review of the north slope oil properties relevant to ... · pdf filereview of the north slope...

Review of the North Slope Oil Properties

Relevant to Environmental Assessment and Prediction

Prepared for

Prince William Sound Regional Citizens’ Advisory Council (PWSRCAC)Anchorage, Alaska

by

Merv FingasSpill Science

Edmonton, Alberta, Canada

June, 2010

4-5 Attachment A

AbstractThis paper is a summary of several oil parameters and the spill behaviour of the new

Alaskan North Slope sample as provided to Environment Canada. The essential parametersincluded in the new analysis are the oil viscosity, density, SARA analysis and distillation curve.Additional data are desirable including: PAH content, naphthalene content, C12-C18 content,and sulfur content.

Oil spill modeling and prediction are important facets of oil spill preparedness. The mostimportant estimation and modeling algorithms are that for oil spill emulsification, evaporation,chemical dispersibility and those that might be used to predict other countermeasures such asrecovery and burning. Oil spill modeling relies on algorithms that require a number of oilproperty inputs

The new sample of ANS oil was found to be atypical of North Slope oils from the past. Itwas much lighter and less viscous than any previous samples.

The environmental behaviour parameters of evaporation, emulsification and dispersibilitywere predicted. These show that they are similar but more indicative of a light oil than previousANS samples.

iii

Executive SummaryThe essential new data are:

The data show that this sample is a much lighter crude oil than any previous samples of ANS or Prudhoe Bay analysed by Environment Canada.

The use of these parameters in assessment and prediction is summarized in Table 1.

The origin and use of these predictions is illustrated in the text.

iv

List of Acronyms

ANS Alaska North Slope - This usually refers to the crude oil mixture at the end of thepipeline.

DLVO Derjaguin Landau Verway Overbeek - This refers to a theory on surfactant stabilization,with each letter referring to the author of the original theory.

EPA U.S. Environmental Protection Agency

FT-IR Fourier transform-infrared

GC Gas chromatograph - This is a chemical analytical technique.

IFT Interfacial tension

NMR Nuclear magnetic resonance

PAH Polynuclear aromatic hydrocarbons

PWSRCAC Prince William Sound Regional Citizens’ Advisory Council

RESIDUUM The heavy part of oil that is not distilled at normal temperatures, usually that above 535 Co

RSD Relative standard deviation

SARA Saturates, aromatics, resins, and asphaltenes

VOCs Volatile organic compounds

v

AcknowledgementsThe author thanks Joe Banta of the Regional Citizens’ Advisory Council of Prince

William Sound, who is the contract manager for this project. Environment Canada and BruceHollebone are acknowledged for their recent oil analysis.

vi

Table of Contents

Abstract.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiExecutive Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ivList of Acronyms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiAcknowledgements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii

1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101.2 Oil Properties and Alyeska Crude.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111.3 A Summary of Oil Composition and Behaviour.. . . . . . . . . . . . . . . . . . . . . . . . . . . 11

1.3.1 Oil Properties. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121.3.2 Behaviour of Oil. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

2. Emulsion Formation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192.1 Asphaltenes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232.2 Resins. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262.3 Waxes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

3. Methodologies for Studying Emulsions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273.1 Dielectric.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273.2 Rheology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283.3 Nuclear Magnetic Resonance (NMR).. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283.4 Interfacial Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293.5 Physical Studies - Structure and Droplet Sizes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303.6 Recent Studies on Stability Classes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

4. Emulsion Formation Modelling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324.1 Fingas 2001 Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354.2 Fingas 2004 Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354.3 Fingas 2005 Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374.3 Fingas 2009 Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

5. Dispersant Effectiveness Prediction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

6. Quick Behaviour Modeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

7. Summary.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

8. References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

vii

List of Tables1 Summary of Prediction of Important Environmental Parameters. . . . . . . . . . . . . . . . . . . V2 Summary of Physical Data on the New North Slope Sample. . . . . . . . . . . . . . . . . . . . . . . 13 Distillation Cut Points. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 PAH Analysis Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Properties of the Alaskan North Slope Oils. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Emulsion Classification for Models.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 577 Classification Scheme for Emulsions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588 Conditions for Type Calculations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

List of Figures1 The Variation in Viscosity over the Past Years. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 The Variation in Viscosity over the Past Years. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 The Variation in Distillation Data over the Past Years. . . . . . . . . . . . . . . . . . . . . . . . . . . 134 The Variation in SARA Data over the Past Years. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 The Variation in ANS Sulphur Content in Past Years.. . . . . . . . . . . . . . . . . . . . . . . . . . . 316 The Variation in Evaporation Amounts with Temperature. . . . . . . . . . . . . . . . . . . . . . . . 547 The Variation in ANS Evaporation Curves over the Years. . . . . . . . . . . . . . . . . . . . . . . . 54

viii

1. Introduction1.1 Background

The objective of this paper is to provide environmental prediction information on a newsample of North Slope Alaskan crude.

1.2 Oil Properties and Alyeska CrudeIt is important to recognize the nature of a crude oil that stems from the inputs into the

pipeline and the changing blends that occur over time. A crude oil sample drawn at one point intime from a pipeline may be completely different than a sample drawn at a later time.

The Alaska crude is an example of this principle. The Trans-Alaska Pipeline begins atpump station number one (International Petroleum Encyclopedia, 2004). At this point, it is amixture of crude oils in varying proportions from several fields. These fields include the majorfields of Prudhoe Bay and Kuparuk and several minor fields including Tarn, Alpine, Fiord,Kabulkic, Milne Point, Point Mcintyre, West Beach, Point Thompson, Endicott, and Sandpiper.The characteristics of these fields vary and thus, as they are blended into pump station number 1at the head of the Alyeska pipeline, the starting crude varies as well. Oil is withdrawn from thepipeline for the refinery at Fairbanks where residual oils are re-injected into the pipeline. Thesequence of this changes the composition of the oil when it arrives in Valdez.

In 2010, a new sample was drawn and the properties were reported by EnvironmentCanada. Tables 2 to 4 give these properties (Hollebone, 2010). Table 5 summarizes the newproperties compared to the older measured properties of similar crudes. It is noted that the newsample is a much lighter oil than those previously reported as being from the region.

1

1.3 A Summary of Oil Composition and BehaviourCrude oils are mixtures of hydrocarbon compounds ranging from smaller, volatile

compounds to very large, non-volatile compounds (Fingas, 2000). This mixture of compoundsvaries according to the geological formation of the area in which the oil is found and stronglyinfluences the properties of the oil. Petroleum products such as gasoline or diesel fuel aremixtures of fewer compounds and thus their properties are more specific and less variable.Hydrocarbon compounds are composed of hydrogen and carbon, which are therefore the mainelements in oils. Oils also contain varying amounts of sulphur, nitrogen, oxygen, and sometimesmineral salts, as well as trace metals such as nickel, vanadium, and chromium.

In general, the hydrocarbons found in oils are characterized by their structure. Thehydrocarbon structures found in oil are the saturates, olefins, aromatics, and polar compounds.The saturate group of components in oils consists primarily of alkanes, which are compounds ofhydrogen and carbon with the maximum number of hydrogen atoms around each carbon. Thus,the term 'saturate' is used because the carbons are ‘saturated’ with hydrogen. Larger saturatecompounds are often referred to as ‘waxes’. The aromatic compounds include at least onebenzene ring of six carbons. Three double carbon-to-carbon bonds float around the ring and addstability. Because of this stability, benzene rings are very persistent and can have toxic effects onthe environment.

The most common smaller and more volatile compounds found in oil are often referred toas BTEX, or benzene, toluene, ethyl-benzene, and xylenes. Polyaromatic hydrocarbons or PAHsare compounds consisting of at least two benzene rings.

Polar compounds are those that have a significant molecular charge as a result of bondingwith compounds such as sulphur, nitrogen, or oxygen. The 'polarity' or charge that the moleculecarries results in behaviour that is different from that of unpolarized compounds, under somecircumstances. In the petroleum industry, the smallest polar compounds are called 'resins', whichare largely responsible for oil adhesion. The larger polar compounds are called 'asphaltenes'because they often make up the largest percentage of the asphalt commonly used for roadconstruction. Asphaltenes often have very large molecules and, if in abundance in an oil, theyhave a significant effect on oil behaviour.

1.3.1 Oil PropertiesThe properties of oil discussed here are viscosity, density, specific gravity, solubility,

flash point, pour point, distillation fractions, interfacial tension, and vapour pressure. Viscosity is the resistance to flow in a liquid (Fingas, 2000). The lower the viscosity, the

more readily the liquid flows. For example, water has a low viscosity and flows readily, whereashoney, with a high viscosity, flows poorly. The viscosity of the oil is largely determined by theamount of lighter and heavier fractions that it contains. The greater the percentage of lightcomponents such as saturates and the lesser the amount of asphaltenes, the lower the viscosity.

As with other physical properties, viscosity is affected by temperature, with a lowertemperature giving a higher viscosity. For most oils, the viscosity varies as the logarithm of thetemperature, which is a very significant variation. Oils that flow readily at high temperatures canbecome a slow-moving, viscous mass at low temperatures. In terms of oil spill cleanup, viscositycan affect the oil's behaviour. Viscous oils do not spread rapidly, do not penetrate soil as readily,

10

and affect the ability of pumps and skimmers to handle the oil.The new ANS oil can be compared to the old ANS oils and from Figure 1 it can be seen

that the new oil is very much different from the old oil. It is much less viscous.Density is the mass (weight) of a given volume of oil and is typically expressed in grams

per cubic centimetre (g/cm ). It is the property used by the petroleum industry to define light or3

heavy crude oils. Density is also important as it indicates whether a particular oil will float orsink in water. As the density of water is 1.0 g/cm at 15 C and the density of most oils ranges3 o

from 0.7 to 0.99 g/cm , most oils will float on water. As the density of seawater is 1.03 g/cm ,3 3

even heavier oils will usually float on it. As the light fractions evaporate with time, the density ofoil increases.

Figure 1 The Variation in Viscosity over the Past Years

Occasionally, when the density of an oil becomes greater than the density of freshwater orseawater, the oil will sink. Sinking is rare, however, and happens only with a few oils, usuallyresidual oils such as Bunker C. Significant amounts of oil have sunk in only about 25 incidentsout of thousands.

Again to compare the new sample to the old data, one can examine Figure 2. This againshows that the new sample is quite different from the old samples.

Another measure of density is specific gravity, which is an oil's relative density comparedto that of water at 15EC. It is the same value as density at the same temperature. Another gravityscale is that of the American Petroleum Institute (API). The API gravity is based on the density

11

of pure water which has an arbitrarily assigned API gravity value of 10 (10 degrees). Oils witho

progressively lower specific gravities have higher API gravities.

Figure 2 The Variation in Density over the Past Years

The following is the formula for calculating API gravity: API gravity = [141.5 ÷ (densityat 15.5EC)] - 131.5. Oils with high densities have low API gravities and vice versa. In the UnitedStates, the price of a specific oil may be based on its API gravity as well as other properties of theoil.

Solubility in water is the measure of how much of an oil will dissolve in the watercolumn on a molecular basis. Solubility is important in that the soluble fractions of the oil aresometimes toxic to aquatic life, especially at higher concentrations. As the amount of oil lost tosolubility is always small, this is not as great a loss mechanism as evaporation. In fact, thesolubility of oil in water is so low (generally less than 100 parts per million) that it would be theequivalent of approximately one grain of sugar dissolving in a cup of water.

The flash point of an oil is the temperature at which the liquid gives off sufficient vapoursto ignite upon exposure to an open flame. A liquid is considered to be flammable if its flash pointis less than 60EC. There is a broad range of flash points for oils and petroleum products, many ofwhich are considered flammable, especially when fresh. Gasoline, which is flammable under allambient conditions, poses a serious hazard when spilled. Many fresh crude oils have anabundance of volatile components and may be flammable for as long as one day until the morevolatile components have evaporated. On the other hand, Bunker C and heavy crude oils are not

12

generally flammable when spilled. Flash point generally tracks many of the other data such asdensity, distillation data, etc.

The pour point of an oil is the temperature at which it takes longer than a specified timeto pour from a standard measuring vessel. As oils are made up of hundreds of compounds, someof which may still be liquid at the pour point, the pour point is not the temperature at which theoil will no longer pour. The pour point represents a consistent temperature at which an oil willpour very slowly and therefore has limited use as an indicator of the state of the oil. In fact, pourpoint has been used too much in the past to predict how oils will behave in the environment. Forexample, waxy oils can have very low pour points, but may continue to spread slowly at thattemperature and can evaporate to a significant degree. Because pour point is not the solidificationpoint of oil, it is not the best predictor of how oil will behave or even more specifically, how itwill move in the environment. Pour point is often used incorrectly as a parameter of oilbehaviour.

Distillation fractions of an oil represent the fraction of an oil (generally measured byvolume) that is boiled off at a given temperature. This data is obtained on most crude oils so thatoil companies can adjust parameters in their refineries to handle the oil. This data also providesenvironmentalists with useful insights into the chemical composition of oils. For example, while70% of gasoline will boil off at 100 C, only about 5% of a crude oil will boil off at thato

temperature and an even smaller amount of a typical Bunker C. The distillation fractionscorrelate strongly to the composition of the oil as well as to other physical properties of the oil.

Figure 3 compares the change in the fraction of oil distilled at 180 C, the point used too

develop evaporation equations. This again illustrates the large change in ANS over the years.

Figure 3 The Variation in Distillation Data over Past YearsThe oil/water interfacial tension, sometimes called surface tension, is the force of

13

attraction or repulsion between the surface molecules of oil and water. Together with viscosity,surface tension is an indication of how rapidly and to what extent an oil will spread on water. Thelower the interfacial tension with water, the greater the extent of spreading. In actual practice, theinterfacial tension must be considered along with the viscosity because it has been found thatinterfacial tension alone does not account for spreading behaviour.

The vapour pressure of an oil is a measure of how the oil partitions between the liquidand gas phases, or how much vapour is in the space above a given amount of liquid oil at a fixedtemperature. Because oils are a mixture of many compounds, the vapour pressure changes as theoil weathers. Vapour pressure is difficult to measure and is not frequently used to assess oilspills.

While there is a high correlation between the various properties of an oil, thesecorrelations should be used cautiously as oils vary so much in composition (Jokuty et al., 1995).For example, the density of many oils can be predicted based on their viscosity. For other oils,however, this could result in errors. For example, waxy oils have much higher viscosities thanwould be implied from their densities. There are several mathematical equations for predictingone property of an oil from another property, but these must be used carefully as there are manyexceptions.

The measurement of oil properties is an important consideration. While there are manystandards for measuring fuel, e.g., ASTM, many of these standards are not applicable to crudeoils and especially not to heavier crudes and residual oils. Similarly, many of the apparatuses formeasurement are only appropriate for lighter fuels. The best examples of inappropriatemeasurements are the use of Brookfield viscometers for measuring heavier crudes (McDonagh etal., 1995). These can result in errors of over 3 orders-of-magnitude (Durrell et al., 1994). Anextensive report on measurement techniques and standards is available in Wang et al. (2004).

1.3.2 Behaviour of OilOil spilled on water undergoes a series of changes in physical and chemical properties

which in combination are termed 'weathering' (Fingas, 2000). Weathering processes occur at verydifferent rates, but begin immediately after oil is spilled into the environment. Weathering ratesare not consistent throughout the duration of an oil spill and are usually highest immediately afterthe spill.

Both weathering processes and the rates at which they occur depend more on the type ofoil than on environmental conditions. Most weathering processes are highlytemperature-dependent, however, and will often slow to insignificant rates as temperaturesapproach zero degrees. The processes included in weathering are evaporation, emulsification,natural dispersion, dissolution, photooxidation, sedimentation, adhesion to materials, interactionwith mineral fines, biodegradation, and the formation of tar balls. These processes are listed inorder of importance in terms of their effect on the percentage of total mass balance, i.e., thegreatest loss from the slick in terms of percentage and what is known about the process.

Evaporation is usually the most important weathering process. It has the greatest effect onthe amount of oil remaining on water or land after a spill. Over a period of several days, a lightfuel such as gasoline evaporates completely at temperatures above freezing, whereas only a smallpercentage of a heavier Bunker C oil evaporates. The rate at which an oil evaporates depends

14

primarily on the oil's composition. The more volatile components an oil or fuel contains, thegreater the extent and rate of its evaporation. Many components of heavier oils will not evaporateat all, even over long periods of time and at high temperatures.

Oil and petroleum products evaporate in a slightly different manner than water and theprocess is much less dependent on wind speed and surface area. Oil evaporation can beconsiderably slowed down, however, by the formation of a 'crust' or 'skin' on top of the oil. Thishappens primarily on land where the oil layer does not mix with water. The skin or crust isformed when the smaller compounds in the oil are removed, leaving the larger compounds, suchas waxes and resins, at the surface. These components seal off the remainder of the oil andprevent evaporation. Stranded oil from old spills has been re-examined over many years and ithas been found that, when this crust has formed, there is no significant evaporation in the oilunderneath. When this crust has not formed, the same oil could be weathered to the hardness ofwood.

The rate of evaporation is very rapid immediately after a spill and then slowsconsiderably. About 80% of evaporation occurs in the first few days after a spill. The evaporationof most oils follows a logarithmic curve with time. Some oils such as diesel fuel, however,evaporate as the square root of time, at least for the first few days. This means that theevaporation rate slows very rapidly in both cases. The properties of an oil can changesignificantly with the extent of evaporation. If about 40% (by weight) of an oil evaporates, itsviscosity could increase by as much as a thousand-fold. Its density could rise by as much as 10%and its flash point by as much as 400%. The extent of evaporation can be the most importantfactor in determining properties of an oil at a given time after the spill and in changing thebehaviour of the oil.

Emulsification is the process by which one liquid is dispersed into another one in theform of small droplets. Water droplets can remain in an oil layer in a stable form and theresulting material is completely different. These water-in-oil emulsions are sometimes called'mousse' or 'chocolate mousse' as they resemble this dessert. In fact, both the tastier version ofchocolate mousse and butter are common examples of water-in-oil emulsions.

The mechanism of emulsion formation is not yet fully understood, but it probably startswith sea energy forcing the entry of small water droplets, about 10 to 25 ìm (or 0.010 to0.025 mm) in size, into the oil. If the oil is only slightly viscous, these small droplets will notleave the oil quickly. On the other hand, if the oil is too viscous, droplets will not enter the oil toany significant extent. Once in the oil, the droplets slowly gravitate to the bottom of the oil layer.Any asphaltenes and resins in the oil will interact with the water droplets to stabilize them.Depending on the quantity of asphaltenes and resins, as well as aromatic compounds thatstabilize asphaltenes and resins in solution, an emulsion may be formed. The conditions requiredfor emulsions of any stability to form may only be reached after a period of evaporation.Evaporation lowers the amount of low-molecular weight aromatics in the oil and increases theviscosity to the critical value.

Water can be present in oil in four ways. First, some oils contain about 1% water assoluble water. This water does not significantly change the physical or chemical properties of theoil. The second way is called 'entrainment', whereby water droplets are simply held in the oil byits viscosity to form an unstable emulsion. These are formed when water droplets are

15

incorporated into oil by the sea's wave action and there are not enough asphaltenes and resins inthe oil or if there is a high amount of aromatics in the oil which stabilizes the asphaltenes andresins, preventing them from acting on the water droplets. Unstable emulsions break down intowater and oil within minutes or a few hours, at most, once the sea energy diminishes. Theproperties and appearance of the unstable emulsion are almost the same as those of the startingoil, although the water droplets may be large enough to be seen with the naked eye.

Semi- or meso-stable emulsions represent the third way water can be present in oil. Theseare formed when the small droplets of water are stabilized to a certain extent by a combination ofthe viscosity of the oil and the interfacial action of asphaltenes and resins. For this to happen, theasphaltene or resin content of the oil must be at least 3% by weight. The viscosity of meso-stableemulsions is 20 to 80 times higher than that of the starting oil. These emulsions generally breakdown into oil and water or sometimes into water, oil, and stable emulsion within a few days.Semi- or meso-stable emulsions are viscous liquids that are reddish-brown or black in colour.

The fourth way that water exists in oil is in the form of stable emulsions. These form in away similar to meso-stable emulsions except that the oil must contain at least 8% asphaltenes.The viscosity of stable emulsions is 500 to 800 times higher than that of the starting oil and theemulsion will remain stable for weeks and even months after formation. Stable emulsions arereddish-brown in colour and appear to be nearly solid. Because of their high viscosity and nearsolidity, these emulsions do not spread and tend to remain in lumps or mats on the sea or shore.

The formation of emulsions is an important event in an oil spill. First, and mostimportantly, it substantially increases the actual volume of the spill. Emulsions of all typescontain about 70% water and thus when emulsions are formed the volume of the oil spill morethan triples. Even more significantly, the viscosity of the oil increases by as much as 1000 times,depending on the type of emulsion formed. For example, a highly viscous oil such as a motor oilcan triple in volume and become almost solid through the process of emulsification.

These increases in volume and viscosity make cleanup operations more difficult.Emulsified oil is difficult or impossible to disperse, to recover with skimmers, or to burn.Emulsions can be broken down with special chemicals in order to recover the oil with skimmersor to burn it. It is thought that emulsions break down into oil and water by further weathering,oxidation, and freeze-thaw action. Meso- or semi-stable emulsions are relatively easy to breakdown, whereas stable emulsions may take months or years to break down naturally.

Emulsion formation also changes the fate of the oil. It has been noted that when oil formsstable or meso-stable emulsions, evaporation slows considerably. Biodegradation also appears toslow down. The dissolution of soluble components from oil may also cease once emulsificationhas occurred.

