port

Coolant

Insulation(outdoor tanksin cold climates)

Slope

Concrete pad

Alternative cooling methods:A. External cooling jacketB. External heat exchanger with recirculationC. Internal cooling coilsD. Internal cooling panels

CIP spray head

A

15 in. x 20 in.access Manway

Recirculation pump

Outlet

B

CAgitator(if required)

D

Coolant


Different cooling systems are available. Manydistilleries use chilled water to cool thefermentors. From a microbiological standpointthe most desirable cooling system by far usesexternal cooling jackets. The least desirableoption employs internal cooling coils becausethey are practically impossible to clean andsterilize from the top-mounted CIP (clean-in-place) spray nozzles. The external recirculationheat exchanger is frequently used in distilleriesand is sometimes shared with other fermentors.This is a poor option as there are times whenmore than one fermentor needs to be cooled.The main disadvantage of this system whenshared among fermentors is bacterial cross-contamination. With such a system, themicrobiological condition of all fermentors willbe determined by the fermentor with the mostcontamination. Even when these recirculatingheat exchangers are not shared, they are difficultto clean. Throughput of at least 7 ft3/sec mustbe obtained to ensure turbulent flow through thetubes.

The agitator is required particularly at the startand at the end of fermentation. Frequently theagitators are badly designed and provide poormixing. A folding action is required to ensureproper mixing of the mash solids and to ensurean even temperature throughout the fermentor.

Carbon dioxide (CO2) is removed through thevent. Fermentors must be fitted with pressurerelief valves and vacuum breakers to avoidserious accidents. The CO2 is frequentlycollected and sold. In any case, CO2 should bescrubbed to remove alcohol, which is returnedto the beer well. The CO2 header manifold canbe a source of contamination as infected mashfrom one fermentor can migrate in the CO2 intoanother non-contaminated fermentor. The CO2

manifold should be designed to be cleanable.The lag phase is a great opportunity for

lactobacilli to become established. Bacteriaunder optimum growth conditions can reproduceevery 20-30 minutes; however yeast, which aremuch larger microorganisms, can onlyreproduce every 3 hrs. A single bacteriumreproducing every 20 minutes would produce apopulation of 256 in 3 hrs. Some yeast strainshave lag phases as long as 12 hrs, during whichtime the lactobacilli can become heavilyestablished. Therefore, it is very important tochoose a yeast strain with a very short lag phase

and to use a conditioning tank so that the mainfermentation starts immediately after the yeastis added.

There is a choice of materials to use in themanufacture of fermentors. Stainless steel isgenerally the best choice as it is easier to cleanand sterilize. It also lasts much longer than mildsteel or lined vessels; and when considered overthe expected lifetime of a fermentor it is by farthe most economical option. If chloride levelsare high, then special stainless steels should beconsidered to avoid stress corrosion.

It is important that the slope of the bottom besufficient for the mash to run out whendischarging. This type of fermentor is muchbetter suited for use with clear or semi-clearmashes than with whole cereal mashes.

Fermentor cleaning is accomplished withclean-in-place (CIP) equipment. CIP sprayheadsare high-pressure devices (100-120 psi), whichusually have automated cleaning cycles. Atypical cleaning cycle would be (minimum):

Pre-rinse with water 10 minutesDetergent circulation 20 minutesPost-rinse with water 10 minutesSterilization 10 minutes

The detergent used would be based on causticsoda, normally with added wetting agent,antifoam and de-scaling agent. The causticstrength should normally be in the 3-5% range.Ideally, the detergent should be hot (80-90°C).The detergent and rinse waters should becontinually drawn off from the fermentor toprevent accumulation of liquid during the cycle.The detergent and post-rinse should at least berecirculated and continually made up to strength.Chlorine dioxide and iodophors make idealsterilizing agents, but many distillers still usesteam to sterilize the fermentors. This is time-consuming and is probably not as effective aschemical sterilization.

It is important to control the build up ofbeerstone (calcium magnesium phosphate andcalcium oxalate). Bacteria can penetratebeerstone, which is insoluble in straight alkalinesolutions. Beerstone thus protects the bacteriafrom detergent and sterilant. EDTA-basedchelating agents added to the detergent helpdissolve beerstone and will control


accumulation. The level of chelating agentneeded can be calculated from the calcium levelin the mash. A problem recently encountered issodium contamination of process condensatewater. During CIP the caustic based detergentcan be neutralized by the CO2 in the fermentorproducing large quantities of soluble sodiumcarbonate that can pollute the processcondensate. Very high sodium levels can weakenthe yeast during fermentation and care must betaken to reduce these sodium levels.

CONTINUOUS FERMENTATION

The most successful continuous fermentationsystem used in distilling is the cascade system(Figure 9). The system illustrated in Figure 9 hasat least two fermentors and a beer well wherethe yeast are recycled. Yeast can only be recycledwhen clear mashes are used. The yeast can thenbe centrifuged, washed and reused. Typically,yeast can be washed using phosphoric acid at apH of 2.2-2.4 and a holding time of 90 minutes.These conditions are sufficient to kill the bacteriawithout doing permanent damage to the yeast.Chlorine dioxide at 40-50 ppm is an alternativeto acid washing, and seems to be gentler on theyeast.

The system in Figure 9 has two fermentors,whereas most cascade plants have five

fermentors and a pre-fermentor. The majority ofthese plants in the US use whole mash fed intothe first two fermentors. Both are aeratedcontinuously with sterile air and are cooled withexternal coolers. The pre-fermentor feeds yeastequally into both. The mash has been previouslysaccharified and enters the fermentors at a pHof 4.0 or slightly less. Usually these systems, ifyeast stress factors are taken into account,produce 13-15% beers in 24-30 hrs althoughclear mash systems operate in 10-14 hrs. Eachfermentor is individually cooled and agitatedeither with air or CO2 or mechanically. The clearmash systems are capable of maintaining a fixedyeast count although they have the potential torecycle.

Typically yeast cell counts in these fermentorsare slightly higher than in batch systems, withcounts in the range of 180-220 million cells/mlfor the first two fermentors and slightly less asthe mash proceeds through the system. Theviability and percentage of budding cellsdecrease from one fermentor to the next.Typically, the beer entering the beer well couldbe down to 120 million cells/ml with a viabilityof around 30%. Alcohol levels could be 10-11%in Fermentors 1 and 2, 12-13% in Fermentor 3and up to 15% in Fermentors 4 and 5.

The main advantages of continuousfermentation (and they are very importantadvantages) are rapid throughput and the fact

Figure 9. Twin vessel continuous stirred fermentor.

VesselNo. 2

VesselNo. 1

Agitator drive

Oxygencolumn

Oxygen in

Sterilizer

Mash in

Pump

Agitator driveFoam breaker

Beeroutlet

Yeastoutlet

CoolingCoil

CO2 outlet


that these fermentors can be run for very longperiods without stoppage. Cleaning costs aretherefore practically eliminated; and vesselutilization is superior to any batch system. Thesecontinuous systems are frequently used for 12months without stopping and are only cleanedwhen the plant has its annual shutdown.

The most serious problem encountered withthe cascade system is infection. The fermentorscan be infected through non-sterile air injectedinto the first two fermentors. In this case, aceticacid bacteria, which convert the ethanol to aceticacid, can be introduced. This can be detectedby a vinegar odor, or more accurately by HPLCanalysis. There is also a possibility of infectionby lactobacilli that usually propagate incontaminated mash coolers or saccharificationtanks.

The new distillers, at least those in the US, haveelected not to use continuous fermentation andare instead using the batch process. Batchsystems are more easily controlled, more flexibleand capable of higher ethanol levels.

Conclusions

Understanding the stress factors that affect yeastand the fermentation equipment will help usmove to the theoretical 23% ethanol that yeastsuch as Thermosacc™ are capable of producing.It is critical, however, that we push the positives:‘spoon-feed’ sugar to yeast, provide yeast withpeptides to ensure high cell numbers in earlystages of fermentation, control temperature andovercome the negatives. These negatives,infection, lactic acid, acetic acid, phytic acid andmycotoxins all prevent yeast, and therefore thedistillery, from achieving potential yields.

Today distilleries must achieve yields of 2.9gallons per bushel. Tomorrow’s distilleries willneed over 3.2 gallons per bushel. To do this,yeast must remain the central focus of all efforts.Doing so will allow 23% ethanol to soon becomethe norm.

Continuous fermentation in the fuel alcohol industry: how does technology affect yeast? 135

Chapter 11

Continuous fermentation in the fuel alcohol industry: How does thetechnology affect yeast?

W.M. INGLEDEW

Applied Microbiology and Food Science Department, University of Saskatchewan, Saskatoon,Saskatchewan, Canada

Introduction

Fuel alcohol manufacture in North America isconducted in approximately 70 (soon to be 78)production plants. The annual output from NorthAmerican fuel alcohol distilleries will soonexceed 3 billion US gallons (11 billion liters).Both continuous and batch fermentationtechnologies are used, with the breakdownbetween plants (Table 1) and on a volume basis(Figure 1) not exactly illustrating the fact thatolder and larger wet mills predominantly usecontinuous technology whereas the newer andsmaller dry mills predominantly use batchfermentation.

Table 1. Fuel alcohol plants using continuous or batchfermentation (2003).

Numbers of US plants (70 total)

Volume Continuous Batch

<11 million gal/year 3 1512-39 million gal/year 9 15>40 million gal/year 14 14

In this article, continuous fermentation, theprolonging of the logarithmic phase of yeast

�

� � �

� � �

� � �

� � �

� � � �

� � � �

� � � �

� � � �

� � � �

� � � �

Billio

ng

allo

ns

Continuous Batch

Figure 1. Volumes of alcohol made in 2003 in 70 US plants by batch and continuous technology.

136 W.M. Ingledew

growth seen in batch culture, will be the focusalong with a discussion on the technology andinformation relating to problems that yeast havein such environments.

Continuous fermentation

Continuous fermentation is said to beadvantageous because of elimination of the lagphase of yeast growth, the yeast being ‘locked’into exponential phase growth where conditionsexist for maximal ethanol formation, long termcontinuous productivity, higher volumetricproductivity, reduced labor costs while in steadystate, reduced downtime for cleaning, filling andsanitation, easier process control and savings inconstruction of smaller fermentors with higheroutput (Cyzewski and Wilkie, 1978; Kirsop,1982; Kosaric et al., 1987; Maiorella et al.,1981; Sinclair and Cantero, 1990). Highertechnological capability to operate such systemsis needed, but the primary disadvantages incontinuous culture are the problems withcontaminating bacteria and wild yeast thatdisturb the balanced nature of the fermentors,problems of maintenance of high fermentation

rates, and genetic stability of the culture used –all of these negating, in part, the advantagesmentioned above. Contamination leads to lossesin yields of ethanol and unplanned shutdownsof the fermentation trains – factors that lead tolower than expected yields of product and lessefficient conversion of sugars (from starch) toethanol (Bayrock and Ingledew, 2001b). In thefuel ethanol industry, most continuousfermentation is done by the wet milling segmentof the industry with the advantage that ethanolproduction can occur over prolonged periodsof time under optimal environmental conditions.

SINGLE STAGE CONTINUOUS FERMENTORS

This most simple form of continuousfermentation is diagrammed in Figure 2. In sucha system there is a continued flow (F) of nutrientsinto the fermentor from the medium reservoir,and a continual flow of product (P) out of afermentor to a harvest reservoir. The fermentoris designed such that the internal volume of‘medium’ remains constant.

The parameter describing this is called thedilution rate (D), measured in hr-1, calculated by

Figure 2. A schematic diagram of a continuous fermentor. The flow rate of medium into the fermentor is controlled, therebydetermining the flow of product into the harvesting reservoir (Waites et al., 2001).

Pump

Fermentor

Air in

Nutrientreservoir

Harvestreservoir

Air out

Sterile air filter

Overflow weir


dividing the medium flow rate (F) by the workingvolume of the vessel (V). The vessel is stirredto ensure continued mixing of the contents. Anadequate, constant flow of the same mediumensures that D is constant, and therefore that theyeast cell growth is constant, the substrate level(and product level into the harvest reservoir) areconstant. At that time, the vessel will bebalanced in a so-called steady state. Steady stateis governed by the level of a limiting nutrient inthe mash medium. It is normally achieved afterthree residence (replacement) times, where T =1/D. In some studies, six residence times areused to ensure steady state conditions.

When changes to dilution rate take place, theparameters above (P and S concentrations) aswell as the cell biomass all change, and time isneeded to bring the vessel back to steady state.The time required depends upon the size of thevessel(s) used and will also depend on any otherenvironmental changes or chemical or biologicalcontamination which can affect the pH, levelsof inhibitory compounds present, temperatureetc. The rate of cell growth is denoted by µ, thespecific growth rate of the yeast. Contaminantshave their own specific growth rates (µ) that, incompetition with the yeast, may result in fasteror slower growth rates than that of the yeast. Insuch a case, one microbe will wash out of thefermentor, normally leaving the other as thedominant organism. Specific alcoholproductivity in a simple continuous fermentoris ordinarily limited by ethanol inhibition andby low cell concentration, necessitating substratefeed concentrations of about 10% w/v (Kosaricet al., 1983).

MULTISTAGE CONTINUOUS CULTURE

In multistage continuous culture in the fuelalcohol industry, four or five fermentors linkedin series are normally used. In such afermentation ‘train’, each fermentor is usuallyof the same working volume, and is ideally fedfrom the one before. The first fermentor is fedby a mash stream that is under continuousproduction and cooled aseptically tofermentation temperature. Interestingly, iffermentors can be operated in balanced growth,there is no need for added yeast, no need fornutrient supplements or mash additions to thesecond to fifth fermentors and little need foranything but monitoring. In such cases, eachfermentor is in balanced growth where thesubstrate level is constant, the product level isconstant, the rate of fermentation is constant andso is the amount of yeast biomass (viable yeast)that, of course, is the catalyst of the biochemicalreaction. However, none of the fermentors in thetrain have the same steady state values asfermentors upstream or downstream. Dilutionrates, however, will be identical if the volumescontained are the same and if no additional mashor other additions are made distal of Fermentor1. Productivity on continuous processes canexceed 6 g ethanol/L fermentation mash/hour(>2.4 to 3.3-fold more than corresponding batchfermentation – Kosaric et al., 1987), and thismeans that for similar plant outputs, vessels canbe 3-fold smaller, and operating and capital costsare reduced by ~50%.

Figure 3 shows a simple engineering diagramof a continuous train. This is a train with four

Carbondioxide

Beer feedto still

Saccharifiedmash

Inoculum

Fermentors

A and B trains possible!"#$%$&#'%(")'*+,-&.

Scrubber to recover 1% of ethanol

!" !# !$ !% &''(&''()'**)'**&''()'**

Figure 3. A continuous multistage fermentation train as designed by Katzen International Inc.

138 W.M. Ingledew

fermentors and a beer well. The inoculum isadded when needed, and saccharified mash iscontinuously provided to Fermentor 1. Coolingis provided to each fermentor and an effluenttakeoff is sent to the second fermentor at thesame rate of addition as that of the mash enteringFermentor 1. Carbon dioxide leaves eachfermentor, and it is scrubbed to remove ethanol(in this case about 1%) and any dissolvedsolubles or entrained particulates. Often in suchdesigns, two identical trains are built. This givesa remarkable opportunity for any changes madein the A train to be compared to the unchangedB train – during and after a period ofequilibration to steady state.

Figure 4 is a photograph of a typically installedmultistage continuous culture. This fermentationtrain has been a model of steady statefermentation where for long periods of time,balanced growth was maintained, and thefermentor during this time did not requireadditional yeast cells to be added.

There are many possible variations in adesigned multistage continuous fermentation.Some involve a pre-fermentor, a yeastpropagator, or additions of nutrient mash streamsinto more than one fermentor. In all suchschemes, balanced S, P, or biomass is moredifficult to maintain in any vessels distal to thelast one where nutrients are added or wherechanges occur.

Another advantage in operation of a multistagetrain is that the first vessel can be used to producecells (viable biomass) that will be used totransform remaining substrate in Fermentors 2to 4 to end product. For example, if the volumeof Fermentor 1 was twice the volume of the otherproduction fermentors, the dilution rate wouldbe lower (and average medium residence timewould be higher) leading to more extensive cellgrowth. Conditions would be imposed onFermentor 1 to ensure maximum growth, whileconditions in Fermentors 2 to 4 (or 5) could bemore optimal for ethanol production. If amedium feed occurs in Fermentor 2 as well asin Fermentor 1 (called distributed feeding), themathematics at steady state for substrateutilization, product formation, dilution rate, etc.will be different and more complex. Amathematical treatment of continuousfermentation can be found in the PhD thesis ofD. Bayrock (2002).

CONTINUOUS CULTURE WITH YEASTRECYCLE

Figure 5 portrays a continuous culture systemwith provision for yeast recycle. This is the so-called Biostil fermentor system where effluentfrom the continuous fermentor is centrifuged,and the collected yeasts are re-added to the

Figure 4. The continuous fermentation train previously operated by Williams Energy in Peoria, IL.


fermentor to provide additional catalytic activityand therefore faster production of ethanol(Danielsson, 1992). The advantages of thisdesign include the continuous removal of yeastby centrifuge, an increase in yeast cell densityin the fermentor of 6 to 10-fold, and a lessersynthesis of new yeast with the concomitantsmall increase in sugar available for ethanolproduction. Unfortunately, the viability ofrecycled Biostil yeast is not good. Solidsaccumulate with the yeast and build up due torecycle (the reason why the design is more ofimportance in sugarcane juice fermentation thanin corn mash fermentations), bacteria and wildyeast increase in number and lead toaccumulation of inhibitory metabolites, andcentrifuge/separation is required. Operatingcosts and capital expenditures increase due tothe need for cell separators and reinoculation.The literature reports ethanol productivities ashigh as 50 g ethanol/L/hour in these fermentationsystems – directly due to increased yeastbiomass. Yeast is the catalyst in the reactionconverting sugar to ethanol, carbon dioxide andglycerol. It is said that control of infection insuch a system takes place due to the low pH,the low content of fermentable sugar (usuallyless than 0.1%), the high osmotic pressure, the

yeast concentration of up to 109 (1000 million)cells/mL, the 6-9% v/v ethanol present, and thefact that internal pasteurization using a heatexchanger can be accomplished. A one vesselsystem is also easier to shut down, sanitize andrestart.

Continuous cultures, regardless of design, mustoperate simply and at maximum capacity andmaximum efficiency. They must produceethanol as well as by-products of consistentquality. As said previously, the multistage systemsin use in fuel alcohol production are oftensomewhat complex in that more than onefermentor is fed, a pre-fermentor or yeastpropagator is often engineered into the system,and the presence of continued infection leads toabnormal stresses that take the fermentors outof balanced growth. Reestablishment of balancedgrowth can take well over 48 hrs, and this meansthat the plant is continually destabilized and thatthe lab and plant staff are continually challengedto find the key to stabilize conditions.

It is especially a challenge to reduce oreliminate bacteria without losing productivity(ethanol made per working volume per hr). It isnormally not possible to clean and sanitize onefermentor without later cross-contaminating it bythe others in the train. This is because all are

Steam

Yeastseparation

Substrate

+',-,*'./'(/0123/345360(1'305427138

Concentrated stillage

Stripper

Primarydistillation

Ethanol

Fermentor

CO2

Figure 5. The Biostil fermentor system (Chematur Engineering, Concept of Continuous Alcohol Production, 1993).

140 W.M. Ingledew

contaminated in the flow of the liquid throughthe train, and when Fermentor 1 is taken downfor cleaning and then restarted, it quicklybecomes contaminated from the CO2 headersthat often connect the train in a countercurrentdirection. Headers are also very difficult toclean. In addition, the mathematics andengineering theory are so complex incomparison to batch fermentation that it is noteasy to obtain adequate advice on eradicationof problems. Moreover, plants often makechanges to the parameters under which the trainoperates, and do not allow enough time to see ifthe train re-establishes productivity (steady state)before even more changes are made.

Another problem inherent in the operation ofcontinuous multistage systems is thatsaccharification of starch must be carried out atthe optimal temperature for the enzyme prior toaddition of the mash medium to Fermentor 1.The saccharification stage is ideally suited forgrowth of very heat tolerant Lactobacillus strainsknown to create the lactic acid in fuel alcoholplants that interferes with yeast metabolism.Simultaneous saccharification and fermentation(SSF) technologies were developed for batchfermentation to eliminate this very importantsource of bacteria and the resultant loss ofethanol yields. SSF is not possible in continuoussystems at this time. Moreover, in a multistagetrain, saccharification, yeast propagation, pre-fermentation, and fermentation are the normalsequence of events. This complicates the totalunderstanding of changes that occur in the trainwith time. As the section below and the chapterin this volume on water usage show, balancedgrowth is very susceptible to wastewater streamsand surges of such wastes (such as sodiumhydroxide (NaOH) in wash water, evaporatorcondensate, thin stillage, still bottoms, orbiomethanator discharge) and this, compoundedby changes in mash composition, temperature,osmotic pressure, and bioproducts ofcontaminants including acetic acid and lacticacid all can lead to perturbations of the train thatare difficult to repair. In addition, our lab hasshown that deficiencies in yeast-usable nitrogenoccur in mashes made from every conceivablegrain substrate. Such constituents vary from loadto load of received grain, making continuousoperation difficult due to the fact that yeast levelsin mash vary with nutrition. Even small changesin dilution rate can lead to washout of yeast

(insufficient time to multiply), which eventuallyleads to lowered productivity.

It is also the belief of this author thatcontinuous systems may never be able toproduce ethanol at levels over 15-16% v/v, dueto washout of cells occurring because the yeastsare continually bathed in ethanol and becausethe levels of other medium ingredients (somerecycled in backset) also increase providingcontinual, synergistic stresses that lowermetabolic rates and decrease ethanol levelsduring the residence time of the fermentors. Assubstrate levels are increased (concomitantlywith alcohol concentrations), the dilution ratesbecome smaller such that for all practicalpurposes, the fermentors approach batch moderather than remaining in a continuous mode.

It should be mentioned from an academicstandpoint that other continuous systems havebeen developed, including: internal recycle(flocculation of cells), tower fermentors,membrane reactors, vacuum fermentation(removing ethanol and carbon dioxide), solventextraction (removing ethanol), and immobilizedcell reactors. None of these have had muchimportance in manufacturing ethanol except inthe brewing industry.

Contamination and stressful conditions ina fermentation plant

The source of microbes in distilleries is not mashprior to its cooling, due to the high temperaturesused for starch gelatinization and liquefaction.All points in the alcohol process after cooking,however, can be foci for infection includingcoolers or heat exchangers, tankage, transferlines, unprocessed waters, yeast, enzymes andother additives, air, fermentor tankage or the beerwell. Viable contaminants and their end productsin recycled waters can all affect subsequentfermentations. Contaminants of concern to fuelalcohol plants mainly include the Lactobacillusbacteria and wild yeasts. It is important thatlactobacilli do not grow well in corn mash andare not able to grow at low pH. However, someare well adapted and grow exceedingly quicklyunder mash fermentation conditions. For thepurpose of this chapter, it is important to realizethat bacteria and wild yeast compete withSaccharomyces yeast for nutrients. Moreover,and more importantly, the end products of the


mash. That of the wild yeast and molds wouldapproximate 2 cells per ml. At such levels, itappears obvious that the bacteria would need tohave a very definite competitive advantage inthe fermentor (or a prolonged time) beforesignificant interference in the process wouldoccur. This is why a batch fermentation processhas little worry about contaminants in ADY. Acontinuous fermentation train could be indifficulty using direct inoculation of ADY if thebacteria, wild yeast or mold present in the ADYgrew faster in the mash provided under theenvironmental and chemical conditions (andtime of operation) of the fermentation train. Inall likelihood the contaminants from equipmentwould exceed these numbers by a considerableamount. The lowering of pH practiced in plantssuffering contamination often lowers the growthrate of the bacteria present such that they washout of the fermentation, leaving the more pHtolerant yeast to multiply and persist. However,the cost of this procedure is lowered alcoholyield.

Yeast repropagation in a fuel alcohol plant isa different story especially when continuouspropagation is practiced. Under such conditions,the very small number of bacteria mentionedabove that are present with the yeast incommercial ADY preparations are provided anopportunity to grow and compete. Becausebacteria normally have significantly shortergeneration times (times required to double),those few bacteria that can endure theconditions of the fermentor (like Lactobacillus)will compete with Saccharomyces yeasts fornutrients and eventually increase in numbers tothe extent that they begin to produce largeamounts of lactic acid and some acetic acidwhich inhibit yeast and help them to take overthe population in the propagator, and in similarfashion, the fermentors. These end products areknown to affect the growth of culture yeast andreduce alcohol output in fermentors when thesevessels become contaminated at levels over 105bacteria (100,000 ) per mL (Narendranath et al.,1997). As acids build up to over 0.1% w/v (lacticacid) and 0.05% w/v (acetic acid), we know thatthe length of the yeasts’ lag time after inoculationwill increase significantly (Figure 6) as doesyeast generation time (Narendranath andIngledew, 2001a). For these reasons, continuouspropagation is not recommended by this author.

metabolism of the bacteria are normally lacticacid and smaller amounts of acetic acid, whereasthe wild yeasts (like Brettanomyces) produceacetic acid (in addition to ethanol and carbondioxide).

It is these contaminant end products that cyclethroughout the plant due to the fact that they arenot as volatile as ethanol and mostly remain inthe thin stillage. For this reason, waste waterssuch as backset and evaporator condensate arecritical components in mash manufacturebecause they contribute increasing amounts ofchemicals that inhibit yeast growth, and thelevels of such chemicals increase with everysubsequent cycle. Many, if not most, problemsin fermentation can be alleviated if a plant caneliminate backset and/or evaporator condensate,replacing them in the mash bill with water forone or more cycles. This emphasizes the needfor excess evaporator capacity in newly designedfacilities.

CONTAMINATION OF ACTIVE DRY YEAST

This subject of entrance of bacteria into mashvia pipes, unclean tanks, or air, etc. is coveredin the chapter by Larson and Power in thisvolume and found in books discussing cleaningand sanitation. A word is needed, however, onlevels of contamination in both active dry yeastand via plant-propagated yeast in order tounderstand how fermentations can becomecontaminated in-house.

Yeast can be a source of contamination. Activedry yeasts (ADY) cannot be grown, harvestedand dried without a level of bacterial, mold andwild yeast contamination. Usual levels are givenin Table 2.

Table 2. Normal contamination in active dry yeast(organisms/g).

Bacteria Wild yeasts and molds Culture yeasts

103-105 <102 - 103 2.2 to 5 × 1010

Considering the levels of contaminants, andusual levels of inoculation of active dry yeastsinto fermentors (2.2 lb/1000 gal US), it can beestimated that the contribution of the bacteria inthe mash after a one-time inoculation of ADY isless than 26 viable bacteria per ml of inoculated

142 W.M. Ingledew

CONTAMINATION CONTROL STRATEGIES

The effects of bacterial end products when morethan one stress is present are additive, or morelikely synergistic, and affect the yeast inassociation with other stressful situations likehigh ethanol, high temperature and low pH. Theyeasts are then in serious trouble (Narendranathet al., 2001a). The mechanism of action of thestressful chemicals is now better known(Narendranath et al., 2001b). For this reason, itis necessary to reduce the numbers of themicrobes present in mash such that the levels ofbacterial end products are too small to affect ratesof ethanol production or ethanol yields. This isnormally done with low pH, antimicrobials orantibiotics.

The use of antibiotics in continuous culturehas been more of an art than a science.Unfortunately, in the past most applications ofantibiotics have been done with the same thoughtprocess that had been in use for batchfermentations. Antibiotics were, therefore, addedall at once into continuous systems with the resultthat a spike of antibiotic concentration tookplace, which decreased over time at a rateprescribed by the set dilution rate (D). In 2001,the published work by Bayrock and Ingledew(2001a; 2001b) provided alternative strategiesfor the provision of antibiotic to continuoussystems. This work was done in a bench scalemultistage continuous culture system, which wasdesigned to model the commercial fermentation

train at Williams Energy in Pekin, IL. The workwas conducted with penicillin G, an antibioticlong used in the industry. Luckily, penicillin isnot degraded enzymatically by Saccharomycesyeasts. The penicillin was therefore added inpulses considering scientifically only itsdegradation rate at the pH of the mash, itsdegradation rate at the temperatures used overthe time course of the continuous system, andthe dilution rate of the fermentor (and eachcomponent in it). It was known that in the systemused only 49.8% of the antibiotic would remainafter 3.2 hrs of incubation in mash at pH 6 at28°C, with only ~1% remaining at 21 hrs. Theregime for addition therefore included antibioticpulses every 6 hrs with an average penicillinconcentration of 2475 U/liter but with spikes oflarger values than the average such that bacteriawould be alternatively inhibited from growing -and perhaps also encouraged to initiate growthagain when concentrations were lowest. This wasfound to be more effective than the continuousaddition of the same antibiotic over time to thesame average value (Bayrock and Ingledew,2003). This technology is being translated toindustrial use. Other single antibiotics areavailable to the industry as are commercialformulations with combinations of antibioticsproviding a broader target range againstcontaminating bacteria. The latter reduce anumber of persistent infections that detract fromethanol yield.

!

" "#$ "#% "#& "#' "#!

$"

$!

%"

"#$

"#%

"#&

"#'

Acetic acid (% w/v)

Stress occurs >0.05%

()*)+,-./*0-.1)%02+3

()*)+,-./*0-.1)402+3

!

$"

" "#% "#' "#5 "#6 $

$!

%"

%!

&"

&!

"#$

"#%

"#&

"#'

Lactic acid (% w/v)

Stress occurs >0.1%

()*)+,-./*0-.1)$#402+3

()*)+,-./*0-.1)$402+3

9/8501:

'5;<

(7=

9/8501:

'5;<

(7=

µ>'(5<(

µ>'(5<(

Figure 6. Specific growth rates decrease and lag times increase exponentially as acids increase in concentration. MIC (acetic) =0.6%, MIC (lactic) = 2.5%. MIC = no growth at 72 hrs. � = specific growth rate, � = lag times; generation time = ln2/µ.


Conclusions

The economic impact of the presence ofcontaminating bacteria in fermentation mashesis profound. When bacterial numbers reach 106

(one million) to 107 (10 million) per mL, lossesin alcohol of 1-3% can be seen (Dolan, 1979;Narendranath and Ingledew, 2001a). A 3% lossin a 19.3 million gal/year (53,000 gal/day)alcohol plant is a loss of 580,000 gallons ofethanol per year! These very real economicconsiderations dictate that bacteria and fast-growing wild yeasts must be controlled incontinuous fermentation.

References

Bayrock, D.P. 2002. Application of multistagecontinuous culture to VHG-based ethanolfermentations: performance and control ofbacteria by pH and pulsed addtion of antibiotic.PhD Thesis. University of Saskatchewan.Saskatoon, SK Canada.

Bayrock, D. and W.M. Ingledew. 2001a.Changes in steady state on introduction of aLactobacillus contaminant to a continuousculture ethanol fermentation. J. Ind. Microbiol.Biotech. 27:39-45.

Bayrock, D. and W.M. Ingledew. 2001b.Application of multistage continuousfermentation for production of fuel alcohol byvery-high-gravity fermentation technology. J.Ind. Microbiol. Biotech. 27:87-93.

Bayrock, D., K.C. Thomas and W.M. Ingledew.2003. Control of Lactobacillus contaminantsin continuous fuel ethanol fermentations byconstant or pulsed addition of penicillin G.Appl. Microbiol. Biotech. (in press). On lineMay 13, 2003.

Cysewski G.R. and C.W. Wilkie. 1978. Processdesign and economic studies of alternativefermentation methods for the production ofethanol. Biotech. Bioeng. 20:1421-1430.

Danielsson, M.A. 1992. Continuousfermentation of concentrated feed stocks in asingle fermentor. Fuel Alcohol Workshop,Wichita, KS.

Dolan, T.C.S. 1979. Bacteria in whiskyproduction. The Brewer. 65:60-64.

Kirsop, B. 1982. Developments in beerfermentation. Topics Enz. Ferment. Biotechnol.6:79-131.

Kosaric, N., A. Wieczorek, C.P. Cosentino andR.J. Magee. 1983. Ethanol fermentation. In:Biotechnology Vol. 3. (H.-J. Rehm and G. Reed,eds). Verlag Chemie, Weinheim. pp 257-385.

Kosaric, N., Z. Duvnjak, A. Farkas, H. Sahm,O. Goebel and D. Mayer. 1987. Ethanol. In:Ullmann’s Encyclopedia of IndustrialChemistry. 5th Edition. Vol A9. (W. Gerhartz,ed), VCH Publishers, N.Y. pp 587-653.

Maiorella, B., C.R. Wilkie, and H.W. Blanch.1981. Alcohol production and recovery. Adv.Biochem. Eng. 20:13-92.

Narendranath, N.V., S.H. Hynes, K.C. Thomasand W. M. Ingledew. 1997. Effects oflactobacilli on yeast-catalyzed ethanolfermentations. Appl. Environ. Microbiol.63:4158-4163.

Narendranath, N.V., K.C. Thomas and W.M.Ingledew. 2001a. Effects of acetic acid andlactic acid on the growth of Saccharomycescerevisiae in a minimal medium. J. Indust.Microbiol. & Biotech. 26:171-177.

Narendranath, N.V., K.C. Thomas and W.M.Ingledew. 2001b. Acetic and lactic acidinhibition of growth of Saccharomycescerevisiae by different mechanisms. J. Am.Soc. Brew. Chem. 59:187-194.

Sinclair, C.G. and D. Cantero. 1990.Fermentation modeling. In: Fermentation: APractical Approach. (B. McNeil and L.M.Harvey, eds). Oxford University Press, LondonUK. pp. 65-112.

Waites, M.J., N.L. Morgan, J.S. Rockey and G.Higton. 2001. Industrial Microbiology: anIntroduction. Blackwell Science Ltd. Oxford,UK.

Understanding near infrared spectroscopy and its applications in the distillery 145

Chapter 12

Understanding near infrared spectroscopy and its applications in thedistillery

DON LIVERMORE1, QIAN WANG2 AND RICHARD S. JACKSON3

1Research Scientist, Hiram Walker & Sons Ltd., Walkerville, Ontario, Canada2NIR Product Manager, Bruker Optics Inc., Billerica, Massachusetts, USA3Applications Manager, Bruker Optics Inc., Billerica, Massachusetts, USA

Introduction

Near-infrared (NIR) spectroscopy has been areliable technique in the Hiram Walker & Sonsdistillery quality assurance laboratory since1997. It has proven to be powerful tool that canbenefit distilleries, breweries, or other producersof other types of ethanol in quality assuranceprograms and research facilities. As a researchtool NIR spectroscopy can be used to evaluateraw materials, yeasts, enzymes, nutritionalsupplements, and production parameters tooptimize conditions for a production plant. As aquality assurance tool it can assist productionby monitoring and maintaining control ofprocesses. If failures are not quickly identifiedand a process becomes out of control, it can costvaluable time, product, and resources. NIRspectroscopy can help eliminate these problems.

What makes NIR technology so attractive tothe ethanol industry as a quality assurance andresearch tool is that it can measure organicsubstances very quickly (5-10 seconds) andconsistently without the destruction of a sample.Traditional analysis can take a lab technicianseveral hours or even days to get reliable results,but with NIR spectroscopy the result is virtuallyinstantaneous and does not use hazardouschemicals. The areas of production where rapidand reliable analysis are most beneficial includemonitoring of incoming grains, profiling

fermentors, distillate analysis, dry houseoperations, and finished product analysis. Theexact locations where NIR spectroscopy is usedin the Hiram Walker & Sons distillery, and theadvantages and disadvantages of the technology,will be discussed in more detail. Initiallybackground material for NIR spectroscopy willbe discussed, to help the reader understand whatit takes to implement NIR spectroscopy in aquality assurance or research program.

History of NIR spectroscopy

Infrared spectroscopy has been a well-established research tool in academia and inindustry for several decades. The firstexperiments conducted with NIR light were byHerschel in 1800, in which he used a glass prismto demonstrate that there was radiation beyondvisible light (Herschel, 1800). NIR spectra arecomplex however, and only recently havecomputers become powerful enough to analyzethe spectra of complex samples. NIR technologyis now also a choice for process monitoringequipment, because there is no samplepreparation required and fiber optics can be usedto transmit the light over significant distances.This has allowed companies in the food and

146 D. Livermore, Q. Wang and R.S. Jackson

10-10

regio

nInteractio

n

Wavelength (m)

10-9

10-8

10-7

10-6

10-5

10-4

10-3

10-2

10-1 1

101

108

12,500 cm1 (800 nm) 4,000 cm

1 (2500 nm)

107

106

105

104

103

102

101

1 10-1

10-2

10-3

γ-Ray

Nudea

Transitions

Inner

shell

electronic

transitions

X-Ray Ultraviolet

Infrared MicrowaveRadio, TV

Waves

Spin

orientation in

magnetic

field

Molecular

rotations

Molecular

vibrations

Valance

electron

transitions

NIR MIR FIR ESR NMR

Vis

ible

Wavelength (cm-4

)

Figure 1. Location of the NIR region of the electromagnetic spectrum in relation to other wavelengths.

beverage industry such as ethanol producers, tomove from low technology to high technologyproduction methods.

NIR background

A detailed discussion of the theory of NIRspectroscopy is beyond the scope of this chapter,but a review of the basics may help the readerunderstand why NIR is a powerful tool inindustry today.

Light consists of electromagnetic waves ofvarying wavelengths (Figure 1). Our eyes canonly sense light at particular wavelengths, calledthe visible spectrum, but the electromagneticspectrum covers a much wider range. The near-infrared region of the spectrum extends from 800to 2500 nm (12500 to 4000 cm-1), lying betweenthe mid-infrared (MIR) region at longerwavelengths and the visible region (VIS) atshorter wavelengths.

When light impinges on matter, eachwavelength present will be selectively absorbed,scattered, reflected, or transmitted. Absorptionis the transition of a molecule from a lowerenergy level to a higher energy level with thetransfer of energy from the light to the molecule.Scattering is the redirection of light due to theinteraction with matter. Scattering may or maynot occur with a transfer of energy and thewavelength of the scattered light may or may

not be the same wavelength as that of theincident light. Reflected light is similar toscattered light, but the wavelength of reflectedlight is always the same as that of the incidentlight. Scattering and reflection of a particle aredependent on the relation between thewavelength of the light and the dimensions andorientation of the particle. Transmitted light isnon-redirected light, which if it interacts withmatter maintains the same wavelength afterinteraction.

Absorption of infrared light by a molecule isdue to interaction of the electromagneticradiation with the vibration of bonds betweenatoms. To absorb infrared light there must be achange in dipole moment during the vibration,so homonuclear diatomic molecules such as H2,O2, and N2 cannot be monitored using infraredspectroscopy. In the mid-infrared light isabsorbed if the frequency of the light is the sameas the fundamental frequency of vibration of themolecular bond. Molecular vibrations areslightly anharmonic, however, and consequentlyhigher frequency light in the NIR can also beabsorbed if its frequency is the same as that ofone of the harmonics of the fundamental. It isalso possible that vibrations can interact andcouple, if the vibrating bonds are joined to asingle, central atom. The absorption bands inthe NIR are thus referred to as either overtonebands, with frequencies of 2, 3 or more timesthe fundamental, or combination bands with


0.0

0.5

1.0

1.5

2.0

2.5

10000 9000 8000 7000 6000 5000 4000

Wavenumber (cm-1)

Ab

sorb

ance

un

its

Water

Ethanol

O-H

C-H

C-H

O-H

Figure 2. Spectra of water and pure ethanol in the NIR.

frequencies that are the sum of the frequenciesof two fundamentals. The NIR absorptionmechanisms are much weaker that those in themid-infrared, and therefore the absorption bandsare also much weaker. The strength of theabsorption will decrease by a factor of 10–100for each step from the fundamental to the nextovertone. Typical pathlengths (the distance thatthe light travels through the material) in the NIRare therefore 1-10 mm, compared with 10-100 µm in the mid-infrared.

For overtone or combination bands to occurin the NIR, the frequencies of the fundamentalvibrations must be high enough, otherwise thesebands occur at wavelengths in the mid-infrared.This is only true for bonds involving hydrogen,for example C-H, O-H, N-H, and S-H. Theexception to this is the carbonyl vibration, C=O,which is strong enough in the mid-infrared thatthe second overtone can be seen in the NIR.

The absorption of light as a function ofwavenumber (or wavelength) is known as aspectrum . An infrared spectrum typicallycontains a large number of peaks, which in theNIR are typically broad and overlapped.Example spectra of water and ethanol are shownin Figure 2. The y-axis indicates absorbance,which is the negative log of the fraction of lighttransmitted by the sample.

NIR instrumentation

TYPES OF NIR SPECTROMETERS

To successfully implement NIR spectroscopy itis essential to have the right instrument. Themost important factors will depend on theintended use for the instrument, but major onesthat should be considered are the flexibility ofthe instrument when changing from liquid to solidanalysis, the user interface and softwareprovided, the sampling method, the cost of theinstrument, instrument specifications, andsupport from the manufacturer.

Filter-based photometers

There are a number of different types of NIRspectrometer. The simplest and cheapest NIRinstruments are filter photometers. Photometersare distinguished from the other types ofspectrometer that will be considered, which areall spectrophotometers, because they do notproduce a spectrum as shown in Figure 2. Aphotometer uses filters mounted in a rotatingwheel to select small ranges of wavelengths inthe spectrum. The filter wavelengths are chosendepending on the desired analysis, for exampleprotein, moisture, and oil. Figure 3 shows howa photometer works.


Source Sample Detector

Filter Wheel

Scanninggrating

Source Sample Slit DetectorSlit

Figure 3. Schematic representation of filter photometer operation.

Figure 4. Schematic representation of the principles of a scanning dispersive spectrophotometer.

Filter photometers lack the flexibility of otherinstrument types, and can be prone to errors ifthe temperature changes. They have theadvantage of low cost, however, and can beuseful for dedicated analysis either in thelaboratory or on-line in a production facility.

Scanning dispersive spectrophotometers

In this type of spectrometer broadband light isdirected to the sample, and the transmitted orreflected light is then passed through a narrowslit. A diffraction grating then disperses the lightinto separate wavelengths, which are scannedacross an exit slit by moving the grating. Thediscrete wavelengths that pass through the exitslit are sequentially measured by the detector.Figure 4 shows the principle of operation of ascanning dispersive spectrometer.

Like the other types of spectrophotometer,these instruments are flexible and offer a fullrange of analytical techniques. Some high-endinstruments with double beams are also capableof very high photometric precision. The majordisadvantage is that the precision and accuracyof the wavelength calibration are limited by

mechanical precision. This can affect instrumentstability, existing calibration models, and maketransfer of a calibration model to a newinstrument difficult.

Detector array dispersive spectrophotometers

In a detector array dispersive spectrometer thescanning grating, exit slit, and detector in Figure4 are replaced by a stationary grating and adetector array. Since different wavelengths fallon different detector elements they are measuredalmost simultaneously, so detector arraydispersive spectrometers can be very fast. Thelimited number of elements in the array,however, means that this type of spectrometeris usually either low resolution, or only covers alimited portion of the spectrum. With no movingparts, these instruments are also very rugged.

Acousto-optical tunable filter (AOTF)spectrophotometers

In an AOTF spectrometer the grating in Figure 4is replaced with an acousto-optic filter. This is acrystal that acts as a diffraction grating when


Source

Detector

Movingmirror

Fixedmirror

Beamsplitter

Laser

S

Figure 5. Schematic representation of an interferometer in a FT-NIR system.

excited by sound waves using a piezo-electrictransducer. The grating can be tuned to differentwavelengths by changing the frequency of thesound. These instruments are fast and rugged,but have limited resolution and signal-to-noiseratio. They are also very temperature sensitive,and therefore typically have built-inthermostatting.

Fourier transform (FT) NIR spectrophotometers

A schematic representation of a Fouriertransform instrument is shown in Figure 5. Thebroadband light source is directed to aMichelson interferometer. The interferometerconsists of a beamsplitter, a fixed mirror, and amirror that moves back and forth very precisely.The beamsplitter reflects half of the light to thefixed mirror, and transmits the other half to themoving mirror. Both the fixed and movingmirror will direct the light back to thebeamsplitter, where they interfere. Thisinterference changes as the moving mirror isdisplaced, and therefore the intensity of the lightat the detector changes. The intensity as afunction of mirror displacement is called aninterferogram, which must then be Fouriertransformed to obtain the spectrum. A detaileddescription of the operation of Fourier transform

spectrometers can be found in Griffiths et al.(1986).

Fourier transform spectrometers offer highresolution, good speed, and high signal-to-noiseratios. Their biggest advantage, however, stemsfrom the fact that the position of the movingmirror is controlled using a HeNe laser. Theinherent wavelength stability of the laser resultsin very high wavelength accuracy and precision.This in turn means a calibration model is verystable over time, and permits easy transfer ofcalibration models between instruments.

FT-NIR has been the instrument of choice inthe Hiram Walker & Sons laboratory because ofthese advantages, especially the wavelengthaccuracy and precision. In the alcohol industrymeasuring spirit strength is very important toregulatory agencies. If NIR were to be approvedby governments as an official method of alcoholdetermination, then the wavelength stability ofFT-NIR systems might be advantageous becauseof the resulting model stability. Currentlyavailable systems also offer a full range ofanalytical techniques (i.e. solids, pastes,powders, liquids, probes, vial holders,integrating spheres) on a single instrument, witheasy computer controlled switching betweendifferent modes.


Source Sample

ltranslo

Detector

Figure 6. Schematic representation of a transmission measurement.

Sample presentation

There is a wide range of possibilities in thealcohol industry for NIR analysis. There arethree main categories of analysis. The first ofthese is off-line measurement, where samplematerials are selected from production to beanalyzed by an instrument in a laboratory setting.This type of sampling places the least demandon the operator developing the calibrationmodels. The second category is at-line sampling,where the instrument is placed close to theoperation in order to decrease analysis time.Depending on the environment, the instrumentmay need to be in an enclosure. Finally, thereare sampling accessories that can be integratedinto the production line without having to collectsamples by hand. This is called in-line analysis.This type of analysis allows the user to integratethe spectrophotometer with existing controlsystems. In each of these categories there are anumber of sampling accessories available, whichmust be carefully chosen. There are three typesof measuring techniques that the accessories canuse: transmission, diffuse reflection, andtransflection.

TRANSMISSION

In transmission measurements light is directedat a sample, where some of the light is absorbedand some is transmitted to the detector. This typeof measurement is used for samples that areliquids or transparent foils (Figure 6).

The main accessories used for transmissionmeasurements are a vial holder or a liquid probe.The benefits of vial holders include atemperature-regulated block, and custom-sizeddisposable glass vials. The fiber optic-based

liquid probe can be used for either off-line, at-line, or in-line measurements. Light from thespectrophotometer is directed through fiber opticcable to a pair of mirrors that direct the lightthrough a cavity in the probe head. The cavityhas a fixed pathlength that enables a liquidsample to enter (Figure 7). Figure 8 shows acommercial probe.

Figure 7. Schematic diagram of a liquid probe, showing thepath of the light.

Figure 8. A commercial hand-held liquid probe for off-lineor at-line analysis.


Detector

Reflected IR Energy

Output Fiber Bundle

Input Fiber Bundle

IR Energy

IR Source

Sample

Figure 9. Schematic diagram of a diffuse reflectance probe.

DIFFUSE REFLECTION

When light is reflected from rough surfaces orpowders, it is referred to as diffuse reflection.Depending on the sample, light may penetratebeyond the surface a significant distance. Thereis therefore the potential of quantifying thecomponents within the sample. There are twotypes of accessories that measure by diffusereflection: the diffuse reflectance probe, or theintegrating sphere.

The diffuse reflectance probe uses a fiber-opticcable with multiple fibers. Half of the fibers inthe bundle are typically used to transmit the lightto the sample, and half to return the reflectedlight to the spectrometer. A schematic diagramof a diffuse reflectance probe is shown inFigure 9.

A second type of diffuse reflection accessoryis an integrating sphere. In an integrating spherelight is directed onto a sample, as shown inFigure 10. The reflected light is measured by adetector, which is mounted in the wall of the

gold-plated sphere. Typical samples measuredby the integrating sphere are distillers driedgrains with solubles (DDGS), whole kernel corn,and corn mash. An integrating sphere is wellsuited to these types of inhomogeneous samplebecause of the large sampling area. If the sampleis very inhomogeneous, as is the case with cornand DDGS, a rotating cup, as shown in Figure11, can be placed on the integrating sphere toprovide further averaging. In the case of cornmash, an electrically powered mixer is neededto stir the sample so the insolubles will not fallto the bottom.

TRANSFLECTION

Transflection is an extension of the reflectiontechnique, as shown in Figure 12. When a mirroris placed behind the sample, the light transmittedthrough the sample is reflected back through thesample and into the diffuse reflectance probe orintegrating sphere. Transflection thus measures


Figure 10. Schematic representation of an integrating sphere.

Figure 11. A rotating cup full of whole kernel corn on an integrating sphere.

Source

Detector Mirror

l trans

l o

Analyte

Figure 12. Schematic diagram of transflection.

S

D

Sample

Gold sphere

Detector

NIR beam


a combination of transmission and reflection.This technique is useful for emulsions, gels, andliquids.

Calibration development

It is very important to acquire the properinstrument and sampling accessories as it directlyaffects the next stage involving NIR technology,which is calibration development. When usinga NIR spectrometer, a large amount of time isinitially spent turning the raw spectra into usefulinformation. The ultimate goal is to be able topredict component concentrations in a matter ofseconds or minutes, so the time spentperforming wet chemical analysis of thecalibration samples is well worth the effort.

THEORETICAL BASICS

When infrared radiation passes through a sample,some of the light is transmitted or reflected,without interacting with the molecules, and theremainder interacts with the molecules and isabsorbed. The wavelengths at which theabsorption will occur and how much light willbe absorbed depend on a given substance. Thisis the basis of qualitative and quantitativeanalysis by spectroscopy, including NIRanalysis. The amount of absorbed light ismeasured in absorbance. The absorbance,generally denoted as A, is defined as:

A = -log (Is/I0)

Where I0 is the light intensity before passingthrough a sample

Is is the light intensity after passingthough the sample

Spectroscopic quantitative analysis is based onthe Beer-Lambert law (also know as Beer’s law),which states that there is a linear relationshipbetween the absorbance and concentration ofabsorbing molecules. Mathematically, Beer’s lawcan be written as:

A = ε C D

where A = Absorbance at a specific wavelengthε = absorption coefficient at a

given wavelengthC = concentration of absorbing

moleculesD = pathlength

Since the pathlength D is often fixed, in practicethe absorption coefficient and the pathlength arecombined into one coefficient k. Equation (1)can be then simplified to:

A= kC (Equation 1)

Solid or slurry samples are often measured indiffuse reflectance mode, using the accessoriesshown in Figures 9 and 10. For reflectancemeasurements log (1/R) is used instead ofabsorbance, where R denotes reflectance. Thisis defined as:

log (1/R) = -log (Is/I0)

Where I0 is the light intensity measured froma reference material

Is is the light intensity measured fromthe sample

The reference material used in the NIR region isusually a material with uniform scatteringproperties and very little absorption over theentire NIR region. Two kinds of material are usedcommercially. One is called Spectralon®, amaterial made of a special type of Teflon®; theother one is diffuse gold, a piece of metal withfinely roughened surface coated with gold.

There are many theories regarding quantitativeanalysis in diffuse reflectance measurements.Empirically, log (1/R) is found to be fairly linearwith the concentrations of chemical componentsin the NIR, hence,

log (1/R)= kC (Equation 2)

The similarities between equations 1 and 2 areapparent.

Quantitative analysis generally has two steps:calibration and prediction. During the calibrationstage, NIR spectra of a set of samples will firstbe measured so that the absorbance values in


Equation (1) can be determined. The concen-trations of the chemical components can beanalyzed with traditional lab methods such aswet chemical or HPLC methods. Once theabsorbance values A and concentrations C areknown, the values of the coefficients k can bedetermined by the least squares method. Oncethe coefficients are determined, the spectrometeris calibrated and ready to predict theconcentrations of unknown samples. Equation1 only works in a simple system where theabsorption at a particular wavelength comes fromonly one component. Such a method is oftencalled univariate calibration.

As described earlier, NIR spectra are highlyoverlapped, meaning that absorbance at aparticular wavelength contains contributionsfrom multiple chemical components in thesample. The univariate calibration can thereforerarely be applied to NIR applications. A moresophisticated multivariate analysis is usuallyrequired to build the calibration model. Whenmultiple components are present, the totalabsorbance at a given wavelength is thesummation of absorbances of each individualcomponent. The absorbance at a givenwavelength can then be expressed as follows:

A = ε1 C1 D + ε2 C2 D + …+ εn Cn D(2)

where

A = Total absorbance at a specific wavelengthε 1 = Absorption coefficient of component 1C1 = Concentration of absorbing component 1D = path lengthε 2 = Absorption coefficient of component 2C2 = Concentration of absorbing component 2ε n = Absorption coefficient of the nth

componentCn = Concentration of absorbing component n

There are four multivariate methods used to buildthe calibration model, Classical Least Squares(CLS), Multiple Linear Regression (MLR),Principle Component Regression (PCR) andPartial Least Squares (PLS). How these methodswork is beyond the scope of this chapter, butthere are excellent review articles describing howall these methods work as well as the pros andcons of each (Haaland, 1988; Martens et al.,

1989). PLS is the most commonly used methodfor building multivariate quantitative models andis the method used in all data analysis presentedin this chapter. This chapter will only deal withpractical aspects of development of calibrationmodels and how to interpret commonterminology used in the PLS method.

CALIBRATION SAMPLE SET, NIRMEASUREMENTS AND REFERENCE ANALYSIS

To build a quantitative calibration model usingNIR data, the user must have a reference analysismethod to determine the concentrations of thecomponents of interest in a sample. Thereference methods commonly used in QClaboratories in the distillery industry today areHPLC, GC, DMA and various wet chemicalanalyses. The accuracy and precision of a NIRcalibration model is strongly affected by thequality of the reference analysis method. A fewexamples are given in the section of this chapterdealing with NIR applications in the distillery.The reference methods should be closelyexamined and understood so that the user doesnot have unrealistic expectations of NIRcalibration models.

When building the calibration model, thereference analysis and NIR measurementsshould be performed as close to the same timeas possible if samples are not stable. Forexample, live fermented mash samples takenfrom a fermentor are unstable. The longer a samplesits between the time of NIR measurement and thetime of the reference measurement, the bigger theerror to be expected.

Preferably, samples from production that spana period of time should be incorporated in thecalibration model, with the goal of incorporatingall possible variations in the production process.One must be cognizant of production parameters,and collect unique samples when they occur. Forexample, DDGS can vary in a whisky distillerywith the many different mash bills, which caninclude grains such as rye, barley, wheat, andcorn. The NIR user must recognize when differentmash bills are being fermented and collectsamples that do not occur very often andincorporate them into a calibration model. Thismeans it may take an extended period of time toobtain a representative set of samples, but it is


important to do this because it makes acalibration model more robust, accurate, andprecise. Another occasion when a user shouldincorporate unique samples is when processingequipment breaks down. Usually these are timeswhen unique or extreme samples occur. Forexample, when a belt conveyer breaks whiledelivering corn to a pre-mix tank it can causemash to be more dilute than usual. The user cancollect the mash in these fermentors for analysis,collect the syrup in the evaporators, and collectthe DDGS at the end, because the brokenconveyer affects all of these processes. In thatway, the next time the conveyer malfunctionsthe NIR user can quickly identify the source ofthe problem. To make the calibration modelmore robust, it is recommended that any oddsamples be included at least three times in thecalibration model (Williams and Norris, 2001).

The calibration samples should also span theconcentration range as smoothly as possible,avoiding large numbers of calibration samplesat just a few concentrations. For example, referto the calibration model for crude protein inDDGS in Figures 13a and Figure 13b. In Figure13a there are not enough samples of DDGS withlow (21 to 22%) and high crude protein (27 to31%). To make a calibration model more robustand precise, the sample distribution should besimilar to that of Figure 13b. This type of sampledistribution, where samples are evenlydistributed over the calibration set, is oftendescribed as a ‘boxcar’ effect (Williams, 2001).

The calibration range should be larger than thespecified range and not too small in view of theerror of the reference method (EMEA 2003). Ifthe required accuracy is greater than 0.1% byweight, generally it can be achieved with NIR.If the required accuracy is between 0.1 to 0.01%by weight, then the NIR application is likely tobe feasible for a simple system with very fewcomponents, an accurate reference method, awide enough concentration range, and a lot ofsamples. If the required accuracy is 0.01% wt toa few ppm, the application is unlikely to befeasible.

BUILDING CALIBRATION MODELS

After the calibration samples are collected,measured by a NIR spectrometer and analyzed

by the reference method, the user can thendevelop and validate the quantitative calibrationmodels. Once the calibration model is validated,the user can finally enjoy the fruits of all thislabor: the prediction of multiple parameters froma single NIR measurement in a matter of secondsto minutes. Validation is the key to thedevelopment of PLS models. Validation is usedboth to optimize many parameters during thedevelopment of the model, and to identifyabnormal samples (outliers). There are two typesof validation: cross validation and externalvalidation or test-set validation.

External validation

External validation is sometimes referred to astest-set validation. The sample set used inexternal validation consists of samples withknown reference data that were not used in thecalibration set for the model development. Thegoal for the external validation is to test thepredictive ability of the model. At the end ofexternal validation, the NIR predicted values arecompared with those from the reference method,often expressed in a plot of NIR predicted vs.reference values. The value of R2 in this plot isone of two important indicators for how goodthe model is. The other one is the Root MeanSquare Error of Prediction (RMSEP), which isused as an indicator of the average error for NIRprediction on future unknown samples.

Cross validation

The advantage of external validation is that it isa true independent validation. The maindisadvantage of external validation is that itwastes the valuable calibration data in order tocheck the predictive ability of the calibrationmodels. When the number of availablecalibration samples is small, a procedure calledcross validation is often used. In a crossvalidation, the first calibration sample is removedfrom the calibration set. PLS models aregenerated using the remaining calibrationsamples. The PLS models are then used topredict the first sample that is left out and theerror of prediction for this sample is calculated.The first sample is then re-included into thecalibration set and the next sample is removedfrom calibration. The whole process is repeated


Crude protein in DDGS (%)

Nu

mb

er o

f sa

mp

les

Non boxcar calibration

model for protein in DDGS

20 21 22 23 24 25 26 27 28 29 30 31

0

1

2

3

4

5

6

7

8

9

10

11

Crude protein in DDGS (%)

Nu

mb

er o

f sa

mp

les

Boxcar calibration model for protein in DDGS

20

0

1

2

3

4

5

6

7

8

9

10

11

21 22 23 24 25 26 27 28 29 30 31

Figure 13a. A calibration set for protein prediction in DDGS.

Figure 13b. A calibration sample set for protein prediction in DDGS.

until the last sample is removed and analyzed.The number of samples left out can be just onesample or a small set of samples. At the end ofcross validation, the Root Mean Squared Errorof Cross Validation (RMSECV) is calculated andused as an indicator of the future predictiveability of the models.

Results of validation

As discussed earlier, the cross validation or

external validation result is often shown in a plotof NIR predicted values vs. reference methodvalues. In an ideal situation, this plot should bea 45o line, where the prediction should equalthe reference value. The closer the R2 value is to1 or 100%, the more accurate the model. Figures14a and 14b show a PLS cross validation resultand an external validation for prediction ofethanol concentrations in corn mash samples.

Errors exist in all measurements and cannotbe avoided. It is important to identify the


Rank: 5

R2

= 99.86

RMSECV = 0.139

0 1 2 3 4 5 6 7 8 9 10 11 12

-1

0

1

2

3

4

5

6

7

8

9

10

11

12

Reference method value

NIR

pre

dic

ted

val

ue

Rank: 5

R2

= 99.86

RMSEP = 0.144

0 1 2 3 4 5 6 7 8 9 10 11 12 13

0

1

2

3

4

5

6

7

8

9

10

11

12

13


NIR

pre

dic

ted

val

ue

Figure 14a. Cross validation result for ethanol prediction in fermented corn mash samples. The values of R2 and RMSECV were0.9986 and 0.139 % v/v, respectively. The number of samples in the calibration set is 111.

Figure 14b. External validation result for ethanol prediction in fermented corn mash samples. The values of R2 and RMSECVwere 0.9986 and 0.144 % v/v, respectively. The number of samples in the validation set is 40.

accuracy that is needed for a plant to keep aprocess under control. For example in proteinin DDGS, there is no need to strive for an errorof 0.05% when an error of 0.5% is satisfactoryfor plant operations; however, a 0.05% error fordetermining alcohol strength in a finished blendcan cost a company money in terms of ethanollosses and taxes.

How many samples in a calibration data set?

It is very important to include an adequate

number of samples in the calibration set. Ingeneral, one should incorporate as many as areavailable. Whether or not there are enoughsamples to build a stable calibration model canonly be determined by performing validationtests with an independent validation set. As arule of thumb, one should start with a minimumof 10 samples for each independent chemicalcomponent or other source of variation. ASTMprovides guidelines that can be used to judgewhether or not a calibration set containssufficient samples (ASTM, 1997).


Calibration model updating

When a calibration model is finished, it is goodpractice to check periodically to ensure thevalidity of the answers (Martens et al., 1989).The checks might be frequent when a calibrationmodel is first used, but they will become moreinfrequent as the model becomes more robustover time. These routine checks are not onlyimportant in the maintenance of the calibrationmodels, but also help build confidence inoperators, technicians, and management, forwhom NIR might be relatively new technology.

Distillery applications for NIR analysis

Distillery processes are ideal situations for NIRanalysis since most compounds in a distilleryare organic and at sufficiently highconcentration. From the beginning, when grainis unloaded from the truck, to the end whereproduct goes into a bottle, NIR analysis canprovide useful information to maintain processcontrol.

INCOMING GRAIN

Unlike fuel ethanol production where corn is100% of the mash bill, in a Canadian whiskydistillery corn is approximately 90% of the mashbill. It is the largest expense to a distillery, yet itgoes largely unmonitored throughout thedistilling industry. The components ofimportance to distillers are moisture, protein, oil,and starch, all of which can be measured by NIRspectroscopy.

Moisture content of incoming grain is of criticalimportance. If moisture becomes too high it cancause mechanical problems with a hammer milland cause downtime in the distillery. Typically,a maximum specification of 15% moistureshould be targeted. If moisture levels are above15%, there is also less starch available to makealcohol. It is up to the distiller to either makeconcessions with the corn broker or to reject theload of grain. This type of monitoring can justifythe expense of a NIR spectrometer.

Protein content of whole kernel corn is anotherparameter that can be analyzed, because it affectscomposition of DDGS, a valuable by-productat the end of distillation. For example, in 1998 a

distillery was having problems meeting theguarantee for protein content in DDGS. At first,it was thought there were issues withfermentations not finishing properly; howeveron investigation it was found that fermentorswere finishing with a good yield. After furtherinvestigation of the NIR crude protein data, itwas found that the average protein level in theincoming corn fell from 7.5–8.0% in 1997 to7.0% in 1998 due to drought conditions. It wasfound that the expected crude protein contentin DDGS could be roughly determined bymultiplying the protein content of incoming cornby 3. A difference of 0.5–1.0% in protein contentof incoming grain will therefore affect theprotein content of DDGS by 1.5–3%. There isnothing a distillery can do to change theincoming crude protein levels of poor qualitycorn. The only option is to lower the crudeprotein guarantee of DDGS.

Oil content can also be monitored by NIRspectroscopy. The reason that it may beimportant to measure oil is that genetic varietiesof corn with high oil contents exist in thecommercial market. Measuring incoming loadsfor oil allows the distiller to prevent high oil cornvarieties from entering the system.

Starch is the component with the greatest effecton overall yield because it is the source offermentable sugar for yeast growth andultimately alcohol production. NIR spectroscopycan allow a distiller to track the starch levels inpurchased corn.

There are several advantages to using NIRspectroscopy as a primary tool for determiningthe quality of corn:

1. The first, and most obvious is the rapidanalysis of essential components. It can takeup to three days to determine the levels ofstarch, protein, moisture and oil in one cornsample by traditional analysis methods.

2. The distiller can track corn suppliers moreclosely to prevent becoming the victim offraud. The way samples are collected isimportant, because it can affect measurementresults. The suggested method is to gather acomposite sample from a truck with a probethat can measure the bottom, middle, and topof the load. Also, it is important to measurethe front, middle, and back end of the truck.


A sample should always be collected by thisconsistent and statistically reliable method.Sampling with this method can also detectloads containing grain that is out ofspecification on the bottom that have been‘topped up’ with good quality grain.

3. Production efficiencies can be tracked moreclosely. Accurate determination ofproduction efficiency is important, becauseit helps the distiller make improvements andreduce costs.

4. More crop data accumulated each year.

5. Variety selection for desired characteristicsis possible.

The calibration model Hiram Walker & Sonscurrently uses to monitor incoming corn wasprovided by Bruker Optics with the purchase ofthe NIR spectrometer. The expected value ofeach of the four major components and thepredictive error of each component expressedin RMSEP are listed in Table 1.

Table 1. Expected values and predicted errors for starch,oil, protein and oil content of incoming corn expressed inRMSEP.

Component Expected result Approximate RMSEP(% as is basis) (+/- %)

Starch 60–62 1.2Oil 2–3 0.5Protein 7-8 0.5Moisture 12-15 0.3

A major advantage of using an FT-NIR systemis that it has high wavelength accuracy andprecision, which allows calibration models to beeasily transferred from instrument to instrumentvia the internet, without having to adjust forslope or bias on each instrument. As a precaution,the downloaded models are validated with a setof 40 corn samples that are not from the originalcalibration set. The predicted error of the 40validation samples measured by the instrumentin the Hiram Walker & Sons plant was roughlythe same as the error in the original model, whichmet requirements.

Another advantage of this calibration modelis that reliable results can be obtained by

measuring incoming corn directly without anysample preparation such as grinding. Since theabsorption coefficient in the NIR region is low,NIR light can penetrate corn kernels and wholekernels can therefore be used for modeldevelopment. The NIR spectrum can becollected with the aid of an appropriate samplingdevice such as a rotating cup or a flow-throughfunnel. Depending on the available samplingdevices and how a particular calibration modelwas developed, some instruments today stillrequire users to grind incoming grain in orderto achieve a reliable measurement.

Calibration models for other grains includingrye, rye malt, barley, and barley malt arecurrently being developed. The components ofinterest include starch, oil, protein, moisture, andenzymatic activity.

FERMENTATION MONITORING

Fermentation optimization can be complex,because many parameters can affect the finalalcohol content and overall yield. With thetraditional analyses available today it can be timeconsuming for laboratory staff to keep up withthe demand for fermentation analysis. Optimumefficiency can be achieved if all parameters arekept in control prior to and at fermentation, butthis does not always happen as equipmentfailures can occur hours before being identifiedand corrected. NIR technology can identify non-optimum conditions in the fermentation almostimmediately, allowing correction of problems.In addition, NIR analysis can aid distillers inevaluating the new enzyme and yeast nutritionalsupplements.

In the industry today, most distilleries use anHPLC for fermentation monitoring, includingmeasurements of sugars, acids, glycerol, andalcohol. All these components are interrelatedand it is very important to keep track of them ateach stage of fermentation, but there aredisadvantages to using HPLC as the primarymonitoring tool:

1. The technician must be highly skilled.

2. Accessories can be quite expensive (i.e.columns, chemicals, filters).

3. The amount of time needed to determine the


correct answers is at least 20 minutes. Thisdoes not include repeat samples, samplepreparation time, technician training time,instrument setup time, wash outs, andcalibration time. This can be frustrating tofermentor operators when quick decisionsare required.

4. The sample preparation for HPLC onlypermits analysis of those components thatare soluble in the liquid fraction. To extractthe components of interest from the mash, itmust be centrifuged or filtered to retain theliquid portion. This will compound the errorbecause the solid portion that is removed hassugars, acids, and alcohols still bound to thesolid particles, thus skewing results.

5. Results are not always as accurate asrequired.

The amount of ethanol produced in fermentationis very important, and the potentialimprovements in fermentation efficiency thatNIR technology can support should not beunderestimated. For example, if a fermentor is200,000 liters and the average fermentationfinishes when alcohol content reaches 9.6%, thismeans that the plant will produce 19,200 litersof absolute alcohol per fermentor. Afterresearching different protocols to optimize

fermentations such as changing enzymes,process parameters, or nutritional supplements,the alcohol content was raised to an average of10.6% per fermentor. This means that 21,200liters of absolute alcohol are now made in eachfermentor. If 1000 fermentors are set per year ina plant to produce a certain amount of alcohol,then one percentage point increase in fermentoralcohol content means that plant can now setonly 905 instead of 1000 fermentors to producethe same volume of alcohol. This meansconsiderable savings in raw materials,processing fuel, steam, labor, maintenance, andequipment. Also, plants strive to be energyefficient, and generating steam can be quiteexpensive. Overall, this can mean millions ofdollars saved annually for a distillery. Figure 15shows the results of a number of fermentationoptimizations monitored by NIR spectroscopyat Hiram Walker & Sons distillery over a four-year period. NIR has proven invaluable inassisting in fermentor optimization and in savingenergy costs.

It is clear from the example above that raisingthe percent alcohol in a fermentor by even 0.1%can increase efficiency and contribute to costsavings for a distillery. There are two referencemethods that can be used to calibrate the NIRmodel, HPLC or distillation - DMA. The NIRmodel should be based on the most accurate

Hours of fermentation

Eth

ano

l (%

ab

v)

0 15 30 45 60 75

2003

90 105 120

0

1

2

3

4

5

6

7

8

9

10

11

12

1999

Figure 15. Fermentation profiles from 1999 to 2003 during which period NIR has assisted in optimizing fermentors and thusimproved the distillery bottom line.


reference data available in order to determinewhen an enzyme, nutritional supplement orother change in fermentation conditions iseffective.

Figures 16a and b show the two calibrationmodels for determining ethanol in corn mash.The first NIR model was built using HPLC asthe reference method (Figure 16a) and the secondwas built using the distillation – DMA referencemethod (Figure 16b). The distillation – DMAmethod is to distill 100 mL of the corn mash,collect 100 mL of the distillate and thendetermine the percent alcohol via the DMA. The

value of RMSECV for the NIR models using theHPLC and distillation-DMA method is +/- 0.67%and 0.18%, respectively. This demonstrates thegreater accuracy of the distillation-DMA methodfor the calibration of NIR models for corn mash.It is recommended that the distillation-DMAprocedure be used for ethanol calibrations incorn mash instead of the HPLC method.

In contrast to alcohol, for sugar calibrationmodels HPLC is recommended as the referencemethod. Even though the prediction error forsugar is roughly +/- 0.5%, it is still acceptable.The only concern to a distiller is that sugar is

Rank: 5

R2

= 96.2

RMSECV = 0.674

0

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

1 2 3 4 5 6 7 8 9 10 11 12 13 14


NIR

pre

dic

ted

val

ue

Rank: 6

R2

= 99.73

RMSECV = 0.176

0 1 2 3 4 5 6 7 8 9 10 11 12

0

1

2

3

4

5

6

7

8

9

10

11

12


NIR

pre

dic

ted

val

ue

Figure 16a. Cross validation result for ethanol content in 30 fermented corn mash samples. The reference method of analysis wasHPLC. The values of R2 and RMSECV were 0.962 and 0.67% v/v, respectively.

Figure 16b. Represents the cross validation result for ethanol content in 30 fermented corn mash samples. The reference methodof analysis was by distillation - DMA. The values of R2 and RMSECV were 0.997 and 0.18% v/v, respectively.


being converted and is used up by the yeast. Ifsugar values are not decreasing then thefermentation parameters must be quicklyinvestigated and proper actions taken to correctthe problem. Figure 17 shows the carbohydrateconcentration over a 140 hr period, monitoredusing NIR spectroscopy.

Another important parameter in cornfermentation that NIR spectroscopy can monitoris the production of lactic acid. The primarymethod of calibrating the NIR for lactic acid isHPLC. Lactic acid is a compound produced byLactobacillus sp., bacteria which compete withyeast for sugar. The more lactic acid produced,the more bacterial contamination is present.Figure 18 illustrates the amount of lactic acidproduced over time in a typical batchfermentation process. It was noted that as thebackset stillage use rate increased in thefermentor, so did the starting level of lactic acid.By monitoring lactic acid, a distiller can thereforeadjust backset stillage rates in order to optimizefermentations. For continuous fermentations, thelactic acid values can be used to determine whenantibacterial products are required or wash outprocedures should commence.

A potential application for NIR technology isoptimization of fermentor cleaning. One can

optimize the amount of wash water anddetergent necessary for clean fermentations bycorrelating the wash out procedure with thelevels of lactic acid produced at the end offermentation, thus determining which wash outprocedure is most cost effective.

NIR spectroscopy is a flexible analytical toolbecause its results are based on the best primarymethods available, distillation – DMA forethanol, and HPLC for sugars and lactic acid.The predictive errors of the calibration modelscurrently used to monitor corn fermentations atthe Hiram Walker & Sons plant are shown inTable 2.

Table 2. RMESCV values for fermentation parameters.

Component Approximate RMESCV

Ethanol 0.14Dextrins 0.50Dextrose 0.46Maltose 0.52Lactic acid 0.11Glycerol 0.07

Troubleshooting is one of the largest paybacksfor an NIR spectrometer. Over the course of one

Figure 17. Typical reduction in carbohydrates in corn fermentations over a 140 hr period.

0.0

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0


Dextrose

Maltose

Dextrins

Car

bo

hyd

rate

(%

)


year, one lab technician can measure 5600 cornmash samples with just one NIR spectrometer.There are 30 batch fermentors at the HiramWalker & Sons facility, and all are kept full ofmash until the mash is sent to the beer well.Today Hiram Walker & Sons laboratory monitorsthe fermentors at 12–40 hr intervals and canpredict the final percent alcohol. If ethanolvalues are lower than expected, a course of actioncan be determined. Troubleshooting offermentors involves checking records for correctenzyme or yeast addition, or auditing the plantfor mechanical failures. An example ofmechanical failure is an increase in lactic acidvalues and a decrease in ethanol in onefermentor, in which case the Butterworth washermust be closely inspected to determine if it isrotating properly. As another example, if thealcohol content is too low, check for small leaksin the cooling coils. NIR technology can assistin identifying and fixing problems in the earlystage before yield is compromised.

The use of NIR spectroscopy for fermentationanalysis allows the distillery to rapidly performresearch on fermentation supplements. Oneexample is experimenting with various forms ofnitrogen. Nitrogen is the limiting substrate foryeast growth in fermentations. Figure 19 showsNIR results from nitrogen supplementexperiments. It was determined that nitrogen ata specific level decreased fermentation time(increased rate of fermentation) by 30 hrs. A

Figure 18. Production of lactic acid during the course of corn fermentations.

0.40

0.50

0.60

0.70

0.80

0.90

1.00

1.10

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140


Lac

tic

acid

(%

)

fermentor with a nitrogen supplement addedfinished in 50 hrs while the control fermentorsfinished in 80 hrs. A downside of the nitrogensupplement, however, was that if the controlfermentors were given enough time, they wouldfinish with more alcohol than the nitrogen-fedfermentors. This is because the nitrogen inducedyeast cell growth from the carbohydrate sourceinstead of alcohol production. In the end, it wasnot economical for this distillery to use nitrogenas a supplement, since 80 hrs of fermentationwas acceptable. However, if production capacityever needed to be increased, or if the gravity offermentation was increased, then a nitrogensupplement could be an asset to this operation.The NIR analysis allowed the distillery getanalytical results quickly, utilize personnel totheir maximum potential, and optimizedproduction.

NIR is an asset in determining the efficiencyof fermentations. For example, incoming cornwas monitored by NIR spectroscopy and foundto contain 62.6% starch (as is). This means thatthere were 626 kg of starch per metric tonne ofcorn. The 626 kg of starch should produce 696kg of dextrose if it were completely converted.The theoretical ethanol yield from yeast is 51.1%ethanol; therefore, the theoretical yield from1000 kg of this grain is 451 liters. If scales arein place to measure the amount of corn that entersthe fermentor, then the efficiency offermentations can be calculated. If a 200,000


liter fermentor contains 50 metric tonnes of cornand finishes at 10.0% alcohol, then final yield is400 liters of absolute alcohol per tonne of grain.But theoretical yield is 451; therefore, thisfermentation was 89% efficient.

The efficiency of the distillation column canalso be determined. If the above fermentorcontained 20,000 liters of ethanol, and the tankthat recovered the ethanol contained 20,500 litersat 96.0% ethanol, then the distillation efficiencywas 98.4%.

Calibration models for rye mash, barley mash,and molasses have also been developed atHiram Walker & Sons distillery. Optimizationstudies, efficiency determinations, andtroubleshooting currently are carried out forthese types of fermentations. A future calibrationmodel also possible using NIR technology isdetermination of the percent solids in mash priorto exiting the pre-liquefaction tank or prior tofermentation. This can help quickly identify ifthere are milling problems or belt scaleproblems.

DRYHOUSE OPERATIONS

There are several applications for NIR analysisin the dryhouse. The first application is thedetermination of the percent solids in solubles,

or the condensed syrup discharged from theevaporators. After the third effect of a fallingfilm evaporator, the desired level of solids isaround 40%. If the solids level gets much higher,it can cause fouling and eventual plugging ofthe discharge lines. It is very expensive to cleanpipes plugged with solid material; therefore it ishighly desirable to have a quick and easymethod to determine the percent solids. NIR hasthe ability to measure the percent moisture ofthe syrup in a Pyrex beaker at very warmtemperatures with the integrating sphereaccessory. After the percent moisture isdetermined, simply subtract this value from 100to get the percent solids. At Hiram Walker &Sons, the RMSECV for this calibration model is+/- 0.8%.

The most important application for NIRanalysis in the dryhouse area is DDGS analysis.After the rotary driers, most distilleries willguarantee to their customers minimum levels forprotein and fat; while there are maximum levelsfor moisture, fiber, and ash. Protein levels forDDGS are probably the most importantguarantee for the end customer. The mostaccurate wet chemical method for protein is theKjeldahl method. Figure 20 illustrates the crossvalidation for a calibration model using theKjeldahl method as the reference method.

Moisture analysis is also important for the

0

1

2

3

4

5

6

7

8

9

10

11

0 10 20 30 40 50 60 70 80


Eth

ano

l (%

)

Nitrogen

Supplement

Control

Figure 19. Effect of a nitrogen supplement to corn fermentations on fermentation time.


quality of DDGS. To get optimum results, themost accurate wet chemical procedure formoisture analysis is the Bidwell-Sterling method,as compared to the oven moisture method. Theprocedure for the Bidwell-Sterling methodinvolves refluxing 50 g of DDGS in 200–250mL of xylene for 2 hrs. There is a graduatedside arm that collects the water (more dense)from the DDGS, and the remaining (less dense)xylene falls back into the boiling flask. At HiramWalker & Sons the RMSECV of the NIRcalibration model for moisture using Bidwell-Sterling as the reference method is +/-0.35%.

The three remaining components in DDGS thatcan be measured by NIR are fat, fiber, and ash.A fat extraction unit is used to obtain thereference value for fat content in DDGS, whichis a 2-day test. Similarly, fiber content isdetermined with a fiber extraction unit, whichrequires three days. These two components willtake an NIR technician a much longer time todevelop the calibration models. In addition, thecost of each of these extraction units can be quitehigh and they require hazardous chemicals. Thereference method for ash is a simple test. A 2 gsample is burnt in an oven for 2 hrs; and theweight loss is measured at the end. The RMSECVobtained at Hiram Walker & Sons for eachDDGS component is listed in Table 3.

Table 3. RMSCV for DDGS components.

Component Number of Approximate RMSCVfactors (+/- on a % as is basis)

Protein 7 0.46Moisture 5 0.35Fat 6 0.35Fibre 5 0.49Ash 6 0.13

A future opportunity for NIR analysis in thedryhouse is the determination of lactic acid inthe backset stillage that is recycled to thefermentors. The toxic effects of backset stillageon yeast are well documented. If a distiller canmonitor lactic acid and control the amount ofbackset stillage in the mash bill, then alcoholproduction can be optimized.

Another opportunity for NIR analysis in thedryhouse is to measure the residual starch inDDGS. After some initial preliminary trials withstarch analysis by cooking, converting, andmeasuring sugar levels, it was found that thismethod can be quite complex, lengthy, andinaccurate. If fermentations are good, thenresidual starch will not be a concern. Time isbetter spent optimizing fermentation thancreating calibration models for starch in DDGS.

Rank: 7

R2

= 91.75

RMSECV = 0.462

20 21 22 23 24 25 26 27 28 29 30 31

20

21

22

23

24

25

26

27

28

29

30

31


NIR

pre

dic

ted

val

ue

Figure 20. Cross validation of crude protein content of DDGS.


BLENDING OPPORTUNITIES AND ETHANOLDETERMINATION

In the whisky or liqueur industry it is imperativethat the finished product contain the correctstrength of alcohol. Most products will have atarget specification. For example, most whiskieshave a specification of 40.0% alcohol byvolume, however, governments have toleranceson these specifications. For the Bureau ofAlcohol, Tobacco and Firearms (BATF), thespecification is 40.0% minus 0.15% alcohol byvolume, which gives a very small target rangeof 39.85-40.00%. Excise Canada has a largertolerance window of 40.0% ±0.2%, which is atarget range of 39.8%-40.2%. This leaves a verytiny window for a blender to make sure of thecorrect proof of the final product.

To make matters more difficult, when blendingwhiskies, rums, tequilas, or liqueurs there aresuspended solids that obscure the final percentalcohol when measured by a DMA. Theobscuration has to be determined prior to finalproofing in order to determine the true percentalcohol. In order to compensate for theobscuration, the blender must either distill thesample and proof the distillate by DMA, or ovendry the alcohol and determine the solids byweight and factor it into the ‘as is’ DMA result.Either method can be time consuming, and canhold up production for hours for the true percentethanol reading. If the true percent ethanolreading is outside the tolerance reading, then theblender must make adjustments and repeat theprocess. Laboratory analysis in a blendingdepartment is the bottleneck of the operation,and the longer the product stays in a tank themore money it costs the company. In order toreduce the time in blending, the NIR technologycan be a considerable asset. Currently NIRmeasurements are collected using 8 mmdisposable glass vials in transmission mode;however a fiber optic probe can be used ifdesired.

Another benefit of NIR analysis is that it wasfound that the spectrometer could be calibratedto measure ethanol over various temperatureranges without losing much accuracy. This ispossible only if data taken at differenttemperatures are included in the model. Extrasamples should be included in the calibrationwhen failures in the environmental controls of

the laboratory occur. The temperature effects onNIR spectra and the NIR calibration model forethanol in a liqueur were observed during aperiod when the Hiram Walker & Sonslaboratory temperature control systemmalfunctioned. The normal liqueur samples wereoften detected as abnormal samples (outliers)during this period. To resolve the issue, NIRspectra of the liqueur samples were measuredover a range of temperatures and incorporatedinto the calibration model. This calibration modelis shown in Figures 21a and b. Today, thecalibration can be used to determine percentalcohol of the liqueur accurately even when thelaboratory temperature changes a few degrees.Alcohol content of a liqueur can be calibratedto an RMSEP of +/- 0.036%, which is inside thetolerance range allowed by the BATF.

To illustrate further the importance of thereference method, the alcohol content ofapproximately 40 whisky samples was measuredby three different methods: the DMA, the GC,and HPLC. These data were used to generateNIR calibration models. The results are shownin Figures 22a, b and c. This clearly shows thatthe DMA is the most accurate and reliablemethod for a blending department, and shouldbe used as the reference method for NIRcalibrations.

For the future, the next step is to implementin-line measurements to determine the proof ofthe spirit. Currently in a blending operation, thespirit is brought in house at high proof, and thentested for strength. Components are mixed andagain tested for strength. The sample is thenfiltered for final polishing, then stored in a thirdtank. Water is the last component to be added tobring the product down to the final strength,which requires even larger tanks. If NIR werean approved method of determining alcoholstrength, it could be used in-line as a methodfor alcohol strength in the dilution processimmediately before bottling, thus requiring lesstank space and reducing the associated costs.

An additional benefit of NIR technology is thatdifferent components can be determined fromone measurement. For example, Figure 23 showsthe measurement of titratable acidity for analcoholic spirit with added malic and citric acids.The problem with this product was that when itwas distilled for proofing, it required anadditional prior step of neutralization in order


to achieve accurate results, which addedsignificant time in the laboratory. NIRspectroscopy can measure the titratable acidityand the proof of the product at the same time.

Other examples of components that could beanalyzed with NIR spectroscopy are fructose orsucrose when blending liqueurs.

Rank: 3

R2

= 99.9

RMSEP = 0.038

20 20.4 20.8 21.2 21.6 22 22.4 22.8 23.2 23.6 24 24.4 24.8 25.2 25.6 26

20

21

22

23

24

25

26


NIR

pre

dic

ted

val

ue

Rank: 5

R2

= 99.84

RMSEP = 0.0416

20 20.4 20.8 21.2 21.6 22 22.4 22.8 23.2 23.6 24 24.4 24.8 25.2 25.6 26

20

21

22

23

24

25

26


NIR

pre

dic

ted

val

ue

Figure 21a. External validation of the initial model for prediction of percent ethanol on an independent set of 94 liqueur samples.The values of R2 and RMSEP for the external validation are 0.9990 and 0.038% vv, respectively. The number of calibration

samples is 98.

Figure 21b. The external validation result after incorporating additional variances such as temperature. The values of R2 andRMSEP for the external validation are 0.9984 and 0.042% vv, respectively. The number of samples in the calibration set and

independent validation sets are 123 and 120, respectively.


Rank: 2

R2

= 21.33

RMSECV = 0.65

38.1 38.5 38.9 39.3 39.7 40.1 40.5 40.9 41.3 41.7

38.1

38.5

38.9

39.3

39.7

40.1

40.5

40.9

41.3

41.7


NIR

pre

dic

ted

val

ue

Rank: 2

R2

= 30.31

RMSECV = 0.283

39 39.2 39.4 39.6 39.8 40 40.2 40.4 40.6 40.8 41

39

39.2

39.4

39.6

39.8

40

40.2

40.4

40.6

40.8

41


NIR

pre

dic

ted

val

ue

Figure 22a. Cross validation of 45 whisky samples where the reference percent alcohol was measured by HPLC. The RMESCV+/- was 0.66% v/v and R2 is 0.213.

Figure 22b. Cross validation result of 45 whisky samples where the reference percent alcohol was measured by GC. TheRMSECV was +/- 0.28 and the R2 is 0.303.

Conclusions

NIR is definitely a powerful tool that should notbe ignored in the distilling industry. Water andcommon organic molecules such as starch,protein, oil, fiber, ash, acids, sugars, and ethanolcan be monitored quickly, easily and reliablyby NIR spectroscopy at every stage of the

operation. The quick and reliable NIR analyticalresult can help the distilling industry keepproduction processes in control. It is a tool ofthe future that will replace some of the older andslower techniques. NIR technology can alsoassist in research projects where more analysescan be done with fewer resources. NIR analysisallows distillers to troubleshoot and optimize


production. There are exciting times ahead asthe next generation of NIR spectrometers willcontinue to improve as hardware and softwarebecome more advanced and the precision andaccuracy of NIR technology steadily increase.One of biggest hurdles in implementing NIRmethods is to generate reliable calibrationmodels after the NIR instrument is purchased.

Rank: 2

R2

= 90.51

RMSECV = 0.0663

39.55 39.65 39.75 39.85 39.95 40.05 40.15 40.25 40.35 40.45 40.55 40.65 40.75 40.85

39.55

39.65

39.75

39.85

39.95

40.05

40.15

40.25

40.35

40.45

40.55

40.65

40.75

40.85


NIR

pre

dic

ted

val

ue

Figure 22c. Cross validation result of 45 whisky samples where the reference percent alcohol was measured by DMA. TheRMSECV was +/- 0.067 and the R2 is 0.905.

Method:

Rank: 12

R2

= 99.52

RMSECV = 0.106

5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10 10.5 11 11.5 12

5

6

7

8

9

10

11

12

Pre

dic

ted

to

tal a

cid

s (g

/l)

True total acid content (g/l)

Figure 23. Cross validation of the titratable acidity of a high proof alcohol sample. The RMSECV is +/-0.106 with anR2 of 99.52.

In the foreseeable future, it is possible thatinstrument vendors or companies that specializein NIR model development could provide ‘turn-key’ calibration for the distilling industry, leavingmore time for distillers to focus on optimizationof operations. NIR spectroscopy is here to stay;and companies that do not utilize this technologywill run the risk of being left behind.


References

ASTM. 1997 Annual Book of ASTM Standards.Standard Practices for Infrared, Multivariate,Quantitative Analysis. Vol. 03.06. 827–852.

EMEA 2003. Note for Guidance on the Use ofNIR Spectroscopy by the PharmaceuticalIndustry and the Data Requirements for NewSubmissions and Variations. European Agencyfor the Evaluation of Medicinal Products,London. UK.

Griffiths, P.R. and J.A. de Haseth. 1986. FourierTransform Infrared Spectrometry. John Wiley& Sons, New York.

Haaland, D.M., and E.V. Thomas. 1988. Partialleast squares methods for spectral analyses. 1.Relation to other quantitative calibrationmethods and the extraction of qualitativeinformation. Anal. Chem. 60: 1193 - 1202.

Herschel, F.W., Philos., Trans. R. Soc. London.1800. 90:284-292.

Martens, H., and T. Næs. 1989. MultivariateCalibration. John Wiley & Sons Ltd.Chichester, UK

Williams, P.C. and K. Norris. 2001. Variablesaffecting near-infrared spectroscopic analysis.In: Near-infrared technology in theagricultural and food industries. 2nd ed.American Association of Cereal Chemists. 171- 199.

Williams, P.C. 2001. Implementation of near-infrared technology. In: Near-infraredtechnology in the agricultural and foodindustries. 2nd ed. American Association ofCereal Chemists. 153-170.

Biotechnological applications of nonconventional yeast 171

Chapter 13

Emerging biorefineries and biotechnological applications of nonconventionalyeast: now and in the future

CHARLES A. ABBAS

Director of Yeast and Renewables Research, Archer Daniels Midland, Decatur, Illinois, USA

Introduction

The biotechnology, food, and pharmaceuticalindustries have had a long and profitableassociation with yeast other than Saccharomyces(Abbas, 2001; Kurtzman et al., 2001; Demainet al., 1998; Fleet, 2000; Phaff, 1985). Howeverin contrast to the widely utilized and studiedmembers of the genus Saccharomyces, study ofnonsaccharomyces yeast, frequently referred toas nonconventional yeast (NCY), has laggedbehind – as has recognition of the vital role theyplay in ecological, biotechnological, andindustrial applications. NCY have found manyuses in the production of enzymes,polysaccharides, feed additives, fuels, flavors,chemicals, and most recently pharmaceutical andnutraceutical compounds (Table 1). The rapidgrowth in the use of NCY in industrial andbiotechnological applications has been aided bycontinued advances in genetic engineering andfrom new developments in biochemical andphysiological analyses. These applicationsexploit the wealth of the metabolic repertoire thatthis group of yeast possesses as well as theirunique physiological adaptation to growthconditions and habitats that are challenging toSaccharomyces. This chapter aims to introduceseveral NCY of current or potential interest andencourage the reader to consider the variety ofproducts possible in the distillery.

Taxonomic considerations

Since the taxonomic designations of yeastincluding that of NCY are forever changing, thereader is directed and encouraged to consult thetwo recent yeast taxonomic guides by Kurtzmanand Fell (1998) and by Barnett et al. (2000). Inthe introduction to the guide by Barnett et al.(2000) the authors point out the confusion andmany frustrations to which the discipline of yeasttaxonomy lends itself. This is further illustratedby recent results of comparative genomicanalyses that undoubtedly will force a revisitationof many of the current taxonomic designations.In spite of this, the above two guides provide agood current cross-reference for the majority ofthe currently recognized species of NCY thatwill be discussed. In the early chapters of thesetwo guides the reader is provided with many ofthe tools employed today by moderntaxonomists including the recent greater relianceon molecular and evolutionary approaches todefine new families, orders, genera and species.This is not to say that taxonomists haveabandoned morphological characteristics,reproductive cycles, physiological, ecological,and biochemical profiles as these continue toplay a significant role in establishing yeastrelationships and thereby their nomenclature.The role of morphological characterization anddifferentiation in nonconventional ascomycetous

172 C.A. Abbas

yeast consists of asexual vegetative reproductivegrowth and sexual reproduction as illustrated inFigures 1-3. In Figure 1, obtained with lightmicroscopy, the use of budding andpseudohyphae with blastoconidia for thecharacterization and differentiation of severalnew isolates of Candida is provided (Kurtzman,2001) and in Figure 2 a phase contrastmicroscopic photo of the commonly observedthree cell division types, which consist ofmultilateral budding, bipolar budding, andfission in three different genera of yeast isdepicted (Kutrzman et al., 2001). Sexualreproduction or ascosporulation can be readilydemonstrated for some species of ascomycetousNCY but with great difficulty for others. Theascospores formed exhibit a great variety ofshapes as shown in Figure 3. In this chaptermany of the most common names of NCY andtheir synonyms as designated by Barnett et al.(2000) will be used so the reader is encouragedwhen appropriate to consult with this source foradditional information.

Ecological distribution ofnonconventional yeast

The ecological distribution and natural habitatsof NCY are diverse. NCY have been found worldwide in association with decaying matter and insoil, fruits, cereal grains, dairy products, animalby-products, insects, bats, birds, as well as innumerous plant and animal-derived productssuch as syrups, cheeses, alcoholic beverages,and other fermented foods (Fleet, 2000). Yeastecologists and taxonomists recognized early onthe significance of the distribution of NCY intheir natural habitats and their ability to growon a variety of substrates under various andoccasionally extreme conditions. Thisrecognition paved the way for the exploitationof NCY initially in food production and laterfurther expanded to a variety of industrialapplications through use of strain developmentand selection programs. In the excellent chapterby Lachance and Starmer (1998), the authors

Table 1. Selected current and potential biotechnological and industrial products derived from nonconventional yeast.

Amino acids L-cystine, L-glutamic acid, L-serine, L-lysine, L-threonine, D and L-methionine; L-tryptophan, N-acetyl-glycine, L-phenylalanine

Carotenoids Astaxanthin, carotene, lycopene

Chemicals Acetoin, adipic acid, acetanilide, dicarboxylic acids, 1,3-propanediol, ethylacetate, 2,3-butanediol, glycerol,1-ß-hydoxybutyric acid, α-ketoglutaric acid, lactic acid

Enzymes Invertase, galactosidase, lipase, phytase, pectinases, trehalase, inulinase, glucoamylase, proteases, esterases

Feed additives SCP, SCO, phytase, ß-glucans, riboflavin, amino acids

Fermented foods Bread doughs, miso, soy sauce, ethanol

Flavors 2-phenlyethanol, nucleotides, ethylacetate, acetaldehyde, citric acid, furanones, methyl ketones

Food ingredients Citric acid, acetic acid, lactic acid, ribonucleotides, biosurfactants-sophorolids and additives

Heterologous and Enzymes, pharmaceuticals-vaccines, immunoproteins other proteins

Organic acids Acetic, lactic, citric, iso-citric, formic, pimelic, homogentisic acids

Nutraceuticals, N-acetylglucosamine, inositol, S-adenosyl methionine, melanin, tumor necrosis factor, human serum pharmaceuticals factors, ß-glucans and cosmetics

Polyols and Erythitol, glycerol, xylitol, D-arabitol, L-arabitol, D-mannitol sweeteners

Simple sugars and Xylulose, D-psicose, D-lyxose, mannans, phosphomannans, phosphogalactans, pullulans, mannoproteins, polysaccharides glucans, heteroxylans, exo-and acidic polysaccharides

Vitamins Riboflavin (B2), B12, thiamin, ergosterol, ubiquinone


Figure 1. Morphological differentiation and characterization of six new species of Candida. Budding cells are depicted in A, D,G, I, K,and M. Pseudohyphae with blastoconidia are depicted in B, C, E, F, H, J, L and N. Budding cells are from 5% malt extract

agar after 3 days at 25°C while pseudohyphae are from Dalmau plate cultures on yeast morphology agar following 7 days ofincubation at 25°C. Light microscopic photo was provided courtesy of Dr. Cletus Kurtzman.

174 C.A. Abbas

Figure 2. Cell division in the ascomycetous yeast. A) Multilateral budding in Saccharomyces cerevisiae; B) Bipolar budding inSaccharomycodes ludwigii; C) Fission in Schizosaccharomyces octosporus. Phase contrast photo was provided by Dr. Cletus

Kurtzman.

Figure 3. Representative ascospores of A) Hat-shaped (galeate), Pichia bispora; B) Spheroidal, Saccharomycopsis capsularis; C)Elongate with a whip-like extension of the cell wall, Eremothecium (Nematospora) coryli. Bar=5 microns. Bright field photomi-

croscopic photo provided by Dr. Cletus Kurtzman.


stress the significant contribution of yeastbiology to the study of evolutionary ecologyand the valuable role played by yeast in theinteraction of the microbial communities in theecosystem. As such, one cannot ignore the roleof yeast in climatic changes and the need forthe research community to utilizemultidisciplinary approaches to bridge the gapsin the fundamental knowledge of thesignificance of yeast to ecological studies.

NCY industrial strain developmentprograms

Historically strain development programs withNCY have followed the same path successfullytaken with the yeast Saccharomyces cerevisiae(Benitez et al., 1996; Penttila, 1997; Beckerichet al., 1982; Phaff, 1985). These approacheshave often consisted of classical mutagenesisfollowed by screening and selection, yeasthybridization using mating followed bysporulation and tetrad analysis (ploidydependent) or rare mating (ploidy independent),and the employment of parasexual geneticsusing protoplast fusions (Phaff, 1985; Panchalet al., 1984; Beckerich et al., 1982; Benitez etal., 1996). The advent of molecular biologygave rise to the use of new approaches in NCYstrain development with increased use ofrecombinant DNA to transform yeast viaincorporation of genes on plasmids acting asshuttle vectors or by direct chromosomalintegration. The development of tools for thegenetic manipulation of NCY is described ingreater detail in several recent texts that coverprogress in this area (deWinde, 2003; Wolf,1996; Wolf et al., 2003). Using the listedapproaches in this section numeroushomologous and heterologous genes have beencloned and successfully expressed in most NCYof industrial importance as well as those ofcurrent interest (deWinde, 2003; Wolf, 1996;Wolf et al., 2003). As newer yeast straindevelopment approaches are increasingly morereliant on genomic and metabolic engineeringanalyses, these approaches are also emergingas new tools with NCY as well. Thecombination of genome sequencing with geneexpression profiling using microarray chips, 2D

protein electrophoresis, and biochemical andmetabolic flux analyses has been adopted as partof strain development strategies to improve NCYstrains for commercial applications. While theapplication of these approaches in NCY lagsbehind that of S. cerevisiae, great progress inthis area has been facilitated by the major effortto sequence a number of NCY of industrial andbiotechnological importance as a part of theGenolevures undertaking in France. Thisimpressive sequencing program has sought todelineate and compare the functional genomesof hemiascomycetous yeast using randomsequence tags (RST) to obtain the maximumbiologically relevant data with a minimum ofeffort. The selected list of 13 ascomycetous yeastincluded several NCY yeast of industrial andbiotechnological significance (Figure 4). Adetailed and a complete description of the resultsof the initial phase of this effort was publishedin FEBS Letters in December of 2000 and arecent update of the latest progress was presentedat Yeast 2003 in Göteborg, Sweden (Dujon,2003).

Yeast physiology and biochemistry

NCY yeast physiology and biochemistrycontinue to be active and attractive areas ofresearch (deWinde, 2003; Wolf, 1996; Wolf etal., 2003). The greater recent attention to NCYhas resulted from great strides made in ourunderstanding of the genetics, physiology andbiochemistry of the yeast S. cerevisiae. In S.cerevisiae this has given rise to what is currentlyreferred to as the discipline of system biology.System biology seeks to integrate knowledgegained from genetic, biochemical, structural, andphysiological research to develop acomprehensive approach to our understandingof the metabolism of yeast and other organisms(Nielsen and Olsson, 2002). An essentialcomponent of this approach is to provide adetailed description of the interactions amongall components within a cell. In order to pavethe way for the use of this approach to NCY,research must address many of the areas listedin Table 2. Due to space limitations the reader isurged to consult with many of the excellentreferences provided at the end of this chapter.The successful application of this approach to

176 C.A. Abbas

NCY will undoubtedly have a significant impacton the increase in industrial and biotechnologicaluses of these yeast.

Table 2. List of biochemical, cellular, metabolic,nutritional and physiological topics currently studied inyeast.

