National Academies Press: OpenBook

Spills of Emulsified Fuels: Risks and Response (2002)

Chapter: 2 Behavior and Fates: Summary and Evaluation of Available Information

« Previous: 1 Introduction and Overview
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 14
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 15
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 16
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 17
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 18
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 19
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 20
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 21
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 22
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 23
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 24
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 25
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 26
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 27
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 28
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 29
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 30
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 31
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 32
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 33
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 34
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 35
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 36
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 37
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 38
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 39
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 40
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 41
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 42
Suggested Citation:"2 Behavior and Fates: Summary and Evaluation of Available Information." Transportation Research Board and National Research Council. 2002. Spills of Emulsified Fuels: Risks and Response. Washington, DC: The National Academies Press. doi: 10.17226/10286.
×
Page 43

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

2 Behavior and Fates: Summary and Evaluation of Available Information BEHAVIOR AND FATE Physical Characterization Much of the scientific research on the fate of emulsified fuels has been focused on a particular commercial product called Orimulsion manufactured in Venezuela. The following properties specifically refer to Orimulsion-400, which replaced an earlier version of the product called Orimulsion-100. Orimulsion is a mix of approximately 70 percent natural bitumen (pitch) and 30 percent fresh water, with additives to maintain the emulsion. The additives are 1,100 parts per million (0.11 percent) MEA, an emulsion stabilizer, and 1,350 ppm (0.135 per- cent) AE, a water-soluble nonionic surfactant (Golder and Associates, 2001) that retards the coalescence of bitumen droplets. The designations 100 and 400 refer to the type of nonionic surfactant used in the formulation. (Much of the existing literature discusses studies of Orimulsion-100 thus this report makes the distinc- tion as needed.) With a specific gravity greater than 1.0, Orimulsion is classified for regula- tory purposes as a Group V oil (Stout, 1999). The density of the product is intermediate between fresh and salt water (Figure 2.1). It is comparable in pour point1 and viscosity to an industrial fuel oil. The dynamic viscosity of Orimulsion varies with temperature but is in general lower than that of many Group V oils. 1 The lowest temperature at which a substance, such as oil, will flow under specified conditions. 14

BEHAVIOR AND FATES 15 1.030 1.025 BIT UM 1.020 EN 1.015 DENSITY (g/ml) OR IMU LSI 35 ppt 1.010 ON 30 ppt 1.005 1.000 20 ppt 0.995 10 ppt 0.990 0.985 0 ppt 0.980 0 10 20 30 40 50 60 TEMPERATURE (˚C) FIGURE 2.1 Density-temperature relationship for Orimulsion-400, bitumen, and differ- ent salinity waters. SOURCE: Reprinted from PDVSA–Intevep,1998. Permission grant- ed by Bitor America Corporation. The flash point is > 95°C, and thus Orimulsion is classified as nonflammable by the National Fire Protection Association (NFPA). Pure bitumen has a high viscosity, on the order of a million centipoise at typical ambient water temperatures. The density and the pour point are higher than those of the emulsified Orimulsion product (Figure 2.1). Jokuty et al. (1995) list a pour point of 38ºC for bitumen versus a pour point of 3ºC for fresh Orimulsion. In completely quiescent conditions, dispersed bitumen droplets would be expected to sink or to be neutrally buoyant in water at 15°C and salinity less than 10-15 parts per thousand (ppt)2 (Crosbie and Lewis, 1998a), but they would be expected to float in more saline water. The average bitumen droplet size in fresh Orimulsion is around 20 µm (Ostazeski et al., 1998a, 1998b). Although the droplet size distribution (Figure 2.2) is somewhat bimodal (Stout, 1999), almost all of the droplets range in size from 1 to 80 µm. Exposure to even low-salinity water (>5-7 psi (Practical Salinity Units) collapses the surfactant in Orimulsion (Brown et al., 1995; Crosbie and Lewis, 1998b); releasing the bitumen from the emulsion where it can form 2The documents examined report concentrations in metric units such as mg/L or parts per million. This difference reflects differences in common use between scientific practice within the community. This report may use either or both, depending on use in the literature reviewed. For the concentrations and conditions discussed in this report, the conversion between systems is unwarranted.