The process of emulsion formation is discussed in detail in Section 2 of this report. Thelatest state-of-the-art in emulsion formation is published by Fingas and Fieldhouse (2009).

Natural dispersion occurs when fine droplets of oil are transferred into the water columnby wave action or turbulence. Small oil droplets (less than 20 ìm or 0.020 mm) are relativelystable in water and will remain so for long periods of time. Large droplets tend to rise and largerdroplets (more than 100 ìm) will not stay in the water column for more than a few seconds.Depending on oil conditions and the amount of sea energy available, natural dispersion can beinsignificant or it can remove the bulk of the oil. In 1993, the oil from a stricken ship, the Braer,

16

dispersed almost entirely as a result of high seas off Scotland at the time of the spill and thedispersible nature of the oil cargo.

Natural dispersion is dependent on both the oil properties and the amount of sea energy.Heavy oils such as Bunker C or a heavy crude will not disperse naturally to any significantextent, whereas light crudes and diesel fuel can disperse significantly if the saturate content ishigh and the asphaltene and resin contents are low. In addition, significant wave action is neededto disperse oil. In 30 years of monitoring spills on the oceans, those spills at which oil hasdispersed naturally have all occurred in very energetic seas.

The long-term fate of dispersed oil is not known, although it probably degrades to someextent as it consists primarily of saturate components. Some of the dispersed oil may also riseand form another surface slick or it may become associated with sediment and be precipitated tothe bottom.

Through the process of dissolution, some of the most soluble components of the oil arelost to the water under the slick. These include some of the lower molecular weight aromaticsand some of the polar compounds, broadly categorized as resins. As only a small amount actuallyenters the water column, usually much less than a fraction of a percent of the oil, dissolution doesnot measurably change the mass balance of the oil. The significance of dissolution is that thesoluble aromatic compounds are particularly toxic to fish and other aquatic life. If a spill of oilcontaining a large amount of soluble aromatic components occurs in shallow water and creates ahigh localized concentration of compounds, then significant numbers of aquatic organisms can bekilled.

Gasoline, diesel fuel, and light crude oils are the most likely to cause aquatic toxicity. Ahighly weathered oil is unlikely to dissolve into the water. On open water, the concentrations ofhydrocarbons in the water column are unlikely to kill aquatic organisms.

Dissolution occurs immediately after the spill occurs and the rate of dissolution decreasesrapidly after the spill as soluble substances are quickly depleted. Some of the soluble compoundsalso evaporate rapidly.

Photooxidation can change the composition of an oil. It occurs when the sun's action onan oil slick causes oxygen and carbons to combine and form new products that may be resins.The resins may be somewhat soluble and dissolve into the water or they may cause water-in-oilemulsions to form. It is not well understood how photooxidation specifically affects oils,although certain oils are susceptible to the process, while others are not. For most oils,photooxidation is not an important process in terms of changing their fate or mass balance after aspill.

Sedimentation is the process by which oil is deposited on the bottom of the sea or otherwater body. While the process itself is not well understood, certain facts about it are. Mostsedimentation noted in the past has occurred when oil droplets reached a higher density thanwater after interacting with mineral matter in the water column. This interaction sometimesoccurs on the shoreline or very close to the shore. Once oil is on the bottom, it is usually coveredby other sediment and degrades very slowly. In a few well-studied spills, a significant amount(about 10%) of the oil was sedimented on the sea floor. Such amounts can be very harmful tobiota that inevitably come in contact with the oil on the sea bottom. Because of the difficulty ofstudying this, data are limited.

17

Oil is very adhesive, especially when it is moderately weathered, and binds to shorelinematerials or other mineral material with which it comes in contact. A significant amount of oilcan be left in the environment after a spill in the form of residual amounts adhering to shorelinesand man-made structures such as piers and artificial shorelines. As this oil usually contains ahigh percentage of aromatics and asphaltenes with high molecular weight, it does not degradesignificantly and can remain in the environment for decades.

Oil slicks and oil on shorelines sometimes interact with mineral fines suspended in thewater column and the oil is thereby transferred to the water column. Particles of mineral with oilattached may be heavier than water and sink to the bottom as sediment or the oil may detach andrefloat. Oil-fines interaction does not generally play a significant role in the fate of most oil spillsin their early stages, but can have an impact on the rejuvenation of an oiled shoreline over thelong term.

A large number of microorganisms are capable of degrading petroleum hydrocarbons.Many species of bacteria, fungi, and yeasts metabolize petroleum hydrocarbons as a food energysource. Bacteria and other degrading organisms are most abundant on land in areas where therehave been petroleum seeps, although these microorganisms are found everywhere in theenvironment. As each species can utilize only a few related compounds at most, however,broad-spectrum degradation does not occur. Hydrocarbons metabolized by microorganisms aregenerally converted to an oxidized compound, which may be further degraded, may be soluble,or may accumulate in the remaining oil. The aquatic toxicity of the biodegradation products issometimes greater than that of the parent compounds.

The rate of biodegradation depends primarily on the nature of the hydrocarbons and thenon the temperature. Generally, rates of degradation tend to increase as the temperature rises.Some groupings of bacteria, however, function better at lower temperatures and others functionbetter at higher temperatures. Indigenous bacteria and other microorganisms are often the bestadapted and most effective at degrading oil as they are acclimatized to the temperatures and otherconditions of the area. Adding 'super-bugs' to the oil does not necessarily improve thedegradation rate.

The rate of biodegradation is greatest on saturates, particularly those containingapproximately 12 to 20 carbons (Haus et al., 2004). Aromatics and asphaltenes, which have ahigh molecular weight, biodegrade very slowly, if at all. This explains the durability of roofshingles containing tar and roads made of asphalt, as both tar and asphalt consist primarily ofaromatics and asphaltenes. On the other hand, diesel fuel is a highly degradable product as it islargely composed of degradable saturates. Light crudes are also degradable to a degree. Whilegasoline contains degradable components, it also contains some compounds that are toxic tosome microorganisms. These compounds generally evaporate more rapidly, but in many cases,most of the gasoline will evaporate before it can degrade. Heavy crudes contain little materialthat is readily degradable and Bunker C contains almost none.

The rate of biodegradation is also highly dependent on the availability of oxygen. Onland, oils such as diesel can degrade rapidly at the surface, but very slowly if at all only a fewcentimetres below the surface, depending on oxygen availability. In water, oxygen levels can beso low that degradation is limited. It is estimated that it would take all the dissolved oxygen inapproximately 400,000 L of sea water to completely degrade 1 L of oil. The rate of degradation

18

also depends on the availability of nutrients such as nitrogen and phosphorus, which are mostlikely to be available on shorelines or on land. Finally, the rate of biodegradation also depends onthe availability of the oil to the bacteria or microorganism. Oil degrades significantly at theoil-water interface at sea and, on land, mostly at the interface between soil and the oil.

Biodegradation can be a very slow process for some oils. It may take weeks for 50% of adiesel fuel to biodegrade under optimal conditions and years for 10% of a crude oil to biodegradeunder less optimal conditions. For this reason, biodegradation is not considered an importantweathering process in the short term.

Tar balls are agglomerations of thick oil less than about 10 cm in diameter. Largeraccumulations of the same material ranging from about 10 cm to 1 m in diameter are called tarmats. Tar mats are pancake-shaped, rather than round. Their formation is still not completelyunderstood, but it is known that they are formed from the residuals of heavy crudes and BunkerC. After these oils weather at sea and slicks are broken up, the residuals remain in tar balls or tarmats. The reformation of droplets into tar balls and tar mats has also been observed, with thebinding force being simply adhesion.

The formation of tar balls is the ultimate fate of many oils. These tar balls are thendeposited on shorelines around the world. The oil may come from spills, but it is also residual oilfrom natural oil seeps or from deliberate operational releases such as from ships. Tar balls areregularly recovered by machine or by hand from recreational beaches.

2. Emulsion FormationEmulsification is the process whereby water-in-oil emulsions are formed. These

emulsions are often called “chocolate mousse” or “mousse” by oil spill workers. Whenemulsions form, the properties and characteristics of oil spills change to a very large degree. Forexample, stable emulsions contain from 60 to 80% water, thus expanding the spilled materialfrom 2 to 5 times the original volume. The density of the resulting emulsion can be as great as1.03 g/mL compared to a starting density as low as 0.80 g/mL. Most significantly, the viscosityof the oil typically changes from a few hundred mPa.s to about 100,000 mPa.s, a typical increaseof 1,000. A liquid product is thereby changed to a heavy, semi-solid material.

Many researchers feel that emulsification is the second most important behaviouralcharacteristic of oil after evaporation. Emulsification has a significant effect on the behaviour ofoil spills at sea. As a result of emulsification, evaporation of oil spills slows by orders-of-magnitude, spreading slows by similar rates, and the oil rides lower in the water column,showing different drag with respect to the wind. Emulsification also significantly affects otheraspects of a spill, such as cleanup response. Spill countermeasures are quite different foremulsions as they are hard to recover mechanically, to treat, or to burn.

In terms of understanding emulsions and emulsification, the oil spill industry has not keptpace with the petroleum production industry and colloid science generally. Workers in the spillindustry often revert to old papers published in oil spill literature, which are frequently incorrectand outdated. A basic understanding of the formation, stability, and processes of emulsions isnow evident in literature in both the colloid science and oil spill fields, although some newpapers still appear with references only to 20-year-old literature.

The availability of methodologies to study emulsions is very important. In the past ten

19

years, both dielectric methods and rheological methods have been exploited to study formationmechanisms and stability of emulsions made from many different types of oils (Sjöblom et al.,1994; Fingas et al., 1998). Standard chemical techniques, including Nuclear Magnetic Resonance(NMR), chemical analysis techniques, microscopy, interfacial pressure, and interfacial tension,are also being applied to emulsions. These techniques have largely confirmed findings noted inthe dielectric and rheological mechanisms.

The mechanism and dynamics of emulsification were poorly understood until the 1990s.It was not recognized until recently that the basics of water-in-oil emulsification were understoodin the surfactant industry, but not in the oil spill industry. In the late 60s, Berridge and coworkerswere the first to describe emulsification in detail and measured several physical properties ofemulsions (Berridge et al., 1968). Berridge described the emulsions as forming because of theasphaltene and resin content of the oil. Workers in the 1970s concluded that emulsificationoccurred primarily due to increased turbulence or mixing energy (Haegh and Ellingson, 1977;Wang and Huang, 1979). The oil’s composition was not felt to be a major factor in emulsionformation. Some workers speculated that particulate matter in the oil may be a factor and otherssuggested it was viscosity. Evidence could be found for and against all these hypotheses.

Twardus (1980) studied emulsions and found that emulsion formation might be correlatedwith oil composition. It was suggested that asphaltenes and metal porphyrins contributed toemulsion stability. Bridie and coworkers (1980) studied emulsions in the same year and proposedthat the asphaltenes and waxes in the oil stabilized water-in-oil emulsions. The wax andasphaltene content of two test oils correlated with the formation of emulsions in a laboratory test.Mackay and coworkers hypothesized that emulsion stability was due to the formation of a film inoil that resisted water droplet coalescence (Mackay and Zagorski, 1981, 1982a, 1982b). Thenature of these thin films was not described, but it was proposed that they were caused by theaccumulation of certain types of compounds. Later work led to the conclusion that thesecompounds were asphaltenes and waxes. A standard procedure was devised for makingemulsions and measuring stability. This work formed the basis of much of the emulsionformation theory in the oil spill literature over the past two decades.

Scientists reported that asphaltenes were a major factor in water-in-oil emulsions morethan 40 years ago (Berridge et al., 1968). The specific roles of asphaltenes in emulsions have notbeen defined until recently. The basics of water-in-oil emulsification are now understood to amuch better degree (McLean and et al., 1998; Sjöblom et al., 2003, Sjoblom et al., 2007, Fingasand Fieldhouse, 2009). The basic principle is that water-in-oil emulsions are stabilized by theformation of high-strength visco-elastic asphaltene films around water droplets in oil. Resins alsoform emulsions, but resins do not form stable emulsions, and actually aid in asphaltene emulsionstability by acting as asphaltene solvents and by providing temporary stability during the time ofthe slow asphaltene migration. In all, a wide spectrum of researchers have found out that oilcomposition is a key factor to water-in-oil emulsion formation and that several key factorsinclude the asphaltene, resin, and saturate contents as well as the density and viscosity of thestarting oil.

Recently, Fingas and Fieldhouse (2009) published a major paper on emulsion formationas it relates to spilled oil. They found that four clearly-defined water-in-oil types are formed byoil and petroleum products when mixed with water. This was shown by water resolution over

20

time, by a number of rheological measurements, and by the water-in-oil product’s visualappearance, both on the day of formation and one week later. Some emulsions were observed forover a year, with the same results. The types are named stable water-in-oil emulsions, mesostablewater-in-oil emulsions, entrained, and unstable water-in-oil emulsions. The differences amongthe four types are quite large and are based on at least two water content measurements and fiverheological measurements. More than 300 oils or petroleum products were studied.

Fingas and Fieldhouse (2009) carried out tests of several indices of stability, a singlevalue that could provide good discrimination between water-in-oil types even on day one. It wasfound that all of these indices could be used to some extent to differentiate the four water-in-oiltypes. One index of stability was the ratio of the complex modulus of the product on the first daydivided by the starting oil viscosity. This index was named stability A. Stability B was the elasticmodulus on the first day divided by the starting oil viscosity. Stability C was combination ofthese two indices, the log of the cross product of stability A and B. Another index, the viscosityincrease, the ratio of the oil viscosity after mixing divided by the starting oil viscosity was alsoevaluated. Another test, first day water times stability A was a simple composite of the first daywater content times stability A. The last two parameters evaluated were the tan delta on the dayof formation and the change in tan delta over one week. The index called stability C, proved tobe the index that best differentiated all four water-in-oil types.

Stable emulsions are reddish-brown solid-like materials with an average water content ofabout 80% on the day of formation and about the same one week later. Stable emulsions remainstable for at least 4 weeks under laboratory conditions. All of the stable emulsions studiedremained so for at least one year. The viscosity increase over the day of formation averages afactor of 400 and one week later averages 850. The average properties of the starting oil requiredto form a stable emulsion are: density 0.9 g/mL; viscosity 300 mPa.s; resin content 9%;asphaltene content 5 %; and asphaltene-to-resin ratio 0.6.

Meso-stable water-in-oil emulsions are reddish-brown viscous liquids with an averagewater content of 64% on the first day of formation and less than 30% one week later. Meso-stableemulsions generally break down fairly completely within one week. The viscosity increase overthe day of formation averages a factor of 7 and one week later averages 5. The average propertiesof the starting oil required to form a meso-stable emulsion are: density 0.9 g/mL; viscosity1300 mPa.s; resin content 16%; and asphaltene content 8%; asphaltene-to-resin ratio 0.5. Thegreatest difference between the starting oils for stable and mesostable emulsions are the ratio ofviscosity increase (stable 400, first day and 850 in one week; mesostable 7, first day and 5 in oneweek) and resin content (stable - 9%; mesostable - 16%).

Entrained water-in-oil types are black viscous liquids with an average water content of45% on the first day of formation and less than 28% one week later. The viscosity increase overthe day of formation averages a factor of two and one week later still averages two. The averageproperties of the starting oil required to form entrained water are: density 0.97 g/mL; viscosity60,000 mPa.s; resin content 18%; asphaltene content 12%; and asphaltene-to-resin ratio 0.75.The greatest differences between the starting oils for entrained water-in-oil compared to stableand mesostable emulsions are the viscosity of the starting oil (entrained starting oil averages60,000 mPa.s compared to 200 mPa.s for stable emulsions and 1300 mPa.s for mesostableemulsions and the ratio of viscosity increase (entrained = 2, first day and 2 in one week; stable

21

400, first day and 850 in one week; mesostable 7, first day and 5 in one week). Entrained water-in-oil types appear to be applicable to viscous oils and petroleum products, but not extremelyviscous products.

Unstable water-in-oil emulsions are characterized by the fact that the oil does not holdsignificant amounts of water. There is a much broader range of properties of the starting oil thanfor the other three water-in-oil states. For example, viscosities are very low or very high.Included in this group are light fuels such as diesel fuel and very heavy, viscous oil products.

The stability of an emulsion, as represented by stability C, was correlated with starting oilproperties. The different water-in-oil types could be differentiated by correlating individuallywith starting oil viscosity, density, saturate content, asphaltene content, resin content orasphaltene/resin ratio. Combinations of these oil properties yielded even greater differentiationbetween water-in-oil types. Examples of good separation of water-in-oil types was achieved byplotting asphaltene content times asphaltene/resin ratio, along with starting oil viscosity andstability. Other correlations with similar input also yielded differentiation between water-in-oiltypes.

The data suggest that the water-in-oil types are stabilized by both asphaltenes and resins,but that, for greater stability, resin content should exceed the asphaltene content slightly. Excessresin content (A/R > about 0.6) apparently destabilizes the emulsion. This does not consider thequestion of different types of asphaltenes or resins. A high asphaltene content (typically > 10%)increases the viscosity of the oil such that a stable emulsion will not form. Viscous oils will onlyuptake water as entrained water and will slowly lose much of this water over a period of aboutone week. Viscous oils (typically >1000 mPa.s) will not form stable or meso-stable emulsions.Oils or low viscosity or without significant amounts of asphaltenes and resins will not form anywater-in-oil type and will retain less than about 6% water. Oils of very high viscosity (typically >10,000 mPa.s) will also not form any of these water-in-oil types and thus are classified asunstable. This is probably due to the inability of water droplets to penetrate the oil mass.

Several new models for the prediction of water-in-oil emulsions were recently developed(Fingas and Fieldhouse, 2004a, b; 2005; 2006). These models were the first available to calculatethe formation of emulsions using a continuous function and employing the physical and chemicalproperties of oil. Since this initial model was developed, the emulsification properties of moreoils were measured and the properties of some of the oils in the existing set oils have been re-measured. Some of these data were found to be deficient and were eliminated. This enables themodels to be re-calculated with sound data on over 316 oils.

The basis of these models is the result of the knowledge demonstrated above - namelythat models are stabilized by asphaltenes, with the participation of resins. Other components ofoil, especially aromatics may actually interfere with this process. Findings of this group and othergroups show that the entire SARA (Saturates, Aromatic, Resins, and Asphaltene) componentaffects the formation of emulsions as the prime stabilizers, asphaltenes and secondarily resins,are only available for emulsion formation when the concentration of the saturates and aromaticsare at a certain level and when the density and viscosity are correct (Fingas and Fieldhouse,2009). Figure 4 shows the change in SARA content of the ANS oils over the past 20 years. It canbe noted that the oil contains many more saturates than in previous years.

22

Figure 4 The Variation in SARA Content over Past Years

The very early emulsion formation modelling equations at that time did not use specificknowledge of emulsion formation processes (Fingas and Fieldhouse, 2004a). In the past, the rateof emulsion formation was assumed to be first-order with time. This could be approximated witha logarithmic (or exponential) curve. Although not consistent with the knowledge of howemulsions formed, this assumption has been used extensively in oil spill models. The processesoutlined above were not discovered until many years after the simple process equations were putforward. Furthermore, the presence of different water-in-oil states dictates that one simpleequation is not adequate to predict emulsion formation.

The very early emulsion formation modelling equations at that time did not use specificknowledge of emulsion formation processes (Fingas and Fieldhouse, 2004a). In the past, the rateof emulsion formation was assumed to be first-order with time. This could be approximated witha logarithmic (or exponential) curve. Although not consistent with the knowledge of howemulsions formed, this assumption has been used extensively in oil spill models. The processesoutlined above were not discovered until many years after the simple process equations were putforward. Furthermore, the presence of different water-in-oil states dictates that one simpleequation is not adequate to predict emulsion formation.

23

2.1 AsphaltenesMore than 30 years ago, it was found that asphaltenes were a major factor in emulsion

stability (Berridge et al., 1968). Specific roles of emulsions have not been defined until recently.The Sjöblom group in Norway defined the interfacial properties of asphaltenes in several localoffshore crudes (Nordli et al., 1991). Asphaltenes were separated from the oils using consecutiveseparations involving absorption to silica. Molecular weights ranged from 950 to 1,450 Daltons.Elemental analysis revealed that 99 mole % of the asphaltenes was carbon and hydrogen, whileup to 1% was nitrogen, oxygen, and/or sulphur. The films form monomolecular layers at theair/water interface. Aromatic solvents such as benzylalcohol strongly influence the asphaltenesand will destabilize water-in-oil emulsions. Asphaltenes were shown to be the agent responsiblefor stabilizing the Norwegian crudes tested.

Workers in the same group separated resins and asphaltenes and studied the Fourier-transform infrared spectrum and the emulsions formed by each fraction (Mingyuan et al., 1992).The asphaltenes were separated using pentane precipitations and the resins by desorption fromsilica gel using mixtures of benzene and methanol. The fractions were tested in model systemsfor their emulsion-forming tendencies. Model emulsions were stabilized by both asphaltene andresin fractions, but the asphaltene fractions were much more stable.

Acevedo and coworkers studied the interfacial behaviour of a Cerro Negro crude by aplanar rheology (Acevedo et al., 1993). Distilled water and salt water were used with a 30% anda 3.2% xylene-diluted crude. The elasticity and viscosity were obtained from creep compliancemeasurements. The high values of viscoelastic and elastic moduli were attributed to theflocculation of asphaltene:resin micelles at the interface. The high moduli were associated withthe elastic interface. In the absence of resins, asphaltenes were not dispersed and did not formstable interface layers and then, by implication, stable emulsions.