Cell wall chemistry, organization, and flocculationGrowth kineticsNutritional requirements and metabolism of nutrientsRegulation of C/N metabolismStress responses and signal transductionSugar utilization profilesSugar transport and energeticsProtein synthesis, trafficking, secretion and turnoverYeast organelles biogenesis and functionYeast cell structure, compartmentalization, and organizationYeast cell cycle, control of growth and cellular divisionYeast aging and apoptosisGenomics: structural, functional, comparative, evolutionary

NCY of biotechnological importance

CANDIDA UTILIS (ASEXUAL STATE OF PICHIAJADINII)

Candida utilis (also known as Torula utilis orTorula yeast) has been isolated from human and

animal sources and from distillery and yeastfactories (Phaff, 1985; Kurtzman and Fell, 1998;Barnett et al., 2000). Vegetative reproduction inthis yeast occurs by budding with filaments orsimple pseudohyphae occasionally reported.Sexual reproduction has been recognized; andas a result of this its relatedness to Pichia jadiniiaccepted by taxonomists. Due to some concernsabout the potential pathogenicity of this yeast ithas been classified at a biological safety level I(Barnett, 2000). This is in spite of the long safetyrecord attributed to this yeast and its acceptanceby the US FDA as a safe substance and itsapproval as a food additive (Miura et al., 1999).

C. utilis has been reported to produce a varietyof products ranging from amino acids (L-serine;D-methionine; L-glutamic acid), glutathione, S-adenosyl methionine, RNA (9-ß-D-ribo-fuarnoside-5´-phosphoric acid esters of 2-substituted-6-hydroxypurine), coenzyme A, singlecell proteins (SCP), polyols (xylitol), organic acids(acetic acid, adipic acid), sophorolipids,ethylacetate, methyl ketone, ethanol andacetaldehyde. Most recently this yeast has beengenetically engineered to produce a number ofcarotenoids, which include lycopene, ß-caroteneand astaxanthin as well as for the heterologous

Figure 4. Genolevures: Comparative genomics of 13 hemiascomycetous yeast.

00 2020 4040 6060

% of total

S. cerevisiaeS. bayanusS. exiguusS. servazziiZ. rouxilS. kluyverlS. thermotoleransK. lactisK. marxianus

Y. lipolytica

P. angustaD. hanseniiP. sorbitophilaC. tropicalis

Common genes

Ascomycetes specific genes


production of a number of proteins includingthe enzyme α-amylase and the sweet potatoprotein, monellin (Miura et al., 1999; Miura etal., 1998; Shimada et al., 1998). A number ofenzymes of this yeast have been isolated andcharacterized. These include alcoholdehydrogenase, allantoin racemase, 2,3-butanediol dehydrogenase, dimethylaminemono-oxygenase, dihydroxyacetone reductase,glutamine synthetase, invertase, maltosepermease, trimethylamine mono-oxygenase,trehalase, urate oxidase, and urease just to namea few. C. utilis is known to assimilate inexpensivesugars produced by the hydrolysis of ligno-cellulose feedstocks such as alfalfa wastes,cellulose, hemicellulose, pineapple waste, potatowaste, and sugarcane bagasse. It can also utilizeand degrade petroleum waste, rapeseed oil meal,sauerkraut waste streams and tallow. Since C.utilis, unlike S. cerevisiae, cannot produceethanol under strict aerobic conditions, it canbe grown in continuous culture to a high density(Miura et al., 1999). These are desirable featuresthat have been exploited in SCP, in theproduction of extracellular enzymes and otherproteins as well as in the production ofintracellular proteins and other metabolites suchas carotenoids by this yeast. A recent taxonomicdescription of this yeast can be found in Barnettet al. (2000) on pages 532-533.

CANDIDA TROPICALIS

Candida tropicalis is found in association withhuman beings, soil, water, flowers, and in foodssuch as kefir, miso, dates, sauerkrauts, and otherfermented foods produced by mixed culturefermentations. This yeast has also beenrecovered as a contaminant from vegetable oilprocessing facilities, blackstrap molasses, rottenpineapples, and from citric acid fermentations.Vegetative growth in this yeast occurs bybudding with elaborate or septate hyphalformation. No sexual reproductive cycle isknown for this yeast. C. tropicalis has beenreported to produce SCP, ergosterol, xylitol,several amino acids (L-cystine, L-trptophan; L-methionine; N-acetyl glycine), acetanilide, adipicacid, carnitine, ethanol, glutaric acid, L-iditol,mannans, ribonucleic acids, and ubiqinone.

Tanaka et al. (1971) recognized the ability of C.tropicalis to grow on n-alkanes and most recentlythis yeast has been genetically engineered toefficiently produce long-chain dicarboxylic acidsfrom alkanes and fatty acids (Fabritius et al.,1996; Picataggio et al., 1992). These acids areof commercial interest as they can serve asprecursors of polyamides, fragrances, andpolyhydroxyalkenoates (Fabritius et al., 1996).C. tropicalis has also been widely studied forthe production of xylitol from xylose derivedfrom lignocellulosics, where it has shown greatpromise in achieving high rates of conversion,high yield, and high concentration in pilot plantstudies (Yahashi et al., 1996; Choi et al., 2000;Kim et al., 1999; Oh and Kim, 1998). A numberof enzymes from this yeast have been isolatedand characterized including alcohol oxidase,carnitine acetyl transferase, catalase, and urateoxidase. C. utilis has been shown to utilize and/or degrade hydrocarbons, kraft black liquor fromphenol, D-xylose, petroleum, and paper millwaste. The inclusion of this yeast in the initialGenolevures comparative genomic explorationis described by Blandin et al. (2000a). A detailedupdate taxonomic description is found in Barnettet al. (2000) on pages 270-271.

CANDIDA FAMATA (ASEXUAL STATE OFDEBAROMYCES HANSENII)

Candida famata is among the most widelydistributed of the NCY. This yeast has beenisolated from a wide variety of fermented foodsand beverages as well as from human and animalsources. Vegetative reproduction is by buddingwith filaments and simple pseudohyphae formed.Persistent asci are formed by this yeast thatcontain 1-2 rough, round ascospores (Barnett etal., 2000). Recent taxonomic guides list thisyeast as the asexual state of Debaromyceshansenii. The production of the polyols D-arabitol and xylitol, SCP, the organic acids citricand isocitric acid, production of D-lyxose fromD-glucose, production of D-talitol from D-psicose, production of antimicrobials, and thevitamin riboflavin (B2) have all been reported(Lim et al., 2001; Ahmed et al., 1999; Stahmannet al., 2000; Sasahara et al., 1998; Linardi et al.,1996). Special features of this yeast include itsosmotolerance as indicated by its high glucose

178 C.A. Abbas

and salt tolerances as well as its ability to growon hydrocarbons and the pentose sugars D-xylose and L-arabinose that are present inhydrolysates of lignocellulosics (Roostita andFleet, 1999). Flocculation has been describedfor this yeast in the presence of peptone and itsinvolvement in the formation of biofilms hasbeen noted (Martinez et al., 1996). Recently atransformation system has been developed andoptimized for this yeast (Voronovsky et al.,2002) and the use of PCR to identify anddistinguish this yeast from other Candida hasbeen published (Nishikawa et al., 1997; 1999).Due to the inclusion of D. hansenii in theGenolevures comparative genomic project, thegenomic sequence of this yeast has beendelineated and compared to a number of otherascomycetous yeast (Lepingle et al., 2000;Dujon, 2003). Recently, attempts to characterizeand isolate genes responsible for salt tolerancein D. hansenii have been reported (Ramos et al.,2003). Understanding the underlyingmechanism for salt tolerance in this yeast canhave a significant impact on its furthercommercial exploitation as it has been used forover a decade for the large-scale commercialfermentation production of the vitamin riboflavin(B2) (Abbas, 2001). A complete recentdescription of D. hansensii is provided byBarnett et al. (2000) on pages 336-337.

YARROWIA LIPOLYTICA (ASEXUAL STATE OFCANDIDA LIPOLYTICA)

Yarrowia lipolytica is frequently found inassociation with dairy and with waste streamsderived from vegetable oil processing facilities.Reproduction primarily occurs via budding andfilaments of pseudohyphae are frequentlyobserved microscopically. Sexual spores areformed that resemble rough, oval, hat-, Saturn,or walnut-shaped ascospores, and have beenreported following mixing of sexuallycompatible strains. Recent taxonomic guideshave recognized that C. lipolytica is the asexualstate of Y. lipolytica (Barnett et al., 2000). Theproduction of organic acids such as citric, 2-methyl isocitric, isocitric acid, homogentisic(breakdown products from metabolism of L-phenylalanine), pimelic acid, α-ketoglutaric acidand L-ß–hydroxybutyric acid, the polyols

mannitol, xylitol and erythritol, lipases active atpH 5.5 and 7.5, lipases active at pH 9.0 and60°C, lipase activators, proteases, alkalineprotease, melanin, and single cell protein haveall been reported for this yeast (Kyong and Shin,2000; Delgenes et al., 1998; Silva et al., 1996;Rodrigues et al., 1998). One of the manyinteresting characteristics of this yeast is theability to hydrolyze and reduce zearalenone, aswell as the ability to degrade other aromaticcompounds, fatty acids, and alkanes such asazelaic acid, benzo (α) pyrene, biphenyls,hydrocarbons, naphthalene, petroleum, rapeseedoil, and tallow (Hassan anbd El-Sharkway, 1999;De Felice et al., 1997). The above propertieshave been exploited in the production of organicacids and enzymes using ethanol, glucose, fattyacids, or hydrocarbons as the primary carbonsubstrates (Arzumanov et al., 2000; Antonucciet al., 2001; Wache et al., 2003; Pazouki et al.,2000). More recently the development of anefficient transformation system for Y. lipolyticahas aided further in the expansion of the uses ofthis yeast to the production of proteins such asenzymes and pharmaceuticals (Chang et al.,1998; Barth and Gaillardin, 1996). The inclusionof this yeast in the Genolevures 1 and 2comparative genomics project has aided in itsgenomic analysis and comparison to otherhemiascomycetous yeast (Dujon, 2003;Casaregola et al., 2000). A complete recenttaxonomic description is provided by Barnett etal. (2000) on pages 785-786.

CANDIDA GUILLIERMONDII (ASEXUAL STATEOF PICHIA GUILLIERMONDII)

Candida guilliermondii is commonly found inassociation with human beings, animal productsand by-products as well as in soil, water, dairyproducts, dough, insects, and fruit juices.Vegetative reproduction in this yeast occurs bybudding and filaments. Either none or elaboratepseuodhyphae are observed by microscopy.Four hat-shaped ascospores are formed byconjugation of sexually compatible strains uponmixing. When cultured on corn meal agar for11 days, extensive and long filaments have beenreported. C. guilliermondii is a flavinogenic yeastthat accumulates riboflavin intracellularly. Otherproducts produced by this yeast are SCP, ethanol,


citric acid, xylitol from D-xylose, L-arabitol fromL-arabinose, as well as tea cider (Saha andBothast, 1996). Two enzymes that have beenisolated and well characterized from this yeastare alcohol dehydrogenase and deaminase. Dueto the interest in the production of xylitol by C.guilliermondii, the impact of acetic acid andsugar degradation products present in a numberof lignocellulosic hydrolysates from the twoenzymes involved in D-xylose metabolism inthis yeast, xylose reductase and xylitolreducatase, has been extensively studied.Growth of this yeast in brine and on cellulosehydrolysates has been reported. The recent useof this yeast as a biocontrol agent to reducepostharvest fruit spoilage has been described.Genetic transformation of C. guilliermondii hasbeen reported (Boretsky et al., 1999) and adetailed description of the genetics of this yeastprovided by Sibirny (1996a). A completedescription of the taxonomy of this yeast isprovided in Barnett et al. (2000) on pages 523-524.

KLUYVEROMYCES LACTIS

Kluyveromyces lactis is another milk-sugarfermenting yeast that has been isolated from awide range of sources, which include dairyfactories, wineries, human sources, and plantsources such as the slime of an oak tree. Thisyeast been shown to produce ethanol from wheyproduced from cheese manufacturing. Somestrains of this NCY are known to harbor yeastprotein killer factors. Genetic studies of this yeasthave led to the development of efficienttransformation systems for the production ofproteins such as the enzyme ß-galactosidase andthe proteinase calf chymosin, as well as manyother proteins that are listed in the extensivereview by Wesolowski-Louvel et al. (1996).Most of these proteins are commerciallyimportant in pharmaceutical applications. Themetabolism of galactose, high cell densitycultivation, the pattern of N-proteinglycosylation, and its codon usage choice, whichis similar to that of S. cerevisiae, have been wellstudied (Wesolowski-Louvel et al., 1996).Another recent development with this NCY isits inclusion in the Genolevure comparativesequencing efforts, which has provided further

information to support its relatedness to the otherhemiascomycetous yeast, K. marxianus(Bolotin-Fukuhara et al., 2000; Dujon, 2003).The taxonomic description of this yeast has beendescribed by Barnett et al. (2000) on pp. 417-418.

KLUYVEROMYCES MARXIANUS (SYNONYM K.FRAGILIS AND ASEXUAL STATE CANDIDAKEFYR)

K. marxianus is often associated with spoilageof commercial yogurt and has been commonlyisolated from other dairy products such as milkand cheese, pressed commercial yeast, effluentof sugar refineries, bread doughs, beer,fermenting figs, sewage, infected humans andinfected milk cows, where it causes bovinemastitis. In spite of the association of K.marxianus with infected humans and dairy cows,it is considered GRAS and frequently referredto as the ‘dairy yeast’. Vegetative reproductionin this yeast occurs via budding; and simple toelaborate pseudohyphae are produced. Due tothe production of ascospores, K. marxianus iscurrently classified as the perfect state of C. kefyr,while the closely related K. fragilis is listed asone of its many synonyms (Barnett et al., 2000).When cultured on corn meal or potato dextroseagar for 8-11 days, long filaments are observedmicroscopically. K. marxianus has beenreported to produce ethanol from cheese whey,oligonucleotides for use as flavor enhancers,polyphosphates, metallothionein, glycerol,single cell protein from biomass, severalenzymes for commercial application (lactase andpectinases), pectin extraction from fruits,polyphosphate, and glycerol (Parrondo et al.,2000; Ferarri et al., 1996; Belem et al., 1997;Belem et al., 1998). In addition to lactases (ß-galactosidase) and pectinases, several otherenzymes have been isolated, characterized, andcloned including inulinase, alcoholdehydrogenase, acetoin dehydrogenase, andUDP glucose-4-epimerase (Schwan et al., 1997).Interesting characteristics of this yeast includethe production of killer factors and growth andproduction of ethanol from cheese whey andinulin at 45°C to a final concentration of 7.0%(w/v) (Balletsteros et al., 1993; Singh et al.,1998; Banat et al., 1995). K. marxianus can

180 C.A. Abbas

grow on a wide range of carbon sources andwaste streams such as lactic acid, sauerkrautwaste, orange peel, paper mill sludge, and cocoapulp. A simultaneous saccharification andfermentation process to produce ethanol thatcombines fungal cellulase enzymes with thisyeast has been reported (Lark et al., 1997). Theresults have indicated that further developmentwith the genetic engineering of K. marixanus toferment lignocellulosics to ethanol is feasibleand attractive given this yeast’s ability to growon a variety of feedstocks and its thermo-tolerance. K. marxianus has been shown toaccumulate toxic compounds includingcadmium, copper, and silver, and has also beenused in adhesive testing. The controversy overthe relatedness of this yeast to K. lactis is citedby Wesolowski-Louvel et al. (1996) and Barnettlists K. lactis and K. marxianus separately asseveral lines of evidence clearly distinguishbetween these two yeasts. Both of these yeastswere analyzed as a part of the Genolevurescomparative genomic sequencing and resultsreported (Bolotin-Fukuhara et al., 2000; Llorenteet al., 2000). A complete description of this NCYis provided in Barnett et al. (2000) on pages420-421.

CRYPTOCOCCUS LAURENTII

Cryptococcus laurentii is commonly associatedwith cereal grains, grapes, soil, birds, decayingplant matter, air, seawater, and tropical plants.Vegetative reproduction occurs by budding andfilaments are present or absent with the formationof simple and/or septate hyphae. Sexualreproduction in this yeast has not been reported.C. laurentii is known to produce a number ofexo- and acidic polysaccharides, xylanases andglycosidases, a thermostable glutaminase, killertoxins, L-lysine, and a number of diolcompounds. Other characteristics of this yeastinclude its alkali tolerance and its ability todegrade benzene derivatives. Due to C.laurentii’s ability to produce copious levels ofextracellular polysaccharides, there has beengreat interest in developing food and non-foodapplications for their use (De Baets andVandamme, 1999; Nadezda et al ., 1997).Another area that has received commercialinterest is use of this yeast as a biocontrol agentto retard fruit spoilage (Giuseppe et al., 1998;

Castoria et al., 1997). The involvement of 1,3-ß-glucanases as the mode of action againstspoilage fungi has been implicated as theprotective mechanism. A complete descriptionof this yeast is provided by Barnett et al. (2000)on pages 315-316.

ZYGOSACCHAROMYCES ROUXII

Zygosaccharomyces rouxii is one of the mostcommon food and beverage spoilage yeast.Vegetative reproduction of this yeast occurs viabudding with the presence or absence offilaments or simple pseudohyphae. Persistentasci formation containing up to four rough orsmooth sound ascospores has been noted for thisyeast. Among the many products of this yeastare acetic acid, formic acid, ethanol, alkyl esters,L-glutamine, mannans, miso, 4-hydroxy-furanone, soy sauce, lipases that are active atpH 9.0 at 60°C, UDP-N-acetylglucosamine, and5´-nucleotides (Hayashida et al., 1998).Important characteristics of this yeast are itsresistance to high salt, sugar, sorbic acid, andtolerance to high pressure, which contribute toits survival in and spoilage of preserved foodsand beverages. Due to this property, this yeastis increasingly being targeted forbiotechnological applications (Limtong et al.,1998). The inclusion in the Genolevurescomparative genomics effort will propel furtherrapid developments in the genetic manipulationand exploitation of Z. rouxii (De Montigy et al.,2000; Sychrova et al., 2000). The isolation ofura3 mutants and the development of an efficienttransformation system were recently describedby Pribylova and Sychrova (2003). A completedescription of the taxonomy of this yeast isprovided in Barnett et al. (2000) on page 797.

CANDIDA BOIDINII

Candida boidinii have been isolated from a widerange of geographic locations from a numberof sources such as tanning solutions, fermentedacidic products, swamps and other water habitatsand flowers. Vegetative reproduction occurs bybudding with the formation of simple toelaborate pseudohyphae. No sexual reproductionstage has been described for this yeast. C.boidinii grows on and degrades methanol as well


as pectin and can produce a number of productsof interest such as the polyols xylitol and L-arabitol, xylulose, SCP, nucleotides such as ATP,ethanol, D-amino acids, as well as a number ofpigments (Vandeska et al., 1996; Suryadi et al.,2000; Yamada et al., 2001). Several enzymeshave been isolated and characterized from thisyeast including NAD-linked alcoholdehydrogenase, polyamine oxidase, formatedehydrogenase, diamino acetyltransferase,dihydroxyacetone synthase, and S-formylglutathione hydrolyase (Labrou et al.,2001). Due to C. boidinii’s ability to grow onmethanol, the metabolism of this carbon sourceby this yeast has been extensively studied(Yurimoto et al., 2000, Sakai et al., 1998;Gabrinska-Loniewska, 1997; Aggelis et al.,1999). The development of genetic tools and atransformation system have enabled the cloningand expression of a number of heterologousproteins in this yeast (Sakai et al., 1996;Nakamura et al., 2000). Significant interest hasalso been shown in the use of this yeast for thecommercial production of dicarboxylic acids. Arecent taxonomic description of C. boidinii canbe found in Barnett et al. (2000) on page 145.

PICHIA PASTORIS

Pichia pastoris can be found in exudates ofhardwood trees such as black oak and chestnut.Vegetative reproduction in this yeast is bybudding and no filaments are formed. Theformation of asci containing up to 4 hat-shapedascospores have been noted in this yeast (Barnettet al., 2000). P. pastoris has attracted muchinterest in biotechnological applications thatinvolve heterologous proteins. Its ability to growon methanol as a carbon source has beenextensively characterized and the peroxisomalsystem described in great detail. Thedevelopment of genetic tools for this yeast hasled to the rise in the use of commercial laboratorykits and the demonstration of high-levelproduction of a number of pharmaceutical andother proteins by P. pastoris. In a properlydesigned fermentation process this yeast can becultivated to high density, which is key todemonstrating commercial feasibility ofprocesses that are dependent on optimizing cellmass production (Chung, 2000; D’Anjou andDauglis, 2001; Thorpe et al., 1999; D’Anjou and

Dauglis, 2000; Minning et al., 2001; Trujillo etal., 2001; Rodriguez et al., 1996; Sberna et al.,1996). In addition to its ability to grow onmethanol and other oxygenated hydrocarbons,P. pastoris can degrade hydrogen peroxide andcan be used to produce acetaldehyde and SCP.Due to space limitations, the reader is directedto excellent recent reviews of this yeast thatdescribe in greater detail its biotechnologicalapplications (Gellissen, 2000; Sreekrishna andKropp, 1996). Comparative genomic sequencedata comparison from inclusion of this yeast inGenolevures was reported (Blandin et al., 2000)A recent description of this yeast is provided byBarnett et al. (2000) on page 554.

PHAFFIA RHODOZYMA (ANAMORPH OFXANTHOPHYLLOMYCES DENDRORHOUS)

Phaffia rhodozyma is a carotenogenic yeast thatwas originally isolated from the stump of birchtrees in Japan, Finland and Russia and from analder tree in Japan. This yeast reproducesvegetatively by budding with formation offilaments with simple pseudohyphae present orabsent. Sexual spores or basidiospores are borneon basidia. The current taxonomic classificationof P. rhodozyma as an anamorph of X.dendrorhous has not been accepted by Kurtzmanand Fell (1998). Due to its ability to produce thecarotenoid astaxanthin, which is a valuableadditive in salmon diets, this yeast has been ofimportant commercial interest. The availabilityof improved strains of P. rhodozyma that caneconomically produce astaxanthin intra-cellularly has aided in the development ofindustrial scale high cell density fermentations(Abbas, 2001). The production of othercarotenoids by P. rhodozyma has also beenreported. Cultivation of this yeast to produceastaxanthin from several nonconventionalcarbon sources such as wood hydrolysates, datejuice, glycerol, D-xylose, a number ofdisaccharides, as well as on dried distillers grainsproduced from ethanol dry mill plants has alsobeen noted (Vazguez et al., 1998; Parajo et al.,1998a,b; Fontana et al., 1997; Santopietro et al.,1998; Ramirez et al., 2000; Chan and Ho, 1999;Cruz and Parajo, 1998; Florencio et al., 1998;Kesava et al., 1998). A transformation systemhas also been developed for this yeast andimproved recently (Wery et al., 1998; Visser,

182 C.A. Abbas

2003). A detailed taxonomic description of thisyeast is provided in Barnett et al. (2000) on page784.

HANSENULA POLYMORPHA (SYNONYM PICHIAANGUSTA)

Hansenula polymorphia has been isolated froma number of sources in different parts of theworld, which include spoiled orange juice, maizemeal, the fruit fly Drosophila pseudoobscura,milk from cows with mastitis, and soil irrigatedwith a distillery effluent. Vegetative reproductionoccurs by budding and no filaments are formed.Asci that contain up to 4 hat-shaped ascosporeshave been described for this yeast. H.polymorpha is characterized by its excellentgrowth on short chain oxygenated hydrocarbonssuch as ethanol and methanol as well as othersugar alcohols or polyols, which includeglycerol, erythritol, xylitol, and L-arabitol(Sanchez et al., 1998; Suryadi et al., 2000). Dueto its growth on methanol, the physiology andbiochemistry of this yeast has been well studiedand the metabolism of this carbon source is wellunderstood. Based on the ability of this yeast togrow on methanol and ethanol, alcoholbiosensors have been described that utilizecoupled oxidase-peroxidase enzymes(Vijayakumar et al., 1996). The development ofgenetic tools for H. polymorpha has given riseto the genetic manipulation of this yeast and theproduction of a number of heterologous proteinsof commercial interest such as maltase, L-rhamnosidase, and phytase (Liiv et al., 2001;Mayer et al., 1999; Yanai and Sato, 2000). Thishas led to numerous biotechnologicalapplications of this NCY and due to spacelimitations the reader is urged to consult withexcellent recent reviews cited at the end of thischapter (Hansen and Hollenberg, 1996;Hollenberg et al., 1997; Gellissen, 2000). Adetailed taxonomic description is provided inBarnett et al. (2000) on page 493.

SCHIZOSACCHAROMYCES POMBE (SYNONYMSACCHAROMYCES POMBE)

Schizosaccharomyces pombe was originallyisolated from East African millet beer and sincethen from a variety of alcoholic beverages such

as palm wine, arak, beer breweries, and fromsulphited grape juice, grapes, cane sugar, andmolasses. This yeast is fairly distinguishableusing microscopy by its large size and by itsvegetative reproduction by splitting or fission.The production of filaments in this yeast has notbeen observed. Sexual reproductive asci areproduced with up to four smooth, oval, or roundascospores reported (Barnett et al., 2000). Dueto its unique morphological, biochemical,physiological, and metabolic features, S. pombehas received much recent attention forbiotechnological applications. Theseapplications have taken advantage of its highosmotolerance, high temperature tolerance andthe development of genetic tools for itsexploitation (Dhamija et al., 1996; Chaudhariand Chincholkar, 1999). Among the manyproducts that can be produced with S. pombeare ethanol, SCP, and heterologous proteins foruse in food and pharmaceutical applications(Amanchukwu et al., 1989; Fukuda et al.,1984). An excellent recent text by Giga-Hama(1997) describes many of the heterologousproteins that can be produced by this yeast. Adetailed taxonomic summary is provided byBarnett et al. (2000) on pages 678-679.

OTHER NONCONVENTIONAL YEAST OFINTEREST

In addition to the NCY described in detail in theprevious section there are many others that areof current interest to the biotechnology industry.These include several that can produce ethanolfrom sugars present in lignocellulosichydrolysates such as Candida shehatae,Candida wickerhamii, Pichia stipitis, andPachysolen tannophilus (Sreenath and Jeffries,1999; Sanchez et al., 1999; Amartey and Jeffries,1996; Possoth et al., 1996; Kruse and Schuegerl,1996; Skory et al., 1996; Freer et al., 1998). Theamylolytic NCY Debaromyces (Schwann-iomyces) occidentalis, Lipomyces starkeyi andAruxla adeninivorans have been recently tappedfor their ability to grow on starch and have otherattributes that can be exploited further for a widerange of biotechnological applications (Kunzeand Kunze, 1996; Wartmann and Kunze, 2000;Laires et al., 1983). The yeast Candidabombicola, C. rugosa, and C. maltosa can grow


on n-alkanes, fatty acids, waste animal fat orwaste vegetable oils to produce products suchas single cell oils, biosurfactants, and 1,3-butanediol (Sharyshev et al., 1997; Montesinoset al., 2003; Matsuyama et al., 2001; Deshpandeand Daniels, 1995; Rau et al., 1999; Daniel etal., 1998; Rau et al., 2001; Lang et al., 1996;Daniel et al., 1999; Mauersberger et al., 1996).The NCY Candida krusii, Candida magnoliaand Pichia farinosa can produce polyols suchas glycerol and erythritol (Abbas, unpublished;Sung-Wook et al., 1999; Nakano et al., 2000;Yamazaki et al., 2000). The methylotrophic yeastPichia methanolica can grow on methanol(Sibirny, 1996). The yeast Pichia fermentans canproduce the flavoring compound, 2-phenylethanol (Huang et al., 2000). SeveralNCY, which include Rhodotorula aurantiaca,R. rubra, and R. glutinis have been evaluated asbiocontrol agents to prevent fruit harvest decaywhile the yeast Pichia anomala has been shownto inhibit mold growth during cereal grainstorage (Chand-Goyal et al., 1996; Sugar andSpotts, 1999; Lima et al., 1998). Rhodotorulaglutinis and R. rubra are also known for theirproduction of the carotenoids ß-carotenes andlycopene (Bhosale and Gadre, 2001; Matelli etal., 1990; Buzzini, 2000; 2001). Several NCYare well known pathogens such as Candidaalbicans and Filobasidiella (Cryptococcus)neoformans and therefore have received formany decades much attention from the medicaland the pharmaceutical industries (Demain et al.,1998). Finally, many NCY have been recognizedfor their role in food spoilage. Some of the manyNCY that are involved in food spoilage aremembers of the genera Brettanomyces, Pichia,Candida, Torulaspora, Metschnikowia, andHansenisapora (Grbin and Henschke, 2000).

Future predictions and conclusions

Over the next decade the commercial value ofall products derived from the biotechnologicalapplications of NCY will surpass all thosederived from the current uses of food/fuel/feed/chemical as well as pharmaceutical andnutraceutical uses of S. cerevisiae (Abbas, 2001).Rapid advances in metabolic engineering incombination with bioinformatics and improvedfermentation processes from the greater

employment of metabolic controls aided byincreased automation will fuel much of theprogress in the uses of NCY in industrial andbiotechnological applications. With thesedevelopments in mind, the next decade promisesto be an exciting era for the biotechnologicaland industrial uses of NCY.

References

Abbas, C.A. 2001. Industrial applications ofnonconventional yeast. Plenary presentationabstract, ISSY 2001, Lviv, Ukraine.

Aggelis, G., S. Fakas, S. Melissis and Y.D. Clonis.1999. Growth of Candida boidinii in amethanol-limited continuous culture and theformation of methanol-degrading enzymes. J.Biotech. 72(1/2):127-139.

Ahmed, Z., H. Sasahara, S.H. Bhuiyan, T. Saiki,T. Shimonishi, G. Takada and K. Izumori.1999. Production of D-lyxose from D-glucoseby microbial and enzymatic reactions. J.Bioscience and Bioeng. 88(6):676-678.

Amanchukwu, S.C., A. Obafemi and G.C.Okpokwasili. 1989. Single-cell proteinproduction by Schizosaccharomyces pombeisolated from palm wine using hydrocarbonfeedstocks. Folia Microbiol. (Prague)34(2):112-119.

Amartey, S. and T. Jeffries. 1996. Animprovement in Pichia stipitis fermentation ofacid-hydrolyzed hemicellulose achieved byoverliming (calcium hydroxide treatment) andstrain adaptation. World. J. Microbiol.Biotechnol. 12(3):281-283.

Antonucci, S., M. Bravi, R. Bubbico, A.D.Michele and N. Verdone. 2001. Selectivity incitric acid production by Yarrowia lipolytica.Enzyme and Microbial Tech. 28(2/3):189-195.

Arzumanov, T.E., N.V. Shishkanova and T.V.Finogenova. 2000. Biosynthesis of citric acidby Yarrowia lipoltyica repeat-batch culture onethanol. Appl. Microbiol. Biotechnol. 53(5):525-529.

Balletsteros, J., J.M. Oliva, M. Ballesteros andJ. Carrasco. 1993. Optimization of thesimultaneous saccharification andfermentation process using thermotolerantyeasts. Appl. Biochem. Biotech. 39/40:201-211.

184 C.A. Abbas

Banat, I.M. and R. Marchant. 1995.Characterization and potential industrialapplications of five novel, thermotolerant,fermentative, yeast strains. World J. Microb.Biotech. 11:304-306.

Barnett, J.A., R.W. Payne and D. Yarrow. 2000.Yeasts: characteristics and identification.University Press, Cambridge, U.K.

Barth, G. and C. Gaillardin. 1996. Yarrowialipolytica. In: Nonconventional Yeasts inBiotechnology (K. Wolf, ed). Springer-VerlagBerlin Heidelberg, New York, pp. 313-388.

Beckerich, J.M., P. Fournier, C. Gaillardin, H.Heslot, M. Rochet and B. Treton. 1984. Yeasts.In: Genetics and breeding of industrialmicroorganisms (C. Ball, ed). CRC Press Inc.,Boca Raton, Florida, USA.

Belem, M.A.F., B.F. Gibbs and B.H. Lee. 1997.Enzymatic production of ribonucleotides fromautolysates of Kluyveromyces marxianusgrown on whey. J. Food Sci. 62(4):851-854.

Belem, M.A.F. and B.H. Lee. 1998. Productionof bioingredients from Kluyveromycesmarxianus grown on whey: an alternative.Critical Rev. Food Sci. Nutr. 38(7):565-598.

Benitez, T., J.M. Gasent-Ramirez, F. Castrejonand A.C. Codon. 1996. Development of newstrains of yeast for the food industry.Biotechnol. Prog. 12:149-163.

Bhosale, P.B. and R.V. Gadre. 2001. Productionof ß-carotene by a mutant of Rhodotorulaglutinis. Appl. Microbiol. Biotechnol. 55:423-427.

Blandin, G, O. Ozier-Kalogeropoulos, P. Winker,F. Artinguenave and B. Dujon. 2000a.Genomic exploration of the hemiasco-mycetous yeasts: 16. Candida tropicalis. FEBSLetters 487(1):91-94.

Blandin, G., B. Llorente, A. Malpertuy, P.Wincker, F. Artinguenave and B. Dujon. 2000b.Genomic exploration of thehemiascomycetous yeasts: 13. Pichia angusta(FEBS 24377) FEBS Letters 487(1):76-81.

Bolotin-Fukuhara, M., C. Toffano-Nioche, F.Artinguenave, G. Duchateau-Nguyen, M.Lemaire, R. Marmeisse, R. Montrocher, C.Robert, M. Termier, P. Wincker and M.Wesolowski-Louvel. 2000. Genomicexploration of the hemiascomycetous yeasts:11. Kluyveromyces lactis (FEBS 24375). FEBSLetters 487(1):66-70.

Boretsky, Y., A. Voronovsky, O. Liuta-Tehlivets,M. Hasslacher, S.P. Kohlwein and G.M.Shavlovsky. 1999. Identification of an ARSelement and development of a high efficiencytransformation system for Pichiaguilliermondii. Curr. Genet. 36(4):215-221.

Buzzini, P. 2000. An optimization study ofcarotenoid production by Rhodotorula glutinisDBVPG 3853 from substrates containingconcentrated rectified grape must as the solecarbohydrate source. J. Industrial Microb.Biotech. 24:41-45.

Buzzini, P. 2001. Batch and fed-batch carotenoidproduction by Rhodotorula glutinis-Debaromyces castelli co-cultures in cornsyrup. J. Appl. Microb. 90:843-847.

Casaregola, S., C. Neuveglise, A. Leplinge, E.Bon, C. Feynerol, F. Artiguenave, P. Winckerand C. Gaillardin. 2000. Genomic explorationof the hemiascomycetous yeasts: 17. Yarrowialipolytica. FEBS Letters 487(1):95-100.

Castoria, R., F. De Curtis, G. Lima and V. DeCicco. 1997. Beta-1, 3-glucanase activity oftwo saprophytic yeasts and possible mode ofaction as biocontrol agents against postharvestdiseases. Postharvest Biol. Tech. 12(3):293-300.

Chan, H.Y. and K.P. Hon. 1999.Growth andcarotenoid production by pH-stat cultures ofPhaffia rhodozyma. Biotech. Letters21(111):953-958.

Chand-Goyal, T. and R.A. Spotts. 1996. Controlof postharvest pear diseases using naturalsaprophytic yeast colonists and theircombination with a low dosage ofthiabendazole. Postharvest Biol. Tech. 7(1/2):51-64.

Chang, C-C., C-S. Park and D.D.Y. Ryu. 1998.Improvement of heterologous proteinproductivity through a selected bioprocessstrategy and medium design-a case study forrecombinant Yarrowia lipolytica fermentation.Appl. Biochem. Biotech. 74(3):173-189.

Chaudhari, A.B. and S.B. Chincholkar. 1999.New osmotolerant Schizosaccharomyces forethanol production. J. Food Sci. Techn., India36(2):166-169.

Choi, J-H., K-H. Moon, Y-W. Ryu and J-H. Seo.2000. Production of xylitol in cell recyclefermentations of Candida tropicalis. Biotech.Letters 22:1625-1628.


Chung, J.D. 2000. Design of metabolic feedcontrollers: application to high-densityfermentations of Pichia pastoris. Biotech.Bioeng. 68(3):298-307.

Cruz, J.M. and J.C. Parajo.1998. Improvedastaxanthin production by Xanthophyllomycesdendrorhous growing on enzymatic woodhydrolysates containing glucose andcellobiose. Food Chem. 63(4):479-484.

Daniel, H-J, R.T.Otto, M. Reuss and C. Syldatk.1998. Sophorolipid production with highyields on whey concentrate and rapeseed oilwithout consumption of lactose. Biotech.Letters 20(8):805-807.

Daniel, H-J., R.T. Otto, M. Binder, M. Reuss andC. Syldatk. 1999. Production of sphorolipidsfrom whey: development of a two-stageprocess with Cryptococcus curvatus ATCC20509 and Candida bombicola ATCC 22214using deproteinized whey concentrates assubstrates. Appl. Microbiol. Biotechnol. 51:40-45.

D’Anjou, M and A.J. Daugulis. 2000. Mixed-feed exponent ional feeding for fed-batchculture of recombinant methylotrophic yeast.Biotech. Letters 22:341-346.

D’Anjou, M and A.J. Dauglis. 2001. A rationalapproach to improving productivity inrecombinant Pichia pastoris. Biotech. Bioeng.72(1):1-11.

De Baets, S. and E.J. Vandamme. 1999. Yeastsas producers of polysaccharides with novelapplication potential. SIM News 49(6):321-328.

De Felice, B., G. Pontecorvo and M. Carfagna.1997. Degradation of waste waters from oliveoil mills by Yarrowia lipolytica ATCC 20255and Pseudomonas putida. Acta Biotechnol.17(3):231-239.

Delgenes, J.P., M.C. Escare, J.M. Laplace, R.Moletta and J.M. Navarro. 1998. Biologicalproduction of industrial chemicals, i.e. xylitoland ethanol from lignocelluloses by controlledmixed culture systems. Industrial Crops andProducts. 7(2/3):101-111.

Demain, A.L., H.J. Phaff and C.P. Kurtzman.1998. The industrial and agriculturalsignificance of yeasts. In: The Yeasts, ATaxonomic Study (C.P. Kurtzman and J. W.Fell, eds). 4th edition, Elsevier, Amsterdam, TheNetherlands, pp. 13-19.

De Montigy, J., M-L. Straub, S. Potier, F. Tekaia,B. Dujon, P. Wincker, F. Artiguenave and J-L.Souciet. 2000. Genomic exploration of thehemiascomycetous yeasts: 8.Zygosaccharomyces rouxii. FEBS Lett.487(1):52-55.

Deshpande, M. and L. Daniels. 1995. Evaluationof sophorolipid biosurfactant production byCandida bombicola using animal fat.Bioresource Technology 54:143-150.

De Winde, J.H. (ed.). 2003. Topics in CurrentGenetics: Functional Genetics of IndustrialYeasts. Springer-Verlag, Berlin, Germany, pp.1-367.

Dhamija, S.S., D. Bhaskar and R. Gera. 1996.Thermotolerance and ethanol production (from15% sucrose at 37°C) are at variance in mutantsof Schizosaccharomyces pombe. Biotech.Letters 18(11):1341-1344.

Dujon, B. 2003. Comparative genomics ofHemiascomycetous yeasts: the systematicsequencing of Candida glabrata,Kluyveromyces lactis, Debaromyces hanseniiand Yarrowia lipolytica. Plenary lecture 2-5abstract.

Fabritius, D., H.J. Schaefer and A. Steinbuechel.1996. Identification and production of cis-3-hydroxy-1, 18-octadec-9-enedioic acid bymutants of Candida tropicalis. Appl.Microbiol. Biotechnol. 45(3):342-348.

Ferrari, M.D., L. Loperena and H. Varela. 1996.Ethanol production from whey permeate usingfed-batch culture of Kluyveromyces fragilis.Biotech. Letters 16(2):205-210.

Fleet, G. 2000. The biodiversity of yeasts in theproduction of foods and beverages. Plenarypresentation abstract, ISY 2000, Arnhem, TheNetherlands.

Florencio, J.A., C.R. Soccol, L.F. Furlanetto andT. Bonfim.1998. A factorial approach for asugarcane juice-based low cost culturemedium: increasing the astaxanthin productionby the red yeast Phaffia rhodozyma.Bioprocess Eng.19(3):161-164.

Fontana, J.D., M.B. Chocial, M. Baron, M.F.Guimaraes, M. Maraschin, C. Ulhoa, J.A.Florencio and T.B. Bonfim. 1997.Astaxanthinogenesis in the yeast Phaffiarhodozyma-optimization of low-cost culturemedia and yeast cell-wall lysis. Appl. Biochem.Biotech. 63-65:305-314.

186 C.A. Abbas

Freer, S.N., C.D. Skory and R.J. Bothast. 1998.Production of fuel ethanol from cellulosicbiomass. Recent Res. Dev. Microbiol.2(1):201-210.

Fukuda, H., T.Fujii and T. Ogawa. 1984.Microbial production of C-2 hydrocarbons,ethane, ethylene and acetylene. Agric. Biol.Chem. 48(5):1363-1365.

Gabrinska-Loniewska, A., E. Pajor and E.Slavikova. 1997. Comparative analysis ofyeast-like species potentially useful toenvironmental biotechnology. 1997. ActaMycologica 32(1):99-106.

Gellissen, G. 2000. Heterologous proteinproduction in methylotrophic yeasts. Appl.Microb. Biotech. 54(6):741-750.

Giga-Hama, Y. 1997. Fission yeast Schizo-saccharomyces pombe: an attractive host forheterologous protein production. In: ForeignGene Expression in Fission Yeast:Schizosaccharomyces pombe (Y. Giga-Hamaand H. Kumagai, eds). Springer-Verlag,Germany, pp. 3-28.

Giuseppe, L., D.C. Filippo, C. Raffaello and D.C.Vincenzo. 1998. Activity of the yeastsCryptococcus laurentii and Rhodotorulaglutinis against post-harvest rots on differentfruits. Biocontrol Sci. Tech. 8(2):257-267.

Grbin, P.R. and P.A. Henschke. 2000. Mousyoff-flavour production in grape juice and wineby Dekkera and Brettanomyces yeasts. Aust.J. Grape Wine Res. 6(3):255-262.

Hansen, H. and C.P. Hollenberg. 1996.Hansenula polymorpha (Pichia angusta). In:Nonconventional Yeasts in Biotechnology (K.Wolf, ed). Springer-Verlag, Berlin, Germany,pp. 293-311.

Hassan, M.A. and S.H. El-Sharkway. 1999.Microbial production of useful fatty acids fromused sunflower oil. Zagazig J. Pharm. Sci.8(2):46-49.

Hayashida, Y., K. Nishimura and J.C. Slaughter.1998.The importance of the furanones HDMKand HEMF in the flavour profile of Japanesebarley miso and their production duringfermentation. J. Food Sci. Agric. 78(1):88-94.

Hollenberg, C.P., and G. Gellissen.1997.Production of recombinant proteins bymethylotrophic yeasts. Current Opinions inBiotechnology 8:554-560.

Huang, C. Jr., S.L. Lee and C.C. Chou. 2000.Production and molar yield of 2-phenylethanolby Pichia fermentans L-5 as affected by somemedium components. J. Biosc. Bioeng.90(2):142-147.

Kesava, S.S., G-H. An, C-H. Kim, S-K. Rheeand E-S. Choi. 1998. An industrial mediumfor improved production of carotenoids froma mutant strain of Phaffia rhodozyma.Bioprocess Eng. 19(3):165-170.

Kim, J-H., Y-W. Ryu and J-H. Seo. 1999.Analysis and optimization of a two-substratefermentation for xylitol production usingCandida tropicalis. J. Indust. Microb. Biotech.22(3):181-186.

Kruse, B. and K. Schuegerl. 1996. Investigationof ethanol formation by Pachysolentannophilus from xylose and glucose/xyloseco-substrates. Process Biochem. 31(4):389-407.

Kunze, G. and I. Kunze. 1996. Arxulaadeninivorans. In: Nonconventional Yeasts inBiotechnology (K. Wolf, ed). Springer, Berlin,Germany, pp. 389-409.

Kurtzman, C.P. and J.W. Fell. 1998. The Yeasts,A Taxonomic Study (C.P. Kurtzman and J. W.Fell, eds), 4th edition. Elsevier, Amsterdam, TheNetherlands.

Kurtzman, C. 2001. Six new anamorphicascomycetous yeasts near Candida tanzawaensis. FEMS Yeast Research 1:177-185.

Kurtzman, C. and J. Sugiyama. 2001.Ascomycetous yeasts and yeastlike taxa. In:The Mycota VII Part A. Systematics andEvolution (McLaughlin, McLaughlin andLemke, eds.) Springer-Verlag, Berlin,Germany, pp. 179-200.

Kyong, S.H. and C.S. Shin. 2000. Optimizedproduction of L-ß-hydroxybutyric acid by amutant of Yarrowia lipolytica. Biotechnol. Lett.22(13):1105-1110.

Labrou, N. and D.J. Rigden. 2001. Active-sitecharacterization of Candida boidinii formatedehydrogenase. Biochem. J. 354(2):455-463.

Lachance, M-A. and W.T. Starmer. 1998.Ecology and yeasts. In: The Yeasts, ATaxonomic Study (C.P. Kurtzman and J. W.Fell, eds), 4th edition. Elsevier, Amsterdam, TheNetherlands, pp. 21-30.


Laires, A., I. Spencer-Martins and N. van Uden.1983. Use of D-glucosamine and 2-deoxyglucose in the selective isolation ofmutants of the yeast Lipomyces starkeyiderepressed for the production of extracellularendodextranase. Zeitschrift fur AllgemeineMikrobiologie 23(9):601-603.

Lang, S., A. Brakemeier, U. Rau and F. Wagner.1996. Biotechnological production ofinterfacial active glycolipids. Oils-Fats-Lipids1995, Proc. World. Congr. Int. Soc. Fat. (21st)1:53-54.

Lark, N., Y. Xia, C-G. Qin, C.S. Gong and G.T.Tsao. 1997. Production of ethanol fromrecycled paper sludge using cellulase andyeast, Kluyveromyces marxianus. Biomass andBioenergy 12(2):135-143.

Lepingle, A., S. Casaregola, C. Neuveglise, E.Bon, H-V. Nguyen, F. Artiguenave, P. Winckerand C. Gaillardin. 2000.Genomic explorationof the hemiascomycetous yeasts: 14.Debaromyces hansenii var hansenii. FEBSLetters 487(1):82-86.

Liiv, L., P. Parn and T. Alamae. 2001. Cloningof the maltase gene from methylotrophic yeast,Hansenula polymorpha. Gene 265(1/2):77-85.

Lim, S.H., J.S. Choi and Y.E. Park. 2001.Microbial production of riboflavin usingriboflavin overproducers, Ashbya gossypii,Bacillus subtilis and Candida famata. Anoverview. Biotechnol. Bioprocess. Eng.6(2):75-88.

Lima, G., F. De Curtis, R. Castoria and V. DeCicco. 1998. Activity of the yeastsCryptococcus laurentii and Rhodotorulaglutinis against post-harvest rots in differentfruits. Biocontrol Sci. Tech. 8(2):257-267.

Limtong, S., S. Deejing and W. Santisoparsi.1998. Construction of high ethanolhalotolerant hybrid by intergeneric protoplastfusion of Saccharomyces cerevisiae andZygosaccharomyces rouxii. Kaseterat J. Nat.Sci. 32(2):213-223.

Linardi, V.R., J.C.T. Dias and C.A. Rosa. 1996.Utilization of acetonitrile and other aliphaticnitriles by a Candida famata strain. FEMSMicrobiol. Lett. 144(1):67-71.

Llorente, B., A. Malpertuy, G. Blandin, F.Artinguenave, P. Wincker and B. Dujon. 2000.Genomic exploration of the

hemiascomycetous yeasts: 12. Kluyveromycesmarxianus var. marxianus (FEBS 24376) FEBSLetters 487(1):71-75.

Martinez, X.C., A. Narbad, A.T. Carter and M.Stratford. 1996. Flocculation of the yeastCandida famata (Debaromyces hansenii): anessential role for peptone. Yeast 2(5):415-423.

Matelli, H.L., I.M. da Silva, N.O. Souza and D.Pomeroy. 1990. Production of ß-carotene bya Rhodotorula strain grown on sugar canejuice. Biotechnology Letters 12(3):207-208.

Matsuyama, A., H. Yamamoto, N. Kawada andY. Kobayashi. 2001. Industrial production ofR-1,3-butanediol by new catalysts. J. Mol.Catal. B: Enzym. 11:4-6.

Mauersberger, S., M. Ohkuma, W-H. Schunckand M. Takagi. 1996. Candida maltosa. In:Nonconventional Yeasts Biotechnol (K. Wolf,ed.). Springer, Berlin, Germany, pp. 411-580.

Mayer, A.F., K. Helmuth, H. Schlieker, R. Lopez-Ulibarri, S. Oertel, U. Dahlems, A.W.M.Strasser and A.P.G.M. van Loon. 1999. Anexpression system matures: a highly efficientand cost-effective process for phytaseproduction by recombinant strains ofHansenula polymorpha. Biotech. Bioeng.63(3):373-381.

Minning, S., A. Serrano, P. Ferrer, C. Sola, R.D.Schmid and F. Valero. 2001. Optimization ofthe high-level production of Rhizopus oryzaelipase in Pichia pastoris. J. Biochem. 86(1):59-70.

Miura, Y., K. Kondo, T. Saito, H. Shimada, P.Fraser and N. Misawa. 1998. Production of thecarotenoids lycopene, ß-carotene andastaxanthin in the food yeast Candida utilis.Appl. Envt. Microbiol. 64(4):1226-1229.

Miura, Y., K. Kondo, H. Shimada, T. Saito, K.Nakamura and N, Misawa. 1998. Productionof lycopene by the food yeast, Candida utilisthat does not naturally synthesize carotenoid.Biotech. Bioeng. 58(2/3):306-308.

Miura, Y., M. Kettoku, M. Kato, K. Kobayashiand K. Kondo. 1999. High level productionof thermostable α-amylase from Sulfolobussolfataricus in high-cell density culture of thefood yeast Candida utilis. J. Mol. Microbiol.Biotechnol. 1(1):129-134.

Montesinos, J.L., E. Dalamu and C. Casas. 2003.Lipase production in continuous culture of

188 C.A. Abbas

Candida rugosa. J. Chem. Tech. Biotech.78:753-761.

Nadezda, K., M. Maria and C. Peter. 1997.Structure of glucomannan-protein complexfrom the yeast Cryptococcus laurentii. J.Carboh. Chem. 16(5-4):609-623.

Nakamura, T., T. Suzuki, J. Junko, N. Kato, YSakai, D. Mochizuki and H. Takahashi. 2000.Production of recombinant phytase by fed-batch fermentation of Candida boidinii. USPatent 6,140,077.

Nakano, K., R. Katsu, K. Tada and M.Matsumura. 2000. Production of highlyconcentrated xylitol by Candida magnoliaeunder a microaerobic condition maintained bya simple fuzzy control. J. Bios. Bioeng.89(4):372-376.

Nielsen, J. and L. Olsson. 2002. An expandedrole for microbial physiology in metabolicengineering and functional genomics: movingtowards systems biology. FEMS YeastResearch 2:175-181.

Nishikawa, A., T. Sugita and T. Shinoda. 1997.Differentiation between Debaromyceshansenii/Candida famata complex andCandida guilliermondii by polymerase chainreaction. FEMS Immunol. Med. Microbiol.19(2):125-129.

Nishikawa, A., T. Sugita and T. Shinoda. 1999.Rapid identification of Debaromyces hansenii/Candida famata by polymerase chain reaction.Med. Mycol. 37(2):101-104.

Oh, D-K. and S-Y. Kim. 1998. Increased xylitolyield by feeding xylose and glucose inCandida tropicalis. Appl. Microbiol. Biotech.50(4):419-425.

Panchall, C.J., I. Russell, A.M. Sills and G.G.Stewart. 1984. Genetic manipulation ofbrewing and related yeast strains. Food Tech.,February, pp. 99-111.

Parajo, J.C., V. Santos and M. Vasquez. 1998a.Optimization of carotenoid production byPhaffia rhodozyma cells grown on xylose.Process Biochem. 33(2):181-187.

Parajo, J.C., V. Santos and M. Vazquez. 1998b.Production of carotenoids of Phaffiarhodozyma growing on media made fromhemicellulosic hydrolyzates of Eucalyptusglobules wood. Biotechnol. Bioeng. 59(4):501-506.

Parrondo, J., L.A. Garcia and M. Diaz. 2000.Production of an alcoholic beverage byfermentation of whey permeate withKluyveromyces fragilis. I. Primary metabolism.J. Inst. Brewing 106(6):367-375.

Pazouki, M., P.A. Felse, J. Sinha and T. Panda.2000. Comparative studies on citric acidproduction by Aspergillus niger and Candidalipolytica using molasses and glucose.Bioprocess Eng. 22(4):353-361.

Penttila, M. 1997. In: Biotechnology in the foodchain. VTT Symposium 177. pp. 109-118.

Phaff, H.J. 1985. Biology of yeasts other thanSaccharomyces. Biotech. Series 6:537-562.

Picataggio, S., T. Rorer, K. Deanda, D. Lawning,R. Reynolds, J. Mielenz and L.D. Dudley.1992. Metabolic engineering of Candidatropicalis for the production of long-chain fattyacids. Biotech. 10:894-898.

Possoth, V., M. Zimmermann and U. Klinner.1996. Peculiarities of the regulation offermentation and respiration in the Crabtree-negative, xylose-fermenting yeast Pichiastipitis. Appl. Biochem. Biotechnol. 57/58:201-212.

Pribylova, L. and H. Sychrova. 2003. Osmoticstress resistant yeast Zygosaccharomycesrouxii-isolation of auxotrophic mutants andefficient transformation by electroporation.Abstract 11-7. XXIst International Conferenceon Yeast Genetics and Molecular Biology,Gothenburg, Sweden.Yeast 20(S1).

Ramirez, J., M.L.Nunez and R. Valdivia. 2000.Increased astaxanthin production by a Phaffiarhodozyma mutant grown on date juice fromYucca fillifera. J. Indust. Microb. Biotech.24(3):187-190.

Ramos, J. V. Montiel, C. Prista, R. Garcia-Salcedo and M.C. Loureiro-Dias. 2003.Abstract 11-46. Yeast 20(S1), XXIInternational Conference on Yeast Geneticsand Molecular Biology, Gothenburg, Sweden.

Rau, U., R. Heckmann, V. Wray and S. Lang.1999. Enzymatic conversion of a sophorolipidinto a glucose lipid. Biotech. Letters 21:973-977.

Rau, U., S. Hammen, R. Heckmann, V. Wrayand S. Lang. 2001. Sophorolipids: a sourcefor novel compounds. Industrial Crops andProducts 13:85-92.


Rodriguez, E., K. Sanchez, H. Roca, B. Garcia,J. Cremata and J. Delgado. 1996. Procedurefor the high-level recombinant dextranaseproduction. Biotecnol. Apl. 13(3):201.

Rodrigues, D.C.G.A., S.S. Silva, A.M.R. Prataand M.G.A. Felipe. 1998. Biotechnologicalproduction of xylitol from agroindustrialresidues. Evaluation of bioprocesses. Appl.Biochem. Biotech. 70-72:869-875.

Roostita, R. and G.H. Fleet. 1999. Growth ofyeasts isolated from cheese on organic acidsin the presence of sodium chloride. FoodTechnol. Biotechnol. 37(2):73-79.

Saha, B.C. and R.J. Bothast. 1996. Productionof L-arabitol from L-arabinose by Candidaentomaea and Pichia guilliermondii. Appl.Microbiol. Biotechnol 45(3):299-306.

Sakai, Y, M. Akiyama, H. Kondoh, Y. Shibanoand N. Kato. 1996. High-level secretion offungal glucoamylase using the Candidaboidinii gene expression system. Biochimicaet Biophysica Acta 1308(1):81-87.

Sakai, Y., Y. Tani and N. Kato. 1998.Biotechnological application of cellularfunctions of the methylotrophic yeast. J.Molecular Catal. B Enzymatic 6(3):161-173.

Sanchez, S., V. Bravo, E. Castro, A.J. Moya andF. Camacho. 1998. The production of xylitolfrom D-xylose by fermentation withHansenula polymorpha. Appl. Microb. andBiotech. 50(5):608-611.

Sanchez, S., V.Bravo, E. Castro, A.J. Moya andF. Camacho. 1999. Comparative study of thefermentation of D-glucose/D-xylose mixtureswith Pachysolen tannophilus and Candidashehatae. Bioprocess Eng. 21(6):525-532.

Sasahara, H., M. Mine and K. Izumori. 1998.Production of D-talitol from D-psicose byCandida famata R28. J. Ferment. Bioeng.85(1):84-88.

Santopietro, L.M.D., J.F.T. Spencer and F.Sineriz. 1998. Fed-batch and continuousculture of Phaffia rhodozyma (Xantho-phyllomyces dendrorhous). Folia Microbiol.43(2):169-172.

Sberna, G., R. Cappai, A. Henry and D.H. Small.1996. Advantages of the methylotrophic yeastPichia pastoris for high-level expression andpurification of heterologous proteins.Australas. Biotechnol. 6(2):82-87.

Schwan, R.F., R.M. Cooper and A.E. Wheals.1997. Endopolygalacturonase secretion byKluyveromyces marxianus and other cocoapulp-degrading yeasts. Enzyme and MicrobialTech. 21:234-244.

Sharyshev, A.A., E.V. Peskova and G.N.Komarova. 1997.Effect of cultivationconditions on the level of enzymes of n-alkanemetabolism in Candida maltosa cells.Microbiology (Moscow) 66(6):652-656.

Shimada, H., K. Kondo, P. Fraser, Y. Miura, T.Saito and N. Misawa. 1998. Increasedcarotenoid production by the food yeastCandida utilis through metabolic engineeringof the isoprenoid pathway. Appl. Envt.Microbiol. 64(7):2676-2680.

Sibirny, A.A. 1996a. Pichia guilliermondii. In:Nonconventional Yeasts in Biotechnology (K.Wolf, ed). Springer-Verlag, Berlin, Heidelberg,New York, pp. 255-275.

Sibirny, A. 1996. Pichia methanolica (Pichiapinus MH4). In: Nonconventional yeasts inbiotechnology (K. Wolf, ed.) Springer-Verlag,Berlin, Germany, pp. 277-291.

Silva, S.S., I.C. Roberto, M.G.A. Felipe and I.M.Mancilha. 1996. Batch fermentation of xylosefor xylitol production in a stirred tankbioreactor. Process Biochem. 31(6):549-553.

Singh, D., P. Nigam, I.M. Banat, R. Marchantand A.P. McHale. 1998. Review: ethanolproduction at elevated temperatures andalcohol concentrations: Part II-use ofKluyveromyces marxianus IMB3.World J.Microb. Biotech. 14(6):823-834.

Skory, C.D., S.N. Freer and R.J. Bothast. 1996.Properties of an intracellular beta-glucosidasepurified from the cellobiose-fermenting yeastCandida wickerhamii. Appl. Microbiol.Biotechnol. 46(4):353-359.

Sreekrishna, K. and K.E. Kropp. 1996. Pichiapastoris. In: Nonconventional Yeasts inBiotechnology (K.Wolf, ed). Springer-Verlag,Berlin, Germany, pp. 203-253.

Sreenath, H.K. and T.W. Jeffries. 1999.Production of ethanol from wood hydrolyzateby yeasts. Bioresour. Technol. 72(3):253-260.

Stahmann, K-P, J.L. Revuelta and H. Seulberger.2000. Three biotechnical processes usingAshbya gossypii, Candida famata, or Bacillussubtilis compete with chemical riboflavin

190 C.A. Abbas

production. Appl. Microbiol. Biotechnol.53(5):509-516.

Suryadi, H., T. Katsuragi, N. Yoshida, S. Suzukiand Y. Tani. 2000. Polyol production by cultureof methanol-utilizing yeast. J. Biosc. Bioeng.89(3):236-240.

Sugar, D. and R.A. Spotts. 1999. Control ofpostharvest decay in pear by four laboratory-grown yeasts and two registered biocontrolproducts. Plant Disease 83(2):155-158.

Sung-Wook, Y., J-B. Park, N.S. Han, Y-W. Ryuand J-H. Seo. 1999. Production of erythritolfrom glucose by an osmophilic mutant ofCandida magnoliae. Biotech. Letters21(10):887-890.

Sychrova, H., V. Braun, S. Potier and J-L.Souciet. 2000. Organization of specificgenomic regions of Zygosaccharomyces rouxiiand Pichia sorbitophila: comparison withSaccharomyces cerevisiae. Yeast 16(15):1377-1385.

Tanaka, A., R. Yamada, S. Shimizu and S. Fukui.1971. Studies on the formation of vitamins andtheir functions in hydrocarbon fermentations.J. Ferment. Technol. 49(9):792-802.

Thorpe, E.D., M. d’Anjou and A.J. Daugulis.1999. Sorbitol as a non-repressing carbonsource for fed-batch fermentation ofrecombinant Pichia pastoris. Biotech. Letters21:669-672.

Trujillo, L.E., J.G. Arrieta, F. Dafhnis, J. Garcia,J. Valdes, Y. Tambara, M. Perez and L.Hernandez. 2001. Fructo-oligosaccharidesproduction by the Gluconacetobacterdiazotrophicus levansucrase expressed in themethylotrophic yeast Pichia pastoris. Enzymeand Microbial Tech. 28(2/3):139-144.

Vandeska, E., S. Amartey, S. Kuzmanova andT.W. Jeffries. 1996. Fed-batch culture forxylitol production by Candida boidinii .Process Biochem. 31(3):265-270.

Vazquez, M., V. Santos and J.C.Parajo.1998.Fed-batch cultures of Phaffia rhodozyma inxylose-containing media made from woodhydrolyzates. Food Biotechnol. 12(1/2):43-55.

Vijayakumar, A.R., E. Csoeregi, A. Heller andL. Gordon. 1996. Alcohol biosensors basedon coupled oxidase-peroxidase systems. Anal.Chim. Acta 327(3):223-234.

Visser, H., J. Verdoes, G. Sandmann and J. vanden Berg. 2003. Genetical engineering of the

carotenoid biosynthetic pathway inXanthophyllomyces dendrorhous. Abstract 12-46.Yeast 20(S1). XXI International Conferenceon Yeast Genetics and Molecular Biology,Gothenburg, Sweden.

Voronovsky, A., C.A. Abbas, L.R.Fayura, B.V.Kshanovska, K.V. Dmytruk, K.A. Sybirna andA.A. Sibirny. 2002. Development of atransformation system for the flavinogenicyeast Candida famata FEMS. Yeast Res.2(3):381-388.

Wache, Y., M. Aguedo, J-M. Nicaud and J-M.Belin. 2003. Catabolism of hydroxyacids andbiotechnological production of lactones byYarrowia lipolytica. Appl. Microbiol. Biotech.61:393-404.

Wartmann, T. and G. Kunze. 2000. Genetictransformation and biotechnologicalapplication of the yeast Arxula adeninvorans.Appl. Microb. Biotech. 54(5):619-624.

Wery, J., J.C. Verdoes and A.J.J. Van Ooyen.1998. Efficient transformation of theastaxanthin-producing yeast Phaffiarhodozyma. Biotechno. Tech. 12(5):399-405.

Wesolowski-Louvel, M., K. Breunig and H.Fukuhara. 1996. Kluyveromyces lactis. In:Nonconventional Yeasts in Biotechnology (K.Wolf, ed). Springer-Verlag, Berlin, Heidelberg,New York, pp. 139-201.

Wolf, K. 1996. In: Nonconventional Yeasts inBiotechnology. A Handbook. Springer-Verlag,Berlin, Heidelberg, New York.

Wolf, K., K. Breunig and G. Barth (eds). 2003.Non-conventional Yeasts in Genetics,Biochemistry and Biotechnology. Springer-Verlag, Berlin, Germany. pp. 1-494.

XXI International Conference on Yeast Geneticsand Molecular Biology, July 7-12,Gothenburg, Sweden. Published in Yeast20(S1) (J.M. Cherry and S. Hohmann, eds).

Yahashi, Y., H. Horitsu, K. Kawai, T. Suzuki andK. Takamizawa. 1996. Production of xylitolfrom D-xylose by Candida tropicalis: theeffect of D-glucose feeding. J. Ferment.Bioeng. 81(2):148-152.

Yamada, R-H, Y. Kera, H. Toi, T. Hayashi, K.Arimoto, M. Takahashi, I. Iwazaki and S.Yamashita. 2001. Microbial oxidases of acidicD-amino acids. J. Molecular Catalysis B:Enzymatic 12(1-6):93-104.


Yamazaki,T., A. Kubiki, H. Ota, T.M. Mitzel,L.A. Paquette and T. Sugai. 2000.Yamadazyma farinosa IFO 10896-mediatedreduction of 4,4-dimethoxy-2-butanone as thekey step for the preparation of 1,3- diols withunsymmetrical substituents. Synth. Commun.30(16):3061-3072.

Yanai, T. and M. Sato. 2000. Purification andcharacterization of an α-L-rhamnosidase fromPichia angusta X349. Bioscience, Biotech.Biochem. 64(10):2179-2185.

Yurimoto, H., T. Komeda, C.R. Lim, T.Nakagawa, K. Kondo, N. Kato and Y. Sakai.2000. Regulation and evaluation of fivemethanol-inducible promoters in themethylotrophic yeast Candida boidinii.Biochimica et Biophysica Acta 1493(1-2):56-63.

Beverage alcohol production

Production of Scotch and Irish whiskies: their history and evolution 193

Chapter 14

Production of Scotch and Irish whiskies: their history and evolution

T. PEARSE LYONS


Introduction

Whisky is the potable spirit obtained bydistillation of an aqueous extract of an infusionof malted barley and other cereals that has beenfermented with strains of Saccharomycescerevisiae yeast. Various types of whisky areproduced in a number of different countries inthe world. They differ principally in the natureand proportion of the cereals used as rawmaterials along with malted barley, and also inthe type of still used for distillation.

The principal types of whisky are alsocharacteristic of particular geographical regionsof the world. In Scotland, the characteristicproduct is manufactured using only maltedbarley as the raw material; and the fermentedmalt mash is distilled in relatively small pot stills.

The product, known as Scotch malt whisky, isproduced in small distilleries, of which there areover 100 in Scotland. Scotch malt whisky ismarketed both as a straight malt whisky, manybrands of which have recently becomeextremely popular throughout the world, andalso as a blend with another type of whiskyproduced in Scotland known as ‘Scotch grainwhisky’ or (because it is distilled continuouslyin Coffey-type patent stills) as ‘patent-stillwhisky’. Most Scotch whiskies available on theinternational market consist of blends with 20-30% malt whisky and 70-80% grain whisky.

Within the blend, there may be as many as 20-30 individual malt whiskies and grain whiskies.These blends are, by law, matured for at leastthree years but in practice this period is muchlonger. Unblended Scotch malt whiskies areusually matured for a minimum of eight years.

The cereals used in the manufacture of Scotchgrain whisky are malted barley, together with ahigh proportion (up to 90%) of wheat or corn(maize). Currently wheat is the main cereal,chosen on the basis of cost and the attraction ofusing a Scottish-grown cereal. All whiskies arelegally protected and defined, mainly becauseof the huge revenues that governments obtainfrom their sale. The Scotch Whisky Order (1990)and the Scotch Whisky Act (1988), in definingScotch Whisky, state that to be called ScotchWhisky spirits must be:

1. produced at a distillery in Scotland

2. produced from water and malted barley towhich only whole grains or other cereals maybe added

3. processed at that distillery into a mash

4. converted to fermentable carbohydrate onlyby endogenous enzymes

5. fermented only by the addition of yeast

194 T.P. Lyons

6. distilled to an alcohol strength less than94.8% so that the distillate has the aroma andtaste derived from the raw materials

7. matured in Scotland in oak casks of less than700 litres for a minimum of three years

8. be a product that retains the color, aroma andtaste derived from the raw materials

9. produced with no substance other than waterand spirit caramel added

The word ‘Scotch’ in this definition is ofgeographical and not generic significance.

Irish whiskey (spelled with an e) is a distinctiveproduct of either the Republic of Ireland or ofNorthern Ireland. In the Republic of Ireland,definitions were enacted by the Parliament inthe Irish Whiskey Acts of 1950 and 1980. The1950 Act distinguished pot still whiskey fromblends and stated that the title ‘Irish Pot-StillWhiskey’ was reserved solely for spirits distilledin pot stills in the Republic from a mash of cerealgrains normally grown in that country andsaccharified by a diastase of malted barley. The1980 legislation specified that the term ‘IrishWhiskey’ only applied to ‘spirits distilled in theRepublic or Northern Ireland from a mash ofcereals saccharified by the diastase of maltcontained therein, with or without other naturaldiastases’. This meant that unlike Scotch, Irishwhiskey may be produced with the use ofmicrobial enzyme preparations in addition tomalt. The 1980 Act also specified that thewhiskey must be aged for at least three years inIreland in wooden barrels. While not possessingthe ‘smoky’ taste and aroma of Scotch, Irishwhiskey is usually more flavorful and has aheavier body than Scotch. Moreover, thewhiskey is distilled not twice, as in Scotland,but three times to give a very strong spirit of 86o

GL compared with the 71o GL whisky distilledin Scotland.

The tremendous popularity of whiskiesmanufactured in Scotland, Ireland, the US andCanada has prompted several other countries totry manufacturing whiskies, usually onesdesigned to resemble Scotch. Indeed, the numberof countries with minor but neverthelesssignificant whisky-distilling industries must nowbe well over a dozen. Some countries, notablyAustralia and Japan, have whisky-distilling

industries producing a sufficiently acceptableproduct for them to venture into the exportindustry. Other countries, including theNetherlands and Spain, have whisky-distillingindustries that cater mainly, if not exclusively,for home consumption. The measures that someof these industries have taken to imitate Scotchbecome apparent when the spirits are sampled.One of the two whisky distilleries in Spainlocated northwest of Madrid in the GuadarramaHills near Segovia produces a very acceptableScotch-type whisky. Its quality is attributed inpart to the fact that the water used, which comesfrom the surrounding hills, closely resembles thatused in highland Scotch whisky distilleries.

Recommended descriptions of Scotch whiskymanufacture have come from Brander (1975),Daiches (1969), Dunnett (1953), Gunn (1935),Robb (1950), Ross (1970), Simpson (1968),Simpson et al. (1974), Ross Wilson (1959, 1970and 1973) and Philip (1989). There are a fewsmaller books describing individual distilleriesin Scotland. Recommended are those byMcDowell (1975) and Wilson (1973; 1975).Recommended reading on Irish whiskeyincludes an account by McGuire (1973) andshort review articles by Court and Bowers (1970)and Lyons (1974).

History of whisky production

The origins of the art of distilling potable spiritsare obscure, and probably date back to ancientChina. However, the first treatise on distillingwas written by the French chemist Arnold deVilleneuve around 1310. The potency of distilledspirits caused many to be known as the ‘waterof life’, a description that survives today in suchnames as eau de vie for French brandy andakvavit and aquavit for spirits in NorthernEurope. The name ‘whisky’ is a corruption of‘uisgebaugh’, the Gaelic word for water of life.Uisge was corrupted first into ‘usky’, whichfinally became whisky after several centuries.Dr. Johnson sang the praises of this potable spirit,although in his Dictionary of 1755 it is listedunder ‘u’ and not ‘w’.

Much to the chagrin of the Scotsman, it islikely that the first whisky was distilled not inScotland but in Ireland. The spirit was known inIreland when that land was invaded by the


English in 1170. In all likelihood the art ofdistillation was imported into Scotland bymissionary monks from Ireland. Two of today’smain centers of Scotch distilling, namely theisland of Islay and the Speyside town ofDufftown, were the sites of early monasticcommunities.

Whisky, principally Scotch whisky, has formany years been one of the most populardistilled beverages in the world; and it was inScotland rather than in Ireland that its qualitiescame to be extensively appreciated. This hascontinued to the present day and in theintervening period many Scotsmen have feltcompelled to record for posterity their thoughtsand inspirations on the potable spirit. There arenumerous histories of whisky distilling inScotland, some more comprehensive than others.For good general accounts, the reader is referredto the texts by Brander (1975), Daiches (1969),Ross (1970) and Ross Wilson (1970).

Whisky distilling flourished in Scotland notleast because consuming the spirit helped theinhabitants to withstand the climatic rigors ofthis northern region of Britain. The first recordedevidence of whisky production in Scotland isan entry in the Exchequeur Rolls for the year1494. It reads “To Friar John Cor, by order ofthe King, to make aquavitae, eight bolls of malt”.Production of whisky was therefore beingcontrolled; and an Act of 1597 decreed that onlyearls, lords, barons and gentlemen could distillfor their own use. To many Scots of this erawhisky was a medicine, and in 1506 King JamesIV of Scotland had granted a monopoly formanufacture of ‘aqua vitae’ to the Guild ofBarber Surgeons in the City of Edinburgh.

Taxation on whisky production first appearedin the 17th century. Breaches of the monopolyregulations and the need to raise money to sendan army into England to help the EnglishParliament in its war against Charles I led to theAct of 1644, which fixed a duty of two shillingsand eight pence Scot’s on a pint of whisky (theScot’s pint was then about 1.5 litres). But thetax was short-lived and was replaced by a malttax that later was also repealed.

At the time of the Treaty of Union betweenScotland and England in 1707 there was a taxon malt in England, but not in Scotland. TheEnglish were irate; and in 1725 when LordWalpole’s administration decided to enforce thetax in Scotland, the first of a series of Malt Tax

riots occurred. The English, meanwhile, hadcultivated a taste for French brandy, there beingvery little whisky consumed at that time outsideof Scotland. However around 1690 William IIIbegan to wage commercial war against theFrench, and imposed punitive taxes on importsof French brandy into England. The Englishreacted by acquiring a taste for gin, which wasdistilled locally. The scale of drunkenness thatdeveloped with the popularity of gin had to becontrolled by law; and the Acts of 1736 and 1713levied high taxes on gin manufacturers. Both ofthese acts contained clauses exempting Scotland,but not for long. The Parliament in London sawthe prospect of a rich harvest of taxes in thedistilleries of Scotland; and in a series of actsstarting in 1751, production of whisky inScotland was increasingly subjected to taxation.

The outcome of these punitive measures wasnot surprising. An extensive and thrivingbusiness in illicit distillation of whisky grew upin Scotland as described by Sillett (1965).Curiously, illicit production of Scotch hardlyextended over the border into England, althoughthere are a few records of the operation of illicitstills in the Cheviot Hills west of Newcastle-upon-Tyne. Following the Act of 1823, whichintroduced much stiffer penalties for illicitdistillation, and to some extent because of theincreased standards of living in northernScotland, illicit manufacture of whisky declined.Indeed, many erstwhile illicit distillers emergedto become legal and registered distillers ofScotch whisky.

In 1826, Robert Stein of the Kilbagie distilleryin Clackmannanshire, Scotland, patented acontinuously operating still for whiskyproduction. However, this invention wassuperseded in 1830 with the introduction byAeneas Coffey of an improved version of thistype of still. The appearance of continuous stillssparked off a period of turmoil in the Scotchwhisky industry, it being claimed that the productfrom the continuous distillation of a mash thatcontained unmalted grain (described as neutralor ‘silent spirit’) could not be called whisky,since it had not been distilled in the traditionalpot still. The battle was waged for about threequarters of a century; and in 1908 a RoyalCommission decided that malt whisky and grainneutral spirit, when blended, could be labeledwhisky.

196 T.P. Lyons

The major factors which have affected thedevelopment of the whisky distilling industryin Scotland in this century have been economic.The industry has had to endure the privations oftwo world wars, the economic depression inGreat Britain during the 1920s and prohibitionin the United States from 1920 to 1933, whichgreatly affected export of Scotch to NorthAmerica. Since 1945, however, the industry inScotland has consolidated and expanded. The20th century has also witnessed a considerableimprovement in the quality of whisky distilledand blended in Scotland as a result of theacceptance of blending malt whiskies with grainwhisky and the amalgamation of several smallerdistilleries into combines.

Scotch malt whiskies can be divided into‘highland’, ‘lowland’, ‘Islay’ and ‘Campbeltownwhiskies’ (Simpson et al., 1974). The ‘highlandline’, which separates the areas in Scotland inwhich the first two types of spirit are distilled, isa straight line which runs from Dundee in theeast to Greenock in the west (Figure 1). It thenextends southwards, below the Isle of Arran.Any whisky produced north of this line,including those from Campbeltown and Islay,

is entitled to be called a highland malt whisky,while whiskies distilled in areas south of the lineare designated as lowland whiskies. Of the 104malt whisky distilleries in Scotland, 95 arehighland malt whisky distilleries. Of these, nofewer than 49 are situated in an area measuring50 miles east to west and 20 miles southwardsfrom the Moray Firth. This area of Speyside hasbeen called the ‘Kingdom of Malt Whisky’(Cameron Taylor, 1970). Classification of thefour whiskies distilled on the islands of Jura,Orkney and Skye is disputed. Some authoritieslist them along with the Islay whiskies as ‘island’whiskies, others as highland whiskies, which isgeographically correct. There are also eight grainwhisky distilleries in Scotland.

Whiskey distilling in Ireland was, as has beennoted, first recorded in the 12th century. By1556, it had become sufficiently widespread towarrant legislation to control it. A statuteproclaimed that year stated that a license wasrequired to manufacture the spirit, but that peers,gentlemen owning property worth £10 or moreand borough freemen were exempt (McGuire,1973). Taxation of whisky distilling graduallybecame more excessive and collection of taxes

Figure 1. Highland and lowland whisky-producing regions of Scotland.

Highland whisky region

Lowland whisky region

D

ENGLAANDLAND

h SeaNorth


became increasingly efficient. However, in 1779there was an important change in the distillerylaws. An attempt was made to limit the extent ofevasion of spirit duty by prescribing a minimumrevenue to be exacted from the owner of eachstill. The effect of this legislation was dramatic.In 1779, there was said to be 1,152 registeredstills in Ireland. By 1790, this number had fallento 216 and this inevitably fostered widespreadillicit distilling (McGuire, 1973). This legislationlasted until 1823 when it was replaced by lawsthat taxed Irish whiskey on the volume ofproduction, legislation that is essentially still inforce today. Development of the Irish whiskeydistilling industry in the present century hasinevitably been influenced by economiccircumstances and by the political division ofIreland into the Republic and Northern Irelandthat occurred in 1922. Barnard (1887) describedvisits to 28 distilleries in Ireland, but closuresand amalgamations followed such that whenMcGuire (1973) prepared his account, therewere only two whiskey-distilling companies inIreland, one with distilleries in Dublin and Corkin the Republic and the other with plants inBushmills and Coleraine in Northern Ireland.These two companies have since amalgamatedand have concentrated their distillery operationsin Cork and Bushmills. There has also been amove towards production of a lighter Scotch-type whisky in Ireland to replace the heaviertraditional Irish whiskey.

Outline of whisky production processes

Whiskies differ basically in the nature andproportion of the cereals used as raw materialsand on the type of still used in the distillationprocess. These differences in the productionprocess are illustrated in the flow diagram inFigure 2 for production of Scotch malt whisky(production of Irish whiskey is very similar).Detailed accounts of each of the unit processesin whisky production are given in subsequentsections of this chapter.

A characteristic of Scotch malt whisky is thatthe only cereal used in its manufacture is maltedbarley (Table 1). After milling, the meal ismashed in a mash tun (Figure 2) similar to thatused in breweries for beer production. Duringmashing or conversion, enzymes in the maltcatalyze the hydrolysis of starch into fermentablesugars. In the manufacture of Scotch grainwhisky and Irish whiskey, other cereals are usedalong with malted barley to provide additionalstarch in the mash tun (Table 1). Owing to thehigh gelatinization temperature of their starches,unmalted cereals must be precooked before theyare incorporated into the mash.

The wort, or clear mash, leaving the mash tunis cooled and fed into a vessel where it is mixedwith yeast. In Scotland and Ireland thesefermentation vessels have a relatively smallcapacity and are known as ‘washbacks’ (Figure2).

Table 1. Raw materials and unit processes used in the production of Scotch and Irish whiskies.

Scotch malt Scotch grain Irish

Raw materials Peated and unpeated Wheat or corn and a small Unmalted barley and unpeated,malted barley proportion of malted barley malted barley

Conversion Infusion mash Mash cook followed by Infusion mashconversion stand

Fermentation Distillers yeast and Distillers yeast Distillers yeastbrewers yeast

Distillation Two pot stills Continuous still Three pot stills

Maturation At 62o GL in used charred Up to 67o GL in used charred At 70o GL in used charred oakoak bourbon whiskey oak bourbon whisky barrels bourbon whisky barrels or sherrybarrels or sherry casks or sherry casks for at least casks for at least three yearsfor at least three years three years

198 T.P. Lyons

Fermentation is conducted with strains of theyeast Saccharomyces cerevisiae that are usuallyspecially propagated for the purpose, althoughScotch malt whisky distillers may use somesurplus brewers yeast (Table 1). The process isallowed to proceed to a point at which thespecific gravity of the fermented mash hasusually dropped to below 1.000. In pot stilldistilleries, the fermented mash or ‘wash’ (beer)is fed directly to a still known as the wash still,from which the distillates are redistilled in thesecond or low wines still. In Ireland, and in oneScotch malt whisky distillery, a third distillationis carried out. Finally, the freshly distilled whiskyis stored in charred oak barrels for minimumperiods of time that depend on the legislation inthe producing country (Table 1). Scotch malt andIrish whiskies are customarily matured for muchlonger than the legal minimum period.

Individual operations

RAW MATERIALS

Malted cereals

Malted barley is the principal malted cereal usedin whisky production. Like the brewer, thewhisky distiller uses barley cultivars of the

Figure 2. Flow diagram showing the principal operations during production of Scotch malt whisky.

Wash still(pot still)

Lowwines

Low winesstill

Maturing warehouse

Oak casks(barrels)

Spirit receiverFeints and low

wines

For

esho

t and

fein

ts(H

ead

& ta

ils)

Spi

rit o

r m

iddl

e ru

n

Condensers

Yeast

Washback(fermenter)

Hotwater

Maltstorage

bins

Maltmill

Mashingmachine(mixer)

Heatexchanger

Draft(grainsolids)

Mash tub(cooker)

Wort

species Hordeum vulgare L. and Hordeumdistichon (Hough et al., 1971). Malted barley isemployed as a source of enzymes (principallyamylolytic) that catalyze the hydrolysis ofstarches and in some instances serves as a sourceof starch that is converted ultimately into ethanol.These two demands must be finely balanced. Inthe manufacture of Scotch malt whisky, whereonly malted barley is used, care must be takenwhen the grain is sprouting during the maltingprocess to ensure that only a limited amount ofenzyme activity is produced. This is becauseenzyme is produced at the expense of thefermentable materials in the grain (referred toas ‘extract’). However in the manufacture ofother types of whisky, malted barley is oftenused as the only source of amylolytic enzymein a mash bill that contains a high proportion ofunmalted grain. In this type of whisky productionthe enzyme activity of the malted barley mustbe greater than that used in Scotch malt whiskymanufacture.

Traditionally, barley used in the production ofScotch malt whisky was malted on the distillerypremises using a floor malting system and driedover fires of coke and peat in the pagoda-shapedkilns which are still a feature of these distilleries.To a large extent, this system has beensuperseded by mechanical maltings, which


produce malt for groups of distilleries. In ordernot to destroy the enzyme activity developedduring malting, a balance must be achieved inthe kiln between drying the green malt to asuitably low moisture level for storage, curingto give it the appropriate flavor and retainingsufficient enzyme activity (Simpson, 1968). Inmaltings attached to the distillery, the kilntemperature is increased slowly over a 48 hrperiod to achieve an even rate of drying and thedesired flavor. The latter character is achievedby fuelling the furnace with peat during the earlypart of the kilning period when the green malt ismoist and readily absorbs the peat smoke or‘reek’. In mechanical maltings the green malt isdried at a faster rate with a forced-air draught,but a supplementary peat-fired kiln is often usedto produce flavored malts. The amount of peatused varies with different maltings. Some of thedistilleries on Islay in Scotland specialize inproducing a whisky with a very pronounced peatflavor, and they therefore use heavily-peatedmalts. Malted barleys used in the manufactureof Scotch grain and Irish whiskies are not driedover a peat fire. They generally have a greaterenzyme activity, so the relatively smallproportion of malted barley used in the mashcontains sufficient enzyme activity to convertall of the starch in the mash (principally suppliedby unmalted cereals) into fermentable sugars.The greater enzyme activity in these maltedbarleys is reflected in their nitrogen content.Malts used in production of Scotch grain whisky

have a nitrogen content of 1.8% or higher(compared with a brewer’s malted barley with anitrogen content in the region of 1.5%) (Houghet al., 1971). Malting aids, such as gibberellicacid and bromate, are not normally used inproduction of malted barley for whiskymanufacture.

Because of the high cost of malted cereals,considerable effort has been expended by thewhisky distiller in attempting to devise methodsthat allow prediction of the yield of alcoholexpected using different proportions of maltedbarley in the mash bill. Unfortunately, themethods customarily used by the brewer, suchas those recommended in Great Britain by theInstitute of Brewing Analysis Committee (1975),have proved of limited value. The brewer hasused measurements of diastatic power,expressed as the Lintner value, which is ameasure of the extent of saccharification ofsoluble starch present in a cold water extract ofthe malt (Lloyd Hind, 1948) as an indicator ofmalt quality. Diastatic activity measured in thisway includes contributions from both α- and ß-amylases. However, Preece and his colleague(Preece, 1947; 1948; Preece and Shadak-sharaswamy, 1949 a,b,c) have shown that highß-amylase activity, as determined by the Lintnervalue, is not always accompanied by high α-amylase activity. Pyke (1965) showed that theLintner value of a malt is only useful forpredicting the spirit yield in manufacture ofScotch grain whisky when the proportion of

300

325

350

375

100 20 30

Malt enzyme activity (°Lintner)

Sp

irit

yie

ld (

Lit

ers

abso

lute

p

er t

on

ne

of

gra

in)

Figure 3. Relationship between the diastatic activity of a Scotch grain whisky mash bill and spirit yield. Laboratory fermentationswere conducted using mash bills containing different proportions of malt mixed with corn, and therefore with different Lintnervalues. It can be seen that only when the proportion of malt is rate-limiting can spirit yield be correlated with the Lintner value

(Pyke, 1965).

200 T.P. Lyons

malted barley in the mashbill is low (Figure 3).Determination of α-amylase activities of the maltgave a less satisfactory correlation than theLintner value, an observation which agrees withthat made earlier by Thorne et al. (1945). Furtherevidence for the unsuitability of employingtraditional malt specifications for predictingperformance in whisky manufacture has comefrom Griffin (1972).

Although the degree to which a malted barleyhas been peated can to some extent be assessedby smell, such is the importance of this characterin malt that it must be determined in a morerigorous fashion. Peat smoke or ‘reek’ containsa wide range of compounds, but it is generallyheld that the peaty character is imparted to themalt largely as a result of absorption of phenols.For some years, Scotch distillers used a methodbased on a reaction of phenols with diazotizedsulphanilic acid. A lack of specificity in thismethod, coupled with the instability of thediazonium salt, prompted MacFarlane (1968) torecommend an alternative method involvingextraction of phenols from malt with diethyletherunder acid conditions with absorptiometricmeasurement of the color developed when thephenols are reacted with 4-aminophenazone.MacFarlane applied the method to a wide varietyof malts, both peated and unpeated, and reportedvalues ranging from zero (for an unpeated grain)to as high as 9.4 ppm for a malt produced onIslay in Scotland. (It has been calculated that toobtain a malt with 10 ppm phenols, one tonneof peat must be used for drying each tonne ofmalted barley).

Scotch whisky distillers have been concernedby the possibility that colorimetric methods, suchas those recommended by MacFarlane (1968)and Kleber and Hurns (1972), may not assay allof the organoleptically-important compoundsthat malt acquires as a result of peating. Toexamine this possibility, MacFarlane et al.(1973) produced peat smoke condensate on alaboratory scale and separated the oil andaqueous phases from the wax fraction, as onlythe two former would contain components thatmight appear in the distilled whisky. Sixcompounds, namely furfural, 5-methylfurfural,guiacol, phenol, ρ-cresol and 5-xylenol, weredetected in the aqueous fraction by gas liquid

chromatography (GLC). The peat smoke oil wasmore complex, and no fewer than 30 peaks,some of them created by more than onecompound, were obtained by GLC. Thecompounds included hydrocarbons, furfuralderivatives, benzene derivatives and phenols.The authors stressed that using their GLCtechniques, 3,5-xylenol is masked by the peaksof m-ethylphenol and ρ-ethylphenol, twocompounds thought to make an importantcontribution to peat aroma and taste. Figure 4shows gas liquid chromatograms of extracts ofa peated and a unpeated malt.

Unmalted cereals

Fewer problems are encountered in arriving atspecifications for the unmalted cereals used inwhisky manufacture, namely wheat, corn, ryeand barley. The corn (varieties of Zea mays) usedin mash bills for manufacturing Scotch grainwhiskies is usually of the yellow dent type,generally obtained from France. Occasionallywhite corn is used, and it is reputed to give ahigher alcohol yield. Corn is a popular grainbecause it has a high content of starch that isreadily gelatinized and converted intofermentable sugars. The US has imposed controlson the quality of corn used for whiskymanufacture. In the US there are three grades ofcorn, with only grades 1 and 2 being approvedfor spirit manufacture. In Great Britain, on theother hand, the corn used is normally grade 3on the US scale.

Unmalted barley used in manufacture of Irishwhiskey has a quality intermediate to that usedfor malting and that used for cattle feed. In thisway, the maltster can select the best barleyavailable on the market at the time of purchase.For many years, a small percentage (about 5%of the total) of unmalted oats (Avena spp.) wasincluded in mash bills for manufacturing Irishwhiskey. It was contended that these grains, withtheir large husks, improved the texture of thegrain bed in the mash tun (to assist in strainingoff the clear wort from the mash solids) and thatoats influenced the flavor of Irish whiskey.Whether either or both of these effects wereimportant will probably never be known, for oatsare no longer used in the production of Irishwhiskey (Court and Bowers, 1970).


Unpeated malt

Peated malt

ρ-E

thylp

henol

ρ-E

thylp

henol

ρ-C

resol +

Phenol +

Guaic

aol

Phenol +

ρ-C

resol +

Guaic

aol

ρ-C

resol

Figure 4. Gas-liquid chromatograms of extracts of unpeated and peated Scoth barley malts showing the contribution peatingmakes to the content of phenols in the malt. Peating leads to an increase in the size of the peak corresponding to phenol, ρ-cresol

and guiacol, and of the peaks which lie to the right of the mixed phenol peak (attributable to furfurals and hydrocarbons). ρ-cresolis also detectable on the chromatogram of extracts of peated malt. The ratio of the area of the total phenol peaks to that of ρ-

ethylphenol is used as an indication of the peatiness of the malt.

202 T.P. Lyons

MASHING AND COOKING

Mash production in Scotch and Irish distilleriesinvolves a process not unlike that used inbreweries to prepare wort for beer manufacture.However, where cereals other than malted barleyare used, the malt mashing is preceded by a hightemperature cooking process.

Mashing

Regardless of whether the mash bill containscereals other than malted barley, the mainbiochemical changes that take place duringmashing are hydrolysis of starch, protein andother biopolymers in the meal to produce watersoluble low molecular weight compounds thatform a fermentable substrate (or wort). The majorstarch-liquefying and saccharifying enzymes areα- and ß-amylases, while the limit dextrinsformed by action of amylases on amylopectinare further hydrolyzed by limit dextrinases.

Barley malt used in the manufacture of Scotchmalt whisky is coarsely ground in a roller milladjusted to give a grind no finer than the maltwarrants. Too fine a grind can give rise to a ‘setmash’ that settles on the bottom of the mash tunto block and impede drainage of the liquid(Simpson, 1968). The mash tun is usuallypreheated with water and the perforated bottomplates are flooded before mashing commences.

The meal from overhead bins is mixed with hotwater at 60-65oC in the proportion of one partmeal to four parts water, and the mash ishomogenized by action of revolving rakes. Inthe traditional mashing process, the mash isusually loaded into the tun to a depth of aboutone meter and allowed to stand for about onehour, after which the wort is drained off fromunder the grain bed. This liquid extract, whichhas a specific gravity (SG) of 1.070-1.060(Figure 5) is collected in an intermediate vesselknown as an ‘underback’. After being cooled toaround 25oC in a heat exchanger, the wort ispumped into the fermentation vessel. The bedof grains in the tun is then re-suspended in waterat 75oC and a second batch of wort is drawn offat a specific gravity of around 1.030 and passedinto the underback. This process is known asthe first aftermash and is repeated twice more;except that the dilute worts drawn off are notpassed to the underback, but are returned to thehot water tank to be used in the next mash. Thewort in the underback has a pH value of about5.5, a specific gravity of 1.045-1.065, and anamino nitrogen content of about 150-180 mg/L. The spent grain residue or ‘draff’ is removedfrom the mash tun and sold as animal feed. Manydistilleries have now reduced this process to justthree water additions. Larger distilleries are nowinstalling semi-lauter tuns in place of thetraditional mash tuns. These have vertical knives,

Wort

(SG 1.060-1.070)

Weak wort

(SG about 1.030)

First

aftermash

Fermentation

Malt meal

MashThird

aftermash

Second

aftermash

Weak wort

(SG about 1.010)

Weak wort

(SG about 1.006)

Weak wort for next mash

Figure 5. Flow diagram showing the mashing cycle in a Scotch malt whisky distillery.


enabling the structure of the mash bed to bemaintained whilst speeding wort draining, andsparging rings to allow simultaneous drainingand sparging. The result is a faster more efficientextraction without loss of clarity or changing thecomposition of the wort.

Mash preparation in the manufacture of Irishwhiskey is very similar to that used in Scotchproduction, but there are certain differences. Useof a high percentage of raw barley in the mashbill (up to 60%) has necessitated the use of stonemills or hammer mills to achieve the requiredgrind. The unmalted barley is sprayed with waterto give a moisture content of about 14% andthen dried to around 4.5% moisture beforegrinding. The malted barley used is roller-milled.In recent years a plant has been installed inIreland which uses a wet milling process thateliminates the need for watering and drying ofthe unmalted barley. Mash tuns used in Irishwhiskey distilleries differ from those used tomake Scotch in that they are larger, simplybecause these distilleries have stills with a greatercapacity than those in Scotland. (This differencedates back to when there was a flat rate of taxper still in Ireland, rather than a tax per gallonof product.) Moreover, the mashing cycle differsin that the weak wort from the first aftermash isnot passed to the underback, but is mixed withworts from the second and third aftermashes tobe used in the subsequent mash (Lyons, 1974).

Cooking followed by malt mashing

The wheat, corn and rye used in production ofScotch grain whisky must be cooked beforebeing added to the mash containing maltedbarley in order that the starch in these grainscan become gelatinized and accessible to themalt enzymes. Traditionally, cooking was carriedout as a batch process, but it has to some extentbeen superseded by continuous processes.

For batch cooking, the grain is freed fromextraneous material by passage through screensand then ground either in a pin-type mill or ahammer mill. It is then conveyed to the cookerand subjected to a cooking cycle involving hightemperatures and pressures designed to bringabout complete gelatinization of the starch.Inadequate cooking will sometimes leave starchgranules intact in the mash, while excessiveheating can cause caramelization and therefore

loss of sugar and decreased spirit yield. Pyke(1965) has provided an account of the cookingof corn in its production of Scotch grain whiskyand a similar process is still used today. Theconventional cooker is a horizontal cylindricalvessel capable of working at pressures up to 90psi (63 x 104 Pa), and fitted with stirring gear. Atypical cycle in cooking corn might consist of1-1.5 hrs of heating with live steam injection toreach a temperature of 120oC and a pressure of15 x 104 Pa. The mash is held at this temperatureand pressure for a further 1.5 hrs. The liquidused for the mash is often that from the thirdaftermash mentioned earlier. At the end of thecooking time, the pressure is released and thehot cooked corn mash is blown directly into asaccharification vessel containing the maltedbarley suspended in hot water. Cold water is thenadded to bring the temperature in the mash tunto around 63oC. Good mixing is also essentialat this stage; and failure to achieve it can lead tothe entire mash solidifying. The need for rapidcooling of the corn mash was emphasized byMurtagh (1970), who showed that with slowcooling or holding at a temperature around 82oClipid-carbohydrate complexes could be formedthat are not fermentable and can lead to a lossof 1-2% in spirit yield.

After a suitable mashing period, the mash maybe filtered as described previously. Morecommonly, the entire mash may be pumped viaa heat exchanger into a fermentor. Pyke (1965)has published the composition of a typicalScotch grain whisky mash (Table 2).

In the manufacture of Scotch grain whisky, theproportion of malted barley in the mash bill isusually around 10-15%, a proportion far greaterthan that required merely to provide a source ofamylolytic enzymes. It is a practice of longstanding, and is done to obtain the required maltflavor in the grain whisky. This was noted asearly as 1902 by Schidrowitz and promptedfurther comment from Valaer in 1940.

Continuous cooking has been adopted inrecent years in Scotland. Stark (1954) listed theadvantages of using continuous cooking. Whilepractices vary to some extent in differentdistilleries, essentially the cereal is slurried ataround 50oC and then pumped through a steam-jet heater into the cooking tubes where theresidence time is normally 5-10 minutes. Thecontinuous cooker has a narrow tube (6-16 cm

204 T.P. Lyons

in diameter) which reduces the incidence ofcharring and carbon deposition. The mashpasses through the tube at about 20 meters/minute, with the temperature reaching near 65oCand the pressure 65 psi (40 x 104 Pa).

Table 2. Chemical composition of a typical scotch grainmash*.

%

Total soluble carbohydrate (as glucose) 9.00Insoluble solids 2.20Fructose 0.13Glucose 0.29Sucrose 0.28Maltose 4.65Maltotriose 0.96Maltotetraose 0.45Dextrin 2.54Amino nitrogen (as leucine) 0.09Ash 0.27 containing P2O5 0.09 K2O 0.09 MgO 0.02

µg/mlThiamin 0.46Pyridoxine 0.61Biotin 0.01Inositol 236Niacin 11.1Pantothenate 0.71

*Pyke, 1965.

Fermentation

The objectives in fermenting a whisky distillerymash with strains of Saccharomyces cerevisiaeare to convert the mash sugars to ethanol andcarbon dioxide while producing quantitativelyminor amounts of other organic compoundswhich contribute to the organoleptic qualities ofthe final distilled product. Fermenting vesselsvary considerably in volume, depending on thedistillery. The small Scotch whisky distillery atEdradour near Pitlochry, for example, hasfermentors with a capacity of only 4,500 liters.In contrast, other Scotch and Irish pot whiskeydistillers have fermentors with a capacity in therange 50,000 to 150,000 liters. Much largerfermentors are found in Scotch grain whiskydistilleries. Traditionally, the smaller pot distillery

fermentors were made of wood, usually of larchor Oregon pine, but in recent years they havebeen constructed of steel or aluminum. In someScottish distilleries, timber is still used as acovering material.

When the fermentor is partly filled, the mashis inoculated with a suspension ofSaccharomyces cerevisiae. The source of yeastvaries with the location and size of the distillery.Distilleries in Scotland and Ireland, particularlythe pot whisky distilleries, rarely have their ownyeast propagation equipment. They rely onspecially-propagated pressed yeast for use infermentations. In Scotch malt whisky distilleries,this pressed yeast is augmented, usually on anequal weight basis, with surplus brewery yeasttypically supplied in compressed form frombreweries that may be as far as 500 km from thedistillery.

The requirements for a strain of Saccharo-myces cerevisiae yeast used in distillery practicehave been described by Harrison and Graham(1970). Apart from the obvious need to select astrain that maintains very high viability in thepressed state (containing 25% dry weight), othervery important properties of these strains areability to tolerate concentrations of ethanol ofthe order of 12-15% (v/v) and the capacity tometabolize oligosaccharides such as maltotrioseand maltotetraose in order to maximize theconversion of starch into ethanol and carbondioxide.

Whisky worts usually have a specific gravityin the range 1.050-1.080, a pH value of around5.0, a total acid content of 0.1% and an opticalrotation of +30o. After inoculation, the yeastcontent is 5-20 million cells/ml. The bacterialcount varies with the cleanliness of the plant andthe extent to which the raw materials wereendowed with a microbial flora. Scotch grainwhisky fermentations create little if any foambecause of their large content of suspendedsolids. However, in most Scotch malt whiskyfermentations only a small proportion of thesuspended solids in the mash is retained in thefermentation vessel. These fermentations tendto foam and the distillers have resorted to theuse of various types of antifoams.

