white paper selective non-catalytic reduction (sncr) for controlling

30
WHITE PAPER SELECTIVE NON-CATALYTIC REDUCTION (SNCR) FOR CONTROLLING NO X EMISSIONS PREPARED BY: SNCR COMMITTEE INSTITUTE OF CLEAN AIR COMPANIES, INC. February 2008 Copyright © Institute of Clean Air Companies, Inc., 2008. All rights reserved.

Upload: truongthuy

Post on 13-Jan-2017

225 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: white paper selective non-catalytic reduction (sncr) for controlling

WHITE PAPER

SELECTIVE NON-CATALYTIC

REDUCTION (SNCR)

FOR CONTROLLING

NOX EMISSIONS

PREPARED BY:

SNCR COMMITTEE

INSTITUTE OF CLEAN AIR COMPANIES, INC.

February 2008

Copyright © Institute of Clean Air Companies, Inc., 2008. All rights reserved.

Page 2: white paper selective non-catalytic reduction (sncr) for controlling

ICACThe Institute of Clean Air Companies

(ICAC) is the national association of com-

panies that supply stationary source air

pollution monitoring and control systems,

equipment, and services. It was formed in

1960 as a nonprofit corporation to pro-

mote the industry and encourage improve-

ment of engineering and technical

standards.

The Institute’s mission is to assure a

strong and workable air quality policy

that promotes public health, environmen-

tal quality, and industrial progress. As the

representative of the air pollution control

industry, the Institute seeks to evaluate

and respond to regulatory initiatives and

establish technical standards to the benefit

of all.

2

Page 3: white paper selective non-catalytic reduction (sncr) for controlling

Table of Contents

page

PURPOSE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

EXECUTIVE SUMMARY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

SELECTIVE NON-CATALYTIC REDUCTION (SNCR)

FOR CONTROLLING NOx EMISSIONS . . . . . . . . . . . . . . . . 4

What is SNCR?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

How much NOx can SNCR remove? . . . . . . . . . . . . . . . . . . 5

Is SNCR a new technology? . . . . . . . . . . . . . . . . . . . . . . . . . . 6

Is SNCR commercially deployed?. . . . . . . . . . . . . . . . . . . . . 6

Are there applications for which SNCR is particularly

suited? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

How much does SNCR cost? . . . . . . . . . . . . . . . . . . . . . . . . . 7

What about ammonia slip? . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Does SNCR have other limitations? . . . . . . . . . . . . . . . . . . . 8

What are the common misconceptions regarding

SNCR? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

Can SNCR be used in combination with selective

catalytic reduction (SCR)? . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

How can SNCR be used to best advantage?. . . . . . . . . . . . 9

What are the water quality considerations where

urea or aqueous ammonia is utilized for SNCR? . . . . . . 10

APPENDIX 1: Selected Applications of Urea-Based SNCR,

by Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

APPENDIX 2: Selected Applications of Ammonia-Based

SNCR, by Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

3

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 4: white paper selective non-catalytic reduction (sncr) for controlling

PURPOSE

To comply with federal, state, and local acid rain andozone non-attainment rules, both regulators and regu-lated industry seek nitrogen oxide (NOx) controlswhich offer the greatest reliability and effectiveness atthe least cost. One such NOx control technology is se-lective non-catalytic reduction (SNCR). Although SNCRwill not be universally applicable, or always the mostcost effective control strategy, in many cases it willmeet the dual requirements of high performance andlow cost, and so should be considered by affectedsources and permitting authorities. To date, SNCRtechnology has been installed on 90 units in the powergeneration industry and on more than 300 industrialunits (see Appendix 1 for a partial installation list).

The SNCR Committee of the Institute of Clean AirCompanies, Inc. (ICAC) prepared this white paper toeducate all interested parties on the capabilities, limita-tions, and cost of SNCR.

ICAC is the nonprofit national association of com-panies which supply stationary source air pollutionmonitoring and control systems, equipment, and ser-vices. Its members include suppliers of SNCR systems,and of competing NOx control technologies.

EXECUTIVE SUMMARY

Selective non-catalytic reduction (SNCR) is a chemicalprocess for removing nitrogen oxides (NOx) from fluegas. In the SNCR process, a reagent, typically urea oranhydrous gaseous ammonia, is injected into the hotflue gas, and reacts with the NOx, converting it to nitro-gen gas and water vapor. No catalyst is required forthis process. Instead, it is driven by the high tempera-tures normally found in combustion sources.

SNCR performance depends on factors specific toeach source, including flue gas temperature, availableresidence time for the reagent and flue gas to mix andreact, amount of reagent injected, reagent distribution,uncontrolled NOx level, and CO and O2 concentrations.However, reductions in emissions of 25-75 % are com-mon. Using appropriately designed SNCR systems,these levels of control are not accompanied by exces-sive emissions of unreacted ammonia (ammonia slip)or of other pollutants, particularly using recent designupgrades demonstrated on commercial systems.Further, SNCR does not generate any solid or liquidwastes.

SNCR also may be combined with low NOx burn-ers (LNB), over-fired air (OFA), neural networks, richreagent injection (RRI), and selective catalytic reduc-tion (SCR) systems or with gas reburn technologies toprovide deeper emissions reductions for moderate cap-ital investment. A combined SNCR/SCR system can bedesigned to use substantially less catalyst (typically in-stalled “in-duct”) than a conventional SCR, allowing

higher overall NOx reduction than SNCR alone andlower ammonia slip, but with a relatively moderate in-crease in capital cost. A combined SNCR/SCR systemcan also be designed, particularly in the case of moder-ate duty boilers, to have a SNCR/SCR system to per-form equivalently to a full SCR system to smoothenNOx reduction at lower loads.