Mohammed and coworkers studied surface pressure, as measured in a Langmuir filmbalance, of crude oils and solutions of asphaltenes and resins (Mohammed et al., 1993a, 1993b).They found that the pseudo-dilational modulus has high values for low resin-to-asphaltene ratiosand low values for high resin-to-asphaltene ratios. They suggest that low resin-to-asphalteneratios lead to more stable emulsions and vice-versa.

Chaala and coworkers studied the flocculation and the colloidal stability of crudefractions (Chaala et al., 1994). Stability was defined as the differential in spectral absorptionbetween the bottom and top of a test vessel. The effects of temperature and additions of waxesand aromatics on stability were noted. Increasing both waxes and aromatics generally decreasedstability. Temperature increased stability up to 60 C and then stability decreased.o

In another study, the resins and asphaltenes were extracted from four crude oils byvarious means (Schildberg et al., 1995). It was found that different extraction methods resulted indifferent characteristics as measured by FT-IR spectroscopy as well as different stabilities whenthe asphaltenes and resins were used as stabilizers in model systems. It was concluded that theinterfacially active components in crude oil were interacting and were difficult to distinguish.Both the resins and asphaltenes appeared to be involved in interfacial processes.

Urdahl and Sjöblom (1995) studied stabilization and destabilization of water-in-crude oilemulsion. It was concluded that indigenous, interfacially active components in the crude oils areresponsible for stabilization. These fractions would be the asphaltenes and resins. Model systems

24

stabilized by extracted interfacially active components had stability properties similar to thecrude oil emulsions. The same group studied the ageing of the interfacial components (Sjöblomet al., 1995). Resins and asphaltenes were extracted from North Sea crudes and exposed toageing under normal atmospheric and ultraviolet conditions. The FT-IR spectra show that thecarbonyl peak grew significantly as indicated by the C=C mode. Spectra also showed thatcondensation was occurring. The interfacial activity increased in all fractions as the ageingprocess proceeded. In the case of two crude oils, the ageing was accompanied by an increase inthe water/oil emulsion stability.

McLean and Kilpatrick (1997a) studied asphaltene aggregation in model emulsions madefrom heptane and toluene. The resins and asphaltenes were extracted from four different crudeoils - two from Saudia Arabia, Alaskan North Slope, and San Joaquin Valley crudes. Theasphaltenes were extracted using heptane and the resins using open-column silica columns.Asphaltenes dissolved in heptol consisting of only about 0.5% asphaltenes generated more stableemulsions than those generated by the originating crude oils. Although some emulsions could begenerated using resins, they were much less stable than those generated by asphaltenes. Themodel emulsions showed that the aromaticity of the crude medium was a prime factor. This wasadjusted by varying the heptane:toluene ratio. It was also found that the concentration ofasphaltenes and the availability of solvating resins were important. The model emulsions weremost stable when the crude medium was from 30 to 40% toluene and with low resin:asphalteneratios.

McLean and Kilpatrick (1997b) put forward the thesis that asphaltenes were the mosteffective in stabilizing emulsions when they are near the point of incipient precipitation. It wasnoted that there are specific resin-asphaltene interactions, as differing combinations yieldeddifferent results in the model emulsions. The resins and asphaltenes were characterized byelemental and neutron activation analyses. The most effective emulsion stabilizers of the resinsand asphaltenes were the most polar and the most condensed. McLean and Kilpatrick concludedthat the solubility state of asphaltenes is the most significant factor in emulsion formation.

In 1998, Mouraille and coworkers studied the stability of emulsions usingseparation/sedimentation tests and high voltage destabilization. It was found that the mostimportant factor was the stabilization state of the asphaltenes. The wax content did not appear toaffect the stability except that a high wax content displayed a high temperature dependence.Resins affected the solubilization of the asphaltenes and thus indirectly, the stability.

In the same year, McLean and coworkers reviewed emulsions and concluded that theasphaltene content is the single most important factor in the formation of emulsions (McLean etal., (1998). Even in the absence of any other synergistic compounds, i.e., resins, waxes, andaromatics, asphaltenes were found to be capable of forming rigid, cross-linked, elastic filmswhich are the primary agents in stabilizing water-in-crude oil emulsions. It was noted that theexact conformations in which asphaltenes organize at oil-water interfaces and the correspondingintermolecular interactions have not been elucidated. McLean and colleagues suggest that theintermolecular interactions must be either ð-bonds between fused aromatic sheets, H-bondsmediated by carboxyl, pyrrolic, and sulfoxide functional groups, or electron donor-acceptorinteractions mediated by porphyrin rings, heavy metals, or heteroatomic functional groups.

It is suggested that specific experimental designs to test these concepts are needed to

25

understand the phenomenon on a molecular level. Such knowledge would aid in the design ofchemical demulsifiers. The oleic medium plays an important role in the surface activity ofasphaltenic aggregates and in the resulting emulsion stability. It is noted that the precise role ofwaxes and inorganic solids in either stabilizing or destabilizing emulsions is not known.Emulsions are stabilized primarily by rigid, elastic asphaltenic films.

Recently, Singh and coworkers studied the effect of the fused-ring solvents includingnaphthalene, phenanthrene, and phenanthridine in destabilizing emulsions (Singh et al., 1999).They note that the primary mechanism for emulsion formation is the stability of asphaltene filmsat the oil-water interface. They suggest that the mechanism is one in which planar, disk-likeasphaltene molecules aggregate through lateral intermolecular forces to form aggregates. Theaggregates form a viscoelastic network after absorption at the oil-water interface. The network issometimes called a film or skin and the strength of this film correlates with emulsion stability.The strength of the film can be gauged by shear and elastic moduli. Singh and coworkers probedthe film-bonding interactions by studying the destabilization by aromatic solvents (Singh et al.,1999). It was found that fused-ring solvents, in particular, were effective in destabilizingasphaltene-stabilized emulsions. It is suggested that both ð-bonds between fused aromatic sheetsand H-bonds play significant roles in the formation of the asphaltene films.

Sjöblom and coworkers (1999) used dielectric spectroscopy to study emulsions over aperiod of years. It is concluded that the asphaltenes, not the resins, are the stabilizing fraction inwater-in-oil emulsions. It was noted, however, that some resins must be present to give rise tostability. It is suggested that the greater mobility of the resins is needed to stabilize the emulsionsuntil the asphaltenes, which migrate slowly, can align at the interface and stabilize the emulsions.

2.2 ResinsNeuman and Paczynska-Lahme (1996) studied the stability of petroleum o/w emulsions

and found that they are stabilized by ‘thick films’ which appeared to be largely composed ofpetroleum resins. These thick films demonstrate elasticity and thus increase stability.Temperature increases showed increasing structure formation of the films. Isolated petroleumresins showed structure formation as well.

Rønningsen and coworkers studied the ageing of crude oils and its effect on the stabilityof emulsions (Rønningsen et al., 1995). The oil was exposed to air and light and it was found thatthe interfacial tension of the oil toward formation water decreased as a result of the ageing. Thiswas caused by the formation of various oxidation products, mainly carbonyl compounds. Ingeneral, the emulsions became more stable. In some cases, however, the opposite was observed,namely, that although the interfacial tension was high, the emulsion stability was lower,presumably because new compounds with less beneficial film properties are formed. Presumably,the compounds that were formed could be loosely classified as resins.

2.3 WaxesJohansen and coworkers studied water-in-crude oil emulsions from the Norwegian

continental shelf (Johansen et al., 1988/89). The crudes contained a varying amount of waxes (2to 15%) and few asphaltenes (0 to 1.5% by weight). Emulsion stability was characterized byvisual inspection as well as by ultracentrifugation at 650 to 30,000 g. Mean water droplet sizes of

26

10 to 100 ìm were measured in the emulsions. It was found that higher mixing rates reduced thedroplet size and a longer mixing time yielded a narrow distribution of droplet size. The emulsionstability correlated with the emulsion viscosity, the crude oil viscosity, and the wax content.

McMahon (1992) studied the effect of waxes on emulsion stability as monitored by theseparation of water over time. The size of the wax crystals showed an effect in some emulsionsbut not in others. Interfacial viscosity indicated that the wax crystals form a barrier at thewater/oil interface that retards the coalescence of colliding water droplets. Studies withoctacosane, a model crude oil wax, show that a limited wax/asphaltene/resin interaction occurs.A wax layer, even with absorbed asphaltenes and resins, does not by itself stabilize an emulsion.McMahon concludes that the effect of wax on emulsion stability does not appear to be throughaction at the interface. Instead, the wax may act in the bulk oil phase by inhibiting film thinningbetween approaching droplets or by a scavenging demulsifier. It was found that the asphaltenesand resins affected stability via interfacial action and are primarily responsible for the emulsionformation.

Puskas and coworkers extracted paraffinic deposits from oil wells and pipelines. Thishydrophobic paraffin derivative had a high molar mass and melting point and contained polar endgroups (carbonyls) (Puskas et al., 1996). This paraffinic derivative stabilized water-in-oilemulsions at concentrations of 1 to 2%.

3. Methodologies for Studying Emulsions3.1 Dielectric

Dielectric spectroscopy is one of the methods used to study emulsions. The permittivityof the emulsion can be used to characterize an emulsion and assign a stability (Sjöblom et al.,1994, 1997, 1999; Sjöblom and Førdedal, 1996; Førdedal et al., 1996a; Gestblom et al., 1994;Skodvin et al., 1994a, b; Skodvin and Sjöblom, 1996).

The Sjöblom group has measured the dielectric spectra using the time domainspectroscopy (TDS) technique. A sample is placed at the end of a coaxial line to measure totalreflection. Reflected pulses are observed in time windows of 20 ns, Fourier-transformed in thefrequency range from 50 MHz to 2 Ghz, and the complex permittivity calculated. Water or aircan be used as reference samples. The total complex permittivity at a frequency (ù) is given by:

swhere: å is the static permittivity,

öå is the permittivity at high frequencies,ù is the angular frequency, andô is the relaxation time.

The measuring system used by the Sjöblom group includes a digital samplingoscilloscope and a pulse generator. A computer is connected to the oscilloscope and controls themeasurement timing as well as performing the calculations.

The data are used to indicate stability and the geometry of the droplets. Flocculation ofthe emulsion can be detected. In tests of flowing and static emulsions, it was shown that the

27

flowing emulsions have lower static permittivities, which was interpreted as indicatingflocculation in the static emulsions (Skodvin et al., 1994a).

Skodvin and Sjöblom (1996) used dielectric spectroscopy in conjunction with rheology tostudy a series of emulsions. A close connection was found between the viscosity and dielectricproperties of the emulsions. The large effects of shear on both the static permittivity and thedielectric relaxation time for the emulsion was ascribed, at least in part, to the degree offlocculation in the emulsion system. At high shear rates, at which emulsions are expected to havea low degree of flocculation and high stability, the dielectric properties still varied from thoseexpected from a theoretical model for spherical emulsion droplets.

Førdedal and coworkers used dielectric spectroscopy to study several real crude oilemulsions and model systems stabilized with either separated asphaltenes and resins from crudeoil or by commercial surfactants (Førdedal et al., 1996a,b). Emulsions could be stabilized by theasphaltene fraction alone, but not by the resin fraction alone. A study of a combination ofmixtures shows an important interaction between emulsifying components. Førdedal andcoworkers used dielectric spectroscopy to study model emulsions stabilized by asphaltenesextracted from crude oils. Analysis showed that the choice of organic solvent and the amount ofasphaltenes, as well as the interaction between these variables, were the most significantparameters for determining the stability of the emulsions.

Ese and coworkers found similar results on model emulsions stabilized with resins andasphaltenes extracted from North Sea oil (Ese et al., 1997). The dielectric spectroscopy resultsshowed that the stability of model emulsions could be characterized. Stability was found todepend mainly on the amount of asphaltenes, the degree of aging of asphaltenes and resins, andthe ratio between asphaltenes and resins.

3.2 RheologyIn 1983, Steinborn and Flock studied the rheology of crude oils and water-in-oil

emulsions (Steinborn and Flock, 1983). Emulsions with high proportions of water exhibitedpseudo-plastic behaviour and were only slightly time-dependent at higher shear rates. Omar andcoworkers also measured the rheological characteristics of Saudi crude oil emulsions (Omar etal., 1991). Non-Newtonian emulsions exhibit pseudo-plastic behaviour and followed a power-law model. Mohammed and coworkers studied crude oil emulsions using a biconical bobrheometer suspended at the interface (Mohammed et al., 1993b). More stable emulsionsdisplayed a viscoelastic behaviour and a solid-like interface. Demulsifiers changed the solid-likeinterface into a liquid one.

Tadros (1994) summarized the fundamental principles of emulsion rheology. Emulsionsstabilized by surfactant films (such as resins and asphaltenes) behave like hard spheredispersions. These dispersions display viscoelastic behaviour. Water-in-oil emulsions show atransition from a predominantly viscous to a predominantly elastic response as the frequency ofoscillation exceeds a critical value. Thus, a relaxation time can be determined for the system,which increases with the volume fraction of the discontinuous phase. At the critical value, thesystem shows a transition from predominantly viscous to predominantly elastic response. Thisreflects the increasing steric interaction with increases in volume of the discontinuous phase.

In 1996, Pal studied the effect of droplet size and found it had a dramatic influence on

28

emulsion rheology (Pal, 1996). Fine emulsions have much higher viscosity and storage modulithan the corresponding coarse emulsions. The shear thinning effect is much stronger in the caseof fine emulsions. Water-in-oil emulsions age much more rapidly than oil-in-water emulsions.More recently, Lee and coworkers and Aomari and coworkers examined model emulsions andfound that a maximum shear strain existed that occurred around 100 s (Lee et al., 1997; Aomari-1

et al., 1998).

3.3 Nuclear Magnetic Resonance (NMR)Urdahl and coworkers studied crude oils and the silica-absorbed compounds (asphaltenes

and resins) using C Nuclear Magnetic Resonance (NMR) techniques (Urdahl et al., 1992). It13

was found that the asphaltenes and resins are enriched in condensed aromatics compared to thewhole crude oils. There was strong indications of a long, straight-chain aliphatic compoundcontaining a heteroatom substituent which is abundant in paraffinic oils. There was also reason tobelieve that this compound was active in the formation of stable water-in-crude oil emulsions.The range from 130 to 210 ppm in the C NMR was of particular interest as this region13

represents quaternary aromatic and heteroatom-bonded carbons.Balinov and coworkers studied the use of C NMR to characterize emulsions using the13

NMR self-diffusion technique (Balinov et al., 1994). The technique uses the phase differences inconsecutive pulses to measure the diffusion length of the target molecules. As such, thetechnique indicates the relative particle size in an emulsion. The NMR technique was comparedto optical microscopy and showed good correlation over several experiments involving ageingand breaking of the emulsions.

LaTorraca and coworkers used proton NMR to study oils and emulsions (LaTorraca,1998). They found that the amount of hydrogen was inversely proportional to the viscosity. Theamount of water could be determined in an emulsion because of the separation downfield of theproton on water and on hydrocarbons. The viscosity and water content of emulsions could besimultaneously determined.

3.4 Interfacial StudiesSjöblom and coworkers studied model emulsions stabilized by interfacially active

fractions from crude oil (Sjöblom et al., 1992a). A good correlation was found betweeninterfacial pressure of the fractions, as measured in a Langmuir trough, and the stability ofemulsions as measured by the amount of water separated with time. Surface tension, as measuredby the drop volume method, did not show a surfactant-like behaviour for the asphaltenes andresins.

Børve and coworkers studied the pressure-area isotherms and relaxation behaviour in aLangmuir trough (Børve et al., 1992). In one study, model polymers, styrene, and allyl alcohol(PSAA, molecular weight 150 g mol ) and mixtures of PSAA with 4-hexadecylaniline or-1

eicosylamine were used as comparative polymers to the natural surfactants in oils. The mixturesof PSAA with the amines reproduce the ð-A isotherms and relation properties shown by theinterfacially active fractions of crude oils. The main difference was found to be a lowermaximum compressibility and a higher surface activity. The crude oil fractions are well modelledby a relatively low-molecular weight aromatic, weakly polar, water-insoluble hydrocarbon

29

polymer to which long-chain amines have been added.In a similar study, Ebeltoft and coworkers mixed surfactants (sodium dodecylsulphate,

and cetyltrimethlyammonium bromide or cetylpyridinium chloride) with PSAA and studied thepressure-area isotherms (Ebeltoft et al., 1992). All the surfactants appeared to interact with thePSAA and in similar ways. It was concluded that PSAA monolayers are good models foremulsion-stabilizing monolayers in Norwegian crude oils. Monolayers of both PSA and crude oilinterfacially active fractions responded similarly to the presence of ionic surfactants, indicatinganalogous dissolution mechanisms.

Hartland and Jeelani (1994) performed a theoretical study of the effect of interfacialtension gradients on emulsion stability. Dispersion stability and instability were explained interms of a surface mobility that is proportional to the surface velocity. When the interfacialtension gradient is negative, the surface mobility is negative under some conditions, whichgreatly reduces the drainage so that the dispersion is stable. This is a normal situation assurfactant is present at the interface. Demulsifier molecules penetrate the interface within thefilm, thereby lowering the interfacial tension sufficiently and causing a positive interfacialtension gradient. Drainage is subsequently increased and the emulsion becomes unstable.

Ese and coworkers studied the film-forming properties of asphaltenes and resins using aLangmuir trough (Ese et al., 1998). Asphaltenes and resins were separated from different crudeoils. It was found that the asphaltenes appear to pack closer at the water surface and form a morerigid surface than the resins. The size of asphaltene aggregates appears to increase when thespreading solvent becomes more aliphatic and with increasing asphaltene bulk concentration.Resin films show high compressibility, which indicates a collapse of the monomolecular film. Acomparison between asphaltenes and resins shows that resins are more polar and do notaggregate to the same extent as the asphaltenes. Resins also show a high degree of sensitivity tooxidation. When resins and asphaltenes are mixed, resins begin to dominate the film propertieswhen resins exceed about 40 wt %.

3.5 Physical Studies - Structure and Droplet SizesEley and coworkers studied the size of water droplets in emulsions using optical

microscopy and found that the droplet sizes followed a log-normal distribution (Eley et al.,1988). The number mean diameters of the droplets varied from about 1.4 to 5.6 ìm. Paczynska-Lahme (1990) studied several mesophases in petroleum using optical microscopy. Petroleumresins showed highly organized laminar structures and water-in-oil emulsions were generallyunstructured but sometimes hexagonal.

Toulhoat and coworkers studied asphaltenes extracted from crude oils of various originsusing atomic-force microscopy (Toulhoat et al., 1994). The asphaltenes were dried onto freshlycleaved mica and in some cases were present in discoids of approximately 2 nm × 30 nm. It wasnoted that these dimensions were similar to those measured using neutron scattering ofasphaltenes in solvents. An increase in discoid size was observed with increasing sulphur contentof the asphaltenes, but no correlation in size was observed with increasing asphaltene molecularweight. Absorption of asphaltenes from unfiltered solutions revealed fractal-like asphalteneclusters a few micrometers long, 1 ìm wide, and 10 to 20 m thick. Figure 5 shows the change inoil sulphur content over the years.

30

Figure 5 Variation in ANS Sulphur Content in Past Years

Balinov and coworkers studied the use of C NMR to characterize emulsions using the13

NMR self-diffusion technique and compared this to optical microscopy (Balinov et al., 1994).The optical microscopy showed an average droplet size of about 4 ìm with a volume mean ofapproximately 8 ìm (estimated by the present author).

Eley and coworkers studied the formation of emulsions and found that the rate was firstorder with respect to stirring time (Eley et al., 1988). As the asphaltene content increased, the rateconstant decreased.

3.6 Recent Studies on Stability ClassesThe most important characteristic of a water-in-oil emulsion is its ‘stability’. The reason

for this is that one must first characterize an emulsion as stable (or unstable) before one cancharacterize its properties. Properties change significantly for each type of emulsion. Untilrecently, emulsion stability has not been defined (Fingas et al., 1995). Studies were thereforedifficult because the end points of analysis were not defined.

This section summarizes studies to measure the stability of water-in-oil emulsions anddefine characteristics of different stability classes. Four ‘states’ in which water can exist in oilwill be described. These include: stable emulsions, meso-stable emulsions, unstable emulsions(or simply water and oil), and entrained water. These four ‘states’ are differentiated by visualappearance as well as by rheological measures.

Studies in the past three years have shown that a class of very ‘stable’ emulsions exists,

31

characterized by their persistence over several months. The viscosity of these stable emulsionsactually increases over time. These emulsions have been monitored for as long as 3 years in thelaboratory. ‘Unstable’ emulsions do not show this increase in viscosity and their viscosity is lessthan about 20 times greater than the starting oil. The viscosity increase for stable emulsions is atleast three orders-of-magnitude more than the starting oil.

Fingas and coworkers have studied emulsions for many years (Bobra et al., 1992; Fingaset al., 1993, 1995, 1996, 1997, 1998, 1999, 2000a, 2000b, 2001a, 2002b; Fingas and Fieldhouse,1994, 2009). The last two of these references as well as Fingas et al. (1995) describe studies todefine stability. The findings of these earlier studies are summarized here. Both on the basis ofthe literature and experimental evidence, it was concluded that certain emulsions can be classedas stable. Some, if not all or many, stable emulsions increase in apparent viscosity with time, i.e.,their elasticity increases. It is suspected that the stability derives from the strong viscoelasticinterface caused by asphaltenes, perhaps along with resins. Increasing viscosity may be caused byincreasing alignment of asphaltenes at the oil-water interface.