Changes over the time course of a typicalfermentation in a Scotch malt whisky distilleryare depicted in Figure 6. Fermentation proceedsvigorously for the first 30 hrs, during which time


the specific gravity falls to 1.000 or below andthe optical rotation to around zero. The sugarsin the wort are utilized in a particular sequencewith glucose and fructose being fermented first,followed by maltose and then maltotriose. Theremoval of sugars during fermentation of aScotch grain whisky mash is shown in Figure 7

(Pyke, 1965). Over the first 30 hrs the pH value,after declining to around 4.2, rises to about 4.5.During the first 30 hrs the specific gravity dropsat a rate of about 0.5o per hour accompanied bya massive evolution of heat. While many of thelarger grain whisky distilleries have fermentorsfitted with cooling coils, these are absent, or if

Figure 6. Changes in specific gravity, optical rotation, pH value and acidity during fermentation of a mashin the production of Scotch malt whisky (Dolan, 1976).

5.0

4.5

4.0

3.5

3.0

2.5

2.0

1.060

1.050

1.040

1.030

1.020

1.010

00.995

+30

+25

+20

+15

+10

+5

0

0 10 20 30 40 50

pH

Sp

ecif

ic g

ravi

ty0.30

0.25

0.20

0.15

0.10

0.05

0

Optic

al rotatio

n (

a °

D)

Tota

l aci

d (

%)


pH

Acidity

Specific

gravity

Optical rotation

Maltotriose

Maltose

Glucose

0

10 30 50

10

30

50

Fermentation time (hrs)

Co

nce

ntr

atio

n o

f fe

rmen

tab

le s

ug

ar (

g/l)

Figure 7. Time course removal of fermentable sugars from a Scotch grain whisky mash (Pyke, 1965).

206 T.P. Lyons

fitted are relatively inefficient in most maltwhisky distilleries where temperature can riseby the end of the fermentation to as high as 35-37oC. The distiller is concerned about thetemperature rise during fermentation since thiscan cause the fermentation to stop or become‘stuck’. Temperature rise can be controlled byusing a lower starting temperature or, becauseglycolysis of sugar is a heat-producing process,by using a lower initial concentration. Strains ofSaccharomyces cerevisiae are well suited formalt whisky distillery fermentations since theycan ferment efficiently over a wide temperaturerange. Fermentation is usually continued for atleast 36 hrs and frequently longer, at which timethe ethanol content of the wash is 7-11% (v/v).In larger distilleries, particularly those in the US,the carbon dioxide evolved is collected, liquefiedand sold. Smaller distilleries, particularly the maltwhisky distilleries in Scotland, usually do nothave this facility.

It should be noted that mashes in malt whiskydistilleries are not boiled, so any enzyme activitymanifested at the temperature of the mash andany microorganisms that can survive at thattemperature will continue to be active during thefermentation. The continued activity of limitdextrinases in unboiled distillery mashesincreases the concentration of sugars availablefor fermentation by the yeast. Hopkins andWiener (1955) calculated that with amylasesalone the yeast cannot metabolize the equivalentof the final 12-16% of the starch.

Another important consequence of using non-sterile conditions in distillery fermentations isthe activity of bacteria that pass through in themash, which are encouraged to some extent bythe relatively high temperatures to which thefermentations can rise. In addition to lactic acidbacteria, the flora can include other Gram-positive as well as Gram-negative strains. Theconcentration of the flora depends on a numberof factors including the extent to which the lacticacid bacteria grew during yeast propagation, theextent of the flora on the cereal raw materialsand on the standard of hygiene in the distillery.There is no doubt, however, that the controlledactivity of this bacterial flora, and particularlyof the lactic acid bacteria, is accompanied byexcretion of compounds that contribute to theorganoleptic qualities of the final whisky(Geddes and Riffkin, 1989).

During the first 30 hrs or so of malt whiskyfermentation there is vigorous fermentation andthe majority of the aerobic bacteria die. This,however, provides ideal conditions for growthof anaerobic or microaerophilic bacteria,principally lactic acid bacteria (mainly strainsof Lactobacillus brevis, L. fermenti andStreptococcus lactis) with the result that theconcentration of lactic acid in the fermented mashcan be as high as 30 mg/L (MacKenzie andKenny, 1965). A wide range of lactobacillusspecies have been identified in Scotch whiskyfermentations including L. fermentum, L. brevis,L. delbrueckii, L. plantarum, L. casei and abacterium resembling L. collinoides, in additionto Leuconostoc spp., Streptococcus lactis andPediococcus cerevisiae (Bryan-Jones, 1976).More recently Barbour (1983) isolated manyspecies that did not conform to recognizedspecies of lactic acid bacteria, a pointemphasized by Walker et al. (1990) who usedDNA hybridization techniques to classifydistillery bacteria. Growth of lactic acid bacteriais probably enhanced by yeast excretion ofnitrogenous nutrients at the end of a vigorousfermentation. Kulka (1953) demonstrated theideal nature of yeast autolysate for growth oflactobacilli. Bacterial activity in the fermentingwort also leads to removal of some acids.Actively growing yeasts secrete citric and malicacids, but MacKenzie and Kenny (1965)attribute the lower concentrations of these acidsin malt distillery worts (as compared to breweryworts) to their partial removal by bacteria.

Occasionally, the extent of the bacterial florain the fermenting wort can become too large.This causes problems due to sugar utilizationby the bacteria that lead to an overall decreasein spirit yield. In addition, the bacteria mayproduce organoleptically-undesirablecompounds and also release hydrogen ionscausing the pH value of the wort to fall too low,thereby providing suboptimal conditions foraction of certain enzymes. Examples ofundesirable compounds that may be excretedby bacteria are hydrogen sulfide and othersulfur-containing compounds (Anderson et al.,1972). Lactobacilli can also metabolize glycerol(excreted by the yeast during fermentation) toproduce ß-hydroxypropionaldehyde, whichsubsequently breaks down on distillation to giveacrolein (Harrison and Graham, 1970). Acrolein


imparts a pungent, burnt and often peppery odorto the whisky (Lyons, 1974). In a later paper,Dolan (1976) concentrated on the problemsarising in malt whisky distilleries when there isan unacceptably high concentration of bacteriain the mash. Table 3 shows changes in theconcentrations of Gram-negative and Gram-positive bacteria and (separately) of lactobacilliduring fermentation of a minimally-infectedmash and of a heavily-infected mash. The timecourse of fermentation of an unacceptably-infected malt distillery mash (Figure 8) shows,in comparison with similar data for fermentationof an acceptable mash (Figure 6), a greater risein the acid content of the mash after about 35

hrs and a lower optical rotation of the mash afterabout 40 hrs. In the fermentations there is oftena difference of up to 4 hrs from the time a rise inthe acid content is detected to the point whenthe pH value of the fermentation begins to fall.Dolan (1976) attributes this to the bufferingcapacity of the mash. The data in Table 4 showthe effect of different levels of infection after 30hrs fermentation of a malt distillery mash onspirit yield and the associated financial losses tothe distiller. Dolan (1976) recommends upperlimits of 1,500 bacteria, 50 Gram-positive and10 lactic acid-producing bacteria per millionyeast cells in the mash at the start of fermentation.

Table 3. Changes in the concentration of bacteria during fermentation of a minimally infected and a heavily infectedScotch malt whisky wort.*

Minimally-infected wort Bacteria/ml Heavily-infected wort

Age Gram- Gram-negative Lactobacilli Gram-positive Gram-positive Lactobacilli(hours) negative rods cocci rods cocci

At setting 2,000 3,700 n.d. 52,000 60 0.6 x 106

10 150 n.d. n.d. n.d. n.d. 2.3 x 106

20 n.d.** n.d. 1.53 x 106 n.d. n.d. 18.8 x 106

30 n.d. n.d. 10.2 x 106 n.d. n.d. 96 x 106

40 n.d. n.d. 10.2 x 106 n.d. n.d. 502 x 106

50 n.d. n.d. 50 x 106 n.d. n.d. 1 x 109

*Dolan, 1976.**None detected.

1.0605.0

4.5

4.0

3.5

3.0

2.5

2.0

+30

+25

+20

+15

+10

+5

0

-2.5

0 10 20 30 40 50

1.050

1.040

1.030

1.020

1.010

00.995

0.30

0.25

0.20

0.15

0.10

0.05

0

Sp

ecif

ic g

ravi

ty

pH

Op

tica

l ro

tati

on

(a

°D)


pH

Acidity

Optical rotationSpecific

gravity

Tota

l aci

d (

%)

Figure 8. Changes in specific gravity, optical rotation, pH value and acidity during fermentation of a Scotch malt whisky mashcontaining an unacceptably high concentration of bacterial infection (Dolan, 1976).

208 T.P. Lyons

Much less has been published on the effect ofretaining solid material in the fermenting mash.However, marine microbiologists have longknown that the presence of solid particles in aliquid medium can affect bacterial growth,probably because of the concentration ofnutrients at the solid-liquid interface (Heukele-kian and Heller, 1940; Zobell, 1943). Moreover,Cromwell and Guymon (1963) found that form-ation of higher alcohols during fermentation ofgrape juice is stimulated by the presence of grapeskins or inert solids. Beech (1972) made similarobservations on cider fermentations. Merritt(1967), in the only detailed report on the role ofsolids in whisky distillery fermentations, statesthat a dry solid concentration of 50 mg/100 mlmight typically be expected, although much willclearly depend on the design of the mash tunsused in individual distilleries. Merritt went on toreport that a concentration of dry solids as lowas 5 mg/100 ml causes an increase in yeastgrowth, and that solids also enhance the rate ofproduction of ethanol and glycerol. There wasalso an effect on production of higher alcoholsby the yeast (Table 5). With the possibleexception of n-propanol, production of all of the

major higher alcohols was increased in thepresence of solids, the effect being particularlynoticeable with isobutanol and 2-methylbutanol.

The effect of low insoluble solids content is afactor relevant to congener levels in malt whiskyfermentations. In grain whisky production,where ‘all-grains-in’ fermentations are generallyused, the degree of rectification duringdistillation is the principle determinant of higheralcohol levels in the spirit.

DISTILLATION

Whether the fermented beer is distilled in a potstill, as in production of Scotch malt and Irishwhiskies, or in a continuous still based on theCoffey design as in the manufacture of Scotchgrain whiskies, the objectives are the same:selective removal of the volatile compounds(particularly the flavor-producing congeners)from the non-volatile compounds and to createadditional flavor-producing compounds as aresult of chemical reactions that take place inthe still. Nevertheless, it is still most convenientto discuss whisky distillation under the separateheadings of pot still and continuous distillation.

Table 4. Effect of the level of bacterial infection on the loss of spirit incurred following fermentation of a Scotch maltwhisky mash.*

Infection Bacteria/ml in 30 hr old Approximate loss in Approximate financial loss inrating mash (million) spirit yield (%) US dollars (thousands)

At filling Duty paida 0 - 1 <1 <26 <369b 1 - 10 1-3 26 - 79 369 - 1106c 10 - 100 3-5 79 - 131 1106 - 1844d >100 >5 >131 >1844

*Dolan, 1976.

Table 5. Production of higher alcohols (mg/100 ml).

Insoluble solids content 2-methyl butanol 3-Methyl butanol

Mash (mg/100 ml) n-Propanol Isobutanol (amyl alcohol) (isoamyl alcohol) Total

1 0 1.5 4.2 3.0 8.0 16.725 1.5 6.5 4.4 8.8 21.2

2 0 2.8 4.5 2.7 9.5 19.535 3.3 6.6 3.7 10.1 23.7

3 0 1.7 4.8 3.1 8.0 17.650 1.7 6.8 4.6 9.5 22.6


Pot still distillation

The copper pot still, which is the featuredominating any Scotch malt or Irish whiskeydistillery, has changed hardly at all over thecenturies, except of course in size. Traditionally,the onion-shaped stills were fired from beneathand had the vapor pipe or ‘lyne arm’ from thestill projecting through the distillery wall toconnect with a condenser in the form of a coilimmersed in a water tank fed from a local stream(Figure 9). Internal steam-heated calandria arenow preferred to direct firing because thisdecreases the extent of pyrolysis of the stillcontents and results, for example, in a lowerconcentration of furfural in the whisky. Variationsin still design include expansion of the surfacearea of the column into a bulbous shape, waterjacketing and return loops from the first stageof the condensation. (Nettleton (1913) provideda valuable account of early still design). In manydistilleries, condenser coils have been replacedby tubular condensers that have an advantagein that they are designed to conserve the heatextracted from the distillates. Yet other pot stillsare fitted with ‘purifiers,’ which consist of acircular vessel cooled by running waterinterposed between the neck of the still and the

condenser. In Irish pot stills this purifier functionis effected by a trough fitted around the lynearm through which running water is circulated.Pot stills in Scotch and Irish distilleries aretraditionally constructed of copper. The reasonfor this adherence to copper is more thantradition. It has been established that copperfixes some volatile sulfur-containingcompounds that are produced duringfermentation but undesirable in the distilledspirit.

Early whisky distillers realized that althoughthe objective of distilling was to separate volatileconstituents from the beer, collecting thedistillate not as a whole but in several fractionsand combining certain of these fractions gave amuch more acceptable product. Pot stilldistillation in Scotland and Ireland differ not onlyin the size of the stills (25,000-50,000 liters inScotland vs 100,000-150,000 liters in Ireland),but also in the different ways in which fractionsare collected from the stills.

In Scotland, the beer is subjected to twodistillations. In the first, carried out in the beerstill, the beer is brought to a boil over a periodof 5-6 hrs and the distillate is referred to as lowwines. This distillation effects a three-fold

Pot still

Heatingcoils

Rousinggear

Condensatereturn pipe

Lyne arm

Lyne arm tank

Wormrub

Product

Figure 9. Diagram of an Irish distillery pot still. Designs used in Scotch malt whisky distilleries are similar except that the pot isonion-shaped and the still usually has a shorter lyne arm not surrounded by a lyne-arm tank (Lyons, 1974).

210 T.P. Lyons

concentration of the alcohol in the beer (Figure10). The residue in the wash still, known as ‘potale’, is either discharged to waste or evaporatedto produce an animal feed (Rae, 1967).Distillation of the low wines in the spirit is stillmore selective. The first fraction, which containslow boiling point compounds, is rejected as ‘fore-shot heads’. At a stage determined by continuedhydrometric monitoring, which usually occurswhen the distillate has an alcohol content ofapproximately 70-73o GL, the distillate isswitched from the fore shots tank to the whiskyreceiver tank. This switch has traditionally beenmade at the discretion of the distiller; and hehas been aided by the disappearance of a bluishtinge when water is added to the distillate. Themonitoring process takes place in a spirit safesecured with a Government Excise lock.Collection of the distillate is terminated whenthe alcohol content has fallen to a specified value,although distillation of the feints or tails iscontinued until all of the alcohol has beenremoved from the low wines. The residueremaining in the spirit still is known as ‘spentlees’, and like pot ale is either run to waste orevaporated to manufacture animal feed. The

whisky distilled over in the middle fraction hasan alcohol content of 63-70o GL.

The manufacture of Irish whiskey involvesthree rather than two distillations (Figure 11).The fermented beer is heated in the wash still,and the first distillation (‘strong low wines’) iscollected until the distillate reaches apredetermined specific gravity. The distillate,then known as ‘weak low wines’, is switchedinto a separate vessel. The weak low wines arepumped into the low wines still and are re-distilled to produce two fractions termed ‘strongfeints’ and ‘weak feints’. Strong feints are mixedwith the strong low wines in a spirit still.Distillates from this are collected in the samefashion as in production of Scotch. The whiskeycollected usually is about 89-92o GL.

Continuous distillation

No fundamental changes have been introducedinto the design of the patent or Coffey still overthe past century. Automation, particularly of thebeer feed, is now commonplace, as is continuousmonitoring of other stages in the distillationprocess. Nevertheless, many Scotch and Irish

Foreshots(heads)

Fermented beer(about 8°GL)

Wash Still

Low Wines Still(about 24°GL)

Spirit Still

Whiskey(about 63-70°GL)

Feints(tails)

Pot Ale

Spent Lees

Rec

ycle

Rec

ycle

Figure 10. Flow diagram showing the stages in distillation of Scotch malt whisky.


producers of grain whiskies continue to use astill which, like the original Coffey still, has justtwo columns: a beer stripper (or analyzer) and arectifier.

A description of the operation of two columncontinuous stills in the manufacture of Scotchgrain whisky has come from Pyke (1965). Inorder to obtain whisky of high quality from thesestills, they must be operated such that the alcoholconcentration of the spirit at the spirit draw trayin the rectifier is not less than 94.17o GL. Themanner in which the precise control of stilloperation can affect the composition of thewhisky is shown in Figure 12. As illustrated, ifconditions are changed in either direction on theabscissa, the concentration of congeners willalter with a possible adverse effect on finalproduct quality.

MATURATION AND AGING

Freshly distilled whisky of any type is verydifferent from the spirit that is later bottled, eithersingly or blended. The transformation is brought

about by storing the whisky in oak barrels forperiods of time that depend on traditionalpractice and legal requirements. In general,whiskies are matured for far longer than thelegally-required period of time. The raw spirit istaken by pipeline from the distillation plant tothe tank house where it is diluted with water tothe required strength and then transferred intobarrels.

Maturation in barrels is accompanied by a lossof liquid by evaporation, and the relative ratesof loss of water and of alcohol determinewhether the aged whisky has a higher or loweralcoholic strength than that at filling. In Scotland,where the barrels of whisky are stored in cool,unheated, but humid warehouses, the alcoholicstrength decreases (Valaer, 1940). In contrastValaer and Frazier (1936) reported that in theUS storage conditions cause an increase inalcoholic strength. Maturation in barrels is alsoaccompanied by changes in the chemicalcomposition of the whisky. These changes areattributable to extraction of wood constituentsfrom the barrel, oxidation of components presentin the original whisky as well as those extracted

Figure 11. Flow diagram showing the stages in distillation of Irish whiskey.

Fermented beer(about 8°GL)

Wash Still

Low Wines Still

Spirit Still

Whiskey(about 89-92°GL)

Feints(tails)

Heads

Strong Feints Weak Feints(tails)

Strong low wines Weak low wines(tails)

212 T.P. Lyons

0

10-4

10-3

10-2

10-1

102

1

100

90

80

70

60

50

40

30

20

10

0

10

5 10 15 20 25 30

Tray number

Furfural

Total esters

n-Butanol

Ethanol

IsobutanolAmyl alcohol

n-Propanol

Total aldehydes

Total acids

Co

mp

osi

tio

n (

% w

/w)

Eth

ano

l co

nce

ntr

atio

n (

%w

/w)

Figure 12. Changes in composition of the vapor at different trays in a Coffey still rectifier used in the manufactureof Scotch grain whisky (Pyke, 1965).


from the wood, reactions between componentsin the whisky and removal and oxidation ofhighly volatile sulfur components by the carbonchar on the inner surface of the barrel.

Some of the earlier investigators reported onchanges in the composition of the major classesof organoleptically-important compoundsduring maturation in barrels. Thus, Schidrowitzand Kaye (1905) found increased concentrationsof volatile acids in aged whiskies, a trend alsodescribed in the report of the Royal Commissionon Whisky and Other Potable Spirits (1909).Several reports followed. Liebmann and Scherl(1949), for example, reported increasedconcentrations of acids, furfural, tannins andcolor with maturation in barrels. The arrival ofthe gas liquid chromatograph and HPLC greatlyaccelerated research on this topic; and morerecent data on chemical changes that take placeduring maturation are described later in thischapter.

Maturation of whisky in oak barrels is anexpensive process; and it is hardly surprisingthat consideration has been given to methodsfor acceleration. Jacobs (1947) described severalof these methods including pretreatment of thebeer with activated carbon, chemical treatmentof the whisky to convert aldehydes into estersand use of oxidation treatments. Such techniquesare not used in the industry, which adheres tothe traditional nature of the whisky-producingprocess.

BLENDING AND COLORING

Whisky manufacturers take great pride in thequality of their straight or blended whiskies andin their ability to maintain the quality of aparticular product over the years. Blending isconducted by experts who advise manufacturerson mixing or blending large volumes of whiskyafter appropriate sniffing and tasting sessions onexperimental blends. Blended whiskies are filledinto re-used barrels after dilution with water (ornot) and stored for a further period of timereferred to as ‘marrying’. The water used fordilution is softened or demineralized, since watercontaining an appreciable concentration of saltscan cause hazes in the whisky (Warricker, 1960).After marrying, and dilution if necessary, thecolor of the whisky may be adjusted to the

desired value by adding caramel. Somebrownish pigment is extracted from the casks,but this may not be sufficient to provide thedesired color. Finally, the whisky is clarified forbottling by filtration through sheets of cellulose(Simpson, 1968). Chill filtration may also bepracticed, since it removes tannin material fromthe whisky and prevents subsequent appearanceof haze.

EFFLUENT DISPOSAL AND SPENT GRAINSRECOVERY

Traditionally effluents from whisky distillerieswere disposed of in the most convenient manner.Spent grains were retailed, often quite cheaply,to local farmers as animal fodder while pot aleand spent lees were simply discharged into thelocal sewer, stream or river. This is no longerthe case, mainly because of the distiller’sawareness of the nutritional value (largely theprotein content) of some of these effluents, andpublic awareness of environmental problemsarising from uncontrolled disposal of effluentsinto waterways. Simpson has described theproduction of ‘distillers dark grains’. Theseprocesses are widely used to dispose of effluentsfrom whisky distilleries, especially where thetraditional methods of disposal are forbidden oruneconomic.

Organoleptically important componentsof whisky

Modern analytical techniques have enabledmajor advances in the understanding of thecompounds responsible for the organolepticproperties of whiskies. However, the first reportson the nature of flavor-producing compoundsin whiskies antedate the era of gas liquidchromatography by nearly half a century. Twopublications by Schidrowitz (1902) andSchidrowitz and Kaye (1905), dealingexclusively with Scotch whiskies, reported onthe higher alcohol, acid and ester contents ofsome 50 different brands. They reportedanalyses of several Campbeltown Scotch maltwhiskies. A report by Mann (1911) published afew years later also quoted values for acidityand levels of furfural, aldehydes, esters and

214 T.P. Lyons

alcohols in Scotch whiskies imported intoAustralia.

CONCENTRATIONS OF ORGANOLEPTICALLYIMPORTANT COMPOUNDS

Since the introduction of GLC into distillerylaboratories, several reviews have beenpublished on the composition of the majorflavor-producing compounds in whiskies(namely higher alcohols, esters, carbonylcompounds, organic acids, aromaticcompounds) and on the identification ofindividual compounds that make up thesefractions. Higher alcohols, which are stillroutinely determined with GLC using a polarstationary phase (Aylott et al., 1994), arequantitatively the most important. Scotch maltwhiskies are richest in higher alcohols, withcontents often well over 2 g/L. Free fatty acidsare relatively volatile and make a majorcontribution to the organoleptic qualities ofwhiskies. Concentrations of acids in some Scotchmalt whiskies can be as high as 0.4-1.0 g/Labsolute alcohol (Duncan and Philip, 1966).Roughly comparable concentrations of esters arefound in whiskies, although those produced withpot stills generally have higher concentrationsthan those from continuous stills. Esterconcentrations in Scotch and Irish pot stillwhiskies have been reported in the range of 0.27-0.87 g/L absolute alcohol (Valaer, 1940). Lighterwhiskies contain lower concentrations ofcarbonyl compounds, although theconcentration varies with the brand. Grainwhisky may have as little as 20 mg/l ofaldehydes, while in a mature Scotch malt whiskythe concentration may be as high as 80 mg/L(Duncan and Philip, 1966).

CHEMICAL NATURE OFORGANOLEPTICALLY-IMPORTANTCOMPOUNDS

Duncan and Philip (1966) reviewed chromat-ographic and other methods used to separate thevarious organoleptically important compoundsfrom whiskies. Even though analytical methodssuch as capillary column gas chromatographylinked to a mass selective detector (GC-MS)(Aylott et al., 1994) have developed in terms of

sensitivity and discrimination, a major problemin analyzing whisky by GLC is theoverwhelming preponderance of ethanol andwater. Only one volatile compound, namelyisoamyl alcohol, is likely to be present in aconcentration exceeding 0.01%; while most ofthe others are present in concentrations thatrarely exceed 50 ppm. Indeed many compoundsnow understood to have an important impact onwhisky flavor are present at ppb levels.Carter-Tijmstra (1986) described a technique formeasuring dimethyl trisulfide, a compound witha threshold of only 0.1 ppb present in whisky atconcentrations below 50 ppb. Analyses are mostconveniently conducted on extracted fractionsof the different classes of compounds. Whendirect analysis of whiskies has been employed,only a limited number of components have beendetermined (Morrison, 1962; Bober andHaddaway, 1963; Singer and Stiles, 1965).

Recent developments in headspace analysisusing trapping and thermal desorptiontechniques (Pert and Woolfendon, 1986) haveenabled analysis of the more volatilecomponents of whisky flavor whilst avoidinginterference from high concentrations of othercongeners. Headspace analysis using trap andpurge techniques would now be the method ofchoice for measuring highly volatile sulfurcompounds such as dimethyl trisulfide.

Some idea of the variety of compoundsdetected in whiskies came from a compilationof both published and unpublished sources(Kahn 1969, Kahn et al., 1969). Of the some200 compounds listed there are 25 higheralcohols, 32 acids, 69 esters and 22 phenoliccompounds. Undoubtedly, this list could nowbe extended quite considerably. Of the higheralcohols, isoamyl alcohol and optically-activeamyl alcohol predominate, accompanied bylower concentrations of isobutanol and n-propanol. Characteristically, there are usuallyonly low concentrations of n-butanol andsecbutanol. The principal organic acid inwhiskies is acetic acid, which can account forbetween 50 and 95% of the total content ofvolatile acids determined by titration. Of theremaining acids, caprylic, capric and lauric arequantitatively the most important (Suomalainenand Nykänen 1970a). Some of the characteristicflavor and aroma of Irish whiskies may beattributed to somewhat higher concentrations of


the odoriferous butyric acid. Compared withother types of whisky, Scotch whiskycharacteristically contains more palmitoleic acid(and its ethyl ester) than palmitic acid.Suomalainen (1971) has suggested that thetypical stearin-like smell of Scotch malt whiskymay be attributed to long chain fatty acid ethylesters. It is not surprising to find that ethyl acetateis the major whisky ester in view of theprevalence of acetic acid in the distillate.Concentrations of 95 mg/L have been detectedin a blended Scotch whisky (de Becze, 1967).Other esters such as ethyl caprate are present inmuch lower concentrations, on the order of 2-10 mg/L. Of the carbonyl compounds,acetaldehyde is the principal componenttogether with a range of other short chainaldehydes. Furfural, with an aroma resemblingthat of grain, also occurs with as much as 20-30mg/L in Scotch malt whiskies (Valaer, 1940).Acrolein, a pungent and lachrymatorycompound, is also present; and it has beensuggested that it may contribute to the ‘peppery’smell of whisky.

A variety of other organoleptically importantcompounds has also been detected in differentwhiskies, many of which result from maturingthe whisky in charred oak barrels. Scopoletinand other aromatic aldehydes, including vanillin,were detected in bourbon by Baldwin and hiscolleagues (1967). These compounds werepreviously identified by Black et al. (1953) inethanolic extracts of plain and charred Americanwhite oak from which the bourbon whiskeybarrels subsequently used for the maturation ofScotch and Irish whiskies are constructed. (Itshould be noted that the US regulations forbourbon whiskey production require that newcharred oak barrels be used for maturation; sothere is a continuous supply of once-usedbourbon barrels that are shipped to Ireland andScotland for re-use.) A lactone dubbed ‘whiskylactone’, also appears in whisky followingstorage in oak barrels. This compound, ß-methyl-octalactone, was first isolated from Scotchwhisky by Suomalainen and Nykänen (1970b).It has since been reported that both cis and transdiastereomers of the compound occur in whisky(Nishimura and Masuda, 1971). Othercompounds detected in whisky include phenols(Salo et al., 1976), glycerol and erthritol (Blackand Andreasen, 1974), pyridine, α-picoline andvarious pyrazines (Wobben et al., 1971).

CONTRIBUTION OF COMPOUNDS TOORGANOLEPTIC PROPERTIES

Published information on compoundsresponsible for the organoleptic qualities ofwhiskies is meagre and very largely confined toreports by Suomalainen and his colleagues fromthe State Alcohol Monopoly in Helsinki, Finland.In order to assess the contributions made bywhisky components to the odor of these spirits,Salo et al. (1972) concocted a synthetic whiskywith components that chromatographic analysishad revealed were present in a light-flavoredScotch whisky. The synthetic whisky was madeusing 576 g of a mixture of higher alcohols, 90mg of acids, 129 mg esters and 17.4 mgcarbonyl compounds in highly-rectified grainspirit diluted to 34o GL in water that had beenion-exchanged and treated with activatedcharcoal. This imitation whisky contained 13alcohols in addition to ethanol, 21 acids, 24esters and 9 carbonyl compounds. Caramelcoloring was used to give it the color of a distilledand matured whisky. Odor thresholds of theindividual compounds and groups ofcompounds were determined as described bySalo (1970). Experienced taste panel participantswere easily able to distinguish the imitationwhisky from a blended Scotch whisky; but whenthe concoction was mixed with an equal amountof the Scotch, only 6% correct judgments abovechance were made. This suggested that theconcentrations of, and interactions between,components of the synthetic whisky were notgreatly dissimilar from that in the Scotch usedfor comparison.

Odor threshold determinations of individualcomponents in the imitation whisky revealed thatthe contributions made by the mixture ofalcohols and acids accounted for only about 10%of the total odor intensity, despite the fact thatthe alcohols themselves accounted for over 70%of the total concentration of organoleptically-important compounds in the concoction. Estersand carbonyl compounds had a much greaterinfluence, particularly butyraldehyde,isobutyraldehyde, isovaleraldehyde, diacetyl,and the ethyl esters of acetic, caproic, caprylic,capric and lauric acids. Since just three of themost important carbonyl compounds couldsubstitute for the whole carbonyl fraction, therewould seem to be considerable homogeneity inthe odor contributions made by these

216 T.P. Lyons

compounds. Interestingly, the relativecontributions made by the different classes ofcompounds are not very different from thecontributions Harrison (1970) reported that theymake towards the taste of beers.

Threshold values can be assessed not only forindividual components in whisky, but for thetotal aroma of the beverage. Salo (1975)examined this by diluting different whiskies withwater until the characteristic whisky aroma couldonly just be recognized. Values for the thresholddilution of several commercial whiskies shownin Table 6 reflect the differences in aromastrength for several different commercialwhiskies.

An excellent review of how thesecharacteristics can be evaluated from anorganoleptic point of view comes from Jack(2003). Sensory data are only as good as thesample preparation and guidelines were set asapplicable to the Scotch whisky industry. It isimportant that the line between samplepreparation and assessment should be minimizedsince time and the temperature of nosing have amajor impact on sensory character. Astemperature increases, a greater amount of theflavor volatiles are released, thus altering thesensory profile. Effects of changes in roomtemperature on flavor scores determined by asensory panel for a blended Scotch whisky areillustrated in the form of a spider diagram inFigure 13.

ORIGIN OF ORGANOLEPTICALLY IMPORTANTCOMPOUNDS

The two main sources of the organolepticallyimportant compounds in whisky are the yeast

used to ferment the wort and the charred oakbarrels in which the whisky is matured.Suomalainen and Nykänen (1966) fermented anitrogen-free medium containing sucrose witha strain of Saccharomyces cerevisiae and distilledthe fermented medium either after removing theyeast by centrifugation or with the yeastremaining in the medium. Gas chromatographicanalyses of these distillates are reproduced inFigure 14, which also shows a chromatogramof Scotch whisky for comparison. There isclearly a similarity among the three analyses;although differences, such as the higherproportion of isoamyl alcohol in the distillatefrom the fermented medium, can be detected.Also worth noting is the greater concentrationof ethyl caprate in the distillate from the spent

Table 6. Threshold dilution levels for nine different types of whisky*.

Whisky Threshold dilution (x 10-4) One standard deviation range (x 10-4)

Scotch malt 0.56 0.2 - 1.3Scotch, old blend 0.87 0.5 - 1.5Scotch, blend 1.20 0.3 - 4.2Irish 1.30 0.4 - 2.0Bourbon 2.40 0.6 - 11.5Irish 4.50 2.7 - 7.5Canadian 10.40 3.0 - 37.0

*Salo, 1975.

Figure 13. Impact of temperature on flavor profile of Scotchwhisky (adapted from Jack, 2003).

Stale

Sulfury

Soapy

Sour

Oily

Woody Sweet

Estery

Aldehydic

Cereal

Feinty

PhenolicPungent

2.5

1.5

1

0.5

0

2

5°C Room temperature 30°C


medium containing yeast as compared with thatobtained by distilling the medium from whichyeast had been removed. It would be interestingto learn of the importance of yeast strain inproduction of organoleptically-importantcompounds in whisky. Unfortunately, there is alack of published data on this matter.

Iso

am

yl A

lco

ho

lIs

oam

yl A

lco

ho

lIs

oam

yl A

lco

ho

l

Eth

yl C

ap

ryla

te

Eth

yl C

ap

ryla

te

Eth

yl C

ap

ryla

te

Eth

yl L

au

rate

Eth

yl L

au

rate

Eth

yl L

au

rate

Eth

yl P

alm

itate

Eth

yl P

alm

itate

Eth

yl P

alm

itate

Eth

yl C

ap

rate

Eth

yl C

ap

rate

Eth

yl C

ap

rate

Scotch whisky

Sugar fermentationYeast separated

before distillation

Sugar fermentationDistilled with yeast

65

80 120 160 200 210

Temperature (°C)

Figure 14. Gas liquid chromatograms of aroma compoundsproduced by yeast in a nitrogen-free sugar fermentation,

with a trace for comparison of the aroma compoundsdetected in a sample of Scotch whisky (Suomalainen

and Nykänen, 1966).

The complexity of the processes for determiningthe presence and concentration of compoundsin whisky may be illustrated by looking at onecompound for which this process has beenelucidated in detail. In the 1980s ethyl carbamate,a naturally-occurring component of manyalcoholic drinks, was identified as beingundesirable. Canada specified by regulation amaximum limit for ethyl carbamate in whiskyspirit of 150 ppb and the US set guidelines of120 ppb. Control of this compound was onlypossible after the processes resulting in itspresence in spirit were fully understood.

Extensive research in Scotland revealed themechanism of ethyl carbamate production andfacilitated the introduction of very effectivecontrol measures. These measures maintainlevels close to zero and always less than 30 ppbin distilled whisky spirit.

Cook (1990) reviewed the outcome of thisresearch and the resulting control procedures.When barley sprouts during malting, aglycoside, epiheterodendrin (EPH) is present inthe acrospire. This glycoside, which surviveskilning and mashing, is extracted into the wortand converted by yeast enzymes to glucose andisobutyraldehyde cyanohydrin (IBAC). TheIBAC is stable during fermentation, but whenheated above 50oC at distillation it breaks downto form volatile nitriles. There are a multiplicityof potential routes nitriles can follow includingreaction with other beer components or complexreactions often mediated indirectly by copper.Some of the volatile nitriles may escape thesereactions and pass into the distillate. A numberof reactions can remove the nitriles from thespirit, one being the reaction with ethanol to formethyl carbamate.

Two control strategies may be used to preventethyl carbamate reaching the final spirit. Firstly,some varieties of barley produce low levels ofthe glycoside EPH (Cook, 1990), and plantbreeders are now concentrating on incorporatingthis character in all varieties for the distillingindustry. The second strategy relies on the lowvolatility of ethyl carbamate. Control ofdistillation conditions eliminates all the volatileprecursor formed by copper mediated reactions.Non-volatile substances are formed prior to thefinal distillation in malt distillation or beforerectification in grain distillation. Thus, ethylcarbamate can be minimized in the final spirit.

A considerable number of studies have beenpublished on those organoleptically importantcompounds arising either directly or indirectlyfrom the oak barrels in which whisky is matured.The increase in coloring, tannin, dissolved solidsand acid concentrations are not observed whenwhisky is stored in glass, which is proof of theimportance of the oak barrels in the maturationprocess. An analysis of heartwood of theAmerican oak (Quercus albus) gave cellulose(49-52%), lignin (31-33%), pentosans (orhemicelluloses, 22%) and compounds extractedwith hot water and ether (7-11%; Ritter and Fleck,

218 T.P. Lyons

1923). However, when charred oak sawdust wasdirectly extracted with water or 96o GL ethanol,the extracts obtained differed markedly in odorfrom aged whisky. Moreover, none of thevarious fractions of ethanol-soluble oakextractives contained flavors that resemblemature whisky (Baldwin et al., 1967). As aresult, it is now generally held that thematuration process involves not only extractionof compounds from the oak but also chemicalmodifications of at least some of the compoundsextracted from the wood.

For a long time, most of the work reported onthis aspect of maturation of whisky came fromthe laboratories of Joseph E. Seagram and Sonsin the US. More recently, accounts of themechanisms of Scotch whisky maturation havebeen given by Philip (1989) and Perry (1986)and on the maturation of whisky generally, byNishimura and Matsuyama (1989). A theme fromall this work has been the identification of anumber of mechanisms of maturation commonto all whiskies. These divide into addition,subtraction and modification by reaction. Thereis addition of components from the oak wood,including those derived from lignin, tannins andoak lactones. There is the subtraction of volatilecompounds from the maturing whisky by

evaporation and adsorption on the charredsurface of the barrel (Perry, 1986). Lastly thereare reaction processes including establishmentof equilibria among acetaldehyde, ethanol andacetal (Perry, 1986), polymerization reactions(Nishimura and Matsuyama, 1989) andoxidation-reduction reactions (Perry, 1986;Connor et al., 1990). Many of the reactionsinvolve, and are indeed dependent on, thecomponents extracted from the wood.

Looking at several of these reactions in moredetail will serve to illustrate the complexity ofthe maturation process. The work at the Seagramlaboratories, whilst focused on bourbon, isdirectly relevant to Scotch and Irish whiskies forwhich once-used bourbon barrels are extensivelyutilized for maturation. Changes in theconcentrations of organoleptically-importantcompounds during a 12 year storage of a 109o

US proof (54.5o GL) bourbon, on a 100o proof(50o GL) basis, are shown in Figure 15. Thenature and origin of some of these compoundshave been examined in some detail. Among thealdehydes, scopoletin and the aromaticaldehydes syringaldehyde, sinapaldehyde,coniferaldehyde and vanillin are important.According to Baldwin et al. (1967), thesecompounds could be formed by ethanol reacting

Total solids

Co

ncen

trati

on

(g

/100L

) at

100°

pro

of

Ab

so

rban

ce a

t 430 n

m a

t 100°

pro

of

Aldehydes

Furfural

Color

Esters

Tannins

Total acids

Age (years)

1

0

20

40

60

80

100

120

140

160

0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8180

200

220

240

260

2 3 4 5 6 7 8 9 10 11 12

Figure 15. Changes in composition of the vapor at different trays in a Coffey still rectifier used in the manufactureof Scotch grain whisky (from Pyke, 1965).


with lignin in the oak wood to produce coniferylalcohol and sinapic alcohol. Under mildlyoxidizing conditions, these alcohols could beconverted into coniferaldehyde andsinapaldehyde, respectively. Vanillin could thenarise from coniferaldehyde, and syringaldehydefrom sinapaldehyde. The increase in aldehydecontent during maturation is also attributable inpart to formation of acetaldehyde by oxidationof ethanol. Formation of ethyl acetate probablyaccounts for the steady rise in the ester contentof whisky during maturation.

Several other groups of compounds notdescribed in Figure 14 are also important in thematuration process. Monosaccharide sugars arefound in mature whisky, and probably arise fromthe pentosans and other polysaccharides in theoak wood. Otsuka et al. (1963) reported that amature Japanese whisky contained xylose,arabinose, glucose and fructose, while Black andAndreasen (1974) added rhamnose to this listwhen they analyzed a mature bourbon. The latterworkers found that the concentrations ofarabinose and glucose increased at a faster ratethan those of xylose and rhamnose over a 12year maturation period. Salo et al. (1976) alsodetected low concentrations of mannose andgalactose in a matured Scotch malt whisky inaddition to the sugars already noted. Theconcentrations of sugars in mature whiskies (ofthe order of 100 mg/L) are too low to suggestany sweetening effect on the beverage. Phenolsare also detectable in mature whisky, althoughsome of these probably arise during mashing(Steinke and Paulson, 1964) or from maltproduced using peat-fired kilns (MacFarlane,1968). However, Salo et al. (1976) reported anincrease during a one year maturation of aScotch malt whisky in the concentration ofeugenol, which is a major phenol extracted fromoak chips by ethanol (Suomalainen andLehtonen, 1976). Also present in maturewhiskies are sterols, which may precipitate inbottled whisky stored at room temperature.Black and Andreasen (1973) found campesterol,stigmasterol and sitosterol in mature bourbon,in addition to sitosterol-D-glucoside, althoughthe possibility that some of these were formedduring mashing cannot be excluded. Finally,reference has already been made to the whiskylactone, ß-methyl-octalactone, and its origin.

Not surprisingly, the nature and amounts of

compounds extracted from charred oak wooddepend on the ethanol concentration of thewhisky. It is to some extent an advantage tomature whisky at a high proof, since this requiresfewer barrels and saves on storage space. Until1962 the US Treasury Department limited thebarrelling proof of whisky to a maximum of 110o

US proof (55o GL). In anticipation of this limitbeing raised to 125o US proof (62.5o GL),Baldwin and Andreasen initiated a series ofexperiments in 1962 to establish the importanceof barrelling proof on changes in color andconcentrations of organoleptically-importantcompounds during maturation of bourbonwhiskies. Their report in 1973 indicated thatcolor intensity and congener concentration ofwhiskies matured for 12 years decreased as thebarrelling proof was raised from 109o US proof(54.5o GL) to 155o US proof (77.5o GL). Theone exception was the higher alcohol content,which remained approximately constant.

Is it really Scotch?

Analysis of the variety of compounds in spirit isalso of interest for reasons of identifying orauthenticating product origins or types. Adamet al. (2002) investigated whether malt, blendedand grain whiskies could be differentiated basedon content of various metals. While it was notpossible to define a ‘metal fingerprint’ that wouldidentify a whisky as to origin, malt whiskies hadmarkedly higher concentrations of copper thanblended or pure grain whiskies. Differences weresignificant with malt whiskies containing 385-480 ng Cu/mL and grain and blended whiskiescontained 131-242 ng/mL. Since malt Scotch isproduced in traditional copper pot stills whilegrain whisky used in blending is made incontinuous patent stills, copper content wassuggested as a means of distinguishing a maltScotch from a blended product.

Another approach to distinguishing betweenmalt and blended products is based on thedifferences between barley, which is used toproduce malt Scotch and the corn (maize) thatis typically used in production of grain whisky.Parker et al. (1998) found that the ratio of 12Cand 13C isotopes in volatiles such asacetaldehyde, ethyl acetate, n-propanol andothers differed among whiskies. This was a

220 T.P. Lyons

reflection of the different types of photosyntheticmetabolism of barley, a cool season plant withC3 metabolism, versus maize, which beingadapted to hotter climates has C4 metabolism.

There is no doubt value in being able tochemically define the special characteristics thatdistinguish an ordinary whisky from malt Scotch.Authentication is often needed; and a thoroughunderstanding of the chemical basis oforganoleptic properties might pave the way toproduction of even finer whiskies than those weenjoy most today. For the moment, however, themagic remains. The distiller’s art continues toelude the chemist.

Acknowledgments

The author wishes to thank his friends andcolleagues in whisky production in manycountries of the world for their cooperation andwillingness to supply information on theprocesses of whisky production. Particularthanks are due to Dr. Gareth Bryan-Jones ofUnited Distillers, Menstrie, Scotland, for hisassistance in providing research reports.

References

Anderson, R.J., G.A. Howard and J.S. Hough.1972. ‘Proceedings of the European BrewingConvention, Vienna. p 253.

Aylott, R.I., A.H. Clyne, A.P. Fox and D.A.Walker. 1994. Analyst 119:1741-1746.

Baldwin, S. and A.A. Andreasen. 1973. Journalof the Association of Official AnalyticalChemists 57:940.

Baldwin, S., R.A. Black, A.A. Andreasen andS.L. Adams. 1967. Journal of Agriculture andFood Chemistry 15:381.

Barbour, E.A. 1983. Taxonomy of some lacticacid bacteria from scotch whisky distilleries.PhD Thesis, Heriot-Watt University.

Barnard, A. 1887. The whisky distilleries of theUnited Kingdom. Proprietors of HarpersWeekly Gazette, London, pp. 457.

de Becze, G.I. 1967. Encyclopedia of IndustrialAnalysis. Vol. 4:462. Interscience Public-ations, New York.

Beech, F.W. 1972. Progress in Industrial Micro-

biology I I:178.Black, R.A. and A.A. Andreasen. 1973. Journal

of the Association of Official Analytical Che-mists 56:1357.

Black, R.A. and A.A. Andreasen. 1974. Journalof the Association of Official Analytical Che-mists 57:111.

Black, R.A., A.A. Rosen. and S.L. Adams. 1953.Journal of the American Chemical Society75:5344.

Bober, A. and S.W. Haddaway. 1963. Journalof Gas Chromatography 1 12: 8.

Brander, M. 1975. A Guide to Scotch Whisky.Johnson and Bacon, Edinburgh. p. 96.

Bryan-Jones, G. 1976. Lactic Acid Bacteria inBeverages and Foods. Proceedings of theFourth Long Ashton Symposium. (J.G. Carr,C.V. Cutting and G.C. Whiting, eds). Acad-emic Press, London. pp. 165-175.

Cameron Taylor, I.B. 1970. Whisky Export, 2ndissue, June-Sept. Aldiffe Publishing Co. Ltd.,Wilmslow, Cheshire.

Carter-Tijmstra, J. 1986. Proceedings of SecondAviemore Conference on Malting, Brewingand Distilling. pp. 413-416.

Connor, J.M., A. Paterson and J.R. Piggot. 1990.Proceedings of Third Aveimore Conference onMalting, Brewng and Distilling. pp. 460-463.

Cook, R. 1990. Proceedings of Third AveimoreConference on Malting, Brewing and Dist-illing. pp. 237-243.

Court, R.E. and V.H. Bowers. 1970. ProcessBiochemistry 5:17.

Cromwell, E.A. and J.F. Guymon. 1963.American Journal of Enology and Viniculture12:214.

Daiches, D. 1969. Scotch whisky, its past andpresent. Andre Deutsch, London. p. 168.

Dolan, T.C.S. 1976. Journal of the Institute ofBrewing. 82:177.

Duncan, R.E.B. and J.M. Philp. 1966. Journalof the Science of Food and Agriculture 17:208.

Dunnett, A. 1953. The land of scotch. ScotchWhisky Association, Edinburgh. p. 170.

Geddes, P.A. and H.L. Riffkin. 1989. DistilledBeverage Flavour (ed. Piggot, J.R. Ellis,Horwood, London. pp. 193-199.

Griffin, O.T. 1972. Process Biochemistry 7:17.Gunn, N.M. 1935. Whisky and Scotland. Rout-

ledge, London. p 198.


Harrison, G.A.F. 1970. Journal of the Instituteof Brewing 76:486.

Harrison J.S. and J.C.J. Graham. 1970. In: Theyeasts. (A.H. Rose and J.S. Harrison, eds.). Vol.3:283-348. Academic Press, London.

Heukelekian, H. and A. Heller. 1940. Journal ofBacteriology 40:547.

Hopkins, R.H. and S. Wiener. 1955. Journal ofthe Institute of Brewing 61:493.

Hough, J.S., D.F. Briggs and R. Stevens. 1971.Malting and brewing science. Chapman andHall Ltd., London. p. 678.

Institute of Brewing Analysis Committee,Recommended methods of analysis. 1975.Journal of the Institute of Brewing 81:368.

Jacobs, M.B. 1947. American Perfumery andEssential Oil Review. p 157.

Jack, F. 2003. Development of guidelines forthe preparation and handling of sensorysamples in the Scotch whisky industry. J. Inst.Brewing 109:114-119.

Kahn, J.H. 1969. Journal of the Association ofOfficial Analytical Chemists 52: 1166.

Kahn, J.H., P.A. Shipley, E.G. Laroe and H.A.Conner. 1969. Journal of Food Science34:587.

Kleber, von W. and N. Hurns. 1972. Brauwiss-enschaft 25: 98.

Kulka, D. 1953. Journal of the Institute ofBrewing 59: 285.

Liebmann, A.J. and B. Scherl. 1949. Industrialand Engineering Chemist 41:534.

Lloyd Hind, H. 1948. Brewing Science andPractice, Vol. 1:505. Chapman and Hall,London.

Lyons, T.P. 1974. The Brewer. p 634.McDowell, R.J.S. 1975. The Whiskies of

Scotland. 3rd edit. John and Murray, London.p. 166.

MacFarlane, C. 1968. Journal of the Institute ofBrewing 74: 272.

MacFarlane, C., J.B. Lec and M.B Evans. 1973.Journal of the Institute of Brewing 79:202.

McGuire, E.B. 1973. Irish Whiskey. Gill andMacMillan, Dublin. p. 462.

MacKenzie, K.G. and M.C. Kenny. 1965.Journal of the Institute of Brewing 71:160.

Mann, E.A. 1911. Government Analyst for West-ern Australia. Perth, W. Australia. pp. 1-12.

Merritt, N.R. 1967. Journal of the Institute ofBrewing 73:484.

Morrison, R.L. 1962. American Journal ofEnology and Viniculture 13:159.

Murtagh, J.E. 1970. M.Sc. Thesis: UniversityCollege of North Wales, Bangor.

Nettleton, J.A. 1913. The manufacture of whiskyand plain spirit. Cornwall and Sons, Aberdeen.p. 616.

Nishimura, K. and M. Masuda. 1971. Journalof Food Science 36:819.

Nishimura, K and R. Matsuyama. 1989. TheScience and Technology of Whiskies. (J.R.Piggott, R. Sharp and R.E.B. Duncan, eds.).Longman Scientific and Technical, Harlow. pp.235-263.

Otsuka, K. and K. Morinaga and S. Imai. 1963.Nippon Jozo Kyokai Zasshi 59:448.

Parker, I.G., S.D. Kelly, M. Sharman, M.J.Dennis and D. Howie. 1998. Investigations intothe use of carbon isotope ratios (13C/12C) ofScotch whisky congeners to establish brandauthenticity using gas chromatography -combustion isotope ratio mass spectrometry.Food Chem. 63:423-428.

Perry, D.R 1986 Proceedings of Second Avie-more Conference on Malting, Brewing andDistilling. pp. 409-412.

Pert, D. and E. Woolfendon. 1986. Proceedingsof Second Aviemore Conference on Malting,Brewing and Distilling. pp. 417-424.

Philip, J.M. 1989. The Science and Technologyof Whiskies. (J.R. Piggott, R. Sharp and R.E.B.Duncan, eds). Longman Scientific andTechnical, Harlow. pp. 264-294.

Preece, I.A. 1947. Journal of the Institute ofBrewing 53:154.

Preece, I.A. 1948. Journal of the Institute ofBrewing 54:141.

Preece, I.A. and M. Shadaksharaswamy. 1949a.Journal of the Institute of Brewing 55:298.

Preece, I.A. and M. Shadaksharaswamy. 1949b.Biochemical Journal 44:270.

Preece, I.A. and M. Shadaksharaswamy. 1949c.Journal of the Institute of Brewing 55:373.

Pyke, M. 1965. Journal of the Institute ofBrewing 71:209.

Rae, I.L. 1967. Process Biochemistry I, 8:407.Ritter, G.J. and L.C. Fleck. 1923. Industrial and

Engineering Chemistry 15:1055.

222 T.P. Lyons

Robb, J.M. 1950. Scotch Whisky. W. and R.Cham-bers, London. p 197.

Ross, J. 1970. Whisky. Routledge and KeganPaul, London. p. 158.

Royal Commission on Whisky and Other PotableSpirits. 1909. Appendix Q, p. 229.

Salo, P. 1970. Journal of Food Science 35:95.Salo, P. 1975. Proceedings of the International

Symposium on Aroma Research. CentralInstitute for Nutrition and Food Research,TNO, Zeist, Netherlands. p 121.

Salo, P., L. Nykänen and H. Suomalainen. 1972.Journal of Food Science 37:394.

Salo, P., M. Lehtonen and H. Suomalainen. 1976.Proceedings of the Fourth Symposium onSensory Properties of Foods. Skövde, Sweden.

Schidrowitz, P. 1902. Journal of the ChemicalSociety. p 814.

Schidrowitz, P. and F. Kaye. 1905. Journal ofthe Chemical Society. p. 585.

Scotch Whisky Order. 1990. Stationery Office,London.

Sillett, S.W. 1965. Illicit Scotch. Beaver Books,Aberdeen. p. 121.

Simpson, A.C. 1968. Process Biochemistry 3,1:9.

Simpson, B., A. Troon, S.T. Grant, H. Mac-Diarmid, D. MacKinlay, J. House and R.Fitzgibbon. 1974. Scotch Whisky. Mac-Millan,London. p. 120.

Singer, D.D. and J.W. Stiles. 1965. Analyst90:290.

Stark, W.H. 1954. Industrial Fermentations (L.A.Underkofler and R.J. Hickey, eds). ChemicalPublishing Company, New York. Vol. 1, pp.17-72.

Steinke, R.D. and M.C. Paulson. 1964. Journalof Agricultural and Food Chemistry 12: 381.

Suomalainen, H. 1971. Journal of the Instituteof Brewing 77:164.

Suomalainen, H. and M. Lehtonen 1976.Kemia-Kemi 3, 2:69.

Suomalainen, H. and L. Nykänen. 1966. Journalof the Institute of Brewing 72:469.

Suomalainen, H. and L. Nykänen. 1970a.Naeringsmidelidustrien 23:15.

Suomalainen, H. and L. Nykänen. 1970b.Process Biochemistry. July 1.

The Scotch Whisky Order. 1990. StatutoryInstruments No 998. HM Stationery Office,London.

Thorne, C.B., R.L. Emerson, W.J. Olsen andW.H. Paterson. 1945. Industrial andEngineering Chemistry 37:1142.

Valaer, P. 1940. Industrial and Engineering Che-mistry 32:935.

Valaer, P. and W.H. Frazier. 1936. Industrial andEngineering Chemistry 28:92.

Walker, J.W., D.M. Vaughan and H.L. Riffkin.1990. Proceedings of Third Aveimore Conf-erence on Malting, Brewing and Distilling. pp.421-426.

Warricker, L.A. 1960. Journal of the Science ofFood and Agriculture 11:709.

Wilson, J. 1973. Scotland’s malt whiskies, adram by dram guide. Famedram Ltd.,Gartochan, Dunbartonshire, Scotland. p. 109.

Wilson, J. 1975. Scotland’s distilleries. a visitor’sguide. Famedram Ltd., Gartochan, Dunbart-onshire, Scotland. p. 104.

Wilson, Ross. 1959. Scotch made easy. Hutchin-son, Ltd., London. p 336.

Wilson, Ross. 1970. Scotch: the formative years.Constable, London. p 502.

Wilson, Ross. 1973. Scotch, its history andromance. David and Charles, Newton Abbot,Devon, England. p 184.

Wobben, H. J., R. Timmer, T. ter Heide and P.de Valois, P. 1971. Journal of Food Science35:464.

Zobell, C.E. 1943. Journal of Bacteriology46:39.

Tequila production from agave 223

Chapter 15

Tequila production from agave: historical influences and contemporaryprocesses

MIGUEL CEDEÑO CRUZ

Tequila Herradura, S.A. de C.V. Ex-Hda San Jose del Refugio Amatitán, Jalisco, México

Introduction: agave beverage alcoholproducts

Tequila is classically associated with Mexico andhas been long identified as more than just thenational drink, but a symbol of our culture andenvironment. As the reach of the global tradehas extended, tequila is now consumed aroundthe world.

This alcoholic beverage is obtained from thedistillation of fermented juice of the agave plant,Agave tequilana Weber var. azul. A. tequilanais often confused with various cacti growing insame desert environment, but they belong todifferent botanical families: agaves to Agavaceaeand cacti to Cactaceae families, respectively.

There are two classifications of tequila basedon agave content: 100% tequila, produced usingonly agave, and 51% tequila, which containssugars from other sources such as cane and cornadded in the fermentation step, at up to 49% byweight. It is important to note that only tequila51% can be exported in bulk outside Mexicosince 100% tequila must be state-bottled.

Based on maturation, there are fourclassifications of tequila: silver, withoutmaturation, gold, containing permitted additivesand colors (generally caramel color), agedtequila (Reposado) matured at least two monthsin white oak tanks with a maximum capacity of

159 gallons or barrels, and extra-aged (Añejo)matured at least 12 months in white oak barrels.

More than 300 brands (labels) of tequila areproduced in the 104 distilleries registered inMexico by the Tequila Regulatory Council (CRTby its acronym from the Spanish: ConsejoRegulador del Tequila), which verifiescompliance with the Mexican official standard(Secofi, 1994) through routine on-siteinspections.

Tequila is differentiated from the beverageknown as ‘mezcal’ by the type of agave used inits production. Mezcal is made from Agavepotatorum, which grows in the states of Oaxaca,San Luis Potosi, Guerrero and Zacatecas. Mostmezcal producers use a rudimentaryfermentation and distillation process (Sánchez,1991). There is no technical reason for, or anyimprovement in the organoleptic characteristicsfrom, the worm inside the bottles of some mezcalbrands. The worm is primarily a commercialploy. Worms are grown in agave plants andintroduced manually in the bottling line.

‘Pulque’ is another beverage obtained byfermentation of the juice obtained from severalspecies of agave, A. atrovirens and A. salmianaamong others, by a complex succession of yeast

224 M. Cedeño Cruz

and bacteria that produce ethanol, a diversity ofchemical compounds, and some polymers thatgive a sticky consistency to the final product(Rzedowski, 1978; Sánchez Marroquin andHope, 1953). Pulque is sometimes mixed withfruits or vegetables, but has poor stability as it isneither distilled nor pasteurized.

Agave plants still serve as food in some statesof Mexico; and other fermented regionalbeverages are produced (e.g., ‘Sotol’ in the stateof Chihuahua and ‘Bacanora’ in the state ofSonora), but only tequila and more recentlymezcal have reached international recognition.Another difference between tequila and mezcalor other regional drinks is that both are subjectto an official standard that for tequila is NOM-006-SCFI-1994 (Secofi, 1994), and productionis supervised by the Mexican government. Table1 shows the main differences among the regionalbeverages.

Origin and history of tequila

The word ‘tequila’ is believed to originate fromthe tribe of ticuilas who long ago inhabited thehillside of a volcano bearing the same namelocated near the city of Tequila. Another possibleorigin is the Nahuatl word tequitl, which meanswork or employment, and the word tlan, whichmeans place. Therefore ‘tequila’ would meanplace in which labor or work is done.

The oldest reference to the existence of agaveand its different uses is from the era before theSpaniards in several codices preserved to thepresent time (a codex, from the Latin codexmeaning board or writing tablet, is a manuscriptvolume, especially of a classic work or scripture).The most important is the Tonalmatlnahuatl

codex, which notes that certain tribes had learnedto cook agave plants and used them as food andto compensate for the lack of water in desertlands. Also, these tribes discovered that cookedagave when soaked in water would ferment,producing a much appreciated beverage. In fact,this primitive and rudimentary method was usedfor centuries to produce beverages from agave,considered a sacred plant possessing divineproperties. In other codices, such as Nutall,Laud, Borgia and Florentine, there are manyreferences to uses of the agave plant for soapmanufacture, a source of fiber, footwear,medicine and sewing needles as well as thread,paper and rope. In fact, Indians distinguishedthe different species of agave by color, size, stem,leaf width and the different uses given each plant(Muria, 1990). The great religious importanceof agave was apparent in those codices, as onlywarriors and priests used fermented drinks inritual ceremonies.

In pre-hispanic Mexico the general name forall species of agave (or mezcal as it is alsoknown) was Metl, which is a representation ofthe goddess Mayahuel (Figure 1). The alcoholicdrink produced was called Iztac octli (whitewine). The first Spaniards to arrive in Mexicoreferred to the plant as ‘maguey’, a name usedfor an identical plant they had seen in theCaribbean Islands, where they first encounterednew world plants and animal life (Bottorff,1971). It was not until the arrival of theSpaniards, who brought knowledge ofdistillation techniques, that tequila took itspresent form (Goncalves, 1956). Tequila isconsidered the first distilled beverage in NorthAmerica and has been given several differentnames including agave wine, mezcal wine andfinally tequila.

Table 1. Agave alcoholic beverages produced in Mexico.

Beverage Agave Distillation Producing Fermentation:species process region main microorganisms

Tequila tequilana Yes Jalisco YeastMezcal potatorum, salmiana, angustifolia Yes Oaxaca, San Luis Potosi YeastBacanora pacifica, palmieri, angustifolia Yes Sonora YeastSotol angustifolia, dasylirion Yes Chihuahua YeastPulque atrovirens, salmiana No Central Yeast, bacteria


Figure 1. Representation of the goddess Mayahuel.

There are specific regions in Mexico whereagave should be grown, where tequila must beproduced and where 100% tequila must be state-bottled. This area comprises the municipalities

of the following states: Jalisco (all), Nayarit (8),Michoacan (30), Guanajuato (7) and Tamaulipas(11) (Figure 2). The oldest, the Tequila-Amatitánregion that comprises the Amatitán village,developed at the end of the 17th century. Thesecond region, the Jalisco Highlands, appearedin the last decade of the 19th century (Luna,1991).

The first tequila production process with acommercial purpose was established in the cityof Tequila around the end of the 18th century.The main consumers were in the mining zoneslocated in the state of Jalisco. The Spaniards triedto suppress consumption of tequila in order toreduce competition for brandy and other winesimported from Spain with a decree signed byCarlos III forbidding sale and production oftequila under the pretext that its consumptionwas the cause of several illnesses. The resultswere negative. The governor of the region laterissued a decree imposing a tax on tequila in orderto enrich the royal coffers, thus permitting itssale in all of New Spain. By the end of the 19thcentury, expansion of the tequila industry, helpedby the railway, was evident. However it was notuntil the first casks were exported to the US thattequila was known beyond Mexico’s borders.

Tamaulipas (11)

Guanajuato (7)

Michoacán (30)

Jalisco (all)

Nayarit (8)

Figure 2. States and municipalities for agave cultivation and tequila production in Mexico.

226 M. Cedeño Cruz

The agave plant

According to Granados (1985), the genus Agave,which means ‘noble’ in Greek, was defined byLinneaus in 1753 when he described the plantA. americana as the first agave species knownto science. Agave plants, which are oftenconfused with cacti, belong to the familyAgavaceae and are succulent plants withspirally-arranged leaves forming a rosette. Somehave definite trunks, but more often they arenearly stemless. The leaves are bluish green incolor, over 1 m long in mature plants, and endin a sharpened brown thorn. As Backman (1944)pointed out, the widespread distribution of some300 species, combined with the fact that theplants require approximately 8-12 years tomature and hybridize very easily, make thetaxonomic and phylogenetic study of the genusAgave extremely complicated and very difficult.The family Agavaceae includes 20 genera andnearly 300 species. Of these, around 200 arefound in Mexico. In Figure 3, a diagram of anagave plant in soil shows the main structuralfeatures.

Figure 3. Structure of Agave tequilana Weber var. azul.

Agave has commercial importance as a fibersource; and this industry developed in the 19thcentury in Mexico mainly in the Yucatan statewith the cultivation of henequen, A. fourcroydes,to export fiber to Europe. However, thePhilippines, Indonesia and Tanzania in Africacompeted in this market with fiber produced byA. sisalana. The use of agave as a forage forcattle is a recent development for milkproduction, but must be enriched with a proteinsource. Other agaves such as A. americana, A.victoria regina, A. filifera and A. stricta amongothers are used as ornamentals.

In the beverage industry, A. salmiana and A.atrovirens are valued for pulque production, A.potatorum, A. salmiana and A. angustifolia formezcal production, A. pacifica, A. palmieri andA. angustifolia for the elaboration of bacanoraand A. dasylirion and A. angustifolia for sotolproduction. Finally, A. tequilana Weber var.azul, named around 1900 by the Germanbotanist Weber (Diguet, 1902), is used to producetequila.

CULTIVATION AND HARVEST

The blue agave, as it is known, is the only speciesout of hundreds of Agavaceae with theappropriate characteristics for tequilaproduction. These include a high inulinconcentration, low fiber content and thechemical compounds present in the plant thatcontribute to the final taste and flavor of tequilato give the beverage its particular character.Some have attempted to produce tequila in otherstates with other agave species, but withoutsuccess. The blue agave is mainly cultivated inthe state of Jalisco in the two regions with theright climate and soil composition for its growth,namely the Highlands and the Tequila-Amatitánregion. The temperature conditions for goodagave yields are a minimum of 3°C, an optimumof 26°C and maximum of 47°C. Soil should befertile but not very deep, 30 to 40 cm. Gooddrainage is required to avoid effects of flooding,which are very harmful for development of theagave. Agave are planted at 800 to 2500 mabove sea level where the annual rainfall is about800 to 900 mm. The correct planting time is


immediately before the rainy season, from Juneto September, so that the plants do not sufferfrom water stress during the first year of growth.

Propagation is accomplished by vegetativemeans in agave. Sexual reproduction via seedsis not usual. Asexual bulbils develop in theinflorescence at the base of the flowers,producing small plants that after some timedetach themselves from the floral peduncle andfall to the soil where they root. Another modeof asexual propagation is by suckers, which area characteristic type of lateral bud or branchdeveloping at the base of the main stem. Plantsdeveloping near each mother plant are separatedat the age of 3-4 years. These baby plants arecalled ‘first-class seed’ because they are betterand healthier (Sánchez, 1991). In practice, peopleuse the word ‘seed’ to refer to such youngplants, but from a botanical point of view theseare rhizome shoots or suckers (Valenzuela,1992). Plant cell culture is used by some tequilacompanies in order to maintain plants ofconsistent quality, reduce the cost of labor forsowing the plants, eliminate use of fungicidesand antibiotics during sowing, and have biggerplants with higher inulin content at the end ofthe cultivation period.

Land for agave cultivation must be cleared anddeep-ploughed, sometimes twice. Agave isplanted approximately 2-4 m apart in straightlines called ruts. Sowing is done by hand in holes15 cm in depth. Plant density is around 2000-4000 plants per hectare, depending on theplantation system used; and yields can bebetween 30,000 and 200,000 kg/ha, assumingthat the weight of a harvested plant varies from15 to 50 kg. This variation is caused bydifferences in soil conditions, quality of plantssown, rainfall, pests and fertilization. Sometimesagave is interseeded with nitrogen-fixing cropssuch as peanuts, beans, chickpeas or soybeans.After the agave has been in the soil for a year,visual inspection is conducted and sick or deadplants are replaced. This operation is called re-seeding. The percentage of dead plants dependson soil and plant characteristics, but is generallyfrom 8 to 15% (Pérez, 1990).

Agave plants regularly host borer insects thatlive in the stems, leaves and fruits during thelarval stage. These include butterflies of thefamily Megathymidae and moths of the familyProdoxidae. Also, the fungi Diplodia

theobromae and Colletotrichum agavae maycause serious damage to agave leaves (Halffter,1975; Agricultural Research Service, 1972).

Agave plants, adapted to hot and very aridconditions, employ crassulacean acidmetabolism (CAM). Crassulacean acidmetabolism is similar to photosynthetic processesin other hot temperature-adapted plants in thatthe enzyme phosphoenolpyruvate (PEP)carboxylase is used to fix CO2 into a 4-carboncompound (C4 metabolism). However in mosthot season plants the light-dependent reactionsand CO2 fixation are separated in different cellsof the leaf. In contrast, in plants with crassulaceanacid metabolism such as cacti and agave thesereactions occur at different times of the day.Carbon dioxide (CO2) enters the leaf throughstomata with the simultaneous loss of water. Byonly requiring the stomata to be open during thecooler night for CO2 uptake and keeping themclosed during the hotter day, the agave plantslose much less water (Ehrler, 1967). Malic acidformed by CO2 fixation using PEP carboxylaseaccumulates during the night, and is broken downduring the day (Figure 4). The CO2 that isreleased can be re-fixed by the normalphotosynthetic Calvin cycle (C3 metabolism) toproduce sugars and then inulin.

The thick cuticle overlying the epidermis,which is quite evident in tequila agave,apparently prevents damage to the leaf from hightemperatures (Casado and Heredia, 2001). Thiswaxy cuticle produces turbidity in tequilabecause it dissolves in the distillation step andproduces a haze in the final product when it isdiluted or cooled. One way to avoid this is totreat tequila with activated charcoal and to filterit through pure cellulose filter pads. This,unfortunately, results in the loss of some aromas.

Agave fertilization is determined by soilcomposition, plant age, and the type of chemicalcompound used. The normal procedure is to useurea as a nitrogen source in amounts of 30-70 gper plant incorporated directly into the soil. Insome areas, phosphorus and potassiumfertilization is also required. As some agaveregions are also involved in cattle, swine, andchicken production, manure is sometimesemployed for fertilization (GEA, 1992).

The average maturity time for agave is sevenyears; and at this time the content of inulin is atits highest level (Mendez, 1999). Every year

228 M. Cedeño Cruz

following planting, fields are cultivated to loosenthe soil and weed and pest control is carried out.Each plant matures individually. Harvest beginsin the seventh year. The leaves are cut from thebase and left in the field to recycle nutrients.The harvested plants free of leaves look like largepineapples and weigh from 20 to 90 kg. Theyare transported to the distillery. Only the betterplants, meaning those of good size and highinulin content (measured as reducing sugars),

are harvested; but on the 12th year, all plantsremaining (the weakest plants) are cut. Theseare called ‘drag’ and are generally discarded.Agave composition varies seasonally, but anaverage would be (wet basis) 27% (w/w)reducing sugars, a juice content of 0.572 ml/gand a pH of 5.2.

In Figure 5 the amount of agave harvested fortequila production is presented for the period1995-2002. These figures show, first of all, a

CO2

CO2

H2O

CO2 fixation

C4 acids

PEP

Air

NightDay

Leaf

Glycerol

Calvin

Cycle

Growth

InulinStomata (open)

Figure 4. Schematic representation of carbon dioxide fixation by agave (PEP: Phosphoenolpyruvate).

Figure 5. Agave harvested for tequila production (CRT, 2002).

0

To

ns

of

agav

e('0

00s)

Tequila100%

Tequila

Total

800

1995 1996 1997 1998 1999 2000 2001 2002

700

600

500

400

300

200

100


decrease in total agave production. This reflectsthe shortage caused by disease in agaveplantations in the year 2000, which was mainlydue to a bacteria Erwinia carotovora and afungus Fusarium oxysporum, which destroyedmore than 25% of the total agave plants fortequila. From this graph, it is evident that theagave utilization for 100% tequila is growing,since that beverage is being recognized as apremium product not only in Mexico but alsooverseas.

Tequila production

Tequilas differ greatly depending on agavequality and origin (Highlands or Tequila-Amatitán regions). The production process alsostrongly influences quality of the final product.Some distilleries still employ rudimentaryproduction methods, just as they did severaldecades ago (Pla and Tapia, 1990). Mostcompanies, however, employ technological

advances that improve process efficiency andconsistency; and some have implemented a ISO-9000 standard. Whatever the process used,tequila manufacture comprises four main steps:cooking, milling, fermentation and distillation.An extra step, maturation, is required for agedproducts. Figure 6 shows a flow diagram for thetequila production process, which is explainedin detail in the following paragraphs.

Processing harvested agave

Agave is transported from the fields to thefactories as soon as possible to avoid weightlosses, because today most distilleries pay byweight and not by inulin content. The heads areunloaded from the truck in the receiving area ofthe factory and must be protected from the sunand rain in order to avoid withering and fungalgrowth. Although the agave has already beeninspected during growth and at harvest, it isexamined again to reject visually unacceptable

Figure 6. Flow diagram for the tequila production process.

AGAVE

MATURATION

BOTTLING

DILUTION

2ndDISTILLATION

COOKING MILLING BAGASSE

Autoclave

Yeast

Oven

1stDISTILLATION

STILLAGE STILLAGE

FERMENTATION

DISSOLUTION OFOTHER SUGARS

230 M. Cedeño Cruz

plants or those damaged by pests. At this point,a representative sample of agave is taken forlaboratory analysis. Modified AOAC (1990)procedures are used to determine reducing sugarcontent (after acid hydrolysis of inulin) alongwith pH, moisture, dry weight, juice and ashcontent.

Agave heads usually weigh between 20 and60 kg, although some can reach 100 kg. Theyare cut to sizes that facilitate uniform cookingand handling. Different agave cutting systemsexist, but the use of axes and a specialized toolcalled ‘coa’ are the most popular. The heads arecut in halves or quarters, depending on theweight, and the pieces are arranged manually inan oven or autoclave. Band saws can also beused to cut the agave heads, since they are fasterand less labor-intensive than the manualprocedure, but the belts break frequently becauseof the resinous consistency of agave. Somefactories tear uncooked agave first with a knifeand place the resulting pieces mixed with waterinto autoclaves to be cooked.

OTHER RAW MATERIALS FOR TEQUILAPRODUCTION

When producing 100% agave tequila, the onlysource of carbohydrate is the inulin hydrolyzedfrom agave in the cooking step. Inulin, a polymerof fructose with a terminal glucose, belongs tothe family of fructans (Figure 7). After starch,fructans are the most abundant non-structuralpolysaccharide found in nature.

For other kinds of tequila the law permits useof other sugars in amounts up to 49% by weightin wort formulation. There are no legalspecifications regarding the type of adjunct sugarsources to be used in tequila manufacture; andtheoretically any kind of fermentable sugar isallowed in wort formulation. In practice and froman economic point of view, only cane sugar,Piloncillo, cane molasses and acid- or enzyme-hydrolyzed corn syrup are employed. Canesugar is received in 50 kg bags and stored in adry, cool place for subsequent utilization.Piloncillo consists of brown cones of crystallizedcomplete cane juice, sometimes individuallywrapped in corn or cane leaves and packed insacks. Cane molasses is also used but it is difficultto handle and there is risk of spoilage over along storage period. Acid- or enzyme-

hydrolyzed corn syrup may be used to formulatethe wort. All sugars used in wort formulationare routinely analyzed by measuring solidscontent and reducing and fermentable sugarcontent.

CH2OH

O

HOH

H

H

OH

H H

OH

CH2OH O

H

OH

OH

H

O

O

CH2

CH2OH

H

CH2OH O

H

OH

OH

H

H

Glucose

Fructose

n

Figure 7. Structure of inulin, in which chain length, n, variesfrom 3 to 19, depending on the maturity of the agave plant

(Mendez, 1999).

The cooking step: hydrolysis of inulin

Cooking the agave serves three purposes. First,the low pH (4.5) together with the hightemperature hydrolyzes inulin and othercomponents of the plant. A recent study (Pinal,2001) showed that the concentrations of furfural,5-hydroxymethyl furfural, 5-methylfurfural and2-acetylfuran increase with length of cookingtime. These components are important as theymay influence fermentation rate as well as theorganoleptic characteristics of the final tequila.For this reason, time control during cooking isvery important. In addition, cooked agave has asoft consistency that facilitates the millingoperation.

In the pre-Hispanic era, agave was cooked inholes filled with stones heated using wood forfuel. The stones retained the heat for the timeneeded to cook the agave. Nowadays some


additional steam, cooking slowly in theremaining heat. This step produces a syrup witha high sugar concentration (>10% by weight)that is later used to formulate the initial wort. Tocalculate the yield and efficiency of this step,the amount of cooking honey and its reducingsugar content as well as the cooked agave aremeasured. Figure 8 illustrates the temperatureprofile over time for cooking agave in anautoclave compared with an oven.

The main difference between autoclaved andoven-baked agave is that careful control ofcooking time, temperature and steam pressuremust be maintained in autoclaves to preventovercooking or burning the agave. Overcookinggives a smoky taste to the tequila, increases theconcentration of furfural in the final product andreduces ethanol yield due to the caramelizationof some of the fermentable agave sugars. Thisis why some factories with both cooking systemsreserve the ovens for their better-qualityproducts. Although it is easier to obtain well-cooked agave in an oven than in an autoclave,there is no difference in terms of flavor andfermentability between agave cooked inautoclaves or in ovens if both are correctlycontrolled.

Extraction of agave juice: milling

Milling has gone through three historical stages.In ancient days cooked agave was crushed with

distillers have replaced the heated stone holeswith brick ovens and heating is accomplishedby steam injection after the chopped raw agavehas been introduced into the oven. Oven cookingis slow, and steam injection lasts around 36-48hrs to obtain temperatures of 100°C. After thatperiod, the steam is shut off and the agave is leftin the oven for a further two days to completethe cooking process. During this step, a sweetliquid called ‘cooking honey’ is collected andused later as a source of free sugars, mainlyfructose. Also during this step some of the sugarsare caramelized; and some of the compoundsthat contribute significantly to the aroma andflavor in wort formulation are due to its highcontent of fermentable sugars (>10% w/v).Finally, the oven door is opened to allow thecooked agave to cool. The agave is then readyfor milling.

In most distilleries brick ovens have beenreplaced by steel autoclaves. Autoclaves havesuperior efficiency and allow good pressure andtemperature control, enabling homogeneous andeconomic cooking. In a typical autoclavecooking operation, steam is injected for 1 hr sothat steam washes the agave. This condensedliquid is called ‘bitter honey’ and is discardedbecause it contains waxes from the agave cuticleand has a low sugar content (<1 % w/w). Steamis injected for an additional 6 hrs to obtain apressure of 1.2 kg/cm2 and a temperature of121°C. At the end of that time the agave remainsin the autoclave for another 6 hrs without

0

0 20 40 60 80

Time (hours)

Tem

per

atu

re(°

C)

Autoclave

Oven

140

120

100

80

60

40

20

Figure 8. Temperature profile for agave cooking in oven and autoclave.

232 M. Cedeño Cruz

wood or steel mallets to extract the juice. Later,a rudimentary mill consisting of a large circularstone 1.3 m in diameter and 50 cm thick wasused. Driven by animals, the stone turned in acircular pit containing cooked agave andextracted the juice. The resulting juice wascollected by hand in wooden basins and carriedto fermentation tanks. By the 1950s modernsystems were implemented in which cookedagave was passed through a cutter to beshredded (except in factories that did thisoperation before cooking); and with acombination of milling and water extraction,sugars were extracted. The mills used for agaveare similar to those used in the sugarcaneindustry but are smaller in size (normally 50 cmwide). This system is still employed in mostdistilleries.

In recent years the tequila industry has beenusing a technology to extract fructose in cookedagave or inulin in raw agave based oncountercurrent extraction by means of a pieceof equipment called a diffusion band. With theuse of this technology, used for many years inthe extraction of sugar and alcohol in grapes inSpain, the process has been improved and thebagasse at the end of the extraction has nearlyno residual sugar.

Juice obtained in milling is mixed with thesyrup obtained in the cooking step and with asolution of sugars, normally from sugarcane (ifthe tequila to be produced is not 100% agave),and finally pumped into a fermentor. Althoughthe amount of sugar employed as an adjunct isregulated by law and must be less than 49% byweight at the beginning of the fermentation, eachfactory has its own formulation.

The milling step generates a by-product calledbagasse, which represents about 40% of the totalwet weight of the milled agave. Bagassecomposition (dry weight basis) is 43% cellulose,19% hemicellulose, 15% lignin, 3% totalnitrogen, 1% pectin, 10% residual sugars and9% other compounds. The bagasse, mixed withclay, is used to make bricks; but it is also thesubject of research to find alternative uses.Examples are use of bagasse as an animal feedor as a substrate on which to grow edible fungi(Iñiguez et al., 1990; 2001). Attempts are alsounderway to recover bagasse components(cellulose, hemicellulose and pectin) using high-efficiency thermochemical reactors (Alonso etal., 1993), to obtain furfurals, make particle

board, or enzymes (cellulase and pectinase).Many of these projects are at the laboratory stage,and there is not enough information to evaluatefeasibility.

In milling, as in all steps of the tequila process,a sugar balance is computed to determine theyield. If the yield decreases, the extractionpressure in the mill and the water/agave ratio areincreased to improve efficiency.

Fermentation

WORT FORMULATION

To produce 100% agave tequila, only agave maybe used and the initial sugar concentration rangesfrom 4 to 10% w/v, depending on the amount ofwater added in milling. When other sugars areemployed, they are previously dissolved andmixed with agave juice to obtain an initial sugarconcentration of 8-16%, depending on sugartolerance of the yeast strain. Wort formulationin most distilleries is based solely on previousexperience. A few distilleries base wortformulation on composition of raw materials andnutritional requirements for yeast growth andfermentation. In these distilleries responsesurface methodology is the preferred method tooptimize nutrient concentrations, usingfermentation efficiencies and taste of theresulting tequila as responses (Montgomery,1984). To correct nutritional deficiencies ofagave juice and sugars employed in the growthand fermentation steps, urea, ammonium sulfate,ammonium phosphate or magnesium sulfate canbe added. Because pH of the agave juice isaround 4.5, there is no need for adjustment andthe same wort composition is used for bothinoculum growth and fermentation.

YEASTS

Some companies do not inoculate a specificstrain of S. cerevisiae and instead allow naturalfermentation to proceed. Others inoculate thewort with fresh packages of baker’s yeast or acommercial dried yeast to obtain initialpopulations of 20-50 x 106 cells/ml. The driedyeasts were originally prepared for wine, beer,whisky or bread production; and sometimes thequality of the tequila obtained using these yeastsis not satisfactory, with large variations in flavor


and aroma. To achieve high yields and maintaina constant quality in tequila, some companiesuse yeast strains isolated from a naturalfermentation of cooked agave juice. Nutrientsare added and special conditions such as a highsugar concentration or temperature aremaintained. These isolated and selected yeaststrains have been deposited in nationalmicrobial culture collections, the most importantbeing the Biotechnology and BioengineeringDepartment Culture Collection of CINVESTAV-IPN, located In México City.

Toward the goal of improving fermentationproductivity, one alternative is to use a strain ofyeast capable of both efficiently converting wortsugars into alcohol and of producing theappropriate organoleptic compounds that impart

a pleasant aroma to the final product. Thedifference in alcohol yield in tequila productionusing a selected commercial strain of S.cerevisiae compared with wild yeast is presentedin Figure 9. The higher alcohol concentration atthe end of the fermentation process clearlymeans better productivity and a lowerproduction cost.

Fermentation temperature is important intequila production since most of the distilleriesare located in regions where ambienttemperatures are normally warm, especiallyduring the summer, and may be as high as 37ºC.Not all strains of S. cerevisiae are resistant tosuch high temperatures; and the selection of acommercial strain is very important as it canrepresent an advantage in process yield. Figure

0. 00

0. 50

1. 00

1. 50

2. 00

2. 50

3. 00

3. 50

4. 00

0 20 40 60 80 10 0

Fermentation time (hours)

Alc

oh

olv

olu

me

(%)

Wild yeast 26°C

Alltech Thermosacc 37°C

Figure 9. Alcohol concentration in a fermentation for tequila production using wild yeast compared witha commercial strain of yeast.

0. 00

0. 50

1. 00

1. 50

2. 00

2. 50

3. 00

3. 50

4. 00

4. 50

5. 00

0 20 40 60 80 10 0

Fermentation time (hours)

Alc

oh

olv

olu

me

(%)



Figure 10. Alcohol concentration in tequila fermentation using a thermotolerant commercial strain of yeastat temperatures of 26 and 37ºC.

234 M. Cedeño Cruz

10 illustrates the fact that alcohol volume in thefermentation process for a thermotolerant strainis better at 37ºC than at 26ºC.

INOCULUM GROWTH

When an inoculum is used, it is grown in thelaboratory from a pure culture of S. cerevisiaemaintained on agar slants in lyophilized formor frozen in liquid nitrogen. All laboratorypropagation is carried out under asepticconditions using a culture medium with the sameingredients used in the normal process butenriched to promote cellular growth. Theinoculum is scaled-up with continuous aerationto produce enough volume to inoculatefermentation tanks at 10% of the final volume.Populations of 200-300 x 106 cells/ml arenormally achieved. Strict cleanliness ismaintained in this step as bacterial contaminationis highly undesirable. When contamination isdetected, antibiotics or ammonium bifluoride areused as antimicrobial agents. Once an inoculumis grown, it is maintained by mixing 10% of thevolume of an active culture with fresh agavejuice and nutrients.

Although inoculation with commercial yeastgreatly improves yield and turnover time, somecompanies prefer a more complex (in terms ofthe microbial diversity) fermentation. Whileyields might be lower and turnover time higher,the range of microorganisms produces morecompounds contributing to a more highlyflavored tequila. It is also important to recognizethat a change in taste and flavor could negativelyaffect the market for a particular brand of tequila.For this reason it is very important that when acommercial strain of yeast is used in thefermentation process, the final aroma profile intequila must be evaluated using a sensoryevaluation panel together with gas chromat-ography analysis.

FERMENTATION OF AGAVE WORT

Once a wort is formulated with the requirednutrients and the temperature is around 30°C, itmay be inoculated with 5 to 10% (volume) of apreviously grown S. cerevisiae culture with apopulation of 100-200 million cells/ml.

Otherwise, microorganisms present in the wortcarry out the fermentation. If an inoculum is notadded, the fermentation could last as long asseven days. With an inoculum the fermentationtime ranges from 20 hrs in the faster process tothree days in the slower one.

Production of ethyl alcohol by yeast isassociated with formation of many fermentationcompounds that contribute to the final flavor ofthe tequila. These are organoleptic compoundsor their precursors produced either in subsequentmaturation of the wort before it goes to thedistillation step, in the distillation process, or inthe barrels if tequila is aged. The factorsinfluencing formation of the organolepticcompounds in alcoholic beverages have beenreviewed by many authors (Engan, 1981; Berry,1984; MacDonald et al., 1984; Ramsay andBerry, 1984(a); Geiger and Piendl, 1976).Experience in the tequila industry is that theamount of organoleptic compounds producedis lower in fast fermentations than in slowfermentations. As a consequence, the flavor andgeneral quality of tequila obtained from wortsfermented slowly is best. The rate offermentation depends mainly on the yeast strainused, medium composition and operatingconditions, as previously explained. The wortsugar content decreases from an initial value of4-11% to 0.4% (w/v) reducing sugars if anefficient yeast strain is employed. Otherwise, theresidual sugar content could be higher,increasing production costs.

Fermentation vessels vary considerably involume, depending on the distillery. Capacityranges from 12,000 liters for small tanks to150,000 liters for the largest ones; and they areconstructed of stainless steel in order to resistthe acidity of the wort. Ethanol production canbe detected almost from the onset; and a pH dropfrom 4.5 to 3.9 is characteristic of thefermentation. The alcohol content at the end offermentation is between 4 and 9% v/v,depending on the initial sugar concentration. Inorder to increase the fermentation yield, inaddition to selection of a good yeast strain,another option is the use of enzymes or enzymecomplexes, to convert residual polymers fromagave into fermentable sugars, which areconverted mainly into alcohol improving theproductivity of tequila production. Figure 11shows the results of an experiment using


different doses of a commercial enzyme complexin the alcohol concentration of the wort.

Alcohol losses may be significant becausemany fermentation tanks are open, allowingevaporation of alcohol with carbon dioxide.Some of the largest distilleries have a coolingsystem that keeps fermentation temperaturewithin a tolerable range for yeast, but smallproducers do not have these systems. Thefermentation temperature can exceed 40°C,causing the fermentation to stop with anaccompanying loss of ethanol and flavors thatconsequently decreases yields and affects thequality of the tequila. Fermentations carried outwith pure agave juice tend to foam, sometimesrequiring the addition of silica. In worts withadded sugars, foaming is usually not a problem.

Non-aseptic conditions are employed infermentation, and in consequence bacterialactivity may increase. The size of the bacterialflora depends on a number of factors includingthe extent to which bacteria grow during yeastpropagation (if used), the abundance of bacteriaon the raw materials and hygiene standards inthe distillery. There is no doubt that the activityof these bacteria contributes to the organolepticcharacteristics of the final product. Occasionally,the size of the bacterial population in fermentingwort may become too large (>20 x 106 cells/ml), in which case the bacteria use the sugars,decreasing ethanol yields and sometimesexcreting undesirable compounds. The same

compounds used in the propagation step maybe used here to decrease common bacterialcontaminants found in tequila worts.Lactobacillus, Streptococcus, Leuconostoc, andPediococcus are the most commoncontaminants, but Acetobacter may be found infermented worts that are left inactive for a longtime prior to distillation.

In contrast to other distilled beverages, theorganoleptic characteristics of tequila come fromthe raw material (cooked agave) as well as fromthe fermentation process. In most of the processesused in the tequila industry, fermentation isspontaneous with the participation ofmicroorganisms from the environment, mostlyyeasts and a few bacteria. This peculiarity bringsabout a wide variety of compositions andorganoleptic properties. However, the specialcharacteristics of the wort make it a selectivemedium for the growth of certain yeasts such asSaccharomyces and to a lesser extent, acetic andlactic bacteria. In experiments isolating microbialflora from musts of various origins, differentmicroorganisms were isolated. In most cases,this difference in flora is responsible for the widevariety of organoleptic characteristics of tequilabrands (Pinal, 1999). Where a single purifiedstrain is used, the final flavor and aroma aremore neutral since the bouquet created by thecontribution of several stains is richer than thatobtained from only one type of yeast. Moreover,when yeast produced for bakery applications is

0

1

2

3

4

5

0 20 40 60 80 10 0 12 0 14 0

Time (hours)

Alc

oh

olv

olu

me

(%)

Control 35°

0.25% Rhizozyme 35°

0.5% Rhizozyme 35°

1% Rhizozyme 35°

Figure 11. Alcohol volume in the wort at different concentrations of an enzyme complex (RhizozymeTM)for tequila production at 35ºC.

236 M. Cedeño Cruz

used, the final product is also more neutral. It isalso recognized that when non-100% agavetequila is made, a poorer bouquet is obtainedbecause a more defined medium yields a moredefined product.

ORGANOLEPTIC COMPOUNDS GENERATEDDURING FERMENTATION

Fusel oil

As in many other alcoholic fermentationprocesses, higher alcohols are the most abundantcompounds produced along with ethanol. Wehave found (in decreasing order of abundance)isoamyl alcohol, isobutanol, active isoamylalcohol and phenylethanol. It has beenestablished that production of isoamyl andisobutyl alcohols begins after the sugar level islowered substantially and continues for severalhours after the alcoholic fermentation ends. Incontrast, ethanol production begins in the firsthours of the fermentation and ends withlogarithmic yeast growth (Pinal et al., 1997).

The most important factor influencing theamount of isoamyl alcohol and isobutanol is theyeast strain. It was found that a native strainisolated from tequila must produces a higheramount of such compounds when comparedwith a strain usually employed in bakeries. Theseresults agree with those reported for Scotch

whisky (Ramsay and Berry, 1984(b)) and beer(García et al., 1994).

The carbon:nitrogen ratio also has a significantinfluence on higher alcohol production. Intequila musts, which contain mainly fructose(≈95%) as a carbon source and an inorganicnitrogen source (ammonium sulfate), it wasfound for both native and bakery yeast strainsthat low carbon:nitrogen ratios result in lowamounts of isoamyl alcohol: 19 mg/L in bakerystrains and 30 mg/L for native strains vs 27 and64 mg/L, respectively, for high carbon:nitrogenratios. A similar relationship exists for isobutylalcohol production (Figures 12 and 13).

Temperature is a third factor affecting isobutyland isoamyl alcohol production with highertemperatures (e.g. 38 vs 32°C) yielding higherconcentrations of those alcohols. On statisticalanalysis it was found that in addition to the directeffects of yeast strain, carbon:nitrogen ratio andtemperature, the interaction of these three factorsalso had an impact on higher alcoholconcentration in tequila (Pinal et al., 1997).These results are consistent with the fact that withhigh carbon:nitrogen ratios there is a tendencyto use amino acids as a nitrogen source, whichimplies the production of fusel oil as a by-product (by the Erlich pathway). On the otherhand, variables such as the type of nitrogensource (urea or ammonium sulfate) or theamount of inoculum used for fermentation had

8.4

8.0

7.2

0 5 10 15 20 25 30 35

0

20

40

60

80

100

120

140

Time (hrs)

Log cells/ml

Reducing sugars, g/L

Ethanol, g/L

Isoamyl + isobutyl alcohols, mg/L

Figure 12. Higher alcohol production in tequila wort by a baker’s yeast strain.


little or no effect on the production of higheralcohols. Pareto diagrams involving all thevariables tested are depicted in Figures 14 and15.

Methanol

Another characteristic compound present intequila is methanol. It is generally thought thatmethanol is generated through hydrolysis ofmethylated pectins present in the agave plant.Nevertheless, it is also believed that some yeaststrains, natural or inoculated, have pectin methylesterases. Some authors describe various

amounts of methanol in musts with the samecomposition but fermented with different strains(Téllez, 1999).

Aldehydes

Along with the production of ethyl acetate, theoxidation of ethanol also generates acetal-dehyde, an intermediate in the production ofacetic acid. It is well known in commercialpractice that an oxidation process instead offermentation begins after the sugar concentrationdeclines, thus provoking the increase inacetaldehyde levels. However, there is no formal

8.4

8.0

7.2

0 5 10 15 20 25 30 35

0

20

40

60

80

100

120

140

Time (hrs)

Log cells/ml

Reducing sugars, g/L

Ethanol, g/L

Isoamyl + isobutyl

alcohols, mg/L

Figure 13. Higher alcohol production in tequila wort by a native yeast strain.

-26.6821.61

-15.5610.32

8.34

-7.38-6.71-6.41

5.61

-5.10-4.65

3.70-3.09

-2.532.50

-1.931.00

A: StrainB: TemperatureC: Nitrogen sourceD: Carbon:nitrogen ratioE: Inoculum amount

0 10 20 30 40 50

E

ACE

CDE

BE

ABE

ABD

BD

AD

A

Fac

tors

Effect level

Figure 14. Pareto diagram for isoamyl alcohol production in tequila.

238 M. Cedeño Cruz

report of such phenomena in tequila, as isdescribed in beer (Hammond, 1991).

Organic acids

Small organic acids (up to six carbons) and largermolecules (fatty acids) are produced duringfermentation. The smaller molecules can beproducts of intermediate metabolism of thenormal microbial flora; and their productiondepends on the presence of oxygen. The largerfatty acids are synthesized for membranestructures during cell growth and can alsoappear at the end of the fermentation when lysistakes place. The presence of octanoic anddecanoic acids in the final product has beendescribed particularly for tequila.

Esters

Esters are very important compounds in theirparticular contribution to flavor and aroma, sincethey have the lowest organoleptic thresholdvalues (Ramsay and Berry, 1984a). In particular,ethyl acetate is the most abundant and importantcompound of this family. Ethyl acetate has beenreported to be the second most abundantcompound in tequila after isoamyl alcohol. Thequantity of this compound present in the finalproduct can vary widely, since it is synthesizedfrom acetic acid (in the form of acetyl CoA) and

ethanol. Acetic acid can also be produced bythe oxidation of ethanol when the fermentationhas ceased and an oxidative process starts onthe surface of the fermentation tank bySaccharomyces and many other yeasts such asBrettanomyces. Therefore, long fermentationperiods (a current practice in the tequilaindustry) yield high ethanol oxidation. Inaddition, in open fermentation tanks with wortsat low pH containing alcohol, ethanol is alsotransformed to acetic acid (itself a precursor ofethyl acetate) by bacteria of the generaAcetobacter. Besides ethyl acetate, the presenceof several other esters has been describedincluding ethyl and isoamyl esters. Some of themost important esters found in silver tequila arelisted in Table 2.

Table 2. Most abundant esters found in silver tequila.1

Ester %

Ethyl acetate 17.77Ethyl decanoate 2.78Ethyl lactate 2.74Ethyl octanoate 1.92Ethyl dodecanoate 0.95Ethyl butanoate 0.63Isoamyl acetate 0.58Ethyl propanoate 0.57Ethyl hexanoate 0.48Ethyl hexadecanoate 0.481Estarrón et al., 1999.

Figure 15. Pareto diagram for isobutyl alcohol production in tequila.

0 10 20 30 40 50

E

ACE

CDE

BE

ABE

ABD

BD

AD

A

Fac

tors

Effect level

-41.85

16.33

-13.4112.94

-12.73-8.47

8.137.35

6.99-6.56

6.56

-6.12-6.03

-6.015.66

3.841.55

A: StrainB: TemperatureC: Nitrogen sourceD: Carbon:nitrogen ratioE: Inoculum amount


Distilling

Distillation involves the separation andconcentration of the alcohol from the fermentedwort. In addition to ethanol and other desirablesecondary products, wort contains solid agaveparticles consisting mainly of cellulose, pectin,and yeast cells in addition to proteins, mineralsalts and some organic acids. Although manytypes and degrees of distillation are possible, themost common systems used in the tequilaindustry are pot stills and rectification columns.The pot still is considered the earliest form ofdistilling equipment. It is of the simplest design,consisting of a kettle to hold the fermented wort,a steam coil, and a condenser or a plate heatexchanger. Pot stills are often made of copper,which ‘fixes’, according to Thorne et al. (1971),malodorous volatile sulfur compounds producedduring fermentation. Batch distillation using potstills is carried out in two steps. First thefermented wort is distilled to increase the alcoholconcentration to 20-30% by volume, separatingout the first fraction called ‘heads’, and the lastfraction, called ‘tails’. Composition of thesefractions varies depending on many factorsincluding the yeast strain employed, wortnutrient composition, fermentation time anddistillation technique; but in general, heads arerich in low boiling point compounds such asacetaldehyde, ethyl acetate, methanol, 1-propanol, 2-propanol, 1-butanol, and 2-methylpropanol, which give a very pleasant flavor andtaste to tequila. Heads are normally mixed withthe wort being distilled. The tails contain highboiling point components such as isoamylalcohol, amyl alcohol, 2-furaldehyde, acetic acidand ethylactate, giving a strong taste and flavorto the tequila; and when the concentration isabove 0.5 mg/ml, the final product becomesunpleasant. This fraction is not used.

In the second step, the liquid obtained fromthe first stage is re-distilled in a similar pot stillin order to obtain a final product that is 110°proof if it is sold in bulk (reducing transportcosts) or 80° proof if it is to be bottled. Somecompanies obtain high proof tequila and diluteit with demineralized water or water purified byreverse osmosis.

In continuous distillation systems, thefermented wort enters the feed plate of thecolumn and flows downward, crossing a series

of trays. Steam is injected from the bottom in acoil and strips the wort of its volatilecomponents. Vapors condense higher in thecolumn, depending on component volatility,allowing liquids to be drawn off or recycled atthe various plates as appropriate. Sometimestequila obtained in this way is mixed with tequilafrom pot stills to balance the amount oforganoleptic compounds because in general,tequila obtained through continuous columnshas less aroma and taste than tequila obtainedfrom pot stills.

The presence of methanol in tequila is still asubject of discussion because whether methanolis produced only by a chemical reaction or incombination with a microbial hydrolysis has notbeen satisfactorily demonstrated. The chemicalreaction is demethylation of agave pectin by thehigh pH during cooking and the first distillationsteps. The microbial reaction could be thehydrolysis of agave pectin by the enzyme pectinmethyl esterase produced by somemicroorganisms during the fermentation step, butthis has not yet been demonstrated. Preliminaryresults favor the first theory, but research is stillneeded on this matter because it is important tomaintain the methanol concentration within thelimits established by the official standard, whichis 300 mg per 100 ml of anhydrous ethyl alcohol.Table 3 shows the specifications for tequilaaccording to the standard of quality from theMexican Official Norm (Secofi, 1994).

Effluent disposal

The discharge from pot stills or distillationcolumns is known as stillage, slops or vinasse;and in a typical tequila distillery 7 to 10 liters ofeffluent are produced per liter of tequila at 110ºproof. Tequila stillage has a biological oxygendemand (BOD) of 25 to 60 g/L. In addition tothe dissolved salts (mainly potassium, calcium,and sulfate ions) and the low pH (<3.9) of thestillage, there are significant disposal or treatmentproblems. A general solution to the disposalproblem does not exist because every factoryhas its own production process and is locatedeither in a city or near agave fields. As a resultof the difficulties of treating vinasses and due to

240 M. Cedeño Cruz

Table 3. Tequila specifications according to NOM.

Tequila blanco Tequila joven u oro Tequila reposado Tequila añejo(White tequila) (Gold tequila) (Aged) (Extra-aged)

min max min max min max min max

Alcohol, % bv at 20ºC 38.0 55.0 38.0 55.0 38.0 55.0 38.0 55.0Dry extract, g/L 0 0.20 0 5.0 0 5.0 0 5.0

Values expressed as mg/100 ml pure ethanol

Higher alcohols 20 400 20 400 20 400 20 400 (as amyl alcohol1)Methanol2 30 300 30 300 30 300 30 300Aldehydes 0 40 0 40 0 40 0 40Esters 2 270 2 350 2 360 2 360Furfural3 0 1 0 1 0 1 0 1

1Using a Gas Chromatograph the value could be up to 500 mg/100 ml.2Minimum could be less than this value if the distillery could demonstrate the use of a technology to reduce methanol concentration.3Using chemical analysis the maximum value could be up to 4 mg/100 ml.

their high concentrations of dissolved matter, ahost of utilization schemes have been proposed.Some of the methods indicated below are underinvestigation and others are in use.