16 SPILLS OF EMULSIFIED FUELS 15 Orimulsion 400 1998 10 Percent 5 0 0.5 1.6 2.4 3.5 5.2 7.8 11. 17. 25. 37. 56. 83. 6 2 5 8 1 3 Diameter (µm) FIGURE 2.2 Histogram showing average bitumen droplet size in fresh Orimulsion-400. SOURCE: Stout, 1999. droplets. If the droplets collide, the bitumen may coalesce and form larger droplets, which (based on Stokes equation, Figure 2.3), increases the likelihood that the droplet will either surface or sink, depending on its relative buoyancy compared to the surrounding water. Because Orimulsion is not a homogeneous fluid, its properties may change significantly during the course of an accidental release. According to Febres et al. (1995), at a concentration greater than 20,000 ppm the product retains its emulsion properties, while at a concentration less than 10,000 ppm the material is expected to behave like dispersed bitumen droplets. Such high concentrations would exist only for an open-water spill within the immediate vicinity of the spill source and for a very short time. The exception would be a spill scenario in which mixing with the ambient water and subsequent dilution of the product were restricted. For the most part, predicting Orimulsion spill behavior becomes a matter of predicting the behavior of the dispersed bitumen cloud. Chemical Characterization The Cerro Negro bitumen used to produce Orimulsion comes from the Orinoco Belt in the Eastern Venezuelan Basin (Bitumenes Orinoco). This bitu- men is highly weathered (degraded) in nature and consists primarily of high- molecular-weight, multi-ring aromatic hydrocarbons and resins that account for

BEHAVIOR AND FATES 17 0.01 Vertical velocity (cm/sec) 0.005 0 50 100 Bitumen diameter (microns) FIGURE 2.3 Re-float velocity of bitumen droplets (assumed to be spherical) by Stokes equation in seawater (35 ppt) at 15ºC. 63 to 69 percent of the fuel (Ostazeski et al., 1998a, 1998b; Brown et al., 1995; Jokuty et al., 1999; Stout, 1999). Figure 2.4 presents flame ionization detector gas chromatographic profiles. These profiles show neat Orimulsion (sample 3), an oil-in-water dispersion of 18,250 mg/L Orimulsion (dissolved phase) filtered through a 1-µm (micron) membrane (sample 4), and an oil-in-water dispersion of 5,475 mg/L Orimulsion that has not been filtered (sample 8), (Wang and Fingas, 1996). The neat Orimulsion is characterized by a bimodal unresolved complex mixture (Fig. 2.4A). Very few of the individual constituents present in the Orimulsion are partitioned into the dissolved phase (Fig. 2.4B). The unfiltered sample shows the total petroleum hydrocarbons clearly associated with the bitu- men droplets as evidenced by the pattern which is almost identical to the whole Orimulsion pattern (Fig. 2.4C). Orimulsion has very low concentrations of BTEX that are at least an order of magnitude lower than in the typical No. 6 fuel oil it is likely to replace (Table 2.1). Individual and total polynuclear aromatic hydrocarbon (PAH) concentra- tions in Orimulsion are up to one order of magnitude below those typically found in crude oils and refined products (Table 2.2). Orimulsion is relatively high in sulfur, nickel, and vanadium (Table 2.1), with the latter two constituents tied up as metalloporphyrins, which make them biologically unavailable. The additives used in the production of Orimulsion-400 are a water-soluble nonionic surfactant (a narrow distillate cut of widely used AE) and an emulsion

18 SPILLS OF EMULSIFIED FUELS 4.0e4 3 3.0e4 2.0e4 1.0e4 0 0 10 20 30 40 50 60 4.0e4 4 3.0e4 2.0e4 1.0e4 0 0 10 20 30 40 50 60 4.0e4 8 3.0e4 2.0e4 1.0e4 0 0 10 20 30 40 50 60 FIGURE 2.4 Flame ionization detector gas chromatographic profiles for neat Orimulsion (sample 3), an oil-in-water dispersion filtered through a 1-µm membrane (sample 4), and an oil-in-water dispersion that has not been filtered (sample 8). SUR and IS represent surrogate and internal standards, respectively. SOURCE: Wang and Fingas, 1996. Copy- right 1996 American Chemical Society.