Meso-stable emulsions are emulsions that have properties between stable and unstableemulsions (really oil/water mixtures) (Fingas et al., 1995). It is suspected that meso-stableemulsions either lack sufficient asphaltenes to render them completely stable or still contain toomany destabilizing materials, perhaps some aromatics and aliphatics. The viscosity of the oil maybe high enough to stabilize some water droplets for a period of time. Meso-stable emulsions maydegrade to form layers of oil and stable emulsions. Meso-stable emulsions can be red or blackand are probably the most commonly formed emulsions in the field.

Unstable emulsions are those that decompose (largely) to water and oil rapidly aftermixing, generally within a few hours. Some water, usually less than about 10%, may be retainedby the oil, especially if the oil is viscous.

The most important measurements to characterize emulsions are forced oscillationrheometry studies. The presence of significant elasticity clearly defines whether a stable emulsionhas been formed. The viscosity by itself can be an indicator of the stability of the emulsion,although it is not necessarily conclusive unless one is absolutely certain of the viscosity of thestarting oil. Colour is an indicator, but may not be definitive. This laboratory’s experience is thatall stable emulsions were reddish. Some meso-stable emulsions had a reddish colour andunstable emulsions were always the colour of the starting oil.

Water content is not an indicator of stability and is error-prone because of ‘excess’ waterthat may be present. It should be noted, however, that stable emulsions contain more than 70%water and unstable emulsions or entrained water-in-oil generally contain less than 50% water.Water content after a period of about one week is found to be more reliable than initial watercontent because separation will occur in emulsions that are less stable.

4. Emulsion Formation ModellingThe early emulsion formation theories were translated into modelling equations at that

time. Unfortunately, the processes described above were not apparent until about 10 years agoand have not yet been translated into modelling equations. Furthermore, the presence of differentwater-in-oil states dictates that one simple equation is not adequate to predict formation. Information on the kinetics of formation at sea and other modelling data was less abundant in the

32

past than it is now. It is now known that emulsion formation is a result of surfactant-likebehaviour of the polar asphaltene and resin compounds. These are similar compounds and bothbehave like surfactants when they are not in solution, but asphaltenes form much more stableemulsions. Emulsions begin to form when the above chemical conditions are met and when thereis sufficient sea energy.

In the past, the rate of emulsion formation was assumed to be first-order with time. Thiscan be approximated with a logarithmic (or exponential) curve. Although not consistent with theknowledge of how emulsions formed, this assumption has been used extensively in oil spillmodels. Most models that incorporate the phenomenon use the estimation technique developedby Mackay and coworkers (Mackay and Zagorski, 1982a,b; Mackay et al., 1980a,b, 1982) or avariation of this.

Mackay suggested an equation to model water uptake:

a bÄW = K (U + 1) (1 - K W)Ät, (2)2

where: ÄW is the water uptake rate, W is the fractional water content,

aK is an empirical constant,U is the wind speed,

bK is a constant with the value of approximately 1.33, andt is time.

Because eqn (1) predicts that most oils will form emulsions rapidly given a high windspeed, most users have adjusted the equation by changing constants or the form slightly.

Mackay and Zagorski (1982a,b) proposed two relationships to predict the formation ofemulsions on the sea. They proposed that the stability could be predicted as follows:

where: S is the stability index in relative units, high numbers indicate stability,

a x is the fraction of asphaltenes,

aã is the activity of asphaltenes,

aoK is a constant which here is 3.3,

wx is the fraction of waxes,

awK is a constant which is 200 at 293 K, andT is the temperature in kelvin.

Water uptake was given as:

T L sÄW = ÄW + ÄW

f l l= ÄT[k -k W ], (4)

Twhere: ÄW is the total change in water content,

LÄW is the change in water content for large droplets,

33

sÄW is the change in water content for small droplets,

f lÄT is time, k is the rate constant for formation, typically 1 h , k is the rate constant for-1

large droplet formation and is about 3 h , and-1

lW is the fraction of large droplets which is typically 3 to 4.Kirstein and Redding (1988) used a variation of the Mackay equation to predict

emulsification:

2where: k is a coalescing constant which is the inverse of the maximum weight fraction waterin the mixture,W is the weight fraction water in the mixture,

1k is the Mooney constant which is 0.62 to 0.65,

5k is the increase in mousse formation by weathering,

3k is the lumped water incorporation rate constant and is a function of wind speed inknots, andt is the time in days.

The change in viscosity due to mousse formation was given by:

where: ì is the resulting viscosity,

0ì is the viscosity of the starting oil, andthe remainder are identical to the above.

Reed (1989) used the Mackay equations in a series of models. The constants were adjustedto correspond to field observations:

wcwhere: dF /dt is the rate of water incorporation,W is the wind speed in m/s,

wcF is the fraction of water in oil, and

3C is the rate constant equal to 0.7 for crude oils and heavy fuel oils.The viscosity of the emulsion was predicted using the following variant of the Mooney

equation:

34

where: ì is the viscosity of the mixture,

0ì is the viscosity of the starting oil, and

wcF is the fraction of water in oil.The effect of evaporation on viscosity is modelled as:

0 4 evapì = ì exp (C F ), (9)

where: ì is the viscosity of the mixture,

0ì is the viscosity of the starting oil,

4C is a constant which is 1 for light fuels and 10 for heavy fuels, and

evapF is the fraction evaporated from the slick. All of the above work has a basis in the Mackay equations, which were developed beforeextensive work on emulsion physics took place. The present author suggests that both thetendency and the formation of emulsions could be predicted with a greater degree of accuracyusing empirical data as described below.

4.1 Fingas 2001 ModelIn 2001, Fingas proposed that a characteristic table be used to predict the formation of

water-in oil emulsions (Fingas et al., 2001a, b). A table of threshold properties relevant topredicting the formation of emulsions is used. Starting at the bottom of this table, it is presumedthat unstable emulsions are the most common state, which they are. The properties of the oil arethen compared, up to the final state or stable emulsions. The properties of the starting oil changeas it weathers and most models already incorporate formulae to estimate density and viscosity asthe oil weathers. Since resins and emulsions do not evaporate, it can be presumed that theyincrease proportionately to the loss by weathering. For example, if an oil containing 5% resinsand 5% asphaltenes weathers to 50%, the resin content is now 10% and the asphaltene content is10% as well.

4.2 Fingas 2004 ModelIn 2004, Fingas proposed a continuous model to predict any of the four emulsion states

(Fingas and Fieldhouse, 2004a,b,c). The formation data were used to develop mathematicalcorrelation equations. The value for each parameter was correlated in a series of models usingDataFit (Oakdale Engineering), which calculates linear models. A two-step process wasnecessary as DataFit is not able to calculate the specific mathematical function with more than 2variables due to the large number of possibilities. Thus, the function, e.g., linear, square, log,were calculated using a two-way regression (TableCurve) and these functions were in turn usedto develop a predictor model for emulsification.

The steps to produce the model are summarized here. First the parameters available werecorrelated one at a time. Regression coefficients were optimized by adjusting the class criteriafrom a starting value of 1 to 4 to a logarithm of this value. This was performed on a trial-and-error basis to yield the highest regression coefficient. The resulting criteria are: - 0.22 is unstable,0.69 is entrained, 1.1 is meso-stable, and 1.38 is stable. Several of these parameters can have azero value which causes calculation problems. If this is the case, the 0 is adjusted to either delete

35

these values or to adjust it to the typical high value for the parameter. This is called the ‘firsttransform’. A second transformation is performed to adjust the data to a singular increasing ordecreasing function. Most parameters have an optimal value with respect to class, that is thevalues have a peak function with respect to stability or class.

The values of the second correction were then correlated using the package Data Fit.Several models were developed. It had been noted in earlier work that heavy oils were somewhatdifferent in emulsion formation than were light oils (Fingas and Fieldhouse, 2003). For thisreason, trial models were created separately for light and heavy oils. These were actually poorerthan the other combined models and so ultimately were not used. The best model was one thatincluded only five parameters: density, viscosity, saturates, resins, and asphaltenes. It was foundthat the regression coefficients of class with aromatic content, asphaltene/resin ratio, and waxeswere too low to include in the model. It should be noted that there are some problems with thefundamental process of categorizing water-in-oil states at the onset. Some crude oils areenhanced when emulsion preventors, also called asphaltene suspenders, are added directly at thewell-head because they are very emulsion-prone. Some emulsion-prone oils may therefore notform emulsions during the laboratory or field tests because of these emulsion-preventingmaterials. Although attempts are made to obtain oils that do not contain these emulsion-preventing materials, it is impossible to know this fact in every case.

The oils and resulting water-in-oil states used for this correlation were studied to yield theaverage water content and increase in the viscosity from the starting oil to the water-in-oil class.This will be used to predict the water content and the viscosity, given the known class of water-in-oil mixture formed.

The kinetics of emulsion formation have been studied and data are available to computethe time to formation. A kinetics study has shown the times to formation for stable emulsions areparticularly rapid and that of entrainment is also rapid - both in a matter of minutes (Fingas et al.,2000b). This study yields data in terms of relative formation time and energy (rpm) of the mixingapparatus. This particular data set is thought to be especially accurate. A study in a large test-tankhas yielded data on the formation time of the various water-in-oil states (Fingas et al., 2002b).The data of the relative formation times and the wave height are available. The average data over25 runs was used to make the calculations. The formation time is taken as that time at which 75%of the maximum stability measured occurs. The conditions under which these tests took placeand the measurements taken are described in the literature (Fingas et al., 2002b). The waveheight for each experiment was measured and used to indicate relative sea energy, taken for afully developed sea. The laboratory data was converted from relative rotational energy to waveheight by equating formation times and then using this multiplier to calculate the equivalentwave height. Formulae were fitted to each of the three categories and the common formulaamong all three relevant classes, was found to be 1/x . Application of the equations will then1.5

provide a user with a time to formation of a particular water-in-oil state, given the wave height.The first step in using the model is to obtain or estimate the oil properties as they are at the

weathering condition of concern. The properties needed are the density, viscosity, and saturate,resin and asphaltene contents. These values require transformation as noted below.

Density - If the density is less than 0.96, then the density parameter is 0.96 less the densityand if it is greater than 0.96, it becomes the density less 0.96. The value used in the equation is

36

then the exponential of this transformed value. (10)Viscosity - Take the natural logarithm (ln) of the viscosity. If the natural log of the

viscosity is less than 7.7, then the viscosity parameter is 7.7 less the viscosity natural log and if itis greater than 7.7, it becomes the natural log of viscosity less 7.7. The value used in the equationis this transformed value. (11)

Saturate Content - If the saturate content (in percent) is less than 39, then the saturatecontent parameter is 39 less the saturate content and if it is greater than 39, it becomes thesaturate content less 39. The value used in the equation is this transformed value. (12)

Resin Content - If the value of the resins is zero, then set this value to 20. If the resincontent is less than 2.4, then the resin content parameter is 2.4 less the resin content and if it isgreater than 2.4, it becomes the resin content less 2.4. The value used in the equation is thenatural logarithm of this transformed value. (13)

Asphaltene Content - If the value of the asphaltene content is zero, then set the value to30. If the asphaltene content is less than 15.4, then the asphaltene content parameter is 15.4 lessthe asphaltene content and, if it is greater than 15.4, it becomes the asphaltene content less 15.4.The value used in the equation is then the exponential of this transformed value. (14)

The class of the resulting emulsion is then calculated as follows:

Class = -1.36 + 2.62*Dt - 0.18*Vt -0.01*St +0.02*Rt - 2.25 10 *At (15)-7

where: Class is the numerical index of classification,Dt is the transformed density as calculated in equation (10),Vt is the transformed viscosity as calculated in equation (11),St is the transformed saturate content as calculated in equation (12),Rt is the transformed resin content as calculated in equation (13), andAt is the transformed saturate content as calculated in equation (14).

The second step in calculating the emulsion formation and its properties is to apply thenumeric class value as yielded from equation (15). This is simply accomplished by using a giventable (Fingas and Fieldhouse, 2004c).

The third step is to predict the properties of the resulting water-in-oil emulsion. Theaverage water content and increase in viscosity of the various types of emulsions is taken fromformer data.

The fourth step is to predict the time to formation after the oil is weathered to the statedpercentage. This calculation can be made using the equations given in the literature:

Time to formation (min) = a + b/Wh (16)1.5

where: a is a constant and is 27.1 for a stable emulsion formation, 47 for meso-stable, and30.8 for an entrained water-in-oil class,b is a constant and is 7,520 for a stable emulsion formation, 49,100 for meso-stable,and 18,300 for an entrained water-in-oil class, andWh is the wave height in cm.

37

4.3 Fingas 2005 ModelThe 2004 model was totally revised in 2005 to provide a model of greater accuracy (Fingas

and Fieldhouse, 2005). The development methodology was similar to that above. Themathematical procedures still required transformations of the data for the reasons stated above.These transformations of input values are summarized below.

Density - If the density is less than 0.97, then the density parameter is 0.97, less the densityand, if it is greater than 0.97, it becomes the density, less 0.97. The value used in the equation isthis transformed value. (17)

Viscosity - Take the natural logarithm (ln) of the viscosity. If the natural log of theviscosity is less than 8.7, then the viscosity parameter is 8.7 less the viscosity natural log, and if itis greater than 8.7, it becomes the natural log of viscosity less 8.7. The value used in the equationis this transformed value. (18)

Resin Content - If the value of the resins is zero, then set this value to 20. If the resincontent is less than 5.4, then the resin content parameter is 5.4, less the resin content and if it isgreater than 5.4, it becomes the resin content, less 5.4. The value used in the equation is thistransformed value. (19)

Asphaltene Content - If the value of the asphaltene content is zero, then set the value to30. If the asphaltene content is less than 12, then the asphaltene content parameter is 12 less theasphaltene content and, if it is greater than 12, it becomes the asphaltene content less 12. Thevalue used in the equation is this transformed value. (20)


Class = 0.738 - 0.197*Dt - 0.0126*Vt -0.0007*Rt - 0.00358*At (21)

where: Class is the numerical index of classification,Dt is the transformed density as calculated in equation (17),Vt is the transformed viscosity as calculated in equation (18),Rt is the transformed resin content as calculated in equation (19), andAt is the transformed asphaltene content as calculated in equation (20).

The second step in calculating the emulsion formation and its properties is to apply thenumeric class value as yielded from equation (21). This is simply accomplished by using Table 2.

The third step is to predict the properties of the resulting water-in-oil emulsion. Theaverage water contents and increase in viscosity are given in the paper (Fingas and Fieldhouse,2005).

The fourth step is to predict the time to formation after the oil is weathered to the statedpercentage. This calculation can be made using the equations in the reference above:


where: a is a constant and is 27.1 for a stable emulsion formation, 47 for meso-stable, and30.8 for an entrained water-in-oil class,b is a constant and is 7,520 for a stable emulsion formation, 49,100 for meso-stable,and 18,300 for an entrained water-in-oil class, and

38

Wh is the wave height in cm.

4.4 Fingas 2009 ModelThe older modeling method as noted in (3) above, used an assigned number for each

water-in-oil type and then this was regressed with the various oil starting properties. Further, thefunctional relationship (square, log, exponential, etc.) of the relationship of the starting oilproperties was determined before hand and then correlated with a multi-parameter linear methodto yield a predictor model. Since a stability index has been established, this can be used as thetarget of the correlation (Fingas and Fieldhouse, 2009). This new modeling has a differentapproach than the old one. Furthermore, this new approach used a multi-regression programdirectly, using various multi-functional transformations of the input oil property data. This willallow the software of the regression model to assign portions of the functions necessary toachieve the highest correlation factor. Other differences are also that the transformations areautomated using a peak function fit, yielding slightly different transformation values than the oldvalues; use of a newer data set; and incorporation of several functional input parameters,dependent on the statistics of the model, into the final equation.

A transformation is needed to adjust the data to a singular increasing or decreasingfunction. Regression methods will not respond correctly to a function that varies both directlyand inversely with the target parameter. Most parameters have an optimal value with respect toclass, that is the values have a peak function with respect to stability or class. The stability asused in this formulation is called Stability C.

The arithmetic to perform the transformation is: if the initial value is less than the peakvalue, then the adjusted value is the peak value less the initial value; and if the initial value ismore than the peak value, the adjusted value is the initial value less the peak value. It should benoted that the exponential of density was used and the natural log of the viscosity. Previousmodeling work had shown that these mathematical changes are necessary to achieve highercorrelations.

Having the transformed values, the new model proceeded directly to fit a multiplelinear equation to the data. The choice of functions was achieved by correlating the stabilityfunction directly with the data and taking the best of the functions (eg. square, log, etc) furtherinto the regression process. The functionalities of square, logarithmic or exponential curves areachieved by correlating the straight value of the input properties plus their expanded values,taken here as the cube of the starting parameter as well as the square of the exponential of thestarting value; and their companded values, the natural log (ln) and the logarithm (base 10) ofparameter divided by the square of the value. Thus each parameter is correlated with the stabilityC index in five sets of mathematical statements. This is similar to the standard Gaussianexpansion regression technique. In this method the regression is expanded to functionalitiesabove and below linear until the entire entity is optimized. For example a linear function wouldbe included, then a square and then a square root and so on until tests of the complete regressionshow that there are no more gains in increased expansions. Using this technique, six inputparameters: exponential of density, ln of viscosity, saturate content, resin content, asphaltenecontent and the asphaltene/resin ratio (A/R) were found to be optimal. Thus with 4transformation and the original values of these input parameters, there are six times five or 30

39

input combinations.Using Datafit, a multiple regression software, a maximum of 20 of these could be taken at

a time to test the goodness-of-fit. Values that yield Prob(t) factors of greater than 0.9 weredropped until all remaining factors could be calculated at once. The Prob(t) is the probability thatinput can be dropped without affecting the regression or goodness-of-fit. Over twenty regressionswere carried out until the resulting model was optimal. The r , the regression coefficient, was2

0.75, which is quite high considering the many potential sources of error, etc. The procedures forcalculating, using the new model are given below.

The first step is to transform the input data so that it forms a continuous declining orincreasing function.

Density: Take the exponential of the density. If the exponential of density is less than2.5, then the density parameter is 2.5 less the density and if it is greater than 2.5, it becomes thedensity less 2.5. The value used in the equation is this transformed value. (23)

Viscosity: Take the natural logarithm (ln) of the viscosity. If the natural log of theviscosity is less than 5.8, then the viscosity parameter is 5.8 less the viscosity natural log and if itis greater than 5.8, it becomes the natural log of viscosity less 8.7. The value used in the equationis this transformed value. (24)

Saturate Content: If the saturate content is less than 45, then the saturate content parameteris 45 less the saturate content and if it is greater than 45, it becomes the saturate content less 45.The value used in the equation is this transformed value. (25)

Resin Content: If the value of the resins is zero, then set this value to 20. If the resincontent is less than 10, then the resin content parameter is 10 less the resin content and if it isgreater than 10, it becomes the resin content less 10. The value used in the equation is thistransformed value. (26)

Asphaltene Content: If the value of the asphaltene content is zero, then set the value to 20.If the asphaltene content is less than 4, then the asphaltene content parameter is 4 less theasphaltene content and if it is greater than 4, it becomes the asphaltene content less 4. The valueused in the equation is this transformed value.

(27)A/R or Asphaltene/Resin Ratio: This is taken as the direct value of the asphaltene content

in percent (untransformed) divided by the resin content in percent (again untransformed). If A/Ris less than 0.6, then the A/R parameter is 0.6 less the A/R and if it is greater than 0.6, it becomesthe A/R less0.6. The value used in the equation is this transformed value. (28)


Stability C = 12.3 + 0.259*St - 1.601*Rt - 17.2*A/Rt - 0.50*Vt + 0.002*Rt 3 3

+ 0.001*At + 8.51*A/Rt - 1.12*lnVt + 0.700*lnRt + 2.97*lnA/R3 3

+ 6.0X10 *ExplnVt -1.96*ExpA/Rt - 4.0X10 *logExpDt/expDt-8 2 2 -6 2

- 1.5X10 *logA/Rt/A/Rt (29)-4 2

Where Class is the Stability C of the resulting water-in-oil type,St is the transformed saturate content as calculated in equation (25), abbreviated A

here,

40

Rt is the transformed resin content as calculated in equation (26), abbreviated B,

A/Rt is the transformed asphaltene/resin ratio as calculated in (28), abbreviated C,

Vt is the cube of the transformed ln viscosity as calculated in (24),3

abbreviated D,Rt is the cube of the transformed resin content as calculated in equation (26),3

abbreviated E,At is the cube of the transformed asphaltene content as calculated in equation (27),3

abbreviated F, A/Rt is the cube of the transformed A/R ratio as calculated in equation (28),3

abbreviated G,lnVt is the natural logarithm (ln) of the transformed viscosity as calculated in equation

(24), abbreviated H,lnRt is the natural logarithm (ln) of the transformed resin content as calculated in

equation (26), abbreviated I,lnA/R is the natural logarithm (ln) of transformed asphaltene/resin ratio as calculated

in (28), abbreviated J,ExplnVt is the is the exponential of the transformed viscosity - squared - as2

calculated in equation (24), abbreviated K, ExpA/Rt is the is the exponential of the A/R ratio - squared - as calculated in2

equation (28), abbreviated L,logExpDt/expDt is the is the logarithm (base 10) of exponential of the density -2

divided by the square of the transformed density - the transformed density ascalculated in equation (21), abbreviated M,

logA/Rt/A/Rt is the is the logarithm (base 10) of exponential of the A/R ratio -2

divided by the square of the A/R ratio - the transformed A/R ratio as calculated inequation (28), abbreviated N,

A simplified version of the equation is then:Stability C = 12.3 + 0.259A - 1.601B - 17.2C - 0.50D + 0.002E + 0.001F + 8.51G - 1.12H + 0.700I + 2.97J + 6.0X10 K -1.96L - 4.0X10 M - 1.5X10 N (30)-8 -6 -4

Where the parameters A to N are defined as above.