SNCR is a proven and reliable technology. SNCRwas first applied commercially in 1974, and significantadvances in understanding the chemistry of the SNCRprocess since then have led to improved NOx removalcapabilities as well as better ammonia slip control. As aresult, approximately 400 SNCR systems have been in-stalled worldwide. Applications include utility and in-dustrial boilers, process heaters, municipal wastecombustors, and other combustion sources.

SNCR is not a capital-intensive technology. Lowcapital costs, ranging from $5-20/kWe on power gener-ation units, make SNCR particularly suitable for use onlower capacity factor units, on units with short remain-ing service lives and for seasonal control. SNCR also iswell suited for NOx “trimming” and for use in combi-nation with other NOx reduction technologies. SNCRcan provide 10-25 % reductions in power generationboiler NOx emissions for total costs below 1 mill/kWh.Removal cost effectiveness values for SNCR centeraround $1,500-2,500 per ton of NOx removed.

The performance and cost of SNCR make thistechnology attractive for export, including to develop-ing and former Soviet Union countries.

SELECTIVE NON-CATALYTICREDUCTION (SNCR) FORCONTROLLING NOX EMISSIONS

What is SNCR?

Selective non-catalytic reduction (SNCR) is achemical process that changes nitrogen oxides (NOx)into molecular nitrogen (N2), carbon dioxide (CO2)(if urea is used), and water vapor. A reducing agent,typically anhydrous gaseous ammonia or liquid urea,is injected into the combustion/process gases. At suit-ably high temperatures (1,600 - 2,100 F)11, the de-sired chemical reactions occur.

Conceptually, the SNCR process is quite simple. Agaseous or aqueous reagent of a selected nitrogenouscompound is injected into, and mixed with, the hot fluegas in the proper temperature range. The reagent then,without a catalyst, reacts with the NOx in the gasstream, converting it to harmless nitrogen gas, carbondioxide gas (if urea is injected), and water vapor. SNCRis “selective” in that the reagent reacts primarily withNOx,. A schematic depicting the SNCR process isshown in Figure 1.2

4

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 5: white paper selective non-catalytic reduction (sncr) for controlling

No solid or liquid wastes are created in the SNCRprocess.

While either urea or ammonia can be used as thereagent, for most commercial SNCR systems, urea hasbecome the prevalent reagent used. Urea is injected asan aqueous solution while ammonia is typically in-jected in either its gaseous or anhydrous form usingcarrier air as a dilutive and support medium.

The principal components of the SNCR system arethe reagent storage and injection system, which in-cludes tanks, pumps, injectors, distribution modules,and associated controls. Given the simplicity of thesecomponents, installation of SNCR is easy relative to theinstallation of other NOx control technologies. SNCRretrofits typically do not require extended source shutdowns.

How much NOx can SNCR remove?

While SNCR performance is specific to eachunique application, NOx reduction levels rangingfrom 30 % to more than 75 % have been reported.

Temperature, residence time, reagent injectionrate, reagent distribution in the flue gas, uncontrolledNOx level, and CO and O2 concentrations are importantin determining the effectiveness of SNCR.3 In general,if NOx and reagent are in contact at the proper temper-ature for a long enough time, then SNCR will be suc-cessful at reducing the NOx level.

SNCR is most effective within a specified tempera-ture range or window. A typical removal effectivenesscurve, as a function of temperature within this win-dow, is shown in Figure 2. At temperatures below the

window, reaction rates are extremely low, so that littleor no NOx reduction occurs. As the temperature withinthe window increases, the NOx removal efficiency in-creases because reaction rates increase with tempera-ture. Residence time typically is the limiting factor forNOx reduction in this range. At the plateau, reactionrates are optimal for NOx reduction. A temperaturevariation in this range will have only a small effect onNOx reduction.

5

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Figure 1. SNCR Process Schematic. Source: Fuel Tech

Figure 2. Typical SNCR Temperature Ranges

A further increase in temperature beyond theplateau decreases NOx reduction. On the right side ofthe curve, the oxidation of reagent becomes a signifi-cant path and competes with the NOx reduction reac-tions for the reagent. Although the efficiency is lessthan the optimum, operation on the right side is prac-ticed and recommended to minimize byproduct emis-sions. On the left side of the curve, there is also greaterpotential for ammonia slip for a given NOx removaland residence time.

The effective temperature window becomes wideras the residence time increases, thus improving the re-moval efficiency characteristics of the process. Longresidence times (�0.3 second) at optimum tempera-tures promote high NOx reductions even with less thanoptimum mixing.

Normal stoichiometric ratio (NSR) is the term usedto describe the N/NO molar ratio of the reagent in-jected to the uncontrolled NOx concentrations. In gen-eral, one mole of ammonia species will react with onemole of NO in the reduction reaction. If one mole ofanhydrous ammonia is injected for each mole of NOx

in the flue gas, the NSR is one, as one mole of ammo-nia will react with one mole of NOx. If one mole ofurea is injected into the flue gas for each mole of NOx,the NSR is two. This is because one mole of urea con-tains two ammonia radicals and will react with twomoles of NOx.3 For both reagents, the higher the NSR,the greater the NOx reduction. Increasing NSR beyonda certain point, however, will have a diminishing effect

Page 6: white paper selective non-catalytic reduction (sncr) for controlling

on NOx reduction with a resultant increase in ammo-nia slip and reagent cost.

Is SNCR a new technology?

No. Commercial installations using SNCR havebeen in existence for more than 30 years.