BEHAVIOR AND FATES 19 TABLE 2.1 Physical Properties and Chemical Composition of Orimulsion-400 and No. 6 Fuel Oil Parameter Orimulsion-400 No. 6 fuel oil Density (g/ml at 15ºC) 1.01-1.02 0.94-1.02 Pour point (ºC) 0-3 −17-36 Viscosity (cP) 200-350 (at 30 ºC) 325-47,000 (at 15 ºC) Mean particle size (µm) 14-20 Not applicable Sulfur content (wt%) 2.85 0.7-3.0 Hydrocarbon groups (wt%) Saturated 14 11 Aromatic 47 55 Resins 22 20 Asphaltenes 17 14 Total BTEX (ppm) 36 464 Benzene 0 17 Toluene 4 100 Ethylbenzene 19 47 Xylene 13 300 Total PAH (µg/g bitumen) 3,040 317,627 Naphthalenes (C0-C4) 474 46,600-106,000 Phenanthrenes (C0-C4) 854 50,200-113,000 Dibenzothiophenes (C0-C4) 1,330 18,713 Fluorenes (C0-C4) 348 12,800-30,700 Chrysenes (C0-C4) 168 21,600-48,700 Metals (ppm) Nickel 55 37 Vanadium 310 32 Zinc 19 45 SOURCE: Bitumenes Orinoco, S.A., b; National Oceanic and Atmospheric Administration ADIOS Model Database. stabilizer, MEA. The purpose of these components is to maintain the stability of the bitumen droplets in the emulsion by preventing particle-particle agglomera- tion and coalescence. The AE that make up the surfactant are composed of a long-chain fatty (alkyl) alcohol (hydrophobic) and an ethylene oxide (EO) chain (hydrophilic), connected by an ether linkage. The nomenclature of AE is determined by the average number of carbons in the alkyl chain of the alcohol and the number of EO groups in the hydrophilic moiety (e.g., C10-12EO8 represents an alcohol with 10 to 12 carbons attached to polyethylene oxide with 8 EO units). The alcohol ethoxylate used in Orimulsion-400 is known as GENAPOL X 159 and is complexly branched. It contains a mixture of highly branched C12 (22- 30 percent) and C13 (70-78 percent) fatty alcohols with anywhere from 9 to 22 (EO9 to EO22) EO groups (Bjornestad et al., 1998; Bowadt et al., 1998). The

20 TABLE 2.2 PAH Concentrations in Orimulsion-400, Crude Oils, and Petroleum Products Average Heavy Average Average Light Crude Prudhoe Bay Crude Arab. Med. Bunker C Cold Lake (average of 19 oils) crude (average of 6 oils) Crude Diesel No. 2 (average of 4) Orimulsion Bitumen Sample Type PAHs (µg/g oil) (ug/g oil) (µg/g oil) (µg/g oil) (µg/g oil) (µg/g oil) (µg/g oil) (µg/g oil) Naphthalene C0-N 45.6 46 21.0 21.0 1404.6 28.2 9.3 47.0 C1-N 540.4 540 148.9 148.9 5174.0 87.6 47.2 128.0 C2-N 1696.9 1697 427.0 427.0 7031.6 327.8 165.7 490.0 C3-N 2082.9 2083 602.4 602.4 5591.4 541.0 243.0 839.0 C4-N 990.6 991 313.2 313.2 2963.6 507.0 309 705.0 Sum 5356 5356 1513 1513 22165 1492 774 2209 Phenanthrene C0-P 41.4 41 80.4 80.4 455.8 68.4 45.7 91.0 C1-P 348.3 348 324.2 324.2 657.5 210.2 133.4 287.0 C2-P 462.3 462 405.5 405.5 223.5 406.3 306.5 506.0 C3-P 380.3 380 356.4 356.4 33.0 472.2 425.4 519.0 C4-P 262.2 262 225.8 225.8 6.3 202.3 228.6 176.0 Sum 1494 1494 1392 1392 1376 1359 1140 1579 Dibenzothiophene C0-D 185.8 186 37.4 37.4 367.4 34.9 16.7 53.0 C1-D 623.5 624 124.9 124.9 414.9 136.8 67.6 206.0 C2-D 1092.9 1093 212.3 212.3 160.5 337.5 222.9 452.0 C3-D 967.5 968 192.5 192.5 30.6 405.0 364.0 446.0 Sum 2870 2870 567 567 973 914 671 1157 Fluorene C0-F 43.1 43 18.6 18.6 179.4 21.1 10.1 32.0 C1-F 120.3 120 60.6 60.6 404.5 54.8 38.6 71.0 C2-F 240.6 241 98.0 98.0 375.0 135.8 126.7 145.0 C3-F 225.8 226 109.3 109.3 221.3 152.2 134.4 170.0 Sum 630 630 286 286 1180 364 310 418 SPILLS OF EMULSIFIED FUELS Chrysene C0-C 17.1 17 32.0 32.0 0.5 21.4 14.9 28.0 C1-C 26.3 26 82.9 82.9 0.0 39.8 29.5 50.0 C2-C 37.7 38 139.8 139.8 0.0 62.0 53.9 70.0