Recycling reduces the volume of waste to betreated. Stillage can be recirculated, mixing 5 to10% of the total volume of the waste obtainedwith clean water to substitute for the dilutionwater used to prepare the initial wort. This canbe carried out for a number of cycles, usuallyno more than five, because the concentration ofdissolved salts increases and could affect thefermentation process. Also, great care must betaken with the final taste and flavor of the tequilabecause some components present in stillagecould affect the organoleptic characteristics ofthe final product. Currently, only one tequilacompany uses this system.

Direct land application as irrigation water andfertilizer in agave fields is under carefulevaluation to determine the optimum loadingrates and the effects on the agave over the longtime needed to reach maturity. Evaporation orcombustion of stillage could provide fertilizeror potash, but the high cost of such a process isa serious limitation (Sheenan and Greenfield,1980). The production of biomass andbiochemicals including fodder yeast is apossibility (Iñiguez et al., 1996), but theremaining liquor still has a high BOD (Quinn

and Marchant, 1980). Stillage may be used as afood supplement for cattle, but it has anundesirable laxative effect on animals.Biological, aerobic, or anaerobic treatment offersa real means of disposal, but the cost is likely tobe as high as the fermentation costs themselves(Speece, 1983; Maiorella, 1983). Ultimately,tequila vinasses should be viewed as a rawmaterial rather than a waste, and a strategyshould be devised that maximizes economic andsocial benefits and reduces recovery costs.

Maturation

Distillation is the final stage of tequila productionif silver or white tequila is the desired product.For rested (reposado) or aged (añejo) tequila,maturation is carried out in 200 liter white oakcasks or in larger wood tanks. The time legallyrequired is two months for rested tequila and 12months for aged tequila. Tequila is generallymatured for longer periods, depending on thecharacteristics each company desires for itsparticular brand.

As tequila ages in barrels, it is subject tochanges that will determine its final quality.Thickness and quality of the stave, depth of thechar, temperature and humidity in the barrellingarea, entry proof (40-110° proof), length of


storage and number of cycles for the barrel (inMexico barrels may be re-used several times)all affect the final taste and aroma of the tequila.Fusel oil content decreases during maturationowing to the adsorbant nature of the char,smoothing the final product. Complex woodconstituents are extracted by the tequila,providing color and the particular taste.Reactions among certain tequila compoundsyield new components; and oxidation reactionschange some of the original components intequila and those extracted from the wood. As aresult of all of these changes, the concentrationof acids, esters and aldehydes is increased, whilethe concentration of fusel oils decreases astequila reposes in barrels.

After aging and dilution with demineralizedwater (if necessary) the color of the tequila maybe adjusted to the desired value by the additionof caramel. Alternatively, some companies blenddifferent batches of tequila to obtain astandardized final product.

Government inspectors supervise the entireaging process. Prior to bottling, tequila is filteredthrough cellulose filter pads or polypropylenecartridges. Sometimes afterwards a pretreatmentwith charcoal is used to eliminate turbidity.

Production statistics

Tequila production has grown since 1995 at anaverage rate of more than 10% annually andreached its maximum production of 190.6

million liters of tequila (at 40% alcohol byvolume) in 1999 (Figure 16). After that, ashortage of agave occurred due to a combinationof environmental and market factors. A bacterialinfection (Erwinia carotovora) and a fungus(Fusarium oxysporum), affected agave harvests,while the growing number of distilleries,development of many new brands in the market,and an increase in tequila consumption in thedomestic and export market affected demand.The drop in production began in 2000 with areduction to 181.6 million liters and in 2002declined to 141 million liters.

Production of 100% agave tequila had the mostpronounced decrease due to the agave shortage,because 6.0 kg of agave are required to produce1 liter of tequila (at 55% abv) whereas othertequilas require only 3.0 kg of agave per liter.Most tequila producers moved into productionof other tequila types and reduced amounts of100% agave tequila.

The outlook for the future appears promisingfor tequila production, and a return to previousindustry growth rates is expected by the end of2003.

In the export market the scenario was verysimilar but the effect of the agave shortage lastedthrough 2002 since the main export companiessell tequila in bulk (Figure 17). Total exportvolume was at its lowest in 2001 (75.6 millionliters of tequila at 40% abv) with the beginningof recovery noted in 2002. Exports of 100%agave tequila became almost stable at a volume

0

Tequila100%

Tequila

Total

200

150

100

50

1995 1996 1997 1998 1999 2000 2001 2002

Figure 16. Tequila production in Mexico in millions of liters at 40% abv (CRT, 2002).

242 M. Cedeño Cruz

near 8 million liters. Again, it should be notedthat this type of tequila can only be exportedbottled and not in bulk.

Although the European and other markets aregrowing, nearly 80% of all tequila exportedcontinues to go to the US market with volumesthat in 2002 represented over 69 million litersof tequila at 40% abv, which is more than the 53million liters consumed in the domestic market(Figure 18). Finally, even though only 25% ofthe tequila is exported bottled during 2002 (22

million liters of tequila at 40% abv), this has beengrowing since 1995 when 6.5 million liters werestate-bottled (Figure 19).

Future developments

Research and future developments in theproduction of tequila and agave cultivation cantake many directions, but the implementation ofany change must allow quality of the final

0

20

40

60

80

10 0Tequila100%

Tequila

Total

1995 1996 1997 1998 1999 2000 2001 2002

Figure 17. Exports of tequila by type in million liters at 40% abv (CRT, 2002).

0

20

40

60

80

10 0 USA Europe Other Total

1995 1996 1997 1998 1999 2000 2001 2002

Figure 18. Exports of tequila by destination in million of liters at 40% alcohol in volume (CRT 2002).


product to be maintained and provide substantialimprovement in the process. There are severalkey areas where important developments couldtake place. Development of new varieties ofagave endowed with resistance to pests orextremely dry environments, use ofmicropropagation techniques, satellite remoteinspection to evaluate health status of agaveplantations, higher inulin content, low waxcontent in the leaf cuticle and superior growthrates would all be of benefit. Mechanizationwould improve aspects of cultivation and harvestof agave. In addition, optimization of thecooking and fermentation steps would improveyields and reduce the amount of waste, whileyeast strain selection could improve ability toferment musts with high concentrations of sugars.Another area of opportunity is the use of isotope(H2 and O18) measurement in tequila to authenticatethe beverage. Finally, low cost alternatives forwaste (bagasse and stillage) treatment areneeded.

Summary

Tequila is a beverage obtained from thedistillation of fermented juice from the agaveplant (Agave tequilana. Weber var. azul). Theuse of agave to produce drinks in Mexico datesfrom long ago when certain tribes used themfor religious ceremonies. It was not until thearrival of the Spaniards, who brought

knowledge of distillation techniques, that tequilaacquired its present form. The agave plant, whichis often confused with cacti, is propagatedthrough traditional methods or by means of plantcell culture. It is grown for 7 to 8 years before itcan be harvested and sent to the distillery.

The tequila production process comprises fivestages. In the first step, agave is cooked tohydrolyze the polymers present in the plant,mainly inulin, into fermentable sugars. In somefactories this step is accomplished using stoneovens and in others it is carried out in autoclaves.The second stage is sugar extraction from cookedagave through milling; and the agave juiceobtained in this step can be mixed with sugarsfrom other sources, normally sugar cane, if 100%agave tequila is not desired. The third and mostimportant stage is fermentation in which sugarsare transformed into ethanol and othercompounds such as esters and organic acids.These, along with other substances derived fromthe cooked agave, give the characteristic flavorand taste to tequila. It is of great importance tohave a good yeast strain and nutritionallybalanced wort for tequila production, as lossescan be as high as 35% of the total production ifinefficient yeast is used or nutrients are notpresent in the right proportions. In the fourthstage, fermented wort is distilled, normally usingpot stills or in some cases rectification columns,to obtain the final product. At the end of thedistillation process white tequila is obtained.Maturation, the last stage, in white oak barrels

Figure 19. Exports of tequila in millions of liters at 40% abv (CRT, 2002).

0

20

40

60

80

10 0Bulk

Bottled

Total

1995 1996 1997 1998 1999 2000 2001 2002

244 M. Cedeño Cruz

is required for rested or aged tequila. Theminimum maturation times are 2 and 12 months,respectively, for rested and aged tequila asrequired by government regulations. At everystep of the production process, most companiesemploy several quality control analyses in orderto ensure the quality of the product and theefficiency of the process. Some major producersof tequila are certified through an ISO-9000standard. Tequila production is governed by theofficial norm NOM-006-SCFI-1994, which mustbe followed by all tequila producers to guaranteea good quality final product.

Acknowledgement

The author would like to express his gratitudeto Tequila Herradura, S.A. de C.V., for its supportin writing this chapter.

References

Agricultural Research Service. 1972. The Agavefamily in Sonora. Agriculture HandbookNo.399, US Department of Agriculture, p. 195.

Alonso, G.S., L. Rigal and A. Gaset. 1993.Valoración química de bagazo de agave de laindustria tequilera. Rev. Soc. Quim. Mex.3(6):19.

AOAC 1990. Distilled liquors. Official Methodsof Analysis of the Association of OfficialAnalytical Chemists (H. Kenneth, ed) Assoc-iation of Official Analytical Chemists,Arlington, Virginia, p. 690.

Backman, G.E. 1944. A karyosystematic studyof the genus Agave. Am. J. Bot. 35:283.

Berry, D.R. 1984. Physiology and microbiologyof the malt whisky fermentation. In: Progressin Industrial Microbiology. (M.E. Bushell, ed.)Elsevier, Amsterdam, p. 189.

Bottorff de B., V. 1971. A guide to tequila,mezcal and pulque. Minutae Mexicana,Mexico, D.F.

Casado, C.G. and A. Heredia. 2001. Self-association of plant wax components: athermodynamic analysis. Biomacromolecules2:407-409.

CRT. 2002. Consejo Regulador del Tequila.Informe de Actividades.

Diguet. L. 1902. Estudio sobre el maguey detequila. Generalidades e historia. El Prog. Mex.9:424.

Ehrler, W.H. 1967. Agave plant, efficient wateruser. USDA Agr. Res. 16(4):11.

Engan, S. 1981. Beer composition: volatilesubstances. In: Brewing Sciences, Vol. 2(J.R.A. Pollock, ed.) Academic Press, London,p. 98.

Estarrón, M., T. Martín del Campo and R. Cosío.1999. Identificación de los componentesvolátiles que caracterizan la huellacromatográfica distintiva de tequilas. TechnicalReport for Tequila Herradura S.A.

García, A.I., L.A. García and M. Díaz. 1994.Fusel alcohol production in beer fermentationprocesses. Process Biochem. 29:303-309.

GEA. 1992. La fertilización, La Jima. BoletinInformativo, Gerencia de extensión agrícola.Tequila Sauza, 7(50).

Geiger, E. and A. Piendl. 1976. Technologicalfactors in the formation of acetaldehyde duringfermentation, MBAA Tech. Q. 13:51.

Goncalves de L. O. 1956. El Maguey y elPulque, Fondo dc Cultura Económica, México.

Granados, S.D. 1985. Etnobotanica de losagaves dc las zonas áridas y semiáridas. In:Biología y Aprovechamiento Integral delHenequén y Otros Agaves, CICY, A.C.,México, p. 127.

Halffter, G. 1975. Plagas que afectan a lasdistintas especies de agave cultivado enMéxico. Secretaría de Agricultura y Ganadería,México, D.F.

Hammond, J.R.M. 1991. Brewer’s Yeast. In: TheYeasts, Vol. 5. Yeast Technology (A.H. Roseand J.S. Harrison, eds.) 2nd. ed. AcademicPress, Redding, U.K.

Iñiguez, C.G., I.J.A. Cuaron, G.P. Perez, M.M.De la Torre and P.I. Magaña. 1990.Fermentation characteristics, digestibility andperformance of ensiled swine waste, wheatstraw and cane molasses fed to sheep. Biol.Wastes 34(4):281-299.

Iñiguez, C.G., G.Ma. de J. Franco and O.G.Lopez. 1996. Utilization of recovered solidsfrom tequila industry vinasse as fodder feed.Bioresource Technology 55(2):151-155.

Iñiguez, C.G., S.E. Lange and R.M. Rowell.2001. Utilization of byproducts from the


tequila industry: Part 1: Agave bagasse as araw material for animal feeding and fiberboardproduction. Bioresource Technology 77:25-32.

Luna, Z.R. 1991. La historia del tequila, de susregiones y de sus hombres. Consejo Nacio-nal para la Cultura y las Artes, México, D.F.

MacDonald, J., P.T.V. Reeve, J.D. Ruddlesdenand F.H. White. 1984. Current approaches tobrewery fermentations. In: Progress inIndustrial Microbiology. Vol. 19, ModernApplications of Traditional Biotechnologies,(M.E. Bushell, ed.) Elsevier, Amsterdam.

Maiorella. B.L., H.W. Blanch and C.R. Wilke.1983. Distillery effluent treatment and by-product recovery. Proc. Biochem. 8(4):5.

Mendez, R.M.D. 1999. Acumulación defructanas en Agave tequilaza Weber var. azulcultivado en campo. MS Food Science Tesis.ENCB-IPN pp. 80-87.

Montgomery, D.C. 1984. Design and Analysisof Experiments, 2nd Ed., John Wiley & Sons,New York. p. 445.

Muria, J.M. 1990. El tequila. Boceto históricode una industria. Cuad. Difus. Cient. 18:13.

Pérez, L. 1990. Estudio sobre el maguey llamadoMezcal en el estado de Jalisco. Programa deEstudios Jalisciences, Instituto del Tequila,A.C., México.

Pinal, Z.L.M. 1999. Tesis de Maestría enProcesos Biotecnológicos. Universidad deGuadalajara, Guadalajara, México.

Pinal, Z.L.M. 2001. Influencia del tiempo decocimiento sobre la generacion de compuestosorganolepticos en las etapas de cocimiento yfermentación en la elaboración de tequila. MSBiotechnology Tesis. U. de G. Cucei pp. 85-91.

Pinal, L., M. Cedeño, H. Gutiérrez and J. Alvarez-Jacobs. 1997. Fermentation parametersinfluencing higher alcohol production in thetequila Process. Biotechnol. Lett. 19(1):45-47.

Pla, R. and J. Tapia. 1990. El agave azul, de lasmieles al tequila. In: Centro de EstudiosMexicanos y Centroamericanos, InstitutoFrancés de America Latina y el InstitutFrancais dc Recherche Scientifique pour le

developement en Cooperation, México, p. 61.Quinn, J.P. and R. Marchant. 1980. The treatment

of malt whisky distillery waste using thefungus Geotrichum candidum. Water Res.14:545.

Ramsay, C.M. and D.R. Berry. 1984a. The effectof temperature and pH on the formation ofhigher alcohols, fatty acids and esters in maltwhisky fermentation. Food Microbiol. 1:117.

Ramsay, C.M. and D.R. Berry. 1984b.Physiological control of higher alcoholformation in Scotch whisky fermentation. In:Current Developments in Yeast Research (G.G.Stewart and I. Russell, eds) Pergamon Press,London, Canada.

Rzedowski, J. 1978. Vegetación de México, Ed.Limusa. México.

Sánchez, A.F. 1991. Comparación demetodologías de micropropagación de Agavetequilana Weber. Tesis de Ing. Agrónomo,Facultad de Agronomía, Universidad deGuadalajara, Guadalajara, México.

Sánchez-Marroquín, A. and P.H. Hope. 1953.Fermentation and chemical compositionstudies of some species. Agric. Food Chem.246:1.

Secofi. 1994. Norma Oficial Mexicana. NOM-070-SCFI-1994. Bebidas alcoholicas-Mezcal-Especificaciones. Diario oficial 4 de Julio.

Sheenan, G.J. and P.F. Greenfield. 1980.Utilization, treatment and disposal of distillerywastewater. Water Res. 14:257.

Speece, R.E. 1983. Anaerobic biotechnology forindustrial wastewater treatment. Environ. Sci.Technol. 17(9):416A.

Téllez, P. 1999. Tesis de Maestría en ProcesosBiotecnológicos. Universidad de Guadalajara,Guadalajara, México.

Thorne, R.S.W., E. Helm and K. Svendsen.1971. Control of sulfur impurities in beeraroma, J. Inst. Brew. 77(2):148.

Valenzuela, Z.A. 1992. Floración y madurez delagave, La Jima. Boletin Informativo, Gerenciade extensión agrícola. Tequila Sauza 5:33.

Production of heavy and light rums: fermentation and maturation 247

Chapter 16

Production of heavy and light rums: fermentation and maturation

ROBERT PIGGOT


Rum defined

Rum differs from Scotch, bourbon or cognac inthat it is produced in many countries. In factmany countries import bulk rum, mature andblend it then market it either locally orinternationally. The country where it is sold setsthe definition of rum, which means that it is notthe producing country that defines the ingredientsand processes involved in making rum.

In the US the BATF definition of rum is:

“An alcoholic distillate from the fermentedjuice of sugar cane, sugar cane syrup, sugarcane molasses, or other sugar cane by-products, produced at less than 190° proof insuch manner that the distillate possesses thetaste, aroma and characteristics generallyattributed to rum, and bottled at not less than80° proof; and also includes mixtures solelyof such distillates.”

The EU definition is essentially the same:

“A spirit drink produced exclusively byalcoholic fermentation and distillation, eitherfrom molasses or syrup produced in themanufacture of cane sugar or from sugar canejuice itself, and distilled at less than 96% vol,so that the distillate has the discerniblespecific organoleptic characteristics of rum.”

The two key points in both these definitions isthat rum must be made with products derivedfrom sugar cane (beet molasses products cannotbe called rum and have a very different flavorprofile) and it must have taste and aromaassociated with rum. This part of the definitionhas been stretched in recent years with the verylight white rums. In some blended drinks thespirit is referred to as ‘molasses spirit’ instead ofrum for this reason. These definitions mayappear vague, which is due to the wide range ofstyles and production methods that go intoproducing quality rum. For this reason thischapter is intended to give an overall guide torum production and should be read inconjunction with the chapters on molasses as afeedstock and beverage alcohol distillation alsoin this volume.

Types of rum

Rum, along with its sister spirits such as cachaça(ka-shah’-sa) from Brazil (1.3 billion litresproduced per annum according to Braziliangovernment figures), cane spirit from SouthAfrica as well as much of the world’s vodkaspirit are all made from sugar cane. This meansthat a large percentage of the world’s beverage

248 R. Piggot

alcohol is made from the cane grass. Rum canbe loosely divided into three groups: dark, amberand white. White rum is produced by continuousdistillation. It is not aged, except where requiredby law. When aged, it is aged in old well-usedbarrels and charcoal treated before bottling. Ithas a light and delicate nose and flavor. Amberrum is aged in wood for a short period, from sixmonths to two years. Amber rum may containsome pot still rum for flavor. Dark rum is morefully bodied. It is aged in wood for longerperiods, up to 12 years.

History

The beginnings of rum go back to the mid1600s, possibly originating in Barbados, whereit was known as a ‘hot and hellish’ drink. Itcarried nicknames such as ‘kill devil’ and‘rumbullion’, the latter meaning clamor or noise.Today rum is associated with warm tropicalislands, images of swashbuckling pirates and theBritish Navy. The link between rum and theBritish Navy runs deep:

• Vice Admiral William Penn first gave it in1655 after the capture of Jamaica, which didnot have brewing facilities. It graduallyreplaced beer as it could be stored for longerperiods.

• It became the official drink of theBritishNavy in 1731; and each sailorreceived daily a half pint (tot) of undilutedrum. The previous ration was one gallon ofbeer or wine a day. In their days, Hawkinsand Frobisher said they could cruise as longas the beer lasted.

• In order to reduce drunkenness among thecrew, on August 21, 1740 Vice AdmiralVernon ordered the daily rum ration dilutedwith water. The admiral’s nickname, Grog,was given to the diluted mixture.

• After the battle of Trafalgar in 1805 the bodyof Lord Nelson was preserved in rum, givingit the nickname ‘Nelson’s Blood’.

• The tot was reduced to 1/8 pint on January 1,1851.

• July 30, 1970 ‘Black Tot Day’ was the lastday the traditional tot was given in the BritishNavy.

Dark rum was the only type of rum availablefor many years. It was generally built on theimage of British Navy Rum. As tastes changedit declined in popularity. The Bacardi revolutionof the 1960s, with its amber and light rums,quality control and marketing changed theimage of rum and helped make it the popularbeverage it is today. Even quality dark rum,which has all the nuances of a fine cognac, ismaking a strong comeback. Today the rumindustry employs 50,000 people in the Caribbeanbasin and generates $260,000,000 in exportearnings.

Equipment and feedstocks

EQUIPMENT

Rum is made with a wide variety of equipment,both old and modern. Until recently, an originalwood-sided Coffey still was still in use inGuyana. The art of fine rum making is not inthe equipment; it is in understanding andoperating the equipment to produce the desiredcharacteristics.

SUBSTRATES

Molasses

The use of molasses as a feedstock for alcoholproduction has been covered in an earlier chapterin this volume; therefore only a few commentsare necessary with specific reference to rum.

The source of the molasses can have a stronginfluence on the aromatic quality of rum. Thishas been demonstrated in trials on making rumfrom beet molasses which showed that it wasnot possible to obtain the same characteristicaromas as obtained from cane molasses (Arroyo,1948). Arroyo (1941, 1942) reported that freshblackstrap molasses with low viscosity, high totalsugars, nitrogen, phosphorus and a low ash andgum content was preferable in the productionof rums with desirable odors and tastes. Goodmolasses for rum production should be low inash and sulfur and high in fermentable sugarsand free amino nitrogen.

Cane juice

Most rum is produced from blackstrap molasses,


although cane juice is used in the FrenchCaribbean islands. This spirit, known as‘agricultural rum’ has a very high congener value(over 2,250 ppm) (EU regulations). Using canejuice for rum production has both advantagesand disadvantages. The main advantage is thatno additional processing is required. The caneis simply crushed and the juice fed directly intothe fermentors. Another advantage is that canejuice does not have a high content of dissolvedsalts and therefore does not cause as muchscaling and blocking of distillation columns asmay occur using molasses.

One of the main disadvantages of using canejuice is that it normally contains only about 12-16% w/w sugar, so that the alcohol content ofthe fermented material is limited to about 6-8%v/v as compared to 10% v/v obtainable frommolasses. Another significant disadvantage isthat cane juice cannot be stored in bulk. Canejuice normally becomes heavily infected withbacteria and yeast in the crushing process. Thiscontamination may significantly reduce alcoholyields and may cause it to start to fermentspontaneously and very rapidly. This means thatthe distillery must be located in close proximityto the cane mill. It also means that the cane juiceis only available during the cane-crushingseason, which generally does not exceed sixmonths per year. Cane juice may also have costdisadvantages in that it is a primary, or at leastan intermediate product in the production ofsugar. Sugar production is usually heavilysubsidized and so it is to the producer’s benefitto produce as much as possible. In contrast,molasses is generally a by-product of sugarproduction, considered to be of lower value.Cane juice can be blended with molasses toimprove the FAN content and reduce the amountof ash.

Fermentation and yeast

HEAVY RUM

Flavors present in the final product are producedduring fermentation. Distillation refines theflavor, but does not create the base components.In order to get the desired flavor profile manycompanies use their own yeast culture, whichthey grow up from a slant. This yeast must be

grown under sanitary conditions with a smallquantity of sterile air and yeast food to establishhealthy yeast growth. The solution is usually amolasses solution of about 16°Brix. The volumeof the final propagator should be about 10% ofthe fermentor volume and have a cell count ofat least 200 million cells/ml (Murtagh, 1999). Inpractice the yeast count in the propagation phaseis often lower than this. It is important that theyeast in the propagator is still in the exponentialgrowth phase when added to the fermentor. Ifkept too long, the yeast will enter the stationaryphase and bacteria present in the fermentor willflourish on the sugar present. Some rumproducers rely on the fermentation to startspontaneously from naturally occurring flora.This procedure results in a relatively inefficientfermentation with low yields and varying quality.The high level of bacterial contamination in thesefermentations results in many desirablecongeners, especially acids and esters. Jamaicain particular is renowned for its production ofhigh ester rums. Most of these are used asblending stock to add flavor to dark and amberrum.

Arroyo (1945) has shown that the best rumyield and aroma were achieved with ayeast:bacteria ratio of 5:1. In order to create anon-yeast congener profile in a more controlledmanner, a pure culture of bacteria such asClostridium saccharobutyricum (its presence hasbeen shown to accelerate fermentation (Arroyo,1945)) may be added after 6-12 hrs of the yeastfermentation. Usually the bacterial inoculumamounts to about 2% of the fermentor capacity;and the pH of the fermenting mash is adjustedupwards to about pH 5.5 before addition to givemore suitable conditions for the bacterialpropagation. The bacteria produce a mixture ofacids, predominantly butyric, together withothers such as acetic, propionic, and caproicacids. These acids in turn react with the alcoholto produce desirable esters (Murtagh, 1999).

FERMENTATION FOR LIGHT RUMS

In light rum production, the emphasis isgenerally on maintaining clean, rapidfermentations to minimize development ofundesirable congeners and to maximizefermentation efficiency. Many distilleries use

250 R. Piggot

commercial distillers yeast. Good yeast for theproduction of light rum should be tolerant ofhigh temperatures (most rum is made in thetropics and the fast fermentations make coolingespecially difficult), fast (to use the sugars beforebacteria metabolize them thereby reducingyield), tolerant of high osmotic pressure andresistant to acid stress. The yeast selected shouldalso provide a light congener profile. Unlikewhole grain fermentations, it is possible torecycle the yeast from molasses fermentationseither by flocculation or centrifugation of thedead wash. Traditionally recycled yeast has been‘acid washed’, i.e., reduced to a pH of about2.3 using phosphoric acid and added to the nextfermentor; or it may be treated withantimicrobials, or with about 15 ppm Cl fromchlorine dioxide at a pH of about 3.5 (Murtagh,1999). The objective of yeast recycling is to takeadvantage of sugar that would otherwise beneeded for yeast growth and to ensure a veryhigh cell count in the fermentor inoculum.Recycling of yeast is generally notrecommended. The yeast is in the stationaryphase, and therefore 30 times the amount of yeastmust be added in order to get the same rate ofalcohol production as fresh budding yeast. Alsoit has been shown that bacteria can becomeresistant to acid washing, so bacteria are alsogetting recycled. It is recommended that whenrecycled yeast is used it is dumped and renewedon a regular basis.

Distillation

The method of distillation used has aconsiderable effect on the nature of the rumproduct. Heavy rums are usually produced bybatch distillation or in a 2 column continuousstill, while light rums are normally produced bycontinuous distillation with a hydrofiningcolumn.

POT DISTILLATION

There are a wide variety of pot stills in use inthe rum industry, which vary from the straightpot still to the pot still with a distillation head asdescribed in the chapter on beverage alcoholdistillation. One type of pot still common in theCaribbean rum industry is the double retort

(Figure 1). This still can be operated eitherslowly with very low steam input or quickly witha high steam input. The still is made of copperand should be inspected on a regular basis asthe copper takes part in the chemical reactionsduring the distillation process. Wash at about 8%abv is pumped into the pot. The steam vaporizesthis mixture; and the vapor is approximately 50%abv. Some of this condenses in the first (lowwine) retort. The vapor passing out of the lowwine retort is enriched further to about 75% abv.The same process is repeated in the second (highwine) retort. The vapor leaving the high wineretort is about 85% alcohol. After the finalstrength drops below a fixed value, the still isboiled out to feints. These may be used to chargethe retorts for future distillations, and the processrepeated.

TWO COLUMN DISTILLATION

A 2-column still is basically the Coffey still.While it does not yield a neutral spirit, it is‘cleaner’ than a pot distillation. Its mainadvantage however is that it can be ‘tuned’ togive the desired congener profile. This is doneby deliberately overloading the rectifyingcolumn to varying degrees meaning a greaterproportion of the congeners come over in theproduct. For example a light rum may have afusel oil draw of 1000 mL/min and a heads drawof 600 mL/min, a medium rum may have a fuseloil draw of mL/min and a heads draw of 400mL/min and a heavier rum may have a fusel oildraw of 300 mL/min and a heads draw of 200mL/min. This gives the blender a full pallet toproduce the desired blend.

PRODUCTION OF LIGHT (WHITE) RUM

The production of light beverages andhydrofining is discussed in the chapter onbeverage distillation. Molasses has someimportant differences from other feedstocks inthat it has very little pectin to break down intomethanol. As such, it is generally not necessaryto install a demethylizing column. The problemswith mud and scaling are covered in the chaptercovering molasses as a feedstock. It is importantthat the wash column is designed to handle mudand scale. Sulfur is often a problem in molassesfermentations. The distillation system must have


plenty of copper contact. A heads concentratingsystem such as a barbet head or a separate headscolumn may be needed in some cases.

Maturation and flavor developement

Esterification is very important in developingthe flavor profile of rum. It is a chemical reactionand so is affected by time, temperature and thecomponents present. It takes place at a relativelysteady rate during the maturation of rum.(Reazin, 1983) Some esters are also formedduring fermentation. It is a reversible reactionbetween an alcohol and an organic acid. Forexample:

Ethanol + Acetic acid → Ethyl acetate

MATURATION

Maturation of rum is a complex subject; and thefollowing is intended only as a brief description.

Steam

Low wine

retort

Liquid

level

Vapor

Feints Room

Condensate

Condenser

As mentioned previously, white rum is generallynot wood-aged. Amber and dark rums are agedin wood barrels, most commonly once-usedbourbon barrels of American white oak. Someof the barrels are used charred, and others arede-charred before use. Flavor components suchas syringic acid, vanillin, syringaldehyde andgallic acid all have their origin in the ligninmaterial of the oak barrels (Lehtonen, 1982).Figure 2 compares the barrel aging componentsin Scotch, cognac and dark rum. These can beaffected by many things including the numberof times the barrel has been used, degree oftoasting, depth of char and maturationtemperature. In some Asian countries localwoods are used for the barrels, imparting aunique flavor to the spirit produced in that region.Rum, in contrast to whisky, is generally aged ina warm climate. This changes the maturationprocess in that some reactions speed up whileothers progress at the same rate as the cooler-maturing spirits. As can been seen from the graphin Figure 3, the greatest increase in color, tannins,

Figure 1. A double retort pot still.

252 R. Piggot

acids and solids occurs in the first year whileesters and aldehydes form at a steady ratethought the maturation process (Reazin, 1983).

References

Arroyo, R. 1941. The manufacture of rum.Sugar 36(12).

0

1

2

3

4

5

6

pp

m

Scotch

Dark Rum

Cognac

Pheno

l

Guaiac

ol

p-Et

hyl p

heno

l

p-Et

hyl g

uaiac

ol

Euge

nol

p-(n

-Pro

pyl)

guaia

col

Gallic

acid

Vanilli

cac

id

Syringic

acid

Vanilli

n

Syringa

ldeh

yde

0

100

200

300

0 1 2 3 4 5 6 7 8 9 10 11 12

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0Solids Tannin

Esters

Acids

Co

ng

en

er

co

ncen

trati

on

g/1

00

PL

Co

lor

(ab

so

rban

ce

at

430

nm

)

Years

Color

Aldehyde

Figure 2. Comparison of aging components in cognac, dark rum and Scotch whisky.

Figure 3. Increase in congeners with aging.

Arroyo, R. 1942. The manufacture of rum.Sugar 37(1), (5) (7).

Arroyo, R. 1945. The Production of heavybodied rum. International Sugar J. 11(8).

Arroyo, R. 1948. The flavour of rum - recentchromatographic research. International SugarJ. 50:210.

Lehtonen, M. 1982. The chromatic determinationof some organic trace compounds in alcoholic


beverages. Presentation, Helsinki.Lehtonen, M. and Jounlea-Eriksson. 1983.

Flavour of distilled beverages (J.R. Piggot, ed).Ellis Horwood Ltd.

Murtagh, J. 1999. Feedstocks, fermentation anddistillation for production of heavy and lightrums. In: Alcohol Textbook, 3rd ed. (K.A.Jacques, T.P. Lyons and D.R. Kelsall, eds),Nottingham University Press, UK.

Reazin, G. 1983. In: Flavour of DistilledBeverages (J.R. Piggot, ed). Ellis Horwood Ltd.

From pot stills to continuous stills: flavor modification by distillation 255

Chapter 17

From pot stills to continuous stills: flavor modification by distillation

ROBERT PIGGOT


Introduction

Since at least the early 8th century men have beenincreasing the strength of their fermentations.The first attempts were freezing the mash andremoving the ice. Later distillation wasdeveloped. Distillation is the separation ofcomponents by taking advantage of their relativevolatilities. When a liquid is boiled its vapor willbe richer in its more volatile components. Whenthis vapor is condensed, the resulting liquid willhave a higher concentration of the more volatilecomponent. This process can then be repeated.

Pot distillation

Until the advent of practical continuousdistillation by Stein in 1827 and Aeneas Coffeyin 1830, all distillations were carried out usingpot stills. Although simple, pot stills continue tobe used to produce the finest Scotch maltwhisky, cognac and Irish whiskey. They haveevolved from simple small retorts to uniquecomplex pieces of technology (Nicol, 1989).Development took place in Scotland during thenineteenth century as each distiller tried todifferentiate himself from the distiller next door.Once set, however, the still design, even to dents,and production methods are not changed(Bathgate, 2003). Pot stills are also used inproduction of extracted flavor drinks such as gin.While the apparatus is simple it takes skill to

produce fine beverages with this type ofdistillation. The pot still is made of copper.Originally copper may have been chosenbecause the metal was durable and easy to workand a good conductor of heat. It has proved tobe an important part of the distillation process.It is more recently that another attribute, itsability to influence the flavor of the distillate,has been fully appreciated (Nicol, 1989). Copperaids in the removal of undesirable components,especially sulfur, which imparts an off flavor tothe distillate.

There are basically two ways to run a pot still.In the first the wash is added to the pot, andthen steam is applied to the coils to start the liquidboiling. The collection of the alcohol enhancedcondensate is continued until the alcoholremaining in the pot is not economical toremove. To prevent solids burning on the coils,it is important that the liquid is always kept abovethe steam coils. The spirit derived from this typeof distillation is normally very rough as itcontains most of the congeners from thefermentation and it is often redistilled to removesome of these congeners. The other way tooperate a pot still is to take three fractions fromthe distillation. The first spirit off, called theforeshot or heads cut, is removed and storedseparately. The mid-run is the next spirit off, andit is kept for further distillation. The last spiritoff, called the tails cut, contains the heavy fusel

256 R. Piggot

oils and is removed at the end of the distillation.Distillation is continued (steam is normallyincreased at this time to speed the process time)until all practical alcohol has been removed (pottemperature approximately 98°C) the distillationis complete and the spent wash is dischargedfrom the base of the still. The mid-cut from thefirst distillation, called low wines, (20-25%alcohol) is distilled again in similar fashion. Irishpot distilled whiskey is distilled three times,cognac twice. The result is spirit containing 65-75% alcohol that has the characteristics that thedistiller deems right for aging. This sounds likea very simple process, but there are many factorsthat can influence the characteristics of potdistilled spirit.

WASH

The wash must be from a fermentation that iscomplete, free from infection and in which theyeast has not been stressed or produced excessfusel oils. The ingredients used to make up thewash will add their own individual characteristicsto the final distillate. Water should be soft, clear

and free from off odors or tastes. Wine used inbrandy production should be as low as possiblein sulfur compounds. While it is obvious thatdifferent sugar sources, molasses, fruit or grainwill produce different spirits, more subtleinfluences such as fermentation temperature willalso affect the final product.

HEADS CUT

The wash strength and the cutoff points for theheads and tails cuts are only changed after veryserious consideration as such changes willprobably alter the flavor profile of the resultingspirit. The amount of heads removed at the startof distillation will affect the volatile top note ofthe distillate and remove the ‘low boilers’(compounds with low boiling points) such asmethanol and acetone from the product. The firstcut has a role often overlooked, which is toremove the heavy oils and fatty acids in the swanneck, lyne arm and condenser left over from theprevious distillation. The historic way to changefrom the heads cut to the mid-cut in Scotland isto perform a haze test, when it no longer turns

Condenser

Cooling water out

Spirit receiver

Low pressure

steam in

CrownCondensate to boiler

Cooling water in

Spirit out

Head

Temperature

Lyne arm

Swan neck

Manway

Sight

glass

Charge

Shoulder

Vacuum

breaker

Temperature

VaporsTo

feints

Spirit safe

Figure 1. Schematic of a wash still.


cloudy when mixed with water. These days thechange is based on time or volume measurement.

TAILS CUT

While some of the heavier alcohols are neededto give the spirit some body, too much results inan oily, unpleasant character and may cause hazeproblems when diluted to bottling strength.

STILL DESIGN

The design of a still will affect the characteristicsof the product. Length and configuration of theswan neck will influence the amount of reflux.Another variation on the pot still is to put a head,either packed or with cross flow trays, in thevapor path (Figure 2). This allows a cleaner,more consistent product. It is popular in smalldistilleries that cannot afford or justify acontinuous still. There are also some very largepot stills used by major beverage alcoholproducers for flavor production.

DISTILLATION RATE

The amount of steam applied to a pot distillationwill affect the final spirit quality. When thedesired final result is obtained, the staging andamount of the steam addition should be keptconstant. Too much steam can result in foaming,liquid carryover, and very little internal reflux.A light steam load can give the product a‘stewed’ character. The large internal reflux willproduce a spirit lighter in character, but willresult in more time for chemical combinationsto occur during distillation. The resulting energylosses raise the distillation cost.

CONTROL AND AUTOMATION

Traditionally, the changes from heads to mid-cut and then to tails have been accomplished bymeasuring spirit density after the condenser asit flows through the ‘safe’. The safe is a devicedesigned to let the distiller control the processwithout actually removing a sample. Use of aspirit safe allowed distillation to proceed withoutthe presence of customs and revenue personnel

Temperature

Temperature

5 - 8 trays

or packed

column

Vent

Water cooled

condenser

Product

Steam

Recycle

Condensate

Figure 2. Schematic of a pot still with a distillation head.

258 R. Piggot

who are very sensitive about the removal of spiriton which the excise duty has not been paid.

Today many pot stills are automated. As canbe seen in Figures 3 and 4, if we know thetemperature of either the liquid or vapor phasewe can know the composition. This is generallythe way that pot distillations are automated.Since the feedstock can vary in strength and flowcan be erratic at the start of a distillation, it maybe better to take the heads cut on a volume basis.

At the start of distillation the flow is valved to acontainer. When a level probe is activated bythe distillate, flow switches to the productreceiver. A valve opens, dumping the containerinto the feints tank. When the temperatureindicates that the mid-cut is finished, the flow isreturned to the feints tank until practically allalcohol has been removed. This temperature isabout 98°C. Above that temperature, the steamcost is worth more than the alcohol.

100

95

90

85

80

75

70

65

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Tem

pera

ture

(°C

)Vapor composition

Liquid composition

Mole fraction alcohol

0

20

40

60

80

100

Mole fraction

Perc

en

t

0.05 0.20 0.40 0.60 0.80 1.00

Volume Weight

Figure 3. Relationship of temperature and mole fraction of alcohol.

Figure 4. Converting between mole fraction and percent alcohol.


Production of flavored beverages

There are several products which extract flavorby redistillation in a pot still, gin being the mostcommon. While gin may be made by addingflavors to neutral spirit, the best gins are distilledwith botanicals to extract the subtlecharacteristics of these flavored ingredients. Thefollowing is an excerpt from the EU regulationson gin:“The drink may be called ‘distilled gin’, if it isproduced solely by redistilling organolepticallysuitable ethyl alcohol of agricultural origin ofan appropriate quality, with an initial alcoholicstrength of at least 96% vol, in stills traditionallyused for gin, in the presence of juniper berriesand of other natural botanicals, provided thatthe juniper taste is predominant. The term‘distilled gin’ may also apply to a mixture of theproduct of such distillation and ethyl alcohol ofagricultural origin with the same composition,purity and alcoholic strength. Natural and/ornature-identical flavoring substances and/orflavoring preparations as specified at (a) mayalso be used to flavor distilled gin. ‘London Gin’is a type of distilled gin.

Gin obtained simply by adding essences orflavoring to ethyl alcohol of agricultural originshall not qualify for the description ‘distilledgin’.”

DISTILLED GIN PRODUCTION

A measured amount of good quality neutralspirit is added to the pot still; and sufficient waterwith no odor or taste is added to bring thecontents to about 60% alcohol by volume.Botanicals are included in weighed amounts.While juniper berries must be included, otherbotanicals are also used to enhance the flavorprofile. Common botanicals used are corianderseed, cassia bark, citrus peel, angelia root, orrisroot and fennel seed. There are several ways toextract the flavor from the botanicals duringdistillation. They may be suspended in muslinbags in the swan neck, spread out on a rackabove the alcohol level or placed into the alcoholmixture. A standard pot still distillation is thencarried out. Steam is added, and the flavor andaroma of the botanicals is carried with the spiritinto the receiver. A heads and tails cut is takenas normal. Usually the cut to tails occurs whenthe vapor reaches 55% by volume.

Continuous distillation

THE COFFEY STILL

The first continuous still was designed andpatented by Robert Stein in 1828. It was installedin the Kirkliston distillery near Edinburgh. It wasvery complex and not a commercial success. In1830 Aeneas Coffey, an Irish excise officer,developed the more efficient Patent still inDublin. This design was simpler, more efficientand well suited for distilling whole grain wortfirst (Bathgate, 2003). The design took thedistilling world by storm and was an immediatesuccess (Table 1, Figure 5) (Robson, 2001). Thisstill had metal trays and wooden walls on thecolumn. A few of these early stills are stillfunctional.

Table 1. Coffey stills in operation 1820-1860.

Year England Scotland Ireland Total

1820 0 1 2 31840 4 2 13 191860 8 12 8 28

Variations on Coffey’s design are still in use inthe Caribbean rum and the Scotch whiskyindustries. In several of the rum plants they willvary the flavor of the final product by adjustingthe fusel oil and heads draw by a fixed ratio.

Neutral spirit and brown spirits are two mainproduct types from continuous distillation.Neutral spirit can be made from any feedstockbut is usually made from grain or molasses. Thisspirit has very low odor and taste, and is usedfor non-aged products (known as white spirits)such as gin and vodka. It may also be used toblend with a highly flavored product and agedin wood barrels. This blended product usuallyhas a lighter flavor than its pot or single columnstill counterpart.

Brown spirits, so called because they developa color during maturation in wood barrels, havea much stronger flavor. These spirits are distilledto leave some congeners in the final product.This spirit more resembles pot distillation andgives us bourbon, Armangac, some fine rumsand Canadian whisky.

260 R. Piggot

FLAVORED SPIRIT PRODUCTION ON ACONTINUOUS STILL

As with the pot distillation, it is important thatthe fermented beer or wash is sound. Off odorsor taste from poor quality feedstock orcontaminated fermentations can carry over intothe final product. In the production of Canadianwhisky and bourbon, the beer is distilled in asingle column that is really a stripping andcondensing column stacked together. Someplants use a doubler or thumper (a sort ofcontinuous pot still) after this column.

Operation of a Canadian whisky flavor still

A Canadian whisky flavor still is illustrated inFigure 6. The beer (normally 8 to 10% alcoholby volume, but it can be as high as 15%) ispumped from fermentation though a preheater

Figure 5. Schematic of a Coffey still, circa 1840.

Analyser Rectifier

Wash feed

Heated wash

Perforated

trays

Wash being

stripped of alcohol

Stillage

to disposal

Steam

Vapor feed

Low strength/quality

spirit recycled to analyzer

High strength alcohol

Low boiler

recycle

Reflux as spirit condenses, heating wash

where it is heated with the overhead vapors fromthe column. It passes through a tank that ventsthe liberated carbon dioxide (CO2) to theatmosphere and then into the column.

Steam enters at the base of the column, eitherthrough a reboiler or by direct injection. Thecolumn below the feed tray (stripper section)removes the alcohol. The stillage passing outthe base should be less than 0.07% alcohol. Sinceit is a whole mash, it is important that thestripping section uses trays that are not easilyfouled by the grain solids or mineral buildup.

The alcohol vapors pass through a solidsseparation tray into the concentrating section.This section concentrates the alcohol from 55to 80% by volume. The draw strength has a greateffect on the flavor profile of the spirit. Spiritstrength is controlled by a temperature readingat the top of the column. This controls either the


Figure 6. Typical Canadian whisky flavor still.

Vent

5 - 8

Concentrating trays

Side draws may be

used for product

Solid separation tray

20 -35

Stripping trays

Vapor from column

Vent

Cooler

CO2 removal

Beer

pre-heater

Beer:

8-10%

alcohol

Hot H20

out

H20

H2O

Temperature

control

Product:

55-80%

alcohol

Low pressure

steam

Condensate

Stillage to dryhouse

<50 ppm alcohol

Condenser

Vent

Con.

Recycle

to c

olu

mn

Reboiler

Copper demister pad

Temperature control loop

flavor in the final product, while the specificationsfor good quality vodka demand no odor or taste.Specifications will also indicate the amount ofcongeners allowed to be present. These can beless than 1 ppm. Depending on the degree ofneutrality required, the rectifying section in abeverage plant can consist of two, three or fourcolumns.

SPIRIT PREPARATION FOR THE RECTIFYINGSYSTEM

Many whisky plants in North America also

amount of reflux or product draw. If temperatureis below set point, then spirit is too strong andrecycle is reduced. If too high, then spirit is tooweak and recycle is increased.

Spirit straight from the still is raw. It is agedfor a minimum of three years in small oak barrels.

Production of light and neutral spirit

The product of neutral spirit in a continuous stillrequires the removal of all components that haveodor or taste components over a set threshold.Some products such as light rum require some

262 R. Piggot

produce a neutral spirit. These plants generallyuse the flavor column, or one of similar design,as the base alcohol to feed the rectifying system.The strength of this product is usually 70-80%.Due to this low strength, the fusel oils have notbeen removed. Some heads will be vented orsent to a heads column.

In plants designed to just produce neutral spirit,the spirit is often taken to a higher strength (over95%). This allows the removal of some fusel oilsand heads, similar to the rectification column,before it is sent to the rectifying system. Anotherpossibility is the use of a Barbet head on thecolumn to concentrate and remove the heads(Figure 7).

Extractive distillation or hydrofining

The most common way to remove fusel oils isknown as extractive distillation, or hydrofining.The operating principle behind this column isaddition of water to the spirit. Ethanol is verysoluble in water while the fusel oils are not(Figure 8). The alcohol is added to the columnabout 2/3 of the way up. The water solublecomponents (ethanol and methanol) willcombine with the water, forming a solution witha higher vapor pressure and hence move downthe column. The components that are less watersoluble will rise to the top of the column wherethey are removed. This column requires threecomponents (alcohol feed, steam and water) tobe properly balanced. Generally at least eightparts of water for each part of alcohol must beused. In practice much higher ratios (10 to 15)are normally used for good separation.

In order to balance the steam, one must firstunderstand what is happening in the column.The water coming from the top of the columnwill push the ethanol down the column, whilesteam coming from the base of the column willtry to push it up the column. The result is thatthe alcohol has a peak in concentration, knownas the pinch point, about 16 trays from the baseof the column. Keeping this pinch in the correctposition and strength is a key to the properbalancing of this column. The concentrated fuseloil removed from the top of the column can besent to a fusel oil washer. Strength at the top ofthe column will be in the 60 to 68% range. Ascan be seen in Figure 9, propanol is the most

difficult component to remove and is thereforea good indication of how well the column isworking. This cleaner low strength spirit is sentto the rectifying column.

Rectification

The rectifying column takes the low strengthspirit, removes the last traces of impurities, fuseloils and heads, and takes the strength up to theazeotropic point. One difference in the rectifyingcolumn is the large number of trays: over 60trays is common. This enables a large amountof internal reflux inside the column, which helpsconcentrate the congeners for better removal(Figure 9). Fusel oils concentrate on the traywhere alcohol concentration is 65%; propanolconcentrates on the tray with 80% alcohol. It isimportant that the column is kept stable and thatthe draws are sufficient to prevent the columnbecoming loaded with impurities. The cleanerproduct is removed about five trays from thetop of the column. While they concentrate at thetop of the column where they are removed, itshould be noted that the heads must pass theproduct draw tray to reach the top of the column.It is therefore not possible to remove all the headsfrom the product in the rectifying column. Toimprove alcohol recovery and to increase puritylevels, other columns are usually added to thesystem for demethylization, heads and fusel oilseparation. In a well-designed and run extractivedistillation system, components other thanethanol will be less than 1 ppm.

FUSEL OIL AND HEADS COLUMNS

These columns take the fusel oil and heads drawsfrom various parts of the process, concentratethem for removal and return cleaned spirit tothe process. They are generally small columnswith low feed rates that have trays or are packed.

DEMETHYLIZING COLUMN

Certain feedstocks such as fruit and potatoes canend up with very high concentrations ofmethanol that need to be reduced, or perhaps aproduct very low in methanol is required. Oneway to do this, especially if methanol cannot be


10

1

0.1

0 20 40 60 80 100

Acetaldehyde

Ethyl acetateDiacetyl

Acrolein

Methanol

Isobutanol

Isoamyl

Ethanol % (v)

Propanol

Figure 8. Relative volatility of congeners and ethanol.

Figure 7. Concentrating neutral spirit: Barbet head on the column to concentrate and remove the heads.

Fusel oil decanter

Condenser

Vent Vent

Heads

95° GL

Intermediate product

Propanol draw

Fusel oil

Water

Surge

tank

Heads

Beer

feed

Steam

Pre-strip

per

Beer strip

per

Liq

uid

Vapor

36

30

24

55

60

28

25

21

Heads

concentrator

Fusel draw 65° GL

Steam

Concentrator

Bottom

strip

per

Condenser

Condenser

264 R. Piggot

removed at the beginning of the process, is toremove it after the rectifying column. Steam tothis column is via a reboiler so as not to addwater to the system. The alcohol is added to thecolumn about 1/3 to 1/2 of the way up. Thelighter methanol and any other heads presentare driven up the column and removed. Thepurified ethanol is removed at the base (Figure10).

PACKED COLUMNS

Packed columns are known as continuouscontact columns. They have an advantage ofcontinuous vapor-liquid contact. Because of this,they may be an advantage when a small columnneeds an increase in capacity. However, they areonly used on clear, non-fouling tasks and as the

column size increases they become increasinglycomplex and not cost effective.

Energy efficiency

Distillation uses a lot of energy, however withgood design the energy requirements can bereduced or recovered for other operations.

REBOILERS

Reboilers are simply heat exchangers that takea hot stream and use it to generate steam foranother part of the process. This is illustrated inthe flavor column (Figure 10). The advantageof this is that the condensate can be returned to

Figure 9. Fusel oil and heads removal in the rectifying column.

The dotted lines indicate the spread

of the peaks at higher congener

concentrations.

Heads removed

from recycle

Temperature control loop

Increasing concentration

Propanol draw (~80% ethanol)

Fusel draw (~65% ethanol)

Heads

Propanol

Ethanol

Fusel oil

Product


Figure 10. Three column rectification system used in neutral spirit production.

the boiler instead of creating extra liquid thatmust be disposed of in the stillage.

PRESSURE DISTILLATION

One or more columns are run under pressure.Instead of the heads being condensed by coolingwater, they are passed through a reboiler thatproduces steam for another column. This reducesthe total distillation steam required.

VACUUM DISTILLATION

In this configuration, one column (usually thestripping column) is run under vacuum. Thismeans heads from an atmospheric column canpower the reboiler on the vacuum column. Thisis becoming common in distilleries using

Surge

tank

95°GL

Intermediate product

Steam

Vent

Vent

ED

heads

Steam

Excess

water Steam condensate

Final product

96° GL

Propanol draw

Extractiv

e dis

tilla

tio

n colu

mn

Rectif

ier

Dem

ethylizer

Strip

pin

g

sectio

n

Hot w

ater

Condenser

30

40 75

70

50

30

23

20

15

Condenser

Heads

Condenser

Vent

Heads

Condenser

Condenser

Reboiler

molasses as a feedstock. The lower temperaturesin the stripping section reduce calcium foulingof the trays.

THERMOCOMPRESSORS

Thermocompressors are simple, low maintenancedevices that can save a considerable amount ofenergy. High pressure steam is passed througha venturi-type nozzle. This draws a vacuum inthe vessel containing the hot liquid causing it toboil. This steam combines with the now-low-pressure steam for use. About 25% (dependingon design) of the steam out of the thermo-compressor is recovered heat from the liquid.Mechanical recompression is another possibleoption, but it is costly to install and run so thesavings must make it cost effective.

266 R. Piggot

HEAT RECOVERY

The condensers should be viewed as a sourceof heat. This could be used for feed preheating,heating boiler water feed or any other processthat requires hot water such as cooking or CIPsystems.

INSULATION

The heat savings obtained by simply insulatingcolumns and other hot vessels to retain heat isoften overlooked, but usually gives an excellentreturn on investment.

Quality assurance

In order to be sure the correct quality productreaches the receiver, a system of predictivequality control must be implemented. This caninclude such things as a check on the fermentorfor odor and taste before it is pumped to the still.Samples of the fusel oil and propanol draws onthe rectifying column can be obtained to see ifthere is a buildup before it gets into the finalproduct. Whatever works in a particular system,the trick is to detect potential problems beforethey become full-blown problems that affectfinal spirit quality.

While many countries including the EU havespecifications for neutral spirit (the US does not)the companies using the spirit usually have theirown much tighter specifications that must be metby the producer.

In these days of large corporations andimproved technology there is an increasing trendto use instruments, especially the gas chromato-graph, to set standards for beverage alcohol.While these procedures can be of great use indoing predictive testing of the distillation system,they cannot duplicate the human senses when itcomes to assessing a product. To do this, it isimportant that a distillery has a trained staff,proper facilities and procedures for organoleptictesting. Remember that the final customer willnot have a gas chromatograph on the bar, buthe or she will have a nose and taste buds. Thecustomer is the final test of product quality.

References

Bathgate, G. 2003. History of the developmentof whisky distillation. In: Whiskey Technology,Production and Marketing (I. Russell, ed).Academic Press.

Murtagh, J. Feedstocks, fermentation anddistillation for production of heavy and lightrums. In: Alcohol Textbook, 3rd ed. (K.A.Jacques, T.P. Lyons and D.R. Kelsall, eds),Nottingham University Press, UK.

Nicol, D. 1989. Batch distillation. In: The Scienceand Technology of Whiskies, LongmanScientific and Technical.

From liqueurs to ‘malternatives’: the art of flavoring and compounding alcohol 267

Chapter 18

From liqueurs to ‘malternatives’: the art of flavoring and compoundingalcohol

ANDY HEAD AND BECKY TIMMONS


Introduction

Using flavor in alcoholic beverages expands thevariety of consistent, high quality productsavailable to the consumer. From high proofspirits that have a twist of fruit flavor or spice,to medium strength cordials, to low alcoholpremixed cocktails and ‘malternatives’, theinclusion of flavors gives the beverage producerlimitless flexibility in creating products to satisfythe constantly changing beverage alcoholmarket.

Historically, flavors were added to alcohol tocover up poor quality and bad taste. Gin wasfirst used for medicinal purposes in the 17th

century. Juniper berries were added prior todistillation to improve the taste. It was commonpractice to use poor quality corn and barley andto add a flavor to make the product morepalatable. In the 18th century, production offlavored vodkas became popular. This was doneto improve marketability of poor quality spirits.

Today, flavors are often added to give amarketing edge to a competitive industry.According to Adams Handbook Advance, “TheUS distilled spirits industry was up for the 5th

consecutive year in 2002”. The number offlavored alternatives is credited with this increase.

Flavor perception

Our tongues sense the four basic tastes of

sweetness, sourness, saltiness and bitterness.Additionally our mouths detect textures referredto as ‘mouthfeel’. In beverages these includeviscosity and slickness often described asoiliness. Other sensations are drying andtrigeminal sensations or salivation stimulation.Some tastes, especially bitterness, tend to lingeror are perceived less quickly.

Most of what we describe as flavor perceptionis actually aroma. Aroma perception occurs inthe nose and can occur without tasting.Thousands of aromas are distinguishable. Inorder for a chemical to be smelled, it mustbecome airborne so that it can reach the air inthe nose and be detected. Holding a beveragein the mouth warms it and more of the aromaticcompounds are released. Most aroma chemicalsare not soluble in water. Alcoholic beverageshave an advantage in that rich, intense flavorsare produced because alcohol is a natural solventthat is exceptionally volatile. The aromacompounds release more readily from analcoholic beverage than from a soft drink.

Aesthetics of flavored beverages

Because people ‘drink with their eyes’, theappearance of these lower proof products isimportant. Approved artificial colors andcaramel can produce virtually any shade of

268 A. Head and B. Timmons

color. Adding a homogenized clouding flavorcan simulate the inclusion of juice. Thecomplexity of the interaction of these taste, smelland visual sensations is what provides nuanceand character to a beverage.

Types of flavored beverages

HIGH PROOF SPIRITS

Although originally considered an inferiorsubstitute for the distiller’s art or a mask forproduct deficiencies, flavor-added spirits arenow considered to be the equal of unflavoredspirit. As previously mentioned, one of theearliest examples of added flavor in high proofspirits was improvement of inferior spirit flavorby adding juniper berries during distillation.These were the earliest gins. Now, many ginproducts are made by adding flavor from therefined oil of juniper berries, along with otheressential oils, directly to high proof neutralspirits. Refined essential oils are much moreconsistent than raw botanicals, which can beaffected by varying growing conditions.

Blended American and Canadian whiskies useflavors to provide a fuller, more balanced flavorprofile while reducing the ingredient costs ofmore expensive spirits. American whisky blendsare made by combining a straight bourbonwhisky with neutral alcohol. A well-designedflavor can mimic the flavor compounds lostfrom the neutral spirit. Vanilla can be added toreplace that which would have been extractedfrom the oak barrel. Higher alcohols and othercongeners can be added back in amounts andproportions that are more desirable than theoriginal product. Oak flavor from oak wood isnot permitted for flavoring American blendedwhiskies.

Producers in the crowded vodka and rummarket have found that flavor and a little sugarcan add an ‘extra’ that sets a product apart. Citrusextracts and pepper or spice flavors are quitepopular in flavored vodkas, which have founda new niche in trendy martini cocktails. Spicedrums, which are high in vanilla flavor, andtropical rums, reminiscent of piña coladas, havealso become the base of many popular drinks.

CORDIALS AND LIQUEURS

Cordials, and other medium strength sweetened

beverages, are an extremely important segmentof the alcoholic beverage market. Coffee,chocolate, vanilla, almond, cherry, peach,orange, cinnamon, hazelnut, anise and evengold leaf are just the beginnings of a list forpopular flavors used in this broad category.Sweetened to between 12 and 40% sugar, theseproducts can be consumed by themselves, butare frequently used as ingredients in cocktails,in cooking or even as toppings for deserts.

PREMIX COCKTAILS

Shelf-stable cream-type beverages can beproduced using non-dairy cream substitutes. Thisallows for a whole array of milkshake typeproducts, with names such as mudslides,creamsicles and grasshoppers. Margaritas andsimilar premixed cocktails give the consumerthe convenience of a complete spirits drinkwithout the inconvenience of buying multipleingredients and the hassle of complex and timeconsuming preparation.

MALTERNATIVES

‘Malternatives’ and ‘Alcopops’ are beveragesthat contain alcohol at the same level as beer.Malternatives, which are produced in the US,use as a base a malt product with an intentionallylow flavor. Flavor and sweetener are added toproduce a drink similar to a cocktail from adistilled spirit. Many are marketed using thebrand name of a spirit product, e.g. SMIRNOFFICE®, BACARDI SILVER™, etc. This categoryhas been the most successful of recentlyintroduced product lines. Among the marketingadvantages these products have is unlimitedflexibility for product design. Virtually any typeof cocktail can be developed in this base product.These products can be sold at the same retailoutlets as standard beer products, which is atremendous advantage over spirit-basedbeverages. Crossover advertising can benefit thespirit product, as well. Distilled spirits havetraditionally not been advertised on televisionin the US, however malternatives with the samebrand name can be advertised (e.g. SchmirnoffIce). The production costs are much lowerbecause of the differential between beer ($18.00per 31 gallon barrel regardless of alcohol level),wine and flavoring ($2.00 per gallon of alcohol)


and distilled spirits excise taxes ($27.00 pergallon).

Alcopops are similar products that are actuallymade with spirits set to a comparable alcohollevel. Smirnoff Ice sold outside the US willcontain vodka as the alcohol source in place ofa malt base.

Types of flavors

ARTIFICIAL VS VATURAL

For the purposes of the Codex Alimentarius“Natural flavors and ‘natural flavoringsubstances’ are preparations and singlesubstances, respectively, acceptable for humanconsumption, obtained exclusively by physicalprocesses from vegetable and sometimes animalraw materials either in their natural state or astraditionally processed for human consumption.”‘Natural’ does not mean coming from the plantthat characteristically produces the flavor. Theclassic example is benzaldehyde derived frompeach pits and used in almond or cherry flavors.‘Natural’ does not mean more healthful, saferor higher in quality than artificial. It is merely adescription of the source.

‘Nature–identical flavoring substances’ are“substances chemically isolated from aromaticraw materials or obtained synthetically. Theyare chemically identical to substances presentin natural products intended for humanconsumption, either processed or not, as definedabove.”

Another group of ingredients called ‘nature-identical’ is comprised of “substances that arechemically identical to flavoring substancesnaturally present in vegetable or animal rawmaterials that are not normally considered ashuman food.”

The same ingredient can be available in bothnatural and nature-identical forms (i.e. naturalvanilla versus vanilla that is nature-identical).The natural version is typically considerablymore expensive and less concentrated. However,there can be valuable reasons for selecting thenatural form for a beverage flavor.

‘Artificial flavoring substances’ are “thosesubstances which have not yet been identifiedin natural products intended for humanconsumption, either processed or not.”

In addition to these classifications the USBureau of Alcohol, Tobacco and Firearms(BATF) has placed restrictions on the inclusionand labeling of certain ingredients. Artificialflavor substances can be included in an alcoholicbeverage product without being labeledartificially flavored if the BATF-approved flavorcontains 0.1% or less of an artificial ingredient.There are also restrictions on the level of certainingredients, whether or not they are artificial.

KOSHER

Foods, beverages, raw materials and processingaids are kosher if they are produced inaccordance with the Jewish dietary laws. Thereare several rabbinical services that will, for a fee,certify processes, equipment and ingredients askosher and allow their seal to be placed onproduct packaging. Although these certificationsare useful for marketing purposes, Kosher doesnot mean more healthful, safer or superior inquality to non-Kosher.

BOTANICALS

These are the starting point for most naturalflavors. Botanicals are parts of a plant that containaromatic chemicals that are desirable as flavoringagents. All parts of plants have been used forflavor. A few examples are juniper berries, hopflowers, leaves and stems, rose petals and hips,cinnamon leaf and bark, whole peppers andpepper corns, clove buds, angelica root and citrusrinds. Non-plant derived natural flavoring agentsinclude beeswax, castoreum and fermentationcongeners.

EXTRACTS

It is frequently useful to remove the aromaticcompounds from the whole plant material priorto use. Essential oils, absolutes and concretesare the fragrant chemicals found in the plantmaterial. There are various extraction techniques.Expressed oils are squeezed out in presses.Carbon dioxide, propylene glycol, steam andethanol are commonly used for extraction ofessential oils. Further purification can beachieved by distillation into various fractionsbased on boiling points.


Flavor delivery systems

Most aroma chemicals are hydrophobic, notsoluble in water. In order to use them in abeverage they must be made soluble. The threecommonly used delivery systems for puttinghydrophobic flavors into beverages aresolutions, emulsions and extracts. Solutions aresimple mixtures made by dissolving flavoringredients in solvents that can be mixed intowater. Typical solvents are ethanol andpropylene glycol. Extracts or hydro-alcoholicextracts take advantage of the different solubilitylimits of the different compounds in essentialoils, especially citrus oils. Citrus extracts aremade by dissolving a citrus oil or blend of oilsin high proof alcohol and then combining or‘shocking’ with water. The less solublecompounds will come out of solution and riseto the top and can be decanted. Fortunately, withcitrus oils the compounds that are soluble in thealcohol/water portion are the moreorganoleptically pleasing compounds. Theinsoluble terpenes, which have been removed,are typically astringent and bitter. The remainingextracted portion is water-soluble.

Emulsions are made by homogenizing theflavor ingredients in a solution of a gum. A gumis a high molecular weight carbohydratepolymer. The oil droplets are trapped in the gummatrix and will not re-coalesce when mixed intothe beverage. Another use of emulsions is as aclouding agent. If a clouded product is desired,such as to simulate limejuice in a margaritapremix, a homogenized ester gum flavor canproduce the desired effect. Clouding agents canalso be used to add variation to a product line,such as with Gatorade’s Frost line of products.

Tax benefits of flavored beverages in theUS

The alcohol portion of a flavored beverage isusually provided by neutral spirits, but whiskyor other spirit can be used. Alcohol used in theflavor contributes to the total alcohol as well.Many of these products take advantage of theso-called ‘wine option’. This is a tax savingsavailable by using a neutral wine, which can beincluded at up to 49% of the alcohol in theproduct after the alcohol from the flavor is taken

into account (example formulations in Tables 1and 2).

A significant, but frequently overlooked,method of cost reduction available in the distilledspirits industry is the use of flavors to reducethe level of taxable alcohol. Alcohol used inflavors is taxed at a substantially lower rate thandistilled alcohol used in beverages. When aflavor is added to a spirits beverage, the alcoholit contains is included in the total proof.However, alcohol contained in the added flavor,when kept under 2.5% of the total proof of thefinished beverage, is not taxed as beverage spiritsalcohol. Because the tax burden on beveragespirits is so high relative to non-beverage alcohol($13.50 per proof gallon or $27.00 per gallonof pure alcohol for beverage distilled spirits asof this writing), it is possible to add a flavor thatcosts less than the tax. Flavor producers are ableto formulate alcohol-containing flavors that willnot alter the flavor profile of the finished product,thus allowing the full savings available from thistax benefit. Flavoring substances must becompounds judged by BATF to be ‘unfit’ forhuman consumption at full strength, but can bedeveloped such that when diluted and/orcombined with other flavors, the desired flavorprofile is achieved. A judicious use of flavorscontaining alcohol can reduce costs in alcoholicbeverages containing distilled spirits (See Tables3-5).

Building a beverage

Creating a new alcoholic beverage requiresdecisions at every step. One first must decideon the alcohol level and source. In addition tobeverage type, tax rates and marketingregulations affect this decision. For example,there might be a marketing benefit to using rumto make a premixed cocktail, but the cost mightbe prohibitive. A combination of rum andneutral spirits might be the compromise. Maltand other than standard wine can only be usedin products that are relatively low in alcohollevel, i.e. cordials and malternatives. They alsohave characteristic flavor and so must be usedin heavily flavored products.

Next, the flavor is selected. This is typicallythe bottleneck of the project. Choosing theperfect flavor, or combinations of flavors to give


Table 1. Vodka with full ‘drawback’1.

Vodka @ 80 proof (full drawback) Vodka @80 proof

Wine gallon Proof gallon Wine gallon Proof gallon

Vodka (@ 190 proof) 41.05263 78 42.10526 80Neutral blender (@180 proof) 1.11111 2 0.0 0Water 95 0 95 0Total 100.00 80 100.00 80

TaxesTax paid at $13.50/proof gallon for distilled spirits $1053.00 $1080.00Tax paid at $1.00/proof gallon for approved neutral blender $1.11 $0.00Total $1054.11 $1080.00

Tax difference = $1080.00-$1054.11 = $25.89/100 wine gallons (finished goods)One case of 12 x 750 ml bottles = 2.38 wine gallons2.38 wine gallons/case x $0.2589 tax reduction/wine gallon = $0.61 tax savings per case.1Full drawback is the term describing the maximum advantage taken of the lack of excise tax on alcohol in flavors (i.e. 2.5% of total alcohol). In a flavored 80 proof vodka, 2 proof degrees or 1/40th of the alcohol is non-taxable.

Table 2. Margarita premixed cocktails.

25 proof 25 proof with wine option and drawback

Wine gallons Proof gallons Wine gallons Proof gallons

High proof tequila (@ 150 proof) 16.63 24.945 8.287 12.431O.T.S. wine @21% aav2 28.428 11.944High fructose corn syrup (72°Brix) 15.00 15.00Neutral blender @ 180 proof 0.317 0.57Orange ExtractTM1 @50% aav 0.01 0.005 0.01 0.005Lime ExtractTM1 @ 50% aav 0.1 0.05 0.1 0.05Neutral Cloud FlavorTM1 0.05 0.05Citric acid, lbs 4.2 4.2FD&C Yellow 5, lbs 0.1 0.1FD&C Yellow 6, lbs 0.02 0.2Water QS QSTotal 100 25 100.0 251Alltech Inc.2Absolute alcohol by volume

the intended effect, can take time. Any changesin the other parameters can put you back to thedrawing board with the flavor.

Sweetness level is based on the beverage typeand flavor. As little as 1% sugar can smooth theroughness of a high proof spirit. Premixesblended with ice need the sugar levelsufficiently high to prevent the productbecoming watered down after dilution. Fruitflavored beverages may have differentsweetness levels depending on the fruit flavor.Sugar also adds viscosity to a beverage. Someproducts such as eggnogs can be made moreviscous by adding gums such as xanthan.

Acids (usually citric and sometimes withsodium citrate) are used in most cocktails to

provide tartness and simulate juice content. Thecorrect acid/sugar ratio is important for givingthe right taste and mouth feel to a drink. Coffeeand vanilla beverages require little or no acidaddition. Cream drinks are incompatible withacid, so the flavor selected for these productsshould not need tartness to enhance flavor. Malt,which is fermented in an acidic pH range, isalso not suitable for cream drinks.

After the flavor profile is completed theaesthetic issues of color and cloud areaddressed. Many citrus flavored drinks areclouded to simulate juice content. Others suchas a melon cordial would be shiny with onlysome color added. Deep color suggests richnessand enhances the flavor experience.


Table 3. Irish cream cordial with full drawback @ 50 proof.

Wine gallon Proof gallon

Irish whiskey @ 65% aav1 37.5 48.75Neutral blender TM2 @ 90% aav 1.389 1.25High fructose corn syrup @72°brix 12.0Non-dairy cream substitute, lbs 200Water (homogenize with cream substitute) 25.00Non-alcoholic vanilla flavorTM2 0.1Caramel color, wine gallons 0.01Water QSTotal 100 501Absolute alcohol by volume2Alltech Inc.

Table 4. Strawberry malternative @ 6% a.a.v.

Wine gallon Proof gallons

Beer @ 10% aav1 30.0 6.0Strawberry FlavorTM2 @ 60% aav 5.0 6.0High fructose corn syrup @ 72°brix 12.0Citric acid, lbs 2.5FD&C Red 40, lbs 0.5Neutral Cloud Flavor TM2 0.1Water QSCarbonate to 3.5 volumes 100.0 12.0

1Absolute alcohol by volume2Alltech Inc.

Table 5. Conversion tables for computation of taxable quantity of spirits.

Bottle size Equivalent fluid ounces Bottles per case Liters per case US gallons per case Corresponds to

1.75 L 59.2 6 10.50 2.773806 1/2 gallon1.00 L 33.8 12 12.00 3.170064 1 quart

750 mL 25.4 12 9.00 2.377548 4/5 quart375 mL 12.7 24 9.00 2.377548 4/5 pint200 mL 6.8 48 9.60 2.536051 1/2 pint100 mL 3.4 60 6.00 1.585032 1/4 pint

50 mL 1.7 120 6.00 1.585032 1, 1.6 & 2 oz

Official conversion factor: 1 Li = 0.264172 US gallonTTB, 2003

Table 6. Conversion tables for computation of taxable quantity of wine.

Bottle sizeEquivalent fluid ounces Bottles per case Liters per caseUS gallons per case Corresponds to

3.0 L 101 4 12.00 3.17004 4/5 gallon1.5 L 50.7 6 9.00 2.37753 2/5 gallon1.0 L 33.8 12 12.00 3.17004 1 quart

750 mL 25.4 12 9.00 2.37753 4/5 quart500 mL 16.9 24 12.00 3.17004 1 pint375 mL 12.7 24 9.00 2.37753 4/5 pint187 mL 6.3 48 8.976 2.37119 2/5 pint100 mL 3.4 60 6.00 1.58502 2, 3 & 4 oz

50 mL 1.7 120 6.00 1.585032 1, 1.6 & 2 oz

Official conversion factor: 1 L = 0.264172 US GallonTTB, 2003


Conclusions

The beverage alcohol industry is increasinglycompetitive. Producers are in a continuousstruggle to increase profits and market share andconsumers are looking for the latest trend andmore convenience. The art and practice offlavoring beverage alcohol provides shortproduct development times for a wide varietyof products, improves consistency and qualityof products and can reduce tax liability andimprove profits.

Table 7. Tax and fee rates.

Product Tax Tax per package (usually to nearest cent)

Beer Barrel (31 gallons) 12 oz. can

Regular rate $18 $0.05Reduced rate $7 on first 60,000 barrels for brewer who

produces less than 2 million barrels $0.02

Wine Wine gallon 750 ml bottle

14% & under $1.071 $0.21Over 14 to 21% $1.571 $0.31Over 21 to 24% $3.151 $0.62Naturally sparkling $3.40 $0.67Artificially carbonated $3.301 $0.65Hard cider $0.2261 $0.041 $0.90 credit, or for hard cider $0.056, for first 100,000 gallons removed by a small winery producing not more than 150,000wine gallons per year. Decreasing credit rates for winery producing up to 250,000 wine gallons per year.

Distilled spirits Proof gallon 750 ml bottle

All $13.50 less any credit for wine and flavor content. $2.14 (at 80 proof)

This was last updated on November 23, 1999 (www.ttb.gov)

References

Code of regulations- http://www.ttb.gov/regulations/27cfr5.html

TTB. 2003. Tobacco Tax and Trade Bureau,BATF, Dept. of the Treasury. www.ttb.gov.

2003 Adams Handbook Advance, AdamsBeverage Group, Norwalk, CT

Production of American whiskies: bourbon, corn, rye and Tennessee 275

Chapter 19

Production of American whiskies: bourbon, corn, rye and Tennessee

RON RALPH

Ron Ralph & Associates Inc., Louisville, Kentucky, USA

Introduction: definitions of bourbon,corn, rye, wheat, and Tennessee whiskies

The US Bureau of Alcohol, Tobacco andFirearms (BATF) has set specific guidelines todefine all types of alcoholic beverages producedin the United States. The general definition ofwhisky is ‘a spirit aged in wood, obtained fromthe distillation of a fermented mash of grain’.This spirit can be produced from any grain orcombination of grains; but corn, rye and maltedbarley are the principle grains used. Whisky isan alcohol distillate from a fermented mashproduced at less than 190o proof in such amanner that the distillate possesses the taste,aroma and characteristics generally attributed towhisky; stored in an oak container, and bottledat not less than 80o proof. Also, whisky maycontain mixtures of other distillates for whichno specific standards of identity are noted.

Bourbon, rye and wheat whiskies are produced(distilled) at a proof no higher than 160o from afermented mash of not less that 51% corn, ryeor wheat and aged in new, charred oak barrelsat a proof no greater than 125o. Also, thesewhiskies may include mixtures of whiskies ofthe same type. Corn whisky differs in that itmay be aged in used or uncharred new oakbarrels. Also, corn whisky may include a

mixture of other whiskies. Tennessee whiskyhas the same definition as the other four whiskytypes, but to be labeled ‘Tennessee’ it must beproduced and aged in wood in the state ofTennessee.

All whiskies conforming to Section 5.22 of theBATF regulations must be aged a minimum oftwo years. To be designated a ‘straight’ whisky,it must conform to all regulations for its typeand be aged not less than two years. ‘Light’whisky is another type of whisky produced inthe US. It is distilled at more than 160o but lessthan 190o proof and aged at least two years inused or uncharred new oak barrels.

In the US regulations, neutral spirits, vodka,Scotch whisky, Irish whiskey, and Canadianwhisky are further defined. Neutral spirits aredistilled spirits produced from any materialdistilled at or above 190o proof and bottled atnot less than 80o proof. Vodka is a neutral spiritdistilled and treated with charcoal or othermaterials to be without distinctive character,aroma, taste or color. Scotch, Irish and Canadianwhiskies are defined as ‘distinctive products ofScotland, Ireland, and Canada, respectively, andproduced and distilled under the laws of thosecountries’.

276 R. Ralph

History of North American whiskyproduction

Whisky production began in the US in 1733when the British government passed theMolasses Act. Until that time, the colonistsproduced distilled spirits from molasses. TheMolasses Act imposed a duty on molasses ofnon-British origin. Since the American colonistsimported most of their molasses from the Frenchand Spanish islands, they were greatlyconcerned. Since non-British molasses wascheaper and more abundant, smuggling andignoring the Molasses Act (and the later SugarAct) was the basis for much of the ‘Spirit of ‘76’.

Pre-revolution grain whisky production wassmall; although history notes that settlers inwestern Maryland and Pennsylvania producedrye whisky from their abundant rye grain cropsand that rye whisky began to replace the popularmolasses-based rums. After the Revolution, theEmbargo Act cut off the supply of molasses; andwith abolition of the slave trade by the newCongress, both molasses and slaves weresmuggled into the US. These events increasedthe cost of molasses and accelerated the declineof rum.

THE WESTWARD MIGRATION

Early settlers crossing the Allegheny Mountainsincluded many Scots and Irish immigrants whowere grain farmers and distillers with knowledgeof pot still operation from their homelands.They produced the rye whisky that became thefirst ‘American’ whisky. When AlexanderHamilton needed money to pay the debtsincurred during the American Revolution, hepushed an excise tax levied on distilled spiritsthrough Congress. As news of the tax spread,the uproar and public outrage was so intensethat President Washington sent 13,000 troopsinto western Pennsylvania to quell the ‘WhiskyRebellion’. As the troops entered from the east,many farmer-distillers packed their stills andheaded west to Kentucky to avoid both the taxand the army.

The farmers found Kentucky soils not assuitable for rye and wheat crops as soils inPennsylvania and Maryland. They discoveredthat corn was much easier to cultivate. The first

writing that expounded on corn growing inKentucky comes from the Jesuit HierosmLalemont. He noted ‘to mention the Indian Cornonly, it puts forth a stalk of such extraordinarythickness and height that one could take it for atree, while it bears ears two feet long with grainsthat resemble in size our large Muscatel grapes’(Carson, 1963).

Whisky production grew rapidly in the earlyfrontier areas as the settlers found in whisky ameans of moving grain to market. A pack horsecould carry only four bushels of corn, rye orwheat; but that same horse could carry 24bushels of grain that had been mashed anddistilled into two kegs of whisky. Also, the priceof whisky was more than double the price thefarmer could get for grain.

BOURBON’S ‘ACCIDENTAL’ HISTORY

As settlements moved westward, the demand forspirits increased. Riverboats had become ameans for shipping barrelled whiskies to theirdestinations. A version of ‘bourbon history’recounts how a Baptist minister, Elijah Craig,burned or ‘charred’ the inside of fish barrels torid them of the fishy smell so he could fill thebarrels with whisky to be shipped by raft downthe Mississippi River to New Orleans. Thewhisky from the charred oak barrels ‘aged’during shipping and storage. This agingimproved the character of the whisky, gave itcolor and smoothed the taste (Carson, 1963).

Another version of this history tells of acareless cooper who accidentally let the stavescatch fire (char) when heating them for pliabilityto make into barrels. Not wanting to losemoney, he did not tell his distiller customer aboutthe charred staves in the barrels. Months later,after the distiller filled the barrels and shippedthe whisky downriver, the distiller heard pleasingcompliments about his whisky. Afterdiscovering the cooper’s ‘mistake’, the distillerasked him to repeat the charring process for allof his barrels.

Contrary to popular belief, none of this‘history’ occurred in Bourbon County,Kentucky, and no one really knows howbourbon whisky was first made. The onlyhistorical evidence indicating Bourbon Countyas the source of bourbon whisky comes from a


1787 indictment of James Garrad (later aKentucky governor) and two others by aBourbon County grand jury for retailing liquorwithout a license. The only certainty in any ofthe lore is that Kentucky has a county namedBourbon and produces a whisky by the samename (Connelley and Coulter, 1922).

ESSENTIAL TRADITIONS

Despite uncertainty about origins of the bourbonname, a tradition of good whisky making washanded down from fathers to sons forgenerations. Formulas, mash bills, yeastingmethods and skills for operating the stills werepassed along, even though many farmer-distillerscould not read or write. They did not knowacrolein from fusel oil, but they did have thespecial knack for making the ‘cuts’. They knewgood, clean yellow corn and plump rye. Theyfaithfully guarded their yeast and yeast methodsthough many could not have said whether yeastbelonged to the animal or vegetable kingdoms.Two exceptional bourbon whiskies of the 19thcentury were Old Taylor and Old Crow. OldOverholt was reportedly the best of the ryewhiskies.

During the 19th century, the pot still evolvedinto the continuous column ‘beer still’ with adoubler or thumper. The continuous stilloperations allowed distillers to move to largerfermentors, more and larger cookers andautomated grain handling. As the distilleryoperations grew and became increasinglyautomated, some of the smaller distilleries fellby the wayside or promoted their brands asbetter as a result of their ‘old time, small distillerytradition and quality’. They touted this traditionand sold it as part of the product. The well knownMaker’s Mark bourbon is a prime example oftradition and sound practice bottled andsuccessfully marketed to modern consumers.

As grain whisky and bourbon production grewin the 19th century, the US governmentincreased the excise tax and the number ofregulations. Costs passed on to the consumersdampened their enthusiasm for drink; butgovernment regulation incensed leaders of theTemperance Movement because the tax andregulations drew attention to the production andsales of liquor. Most whisky at that time was

sold ‘from the barrel’, and quality standards werealmost nonexistent. It was not until the end ofthe century that consumers could purchasewhisky sold in a corked and sealed bottle. OldForester was the first product with ‘guaranteedquality’ put on the label. Old Heritage was thefirst bourbon with a strip stamp over the cork,thereby becoming the first ‘bottled-in-bond’bourbon.

THE PROHIBITION ERA

All of the improvements in the ‘character’ ofwhisky production and consumption were to noavail when at 12:01 a.m. on Saturday, January17, 1920 all beverages containing more than0.5% alcohol were outlawed in the US(Tennessee, the state with the first registereddistillery in the country, Jack Daniel Distillery,became the third state to vote to go dry in 1910,ten long years before the passing of the Volstead(Alcohol Prohibition) Act). The 18thamendment to the Constitution, which prohibitedthe production and sale of alcoholic beverages,sounded a death knell for many distilleries. Adrive through the countryside revealed closeddistilleries choked with weeds with facilities inruins. Brown-Forman Distillery in Louisville,Kentucky, was one of the few that survivedbecause it produced its Old Forester Bourbon‘for medicinal purposes only’. Other distillerieslay in wait for ‘the experiment’ of prohibition toend.

Passage of the 21st amendment endedProhibition after 13 years. At midnight on April7, 1933 wines and beer were again legally sold;and on December 5 of that year at 5:32 p.m.bourbon was again on the market. TheAmerican distilled spirits industry surged intoproduction, building larger, more moderndistilleries. The resumed legal production andsales were reinforced by the federal governmentwhen it passed the Federal Alcohol Control Act,which eliminated the sale of bulk whiskies tothe wholesale and retail trades. This Act alsoformed the Alcohol and Tobacco Tax Divisionof the Internal Revenue Service. Thoughcreating bureaucracy and taxation, the new Actalso set regulations and definitions for producingspirits in the United States.

278 R. Ralph

DISTILLATION’S NEW ERA

The new regulations defined the production ofbourbon, rye, corn and blended whiskies as wellas gin, brandies, rums, cordials and vodkaaccording to their spirit type. The Act also notedthe use of geographical designations with anorigin defined for Scotch, Canadian, and Irishwhiskies. This evolution of the industry led tothe organization of companies that had goals ofproducing top quality spirits within the newregulations. New companies such as NationalDistillers and Schenley joined the Americandistilleries that survived prohibition, Brown-Forman and Jim Beam, along with the Canadiandistillers, Seagrams and Hiram Walker. Thesedistillers based their production methods onprecisely defined procedures from yeasting tomaturation.

During World War II, North Americandistilleries ceased whisky production and beganmanufacturing industrial alcohol for the wareffort. Distillers gained the resources for furthertechnical improvements; and at the War’s end,the industry in the United States was technicallyready to produce better bourbons, ryes, andTennessee whiskies than ever before.

While the basis for modern American whiskyproduction was developed during World War II,today’s modern American distillery operates withrecent innovations. Even with all of the moderntechnology, the US distiller still carefully controlshis yeasts, mash bills, distillation methods andmaturation criteria, essential factors for goodquality products initiated by the early pioneerdistillers.

Production and maturation operations

In the production of American whiskies, sixfactors determine the character and flavor foreach type of whisky:

1) Grain proportions in the mash bill

2) Mashing technique

3) Strain of yeast

4) Fermentation environment

5) Type and operation parameters of distillationequipment

6) Type of barrel used and the maturationprocess

These factors were recognized and carefullycontrolled in the distilling operations of the earlysettlers (Lyons, 1981). Families and companiesthat ensured consistency and control over thesefactors remain in business today. Those whofailed to strictly adhere to a regimen controllingeach factor have fallen by the wayside.

A MATTER OF DISTINCTION

Popular bourbon producers use specific processdifferences to make their products distinct. TheJack Daniel Distillery continues to produce itswhisky with the same processes used more than100 years ago. Their strict adherence to traditionalong with a down-home, folksy image haveproven successful marketing tools for thewhisky. At Maker’s Mark, grain selection andproportions are used to produce a superiorwhisky. They use wheat instead of rye in themash bill. The wheat and good, consistentcontrol of all factors, especially the barrels, makeMaker’s Mark the smoothest of the bourbons.Their claim of ‘handmade’ is justified becauseof their intense attention to productionparameters.

All American whiskies maintain standards forthe six production factors; and the variationsamong distilleries in adherence to standards forthese factors determine flavor and costdifferences. All American distillers start with acareful grain purchasing program. Though priceis a criterion, they all use No. 1 or No. 2 yellowcorn, No. 1 plump northern rye, and choicenorthern malted barley. Any off odors or below-grade grain are rejected at the distillery. Verystringent grain standards are a common featureof all American distilleries (Table 1).

MASH BILL

The mash bill may vary, with a typical bourbonhaving a mash bill of 70% corn, 15% rye and15% malt. A typical Tennessee whisky mayhave 80% corn, 10% rye and 10% malt while atypical rye whisky will have a mash bill of 51%rye, 39% corn and 10% malt. All grains are


ground, with the hammer mill being the mostcommon type of processing; however someroller and attrition mills are still in use. Themilling (Figure 1) is checked for grind by a sieveanalysis. A typical sieve analysis for grains in abourbon mash bill is shown in Table 2.

Table 2. Typical sieve analysis for grains in a bourbonmash bill.

US Sieve # Corn Rye Malt Wheat

16 15 22 2 2020 21 25 8 2630 17 13 14 1240 13 9 16 850 10 7 13 660 3 2 8 2Through 60 21 22 34 22

Mashing

Mashing techniques vary considerably, but themajor difference is whether pressure or

atmospheric batch cooking is used. Bourbon,rye, wheat, Tennessee and corn whisky aremashed using batch cookers. Only the ‘blend’or ‘light’ whisky producers use continuouscookers. Pressure cooking is usually done at124oC while atmospheric cooks are done at100oC. Cooking time varies from 15 minutes to1 hr. Conversion time and temperature are veryconsistent among distilleries. Malt is neversubjected to temperatures greater than 64oC; andconversion time is usually less than 25 minutesto minimize contamination. All distillers usebackset (centrifuged or screened stillage fromthe base of the still), but the quantity of backsetwill vary based upon the beer gallonage (gallonsof water per 56 lb distillers bushel of grain) tobe used. American whiskies have beergallonages in the 30-40 gallon range. Highenergy costs for by-product recovery haveencouraged some distillers to use lower beergallonage ratios for spirits. However, bourbonand Tennessee whisky producers continue to use30-40 gallon beers. The cooling of cooked mash

Table 1. Specifications and analyses of corn, rye, wheat and malt.

Specification Typical analysis

Corn (No. 2 recleaned)Grain odor No musty, sour or off odor ‘Meets spec’Moisture, % 14.0 (maximum) 12-14Cracked grains and foreign material, % 2.0 (maximum) 1-2Damaged kernels, % 3.0 (maximum) 0-1.5Heat-damaged kernels, % 0.2 (maximum) 0-0.1Bushel weight, lbs 55.0 (minimum) 55-60

Rye (No. 1 plump)Odor None ‘Meets spec’Moisture, % 14.0 (maximum) 10-14Thins, % 2.0 (maximum) 1-2Dockage, % 2.0 (maximum) 1-2Bushel weight, lbs 56.0 (minimum) 56-60

MaltBushel weight, lbs 35.0 (minimum) 35-38Moisture, % 6.0 (maximum) 4-6α-amylase 60.0 (minimum) 60-64Diastatic power 22.0 (minimum) 22-26Bacteria count, CFU/g 1,000,000.0 (maximum) 400-500,000

WheatOdor None ‘Meets spec’Moisture, % 14.0 (maximum) 10-14Thins, % 2.0 (maximum) 1-2Dockage, % 2.0 (maximum) 1-2Bushel weight, lbs 56.0 (minimum) 56-60

280 R. Ralph

to fermentation temperature is achieved usingvacuum (barometric condensers) or coolingcoils.

YEASTING

All whisky producers use Saccharomycescerevisiae, however the yeasting techniques varytremendously between the ‘modern’ and‘traditional’ distillers. The modern distillers haveelaborate yeast laboratories and will propagatea new yeast from an agar slant every week. Theyare very aseptic and accurate, assuring continuityof the same flavor. The ‘traditional’ distillersuse yeast stored in jugs; and though theybackstock weekly, the potential for gradual yeastculture changes and contamination can lead toflavor variances. These distillers take extra effortand care to ensure that their yeasting does notcause ester, aldehyde or fusel oil variances inthe distillate.

The most common grains used for yeastingare small grains, rye and malted barley. Thesegrains are cooked in a separate cooker to about63oC, and the pH is adjusted to 3.8 with lacticacid bacteria grown in the yeast mash. Lactic

acid production is then stopped by increasingthe temperature to 100oC for 30 minutes to killthe bacteria. This aseptic, sterile mash is thenready for the yeast from the dona tub grown inthe laboratory (Figure 2). The yeastfermentation temperature is controlled at 27-30oC; and the yeast propagates until the Ballingdrops to half the original 22o Balling reading.This yeast mash will have a yeast concentrationof 400 million cells/ml. Both modern andtraditional distillers regularly have clean, sterileyeasts free of bacterial contamination that maycause side fermentations and unusual congenersin the distillate. The ‘lactic souring’ and thealcohol content of the finished yeast mash (8%),along with sterile dona and yeast tank methodscontribute to the excellent reputation Americanwhiskies have for fermentation congenerconsistency. The advantage of using smallgrains are: preservation of enzymes forsecondary conversion, low steam requirementsand shorter processing time. Also, because ofits nutrient value, barley malt is the mostimportant constituent of yeast mashes. Corn isnot used in a yeast mash because it does notcontain the growth factors required for yeast andlactic bacteria growth.

Figure 1. Typical grain-handling and milling facility.

Grain

cleaner

Corn

Grain

receiving pit

Rye Malt

Grain

storage

bins

Meal

holding

bins

Meal

scale bin

Hammer mill

To cooker

Corn Rye Malt

Cyclone


Figure 2. Stages in yeast production.

200 gallon

dona tank

4 x 3 liter

Erlenmeyer

flasks

2 x 400 ml

Erlenmeyer

flasks

2 x 15 ml

test tubes

As required

2000 gallon

yeast tub

To fermentor

FERMENTATION

Fermentation is the simplest part of theproduction process, but requires more controlwith efficient equipment in order to have stable,consistent results. After the two or three cooksrequired are completed, cooled, and transferredto the fermentor, the fermentor is ‘set’. Settingthe fermentor means filling the fermentor withcooked mash, inoculated yeast and backset. Theyeast mash is pumped in as soon as the first cookis added to the fermentor. The addition ofbackset and/or water is done at the end of filling

to bring the fermentor to the desired final beergallonage. Most distillers use a 30-36 gallonmash, and the water:backset ratio determines theset pH. Set pH values of 4.8-5.2 are consideredto be the best starting point. The moderndistillers have closed-top fermentors withcooling coils or external heat exchangers tocontrol fermentation temperature. They usuallyset fermentors at 27-29oC and control at 30-31oC.Traditional distillers will have metal or woodopen-top fermentors without any device tocontrol fermentation temperature. They usuallyset their fermentors as cool as they can (18-21oC)

282 R. Ralph

and let the fermentors work up to 31-32oC. Allof this is controlled by mash cooker cooling,addition of cold water and weather factors.Contamination is controlled by cleaningfermentors, ensuring no mash pockets in pipesand regular steaming of fermentors.

Both traditional and modern distillers fermenttheir beers for at least 72 hrs, and some for aslong as 96-120 hrs. Three and five dayfermentations are the norm. During theseperiods the Balling will drop to 0.0 and the pHfrom 5.0 to 3.8 while the alcohol concentrationrises to 8-10%. All the changes that happenduring fermentation are checked daily byperforming ‘beer chemistry’. Balling, pH, acids,and fermentor temperature are monitored andrecorded daily. The pH is the main indicator ofcontamination and potential fermentationproblems and is regularly measured bytraditional and modern distillers (Table 3).

DISTILLATION

Upon completion of the fermentation process,the beer with 8.0-10.0% alcohol is transferred

from the fermentor to the beer well. The beerwell is a holding tank for the fermented beer,such that a continuous feed to the beer still canbe maintained. Beer wells are usually 1-1.5times the size of a fermentor. They also havecontinuous agitation to prevent solid grainparticles from settling to the bottom of the vessel.All American whisky producers use a continuousstill (Figure 3), though some have a seconddistillation ‘doubler’ or ‘thumper’ (Figure 4).The basic difference between a doubler and athumper is whether the unit is operated with aliquid level (doubler) or essentially dry(thumper). Both the doubler and the thumperprovide a second distillation.

The beer is pumped into the upper section ofthe first continuous column, the beer still, six toten plates from the top. Live steam is introducedat the bottom. Beer stripping plates 1-18 haveperforations, a downcomer from above and adish on the plate below to hold the beer liquid ata set level so the plates are never dry, as thebeer moves back and forth across each plate.The steam passing up and through theperforations, controlled by pressure or flow rate,strips the lighter, more volatile alcohol from the

Table 3. Typical analysis of beers from bourbon, rye and corn whisky production.

Bourbon Rye Corn

Set sampleBalling 13.4 13.4 12.3Titratable acidity 4.5 3.6 2.8pH 4.5 5.0 5.2Temperature, °C 27.0 24.0 27.0

24 hr sampleBalling 2.6 4.0 3.6Titratable acidity 5.1 4.6 4.2pH 4.2 4.4 4.3Temperature, °C 30.0 31.0 29.0

48 hr sampleBalling 2.4 3.6 1.0Titratable acidity 7.8 7.5 6.1pH 3.8 3.9 3.8Temperature, (°C) 30.0 30.0 30.0

Drop sampleBalling 0.4 1.5 -0.4Titratable acidity 8.2 7.9 7.1pH 3.8 3.9 3.8Temperature, °C 30.0 30.0 30.0Alcohol, % by volume 6.73 5.8 6.8Residual carbohydrates, % 8.0 8.4 4.2Residual carbohydrates, % maltose 0.73 0.6 0.46


Figure 3. Bourbon whisky still.

Low

wine

tank

Stillage

Steam

Whisky

tank

Recycled To fermentors

To Maturation

Rotam

eter

Whis

ky

coole

r

Beer f

eed

18

sie

ve pla

tes

Beer strip

pin

g

sectio

n

4-8

bubble

cap

pla

tes

Concentratin

g

Sectio

n

Beer

pre-heater

Dephlegmator

Vent

condenser

Doubler

Steam

First

distillation

Second

distillation

Whisky

tank

Whisky

tank

Low

wine

tank

(Cistern room)

New charred

white oak

barrels

Hig

h w

ine

condenser

Stillage

Steam

Rotam

eter

Whis

ky

coole

r

coole

r

Beer f

eed

(1

8 sie

ve pla

tes)

Beer strip

pin

g

sectio

n

(4

-8

bubble

cap

pla

tes)

Concentratin

g

sectio

n

Beer

pre-heater

Low wine

condenser

Vent

Figure 4. Bourbon whisky distillation system, including doubler.

284 R. Ralph

water/grain mixture on the plates. When thisalcohol gets to the top 4-8 bubble cap plates it isconcentrated to about 100o proof (50o GL). Asthe vapors continue up into the beer preheater(a heat exchanger to heat beer going into thecolumn), some alcohol is condensed andrefluxed to the top of the beer column. The restflows to the doubler or thumper as vapor. Fromeither of these pot still-type chambers, the vaporgoes to the condenser where the product isdrawn off at 130-140o proof (65-70o GL).Bourbon cannot be distilled above 160o proof(80o GL). If it comes off at more than 160o proof,it must be called ‘light’ whisky. The stillage fromthe base of the beer still is pumped to the dryerhouse to be processed. The ‘high wine’ fromthe stills is pumped to the cistern room where itis held, tested and reduced to barrelling proof(Table 4).

BY-PRODUCT RECOVERY

The by-product recovery does not usuallyreceive the same care that the other processesdemand. The only essential for this process isensuring that the backset stays hot, around 99oC,so that it is absolutely sterile when used in themash tubs, yeast tubs and fermentors. (Mostdistillers use 20-30% backset in their totalprocess.) Furthermore, the hotter the backsetremains, the greater the savings in energy costsduring the cooking process. The stillage fromthe base of the beer still has about 7-10% totalsolids. When used for backset it is screened or

centrifuged to prevent solids accumulation in thecooker and fermentors. Nearly all distillers havea dryer house, though a couple of traditionaldistillers continue to sell their ‘slop’ to nearbyfarmers. For a while some, like the Jack DanielDistillery, fed the stillage to cattle in wet-feedingoperations; but environmental restrictions havegenerally eliminated such practices. Also,modern dryer house operations have becomevery profitable and the profit helps lower thecost of making whisky.

COOPERAGE AND MATURATION

Cooperage and maturation are the processingfactors that distinguish bourbon and Americanwhisky from the other whiskies of the world.Only bourbon whisky regulations require that itbe matured in a new charred, white oak barrel.Other whiskies of the world may require the useof small wooden barrels, but no other whiskybut bourbon goes through the care and expenseof requiring small white oak barrels to be freshlycharred.

The making of the bourbon whisky barrel is avery traditional, but exact science and craft.Staves and heading are quarter-sawed frommature, white oak timber. Actually, somephysical variance exists within a single tree, butno more than between trees. After being quarter-sawed with the medullary rays not less than 45degrees to the stave surface, the staves andheading are air-dried in a stave yard. The moretraditional distilleries require wood to be air-

Table 4. Typical operating data for whisky distillations.

Bourbon Rye

Product proof 130° 130°Still pressure (inches of water) 48 42Steam rate, lb/hr 12,000 12,000Beer feed rate, gallons/minute 120 117Reflux rate from beer preheater, gallons/hr 350 300Reflux rate from dephlegmator, gallons/hr 700 750Reflux rate from vent condenser, gallons/hr 30 65Draw-off of product, gallons/minute 12.5 10.0Still losses, % 0.0004 0.00025Water temperature to vent condenser, °C 21 21Water temperature from dephlegmator, °C 79 79Beer solids, % 0.06 0.06


dried at least one year. The modern distilleryusually has no air-drying specifications andallows its barrels to be produced from wood thathas been air-dried for six months or less. All ofthe wood is kiln-dried at the cooperage, withstaves and heading dried to 12 and 10%moisture, respectively. The kiln-drying isessential to prepare the wood for the planing,milling, edging and joining operations thatcannot be done on wet wood. More important,however, is that proper drying (air-drying a yearfollowed by kiln-drying), makes the woodchemistry satisfactory for flavoring the whiskyduring maturation.

As the whisky goes through maturation (3-4years for the modern distillers and 4-8 years forthe traditional distillers), two distinct types ofreactions occur: reactions between the distillatecomponents (regardless of the barrel); andreactions that occur when the distillate extractschemical compounds from the wood. A majorfactor differing among the distillers is thewarehouse environment. Most modern distillersheat their warehouses in winter. One particulardistiller even controls the heat cycles to ensureconstant aging. The traditional distillery seldomhas heated warehouses and depends strictly onMother Nature to determine the number andrange of its heating and cooling cycles.