The first commercial application of SNCR was inJapan in 1974.4 This installation used anhydrous am-monia. At about the same time, the anhydrous ammo-nia injection process was patented in the U.S. by ExxonResearch and Engineering Co. This process is com-monly known as the Thermal DeNOx process.

Fundamental thermodynamic and kinetic studiesof the NOx-urea reaction occurred during 1976-1981under the direction of the Electric Power ResearchInstitute (EPRI). Patents granted to EPRI for thisprocess were licensed to Fuel Tech which, with its im-plementors and sub-licensees, has marketed the urea-based NOxOUTR process with improvements to theoriginal patents.

Is SNCR commercially deployed?

SNCR systems are in commercial operation inthe United States, as well as in Europe and Asia andis one of the key technologies used for compliancewith the NOx SIP Call program.

SNCR is a fully commercial NOx reduction tech-nology, with successful application of the urea- andammonia-based processes at approximately 400 instal-lations worldwide (see Appendix 1 and 2), covering awide array of stationary combustion units firing anequally diverse types of fuels.

In the U.S., commercial installations or full-scaledemonstrations include virtually every boiler configu-ration and fuel type, as well as other major NOx emit-ting process units, such as cement kilns andincinerators. Urea-based SNCR has been applied com-mercially to sources ranging in size from a 60MMBtu/hr (gross heat input) paper mill sludge incin-erator to a 640 MWe pulverized coal-fueled, wall-firedelectric utility boiler. The earliest commercial urea-based SNCR system in the U.S. was installed in early1988 on a 614 MMBtu/hr CO boiler in a SouthernCalifornia oil refinery. This SNCR system reduces NOx

emissions 65 % from a baseline of 90 ppm.Industrial boilers, process units, municipal and

hazardous waste combustors, and power boilers makeup the largest share of commercial SNCR installationsin the U.S. This distribution is determined more byNOx control regulations than by SNCR process limita-tions. To illustrate the breadth of deployment of SNCR,the following examples of commercial installations include:

• Two 500 MWe cyclone-fired boilers at Amerenutilize a combination of SNCR with RRI®, whichis an offshoot of SNCR technology under licensewith EPRI.

• Two 75 MWe pulverized coal tangentially firedpower boilers in California equipped with lowNOx burners and overfire air required the instal-lation of SNCR to meet a 165 ppm permit limit.5

• SNCR systems installed on the coal-burning,wall-fired Dominion Energy’s Salem HarborStation Units 1, 2 (84 MWe each) and 3 (156MWe) in 1993, together with LNBs, can reduceNOx emissions 50-75 % from a baseline of 0.85-1.12 lb/MMBtu.

• Commercial SNCR systems retrofit on 320 MWewet-bottom, twin furnace boilers in New Jerseyprovide 30-35 % NOx reductions.6

• Commercial SNCR systems retrofit on cyclone-fired boilers in New Jersey reduce NOx emis-sions by 35-40 %.

• SNCR is achieving compliance with RACT limitsat coal-fired boilers in Massachusetts7 andDelaware.8

• A SNCR system installed on a 640MW supercriti-cal boiler is achieving 25 % removal efficiencyusing only wall injectors. This option offerslower cost (about $6/KW including installation)than utilizing multi-nozzle lances.

• SNCR systems at Duke Energy’s Marshall Stationon 600 MWe boilers incrementally reduced NOx

by 25 % above the reductions being obtainedwith LNB.

• An SNCR system installed on a circulating flu-idized bed boiler designed to produce 350,000lb/hr of steam can reduce NOx emissions from abaseline of 0.2-0.35 lb/MMBtu to below 0.15lb/MMBtu over a load range of 40-100 %.9

• Among significant applications in the U.S.:

• A SNCR system on a 600 MW coal-fired boilerfiring 3.5 % sulfur coal reduced NOx by 30 %across the load range while maintainging ammo-nia slip near 5 ppm. The unit experienced veryfew operational difficulties.10

• SNCR, in conjunction with combustion temper-ing, is achieving NOx reductions of nearly 60 %on a 244 MWe gas-fueld cyclone boiler.11

• SNCR, in conjunction with burner optimizations,reduced NOx on coal over 70 % on coal fired boilers.12

• SNCR provided an 80� % reduction from uncon-trolled emissions of 3.5-6.0 lb NOx per ton ofclinker in a demonstration at a West Coast ce-ment kiln.

• A SNCR system in combination with a modifiedreburn process is meeting 0.2 lb/MMBtu on a600 MW boiler firing Powder River Basin coal.

SNCR also has been commercially installed anddemonstrated in Asia. For example, an SNCR systeminstalled on a 331 MMBtu/hr pulverized coal-fired in-dustrial boiler in Kaohsuing, Taiwan, in 1992 reduced

6

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 7: white paper selective non-catalytic reduction (sncr) for controlling

NOx emissions from this front-fired boiler from 300 to120 ppm.

In addition, SNCR has been commercially installedthroughout Europe. Installations include coal-fueleddistrict heating plant boilers, electric utility boilers, mu-nicipal waste incinerators, and many package boilers.

In Germany, commercial SNCR systems installedon municipal waste incinerators in Hamm, Herten, andFrankfurt reduce NOx emissions 40-75 % from base-lines of 160-185 ppm. SNCR also has been installed onmore than 20 heavy oil-fired Standardkessel packageboilers.