The Stability C of the resulting product is calculated using the rheological measurementsof the water-in-oil product formed (Fingas and Fieldhouse, 2009).

The basic uncorrected Stability C or cross product is:Xpr = Complex Modulus X Elastic Modulus (31)

Starting Oil Viscosity Starting Oil Viscosity(Stability A) (Stability B)

The corrected stability C is:

41

Stability C = ln (Xpr*xpr)/1000 (32)

Where the Xpr is the value from equation (31).

Or equivalently the Stability C can be calculated from:Stability C = ln (Stab A*Stab B) /10 (33)2 11

The first point to be discussed here will be the error or goodness-of-fit. An analysis oferror shows the error percentage varies by water-in-oil type. It is noted that there is little error formore stable types, but more error for the unstable types. This was noted in past modeling as well(Fingas and Fieldhouse, 2004a, 2005). It is suspected that the reasons for this are:

a) Unstable types generally consist of three widely separate classes of oils or fuels, verylight oils such as the fuels which have little or no resins or asphaltenes; those very heavy oilswhich are so viscous that they will not uptake water; and those oils that have the incorrect ratioor amounts of resins or asphaltenes. It is difficult to mathematically incorporate all three of thesevariances into one grouping.

b) Some of the oils may be able to form different water-in-oil types, but emulsioninhibitors or asphaltene suspenders have been added to the products. These types of oils makeprediction very difficult as they mislead the regression process. and

c) There are many different asphaltenes some of which make much more stable emulsionsthan other. Recent work has shown that there are hundreds of asphaltene sub-components varyingvery much in composition and molecular size (Groenzin and Mullins, 2007). Thus the percent ofasphaltenes (or resins) certainly does not tell the whole story about the emulsion-stabilizers in theemulsions.

It is important to recognize the properties and behaviours of the four water-in-oil types arevery different from each other. It should be noted that on the day of formation, only meso-stableand stable may be somewhat similar, this similarity disappears after one week when the meso-stable emulsions break down. The other significant point is that the change in water content ofthe entrained types is the slowest of all. In summary, the four water-in-oil types differ by as muchas orders-of-magnitude in viscosity and water content at one day, one week and one year.

A remaining question is how good is this model compared to other models? Table 2shows a comparison for several oil types. These oil types were chosen as common types that arein commerce or as they are well known for their properties. The first six columns give the oilidentification and its actual properties of water content and viscosity on the first day offormation. The first model compared shows the traditional exponential water uptake as predictedusing equations in Section 2 above. The shaded and boxed values are those predictions in error.A criteria of at least 25% in water content or incorrect type - if predicted. The number of casespredicted correctly are given at the bottom and the average orders-of-magnitude different fromthe actual value. It can be seen from Table 2 that the old exponential water uptake modelpredicted the correct water content only 7% of the times in this set of oils and the order ofmagnitude of the magnitude of viscosity was over two order-of-magnitude from the actualviscosity on the day of formation. This comparison would indicate that exponential water uptakemodels are not useful and are in-fact, misleading.

42

The second model compared in Table 2 is that of ADIOS, a widely-available model fromNOAA. The water content is predicted within about 25% in about 49% of cases and the viscosityaverages about 1.6 order-of-magnitude in error from the actual value. The third model comparedis the older model by the present author. It predicted the water content correctly in 72% of casesas well the correct water-in-oil type. The last model compared is the one presented for the firsttime in this paper. This model was excellent and predicted the water content and type 100%correct for all types and the viscosity error average 0.2 orders-of-magnitude.

5 Dispersant Effectiveness PredictionOil properties can be correlated with dispersant effectiveness to estimate the amount of oil

dispersion. Such correlation could be used to indicate which oil properties, such as asphaltenecontent, might inhibit or facilitate oil dispersion.

In an earlier study, the dispersant effectiveness data on 295 oils and 29 of their chemicaland physical properties were correlated to develop a prediction equation and to examine theprocess of chemical dispersion (Fingas et al., 2003b). The dispersibility of Corexit 9500 in theswirling flask apparatus was used as the key parameter. The highest correlation parameters wereobserved with the content of nC12, naphthalenes, inversely with C26, the PAHs, and the sum ofC12 to C18 hydrocarbons. This is highly indicative that the smaller aliphatic hydrocarbons up toC18 and the PAHs are the most dispersible components of oil. An inverse relationship wasobserved with aliphatic hydrocarbons greater than C20. This indicates that these hydrocarbonsresist dispersion.

Thirteen statistical models were constructed to predict the chemical dispersibility of oils. The simplest and best model is: Corexit 9500 dispersibility (%) = -11.1 -3.19(lnC12 content) +0.00361(naphthalene content in ppm) - 7.62(PAH content squared) + 0.115(C12 to C18 contentsquared) + 0.785(%fraction oil boiling below 250 C). (34)o

Statistical models ranged from simple predictors involving only two parameters such asviscosity and density to 14-parameter models. The models developed were analysed statisticallyand the effectiveness was calculated for several dispersants. The more sophisticated models areable to predict dispersant effectiveness with high accuracy.

It should be noted that the parameters that correlate most highly with the suite ofparameters are those composition parameters that relate to smaller compounds in the oil. Theseinclude n-C12, naphthalenes, and the sum of the C12 to C16 components. Those that relate to thelarge compounds in the oil relate negatively to the dispersibility, including C26 and resins. Thiswill be discussed in greater detail later, however, it indicates that dispersion primarily affectsonly the smaller components of the oil.

The highest correlation was achieved with the n-C12 component. The regressioncoefficient was 0.79, which indicates that C12 is highly dispersible. It should be noted that onlyabout 15 of the 299 values used, which were correlated, had data for C-12 and some of the otherspecific component data. The next highest correlation coefficient was 0.76 for the naphthalenecontent. This also indicates a high dispersibility for naphthalene. The third highest correlation isfor n-C26 and this is an inverse correlation. This indicates that the more n-C26, the lessdispersion, especially in the initial amount. There are two points at which the C26 is very lowand the dispersibility is very high, yielding a sharp initial decrease in the initial portion of the

43

curve. This indicates that components of the size of C26 and greater are poorly dispersed and infact inhibit dispersion. The fourth highest correlation is the PAH content and this correlatespositively, namely that the higher the PAH content, the higher the dispersion. This is somewhatsurprising since the PAH compounds, especially the larger PAHs such as phenanthrene andchrysene, were not thought to be dispersible. This high correlation indicates that most of thePAHs are dispersible, possibly because most of the PAHs consist of smaller-ring compounds.The fifth highest correlation is that of the sum of the C12, C14, C16, and C18 components. Thiscorrelation is highly indicative that alkanes up to C18 are the prime components dispersed alongwith the PAHs. The fact that C12 correlates the highest of these n-alkanes and that thiscorrelation rapidly drops off to C18 with no useable correlation for C20 indicates that primarilyhydrocarbons up to C18 disperse and that past C20 compounds disperse poorly.

Although viscosity correlates moderately, it would not be a good predictor by itself.Viscosity has a tendency to range over orders of magnitude and higher viscosity oils over about5000 mPa.s have no dispersibility. The problem with using viscosity alone is that some of theoils in any test set can have viscosity as much as 4 orders-of-magnitude above that which wouldstill achieve dispersant effectiveness. This results in lack of continuity in dispersant effectivenessover the typical viscosity range.

The oil fraction that boils below 250 C has a correlation coefficient of 0.62 and thuso

shows that this component of the oil is relatively-well dispersed using a chemical dispersant.Some models use this fraction (BP < 250 C) as the amount that evaporates within the first fewo

hours after a spill. In fact, some algorithms match this fraction with the percent that wouldevaporate in 2 days. This indicates that chemical dispersion is strongly competitive toevaporation in that the same fraction is subject to either process.

The n-alkane 14 and 16 correlation with Corexit 9500 dispersion with a correlationcoefficient of 0.61 and 0.56 show that these components of the oil are preferentially dispersedusing a chemical dispersant. It should be noted that the correlation coefficient declinesprogressively from C12 to C20 and then rises inversely to C26.

The correlation of the oil density with chemical dispersibility yields a correlationcoefficient of 0.54. This correlation may be quite useful since the density of the oil is usuallyknown and the correlation is relatively good and continuous throughout the density range. Thiscorrelation can be used when little else is known about the oil.

The correlation of the resin content with the Corexit 9500 dispersibility is poor. The resinsshow the highest correlation coefficient of the SARA fraction (Saturates, Aromatics, Resins,Asphaltenes). It was thought that the SARA analysis would yield a good simple prediction ofdispersibility. This study shows, however, that the SARA fraction actually is a poor predictor ofdispersibility.

The correlation of the saturates, aromatics, and asphaltene components yield correlationcoefficients of 0.36, 0.18, and 0.24, respectively. These three components display an even greaterscatter than the resins with the corresponding low correlation coefficients. The reason for thepoor fit of the SARA components, particularly the saturates and aromatics, is that compoundsgrouped in these categories have a wide range of masses. The lighter fractions disperse readilywhile the heavier fractions do not. For example, the C12 to C18 group as described above aresaturates and are highly dispersible. On the other hand, the C20 fraction and above is not

44

dispersible as already noted, and is also part of the saturate group. The same situation exists forthe aromatics group.

The correlation of pour point with Corexit 9500 is poor with dispersibility, R = 0.25. This2

latter correlation is poor and is not useful for prediction. Pour point is not a continuous functionand thus becomes a poor predictor of physical behaviour.

The correlation of sulphur content (R = .23) with Corexit 9500 dispersibility is poor. The2

sulphur content does not show significant relationship to dispersibility, as might be expected, andmost values cluster around the 0 to 10% sulphur content.

The total VOC and the C18 content yield correlation coefficients of 0.33 and 0.32,respectively. The total VOC content displays a large scatter with dispersibility. This is probablythe result of rapid loss of some of the VOC components before dispersion. The C18 content isthe largest n-alkane to show a positive correlation with the dispersion. This indicates that C18 isprobably the largest n-alkane to undergo chemical dispersion. The next member chosen, C20,shows little correlation.

The factors that were correlated and show little correlation include the Reid vapourpressure, flash point, waxes, and surface tension (and interfacial tension with water). There is noreason to believe that any of these are related to chemical dispersibility.

Model 1 uses the four of the higher correlating parameters of C12, naphthalene, PAHs,C12 to C18, and the fraction that boils at less than 250 C. The regression coefficient achievedo

was 0.98.The model is:

Corexit 9500 dispersibility (%) = -11.1 -3.19(ln C12 content) + 0.00361(naphthalenecontent in ppm) - 7.62(PAH content squared) + 0.115(C12 to C18 content squared) +0.785(%fraction oil boiling below 250 C) (35)o

The Prob(t) shows that all factors are relevant and are needed to form a reliable prediction. Onlyboiling point has a low Prob(t) value. It should be noted that only 15 oils have the full data set toform this prediction set.

Model 2 uses the six highest correlating parameters of C12, naphthalene, PAHs, C12 toC18, the C26 fraction (negative correlation), and the fraction that boils at less than 250 C. Theo

regression coefficient achieved was 0.98.The model is:

Corexit 9500 dispersibility (%) = -10.7 -2.75(ln C12 content) + 0.00354(naphthalenecontent in ppm) + 0.113/C26 content - 7.48(PAH content squared) + 0.0107(C12 to C18 contentsquared) + 0.761(% fraction oil boiling below 250 C) (36)o

The Prob(t) shows that most factors are relevant and are needed to form a reliable prediction.This prediction set is very similar to Model 1 and the predictions are similar, but slightly moreaccurate.

Model 3 uses the 5 highest correlating parameters of C12, naphthalene, PAHs, C12 toC18, the C26 fraction (negative correlation), and the viscosity of the oil rather than the fractionthat boils at less than 250 C. The regression coefficient achieved was 0.94. The model is:o

Corexit 9500 dispersibility (%) = -2.93 -1.29(ln C12 content) + 0.00368(naphthalenecontent in ppm) - 0.0185/C26 content - 8.65(PAH content squared) + 0.0144(C12 to C18 contentsquared) + 100/viscosity (37)

45

The Prob(t) shows that there may be redundancy in adding viscosity. This prediction set is verysimilar to Model 2 and the predictions are similar but less accurate as viscosity is not as good apredictor as the values associated with the fraction boiling below 250 C.o

Model 4 is a simple 2-parameter predictor using only density and viscosity. The regressioncoefficient is 0.71. The model is:

(38)Corexit 9500 dispersibility (%) = -77.6 + 214e + 60/viscosity -density 0.5

While this model produces a poorer prediction than most, it requires very little input data and thisdata, the density and viscosity, are readily available. The overall standard deviation is 4.6 as anaverage, but the maximum standard deviation is 32.

Model 5 is also a simple 2-parameter predictor using only density and the fraction boilingbelow 250 C. The regression coefficient is 0.7. The model is:o

(39)Corexit 9500 dispersibility (%) = -68.8+67.4/density + 0.787BP 1.5 1.5

This model produces a poorer prediction, similar to Model 4, but requires very little input dataand this data, the density and fraction that boils at less than 250 C, are commonly available. Theo

overall average standard deviation is 5 and the maximum standard deviation is 30. Both theaccuracy and other features of Model 5 are similar to Model 4, although the maximum deviationswith Model 5 are less. It should be noted that as many as 295 data points were used to generateboth Models 4 and 5.

Model 6 uses the SARA parameters of saturates, aromatics, resins, asphaltenes, and theviscosity of the oil. The regression coefficient achieved was 0.68. The model is:

Corexit 9500 dispersibility (%) = -7.78 + 0.315(saturate content) + 3.44(square root ofaromatic content in percent) - 4.32(ln resin content) - 1.81(ln asphaltene content) + 58.9/viscosity

(40)The Prob(t) shows that there is little redundancy but none of the factors are highly significant. Asnoted above, it was thought that the SARA analysis would yield a good simple prediction system,however this analysis shows that the SARA fraction actually is a poor predictor of dispersibility.The reason for the poor correlation achieved with SARA components, particularly the saturatesand aromatics, is that compounds grouped in these categories have variable dispersibility. Forexample, the C12 to C18 group as described above are saturates and are highly dispersible. Onthe other hand, the C20 fraction and above are not dispersible, but are also saturates. The samesituation exists for the aromatics group. Model 6 does not show good predictability as shown inTable 4 and Table A2. Model 6 has the second poorest correlation coefficient of all of the 13models described in this study.

Model 7 uses the SARA parameters of saturates, aromatics, resins, asphaltenes, and thesum of the C12 to C18 components. The regression coefficient achieved was 0.95. The model is:

Corexit 9500 dispersibility (%) = 296 - 1.86(saturate content) - 18.2(square root ofaromatic content in percent) - 33.6(ln resin content) - 9.03(ln asphaltene content) +0.0065(square of the C12 to C18 content in ppm) (41)The Prob(t) shows that there is little redundancy in any input parameter but also that none of theparameters are highly significant. This model is very much better in terms of fit and accuracythan the very similar Model 6. This is because the C12 to C18 component provides theinformation to the model as to what is being dispersed. In Model 6, this parameter was viscosity,which is a less powerful predictor.

46

Model 8 is similar and uses the SARA parameters of saturates, aromatics, resins,asphaltenes, and the VOCs. The regression coefficient achieved was 0.71. The model is:

Corexit 9500 dispersibility (%) = 73.4 - 0.0298(saturate content) - 2.24(square root ofaromatic content in percent) - 12.2(ln resin content) - 4.873(ln asphaltene content) +0.000681(VOC content in ppm) (42)The Prob(t) shows that none of the parameters are more significant compared to the other factors.Model 8 does not show good predictability. The VOC content does not substitute for the highpredictability of the C12 to C18 content as used in Model 7.

Model 9 uses only the SARA parameters of saturates, aromatics, resins, and asphaltenes.The regression coefficient achieved was 0.68, the poorest of the 13 models described in thisanalysis. The model is:

Corexit 9500 dispersibility (%) = 62.7 - 0.103(saturate content) - 0.678(square root ofaromatic content in percent) - 13.3(ln resin content) - 4.38(ln asphaltene content)

(43)The Prob(t) shows that there is redundancy in input parameters except somewhat for the saturatecomponent, which is a better predictor than the other input parameters. Model 9 shows theSARA component does not provide good information upon which to build a dispersibility model.

Model 10 is a larger model and uses all the composition components for which data hadbeen collected, including the SARA parameters of saturates, aromatics, resins, asphaltenes, andthe VOCs, the C12 to C18 component, and the C12, C14, C16, C18, C26 naphthalene and PAHcomponents. The regression coefficient achieved was 0.998. This is the second best modeldeveloped in this study. The model is:

Corexit 9500 dispersibility (%) = 368 - 2.25(saturate content) - 15.4(square root ofaromatic content in percent) - 42.6(ln resin content) - 14(ln asphaltene content) + 0.000472(VOCcontent in ppm) + 0.074(C12 to C18 content squared) - 1.71(ln(C12 content) - 8.34(ln C14content) - 17(C16 content) + 8.87(C18 content) + 0.821/C26 content + 0.00156(naphthalenecontent in ppm) - 1.36(PAH content squared). (44)The Prob(t) shows that there is significance to all input parameters, especially the C12 and C14parameters. Model 10 shows good predictability.

Model 11 is the largest model described in this study and uses many of the compositioncomponents including the SARA parameters of saturates, aromatics, resins, asphaltenes, and theVOCs, the C12 to C18 component, the C12, C14, C26 naphthalene, but physical componentswere substituted for those component parameters that showed high redundancy in Model 10. Thephysical components added were density, viscosity, and the fraction that boils at less than 250 Co

and less than 200 C. The regression coefficient achieved was 0.998. This is the best modelo

developed in this study, however, the fit is only marginally better than Model 10. The model is:Corexit 9500 dispersibility (%) = 855(1/density) -250/viscosity - 7.09(saturate content) -

72.6(square root of aromatic content in percent) - 69.7(ln resin content) - 11.6(ln asphaltenecontent) + 0.00045(VOC content in ppm) - 6.82(%fraction oil boiling below 200 C) +o

4.96(%fraction oil boiling below 250 C) - 0.0226(C12 to C18 content squared) + 11.4(ln(C12o

content) + 2.8(ln C14 content) + 0.299(1/C26 content) - 0.00414(naphthalene content in ppm) (45)

The Prob(t) shows that there is significance to all input parameters, especially the C12, C14, and

47

C26 parameters. Model 11 also shows good predictability.Model 12 is based on physical measurements. The physical components used were density,

viscosity, and the fraction that boils at less than 250 C and less than 200 C. The regressiono o

coefficient achieved was 0.71. The model is:Corexit 9500 dispersibility (%) = - 95.6 + 90(1/density) + 22.9/viscosity -

0.443(%fraction oil boiling below 200 C) + 0.855(%fraction oil boiling below 250 C) o o

(46)The Prob(t) shows that there is little significance to any input parameters compared to otherinput parameters. Physical measurements provide poor predicability.

Model 13 is based on the same physical measurements as Model 12, but pour point wasadded. The physical components used were density, pour point, viscosity, and the fraction thatboils at less than 250 C and less than 200 C. The regression coefficient achieved was 0.69. Theo o

model is:Corexit 9500 dispersibility (%) = - 124 + 121/density - 0.00071(pour point squared)+

15.3/viscosity - 0.488(%fraction oil boiling below 200 C) + 0.732(%fraction oil boiling belowo

250 C) (47)o

The Prob(t) again shows that there is all input parameters have low significance. The model ispoorer than Model 12 which includes the same parameters without pour point. This shows thatthe addition of pour point actually decreases the accuracy of the model. As discussed above, pourpoint is a very poor predictor and is not a continuous variable.

The work presented above used the dispersibility with Corexit 9500 as the primeparameter. This was carried out as the Corexit 9500 data was the newest and most accurate.Using the program TableCurve and the data in Table A1, predictor equations were developed forthe dispersibility of other dispersants with the various oils.

The equation for the prediction of Corexit 9527 dispersibility is:Corexit 9527 dispersibility (%) = - 0.35 + 0.80(Corexit 9500 dispersibility)

(48)The equation for the prediction of Dasic LTS dispersibility is:Dasic LTS dispersibility (%) = 1.5 + 0.42(Corexit 9500 dispersibility) (49)

The equation for the prediction of Enersperse 700 dispersibility is:Enersperse 700 dispersibility (%) = 1.9 + 0.55(Corexit 9500 dispersibility)

(50)

The development of these models reveals essentials of chemical dispersion. The resultsshow that small n-alkanes are prone to dispersion and that this ends at about C20 andhydrocarbons as large as C26 resist dispersion. There is a steady progression downwards ineffectiveness beginning at C12 and crossing 0 at about the C20 carbon number. The aromaticcomponent may show a similar tendency. The naphthalene component showed a high regressioncoefficient (R = 0.76) and the total PAHs were relatively high (R = 0.67). This indicates that the2 2

PAHs are relatively dispersible and that the smaller ones (naphthalenes) are highly dispersible.The development of the model shows that certain parameters are very good predictors of

chemical dispersibility. These include the specific chemical composition indicators such as the n-alkane values of C12, C14, naphthalenes, etc. The group composition indicators such as SARA

48

are poor predictors. The physical properties are poor predictors of chemical dispersibility. Thereare some properties that have no or very little dispersibility prediction indication and theseinclude: wax content, interfacial tension, and flash point.