C1-F 120.3 120 60.6 60.6 404.5 54.8 38.6 71.0 C2-F 240.6 241 98.0 98.0 375.0 135.8 126.7 145.0 C3-F 225.8 226 109.3 109.3 221.3 152.2 134.4 170.0 Sum 630 630 286 286 1180 364 310 418 Chrysene C0-C 17.1 17 32.0 32.0 0.5 21.4 14.9 28.0 C1-C 26.3 26 82.9 82.9 0.0 39.8 29.5 50.0 C2-C 37.7 38 139.8 139.8 0.0 62.0 53.9 70.0 C3-C 34.5 35 72.5 72.5 0.0 44.6 46.3 43.0 Sum 116 116 327 327 1 168 145 191 TOTAL PAH 10466 10466 4086 4086 25696 4297 3040 5554 Other PAHs BEHAVIOR AND FATES Biphenyl 26.9 26.90 2.7 2.7 363.4 11.2 5.4 Acenaphthalene 10.2 10.22 0.7 0.7 31.7 2.7 1.6 Acenaphthene 7.5 7.50 1.0 1.0 24.5 3.2 11.2 Anthracene 3.1 3.06 0.8 0.8 59.5 6.5 2.7 Fluoranthene 3.2 3.19 0.7 0.7 0.1 3.4 3.2 Pyrene 4.3 4.26 4.8 4.8 0.3 1.8 6.5 Benz(a)anthracene 0.9 0.86 5.2 5.2 0.0 0.3 3.4 Benzo(b)fluoranthene 1.3 1.26 3.2 3.2 0.0 2.2 1.8 Benzo(k)fluoranthene 0.4 0.43 0.7 0.7 2.1 0.3 Benzo(e)pyrene 3.1 3.13 12.2 12.2 0.0 6.7 2.2 Benzo(a)pyrene 0.5 0.46 2.4 2.4 0.1 0.3 2.1 Perylene 0.1 0.09 0.8 0.8 0.0 0.2 6.7 Indeno(1,2,3cd)pyrene 0.1 0.06 0.2 0.2 0.1 1.9 0.3 Dibenz(a,h)anthracene 0.2 0.21 1.4 1.4 0.0 49.5 0.2 Benzo(ghi)perylene 0.6 0.58 7.0 7.0 0.0 48.6 1.9 TOTAL 62 62 44 43.8 480 141 49 Source: Environment Canada, 2001. 21

22 SPILLS OF EMULSIFIED FUELS selected group of C12 and C13 branched AE was used in Orimulsion-400 rather than nonylphenol ethoxylates used in Orimulsion-100, because of the latter’s potential for endocrine disruption and the fact that the degradation metabolites were more toxic and persistent than the parent compounds (Harwell and Johnson, 2000). Because of their widespread use in household products, much of the research on microbial degradation of AE has been based on their removal efficiencies in wastewater treatment facilities. For various types of treatments, 86-99 percent of the AE in wastewater influents are degraded to some intermediate form (Talmage, 1994). It has long been held that the major, if not only, route for linear AE biodegradation is that of cleavage at the ether bridge between the alkyl chain and the EO moiety (Swisher, 1987). After that, biodegradations of fatty alcohols and polyethylene glycols (PEG) were believed to proceed independently and more slowly. Branching of the alkyl chain in the vicinity of the central ether bridge ap- pears to inhibit central ether cleavage (Di Corcia et al., 1998). In addition, the same study (Di Corcia et al., 1998) reported that bacterial attack on the ethoxy chain produced metabolites with the EO either shortened or, to a lesser extent, oxidized to a terminal carboxylic acid group. They also reported end-of-chain oxidation of both alkyl side chains in branched AE to form very polar di- and tri- carboxylic acids. Marcomini et al. (2000a,b) reported fast biomediated ether- linkage cleavage of linear and short-chain (methyl or ethyl) C2-monobranched AE with slower biodegradation of the released PEG by both hydrolytic shorten- ing and oxidative hydrolysis to form shorter PEG oligomers and carboxylated PEG. Taken together, these studies suggest that the AE mixtures used in Orimulsion-400 formulations are capable of slow aerobic biodegradation. In fact, one study on the rate of biodegradation of GENAPOL X 159 stated that “the toxicity of the AE was still 100 percent even after 56 days of biodegradation.” This is in accordance with the results of Bjornestad et al. (1998) who found biodegradation of only 35 percent of the AE. This rate of biodegradation is at variance however with oxygen consumption studies on the specific AE mixture used for Orimulsion-400 that have shown a biodegradability (compared to com- plete chemical oxygen demand with potassium dichromate) of 79 percent in seawater over a 28-day period (VKI, 1997a). MEA, the substance used to stabilize Orimulsion, is widely used in healthcare products and the surfactant industry. It is not considered mutagenic or carcino- genic (Johnson, 1998) and is utilized and metabolized by plant, animal, and microbial cells. As such, it is quickly transformed in the environment with a half- life of days to weeks (Johnson, 1998). Oxygen consumption studies on MEA have shown a biodegradability (compared to complete chemical oxygen demand with potassium dichromate) of 74 percent in seawater over a 28-day period (VKI, 1997b). Because of its high water solubility, MEA has a low potential for bioaccumulation. This is discussed in greater detail in Chapter 3.