In Sweden, a commercial SNCR system on a 275MMBtu/hr coal-fueled, stoker-fired boiler at theLinkoping P1 district heating plant reduces NOx emis-sions 65 % from a baseline of 300-350 ppm. At theNykoping demonstration on a 135 MMBtu/hr coal-fu-eled circulating fluidized-bed boiler, SNCR achieves a70 % NOx reduction from a 120-130 ppm baseline.Demonstrations of SNCR, in addition to municipalwaste incinerators and wood- and coal-fueled districtheating plant boilers, included a pulp and paper millkraft recovery boiler, where a 60 % reduction from un-controlled emissions of 60 ppm was attained.13

To meet new environmental demands in EasternEurope, SNCR systems were installed on five coal-firedindustrial boilers in the Czech Republic since 1992.

Are there applications for which SNCR is par-ticularly suited?

Yes. Some applications have combinations oftemperature, residence time, unit geometry, and un-controlled NOx level, and operating modes whichmake them especially well-suited for cost-effectivereduction of NOx by SNCR.

Certain applications are technically well-suited forthe use of SNCR. These include combustion sourceswith exit temperatures in the 1550-1950 ºF range andresidence times of one second or more, examples ofwhich are many municipal waste combustors, sludgeincinerators, CO boilers, and circulating fluidized bedboilers. Furnaces or boilers with high NOx levels orwhich are not suited to combustion controls, e.g., cyclone-type or other wet bottom boilers and stokersand grate-fired systems, also are good candidates forSNCR.

Other applications are well-suited to the use ofSNCR for economic reasons. For these applications,controls with reduced capital cost, even at the expenseof somewhat higher operating costs, may be the leastexpensive to operate. Applications meeting these crite-ria include units with lower capacity factors, such aspeaking and cycling boilers, units requiring limitedcontrol, e.g., additional “trim” beyond combustion con-trol or seasonal control.

How much does SNCR cost?

The capital cost of a selective non-catalytic re-duction system is among the lowest of all NOx re-

duction methods. Recent innovations in the controlof reagent injection make SNCR operating costsamongst the lowest of all NOx reductionmethods.

SNCR is an operating expense-driven technology,so that the absolute cost of applying SNCR varies di-rectly with the NOx reduction requirements.

Typical SNCR capital costs for utility applicationsare $5-15/kW, vendor scope, which corresponds to amaximum of $20/kW if balance-of-plant capital re-quirements are included. For example, the total capitalrequirement for the commercial installation of SNCR atNew England Electric’s Salem Harbor Station (threepulverized coal-fired boilers) was $15/kW.14 Similarly,total capital requirements for Public Service Electricand Gas’ Mercer Station Unit 2 and B.L. EnglandStation Unit 1 were $10.6/kW and $15/kW,respectively.15 Southern California Edison reported aneven lower capital requirement of $3/kW for installing“urea injection” on 20 units totaling 5600 MW16.

On an updated turn-key basis, a typical SNCRwould range from $1325-7000/MMBtu/hr dependingon the process category. An example of a high cost ap-plication, might apply to an all-in cost for an extremelysmall rotary kiln. On the opposite end, the lower costapplications are typically large hazardous waste incin-erators and large bubbling bed/fluidized bed boilersand large wood-fired stokers.

For similar type sources, the installed capital costper unit of output (e.g., $/kWe) decreases as the sourcesize increases, i.e., due to economy of scale, total capi-tal outlay increases less than linearly with increasingboiler capacity.

Given such low capital requirements, most of thecost of using SNCR will be operating expense. A typicalbreakdown of annual costs for utilities will be 25 % forcapital recovery and 75 % for operating expense. Forindustrial sources, annual costs will be 15-35 % forcapital recovery and 65-85 % for operating expense.For an operating expense-driven technology, little costwill be incurred if the source is not operating, and costeffectiveness (the cost per ton of NOx removed) will berelatively insensitive to capacity factor or duty cycle.This makes SNCR particularly attractive for seasonalcontrol of NOx emissions. (For capital-intensive tech-nologies, cost effectiveness becomes worse with de-creasing capacity factor.)

Demonstrated cost-effectiveness values for SNCRare low, ranging from $400 to $3,000 per ton of NOx re-moved, depending upon site-specific factors. For exam-ple, the cost effectiveness of SNCR at DominionEnergy’s Salem Harbor Station unit 2 is $670/ton.17 Thewide range exists because of differing conditions foundacross different facilities, even with in the same indus-try. For utility boilers alone, cost effectiveness varieswith factors such as uncontrolled NOx level, requiredemission reduction, unit size, capacity factor (or dutycycle), heat rate (or thermal efficiency), degree ofretrofit difficulty, and economic life of the unit.

7

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 8: white paper selective non-catalytic reduction (sncr) for controlling

Of primary interest to electric utilities is the cost ofpollution controls per unit of electricity generated, ex-pressed on a busbar basis (mills/kWh). For SNCR, thebusbar cost varies directly with the amount of NOx tobe removed. Costs range from less than 1.0 mill/kWhfor “trim reduction” on a coal-fired unit or RACT-levelreduction on an oil-fired unit, to 3.5 mills/kWh for a 75% reduction on a unit with uncontrolled emissionsgreater than 1 lb NOx/MMBtu. A commercial installa-tion of urea-based SNCR on a Dominion Energy’s unithas a busbar cost of 2.7 mills/kWh, and a cost effective-ness of approximately $1,000/ton. (To convert the bus-bar costs of SNCR to a cost increment relative to fuelprice, 0.5-3.5 mills/kWh is roughly equivalent to $0.05-$0.35/MMBtu.)