Whether the heat cycle is natural or forced,the greatest rate of change or formation ofcongeners occurs in the first 12-16 months. Onlyester formation occurs at a fairly constant rate.Proof increases at a fairly constant rate of 4-5%each year of aging. Other specific changesoccuring when the distillate reacts with thecharred wood are: a) aldehyde formation,specifically acetaldehyde, which comes from thealcohol via oxidation, b) acetic acid formationwith greatest activity in the first year ofmaturation, and c) ester formation (ethyl acetate)

from the alcohol via oxidation. The componentscoming from the wood are tannins, sugars,glycerol and fructose. The hemicellulose in thewood appears to be the source of the sugarsfound in aged American whiskies.

The depth of charring and ‘toast level’ in thebarrel determines the color of the whisky. Colorformation is almost instantaneous when thedistillate is put into the charred barrel, with 25-30% of the color formed in the first six months.Some color development occurs each year ofmaturation until the whisky is dumped from thebarrel. The final product is then filtered, its proofreduced with demineralized water and bottled.Compared to the sweet rums, ‘breathless vodka’,fruity gins, light Canadians and Scotches, theAmerican whiskies have flavor and bravado witha balance that is pleasant to the taste. Americanwhisky produced and matured as described hasa big, pungent aroma that leaves no doubt thatit is bourbon or Tennessee whisky.

Though the modern and traditional distillershave different levels of technology, they bothuse the same basic processes for makingbourbon or other American whiskies. Individualplant nuances produce different flavors, but theyall have an American whisky bouquet and taste.

References

Carson, G. 1963. The social history of bourbon.Dodd, Mead & Company, New York.

Connelly, W.E., and E.M. Coulter. 1922. Historyof Kentucky, Vol. II. The American HistorySociety, Chicago and New York.

Lyons, T.P. 1981. Gasohol, A Step to EnergyIndependence. Alltech Technical Publications,Nicholasville, Kentucky.

Contam

ination and hygiene

Bacterial contamination and control in ethanol production 287

Chapter 20

Bacterial contamination and control in ethanol production

N.V. NARENDRANATH


Introduction

Bacterial contamination is a major cause ofreduction in ethanol yield during fermentationof starch-based or sugar-based feedstocks bySaccharomyces cerevisiae. The sugar consumedby bacteria is diverted away from alcoholproduction and is converted into by-products.In addition to reducing the yield, the presenceof bacterial metabolites in the fermentationmedium is inhibitory to yeast growth andmetabolism. In distilleries, cleaning andsanitizing are much less rigorous compared tobreweries, and mashes are subjected to less heatand are not sterile. Contaminants can arise fromtankage, transfer lines, heat exchangers, rawmaterials, active dry yeast, poorly storedbackset, or yeast slurry used as inoculum.Microbial numbers can be significantly reducedby cleaning and sanitizing the equipment, bymaintaining backset at a temperature over 70°C,by pasteurizing or chemically sterilizing thesubstrates, and by adding antibiotics tofermentors (Ralph, 1981). In a distillery, it isnecessary to recognize the potential sources ofcontamination and know the most commonlyencountered contaminants so that appropriatemeasures can be taken to minimize seriouslosses.

Commonly encountered contaminants

The bacterial contaminants encountered duringalcohol production include both Gram-positiveand Gram-negative species. Figure 1 summarizesthe types of bacteria that can occur in a distilleryand their characteristics. Among the bacterialcontaminants encountered, lactic acid bacteriaare the most troublesome because of theirtolerance to high temperature and low pH andtheir ability to grow rapidly and survive underethanol production conditions. Therefore,discussion in this chapter will primarily focuson the effects of lactobacilli, the predominantorganism in distilleries and fuel ethanol plantsand their control.

ACETIC ACID-PRODUCING BACTERIA

While Gram-positive lactobacilli comprise themost important single group of bacterialcontaminants, certain Gram-negative bacteriasuch as acetic acid bacteria cannot be ignored.The genera Acetobacter and Gluconobactercomprise bacteria that produce acetic acid as thedominant end product of metabolism.Acetobacter oxidizes acetic acid to CO2 andwater whereas Gluconobacter just produces

288 N.V. Narendranath

Dia

gnos

tic

char

acte

rist

ics

App

eara

nce

on L

MD

A

Hab

itat

Dis

tinc

tive

ph

ysio

logy

Alt

erna

tive

na

me

Pre

sent

nam

e

Ana

erob

ic

No a

cetic

acid

productio

n

Acetic

acid

produced

from

ethanol

Grow

s o

nly

in

hig

h O

2

Lactic

acid

productio

n

Lactic

acid

productio

n

Lactic

acid

productio

n,

som

etim

es a

cetic

acid

No dis

tin

guis

hin

g

characteris

tic

s (

on

sulfite m

edia

colo

nie

s a

re

bla

ck o

n u

ndersid

e)

Yellow

zone,

green u

ndersid

e

Pro

du

ctio

nfe

rmen

tor

Are

as w

ith

et

han

ol a

nd

O2

Elo

ngated y

east,

multip

le s

hapes

Cocci

Cocci,

som

e t

etrads

Cocci,

chain

s

Rods, o

ften in

chain

s, n

o s

pores

Yellow

zone

No d

istin

guis

hin

g

characteris

tic

s

Yellow

zone

Yellow

zone

Yellow

zone

Pro

du

ctio

nfe

rmen

tor

Co

mm

on

airb

orn

e co

nta

min

ant

Pro

du

ctio

nfe

rmen

tor

Pro

du

ctio

nfe

rmen

tor

Pro

du

ctio

nfe

rmen

tor

(esp

. w/s

ucro

se)

Requir

es

sim

ple

sugars

Does n

ot

further o

xid

ize

acetic

acid

Slo

wly

oxid

izes

acetic

acid

Produces

acetic

acid

Does n

ot g

row

in f

erm

entor

Produces C

O2

Sac

char

om

on

asA

ceto

mo

nas

'Wild

yea

st'

Bre

ttan

om

yces

ZymomonasGluconobacterAcetobacter

Dekkera

Micrococcus

PediococcusLeuconostoc

Lactobacillus

Aer

obic

Cat

alas

e (+

)

GR

AM

PO

SIT

IVE

YE

AS

TG

RA

M N

EG

AT

IVE

Cat

alas

e (-

)

Fig

ure

1. C

lass

ific

atio

n an

d ch

arac

teri

stic

s of

com

mon

con

tam

inan

ts e

ncou

nter

ed in

a d

isti

ller

y.


acetic acid from sugar substrates and does notfurther oxidize the acetic acid produced.

Acetic acid bacteria are rod-shaped cells andare unable to survive and grow in the absenceof oxygen. Therefore, these organisms are nota threat in the fermentor where the conditionsare anaerobic. However, acetic acid bacteria canoccur in the yeast propagator or yeastconditioning tank, which is subjected to highagitation and aeration to promote yeast growth.Once the inoculum is pumped to the fermentorand an anaerobic system is established, aceticacid bacteria will usually die; however sufficientacetic acid to slow or inhibit yeast growth mayalready be present. It is easier to control thesebacteria in batch propagation systems where thetank is emptied, cleaned and sanitized betweenbatches.

LACTIC ACID BACTERIA

Lactic acid bacteria (LAB) are Gram-positive,catalase negative, microaerophilic oraerotolerant anaerobes. They are either rod-shaped or cocci-shaped cells that produce lacticacid as a major end product of carbohydratemetabolism. LAB have complex nutritionalrequirements (Kandler and Weiss, 1986). Thesebacteria can grow in a wide range of temperatures,

from 2 to 53°C, with an optimum temperaturegenerally between 30 and 40°C. Optimal pH forgrowth is 5.5-6.0. Growth generally occurs atpH 5.0 or less, but growth rate is significantlyreduced. Under optimal growth conditions, thedoubling time for lactic bacteria is much quickerthan that of yeast.

Metabolically, LAB possess efficientcarbohydrate fermentation pathways. The mainfermentation pathways for glucose are theEmbden-Meyerhof pathway, converting 1 moleof glucose to 2 moles of lactic acid (homolacticfermentation) and the 6-phosphogluconatepathway, resulting in 1 mole of CO2, 1 moleethanol (or acetic acid) and 1 mole lactic acid(heterolactic fermentation). The homoferment-ative LAB produce virtually a singlefermentation product, lactic acid, whereas theother LAB, called heterofermentative, produceother products, mainly ethanol (or acetic acid)and CO2 as well as lactic acid. The abbreviatedpathways for the fermentation of glucose byhomo- and heterofermentative organisms areshown in Figure 2. The differences observed inthe fermentation products are determined by thepresence or absence of the enzyme aldolase, oneof the key enzymes in glycolysis.

Lacking the aldolase enzyme, hetero-fermenters instead oxidize glucose-6-phosphateto 6-phosphogluconate and then decarboxylate

Fructose 1,6bisphosphate

Triose-3-phosphate

Pyruvate + ATP

Aldolase

Lactic acid

GLUCOSE Glucose6-phosphate

6-phosphogluconate

Xylulose-5-P + CO2

Triose-3-P + Acetyl CoA

Pyruvate + ATP

Lactic acid

AcetateEthanolATP

Phosphoketolase

Hom

ofer

men

tatio

nH

eterofermentation

Figure 2. The fermentation of glucose in homofermentative and heterofermentative lactic acid bacteria.


this to pentose phosphate, which is then brokendown to triose phosphate and acetylphosphateby means of the enzyme phosphoketolase.Whereas the hetero-fermentative lactobacillipossess ketolase but no aldolase, homo-fermentative species possess aldolase but nophosphoketolase. Obligate homofermenters arethus unable to ferment pentoses, which arebroken down by the heterofermenters viaphosphoketolase, yielding equimolar amountsof lactic acid and acetic acid. However, onegroup of homofermentative lactobacilli possessesan inducible phospho-ketolase with pentosesacting as inducers. They are thus able to fermentpentoses upon adaptation to lactic acid and aceticacid, while hexoses are homofermentativelymetabolized. Therefore, these lactic bacteria arecalled facultative heterofermenters. Table 1shows some examples of the different classesof lactic acid bacteria. Isolates of LAB fromdistilleries are well-adapted to the conditionsexisting in such fermentations (Bryan-Jones,1975). In fact, enumeration of bacteria in manydistilleries is often limited to the detection oflactic acid bacteria because aerobes andfacultative anaerobes with little pH tolerance arenot considered serious threats to product qualityor production efficiency.

Lactobacilli effects on ethanol productionby yeast

Yeasts and LAB are often encountered togetherin natural ecosystems and may be in competitionfor the same nutrients (Alexander, 1971). Thegenus Lactobacillus is of major concern todistilleries and fuel ethanol plants. Chin andIngledew (1994) reported that Lactobacillusfermentum inoculated at approximately 108 CFU/ml did not seriously affect ethanol productivityin the fermentation of diluted (14°Plato) wheatmash. Only moderate growth (2 to 8-fold

increases) of the bacteria occurred. Otherscientists, however, have reported that whenbacterial numbers exceeded 108 CFU/ml at 30hrs of fermentation, alcohol loss wasapproximately 5% (Barbour and Priest, 1988;Dolan, 1979). According to Makanjuola et al.(1992), reduced ethanol yields, lower yeastnumbers, reduced carbohydrate utilization, andan increase in acidity were all caused by thebuild up of lactic acid produced by lactobacilli.They found that a bacterial count of 4.5 × 108

CFU/ml at 30 hrs resulted in a 17% reduction inethanol yield due to a stuck fermentation.

Initial bacterial contamination of mash withapproximately 107 CFU/ml led to as much as0.6-1% v/v reduction in ethanol depending onthe strain of bacteria (Narendranath et al., 1997).These authors were the first to report that bothfinal lactic acid concentrations and decreases inethanol yields at the end of fermentation weredirectly correlated with initial numbers of viablebacteria in mash (Figure 3). Apart from thediversion of glucose for growth, the productionof end products such as lactic and acetic acid,and a suspected competition with yeast cells foressential growth factors in the fermentingmedium are the major reasons for the reductionsin yeast growth and final ethanol yield whenlactic bacteria are present.

EFFECTS OF BACTERIAL END PRODUCTS ONYEAST GROWTH

The end products of metabolism of lactobacilli,lactic acid and acetic acid, are inhibitory to yeastgrowth and metabolism. The concentrations ofacetic acid (a minor end product ofheterofermentative lactic acid bacteria and wildyeasts or a major end product of aerobic bacteriasuch as Acetobacter spp.) and lactic acidinhibitory to the growth of S. cerevisiae were0.5-9 g/L and 10-40 g/L, respectively, and an

Table 1. Three classes of lactic acid bacteria.

Type Examples

Obligately homofermentative Lactobacillus delbrueckii, L. acidophilus, Pediococcus damnosusObligately heterofermentative L. brevis, L. buchneri, L. fermentumFacultatively heterofermentative L. plantarum, L. casei, L. pentosus, L. sake


80% reduction in yeast cell mass occurred atconcentrations of 7.5 g and 38 g/L, respectively(Maiorella et al., 1983). The effects of theseacids on the specific growth rate of yeast aredetailed in the chapter on continuousfermentation in this volume. The minimuminhibitory concentration (MIC) of acetic acid foryeast growth is 0.6% w/v (100 mM) and that oflactic acid is 2.5% w/v (278 mM). However,acetic acid at concentrations as low as 0.05-0.1%w/v and lactic acid at concentrations of 0.2-0.8%w/v begin to stress yeast as seen by decreasedgrowth rates and decreased rates of glucoseconsumption (Narendranath et al., 2001a).

The inhibitory effect of weak organic acidssuch as acetic acid and lactic acid on yeastgrowth depends on the pH of the medium, thedissociation constant of the acid, and its molarconcentration. In solution, a weak acid exists ina pH-dependent equilibrium between dissociatedand undissociated states, as described by theHenderson-Hasselbach equation (pH = pKa +log[A-]/[HA] where A- and HA are thedissociated and undissociated species,respectively). This equation indicates that at apH above its pKa value, more than 50% of theacid is dissociated and that the concentration ofundissociated acid increases logarithmically asthe pH declines. Since organic acids are

0.0

0.3

0.6

0.9

1.2

1.5

1.8

Initial bacteria in mash (CFU/ml)

La

cti

ca

cid

(%w

/v)

L. paracasei Lactobacillus #3 L. rhamnosus L. plantarum L. fermentum

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

104

105 106

107 108

109

1010

Initial bacteria in mash (CFU/ml)

Red

ucti

on

ineth

an

ol(%

w/v

)

104

105 106

107 108

109

1010

Figure 3. Effect of the initial numbers of lactobacilli on the final lactic acid concentration and on the reduction in final ethanolconcentration compared to the control with no bacterial inoculation. (Adapted from Narendranath et al., 1997).

generally more toxic to microorganisms at lowpH, it is assumed that the antimicrobial activityof these acids is the result of an increasedproportion of undissociated molecules (Salmondet al., 1984). Therefore, acetic acid (pKa = 4.74)with a higher pKa value is more toxic to yeastthan lactic acid (pKa = 3.86) at any given pH ofthe medium. Acetic acid has between two andfour times more molecules in the undissociatedform over a pH range between 4.0 and 4.6compared to lactic acid (Lindgren andDobrogosz, 1990). Undissociated weak acidsdiffuse passively into the microbial cell untilequilibrium is established across the membrane.As molecules enter the cytoplasm, they dissociateat the higher intracellular pH. The protonsliberated are either pumped out of the cell inexchange for cations or are neutralized by thebuffering capacity of the cytoplasm (Booth andKroll, 1989). Figure 4 illustrates how the anions(A-) and undissociated acids (HA) are distributedinside and outside the yeast cell for a givenconcentration of acetic acid and lactic acid at aparticular pH of the medium. However,Narendranath et al. (2001b) found that acetic andlactic acids inhibit yeast growth by differentmechanisms.

Therefore, the two major reasons for reductionin ethanol yield due to lactobacilli contamination


are: 1) Glucose that could be available for yeastto produce ethanol is taken up by these bacteria,and 2) the end products of metabolism, aceticacid and lactic acid, inhibit yeast growth andconcomitantly ethanol production.

As little as 1% decrease in ethanol yield ishighly significant to distillers of fuel alcohol(Makanjuola et al., 1992). In large plants withoutputs of 400 million to 1100 million liters ofethanol per year, such a decrease would reduceincome by 1 to $3 million USD annually.

Management of lactobacilli in distilleries

Despite cleaning and sanitation procedures(covered in detail in elsewhere in this volume),bacteria may find their way into distillery

processes. These bacteria must be removed asquickly as possible. The methods used in theethanol industry to control contaminant bacteriainclude stringent cleaning and sanitation, acidwashing of yeast destined for reuse (in case ofbreweries), adjustment of mash pH, and the useof antibiotics during fermentation. Themethod(s) used depends to a large degree onthe end use of the alcohol.

USE OF ANTIBIOTICS

To control lactobacilli during fermentation,antibiotics are used in fuel ethanol plants.Antibiotics are compounds produced bymicroorganisms, which at low concentrationsinhibit the growth of other microorganisms.

CH3COOH CH3COO- + H+

CH3CHOHCOOH CH3CHOHCOO- + H+

LACTIC ACID (pKa = 3.86)

ACETIC ACID (pKa = 4.74)

pH in cell: 5.4

pH outside

cell: 2.5

H+ATP

ATP

H+

Plasmamembrane

CH3COOH CH3COO- + H+

CH3CHOHCOOH CH3CHOHCOO- + H+

40 mM added (0.36 % w/v)

95.8% undissociated (38.3 mM)

at extracellular pH

40 mM added (0.24 % w/v)

99.42% undissociated (39.8 mM)

at extracellular pH

Inside the cell 97.2% is dissociated!

Equals 1.33 M or 12% lactic acid!

Inside the cell 82.05% is dissociated!

Equals 181.8 mM or 1.1% acetic acid!

Only undissociated acid

can enter the cell

Because of dissociation occuring inside the

cell at the higher pH, enormous concentrations

of dissociated acid and hydrogen ion build up

until the undissociated acid is in equilibrum with

the external concentration.

Figure 4. Illustration of the concentrations of anions and undissociated acids that would be present in the medium and inside thecell (based on the pHout and pHin) when 40 mM of either acetic acid or lactic acid is present. The concentrations were calculated

based on the Henderson-Hasselbach equation. The concentrations differ based on the molar concentration of the acids in themedium and the pH

out and pH

in values. The concentrations indicated do not remain steady because there is constant pumping out

of the excess protons (by H+-ATPase) to raise the intracellular pH, resulting in further penetration of weak acid molecules into thecell that re-acidify the cytoplasm. In most situations, the pH of the medium also falls, affecting results. The concentrations ofanions could reach 181.8 mM for acetate and 1.33 M for lactate under the stated conditions when 40 mM acid is added to the

medium (adapted from Narendranath et al., 2001b).


strains is a global issue, and one that ethanolproducers must understand because of DDGSuse in food animal diets. Penicillin G isinactivated during the alcohol fermentation(Islam et al., 1999). Inactivation of penicillin Gis significant at 35°C. At pH 4.8, the biologicalhalf-life of penicillin G is 24 hrs at 25°C and 4hrs at 35°C, whereas at pH 3.8, it is <4 hrs forboth temperatures. Moreover, any penicillin Gstill active at the end of fermentation is definitelydestroyed in the still since penicillin Gdecomposes rapidly at temperatures over 52°C(Kheirolomoom et al., 1999). In contrast,virginiamycin is not significantly altered duringalcohol fermentation after 72 hrs at 35°C (Islamet al., 1999). Research has shown that usingover 2 ppm virginiamycin to control bacteriasuppresses the rate of fermentation. Upondistillation (100°C for 30 min) virginiamycinactivity is reduced to 2.6% of the original(Hamdy et al., 1996). Residues can therefore bepresent in DDGS; which is a potential problemin markets where this antibiotic has been bannedin animal feeds.

SYNERGY: ANTIBIOTIC COMBINATIONS

Using single antibiotics such as penicillin G orvirginiamycin repeatedly can lead todevelopment of resistance in microorganisms.Figure 5 shows the development of resistant

There are different kinds of antibiotics and theycan be grouped according to their mechanismof action against bacteria.

Penicillin G has been widely studied (Day etal., 1954) and used in the alcohol productionindustry during fermentation since the 1950s.Penicillin G is primarily active against Gram-positive bacteria. The extensive use of thisantibiotic, which acts by inhibiting bacterial cellwall synthesis, has led to the emergence ofresistant microflora. For this reason and due tothe instability of penicillin G pH levels below 5,various other antimicrobials have beenintroduced or investigated for applicationincluding virginiamycin (Hynes et al., 1997),streptomycin, tetracycline (Aquarone, 1960; Dayet al., 1954) and monensin (Stroppa et al., 2000).Tetracycline is a broad spectrum antibiotic,where as penicillin V, monensin, andvirginiamycin are more active against Gram-positive bacteria. The other antibiotics that havebeen tried such as streptomycin and polymixinare more active against Gram-negative bacteria.The typical usage rate for these antibiotics inthe fermentation is 2-5 ppm. The mechanismsof action and the spectrum of these antibioticsare summarized in Table 2.

Although many antibiotics have been testedat a laboratory scale, the most widely used inthe ethanol industry are penicillin G andvirginiamycin. Antibiotic residues andestablishment of antibiotic-resistant bacterial

Table 2. List of some of the antibiotics used in the ethanol industry, their modes of action and activity spectrum.

Antibiotic Mechanism Bactericidal/static Spectrum

1a. Penicillin G Inhibits cell wall synthesis Bactericidal Gram(+) bacteria1b. Penicillin V*

2. Bacitracin Affects cell wall Bactericidal Gram(+) bacteria

3. Tetracycline Protein synthesis inhibitor Bacteriostatic Gram(+) & Gram(-) bacteria4. Streptomycin Protein synthesis inhibitor Bactericidal Gram(+) & aerobic

Gram(-) bacteria

5. Erythromycin Protein synthesis inhibitor Bacteriostatic, cidal at high doses Gram(+) & Gram(-) bacteria6. Polymixin Affects cell membrane Bactericidal Gram(-) bacteria

7. Virginiamycin Protein synthesis inhibitor Bactericidal Gram(+) bacteria

8. Monensin Affects cell membrane Bactericidal Gram(+) bacteria

9. Chloramphenicol Protein synthesis inhibitor Bactericidal (or) static Gram(+) & Gram(-) bacteria.Good against anaerobes.Heat stable?

* Has the same properties as penicillin G but is more stable at acidic pH.


mutants to single antibiotics such as penicillinor virginiamycin. One of the strategies to reducethe risk of resistance is to use a combination ofantimicrobials, including non-antibioticsubstances, thereby taking advantage ofsynergism in activity as well. LactosideTM, aproprietary blend of antimicrobials, reduced riskof resistance while increasing spectrum ofactivity and pH range. Following five years ofuse, no build up of resistance has been reported.

The advantage of balanced mixtures ofantibiotics has been demonstrated in severalpractical situations. A recent experimentexamined corn mash (30% dry solids; pH 5.6)contaminated with an aggressive Lactobacillusparacasei strain. Yeast was inoculated at 30million cells/ml. Just prior to yeast inoculation,test fermentors were contaminated (inoculated)with L. paracasei at ~107 CFU/ml (in thetreatments with bacteria). The antibiotics used,penicillin G, virginiamycin, and Lactoside™,were added from 10,000 ppm stock solutions.Virginiamycin was made up in HPLC-grademethanol. The temperature was maintained at31.1°C (88°F) throughout the fermentation.Samples were withdrawn for analysis from eachfermentor at 8, 24, 32, 48, 56, and 72 hrs. Viablecell counts were monitored by the spread platetechnique. For enumeration of bacteria, the MRSagar plates (with 10 ppm cycloheximide) wereincubated in anaerobic jars at 30°C. The platingwas done in duplicate for each dilution. Results

indicated that LactosideTM significantly reducedthe viable numbers of bacteria. Virginiamycinwas the least effective antibiotic at the dosageused in controlling L. paracasei (Figure 6).

In experiments performed to determine theminimum inhibitory concentrations (MIC) forvarious antibiotics against lactobacilli at variousstages of their growth, LactosideTM at lowconcentrations killed L. plantarum at early, midand late log phases of growth (Table 3).

Table 3. Minimum inhibitory concentrations fordifferent antibiotics and LactosideTM (expressed in ppm)for L. plantarum at different stages of its growth.

Antibiotic MICs when antibiotic added atvarious stages of growth

0 hrs Early log Mid log Late log

Penicillin G 0.2 12.8 >12.8 >12.8Penicillin V 0.2 12.8 >12.8 12.8Virginiamycin 0.2 >12.8 >12.8 >12.8LactosideTM 0.2 0.2 0.2 0.8Oxytetracycline >6.4 >6.4 >6.4 >6.4Erythromycin 12.8 >12.8 >12.8 >12.8Streptomycin >12.8 >12.8 >12.8 >12.8

Similar results were obtained with four otherspecies of lactobacilli. Figure 7 shows the regionsof a typical growth curve for a microbe wherethe various antibiotics when used at lowconcentrations would be effective in killing themicrobe. This information is important for a

Figure 5. A typical disc assay performed to test the antimicrobial activity of penicillin G and virginiamycin against Lactobacillusparacasei. Note the growth of resistant mutants in the zone of inhibition.


0

2

4

6

8

10

0 10 20 30 40 50 60 70 80

Time (h)

Via

ble

bac

teri

a(l

og

CF

U/m

)10

No antibioticsVirginiamycin(5 ppm) Lactoside (5 ppm)No antibiotics, no yeast

Penicillin G (5 ppm)

Figure 6. Survival of L. paracasei in the fermentation of corn mash (30% DS, pH 5.6) at 31.1°C (88°F) in the presence orabsence of various antibiotics. Values are means of duplicate fermentations that had a CV of <8%.

Figure 7. A typical growth curve of a microorganism showing at what stages antibiotics are effective at low concentrations.

distillery since contamination can be detectedas early as 6-10 hrs from the start of filling afermentor. By this time, the contaminants areapproaching the mid-log phase of growth.Therefore, an antibiotic or antimicrobial effectiveat a low concentration at this stage would be anideal choice.

The ideal antimicrobial should, however, possessall the following characteristics:

1. Broad-spectrum activity: effective againstboth Gram-positive and Gram-negativebacteria.

2. Effective over a wide pH range.

0.000

0 12 24 36 48 60

0.200

0.400

0.600

0.800

1.000

1.200

Lactoside�

Penicillin GVirginiamycin

Time (hrs)

Ab

sorb

ance

(62

0 n

m)


3. Control the bacteria at various growth stages.

4. Effective at low concentrations.

5. Bactericidal rather than bacteriostatic.

6. Able to be combined with other antibioticsor antimicrobials to avoid resistancedevelopment in bacteria.

FORMULATION OF ANTIBIOTICCOMBINATIONS

Factorial experiments are conducted to testinteraction effects (either synergistic orantagonistic) among antibiotics used incombination to control growth of a test lacticacid bacteria. These experiments are conductedin corn mash rather than laboratory media sothat the results can be translated to the field. Fromthese experiments, the individual effects of theantibiotics can also be determined. Lactic acidis the parameter measured since lactic acidproduced is directly proportional to the viablecell numbers of bacteria present in mash(Narendranath et al., 1997). The results of theinteractions are plotted in 3-D graphs with thetwo test antibiotics on x- and y-axes, and lacticacid on the z-axis. The graphs are created basedon the response surface regression analysis ofthe data using the SAS program. Figure 8 showsthe model created by using the equation basedon the data obtained in an experiment. Thisequation can then be used to make formulations

with correct proportions of each antibiotic.Experimental designs are available to generatesuch equations for studying the combination ofmore than two antibiotics.

It is becoming a common practice in someindustrial operations to either under- or overdoseantibiotics for controlling contaminatingbacteria. Under-dosing antibiotics leads to thedevelopment of resistance in microorganisms.Overdosing antibiotics can affect fermentationrate by yeast (Hamdy et al., 1996) and increasethe chances that all the antibiotic will not beinactivated by the distillation process. Therefore,it is always advisable to use the antibiotics atthe dosage recommended by the supplier/manufacturer.

LOWERING MASH pH CAUSES ETHANOL LOSS

Since the growth rate of lactic acid bacteria isreduced significantly at pH levels <5.0 (Kandlerand Weiss, 1986), most ethanol plants tend toreduce the pH of the mash with sulfuric acid to<4.5 at the start of fermentation. Somecontinuous fermentation plants even startfermentations at pH of 4.0 or less. Loweringmash pH may reduce the growth rate ofcontaminant bacteria but it significantly reducesethanol yield. In an experiment using mashesof various dissolved solids concentrationsadjusted to pH values from 4 to 5.5, significantlyhigher ethanol was obtained when the initial

00. 5

12

0

0. 5

1

2

0. 0

0. 2

0. 4

0. 6

0. 8

1. 0

1. 2

Lact

icac

id (%

w/v

)

Penicillin

Monensin

1.0-1.2

0.8-1.0

0.6-0.8

0.4-0.6

0.2-0.4

0.0-0.2

Figure 8. Lactic acid produced by L. plantarum in corn mash (pH 5.5, D.S. 24%) at 30°C as affected by various concentrations(ppm) of penicillin and monensin in combination in a 4 × 4 factorial experiment. Model was significant (P<0.001) with R2 = 0.9993.


mash pH was 5.5 (Figure 9). Moreover, yeastcan tolerate higher acetic acid (0.1-0.2%w/v) andlactic acid (up to 3%w/v) when the starting pHis 5.5, since at this pH these acids are primarilyin the dissociated form. By the end offermentation, however, the pH drops 1 unit to4.5 under normal ‘contaminant-free’ conditions.Similar observations were also made with wheatand milo mashes.

The occurrence of bacterial contaminants inan industrial-scale ethanol production processis unavoidable, especially when the pH of themash at set (start of fermentation) is between5.0 and 5.5. Therefore, a good cleaning andsanitation regime combined with an effectiveantimicrobial program will reduce bacterialnumbers and significantly increase ethanolyield, which ultimately results in increased profitfor ethanol producers.

References

Alexander, M. 1971. Microbial ecology. JohnWiley & Sons, Inc., London, UK.

Aquarone, E. 1960. Penicillin and tetracyclineas contamination control agents in alcoholicfermentation of sugarcane molasses. Appl.Microbiol. 8:263-268.

Barbour, E.A. and F.G. Priest. 1988. Some effectsof Lactobacillus contamination in scotch

whisky fermentations. J. Inst. Brew. 94:89-92.Booth, I.R. and R.G. Kroll. 1989. The

preservation of foods by low pH. In:Mechanisms of action of food preservationprocedures (G.W. Gould, ed). Elsevier,London, UK, pp. 119-160.

Bryan-Jones, G. 1975. Lactic acid bacteria indistillery fermentation. In: Lactic acid bacteriain beverages and foods. Proceedings of the IVLong Ashton Symposium (J.G. Carr, C.V.Cutting and G.C. Whiting, ed). AcademicPress, London, UK, pp. 165-175.

Chin, P.M. and W.M. Ingledew. 1994. Effect oflactic acid bacteria on wheat mashfermentations prepared with laboratorybackset. Enzyme Microb. Technol. 16:311-317.

Day, W.H., W.C. Serjak, J.R. Stratton and L.Stone. 1954. Antibiotics as contaminationcontrol agents in grain alcohol fermentations.J. Agric. Food Chem. 2:252-258.

Dolan, T.C.S. 1979. Bacteria in whiskyproduction. Brewer 65:60-64.

Hamdy, M.K., R.T. Toledo, C.J. Shieh, M.A.Pfannenstiel and R. Wang. 1996. Effects ofvirginiamycin on fermentation rate by yeast.Biomass Bioenergy 11:1-9.

Hynes, S.H., D.M. Kjarsgaard, K.C. Thomas andW.M. Ingledew. 1997. Use of virginiamycinto control the growth of lactic acid bacteria

7

8

9

10

11

12

13

14

15

16

20 25 30 35

Dissolvedsolids (%w/v)

pH=4.0 pH=4.5 pH=5.0 pH=5.5

Eth

ano

l (%

v/v)

Figure 9. Ethanol produced by S. cerevisiae in 48 hrs at 30°C at different concentrations of dissolved solids in the medium at fourpH levels. Coefficient of variation among the duplicates was <2%.


during alcoholic fermentation. J. Ind.Microbiol. Biotechnol. 18:284-291.

Islam, M., R. Toledo and M.K. Hamdy. 1999.Stability of virginiamycin and penicillin duringalcohol fermentation. Biomass Bioenergy17:369-376.

Kandler, O. and N. Weiss. 1986. Regular, nonsporing gram-positive rods. In: Bergey’smanual of systematic bacteriology, Vol. 2(P.H.A. Sneath, N.S. Mair, M.E. Sharpe andJ.G. Holt, ed). The Williams & Wilkins Co.,Baltimore, MD, pp. 1208-1234.

Kheirolomoom, A., A. Kazemi-Vaysari, M.Ardjmand and A. Baradar-Khoshfetrat. 1999.The combined effects of pH and temperatureon penicillin G decomposition and its stabilitymodeling. Process Biochem. 35:205-211.

Lindgren, S.E. and W.J. Dobrogosz. 1990.Antagonistic activities of lactic acid bacteriain food and feed fermentations. FEMSMicrobiol. Rev. 87:149-164.

Maiorella, B., H.W. Blanch and C.R. Wilke.1983. By-product inhibition effects onethanolic fermentation by Saccharomycescerevisiae. Biotechnol. Bioeng. 25:103-121.

Makanjuola, D.B., A. Tymon and D.G. Springham.1992. Some effects of lactic acid bacteria onlaboratory scale yeast fermentations. EnzymeMicrob. Technol. 14:351-357.

Narendranath, N.V., S.H. Hynes, K.C. Thomas andW.M. Ingledew. 1997. Effects of lactobacilli onyeast-catalyzed ethanol fermentations. Appl.Environ. Microbiol. 63:4158-4163.

Narendranath, N.V., K.C. Thomas and W.M.Ingledew. 2000. Urea hydrogen peroxidereduces the numbers of lactobacilli, nourishesyeast and leaves no residues in the ethanolfermentation. Appl. Environ. Microbiol.66:4187-4192.

Narendranath, N.V., K.C. Thomas and W.M.Ingledew. 2001a. Effects of acetic acid andlactic acid on the growth of Saccharomycescerevisiae in a minimal medium. J. Ind.Microbiol. Biotechnol. 26:171-177.

Narendranath, N.V., K.C. Thomas and W.M.Ingledew. 2001b. Acetic acid and lactic acidinhibition of growth of Saccharomycescerevisiae by different mechanisms. J. Am.Soc. Brew. Chem. 59:187-194.

Ralph, R.J. 1981. Practical aspects of operatingan alcohol plant. In: A Step to energyindependence - a text book for fuel alcoholproduction (T.P. Lyons, ed.). Alltech TechnicalPublications, Lexington, KY, pp. 255-265.

Salmond, C.V., R.G. Kroll and I.R. Booth. 1984.The effect of food preservatives on pHhomeostasis in Escherichia coli. J. Gen.Microbiol. 130:2845-2850.

Stroppa, C.T., M.G.S. Andrietta, S.R. Andrietta,C. Steckelberg and G.E. Serra. 2000. Use ofpenicillin and monensin to control bacterialcontamination of Brazilian alcoholfermentations. Int. Sugar J. 102:78-82.

Managing the four Ts of cleaning and sanitizing 299

Chapter 21

Managing the Four Ts of cleaning and sanitizing: time, temperature,titration and turbulence

JIM LARSON AND JOE POWER


Introduction

Cleaning and sanitizing is an integral part of theoperation in many processing facilities includingdistilleries, ethanol plants and breweries. Thecleaning and sanitizing effort represents asignificant investment of time, labor andoperating costs. At the same time there is usuallya positive payback for this effort in terms ofyields, efficiencies, safety and product quality.

In a clean plant that is running efficiently withno infection problems, there is always thetemptation to reduce operating costs by trimmingback cleaning times and chemical usages. Thistype of ongoing evaluation is a worthwhile partof plant management. We should always striveto clean and sanitize to the extent required bythe process and the equipment but overkill inthis area should be avoided. If a little bit is good,a lot is not necessarily better.

To achieve the proper degree of cleaning andsanitizing at an acceptable cost, there are amyriad of approaches and tools that can be used.Sometimes called the ‘cleaner’s bag of tricks’,these tools can be grouped in four categoriesknown as the ‘Four Ts’ of Cleaning andSanitizing. The Four Ts are:

• Time

• Temperature

• Titrations

• Turbulence

Time refers to the frequency of running thecleaning operation, for example weekly, daily,or after each batch. In a given cleaning cycle italso refers to the length of each step. For examplea 5 min rinse, a 1 hr caustic wash, etc.

Temperature refers to the temperature of thecleaning solutions used in each step of thecleaning cycle. The first rinse may be at ambienttemperature while the caustic wash may bespecified between 150 and 160oF.

Titrations refers to the chemistry of the cleaningsolutions. This includes selecting the rightcleaning chemical for the job being done. It alsoincludes the concentration of chemicals used inthe cleaning solutions.

Turbulence is the mechanical action in thecleaning program that will physically scrub soilfrom dirty surfaces. Turbulence encompasses theuse of scrub brushes and high-pressure nozzlesas well as pumping cleaning solutions at highvelocities through dirty pipelines.

300 J. Larson and J. Power

For any cleaning job, many combinations oftime, temperature, titrations and turbulence canbe used to achieve effective cleaning at areasonable cost. There are also manycombinations that will not work because of adeficiency in one or more of the Four Ts. Forexample, cleaning time may be too short or tooinfrequent. The detergent may be too weak, itmay be at too low a temperature or it may notbe the right choice for the soil type present(Figures 1 and 2).

If cleaning is inadequate, it is usually possibleto correct it by increasing the intensity of one of

the Four Ts (Figure 3). If the turbulence createdby the velocity of cleaning solution flowing in apipeline is inadequate, it may be possible tocompensate by cleaning for a longer time, byincreasing the concentration of the cleaningchemicals or by raising the cleaning solutiontemperature.

Our challenge is to achieve an acceptabledegree of cleaning and sanitizing in specificprocesses. We sometimes go through elaborateprocedures but in the end still have infections.Why?

Figure 2. Some combinations do not work; and dirt remains.

TemperatureTime

TitrationTurbulence

TemperatureTime

Turbulence

Titration

TemperatureTime

Titration

Tu

Figure 1. Some combinations work.

Figure 3. Deficiencies in one of the Four Ts can be corrected or compensated by changing another.

Temperature Temperature

Temperature


This paper examines the tools at our disposalfor effective cleaning and sanitizing – ourcleaning bag of tricks. Proper attention to theFour Ts in the toolkit and how they arecombined is needed to make cleaning andsanitizing effective.

Definitions

Cleaning is reducing the amount of soil to anacceptable level. Cleaning is usually done witha combination of water and added detergents.Undercleaning means that we have not done thejob properly. Gross overcleaning means wastedresources – time, chemicals and manpower. Anacceptable level of cleaning in a pharmaceuticalplant would probably be overcleaning in anethanol distillery.

Sanitizing is reducing the population of viableorganisms on a clean surface to an acceptablelevel. Again, ‘acceptable’ has different meaningsin different operations. Within the ethanolindustry it may even vary from plant to plant.Heat and/or sanitizing chemicals are used forthis step.

Disinfection is the destruction of all vegetativemicroorganisms, but not necessarily spores.

Sterilization is the complete destruction of allorganisms including spores and viruses.

Soil is any unwanted material left on a surfacethat needs to be clean. Soils can be composedof sugars, salts, fats, proteins, microbes, scalesand other mineral deposits. Knowing the natureof the soil is necessary when planning thecleaning procedure used to remove it. Soils lefton surfaces harbor contaminating bacteria,reduce efficiency in heat exchangers and plugpipes and other passageways.

CIP is the acronym for cleaning-in-place.Pipelines, tanks, process equipment andaccessories are cleaned by pumping cleaningsolutions through them without disassembly ormanual cleaning. CIP processes are usuallyautomated.

COP is cleaning-out-of-place. COP involvesmanual disassembly and cleaning by hand.

Therefore, cleaning is defined as the removal ofsoil while sanitizing is the reduction of viableorganisms remaining on the clean surface. Animportant concept in cleaning and sanitizing isthat first we remove the dirt, then we sanitize.Sanitizing dirt is not in the program!

Time

If you don’t get it clean the first time, wash itagain. Spend more time cleaning and you willget better results. This sounds like a simplesolution, but a dilemma arises because time spentcleaning could be time spent producing product.

Figure 4 illustrates this in the operation of abatch ethanol fermentor. In the fermentationcycle, a period of time is set aside for cleaningand sanitizing between batches. The ethanolconcentration curve, however, shows that we canproduce more ethanol in the fermentor if we omitall or part of the cleaning time. But what will thelonger-term results be if we take shortcuts incleaning? We must find a happy medium, whichis the optimum amount of time taken fromproduction to keep the process equipment clean.The other three Ts of cleaning and sanitizing canhelp us do this.

Time refers to the frequency of cleaning as wellas the length of the cleaning cycle. We learn fromexperience that some parts of the process arevery sensitive to infection and require frequentcleaning. An example is the yeast propagationsystem, which is thoroughly cleaned andsanitized after every use. Other parts such as theslurry tank or liquefaction tank are less criticaland may be cleaned every 1-6 months.

Continuous fermentation systems presentspecial problems compared to batch fermentors.A continuous fermentation train may be shutdown only one or two times a year for cleaning.This shutdown requires a total stoppage ofproduction, so it has significant cost implications.Because of the continuous flow through thefermentation train, there is some degree offlushing of the infecting organism. Frequentlythe culture yeast grows more slowly than thecontaminant, and reduced ethanol yield requiressome action. Antibiotics such as Lactoside ™and Allpen™ and antimicrobials may be addedto the mash to kill bacteria without requiring ashutdown. For a wild yeast infection it is usually


necessary to shut down and clean becauseagents that kill the wild yeast will also kill theculture yeast.

Organizing a cleaning schedule for every pieceof equipment in the plant is an important part ofthe CIP program. An example of such a scheduleis shown in Table 1.

Temperature

Warm cleaning is more effective than coldcleaning. Within limits, this is usually true incleaning and may or may not be true insanitizing. Caustic soda (sodium hydroxide)cleaning solutions can be used at ambienttemperatures but they are normally heated, ashigh as 82°C (180°F). Reactions proceed fasterat higher temperatures and many soils are moresoluble in hot water than in cold water. Thecleaning process can therefore be faster and moreeffective if hot solutions are used.

Table 2 shows the results of a cleaning studycomparing warm and cold detergent solutionsto clean dirty beer lines. The amount of soilpresent was measured with an ATP-detectingluminometer. Starting from a ‘soil overload’condition (>500,000), the cold solution left 300times as much soil as the warm solution.

Table 2. Effect of temperature on amount of soil presentbefore and after use of cold or warm cleaning solutions.

Cold cleaning Warm cleaning23°C (74°F) 57°C (135°F)

Before cleaning >500,000 >500,000After cleaning 18,242 59

When raising the temperature of a cleaningsolution it is important to check the chemicalcompatibility of the equipment material ofconstruction with the hot cleaning solution.Some hot acid solutions, for example, may betoo corrosive for metals such as mild steel orbrass.

Most sanitizing solutions lose effectivenesswhen the temperature is raised above a thresholdlevel. For example, cool or tepid water shouldbe used for preparing solutions of IotechTM andother iodophor sanitizers and they should bestored below 38°C (100°F). At highertemperatures the vapor pressure of the activeingredient causes it to vaporize.

Temperature can be an effective tool in the‘cleaning bag of tricks.’ For example, if we wantto reduce the time spent cleaning a product transferline without sacrificing cleaning effectiveness, a

Figure 4. The dilemma between cleaning time vs. production: time = money!

%Ethanol

Fermentoruse

Money

Fill

Time

Time

Fill

Ferment

Empty Empty

Ferment

CIP


Tab

le 1

. Con

trol

roo

m C

IP s

ched

ule.

Mon

day

Tues

day

Wed

nesd

ayT

hurs

day

Frid

aySa

turd

aySu

nday

DAY

TIM

ESt

art

DAY

TIM

ESt

art

DAY

TIM

ESt

art

DAY

TIM

ESt

art

DAY

TIM

ESt

art

DA

YTIM

ESt

art

DAY

TIM

ESt

art

time

time

time

time

time

time

time

Dis

tilla

tion:

Dis

tilla

tion:

Dis

tilla

tion:

Dis

tilla

tion:

Dis

tilla

tion:

Dis

tilla

tion:

Dis

tilla

tion:

Spir

al h

eate

r:Sp

iral

hea

ter:

Spir

al h

eate

r:Sp

iral

hea

ter:

Spir

al h

eate

r:Sp

iral

hea

ter:

Spir

al h

eate

r:

NIG

HT

NIG

HT

NIG

HT

NIG

HT

NIG

HT

NIG

HT

NIG

HT

Exc

h:13

02 A

or B

Exc

h: 1

400

Exc

h: 1

402

Exc

h: 1

404

Exc

h: 1

302

A o

r BE

vapo

rato

rE

xch.

130

3E

xch:

140

1E

xch:

140

3E

xch:

130

1BE

xch:

130

2CE

xch

1302

C E

xch:

140

6E

xch:

103


warmer or hotter cleaning solution could be theanswer.

Titrations

Titrations, the third T in the cleaning andsanitizing bag of tricks, refers to the chemistryof cleaning and sanitizing chemicals andsolutions. It includes matching the type ofchemical cleaner to the type of soil present andusing that chemical at a strength that is botheffective and affordable. It is therefore essentialto have as much information as possible aboutthe soil we are trying to remove (McCabe, 1999;Kretsch, 1994).

THE NATURE OF SOILS

Any of the compounds present in raw materialsor in the water supply used for fermentation maybecome insoluble and be deposited onequipment surfaces as an unwanted soil.Methods used to remove these soils depend onthe specific nature of the material involved.

Carbohydrates

The soluble sugars present during fermentationmay be left on surfaces, especially the non-fermentable sugars. A much bigger problem isthe higher molecular weight carbohydrates suchas gums, which can form tough deposits onsurfaces. Retrograded starch can form whenstarch is not sufficiently digested by α-amylaseenzymes. Depending on the type of grains used,gums such as ß-glucans or pentosans maybecome soluble in the mash and then bedeposited later on. Dextrans produced byLeuconostoc bacteria from sucrose can beproduced in plants fermenting sucrose-containing materials.

Proteins

Most of the protein present in grains is insolubleand passes through mashing and fermentationin the insoluble form. The soluble portion of theprotein may become denatured by conditionssuch as temperature changes or pH shifts. Asthe protein becomes insoluble it may bedeposited onto various surfaces. If tannins are

present, they may react with proteins in a reactionsimilar to the tanning of leather, forming tough,water-resistant deposits.

Oils

Oils are liquid fats and are intrinsically insoluble.Much of the oil present in grains is broken intomicroscopic droplets or emulsions that aresuspended in the mash. As conditions change,the emulsions may coalesce forming largerdeposits of insoluble oil.

Inorganic compounds

Various inorganic compounds may becomeinsoluble during mashing or fermentation. Mostcommonly calcium in water reacts with theoxalate present in grains to form insolublecalcium oxalate, commonly called beerstone.Carbonates present in water may combine withcalcium to form calcium carbonate scale.Magnesium hydroxide tends to co-precipitate aspart of the carbonate scale. Occasionally waterwith high silicate content may generate a silicascale as the pH drops. Soluble silicate isconverted by acid to insoluble silica (SiO2).

Biological deposits

Many microorganisms tend to grow better onsurfaces than when freely suspended in a liquid.The layer of microorganisms that forms togetherwith other materials such as non-living soil andexternal polysaccharides produced by themicroorganisms is called a biofilm (Czechowskiand Banner, 1992). A biofilm may be observedas a slimy layer on tanks used to store water.Fortunately, most microorganisms found infermentation do not have a pronouncedtendency to form a biofilm, but films of yeast orbacteria growing inside a surface scale arefrequently observed. These films are a potentsource of contamination.

MATERIALS USED FOR CLEANING

Water

Water is a good solvent for many compounds,especially polar compounds. Materials such assimple sugars, soluble proteins and the soluble


products of fermentation are easily removed withwater. Most compounds are more soluble in hotwater than cold, so heating water helps make ita better cleaner. Exceptions are the waterhardness compounds, particularly those oftemporary hardness. These become less solubleat higher temperature leading to the depositionof calcium carbonate-based water scales. Animportant consideration in the use of hot waterfor cleaning is to prevent formation of scales bysoftening the water. Many cleaners containwater-softening ingredients.

Pure water has a relatively high surfacetension, as shown by the fact that water tends toform round drops when placed on a surface.Water is described as not being good at wettingbecause the water forms droplets and does notwet the entire surface. Compounds thatconcentrate at the surface of water often decreasethe surface tension of water droplets, making iteasier for the water to spread out and cover moreof the surface. These compounds are calledwetting agents.

Alkali

An alkali is something that raises the pH ofwater. The pH depends on the concentration ofhydroxide ions: the more hydroxide the higherthe pH. The most common alkali used forcleaning is caustic soda (sodium hydroxide),which is an excellent source of hydroxide ionsand is relatively inexpensive. Unfortunately,caustic soda has poor rinsability; it does notrinse away very easily after it is used for cleaning.Potassium hydroxide is also a strong alkali andit rinses more easily than sodium hydroxide, butit is significantly more expensive. Addition ofsurfactants to cleaning compounds with causticsoda will improve rinsability.

Other alkaline compounds include silicates,phosphates and carbonates. Sodium metasilicateis a commonly used alkali. It provides a highpH, although not as high as caustic. It tends tokeep removed soil in suspension, providinggood removal from the area cleaned. Silicatedoes not tend to corrode soft metals such asaluminum and copper, while caustic attacksthese metals. Silicates are also less dangerous tohandle.

Alkaline phosphates, especially trisodiumphosphate, provide significant alkalinity, but less

than caustic or silicates. It is safe for use in handcleaning applications. It helps to soften waterby forming an easily removed precipitate withhardness compounds. It can be combined withsodium hypochlorite to form a co-crystal, whichis a convenient, stable source of chlorine.

Sodium carbonate or washing soda is a mildlyalkaline compound that helps soften water aswell as provide alkalinity. Its main use is in amixture with solid caustic soda where its anti-caking activity helps prevent handling problems.

When alkaline compounds raise the pH level,they increase the amount of negative hydroxideions present. This has two main effects. Bondsbetween the amino acids of proteins can behydrolyzed or broken by reaction withhydroxide. This makes the proteins smaller andmore soluble so that they can be removed. Acidcompounds can have their acidic hydrogenremoved by neutralization with hydroxide. Thecompounds are converted to their negativelycharged ionic form. The negative ions repel eachother, leading to a breakup of large aggregatesof soil to smaller ones. This breakdown is calledpeptizing. The negative ions are also moresoluble in water, leading to their removal fromthe surface. Proteins, composed of amino acids,are very effectively cleaned by alkali. If the alkaliis strong enough, carbohydrates can also beconverted into negative ions leading to theireasier removal.

Alkali will also break down fats and oils totheir component glycerol and free fatty acids.This is the basis for making soap from fat andalkali, and the process is called saponification.Provided that hardness ions are not present toprecipitate them and that the high pH keeps thefatty acids in the ionized form, fats and oils areeffectively removed by alkali.

Most inorganic scales are not effectivelyremoved by alkali.

Chelating agents and sequestrants

Chelation is the binding of ions, mostimportantly the hardness ions calcium andmagnesium, so that the hardness does not causeprecipitation and scale formation. Commonchelating agents are EDTA (ethylene diaminetetraacetic acid) and NTA (nitrilo triacetic acid).Sequestrants, especially the polyphosphates,have a similar action. Sodium gluconate acts as


a chelator under strongly alkaline conditions andis often combined with caustic soda.

A strong chelating agent can also help dissolveinorganic scale that has already formed onsurfaces. Chelators such as EDTA are relativelyexpensive and relatively large amounts areneeded to remove scale where it has acumulatedto significant depths. It is a good idea to measurethe amount of free EDTA remaining in solutionwhen it is being used for scale removal to makesure that not all of the compound has formedcomplexes with hardness ions.

Manufacturers of EDTA do not recommendusing EDTA at high pH such as a mixture withcaustic soda. The reason is because it loses itseffectiveness at chelating magnesium at the highpH, but its chelation of calcium is not affected.Since cleaning difficulties are more associatedwith calcium than with magnesium, this cautioncan often be ignored.

Surfactants

Surfactants concentrate at the surface of water.They have both hydrophobic (water-fearing)and hydrophilic (water-loving) properties. Bothproperties can be satisfied at the same time atthe surface where the hydrophilic portion canremain in the water and the hydrophobic portioncan exit the water. The surface of the water isstabilized by surfactants. This means that watercan more easily spread out and wet larger areasof surface. By increasing the wetting ability ofcleaning solutions, surfactants allow the cleanerto spread out and clean more of the surface. Bya similar mechanism it also helps the cleanerpenetrate small crevices in the soil, increasingits effectiveness.

Surfactants may also congregate at the interfacebetween water and small particles of soil. Theycan form a stabilizing sphere around the particlethat keeps it in suspension and allows it to beefficiently swept away from the surface.Particles stabilized in this way are called micelles.

The oldest surfactant used for cleaning is soap.Soap is free fatty acid. As a rather insoluble acid,it only remains soluble if the pH is high(alkaline), stabilizing the acid in water. Soapsalso form insoluble salts with the hardness ions,calcium and magnesium (bathtub ring) so theywork poorly in hard water.

Newer synthetic detergents such as alkyl (fromoil) benzene sulfonates overcome the problems

of soap. The sulfonic acid portion is more easilyionized, so the pH does not need to be so high.The acid also does not tend to form salts withwater hardness, so these detergents can be usedin hard water. The drawback with these syntheticcompounds, called anionic detergents, is thatthey are so surface active that they form a heavylayer of foam when sprayed onto surfaces inclean-in-place applications. Their use is thereforelimited in industrial cleaning.

Another class of synthetic detergents, non-ionic detergents, does not have a strongtendency to form foam. They therefore can beused in clean-in-place applications. Most ofthese are polyoxyethylene compounds.

Acids

Acids are used in cleaning especially whereremoval of inorganic scales is necessary.Calcium compounds do not dissolve well inalkali, but are often quite soluble in acid. Mostcommonly, phosphoric acid is used or acombination of phosphoric and nitric acids.Sulfamic acid is also occasionally used.

An acid cleaning can follow an initial cleaningwith a caustic soda-based alkaline detergent,which removes the organic components of thesoil. The acid is then used to remove inorganicresidue.

Broader range detergents based on acid withadditional ingredients to remove organic soilshave been used in some applications, but theiruse is not widespread.

Bleach

Oxidizing bleaches are used to partially oxidizecomponents of the soil. Bleach can oxidizesubunits of long chain proteins orpolysaccharides, producing smaller and moresoluble fragments, which can be more easilyremoved. Bleaches are usually added to othercomponents of a formulated detergent toenhance overall cleaning.

Chlorine is the most commonly used bleach.Sodium hypochlorite, produced by addingchlorine to a caustic solution, can be added insmall amounts to alkaline cleaners, such ascaustic soda. Chlorinated trisodium phosphate,a stable solid, is also commonly used.

Typically about 200 ppm of chlorineequivalent is present in the cleaning solution.


Chlorine can be corrosive to stainless steel if leftin contact for long periods of time, so theequipment should be thoroughly rinsedimmediately after cleaning. Chlorinated alkalicleaners are very effective at removing soils withhigh protein content and also beerstone scales.Probably the beerstone is stabilized by organiccompounds between the calcium oxalatecrystals, and the cleaner removes the organicpart very efficiently, making it easier to removethe exposed crystals.

The chlorine in cleaners is diluted to a muchlower concentration than in the original sodiumhypochlorite; and diluted hypochlorite is muchmore stable than concentrated. Diluted cleanershave been used at temperatures as high as 80ºC(176°F).

Hydrogen peroxide is also effective as a‘bleach’ for cleaning. Stabilized peroxides suchas sodium percarbonate and sodium perborateare available in powdered form. These are addedto cleaner or blended with solid cleaningproducts.

SANITIZERS

The use of a sanitizer after effective cleaningcan provide nearly sterile equipment. It shouldbe emphasized that the cleaning must beextremely thorough in order to achieve this levelof sanitation. Cleaning removes most dirt fromsurfaces and a part of this soil will be themicroorganisms left behind after previous use.Most of the removal of microorganisms willoccur during the cleaning step. Sanitizersremove most of the last organisms remainingbehind.

Chlorine-releasing materials

The most commonly used ‘chlorine’ solution isactually a solution of sodium hypochlorite.Adding chlorine gas to a caustic solutionproduces sodium hypochlorite, and under theright conditions the chlorine is slowly releasedto do its sanitizing job. The sodium hypochloritecan exist in two forms, hypochlorite ion andhypochlorous acid. It is the hypochlorous acidthat attacks microorganisms and provides thesanitizing effect. When the sodium hypochloriteis mixed with water, the pH of the solutiondecreases and hypochlorous acid begins to form.

The lower the pH, the more hypochlorous acidis formed and the more effective sanitation willbe. Unfortunately, if the pH drops too low, freechlorine gas begins to be released, leading to avery dangerous situation. Therefore, for effectiveuse, the pH should be below 8.2 but not lowerthan 7.0. There is a very narrow range wherehypochlorite should be used.

Dual halogen sanitizers have been developedthat extend the effective range of chlorine-basedsanitizers. A source of bromine is added tohypochlorite, allowing hypobromite to beformed. The dual halogens are effective at aboutone pH unit higher than plain hypochlorite.

It may seem paradoxical that hypochlorite asa bleach can be used in high pH alkaline cleaningformulations, but that it should be used at nearlyneutral pH as a sanitizer. The bleaching actionis indiscriminate, while the antimicrobial actionis more specific, appearing to depend on theability of the non-ionized form to penetrate andexert killing action inside living cells.

Chlorine in the form of hypochlorite is the leastexpensive sanitizer available. It is effectiveagainst a wide range of microorganisms,including spores. Its action is not diminished inhard water. If residual soil is present, thehypochlorite will bleach or oxidize the soil andbe lost. It is important that equipment bethoroughly cleaned before a chlorine sanitizeris used or it will be inactivated before it can exertits sanitizing effect. Chlorine is also corrosiveto many metals. It may be used on stainless steelat low concentrations, but it must be thoroughlyremoved by rinsing after use.

Chlorine dioxide

Chlorine dioxide (ClO2), is a relatively newsanitizer based on chlorine. Chlorine dioxide willkill organisms more quickly and at lowerconcentrations than hypochlorite. Typical userates are 2 to 10 ppm, measured as availablechlorine, versus 25 to 200 ppm for hypochlorite.When residual soil is present, chlorine dioxidereacts more slowly than does hypochlorite sothe need for thorough cleaning before use isreduced. It also can be used over a wide pH rangesince it dissolves as a gas, not forming an acidor salt in water. It is less corrosive thanhypochlorite. The biggest disadvantage ofchlorine dioxide is that it is relatively expensive.


It is a gas and will not remain in water solutionfor long periods of time; it must be generatedand used on the same day. Many methods ofgeneration are available. Acid conversion ofchlorite is used on a small scale, but is inefficient.Larger scale generators may use dangerouschemicals such as chlorine gas or hydrochloricacid. Electrolytic generators are a newerdevelopment that promise greater efficiency andsafety.

Iodine-releasing materials

Iodine is a very efficient sanitizer when used ata concentration of about 25 ppm at coldertemperatures, and at an acid pH. Iodine tincturesin water are not very stable and actively stainmaterials with which they come into contact.Industrial use of iodine is usually in the form ofa complex with detergents and acid called aniodophor. When concentrated iodophor is dilutedin water, the iodine is freed from its complexwith detergent. The acid in the iodophor reducesthe pH of the diluted solution to the acid rangewhere iodine is most effective. The acid can alsohelp dissolve mineral scale left behind afteralkaline cleaning. Because of the detergents inthe iodophor, the sanitizer tends to foam whenused in spray applications. The recentdevelopment of iodine stabilized with chlorine,‘stabilized iodine’, has eliminated foaming as aproblem and requires about half as much iodineto do the same sanitizing job. Since stabilizediodine is relatively new, it has not been legallycleared for use as a food equipment sanitizer inall areas.

Iodine must be used at an acid pH to be mosteffective. If an iodophor is used at higher thanrecommended dilution, the pH of the mixedsanitizer should be checked to make sure thatthe pH of the diluted solution is not too high.Since iodine tends to stain, be certain that stainingcan be tolerated before using an iodine basedcompound. There is little tendency to corrodemetal equipment unless the metal is not resistantto the acid contained in iodophors.

Anionic surfactants

The anionic surfactants are not very commonlyused as sanitizers, but they do have propertiesthat make them the choice for some specialapplications. They are surface-active compounds

and are active at low pH, so they must becombined with acid. While hypochlorites,iodophores and chlorine dioxide all dissipate orevaporate in a relatively short period of time,anionics are completely stable. They will last forweeks or months and are therefore the bestchoice for applications where equipment is tobe stored in water or filled with water for longerthan a few days. The practice of storingequipment soaking in plain water is extremelyunsanitary and is a very common but badpractice. Bacteria will grow in the water,especially if some residue is left on theequipment. Slimy layers that often form on theequipment and storage containers are layers ofbillions of bacteria or a biofilm. Store theequipment with mild acid and an anionicsurfactant and it will remain clean for months.

Being detergents, some of the anionicsurfactants are also fairly good cleaningcompounds and acid cleaning can also beaccomplished with some of these compounds.

Peracetic acid

Peracetic acid is a complex of acetic acid withhydrogen peroxide. The peroxide is the sanitizer.It is effective at about 200 ppm of peroxide andat colder temperatures. It is expensive and is usedmainly in special situations where residue ofchloride or iodide from less expensive sanitizersor some of their reaction products cannot betolerated, especially some food applications.

Quarternary ammonium compounds

Quarternary ammonium compounds, called‘quats’, are stable, effective and long lasting,tending to leave a residue. They adsorb tosurfaces and have a residual killing effect forrelatively long periods of time. Usually theresidue is not desirable in vessels such asfermenting tanks since the residue will also killsome yeast. If a lot of yeast are present, thenumber killed may not be too significant, soquats are occasionally used to sanitize thesevessels. A better application is to use quats onareas where yeast should not be present. Theyare used on the outside of fermenting tanks andon floors where spills are likely to occur to keepdown the growth of undesirable organisms suchas molds on these surfaces.


Hot water

Hot water, over 180ºF (82°C), is a very effectivesanitizer because the heat can be transferred toareas that are not penetrated by chemicalsanitizers. Cost of fuel to generate hot water is afactor that limits its use as a sanitizer.

Hot caustic

Caustic (NaOH) by itself is not a sanitizer.Bacteria are often found in high numbers incaustic solutions stored at room temperature.Hot caustic, though, will very effectively killmicroorganisms. The temperature must be atleast 110ºF and temperatures above 130ºF givegood sanitizing in a matter of a few minutes.The exact effect depends on the time,temperature and caustic concentration used.Many areas have legal requirements for washingand sanitizing reusable containers with causticsolutions.

Turbulence

Turbulence refers to the scrubbing and scouringaction of cleaning solutions as they flow throughpipelines or onto the dirty surfaces inside tanksand other equipment. More turbulence improvescleaning. Sometimes the issue is more basic thanthe degree of turbulence: does the cleaningsolution even contact the soil? Large diameterhorizontal pipes will not completely fill withliquid if the flow rate is too low. ‘Shadow areas’can also prevent cleaning solution fromcontacting the soil.

For pipelines, effective cleaning requiresvelocities of 1-1.5 to 3 m/s (5-10 ft/s). We aremore used to thinking in terms of flow rates likecubic meters per hour or gallons per minute. Thisrequirement means that bigger pipes need higherflow rates for cleaning. Table 3 shows the flowrates required for 5 ft/s (1.5 m/s) in pipes withdiameters from 75 mm to 300 mm. It isimportant to note that when the pipe diameter isdoubled, the required flow rate for cleaning mustbe four times as large. This has a large effect onthe size of pumps needed for CIP.

Table 3. Turbulence in pipelines: flow rates required foreffective cleaning as pipe diameter increases1.

Pipe size Flow rate for 5 ft/sec (1.5 m/sec)

Inches Millimeters US gpm Liter/min

3 75 102 3864 100 182 6896 150 410 15528 200 728 2755

12 300 1620 6130

1Anon (1992).

A rule of thumb for cleaning process equipment(e.g., a heat exchanger) is that the cleaningsolution flow rate should be 1.5 times the normalprocess flow rate to achieve good turbulence.An effective trick is to clean the equipment inboth the forward-flow and the reverse-flowdirections (Figure 5).

SANITARY DESIGN

One of the greatest impediments to achievingturbulence in cleaning is the presence of deadends in process pipelines and other equipment(Peter, 1983). A dead end can be a length ofpipe that is wetted by the cleaning solution butwithout sufficient velocity to clean the surfaces.A common example is the bypass line, shownin Figure 6. When the heat exchanger is cleaned,solutions cannot flow through the dirty bypassunless someone opens the bypass valve. Thismay or may not happen. It should also be notedthat when the bypass valve is open, the totalsolution flow is split in two directions: throughthe equipment and through the bypass line. Theflow rate through each is therefore reduced andthe velocity of the cleaning solution iscompromised.

Figure 7 shows some design tricks that caneliminate the creation of dead ends before theyare built into the system. For example, branchlines from the main line should have the valveinstalled not more than one diameter from thewall of the main pipeline. This ensuresturbulence in the branch – that it will not be adead end for cleaning. Tees in the main lineshould be manufactured tees made with smoothrounded contours that are sanitary and cleanable.


Fitted joints, that is homemade or fabricated-in-the-field tees, are not sanitary and should not beaccepted.

Figure 8 shows an innocent looking tee with avalve on the branch line. This section of pipewas opened for inspection and the branch pipewas packed with soil. The soil buildup was notremoved during CIP because the cleaningvelocity was insufficient or because the dead endwas more than one pipe diameter.

During construction, piping should be weldedusing sanitary welding procedures including (forstainless steel) butt welding with an inert gaspurge. Pipelines should be self-draining andwithout high spots or domes where air can

Plateheat

exchanger

Rule of thumb:

1.5 times the normal process flow rate

Figure 5. Turbulence in process equipment.

Plate heat exchanger

Figure 6. Dead ends and turbulence: did the bypass get cleaned?

become trapped and prevent cleaning solutionsfrom touching those parts of the pipe.

For processing equipment the American MeatInstitute created a task force to improveequipment design with the goal of eliminatingharborage areas where microorganisms cancollect and grow. The result is a set of guidelines,the Ten Principles of Sanitary Design (Stout,2003). These principles, for food processingequipment, are as follows:

1. Cleanable to microbiological level. A pieceof equipment must be more than just visuallycleanable – over the entire life of theequipment.


Figure 7. Eliminating dead ends with sanitary design.

4" x 4" x 2" Sanitary Tee

4" 2"

45° Elbow

UNSANITARY TEE ARRANGEMENT

CORRECT SANITARY TEEARRANGEMENT

Critical dimension: Less than

one diameter of the branch

Fitted joint

Figure 8. Buildup of soil in the dead end branch of a tee.


2. Made of compatible materials. Constructionmaterials must be compatible with theproduct, the environment and with thecleaning materials.

3. Accessible for inspection, maintenance andcleaning. A person without tools must be ableto inspect and clean the equipment.

4. No product or liquid collection. Theequipment must be self draining with noareas where water or product canaccumulate.

5. Hollow areas hermetically sealed. Hollowareas must be eliminated or hermeticallysealed by continuous seam welding toeliminate any area where soil couldaccumulate.

6. No niches. There may be no harboragepoints: cracks, crevices, pits, dead ends, orniches. This encompasses sanitary weldingin process pipelines.

7. Sanitary operational performance. Theequipment must perform so that it does notharbor bacteria or contribute to micro-biological contamination.

8. Hygienic design of maintenance enclosures.Control panels, junction boxes, belt guards,gear enclosures etc. are designed to ensurethat neither water nor product accumulate inor on the enclosure.

9. Hygienic compatibility with other plantsystems. Any new piece of equipment mustbe compatible with steam, air, water and otherutilities and systems in the plant.

10. Validate cleaning and sanitizing protocols.The equipment manufacturer and equipmentpurchaser must work together during theequipment design stage to design cleanabilityinto the machine before it is built.

THE CLEANING LOOP

A cleaning loop is the path that the cleaningsolution flows during a CIP cycle. It begins atthe CIP system and ends at the CIP system(Figure 9). Designing and planning cleaningloops is essential to achieving effective cleaningalong the entire path that product followsthrough the plant.

CIP

System

Process

Figure 9. The cleaning loop.

Sanitary design is not just about individual tanksor pieces of equipment, it is about processingsystems. Cleaning a fermentor, for example,encompasses much more than the tank itself. Itincludes the external heat exchanger, the productinlet and product outlet pipes, CO2 collectionlines, additive addition lines, pressure reliefvalves, and more. “How do we clean it?” and“what cleaning loops are necessary?” are keyquestions that must be asked early in the designstage.

Completely cleaning a fermentor systemrequires not one, but several cleaning loops(Figures 10a-10d). Some of these loops and theirseparate requirements include:

• Mash filling pipelines (minimum 3 m/s CIPvelocity).

• Fermentor CIP sprayhead (pressurerequirement).

• External heat exchanger loop (1.5 timesnormal process flow).

• Fermented product emptying pipeline(minimum 3 m/s CIP velocity).

The loops must be planned such that there areno lengths of pipe that lie between two loopsand are therefore not included in either loop.Ideally, loops will overlap.

Validation of cleaning

When the cleaning process is complete we mustknow if the equipment got cleaned properlybefore it is committed to a new batch of product.If the final product is infected, we know thatcleaning was not effective but it is then too late.


Coolant

MashSystem

Concrete pad

Return to CIP

CIP supply

Coolant

CIPSystem

ProductReceiver

Coolant

MashSystem

Return to CIP

CIP supply

Coolant

CIPSystem

ProductReceiver Concrete pad

Coolant

MashSystem

Return to CIP

CIP supply

Coolant

CIPSystem

ProductReceiver Concrete pad

Coolant

MashSystem

Concrete pad

Return to CIP

CIP supply

Coolant

CIPSystem

ProductReceiver

Figure 10a. The mash filling pipeline loop. Figure 10b. The fermentor CIP sprayhead loop.

Figure 10c. The external heat exchanger loop. Figure 10d. The fermented product loop.


Temperature, pressure and flow records canbe used to document that the CIP programmeran successfully. If automation does not providethese reports, the parameters can be checked andrecorded by the operators. These records areoften part of a formal HACCP program.

Visual inspection may be possible frommanways, inspection ports or windows. Thisimmediately provides an indication of grossdeficiencies in the cleaning process andappropriate measures can be taken before thenext batch is started.

Microbiological swabs and plating, along withATP bioluminescence swabs, are more specificindicators of successful cleaning and sanitizing.An ATP bioluminescence swab indicates theamount of soil left on a surface that may look‘clean’ (Ehrenfeld et al., 1996). A big advantageis that the result is known in less than oneminute. Again, appropriate action (repeat the CIPprocess) can be taken. Microbiological swabsand plating can tell us how much of whatorganisms are present, but results are not knownfor 2-5 days. Newer rapid methods areshortening this time.

Examples of CIP cycles

A TYPICAL CIP CYCLE FOR TANKS ANDPIPELINES

A. Prerinse to the sewer. Removes grossamounts of soil before the main wash. It iscommon to save the water rinse (item C) andreuse it for this step. This practice conserveswater, and the rinse water usually has a smallamount of detergent in it that enhances theeffectiveness of the prerinse.

B. Alkaline (caustic) wash. For organic and oilysoils, caustic soda is most commonly used.It often has additives to improve wetting andrinsing properties. Strengths from 1 to 3 %by weight are common and it is used attemperatures from ambient up to 180°F. It isa more effective cleaner at the warmertemperatures.

C. Water rinse. Rinses out the washing water. Itis often saved for reuse as the prerinse. Thesource should be fresh, clean water so as notto recontaminate.

D. Acid wash and water rinse (optional). Thisadditional cleaning is done to prevent scaleformation, for example in beer kegs, or toremove scale deposits.

E. Sanitizing rinse. A sanitizing agent may beinjected into the clean rinse water to reducemicroorganisms on the clean surfaces to anacceptable level.

CLEANING PROCEDURE FOR CO2 SCRUBBER

A. Stop flow of CO2 through the scrubber andremove CO2 from inside the scrubber byventing or rinsing with fresh water. ResidualCO2 in the scrubber will neutralize the causticsolution.

B. Make a caustic soda solution 2 to 3% byweight at ambient temperature or preferablywarmed to 100 to 120°F. Make enough tomaintain a level in the bottom of the scrubberwhen recirculating.

C. Circulate the caustic solution through thescrubber sprayer for at least 1 hr. Severalhours may be required depending on the soilload.

D. After 15 to 30 minutes take a sample of thecaustic and titrate to make sure it has notbeen neutralized by residual CO2 in thescrubber. If less than 2% strength, add morecaustic to get 2 to 3%.

E. Drain caustic, then rinse using burst rinse(water for 30 sec, drain for 3 min.). Repeatburst rinse cycle until rinse water shows nocaustic when tested with phenylphthalein orlitmus paper.

F. Prepare acid wash, 1 to 2%, using ScalebiteTM

at ambient temperature. Circulate for 1 hr.

G. Drain acid solution and rinse using burstrinse, three cycles.

H. Prepare IotechTM solution, 25 ppm at ambienttemperature, and recirculate for 30 mins.

I. Drain IotechTM solution and rinse using burstrinse, three cycles.

J. Scrubber is ready to bring back on stream.


PROCEDURES FOR DE-SCALING FERMENTORS

The following is an approach to descaling tanksand equipment after beerstone or scale hasaccumulated over a period of time. Since scaleamounts and compositions vary from time totime and place to place, some trial and error isusually required.

A. Prerinse the tank to remove heavy soil. Usefresh water or water with a little caustic, likethe saved final rinse water from washinganother tank. Do not save this liquid in theCIP system.

B. Main caustic wash. Caustic strength at 3 to5%, temperature 120 to 190°F. This will besimilar to the normal caustic wash and whenfinished, the tank should be clean except forthe scale or beerstone. Addition of sodiumhypochlorite (bleach) to caustic solutionsmay help to attack the scale. It is used atconcentrations of 200 to 400 ppm as activechlorine. CAUTION – never mixhypochlorite (bleach) with acid cleaners.

C. Scale-removing caustic wash. Use caustic asin step 2 but with addition of ScalebanTM asa supplement. For infrequent de-scaling, add5 oz ScalebanTM for every gallon of thecaustic solution. Same temperatures, 120 to190°F. Warmer is usually better. Circulateseveral hours if possible. Use kit to titrateEDTA, the active ingredient in ScalebanTM.Add more ScalebanTM if EDTA gets depleted.For routine scale control, 1 to 3 oz ofScalebanTM is added to each gallon of thenormal caustic solution.

D. Rinse with clean water. Inspect visually. Note:when tank is wet you may not see the scale.When it dries, the scale is more visible.

E. Scale-removing acid wash. If scale remainsafter the ScalebanTM wash, an acid cleanermay be required. ScalebiteTM is a formulatedacid cleaner that is used between 1 and 3%in water. It is recirculated several hours. Note:ScalebiteTM is safe with stainless steel but doesattack mild steel, especially at warmertemperatures. Sulfamic acid is also effectiveagainst some scales. Like ScalebiteTM, it isused in concentrations between 1 and 3 %,temperatures up to 140°F.

Problem areas

When a contamination problem develops, oneof the first questions asked is ‘How did theinfection develop?’ One of the most difficultchallenges in industrial sanitation is in keepinga continuous fermentation system free ofinfection. Doug Banks from New Zealand’sDominion Breweries, which has successfullyoperated breweries with continuous fermentationof beer and no infection for many years,classifies sources of contamination into twogroups, point sources and dead ends, based onyears of experience:

Point sources are sources that occur only fromtime to time. They may appear and thendisappear pretty much at random. Some pointsources include:

• Bad pump seals that allow contamination tobe drawn in from the outside environmentby a Venturi effect of the flowing liquid.

• A piece of equipment that was missed duringregularly scheduled cleaning.

• An employee with dirty boots, etc. bringscontamination into the system by climbinginto a clean tank or some other area to makesome final adjustment or repair.

• A cleaning sprayball has become plugged.

• Sample ports have been forgotten duringcleaning and have become infected.

• A solution used for sanitizing small piecesof equipment out of place has stood for days,losing its effectiveness.

• The cleaning solution used was too dilute.

• The wind blows contaminated air or anaerosol into equipment when it istemporarily opened after cleaning and beforeuse.

• Insects carry contamination into a piece ofequipment temporarily left open.

Dead ends are sources of contamination that arealways in contact with the process stream.Usually these are the most serious causes ofcontamination. Some examples are:


• Dead ends in piping that cannot bethoroughly cleaned.

• Scale built up inside fermenting tanks orcoolers. Scale cannot be cleaned because thecleaning compounds cannot penetrate thescale once it is thick enough.

• Carbon dioxide vent lines on fermentors intowhich fermenting liquid has flowed or beensprayed as an aerosol.

• Faulty check valves

• Contaminated yeast

• Poor drainage of cleaning solution from atank while it is being cleaned.

• Seats on valves that have worn and begunto develop cracks

• Non-sanitary pipe connections, especiallypipe threads. These will trap liquid in thethreads that cannot be removed by cleaning.

Some of the likely ultimate sources ofcontamination include:

• Grains being brought into a plant carry largenumbers of a variety of microorganisms

• People and the food they bring into the plant

• Insects

The most likely sources of contamination,though, are within the plant itself. Infectionsources are often found in scale, dead ends, evenin coolers, spilled material left on the ground,material in sewers that can be splashed into theair as an aerosol, dirty and contaminated carbondioxide lines and aerosols created when openingcontaminated equipment such as perhaps thebeer well.

Hidden issues related to chemistry occur whencleaning tanks that have a CO2 atmosphere. TheCO2 will react with caustic to form sodiumcarbonate, a cleaner that is much weaker thancaustic soda (Figure 11). However, if thisreacted solution is analyzed by the normal singleend-point titration, the effect is not detected.Carbonate titrates just like the hydroxide. Thisproblem can be overcome by replenishing thereacted caustic or by venting the vesselcompletely before cleaning. The correct titrationuses barium chloride first, then acid. The method

can be found in Alltech’s Laboratory Manual(Alltech, 2001).

2 NaOH + CO2 Na2CO3 + H2O

Sodium hydroxide

(an excellent cleaner)

Sodium carbonate

(a fairly good cleaner)

Figure 11. Dilution of caustic by carbon dioxide.

A CASE STUDY: CLEANING A FERMENTOR

Cleaning the inside of a fermentor with asprayhead and cleaning the tank outlet (Figure12)

The normal flow through a sprayhead is around100 gallons/min (about 380 liters/min). For a 6-inch fermentor outlet pipe, the minimum velocityfor cleaning is not reached until the flow rate is410 gallons/min. This is a serious cleaningdeficiency and it is common to see grossaccumulations of soil on the top half of theseoutlet pipes because they do not even fill withcleaning solution when the sprayhead is beingused.

The cleaning and sanitizing program

A routine and effective microbiological samplingprogram should be in operation. Samplingshould be planned so as to provide a complete,overall picture of the state of sanitation in theplant at any time. Special attention should bepaid to an evaluation of the cleaning proceduresin place. Infections usually develop in areaswhere cleaning is not effective or is missed.Swabs should be taken from cleaned equipment.An ATP detecting luminometer can provide moreimmediate evaluation on the effectiveness ofcleaning. Strengths of cleaning and sanitizingsolutions should be measured routinely, as wellas temperatures of cleaning solutions.

Sanitation Quality Control is like preventivemaintenance for the product. It is troubleshootingdesigned to identify potential problems so thatthey can be corrected before damage is done tothe product. There are four basic componentsin an effective sanitation quality controlprogram:


Inspection light

Vent

From CIPMash inYeast/enzymes in

20 in. Manway/sampling port

Coolant

Slope

Concrete pad

Alternative cooling methods:A. External cooling jacketB. External heat exchanger with recirculationC. Internal cooling coilsD. Internal cooling panels

CIP spray head100 psi100 gpm

A

15 in. x 20 in.access Manway

Recirculation pump

Returnto CIP

5 - 10 ft/sec.

B

CAgitator(if required)

D

Coolant

Figure 12. Cleaning a fermentor.

A. The Sanitation Plan

1. Master sanitation schedule

2. Times, temperatures, materials, strengths forevery cleaning job

3. Detailed written instructions for everycleaning job

B. Measurements and Standards

Every cleaning and sanitizing process shouldhave, as an objective, the achievement of an endresult that can be measured and compared to apredetermined standard.

1. Visual inspection - gives a ‘go/no go’ resultimmediately. Wet surfaces can appear cleanwhen they are actually dirty, so the visualinspection is best done when the surface isdry. This should also include ‘white glove’checks.

2. ATP-detecting luminometer - immediatequantitative results for presence of ATP,which indicates cleaning was not complete.

3. Microbiological - a time of 24 hrs to 1 weekis required before results are known. Testsinclude forcing tests, swabs, pour plates andmillipore filtering. Different tests are


specified at different stages of the processand each test has standard results or controllimit.

C. Records

These should include: when was the cleaningreally done (vs schedule), automatic or manualrecording of temperatures and strengths, whenwas maintenance done (UV lights renewed, eg),problems encountered when the job was done.

D. Results

Immediate reporting of microbiological resultsand charting trends for statistical quality controlare of key importance.

Summary

For every cleaning and sanitizing problem thereis a solution, in fact there can be many solutions.Unfortunately the solutions are not available asdefined recipes in a cookbook. The Four Ts, ourcleaning and sanitizing bag of tricks, experienceand trial and error are the tools at our disposal.We know that the program is to first remove thedirt, and then sanitize. Sanitizing dirt may workfor a short time but it is not a permanent solution.We must also determine the acceptable levels ofcleanliness and sanitation and then avoidexcesses that are hard on the equipment andcostly in terms of time, money and utilities.

We must use our knowledge of the Four Ts ofcleaning and sanitizing to find a combination oftime, temperature, turbulence and titration thatis both effective and economical. Finally, wemust realize that while CIP is the ideal, COPcombined with CIP is usually the reality.

“Regardless of technology, the mostimportant food plant sanitation approachis getting back to basics, which meanshaving motivated, passionate people whoknow both the importance of and how toclean effectively. It comes down to trainingand motivation; simply put, how to usemechanical action, with the rightdetergent, at the right concentration forthe right time at the right temperature withthe right attitude, and giving people creditfor the good work they do.”

(Stout, 2002).

Cleaning and sanitation is an area where thequality control principle of ‘continuousimprovement’ can be effectively applied. Noplant is perfectly clean. There is always roomfor improvement, and small improvements canaccumulate into major improvements in theoverall efficiency and profitability of the plant.

References

Alltech Inc. 2001. Laboratory Methods inBrewing and Distilling. Alltech TechnicalPublications, pg 3-16.

Anon. 1992. 3-A Accepted Practices ForPermanently Installed Product And SolutionPipelines And Cleaning Systems Used In Milkand Milk Product Processing Plants, Number605-04. Dairy, Food and EnvironmentalSanitation, Volume 12, No. 2.

Czechowski, M.H. and M. Banner. 1992. Controlof biofilms in breweries through cleaning andsanitizing.Tech. Q. Master Brewers Assoc. ofAm. 29(3):86-88.

Ehrenfeld, E.E., J. Scheld, S.A. Miller and C.R.Carpenter. 1996. Use of ATP-bioluminescencein the brewing industry. Tech. Q. MasterBrewers Assoc. Am. 33(1):59-62.

Kretsch, J. 1994. Practical considerations forbrewery sanitation. Tech. Q. Master BrewersAssoc. Am. 31(4):124-128.

McCabe, J.T. 1999. The Practical Brewer, 3rdEdition. Master Brewers Association of theAmericas.

Peter, W.A. 1983. Designing for automation.MBAA Technical Quarterly 20(1):22.

Stout, J. 2002. Perspective on practices in foodplant sanitation and hygiene. Food SafetyMagazine, April/May pg.20.

Stout, J. 2003. 10 Principles of equipment designfor ready to eat food processing operations.Food Safety Magazine, June/July, pg. 18.

Recovery

Ethanol distillation: the fundamentals 319

Chapter 22

Ethanol distillation: the fundamentals

P.W. MADSON

KATZEN International, Inc., Cincinnati, Ohio, USA

Fundamentals of a distilling system

Certain fundamental principles are common toall distilling systems. Modern distillation systemsare multi-stage, continuous, countercurrent,vapor-liquid contacting systems that operatewithin the physical laws that state that differentmaterials boil at different temperatures.

Represented in Figure 1 is a typical distillationtower that could be employed to separate an idealmixture. Such a system would contain thefollowing elements:

a. a feed composed of the two components tobe separated,

b. a source of energy to drive the process (inmost cases, this energy source is steam, eitherdirectly entering the base of the tower ortransferring its energy to the tower contentsthrough an indirect heat exchanger called areboiler),

c. an overhead, purified product consistingprimarily of the feed component with thelower boiling point,

d. a bottoms product containing the componentof the feed possessing the higher boilingpoint,

e. an overhead heat exchanger (condenser),normally water-cooled, to condense the vapor

resulting from the boiling created by theenergy input. The overhead vapor, aftercondensation, is split into two streams. Onestream is the overhead product; the other isthe reflux which is returned to the top of thetower to supply the liquid downflow requiredin the upper portion of the tower.

The portion of the tower above the feed entrypoint is defined as the ‘rectifying section’ of thetower. The part of the tower below the feed entrypoint is referred to as the ‘stripping section’ ofthe tower.

The system shown in Figure 1 is typical forthe separation of a two component feedconsisting of ideal, or nearly ideal, componentsinto a relatively pure overhead productcontaining the lower boiling component and abottoms product containing primarily the higherboiling component of the original feed.

If energy was cheap and the ethanol-watersystem was ideal, then this rather simpledistillation system would suffice for theseparation of the beer feed into a relatively pureethanol overhead product and a bottoms productof stillage, cleanly stripped of its ethanol content.Unfortunately, the ethanol-water (beer) mixtureis not an ideal system. The balance of this chapter

320 P.W. Madson

will be devoted to a description of themodifications required of the simple distillationsystem in order to make it effective for theseparation of a very pure ethanol product,essentially free of its water content.

Figure 2 expands on Figure 1 by showing someadditional features of a distillation tower. Theseare:

a. The highest temperature in the tower willoccur at the base.

b. The temperature in the tower will regularlyand progressively decrease from the bottomto the top of the tower.

c. The tower will have a number of similar,individual, internal components referred toas ‘trays’ (these may also be described asstages or contactors).

d. Vapor will rise up the tower and liquid willflow down the tower. The purpose of thetower internals (trays) is to allow intimatecontact between rising vapors anddescending liquids correlated with separationof vapor and liquid.

Figure 3 shows a vapor-liquid equilibriumdiagram for the ethanol-water system atatmospheric pressure. The diagram shows molepercent ethanol in the liquid (X axis) vs molepercent ethanol in the vapor (Y axis). The plotcould also be made for volume percent in theliquid vs volume percent in the vapor and theequilibrium curve would only be slightlydisplaced from that shown in Figure 3. Molepercent is generally used by engineers to analyzevapor/liquid separation systems because it

Vapor

Reflux

Cooling water

Low boiling product

Feed mixture

Steam (energy)

Strip

pin

gR

ectif

yin

g

High boiling product

Condenser

Figure 1. Ideal distillation system.


relates directly to molecular interactions, whichmore closely describe the process occurring ina distillation system.

Analysis of the ethanol-water distillationsystem is mathematically straightforward whenusing molar quantities rather than the morecommon measurements of volume or weight.This is because of an energy balance principlecalled ‘constant molal overflow’. Essentially, thisprinciple states that the heat (energy) requiredto vaporize or condense a mole of ethanol isapproximately equal to the heat (energy)required to vaporize or condense a mole ofwater; and is approximately equal to the heat(energy) required to vaporize or condense anymixture of the two. This relationship allows thetower to be analyzed by graphic techniquesusing straight lines. If constant molal overflowdid not occur, then the tower analysis would

become quite complex and would not lend itselfeasily to graphic analysis.

Referring to Figure 3, a 45o line is drawn fromthe compositions of 0, 0 to 100% and 100%.This 45o line is useful for determining ranges ofcompositions that can be separated bydistillation. Since the 45o line represents thepotential points at which the concentration in thevapor equals the concentration in the liquid, itindicates those conditions under whichdistillation is impossible for performing theseparation. If the equilibrium curve contacts the45o line, an infinitely large distillation towerwould be required to distill to that compositionof vapor and liquid. Further, if the equilibriumcurve crosses the 45o line, the mixture hasformed an azeotrope. This means that even ifthe tower were infinitely large with an infiniteamount of energy, it would be impossible todistill past that point by simple rectification.

Condenser Cooling water

Overhead product

Reflux liquid

Vapor

Feed

Hig

her

tem

perature

Low

er

tem

perature

Thermal energy

Vapor

Boiling liquid

Trays (contacting devices)

Bottoms product

Figure 2. Typical distillation relationships.

322 P.W. Madson

Consider a very simple system consisting of apot filled with a mixture of ethanol in water (abeer) containing 10% by volume ethanol (3.3mole %). This composition is identified in thelower left portion of Figure 3. A fire could bekindled under the pot, which would add thermalenergy to the system. The pot would begin toboil and generate some vapor. If we gathered asmall portion of the vapor initially generated andmeasured its ethanol content; we would findabout 24 mole % ethanol (53 volume %). If wecondense this vapor (note: there will be only asmall amount of this vapor), boil it in a secondpot and again collect a small amount of the initialvapor generated, this second vapor wouldcontain about 55 mole % (83 volume %) ofethanol (see Figure 3). If we should continuethis simplified process to a third and fourth

collection of small amounts of vapor, analysiswould reveal that each successive portion ofvapor would become richer in ethanol.

Thus we have created a series of steps bywhich we keep increasing the ethanol contentof the analyzed sample, both liquid and vapor.Unfortunately, this oversimplified process isidealized; and practically speaking, isimpossible. However if we had supplied ouroriginal pot with a continuous supply of ethanol-water feed and vapor generated in the first potwas continuously condensed and supplied to thesecond pot, etc. then the process becomes similarto the industrial distillation tower operationshown in Figure 2.

How far can this process be extended? Couldwe produce pure ethanol by continuouslyextending our process of boiling and reboiling?

100

0

Azeotrope

194° proof

(10 Volume %)

10 20 30 40 50 60 70 80 90 100

90

80

70

60

50

40

30

20

10

0

Mole % ethanol in liquid

Mole

%

ethanol in

vapor

Equilibrium Curve

Figure 3. Vapor/liquid equilibrium for the ethanol/water system at atmospheric pressure.


The answer is, no! We would finally reach apoint in one of the downstream pots, where thevapor boiling from the liquid was of the samecomposition as the liquid from which it wasbeing generated. This unfortunate consequencelimits our ability to produce anydrous ethanolfrom a dilute ethanol-water feed. What we finallyencounter in our simplified process is theformation of an azeotrope. This is a concentratedsolution of ethanol and water that when boiledproduces a vapor with a composition identicalto the composition of the liquid solution fromwhich it originated.

In summary then, we are limited in ethanol-water purification in any single multistagedistillation tower to the production of azeotropicethanol-water mixtures. These azeotropic

solutions of ethanol and water are also knownas constant boiling mixures (CBM) since theazeotropic liquid will have the same temperatureas the azeotropic equilibrium vapor being boiledfrom itself. Without some sort of drastic processintervention, further ethanol purificationbecomes impossible. The question thenbecomes: What can we do to make it possible toproduce anhydrous ethanol? Methods of doingso will be covered later in this chapter.

Figure 4 depicts the structure of the distillationprocess by dividing the vapor/liquid equilibriuminformation into three distinct zones of processand equipment requirements: stripping, rectifyingand dehydration. This division is the basis forthe design of equipment and systems to performthe distillation tasks.

0

0 10 20 30 40 50 60 70 80 90 100

10

20

30

40

50

60

70

80

90

100


Azeotrope 194° proof

(10 volume %)

Mole

%

ethanol in

vapor

Equilibrium curve

Rectifying system

Stripping system

Stilla

ge

Spir

it

Anhydrous

product

Dehydration

system

Figure 4. Structuring the distillation system strategy.

324 P.W. Madson

Considerations in preliminary design

The engineer, given the assignment of designinga distillation tower, is faced with a number offundamental considerations. These include:

a. What sort of contacting devices should beemployed? (e.g. trays or packing). If traysare chosen, what type will give the mostintimate contact of vapor and liquid?

b. How much vapor is needed? How muchliquid reflux is required? (What ratio of liquid:vapor is required?)

c. How much steam (energy) will be required?

d. What are the general dimensions of thedistillation tower?

DISTILLATION CONTACTORS

Trays are the most common contactor in use.What are the functions expected of traycontactors in the tower? Figure 5 depicts a singletray contactor in a distillation tower and showsthe primary functions desired:

- mixing rising vapor with a falling fluid

- allow for separation after mixing

- provide path for liquid to proceed down thetower

- provide path for vapor to proceed up thetower

Enriched

vaporLiquid

Stripped

liquid

Vapor

V/L mixing

V/L separation

Figure 5. Distillation tray functions.

Figure 6 depicts a perforated tray contactor withcertain accoutrements required to control theflow of liquid and vapor and to assure theirintimate contact. Another type of tray contactingdevice, the disc-and-donut or baffle tray isshown in Figure 7. The characteristics of thistype of contactor make it especially useful fordistilling materials such as dry-milled grain beer,which would foul ordinary trays such asperforated, valve, etc.

ENERGY ANALYSIS

In addition to the selection of the basic contactingdevice, the energy requirement must be

Inlet weir

Outlet weir

Downcomer

area

Downcoming

liquid

Stripped

liquid

Enriched

vapor

Vapor

V/L mixing

V/L separation

Perforated

area

Figure 6. Perforated trays.


established. This is accomplished by analyzingthe vapor/liquid equilibrium data from Figure4, for the liquid:vapor flow ratio to perform acontinuous series of steps within the limits ofthe equilibrium curve. Table 1 demonstrates a

Disc

Donuts

Descending

liquid

Stripped stillage

Vapor

Enriched

vapor

Descending

liquid

simplified procedure to calculate theapproximate energy requirement from theliquid:vapor flow ratio that will be employed inthe tower design. Repetition of this type ofcalculation for different conditions produces a

Figure 7. Disc-and-donut trays.

Table 1. Simplified calculations for steam requirements for ethanol distillation.

Example 1. Calculate the steam required (lbs/gallon of product) to strip 100 gpm of a 10% volume beer (90 gpm water/10 gpm ethanol).

L/V* = 5.0 (typical for a 10% volume beer) or L = 5 V

L = 90 gpm • 500 lbs/hr = 45,000 lbs/hr = 5 • V gpm

Therefore, V = 9,000 lbs/hr (steam)

And: 9,000 lbs/hr (steam) • hr = 15 lbs steam/gallon of product 10 gpm (product) 60 min

Example 2. Calculate the steam required to strip 100 gpm of a 5% volume beer (95 gpm water/5 gpm ethanol).

L/V = 6.33 (typical for a 5% volume beer) or L = 6.33 • V

L = 95 gpm • 500 lbs/hr = 47,500 lbs/hr = 6.33 • V gpm

Therefore, V = 7,500 lbs/hr (steam)

And: 7,500 lbs/hr (steam) • hr = 25 lbs steam/gallon of product 5 gpm (product) 60 min

*L and V are liquid and vapor flow rates, respectively, expressed in lb-mole per hr. Note: at base of column use simplifying assumption of water (L) and steam (V). Therefore: L (lb-mole/hr) = L (lbs/hr)

V(lb-mole/hr) V(lbs/hr)

326 P.W. Madson

design chart like that shown in Figure 8 for theethanol-water system. Such a graph is usefulwhen calculations are needed to ascertaintechnical and economic feasibility andpreliminary conditions for the design.

Figure 9 demonstrates how the liquid:vaporflow ratio, in connection with the number ofstages (theoretically ideal trays) required for aspecified separation between ethanol and water,is graphically determined. Note that the stagesare constructed by drawing straight linesvertically and horizontally between theequilibrium curve (previously determinedexperimentally) and the operating lines. For anethanol stripper/rectifier, there are two operatinglines: one for the rectification section and onefor the stripping section. The operating linesrepresent the locus of concentrations within thedistillation tower of the passing liquid and vaporstreams. The operating lines for a given tower

are based on the energy input, as calculated andrepresented in Figure 8. Because of the principleof constant molal overflow, the operating linescan be represented as straight lines. If constantmolal overflow was not valid for the ethanol/water distillation, then these lines would becurved to represent the changing ratio of liquidflow to vapor flow (in molar quantities)throughout the tower. The slope of the operatingline (the ratio of liquid flow to vapor flow) isalso called the internal reflux ratio. If the energyinput to a tower is increased while the beer flowremains constant, the operating lines will movetoward the 45o line, thus requiring fewer stagesto conduct the distillation. Likewise if the energyinput is reduced (lowering the internal refluxratio), the operating lines will move toward theequilibrium curve, reducing the degree ofseparation achievable in each stage and therefore

Beer concentration (volume %)

Lbs steam

/gallon (anhydrous ethanol basis

)

30

25

20

15

10

5

0

0 5 10 15

Constraints:

1. 190° proof product

2. 0.02% (wt) bottoms

3. Saturated feed

4. Direct steam

Theoretical minimum

Practical minimum

Figure 8. Steam requirements ethanol stripper/rectifier.


requiring more stages to conduct the distillation.The calculations underlying the preparation ofFigure 9 go beyond the scope and intent of thistext, but have been included for continuity. Thedashed lines represent the graphical solution tothe design calculations for the number oftheoretical stages required to accomplish adesired degree of separation of the feedcomponents. Figure 9 is referred to as a McCabe-Thiele diagram. For further pursuit of thissubject, refer to the classical distillation textbookby Robinson and Gilliland (1950).

TOWER SIZING

The goal of the design effort is to establish thesize of the distillation tower required. Table 2shows the basic procedure to determine the

diameter required for the given distillation tower.Since all of the distillation ‘work’ is done by thetrays, the tower is actually the ‘container’ tosurround the vapor and liquid activity that is‘managed’ by the trays. Tower diameter designis, therefore, actually the design of the necessarytray diameter for proper vapor/liquid interactionand movement.

The f factor (vapor loading) is an empiricallydetermined factor that depends primarily upontray type and spacing, fluid physical properties,froth stability and surface tension at theoperating conditions of the system. The propervalues for f are determined by field observations.In summary then, the f factor can be describedas an adjusted velocity term (units are ft/sec) thatwhen multiplied by the square root of the densityratio of liquid to vapor, will give the allowablevapor velocity in the empty tower shell, such

14

100

0 10 20 30 40 50 60 70 80 90 100

90

80

70

60

50

40

30

20

10

0

Typical L / V=5.0

Typical: 14 rectifying stages

8 stripping stages

1

2

3

4

5

6

7

8 1110 12 139

Equilibrium curve

Operating line

rectification

Operating line

stripping

190° proof


Stillage

(0.02 wt %)

Beer feed

(10 volume %)

Mole %

ethanol in vapor

Figure 9. Vapor/liquid equilibrium stage analysis.

328 P.W. Madson

Table 2. Calculations for tower sizing (base of stripper).

Example: Calculate tower diameter required for a 10% volume beer at 100 gpm (15 lbs steam/gallon of product)

W Vapor flow rate = 9,000 lbs/hr (steam)P Operating pressure at base = 1.34 ATMMAVG Average MW of vapor = 18 lbs/lb-moleρL Liquid mixture density = 59.5 lbs/ft3 (227 °F)T Absolute operating temperature = 687 °RD Tower inside diameter in inchesf Vapor loading factor = 0.05-0.3

Sizing equation: f

16.45=

687

59.5•18•1.34•f

9000•0.2085=

T

r•M•P•f

W•0.2085=D

LAVG

Assuming f = 0.16 (specific to tray design and spacing), the tower diameter is:

inches41.125=0.16

16.45=D

Tower diameter calculation utilizes the following equation. Values for the terms are indicated above.

T

r•M•P•f

W•0.2085=D

LAVG

The final design equation can be derived beginning with the fundamental equation:

ɟɟ-ɟ•=u

V

VLf

Where u = average vapor velocity in empty tower shell (ft/sec)ρL = liquid density (lbs/ft3)ρV = vapor density (lbs/ft3)f = tower vapor loading factor (ft/sec)

(Note: For most cases ρL is much greater than ρV, so that ρL – ρV ≅ ρL. For example, water (steam) at 212oF and atmosphericpressure: ρL = 59.8 lbs/ft3 and ρV = 0.0373 lbs/ft3. Then ρL – ρV = 59.7627 lbs/ft3 ≅ 59.8 lbs/ft3, which results in a negligible0.06% error.)