6. Quick Behaviour ModelingAn important tool in oil spill planning and response is oil spill modeling. Oil spill

modeling relies on algorithms that require a number of oil property inputs. Sometimes theseinputs are not readily available, especially at the time of a spill. It is the focus of this section toreport on the development of simple estimators to predict oil evaporation, emulsification andchemical dispersion, using only density and viscosity. Density and viscosity are often the onlydata available at the time of the spill. The algorithm development will be described with thebackground of the significance of spill processes and more detailed behaviour predictionalgorithms, sometimes using more sophisticated chemical data inputs.

The behaviour of oil in the environment is largely dictated by its chemical composition.Oil composition data that are relatively available include the SARA or Saturates, Aromatics,Resin and Asphaltene contents. As will be shown later, sometimes these data are not sufficient toyield a truly accurate prediction. More detailed compositional data are available for only a fewoils (Oil Catalogue, 2009).

The most important behavioural aspect is felt to be evaporation. Evaporation isresponsible for the largest mass balance change in oils except for those rare times when oilsdisperse naturally. Evaporation is generally necessary before emulsification occurs and is acompeting process to natural dispersion. The chemistry of crude and refined oils defines their behaviour at sea. The chemistry ofcrude and refined oils is in turn, dominated by the fact that they are composed of many, in somecases, hundreds, of different compounds, each with their own behaviour and properties. Thecombination of these compounds can be viewed from a number of different perspectives; someview it as a combination of selected typical compounds, others view oil as being represented by asingle compound and still others view each oil as a unique substance. All of these points of viewhave disadvantages in modelling. If an oil is viewed as a combination of a few markercompounds, some of the behavioural aspects are not accounted for - eg. asphaltenes and their rolein water-in-oil emulsification. Taking the oil as a single compound removes many behaviouralcharacteristics of an oil and thus is not useful for spill modelling if behaviour is an importantfactor. Taking an oil as a unique combination complicates spill modelling in that new oils aredifficult to add. The importance of understanding the composition and overall behaviour of aspecific oil cannot be over-emphasized.

A number of oil properties are important to the understanding of oils and their behaviour. An important property is viscosity or resistance to flow. Technically defined, viscosity is theshear stress (force/area) divided by the shear rate (s ). The viscosity of an oil also defines its-1

spreading rate and is related in some way to most behavioural aspects. Density is the weight ofspecific volume of oil and in terms of behaviour defines how an oil will float on water. Densityis strongly correlated to viscosity for most oils. American Petroleum Institute (API) gravity is adifferent way of stating density (API = 141.5/d -131.5). Water content is an importanto

measurement for oils that have emulsified or formed water-in-oil emulsions. This is an indication

49

of the expansion of spill volume resulting from water incorporation and to a lesser degree of thestability of the resulting emulsion. Surface and interfacial tension are often cited as beingimportant factors. The interfacial tension is the energy difference between the oil layer and air orwater. Interfacial tension is not as important as some might stress since it cannot be used inisolation to predict any behavioural aspects of the oil. Furthermore it varies little between oils.Pour point is the temperature below which the oil ceases to pour as a fluid. Pour point is adifficult measurement and subject to variance, furthermore with a multi-component fluid such ascrude oil and many of the distilled products, pour point is not a meaningful measurement forpredicting or understanding oil behaviour.

EvaporationEvaporation is a very important process for most oil spills. In a few days, typical crude oils

can lose up to 40% of their volume. Most oil spill behaviour models include evaporation as aprocess and as an output of the model. Despite the importance of the process, relatively littlework had been conducted on the basic physics and chemistry of oil spill evaporation untilrecently. The difficulty with oil evaporation is that oil is a mixture of hundreds of compoundsand this mixture varies from source to source and even over time. Much of the work described inthe literature focuses on ‘calibrating’ equations developed for water evaporation. Scientific andquantitative work on water evaporation is decades old. The basis for the oil work in the literatureis water evaporation. There are several fundamental differences between the evaporation of apure liquid such as water and that of a multi-component system such as crude oil. Mostobviously, the evaporation rate for a single liquid such as water is a constant with respect to time. Evaporative loss, by total weight or volume, is not linear with time for crude oils and other multi-component fuel mixtures.

Evaporation of a liquid can be considered as the movement of molecules from the surfaceinto the vapour phase above it. The layer of air above the evaporation surface is known as theboundary layer. The characteristics of this air layer or boundary layer can influence evaporation. In the case of water, the boundary layer regulates the evaporation rate. Air can hold a variableamount of water, depending on temperature, as expressed by the relative humidity. Underconditions where the boundary layer is not moving (no wind) or has low turbulence, the airimmediately above the water quickly becomes saturated and evaporation slows or ceases. Inpractice, the actual evaporation of water proceeds at a small fraction of the possible rate becauseof the saturation of the boundary layer. The boundary layer physics is then said to regulate theevaporation of water. This regulation manifests itself in the sensitivity of evaporation to wind orturbulence. When turbulence is weak or absent, evaporation can slow down by orders-of-magnitude. The molecular diffusion of water molecules is at least 10 times slower than3

turbulent diffusion.If the evaporation of oil was like that of water and was boundary-layer regulated one could

write the mass transfer rate in semi-empirical form (also in generic and unitless form) as:

50

where E is the evaporation rate in mass per unit area, K is the mass transfer rate of theevaporating liquid, presumed constant for a given set of physical conditions, sometimes denoted

gas k (gas phase mass transfer coefficient, which may incorporate some of the other parameters

unoted here), C is the concentration (mass) of the evaporating fluid as a mass per volume, T is afactor characterizing the relative intensity of turbulence, and S is a factor that relates to thesaturation of the boundary layer above the evaporating liquid. The saturation parameter, S,represents the effects of local advection on saturation dynamics. If the air is already saturatedwith the compound in question, the evaporation rate approaches zero. This also relates to thescale length of an evaporating pool. If one views a large pool over which a wind is blowing,there is a high probability that the air is saturated downwind and the evaporation rate per unitarea is lower than for a smaller pool. It should be noted that there any many equivalent ways ofexpressing this fundamental evaporation equation.

Much of the pioneering work for evaporation work was performed by Sutton. Suttonproposed an equation based largely on empirical work:

swhere C is the concentration of the evaporating fluid (mass/volume), U is the wind speed, d isthe area of the pool, Sc is the Schmidt number and r is the empirical exponent assigned valuesfrom 0 to 2/3. Other parameters are defined as above. The terms in this equation are analogous tothe very generic equation, (51), proposed above. The turbulence is expressed by a combinationof the wind speed, U, and the Schmidt number, Sc. The Schmidt number is the ratio of kinematicviscosity of air (í) to the molecular diffusivity (D) of the diffusing gas in air, i.e., a dimension-less expression of the molecular diffusivity of the evaporating substance in air. The coefficient ofthe wind power typifies the turbulence level. The value of 0.78 (7/9) as chosen by Sutton,represents a turbulent wind whereas a coefficient of 0.5 would represent a wind flow that wasmore laminar. The scale length is represented by d and has been given an empirical exponent of-1/9. This represents, for water, a weak dependence on size. The exponent of the Schmidtnumber, r, represents the effect of the diffusivity of the particular chemical, and historically wasassigned values between 0 and 2/3.

This expression for water evaporation was subsequently used by those working on oilspills to predict and describe oil and petroleum evaporation. Much of the literature follows thework of Mackay. Mackay and co-workers adapted the water equations for hydrocarbons using theevaporation rate of cumene (Fingas, 1995). Data on the evaporation of water and cumene wereused to correlate the gas phase mass transfer coefficient as a function of wind-speed and poolsize by the equation:

m K = 0.0292 U X Sc (53) 0.78 -0.11 -0.67

mwhere K is the mass transfer coefficient in units of mass per unit time and X is the pooldiameter or the scale size of evaporating area. Stiver and Mackay subsequently developed thisfurther by adding a second equation:

51

m N = k AP/(RT) (54)

mwhere N is the evaporative molar flux (mol/s), k is the mass transfer coefficient at the prevailingwind (m/s), A is the area (m ), P is the vapour pressure of the bulk liquid (Pascals), R is the gas2

constant [8.314 Joules/(mol-K)], and T is the temperature (K). Thus, boundary layer regulation was assumed to be the primary regulation mechanism for

oil and petroleum evaporation. This assumption was not tested by experimentation, as revealedby the literature search (Fingas, 1995). The implications of these assumptions are thatevaporation rate for a given oil is increased by:

- increasing turbulence- increasing wind speed, and- increasing the surface area of a given mass of oil.Experiments on evaporation were conducted from 1992 to 1998 (Fingas, 2004). The

experimental series are summarized below:1. Wind Experiments Experiments on the evaporation of oil with and without wind were conducted with ASMB

(Alberta Sweet Mixed Blend), Gasoline, and with water. Water formed a baseline data set sincemuch is known about its evaporation behaviour. Regressions on the data were performed and theequation parameters calculated. Curve coefficients are the constants from the best fit equation[Evap = a ln(t)], t = time in minutes, for logarithmic equations or Evap = a /t, for the square rootequations. Oils with few components evaporating at one time or components with similarevaporation behaviour, have a tendency to fit square root curves (Fingas, 2004). A comparison ofoil and water evaporation clearly shows that water evaporation rate increased, as expected, withincreasing wind velocity. The oils, ASMB and gasoline, did not show a significant increase withincreasing wind speed. Any increase may only be a result of the increase in evaporation in goingfrom the 0-wind level to the other levels. In any case, they do not show the U relationship that0.78

water shows, as in equation (3). These data show that oil is not boundary-layer regulated.2. Evaporation Rate and AreaA light crude oil was also used to conduct a series of experiments with varying

evaporation area. The mass of the oil was kept constant so that the thickness of the oil would alsovary. However, the greater the area, the lesser the thickness and both factors would increase oilevaporation if it were boundary-layer regulated. Because of the lack of correlation between areaand evaporation rate, one can conclude that evaporation rate is not highly correlated with areaand that the evaporation of oil is not boundary-layer regulated.

3. Study of Mass and Evaporation RateLight crude oils were again used to conduct a series of experiments with volume as the

major variant. Alternatively, thickness and area were held constant to ensure that the strictrelationship between these two variables did not affect the final regression results A strongcorrelation between oil mass (or volume) and evaporation rate was shown. This suggests no orlittle boundary-layer regulation.

4. Study of the Evaporation of Pure Hydrocarbons - With and Without WindA study of the evaporation rate of pure hydrocarbons was conducted to test the classic

52

boundary-layer evaporation theory as applied to the hydrocarbon constituents of oils. Thisexperiment showed that the evaporation rates of the pure hydrocarbons have a variable responseto wind. Heptane (hydrocarbon number 7) shows a large difference between evaporation rate inwind and no wind conditions, indicating boundary-layer regulation. Decane (carbon number 10)shows a lesser effect and hexadecane (carbon number 16) shows a negligible difference betweenthe two experimental conditions. This experiment shows the extent of boundary-regulation andthe reason for the small or negligible degree of boundary-regulation shown by crude oils andpetroleum products. Crude oil contains very little material with carbon numbers less than decane,often less than 3% of its composition. Even the more volatile petroleum products, gasoline anddiesel fuel only have limited amounts of compounds more volatile than decane, and thus are alsonot strongly boundary-layer regulated.

Another consideration of evaporation regulation is that of saturation concentration, themaximum concentration soluble in air. In fact saturation concentration of water is less than thatof many common oil components (Fingas, 2004). The saturation concentration of water is infact, about two orders of magnitude less than the saturation concentration of volatile oilcomponents such as pentane. This further explains why oil has no boundary layer limitationcompared to that of water.

Several studies showed that unique equations for each oil at each temperature can bedeveloped (Fingas, 2004).

Further examination of the temperature behaviour of oil evaporation was conducted bydetermining the equations by which the evaporation rates, or equation parameters, change withtemperature. A series of correlations was performed, between the evaporation rates, by bothpercentage and weight loss, using a linear equation.

The empirically measured parameters at 15 C were correlated with both the slopes and theo

intercepts of the temperature equations. Full details of this correlation are given in the literature(Fingas, 1996). The resulting equations are:

Percentage evaporated = [B + 0.045(T-15)] (55)

where B is the equation parameter at 15 C and T is temperature in degrees Celsius.o

The finding that unique equations may be needed for each oil is a significant disadvantageto practical end use, and a way to accurately predict evaporation using other readily-availabledata is necessary. One way to do so was to use distillation data which are very common. Newprocedures to measure distillation data are very simple, fast and repeatable (Jokuty et al., 1999).

Distillation data were directly correlated to the evaporation rates determined byexperimentation (Fingas, 2004). The percentage distilled at each temperature was correlated withthe equation parameter (sometimes referred to here as the evaporation rate). Detailed correlationdata are given in the literature. The percent mass distilled at 180 degrees was used to calculatethe relationship between the distillation values and the equation parameters. The equations usedwere derived from correlations of the data.

The data from those oils that were better fitted with square root equations, such as diesel,Bunker C light and FCC Heavy Cycle, were calculated separately. The equations derived for thepredication are:

53

For oils that follow a logarithmic equation:

Percentage evaporated = [.165(%D) + .045(T-15)]ln(t) (56)

For oils that follow a square root equation:

Percentage evaporated = [.0254(%D) + .01(T-15)]/t (57)

where %D is the percentage (by weight) distilled at 180 C.o

Since the distillation data on the new oil is known and at 180 C, 22.3 % is distilled, theo

evaporation equation for ANS now becomes:

Percentage evaporated = [3.63 + 0.45(T-15)]ln(t) (58)Figure 6 shows this equation applied to predict evaporation at various temperatures.

Figure 6 The Variation in Evaporation Amounts with Temperature

54

Figure 7 shows the comparison to the older empirical equations measured for ANS orANS types. As can be seen from Figure 7, the latest sample is much lighter and would evaporatemuch more extensively at a given temperature.

Figure 7 The Variation in Evaporation Curves in ANS Samples over the Years

In the past, a large number of experiments were performed on oils to directly measure theirevaporation curves. The empirical equations that result are given in the literature and some willbe summarized here as well (Fingas, 2004).

The purpose of this analysis is to present findings of another simple means to calculate theevaporation rate or parameter B in equation (55). The evaporation parameters (beforetemperature correction) are the simplest way to proceed. From the work described above, theseparameters are given as the B above (equation 55)) and -0.675 for oils that follow a logarithmiccurve and -0.195 for oils that follow a square root curve. Data on a number of oil evaporationrates, done experimentally and basic properties of oil were collected, primarily from a previouspaper, and also from literature sources (Jokuty et al., 1999; Oil Catalogue, 2009). These data areshown in Table 1.

While it is known that density and viscosity are not necessarily good indicators of other oilproperties, these may be the only data available at the site of a spill, at least until some timepasses (Jokuty et al., 1995). The data were correlated with the evaporation parameter with theobjective of deriving a simplified model using oil density and viscosity or SARA data. Thecurve-fitting statistical programs, TableCurve 3D and Datafit, were used. Models were judged onthe basis of goodness-of-fit (regression coefficient or r ) for the simplest equation. It should be2

55

noted that one can sometimes fit large-polynomial equations to even unrelated data, therefore it isessential to use judgement and chose a realistic equation out of the many that are available.

The curve fitting exercise with the logarithmic evaporation data and density and viscositydata resulted in a best-fit simplest equation:

vapE = 15.4 -14.5*density + 2.58/viscosity (59)

vapWhere: E is the evaporation equation parameter and as in equation (55) is B - 0.675 tocompensate for the temperature parameter as explained above.

The results of the correlation and the prediction equation are shown in Table 2. Theregression coefficient (r ) was 0.95 and the average standard deviation between the input2

equation parameters and those predicted was 0.145. Similarly the data for those oils that evaporate as a square root with time were correlated

with density and viscosity. The regression coefficient (r ) was 0.79 and the average standard2

deviation between the input equation parameters and those predicted was 0.37. It should be notedthat the fit with the oils that evaporate as a square root with time, is not as good as with the otheroils. The reason for this is not apparent.

The equation for predicting those oils that evaporate as a square root of time is:

vapE = 2.3 -1.47*density - 0.073*ln(viscosity) (60)

vapWhere: E is the evaporation equation parameter.This process will not be used here because it is somewhat less accurate than equation (58)

above, however it does yield a similar equation or:Percentage evaporated = [3.09 + 0.45(T-15)]ln(t) (61)Other data that may be available are SARA or Saturate Aromatic Resins and Asphaltene

data. The saturate content correlates somewhat with the equation parameter. Statistical analysisshows that only the saturate content correlates and this does not so well (r < 0.6). The addition of2

the other SARA data only marginally assists in the correlation. However, an equation for theprediction of the evaporation parameter using SARA data is:

vap s E = 0.048*S +0.006*A -0.025*R -0.0003A + 0.27 (62)

vapWhere: E is the evaporation equation parameter,S is the saturate content in percent,A is the aromatic content in percent,R is the resin content in percent, and

sA is the asphaltene content in percent.

Using this prediction scheme one gets the following evaporation equation: Percentage evaporated = [2.86 + 0.45(T-15)]ln(t) (63)

This equation is lower than (58) above and would be less accurate, however it is within anorder-of-magnitude of the better equations.

56

EmulsificationEmulsification is the process whereby water-in-oil emulsions are formed. These emulsions

are often called “chocolate mousse” or “mousse” by oil spill workers. When emulsions form, thephysical properties and characteristics of oil spills change extensively. Many researchers feel thatemulsification is the second most important behavioural characteristic of oil after evaporation.Emulsification has a significant effect on the behaviour of oil spills at sea. As a result ofemulsification, evaporation of oil spills slows by orders-of-magnitude, spreading slows bysimilar rates, and the oil rides lower in the water column, showing different drag with respect tothe wind. Emulsification also significantly affects other aspects of a spill, such as cleanupresponse. Spill countermeasures are quite different for emulsions as they are hard to recovermechanically, to treat, or to burn.

The basics of water-in-oil emulsification are now understood and well-established (Fingasand Fieldhouse, 2009). The basic principle is that water-in-oil emulsions are stabilized by theformation of asphaltene layers around water droplets in oil. The layer or film is characterized byits high-strength visco-elastic properties. Resins also form emulsions, to a certain extent, butresins do not form stable emulsions, and may aid in asphaltene emulsion stability by acting asasphaltene solvents and by providing temporary stability during the time of the slow asphaltenemigration. The extent of knowledge on water-in-oil emulsion formation to this date is stronglylimited by the amount of knowledge of asphaltenes and resins themselves. There is still verylimited knowledge on the structure and chemistry of asphaltenes (Fingas and Fieldhouse, 2009).

Fingas and Fieldhouse (2006) proposed a model that takes into account severalcompositional factors in the oil. The first step in this model is to obtain or estimate the oilproperties as they are at the weathering condition of concern. The properties needed are thedensity, viscosity, and saturate, resin and asphaltene contents. These values requiretransformation as summarized below:

Density: If the density is less than 0.97, then the density parameter is 0.97 less the densityand if it is greater than 0.97, it becomes the density less 0.97. The value used in the equation isthis transformed value. (64)


Resin Content: If the value of the resins is zero, then set this value to 20. If the resincontent is less than 5.4, then the resin content parameter is 5.4 less the resin content and if it isgreater than 5.4, it becomes the resin content less 5.4. The value used in the equation is thetransformed value. (66)

Asphaltene Content: If the value of the asphaltene content is zero, then set the value to 30.If the asphaltene content is less than 12, then the asphaltene content parameter is 12 less theasphaltene content and if it is greater than 12, it becomes the asphaltene content less 12. Thevalue used in the equation is this transformed value. (67)


57

Class = 0.738 - 0.197*Dt - 0.0126*Vt -0.0007*Rt - 0.00358*At (68)

Where Class is the numerical index of classificationDt is the transformed density as calculated in equation (64)Vt is the transformed viscosity as calculated in equation (65)Rt is the transformed resin content as calculated in equation (66)At is the transformed asphaltene content as calculated in equation (67)

The class is then assigned according to the following scheme (Table 6):

Table 6 Emulsion Classification for Model

category class number

unstable <0.615

mesostable >0.615 <0.66

stable >0.66

entrained >.64 <.72

but density >0.96 or viscosity >10,000

Using this classification scheme, one obtains for the fresh oil, a value of 0.60 which meansthat the fresh emulsion would be unstable. Using estimations of the properties after weathering,one estimates that a meso-stable emulsion is then produced.

Note that the entrained state overlaps the other states, but given that the density of the oilis greater than 0.96 mg/mL or that the viscosity of the oil is greater than 10,000 mPa.s, then theconditions for the entrained state apply.

Applying this equation to the new sample, The kinetics can be modelled as the time to formation after the oil is weathered to the

stated percentage:


Where: a is a constant and is 27.1 for a stable emulsion formation, 47 for meso and 30.8for an entrained water-in-oil class

b is a constant and is 7520 for a stable emulsion formation, 49100 for meso and 18300for an entrained water-in-oil classWh is the wave height in cm

It is difficult to model emulsification using only density and viscosity, however, this isdefinitely better than using no estimator or an old water-uptake model. To perform thiscorrelation the same data set was used as above (Fingas and Fieldhouse, 2006). The procedurewas also the same, using only the values of density and viscosity. Many iterations are required toachieve the greatest amount of discrimination between states.


Class = -21.4 + 28*D -0.00000073*V (70)

58

Where Class is the numerical index of classificationD is the densityV is the viscosity

The class is then assigned according to the following scheme (Table 7):

Table 7 Classification Scheme for

Emulsions

Category Class Value

Unstable <2.3

Mesostable 2.3 to 3.6

Stable >3.6

Entrained density >9.6

viscosity >10,000

The fit statistics on this estimation technique are not as good as on the full modeldescribed above. The state mis-classifications are 58% for the simple model compared to onlyabout 20% for the full model. The regression coefficient for the correlation was 0.81 for thesimple model.