Innovations in SNCR control systems and contin-ued system optimization during operation have re-duced reagent usage at commercial installations, thusdecreasing operating costs further. At one coal-firedutility boiler, a control upgrade, including continuousammonia and temperature monitors, improved controlhardware and software, and additional injector pres-sure controls, allow over a 50 % decrease in reagentuse from baseline levels.18 At a second coal- and oil-fired unit, system optimization after start-up has low-ered reagent consumption 35 % below predictedlevels.19 Given that the reagent dominates SNCR oper-ating cost, such large reductions in reagent use trans-late to significant reductions in operating cost.

What about ammonia slip?

Ammonia slip, or emissions of ammonia whichresult from incomplete reaction of the NOx reduc-ing reagent, typically can be limited to low levels.

Ammonia slip may result in one or more prob-lems, including:

• Formation of ammonium bisulfate or other am-monium salts which can plug or corrode the airheater and other downstream components;

• Ammonia absorption on fly ash, which maymake disposal or reuse of the ash difficult;

• Formation of a white ammonium chloride plumeabove the stack; and,

• Detection of an ammonia odor around the plant.

Ammonia slip is controlled by careful injection ofreagent into regions of the furnace or other sourceswhere proper conditions (temperature, residencetime, and NOx concentration) for the SNCR reactionexist. If the reagent reacts in a region where the tem-perature is too low for the NOx-reducing reaction tooccur in the available residence time, then some un-reacted ammonia will be emitted. Further, if reagentis injected in such a way that some regions of the fur-nace are over treated, the excess reagent can lead toammonia slip. Thus, it is critical that the SNCR injec-tion system be designed to provide the appropriatereagent distribution.

The difficulty in controlling ammonia slip will varyfrom application to application. At many commercialinstallations, particularly in electric utilities, units haveoperated with ammonia slip levels of equial to or lessthan 5 ppm upstream of the air heater to meet the re-quirements of owners or permitting authorities. This isa far more stringent criterion than stack emissions. Inany case, ammonia concentrations at ground level willbe well below thresholds for both odor and toxicity.

Control system upgrades and process optimizationafter installation can lower slip below guaranteed lev-els. Thus, at a commercial SNCR system on a coal-fired boiler, improved controls have lowered ammoniaslip from 10-15 ppm to below 5 ppm, and have reducedammonia on the fly-ash by half.

Use of a down-sized SCR downstream of a SNCRalso optimizes the integration to ammonia-sensitiveunits.

Does SNCR have other limitations?

As do all pollution control technologies, SNCRhas limitations which must be understood in orderto use it properly to optimize the control of NOx

emissions.

High temperature and critical NOx concentra-tion. As temperature increases, the “critical” or equilibrium NOx concentration at a given oxygen con-centration increases. At high enough temperatures, anyreduction of NOx to below the critical level by SNCR orother means will be counteracted by the rapid oxidationof nitrogen to re-form NOx. For this reason, at suffi-ciently high temperatures and baseline NOx levels be-low the critical concentration, injection of ammonia orurea into the flue gas will result in increased NOx levels.If, however, the baseline NOx concentration is above thecritical level, NOx reduction will result. For typical coal-and oil-fired steam boilers, critical NOx levels are 70-90ppm (ca. 0.1 lb/MMBtu) in the upper furnace.

High furnace carbon monoxide concentration.High CO concentrations can shift the temperature win-dow of the SNCR process. When CO concentrations inthe region of reagent injection are above 300 ppm, thecritical NOx level and SNCR reaction rate will increaseabove what they would have been had little CO beenpresent, as if the temperature were slightly higher.Therefore, in some furnaces with high CO levels, it ispreferable to inject reagent at lower temperatures to ef-fect good NOx control.

Carbon monoxide emissions. In a well-controlledurea-based SNCR system, the carbon contained in theurea is fully oxidized to carbon dioxide. Normally,steps taken to control ammonia slip impose sufficientrestrictions on reaction temperature to prevent sub-stantial emissions of CO.

What are common misconceptions regardingSNCR?

In earlier days, several common misconceptionsinitially slowed the acceptance of SNCR by utilities.

8

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 9: white paper selective non-catalytic reduction (sncr) for controlling

Misconception: As boiler size increases, SNCRefficiency decreases. As long as reagent can be distrib-uted, there is no technical limitation to the size of boil-ers on which SNCR will be effective. Thismisconception arose in part from the earliest experi-ences at large utility boilers in California. These boilerswere equipped with low NOx combustion systems, hadhigh furnace exit gas temperatures, and very rapid cool-ing of the gases in the boiler convective regions. Lowbaseline NOx levels resulting from these natural gas-fired boilers and rapid cooling led to low NOx controlefficiencies and high ammonia slips using SNCR.Increased technical knowledge and experience have al-lowed better delineation of the limitations of the SNCRprocess, which since then has been used to achieve over60 % NOx reductions on some electric utility boilers.

The commercial development of retractable multi-nozzle lances as well as advances in feed-forward con-trols has extended the applicability of urea-basedSNCR technology. These advances enable delivery ofreagent across the boiler, as has been demonstratedboth in the U.S. and abroad. Today, there are severalfacilities utilizing SNCR on units of greater than 600MW capacity.

Misconception: SNCR cannot be used on boilersequipped with low NOx combustion controls. SNCRhas been installed commercially on boilers equippedwith low NOx burners, overfire air, and flue gas recir-culation, and has been shown to operate effectivelywith all of these technologies.20 Typically, SNCR re-duces NOx an additional 20-30 % above LNB/combus-tion modifications.