Consequently, ɟɟ•u

V

Lf@

Imposing the equation of continuity: W = A•ρV•u or u = W/A•ρV

Where A = tower cross-section area (ft2)ρV = vapor density (lbs/ft3)u = average vapor velocity in empty tower shell (ft/sec)

and W = vapor mass flow (lbs/sec)

Substituting for u in the equation above:

ɟɟ•

ɟV•A

W

V

Lf@ then ɟ•ɟ•

WAorɟ•ɟ•

A

W

vL

vLf

f @@ɟV ɟL ɟV

ɟL ɟV


that liquid entrainment and/or vapor phasepressure drop in the tower will not be excessive.

Excessive vapor velocity will first manifestitself by causing excessive liquid entrainmentrising up the tower, causing loss of separationefficiency. Ultimately the excessive entrainmentand pressure drop will cause tower flooding.

To achieve a well-balanced tower design, theforegoing analysis must be performed at eachstage of the tower, from bottom to top.Composition changes, feed points, draws, etc.,each can cause a different requirement. Thetower must be examined to locate the limitingpoint.

Similar analyses, with empirically-observedperformance coefficients, are applied to vaporpassing through the trays and through the liquid,and to the movement and control of liquidpassing through downcomers and across thetrays. These analytical procedures are beyondthe scope of this text. Reference should be madeto the aforementioned text by Robinson andGilliland for further information.

Considerations in optimizing distillationsystem design

Optimizing the technical and economic designof distillation equipment and similar gas and

vapor/liquid mass transfer systems involves anumber of interrelated parameters. The positive/negative balance of a variety of contactingdevices with different capacities and efficienciesfor promoting vapor/liquid mass transfer mustbe taken into consideration. Along with thetechnical issues considered in such designs,economical operation is essential not only in thereduction of energy and other direct costs, butalso in relation to investment and return oninvestment from the operation being considered.In this respect, distillation towers are notindependent process-wise, as consideration mustalso be given to other auxiliaries such asreboilers, condensers, pumps, controls andrelated equipment.

SIZING TOWERS

In determining optimum diameter and height oftowers for distillation, absorption, stripping andsimilar mass transfer operations, design factorsare affected by whether the installations will beindoors or outdoors. With indoor installations,building height limitations, as well as floor levelaccessibility, are an important factor in thedesign. Where there are height limitations, towersmust be increased in diameter to provide forreduced tray spacing, which in turn will require

Use the Ideal Gas Law to express the vapor density.

T•RM•P

=ɟ AVGV Then by substitution one obtains

T•Rɟ•M•P

•

W=A

LAVGf

Using the Universal Gas Constant R = 0.73 (ft3)(atm)/(lb-mole)(oR) the equation becomes:

Tɟ•M•P

••1.17

W=A

LAVGf Now D•0.7854=4D•ˊ

=A 22

andTɟ•M•P

•

W•1.0879=

Tɟ•M•P

••0.9192

W=D

LAVGLAVG

2

ff

Adjusting units:

Tɟ•M•P

•

W•

60

12•1.043=D

LAVGf

Therefore,

Tɟ•M•P

•

W•0.2085=(inches)D

LAVGf

M•P AVG

r•M•P LAVG D2•

r•M•P LAVG

r•M•P LAVG r•M•P LAVG

r•M•P LAVG r•M•P LAVG

330 P.W. Madson

lower vapor velocities. With outdoorinstallations, literally the ‘sky is the limit’, andrefinery and petrochemical towers of 200 feetin height are not uncommon.

In either case, indoors or outdoors, theinterrelated tower diameter and tray spacing arelimited by allowable entrainment factors (ffactors) (Katzen, 1955). If outdoors, towerheights and diameters must be related tomaximum wind loading factors in the specificplant location and may be complicated byallowance for earthquake factors.

TRAY AND PACKING SELECTION

Vapor/liquid contacting devices may be of twodistinct types, namely packed or tray (staged)towers. In packed towers, the transfer of materialbetween phases occurs continuously anddifferentially between vapor and liquid through-out the packed section height. By contrast, intray towers, the vapor/liquid contact occurs onthe individual trays by purposely interruptingdown-flowing liquid using downcomers toconduct vapor-disengaged liquid from tray totray and causing the vapor/liquid contact to occurbetween cross-flowing liquid on the tray withvapor flowing up through the tray. In otherwords, the vapor/liquid contact is intermittentfrom tray to tray, and is therefore referred to asbeing stagewise. Thus, for any given separationsystem, the degree of vapor/liquid contact willbe greater with a greater height of the packedsection, or in the case of tray towers, a greaternumber of trays used.

It is generally considered that packing-typeinternals may be used with relatively clean vaporand liquid systems where fouling is not aproblem. Economics indicate that packing isapplicable in small and modest sized towers. Asthe towers become larger, packing becomescomplicated by the need for multiple liquidredistribution points to avoid potential vapor/liquid bypassing and reduction in efficiency.Structured packings (Fair et al., 1990; Bravo etal., 1985) are designed to minimize theseproblems by reducing the height requirement andcontrolling, to some extent, the distribution ofliquid. However, high fabrication andspecialized installation costs would indicate thatthese are applicable only for relatively low-volume, high-value product processing.

Trays of various types are predominant invapor/liquid contacting operations, particularlyon the very large scale encountered in thepetroleum and petrochemical industries, in largescale operations of the chemical processindustries and in the large scale plants of themotor fuel grade ethanol industry.

The venerable bubble cap tray, with a widevariety of cap sizes, designs and arrangementsto maximize contact efficiency, has fallen outof favor during the past few decades because ofthe relatively high cost of manufacture andassembly. Valve trays of several types have takenover in operations requiring a relatively widevapor handling capacity range (turndown). Thishas been extended by use of different weightsof valves on the same tray. Specialty trays suchas the Ripple, Turbogrid, tunnel cap and othersdesigned to improve contact under certainspecific circumstances have been used to alimited extent.

The long established perforated tray is acontacting tray into which a large number ofregularly oriented and spaced small circularopenings have been drilled or punched. Thesetrays are commonly referred to as ‘sieve trays’because of the original practice of putting themaximum number of holes in any given trayarea. This original design produced a fairlyinefficient operation at normal loading, and avery inefficient operation with decreased vaporloading. About 50 years ago, engineers beganto suspect that the design approach had been inerror, and that the hole area in the trays shouldbe limited by the hole velocity loading factor toobtain maximum contact by frothing, asindicated in Figure 10.

The hole vapor loading factor (or perforationfactor) is defined as the vapor velocity throughthe perforations adjusted by the square root ofthe vapor density at the specific tower locationof a given perforated tray. With the paralleldevelopment of separation processes in thepetroleum refining, chemical and ethanolindustries, the modern approach has developedto what is now called ‘perforated tray’ design.The Fractionation Research Institute of theAmerican Institute of Chemical Engineersdiverted its efforts from bubble cap studies toperforated tray testing; and have established abasis for the design of perforated trays with highefficiency and wide capacity range (RaphaelKatzen Associates, 1978).


Where foaming or tray fouling (caused bydeposition of solid materials in the tower feed)can be an operational problem, novel designssuch as the baffle tray may permit extendedoperating time between cleanings. Baffle traysmay take a number of different forms. They canbe as simple as appropriately spaced,unperforated, horizontal metal sheets coveringas much as 50-70% of the tower cross sectionalarea; or they may take the form of a series ofvertically spaced, alternating, solid disc-and-donut rings (see Figure 7). Towers up to 13 ftdiameter are in operation using this simple disc-and-donut design concept.

Although system-specific data have beendeveloped for each type of tray, it is difficult tocorrelate tray loading and efficiency data for awide variety of trays on a quantitative basis.Each system must be evaluated based uponempirically-derived loading factors for vaporand liquid operations within the tower.

AUXILIARIES

Energy input is of prime importance in towerdesign, particularly in ethanol stripping andrectification units. In aqueous and azeotrope-forming systems, direct steam injection has been

common practice to maintain simplicity.However, the need to reduce the volume ofwaste going to pollution remediation facilitieshas minimized use of this simple steam injectiontechnology to avoid the dilution effect of thesteam being condensed and added to the stillage.Direct steam injection transfers both the energyand the water into the process. By imposing aheat exchanger (reboiler) between the steam andthe process, only the energy is transferred intothe tower. The condensate water is returned in aclosed loop to the boiler, thus reducing thebottoms outflow from the process. Reboilers arethus growing in acceptance, and several typesmay be employed. Kettle and thermosyphonreboilers are preferred where fouling is not aproblem. Where fouling can occur, high velocity,forced-circulation, flash heating reboilers arepreferred. Figure 11 depicts the reboiler energytransfer by a forced-circulation reboiler ascompared to Figure 1 which depicts direct steaminjection.

Thermocompression injection of steam hasalso been utilized where low pressure vapors areproduced from flash heat recovery installationsand where higher pressure motive steam is alsoavailable.

Condenser design would appear to be simple.However, in many cases, water limitationsrequire adapting condenser designs to the use

Inlet weir

Outlet weir

Downcomer

area

Downcoming

liquid

Stripped

liquid

Enriched

vapor

Froth zone

Vapor

V/L mixing

V/L separation

Figure 10. Perforated tray frothing.

332 P.W. Madson

of cooling tower water with limited temperaturerise and minimal scale-forming tendencies. Onthe other hand, where water is extremely scarce,air-cooled condensers are used.

Energy conservation

The increasing cost of thermal energy, whetherprovided by natural gas, fuel oil, coal or biomass,is fostering an increased emphasis on heatrecovery and a reduction in primary thermalenergy usage (Fair, 1977; Petterson et al., 1977;Mix et al., 1978). Conventional bottoms-to-feedheat exchangers are now being supplementedwith recovery of overhead vapor latent heat bypreheating feed streams and other intermediateprocess streams. Techniques of multistage

distillation (similar to multiple effectevaporation) are also practiced. Pressure-to-atmospheric, atmospheric-to-vacuum, orpressure-to-vacuum tower stages are utilized,with the thermal energy passing overhead fromone tower to provide the reboiler heat for thenext one. Two such stages are quite commonand three stage systems have also been utilized(Katzen, 1980; Lynn et al., 1986).

Furthermore, the modern technique of vaporrecompression, commonly used in evaporationsystems, is also being applied to distillationsystems. Such a system can provide forcompression of overhead vapors to a pressureand temperature suitable for use in reboiling alower pressure stripping tower. However, thecompression ratios required for such heatrecovery may consume almost as much

Condenser Cooling water

Low boiling product

Note: See Figure 1 for

comparison to direct

steam injection

Reflux

Vapor

Feed mixture

Rectif

yin

gS

trip

pin

g

Steam (energy)

High boiling product

Condensate

return to boiler

Reboiler

Figure 11. Energy transfer by a forced-circulation reboiler.


electrical energy as would be saved in thermalinput. Alternative systems, using vaporrecompression as an intermediate stage devicein the distillation system, have also beenproposed.

Control systems

Control systems can vary from manual control,through simple pneumatic control loops to fullyautomated distributed control (Martin et al.,1970). High level computer control hasfacilitated the application of sophisticatedcontrol algorithms, providing more flexibility,reduced labor and higher efficiency with lowercapital investment. Such systems, when properlyadapted to a good process design, have provenmore user-friendly than the control techniquesutilized in the past.

Economic design

In integrating the technology discussed, the finalanalysis must be economic. Alternative systemsmust be compared on the basis of investmentrequirements, recovery efficiency and relativecosts of operation. Thus, any heat exchangersinstalled for heat recovery must show asatisfactory return on the investment involvedin their purchase and installation. In comparingalternative separation systems, the overallequipment costs must be compared againstenergy and other operating costs to determinewhich system offers the best return. Moderncomputer-assisted designs incorporateeconomic evaluation factors so economicoptimization can be determined rapidly.

Ethanol distillation/dehydration: specificsystems technology

Proven industrial technologies are available fordistillation of various grades of ethanol fromgrain, sugarcane, molasses and other feedstocks.Improvements have been made over the years,particularly during development of the motorfuel grade ethanol industry. In such installations,a key requirement is the minimization of totalenergy usage.

The operation that has been most subject tocritical comment is the distillation process. Manyrelatively new ‘authorities’ in the field havebased their criticism on technologies that goback 50-60 years, and have created anunwarranted condemnation of distillation as aviable process for low energy motor fuel gradeethanol production. Systems developed over theyears will be described to show that much ofsuch criticism is unwarranted and unjustified.

PRODUCTION OF INDUSTRIAL ETHANOL

Prior to the emphasis on motor fuel gradeethanol, the major ethanol product utilizedworldwide was high purity, hydrous industrialethanol, which is generally produced at astrength of 96o GL (192o US proof) (oGL =degress Gay Lussac = % by volume ethanol; USproof = 2 x % by volume ethanol). Efficientsystems have been in commercial operation formany years for the production of such highgrade ethanol from ethylene, grain, molasses andsulfite waste liquor. The basic distillation systemis shown in Figure12.

In the case of synthetic ethanol (outside thescope of this publication), the beer strippingtower is not required and the refining system isa simple three tower unit, which achieves 98%recovery of the ethanol in the crude feed as afirst grade product. The final product maycontain less than 5 ppm total impurities and hasa ‘permanganate time’ of more than 60 minutes.

For the production of industrial or beveragespirit products made by fermentation of grain,molasses or sulfite liquor, the system utilizes thefull complement of equipment shown in Figure12. The beer feed is preheated from the normalfermentation temperature in several stages,recovering low level and intermediate level heatfrom effluent streams and vapors in the process.This preheated beer is degassed and fed to thebeer stripper, which has stripping trays belowthe beer feed point and several rectifying traysabove it. The condensed high wines from thetop of this tower are then fed to the extractivedistillation tower, which may operate at apressure of 6-7 bars (87-101.5 psi). In this tower,most of the impurities are removed and carriedoverhead to be condensed as a low grade ethanol

334 P.W. Madson

stream, from which a small purge of heads(acetaldehyde and other low boiling impurities)may be taken while the primary condensate flowis fed to the concentrating tower. The purified,diluted ethanol from the bottom of the extractivedistillation tower is fed to the rectifying tower,which has an integral stripping section. In thistower, the high grade ethanol product, whetherindustrial or potable, is taken as a side draw fromone of the upper trays. A small heads cut isremoved from the overhead condensate. Fuseloils (mixtures of higher alcohols such as propyl,butyl, and amyl alcohols and their isomers, whichare fermentation by-products or ‘congeners’) aredrawn off at two points above the feed tray butbelow the product draw tray to avoid a buildupof fusel oil impurities in the rectifying tower. Theoverhead heads cut and the fusel oil draws arealso sent to the concentrating tower.

It should be noted that the rectifying tower isheated by vapors from both the pressurizedextractive distillation tower and the pressurizedconcentrating tower.

In the concentrating tower, the various streamsof congener-containing draws are concentrated.

A small heads draw is taken from the overheadcondensate, which contains the acetaldehydefraction along with a small amount of the ethanolproduced. This may be sold as a by-product orburned as fuel. A fusel oil side draw is taken athigh fusel oil concentrations through a coolerto a washer. In the washer, water is utilized toseparate the ethanol from the fusel oil, with thewashings being recycled to the concentratingtower. The decanted fusel oil may be sold as aby-product. The ethanol recovered from thecrude streams is taken as a side draw from theconcentrating tower and fed back to theextractive distillation tower for re-purificationand recovery of its ethanol content.

In an early version of this system, installedmore than 60 years ago for the production ofpotable ethanol from grain and from molasses,all towers were operated at atmospheric pressure.However, installations made within the past 40years utilize the multistage pressure system toreduce energy consumption to a level of about50% of the all-atmospheric system.

The commercial installations utilizing themultistage pressure, or ‘pressure cascading’

Condenser

CW

Beer feed

Extraction

water

Condenser/

reboiler

Condenser/

reboiler

Preheater/

condenser

Reflux

Waste water

Washwater

Fusel oil

by-product

Fusel oil

decanter

Fusel oil

cooler

CW

CW

Heads by-product

Product ethanol

to storage

Stillage

SteamSteam

Steam

Reboiler Reboiler Reboiler

Atm

ospheric

rectif

yin

g tow

er

Pressuriz

ed h

eads

and f

usels

concentratin

g t

ow

er

Pressuriz

ed extractiv

e tow

er

Beer strip

per

Figure 12. Low energy-consuming high grade hydrous ethanol distillation.


technique operate with a steam consumption of3.0-4.2 kg of steam/liter (25-35 lb/gallon) of 96o

GL ethanol. This may be compared to about 6kg of steam/liter for earlier conventionaldistillation systems.

Production of anhydrous ethanol

Systems have been designed and installed forproduction of extremely dry and very pureanhydrous ethanol for food and pharmaceuticaluse, primarily in aerosol preparations. Thesesystems, as shown in Figure 13, yield ethanolcontaining less than 200 ppm water (99.98o GL),less than 5 ppm total impurities and more than60 minutes permanganate time.

The two tower dehydrating system has beenoperated in two super-anhydrous plants inCanada, and was used to produce motor fuelgrade ethanol (99.5o GL) in four installations inCuba (prior to the advent of the Castro regime).The dehydrating tower and the entrainer-recovery tower are operated at atmosphericpressure. Thus, they may utilize either lowpressure steam, hot condensate or hot wastestreams from other parts of the ethanol processto minimize steam usage. To simplify equipment

and minimize investment, a common condensingand decanting system is used for the two towers.

The entrainer used to remove water as a ternary(three component) azeotrope may be benzene,heptane (C6-C8 cut), cyclohexane, n-pentane,diethyl ether or other suitable azeotropic agents.The entrainer serves to create a three componentazeotrope that boils at a temperature lower thanany of the three individual components andlower than the ethanol/water binary (twocomponent) azeotrope. Therefore the ternarymixture will pass overhead from the tower,carrying the water upward. Upon condensing,the mixture separates in a decanter into anentrainer-rich layer and a water-rich layer.

The hydrous ethanol feed enters thedehydrating tower near the top. The feedcontacts the entrainer in the upper section of thetower. The three component mixture in thissection of the tower seeks to form its azeotrope,but is deficient in water and contains moreethanol than the azeotrope composition.Therefore, the ethanol is rejected downward inthe liquid and is withdrawn as an anhydrousproduct from the bottom of the tower. The waterjoins the entrainer, passing upward as vapor toform a mixture that is near the azeotrope

190° proof ethanol feed

from distillation

Dehydration

tower

Reboiler

Condenser

Entrainer

recovery tower

Anhydrous ethanol

to storage

CW

CW

Reboiler

Waste

water

Cooler

Decanter

Low pressure steam

or hot condensate

Condensate return

CW

Cooler

Figure 13. High grade anhydrous ethanol system.

336 P.W. Madson

composition for the three components. Thecondensed mixture separates into two layers inthe decanter and the entrainer-rich layer isrefluxed from the decanter back to the top ofthe tower. The aqueous layer is pumped fromthe decanter to the entrainer-recovery tower, inwhich the entrainer and ethanol are concentratedoverhead in the condenser-decanter system. Thestripped water, emerging from the base of thetower, may go to waste. If it has substantialethanol content, it may be recycled to the spiritunit, but this introduces the risk of traces of theentrainer in the hydrous ethanol which may notall be sent to the dehydration system. This systemoperates with a steam consumption of 1-1.5 kg/liter (8.3-12.5 lb/gallon) of anhydrous ethanoldepending on the quality of product required.As indicated above, a major part of theequivalent steam energy can be provided by hotcondensate and hot waste streams from the spiritunit.

References

Bravo, J.L. et al. 1985. Hydrocarbon Proc. Jan.p.91.

Katzen, R. 1955. Chem. Eng. Nov. p. 209.Fair, J.R. 1977. Chem. Eng. Prog. Nov. p. 78.Fair, J.R. et al. 1990. Chem. Eng. Prog. Jan. p.

19.Katzen, R. 1980. Low energy distillation

systems. Bio-Energy Conference, Atlanta, GA,April.

Lynn, S. et al. 1986. Ind. & Eng. Chem. 25:936.Martin, R. L. et al. 1970. Hydrocarbon Proc. March

1970, p. 149.Mix, T.J. et al. 1978. Chem. Eng. Prog. April p.

49.Petterson, W.C. et al. 1977. Chem. Eng. Sept. p.

79.Raphael Katzen Associates. 1978. Grain Motor

Fuel Alcohol Technical and EconomicAssessment Study. Prepared for: USDepartment of Energy HCG/J6639-01.

Robinson and Gilliland. 1950. Elements ofFractional Distillation, McGraw Hill Book Co.,Inc.

Development and operation of the molecular sieve: an industry standard 337

Chapter 23

Development and operation of the molecular sieve: an industry standard

R.L. BIBB SWAIN

Delta-T Corporation, Williamsburg, Virginia, USA

Introduction: the early days

AZEOTROPIC DISTILLATION

The equipment used during the dawn of the fuelethanol program was built using a strangecombination of technologies borrowed fromdifferent industries. Most of the basic productionexpertise came from the beverage alcoholindustry, but there was no need to remove allthe water from beverage alcohol. To preventphase separation when blended with gasoline,fuel grade ethanol needed to be almostcompletely dry. Standard distillation still leavesover 4% water in the ethanol, so a process called‘azeotropic distillation’ was used in all the earlycommercial ethanol plants to remove the finalwater from (dehydrate) the ethanol. Azeotropicdistillation uses a third component, typicallybenzene or cyclohexane, to ‘break’ the azeotrope(the composition above which standarddistillation becomes ineffective).

Azeotropic distillation systems tended to bequite expensive, difficult to operate and adjust,and consumed a significant amount of energy.Attempts were made to reduce the operatingdifficulties and energy consumption, but the twogoals were incompatible. Multi-pressuretechniques could reduce energy consumption,but at the expense of more difficult operationand considerably higher capital cost. A processupset could easily contaminate the ethanol

product with benzene; and there was no way tototally protect plant workers from exposure toknown or expected carcinogens.

SYNTHETIC ZEOLITES AND EARLYMOLECULAR SIEVE DEHYDRATORS

An inventor named Skarstrom received a patentin 1957 for a device that utilized a syntheticzeolite adsorbent that selectively removed waterfrom air and some other gasses and vapors(Ruthven, 1984). A different zeolite could evenbe used in virtually the same device to separateoxygen and nitrogen from air (Ruthven, 1984).These synthetic zeolites became known as‘molecular sieves’ due to the very precise poresize that enabled them to select and remove onemolecule size from a bulk mixture containingmolecules with a larger size or lower polarity. Aproperly designed system could dry air to a dewpoint of -100oF; and drying air became the firstmajor application for molecular sieves. By theearly 1980s, at least a dozen companies offeredmolecular sieve air dryers.

If molecular sieves could be used to dry air,why not ethanol? Experiments were under way

338 R.L. Bibb Swain

at that time by several inventors and companiesto prove efficacy of that concept. In the August1981 issue of Gasohol USA (the only tradejournal at the time), two companies advertisedmolecular sieve dehydrators that dried ethanolin the vapor phase. Ad-Pro Industries, Inc. ofHouston, Texas, and Anhydrous Technology,Inc. from Sugarland, Texas, both offered smallskid-mounted systems for drying up to 200gallons per hour. Two other companies, PallPneumatics and Shroeder Farms, were marketingmolecular sieve ethanol dehydrators that driedthe ethanol in the liquid phase. The liquid phasesystems used too much energy, had severeproblems with desiccant fouling and finally werejust not able to compete with the vapor phaseunits. The two companies eventually soldapproximately a dozen of the small vapor phasedehydrators. Since the devices were relativelysimple, many entrepreneurs built unauthorizedcopies for themselves and others.

The early vapor phase dehydrators hadproblems of their own. Aside from a general lackof reliability, the molecular sieve desiccantwould deteriorate within a year; and all the dustthat resulted would contaminate the ethanolproduct. The commercial molecular sievedesiccant was manufactured by mixing themolecular sieve crystalline powder (much likeglass dust) with a clay binder, and forming theslurry into beads or extruded pellets about 1/8inch in diameter. The finished product had thelook and toughness of limestone. Any movementof a bed of the material resulted in abrasionbetween the particles. The resulting dustaccumulated in the bed causing vaporchanneling and product contamination. Bedmovement could be caused by excessivetemperature swings (thermal expansion andcontraction), excessive pressure drop across thebed (fluidization) and local disturbance by high-speed vapor entering the bed during cycletransition (jetting). By 1987, Ad-Pro Industriesand Anhydrous Technology, whose designs weresubject to all these problems, were both out ofbusiness.

NEW TECHNOLOGY BECOMES PROVEN ONLARGE SCALE

In 1984, Delta-T Corporation was formed inWilliamsburg, Virginia, to build and market a

molecular sieve dehydrator design that had beendeveloped over the past three years by Delta-T’s founder. The new design eliminated thereliability and desiccant deterioration problemsthat plagued earlier designs. Delta-T built a 150-gallon/hr prototype at a small ethanol plant inCharles City, Virginia. After conducting a batteryof tests to prove the concept and optimize thedesign, the prototype continued to produceethanol commercially at the plant until theVirginia ethanol incentive expired many yearslater. Delta-T sold six dehydrators during its firsteight months of operation, the largest of whichwas 800 gallons/hr.

In the meantime, larger ethanol plantscontinued to be built using azeotropic distillationdehydration technology. Molecular sievedehydration was a relatively new technologywith a somewhat troubled past; and Delta-T hadno large reference unit to show a customer. Inaddition, the firms providing technology for thebalance of the plant had their own azeotropicdistillation technologies to sell the customer.They were quick to point out both the difficultiesof getting uniform vapor flow through large bedsand that the technology was not proven at a largerscale. Both were correct, but misleading.

Not able to sell a large dehydrator to anAmerican ethanol plant, Delta-T formed a joint-venture with an Italian engineering firm to builda 30 million gallon/year dehydrator for thelargest distillery in Europe, located in Sicily.Design specifications called for the feedstockethanol to be as low as 160 proof (80% ethanolby volume), with a 199.4 minimum productproof. Performance guarantees were stringent,but Delta-T badly needed a large reference unit,so it contracted to provide the unit. Happily forall project participants, the new world-scaledehydrator worked just as promised, and Delta-T finally had proven the design at large scale.

Molecular sieve dehydrators had alwaysoffered the industry lower capital costs, loweroperating costs, and greater personnel safety thanazeotropic distillation. Now that a large unit wason line and performing well, the last obstacle toindustry acceptance was overcome. At first,ethanol plant developers would get their planttechnology from the older process design firms,but they would come directly to Delta-T for theirmolecular sieve dehydrator. When the processtechnology firms realized they would no longerbe able to sell azeotropic distillation dehydration


to their customers, they reluctantly adoptedmolecular sieve technology. Today, all newethanol plants are built with molecular sievedehydrators.

How does a molecular sieve dehydratorwork?

PRESSURE SWING ADSORPTION

Most modern molecular sieve dehydrators usea process known in the industry as ‘pressureswing adsorption’ to remove water from avaporized feed stream. We will concentrateherein on the removal of water from an ethanolvapor stream, but many other applications arepossible with virtually the same device. The term‘pressure swing’ refers to the fact that thedehydrator uses a relatively high pressure whenwater is being removed from the feed streamand a relatively low pressure when the molecularsieve desiccant is being regenerated (havingwater removed from the desiccant). Mostcommercial designs have two or more beds ofdesiccant and cycle the vapor flow through thebeds to provide continuous operation. While onebed is on-line drying the feed vapor, anotherbed is being regenerated. In some designs oneor more additional beds are being depressurizedor repressurized in preparation for the next cycle(Le Van, 1996).

HOW DOES ADSORPTION WORK?

The ‘macroscopic’ principles of adsorption arequite simple, but attempts to accurately explainhow adsorption works often fail because theauthor gets bogged down in technical jargon.The basic characteristic of an adsorbent materialis a stronger affinity for one type of atom ormolecule than for the other types in the vaporstream. In the case of a molecular sieve ethanoldehydrator, we select an adsorbent with a strongaffinity for water and little affinity for ethanoland the other impurities typically contained inthe ethanol feed stream. As wet ethanol vaporpasses through the bed, the desiccant adsorbsthe water molecules but not the ethanolmolecules. This process cannot continueindefinitely, because the desiccant has a finitecapacity for water; and that capacity is affected

by the operating temperature and pressure.Moisture capacity increases as pressure increasesor temperature decreases, and vice versa. Sincea pressure swing adsorption dehydrator operatesat almost constant temperature (Ruthven, 1994),we can desorb the water adsorbed on theprevious cycle by lowering the operatingpressure and passing a purge vapor through thebed in the opposite direction to sweep the watermolecules from the free space surrounding thedesiccant beads.

The class of materials known as zeolites occursnaturally, but most commercial molecular sievesare man-made. Synthetic zeolites have acrystalline lattice structure that contains openings(pores) of a precise size, usually measured inangstroms (Å). Molecular sieves can bemanufactured with different pore sizes by usingdifferent chemistry and manufacturing methods.A synthetic zeolite of type 3Å is used in mostethanol dehydrators, because the pores are 3Åin diameter while water molecules are 2.8Å andethanol molecules are 4.4Å. Therefore, watermolecules are strongly attracted into the poresbut ethanol molecules are excluded. Many otheradsorbent materials are available that have anaffinity for water such as activated alumina.However, other adsorbents have a wide rangeof pore sizes and therefore are much lessselective than molecular sieves.

Water is so strongly attracted to type 3Åmolecular sieve that for each pound adsorbed,1,800 BTUs of heat are released. This effect isreferred to as the heat of adsorption. When youremove that same pound of water duringregeneration, you must supply 1,800 BTUs ofheat (referred to as the heat of desorption). Thus,as feed vapor is dehydrated the bed of desiccantheats up; and as the bed is regenerated it coolsdown. Type 3Å molecular sieve is capable ofadsorbing up to 22% of its weight in water, butpressure swing adsorption dehydrators operateat a much lower moisture loading and limit thedehydration period to prevent an excessivetemperature swing.

Operational considerations

ENERGY NEEDS, PRODUCTION CAPACITIES

By limiting the adsorption period, temperatureswings are kept to a reasonable value and the

340 R.L. Bibb Swain

heat of adsorption can be effectively stored inthe bed of desiccant as sensible heat. The heatis then available to supply the necessary heat ofdesorption during the regeneration period. Theability to recover the heat of adsorption is whymolecular sieve dehydrators are so energyefficient. Assuming the feed vapor comes directlyfrom the rectifier column, the only energy useddirectly by the molecular sieve dehydrator is thesteam required to superheat the vapor to bedoperating temperatures (0.1 to 0.2 lbs of steamper gallon of anhydrous ethanol) and theelectricity to operate the pumps (0.02-0.03 Kwhr per gallon). Many companies forget thatadditional energy is required to redistill theliquid that results during regeneration as well asthe electricity to operate the cooling tower andair compressor. A properly designed system willhave a totally inclusive energy consumption ofabout 4,000 BTUs per gallon of anhydrousethanol, assuming the feed ethanol contains 5%water and the product is dried to 0.25% water.Additional energy is required if the feed ethanolis wetter or if the product is dryer.

Molecular sieve dehydrators can accommodatea wide range of specifications, and can be builtfor any desired production rate. Some arecurrently in service drying ethanol containingas much as 20% water and some can produceanhydrous ethanol with as little as 20 ppm waterin the final product. The smallest units aredesigned for a product rate of about two gallonsper hour (mostly for lab and test work), whilethe largest can produce 100 million gallons peryear.

A TYPICAL OPERATING CYCLE DESCRIBED

The following describes a typical operating cyclefor a 2-bed dehydrator drying ethanol containing5% water. We begin with Bed 1 on-line (dryingthe feed vapor) and Bed 2 being regenerated(see Figure 1). Wet ethanol vapor enters the topof Bed 1 under a modest pressure, passesdownward through the bed and exits the bottomas dry vapor. 60-85% of the vapor leaves thesystem as anhydrous ethanol product; and the

Figure 1. Operation of two bed molecular sieve dehydrator.

Dry ethanol vapor

Wet

ethanol vapor

Very wet

ethanol vapor

Slip stream of dry vapor

RegeneratingBed

(Low Pressure)

DehydratingBed

(High Pressure)


remainder is fed to Bed 2 as a regeneration purgestream. Bed 2 is maintained under a vacuumthat shifts the equilibrium conditions so that theadsorbed water is desorbed. Operatingpressures, temperatures and flow rates areadjusted so that regeneration is complete at thepoint in the cycle when the beds change position(valves are positioned to reverse the roles of thetwo beds). One bed typically stays on-linedehydrating the feed stream for 3-10 minutesbefore being regenerated. When the bedstransition from dehydrate to regenerate mode,the pressure in the inlet header tends to sag,which can upset the rectifier column ifprovisions are not made to mitigate the pressuredrop. After the beds switch, the bed that wasbeing regenerated ‘pressures up’ to thedehydrating pressure and the vaccuum systemlowers the pressure in the other bed to theregenerating pressure. The cycle repeats itselfcontinuously. Since one bed or the other isalways receiving and drying the inlet vapor, theprocess seems to be continuous; but in reality itis more properly termed as ‘cyclic batch’.

While a molecular sieve dehydrator canoperate on a simple timed cycle, the bestsystems are managed by an ‘intelligent’

controller (computer or PLC) to provide greateroperating safety, reliability and energy efficiency.Once adjusted to standard operating conditions,the molecular sieve dehydrator requires littleoperator interface and can tolerate reasonablevariations in feed rate or quality without needfor readjustment. Typical quality controlmeasures include sampling the product, feed andregeneration liquid at regular intervals to assurethe product meets specifications and that thesystem is operating efficiently.

References

LeVan, M.D. 1996. Fundamentals of adsorption.Proceedings of the 5th International Conferenceon Fundamentals of Adsorption. KluwerAcademic Publishers, Boston, MA.

Ruthven, D.M. 1984. Principles of adsorptionand adsorption processes. John Wiley & Sons,New York.

Ruthven, D.M., S. Farooq and K.S. Knaebel.1994. Pressure swing adsorption. VCHPublishers, Inc., New York.

Engineering ethanol ferm

entations

Water reuse in fuel alcohol plants: effect on fermentation 343

Chapter 24

Water reuse in fuel alcohol plants: effect on fermentationIs a ‘zero discharge’ concept attainable?

W.M. INGLEDEW

Applied Microbiology and Food Science Department, University of Saskatchewan, Saskatoon,Saskatchewan, Canada

The subject of water relationships in a fuelalcohol distillery is an interesting and complexissue. Stringent emission controls are in partforcing engineers to design and run newerdistilleries to be virtually free of organics in stackemissions and in any water effluents (Bryan,2003). This dictates that most water outputs inthe distillery (except for the cooling towerwaters) are recycled inside the plant, and thatorganic materials in those waters are thereforenot dispersed onto fields or into water bodieswhere they have a BOD (biological oxygendemand) or COD (chemical oxygen demand)impact on the surrounding environment. ‘Zerodischarge’ is a laudable goal in this industry, butsuch design schemes have created significantproblems for microbiological fermentation.Incomplete understanding of such issues led toa comment I first made at the 1999 Fuel AlcoholWorkshop: “A fermentor is not a garbage can”.This comment, made due to concerns in certainfuel alcohol plants, became a catalyst fordiscussions among engineers, biochemists andmicrobiologists associated with thismultidisciplinary industry, and is beginning tolead to extensive analyses of water effluentsfrom all conceivable sites in modern alcoholplants.

The pre-cook section of a generic dry millingfuel alcohol plant is given in Figure 1. It is inthis area or at the fermentor that waters are

reincorporated into the new mash. In Figure 1,input water entering the system via the grain isrelatively easy to quantitate, as the averagemoisture levels of corn or wheat are normallyknown. Well water (part of the makeup waterused in the cook), recycled water such asbackset, carbon dioxide scrubber water, sidestripper bottoms recycle water, and methanatorrecycle water, are all included. Some of thevolatile compounds in these streams, ethanolincluded, would be lost in cooking due to thetemperature employed and vented in part intothe atmosphere. However, a significant portionof the ethanol remains in the pressurized slurryduring the cook and hold cycle, and then flashesoff. The energy and steam vapor content is usedto run the distillation columns.

It should be noted that most of the existingplants do not incorporate biomethanatortechnology, and therefore evaporator condensatewater is added directly to the cook water tank.When biomethanation is accomplished, a sidedischarge line is made available for removal ofsome of the water from the plant when/ifnecessary. If methanator water is discharged, wellwater volumes are increased correspondingly.Minor input streams of much less concern in theplant include addition of enzymes (α-amylaseat the mingler and at liquefaction; glucoamylaseand urea at the fermentor; or ammonia/ammonium hydroxide at the mingler. Yeast and

344 W.M. Ingledew

yeast nutrients are added at the fermentor. Anapproximate mash bill in a dry mill plant is givenin Table 1.

Table 1. An approximate mash bill for 50 tons/day ofmash in a dry mill fuel alcohol plant.

Component in mash Tons of water

Grain 13.2 tons @ 15% 2.0Plant water 5.3*Evaporator condensate 15.9*Flash tank condensate 1.4*Steam ~1.2*Calcium chloride 0.03Wash waters with NaOH etc. Volume not determined*Backset 12.9*Ammonia 0.06Yeast Variable**

*Components that should be monitored for chemical content.**Components that should be monitored for microbial content.

Output water is more difficult to quantitate. Wetdistillers grain, distillers modified wet grains

(DMWG) and distillers dried grains with solubles(DDGS) contribute water output values inrelation to the degree of drying of the solids.Less well quantified is stack moisture resultingfrom drying of the grain. Other outputs includeevaporator syrup (if marketed separately fromwet or dried grains), dryer water emissions, andstack emissions.

The theme of this chapter is that substrate iscontinually used and ethanol and carbon dioxideare continually produced in the system whetherit operates by batch or continuous technologies.All other components, either unused in themedium or made by the yeasts (or bycontaminants in the fermentation) are recycled.Some waters that recycle are continuallyincreasing in concentration of the compoundspresent until the most inhibitory of theseeventually affects fermentation rate by slowingthe growth rate of the yeast, killing the cells oraltering yeast metabolism. This leads to loss ofethanol concentration, reduced yields, sugar(substrate) ‘bleed’ into the beer well and a generallack of performance of the plant. It should be

Figure 1. Water inputs and outputs from the pre-cook system of a typical dry mill plant (numbers are in US gallons per minute).

Carbondioxide

scrubberwater

Discharge?

Cook

water

Well

water

Corn

flour

Backset

Side

stripper

water

20

56

80

4

13

55

1.5

90 40

290 gpm

Biomethanator (if any)

Mash mingler Slurry

NaOH wash water

Evaporator condensate

Cook systemfor a plant producing 20 million gallons/year


noted that corn wet mill plants – for examplethose that use starch slurry, corn steep liquorand backset – have the same generalconsiderations as dry mill plants. A completeand accurate audit of water in and water out will,however, have to await a thorough examinationof all systems in selected facilities that can thenbe used as models.

In most plants, the volumes of water, water-based chemicals and the water contents ofby-products can be read off computer screensor estimated by the weight of the products andanalytical data. What these plants lack, however,is an exact quantitation on composition of thesestreams, and the variability of contents andvolumes over time, as well as information onspecial purges of water such as seen with sodiumhydroxide cleaning solutions or sanitizers. Amajor concern with process input streams is toxicor inhibitory components of the streams thataffect yeast and therefore the rate and extent ofethanol manufacture. Sodium ion and othercomponents will be discussed later in thiscontext. These considerations put in the contextof continuous fermentation systems are evenmore problematic. Purges of waste streams ofundesirable content into fermentors can createan upset to continuous fermentation trains,causing cascading changes from the upstreamsite of entry throughout the rest of the train.Moreover, the maximum concentration of suchchemicals is always in the fermentor with themost crucial need for ideal conditions – the firstfermentor (or pre-fermentor) – where growth ofyeast is an absolute requirement. More than aweek may be required to regain the balancedgrowth conditions characteristic of continuousfermentation. It is no wonder then that someplants never really operate in a constant orconsistent mode (in balanced growth) for anyextended time!

Some of the input components with potentialproblems for fermentation are listed and detailedbelow. Compositions and variabilities of thecomponents found in each recycled source areall not yet well understood.

Light corn steep water

A representative composition of steep water fromthe wet milling of corn is recorded in Table 2.

Steep water is used as the primary source ofnutrients in fermentation of the ~40% starchslurry obtained as a by-product in corn wetmilling plants. Although used only once intheory, steepwater residues recycle in thin stillageor backset, and this thin stillage thereforecontains all components that were not used byyeast or that were produced in fermentation butnot distilled out of the system. During eachrecycle, the concentrations of the materialsrecycled will increase. Note that Table 2 doesnot provide ion analyses.

Table 2. Composition of light steep water.

Characteristics

pH 3.98Specific gravity (density meter) 1.0527

(13.7 g/100 mL)Dry weight, g/100 mL 11.8N (Kjeldahl method), g/100 mL 0.80α-amino N, mg/L 756*Total amino N, mg/L 1035*Carbohydrate, g/100 mL 5.0*Lactic acid, g/100 mL 3.82*Acetic acid, g/100 mL 0.21*

*HPLC analysisThomas and Ingledew (unpublished data)

Table 3 provides an estimate of the amino acidcontent in unused corn steep liquor, andtherefore the value of this component in thenutrition of yeast. Yeasts only use amino acids,small peptides (<3 amino acids long),ammonium ion, and urea as nitrogen sources.The free amino acids are preferentially utilized,but in a distinct order.

Table 3. Free amino acids in light steep water.

α-amino acid nitrogen (µmole/mL)

Asp 0.93 Tyr 0.94Glu 1.57 Val 3.91Ser 5.22 Met 1.90Gly 2.62 Cys 0.19His 0.73 Ile 1.98Thr 4.03 Leu 8.5Ala 11.68 Phe 3.08Arg 2.03 Trp 0.57Pro nd Orn 1.38

Lys 2.74

Thomas and Ingledew (unpublished data)

346 W.M. Ingledew

These data show that the amino acid spectrumin this light steep water is broad and varied,providing good amino acid nutrition to the mashin which it is used (note that a major dilution ofall compounds would be made on addition ofthe corn steep water to the starch slurry used inwet mills to make the mash). Steepwater isnormally used at about 12% of the mash volume(Table 4).

Table 4. Approximate makeup of mash in a 1000 gpmmash continuous fuel alcohol distillery, showing majorinput waters and the ~40% starch slurry.

Components Gallons per minute

Starch slurry 252Stillage 415*Stripper bottoms 68*Steepwater 123*Enzymes, ammonia or urea, soda ash 5Well water 138*Yeast (depends on balance of the process) ??**

*Components that should be monitored for chemical content.**Components that should be monitored for microbial content.

Usable nitrogen is deficient in all grain samplesthat could possibly be used for fuel alcoholproduction worldwide (Table 5). Nitrogendeficiency is a main cause of stuck and sluggishfermentations in this industry. For this reason,virtually all fuel alcohol plants add liquid or solidurea, ammonia, or ammonium hydroxide,although the amounts may not be optimized.

Table 5. FAN levels in typical mashes from grains usablefor fuel alcohol production.

Grain Free amino nitrogen (mg/L mash*)

Total FAN Usable FAN

Wheat 82 64Barley 84 62Hulless barley 124 100Oats 193 159Hulless oats 184 130Rye 103 83Molasses 267 141Corn 70 58Starch slurry ~ 0 ~ 0

*All mashes normalized to 22% total solids for comparison.Thomas and Ingledew, unpublished data.

Addition of extra usable nitrogen has beenshown to improve the growth of yeast and therate of fermentation of mashes (Ingledew, 1995;1999). More importantly, higher concentrationsof alcohol can then be made. This is the basicprincipal behind very high gravity (VHG)fermentation technology (developed in this lab)being used today in the fuel alcohol industrywhere concentrations of ethanol well above 16%v/v can now be made under routine conditionsin a fuel alcohol plant.

The quantity of lactic acid in corn steepwateris noteworthy. It originates from the bacteria thatare believed to play an important part in the wetmilling of corn – bacteria which grow andmetabolize during the steeping process. Aftersome recycling of thin stillage throughsequential fermentations, the lactic acid levelscan become significant, inhibiting yeast growthand metabolism (Narendranath et al., 1997).Lactic acid content limits the amount of cornsteep liquor (and perhaps also thin stillage)possible in fermentation mash makeup, as dothe foaming properties of these streams (leadingto entrainment of liquid in the carbon dioxideenvironment in the headspace), which has beenknown to cause ‘blowing’ of pressure reliefvalves or plates on the tops of fermentors,creating an immediate and dangerous carbondioxide problem for personnel in the plant.

Thin stillage

Thin stillage is used in both dry and wet millplants. Although it varies from plant to plant, itis generally low in utilizable nitrogencompounds and in total nitrogen. In fact, thereare few, if any, positive reasons (other than waterrecycling) to use thin stillage. Table 6 providesan analysis of thin stillage along with a groupedestimate of the major minerals in the corn (P, S,K, Ca, Mg, Cu, Fe, Mn, and Zn only). Phosphoruslevels ranged from 815 to 1762 mg/L whilepotassium ranged from 705 to 2643 andmagnesium levels in all stillages examined were200-721 ppm. Other elements were much lower,including Zn, Fe, Cu, Mn. Sodium ion was notmeasured at that time.

The lactic acid concentration in a wet millalcohol plant can also be problematic. In theseplants, it can be shown that the concentration of


lactic acid contributed by the light steep waterand the stillage can easily exceed the 0.8% (8g/L) minimal concentration reported to affectyeast (Narendranath et al., 1997; 2001a; 2001b).In this respect, the valuable contribution madeto yeast nutrition by light steep water issomewhat negated by the lactic acid present init and in thin stillage – acid which inhibits thegrowth and metabolism of yeast. The extent ofthis contribution is shown in Table 7.

Table 6. Proximate composition of corn thin stillage.

Components g/L

Total solids 48Total nitrogen 0.53Crude protein 3.0Free amino nitrogen (mg/L) 75Glucose 0.60Lactic acid 11pH 3.7Minerals 3.64

Jones and Ingledew, 1994

Table 7. Lactic acid concentrations in fermentors in wetmills entering the system from stillage and steep water.

Mash constituent L/1000 Lactic acid Final mashL mash (g/L) concentration

(g/L)

Light steep water 134 38.2 4.7Stillage < 415 11.0 4.57Total 9.27

Again, a complete audit of water composition(including recycled waters) in all areas of thedistillery would allow us to determine the exactinfluence that each water stream (with its ownparticular chemical composition) might have onthe fermentation process. As not all of the abovematerials have been repeatedly analyzed over abroad range of times and conditions so that thesignificance of their contents is known, thischapter can only discuss what we do knowabout some of the water components.

A few years ago, we analyzed six samples ofbackset from batch fermentation plants. Half thesamples were from wheat plants and half fromcorn plants. These results are shown in Table 8and were published by Jones and Ingledew,1994.

Table 8. Backset samples from six fuel alcohol plants.

Constituent Wheat Wheat Wheat Corn Corn Corn

Total solids, % w/vSolubles 4.6 6.3 3.4 4.5 2.8 2.4Insolubles 8.0 7.8 2.3 3.2 2.0 2.0

Total N, % w/v 0.18 0.11 0.13 0.03 0.05 0.54FAN, mg/L 96 79 51 118 75 88Glucose, mg/mL 1.6 33 1.0 0.7 0.6 0.2Lactic acid, g/L 11 17 1.5 7 11 6pH 4.2 3.6 4.7 4.7 3.7 4.0Minerals, mg/L 5648 2085 2893 4149 3626 2946

A number of differences were noted among thesesix samples chosen at random from industry. Notethe levels of lactic acid, which we know beginto exert a strong effect on yeast at 8 g/L. Notealso that the free amino nitrogen values (FAN)are quite low, because any usable FAN in thestillage had already been made available to theyeast in one or more prior fermentations. FAN isa measure of the free amino groups in aminoacids, small and large peptides and solubleproteins. As yeasts are unable to use any peptideslarger than a tripeptide (and the latter only slowly,Patterson and Ingledew, 1999; Ingledew andPatterson, 1999), there will always be residualFAN in the backset, but there is little usable FAN!This is the basis of the comment madepreviously that the quality of FAN (and thus thequality of backset) is of less value than many inthe industry would believe.

The glucose content of these backset samplesis also interesting. It can be concluded that onewheat plant was having considerable problemsin fermentation at the time of the backsetprovision. Over 3% glucose was found in thisbackset; and this plant would have been sufferinga loss of ethanol yield of about 15 g/L, a darkerDDG after drying, and a reduced level ofnitrogen in reused backset caused by Maillardreaction complexing of sugars with nitrogensources during heating of the subsequent mashin the jet cooker.

As noted above, the concentrations of chemicalcomponents in thin stillage will increase as thebackset recycles through successivefermentations. This was well recorded by Chinand Ingledew (1993; 1994) where a gradualincrease in lactic acid to 1.4% was seen in infectedmashes and where losses in yeast viability of up

348 W.M. Ingledew

to 60% were shown. This occurred late infermentation after ethanol was totally made tothe rather low alcohol levels of 7-7.5% v/v pre-determined by the mash used in Chin’s work.

The impact of backset on the fermentation isbecause it contains any bacterial end productsthat are not removed in distillation (lactic acidand glycerol, for example), ions that were nottaken up by the yeast in the previousfermentation(s), unused solubles from theoriginal mash, insolubles from previous mashes,if any, and small amounts of volatiles which werenot completely distilled. As the backsetrepeatedly recycles, the concentrations of someof these materials increase and some can becomeinhibitory.

Evaporator condensate water

Evaporators have been extensively used in thefood industry in order to remove water from afluid to obtain a more stable or reduced volumeproduct, which is often later dried using otherequipment. The condensed water from thisprocess is hot, and can contain small amountsof the product, which during evaporation aredistilled over in the concentrate as tiny particlesin the mist. Therefore, evaporator concentrateis only pure in relative terms. It contains smallamounts of a number of organic materials thatare able to support microbial growth if theenvironment is conducive for the microbesbound to be present. In the food industry, thewater can be recovered by reverse osmosis(purification of a water stream by forcing it underpressure through a membrane not permeable tothe impurities it contains), providing savings inenergy, water purchase, water processing ondisposal, and softened water suitable for plantuse. In fuel alcohol plants, the evaporatorcondensate is not processed by reverse osmosis.It is instead recycled to makeup water (withoutfurther processing) at the mash ‘mingler’, withthe idea that a heat process in gelatinization/liquefaction will sterilize the mash and kill anymicroorganisms. Little thought is given to thefact that microbial end products may already bepresent, and that they will survive the up-frontprocessing of substrate. On arrival at thefermentor, they can be antagonistic to the yeast.Evaporator condensate components (those

measured in the plants that collect data) arefound in Table 9.

Table 9. Analytical data on evaporator condensate.

Component* g/100 mL

Ethanol ND**Methanol (or compound with same mobility)*** 0.169Sugars DP3+ (MW of maltotriose and larger) 0.017 DP2 (MW of maltose) 0.013 DP1 (MW of glucose) 0.065Acids Lactic ND Acetic 0.037 Succinic NDGlycerol 0.015

*2 samples analyzed.**Not detected. The heat of the process strips the ethanol if any remained. Note that other components present may not have been detected due to use of the routine fermentation column (Aminex HPX-87H).***May in part or in whole be 2, 3 butanediol.

Carbon dioxide scrubber water

Carbon dioxide exiting fermentation is scrubbedby counter-current flowing water that removesethanol and other compounds as well as anyparticulate materials entrained in the gas. Thewater is recycled - normally to slurry the corn –and is a substantial part of water input. Thisrecycled wastewater stream can containsignificant ethanol (Table 10) which, if heatedtoo high in the mingler or slurry tank, would belost to the alcohol plant. Some is lost inliquefaction, but an estimated 80% of the ethanolremains with the slurry during cooking andflashes off (to distillation towers) with the watervapors, or, in some plants, at the cooling tower.Some of the components identified in carbondioxide scrubber water are shown in Table 10.

Side stripper bottom water

Side strippers handle the rectifier bottoms, andare engineered to strip the last concentrationsof ethanol away from insolubles and less volatileresidues exiting the rectifier. If the side stripper


is engineered with direct steam injections (ratherthan as a reboiler), the side stripper bottoms willcontain small amounts of organic residues. Table11 shows the composition of compounds so faridentified in side stripper bottoms. This wateralso shows the presence of a number ofcomponents that have been entrained in thewater processed through the stills. The amountsare not large. It should be noted that theassessment of this sample has been done to dateonly with the Aminex column and refractiveindex HPLC (system commonly used forfermentation monitoring). The possibility forother compounds like fusel alcohols, and otherorganics not measured or identified by thiscolumn is not ruled out.

Table 10. Components of CO2 scrubber water.

Component* g/100 mL

Ethanol 2.52Methanol (or compound with same mobility) 0.115Sugars DP3+ (MW of maltotriose and larger) 0.044 DP2 (MW of maltose) 0.011 DP1 (MW of glucose) 0.041Acids Lactic ND** Acetic ND** Succinic ND**Glycerol 0.006

*5 samples analyzed**Not detected

Table 11. Components of side stripper bottoms.

Component* g/100 mL

Ethanol 0.132Methanol (or compound with same mobility) ND**Sugars DP3+ (MW of maltotriose and larger) 0.055 DP2 (MW of maltose) 0.022 DP1 (MW of glucose) 0.040Acids Lactic 0.013 Acetic 0.019 Succinic 0.002Glycerol 0.035

*22 samples analyzed**Not detected

Interestingly, side stripper water also has beenfound to contain mineral content ranging from

9 ppm sodium (Na) to over 20 ppm phosphorus(P) and sulfur (S). Other ions including calcium(Ca), iron (Fe) and magnesium (mg) were allbelow 0.016 ppm. Extensive analyses have notbeen available to this author.

Implications of wastewater on yeastfermentation

Consideration of the inhibitory compoundsfound in the above recycled waters led us toconsiderable research and a diagram of thestresses that yeasts must endure. We now knowthat many of the stress factors act in a synergisticmanner when more than one agent is found atinhibitory concentrations (Narendranath et al.,2001). We also know that the presence of solidsin the medium and the medium bufferingcapacity both have an influence on the inhibitionof the growth of yeast by organic acids (Thomaset al., 2002), as does the temperature of thefermention. Figure 2 (Ingledew, 1999) showsthat a number of possible stress agents find theirway into fuel alcohol fermentors via backset andby recycling of other wastewater streams withinthe plant. When concentrations increase to levelsnear or above the levels indicated (as singleinhibitors), the yeast will not grow or metabolizeat the same rate as is possible under idealfermentation conditions (Thomas et al., 2001).In some cases, the levels of compounds such asacetic acid, lactic acid and ions like sodium areso high in fermentation that normal ethanolproduction is retarded. The immediateconsequence in both continuous and batchfermentation is that unused sugars are sent tothe beer well. This ensures that lower ethanolyields than those expected under idealconditions are seen.

The concentrations of stress agents listed inFigure 2 have, for the most part, been determinedas accurately as possible. Unfortunately, whenmedium conditions (mash type andconcentration, pH, presence of additionalstresses) change, so will the level of each‘inhibitor’ examined that affects yeasts. We nowknow that the composition of the medium (i.e.corn mash) and the environmental conditionsimposed (pH, temperature etc.) will influencethe concentrations of any one inhibitor, whichcan affect yeast growth and metabolism. In fact,

350 W.M. Ingledew

inhibition by lactic acid and acetic acid is pHdependent (Narendranath, unpublished; Thomaset al., 2002). As the knowledge of synergisticeffects among inhibiting compounds and themagnitude of the effect of concentration allimprove with further research, this stressdiagram will be amended. The values shown arethose we believe at present to be the minimumlevels where effects on yeast can be measured.

Biomethanators

A more recent addition to the technology of fuelalcohol production is the anaerobic biomethanator.A process flow diagram of the ICM/Phoenixwastewater treatment system is reproduced inFigure 3 (Anon., 2003).

A biomethanator, in part, is a special type offermentor, designed in this case to reduce by90% the variable BOD/COD levels in evaporatorcondensate and other wastewaters. It producesmethane and CO2, the former of which is usedto reduce boiler or dryer operation costs. Sulfurcan also be removed from treated wastewatersby conversion to hydrogen sulfide. This is insome respects an exciting innovation with great

potential to the industry: recycling treated waterback to fermentors.

The anaerobic bioreactor in the biomethanatormust be ‘fed’ a mix of nutrients needed tocontinue life of the bacteria on which thebiomethanator is based – providing the food andenvironmental conditions that the bacterialpopulation in the sludge require to grow. Inaddition, pH is adjusted with tank wash watersrich in sodium ions (used as NaOH for tank andline cleaning). Much of this sodium ion isconverted to sodium carbonate (Na2CO3) by thecarbon dioxide in the system. Regardless, thesodium ion, all unused nutrients, and all endproducts produced from the biomethanator thatwere not converted to methane and carbondioxide move through the biomethanatorunabated. This results in high levels of sodiumion and perhaps a series of yet-to-be-identifiedcompounds in subsequent fermentors. It isuncertain, as yet, whether the levels of sodiumexperienced in plants both with and withoutmethanators are high enough to inhibit yeasts.Evidence accumulated in this lab, however, usingcorn mash similar to that made in dry mills,would indicate that large amounts of sodium ionwould lead to a purging requirement to eliminate

Figure 2. Stress factors that influence growth and metabolism of yeasts. Much of this work was done with a synthetic fermenta-tion medium and may represent the smallest values of a single stress factor that is inhibitory to yeast.

25 - 35°C*

23% but less

for growth

>.05% w/v

>500 mg/L

3.0 to 4.0

>0.8%

38% w/v

>100 mg/L

(varies with strain)

*Depends on solids and stressYeast membrane

Acetic acid

Effects ofmycotoxins and other ions

Ethanol

Carbohydratelevel (not sugar)

Temperature

Sodium ion

Sulfite

pH

Lactic acid

MASH


ion buildup in water that would otherwise re-enter the plant via the methanator.

The levels of sodium needed to inhibitlaboratory fermentation of corn mash when noother stressful agents are present would appearto be between 2500 and 5000 ppm (Figure 4).This should allow plenty of leeway for the useof methanator effluent in fermentation (at leastfrom the sodium standpoint). It should not beforgotten, however, that when more than onestressful agent is present, the effects aresynergistic, and at this point, much lowerconcentrations of sodium (near 500–600 ppm)have been seen to have an effect on the yeast.

When biomethanators are not used, it iscommon practice to use a combination ofevaporator condensate water, carbon dioxidescrubber water, side stripper bottoms water, andwell water for the cooking process. Note thatsodium hydroxide washwater is also unevenlyrecycled to fermentors in volumes averagingabout 0.5% of plant flow rate. It is this volume

Figure 3. The ICM/Phoenix biomethanator.

Anaerobicbioreactor

Effluent

Recyclepump

Blower

Gasexchanger

Recycletank

NutrientsIron

Caustic

Systemfeed

pumpSump

Biogas scrubberH2S removalBiogas

Air scrubberH2S removal

Biogas to boiler, burner,flare or hot water heater

Equalization tanks

Waste water

High COD >5,000 mg/L

TSS <600 mg/L

air, CO2, H2S

Te

FM

pHc

Tc

FC

pHc

of waste sodium ion that is potentially dangerousif added to the cook or directly to fermentors.The disposal of sodium ion even through thedryers via incorporation into DDG may requireassessment.

Anecdotal evidence received from productionplants in the past (prior to methanatortechnology) indicated that levels of sodium ionabove 500 mg/L would inhibit yeast growth andfermentation; and this was published in the stressdiagram (Figure 2) found in the last edition ofThe Alcohol Textbook (Ingledew, 1999). In thepast, evidence published by Maiorella et al.(1984) showed that a sodium concentration of~5500 mg/L sodium was required to demonstratenegative effects on yeast as manifested by >20%inhibition. Moreover, Watson (1970) showedthat in batch culture, 0.25 to 1.5 M NaCl (5750to 33700 ppm Na) was needed to affect growthrate and the yield of cells grown on glucose.Jones and Greenfield (1984) stated that for mosttolerant strains of yeast, total inhibition of growth

352 W.M. Ingledew

occurs above 1-2 M (23,000-46,000 ppm Na),the level needed for non-sensitive yeasts. Forsensitive strains, 0.1 M or 2300 ppm wasneeded. These latter values are in similar rangeto Figure 4. Insensitivity to sodium is said to bevia the diversion of energy needed to maintainan electrochemical gradient of sodium ionagainst passive or facilitated diffusion into thecells. The effect was not the same with differentyeasts measured. Explanation of the discrepancybetween the numbers that would definitely beinhibitory and literature values may be found inthe work of Jones and Gadd (1990), whoindicated that the ratio between sodium andpotassium is very important; and that at a pH of5.0 when the Na:K ratio is about 20, potassiumuptake is inhibited by 70%. At pH 4, sodiumuptake becomes negligible; so they indicated thatin an industrial medium containing high levelsof sodium, a pH around 4.0 would be requiredto overcome the deleterious effects of thesodium ion. The optimal concentrations of K+

ion of 2-4 mM may be in part due to the presenceof sodium in mash.

The differences seen in an industrial plantbetween organic acid problems (caused bybacteria and recycle of stillage) and ionicproblems (probably caused by recycle of NaOHwash waters by stillage and biomethanator

action) are shown in Table 12 in comparison tocontrol fermentations where inhibitory levels ofneither type of inhibitor were present.

Table 12. Plant data demonstrating the effects of bothorganic acids and sodium ion on batch fermentations.

Tank Ethanol Glucose Lactic Acetic Sodium Ethanolno. acid acid loss

(% w/v) (% w/v) (% w/v) (% w/v) (ppm) (%)

A 12.4 0.03 0.35 0.02 <290 -B 12.8 0.24 0.29 0.04 <290 -H 10.8 4.70 1.0 0.32 <290 14.3K 9.5 5.14 1.0 0.41 <290 26.2M 11.5 1.40 0.25 0.0 660 8.7N 10.8 1.29 0.17 0.01 720 14.3

It can be seen from Table 12 that under normaloperational conditions (A and B), this alcoholplant was capable of making over 12% alcoholunder conditions where lactic and acetic acidswere low and sodium was not present at alarminglevels. When organic acids were present atinhibiting levels (H and K), ethanol yieldsdropped, sugar ‘bleed’ at end fermentation washigh, and the acids circulated into otherfermentors in backset. This evidence stronglysupports the synergistic or at least concertedinhibitory effects of lactic and acetic acids. Underconditions of high sodium in Tanks M and N,

0

5

10

15

20

25

0 50 100 150 200

0 ppm Na

2500 ppm Na

5000 ppm Na

7500 ppm Na

10000 ppm NaD

isso

lved

so

lid

s(g

/100m

L)

Time (hrs)

Figure 4. The effect of sodium ion as a single stress factor on the utilization of sugar during fuel alcohol production(Hynes et al., 2003, unpublished data).


where the source of sodium was from tankwashings recirculated with backset, ethanollevels fell by 1-2 g/100 mL, and glucose levelsrose accordingly. Sodium levels of >500 mg/Ltherefore appeared to significantly lower ethanoloutput of this plant. The magnitude of ethanolloss in these examples demonstrates theproblems a plant can experience when there isno opportunity to remove recycled thin stillagefrom the system in order to lower the levels ofrecycled chemicals and preserve the delicatebiological conditions in the fermentor. It is theauthor’s opinion that methanators can lead tosuch elevated levels of ions (but not organicacids) unless there are undetermined toxicbioproducts of methane-producing microbesthat find their way back into the fermentationsystem. This is as difficult a situation to monitorand rectify, as is the present case with continualrecycle of backset.

Conclusions: the importance ofelimination or preventing buildup ofinhibiting chemicals

It has been estimated that the fuel alcoholindustry in general loses from 1 to 4% ofpotential alcohol volume each year. The NorthAmerican industry produces at this writingalmost three billion gallons per year. A 1% lossis 30 million gallons (113 million liters), with avalue of approximately $33 million. A 1% lossin a plant such as the one in Table 12 wouldonly be 0.12% alcohol. Such a loss is unlikelyto be noticed due to variability in chemicalquality control assays and the starch levels inincoming substrate. A 4% loss over the wholeNorth American industry is 120 million gallons(452 million liters) with an approximate valueof $132 million. In certain plants (sometimesunder particular but not infrequent circumstancesas seen in Table 12), losses greatly exceed thesepercentages. A major cause of such losses isbacterial degradation of sugars that should havebeen converted by yeast to ethanol. All effortsshould therefore be made in plants to avoidbacterial action. The action of bacteria or wildyeasts, or the effects of their end products ofmetabolism, which remain after heat or otheragents kill the microbes that made them, aresignificant causes of yield loss. One must not

forget, however, that chemicals outside of thosemade by microbes can also affect the process –sodium ions from tank and line washings are acase in point.

The message for the ethanol industry is that‘zero effluent’ plants (those which do not releasesignificant volumes of water with high BOD orCOD or high concentrations of salts) are noteasily designed – especially when the primemode of water recycle is via the fermentationsystem. Fermentors (biologically fragile reactorsystems) are susceptible to all chemicals addedto them. Engineering to save on water usage iscontrary to the preservation of the physiologicalconditions required by the yeast to ensurecomplete and rapid fermentation.

References

Anonymous, 2003. The ICM/Phoenixbiomethanator. http://www.phiosystems.com/experience.htm.

Bryan, T. 2003. Green Machines. EthanolProducer 9(7):12-15, 52, 55-56.

Chin, P.M. and W.M. Ingledew. 1993. Effect ofrecycled laboratory backset on fermentationof wheat mashes. J. Agric. Food Chem.41:1158-1163.

Chin, P.M. and W.M. Ingledew. 1994. Effect oflactic acid bacteria on wheat mashfermentations prepared with laboratorybackset. Enz. Microb. Technol. 16:311-317.

Ingledew. W.M. 1993. Yeasts for production offuel ethanol. In: The Yeasts Vol. 5. 2nd Edition.Academic Press, N.Y. pp 245-291.

Ingledew, W.M. 1995. Chapter 7. Thebiochemistry of alcohol production. In: TheAlcohol Textbook. Second Edition. (T.P. Lyons,D.R. Kelsall, and J.E. Murtagh, eds).Nottingham University Press. Nottingham, UK.pp. 55-79.

Ingledew, W.M. 1999. Chapter 5. AlcoholProduction by Saccharomyces cerevisiae: ayeast primer. In: The Alcohol Textbook. ThirdEdition.(K. Jacques, T.P. Lyons and D.R.Kelsall, eds). Nottingham University Press.Nottingham, UK. pp. 49-87.

Ingledew, W.M. and C.A. Patterson. 1999. Effectof nitrogen source and concentration on theuptake of peptides by a lager yeast in

354 W.M. Ingledew

continuous culture. J. Amer. Soc. Brew. Chem.57:9-17.

Jones, R.P. and G.M. Gadd. 1990. Ionic nutritionof yeast – physiological mechanisms involvedand implications for biotechnology. Enz.Microb. Technol. 12(6):402-418.

Jones, R.P. and P.F. Greenfield. 1984. A reviewof yeast ionic nutrition. Part 1: Growth andfermentation requirements. Proc. Biochem.19(2):48-60.

Jones A. and W.M. Ingledew. 1994.Fermentation of very high gravity wheat mashprepared using fresh yeast autolysate.Bioresource Technol. 50:97-101.

Maoirella B.L., H.W. Blanch and C.R. Wilke.1984. Feed component inhibition in ethanolfermentation by Saccharomyces cerevisiae.Biotech Bioeng 26:1155-1166.

Narendranath, N.V., S. H. Hynes, K.C. Thomasand W.M. Ingledew. 1997. Effects oflactobacilli on yeast-catalyzed ethanolfermentations. Appl. Environ. Microbiol. 63:4158-4163.

Narendranath, N.V., K.C. Thomas and W.M.Ingledew. 2001a. Effects of acetic acid and

lactic acid on the growth of Saccharomycescerevisiae in a minimal medium. J. Indust.Microbiol. Biotechnol. 26:171-177.

Narendranath, N.V., K.C. Thomas and W.M.Ingledew. 2001b. Acetic acid and lactic acidinhibition of growth of Saccharomycescerevisiae by different mechanisms. J. Amer.Soc. Brew. Chem. 59:187-194.

Patterson, C.A. and W.M. Ingledew. 1999.Utilization of peptides by a lager brewing yeast.J. Amer. Soc. Brew. Chem. 57:1-8.

Thomas, K.C., S.H. Hynes and W.M. Ingledew.2001. Effect of lactobacilli on yeast growth,viability and batch and semi-continuousalcoholic fermentation of corn mash. Journalof Applied Microbiology. 90:819-828.

Thomas, K.C., S.H. Hynes and W.M. Ingledew.2002. Influence of medium buffering capacityon inhibition of Saccharomyces cerevisiaegrowth by acetic and lactic acids. Applied andEnvironmental Microbiology 68:1616-1623.

Watson, T.G. 1970. Effects of sodium chlorideon steady-state growth and metabolism ofSaccharomyces cerevisiae. J. Gen. Microbiol.64:91-99.

Understanding energy use and energy users in contemporary ethanol plants 355

Chapter 25

Understanding energy use and energy users in contemporary ethanol plants

JOHN MEREDITH

President, Ro-Tech, Inc., Louisville, Kentucky, USA

Results of calculations quantifying the thermaland electrical energy needed to convert starchand extract ethanol from grain are oftencontroversial and depend on the assumptionsbehind the calculations. Looking through theliterature and surveying historical operating costsfor beverage distilleries and ethanol plantsreveals a wide range of results reported forenergy use per gallon of ethanol produced. It isalso noteworthy that total energy use per gallonof ethanol produced in beverage plants in thelate 1900s was significantly greater than theenergy use per gallon in today’s new fuel ethanolplants.

Terms defined and assumptions forenergy use calculations

For purposes of this chapter, the criteria belowwill apply to the calculations and conclusionswhich follow:

Gallon of denatured ethanol. One US gallon ofethanol containing 4.76% denaturant and inagreement with ASTM D4806.

Thermal energy. Based on natural gas used at1000 BTU/ft3 at standard conditions (60°F, 1 atmof pressure).

Electrical energy. Kilowatt hours used by theplant. Demand and energy cost components oftypical utility electrical bills are not consideredin this analysis. These two components aremeasured by utility companies and used todetermine monthly electrical bills. This cost isalso a function of a number of factors outsidethe scope of this chapter (plant location, energymarket at any particular time, etc.).

Feedstock. Energy use in ethanol plants isdependent on feedstock. For our purposes,feedstock will be No. 2 corn at approximately15% moisture. Some ethanol plants haveexperimented with using ‘wet corn’ at ~30%moisture as a feedstock. This technique wouldeliminate the energy need for drying the corn to15% moisture at the elevator. However to date,these efforts have been mostly unsuccessful dueto problems in handling and preparing the cornfor cooking.

Other feedstocks require similar energy inputto produce ethanol with some variation. Wheatis virtually identical to corn in energyrequirements except as affected by moisturecontent. Rye and milo (sorghum) require someadditional energy due to higher slurry viscosities,which affect heat transfer and pumping. Sugarcane-based feedstocks use less energy (nodryhouse) compared to grain-based feedstockplants with DDGS facilities.

356 J. Meredith

Boundaries. Energy use for purposes of thischapter starts at grain unloading at the ethanolplant and finishes with product at the point readyto be pumped to a waiting truck or rail car. Somestudies have investigated energy use startingwith corn in the field and finishing with ethanolat a receiving terminal separated from the plant.A recent study by the US Department ofAgriculture stated that ethanol plants deliver 34%more energy to the end user of the fuel than isused in growing, harvesting and distilling corninto ethanol. Ethanol releases 90,030 BTU foreach gallon burned. If 90,030 BTU represents a34% increase, then the total energy required tomake ethanol would be approximately 67,000BTU. Modern plants require approximately36,000 BTU of thermal energy and 1.1 KWH(3800 BTU/gallon) of electrical energy pergallon of denatured ethanol. This totals 39,800BTU/gallon, which leaves over 25,000 BTU/gallon for energy users outside the plantboundaries including farm equipment involvedin planting and harvest, energy to transport thecorn from the field to an elevator for drying andthen to the ethanol plant, energy needed to drycorn to ~15% moisture, and energy expendedin transporting ethanol from plant to receivingterminal.

Despite the focus some place on the overallenergy balance of ethanol production, it is worthremembering that ethanol is not added to fuelfor its energy value. The Clean Air Act of 1990required adding oxygenates to gasoline in somelocations to reduce emissions. Ethanol is anoxygenate, and is compared in Table 1 to otheroxygenates. Methyl tertiary butyl ether (MTBE),the oxygenate of choice until a few years ago,has been found to be a pollutant, which has ledto increased interest in ethanol as a replacement.

Table 1. Comparison of four fuel oxygenates.

Oxygen (%)

Ethanol 34.73Methanol 49.90MTBE (Methanol and isobutylene) 18.15ETBE (Ethanol and isobutylene) 15.66

Energy use for co-product handling, drying andstorage will be included. Although some olderbeverage plants show a net loss in the productionof DDGS when comparing costs to revenues,drying co-products to a DDGS product isconsidered profitable in newly constructed, moreenergy efficient, ethanol plants. Energy use inthe dryhouse is a major component of the plantenergy balance.

Beverage plants versus fuel ethanol plants.While this chapter focuses on energy use in fuelethanol plants, the relationships between energyuse and plant system design apply equally toboth industries. When comparing beverageplants to similarly designed fuel ethanol plants,beverage plants consume more energy per gallonowing to the specifics of producing a definedproduct for human consumption. Distillation isoften more complicated in order to produce aproduct with controlled amounts of congeners.In addition, heat recapture techniques, thickslurries and hot water recycling are not asprevalent due also to quality concerns.

Understanding thermal and electricalenergy usage ratios

BTU (THERMAL ENERGY) PER GALLON OFETHANOL

Modern dry milling mid-sized (40 million gallonsper year) ethanol plants are being constructedin the US based on an expected thermal energyrequirement of 36,000 BTU (natural gas) pergallon of ethanol. This energy requirement isbased on denatured ethanol. In other words, theratio is calculated using a mix of 100 gallons ofundenatured plant product with 5 gallons ofdenaturant (low octane gasoline) yielding 4.76%denaturant in the mix. Some plants look at theratio before denaturing, which would be 36,000BTU/gallon ethanol x 105 gallons denatured/100gallons undenatured ethanol = 37,800 BTU/gallon. Likewise, other plants use ‘steam BTUs’instead of ‘natural gas BTUs’. ‘Steam BTUs’introduce boiler efficiency into the equation asfollows:


ELECTRICAL ENERGY (KILOWATT HOURS)PER GALLON OF ETHANOL

Electrical use in a modern mid-sized plant isexpected to be between 0.7 and 1.2 KWH/gallonethanol. Using 1.1 KWH/gallon as aconservative average for denatured ethanol, andcalculating the value when denaturing isremoved from the equation, the following isobtained:

1.1 KWH 105 gal denatured ethanol 1.155 KWH

gal ethanol gal undenatured ethanol gal undenatured ethanol

x =

RELATIONSHIP BETWEEN THERMAL ANDELECTRICAL ENERGY USE

To further complicate thermal and electricalenergy use statements, the usage ratios can bedramatically changed by plant design techniquesthat transfer energy sources from thermal toelectrical (or electrical to thermal) for certainplant processes. For example, the entiredryhouse evaporator thermal energy load canbe shifted to an electrical load by replacing amultiple effect steam driven evaporator with amechanical vapor recompression type withelectric driver.

Factors affecting energy use andefficiency

MASH DILUTION

Any extra water added to the corn meal slurrysystem before cooking must be heated threetimes in the ethanol plant. Without heat recycle

and recapture, some of the extra water isevaporated twice, requiring over 16,000 BTUfor each gallon of water as it moves through theseprocesses:

• cooking (raising the fluid’s condition from asaturated liquid at ambient temperature to asaturated liquid at approximately 230°F)

• distilling and purifying (changing fluid statefrom a saturated liquid (beer) at 90°F to a partlysaturated vapor at 200 proof and the rest to asaturated liquid at approximately 220°F)

• dryhouse operation (changing conditions froma saturated liquid at 220°F (stillage) to asaturated vapor at 160°F to 230°F)

Modern ethanol plants reduce the water andbackset added to the corn meal in the meal slurrytank to a level acceptable in terms of conversion,yield and general handling characteristics. Thereis, however, a tradeoff between concentratingthe slurry to reduce evaporation costs and thesubsequent increase in enzyme and chemicalcost to maintain the proper fluid viscosity andpH needed for cooking and fermentation.

Keeping the mixture of corn meal and waterin the slurry tank at a ratio of 1 lb corn meal: 2lbs water + backset total provides a concentratedslurry that saves energy and leads to goodconversion and alcohol yield. This 1:2 mixturewill yield an ethanol concentration in the beerof approximately 14% abv. Diluting the mixtureto 1:2.5 gives an ethanol concentration in thebeer of approximately 12% abv. The plantrunning at 12% abv will use as much as 20%more steam in the cooking, distilling anddryhouse operation compared to the plantrunning at beer concentrations of 14% abv.

For denatured ethanol:

36,000 BTU (gas) 0.82 steam BTU 29,520 steam BTUs

gal ethanol 1.0 natural gas BTU gal ethanol

For ethanol before denaturing:

29,520 steam BTU 105 gal 30,996 steam BTUs

gal undenatured ethanol 100 gal undenatured ethanol gal undenatured ethanol

x =

x =

358 J. Meredith

The new high temperature and alcoholresistant yeasts can operate at 95°F and 18%abv (or more) in a fermentor, further decreasingenergy requirements. However, slurry handlingproblems and heat transfer problems increasewith these thicker beers.

NON-ENERGY DEPENDENT YIELD ENHANCERS

Alcohol yield per bushel of corn is calculatedand expressed as follows:

1000 kg corn at 71% starch = 710 kg starch.At 100% efficient conversion 362.8 kg ethanol

would be produced.At 90% efficient conversion 326.5 kg

(718.3 lb) ethanol would be produced.

= 108.8 gallons (6.6 lb/gallon)= 108.8 gallons per 1000 kg corn= 2.77 gallons per bushel (56 lbs) of corn

Typically ethanol plant yields are closer to 2.65gallons/bushel after distillation and finalprocessing. If alcohol yields per bushel of grainincrease due to non-energy related procedures,then thermal and electrical energy ratios willimprove. For example, Alltech’s solid stateenzyme RhizozymeTM increases alcohol yieldper bushel of grain by 10% or more versus theuse of traditional enzymes. This 10% increasein yield per bushel of grain will have a directlyproportional 10% impact on the thermal andelectrical energy ratios before denaturing.

CAPITAL IMPROVEMENTS, EQUIPMENTSELECTION AND ENERGY USAGE RATIOS

The ratios are dependent on capitalimprovements built into plant design that reduceenergy use. Heat recovery, heat interchange,etc. will reduce energy use ratios. Eliminationof the dryer by selling wet distillers grains withsolubles (WDGS) instead of dry distillers grainswith solubles (DDGS) will dramatically improvethe thermal energy ratio. This may not, however,improve the plant profits depending on therelative sale price of WDGS versus DDGS.

HEAT RECLAMATION TECHNIQUES

Listed below are design features presently in use

in ethanol plants. The process designer has achoice in selecting one or more of these featureswith return on investment (operating cost vscapital cost tradeoff) the governing factor.

• Preheat the water used in cooking or userecycled hot water:

- use treated evaporator condensate- use rectifying column bottoms- use distillation column overhead condenser

water- interchange with mash to be cooled after

cooking

• Preheat the beer for distillation:

- interchange with mash to be cooled aftercooking

- interchange with rectifying column bottoms

• Pre-heat or pre-evaporate thin stillage:

- interchange with rectifying columnoverheads

- interchange with molecular sieve overheads

- interchange with evaporator condensate

• Recover the saturated liquid flash from thebottom of the stripping column and returnback to stripping column

• For plants with thermal oxidizers on dryervents

- use regenerative type thermal oxidizer andprovide time for clean up (bake out) in theoperating cycle

- for non-regenerative types, use waste heatboiler on oxidizer discharge and producesteam for the plant

• Interchange molecular sieve inlet and outletstreams

• Use steam turbine drives and use resultinglow pressure steam for distillation anddryhouse systems

Maximizing use of these techniques allows someethanol plants to approach the 36,000 BTU/gallon and 0.7 KWH/gallon ratios noted above.Without these energy saving improvements,thermal energy use will increase dramaticallywith ratios of 75,000 BTU/gallon and higher apossibility.


The

rmal

ener

gy(N

atur

al g

as)

Ele

ctri

cal

ener

gy

Pro

duct

ion

Gra

inR

ecei

ving

and

Mill

ing

Coo

king

and

Liqu

efac

tion

Ferm

enta

tion

Dis

tilla

tion

Mol

ecul

arS

ieve

Ste

am

90%

of

stea

m10

% o

f st

eam

10%

of s

team

70%

of s

team

10%

of s

team

retu

rns

tobo

iler a

ndbu

ildin

gs

5%of

ste

am5%

of s

team

Ele

ctri

cal e

nerg

y (1

00%

)

The

rmal

ene

rgy

(100

%)

Dry

hous

eP

rodu

ctlo

adou

t

G =

Nat

ural

gas

(th

erm

al)

S =

Ste

am (

ther

mal

)

E =

Ele

ctri

cal

P =

Pro

duct

ion

100%

of s

team

EE

E

G

EE

GG

EE

P

G

PP

PP

10%

of

elec

tric

al4%

of

elec

tric

al4%

of

elec

tric

al35

%of

ele

c.

70%

of

gas

30%

of

gas

1% o

fel

ectr

ical

45%

of

elec

tric

al

Note: E

nergy f

or e

mis

sio

n a

batem

ent is n

ot inclu

ded

1% o

fel

ectr

ical

Uti

litie

sB

oile

r, C

hille

r,C

oolin

g to

wer

,B

uild

ing

serv

ices

Fig

ure

1. E

nerg

y fl

ow th

roug

h an

eth

anol

pla

nt.

360 J. Meredith

Distribution of energy use in an ethanolplant

Table 2 details energy use by subsystem in anethanol plant. An overview is depicted in Figure1. Energy use distribution in a beverage plant issimilar. Data in Table 3 were taken from a newgrassroots beverage plant design in 1990.

Table 2. Ethanol plant energy use by subsystem1,2.

Thermal Electrical(% of total) (% of total)

Grain receiving and milling Negligible 10-15Cooking and liquefaction 5-8 1-5Fermentation 1 1-5Distillation/molecular sieve 55-60 1-5Product loadout Negligible 1-5Evaporation and drying (including emissions) 40-45 30-40Utilities 5-7 40-50Buildings 1 1

1Based on data collected by Ro-Tech from plants constructed since 1995.2Emission abatement no included.

Table 3. Beverage plant energy use by subsystem.

Thermal Electrical(% of total) (% of total)

Grain receiving and milling Negligible 16Cooking and liquefaction 7.4 5Fermentation 1.0 1Distillation and product handling 32.1 2Evaporation and drying (including emissions) 53.5 41Utilities 5 34Buildings 1 1

The primary thermal energy use points aredistillation/purification and the dryhousefollowed by the cooking process. Cooking is adirect live steam injection process using acombination of heat, time and enzymes toliquefy the starch. Indirect heat transfer systemshave been tried with little success. Distillationsystems in ethanol plants include stripping andrectifying columns (sometimes combined intoone column) yielding alcohol at a concentrationapproaching the azeotrope of 95.6% ethanol byweight. In most beverage plants, these columns

are heated by live steam while fuel ethanol plantsuse reboilers. Reboilers eliminate beer dilutionassociated with live steam and also permitrecapture of clean condensate heat by the plant’sboiler system.

Purifying the alcohol water mixture afterdistillation from approximately 192 proof to 200proof is accomplished in most ethanol plantsusing molecular sieve technology, whichrequires thermal energy to vaporize the feed tothe sieve and then more energy to superheat thevapor to over 300°F before entering the sieve.Some energy can be saved by sending rectifyingcolumn overhead vapor direct to the sievesuperheater without first condensing therectifying column vapor. However, combiningthe continuous process rectifier with the batchprocess sieves can present control problems forthe ethanol plant operator.

Dryhouses use thermal energy to boil and/ordry the stillage, removing condensed water inthe evaporator and vapor from the dryer stack.Depending on equipment selection, variouscombinations of steam, natural gas, propane orfuel oil are used to provide the thermal energy.Large electrical energy use points (large motors)in an ethanol plant are concentrated in utilityareas (chillers, air compressors, cooling towers,boiler fans). If the evaporator is a mechanicalvapor recompression system, then the dryhousebecomes a major electrical user.

The grain handling and milling areas will usesignificant amounts of power depending on useof pneumatic systems as well as the complexityof the grain receiving, storage, transfer andmilling systems. However, whereas the modernfuel ethanol plant approaches energy usageratios of 36,000 BTU/gallon and 0.7 KWH/gallon, the 1990 beverage plant in Table 3 hada similar electrical ratio but the thermal ratio was73,000 BTU/gallon (calculated as one gallon of200 proof ethanol).

Overall ethanol plant energy balance

Energy inputs and energy exits for an ethanolplant are listed in Table 4. Chemical and physicalproperties (including temperature, pressure andspecific heat) of each item must be known tocalculate the proper enthalpy (BTU/lb) associatedwith entering or exiting energy.


Conclusion

Modern ethanol plant designers are faced witha variety of choices when selecting processsystems and process equipment. After thesystems have been specified and selected,energy usage ratios can be projected.Conservative beverage plant designs have along track record of dependable operation but

Table 4. Energy inputs and energy exits for an ethanol plant.

Energy entering Energy leaving

Corn Denatured ethanol, including heads and fusel oils

Make up water Carbon dioxide

Fuel Wastewater emissions to land application, outfall (stream), or sewer (from make up watertreatment, from process (evaporator, CIP, other), from boiler, scrubbers and cooling tower

Electricity blowdowns)

Denaturant Air emissions and waste heat to atmosphere (from dryer or oxidizer, tank and vessel vents, coolingtowers, baghouses, from thermal losses through exterior vessel walls and building walls, from

Yeast exhaust air (buildings)

Enzymes Waste solids

Make up air Wastewater sludge, CIP screenings, corn screenings (‘overs’) (boiler, dryer, pneumatic

systems, building space) Co-products

Chemicals(for CIP, fermentation nutrients)

often pay an operating cost penalty with thermalratios approaching 75,000 BTU/gallon.Aggressive energy-reducing designs for modernfuel ethanol plants produce ratios approaching36,000 BTU/gallon of ethanol and 0.7 KWH/gallon of ethanol. However, the cost of capitalimprovements to reach these numbers must bejustified; and the reliability of the resulting capitalequipment must be proven.

The dryhouse, co-products and the future

Dryhouse design: focusing on reliability and return on investment 363

Chapter 26

Dryhouse design: focusing on reliability and return on investment

JOHN MEREDITH

President, Ro-Tech, Inc., Louisville, Kentucky, USA

Introduction: an overview of distilleryco-products

Production of ethyl alcohol by fermentation ofnaturally occurring starch or sugar-containingagricultural products is a well-established andcomplex industry today. A variety of feedstocksare utilized, however corn, wheat, rye, milo(sorghum) and sugar cane are the major sourcesof fermentable sugar.

Depending on the industry and application,co-products from these feedstocks include:distillers dried grain with solubles, distillers driedgrain without solubles, distillers solubles, wetdistillers grains with solubles, corn oil, yeastcells, thick wet stillage, thin wet stillage,centrifuge cake, evaporator syrup, corn glutenmeal, corn gluten feed, starch, sugar, carbondioxide, and fusel oils.

Two processes based on corn as a feedstockhave evolved over the last few hundred years.Wet milling, dating back to the mid 1800s,requires wet steeping of the corn kernel beforeprocessing followed by a series of unitoperations aimed primarily at starch recovery.Dry milling, which dates back hundreds of years,requires grinding or ‘milling’ of the corn kernelinto a flour before further processing. Processesgenerating co-products from the dry mill willbe the focus of this chapter.

After fermentation in a dry mill corn feedstockplant, ethyl alcohol is recovered by distillation.Depending on the ethyl alcohol product type(fuel grade ethanol, beverage spirit includingneutral spirits, bourbon, gin, etc.), distillationsystems and equipment vary. However in allcases, the spent residue from the distillationprocess, commonly called whole stillage, is thesource of the dry mill co-products noted inFigure 1. Whole stillage normally passes over ascreen or through a liquid/solid separationdevice and the thin cut (backset) is returned tofermentation. The remainder is collected as eitherthick stillage or additional thin stillage and cake.

Chemical and physical properties of wholestillage are determined by the ethanol productionprocess employed. The feedstock, mashformulation, set fermentor constituents anddistillation system design all are important indetermining the nature of distillery co-products.Quantities of stillage available for recovery arealso determined by process considerationsincluding the grain concentration in thefermentor and the amount of backset recycledto fermentation.

Most dry mill plants in the early 1900s andbefore were small corn milling operations with

364 J. Meredith

Co-

prod

uct

opti

ons

in a

dry

mill

eth

anol

pla

nt

Alc

ohol

Interchange

Wet c

ake

Backset

Thin

stilla

ge

(centrate)

Thic

k

stilla

ge

To s

olid / liq

uid

separatio

n

Superheat

Whole

stilla

ge

6-1

6%

or h

igher

total solids

Fla

sh

evaporatio

n

heat loss

Optio

nal

reboiler

Beer

still

Strip

per

colu

mn

Press

water

Wet c

ake

Backset

Thin

stilla

ge

(C

entrate)

Cake

Cake

Wet c

ake

Wet c

ake

Eductor

recycle

Thic

k

stilla

ge

Syrup

Thin

stilla

ge

Thic

k

stilla

ge

Cen

trif

uge

Sta

tion

ary

Scr

een

Vib

rati

ngsc

reen

Dis

c pr

ess

orS

crew

pre

ss

Sta

tion

ary

scre

en

Cen

trif

uge

Eva

pora

tor

Mix

Dry

Dry

Dry

Ane

robi

cw

aste

disp

osal

Sel

lw

et

Methane

Thin

stilla

ge

(filtrate)

Thin

stilla

ge

(filtrate)

Backset

Backset

Sta

tion

ary

scre

en

Press

water

Solu

ble

s

Sell w

et

Sell w

et

Wet d

istille

rs g

rain

s

wit

h s

olu

ble

s

(W

DG

S)

Dis

tille

rs d

rie

d g

rain

s

wit

h s

olu

ble

s

(D

DG

S)

Dis

tille

rs d

rie

d g

rain

s

(no s

olu

ble

s)

Sell w

et

Heat

recovery

Dis

c pr

ess

orS

crew

pre

ss

Steam

Pad

dle

Scr

een

Fig

ure

1. C

o-pr

oduc

ts o

f eth

anol

pro

duct

ion

in a

dry

mil

l pla

nt.


corn meal as the primary product and beveragespirit as a secondary product. Thick stillagedeveloped into an excellent livestock feedingredient. One gallon of thick stillage contains~0.2 lbs. of protein. In the 1950s, farmerstypically paid $8-10 per 1000 gallons for thickstillage.

Over time, the small corn milling operationswere gradually replaced by larger plantsproducing beverage alcohol as the primaryproduct. Storing and handling large quantitiesof thick stillage introduced problems as thedistilleries increased in size. Whole stillageleaving distillation is a hot, acidic and viscousfluid with limited shelf life (Table 1). Rich inprotein and other nutrients for bacterial growth,stillage will decompose rapidly depending onstorage conditions.

Table 1. Whole stillage properties at the point ofdistillation.

Saturated liquid temperature ≥220°FTotal solids 6-16%Dissolved solids 2-8%pH 4.5 or less (corrosive)General composition Fiber, protein, fats/oil, water,

yeast cells, acetic and lacticacid, other organiccompounds includingglycerol, minerals

For a variety of reasons drying processesdeveloped in the 1990s to remove the waterfrom the thick stillage. Drying eliminates thenecessity to maintain a livestock operation. Shelflife is retained; and the need to handle an acidicand corrosive material precluded. The result isa valuable distillery co-product with less storageand transport volume needed.

Dryhouse design and co-productproduction

Reliability and maximum economic return arethe two key issues that must be addressed indesigning the dryhouse for modern ethanol andbeverage alcohol plants. Process systems mustoperate with minimum downtime for cleaningand repair. Production must not be limited byproblems in the dryhouse. To that end, eachsubsystem or unit operation listed must bereliable:

• whole stillage heat recovery

• whole stillage liquid/solid separation

• backset handling and heat preservation

• cake handling

• evaporation

• syrup handling

• drying

• dry solids handling and storage

• wastewater handling

• emission handling

Dryhouse reliability is also very much a functionof fermentation efficiency in the front of theplant. Unconverted starch and sugar will quicklyfoul heat transfer surfaces in the dryhousecausing capacity reduction (poor heat transfer)and downtime.

Secondly, selection of the process system mustaddress capital cost, revenues and operatingcosts such that the dryhouse remains a profitcenter. The capital cost for building the dryhousedepends on equipment selection as well as designcapacity. Increasing the backset return tocooking and fermentation reduces theevaporative load on the dryhouse and thus thecapital cost needed for construction.

Selection of equipment and the dryhouseprocess

In order to maximize the return on investmentin dryhouse equipment, present value economicanalysis should be completed for the variousdryhouse options. Each co-product option carrieswith it operating costs, as noted in Table 2.

Equally important to operating costs areequipment reliability and longevity. All processmaterials in the dryhouse are corrosive and willattack mild carbon steel. The aggressiveness ofthe attack is somewhat dependent on moisturecontent. For example 30% total solids wet cakeis more corrosive than 90% total solids DDGS.

Stainless steel 304 is the recommendedmaterial for most all applications. Copperbearing steel ‘cor-ten’ can be used in some ofthe less corrosive applications such as DDGS

366 J. Meredith

storage bins. Heated surfaces in some dryers canbe carbon steel without corrosion problems. Insome cases, particularly if sulfur is present,higher grades of stainless steel must be used.

Equipment selection by process system

WHOLE STILLAGE HEAT RECOVERY

Excepting vacuum distillation systems, wholestillage leaving distillation is a superheated liquidat approximately 220°F. If the first processfollowing distillation is screening orcentrifuging, then flash heat (220-212°F) is oftenlost. Available heat recovery processes includeheat interchange with fluids to be heated (beerto stripper column, cook water) and eductorrecycle to the base of the stripper column. Itshould be noted, however, that if the dryhouseincludes evaporators and dryers, cooling of thewhole stillage below the atmospheric boilingpoint causes the need for additional thermalenergy later in the drying process.

LIQUID/SOLIDS SEPARATION

In order to produce any of the co-products(except thick stillage), the thick stillage (or wholestillage, see Figure 1) must be separated into asomewhat diluted liquid that will be partly usedas backset and then can be further processed,and a bulk solid material typically called wetcake. Thick stillage (or whole stillage) total solidscontent in beverage plants typically runs 7-10%compared to 14% and higher for fuel ethanolplants. Beverage plants run ‘thinner’ beerprimarily for quality reasons.

The 8% (or more) total solids in the thickstillage is composed of 3% (or more) dissolvedsolids plus the remaining suspended solids.Dissolved solids are defined as the solids thatwill pass through a No. 4 Whatman filter paperusing a Buchner funnel and vacuum pump.

Good liquid/solid separation is key to runningan energy efficient, low maintenance, dryhouse.As will be discussed later in this chapter,removing water in the dryer uses at least four tofive times more energy than removing water inthe evaporator. The amount of evaporation loadthat can be shifted to the evaporator is almostcompletely dependent on the concentration ofsuspended solids in the evaporator feed (diluteliquid noted above, also called thin stillage orcentrate). Liquid/solid separating equipmentdetermines this concentration.

Early dryhouse designs included combinationsof screens and presses. One common designuses an inclined screen with moving blades orpaddles. Stillage is pumped to the top of thescreen. The thin stillage passes through thescreen while the cake is guided down by thepaddles and exits at the end of the screen, usuallyinto a hopper to be fed into a press. Thesetechniques result in reduced suspended solidscontent in thin stillage compared to the thickstillage. However with these systems, suspendedsolids in the thin stillage remains at levels (2.5%or more) that limit evaporator operation to syrup(evaporator product) levels of 25-28% totalsolids.

Centrifuges

Centrifuges gradually replaced many of thescreen/press systems in the mid 1900s. Various

Table 2. Co-product operating costs.

Co-product Amortization Thermal Electrical Water and Labor Chemicals Marketingenergy energy sewer and supplies and overhead

DDGS Required Required Required Required Required Required RequiredDDG:(dried wet cake) Required Required Required Negligible Required Required RequiredSolubles Required Required Required Required Required Required RequiredThick stillage Negligible Not required Negligible Negligible Negligible Negligible NegligibleThin stillage Required Not required Required Negligible Negligible Negligible NegligibleWet cake Required Not required Required Negligible Required Negligible RequiredSyrup Required Required Required Required Required Required Required


centrifuge configurations were installed;however, the horizontal solid bowl type withseparating forces exceeding 3000 times theforce of gravity performed best on distillerystillage.

Two companies, Sharples/Alpha Laval andWestfalia, have supplied centrifuge equipmentto many ethanol and beverage spirit plants(Figure 2).

There are three continuous actions occurringin the centrifuge (reference Sharplesprocedures):

a. Slurry is introduced into the revolving bowlof the centrifuge through a stationary feedtube at the center of rotation. By centrifugalforce, the solids (more dense than the liquid)are thrown to the wall of the bowl and theliquid forms a concentric inner layer in thebowl.

b. A helical screw conveyor located inside therevolving bowl rotates in the same directionas the bowl but at a slightly lower RPM. Thisconveyor moves the solids to one end of the

bowl, where they are ‘plowed up the beach’(narrowing of helical conveyor) and out ofthe liquid layer into discharge ports.

c. The clarified liquid is continuously removedby overflowing adjustable weirs (plate dams)inside the machine to the discharge point.

It is key to understand that the centrifuge mustbe fine-tuned at each plant location to reachoptimum performance. Various on siteadjustments are required and may take hours ordays to reach design performance:

• The milling system at the front end of the plantaffects particle size and centrifugeperformance. Screen analysis from corn millsshould be obtained and compared to otherplants with good centrifuge performance.

• Depth of pond inside the centrifuge (platedams are adjustable).

• Bowl RPM – example 2600.

Figure 2. Centrifuge manufactured by Westfalia.

368 J. Meredith

• Delta RPM – example 4 to 6 RPM (back drivedetermines delta RPM between bowl andconveyor).

• Beach angle and length.

• Conveyor design including wear partprotection (tiles and carbide wear parts).

Plant operators should monitor total, suspendedand dissolved solids for the process streamsentering and leaving the centrifuge. Lowsuspended solids in the centrate is the firstobjective. Values approaching 1% have beenobtained depending on the plant. Low moisturein the cake is the second objective. Cake from aproperly selected centrifuge typically runs 30%to 31% total solids.

Centrifuge installation

Manufacturers provide detailed and extensiveinstallation instructions. However a few itemsare worth noting. Vents must be provided toatmosphere at each of the two discharge points.Considerable amounts of flash vapor dischargesfrom the centrifuge. The liquid will coolsomewhat (centrifuges are big ‘fans’) fromentrance temperature. Secondly, the supportingstructure for the centrifuge must handle staticand dynamic loads and must be designed by alicensed structural engineer. These large, heavy,fast-rotating machines can impose dynamicloads in excess of three times the static load.Centrifuges are capital cost intensive and highmaintenance machines. Spare parts including aspare bowl assembly are necessary for plantsthat run continuously. Lastly, it is interesting tonote that screen/press systems will give a lowercake moisture (up to 38% total solids) thancentrifuges. However, the high thin stillagesuspended solids for the screen/press systemshifts loads to the dryer and raises overalldryhouse operating costs in most cases.Depending on the application and selection ofco-product types, the centrifuge will not alwaysbe the low cost life cycle option, however.

BULK SOLID CONVEYING AND STILLAGEPUMPING

Equipment must also be selected to move liquids

and bulk solids in the dryhouse from process toprocess. Selecting bulk solids conveyingequipment is the most challenging. Two methodsare common: mechanical conveying andpneumatic conveying.

Mechanical conveyors including screwconveyors, belt conveyors, drag conveyors andbucket elevators are in service in dryhouses.Selection of the type of conveyor is dependenton several factors:

• Length of run

• Elevation change

• Properties of the bulk solids (moisturecontent, bulk density, angle of repose, foulingtendencies)

• Flow rate

Screw conveyors are commonly used for allapplications with the exception of high flow ratesystems. In these high flow systems, cost is afactor and often belt or drag conveyors are lessexpensive than large stainless steel screwconveyors. Bucket elevators have limitedapplication in dryhouses due to the propertiesof most of the bulk solids, which lead to foulingof the buckets. However, bucket elevators havebeen used successfully in handling 90% totalsolids material.

Pneumatic conveying becomes a viablemethod when conveying distances are excessiveand/or elevation change is large, however,certain design issues must be addressed:

• Limit the applications to low moisture bulksolids. DDGS, solubles and DDG (nosolubles) are all successfully transferredpneumatically.

• Dryhouse bulk solids generally contain fiberand are abrasive. Use flat back ells andconservative air:solids ratios.

• Pneumatic conveying introduces the need forair/solids separation at the end point. Thisrequires the use of cyclones, bag houses andvent filters and introduces an air emissionpoint.

These considerations may eliminate viablepneumatic conveying at some locations.