Using this estimation technique for the current oil one obtains a class of 2.74 ormesostable for the fresh oil and similar for the weathered oil. This is a less accurate means ofestimation.

The older modeling method as noted in (68) above, used an assigned number for eachwater-in-oil type and then this was regressed with the various oil starting properties. Further, thefunctional relationship (square, log, exponential, etc.) of the relationship of the starting oilproperties was determined before hand and then correlated with a multi-parameter linear methodto yield a predictor model. Since a stability index has been established, this can be used as thetarget of the correlation (Fingas and Fieldhouse, 2009). This new modeling has a differentapproach than the old one. Furthermore, this new approach used a multi-regression programdirectly, using various multi-functional transformations of the input oil property data. This willallow the software of the regression model to assign portions of the functions necessary toachieve the highest correlation factor. Other differences are also that the transformations areautomated using a peak function fit, yielding slightly different transformation values than the oldvalues; use of a newer data set; and incorporation of several functional input parameters,dependent on the statistics of the model, into the final equation.

A transformation is needed to adjust the data to a singular increasing or decreasingfunction. Regression methods will not respond correctly to a function that varies both directlyand inversely with the target parameter. Most parameters have an optimal value with respect toclass, that is the values have a peak function with respect to stability or class. The arithmeticconverts values in front of the peak to values behind the peak, thus yielding a singular decliningor increasing function. The optimal value of this manipulation is found by using a peak functionas shown by the line in Figure 1. This peak function fit is available from TableCurve, regressionsoftware. In the past this was achieved by trial and error, beginning with the estimated peak from

59

the first correlation. The accuracy of the new method is obviously much greater. It is important to note that the scatter observed in looking at any multi-parameter fit are

typical of scatter observed for a single parameter when a large number of parameters areoperative. If one plotted the total number of parameters simultaneously (eg. in multi-dimensions),then a singular function would appear with much less scatter.

The arithmetic to perform the transformation is: if the initial value is less than the peakvalue, then the adjusted value is the peak value less the initial value; and if the initial value ismore than the peak value, the adjusted value is the initial value less the peak value. It should benoted that the exponential of density was used and the natural log of the viscosity. Previousmodeling work had shown that these mathematical changes are necessary to achieve highercorrelations.

Having the transformed values, the new model proceeded directly to fit a multiplelinear equation to the data. The choice of functions was achieved by correlating the stabilityfunction directly with the data and taking the best of the functions (eg. square, log, etc) furtherinto the regression process. The functionalities of square, logarithmic or exponential curves areachieved by correlating the straight value of the input properties plus their expanded values,taken here as the cube of the starting parameter as well as the square of the exponential of thestarting value; and their companded values, the natural log (ln) and the logarithm (base 10) ofparameter divided by the square of the value. Thus each parameter is correlated with the stabilityC index in five sets of mathematical statements. This is similar to the standard Gaussianexpansion regression technique. In this method the regression is expanded to functionalitiesabove and below linear until the entire entity is optimized. For example a linear function wouldbe included, then a square and then a square root and so on until tests of the complete regressionshow that there are no more gains in increased expansions. Using this technique, six inputparameters: exponential of density, ln of viscosity, saturate content, resin content, asphaltenecontent and the asphaltene/resin ratio (A/R) were found to be optimal. Thus with 4transformation and the original values of these input parameters, there are six times five or 30input combinations.

Using Datafit, a multiple regression software, a maximum of 20 of these could be taken ata time to test the goodness-of-fit. Values that yield Prob(t) factors of greater than 0.9 weredropped until all remaining factors could be calculated at once. The Prob(t) is the probability thatinput can be dropped without affecting the regression or goodness-of-fit. Over twenty regressionswere carried out until the resulting model was optimal. The r , the regression coefficient, was2

0.75, which is quite high considering the many potential sources of error, etc. The statistics onthe new model are shown in Table 3, along with the parameters to create the model. Table 3shows that the 14 remaining parameters all contribute to the accuracy of the final result and thatnone of them can be cut without affecting the outcome of the model. The procedures forcalculating, using the new model are given below.

The first step is to transform the input data so that it forms a continuous declining orincreasing function.

Density: Take the exponential of the density. If the exponential of density is less than2.5, then the density parameter is 2.5 less the density and if it is greater than 2.5, it becomes thedensity less 2.5. The value used in the equation is this transformed value. (71)

60


Saturate Content: If the saturate content is less than 45, then the saturate content parameteris 45 less the saturate content and if it is greater than 45, it becomes the saturate content less 45.The value used in the equation is the this transformed value. (73)

Resin Content: If the value of the resins is zero, then set this value to 20. If the resincontent is less than 10, then the resin content parameter is 10 less the resin content and if it isgreater than 10, it becomes the resin content less 10. The value used in the equation is the thistransformed value. (74)

Asphaltene Content: If the value of the asphaltene content is zero, then set the value to 20.If the asphaltene content is less than 4, then the asphaltene content parameter is 4 less theasphaltene content and if it is greater than 4, it becomes the asphaltene content less 4. The valueused in the equation is this transformed value. (75)

A/R or Asphaltene/Resin Ratio: This is taken as the direct value of the asphaltene contentin percent (untransformed) divided by the resin content in percent (again untransformed). If A/Ris less than 0.6, then the A/R parameter is 0.6 less the A/R and if it is greater than 0.6, it becomesthe A/R less0.6. The value used in the equation is this transformed value. (76)


Stability C = 12.3 + 0.259*St - 1.601*Rt - 17.2*A/Rt - 0.50*Vt + 0.002*Rt 3 3

+ 0.001*At + 8.51*A/Rt - 1.12*lnVt + 0.700*lnRt + 2.97*lnA/R3 3

+ 6.0X10 *ExplnVt -1.96*ExpA/Rt - 4.0X10 *logExpDt/expDt-8 2 2 -6 2

- 1.5X10 *logA/Rt/A/Rt (77)-4 2

Where Class is the Stability C of the resulting water-in-oil type,St is the transformed saturate content as calculated in equation (78), abbreviated Ahere, Rt is the transformed resin content as calculated in equation (71), abbreviated

B,A/Rt is the transformed asphaltene/resin ratio as calculated in (73), abbreviated

C,Vt is the cube of the transformed ln viscosity as calculated in (72),3

abbreviated D,Rt is the cube of the transformed resin content as calculated in equation (74),3

abbreviated E,At is the cube of the transformed asphaltene content as calculated in equation (75),3

abbreviated F, A/Rt is the cube of the transformed A/R ratio as calculated in equation (76),3

abbreviated G,lnVt is the natural logarithm (ln) of the transformed viscosity as calculated in equation

(72), abbreviated H,

61

lnRt is the natural logarithm (ln) of the transformed resin content as calculated inequation (74), abbreviated I,

lnA/R is the natural logarithm (ln) of transformed asphaltene/resin ratio as calculatedin (71), abbreviated J,

ExplnVt is the is the exponential of the transformed viscosity - squared - as2

calculated in equation (72), abbreviated K, ExpA/Rt is the is the exponential of the A/R ratio - squared - as calculated in2

equation (76), abbreviated L,logExpDt/expDt is the is the logarithm (base 10) of exponential of the density -2

divided by the square of the transformed density - the transformed density ascalculated in equation (71), abbreviated M,

logA/Rt/A/Rt is the is the logarithm (base 10) of exponential of the A/R ratio -2

divided by the square of the A/R ratio - the transformed A/R ratio as calculated inequation (76), abbreviated N,

A simplified version of the equation is then:Stability C = 12.3 + 0.259A - 1.601B - 17.2C - 0.50D + 0.002E + 0.001F + 8.51G - 1.12H +2.97J + 0.700I + 6.0X10 K -1.96L - 4.0X10 M - 1.5X10 N (78)-8 -6 -4

The values of Stability C which are assigned to each class are given in Table 8. The viscosity ofthe resulting product can be taken as the average of the types at a given time.

Using this method of estimation, one obtains -8.3 for the fresh oil, and -4 for a weatheredoil. This means that the fresh oil would produce an unstable emulsion and a more weathered on ameso-stable emulsion. It should be noted that these estimations are similar to the very simpleestimator above.

Dispersant EffectivenessIn an earlier work, a large set of data was correlated with over 15 oil composition and

properties characteristics to produce several prediction models (Fingas et al., 2003b). It was

62

found that the parameters that correlate most highly with the suite of parameters are thosecomposition parameters that relate to smaller compounds in the oil. These include n-C12,naphthalenes, and the sum of the C12 to C16 components. Those that relate to the largecompounds in the oil relate negatively to the dispersibility, including C26 and resins. Thisindicates that dispersion primarily affects only the smaller components of the oil.

The highest correlation was achieved with the n-C12 component. The next highestcorrelation was for the naphthalene content. This also indicates a high dispersibility fornaphthalene. The third highest correlation is for n-C26 and this is an inverse correlation. Thisindicates that the more n-C26, the less dispersion, especially in the initial amount. This indicatesthat components of the size of C26 and greater are poorly dispersed and in fact inhibit dispersion.The fourth highest correlation is the PAH content and this correlates positively, namely that thehigher the PAH content, the higher the dispersion. The fifth highest correlation is that of the sumof the C12, C14, C16, and C18 components. This correlation is highly indicative that alkanes upto C18 are the prime components dispersed along with the PAHs. The fact that C12 correlates thehighest of these n-alkanes and that this correlation rapidly drops off to C18 with no useablecorrelation for C20 indicates that primarily hydrocarbons up to C18 disperse and that past C20compounds disperse poorly.

It should be noted that most of the dispersability data are available for Corexit 9500, bothin the past and present study. Some data in the past study were available for Corexit 9527,Enersperse 700, Dasic LTS and a few others.

The past correlation is summarized above. A new correlation was performed using onlydensity and viscosity data. Correlation was performed directly using TableCurve 3D software.One iteration was carried out to remove or correct bad data.

The model achieved is:Dispersibility = 169 - 171*D + 62.4/V (79)

Where Dispersability is the Corexit 9500 dispersibility in percentD is the density, andV is the viscosity.

The average standard deviation was 3.2 and the r for the regression model was 0.82. It2

should be noted that this model uses the same inputs as the old model 4 above, however adifferent equation resulted. Both equations yield roughly the same variance in predictioncapability. The model in equation (79) does not produce as accurate a prediction as many of themodels noted above, however is much better than no model. In this case, the value ofdispersibility predicted is with a 30% error of the actual value. One can compare this to model11, noted above which produces predications within 5%, but requires detailed compositional datainput.

For the new oil this yields a dispersibility of 26%, which is quite a typical number fortests carried out in the past on similar oils.

As noted in the development sections, these simplified models will yield satisfactoryresults until further detailed oil composition data are available. The addition of detailed oilcomposition data improves the accuracy of the predictions.

63

7 SummaryImportant tools in oil spill planning and response is oil spill modeling and prediction.

Decidedly, the most important estimation and modeling algorithms are that for oil spillemulsification, evaporation, chemical dispersibility and those that might be used to predict othercountermeasures such as recovery and burning. This paper showed that the new data for ANScould be used to predict its behaviour.

The evaporation is best predicted by the equation:

Percentage evaporated = [3.63 + 0.45(T-15)]ln(t) (58)

The emulsion formation predictions show that as a fresh oil, it will not produce a water-in-oil emulsion and that when highly weathered would produce a meso-stable emulsion.

The dispersibility of the oil is predicted to be 26%.

These numbers are consistent with the properties of the oil. It should be noted, howeverthat the oil appears to be much lighter and of a slightly different composition than former oilsfrom the same field.

64

8 ReferencesAcevedo, S., G. Escobar, L.B. Gutierrez, and H. Rivas, “Interfacial Rheological Studies of Extra-Heavy Crude Oils and Asphaltenes: Role of the Dispersion Effect of Resins in the Adsorption ofAsphaltenes at the Interface of Water-in-Crude Oil Emulsions”, Colloids and Surfaces A, Vol.71, pp 65-71, 1993.

Allen, A., “Controlled Burning of Crude Oil on Water following the Grounding of the ExxonValdez, Proceedings of the 1991 Oil Spill Conference, American Petroleum Institute,Washington, DC, pp 213-216, 1991.

Aomari, N., R. Gaudu, F. Cabioc’h, and A. Omari, “Rheology of Water in Crude Emulsions”,Colloids and Surfaces A, Vol. 139, pp 13-20, 1998.

Balinov, B., O. Urdahl, O. Soderman, and J. Sjöblom, “Characterization of Water-in-Crude OilEmulsions by the NMR Self-Diffusion Technique”, Colloids and Surfaces A, Vol. 82, pp 173-181, 1994.

Berger, D. and D. Mackay, “The Evaporation of Viscous or Waxy Oils: When is a Liquid-PhaseResistance Significant?”, in Proceedings of the Seventeenth Arctic Marine Oilspill ProgramTechnical Seminar, Environment Canada, Ottawa, ON, pp 77-92, 1994.

Berridge, S.A., R.A. Dean, R.G. Fallows, and A. Fish, “The Properties of Persistent Oils at Sea”,Journal of the Institute of Petroleum, Vol. 54, The Institute of Petroleum, London, UK, pp 300-309, 1968.

Bobra, M., “A Study of the Formation of Water-in-Oil Emulsions”, in Proceedings of theThirteenth Arctic Marine Oilspill Program Technical Seminar, Environment Canada, Ottawa,ON, pp 87-117, 1990.

Bobra, M., M. Fingas, and E. Tennyson, “When Oil Spills Emulsify”, Chemtech, April, pp 214-236, 1992.

Børve, K.G.N., J. Sjöblom, and P. Stenius, “Water-in-Crude Oil Emulsions from the NorwegianContinental Shelf: 5. A Comparative Monolayer Study of Model Polymers”, Colloids andSurfaces, Vol. 63, pp 241-251, 1992.

Bridie, A.L., T.H. Wanders, W. Zegveld, and H.B. Van Der Heijde, “Formation, Prevention andBreaking of Sea Water in Crude Oil Emulsions 'Chocolate Mousses'”, Marine Pollution Bulletin,Vol. 11, pp 343-348, 1980.

Chaala, A., B. Benallal, and S. Hachelef, “Investigation on the Flocculation of Asphaltenes andthe Colloidal Stability of the Crude Oil Fraction (>210 C)”, Canadian Journal of Chemicalo

Engineering, Vol. 72, pp 1036-1041, 1994.

65

Desmaison, M., C. Pierkarski, S. Pierkarski, and J.P. Desmarquest, “Formation et stabilisationdes emulsions inverses eau de mer-petrole : Role de la tension et de la viscosité d’interface”,Revue de l'Institute française du petrole, Vol. 39, No. 5, pp 603-615, 1984.

Durell, G.S., S.A. Ostazeski, A.D. Uhler, I.K. Almas, P.S. Daling, T. Strom-Kristiansen, andA.B. Nordvik, “Evaluation of the Transfer of Crude Oil Weathering Technology: InterlaboratoryComparison of Physico-Chemical Characteristics of Weathered Oils and Emulsions”, inProceedings of the Seventeenth Arctic Marine Oilspill Program Technical Seminar, EnvironmentCanada, Ottawa, ON, pp 39-75, 1994.

Ebeltoft, H., K.G.N. Børve, J. Sjöblom, and P. Stenius, Progress in Colloids and PolymerScience, Vol. 88, pp 131-139, 1992.

Eley, D.D., M.J. Hey, and J.D. Symonds, “Emulsions of Water in Asphaltene-containing Oils: 1.Droplet Size Distribution and Emulsification Rates”, Colloids and Surfaces, Vol. 32, pp 87-101,1988.

Ese, M.-H., J. Sjöblom, H. Førdedal, O. Urdahl, and H.P. Rønningsen, “Ageing of InterfaciallyActive Components and Its Effect on Emulsion Stability as Studied by Means of High VoltageDielectric Spectroscopy Measurements”, Colloids and Surfaces A, Vol. 123-124, pp 225-232,1997.

Ese, M., X. Yang, and J. Sjöblom, “Film Forming Properties of Asphaltenes and Resins. AComparative Langmuir-Blodgett Study of Crude Oils from North Sea, European Continent andVenezuela”, Colloids and Polymer Science, Vol. 276, pp 800-809, 1998.

ETC (Environmental Technology Centre), World Catalogue of Oil Properties, WWW.ETC-CTE.ec.gc.ca, 2009.

Fingas, M.F., “Studies on the Evaporation of Oil Spills”, in Proceedings of the SeventeenthArctic and Marine Oil Spill Program Technical Seminar, Environment Canada, Ottawa, ON, pp189-212, 1994.

Fingas, M.F., “A Literature Review of the Physics and Predictive Modelling of Oil SpillEvaporation”, Journal of Hazardous Materials, Vol. 42, pp 157-175, 1995.

Fingas, M.F., “Studies on the Evaporation of Crude Oil and Petroleum Products: I. TheRelationship between Evaporation Rate and Time”, Journal of Hazardous Materials, Vol. 56, pp227-236, 1997.

Fingas, M.F., “Studies on the Evaporation of Crude Oil and Petroleum Products: II. BoundaryLayer Regulation”, Journal of Hazardous Materials, Vol. 57, pp 41-58, 1998.

66

Fingas, M.F., The Basics of Oil Spill Cleanup: Second Edition, CRC Press, Boca Raton, FL,256 p, 2000.

Fingas, M., “Modeling Evaporation Using Models That Are Not Boundary-layer Regulated”,Journal of Hazardous Materials, Vol. 107, pp. 27-36, 2004.

Fingas, M.F. and B. Fieldhouse, “Studies of Water-in-Oil Emulsions and Techniques to MeasureEmulsion Treating Agents”, in Proceedings of the Seventeenth Arctic and Marine Oil SpillProgram Technical Seminar, Environment Canada, Ottawa, ON, pp 213-244, 1994.

Fingas, M. and B. Fieldhouse, “Water-in-Oil Emulsions and Application to Oil Spill Modelling”,in Proceedings of the Fifth International Marine Environmental Modelling Seminar, SintefApplied Chemistry, Trondheim, Norway, pp 339-351, 2001.

Fingas, M.F. and B. Fieldhouse, “Studies of the Formation Process of Water-in-Oil Emulsions”,Marine Pollution Bulletin, Vol. 47, pp 369-396, 2003.

Fingas, M.F., Z. Wang, B. Fieldhouse, and P. Smith, “The Correlation of ChemicalCharacteristics of An Oil to Dispersant Effectiveness”, in Proceedings of the Twenty-Sixth Arcticand Marine Oil Spill Program Technical Seminar, Environment Canada, Ottawa, Ontario, pp.679-730, 2003.

Fingas, M. and B. Fieldhouse, “Formation of Water-in-oil Emulsions and Application to Oil SpillModelling”, Journal of Hazardous Materials, Vol. 107, pp. 37-50, 2004.

Fingas, M.F. and B. Fieldhouse, “Modelling of Water-in-Oil Emulsions”, in Proceedings of theTwenty-seventh Arctic and Marine Oil Spill Program Technical Seminar, Environment Canada,Ottawa, ON, pp 335-350, 2004a.

Fingas, M.F. and B. Fieldhouse, “A New Approach to Modelling Water-in-Oil Emulsions”, inProceedings of the Seventh International Marine Environmental Modeling Seminar, Washington,DC, SINTEF, Trondheim, Norway, pp. iii 32-53, 2004b.

Fingas, M.F. and B. Fieldhouse, “Formation of Water-in-oil Emulsions and Application to OilSpill Modelling”, Journal of Hazardous Materials, Vol. 107, pp. 37-50, 2004c.

Fingas, M.F. and B. Fieldhouse, “How to Model Water-in-Oil Emulsion Formation” inProceedings of the 2005 International Oil Spill Conference, American Petroleum Institute,Washington, DC, 2005.

Fingas, M. and B. Fieldhouse, “Studies on Crude Oil and Petroleum Product Emulsions: WaterResolution and Rheology”, Colloids and Surfaces A: Physicochemical and Engineering Aspects,Vol. 333, pp. 67-81, 2009.

67

Fingas, M., B. Fieldhouse, M. Bobra, and E. Tennyson, “The Physics and Chemistry ofEmulsions”, in Proceedings of the Workshop on Emulsions, Marine Spill Response Corporation,Washington, DC, 1993.

Fingas, M.F, B. Fieldhouse, L. Gamble, and J.V. Mullin, “Studies of Water-in-Oil Emulsions:Stability Classes and Measurement”, in Proceedings of the Eighteenth Arctic and Marine OilSpill Program Technical Seminar, Environment Canada, Ottawa, ON, pp 21-42, 1995.

Fingas, M.F., B. Fieldhouse, and J.V. Mullin, “Studies of Water-in-Oil Emulsions: The Role ofAsphaltenes and Resins”, in Proceedings of the Nineteenth Arctic and Marine Oil Spill ProgramTechnical Seminar, Environment Canada, Ottawa, ON, pp 73-88, 1996.

Fingas, M.F., B. Fieldhouse, and J.V. Mullin, “Studies of Water-in-Oil Emulsions: StabilityStudies”, in Proceedings of the Twentieth Arctic and Marine Oil Spill Program TechnicalSeminar, Environment Canada, Ottawa, ON, pp 21-42, 1997.

Fingas, M.F., B. Fieldhouse, and J.V. Mullin, “Studies of Water-in-Oil Emulsions: Stability andOil Properties”, in Proceedings of the Twenty-first Arctic and Marine Oil Spill ProgramTechnical Seminar, Environment Canada, Ottawa, ON, pp 1-25, 1998.