Misconception: Use of SNCR on coal-fired plantsresults in fly ash which cannot be sold and the dis-posal of which is expensive. The tendency of fly ash toabsorb ammonia is a function of many factors in addi-tion to the amount of ammonia slip. Ash characteristicssuch as pH, alkali mineral content, and volatile sulfurand chlorine content help to determine whether or notammonia will be absorbed readily by the fly ash. Inmost applications, properly designed SNCR systemswill keep the ammonia slip levels low enough so thatthe salability of the ash should be unaffected.

Can SNCR be used in combination with selec-tive catalytic reduction (SCR)?

Hybrid SNCR-SCR systems have been demon-strated at a number of utility plants, and are beingcommercially installed to meet post-RACT NOx

limits.

SNCR may be combined with selective catalytic re-duction (SCR) using a number of different techniques.NOx control with an SNCR system alone is often lim-ited by ammonia slip requirements. One commerciallyavailable hybrid SNCR-SCR system design generatesammonia slip intentionally as the reagent feed to theSCR catalyst, which provides additional NOx removal.The quantity of catalyst required in a hybrid systemcan be reduced from that of an SCR-only application,

so that the hybrid system could have lower capital requirements.

At two gas-fired utility boilers in SouthernCalifornia, hybrid systems gave emissions reductions of72-91 percent.21 At a wet bottom coal-fired boiler inNew Jersey, a hybrid system reduced NOx emissions byup to 98 percent. In a DOE Clean Coal Technology in-stallation, the combination of SNCR with smaller SCRwill reduce NOx below 0.15 lb/mmBtu at less than two-thirds the cost of full SCR.22 This hybrid approach hasbeen demonstrated in several full-scale utility applica-tions and as a result of the installation at AESGreenidge has been commercially applied. SNCR canalso be applied to units with a conventional SCR sys-tem with a standard ammonia injection grid.

How can SNCR be used to best advantage?

The features of being a low hazard, low capitalcost, expense-driven technology that requires littlespace and little unit down-time to implement sug-gests various appropriate uses to comply with U.S.clean air regulations.

Beyond-RACT Controls for Ozone Attainment.States not meeting the ozone National Ambient AirQuality Standard after application of RACT controlswill require greater NOx reductions from sourceswithin their borders. Many states presume that thesereductions will be based on the addition of post-combustion controls, including SNCR. In some cases,SNCR could be retrofit to units that already have im-plemented combustion modifications. Where SNCR hasbeen used to meet RACT limits, the reagent use ratecould be increased to meet new, lower limits.

Seasonal Controls for Ozone Attainment. In a sea-sonal approach, NOx reductions beyond RACT would berequired only during the “ozone season” (May throughSeptember) when exceedances normally occur. For ex-ample, the states of the northeast Ozone TransportRegion have committed to a plan calling for control ofozone precursors only during the May-September ozoneseason to help meet regional ozone attainment goals.SNCR is particularly well-suited for seasonal control inthat it may provide deep reductions in NOx emissions,but incurs little cost when the system is not in use. Forurea-based SNCR, the incremental cost of control dur-ing the ozone season would be on the order of$0.30/MMBtu on a unit without low-NOx burners, ex-pressed as a fuel cost adder relative to the “off” season.

Acid Rain Control. Under the acid rain provisions(Title IV) of the Clean Air Act Amendments, NOx limitsfor Group 2 coal-fired utility boilers, which include cy-clones, wet-bottom wall-fired boilers, cell-burner-firedboilers, stoker-fired units, and roof-fired boilers werepromulgated in 1996 based upon the capabilities andcosts of available control technologies.

SNCR technology has been successfully installedon cell-, pulverized-coal wet bottom-, cyclone-, andstoker-fired units as well as on circulating fluidizedbed boilers.

9

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 10: white paper selective non-catalytic reduction (sncr) for controlling

Overcontrol. The low capital cost and ease of retro-fit of SNCR suggest its use as an add-on to other NOx

control technologies to provide overcontrol, or controlto below permit limits. Overcontrol can be useful wherethe marginal cost of control on one unit is lower thanon other units, and where averaging or trading emis-sions or emissions reductions is permitted. Trading pro-visions of the proposed NOx SIP Call regulation, theRegional Clean Air Incentives Market (RECLAIM) insti-tuted by the California South Coast Air QualityManagement District, the acid rain NOx rule, and pro-posed rules for generation of emissions reduction cred-its all authorize strategies based on overcontrol.

In an overcontrol strategy, a second SNCR systemmay be used to provide insurance: If the overcontrolledunit in the averaged group is forced out of service, theinsurance system is available to provide the requisiteemissions reductions on a second unit. When the over-controlled unit is in service, the cost of the insuranceSNCR system is limited to a relatively low capitalcharge.

BACT/New Source Controls. SNCR has been uti-lized to fulfill best achievable control technology(BACT) requirements for new stoker units in Maine,Vermont, Massachusetts, Connecticut, and Virginia,among other states. In North Carolina, a new pulver-ized coal-fired unit was permitted recently with SNCRto meet a 0.17 lb/MMBtu NOx emission limit.

What are the water quality considerationswhere urea or aqueous ammonia is utilizedfor SNCR?

Water quality and product handling are importantcomponents of the overall successful operation of theemissions control system when urea or aqueous am-monia reagents are used in post-combustion applica-tions. Water quality is important in order to minimizesystem fouling and corrosion that can result in reducedSNCR system on-line time, higher maintenance costs,and the inability to meet emissions targets. Properproduct handling and storage equipment is necessaryin order to assure that the quality of the reagents havethe optimum characteristics for industrial emissionscontrol applications.