Separator

Long tubevertical

Short tubevertical

Horizontal withremote separator

Vertical withremote separator

(Perferred configuration for falling film systems in dryhouses)

Tubes

Tubes

Separator

Tubes

Tubes

Separator

Separator

Figure 3. Evaporator vessel configuration options.

Stillage pumping systems in the dryhouse donot require exotic, extremely costly equipment.Thick stillage and thin stillage can be pumpedwith centrifugal pumps using an ‘open impeller’.Syrup should be pumped with a positivedisplacement pump (lobe type is suitable).

All piping systems in the dryhouse should befree of internal dead spots. Reduced port ballvalves should not be used. Blind Ts should beavoided. Butterfly valves, ball valves, full portmag-meters and ‘DU-O’ check valves are allsuitable. Rupture discs are preferred to reliefvalves due to fouling.

EVAPORATION

Thin stillage containing 1% or more suspendedsolids and 3% or more total solids can beconcentrated in an evaporator to yield a product(syrup) that ranges from 25 to 50% total solids.Evaporators allow reuse of thermal energy by

taking advantage of the enthalpy of evaporatedvapor at reduced pressure, or by vaporcompression (thermal or mechanical) and thenusing this evaporated vapor for furtherevaporation.

Evaporators find applications in manyindustries. Only certain types of evaporatorconfigurations however, have been usedsuccessfully by ethanol and beverage spiritplants. Choices must be made among:

• Vessel configuration (see Figure 3).

• Forced circulation vs. natural circulation.

• Rising film tubular vs. falling film tubular.

• Plate equivalents of tubular evaporators.

• Multiple effect vacuum with thermalrecompression vs. mechanical vapor recom-pression turbine drive or mechanical vaporrecompression motor drive (Figures 4 and 5).

• Combination evaporator/distillation systems.

370 J. Meredith

Separator

STEAM

FEED

COMPRESSOR

DRIVER

SYRUP

Separator

Condensate Condensate

TUBE

BUNDLE

TUBE

BUNDLE

Figure 5. Falling film mechanical vapor recompression evaporator.

Figure 4. Falling film multiple effect evaporation with thermal recompression.

Separator

FEED

Separator

STEAM

STEAM

T.C.

SYRUP

Separator

Condensate Condensate

Condensate

Condensate

TUBE

BUNDLE

TUBE

BUNDLE

TUBE

BUNDLE

CONDENSER

Water


Rising film and plate equivalents of tubularevaporators have failed and/or caused excessivemaintenance in ethanol and beverage spiritplants and should be avoided. Forced circulationfalling film tubular vessel designs are the mostuser-friendly designs for plant dryhouses. Thekey to falling film systems is the design of thedistribution head at the top entry to the heatexchanger. Each tube should receive the samedesign GPM. Multiple effect and recompressionsystems are equally acceptable for the dryhouse.The final choice depends on life cycleeconomics.

Additional evaporator design considerations

Ethanol plant evaporators are often integratedwith the distillation and molecular sieve systemsto save energy.

Thin stillage reboiler. The first effect of theevaporator becomes the reboiler for the strippingcolumn. This saves energy compared to the useof live steam on the stripper and is more user-friendly (less fouling) than using a whole stillagereboiler on the stripper.

Hot vapor condenser as first effect of evaporator.The first effect of the evaporator replaces oneof the condensers off the rectifying column oroff the molecular sieve – also an energy saver.

Optimum run cycle should be balanced vs. syrupconcentration

Evaporator cleaning is expensive, causing lostproduction and the need for energy input withoutproduct output. Caustic and other chemical costsare also incurred. Each plant should develop theoptimum evaporator run time before cleaningby balancing syrup concentration, operatingcosts and cycle times between cleanups. Overallcosts to produce 35% syrup may exceed coststo produce 30% syrup because of shortenedcycle times at 35% syrup.

Energy costs vary depending on evaporatordesign; and as with other plant systems, mustbe analyzed based on capital cost/operating costtrade offs. Typically, a triple effect thermalrecompression evaporator will evaporate 4.1 lbsvapor per pound of steam consumed. However,

four effect units may approach 6 to 7 lbs vaporper pound of steam. The cost of the fourth effectmust be justified by the increased thermalefficiency. Likewise, depending on heat transfersurface provided, a large mechanical vaporrecompression unit might evaporate 100,000 lb/hr of vapor with a power requirement of 1000kW. Less efficient units would have a lower firstcost.

DRYERS

Dryhouse reliability and operating cost/returnissues hinge as much on the selection of thedryer as any other system in the facility. Dryersare major users of energy and the largest sourceof air emissions in many ethanol and beveragespirit plants. Further, dryer failure will shut theplant down and in some cases dryer failures arenot quickly remedied. Selection of the dryerdepends on various process and operatingconsiderations:

• Energy available (gas, oil, steam).

• Cost of energy options ($/BTU).

• Co-product to be dried (DDGS, syrup).

• Capital cost/operating cost tradeoffs (dryerefficiency vs. dryer cost).

• Corrosion.

• Operating time between shutdowns forcleaning.

• Preheat methods for air including heatrecovery (consider flue gas, steamcondensate, recycle exhaust).

• Environmental issues (dryer emissions)

- scrubber- thermal oxidizer- regenerative thermal oxidizer- remove particulates, VOCs, odor

Types of dryers

As with evaporators, ethanol and beverage spiritplants have used different dryer types dating tothe early 1900s (Table 3).

372 J. Meredith

Table 3. Dryer types, energy sources and co-productsproduced.

Type/description Energy source Typical co-product

Steam tube Steam DDGS (rotating drum)Steam tube Steam DDGS (rotating tube bundle)Direct fire Gas, oil DDGS (rotating drum)Flash dryer Steam, gas, oil DDGSDrum/dehydrator Steam SolublesSpray drier Steam, gas, oil Solubles

Steam tube rotating drum

This dryer is one of the original designs fordistillery dryhouses and many installationsaround the world exist today. Steam tubes(typically 3 or 4 inch pipe) are fixed and locatedin one or more concentric rings inside the drum.Drum diameters may exceed 10 feet, and drumlengths are often 60 feet or longer. Steam entersa distribution head and condensate leaves fromthe same end. As with all dryers, a well mixed,properly conditioned feed is necessary to preventfouling (see following discussion on feedconditioning).

Figure 6. Typical Ronning dryer system.

Each vendor uses in-house design techniques tolift and move the grain through the rotating drum.Efficiencies without air recycle average 0.6 to0.7 lbs evaporated per pound of steam. Inlet airtemperatures may approach 300°F with exittemperatures of approximately 200°F.

Steam tube rotating tube bundle

This is a variation of the steam tube rotatingdrum. In the rotating tube bundle, the drum isfixed and the tube bundle rotates. More heattransfer surface is located in a smaller drumcompared to the rotating drum. Efficiencies aresimilar to the rotating drum. Feed conditioningto prevent fouling is critical to the rotating tubebundle type because of the close clearancesbetween the tubes.

Direct fire rotating drum

Direct fired rotating drum dryers also have a longhistory of service in the ethanol and beveragespirit industries. These dryers have a fixedfurnace section for heating the drying andconveying air stream, followed by the feedintroduction point, which is in turn followed bythe rotating drum. Drum designs vary, with bothsingle pass and multiple pass units available.


Unlike steam tube dryers, direct fire rotatingdrum dryers must include cyclones to separatethe dry product from the conveying air stream.Modern versions of these dryers include airrecycle to capture and reuse energy.

Depending on design, inlet air temperaturesmay vary from 600 to 1200°F. Exit temperaturesare typically 200°F. The elevated inlettemperatures, compared to steam tube systems,will result in additional dryer stack emissions.

Efficiencies are similar to steam tube systems,again depending on design. One direct firerotating drum drying system by RonningEngineering Co., Inc. shows energy use of 1350BTU per pound of water evaporated. A typicalRonning Dryer System is illustrated in Figure 6.A schematic of a direct fire rotating drum systemwith air recycle and incineration of emissionsby ecoDry is illustrated in Figure 7.

Flash dryers

Flash drying technology utilizes a stack or ductsystem instead of a rotating drum for contact

HEAT

RECOVERY

(OPTION)

HEAT

RECOVERY

(OPTION)

WET

PRODUCT

DRYING

DRUM

HEAT

EXCHANGER

BURNER

KILN

CYCLONE

DRY PRODUCT

FAN

Other plant

systems

Other plant

systemsClosed steam

loop

Coupled out gas

FAN

Hot gases

Flue gas

EXHAUST

STACK

Figure 7. Schematic of the ecoDry system for DDGS.

between the hot drying and conveying air streamand the grain. Other features are similar to thedirect fire rotating drum including: fixed furnace,feed entry point, cyclones and possibility of airrecycle. Some vendors use the term ‘ring dryer’instead of flash dryer.

Drying syrup into solubles

Drying evaporator syrup is more challengingthan drying the mixed wet cake/syrupcombination. The syrup is a viscous liquid proneto foul contact surfaces. In contrast, the mixedwet cake/syrup combination is a solid and canbe conditioned with DDGS recycle beforedrying. In addition, dry solubles are extremelyhygroscopic and difficult to handle.

Two technologies, the drum dehydrator andthe spray dryer, have been used to produce drysolubles. In drum dehydrators, the syrup isdeposited as a film on the outside of a heated(steam) rotating steel drum. The resulting drysheet is continuously removed by a blade andthen milled or ground into a powder. Drying

374 J. Meredith

syrup with a spray dryer is not common practice.Difficulties include reaching the very lowmoisture required for solubles (often 4-5%)without fouling the dryer. The resulting productcan be hydroscopic and difficult to handle.However, Alltech has successfully used thistechnology and continues to improve theprocess at two plants.

Feed conditioning for dryers

As noted above, wet cake moisture is typically70% and syrup moisture may be as high as 75%.Feeding this mixture without conditioning tomost dryers will cause rapid fouling and in somecases may lead to dryer fires. The dryer feedshould be conditioned and the moisture contentdecreased by recycling some dry product andmixing it with the wet cake and syrup. Feedmoisture contents of 25 to 30% will enhanceperformance in most DDGS dryers. Theexceptions are drum dehydrators and spraydryers, which are fed syrup ‘as is’.

Selecting the wet cake/syrup/recycle mixer isagain a trade-off between capital cost andoperating performance. Systems that have beenused include:

• U-Trough screw conveyor with mixing blades.

• Pugmill/double screw conveyor with mixingblades (horizontal).

• Vertical enclosed screw conveyor.

• Solid/liquid continuous in-line blenders.

• Ribbon blenders.

Considering mixing efficiency, horsepower andcorrosion/erosion control, the double screwpugmill is a good selection.

WASTEWATER AND EMISSIONS

Wastewater

Evaporator condensate (water removed fromevaporator feed plus in some cases mixed insteam condensate) and CIP (clean-in-place)systems are the two primary sources ofwastewater from ethanol and beverage spiritdryhouses. Many ethanol plants recycle all orpart of the evaporator condensate back to the

cooking system. Beverage plants do not recycledue to quality concerns. The recycled condensatemust be conditioned with chemicals and/or makeup water to prevent contamination andunacceptable pH levels.

Whole stillage, thick stillage and thin stillage(evaporator feed) all have a BOD (BiologicalOxygen Demand) of over 20,000, comparedwith approximately 250 for domestic waste.Evaporator condensate BOD will range from1000 to as high as 5000. The condensate is acidic(usually lactic and acetic acids) with a pH of 4.0or less. EPA direct stream discharge laws usuallylimit BOD to 15 to 30.

Various wastewater treatment systems are inplace in most ethanol plants and virtually allbeverage plants to condition some or all of thewastewater to acceptable levels. Requirementsvary from location to location and also dependon access to municipal sewers. When municipalsewers are an option, dryhouse wastewater ispre-treated before discharge to the sewer.Wastewater pre-treatment requirements alsovary, but a reduction in BOD to 200 would betypical.

Wastewater systems must address thefollowing:

• BOD reduction.

• Suspended solids reduction.

• Temperature.

• Acidity.

• Reduction of other regulated contaminants.

• Sludge handling and disposal.

Wastewater treatment options include:

• Aerobic digestion with lagoons or tanks.

• Sequential batch reactors.

• Anaerobic conversion to methane.

• Land application.

Dryhouse air emissions

Sources for air emission in dryhouses includethe dryer stack, cyclones, bag houses and tankvents. Until about 1999, the US Environmental


Protection Agency (EPA) was primarilyinterested in particulates and the abatementequipment of choice was normally a wetscrubber. However, tests in Minnesota at anethanol plant revealed volatile organiccompound (VOC) and carbon monoxide (CO)emissions from a direct gas fired DDGS dryerthat exceeded permissible levels.

In many states, the EPA now requiresincineration of VOC, CO and particulates usingthermal oxidizers. The design must provide fora specified temperature (usually 1400°F) and aset contact time. Two gas fired oxidizer systemsare being used by the industry:

Regenerative thermal oxidizers This systemutilizes two beds. One bed is online while thesecond preheats the incoming air using heattrapped in the preceding cycle when it wasonline. Vendors state that a 95% recovery ofenergy is possible. However, fouling occurs(dryer stack emissions may include oils andgrease from the corn) and ‘bakeouts’ arenecessary.

Thermal oxidizer with waste heat boiler. Thissystem eliminates the heat retaining bed and usesthe hot air discharge from the oxidizer togenerate steam for the plant with a waste heatboiler. Vendors state that overall thermalefficiency for this system is comparable to pre-oxidizer days when steam was generated by gasfired boilers and dryer stack emissions were notthermally abated.

Storage options for co-products

Wet co-products including thick stillage, wetcake and syrup have a limited shelf life.Inventory at the dryhouse for these co-productsshould not exceed a few days. DDGS andsolubles, however, can be stored for weeks ifconditions are suitable.

Bridging and plugging are problems for DDGSand solubles. Storage in high humidity conditionsaggravates the bridging. Storage in bins andhoppers should be limited to bins with ‘livebottoms’ – typically multi-screw feeder types.Most ethanol plants store DDGS on concretepads within a weather-tight building. Sanitationprocedures must be in place to prevent spoilageand infestation.

Control systems

With the exception of clean up procedures,dryhouse operations are continuous processes.A properly designed, distributed control networkwill provide years of reliable operation.

One operator monitoring the network canresponsibly control centrifuges, evaporators,dryers, conveyors and accessories. Oneadditional operator is usually required to handlebatch operations including CIP, chemicaladditions to wastewater, etc.

Ethanol production and the modern livestock feed industry 377

Chapter 27

Ethanol production and the modern livestock feed industry: a relationshipcontinuing to grow

KATE A. JACQUES


The rapid increase in production of ethanol inNorth America is paralleled by the rise in co-products produced, particularly those from thedry milling ethanol process. While there isundoubtedly great potential for new productsbased on various fractions of stillage, truerealization of that potential will take both timeand considerable industrial creativity. At present,the immediate issue is finding ways to maximizereturn for the co-products currently produced,which is mainly distillers dried grains withsolubles (DDGS).

The primary market for DDGS today islivestock feeding. Of the >3.5 million mT ofDDGS produced in North America (figures forthe year 2000), around 800,000 mT are exportedand the remaining 2.7 million mT are used inUS and Canadian animal feeds (Markham, 2003;Figure 1). If ethanol production from dry mills,and consequently DDGS, production is to doubleas predicted in a few short years, then clearlywe must actively pursue ways to increase itsutilization if DDGS is to maintain its valuablerole in the economics of ethanol production.

0

1

2

3

4

5

6

7

8

Exported DDGS

North American DDGS production

Projected

Mill

ion

met

ric

tonn

es

1990 1995 2000 2005

Figure 1. Production and exports of DDGS in North America (adapted from Markham, 2003).

378 K.A. Jacques

There is a growing amount of animal nutritionresearch in recent years that illustrates that DDGScould play a much larger and more useful rolein modern food and companion animal nutrition.Work with dairy and beef cattle, pigs, poultryand even fish shows that the ‘new’ DDGS hashigher nutrient content, greater flexibility informulations, can be used at higher inclusionlevels and in general provides a cost-effectivemeans of meeting the goals of modern nutrition.While the impact of this work is growing, twokey issues must be addressed if DDGS use inthe feed industry is to increase significantly. Oneis a greater understanding on the part of theethanol industry of what today’s feed industryvalues in ingredients. The second is a change inthe feed industry perception of DDGS.

The contemporary animal feed industry is verydifferent from the industry in the 1970s and1980s when much of the initial nutrition researchwith DDGS was conducted. At that time, theprimary approach to DDGS use was astraightforward replacement of other dietaryprotein sources. Today both the animal and thecontext in which it is fed have changed. Modernlivestock genetics have given us higherperformance expectations for meat, milk and eggproduction; nutritional knowledge has grown,and consequently diet formulations have gottenmore complicated. Feeding goals have changedas well. There is a focus on animal health aspectsof nutrition as well as performance, a generalmove away from animal protein sources infeeding programs in many parts of the worldand the environmental impact of food animalproduction is a critical consideration. Distilleryco-products can have an impact in all of theseareas; and getting that information into themarketplace can increase overall co-productusage to mutual benefit.

‘New generation’ DDGS vs old feedindustry perceptions

Much of the feed industry still perceives DDGSas an ‘alternative protein source’ commoditywith limitations in quality, consistency orcomposition that restrict its use as a feedingredient. Much of this impression, however,stems from the DDGS produced usingtechnologies common 20 to 30 years ago, but isperpetuated by industry references that provide‘book values’ based on ‘old DDGS’. Forexample, energy content in DDGS from modernethanol plants is higher than in ‘old’ DDGS; andboth are higher than values published in the mostrecent edition of Nutrient Requirements of Swine(Table 1, Shurson et al., 2003). Similarly, valuesfor total protein and certain limiting amino acids,particularly lysine, are notably higher in newDDGS (Table 2).

Dated reference values underscore the needfor suppliers of DDGS to provide completeanalytical data. Diet formulation softwareincorporates all nutrient fractions of eachingredient; and reliance on ‘book values’ of anysort by nutritionists is avoided. Part of the qualitycontrol program for feed manufacturers iskeeping ingredient composition updated suchthat fluctuation in nutrient content of aningredient will not result in a change in animalperformance.

The lighter or ‘golden’ color of DDGS frommodern plants is desirable in that it indicates lessheat damage to nutrients, which particularlyaffects amino acid digestibility. The dark colorof earlier DDGS often indicated overheating andcarmelization of nitrogenous components andsugars that escaped fermentation. In modernprocesses less sugar escapes fermentation,stillage viscosity tends to be lower, and more

Table 1. Comparison of energy values for DDGS (dry matter basis)1.

New DDGS New DDGS Old DDGS DDGS(Calculated) (Trial average) (Calculated) (NRC, 1998)

Digestible energy, kcal/kg 3965 4011 3874 3449Metabolizable energy, kcal/kg 3592 3827 3521 3038

Corn: DE (kcal/kg) = 3961, ME (kcal/kg) = 38431Shurson et al., 2003


water can be removed by evaporation, whichreduces heat input in the dryhouse. Ergul et al.(2003) found that color was a good predictor ofamino acid digestibility in chickens and that alighter, more yellow color was associated withhigher lysine, cystine, and threoninedigestibility.

Nutritional value of DDGS

PROTEIN, ENERGY AND PHOSPHORUSCONTENT

Fermentation of the starch in cereal grains,which constitutes about 66% of the weight,concentrates remaining nutrients (fat, proteinand minerals) by approximately a factor of 3.In nutritional terms, this makes DDGS a goodsource of protein, energy and phosphorus, thethree most expensive nutrients added to

livestock diets. For corn, which is normally 9-10% crude protein, resulting DDGS become28-30% (or more) crude protein. Though theenergy in starch is removed during fermentation,the fat remains. Concentration, added to the factthat fat contains three times the energy as starch,brings DDGS energy content nearly equal to thatof corn. This indicates that DDGS can be usedto replace a portion of the corn or other energygrain. The fibrous fraction of corn is concentratedto 7-8% crude fiber in corn DDGS. Phosphorus,notably low in corn (0.28%) is raised to 0.7-0.8%in DDGS. More importantly, the phosphorus inDDGS becomes ‘available’ for digestionfollowing ethanol fermentation. Pigs, poulty andother monogastric species lack the phytaseenzyme needed to hydrolyze phosphorus fromthe phytin form in which it exists in cereal grains(Figure 2). General composition of grain andDDGS for corn, wheat and grain sorghum are inTable 3.

Table 2. Comparison of amino acid composition of DDGS (dry basis) in old generation DDGS, new generation DDGS,and values published in NRC (1998)1,2.

Old generation DDGS New generation DDGS DDGS NRC (1998)

Arginine, % 0.92 (18.7) 1.20 (9.1) 1.22Histidine, % 0.61 (15.2) 0.76 (7.8) 0.74Isoleucine, % 1.00 (9.1) 1.12 (8.7) 1.11Leucine, % 2.97 (12.4) 3.55 (6.4) 2.76Lysine, % 0.53 (26.5) 0.85 (17.3) 0.67Methionine, % 0.50 (4.5) 0.55 (13.6) 0.54Phenylalanine, % 1.27 (8.1) 1.47 (6.6) 1.44Threonine, % 0.98 (7.3) 1.13 (6.4) 1.01Tryptophan, % 0.19 (19.8) 0.25 (6.7) 0.27Valine, % 1.39 (2.3) 1.50 (7.2) 1.401Values in parentheses are coefficients of variation among ethanol plants.2Shurson et al., 2003

Table 3. Composition of grains and DDGS made from them.

Corn Sorghum Wheat

Grain1 DDGS2 Grain1 DDGS2 Grain1 DDGS2

Dry matter, % 89 88.9 89 90.31 88 92.48Crude protein, % 10.3 30.6 9.2 30.3 13.5 38.48Crude fiber, % 8.8 10.7 6.01Crude Fat, % 6.7 10.9 2.9 12.5 2.0 8.27Calcium, % 0.05 0.06 0.03 0.10 0.06 0.15Phosphorus, % 0.43 0.89 0.29 0.84 0.37 1.041NRC, 19982cited at http://www.ddgs.umn.edu/

380 K.A. Jacques

O

H

H

O

O

H

O

H

O

P

O

O O

O Zn

PO O

HO HO

P P OO O

Ca

O

MgH

O

O

H

PO

OFe

OPO

HO

Figure 2. Structure of phytin phosphorus in cereal grains.

CONTRIBUTION OF SOLUBLES VS GRAINS

The amounts and forms of nutrients varybetween grains and solubles; and the differencesin quantities of solubles added back to grains isone potential source of variation in DDGSnutrient content between ethanol plants. Spentgrains are higher in protein than solubles andcontain almost all the fiber. Solubles, in contrast,contain most of the fat, phosphorus and otherminerals. As such, variations in the amounts ofsolubles added can have the biggest potentialimpact on the overall energy and phosphoruscontent of DDGS (Knott et al. (2001).

Knott et al. (2001) examined grains andsolubles from six Minnesota ethanol plants todetermine nutrient variability in these fractionsover a 3-week period (Table 4). Nutrient contentof grains generally varied less than did solubles,and crude fat content of both grains and solublesvaried more than any other nutrient.

VARIATION IN DDGS IMPACT ON NUTRITIONALVALUE

Knott et al. (2001) also noted that some plantshad much more variation in particular nutrientsbetween batches than others; and the authors

calculated the impact this would have on protein,fat or phosphorus of DDGS based on inclusionof 100% of the solubles (Table 5). In the plantwith least variation, crude protein content variedonly 2 percentage points, however it varied by10 percentage points in the plant with mostvariation. Coefficients of variation exceeded20% for fat and phosphorus content in this plantas well. From a nutritionist’s standpoint, the twobatches of DDGS represented by the minimumand maximum from that source vary markedlyin value in terms of cost per nutrient unit orpotential use in the least cost formulation. Whilepercentage differences may appear small, theycan have a large impact on the amounts of other,more expensive ingredients DDGS mightreplace. This, in turn, affects the value of DDGSto the feed manufacturer.

CAN WE INCREASE USE OF DDGS IN DAIRYAND BEEF CATTLE?

Currently beef and dairy cattle consume about80% of the DDGS sold in North America (Figure3, Markham, 2003). DDGS is of value both interms of energy and protein for meat and milkproduction, contributes digestible fiber andcontains nutritionally meaningful amounts ofphosphorus.

Protein nature and quality

Traditionally, most DDGS has been used in cattlefor its ‘by-pass’ protein value, which allowsreduction of more expensive soybean meal (eg)in the diet. Ruminants are unique in that the firstcompartment of the digestive tract is ananaerobic fermentation, which allows thedigestion of fiber by symbiotic bacteria in therumen. As such, part of the animal’s protein/amino acid supply comes from microbial proteinand part from feed protein that escapes or ‘by-passes’ rumen microbes and is digestedenzymatically in the small intestine. Highperformance dairy and beef cattle diets arebalanced to supply both rumen-degradable andrumen escape, or rumen-undegraded protein.The general goal is to let rumen microbesupgrade the biological value of poor qualitynitrogen sources (like yeast, rumen bacteria canuse simple N sources like urea or other FAN) by


Table 4. Average, minimum, maximum, and meannutrient values (dry matter basis), of grains and solublesfractions from six Minnesota ethanol plants1.

Average Minimum Maximum

Grains fractionDry matter, % 34.3 33.7 34.9Crude protein, % 33.8 31.3 36.0Crude fat, % 7.7 2.1 10.1Crude fiber, % 9.1 8.2 9.9Ash, % 3.0 2.6 3.3Calcium, % 0.04 0.03 0.05Phosphorus, % 0.56 0.44 0.69

Solubles fractionDry matter, % 27.7 23.7 30.5Crude protein, % 19.5 17.9 20.8Crude fat, % 17.4 14.4 20.1Crude fiber, % 1.4 1.1 1.8Ash, % 8.4 7.8 9.1Calcium, % 0.09 0.06 0.12Phosphorus, % 1.3 1.2 1.4

1From Knott et al., 2001

Table 5. Predicted nutrient content and coefficients ofvariation in DDGS from plants with the least or mostvariation in protein, fat or phosphorus content1,2.

Plants with Plants withleast variation most variation

Crude proteinMinimum, % 29.0 19.8Maximum, % 31.2 29.3Mean, % 29.9 25.8CV 3.57 20.8

Crude fatMinimum, % 11.0 7.9Maximum, % 12.6 11.5Mean, % 12.0 9.8CV 7.00 43.58

PhosphorusMinimum, % 0.73 0.63Maximum, % 0.82 0.96Mean, % 0.70 0.76CV 7.64 22.44

1From Knott et al., 20012DDGS nutrient content predicted assuming 32 gallons of syrup added/minute, 79.63 lbs of dry syrup added/minute, 40.63% solubles and 59.37% grains in DDGS on a dry matter basis.

Dairy cattle

Beef cattle

Pigs

Poultry

Figure 3. Utilization of DDGS by the North American livestock feed market (Markham, 2003).

382 K.A. Jacques

converting it to microbial protein. On the otherhand, some feed protein sources are higher inbiological value (amino acid profiles useful tothe animal) than microbial protein and of greaterbenefit to the animal if they ‘by-pass’ rumenfermentation. Because it has already beenthrough fermentation, over 50% of DDGSprotein is ‘by-pass’ protein (Schingoethe, 2001),which compares favorably with other availablesources (Table 6).

Table 6. Comparison of the crude protein and rumen-undegradable protein contents of various feedstuffs.

Crude protein Rumen undegradable(%) protein (%)

Forage sourcesAlfalfa hay 19 28Alfalfa silage 19 23Corn silage 9 31

Distillery co-productsCorn DDGS 28 54

Oilseed mealsSoybean meal 47 35Canola meal 42 28Sunflower meal 26 26

Animal sourcesFishmeal 67 60Feather meal 89 71Blood meal 92 82Meat and bone meal 42 61

While substitution of DDGS for other proteiningredients is most common, use of DDGS toreplace energy sources in ruminants mayprovide benefits to rumen/animal health as wellas boost dietary inclusion rates. Starch overloadin the rumen of a dairy cow or beef finishinganimal is a common cause of acidosis, low milkfat content and variation in feed intake andperformance. Distillers grains (wet or dry) aresimilar in energy content to the original grain,but the energy value is contributed primarily byconcentration of the fat along with somedigestible fiber. In addition, bypass protein hasa higher energy value than the same proteindegraded in the rumen. Lower incidence ofacidosis and the impact of energy level and form

have been suggested as the reason that thefeeding value (higher performance andimproved efficiency) of the distillery co-producthas been observed in several trials to be higherthan equal amounts of the original grain(Klopfenstein, 1996; Klopfenstein and Grant,2001).

The fiber in DDGS from corn results in >40%neutral detergent fiber (NDF), which is the sumof cellulose, hemicellulose and lignin. Whileparticle size is too small for the NDF in DDGSto provide the physiological effects needed bycattle from fiber, over 60% of the NDF is in theform of hemicellulose, and thereforenutritionally useful.

Wet distillers grains

Cost of transport and rapid spoilage make useof wet distillers grain practical only where largeconcentrations of cattle are relatively near anethanol plant, but where such opportunities exist,the reduced energy requirements for dryingmake the product attractive for both buyer andseller.

Apart from moisture content, the nutrientcontent of DDGS and wet distillers grains arevery similar, provided that the same amounts ofsolubles are added. Interestingly, some of thestudies with growing cattle have suggested anadvantage in growth performance of feeding wetvs dry distillers grains. Speculation has been thatnormal drying may reduce energy value to someextent, however this has been difficult to prove(Klopfenstein, 1996; Klopfenstein and Grant,2001).

A very real concern when feeding wet grainsis molding and the associated potential formycotoxin contamination, which can bothreduce performance and cause animal healthproblems. Addition of preservatives at thealcohol plant may extend the shelf life of wetgrains, especially during the summer.Experiments with CakeGuard™ (Alltech Inc.),a preservative based on a mixture of organicacids, have shown that use rates of 2 kg/tprevented mold growth for two weeks whenstored at room temperature. In the field,CakeGuardTM is being used with success at 1 kg/tonne to extend stability of wet cake beyond thetypical four days expected.


PIGS, POULTRY AND DISTILLERS GRAINS WITHSOLUBLES

The pig sector arguably represents the largestpotential for increased use of DDGS. Untilrecently, very little DDGS has been used in dietsfed pigs. The primary reasons include variabilityin quality and nutrient content among sources,poor amino acid digestibility due to overheatingduring drying, concerns about the high fibercontent, and cost competitiveness with the corn,soybean meal and dicalcium phosphate(inorganic phosphorus source) they replace(Shurson et al., 2003). Research with all classesof pigs at the University of Minnesota and othermidwestern universities has demonstratedconclusively that the higher energy, digestibleamino acid and available phosphorus contentof DDGS from modern ethanol plants allowsinclusion at levels in excess of 20% in manydiets (Table 7).

Table 7. Use rate recommendations for high qualityDDGS in diets for pigs.

Production Phase Maximum % of diet

Nursery pigs (>7 kg) 25Grow-finish pigs 20Developing gilts 20Gestating sows 50Lactating sows 20Boars 50

Shurson et al., 2003

The higher levels of inclusion listed in Table 7assume that diets are formulated based on thedigestible amino acid (as opposed to totalprotein) and available phosphorus content of dietingredients (Shurson et al., 2003). This is animportant point to note in terms of marketingDDGS. Nutritionists, particularly those feedingpigs and poultry, need to know the amounts ofdigestible amino acids in an ingredient, notsimply crude protein content.

DDGS, like corn, is fairly low in the aminoacid lysine, which is the first limiting amino acidfor growth. Modern ethanol plants, with processcontrols that help maximize starch extraction,fermentation efficiency, and other upstreamoperations, produce less variable DDGS withhigher levels of more digestible lysine (see Table2). As processes further improve and qualitycontrol of co-products reaches higher standards,an increasing number of pig feed manufacturerswill have the confidence in DDGS needed toboost inclusion levels from the currentlyconservative 5 and 10%.

ENVIRONMENTAL IMPACT

Increasing use of DDGS in pig diets canultimately have a large and beneficialenvironmental impact. Owing to the indigestibilityof the phosphorus in corn, the phosphoruscontent of pig manure is a major concern.Phosphorus carried in runoff from land used to

1. 00 E+ 01

1. 00 E+ 02

1. 00 E+ 03

1. 00 E+ 04

1. 00 E+ 05

1. 00 E+ 06

1. 00 E+ 07

Week 1 Week 2

Mo

ld(C

FU

/g)

Control 1 kg/T 2 kg/T 5 kg/T

Figure 4. Effect of CakeGuardTM addition rate on mold growth in wet distillers grains weeks 1-5 through storage.

384 K.A. Jacques

spread manure threatens surface waters, causingalgae blooms in streams. Regulations limitingthe amounts of phosphorus that can be added toa given land area are being put into place invirtually all major agricultural regions, which inturn can limit the number of animals reared andthe overall profitability of the farm. Given boththe expense of adding inorganic phosphorus toanimal feeds and the desire to reduce theenvironmental impact of manure, the highavailable phosphorus content of DDGS hasmuch to offer pig feed manufacturers.

POULTRY

Use of DDGS in chickens and turkeys hashistorically been limited, however its value inproviding ‘unidentified growth factors’ thatinfluence hatchability and efficiency have longbeen recognized. While the fiber content ofDDGS limits its use to a greater extent in poultrythan in other species, research with the productfrom modern plants has shown that when highquality DDGS is used in diets formulated on anavailable amino acid basis, performance andeconomic advantages are the result. Workdemonstrating these responses in turkeys hasresulted in an increasing amount of DDGS beingincluded in turkey diets in Minnesota.

Contaminants: mycotoxins andantibiotics

A quality concern for use of high levels ofdistillery co-products is the threat of mycotoxincontamination. Mycotoxins present in the grainmashed for fermentation are not destroyedduring fermentation and distillation, since heatstability varies among toxins. In addition,moisture conditions in storage are critical forprevention of surface mold. As long as DDGSis dry to <15% moisture and is cool enteringstorage, the potential for mold growth is nogreater than in other feed ingredients. It shouldbe noted, however, that establishing proceduresfor mold prevention (along with mycotoxinanalysis) is something purchasers increasinglook for from ingredient suppliers. Commercialfeed mixes often include mycotoxin adsorbentsas a precaution against mold growth in feed

stored at customer farms. One adsorbent is basedon a yeast cell wall component modified tomaximize the range of toxins adsorbed, whichare thereby prevented from harming the animal(Table 8).

In recent years, the use of in-feed antibioticgrowth promoters has markedly declined. Botha consumer issue and a potential health concernbecause it might contribute to resistance ofpathogens to similar drugs used in humanmedicine, feed manufactures are removing suchcompounds from feeds and ‘antibiotic-free’ is amarketing claim made for some food animalproducts as well as animal feeds. Further, theexclusion of all or some antibiotic growthpromoters (virginiamycin is a relevant example)is a legal requirement in some arts of the world.The possibility of any measurable antibioticresidue, no matter how small the amount, willincreasingly restrict market opportunities forethanol co-products.

Table 8. Mycotoxin binding efficiency of Mycosorb1.

Amount bound2

(%)

Total aflatoxins 85.23 B1+B2+G1+G2)Zearalenone 66.66Citrinin 18.41T2-Toxin 33.39Fumonisin 67.00

1Minus values of control group2Alltech Inc.

Replacing animal proteins: opportunitiesfor export?

Good quality vegetable protein sources are inincreasing demand across the world. While itcan be said that this would be true in any case,one of the changes in animal feeding programsin Europe following BSE was a reduction in useof animal by-product ingredients. Higher pricesand lower availability of fish meal haveincreased use of soybeans as protein sources inpoultry feeds in both Europe and South America.While shipping costs, the acceptability of by-products from genetically modified crop varietiesand other issues affect export patterns, the trend


away from animal by-products may provideopportunities as experience with, and availabilityof, ethanol co-products increases.

Conclusions

Feed industry professionals that know both theNorth American food animal market and thenature and advantages of modern distillery co-products have done the math: the expectedincrease in DDGS could be used in NorthAmerica, and to great advantage to both ethanoland feed manufacturers. To turn thosecalculations into reality, however, the gapbetween the two industries must be bridged withthe kinds of information that will give feedmanufacturers knowledge and confidence in thefeeding value of DDGS. As the sellers, theethanol industry has the job of promoting andensuring consistency of DDGS features andbenefits. As in selling any product, knowing thetarget market and how the product fits into itwill be critical; and attention to nutritional qualitywill be needed.

The ruminant sector has the potential to usemore distillery co-products, particularly bytaking advantage of the opportunity afforded toeliminate starch overload and reduce acidosis.However the biggest increase in DDGS usagecan take place in the pig industry where currentusage is undeveloped and the research is in placeto move forward. Another advantage is that to alarge extent, the pigs in the midwestern US arelocated comparatively close to the ethanolindustry. In contrast, the large dairy cattlemarkets in California, Texas and thesoutheastern US are farther away. The poultryindustry also has the ability to use more DDGS,though inclusion rates are more restricted byDDGS fiber content. However, increasingattention to quality and consistency details thataffect amino acid availability will offer poultryfeed manufacturers an increasingly attractiveingredient option.

References

Ergul, T., C. Martinez Amezcua, C. Parsons, B.Walters, J. Brannon and S. L. Noll. 2003.Amino acid digestibility in corn distillers driedgrains with solubles.

Hancock, J. D. 1995. The value of distillers co-products in swine diets. Proc. Distillers FeedRes. Council. 50:13-20.

Klopfenstein, T. J. 1996. Distillers grains as anenergy source and effect of drying on proteinavailability. Anim. Feed Sci. Technol. 60:201-207.

Klopfenstein, T. and R. Grant. 2001. Uses ofcorn coproducts in beef and dairy rations.Presented at the Minnesota Corn GrowersAssociation Technical Symposium,Bloomington, MN.

Knott, J., J. Shursonand J. Goihl. 2001. Effectsof the nutrient variability of distiller’s solublesand grains within ethanol plants and theamount of distiller’s solubles blended withdistiller’s grains on fat, protein and phosphoruscontent of DDGS. University of MinnesotaDDGS Website: http://www.ddgs.umn.edu/research-quality.htm

Markham, S. 2003. DDGS production. Presentedat the NCGA Regional Distillers GrainsWorkshop, Des Moines, IA. http://www.ddgs.umn.edu/ppt-pqd.htm

NRC. 1998. Nutrient Requirements of Swine.10th revised edition. National Academy Press,Washington, DC.

Schingoethe, D.J. 2001. Using distillers grainsin the dairy ration. National Corn GrowersAssociation Ethanol Co-products Workshop.Nov. 7, Lincoln, NE.

Shurson, J., M. Spiehs, J. Wilson and M.Whitney. 2003. Value and use of ‘newgeneration’ distiller’s dried grains with solublesin swine diets. In: Nutritional Technology inthe Feed and Food Industries, Proceedings ofthe 18th Annual Symposium (TP Lyons and KAJacques, eds). Nottingham University Press,UK. 223-235.

Biorefineries: the versatile fermentation plants of the future 387

Chapter 28

Biorefineries: the versatile fermentation plants of the future

KARL A. DAWSON


Introduction

Traditional ethanol production systems aredesigned to efficiently produce alcohol from simplesugars derived from grains and products from sugarprocessing. In their natural form, these sugars arein the form of polysaccharides that must besubjected to degradative processing before theycan be subjected to the metabolic activities ofmicroorganisms during active fermentation. Sincepolysaccharides are generally derived fromcomplex plant materials, no single processing stepor enzyme treatment will release all of the sugars.As a result, the conversion of plant material touseful sugars is generally incomplete, and a largeproportion of the material initially entering theethanol processes is not converted to ethanol andremains unfermented. One of the key challengesfor the modern alcohol industry is defining systemsfor economically converting these unfermentedmaterials into useful or value-added co-productsthat can contribute to the profitability of thefermentation plant. This chapter will describe someconcepts that can be used to extend thebioprocessing capabilities and flexibility of alcoholproduction plants to make them more sustainableand economically viable.

The biorefinery concept

In the past, processes in traditional alcoholdistilleries have been optimized for the efficient

production of a single product, ethanol. Thevalue of the plant was largely defined on thebasis of its ability to efficiently produce alcohol.There was little consideration given to the ‘by-products’ that are produced, even though thesematerials can account for more than a third ofthe organic materials entering the plant. Themarkets for these materials have often beenconsidered secondary to that of ethanolproduced. With increased emphasis on economicaccountability and the need for controlling wastestreams and environmental impacts, it is nowbecoming critical to update some of the basicprocesses used to produce ethanol. Advances inbiotechnology make these updates particularlytimely, since our new understanding ofbiochemical systems and specific micro-organisms has provided us with many new toolsthat can be used to upgrade the basic processingsystems used in ethanol production.

The term ‘biorefinery’ has been used todescribe chemical production facilities that usebiological systems (microbial fermentations andenzymatic conversions) to efficiently catalyzekey chemical transformations (Abbas andCheryan, 2002). As we have moved into the ageof molecular biology and metabolic engineering,we now see that innovative applications ofbiology can have a tremendous impact on boththe economic and the practical engineering of

388 K.A. Dawson

the ethanol production process. In broaderterms, biorefineries can be seen as very versatilefacilities that are not bound to the production ofan individual product, but can instead use avariety of substrates and many differentprocesses to produce a wide range of products

with minimum amounts of waste (Figure 1).Instead of thinking of traditional processes thatproduce only one or two products from a singlesubstrate (Figure 2a), it is important to thinkabout the value of systems that produce manydifferent products (Figure 2b) from a variety of

OUTPUT• Alcohol• CO2

• Glycerol• Chemicals• Solubles• Yeast• Light grains

Feed and fiber(animal feeds)

INPUT• Corn• Grains• Molasses• Corn stover• Paper waste• Sawdust• Bagasse• Fiber crops• Municipal waste

Figure 1. A biorefinery.

Saccharified Feedstock(1 ton)

Water

Carbondioxide(660 lb)

Ethanol(630 lb)

Animal Feed(710 lb)

Clarification and Drying

Yeast Fermentation

Distillation

Figure 2a. A bioprocessing system that focuses on the efficient production of ethanol.


Feedstock(1 ton)

High value animal feed(550 lb)

Inositol(20 lb)

Animal feed(700 lb)

Secondary fermentation/enzymatic treatment

and extraction

Glycerol(350 lb)

Succinic acid(20 lb)

Ethanol(630 lb)

Carbon dioxide(660 lb)

Carbon dioxide(660 lb)

Plant oils(100 lb)

Special Fermentation

Distillation

Clarification

Evaporation

Separation/refining

Water

Figure 2b. A bioprocessing system that focuses on production of a range of products with minimal waste.

390 K.A. Dawson

substrates. In some cases, this may be at theexpense of ethanol production. Ideally, theeconomic value for the biorefinery would comefrom its versatility. Biorefineries that can use awide variety of feedstocks and produce a varietyof co-products will have the ability to adapt theirprocesses to address external economicpressures.

In addition, there are a number of other basicpressures that make the biorefinery conceptattractive for modern ethanol productionsystems. In particular, environmental issues andthe need for controlling waste streams fromchemical production systems are becoming thedriving force for changes in ethanol production.In this respect, it is becoming possible to usebiotechnology to design biorefineries in a waythat will allow them to be emission-free, efficientin their use of energy and water, and yet flexibleenough to profitably provide useful productsfrom inexpensive substrates.

Feedstocks for the biorefinery

Most modern ethanol production facilities relyon starch or sucrose from grains and sugar caneas the basic substrate for ethanol production.The sugars in these materials are readily releasedfrom polysaccharides for fermentation bydifferent combinations of heat and enzymatictreatment. However, the lignocellulose in plantfiber or biomass is now being seriouslyconsidered as a feedstock for alcohol production(Lynd et al., 1991). This material is the mostabundant source of carbohydrate in the world,but is much more difficult to degrade than thestarches derived from grains. The use of thisfibrous biomass in a biorefinery providesmechanisms for using renewable and sustainablesubstrates as a readily adaptable energy sourceor as a precursor for numerous foods andchemicals (Gong et al., 1999). While majoradvances have been made in convertinglignocellulolosic substrates to more valuablematerials, there are a number of basic limitationsthat have prevented the adaptation of suchtechnologies to large-scale production ofvendable products like ethanol. One majorchallenge with fibrous substrates has been thelack of efficient biochemical mechanisms forpreparing the sugars for fermentation. In recent

years, there has been considerable interest in thedevelopment of specific enzyme systems forproducing fermentable substrates from fibrousplant materials and waste products. Specialprojects have focused on the development ofinexpensive cellulase and xylanase enzymesystems with improved catalytic activities. Majordevelopment efforts using both fiber-digestingenzymes and new thermochemical processeshave allowed for the saccharification of non-traditional lignocellulosic substrates (Kaylen etal., 2000). However, there is also concern aboutthe use of the pentose sugar derived from thehemicellulose fraction of fibrous plant materialbecause these sugars are not used bySaccharomyces cerevisiae in traditional ethanolproduction systems. Feedstocks may alsoinclude waste products from the sugar and starchrefining industry, milk sugars in whey, corn fiber,forest products and even municipal and foodindustry wastes. Innovative applications ofbiotechnology and engineering principles areneeded to obtain maximum benefit from thisarray of substrates.

Traditional co-products from ethanolproduction processes

Traditionally, co-products derived from ethanolproduction processes are those that can beobtained by simple physical or chemicalextractions associated with grain processing. Inmany cases, these products have their ownintrinsic value as products for human and animalfeeds or as feedstocks for other chemicalproduction processes (Table 1). The value ofthese materials is often determined by the priceof competitive products coming from thepetroleum industry.

Probably the most important co-products fromthe alcohol production process are the distillersdried grains (DDG) and distillers dried grainswith solubles (DDGS). This material is theresidue remaining after grain processing,fermentation and the distillation of the alcohol.As a result, the composition of these materialsis based on biological-based transformations thatare brought about by enzyme hydrolysis of thesubstrate and those associated with the yeastduring the fermentation process. Applications ofthese co-products often take advantage of the


high fiber and high protein content that remainsafter the preparation of the readily fermentablesugars and starch for fermentation. Distillersgrains account for the largest single co-productfrom the ethanol industry in the United States.Close to 10 million tons of these DDG isproduced in the United States each year. Becauseof the high fiber content, many of these co-products are typically used as feeds forruminants (beef and dairy cattle). However,many of them also contain useful substrates thatcould provide a potential starting point forfurther biotransformation into high valueproducts. Such transformations provide highervalue feeds for other livestock productionsystems.

Changing the value of traditional co-products in a biorefinery

Distillers grains are considered to be thestandard co-products of the alcohol fermentationprocess in dry mill, ethanol production systems,but can be quite variable in quality. While therehas been some attempt to differentiate distillersgrains obtained from different sources, most ofthese materials are currently sold as commodities

Table 1. Some traditional co-products from various ethanol feedstocksa.

Feedstock Product Uses

Wheat Gluten (proteins) Human foods, building materials, and pharmaceuticalsBran (fiber) Cereal-based food productsGerm Fortification of bakery productsDistillers grains Animal feeds, human food ingredients

Corn Zein (protein) Water resistant films and fibersCorn fiber Animal feedsCorn germ Protein and mineral supplements for baked goodsGluten meal Protein supplementCorn oil Source of gums, lecithin, emulsifiers and antioxidants. Used extensively in human foodsDistillers grains Animal feedsStillage (or steep liquor) Source of nutrients for industrial fermentations

Oats Proteins Protein fortification of human foodsOat starch Cosmetic productionOat hulls Production of adhesives and abrasive filters

Barley Protein Protein for human food and animal feedsBarley fiber Animal feeds and human foodsEnzymes Industrial applications (distilling industry)Distillers grains Animal feedsTocopherols Antioxidant for food products

aAdapted from Torre, 1999.

into the animal feed market. Probably the bestopportunity for developing the biorefineryconcept is related to the innovative use of thedistillers grains. Since distillers grains are mostoften valued as a protein source for foodanimals, strategies that increase the proteincontent, alter the availability the protein, andimprove amino acid composition may haveconsiderable value. Such strategies may alsoimprove the profile of the key nutrients foranimal production and allow for a moreeconomical application of these spent grains inmonogastric animals (pigs) and poultry diets. Anumber of approaches are currently available formodifying distillers grains and improving theiroverall value as co-products. These can besimple supplementation strategies, but many ofthem are true biotransformations that could becarried out in a biorefinery setting.

One approach for improving the value of distillersgrains would be to develop supplementationstrategies that improve nutrient profiles. This canbe accomplished by adding deficient or uniqueamino acids, vitamins, and minerals in a way thatcomplements the basic nutritional value of spentgrains and other dietary components. Since theresults of these strategies are often reflected inimproved animal performance, it is often difficult

392 K.A. Dawson

to evaluate the economic value of suchapproaches based on simple nutrient analysis.Instead, it must be based on the improvedperformance of the livestock (Table 2). Earlystudies with one such supplementation strategyusing proprietary vitamin and mineral mixeshave indicated that it is possible to get a three-to-one return on processing costs when suchstrategies are used to improve milk productionin dairy cattle.

Another approach for improving the value ofthe distillers grain could focus on the relativelypoor digestibility of the protein and fiber fractionsin the spent grains. Current estimates of theavailability of the most limiting amino acids indistillers grains is estimated to be about 65% inswine diets (Dale and Batal, 2002). This suggeststhat a large portion of the protein ingested bythe pig is not subject to digestion and isunavailable as a nutrient. A biotransformationwith an enzymatic treatment that increases theamino acid availability from 65 to 90% couldsignificantly alter the value of spent grains.Currently, biotechnology may provide specificenzyme complexes that can bring about therequired improvements in amino acidavailability. If such modified grains areformulated into a diet for a pig in place ofsoybean meal, savings in the cost of a ton offeed can be demonstrated (Table 2). In this case,biorefinery steps that allow for the pre-treatmentof distillers grains with an enzyme may increasethe basic economic value of the distillers grainas a feed ingredient in pig diets by as much as20 to 30%.

Probably the most effective way to enhancethe value of distillers grains would be to exposethem to a more complex biotransformation usinga secondary fermentation system. This processcould use selected strains of fungi and bacteriain combination with a short fermentation periodto further alter the grain composition. Thisprocess could increase the protein content,increase the lysine content of the protein, andimprove the overall digestibility of the protein.This fermentation process may also decrease thefiber content of the spent grains and increasetheir overall energy availability. Suchmanipulation could provide a much morebalanced protein to the animal in the form of ahighly digestible feed ingredient. Simplefermentations that increase protein content fromaround 30% to up to 45%, increase amino acidavailability from 65 to 90%, and improve lysinecontent (from 1.35% to 2.7%) can improve thevalue of the spent grains as a feed ingredient.This can result in up to a 60% increase in thebasic economic value of the distillers grains asa feed supplement (Table 2). These examplesclearly show the value of strategic applicationsof microbial induced biotransformations in abiorefinery setting.

Products from the biotransformationprocesses

Commercially, a number of products are

Table 2. Potential value of developing fermentation co-products from spent distillers grains for swine.

Strategy Goal of strategy Decrease in feed cost Added value per ton of distillersrelative to soybean meal grainsa

Strategic supplementation Improved or more balanced Depends on cost of Value only obtained through with nutrients or dietary nutrient profile supplement enhanced performance of pigs adjuncts (minerals, 1:3 return on investment has vitamins, amino acids, been reported or enzymes)

Enzymatic treatment Improved digestibility or $2.00 Approx. $20.00 (hydrolytic pre-digestion) improved nutrient availability

Secondary fermentation Improved nutrient content; $6.00 Approx. $60.00 (microbial treatment or improved nutrient profile; ensiling) improved nutrient digestibility

and availabilityaChanges in the value of distillers grains when formulated relative to soybean meal in swine diets


produced using the basic metabolic activities ofliving organisms (Table 3) . Many of theseproducts are produced in well-definedfermentation systems that convert low qualitysubstrates into high value, vendable products orchemical precursors for other manufacturingprocesses. Alternatively, some of thesebiotransformations can be achieved by applyingspecific enzyme systems. By applying newfermentation processes, biotechnology canprovide tools that can be used for the productionof new inexpensive enzyme systems for morecost-effective biotransformations. Innovativeapplications of some of the fermentationprocesses or enzyme treatments into integratedproduction systems will be important to thedevelopment of flexible biorefineries (Figure2b).

Table 3. Some industrial products currently producedusing fermentation technologies that might be producedin a biorefinery using microbial biocatalysts.

EthanolMethanolCitric acidGlycerolXylitol and other sweetenersAstaxanthinLactic acidPolylactide polymersEmulsifiers and surfactantsVitamins (riboflavin)Fatty acids and oilsEnzymesSpecific pharmaceuticals and specific heterologous proteinsSingle cell protein

Microorganism use in biotransformations

The processing power in modern biorefineriesis often limited by the metabolic power ofmicroorganisms or biocatalysts available.Modern biotechnology is now capable ofengineering new metabolic systems that increasethe flexibility of biorefineries. These are basedon both traditional or naturally occurringmicroorganisms and genetically modifiedorganisms. Microorganisms used for bio-transformation fall in to three major groups:classical yeast, fungi, and bacteria. Many typesof organisms have been examined for their

ability to contribute to bioconversion of basicfeedstocks and metabolic intermediates withinbiorefineries.

YEAST

While the metabolic activities of many differenttypes of microorganisms have been examinedfor functional roles in biorefinery systems,currently the vast majority of commercialfermentations depend on active yeast cells. Theseorganisms are the basic biocatalysts that arecentral to the modern fermentation industry. Themost important yeasts in this respect are fromthe species Saccharomyces cerevisiae. There areseveral thousand strains of S. cerevisiaeavailable to the fermentation industry. They canbe differentiated biochemically and by theirbasic fermentation activities. However, a numberof non-conventional yeast have been tested andused as biocatalysts for the production ofspecialized fermentation products (Table 4).Representative strains from these different yeastspecies carry out a variety of biochemicalprocesses that make them attractive for drivingfermentation and biotransformation processes inan integrated biorefinery setting.

Metabolic engineering is a concept that usespowerful genetic engineering approaches andmicrobial physiology in strain developmentstrategies for improving the biocatalyticcapabilities of microorganisms (Olsson andNielsen, 2000). Metabolic engineering of yeastnow allows for the development of fermentationprocesses that integrate much of our knowledgeabout physiological, metabolic, and biochemicalproperties of this group of organisms usingspecific genetic engineering technology toimprove specific bioprocesses. This is becominga very successful way of changing our traditionalfermentation technologies. For example,recently, genetic modification techniques havemade a variety of yeast strains available that havevery specific metabolic activities (Gong et al.,1999, Mouiruzzaman et al., 1997) . Thesemolecular tools are extremely powerful andmake it possible to initiate metabolic engineeringprograms that allow for new processingstrategies. Genetic modification has beenespecially important in developing strains ofyeast that can use the xylose derived from

394 K.A. Dawson

lignocellulose digestion (Ho et al., 1999,Krishnan et al., 1999) and the development ofS. cerevisiae strains that contain ß-glucosidaseand directly produce ethanol from cellulose (Choand Yoo, 1999). When combined with ourunderstanding of molecular genetics, oursophisticated understanding of yeasts and theirphysiological activities make them attractive asbiocatalysts in biorefineries.

FUNGI

The metabolic flexibility and unique substraterange of fungi make them potential contributorsto the modern biorefinery. However, relative tothe yeast, very few species have been evaluatedfor their contributions to commercial productionsystems (Table 5). These organisms can grow

on a variety of substrates and are readilymanipulated to produce a variety of useful endproducts. Fungi are often used to economicallyproduce industrial enzymes. This enzymeproduction often takes advantage of the abilityof these organisms to grow on solid substrates,such as wheat bran, with relatively low moisturecontents (around 50%). This characteristic makesit possible to grow these organisms in surfaceculture systems (solid state fermentation) withexceptionally high yields of enzyme. This hasbeen the basis for developing large scale ‘Koji’or solid state fermentation systems (SSF).Enzyme complexes produced in a SSF systemhave a number of unique characteristics,including a broad range of side or auxiliaryactivities, that make them more powerful thanpurified enzymes obtained from traditional liquidfermentation systems. These types of enzyme

Table 4. Some non-conventional yeast uses available for use in modern biorefineries.

Yeast Applications

Candida boidinii Vendable chemicals and enzymes production from fatty acids, methanol and ethanolCandida butyrii Ethanol production from xyloseCandida guilliermondii Xylitol and citric acid production from pentoses and celluloseCandida shehatae Ethanol production from xyloseCandida tenuis Ethanol production from xyloseCandida tropicalis Xylitol and other vendable chemicals produced from pentoses and petroleum base substratesCandida utilis Wide variety of organic chemicals from fibrous substratesCryptococcus laurentii Enzyme and organic acid production from waste oilsDebaryomyces hansenii Xylitol and citric acid production from xylose and hydrocarbonsGeotrichum spp. Enzyme productionHansenula polymorpha Protein production from methanolKluyveromyces fragilis Thermotolerant ethanol production from fiber hydrolysatesKluyveromyces lactis Ethanol, enzyme and heterologous protein production from wheyKluyveromyces marxianus Ethanol production from whey, xylose, and pectinPachysolen tannophilus Ethanol and xylitol production from xylose from pentosesPhaffia rhodozyma Carotenoid pigment production from corn by-products and wood hydrolysatesPichia pastoris Heterologous protein and enzyme production from methanol and oxygenated hydrocarbonsPichia stipitis Ethanol production from xylose and xylans (hemicellulose hydrolysates)Schizosaccharomyces pombe Ethanol production from simple sugarsSchwanniomyces occidentalis Halophilic biotranformations of starch to organic chemicalsYarrowia lipolytica Citric acid production from plant oils, hydrocarbons, and petroleum products

Table 5. Some fungi that may contribute to biorefinery processes.

Fungal group Recognized functions

Aspergillus niger Citric acid and enzyme production from a variety of substratesPenicillium spp. Enzyme and antimicrobial productionRhizopus oligospora Enzyme production from a variety of substratesTrichoderma reesei Enzyme production from fibrous substrates


systems could be attractive co-products in abiorefinery system where partially dried distillersgrains are available as a substrate. For example,it is possible that DDGS could be used as asubstrate for producing an enzyme such asRhizozyme™, which would make it unique notonly in ability to totally replace othersaccharifying enzymes (saving $0.04-$0.05gallon), but bring about an increase in yield sincehemicellulose/cellulose would be degraded.

It is also possible to think about metabolicengineering of fungal cultures to make themmore useful as tools in a modern biorefinery.While mutant strains of fungi have been usedfor many years to produce commercial enzymes,systems for routinely manipulating their geneticmakeup are still lacking. New frontiers will openup when reliable technologies for metabolicengineering of fungi become readily available.

BACTERIA

Bacteria have a long history of contributing tocommercial fermentation processes and are oftenconsidered in the development of uniqueprocesses in biorefinery systems. The ability ofbacteria to use multiple substrates and theirsusceptibility to genetic manipulation make themattractive candidates for biocatalytic roles.However, bacterial fermentations in commercialsettings generally tend to be less robust than thoseassociated with yeast and fungi. As a result, caremust be taken in preparing substrates andfacilities that will support bacterial fermentations.Many types of naturally-occurring bacteria havebeen examined for their potential role incommercial fermentation systems (Table 6).

Such organisms can provide a wide variety ofbiochemical activities that may be integrated intounique fermentations systems in a biorefinery.

Molecular manipulation and geneticengineering have also allowed for thedevelopment of many new strains of bacteriawith unique fermentation capability. Forexample, strains of bacteria that contain activelignocellulose-degrading enzymes have beendeveloped and can be used to provide innovativemicrobial activities in biorefineries. Geneticmanipulation of Escherichia coli provides oneof the most powerful tools for modernbiotechnology and provides the best studiedsystem for manipulating microbial processes.Manipulated strains of this organism arecommonly used to produce many types ofpharmaceutical products, enzymes, anddiagnostic reagents. Specific strains of this E.coli that produce ethanol have also beendeveloped (Ingram et al., 1999). Some of thesestrains can use a variety of substrates, includingsome of the pentose sugar derived from thedegradation of the hemicellulose fraction of plantfiber (Dien et al., 1999).

In recent years, there has been considerableinterest in using heat resistant thermophilicbacteria in industrial fermentation processes(Scopes, 1997). Many of these bacteria naturallyproduce mixtures of ethanol and organic acids,but tend to be less ethanol tolerant than yeast.These organisms are uniquely adapted to carryout fermentation at temperatures higher than50oC. Application of these types of organismsmay allow for an increase in the temperaturerange for biocatalytic activities and allow forhigh temperature strategies in biorefineries.

Table 6. Some bacteria that may have applications in modern biorefineries.

Bacterial group Application

Arthrobacter spp. Enzyme productionBacillus spp. Enzyme productionClostridium spp. Acetone and butanol productionErwinia spp. Ethanol and organic acid production from hexoses and pentosesEscherichia coli Ethanol and organic acid production from hexoses and pentosesKlebsiella oxytoca Ethanol and organic acid production from hexoses and pentosesLactobacillus spp Used in a wide variety of food productsPseudomonas spp. Enzyme production and complex biotransformationsStreptomyces spp. Organic chemical, antimicrobial and enzyme productionThermoanaerobacter spp. Thermophilic ethanol productionZymomonas mobilis Ethanol from fiber and starch hydrolysates

396 K.A. Dawson

Other tools from biotechnology

In addition to allowing for the development ofnew strains of microorganisms, molecularbiology and metabolic engineering techniquesare also beginning to provide other new systemsand processes that can be used for controllingchemical processes in biorefineries. Specifically,new enzymes and enzyme expression systemshave been strategically designed for specificmetabolic transformations and are currentlybeing produced and evaluated in a number ofapplications. Much of this work is focusing onthe development of cellulases and xylanases thatwill allow for more efficient conversion offibrous biomass into fermentable sugars. Currentresearch programs in these areas not only lookat increasing the yields of enzymes, but also atthe production of more efficient enzymesystems. Tools such as proteomic analysis,molecular flux analysis, genomic analysis anddirected molecular evolution will be importantin developing the next generation of enzymesfor use in biorefineries (Bon and Picataggio,2002). The development of such enzyme-basedsugar production platforms has been a key partof the strategies for economically convertingrenewable fibrous biomass to ethanol. It can beexpected that similar techniques can be used todevelop other enzyme systems to carry out keytransformation in biorefineries. The alcoholindustry can look forward to the application ofmany new enzymes and enzyme complexes thatwill increase the substrate ranges and resistanceto some of the environmental extremesassociated with industrial fermentations. Thesewill improve the efficiency of ethanol productionand improve the value of co-products.

Conclusions

The future of many fermentation plants willdepend on their ability to rapidly adapt tochanges in the cost of substrates and to obtainthe maximum value from the products theyproduce. The concept of developing flexiblebiorefinery systems that allow for the use of avariety of feedstocks and that have the ability toproduce several high value co-products will beimportant when determining the value of alcoholproduction systems. Biotechnology will provide

the basic tools needed to develop these integratedbiological based processing systems for use inbiorefineries.

References

Abbas, C.A. and M. Cheryan. 2002. Emergingbiorefinery opportunities. Appl. Biochem.Biotech 98-100:1147.

Bon, E.P.S. and S. Picataggio. 2002. Enzyme andmicrobial biocatalysts. Appl. Biochem.Biotech. 98-100:163.

Cho, K.M. and Y.J. Yoo. 1999. Novel SSF processfor ethanol production from microcrystallinecellulose using the d–integrated recombinantyeast, Saccharomyces cerevisiae L2612dGC.J. Microb. Biotech. 9(3):340-345.

Dale, N. and A. Batal. 2002. Ingredient analysistable: 2002 Edition. Feedstuff 74:16-22.

Dien, B.S., L.B. Iten and R.J. Bothast. 1999.Conversion of corn fiber to ethanol byrecombinant E. coli strain FBR3. J. Indust.Microb. Biotech. 22:272-581.

Gong, C.S., N.J. Cao, J. Du and G.T. Tsao. 1999.Ethanol production from renewable resources.Adv. Biochem. Engin./Biotech. 65:207-241.

Ho, N.W.Y., Z. Chen, A.P. Brinard and M. Sedlak.1999. Successful design and development ofgenetically engineered Saccharomyces yeastsfor effective cofermentation of glucose andxylose from cellulosic biomass to ethanol.Adv. Biochem. Engin./Bioctech. 65:164-191.

Ingram, L.O., H.C. Aldrich, A.C.C. Borges, T.B.Causey, A. Martinez, F. Morales, A. Saleh, S.A.Underwood, L.P. Yomano, S.W. York, J.Zaldivar and S. Zhou. 1999. Enteric bacterialcatalysts for fuel ethanol production. Biotech.Progress 15:855-866.

Kaylen, M., D.L. Van Dyne, Y.S. Choi and M.Blasé. 2000. Economic feasibility of producingethanol from lignocellulosic feedstocks.Bioresource Tech. 72:19-32.

Krishnan, M.S., N.W.Y. Ho and G.T. Tsao. 1999.Fermentation kinetics of ethanol productionfrom glucose and xylose by recombinantSaccharomyces 1400(pLNH33). Appl.Biochem. Biotech. 77-79:373-388.

Lynd, L.R., J.H. Cushman, R.J. Nichols and C.E.Wyman. 1991. Fuel ethanol from cellulosicbiomass. Sci. 251:1318-1323.


Mouiruzzaman, M., B.S. Dien, C.D. Skory, Z.D.Chen, R.B. Hespell, N.W.Y. Ho, B.E. Dale andR.J. Bothast. 1997. Fermentation of corn fibresugars by an engineered xylose utilizingSaccharomyces yeast strain. World J. Microb.Biotech. 13:341-346.

Olsson, L. and J. Nielsen. 2000. The role ofmetabolic engineering in the improvement ofSaccharomyces cerevisiae: utilization ofindustrial media. Enzyme Microb. Tech.26:785-792.

Scopes, R.K. 1997. Ethanol from biomass: Thepotential use of thermophilic organisms infermentation. Australasian Biotech. 7:296-299.

Torre, P. 1999. Co-products from ethanolfermentation: Alternatives for the future. In:The Alcohol Textbook, 3rd Edition (K.A.Jacques, T.P. Lyons and D.R. Kelsall, eds).Nottingham University Press, Nottingham, UK,pp. 335-346.

The Alcohol Alphabet 399

The Alcohol Alphabet

A glossary of terms used in the ethanol-producing industries

Italic type denotes words (or minor variations of words) defined under separate entries.

A

Å Abbreviation for Angstrom.

Absolute ethanol A pharmaceutical term foranhydrous ethanol. It is generally defined ashaving less than 1% water.

Acetaldehyde Otherwise known as ethanal,acetic aldehyde or ethylaldehyde. A clearflammable liquid with a characteristic pungentodor. Chemical formula CH

3CHO. Boils at

21°C and freezes at -123.5°C. It is misciblein both ethanol and water. It has a narcoticeffect on humans, and large doses may causedeath by respiratory paralysis. It is a congenerin the production of ethanol by fermentation,and is usually a major constituent of the headsfraction removed in rectification.

Acetic acid A colorless liquid with a pungentodor. Flammable at high concentrations.Chemical formula CH

3COOH. Acetic acid

may be produced from ethanol by Acetobacterbacteria under aerobic conditions such aswhen a completed fermentation is agitated oraerated excessively. Vinegar is a solution ofthis acid.

Acetobacter A genus of Gram-negative,aerobic bacteria comprising ellipsoidal to rod-shaped cells as singles, pairs or chains.Otherwise known as acetic acid or vinegarbacteria, they are able to oxidize ethanol toacetic acid. They may be responsible for lossof yield in ethanol production if a fermentedmash is agitated or aerated excessively.

Acetone Otherwise known as 2-propanone,dimethyl ketone or pyroacetic ether. It is avolatile, very-flammable liquid with acharacteristic pungent ‘mousey’ odor and asweetish taste. Chemical formula CH

3COCH

3.

Boils at 56.5oC and freezes at -94oC. It ismiscible with water, ethanol and most oils.Inhalation may cause headaches, and in largeamounts, narcosis. It is a congener in theproduction of ethanol by fermentation,particularly from molasses mashes. It is alsoa by-product of the production of butanol byanaerobic fermentation using the bacteriumClostridium acetobutylicum. It tends toconcentrate in the heads fraction in therectification of neutral spirit.

Acid A compound capable of releasinghydrogen ions.

John MurtaughSept 12, 1936 - Jan 25, 2003

Beginning with his dissertation work on the cooking process through a lifetime of consulting, educating andinspiring others, Dr. John Murtaugh moved the ethanol industry forward. He began this glossary with the2nd edition of the Alcohol Textbook. Like the rest of John’s work, it has an important and very practicalpurpose. As such, it will be updated and continued in this and successive editions of the book.

400 The Alcohol Alphabet

Acidity A quantitative measure of the amountof acid present.

Acid-acid process Term used in starchprocessing when acid hydrolysis is used toaccomplish both the initial liquefaction andthe final saccharification to simple sugars.

Acid-enzyme process Term used in starchprocessing when acid hydrolysis is used toaccomplish the initial liquefaction, and anenzyme such as amyloglucosidase is used forthe saccharification to simple sugars.

Acid hydrolysis The hydrolysis of a polymerby the use of acid. In the case of starchhydrolysis, acids may be used as analternative to enzymes in either (or both) theliquefaction or saccharification processes.

Acid washing A process in which yeastrecovered from a finished fermentation isacidified to pH 2.2 for ~20 minutes using anacid like phosphoric acid, to reduce the levelof bacterial contamination prior to recyclinginto a new fermentation.

Active dry yeast A yeast preparation made fromcompressed yeast by careful and controlledremoval of moisture to ~5% w/v, such that10-50 billion yeast/gram remain viable.Storage for a number of months allows plantsto eliminate risky propagation ‘in-house’.

Aguardiente An unaged alcoholic beverageproduced in Central and South America bythe distillation of beer derived from thefermentation of sugarcane juice or molasses.It is similar to crude rum.

Alcohol A member of a class of organiccompounds containing carbon, hydrogen andoxygen. Considered to be hydroxyl derivativesof hydrocarbons produced by thereplacement of one or more hydrogen atomsby one or more hydroxyl (OH) groups. Underthe International Union of Pure and AppliedChemistry (IUPAC) naming system, the namegiven to an alcohol is derived from the parenthydrocarbon with the final ‘e’ changed to ‘ol’.Thus methane-methanol, ethane-ethanol, etc.

The principal alcohol in fuel and beverageuse is ethanol (otherwise known as ethylalcohol). The lower molecular weight alcohol(methanol) and higher alcohols are moretoxic.

Alcohol Fuel Permit (AFP) A permit issued bythe Bureau of Alcohol, Tobacco and Firearmsallowing the holder to engage in theproduction of ethanol solely for fuel use.

Aldehyde A member of a class of organiccompounds considered to be derived by theremoval of hydrogen atoms from an alcohol.Aldehydes tend to be produced as congeners,or by-products of fermentation; and having alower boiling point than ethanol, they tend tovaporize more readily and may accumulateas a ‘heads’ fraction in the distillation process.

Alkali A compound capable of neutralizinghydrogen ions, usually by generatinghydroxyl ions which combine with hydrogenions to form water.

Alkalinity Specifically in water, a measure ofthe total bicarbonate and carbonate content,determined by titration to an endpoint usingvarious indicators such as phenylphthalein,methyl orange, etc.

Alpha-amylase (ααααα-amylase) An enzyme usedin the liquefaction of starch in the grainmashing process prior to saccharification andfermentation. α-amylase hydrolyzes the long-chain starch molecules into short-chaindextrins. These are more suitable forsubsequent saccharification by other enzymesto fermentable glucose. (α-amylase is anendoenzyme in that it works from the insideof the amylose molecule). In beverage alcoholproduction α-amylase may be derived frommalt (sprouted barley), but in fuel ethanolproduction the enzyme is obtained solely asa bacterial product. The enzyme moleculecontains a calcium atom, which is essentialfor its activity.

American Society for Testing and Materials(ASTM) A scientific and technical


organization with headquarters inPhiladelphia, PA established for ‘thedevelopment of standards on characteristicsand performance of materials, products andservices and the promotion of relatedknowledge’. It sets voluntary consensusstandards through committees representingproducers and users. ASTM has publishedstandards for fuel ethanol and gasoline thathave been adopted by many states.

Amyl alcohol The principal constituent of fuseloil. Otherwise known as pentanol. Chemicalformula C

5H

11OH. Eight isomers exist, the

most common being primary isoamyl alcohol.

Amylase The name given to any enzyme thathydrolyzes (or breaks down) amylose, whichis a major component of starch.

Amyloglucosidase An enzyme, also known asglucoamylase, which hydrolyzes amylose intoits constituent glucose units. It is anexoenzyme in that it works from an outer endof the starch molecule. It is normally used inconjunction with α-amylase for theliquefaction and saccharification of starch inthe grain mashing process, prior tofermentation. Amyloglucosidase is a productof fungal growth produced either in liquidfermentation or surface culture fermentation(See RhizozymeTM).

Amylopectin A major component of starch(together with amylose). The molecule iscomposed of large, branched chains ofthousands of glucose units.

Amylose A major component of starch (togetherwith amylopectin). The amylose molecule iscomposed of straight chains of hundreds ofglucose units. In the grain-mashing processfor ethanol production, amylose may first bebroken down into short-chain dextrins by α-amylase, which are in turn broken down intosingle glucose units by amyloglucosidase.

Anaerobic Literally means ‘without air’. Theopposite of aerobic. For example, yeastpropagation (or multiplication) is more rapid

with aeration (i.e., under aerobic conditions),while yeast produces ethanol under anaerobicconditions of fermentation.

Anaerobic digestion Process of breaking downwaste materials by anaerobic bacterialdegradation. Normally accompanied by theproduction of methane gas.

Angstrom (Å) A unit of length equal to onehundred-millionth (10-8) of a centimeter, usedfor measuring the diameter of chemicalmolecules. Thus, in ethanol dehydration, amolecular sieve material with holes 3Angstroms in diameter may be used toseparate water, which has a 2.5 Å diameter,from ethanol, which has a 4.5 Å diameter.

Anhydrous Literally means ‘without water’. Theterm used for a substance that does not containwater. Ethanol for fuel use is commonlyreferred to as anhydrous, because it has hadalmost all of the water removed. See absoluteethanol.

Antibiotic A chemical substance produced bymicroorganisms that kills or inhibits growthof other organisms. Both single antibiotics andcombinations (See LactocideTM) are used tocontrol bacterial contamination in thefermentor.

Antifoam (or defoamer) A preparationcomposed of substances such as silicones,organic phosphates and alcohols that inhibitsformation of bubbles in a liquid duringagitation by reducing the surface tension.Antifoam may be used in ethanol productionto control the development of foam infermentors. Antifoam may also be used tocontrol foam in beer distillation to increasethe stripping capacity of a column.

Antiscalant (or scale inhibitor) A chemicalcompound or mixture of compounds addedto water or molasses beer to reduce theincidence of scaling in heat exchangers ordistillation columns. Usually, the principalingredients of antiscalants are chelatingcompounds.


Arabinose A pentose sugar comprising a majorconstituent of hemicellulose. Chemicalformula C

5H

10O

5. It is not fermented by

normal strains of distillers yeasts. Also knownas pectin sugar or gum sugar.

Atomic weight The relative mass of an atombased on a scale in which a specific carbonatom (carbon 12) is assigned a mass value of12. Also known as relative atomic mass.

Azeotrope The term used to describe a constantboiling mixture. It is a mixture of two (ormore) components with a lower boiling pointthan either component alone. For examplewater, which boils at 100oC, and anhydrousethanol, which boils at 78.5oC, form aconstant boiling mixture (azeotrope) at78.15oC. The vapor of the mixture has thesame composition as the liquid and thereforeno further concentration can be achieved bynormal distillation. Under normal pressuresit contains approximately 97% by volumeethanol (1940 proof). It is very expensive interms of energy to attempt to reach 1940 proof,so 1900 proof is generally considered to bethe practical, economic azeotrope limit for fuelethanol distillation.

Azeotropic distillation A distillation processin which a liquid compound (entrainer) isadded to the mixture to be separated to forman azeotrope with one or more of thecomponents. Normally, the entrainer selectedis easily separated from the component to beremoved. For example, when benzene is usedin azeotropic distillation to dehydrate ethanol,the overhead condensate phase-separates toyield a water-rich layer that can be withdrawnand a benzene-ethanol layer, which isrefluxed.

B

Backset Recycled thin stillage. It may be addedto the cooker or to the fermentor and mayserve as a partial source of nutrients. It reducesthe water required for mashing and reduces

the volume of liquid residue to be evaporated.Improperly handled, it may be a major sourceof bacterial contamination, and may containhigh levels of chemicals that cause stuck/sluggish fermentations.

Bacteria Microscopic organisms of the kingdomMonera usually characterized by small size,vigorous biochemical activity and lack of atrue nucleus.

Bacterial contamination The conditionoccurring when undesirable bacteria becomeestablished in a fermenting mash and reducethe ethanol yield. The bacteria use availablesugars to produce various compounds(congeners), particularly acids, which mayinhibit yeast activity. In severe situations,bacterial contamination may cause seriouseconomic losses.

Balling (or Brix or Plato) A scale used tomeasure the specific gravity of a liquid inrelation to that of a solution of sugar in water.Each unit on the scale is equivalent to 1% byweight of sugar. Thus a mash of 20o Ballinghas the same specific gravity as a 20% w/wsugar solution. The scale is frequentlyconsidered to indicate % dissolved solids in aliquid, although this is only true of solutionsof pure sucrose. Traditionally, the term‘Balling’ has been used in grain distilleriesand breweries, while ‘Brix’ has been used insugar mills and rum or molasses alcoholdistilleries. The measurement is accomplishedby use of a Balling (or Brix/Plato) hydrometer,refractometer or more recently density meter.

Barbet time See Permanganate time.

Bargeload Generally refers to a river barge witha capacity of 10,000 barrels or 420,000gallons.

Barrel A liquid measure equal to 42 US gallonsor 5.6 cubic feet. Or, a wooden container usedfor the aging and maturation of alcoholicbeverages. Barrels used for whisky maturationare made of oak wood and have a capacity of


about 52 US gallons. Barrels may only beused once for aging bourbon whisky, so thereis a worldwide trade in used bourbon barrelsfor aging other alcoholic products such asScotch whisky and rum. A US beer barrelcontains 31 US gallons.

Base losses The percentage of ethanol lost inthe stillage at the base of a beer strippingcolumn or a rectifying column. It is virtuallyimpossible to achieve zero base losses, and itwould be wasteful of steam. The base lossesare generally monitored and controlled in anoptimal range determined by steam costs, etc.

BATF Abbreviation for US Bureau of Alcohol,Tobacco and Firearms.

Batch cooking Cooking a set amount of grainmeal, water and backset (if any) in a singlevessel as a discontinuous operation. One ormore batch cooks may be used to fill a singlebatch fermentor. Batch cooking is mainlyconfined to the beverage alcohol industry andsmaller plants in the fuel ethanol industry.

Batch distillation Distilling batches of beer ina discontinuous operation. It is not commonlyused in fuel ethanol production, but is usedin the beverage ethanol industry particularlyfor the production of special types of heavily-flavored distillates.

Batch fermentation The fermentation of a setamount of mash in a single vessel in adiscontinuous operation.

Beer The name given to the product of alcoholfermentation. In grain alcohol production,beer may contain from 9->16% ethanol (v/v).

Beer preheater A heat-exchanging device usedfor heating beer before it enters a beer still.Usually, it also serves as the first condenserfor the beer still, so that the overhead vaporsheat the beer.

Beer still The distillation unit used for the initialremoval of ethanol from finished beer. Itgenerally consists of a stripping section that

extracts the ethanol from the beer and aconcentrating or rectifying section, whichnormally takes the ethanol up to 1900 proof(95o GL). Beer stills may consist of a singletall column, or two or more columns standingside by side, linked by vapor pipes.

Beer stripping column See Beer still.

Beer well The holding vessel into which finishedbeer is transferred prior to distillation.

Benzene Colorless, flammable, aromatichydrocarbon liquid. Chemical formula C

6H

6.

Boils at 80.1OC and freezes at 5.4OC. Used asan entrainer for the dehydration of ethanolby azeotropic distillation. Known to becarcinogenic.

Beta-amylase (ß-amylase) An enzyme thathydrolyses the long-chain amylose moleculesin starch into fermentable maltose, the dimer(or double-molecule) of glucose. It is anexoenzyme in that it works from an outer endof the molecular chain. It is found in malt(sprouted barley) in association with α-amylase. With the advent of microbialamyloglucosidase enzymes, malt amylasesare generally only used in the production ofheavily-flavored beverage alcohol.

Betaglucan (ß-glucan) Gum-like polymers ofß-linked glucose as in cellulose instead of theα-linked glucose units as in starch (amylose).Betaglucan is commonly present in barleymashes. It is not broken down by α-amylaseand causes foaming problems due to itsviscous elastic nature.

Betaglucanase (ß-glucanase) An enzyme whichhydrolyses betaglucan. It is frequently usedin barley mashing to reduce foaming andviscosity problems.

Beverage alcohol Any form of distilled ethanolwith or without congeners considered by lawto be fit for human consumption. The laws inmany countries confine beverage alcohol tothat produced by fermentation as opposed toalcohol of synthetic origin.


Binary azeotrope An azeotrope or constantboiling mixture having two components suchas ethanol and water.

Bioethanol See fermentation ethanol.

Biofilm A layer of living microorganisms, theirpolysaccharide by-products, and other soils.Biofilms form on surfaces inside process pipesand equipment and can be very potent sourcesof contamination.

Biogas The gas produced by anaerobicdigestion of wastes. It is mainly methane.

Biological oxygen demand (BOD) A measureof the oxygen-consuming capacity of organicmatter contained in effluents. Duringdecomposition, organic effluents have anoxygen requirement that may deplete thesupply in a waterway and result in the deathof fish and other aquatic life. BOD is normallymeasured on the basis of the weight of oxygenrequired per unit volume of an effluent.

Biomass Any renewable organic matter suchas agricultural crops, crop waste residues,wood, animal and municipal wastes, aquaticplants, fungal growth, etc.

Blackstrap See molasses.

Bleach An oxidizing agent added to detergentsto enhance overall cleaning. Bleaches breakdown long chain protein and polysaccharidemolecules. Sodium hypochlorite andhydrogen peroxide are effective bleachesused in detergent formulations.

Blended whisky Defined by the US Bureau ofAlcohol, Tobacco and Firearms as a mixturecontaining at least 20% of straight whisky ona proof gallon basis, together with otherwhiskies or neutral spirits. When a blendcontains more than 51% of a straight whisky,it may be designated by that specific type,such as ‘blended rye whisky’.

Blender Tax Credit A US federal income taxcredit granted to blenders of ethanol and

gasoline as an alternative to the excise taxexemption. It was originally introduced underthe Crude Oil Windfall Profits Tax Act of 1980,at a level of 40 cents per gallon, and wassubsequently raised to 50 cents by the SurfaceTransportation Assistance Act of 1982, andto 60 cents by the Deficit Reduction Act of1984.

BOD See Biological Oxygen Demand.

Boiler efficiency The thermal efficiency of aboiler in terms of the usable energy output(in the form of steam) in relation to the energyinput in the form of fuel. Boiler efficienciesare commonly in the 70-85% range.

Boiling point The temperature at which thevapor pressure of a liquid equals the totalpressure of the atmosphere above it. Bubblesof vapor can be formed when vapor pressureslightly exceeds atmospheric pressure. Boilingthen takes place as heat is supplied, changingthe liquid phase to the vapor phase. Theboiling point of a pure substance at constantpressure does not change.

Bourbon whisky Defined by the BATF as whiskyproduced at not more than 1600 proof from afermented mash of not less than 51% corn,and stored at not more than 1250 proof incharred, new oak barrels. In practice, thebourbon whisky mash bill is frequently about65% corn, 25% rye, and 10% barley malt.

Brandy Defined by the BATF as ‘an alcoholicdistillate from the fermented juice, mash orwine of fruit, or from the residue thereof,produced at less than 1900 proof, in such amanner that the distillate possesses the taste,aroma and characteristics generally attributedto the product’.

British Thermal Unit (BTU) The amount ofheat required to raise the temperature of onepound of water one degree Fahrenheit underdefined pressure conditions. It is the standardunit for measuring heat energy in the US.

Brix See Balling.


BTU Abbreviation for British Thermal Unit.

BTX Abbreviation for the three related octaneenhancers, benzene, toluene and xylene.

Bubble cap A contacting device used on somedistillation plates. It consists of a cylindricalchimney set in a hole in the plate and coveredby a dome-shaped cap, which deflects thevapors rising up the chimney to cause themto pass through the liquid layer on the topside of the plate.

Bushel A unit of dry volume equal to 2150.42cubic inches or 1.244 cubic feet. When usedto measure grain, bushel weight depends onthe type and condition of the grain. In the caseof corn, bushel volume generally averagesabout 56 lbs by weight. This has led to theuse of a distillers bushel, which represents 56lbs of grain, regardless of type or volume.

Butanol (butyl alcohol) A minor constituent offusel oil. Chemical formula C

4H

9OH. Four

isomers exist. They are all colorless, toxicflammable liquids. N-Butanol may beproduced as a co-product with acetone andethanol by the fermentation of selectedcarbohydrates with the anaerobic bacteriumClostridium acetobutylicum. Butanols areused as solvents and chemical intermediates.

By-products Products that are secondary to theprincipal product of a process. In ethanolproduction, carbon dioxide and distillersdried grains are normally considered by-products; but in certain circumstances theymay be viewed as co-products, in that theymay contribute significantly to the overallprocess economics.

C

Carbohydrate Any of a group of compoundscomposed of carbon, hydrogen and oxygenincluding the sugars, starches, dextrans andcelluloses. They are the most abundant classof organic compounds in nature, constituting

approximately 75% of the dry weight of allvegetation.

Carbon dioxide A colorless non-flammablegas. Composition CO

2. It does not support

human respiration; and in high concentrationsit causes asphyxiation. It is approximately 1.5times the weight of air, and tends toaccumulate in floor drains, pits and in thebottoms of unventilated tanks. It is producedby various means, notably the combustion offuels in an excess of air and is a by-productof yeast fermentation. It may be recoveredfrom fermentations and compressed to a liquidor solid (dry ice), or used to carbonatebeverages.

Carbon monoxide A colorless, odorless,flammable gas (‘coal gas’). Composition CO.It is poisonous if inhaled, as it combines withblood hemoglobin to prevent oxygen transfer.It is very slightly lighter than air. It isproduced by the incomplete combustion offuels with a limited oxygen supply, as in autoengines. It is a major component of urban airpollution, which can be reduced by blendingan oxygen-bearing compound (or oxygenate)such as ethanol into hydrocarbon fuels.

Carbon steel A steel deriving its particularproperties from its content of carbon. Thesesteels may range from ‘low-carbon’, (havingless than 0.25% carbon), to ‘high-carbon’.Low carbon or ‘mild’ steel is easily hot-worked and rolled, and is used in steel platesand beams. It is easily welded, but is readilysubject to rusting and other forms ofcorrosion.

Caribbean Basin Initiative (CBI) The programarising from the Caribbean Basin EconomicRecovery Act passed by the US Congress in1983 to encourage industrial development inthe Caribbean islands and Central America.Under this program, fuel ethanol and otherproducts from the region may enter the USwithout being subject to customs duties.

Cassava A root crop with a high starch content


grown in the tropics and subtropical regions.Known in Brazil as manioc, it is used as analternative to sugarcane as a feedstock forethanol production. It is also processed forfood as ‘tapioca’.

CBI Abbreviation for Caribbean BasinInitiative.

CCC Abbreviation for Commodity CreditCorporation.

CDA Abbreviation for completely denaturedalcohol.

Cell recycle The process of recovering yeastfrom fermented beer to return it to the startingvessel of a continuous fermentation or to anew vessel in a batch fermentation system. Itmay include an acid washing step to reducebacterial contamination.

Cellulase An enzyme capable of hydrolyzinglong-chain cellulose molecules into simplesugars or short-chain polymers.

Cellulose The principal polysaccharide in livingplants. It forms the skeletal structure of thecell wall, hence the name. It is a polymer ofglucose units coupled by ß-type linkages intochains of 2,000-4,000 units. Cellulosenormally occurs with other polysaccharidesand hemicelluloses derived from sugars suchas xylose, arabinose and mannose.

Celsius (or centigrade) A temperature scale inwhich (at normal atmospheric pressure) waterfreezes at zero degrees and boils at 100degrees.

Centrifugal pump A machine for movingliquids by accelerating them radially outwardsby means of a rotating impeller containedwithin a casing. It is the most common andmost versatile type of pump in normalindustrial use. It has an advantage in that itdoes not force a positive displacement ofliquids, so it may be used in systems whereoutput must be throttled or completely closedby control valves. This feature makes

centrifugal pumps unsuitable for movingheavy viscous liquids such as molasses andsyrups.

Centrifuge A machine for separating insolubleliquids or solids from liquids by theapplication of centrifugal force. Two maintypes of centrifuges are basket and decanter.In basket centrifuges, solids are retained in arotating perforated basket while liquid passesthrough the perforations. In decantercentrifuges, the mixture is thrown against asolid-walled cylinder and the heavier solidparticles collect against the wall for removalwhile the lighter liquid collects in the centralpart. Centrifuges are commonly used inethanol plants for yeast recovery and instillage dewatering.

Chelating agent A chemical that combines withand solubilizes water hardness ions to preventprecipitation and scale formation. Chelatingagents are also used in detergent formulationsto remove scale deposits. EDTA (ethylenediamine tetraacetic acid) is a chelating agent.

Chemical oxygen demand (COD) A laboratorytest to determine the oxygen requirements forthe chemical digestion of effluents. It shouldbe distinguished from the BOD test, whichdetermines the total of both chemical andbiological oxygen requirements. COD ismeasured in a rapid test that involves heatingthe effluent in the presence of an oxidizingagent such as potassium dichromate and thendetermining the amount of oxygen absorbedby the effluent.

Chlorine dioxide Chemical formula ClO2. It is

a strongly-oxidizing, yellow-to-reddish-yellow gas at room temperature. It has anunpleasant odor similar to that of chlorine andreminiscent of nitric acid. It is unstable inlight. It reacts violently with organic materialsand is easily detonated by sunlight or heat inconcentrations greater than 10% at atmos-pheric pressure. Boils at 11oC and freezes at-59oC. Chlorine dioxide may be used as asterilant, and may be produced in situ for


sterilizing yeast mashes by addition of sodiumchlorite solution in the presence of acids orchlorine (or hypochlorite solution). It isconsiderably more effective as a sterilant thanstraight chlorine.

Chromatography A method for separating amixture of chemical compounds intoindividual components by selectivedistribution between two immiscible materials(or phases), one stationary and the othermobile. The phases are selected so that themobile phase will carry the variouscomponents through the stationary (or solid)phase at differing rates to give separation. Gaschromatography may be used to separateethanol from the other congeners producedin fermentation and to measure themquantitatively. High performance liquidchromatography may be used to follow starchhydrolysis in grain alcohol production.

CIP Abbreviation for cleaning-in-place system.

Citrus molasses A by-product of the citrus juiceindustry. Citrus residue, mainly peel, is treatedwith lime and then passed through a press.The press liquor is then evaporated to aviscous, dark brown molasses of about 72o

Brix. Citrus molasses is similar to caneblackstrap molasses, having about 45% totalsugars. However, it has more protein and amuch lower ash content. Citrus molasses maybe diluted and fermented for ethanolproduction, but it may need pretreatment toreduce the content of d-limonene (commonlyreferred to as ‘citrus-stripper oil’), which tendsto inhibit yeast growth.

Cleaning In processing systems, cleaning isremoving soils in the equipment to a degreethat is acceptable for the process.

Cleaning-in-place system (CIP) A systemdesigned to permit process equipment to becleaned without disconnecting ordismantling. A sophisticated automatic CIPsystem may provide a sequence of waterflushing cycles, detergent washing and

chemical sterilizing of equipment at the pressof a button.

Cleaning-out-of-place (COP) The manualdisassembly, inspection and cleaning ofprocessing equipment and systems. Somedegree of COP is usually required with CIP(cleaning-in-place).

Closed receiver See Receiver.

CMS Abbreviation for condensed molassessolubles.

Coccus (plural: cocci) A type of bacteria withspherically-shaped cells. They may occur assingle cells, clusters or long chains.

COD Abbreviation for chemical oxygendemand.

Column A vertical cylindrical vessel containinga series of perforated plates or other contactdevices through which vapors may pass toeffect a separation of liquid mixtures bydistillation.

Co-mingled tank The term used for fuel ethanoltanks at refineries or pipeline terminals wheretwo or more suppliers may share the sametank for storing ethanol.

Commodity Credit Corporation (CCC) Anagency of the USDA established to stabilizeand protect farm incomes and prices bymaintaining balanced and adequate suppliesof agricultural commodities.

Completely denatured alcohol (CDA) A termused by the BATF to describe ethanol madeunfit for human consumption by addition ofspecified denaturants such as methyl isobutylketone, kerosene or gasoline.

Compressed yeasts A paste of viable yeasts ofapproximately 30% total solids prepared byfiltration of yeast slurry. Refridgeration isrequired at 4°C, and such yeasts loose viabilityconstantly with a shelf life of less than 3weeks.


Concentrating column A column wherehydrous ethanol is concentrated to decreasewater content. A beer still may consist of astripping column in which dilute ethanol isremoved from beer and a concentratingcolumn, which receives ethanol vapor fromthe stripping column and concentrates it upto about 1900 proof (95oGL). If provision ismade in a concentrating column to removeimpurities such as fusel oil, then it is morecorrectly a rectifying column.

Condensate Liquid condensed from vapor in acondenser.

Condensed molasses solubles (CMS) The termused to describe molasses stillageconcentrated by evaporation. The molassesresidue (after fermentation and distillation)may be concentrated to about 60° Brix (orapproximately 60% solids) to be sold as asubstitute for molasses in animal feeds as acaking agent and dust suppressant. It containshigh concentrations of salts.

Condenser A heat exchange device connectedto the vapor discharge pipe of a column topermit the vapor to be cooled and condensedto a liquid. Condensers are commonlycylindrical vessels containing tubes throughwhich cooling water is passed.

Congeners Chemical compounds producedwith ethanol in the fermentation process. Theyare frequently referred to as impurities.Common congeners are methanolacetaldehyde, esters (such as ethyl acetate)and fusel oils (higher alcohols, particularlyamyl alcohols). Fermentation conditions maybe adjusted to control congener formationdepending on the requirements for the endproduct.

Constant boiling mixture See Azeotrope.

Continuous cooker A system into which a mashof water, grain and enzymes may be fedcontinuously to be cooked and discharged tothe fermentation system. Continuous cookersgenerally consist of a slurry tank connected

by a pump to a steam jet heater (jet cooker), aholding vessel or lengths of piping (to providesome residence time at the cookingtemperature), one or more flash vessels (tocool the cooked mash), a holding vessel forenzymatic liquefaction and a heat exchangerfor final mash cooling. Continuous cookersare more common in fuel ethanol plants thanin beverage alcohol plants.

Continuous fermentation A system into whichcooked mash may be fed continuously to befermented and then discharged to the beer welland distillation system. Continuousfermentation systems generally consist of aseries of interconnected tanks sized to providesufficient residence time for the fermentationto proceed to completion. The ethanolindustry is divided on fermentation systems,with more gallonage produced by continuousfermentation than by batch fermentation.

Continuous distillation A process usingspecially-designed equipment to permit avolatile component such as ethanol to beseparated by distillation from a continuousflow of an aqueous solution such as beer.

Control loop A portion of a process controlsystem that includes a sensing deviceconnected to a signal transmitter, a controller,another signal transmitter and an actuator. Forexample, a pressure control loop on adistillation column may have a pressuresensor connected through a controller to asteam flow regulating valve.

Cooker A device for heating a slurry of grainand water to a sufficiently-high temperaturefor sufficient time to release and gelatinizethe starch in the grain and thereby render itsusceptible to enzymatic hydrolysis. Livesteam is normally used for heating the slurry;and pumps or agitators are used to ensuremixing and even heating. Cooking may beperformed continuously or in a batch mode.

Cooling tower A tower or other type of structurewhere air (the heat receiver) circulates in direct


or indirect contact with warmer water (theheat source) to cool the water. Cooling towersare used in ethanol production plants torecirculate cooling water and to minimize theamount of water used from wells, rivers orpublic sources.

Cooper A person who makes or repairs woodenbarrels.

Cooperage A place where wooden barrels aremade or repaired. Also used to refer to asupply of barrels (i.e., the product of the workof a cooper).

COP The abbreviation for cleaning-out-of-place.

Co-products Where the economics of aproduction process depend on the value ofmore than just the primary product, thesecondary products are referred to as co-products rather than as by-products. Forexample, distillers dried grain may beconsidered a co-product of the production ofethanol from dry milled grain.

Cordials See liqueurs.

Corn steep liquor The concentrated liquid thathas been used for steeping and softening cornprior to wet milling. It contains extracted lowmolecular weight chemicals from the corn,and serves as a nutrient source in subsequentfermentation of the starch stream.

Corn whisky Defined by the BATF as ‘whiskyproduced at under 1600 proof from afermented mash containing not less than 80%corn grain’. If corn whisky is stored in oakbarrels, the BATF stipulates that the proofshould not be more than 125o, and that thebarrels be used, or uncharred if new.Furthermore, the whisky cannot be subjectedto any treatment with charred wood.

Corrosion The destruction, degradation ordeterioration of material (generally metal) dueto the reaction between the material and itsenvironment. The reaction is generallychemical or electrochemical, but there are

often important physical and mechanicalfactors in the corrosion process.

Co-solvent A liquid required to keep anotherliquid in solution in a third liquid. Forexample, in the Dupont Waiver for the use of5% methanol in gasoline, 2.5% ethanol (or ahigher alcohol) is required as co-solvent toimprove the miscibility of the methanol in thegasoline.

Crude Oil Windfall Profits Tax Act of 1980US federal legislation that extended to 1992the excise tax exemption for ethanol-gasolineblends granted under the Energy Tax Act of1978. It also introduced a blender tax creditof 40 cents per gallon of ethanol as analternative to the excise tax exemption.

Cyclohexane A colorless, flammable, alicyclichydrocarbon liquid of chemical formulaC

6H

12, that boils at 80.3oC, and freezes at

6.5oC. It is used as an alternative to benzeneas an entrainer in the dehydration of ethanolby azeotropic distillation.

D

DDG Abbreviation for distillers dried grain.

DDGS Abbreviation for distillers dried grainwith solubles.

DE Abbreviation for dextrose equivalent.

Deadleg A length of mash piping closed eithertemporarily by a valve, or permanently toleave a dormant pocket of mash that maybecome a source of bacterial contamination.

Dealer tank wagon (DTW) Used in referenceto fuel ethanol and gasoline sales, it isgenerally of 7,000-8,000 gallons capacity.

Decanter Vessel used for the separation of two-phase liquids. In a fusel oil decanter, an upperfusel oil phase is separated from a loweraqueous ethanol phase. In a benzene columnreflux decanter, the upper, mainly-benzene,


phase is separated from the lower, mainly-water, phase.

Deficit Reduction Act of 1984 US federallegislation that increased the excise taxexemption on ethanol-gasoline blends from5 cents (as set by the Surface TransportationAssistance Act of 1982) to 6 cents per gallon.It also increased the alternative blender taxcredit to 60 cents per gallon of ethanol.

Defoamer See Antifoam.

Dehydration The process of removing waterfrom a substance, particularly the removal ofmost of the remaining 5% of water from 1900

proof ethanol in the production of absoluteor anhydrous ethanol.

Demethylizing column Occasionally referredto as a supplementary column, it is afractionating column used to removemethanol in the production of neutral spiritand is located after the rectifying column. Thedemethylizing column is heated indirectly viaa reboiler. The impure spirit enters part wayup the column and the methanol is removedin the overhead vapor together with someethanol, while the bulk of the ethanoldescends to be removed at the base of thecolumn as a product relatively free ofmethanol.

Denaturant A substance added to ethanol tomake it unfit for human consumption so thatit is not subject to taxation as beveragealcohol. The BATF permits the use of 2-5%unleaded gasoline (or a large numberedsimilar, specified substances) for use asdenaturants for fuel ethanol. (See alsospecially-denatured alcohol).

Department of Energy (DOE) A departmentof the US federal government established in1977 to consolidate energy-orientatedprograms and agencies. The department’smission includes the coordination andmanagement of energy conservation, supply,information dissemination, regulation,research, development and demonstration.

The department includes an Office of AlcoholFuels.

Dephlegmator Name commonly used for thefirst of two or more condensers attached tothe overhead vapor line of a distillationcolumn. It literally means an entrained liquidseparator.

Desiccant A substance that absorbs water andcan be used for drying purposes. For example,potassium aluminosilicate is used as adesiccant in molecular sieve systems forethanol dehydration.

Detergent package A combination of detergentsand corrosion inhibitors normally added tofuel ethanol to impart a cleaning action toethanol-gasoline blends and thereby reduceengine fuel injector blockages.

Dewatering The removal of water (or otherliquids) from a solid material, particularly theuse of screens or centrifuges in the initialseparation of thin stillage (liquid) from thesolids contained in whole stillage in DDGprocessing.