Fingas, M.F., B. Fieldhouse, and J.V. Mullin, “Studies of Water-in-Oil Emulsions: Energy Threshold of Emulsion Formation”, in Proceedings of the Twenty-second Arctic andMarine Oil Spill Program Technical Seminar, Environment Canada, Ottawa, ON, pp 57-68,1999.

Fingas, M.F., B. Fieldhouse, J. Lane, and J.V. Mullin, “Studies of Water-in-Oil Emulsions:Long-Term Stability, Oil Properties, and Emulsions Formed at Sea”, in Proceedings of theTwenty-third Arctic and Marine Oil Spill Program Technical Seminar, Environment Canada,Ottawa, ON, pp 145-160, 2000a.

Fingas, M.F., B. Fieldhouse, J. Lane, and J.V. Mullin, “Studies of Water-in-Oil Emulsions:Energy and Work Threshold for Emulsion Formation”, in Proceedings of the Twenty-third Arcticand Marine Oil Spill Program Technical Seminar, Environment Canada, Ottawa, ON, pp 19-36,2000b.

Fingas, M.F., B. Fieldhouse, L. Lerouge, J. Lane, and J. Mullin, “Studies of Water-in-OilEmulsions: Energy and Work Threshold as a Function of Temperature”, in Proceedings of theTwenty-fourth Arctic and Marine Oil Spill Program Technical Seminar, Environment Canada,Ottawa, ON, pp 65-77, 2001a.

Fingas, M.F., B. Fieldhouse, J. Lane, and J. Mullin, “Studies of Water-in-Oil Emulsions: Testingof Emulsion Formation in OHMSETT”, in Proceedings of the Twenty-fourth Arctic and MarineOil Spill Program Technical Seminar, Environment Canada, Ottawa, ON, pp 47-63, 2001b.

68

Fingas, M.F., B. Fieldhouse, P. Lambert, Z. Wang, J. Noonan, J. Lane, and J.V. Mullin, “Water-in-Oil Emulsions Formed at Sea, in Test Tanks, and in the Laboratory”, Manuscript Report EE-169, Environmental Protection Service, Environment Canada, Ottawa, ON, 2002a.

Fingas, M.F., B. Fieldhouse, J. Noonan, P. Lambert, J. Lane and J. Mullin, “Studies of Water-in-Oil Emulsions: Testing of Emulsion Formation in OHMSETT, Year II”, in Proceedings of theTwenty-fifth Arctic and Marine Oil Spill Program Technical Seminar, Environment Canada,Ottawa, ON, pp 29-44, 2002b.

Fingas, M.F., B. Fieldhouse and Z. Wang, “The Long Term Weathering of Water-in-OilEmulsions”, Spill Science and Technology Bulletin, Vol. 8, No. 2, pp 137-143, 2003a.

Fingas, M.F., Z. Wang, B. Fieldhouse, and P. Smith, “The Correlation of ChemicalCharacteristics of An Oil to Dispersant Effectiveness”, in Proceedings of the Twenty-Sixth Arcticand Marine Oil Spill Program Technical Seminar, Environment Canada, Ottawa, Ontario, pp.679-730, 2003b.

Førdedal, H., Y. Schildberg, J. Sjöblom, and J-L. Volle,“Crude Oil Emulsions in High ElectricFields as Studied by Dielectric Spectroscopy: Influence of Interaction between Commercial andIndigenous Surfactants”, Colloids and Surfaces, Vol. 106, pp 33-47, 1996a.

Førdedal, H., Ø. Midttun, J. Sjöblom, O.M. Kvalheim, Y, Schildberg, and J-L Volle, “AMultivariate Screening Analysis of W/O Emulsions in High External Electric Fields as Studiedby Means of Dielectric Time Domain Spectroscopy, II”, Journal of Colloid and InterfaceScience, Vol 182, pp 117-125, 1996b.

Friberg, S.E., “Emulsion Stability”, in Emulsions: A Fundamental and Practical Approach, J.Sjöblom (ed.), Kluwer Academic Publishers, Dordrecht, The Netherlands, pp 1-24, 1992.

Friberg, S. E. and J. Yang, “Emulsion Stability”, in Emulsions and Emulsion Stability, JohanSjöblom (ed.), Marcel Dekker, New York, NY, pp 1-40, 1996.

Gestblom, B., H. Førdedal, and J. Sjöblom, “W/O Emulsions in High Electric Fields as Studiedby Means of Dielectric Spectroscopy”, Journal of Dispersion Science and Technology, Vol. 15,pp 450-464, 1994.

Groenzin, H. And O.C. Mullins, “Asphaltene Molecular Size and Weight by Time-ResolvedFluorescence Depolarization”, Chapter 2 in Asphaltenes, Heavy Oils and Petroleomics, O.C.Mullins, E.Y. Sheu, A. Hammami, and A.G. Marshall, Editors, Springer Publications, NY, NY.,pp. 17-62, 2007.

Haegh, T. and T. Ellingsen, “The Effect of Breaking Waves On Oil Spills; 1. Emulsification ofCrude Oil at Sea”, SINTEF, The Foundation of Scientific and Industrial Research at the

69

University of Trondheim, Norway, 23 p, 1977.

Hartland, S. and S.A.K. Jeelani, “Effect of Interfacial Tension Gradients on Emulsion Stability”,Colloids and Surfaces A, Vol. 88, pp 289-302, 1994.

Haus, F., O. Boissel, and G.-A. Junter, “Primary and Ultimate Biodegradabilities of Mineral-based Oils and their Relationships with Oil Viscosity”, International Biodeterioration andBiodegradation, Vol. 54, pp 189-192, 2004.

Hollebone, B., Physical and Chemical Properties of Alaskan North Slope [2010] Crude Oil, areport for PWS RCAC, 2010.

International Petroleum Encyclopedia, PenWell Books, Tulsa, OK, 2004

Jenkins, R.H., S.J.W. Grigson, and J. McDougall, “The Formation of Emulsions at Marine OilSpills and the Implications for Response Strategies”, in Proceedings of the First InternationalConference on Health, Safety and Environment in Oil and Gas Exploration and Production,Vol. 2, Society of Petroleum Engineers, Richardson, TX, pp 437-443, 1991.

Johansen, E.J., I.M. Skjärvo, T. Lund, J. Sjöblom, H. Söderlund, and G. Boström, “Water-in-Crude Oil Emulsions from the Norwegian Continental Shelf: Part 1. Formation, Characterizationand Stability Correlations”, Colloids and Surfaces, Vol. 34, pp 353-370, 1988/89.

Jokuty, P., M. Fingas, S. Whiticar, and B. Fieldhouse, “A Study of Viscosity and InterfacialTension of Oils and Emulsions”, Manuscript Report EE-153, Environmental Protection Service,Environment Canada, Ottawa, ON, 43 p, 1995.

Jokuty, P., S. Whiticar, Z. Wang, M.F. Fingas, B. Fieldhouse, P. Lambert, and J. Mullin,“Properties of Crude Oils and Oil Products, (Volume 1, A-K; Volume 2, L-Z)”, ManuscriptReport EE-165, Environmental Protection Service, Environment Canada, Ottawa, ON, 1999.

Kirstein, B.E. and R.T. Redding, “Ocean-Ice Oil-Weathering Computer Program User's Manual”,Outer Continental Shelf Environmental Assessment Program, Vol. 59, National Oceanic andAtmospheric Administration, Ocean Assessment Division, Alaska, 83 p, 1988.

LaTorraca, G.A., K.J. Dunn, P.R. Webber, and R.M. Carlson, “Low-Field NMR Determinationsof the Properties of Heavy Oils and Water-in-Oil Emulsions”, Magnetic Resonance Imaging,Vol. 16, Nos. 5/6, pp 659-662, 1998.

Lee, H.M., J.W. Lee, and O.O. Park, “Rheology and Dynamics of Water-in-Oil Emulsions UnderSteady and Dynamic Shear Flow”, Journal of Colloid and Interface Science, Vol. 185, pp 297-305, 1997.

70

Mackay, D., “The Fate and Behaviour of Oil in Cold Climates”, Environment CanadaUnpublished Report, Ottawa, ON, 1984.

Mackay, D. and R.S. Matsugu, “Evaporation Rates of Liquid Hydrocarbon Spills on Land andWater”, Canadian Journal Chemical Engineering, Vol. 51, pp 434-439, 1973.

Mackay, D. and W. Zagorski, “Studies of the Formation of Water-In-Oil Emulsions”, inProceedings of the Fourth Annual Arctic Marine Oilspill Program Technical Seminar,Environment Canada, Ottawa, ON, pp 75-86, 1981.

Mackay, D. and W. Zagorski, “Studies of Water-In-Oil Emulsions”, Manuscript Report EE-34,Environmental Protection Service, Environment Canada, Ottawa, ON, 93 p, 1982a.

Mackay, D. and W. Zagorski, “Water-In-Oil Emulsions: A Stability Hypothesis”, in Proceedingsof the Fifth Annual Arctic Marine Oilspill Program Technical Seminar, Environment Canada,Ottawa, ON, pp 61-74, 1982b.

Mackay, D. and M. Nowak, “Water-in-Oil Emulsions: Some Recent Advances”, in Proceedingsof the Seventh Annual Arctic Marine Oilspill Program Technical Seminar, Environment Canada,Ottawa, ON, pp 37-46, 1984.

Mackay, D., S. Paterson, and K. Trudel, “A Mathematical Model of Oil Spill Behaviour”,Manuscript Report EE-7, Environmental Protection Service, Environment Canada, Ottawa, ON,50 p, 1980a.

Mackay, D., I. Buist, R. Mascarenhas, and S. Paterson, “Oil Spill Processes and Models”,Manuscript Report EE-8, Environmental Protection Service, Environment Canada, Ottawa, ON,96 p, 1980b.

McDonagh, M., J.N. Hokstad, and A.B. Nordvik, “Standard Method for Viscosity Measurementof Water-in-Oil Emulsions”, MSRC Technical Report Series 95-003, Marine Spill ResponseCorporation, Washington, DC, 36 p, 1995.

McLean, J.D. and P.K. Kilpatrick, “Effects of Asphaltene Solvency on Stability of Water-in-OilEmulsions”, Journal of Colloid and Interface Science, Vol. 189, pp 242-253, 1997a.

McLean, J.D. and P.K. Kilpatrick, “Effects of Asphaltene Aggregation in Model Heptane-Toluene Mixtures on Stability of Water-in-Oil Emulsions”, Journal of Colloid and InterfaceScience, Vol. 196, pp 23-34, 1997b.

McLean, J.D., P.M. Spiecker, A.P. Sullivan, and P.K. Kilpatrick, “The Role of PetroleumAsphaltenes in the Stabilization of Water-in-Oil Emulsions”, in Structure and Dynamics ofAsphaltenes, Mullins and Sheu (eds.), Plenum Press, New York, NY, pp 377-422, 1998.

71

Mingyuan, L., A.A. Christy, and J. Sjöblom, “Water-in-Crude Oil Emulsions from theNorwegian Continental Shelf: Part VI. Diffuse Reflectance Fourier Transform InfraredCharacterization of Interfacially Active Fractions from North Sea Crude Oil”, in Emulsions - AFundamental and Practical Approach, J. Sjöblom (ed.), Kluwer Academic Publishers,Dordrecht, The Netherlands, pp 157-172, 1992.

Miyahara, S., “On the Formation of Water-In-Oil Emulsions”, in Proceedings of Oil inFreshwater: Chemistry, Biology, Countermeasure Technology, John H. Vandermeulen and SteveE. Hrudey (eds.), Pergamon Press, New York, NY, 1985.

Mohammed, R.A., A.I. Bailey, P.F. Luckham, and S.E. Taylor, “Dewatering of Crude OilEmulsions: 2. Interfacial Properties of the Asphaltic Constituents of Crude Oil”, Colloids andSurfaces A, Vol. 80, pp 237-242, 1993a.

Mohammed, R.A., A.I. Bailey, P.F. Luckham, and S.E. Taylor, “Dewatering of Crude OilEmulsions: 1. Rheological Behaviour of the Crude Oil-Water Interface”, Colloids and SurfacesA, Vol. 80, pp 223-235, 1993b.

Mouraille, O., T. Skodvin, J. Sjöblom, and J.-L. Peytavy, “Stability of Water-in-Crude OilEmulsions: Role Played by the State of Solvation of Asphaltenes and by Waxes”, Journal ofDispersion Science and Technology, Vol. 19, pp 339-367, 1998.

Nesterova, M.P., O.S. Mochalova, and A.B. Mamayev, “Transformation of Oil-in-WaterEmulsion into Solid Tar Balls in the Sea”, Oceanology, Vol. 23, No. 6, pp 734-737, 1983.

Neumann, H.J. and B. Paczynska-Lahme, “Stability and Demulsification of PetroleumEmulsions”, Progress in Colloid and Polymer Sciences, Vol. 101, pp 101-104, 1996.

Nordli, K.G., J. Sjöblom, and P. Stenius, “Water-in-Crude Oil Emulsions from the NorwegianContinental Shelf: 4. Monolayer Properties of the Interfacially Active Crude Oil Fraction”,Colloids and Surfaces, Vol. 57, pp 83-98, 1991.

Omar, A.E., S.M. Desouky, and B. Karama, Journal of Petroleum Science and Engineering, Vol. 6, pp 149-160, 1991.

Paczynska-Lahme, B., “Mesophases in Petroleum”, Progress in Colloid & Polymer Science,Vol. 83, pp 196-199, 1990.

Pal, R., “Effect of Droplet Size on the Rheology of Emulsions”, AIChE Journal, Vol. 42, pp 3181-3190, 1996.

Payne, J.R. and C.R. Phillips, Petroleum Spills in the Marine Environment - The Chemistry andFormation of Water-in-Oil Emulsions and Tar Balls, Lewis Publishers Inc., Chelsea, MI, 1985.

72

Puskas, S., J. Balazs, A. Farkas, I. Regdon, O. Berkesi, and I. Dekany, “The Significance ofColloidal Hydrocarbons in Crude Oil Production: Part 1. New Aspects of the Stability andRheological Properties of Water-Crude Oil Emulsions”, Colloids and Surfaces, Vol. 113, pp 279-293, 1996.

Reed, M., “The Physical Fates Component of the Natural Resource Damage Assessment ModelSystem”, Oil and Chemical Pollution, Vol. 5, pp 99-123, 1989.

Rønningsen, H.P., J. Sjöblom, and L. Mingyuan, “Water-in-Crude Oil Emulsions from theNorwegian Continental Shelf: 11. Ageing of Crude Oils and Its Influence on the EmulsionStability”, Colloids and Surfaces A, Vol. 97, pp 119-128, 1995.

Schildberg, Y., J. Sjoblom, A.A. Christy, J.-L. Volle, and O. Rambeau, “Characterization ofInterfacially Active Fractions and their Relations to Water-in-Oil Emulsion Stability”, Journal ofDispersion Science and Technology, Vol. 16, pp 575-605, 1995.

Singh, S., J.D. McLean, and P.K. Kilpatrick, “Fused Ring Aromatic Solvency in DestabilizingWater-in-Asphaltene-Heptane-Toluene Emulsions”, Journal of Dispersion Science andTechnology, Vol. 20 (1&2), pp 279-293, 1999.

Sjöblom, J., O. Urdahl, H. Høiland, A.A. Christy, and E.J. Johansen, “Water-In-Crude OilEmulsions: Formation, Characterization and Destabilization”, Progress in Colloid & PolymerScience, Vol. 82, pp 131-139, 1990a.

Sjöblom, J., H. Söderlund, S. Lindblad, E.J. Johansen, and I.M. Skjärvo, “Water-in-Crude OilEmulsions from the Norwegian Continental Shelf: Part II. Chemical Destabilization andInterfacial Tensions”, Colloid and Polymer Science, Vol. 268, pp 389-398, 1990b.

Sjöblom, J., O. Urdahl, K.G.N. Børve, L. Mingyuan, J.O. Saeten, A.A. Christy, and T. Gu,“Stabilization and Destabilization of Water-in-Oil Emulsions from the Norwegian ContinentalShelf: Correlation With Model Systems”, Advances in Colloids and Interface Science, Vol. 41,pp 241-271, 1992.

Sjöblom, J., T. Skodvin, T. Jakobsen, and S.S. Dukhin, “Dielectric Spectroscopy and Emulsions:A Theoretical and Experimental Approach”, Journal of Dispersion Science and Technology,Vol. 15, pp 401-421, 1994.

Sjöblom, J., L. Mingyuan, A.A. Christy, and H.P. Ronningsen, “Water-in-Crude Oil Emulsionsfrom the Norwegian Continental Shelf: 10: Ageing of the Interfacially Active Components andthe Influence on the Emulsion Stability”, Colloids and Surfaces A, Vol. 96, pp 261-272, 1995.

Sjöblom, J., T. Skodvin, O. Holt, and F.P. Nilsen, “Colloid Chemistry and ModernInstrumentation in Offshore Petroleum Production and Transport”, Colloids and Surfaces,

73

Vol. 123/124 A, pp 593-607, 1997.

Sjöblom, J., H. Førdedal, T. Skodvin, and B. Gestblom, “Emulsions Characterized by Means ofTime Domain Dielectric Measurements (TDS): Technical Applications”, Journal of DispersionScience and Technology, Vol. 30(3), pp 921-943, 1999.

Skodvin, T., T. Jakobsen, and J. Sjöblom, “Dielectric Properties of W/O Emulsions UnderFlow”, Journal of Dispersion Science and Technology, Vol. 15, pp 423-448, 1994a.

Skodvin, T., J. Sjöblom, J.O. Saeten, O. Urdahl, and B. Gestblom, “Water-in-Crude OilEmulsions from the Norwegian Continental Shelf”, Journal of Colloid and Interface Science,Vol. 166, pp 43-50, 1994b.

Skodvin, T. and J. Sjöblom, “Models for the Dielectric Properties of Flocculated W/O-Emulsions”, Journal of Colloid and Interface Science, Vol. 182, pp 190-198, 1996.

Steinborn, R. and D.L. Flock, “The Rheology of Heavy Crude Oils and their Emulsions”, Journalof Canadian Petroleum Technology, Vol. 22, pp 38-52, 1983.

Stiver, W., M. Novak, W.Y. Shiu, and D. Mackay, “Recent University of Toronto Studies on theFate and Effects of Oil in the Marine Environment”, in Proceedings of the Sixth Annual ArcticMarine Oilspill Program Technical Seminar, Environment Canada, Ottawa, ON, pp 24-29, 1983.

Stiver, W. and D. Mackay, “Evaporation Rate of Spills of Hydrocarbons and PetroleumMixtures”, Environmental Science and Technology, Vol. 18, pp 834-840, 1984.

Tadros, Th.F., “Fundamental Principles of Emulsion Rheology and their Applications”, Colloidsand Surfaces A, Vol. 91, pp 39-55, 1994.

Tambe, D.E. and M.M. Sharma, “The Effect of Colloidal Particles on Fluid-Fluid InterfacialProperties and Emulsion Stability”, Advances in Colloid and Interface Science, Vol. 52, pp 1-63,1994.

Thingstad, T. and B. Pengerud, “The Formation of ‘Chocolate Mousse’ from Statfjord Crude Oiland Seawater”, Marine Pollution Bulletin, Vol. 14, No. 6, pp 214-216, 1983.

Toulhoat, H., C. Prayer, and G. Rouquet, “Characterization by Atomic Force Microscopy ofAdsorbed Asphaltenes”, Colloids and Surfaces A, Vol. 91, pp 267-283, 1994.

Twardus, E.M., “A Study to Evaluate the Combustibility and Other Physical and ChemicalProperties of Aged Oils and Emulsions”, Manuscript Report EE-5, Environmental ProtectionService, Environment Canada, Ottawa, ON, 159 p, 1980.

74

Urdahl, O., T. Brekke, and J. Sjöblom, “ C NMR and Multivariate Statistical Analysis of13

Adsorbed Surface-Active Crude Oil Fractions and the Corresponding Crude Oils”, Fuel, Vol. 71,pp 739-746, 1992.

Urdahl, O. and J. Sjöblom, “Water-in-Crude Oil Emulsions from the Norwegian ContinentalShelf: A Stabilization and Destabilization Study”, Journal of Dispersion Science andTechnology, Vol. 16, pp 557-574, 1995.

Walker, M., “Crude Oil Emulsification: A Comparison of Laboratory and Sea Trials Data”, inFormation and Breaking of Water-in-Oil Emulsions Workshop, Technical Report Series 93-018,Marine Spill Response Corporation, Washington, DC, pp 163-178, 1993.

Walker, M.I., T. Lunel, P.J. Brandvik, and A. Lewis, “Emulsification Processes at Sea - FortiesCrude Oil”, in Proceedings of the Eighteenth Arctic Marine Oilspill Program TechnicalSeminar, Environment Canada, Ottawa, ON, pp 471-491, 1995.

Wang, H. and C.P. Huang, “The Effect of Turbulence on Oil Emulsification”, presented at theworkshop on The Physical Behaviour of Oil in the Marine Environment, Princeton University,Princeton, NJ, pp 81-100, 1979.

Wang, Z., B.P. Hollebone, M. Fingas, B. Fieldhouse, C. Yang, M. Landriault, and S. Peng,“Characteristics of Spilled Oils, Fuels, and Petroleum Products: I Composition and Properties ofSelected Oils”, U.S. Environmental Protection Agency, EPA/600/R-03/072, National ExposureLaboratory, Atlanta, GA, 2004.

75

This page intentionally left blank.

review of the north slope oil properties relevant to ... · pdf filereview of the north slope...

Documents