With urea-based SNCR systems, urea is generallyshipped as a 50 percent solution but is also availableand shipped from a range of concentrations from 32percent to 70 percent solution. Depending on the ureamanufacturer, the water used to ship urea may bedemineralized for quality purposes. Prior to injection,urea solution is diluted in-line anywhere from 50 per-cent to 80 percent. Water quality for this dilution step isimportant to the success of the application becauseurea (as well as ammonia) is highly alkaline in waterand will precipitate hardness and other minerals.Demineralized water will remove any potential sus-

pended solids which may lead to plugging of injectionlances and other components of the SNCR system. Thedilution water for SNCR remains stable if: (1) urea ispurchased from suppliers who supply “NOx grade urealiquor” whereby stabilizers have been mixed into thesolution, or (2) otherwise the dilution water should beof high quality which can be achieved through de-mineralization or reverse osmosis type processes in or-der to provide maximum insurance. Table 1 provides arange of physical properties for varying concentrationsof urea liquor seen in typical SNCR applications.

10

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Urea supply chain and storage is important in or-der to provide the quality assurance and quality controlof the urea liquor used for SNCR systems. The vast ma-jority of anhydrous ammonia and urea manufactured inNorth American is produced for agricultural purposeswhere water quality in the make-up/dilution water isless of an issue. For anhydrous ammonia and urea thatis produced domestically, between 85-90 percent isused for fertilizer. Agricultural applications place ahigher priority on the nitrogen value and certain physi-cal characteristics of the urea to ensure that the fertil-izer is evenly distributed when fertilizing fields. Ureaand anhydrous ammonia that is produced for SNCRgrade applications has a higher standard for the qualityof the water used for the make-up/dilution processes.Although supply is available in most locations in NorthAmerica, the actual distance between point of produc-tion and final use can add up to tens of thousands ofmiles of transport by road, rail, ship, and pipeline in-volving material handling at each step of the deliveryprocess. Some manufacturers have dedicated supplyand storage systems for SNCR grade urea and anhy-drous ammonia in order to ensure that this is no con-tamination between agricultural and industrial gradeproducts. Although not mandatory, minimizing the riskof contamination of the urea or anhydrous ammoniaduring the supply chain will ensure that the supply ofreagent meets the tight quality control requirements de-manded in the air emissions control systems.

Table 1. Range of Properties for SNCR Grade UreaLiquor from Demineralized Water

Characteristic Range

Urea Concentration 32 to 70

Free Ammonia (at loading) �0.2% to �0.5 %

Biuret (at loading) �0.3 % to �0.7 %

Magnesium (Mg) ppm �0.5 to �0.8

Calcium (Ca) ppm �0.5 to �0.8

Phosphates as PO4 ppm �0.5 to �1.5

Iron (Fe) ppm �0.5 to �0.8

Page 11: white paper selective non-catalytic reduction (sncr) for controlling

11

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 12: white paper selective non-catalytic reduction (sncr) for controlling

12

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 13: white paper selective non-catalytic reduction (sncr) for controlling

13

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 14: white paper selective non-catalytic reduction (sncr) for controlling

14

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 15: white paper selective non-catalytic reduction (sncr) for controlling

15

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 16: white paper selective non-catalytic reduction (sncr) for controlling

16

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 17: white paper selective non-catalytic reduction (sncr) for controlling

17

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 18: white paper selective non-catalytic reduction (sncr) for controlling

18

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 19: white paper selective non-catalytic reduction (sncr) for controlling

19

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 20: white paper selective non-catalytic reduction (sncr) for controlling

20

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 21: white paper selective non-catalytic reduction (sncr) for controlling

21

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 22: white paper selective non-catalytic reduction (sncr) for controlling

22

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 23: white paper selective non-catalytic reduction (sncr) for controlling

23

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 24: white paper selective non-catalytic reduction (sncr) for controlling

24

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 25: white paper selective non-catalytic reduction (sncr) for controlling

25

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 26: white paper selective non-catalytic reduction (sncr) for controlling

26

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 27: white paper selective non-catalytic reduction (sncr) for controlling

27

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 28: white paper selective non-catalytic reduction (sncr) for controlling

28

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 29: white paper selective non-catalytic reduction (sncr) for controlling

REFERENCES

1. Smith, D.J., “NOx Emission Control Demands aRange of Solutions”, Power Engineering, July,1992, page 45.

2. Sandell, M.A., and M.T. Hoydick, “Selective Non-Catalytic Reduction of NOx Control”, presented atthe 1992 Spring Gulf Coast Co-GenerationAssociation Meeting, April 21-22, 1992, Houston,Texas.

3. Sun, W.H., and J.E. Hofmann, “Reaction Kinetics ofPost Combustion NOx Reduction with Urea”, pre-sented at the AFRC 1991 Spring Members Meeting,March 18-19, 1991, Hartford, CT.

4. Hurst, B.E., White, C.M. “Thermal De-NOx: ACommercial Non-Catalytic NOx Reduction Processfor Waste to Energy Applications.” Presented at theASME 12th Biennial National Waste ProcessingConference, Denver, June 2, 1986.

5. Comparato, J.R., Buchs, R.A., Arnold, D.S. andBailey, L.K., “NOx Reduction at the Argus PlantUsing the NOxOUT Process,” EPA/EPRI 1991 JointSymposium on Stationary Combustion NOx

Control, Washington, D.C., March 25-28, 1991.6. Huhmann, A.L., Wallace, A.J., Jantzen, T., O’Leary,

J.H. “Evaluation of Retrofitted Post CombustionNOx Control Technology on a Wet Bottom, Coal-Fired Boiler.” Presented at the U.S. Department ofEnergy Conference on Selective Catalytic and Non-Catalytic Reduction for NOx Control, Pittsburgh,May 15-16, 1997.