Dextran A non-fermentable, large, branched-chain polymer of glucose molecules producedfrom sucrose in molasses by bacterialcontamination (mainly Leuconostocmesenteroides). It gives a ropey appearanceto molasses when stirred or poured, andreduces ethanol yield on fermentation.

Dextrin Short-chain polymers of glucosemolecules produced by the partial hydrolysisof starch with α-amylase or acid in the initialstage of conversion to fermentable sugars.

Dextrose An alternative name for glucose.

Dextrose equivalent (DE) A measure of thedegree of hydrolysis of starch. It is no longerconsidered as significant as previously inethanol production with the more generalacceptance of simultaneous saccharificationand fermentation. The use of HPLC tomeasure fermentable carbohydrate directlymakes DE determination necessary.


Diethyl ether The traditional anaesthetic ether,which is employed as an entrainer in someazeotropic distillation processes for fuelethanol dehydration (as an alternative to themore commonly-used benzene). It is acolorless, flammable liquid that boils at34.6oC and freezes at -117oC. Chemicalformula (C

2H

5)

2O. It readily forms explosive

mixtures with air.

Differential pressure cell (DP cell) A devicefor measuring the difference in pressure ofliquid on either side of a restricting orifice. Itis used in flow measurement. (See Orificemeter).

Dimer A compound produced by linkingtogether two molecules of a simplercompound (or monomer). It is the simplestform of polymer. For example, maltose is adimer composed of two linked glucosemolecules.

Disaccharide A compound sugar that yieldstwo monosaccharide units on hydrolysis. Forexample, lactose yields glucose andgalactose, sucrose yields glucose andfructose, while maltose yields two glucoseunits.

Disc and donut column A beer distillation towerpatented by Raphael Katzen with traysalternately consisting of annular shelves withopen centers (donuts) and discs (of a slightlylarger diameter than the donut holes) placedcentrally below the holes. The effect is to givea circular curtain of liquid descending fromtray to tray, through which the rising vaporsmust pass. It has the advantage of beingunaffected by the solids content of the beerfeed or any scaling that may occur in a typicaltray column.

Distilland Material to be distilled, such as beer.

Distillate The portion of a liquid (distilland)removed as a vapor and condensed during adistillation process.

Distillation The process by which thecomponents of a liquid mixture are separated

by differences in boiling point by boiling andrecondensing the resultant vapors. In ethanolproduction, it is the primary means ofseparating ethanol from aqueous solutions.

Distilled spirits permit (DSP) A permit issuedby the BATF allowing the holder to engage inthe production or warehousing ofundenatured ethanol for beverage, industrialor fuel use.

Distillers bushel 56 lbs of any grain, regardlessof volume.

Distillers dried grain (DDG) The dried residualby-product of a grain fermentation process.It is high in protein, as most of the grain starchhas been removed. It is used as an animal feedingredient. By strict definition, DDG isproduced only from the solids separated fromwhole stillage by centrifuging or screening.In practice, the term is commonly used todescribe the entire dried stillage residue,making it synonymous with DDGS.

Distillers dried grain with solubles (DDGS)The product derived by separating the liquidportion (thin stillage or solubles) from grainwhole stillage by screening or centrifuging,then evaporating it to a thick syrup and dryingit together with the grain solids portion.

Distillers dried solubles The product derivedfrom separating the liquid portion (thinstillage) from grain whole stillage,evaporating it to a thick syrup and then dryingit to a fine powder.

Distillers feeds The by-products of fermentationof cereal grains. See DDG and DDGS.

Distillers wet grain (DWG) The wet grainresidue separated from whole stillage byscreening or centrifuging. It has a very limitedstorage life and is generally used only wherecattle feedlots are located near ethanolproduction plants.

Distillery A building or premises where alcoholis distilled.


Distillery run barrels Recently-emptied, usedwhiskey barrels that have not been sorted toremove those with or without defects.

DMA 67Y A detergent produced by Dupont. Itis added to fuel ethanol to ensure that theresultant blend with gasoline will have clean-burning characteristics similar to detergentgasolines supplied by major oil companies.

Downcomer (or downpipe) A device used indistillation columns to allow liquid to descendfrom one plate to another. Downcomers maytake the form of one or more round, oval orrectangular section pipes, or chordal baffles.The downcomer usually has a weir-type sealat the bottom to prevent vapors passingupward.

DP cell See differential pressure cell.

Dry degermination A process for the removalof germ from grain without the need forsteeping and wet milling. It may involve somepretreatment of the grain to raise the moisturecontent before processing. It is used inScotland for corn milling in grain whiskyplants and is used in the production of cornflakes; but is not commonly used in the USethanol production industry.

Dry milling In the ethanol production industrydry milling refers to the milling of whole drygrain where, in contrast to wet milling, noattempt is made to remove fractions such asgerm and bran. Milling may be carried outwith various types of equipment, includinghammer mills and roller mills.

DSP Abbreviation for Distilled Spirits Permit.

DTW Abbreviation for Dealer Tank Wagon.

Dual-flow plate (or tray) A perforated platesimilar to a sieve plate, without downcomers.The perforations are of such a size and openarea that the liquid descends by weepingthrough the holes.

Dunder A Caribbean synonym for vinasse ormolasses stillage. It is commonly used to refer

to vinasse that has been stored for some timeto allow bacterial development prior to beingused as backset in the production of heavily-flavored rums.

Dupont waiver The waiver received by Dupontfrom the EPA in 1985 for a blend of gasolinecontaining a maximum of 5% methanol plusa minimum of 2.5% ethanol or other approvedco-solvent, together with an approvedproprietary corrosion inhibitor.

DWG Abbreviation for distillers wet grain.

E

Ebulliometer A laboratory instrument used tomeasure alcohol concentration in water-ethanol mixtures. The mixture is heated toboiling and the temperature of the vapor ismeasured accurately and corrected forbarometric pressure. The ethanolconcentration is read from a table ofconcentration vs boiling point.

ED Abbreviation for extractive distillation.

Effect In the context of evaporators, the termis used to describe one vessel in a series. Forexample a quadruple effect evaporatorconsists of four linked vessels.

EITC Abbreviation for energy investment taxcredit.

Endoenzyme An enzyme acting on internalportions of a large polymeric molecule ratherthan around the periphery. For example, theα-amylase enzyme hydrolyzes linkages withinamylose and amylopectin molecules. Incontrast, amyloglucosidase acts as anexoenzyme by only hydrolyzing the outermostlinkages.

Energy Policy Act of 1992 US federallegislation that amended the provisions of theOmnibus Reconciliation Act of 1990 to allowthe proration of the excise tax exemption of5.4 cents per gallon on ethanol gasoline blends


that use less than the standard content of 10%ethanol. This was to meet the requirementsof the Clean Air Amendments Act thatrequired that oxygenate blends with 2.1% or2.7% oxygen be used in gasolines indesignated areas suffering from severe airpollution. (These oxygen rates correspond toethanol usages of 6% and 7.7%,respectively).

Energy Security Act of 1980 US federallegislation that supported the emerging fuelethanol industry by establishing anindependent Office of Alcohol Fuels withinthe Department of Energy and authorizedprograms of loan guarantees , priceguarantees and purchase agreements with fuelethanol producers.

Energy Tax Act of 1978 US federal legislationthat instituted the first excise tax exemptionfor gasoline blended with 10% fermentationethanol. It exempted the blends from the taxof 4 cents per gallon. It also created an energyinvestment tax credit (EITC) of 10% thatapplied to equipment for converting biomassto ethanol in addition to the standard 10%investment tax credit.

Entrainer A substance used to assist in thedehydration of ethanol by azeotropicdistillation. Examples are benzene, cyclo-hexane and pentane. (See azeotropicdistillation).

Environmental Protection Agency (EPA) AUS government agency established in 1970.EPA responsibilities include the regulation offuels and fuel additives, including ethanolgasoline blends.

Enzymatic hydrolysis The hydrolysis of apolymer by the use of enzymes. In the caseof starch hydrolysis, an α-amylase enzymemay be used in the initial hydrolysis toachieve liquefaction and an amylo-glucosidase enzyme may be used to completethe hydrolytic saccharification to fermentablesugars.

Enzyme Any of a class of complex proteinaceoussubstances (such as amylases and lactases)produced by living organisms that catalyzechemical reactions without being destroyed.Enzymes may act outside the producingorganism and maybe used in industrialprocesses such as saccharification.

EPA Abbreviation for Environmental ProtectionAgency.

Ester The product derived by the reaction of anacid with an alcohol or other organiccompound having hydroxyl groups. Forexample, ethyl acetate is an ester producedby reacting acetic acid with ethanol. Esterstend to accumulate in distillation in the headsat the top of the tower.

ETBE Abbreviation for ethyl tertiary butyl ether.

Ethanol Otherwise known as ethyl alcohol,alcohol, grain-spirit or neutral spirit, etc. Aclear, colorless, flammable oxygenatedhydrocarbon. Chemical formula: C

2H

5OH. It

has a boiling point of 78.5oC in the anhydrousstate. However, it forms a binary azeotropewith water with a boiling point of 78.15oC at acomposition of 95.57% by weight ethanol.

Ether One of a class of organic compounds inwhich an oxygen atom is interposed betweentwo carbon atoms in the molecular structure.Ethers may be derived from alcohols by theelimination of water. Ethers such as diethylether and isopropyl ether may be used asentrainers in the dehydration of ethanol byazeotropic distillation. Other ethers such asMTBE and ETBE may be used as octaneenhancers in gasoline. Ethers are dangerousfire and explosion hazards. When exposed toair they form peroxides that may detonate onheating.

Ethyl acetate Chemical formula CH3COOC

2H

5.

A clear, volatile, flammable liquid with acharacteristic fruity odor. It has a pleasant,sweet taste when diluted. Boils at 77oC andfreezes at -83oC. It is produced by the reaction


of ethanol with acetic acid and is a majorcomponent of the esters that give rum itscharacteristic odor.

Ethyl alcohol See Ethanol.

Ethyl carbamate Otherwise known as urethaneor carbamic acid ethyl ether. A carcinogeniccompound produced by heating urea withethanol under pressure. Chemical formulaNH

2COOC

2H

5. It is very soluble in ethanol

or water. Traces of ethyl carbamate may beformed in the distillation of beers producedwith the use of urea as a yeast nutrient. It doesnot present a significant hazard if such ethanolis only used for fuel purposes, but maypresent problems in the production ofwhiskies.

Ethylene glycol A slightly viscous, sweet-tasting, poisonous liquid. Chemical formulaCH

2OHCH

2OH. It has a boiling point of

197.6oC, and is considerably hygroscopic. Itis commonly used as an antifreeze. It mayalso be used as an extractant in extractivedistillation processes for the dehydration ofethanol. (See extractive distillation).

Ethyl tertiary butyl ether (ETBE) A colorless,flammable, oxygenated hydrocarbon.Chemical formula C

2H

5OC

4H

9. It may be

produced from ethanol and tertiary butanol(TBA) or isobutylene. It is of similar structureto methyl tertiary butyl ether (MTBE), havingsimilar octane-enhancing properties, but hasa significantly lower effective Reid vaporpressure in blending with gasoline.

Evaporation The process by which a substancein the liquid state is converted into the vaporstate.

Evaporator A device used to evaporate part orall of a liquid from a solution. In the case ofgrain stillage processing, evaporators may beused to concentrate screened thin stillage toa syrup of about 35% solids, which may thenbe fed to a dryer. Evaporators are normally

operated under vacuum (to obtain a lowerboiling point of the liquid) and may consistof a series of interlinked vessels or effectsoperating under differing degrees of vacuum.

Extractant A substance such as ethylene glycolor glycerol be used in extractive distillationprocesses for the dehydration of ethanol.

Extractive distillation A process where anextractant is added to a mixture being distilledto change the volatility of one or morecomponents. The less volatile mixture will thendescend in a continuous distillation columnwhile the more volatile components may beremoved in the condensed overhead vapors.In the use of extractive distillation in thedehydration of ethanol, liquid extractants suchas ethylene glycol, or glycerol, may be used.Salts such as potassium and sodium acetatesmay also be used alone in molten form or inmixtures with glycerol, etc. Anhydrous ethanolis recovered in the overhead condensate whilethe water combines with the extractant toemerge from the bottom of the column. (Thisis the reverse of the situation in azeotropicdistillation). The extractant is then separatedfrom the water in another column (or anevaporator) and is recycled.

In the production of neutral spirit, light rumsor whiskies, extractive distillation may beused to remove fusel oils and some othercongeners in the condensed overhead vapors.In this instance the extractant is water, as someof the congeners with lower volatility thanethanol in a concentrated state may havehigher volatility than ethanol when dilutedwith water and therefore rise up the extractivedistillation column while the ethanoldescends.

Exoenzyme An enzyme restricted to acting onthe outer end of large polymeric moleculesand cleaves molecules one by one. (SeeEndoenzyme).


F

Facultative anaerobe Term used to describe amicroorganism (such as a yeast) that isessentially aerobic (or oxygen-requiring), butcan also thrive under anaerobic (or oxygen-free) conditions.

Fahrenheit scale A temperature scale in whichthe boiling point of water is 212oF and thefreezing point is 32oF. (The zero point wasoriginally established as the lowest pointobtainable with a mixture of equal weights ofsnow and common salt).

Feed plate (or feed tray) The plate or tray ontowhich the distilland (liquid to be distilled) isintroduced in a distillation tower. In theory, itis the point in a tower above whichenrichment or concentration occurs and belowwhich stripping occurs.

Feedstock The raw material used in a process.For example, corn, molasses, whey, etc. maybe used as feedstocks for ethanol production.

Fermentable sugars Simple sugars such asglucose and fructose that can be convertedinto ethanol by fermentation with yeast. Theymay be derived by the hydrolysis of starch orcellulose feedstocks or obtained from othersources.

Fermentation The enzymatic transformation bymicroorganisms of organic compounds suchas sugars. It is usually accompanied by theevolution of gas as in the fermentation ofglucose into ethanol and carbon dioxide.

Fermentation efficiency The measure of theactual output of a fermentation product suchas ethanol in relation to the theoretical yield.

Fermentation ethanol The term used todistinguish ethanol produced by fermentationfrom synthetic ethanol produced fromethylene, etc. The difference is significant inthat only fermentation ethanol qualifies forUS federal and state excise tax exemptions forautomotive fuel use. Furthermore, only

fermentation ethanol may be used forbeverage purposes, and in many countries tomake vinegar.

Fermentor The vessel in which mashfermentation takes place. The vessel may befabricated from stainless steel, fiber glass, etc.and is normally fitted with an internal orexternal cooling system for controllingtemperature of the fermenting mash.

FFV Abbreviation for Flexible Fueled Vehicle.

Flame arrester A device installed on the vaporvents of alcohol storage tanks and distillationcolumns to prevent entry of flames that mightcause an explosion. It normally contains finemetal meshes with holes large enough toallow vapors to escape, but too small to allowflames to pass in the opposite direction.

Flash cooling The rapid cooling achieved whena hot liquid is subjected to a sudden pressuredrop to reduce its boiling point.

Flash point The minimum temperature at whicha combustible liquid will ignite when a flameis introduced. It depends on the volatility ofthe liquid to provide sufficient vapor forcombustion. For example, anhydrous ethanolhas a flash point of 51oF, while 900 proofethanol has a flash point of 78oF.

Flexible fueled vehicle (FFV) A vehicledesigned to operate on a variety of fuels suchas methanol, ethanol or gasoline alone or incombinations without requiring majoradjustments.

Flocculation The aggregation or coalescing offine suspended particles or bodies into looseclusters or lumps. The flocculationcharacteristics of yeast or othermicroorganisms may be important features intheir recovery for cell recycling.

Flow meter A device for measuring the rate offlow of a fluid.

Fluidized bed combustion boiler A boiler inwhich the coal or other fuel particles are kept


in suspension by a rising column of gas ratherthan resting on a conventional grate. Thesystem gives greater heat transfer and higherpotential combustion efficiency. The systemhas the advantage that limestone may beadded to the coal fuel to absorb sulfur gasesto reduce emissions.

Fossil fuel Any naturally-occurring fuel of anorganic nature that originated in a pastgeologic age such as coal, crude oil or naturalgas.

Fractional distillation A process of separatingmixtures such as ethanol and water by boilingand drawing off the condensed vapors fromdifferent levels of the distillation tower.

Fructose A fermentable monosaccharide(simple sugar) of the chemical formulaC

6H

12O

6. Its chemical structure is similar to

that of glucose, but it is sweeter to the taste. Itmay be produced from glucose by enzymaticisomerization, as in the production of highfructose corn syrup (HFCS).

Fuel grade ethanol See motor fuel gradeethanol.

Fuel ethanol Usually denotes anhydrousethanol that has been denatured by additionof 2-5% unleaded gasoline and is intendedfor use as an automotive fuel in blends withgasoline.

Fungible Literally means ‘interchangeable intrade’. Commonly used to denote productssuitable for transmission by pipeline. Ethanol isnot considered fungible in this sense because itwould absorb any water accumulating inpockets in a pipeline.

Fusel oil Term used to describe the higheralcohols, generally the various forms ofpropanol, butanol and amyl alcohol that arecongeners or by-products of ethanolfermentation. Normally, predominantlyisoamyl alcohol. Their presence in alcoholicbeverages is known to be a cause ofheadaches and hangovers. The fusel oils have

higher boiling points than ethanol and aregenerally removed in the distillation processto avoid accumulation in the rectifier. Theymay be subsequently added back into theanhydrous product for motor fuel gradeethanol.

Fusel oil decanter A device used to separateaccumulations of fusel oil from ethanol basedon the fact that fusel oil has a lower miscibilityin water than ethanol and can therefore beremoved by dilution with water. It generallyconsists of a tank with windows or sight glassesto permit the operator to observe and controlthe separation.

G

Galactose A monosaccharide of chemicalformula C

6H

12O

6 that along with glucose is a

constituent of the disaccharide lactose. It isan isomer of glucose, but is less-readilyfermented by yeasts to ethanol.

Gas chromatography (GC) A technique forseparating chemical substances in which thesample is carried by an inert gas streamthrough a tube (or column) packed with afinely-divided solid material. The variouscomponents in the sample pass through thecolumn at differing velocities and emergefrom the column at distinct intervals to bemeasured by devices such as a flameionization detector or a thermal conductivitydetector. (Where the solid material in thecolumn is pretreated with a liquid to achievethe component separation, the process maybe referred to as gas liquid chromatography).The technique may be used for separating andmeasuring the amount of ethanol and thevarious by-products formed in fermentation.

Gas liquid chromatography See gaschromatography.

Gasohol (or gasahol) A trade name registeredby the Nebraska Agricultural ProductsIndustrial Utilization Committee, later


renamed the Nebraska Gasohol Committee.(The committee was responsible for laying thegroundwork for the development of thepresent day US fuel ethanol industry). Thetrade name, in either spelling, covers a blendof anhydrous ethanol ‘derived fromagricultural products’ with gasoline (notnecessarily unleaded). The committee hasfreely granted permission for commercial useof the trade name provided it is not used forblends containing alcohols other than ethanol.

Gasoline A volatile, flammable, liquidhydrocarbon mixture suitable for use as a fuelin internal combustion engines. Normallyconsists of a blend of several products fromnatural gas and petroleum refining togetherwith anti-knock agents and other additives. Itis a complex mixture of hundreds of differenthydrocarbons, generally in the range of 4-12carbon atoms per molecule. The componentsmay have boiling points ranging mainlybetween 30 and 200oC with blends beingadjusted to altitude, season and legalrequirements.

Gasoline extender The term used to describeethanol when it is simply used as a partialreplacement for gasoline without anyconsideration for its value as an octaneenhancer or oxygenate.

Gay Lussac (GL) The name given to a scale ofthe concentration of ethanol in mixtures withwater where each degree is equal to 1% byvolume (i.e., 1o GL is equivalent to 2o USproof). It takes the name from the Frenchchemistry pioneer, Joseph-Louis Gay Lussac.

Gay Lussac equation The equation for thefermentation of sugar by yeast to carbondioxide and ethanol, established by the Frenchchemist Joseph-Louis Gay Lussac in 1815:C

6H

12O

6→ 2CO

2 + 2C

2H

5OH. (See

Stoichiometric yield).

GC Abbreviation for gas chromatograph.

Gear pump A positive displacement rotary

pump containing two intermeshed gearwheels in a suitable casing. The counterrotation of the gears draws fluid between thegear teeth on one side and discharges it onthe other side. It is commonly used for viscousliquids such as molasses and stillage syrup.

Gelatinization In reference to the cooking ofstarchy feedstocks, gelatinization is the stagein which the starch granules absorb water andlose their individual crystalline structure tobecome a viscous liquid gel. Gelatinizationis significant in that it is the preliminaryprocess necessary to render starch susceptibleto enzymatic hydrolysis for conversion tofermentable sugars.

Gin Defined by the BATF as ‘a product obtainedby original distillation from mash, or byredistillation of distilled spirits, or by mixingneutral spirits with or over juniper berries andother aromatics, or with or over other extractsderived from infusions, percolations ormaceration of such materials, and includesmixtures of gin and neutral spirits. It shallderive its main characteristic flavor fromjuniper berries and be bottled at not less than800 proof. Gin produced exclusively byoriginal distillation or redistillation may befurther designated as distilled’.

GL Abbreviation for Gay Lussac.

GLC Abbreviation for gas-liquid chromatography.

Glucoamylase An enzyme that hydrolyses starchinto its constituent glucose units. (SeeAmyloglucosidase).

Glucan See Betaglucan.

Glucanase An enzyme that hydrolyses glucan.(See Betaglucanase).

Glucose A fermentable sugar otherwise referredto as dextrose. It is a monosaccharide and hasthe formula C

6H

12O

6. Glucose is the ultimate

product in the hydrolysis of starch andcellulose, which are both polymers ofhundreds or thousands of glucose units.


Glucose isomerase An enzyme that convertsglucose into its isomer, fructose. It is used inthe production of high fructose corn syrup(HFCS).

Glucosidase See Amyloglucosidase orGlucoamylase.

Glycerol (or glycerine) A clear, colorless,viscous, sweet-tasting liquid belonging to thealcohol family of organic compounds. It hasa chemical formula of CH

2OHCHOHCH

2OH,

having three hydroxyl (OH) groups. It is a by-product of alcoholic fermentations of sugars.It is hygroscopic and may be used as anextractant in the dehydration of ethanol.

Grain alcohol Term used to distinguish ethanolproduced from grain from ethanol producedfrom other feedstocks, or from wood alcohol(methanol).

Grain whisky The term used in Scotland todistinguish the bland, nearly-neutral,continuous distillation product that forms thebase of blended Scotch whisky from the maltwhiskies that contribute most of the flavor inthe blend. Grain whisky is usually producedfrom corn or barley.

Grain sorghum Otherwise known as milo, asorghum grown for grain production, asdistinct from sweet sorghum grown for thesugar content of its stem. It may be used as afeedstock for ethanol production.

Gram-negative See Gram stain.

Gram-positive See Gram stain.

Gram stain A widely-used microbiologicalstaining technique that aids in the identificationand characterization of bacteria. It wasdevised by Danish physician, Hans ChristianGram. Bacteria are described as Gram-positiveif their cell walls absorb and retain the stainand Gram-negative if they do not. Thetechnique is particularly useful in examiningbacterial contaminants in fermentations, asmost Gram-positives are susceptible to controlby penicillin.

H

Hammer mill A type of impact mill or crusherin which materials such as cereal grains arereduced in size by hammers revolving rapidlyin a vertical plane within a steel casing. Ascreen with numerous holes of a selecteddiameter is installed in the casing to controlthe size of particles produced. It is commonlyused for grinding corn as a fermentationfeedstock.

Heads Term used to describe the impuritiesproduced in ethanol fermentations(congeners) that have lower boiling pointsthan ethanol. They include methanol andaldehydes.

Heads concentrating column A distillationcolumn used to concentrate heads removedin the production of neutral spirit, light rumsand whiskies.

Heat exchanger Device used to transfer heatfrom a fluid on one side of a barrier to anotherfluid flowing on the other side of the barrier.Common forms include shell-and-tube heatexchangers and plate-type heat exchangers.Various types of heat exchangers are used inethanol plants for mash cooling, fermentationtemperature control, indirect steam heating(reboilers), overhead vapor condensing, etc.

Heat of condensation The heat given up whena vapor condenses to a liquid at its boilingpoint.

Heat of vaporization The heat input requiredto change a liquid at its boiling point to a vaporat the same temperature.

Hemicellulose Term used to describe non-cellulosic polysaccharide components ofplant cell walls. The most commonhemicelluloses are composed of polymers ofxylose (a 5-carbon sugar, or pentose) togetherwith a uronic acid (a sugar acid), arabinose(another pentose) and mannose (a hexose).Hemicelluloses have no chemical relationshipto cellulose. They frequently surround the


cellulose fibers and increase their bonding andtensile strength. The presence of hemicelluloses(and lignins) in close association with cellulosetends to impede the extraction and hydrolysisof wood cellulose to sugars for ethanolproduction, as well as its degradation innature.

Hexose A class of monosaccharides (simplesugars) containing six carbon atoms in themolecule. They have the formula C

6H

12O

6.

Common examples are glucose, fructose andgalactose.

HFCS Abbreviation for high fructose cornsyrup.

Hiag process A process developed in Germanyin the 1930s for the dehydration of ethanolby extractive distillation using a mixture ofsodium and potassium acetates as theextractant.

High boilers Term used to describe theimpurities produced in ethanol fermentation(congeners) that have higher boiling pointsthan ethanol. Otherwise referred to as tails,they include the higher alcohols, propanols,butanols and amyl alcohols or fusel oils. Inthe context of gasoline, the term refers tocomponents with boiling points considerablyabove the mid-range.

Higher alcohols Alcohols having more than twocarbon atoms. They exist in various isomericforms. As the number of carbon atomsincreases, so does the number of isomers, butat a greater rate. The lower members of thisgroup, namely propanol, butanol and amylalcohol are major constituents of fusel oil.

High fructose corn syrup (HFCS) A productin which a large percentage of the glucosederived from starch hydrolysis has beenconverted into its sweeter-tasting isomerfructose by use of enzymes. It is frequentlyused as a substitute for cane or beet sugar asa sweetener in soft drinks, etc. It is a seasonalalternative to ethanol production in some cornprocessing plants.

High performance liquid chromatography(HPLC) An advanced form of liquidchromatography used for the separation ofcomplex mixtures of non-volatile materialsfor analytical purposes. It involves very smallparticles packed in columns through whichthe liquids flow and high pressures to increasethe rate of flow and shorten the time requiredto achieve the separation. The process iscoupled with other devices to identify theconstituents. It may be used in the analysis ofthe sugars and dextrins produced byhydrolysis of starch.

High test molasses (HTM) See molasses.

Hogshead A wooden barrel with a capacity ofapproximately 66 US gallons (250 liters) usedin Scotland for aging whisky. It is usuallyconstructed from shooked bourbon whiskeybarrels of 52 gallons capacity by usingadditional staves and larger heads and hoops.

HPLC Abbreviation for high performance liquidchromatography.

HTM Abbreviation for high test molasses.

Hydrocarbon A member of a class of organicchemical compounds containing only hydrogenand carbon. It should be clearly differentiatedfrom carbohydrates, which contain oxygen aswell as hydrogen and carbon. The main naturalsources of hydrocarbons are petroleum, coal,natural gas and bitumens.

Hydrolysis Literally means the breakdown,destruction or alteration of a chemicalsubstance by water. In the case of starch andother polymers of glucose, a molecule of wateris divided between two adjacent glucose units,in order to cleave the linkage. For example,maltose (C

12H

22O

11), which contains two

glucose rings, requires the addition of amolecule of water (H

2O) to yield two separate

glucose molecules (C6H

12O

6). The hydrolysis

may be accomplished with the use of acidsor enzymes.


Hydrometer An instrument measuring thedensity, specific gravity or other similarcharacteristics of liquids. It is generallycomprised of a long-stemmed glass tube witha weighted bottom which floats at differentlevels in liquids of different densities. Thereading is taken at the meniscus (where thecalibrated stem emerges from the liquid). Theliquid temperature is normally determinedwhen taking a reading; and reference is madeto hydrometer tables to obtain a correction toa standard temperature. A proof hydrometermeasures the content of ethanol in a mixturewith water. A Brix or Balling hydrometermeasures on a scale equivalent to thepercentage of sugar by weight in an aqueoussolution.

Hydroselection column Synonym for extractivedistillation column.

Hydrous ethanol Term used for ethanol thathas not been subjected to dehydration. It mayrefer to any mixture of ethanol and water, butfrequently is used to denote ethanol at aconcentration of about 190-1920 proof, closeto the azeotropic point.

Hygroscopic Term used to describe a substancewith the property of absorbing moisture fromthe air. Anhydrous ethanol is hygroscopic, andits exposure to moist air should therefore beminimized.

Hydroxyl group A combination of one atomof oxygen and one atom of hydrogen (OH)forming an essential part of any alcohol. Thespare atomic bond of the group is linked to acarbon atom as in ethanol (C

2H

5OH) or

methanol (CH3OH).

I

IDRB Abbreviation for Industrial DevelopmentRevenue Bonds.

Imperial gallon A measure of volume in theBritish system defined in 1824 as the volume

occupied by 10 pounds of water at 62oF and30 inches of barometric pressure. It is theequivalent of 1.2 US gallons, or 4.546 liters.

Industrial alcohol Denotes any ethanol intendedfor industrial uses such as solvents,extractants, antifreezes and intermediates inthe synthesis of innumerable organicchemicals. The term covers ethanol of bothsynthetic and fermentation origin of a widerange of qualities and proofs, with or withoutdenaturants.

Industrial development revenue bonds (IDRB)Debt incurred through local industrialdevelopment authorities in numerous USstates. Such bonds were a popular way tofinance fuel ethanol plants in the past becausethe interest on the bonds was not subject toincome tax and the bonds could be sold withinterest rates substantially below the primerate.

Inoculum The portion of a culture of yeast (orbacteria) used to start a new culture or afermentation.

Inulin A storage polysaccharide found in theroots and tubers of various plants, particularlyJerusalem artichokes. It consists of chains ofan average of 30 fructose units. It is onlyslightly soluble in cold water, but dissolvesreadily in hot water. Unlike starch, it does notgive a color reaction with iodine.

Inulinase An enzyme capable of hydrolyzinginulin to its component fructose units.

IPE Abbreviation for isopropyl ether.

Isoamyl alcohol The principal alcohol in fuseloil. It is an isomer of pentanol, of compositionC

5H

11OH. It is a colorless liquid with a pungent

taste and disagreeable odor. Boils at 132oCand freezes at -117.2oC. It is only slightlysoluble in water but miscible with ethanol. Itmay be recovered by fractionation of fuseloil, and has a wide range of uses in organicsynthesis, pharmaceuticals, photographicchemicals and as a solvent for fats, etc. The


vapors are poisonous and at lowconcentrations may cause headaches anddizziness.

Isomer One of a series of two or more moleculeswith the same number and kind of atoms andhence the same molecular weight, butdiffering in respect to the arrangement orconfiguration of the atoms. For instance,glucose and fructose have the formulaC

6H

12O

6, but different molecular structures.

Isomerase An enzyme that can convert acompound into an isomeric form. For instance,the glucose isomerase enzyme convertsglucose into its sweeter-tasting isomerfructose in the production of high fructosecorn syrup (HFCS).

Isomerization The process of converting achemical compound into its isomer such asconverting glucose to fructose in theproduction of high fructose corn syrup(HFCS). In petroleum refining, the termdescribes methods used to convert straight-chain to branched-chain hydrocarbons, oracyclic to aromatic hydrocarbons to increasetheir suitability for high octane gasoline.

Isopropyl ether (IPE) An ether (otherwiseknown as di-isopropyl ether) used in somefuel ethanol plants as an entrainer in thedehydration process as an alternative tobenzene, etc. It is a colorless, volatile liquidwith chemical formula (CH

3)2CHOCH(CH

3)

2,

which boils at 67.5oC and freezes at -60oC. Itreadily forms explosive mixtures with air.Inhalation of vapors may cause narcosis andunconsciousness.

J

Jet cooker An apparatus for the continuouscooking of grain mashes in which the mash ispumped past a jet of steam that instantly heatsthe mash to gelatinize the starch.

Jobber A gasoline wholesaler. Some jobbersmay operate under contract to one or moremajor oil companies, distributing gasoline tobranded gas stations. Other jobbers mayoperate independently, buying gasoline fromvarious sources for distribution to unbrandedretail outlets. Independent jobbers arefrequently major buyers and blenders of fuelethanol.

K

Karl Fischer titration A method to chemicallydetermine the amount of water present in asample of ethanol and/or other substances.When correctly practiced, the methodprovides an extremely accurate measurementof very small quantities of water in ethanoleven if gasoline denaturant is present. (Seetitration).

Kerosene One of the three permissibledenaturants for fuel ethanol as specified inBATF regulations. It is a refined petroleumfraction used as a fuel for heating, cookingand for jet engines. It has a boiling point rangesomewhat higher than that of gasoline,generally between 180oC and 290oC.

Kjeldahl method An analytical method for thedetermination of nitrogen in organiccompounds. As nitrogen is an essentialelement in protein, the method may be usedfor determining protein in such materials asdistillers dried grains. Kjeldahl protein, orcrude protein, is Kjeldahl N x 6.25.

Kluyveromyces fragilis (or marxianus) Alactose-fermenting yeast used in theproduction of ethanol from cheese whey.

Kubierschky process The first patented processfor the continuous dehydration of ethanol withbenzene. With relatively minor variations, theprocess developed in 1914 based on Young’searlier batch process is still used today in fuelethanol plants.


L

Lactase An enzyme that hydrolyzes lactose intoglucose and galactose. This hydrolysis allowslactose-containing feedstocks such as cheesewhey to be fermented by the commonSaccharomyces cerevisiae yeasts. A principalsource of lactase is the yeast Kluyveromycesfragilis, which can also directly fermentlactose to ethanol.

Lactic acid The organic acid produced in thefermentation of carbohydrates by Lacto-bacillus bacteria. Its production is theprincipal reason for loss of yield incontaminated ethanol fermentations. Purelactic acid is a colorless, odorless, hygroscopic,syrupy liquid with a boiling point of 122oCand a formula CH

3CHOHCOOH.

Lactobacillus A genus of bacteria that producelactic acid as a major product in thefermentation of carbohydrates. Lactobacilliare found extensively in fermenting foodproducts such as souring milk and in graindust. They are the principal cause of loss ofyield in ethanol fermentations. Otherwisereferred to as lactic acid bacteria, they aregenerally Gram-positive and controllable withpenicillin and certain other antibiotics.

LactosideTM A combination of antibioticsformulated to take advantage of synergies inincreasing the spectrum of activity againstlactic and acetic acid-producing bacteria.

Lactose The principal sugar in milk and cheesewhey. It may be fermented by suitable yeaststo ethanol. It is a disaccharide readilyhydrolysed to its two components, glucose andgalactose. Lactose has the formula C

12H

22O

11.

Lag phase Applied to yeast propagation, lagphase refers to the initial period in which theyeast inoculum becomes adapted to the mashprior to the rapid increase in cell numbersreferred to as the logarithmic phase.

Latent heat The quantity of energy absorbedor released when a substance undergoes a

change of state, i.e., from a solid to a liquid(melting) or from a liquid to a vapor, or viceversa. No change in temperature is involved.For example, water requires a large amountof latent heat (measured in BTUs or calories)to convert from the liquid to vapor state(steam) at 100oC, while steam releases thelatent heat again on condensing back to theliquid state.

Lead phase-out The reduction in the amountof lead (normally in the form of tetraethyllead) that could be used as an octaneenhancer in leaded gasoline in the US. Thereduction, which went from 1.1 grams pergallon prior to July 1, 1985 to less than 0.1grams per gallon after January 1, 1986 causedan increase in demand for alternative octaneenhancers such as ethanol and MTBE.

Light whisky Defined by the BATF as whiskyproduced at more than 1600 proof and storedin used or uncharred oak barrels.

Lignin A polymeric, non-carbohydrateconstituent of wood that functions as asupport and binder for cellulose fibers. It maycomprise 15-30% of wood and can only beseparated from the cellulose and hemicellulosecomponents by chemical reaction at hightemperatures. Its presence in wood is a majorbarrier to the hydrolysis of cellulose to sugarsfor fermentation purposes.

Lignocellulose Woody materials made uplargely of lignin, cellulose and hemicelluloses.The chemical bonding between theconstituents renders it resistant to hydrolysis.

Lime A white alkaline powder composed ofcalcium oxide. It is added to grain mash toadjust the pH and to provide calcium ions inorder to prevent the inactivation of the α-amylase enzyme molecule by the loss of itsessential calcium atom.

Liquefaction Conversion of a solid substanceto the liquid state. In reference to starch, it isthe stage in the cooking and saccharificationprocess in which gelatinized starch is partially


hydrolyzed by α-amylase (or occasionally byan acid) to give soluble dextrins. This convertsthe starch mash into a free-flowing liquid,cutting viscosity.

Liqueurs and cordials Defined by the BATF as‘products obtained by mixing or redistillingdistilled spirits with or over fruits, flowers,plants or pure juices therefrom, or any othernatural flavoring materials, or with extractsderived from infusions, percolation ormaceration of such materials, and containingsugar, dextrose or levulose, or a combinationthereof, in an amount not less than 2.5% byweight of the finished product’.

Liter A metric measurement of volume definedas the equivalent of 1000 cubic centimetersor 0.2642 US gallons.

Logarithmic phase Applied to yeastpropagation, it refers to the period in whichcell numbers increase at an exponential rateafter the initial lag phase.

Low boilers In reference to ethanol distillation,the term is applied to the congeners, orfermentation by-products, which boil at alower temperature than ethanol. Morecommonly referred to as heads , thesecompounds are principally aldehydes andmethanol.

LPA Abbreviation for liters of pure alcohol.

M

Macromolecule A giant molecule in which thereis a large number of one or more relativelysimple structural units or monomers.

Malt Barley grains that have been steeped inwater and then allowed to germinate. Thegermination is normally halted by drying thegrains when the sprouts are about the samelength as the grains. At this stage, the malt (or‘malted barley’) contains considerableamounts of α- and ß-amylase enzymes that

saccharify the barley starch and otheradditional starch in a mash to yieldfermentable sugars. (In Scotland, the dryingmay be done by exposing the malt to a flowof peat smoke to impart a smoky odor to themalt). Malt is used in whisky production,where it also contributes to product flavor. Infuel ethanol production the necessarysaccharifying enzymes are normally derivedfrom microbial sources.

Malt whisky In the US, malt whisky is definedby the BATF as a whisky produced at less than1600 proof from a fermented mash containingat least 51% malted barley, and stored at under1250 proof in charred, new oak barrels. InScotland, malt whiskies are made from a100% malted barley mash and may be agedin previously-used oak barrels. Malt whiskiesmay be mixed with grain whiskies to impartmuch of the characteristic flavor of blendedScotch whisky.

Maltase An enzyme capable of hydrolyzingmaltose sugar molecules into their twocomponent glucose units. Occasionally thename is loosely applied to theamyloglucosidase enzyme, which hydrolyzespolysaccharides to glucose.

Maltose A fermentable sugar which is a dimer(or disaccharide) of glucose in that it iscomprised of two linked glucose units. It isthe normal end product of starchsaccharification by the ß-amylase enzyme inmalt.

Manioc See cassava.

Mannose A fermentable 6-carbon sugar (orhexose) which is an isomer of glucose.Chemical formula C

6H

12O

6. It is a constituent

of hemicellulose.

Mash A mixture of milled grain or otherfermentable carbohydrate in water used inthe production of ethanol. The term may beused at any stage from the initial mixing ofthe feedstock in water prior to any cooking


and saccharification through to thecompletion of fermentation, when it becomesreferred to as beer.

Mash bill The percentages of different types ofgrains used in the preparation of a mash inbeverage alcohol production. For example,a typical bourbon whisky mash bill mayconsist of 65% corn, 25% rye and 10% barleymalt.

McCabe-Thiele diagram A graphic method forcalculation of the number of theoretical plates(or trays) required in a distillation column toachieve a desired separation of twocomponents.

Meal The floury or granular product resultingfrom milling or grinding of cereal grains.

Mechanical vapor recompression (MVR) Amethod used in evaporation in which thewater vapor leaving the evaporator isrecompressed and recycled to heat the samevessel. This means that mechanical energy isused rather than heat energy. Therecompressor may be operated by electricityor by a steam turbine (if high pressure steamis available and there is a demand elsewherefor the low pressure turbine exhaust steam).Mechanical vapor recompression can greatlyincrease the steam usage efficiency, but thecapital costs for equipment are also greater.

Metabolism The chemical processes in livingcells by which energy is derived for vitalprocesses, growth and activities.

Methane A colorless, odorless, tasteless, readily-combustible, asphyxiant, lighter-than-air gas.The first member of the paraffin (or alkane)series of hydrocarbons, it has a formula CH

4.

It occurs in high proportions in natural gas,and is produced by decaying vegetation andother organic matter as in landfills andmarshes. The methane produced in sealedlandfills or from the anaerobic digestion ofthin stillage and other wastewaters is used forboiler fuel in some ethanol plants.

Methane digester or methanator (or anaerobicdigester) A system of tanks used for treatmentof organic waste streams such as thin stillageand evaporator condensate. The system isinitially inoculated with a suitable culture ofmethane-producing bacteria and operates ona continuous basis to generate methane foruse as a boiler fuel. Chemical oxygen demandof the water streams are reduced by thisfermentation system.

Methanol (or methyl alcohol) A colorlesspoisonous liquid with essentially no odor andvery little taste. It is the simplest alcohol andhas the formula CH

3OH. It boils at 64.7oC. It

is miscible with water and most organicliquids, including gasoline. It is extremelyflammable, burning with a nearly invisibleblue flame. It is a congener of ethanolfermentations. Having a lower boiling pointthan ethanol, it tends to be a major componentof the heads stream on distillation. Due to itsmiscibility with benzene, its presence in ahydrous ethanol feed may reduce theefficiency of dehydration processes wherebenzene is used as an entrainer. Methanol isproduced commercially by the catalyzedreaction of hydrogen and carbon monoxide.It was formerly derived from the destructivedistillation of wood, which caused it to beknown as wood alcohol. Methanol may beblended with gasoline, but requires a co-solvent such as ethanol or a higher alcohol tomaintain it in solution. (See Demethylizingcolumn).

Methyl tertiary butyl ether (MTBE) Acolorless, flammable, liquid oxygenatedhydrocarbon. Chemical formula (CH

3)

3COCH

3.

It contains 18.15% oxygen and has a boilingpoint of 55.2oC. It is produced by reactingmethanol with isobutylene. It is used as anoctane enhancer in gasoline, but is beingphased out due to pollution.

MFGE Abbreviation for motor fuel gradeethanol.


Microorganism A collective term formicroscopic organisms including bacteria,yeasts, viruses, algae and protozoa.

Milo (or millet, or grain sorghum) Muchsmaller than corn, this cereal grain has asimilar starch content and yields almost thesame amount of ethanol per bushel onfermentation. Milo is more drought-resistantthan corn and is frequently grown in areasunsuited to corn. Milo fermentations may tendto foam more than corn fermentations andwill produce a characteristic surface crust ifnot agitated.

Molar solution A solution of salts or othersubstances in which one molecular weight ofthe substance in grams (one mole) is dissolvedin enough solvent to make up to one liter.Molarity (or molar concentration) is a measureof moles per liter of a solution and is indicatedby the letter M as in 1M, 10M, etc.

Molasses The thick liquid remaining aftersucrose has been removed from the motherliquor (of clarified concentrated cane or beetjuice), in sugar manufacture. Blackstrapmolasses is the syrup from which no moresugar may be removed economically. It hasusually been subjected to at least threeevaporating and centrifuging cycles toremove the crystalline sucrose. Its analysisvaries considerably, depending on manyfactors including sugar mill equipment andoperational efficiency; but it may containapproximately 45-60% fermentable sugars byweight and approximately 10% ash (orminerals). It is commonly used as an ethanolfeedstock when prices are favorable. High testmolasses (HTM) is not a true molasses as it isthe mother liquor from which no crystallinesugar has been removed by centrifugation butwhich has been treated with acid to reducecrystallization. It may contain approximately80% sugars by weight and is very low in ash.Typically HTM is only produced in yearswhen the sugar price does not justify itsrecovery. It may be used as an ethanol

feedstock when prices are favorable, and hasthe advantage over blackstrap of causing lessdistillation column scaling. However, itrequires more nutrients for fermentation. (Seealso citrus molasses).

Mole A gram mole is the quantity of a chemicalcompound or element equal to its molecularweight in grams. By this definition, moles ofdifferent compounds or elements contain thesame number of molecules. Thus in the vaporstate, as ideal gases, moles of differentcompounds occupy the same volume at thesame temperature and pressure. This leadsdistillation equipment designers to base theircalculations on the mole percentages ofdifferent compounds in a mixture to beseparated, such as ethanol and water, ratherthan the percentages by weight or volume.

Molecule The smallest unit into which a puresubstance can be divided and still retain thecomposition and chemical properties of thesubstance. For example the formula for wateris H

2O; and the smallest unit recognizable as

water is the single molecule of two atoms ofhydrogen linked to one atom of oxygen.

Molecular sieve A microporous substancecomposed of materials such as crystallinealuminosilicates belonging to a class knownas zeolites. The size of the pores in thesubstance may vary with its chemicalstructure, being generally in the range of 3 to10 Angstrom units in diameter. With materialhaving a very precise pore size, it is possibleto separate smaller molecules from larger onesby a sieving action. For example, in ethanoldehydration with a potassium aluminosilicatematerial prepared with pores of a diameter of3 Å units, water molecules with a diameter of2.5 Å may be retained by adsorption withinthe pores while ethanol molecules of adiameter of 4 Å cannot enter and thereforeflow around the material.

The term molecular sieve is frequently usedloosely to describe the entire ethanoldehydration apparatus holding the beads of


sieve material and includes the equipment andcontrols necessary to regenerate them whensaturated with water.

Molecular weight The sum of the atomicweights of the atoms in a molecule. Forexample, water molecules are composed oftwo atoms of hydrogen (atomic weight of one)and one atom of oxygen (atomic weight of16) to give a total molecular weight of 18.

MON Abbreviation for motor octane number.

Monomer A single molecule of a substance ofrelatively low molecular weight and simplestructure, which is capable of conversion topolymers by combination with other identicalor similar molecules. For example, glucose isa monomer, which by interlinking ofmolecules can be built into large polymerssuch as the amylose and amylopectincontained in starch.

Monosaccharide A sugar monomer. Examplesare glucose, fructose and galactose. These arethe simplest forms of sugars, which are morereadily fermented by yeasts than theirpolymers (or polysaccharides).

Mother yeasting A system of yeast propagationfrequently used for molasses fermentations inwhich the propagator is not emptied entirelywhen inoculating a fermentor and the portionretained is used for starting another yeastpropagation cycle. By adding acid to maintaina low pH in the propagator it may be possibleto repeat the process for numerous cycleswithout excessive bacterial contamination.

Motor fuel grade ethanol (MFGE) Refers toanhydrous ethanol prior to denaturation tofuel ethanol. The term is used to distinguishit from the various grades of industrial andbeverage alcohol. MFGE is relatively crude,with considerable impurities, but conforms tolegal anhydrous standards. It has notundergone the rectification required to makeit suitable for industrial or beverage uses.

Motor octane number (MON) One of twomethods commonly used to assess the octane

rating of an automobile fuel. (See Octanerating).

MSW Abbreviation for municipal solid waste.

MTBE Abbreviation for methyl tertiary butylether.

Multiple effect evaporator A systemcomprising a series of interlinked evaporatorvessels (or ‘effects’) operating underincreasing degrees of vacuum. As liquids boilat lower temperatures with increasingvacuum, it is possible to use the latent heatof the vapors from one vessel to heat the next(and so on through a series of up to about sixvessels under increasing vacuums) with steamor other external source of heat only beingintroduced to the first vessel. This allowsconsiderable economy in energy consumption.Multiple effect evaporators are commonlyused for concentrating thin stillage solids intoa syrup.

Municipal solid waste (MSW) Regular citytrash or garbage. There have been numerousproposals and some pilot-scale attempts toproduce ethanol from MSW by hydrolyzingand fermenting the cellulosics contained in thematerial. Such efforts have, however,generally run into problems due to the normalwide variability of MSW.

MVR Abbreviation for mechanical vaporrecompression.

Mycotoxin A toxic substance produced bymolds.

N

Near infrared spectroscopy (NIR) NIR analysisemploys light beyond the visible spectrum toprovide analysis of organic substanceswithout the need for sample destruction. Itsadvantages to process management arevirtually instantaneous results versus the hoursor days needed with wet chemistry orchromatography. Applications include


analysis of incoming grains, profilingfermentors, distillate analysis, dry houseoperations, and finished products.

Net energy balance The amount of energyavailable from a fuel by combustion, less theamount of energy taken for its production. Inthe early 1980s a lot of emphasis was placedon the need for a net energy balance in theproduction of fuel ethanol to be used as agasoline extender. The emphasis was laterreduced when it became appreciated that a)much of the energy used in ethanolproduction may come from low qualitysources such as coal, wood or bunker-C oil,which cannot be used as automobile fuels,and b) that ethanol has a value as an octaneenhancer and oxygenate and not just as agasoline extender. In recent work the netenergy balance in corn to ethanol conversionis 1.3 t.

Neutral spirit Defined by the BATF as ‘distilledspirits produced from any material at or above1900 proof’. In practice, neutral spirit ispurified, odorless, tasteless and colorlessethanol produced by distillation andrectification techniques that remove anysignificant amount of congeners. It is used inthe production of beverages such as vodka,gin, cordials and cream liqueurs.

Nitrogen oxides Air-polluting gases containedin automobile emissions, which are regulatedby the EPA. They comprise colorless nitrousoxide (N

2O) (otherwise know as dinitrogen

monoxide or as the anaesthetic ‘laughinggas’), colorless nitric oxide (NO) and thereddish-brown nitrogen dioxide (NO

2). Nitric

oxide is very unstable and on exposure to airit is readily converted to nitrogen dioxide,which has an irritating odor and is verypoisonous. It contributes to the brownish layerin the atmospheric pollution over some metro-politan areas. Other nitrogen oxides of lesssignificance are nitrogen tetroxide (N

2O

4) and

nitrogen pentoxide (N2O

5). Nitrogen oxides

are sometimes collectively referred to as NOx

(or Nox) where x represents any proportionof oxygen to nitrogen.

Normal solution A solution containing oneequivalent weight of a dissolved substanceper liter. One volume of a normal acid willneutralize an equal volume of a normal alkali(or vice versa). This principle is used inmeasuring the acidity in fermentations bytitration with a standardized alkali. The‘normality’ of solutions is indicated by theletter N, as in 0.1N, 2N, etc.

O

Occupational Safety and HealthAdministration (OSHA) An agency of theUS Department of Labor with the mission ofdeveloping and promulgating occupationalsafety and health standards, developing andissuing regulations and conductinginspections and investigations to ensure theircompliance.

Octane A flammable liquid hydrocarbon ofchemical formula C

8H

18 found in petroleum.

One of the eighteen isomers of octane, 2,2,4-trimethylpentane is used as a standard inassessing the octane rating of fuels.

Octane enhancer Any substance such asethanol, methanol MTBE, ETBE, benzene,toluene, xylene, etc., that will raise the octanerating when blended with gasoline.

Octane rating (or octane number) A laboratoryassessment of a fuel’s ability to resist self-ignition or ‘knock’ during combustion in aspark-ignition engine. A standardized-design,single-cylinder, four-stroke engine with avariable compression ratio is used to comparethe knock resistance of a given fuel with thatof reference fuels composed of varyingproportions of two pure hydrocarbons. Onecomponent is the octane isomer 2,2,4-trimethylpentane (sometimes referred to asisooctane), which has a high resistance to


knock and is given the arbitrary rating of 100.The other component, heptane, has a very lowknock resistance and is given the base ratingof zero. For fuels with a rating higher than100 octane, the rating is obtained bydetermining how much tetraethyl lead needsto be added to pure isooctane to match itsknock resistance.

The engine knock tests are performed undertwo sets of operating conditions, the so-called‘motor’ or M method, and the ‘research’ or Rmethod. The average of the two results istaken to be a good indicator of a fuel’sperformance in a typical automobile on theroad. Hence, the use of the (R+M)÷2 octanerating on labels on service pumps.

OFA Abbreviation for Oxygenated FuelsAssociation.

Oligomer A macromolecule formed by thechemical union of two, three or four identicalunits known as monomers. The oligomers oftwo units are dimers, three units are trimersand four units are tetramers. (The prefix oligo-means ‘few’).

Oligosaccharide Short-chain polymers ofsimple sugars (or monosaccharides)generally considered to cover the range of 2-8 units. Short dextrins produced by hydrolysisof starch are included in this category.

OPIS Abbreviation for Oil Price InformationService.

Organic Adjective for chemical compoundscontaining carbon and hydrogen with orwithout oxygen, nitrogen or other elements.

Organoleptic testing The quality controlprocess of checking samples of alcoholicproducts on the basis of odor and taste. It isnormally performed by comparing samplesof new production with older samples ofacceptable quality that have been designatedas standards.

Orifice meter An instrument used to measurefluid flow by recording the differential

pressure across a restriction (or ‘orifice plate’)placed in the flow stream.

OSHA Abbreviation for Occupational Safetyand Health Administration.

Overhead vapors The vapors emerging fromthe top of a distillation column that areconducted to a condenser system.

Oxygenated fuels Literally meaning any fuelsubstance containing oxygen, the term iscommonly taken to cover gasoline-based fuelscontaining such oxygen-bearing compoundsas ethanol, methanol, MTBE, ETBE, etc.Oxygenated fuel tends to give a morecomplete combustion of its carbon to carbondioxide (rather than monoxide) and therebyto reduce air pollution from exhaust emissions.

P

Packed distillation column A column filled witha packing of ceramic, metal or other materialdesigned to increase the surface area forcontact between liquids and vapors.

PADD Abbreviation for Petroleum Administrationfor Defense District.

Pasteurization A method devised by Pasteurto partially sterilize a fluid by heating to aspecific temperature for a specific length oftime. Pasteurization does not greatly changechemical composition, but destroysundesirable bacterial contaminants. Thepractice may be applied to molasses or otherliquid fermentation feedstocks to reduce theinitial level of bacterial contamination.

Patent A certificate or grant by a governmentof an exclusive right with respect to aninvention for a limited period of time. In theUS the effective period is normally 17 yearsfrom the date of granting. Patents may coverprocesses, machines or other apparatus,methods of manufacture, composition ofmaterials and designs. Patents are publishedand freely available and are required to


provide sufficient information on an inventionso that anyone ‘reasonably versed in the art’would be able to implement it. Thus, patentscan be very useful sources of technicalinformation. Regardless of what may be statedin the title, summary or introduction, the legalessence of a patent is in the very preciselyand narrowly-defined claims appearing at theend of the text. Generally, the patent examinerdoes not revise the summary of a patent tocomply with the claims that are finallyapproved; so that the summaries publishedin various journals can be quite misleadingas to the scope of the respective patents.

Pentane A colorless, highly-flammablehydrocarbon liquid with a pleasant odor.Chemical formula C

5H

12, boiling point

37.1oC, freezing point -129.7oC. It is used asan entrainer for the dehydration of ethanolby azeotropic distillation. It presents a severeexplosion risk at low concentrations in air.

Pentanol See amyl alcohol.

Pentose General term for sugars with fivecarbon atoms per molecule such as xylose andarabinose, which are constituents ofhemicellulose. Pentoses are not fermented bynormal strains of distillers yeasts.

Permanganate (or Barbet) time A laboratorytest used for assessing the quality of samplesof industrial or beverage alcohol. It is thetime required for an alcohol sample todecolorize a standard potassiumpermanganate solution. The time is anindication of the reducing (deoxidizing)power of the sample, and is considered to bea crude measure of the presence of congeners.

Petroleum A naturally-occurring complexhydrocarbon, which may be liquid (crude oil),gaseous (natural gas), solid (asphalt, tar orbitumen) or a combination of these forms.

Petroleum Administration for Defense District(PADD) A designation of regions in the USfor the purposes of the Department of Energy’s

presentation of statistics on the usage ofpetroleum-based fuels including diesel fuel,gasoline and gasoline-ethanol blends. Theterm originated in World War II as ‘PetroleumAdministration for War Districts’. Theboundary lines were drawn to reflect thenatural logistical regions of the petroleumeconomy. Thus, the country is divided into 5PADDs: No. 1 (east coast) includes primarilythe area receiving most of its supply of crudeoil by tanker, No. 2 (midwest) and No. 3 (Gulfcoast) are defined in light of normal transportand supply operations, No. 4 (RockyMountains) is largely self-contained, No. 5(west coast) is naturally isolated from the restof the country.

PG Abbreviation for proof gallon.

pH A value measuring the acidity or alkalinityof an aqueous solution. Defined as thelogarithm of the reciprocal of the hydrogenion concentration. Pure water at roomtemperature has a pH of 7.0. Solutions with apH of less than 7 are acidic, and greater than7 are alkaline. As the pH scale is logarithmic,a solution with a pH of 5 has 10 times theacidity of a solution of pH 6. Control of pH isimportant in ethanol production both forobtaining optimal enzymatic activity and incontrolling the growth of bacterialcontaminants.

Phase separation The phenomenon of aseparation of a liquid (or vapor) into two ormore physically distinct and mechanicallyseparable portions or layers. It is used toadvantage in the dehydration of ethanol byazeotropic distillation with an entrainer suchas benzene. When the overhead vapors arecondensed, the liquid separates into twodistinct layers or phases. The upper phasecontains most of the benzene with someethanol and a little water, while the lowercontains most of the water with some ethanoland a little benzene. The lower layer isremoved and redistilled in an entrainerrecovery column to recover the benzene andethanol and to discharge the bulk of the water.


Plate (or tray) A contacting device placedhorizontally at intervals within a distillationcolumn. Plates may be simple perforated discswith or without downcomers as in the sieveplate and the dual flow plate, or they mayhave bubble caps, tunnel caps or varioustypes of floating valves to improve the contactbetween the rising vapor and descendingliquid. Sieve plates and tunnel caps are themost common form of plates used in ethanolproduction facilities.

Plate (or tray) distillation column A distillationcolumn with horizontally arranged contactingplates located at intervals up the column; ascompared to a packed column, which is filledrandomly with contacting devices.

Polymer A macromolecule formed by thechemical union of identical combining unitsknown as monomers. While generallyreferring to a combination of at least fivemonomers, in many cases the number ofmonomers is quite large, such as the averageof 3,500 glucose monomers in cellulose.

Polysaccharide A polymer composed ofnumerous sugar monomers or mono-saccharides. Examples are cellulose and theamylose and amylopectin in starch. Forfermentation purposes, polysaccharides mustbe subjected to hydrolysis to yield theircomponent fermentable monosaccharides.

Positive displacement pump A pump in whichthe displacement is accomplished bymechanically moving a volume of fluid fromthe pump suction to the discharge with nofluid slippage within the pump - no matterhow high the discharge pressure. With everystroke or rotation the same volume of fluid ispumped. Positive displacement pumps maytake various forms such as reciprocatingpiston pumps, gear pumps, lobe pumps anddiaphragm pumps. Generally, they are usedto move relatively low volumes of fluid athigh pressures. As they force a positivedisplacement, they may not be used insystems where the output may be throttled or

closed off without incorporating a pressurerelief or recycle system. Positive displacementpumps are frequently used for movingmolasses and stillage syrups, yeast slurries andfuel oil. They are also effective for meteringsince every rotation or stroke moves a precisevolume from the suction side to the dischargeside regardless of the pressure required.

Potassium aluminosilicate A zeolite. (Seemolecular sieve).

Pot still A simple batch distillation unit usedfor the production of heavily-flavoreddistillates for beverage use. It consists of atank (heated either by an internal steam coilor by an external fire) and an overhead vaporpipe leading to a condenser. It may be usedin the production of heavily-flavored rumsand whiskies.

Pre-fermentor See yeast propagator.

Proof A measure of the absolute ethanol contentof a distillate containing ethanol and water.In the US system, each degree of proof isequal to 0.5% ethanol by volume, so thatabsolute ethanol is 2000 proof. In the Imperialsystem 100 proof is equal to 57.06% ethanolby volume, or 48.24% by weight, whileabsolute ethanol is 75.25 over proof, or175.25 proof.

Proof gallon (PG) The volume of a liquid thatcontains the equivalent of one gallon ofethanol at 1000 proof.

Proof tables Books of tables compiled by theBATF and other organizations for thecalculation of proof of ethanol-containingliquids with adjustments for varyingtemperatures.

Propagation The process of increasing numbersof organisms by natural reproduction. In thecase of yeast, propagation occurs by buddingto form new cells in both aerobic andanaerobic conditions. Yeast companies growyeast under aerobic conditions with high


yield. Fuel alcohol propagation of yeast isnormally under anaerobic conditions (evenwith air supplied) due to the presence of highamounts of sugar. (See yeast propagator).

Propanol (or propyl alcohol) A minorconstituent of fusel oil. Chemical formulaC

3H

7OH. It exists as either of two isomers.

Both are colorless, toxic, flammable liquidswith odors similar to that of ethanol.

Protein Any of a class of high molecular weightpolymer compounds composed of a varietyof linked amino acids. Proteins are an essentialingredient in the diets of animals and humans.Yeast dry matter is approximately 50%protein. Corn DDG generally contains over27% crude protein.

PSI Abbreviation for pounds per square inch.Atmospheric pressure is 14.7 psi, 1 bar is 14.5psi.

PSIG Abbreviation for pounds per square inchgauge.

R

Rack price Used in reference to gasoline andfuel ethanol, it is the wholesale price at thetank-truck loading terminal exclusive of anyfederal, state or local taxes.

Reboiler A device for supplying heat to adistillation column without introducing livesteam. It generally consists of a shell-and-tubeheat exchanger connected to the base of thecolumn with liquid from the column enteringinside the tubes to be heated indirectly bysteam on the shell side.

Receiver A tank into which new distillates flowfrom the still for verifying proof, quantity andquality before transfer to shipping or storagetanks. Where the tank is sealed or has lockson the valves to meet government exciseregulations for preventing unlawful access, itmay be referred to as a ‘closed receiver’.

Recoopered barrels Used wooden barrels thathave been repaired to replace any crackedstaves or heads to be suitable for reuse in theaging of alcoholic beverages.

Rectification The process of concentrating andpurifying ethanol or other materials in arectifying column. The process may beoperated simultaneously with beer distillationwith the rectifying column directly connectedto a beer stripping column. In beverageethanol production the rectification processmay involve concentration and removal offusel oil, esters and heads. In fuel ethanolproduction the process commonly onlyinvolves concentration and the removal offusel oil. In some instances, where benzeneis to be used as an entrainer in the subsequentdehydration process, the heads may also beremoved in the rectification process.

Rectifying column (rectifier, rectificationcolumn or rectifying section) The portion ofa distillation column above the feed tray inwhich rising vapor is enriched by interactionwith a countercurrent descending stream ofcondensed vapor. In the case of ethanolproduction, the rectifying column may havevalves at draw-off points on various trays toallow the removal of accumulated fusel oil andother congeners. In some instances theethanol product is withdrawn from the refluxof the overhead vapor condensate, but inothers the product may be drawn from one ofthe upper trays to permit the heads or lowboilers to be purged from the reflux. (SeeConcentrating column).

Reflux The portion of the condensed overheadvapors returned to a distillation column tomaintain the liquid-vapor equilibrium.

Reflux ratio The ratio of the amount ofcondensate refluxed to the amount withdrawnas product. Generally, the higher the refluxratio the greater the degree of separation ofthe components in a distillation system.

Refractometer An instrument used to measure


the refractive index of liquids and liquidsolutions. Refractometers can be calibrated inconcentration of a solute instead of refractiveindex. In the ethanol industry refractometersare calibrated as degrees Brix or degreesBalling.

Reid vapor pressure (Rvp) A measure of thevapor pressure in pounds per square inch ofa sample of gasoline at 100oF. It is anindication of the volatility of a gasoline. Theblending of ethanol with gasoline tends toincrease Rvp while the blending of MTBE, andmore particularly ETBE, tends to reduce Rvp.

Renewable Fuels Association (RFA) TheWashington DC-based trade association forthe US fuel ethanol industry. Its membershipincludes companies involved in theproduction, blending and marketing ofethanol-blended fuels.

Research octane number (RON) See Octanerating.

Reverse osmosis A technique used in waterpurification and wastewater and stillagetreatment in which pressure is applied to theliquid in a suitable apparatus to force purewater through a membrane that does not allowthe passage of dissolved ions.

RFA Abbreviation for Renewable FuelsAssociation.

RhizozymeTM Saccharifying enzyme complexproduced in surface culture fermentation. Asit is cultured on wheat bran, a range ofcarbohydrases are produced in addition toglucoamylase, which results in moreextensive liberation of fermentable sugarsfrom grain sources.

(R+M)÷2 The formula for calculating octanerating.

Roller mill A mill for crushing or grinding grainor other solid material by passing it betweentwo, and as many as six, steel rollers. Therollers may be smooth, or serrated to shear

the grain, and they may turn at differing speedsto increase the abrasion. Roller mills aresuitable for small grains such as wheat, butdo not perform as well as a hammer mill oncorn.

RON Abbreviation for research octane number.

Rotameter A flow meter with a float and avariable area flow tube.

Rum Defined by the BATF as ‘an alcoholicdistillate from the fermented juice ofsugarcane, sugarcane syrup, sugarcanemolasses, or other sugarcane by-products,produced at less than 1900 proof, in such amanner that the distillate possesses the taste,aroma and characteristics generally attributedto rum’. Unlike the specifications for whiskies,the BATF does not require that rum be agedin oak barrels. British regulations specify thatrum be produced ‘from sugarcane productsin sugarcane-growing countries’.

Rvp Abbreviation for Reid vapor pressure.

Rye whisky Defined by the BATF as whiskyproduced at not more than 1600 proof, from afermented mash of not less that 51% rye andstored at not more than 1250 proof in charred,new oak barrels.

S

Saccharification The process of converting acomplex carbohydrate such as starch orcellulose into fermentable sugars such asglucose or maltose. It is essentially ahydrolysis. The process may be accomplishedby the use of enzymes or acids.

Sacc’ tank (or saccharification tank) A vesselwhere cooked mash is held for thesaccharification process. Sacc’ tanks aregenerally going out of use with the adoptionof simultaneous saccharification andfermentation in the fermentor.


Saccharomyces A genus of unicellular yeastsof the family Saccharomycetaceaedistinguished by the general absence ofmycelium and by their facility to reproduceasexually by budding. This genus includesthe species Saccharomyces cerevisiae, whichis the yeast most commonly used by bakers,brewers, distillers and wine producers.

Sanitizing In processing systems, sanitizing isreducing the population of viable organismsin the equipment to a level that is acceptablefor the process.

Scale A deposit formed on the surface ofequipment due to formation of insolubleinorganic compounds.

Scale inhibitor See Antiscalant.

Scaling The precipitation of salts in a distillationcolumn or heat exchanger which ifuncontrolled may reduce capacity andeventually block the equipment and make itunusable. Scaling occurs because some saltssuch as calcium sulfate and oxalate are lesssoluble at high temperatures and/or in thepresence of ethanol. Scaling problems aremore common in molasses beer distillation,but may occur in grain beer distillation whenconducted under pressure (i.e., at hightemperatures).

Scrubber A device for the removal of entrainedliquid droplets in a gas stream. Scrubbers maybe used to recover ethanol from the carbondioxide vented from fermentors.

SDA Abbreviation for specially denaturedalcohol.

SG Abbreviation for specific gravity.

Shooked barrel (or shook) A used bourbonwhisky barrel that has been dismantled toreduce the space requirements fortransportation.

Sieve analysis A laboratory test made on grainmeal to check that the milling process is beingconducted correctly. The meal is added to the

top of a stack of sieves with increasingly finemesh sizes descending downwards. The sievestack is vibrated for a standard time periodand the weight percentage retained on eachscreen is determined. With hammer mills, thesieve analysis will generally show that themeal gradually becomes more coarse as thehammers wear and need turning orreplacement.

Sieve plate (or sieve tray) A distillation columnplate with perforations of precise number andsize such that the ascending vapor passesthrough the plate vertically to mix with thedescending liquid held on the plate. A sieveplate differs from a dual flow plate in that it isdesigned for the descending liquid tooverflow down a downcomer, instead ofweeping through the perforations.

Simultaneous saccharification andfermentation (SSF) A procedure in whichsaccharification of a cooked (gelatinized andliquefied) starch mash occurs in the fermentor(by addition of glucoamylase) simultaneouslywith the commencement of fermentation (byyeast). This procedure is replacing thetraditional process taken from the whiskyindustry in which there is a specific holdingstage for saccharification with malt ormicrobial enzymes (in a sacc’ tank) before themash goes to a fermentor.

Slurry tank The vessel in which grain meal ismixed into a slurry with water and enzymebefore being pumped through a continuouscooking system.

Small Business Administration (SBA) A USfederal government agency charged withencouraging and assisting the developmentof small businesses. It offers financialassistance such as direct loans, loanguarantees and direct participation loans forvirtually any legitimate small businessdevelopment. The SBA has assisted in thefinancing of many small fuel ethanol facilities.

Soil Any unwanted material that is left on asurface that needs to be clean.


Solute One or more substances dissolved inanother substance (a solvent) to form asolution. The solute is uniformly dispersed inthe solvent in the form of molecules (as withsugars) or ions (as with salts).

Solution A uniformly dispersed mixture at themolecular or ionic level of one or moresubstances (the solute) in one or more othersubstances (the solvent). Ethanol and waterform a liquid-liquid solution, while sugar andwater form a solid-liquid solution.

Solvent A substance capable of dissolvinganother substance (a solute) to form auniformly-dispersed mixture (solution) at themolecular or ionic level.

Specially denatured alcohol (SDA) The termused to describe ethanol denatured with anyformulation of compounds selected from a listapproved by the BATF. The denaturant rendersthe ethanol unfit for beverage purposeswithout impairing its usefulness for otherapplications.

Specific gravity (SG) The ratio of the densityof a material to the density of a standardreference material such as water at a specifictemperature. For example, the specific gravityof ethanol is quoted as 0.7893 at 20/4oC. Thismeans that a given volume of ethanol at 20oCweighs 0.7893 times the weight of the samevolume of water at 4oC.

Spent sulfite liquor (SSL) See sulfite wasteliquor.

Spirit whisky Defined by the BATF as a mixtureof neutral spirits with at least 5% on a proofgallon basis of whisky or straight whisky.

SSF Abbreviation for simultaneous sacchar-ification and fermentation. In the enzymeindustry SSF also is used to indicate solid-state fermentation.

SSL Abbreviation for spent sulfite liquor. (Seesulfite waste liquor).

Starch A mixture of two carbohydrate polymers

(amylose and amylopectin), both of which arecomposed of glucose monomers linked byglycosidic bonds. Starch is a principal energystorage product of photosynthesis and isfound in most roots, tubers and cereal grains.Starch may be subjected to hydrolysis(saccharification) to yield dextrins andglucose.

Sterilization The complete destruction of allorganisms including viruses and spores.

Still An apparatus used in distillationcomprising a vessel in which the liquid isvaporized by heat and a cooling device inwhich the vapor is condensed. It may takethe form of a batch or continuously operableunit.

Stillage The mixture of non-fermentable (ornon-fermented) dissolved solids, insolublegrain fines proteins, dead yeasts and water,which are the residues after removal ofethanol from a fermented beer by distillation.Stillage may be dried to recover the solidmaterial (as DDG in the case of grainfeedstocks).

Stoichiometric yield The theoretical yield of aproduct of a chemical reaction as calculatedfrom the chemical reaction equation. Forexample, in the Gay Lussac equation for thefermentation of glucose to ethanol by yeast,C

6H

12O

6→ 2CO

2 + 2C

2H

5OH, 100 parts (by

weight) glucose should yield 48.89 partscarbon dioxide and 51.11 parts ethanol. Dueto a variety of factors, this yield is notachieved in practice.

Stoichiometry The branch of chemistry thatdeals with the quantities of substances thatenter into and are produced by chemicalreactions.

Stover The dried stalks and leaves remainingfrom a crop, particularly corn, after the grainhas been harvested. It is of interest as apotential source of cellulose feedstock forethanol production.


Straight whisky Defined by the US Bureau ofAlcohol, Tobacco and Firearms as a productconforming to the requirements for the whiskydesignation, which has been stored incharred, new oak barrels for a period of atleast two years. The straight whisky definitionmay include mixtures of straight whiskies ofthe same grain type, produced at the samedistillery, all of which are not less than fouryears old.

Strain See Yeast strain.

Stripping column (or stripping section) Theportion of a distillation column below the feedtray in which the descending liquid isprogressively depleted of its volatilecomponents by the introduction of heat at thebase.

Sub-octane blending The practice of blendingethanol with selected gasoline componentsthat have octane ratings below thecommercially or legally acceptable level. Ittakes advantage of the fact that ethanol canenhance the octane rating of the mixture tobring it to an acceptable level.

Sucrose Common table sugar, derived from beetor cane sources. It is a disaccharidecomprising a monomer of glucose linked to amonomer of fructose. It has a chemicalformula C

12H

22O

11. It is directly fermentable

by common yeasts to produce ethanol.

Sugar Any of a class of water-soluble, simplecarbohydrate, crystalline compounds thatvary widely in sweetness including themonosaccharides and lower oligosaccharides.Sugars may be chemically reducing or non-reducing compounds and are typicallyoptically active. Examples include themonosaccharides glucose, fructose, mannoseand xylose, and the disaccharides sucrose,maltose, lactose, and the trisaccharidesraffinose and maltotriose.

Sulfite waste liquor (SWL) An effluentproduced in the sulfite pulping process used

in some paper mills. It partly consists of adilute solution of sugars produced by the acidhydrolysis of cellulose. It may be used as afeedstock for the production of ethanol byfermentation using selected yeast strains afterstripping out the sulfite (or sulfur dioxide) withsteam.

Supplementary column See Demethylizingcolumn.

Sweet sorghum A plant of the species Sorghumbicolor, closely related to milo or grainsorghum, which is cultivated primarily for thesweet juice in its stem. It is widely used forcattle fodder and silage. The juice containssucrose, which may be extracted by crushingwith modified sugar mill equipment. The juicemay be used in the fermentative productionof ethanol. The plant has an advantage oversugarcane in that it can be grown over a muchwider range of climatic regions.

SWL Abbreviation for sulfite waste liquor.

Synthetic ethanol Ethanol produced by any ofseveral synthetic processes such as thecatalytic hydration of ethylene, the sulfuricacid hydration of ethylene and the Fischer-Tropsch process, in which it is a majorby-product of the synthesis of methanol bycatalytically reacting carbon dioxide andhydrogen. Synthetic ethanol is chemicallyidentical to fermentation ethanol, but does notqualify for US federal or state incentives forblending with gasoline and may not be usedin the production of alcoholic beverages.

T

Tafia An unaged Caribbean or South Americanalcoholic beverage produced by batchdistillation of beers obtained by thefermentation of sugarcane juice or molasses.It is similar to aguardiente and rum.

Tails See high boilers.


TBA Abbreviation for tertiary butyl alcohol. (Seebutanol).

Tequila Defined by the BATF as ‘an alcoholicdistillate from a fermented mash derivedprincipally from Agave tequilana Weber(‘blue’ variety) with or without additionalfermentable substances, distilled in such amanner that the distillate possesses the taste,aroma and characteristics normally attributedto tequila. It is a distinctive product ofMexico, manufactured in that country incompliance with the laws regulating themanufacture of tequila for consumption in thatcountry’. Tequila may be matured in woodencontainers to acquire a light gold color.

Ternary azeotrope An azeotrope or constantboiling mixture made up of three components.For example, a mixture of 74% volumebenzene, 18.5% ethanol and 7.5% of waterforms an azeotrope boiling at 64.9oC.

Tetramer A macromolecule produced bylinking four identical units known asmonomers. (See oligomer).

Tetraethyl lead (or lead tetraethyl) An organo-metallic compound used as an octaneenhancer in gasoline. Its use is regulated bythe EPA to control air pollution. If added togasoline on its own, tetraethyl lead wouldleave combustion residues of lead and leadoxide on engine cylinder walls. To preventaccumulation of these deposits, ethylenedibromide and dichloride are also added toconvert the lead to volatile halides beforerelease into the atmosphere.

Theoretical plate (or tray) A distillationcolumn plate or tray that produces perfectdistillation i.e. it would yield the samedifference in composition between liquid andvapor as that normally existing between theliquid and the vapor when in equilibrium. Adistillation column giving the same separationas 10 simple theoretical distillations is said tohave 10 theoretical plates.

Theoretical yield See stoichiometric yield.

Thermal efficiency The ratio of the energyoutput of a process to the energy input.

Thermal vapor recompression A method usingthe same recompression and recirculationprinciples as in mechanical vaporrecompression (to enhance the efficiency ofevaporation) except that recompression isperformed by passing steam through a venturijet on the vapor line.

Thermophilic Literally meaning ‘heat loving’,it refers to microorganisms such as bacteriathat thrive at warm temperatures, generallyin the range of 35-60oC. Lactobacillus is anexample of a thermophilic bacterium thatmultiplies rapidly at around 55oC. Appreciationof this fact has been a reason why the use ofsacc’ tanks has fallen into disfavor.

ThermosaccTM Stress-tolerant (heat andethanol) strain of Saccharomyces cerevisiaeused in ethanol production.

Thin stillage The liquid portion of stillageseparated from the solids by screening orcentrifuging. It contains some yeast,suspended fine particles and dissolvedmaterial. It is normally sent to an evaporatorto be concentrated to a thick syrup and thendried with the solids portion to give DDGS.

Titration A method of volumetricallydetermining the concentration of a substancein solution by adding a standard solution ofknown volume and strength until the reactionbetween the two is completed, usually asindicated by a change in color due to anadded chemical indicator. For instance, it iscommon practice to perform a titration with0.1 Normal sodium hydroxide solution usingphenol-phthalein as indicator to determine theacidity of a mash or beer.

Toluene An aromatic hydrocarbon compoundthat can be used as an octane enhancer ingasoline. It is a colorless, flammable liquidwith a benzene-like odor. Chemical formula


C6H

5CH

3. It boils at 110.7oC and freezes at -

94.5oC. It is derived from the catalyticreforming of gasoline or the distillation of coaltar light oil. It is known to be carcinogenic.

Total sugars as invert (TSAI) A simple crudeanalytical measure of reducing sugars inmolasses.

Transglucosidase An enzyme with the oppositeeffect of amyloglucosidase in that it bringsabout a polymerization of glucose intopolysaccharides. It may be present as animpurity in commercial enzyme preparations.

Trimer A macromolecule produced by thelinking three identical units known asmonomers. (See oligomer).

Tray See plate.

TSAI Abbreviation for total sugars as invert.

Tunnel cap A contacting device usedoccasionally on distillation plates . It isessentially an elongated bubble cap ,consisting of a chimney portion and a domedcover that deflects the vapors rising throughthe chimney, causing them to pass downthrough the liquid layer on the plate. Tunnelcaps may be made from interlocking metalchannels to form an integral part of the plate.

U

Upgrading A term used loosely to mean thedehydration of hydrous ethanol.

Ultrafiltration A process for the separation ofcolloidal or very fine solid materials, or largedissolved molecules, by filtration throughmicroporous or semi-permeable membranes.The process may be used for removal ofprotein from cheese whey prior tofermentation.

United States Department of Agriculture(USDA) A federal government department

with the mission to improve and maintain farmincome, to develop and expand markets foragricultural products, etc.

Urethane See Ethyl carbamate.

USDA Abbreviation for United StatesDepartment of Agriculture.

US gallon A measure of 231 cubic inches liquidat 60oF. It is the equivalent of 3.785 liters inthe metric system or 5/6 of an Imperial gallon.

V

Vacuum distillation A process that takesadvantage of the fact that liquids boil at lowertemperature under reduced pressure. Thus,distillation under vacuum may be used forsubstances that otherwise have a relativelyhigh boiling point such as ethylene glycol(which may be used in the dehydration ofethanol). The process may also be used in thedistillation of ethanol-water mixtures, as theazeotrope is formed at a higher proof underlower pressures. Vacuum distillation may beused in the production of rum or ethanol frommolasses to reduce the incidence of scalingin the column that occurs at hightemperatures.

Vacuum fermentation A process of operatinga fermentation under vacuum, so that theethanol or other product is vaporized andremoved as it is formed to avoid itsconcentration becoming inhibitory to theyeast. In a patented variation known as theVacuferm process, instead of maintaining theentire fermentor under vacuum, thefermenting beer is circulated through avacuum chamber to flash off the ethanolbefore returning the beer to the fermentor.

Vapor A dispersion in air of molecules of asubstance that is liquid or solid in its normalstate at standard temperature and pressure. Anexample is water vapor or steam.


Vaporization (or Volatilization) Theconversion of a chemical substance from aliquid or solid state to a vapor or gaseous stateby the application of heat, by the reductionof pressure or by a combination of theseprocesses.

Vapor pressure The saturation pressure exertedby vapors when in equilibrium with theirliquid or solid forms.

Vent condenser The final condenser in a seriesof two or more connected to the overheadvapor line of a distillation column. As the finalcondenser, it is the one from which non-condensed (or non-condensable) gases andvapors are vented to the atmosphere or to ascrubber system.

Vinasse The term sometimes applied to thestillage of molasses, grape juice or other liquidethanol feedstocks.

Vodka Currently defined by the BATF as‘neutral spirit so distilled, or so treated afterdistillation with charcoal or other materialsas to be without distinctive character, aroma,taste or color’. It should be noted that thecharcoal treatment is now optional, whereasin earlier BATF regulations it was mandatoryand the minimum time of treatment andamounts of fresh charcoal to be used werespecified.

Volatile Adjective to describe a solid or liquidthat readily converts to the vapor state.

Volatility The tendency of a solid or liquid topass into the vapor state at a giventemperature. With automotive fuels, thevolatility is determined by measuring the Reidvapor pressure (Rvp).

Volatilization See Vaporization.

W

Wash A British synonym for distillers beer.

Wet milling A process in which corn is firststeeped (or soaked) in water containing sulfurdioxide. This softens the kernels and loosensthe hulls. (The liquid when separated isknown as corn steep liquor). The grain isthen degermed by abrasion and liquidseparation. Oil is extracted from the germ,while the remainder of the kernel is groundin impact fiber mills and passed throughstationary screens to separate the starch andgluten from the fibrous portion. The heavierstarch is then separated from the gluten byuse of centrifuges. The starch portion maythen be processed into commercial starch, orit may be used as a feedstock for productionof ethanol or HFCS.

Weeping The condition when droplets of liquidfall through the holes of a sieve plate in adistillation column. It may be caused by a)steam flow that is too low, or b) too low aliquid flow to maintain a level on the plate, orc) having a tilted plate such that liquid depthis uneven.

Whey The serum or watery part of milkseparated from the curd in the process ofmaking cheese. It contains lactose and maybe used as a feedstock for ethanol production.

Whisky Defined by the BATF as ‘an alcoholicdistillate from a mash of grain, produced atless than 190° proof, in such as manner thatthe distillate possesses the taste, aroma andcharacteristics generally attributed to whisky’.With the exception of corn whisky, it shouldbe stored in oak barrels.

Whole stillage The entire stillage emerging froma distillation unit before any removal of solidsby screening or centrifuging.

Wine gallon A US gallon of liquid measure asdistinct from a proof gallon.

Wood alcohol See methanol.

Wort A brewery term used occasionally in


distilleries. It refers to an unfermented mash,particularly if produced from a liquidfeedstock or if the solids have been removedto yield a relatively clear, free-flowing liquid.

X

Xylene One of a group of three isomeric,aromatic hydrocarbons having a formula ofC

6H

4(CH

3)

2. Commercial xylene is normally

a mixture of the o- m- and p- isomers with aboiling point range of 137-145oC. It is a clear,flammable liquid. Derived from catalyticreforming of napthas or the distillation of coaltar, it may be used as an octane enhancer ingasoline.

Xylose A pentose (or 5-carbon sugar) derivedfrom the hydrolysis of hemicellulose. Chemicalformula C

5H

10O

5. It is not fermented by

normal strains of distillers yeasts.

Y

Yeast Any of certain unicellular fungi, generallymembers of the class Ascomycetaceae (a feware members of the class Basidiomycetaceae).Many yeasts are capable of producing ethanoland carbon dioxide by fermentation of sugars.Yeasts are composed of approximately 35%protein and are a rich nutritional source of Bvitamins.

Yeast autolysis The disintegration of yeast cellsby the action of their own enzymes.

Yeast cream Yeast concentrated by centrifugingor decanting from the contents of apropagator or fermentor to be used asinoculum in another fermentor.

Yeast propagator (or pre-fermentor) A tankused for the propagation or development of ayeast culture prior to transfer to a fermentor.It is normally fitted with aeration, agitationand cooling devices and is designed for easeof cleaning and sterilization.

Yeast recycle See cell recycle.

Yeast strain A pure culture of yeast derivedfrom a single isolation. Strains may bespecially selected for certain characteristicssuch as the ability to efficiently produce andtolerate high temperatures or high levels ofethanol (see ThermosaccTM).

Z

Zeolite See Molecular sieve.

Zymomonas A genus of the Pseudomonadaceaefamily of bacteria which are characterized bybeing Gram-negative and non-spore-forming.The genus Zymomonas is distinguished byits fermentation of sugar to ethanol. Theprincipal species being examinedcommercially for fuel ethanol production isZymomonas mobilis. It is, however,considered an undesirable contaminant inbeverage alcohol fermentations in that it tendsto produce hydrogen sulfide from sulfurcompounds in the mash, particularly thatderived from molasses.

Index 441

Index

Acetobacter (see Bacterial contaminants) 235Acidic cleaners 306Active dry yeast 96-97, 111, 121-122, 141Acrolein 103, 206Agave (see also Tequila)

Beverages from 223-224Blue agave cultivation 226-228

Alkali 305Alpha amylase

Bacillus sources 29Calcium 19Cooking systems 16Hydrolysis of starch 14, 29-32In malt 15, 29-31pH ranges 16, 18-19Temperature ranges 18-19, 20

AllcoholaseTM 12, 17, 18, 21AllpenTM 126AllproteaseTM 117AllzymeTM SSFAmmonia 115Amy alcohol (see fusel oils)Amyloglucosidase (see glucoamylase)Amylose

Content of cereals 13-14, 27-28Structure 14, 26-27

AmylopectinContent of cereals 13-14, 27-28Cassava 62Structure 15, 26-27

Anhydrous ethanol 335-336Antimicrobials

Against infection 126, 292-296Residues in DDGS 126, 293Continuous culture 142

Araban 46Arabanase 46Arabinose

Metabolism by fungi and bacteria 48-50AYFTM 82Azeotrope 321-323, 337

BacksetAcidification with 21Amounts added 16Composition 16, 346Contaminants 16, 284

FAN in, 347NIR analysis of 165

Bacterial contaminants 288Acetobacter 235, 287-288, 290Agave fermentation 235Continuous fermentation 140Gluconobacter 287-288Lactic acid bacteria 123-124, 206-208, 288-292

Inhibition of yeast 110, 290-292Loss of spirit due to 102, 110, 126, 208Off flavors 206Practical controls 126, 292-296Saccharification and 21Rum 249

Water recycling and 343-350Barley (see also Malt) 10, 12, 200, 217Barrelling 211, 217, 240-241, 251, 284-285Batch fermentation 130, 135Beerstone 131-132, 304Beta amylase 15, 29-31Biofilm 304Biomass to ethanol (see lignocellulosics)Biomethanator 343-344, 350-351Biorefinery 33, 389-398

Bacterial biocatalysts 397-398Biotransformation of DDGs 394

Biostill fermentor 138-139Bleach 306, 315BOD

Whey 68, 71Wastewater 374

Botanicals (see flavored beverages)Bourbon

History of 276-277Production of 278-285Fermentation 281-282Distillation 278, 282-283

Brettanomyces (see wild yeast)Brix defined 78

CakeGuardTM 382-383Calcium

Enzymes and 11Yeast and 106-108

Canadian whisky 260-261Candida

C. utilis 176

442 Index

C. tropicalis 177C. famata 177C. lipolytica 178C. guilliermondii 178C. kefyr 179C. boidinii 180

Carbon dioxide 122, 131, 140, 391Scrubber CIP 314Scrubber water composition 348-349

Carotenoids 172Cassava

Composition of chips 59Fermentation of 62-63World production 60Ethanol from 60-64

Caustic 309, 314, 316Cellulose

Content in biomass sources 42-43Structure 26-27, 44

Centrifuges 366-368Chelating agents 305-306Chlorine dioxide (see IotechTM) 131, 250, 307-308CIP systems 131-132, 301

Cycle examples 314-316NIR and 162

Citric acid Cycle (see TCA)Clean Air Act 2, 72Cleaning and sanitizing 299-318

Cleaning loops 312-313Four Ts 299Evaluation 317-318Temperature and 302-303Sanitary design 309-312Soil types 304Solutions 304-309

Clostridium saccharobutyricum 249Cocktails 268COD

Whey 68COP 301Coffey still 193, 208, 255, 259-260Congeners (see fusel oils)Continuous fermentation 132-133, 135-143

Cleaning 301Contamination in 136, 139-143Steady state 137

ConversionsMole fraction to % alcohol 258

Cooking 203, 344Agave 230-231Batch 15-16, 19Cassava 62Continuous 17-19, 203-204Jet cooker 18Relative heat requirements 18

Cooling methods 125, 131Cooperage (see Barrelling)

CopperYeast and 106-108Lignocellulosics and 47

Co-products (see also DDGS, Biorefinery) 363-365, 393Cordials 268Corn

Grind size 9-11Gelatinization of 12Yield from 10Sieve analysis 10, 38Starch in 27-28

Crabtree effect 96Cream yeast 96-97Cryptococcus laurentii 180Cylindroconical fermentor 130-131

DDG 392-394DDGS 5, 33-34

Antibiotics in 126, 293, 384Composition 381Dairy and beef cattle use 380-381Mycotoxins and 34-35, 128, 384NIR analysis of 156, 164-165Nutritive value for animals 378Pigs and poultry 383-384Production in NA 377Phosphorus availability 379-380, 383-384Sanitation 39Storage 375

DE 11, 20-21, 29Debaromyces hansenii 177-178Dekkera (see wild yeast)Denatured ethanol 355Dextran 78DextranaseDextrins 11, 14, 29

Limit dextrins 14, 29HPLC and 15

Disaccharides 25Distillation

Azeotropic distillation 321-323, 337Beverages 83, 239, 250-251, 255-266, 282-284Barbet head 262-263Contact devices 324, 330-331Continuous stills 210-211, 239, 259Demethylizer 262-263Energy use 264-265, 324-326, 331-333Extractive distillation 262Fundamentals explained 319-324Fusel oil removal 263-264, 266, 334Heat recovery 265Origins 194Packed column 264Pinch point 262Pot stills 204, 208-210, 255-257Pressure distillation 264-265Reboilers 264, 331Rectifier 250, 261-262, 319

Index 443

Thermocompressors 265, 331Tower sizing 327-330Vacuum distillation 265

Distillers bushel 10Dryers 371-374Dryhouse

Design and equipment 365-375Energy use 360NIR applications 164-165

E85 5Ediesel 5Effluent treatment and disposal 374

Molasses 83-84Tequila 239-240Whey processing 71-72Whisky 213

Emissions 374-375Energy use

Distribution in plant 359-361Factors affecting 357-359Thermal and electrical ratios 356-357

Esters 109-110, 213, 238, 249Esterase 46Ethanolgenic microbes 47Ethyl acetate (see esters)Ethyl caproate (see esters)Evaporators 369-371

Condensate 348, 374

FAN 113-114, 346Fermentation (see Batch, Continuous)

Monitoring with NIR 159Flavor compounds

Distillation and 255-266Lactic bacteria and 249NCY 172Produced by yeast 104-105, 213-220

Flavored beverages 258-266Artificial vs natural 269Botanicals 258-259, 269Color 271Extracts 269Flavor perception 267Tax benefits 270, 271Types of 268-269Vodka 255

Flocculation 105Fuel ethanol industry 1-7

Asia 6Australia 5, 6Brazil 5-6China 5-6EU 6, 72History 1-2India 5Thailand 5US 2-5, 41, 72

Fungi 396-397Fusel oils 108-110, 208, 212, 235-238, 262-263FructoseFruit 262

Galactan 46Galactanase 46Gasohol 1Gelatinization 11-12, 28

Cassava 62Glucoamylase 20, 29-31

Activity 14-15Conventional vs RhizozymeTM 20pH 16, 110Source organisms 21, 29-30

Gin 255, 258-259Glucose

Structure 25-26Glucan

In lignocellulosics 46In yeast cell wall 87-88

Gluconobacter (see Bacterial contaminants)Glycerol 102, 206, 391Glycolysis 23-24, 92Glycogen 26

In yeast 87-88, 112Structure 112

GMPs 33-34Grain handling 33-38, 280

Contaminants 34-35Moisture content 35NIR analysis 158-159Sampling 35Sanitation 37-38

Grain whisky 203-204Grind 9-11, 61, 202Guiacol 201Guy-Lussac equation 93

Hansenula polymorpha 182Hayflick Limit 88Hammer mill 9-10Heat

Control equipment 125Production by yeast 122-123

HemicelluloseComposition 43Structure 44

Hemicellulase 47High gravity fermentation 117Higher alcohols (see fusel oils)High TTM 17, 18HPLC vs NIR 159-162Hydrofining 262Hydroheater 18

Industrial ethanol 333-335

444 Index

Infection (see Bacterial contaminants)Inulin (see also Tequila) 230-231Invert sugars 77IotechTM 314Irish whiskey 194-195, 255-256

vs Scotch 197Production 197-213

Iron 106-107Isoamyl acetate (see esters)Isoamyl alcohol (see fusel oils)Isobutyl acetate (see esters)Iodophors 131, 302, 309

Jet cooker 18

Killer yeast 98-99Kluyveromyces

K. lactis 179K. marxianus 71, 179

Kreb’s Cycle (see TCA)

Lactic acid (see also Bacterial contaminants)Inhibition of yeast 110, 290-292In steep water 345Measurement with NIR 162-163, 165

LactosideTM 63, 126, 294-296Lactose 25, 69

Structure 71Hydrolysis of 70

Lag phase 94Leuconostoc, 206

Molasses contaminant 83Lignin 43

Ligninases 47Lignocellulosics

Composition 41-43Enzymes for 46-47Microbial fermentation of 47-52Processing 44-46Sources 41

Liquefaction 11-21, 62Log phase 96

Magnesium 106-108Maillard reactions 121Malt 25, 199-201, 202, 279

Enzymes 11, 15, 18, 117Malternatives 268-269Maltose 91, 205Maltotriose, 205Manganese 106-108Mannan

In lignocellulosics 46In yeast cell wall 87-88

Mash tun 202Maturation 211-213, 218, 260, 240, 251-252, 284-285Methanol 237, 262Methylene blue stain 101

Milo (see sorghum)Milling 9-11, 37-38

Cassava 61Agave 231-232

Molasses (see also Rum) 23, 75-84Cane vs beet 76Composition of 76-78Dextrans 78Dilution of 79-80Fermentation 80-83Handling 79-80Types of 75-76World production 75

Molecular sieve 337-341Monosaccharides 25MTBE 2, 4-5, 356Mycotoxins 34-35, 128, 384

Neutral spirit 10, 259-263NIR 9, 145-169

Calibration 153-158Dryhouse applications 164-165Ethanol 159-162, 166-169Lactic acid 162Grain quality 158-159Measuring techniques 150-152Spectrometer types 147-149Starch analysis 163-164Sugars 161-162Troubleshooting fermentors 163

Non-conventional yeast (NCY) 171- 190, 396Taxonomy 171-172Products of 172Physiology of 175-177

n-propanol (see fusel oils)

Oats 200Oligosaccharides 25Organoleptic compounds

Scotch 213-220Tequila 235-238

Oxygenates 356

Peracetic acid 308Pichia

P. angusta 182P. stipitis 48-49, 182P. guilliermondii 178-179P. pastoris 181

Phaffia rhodozyma 181-182Phenethyl alcohol (see fusel oils)Phytic acid 129Polysaccharides 25Potassium 106-107Potatoes 262Pullulanase 30, 31Pressed yeast 96-97, 121-122Protease 117

Index 445

Quaternary ammonium compounds 308-309

Reducing sugars 77Reformulated gasoline 2Rice 12Roller mill 10-11, 202-203Rum (see also Molasses)

Types of 247-248Feedstocks 248-249Heavy rum 249Lactic bacteria 249Light rum 249-250Distillation 250-251

RhizozymeTM 9, 10, 12, 17, 18, 21, 117, 126, 128Agave wort 235Cassava 62-63Energy use and 358vs conventional gluc 20, 124-125

Rye 10, 12, 355Whisky 275, 279, 282

Saccharification 11, 14-16, 20Saccharomyces cerevisiae in ethanol production

Acid washing 132Aerobic vs anaerobic metabolism 92-94Amino acid uptake 115Budding 87, 101-102, 124Cell features 86-88Cell growth phases 94, 127Cell wall structure 87-88Composition 88Conditioning yeast 97, 127-128, 234, 280-281Counting 100Energy 94-95Ethanol tolerance 117, 122, 127Ester formation 109-110Flavor compounds 104-105, 109Foods 117Fusel/higher alcohols 108-109Glycolysis in 92-94Glucose tolerance 125-125Growth requirements 89-91Magnesium 107-108Metabolic engineering 395Mineral requirements 106-107Molasses and nutrition of 82Mutations 113Mycotoxins 128Nitrogen and FAN 113-117Oxygen use by 96, 110pH optima 102Phosphorus 107Pitching 100Products of fermentation 102Propagation of 96-97Sucrose hydrolysis by 90Sugar uptake 90-91, 93, 205Sulfur metabolism 104, 107

Size of 86Sterol needs 110-111Strain selection 121-122Stress factors 117, 122-130, 349-350Taxonomy 85Temperature tolerance 99-100, 122Viability 100-102Vitality 100Vitamin needs 105Xylose fermenting 48-52, 395-396

Saccharomyces pombe 182Sanitizing 301-302, 307-309Scale 126, 315Scale-BanTM 126Scale-BiteTM 314Schizosacharomyces pombe 182Scotch whisky 15, 255

Authenticating malts 219-220Blending 213Defined 193-194Maturation 211-213Organoleptic compounds 213-220Production 199-213Regions 196-vs Irish 197Yeast 204

Sieve analysis 10, 38, 279Simultaneous saccharification and fermentation 20, 29, 48,

62, 140Soils 304Sodium

Yeast and 108, 351-352Sorghum 12, 355

DDGS 379Spent grains (see also DDGS) 202Starch 26-32 (see amylose, amylopectin)

Hydrolysis of 12-15, 23-32Steep water composition 345Sterols 110, 111-112Stillage

Thin stillage 346-347Whole stillage 363-365Heat recovery 366

Stuck fermentation 113, 206Sugarbeets (see molasses) 75Surfactants 306Succinic acid 103, 391Sucrase 23Sucrose 23

Structure 77Sugar cane (see molasses)Sulfur

Metabolism by yeast 104Surface culture enzyme (see RhizozymeTM)SuperstartTM 121

TCA cycle 88, 103-104Tequila 223-244

446 Index

Contamination 235-236Cooking 230-232Fermentation 232-236Maturation 240-241Organoleptic characteristics 235Origin and history 224-225Processing agave 229-230Production stats 241-42Yeast 232-236

Thermal oxidizers 375ThermosaccTM 17, 18, 82, 100, 121, 128, 233Thin stillage (see also Stillage, Evaporators) 346-347Torula utilis 176Total sugars as invert 78Trehalose 112-113

Urea 115U-tube cooking system 17-19

Viscosity 11-12, 17, 18-19Molasses 79

Vitamins in yeast metabolism 105-106NCY 172

Vodka 259, 275Flavored 267

Water reuse 343Effect on yeast 349-353

Wet grains 382-383Wheat 12Whey

Composition 65, 70Ethanol production from 70-72Pollution and 65, 68Production of 65, 69Uses of 68-70

Whisky (see Scotch, Irish, Canadian, Bourbon)Wild yeast 98

Xylan 43Xylose

Metabolism by fungi and bacteria 48-50Xlyanase 46

Yarrow lipolytica 178Yeast (see also Saccharomyces cerevisiae other yeast species,

wild yeast, killer yeast, nonconventional yeast)

Zeolites 337Zero effluent design 353Zinc 106-108 (see S. cerevisiae mineral needs)Zygosaccharomyces rouxii 180Zymomonas mobilis 48, 397

Molasses contaminant 83

JACQUESLYONS

KELSALL2003

THEALCOHOLTEXTBO

OK

4THEDITIO

NA

reference for the beverage, fuel and industrial alcohol industries

THEALCOHOLTEXTBO

OK

4THEDITIO

NA

reference for the beverage, fuel and industrial alcohol industries

16683522 alcohol textbook 4th ed

Documents