7. Tsai, T.S., Ariagno, L., Cote, R., Staudt, J.E., Casill,R.P. “Living with Urea Selective Non-Catalytic NOx

Reduction at Montaup Electric’s 112 MWe PCBoiler.” Presented at ICAC Forum ’96, Baltimore,March 19-20, 1996.

8. Ciarlante, V., Romero, C.E. “Design andCharacterization of a Urea-Based SNCR System fora Utility Boiler.” Presented at the EPRI-DOE-EPACombined Utility Air Pollution ControlSymposium, Washington, D.C., August 25-29, 1997.

9. Ellerhorst, R.; Edvardsson, C. “Experience withNOx Control at T.B. Simon CFB Boiler at MichiganState University - Case History.” Presented at theCouncil of Industrial Boiler Owners NOx ControlVIII Conference, Philadelphia, Pennsylvania,March 7-8, 1995.

10. Malone, P.M., Sun, W.H. “Cardinal Unit 1: LargeScale SNCR Demonstration Project.” Presented atICAC Forum 2000, Rosslyn, VA, March 23-24, 2000.

11. Durso, R.A., Trippel, C.E. “Combustion Temperingin Conjunction with SNCR Reduces NOx EmissionNearly 60 % on a Natural Gas Fired CycloneBoiler.” Presented at ICAC Forum 2000, Rosslyn,VA, March 23-24, 2000.

12. Trego, P., St. Laurent, G., Broderick, R.G.,Schindler, E. “Burner Optimizations in

Conjunction with SNCR Reduced NOx EmissionsOver 70 % on Coal-Fired Boilers.” Presented atICAC Forum 2000, Rosslyn, VA, March 23-24, 2000.

13. Lövblad, R., Moberg, G., Olausson, L., Boström, C.“NOx Reduction from a Recovery Boiler byInjection of an Enhanced Urea Solution (NOxOUTProcess).” Presented at the TAPPI EnvironmentalConference, San Antonio, Texas, April 7-10, 1991.

14. Braczyk, E.J.; Sload, A.W.; Arak, L.M.; Johnson,R.A.; Albanese, V.M. “Cost-Effectiveness of NOx

Control Retrofit at Salem Harbor Station.”Presented at PowerGen ’94, Orlando, Florida,December 7-9, 1994.

15. Himes, R.; Hubbard, D.; West, Z.; Stallings, J. “ASummary of SNCR Applications to Two Coal-FiredWet Bottom Boilers.” Presented at the EPRI/EPAJoint Symposium on Stationary Combustion NOx

Control, Kansas City, Missouri, May 19, 1995.16. Utility Generation Report, Power Engineering,

August 1991.17. Braczyk, E.J.; Sload, A.W.; Arak, L.M.; Johnson,

R.A.; Albanese, V.M. “Cost-Effectiveness of NOx

Control Retrofit at Salem Harbor Station.”Presented at PowerGen ’94, Orlando, Florida,December 7-9, 1994.

18. O’Leary, J., Sun, W., Afonso, R., Sload, A. “SNCRReagent Reduction Through Innovative SystemControls at Salem Harbor Station Unit 3.” Presentedat the U.S. Department of Energy Conference onSelective Catalytic and Non-Catalytic Reduction forNOx Control, Pittsburgh, May 15-16, 1997.

19. Hofmann, J.E., von Bergmann, J., Bökenbrink, D.,and Hein, K., “NOx Control in a Brown Coal-FiredUtility Boiler.” Presented at the EPRI/EPASymposium on Stationary Combustion NOx

Control, San Francisco, CA, March 8, 1989;Comparato, J.R., Buchs, R.A., Arnold, D.S., “NOx

Reduction at the Argus Plant Using the NOxOUTProcess.” Presented at the EPRI/EPA Symposiumon Stationary Combustion NOx Control,Washington, D.C., March 1991; Sun, W.H.,Stamatakis, P., and Grimard, F.X., “NOxOUTProcess Demonstration on 325 MW Oil-FiredBoiler.” ENEL, Piombino, Italy (Nalco Fuel Tech -Unpublished).

20. Jantzen, T.M., Zammit, K.D. “Hybrid PostCombustion NOx Control.” Presented at the U.S.Department of Energy Conference on SelectiveCatalytic and Non-Catalytic Reduction for NOx

Control, Pittsburgh, May 15-16, 1997; Nylander, J.,and Krigmont, H.V., “Evaluation of a Full-ScaleHybrid NOx Control System at SDG&E’s EncinaPower Plant.” Presented at the EPRI/EPA JointSymposium on Stationary Combustion NOx

Control, Bal Harbour, FL, May 24-27, 1993. 21. Wallace, A.J., Gibbons, F.X., Roy, R.O., O’Leary,

J.H., Knell, E.W. “Demonstration of SNCR, SCR,

29

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions

Page 30: white paper selective non-catalytic reduction (sncr) for controlling

and Hybrid SNCR/SCR NOx Control Technologyon a Pulverized Coal, Wet-Bottom Utility Boiler.”Presented at ICAC Forum ’96, Baltimore, Maryland,March 19-20, 1996.

22. Roll, D.J., and W.B. Rady, “AES Greenidge Multi-Pollutant Control Project,” Air & Waste

Management Luncheon Meeting, December 13,2006, Rochester, New York, www.netl.doe.gov/technologies/coalpower/cctc/PPII/bibliography/demonstration/environmental/bib_greenidge.html

30

Selective Non-Catalytic Reduction (SNCR) for Controlling NOx Emissions