BEHAVIOR AND FATES 23 TABLE 2.3 Selected Dissolved Aromatic Hydrocarbon Concentrations (Mean and Standard Deviation, µg/L) After Five-Day, Zero-Headspace Equilibrium Exposure Studies Fuel Benzene Toluene Xylene Naphthalene No. 6 fuel oil 73.0 ± 15.8 197.3 ± 29.2 126.5 ± 17.2 142.5 ± 23.6 Orimulsion 05.9 ± 4.2 057.8 ± 5.1 003.8 ± 1.6 004.9 ± 2.5 SOURCE: Brown et al., 1995. A detailed chemical characterization of the water-soluble components con- tained in the 30 percent water added to make Orimulsion and the water-soluble fractions (WSF) generated by dispersions of Orimulsion into fresh and salt water was completed by Potter et al. (1997). In addition, Brown et al. (1995) examined the dissolution behavior of Orimulsion and No. 6 fuel oil spilled in water of varying salinities. These studies concluded that the majority of the surfactants are contained in the aqueous phase and that they would be diluted by the receiv- ing water during an Orimulsion spill. BTEX constituents were not observed in 1:9 (volume:volume) dispersions of Orimulsion in water at a detection limit of 0.2 µg/L (Potter et al., 1997). Table 2.3 presents benzene, toluene, xylene and naphthalene concentrations from zero-headspace, five-day equilibrium exposure studies comparing Orimulsion and No. 6 fuel oil (Brown et al., 1995). In general, levels of volatile and water-soluble components are higher by up to one order of magnitude for the No. 6 fuel oil tested in their studies compared to Orimulsion. To date, no detailed chemical analyses of the dissolved-phase PAH concen- trations in the 30 percent aqueous phase of Orimulsion have been reported. Us- ing equilibrium partition theory and the initial concentrations of PAH in neat Orimulsion, Stout (1999) and French (2000) calculated the initial concentrations of PAH expected in the aqueous phase of Orimulsion; these values are presented in Table 2.4. The highest concentrations of any PAH are for the naphthalenes, and they are only around 2 µg/L—in general agreement with the values measured by Brown et al. (1995), with initial concentrations for other PAH rapidly decreas- ing to values in the range of 0.1-0.8 µg/L. The sum of the PAH, or the total PAH (TPAH) with log Kow values <5.6 is only 14.9 µg/L. Brown et al. (1995) exam- ined the time-series kinetics of dissolution behavior from Orimulsion and No. 6 fuel oil, and concluded that little or no additional environmentally significant PAH dissolution occurs upon release of Orimulsion to the environment. In their dilution studies, however, they did not dilute the Orimulsion fuel enough to allow for continued dissolution of PAH from the bitumen phase. Instead, they con- cluded that the dissolved-phase components measured in their experiments were simply from the dilution of the already near-equilibrium concentrations of PAH in the initial aqueous phase. A similar conclusion was reached by Stout (1999) who examined the dissolution behavior of the high-molecular-weight PAH frac-

24 SPILLS OF EMULSIFIED FUELS TABLE 2.4 Dissolved PAH Concentrations Estimated in the Aqueous Phase of Fresh Orimulsion Molecular Concentration Dissolved Weight Log in Concentration PAH (g/mol) Kow Bitumen (mg/kg) (µg/L) Naphthalene 128 3.37 0015.4 02.363 C1 naphthalenes 142 3.87 043.03 02.263 C2 naphthalenes 156 4.37 136.7 02.464 C3 naphthalenes 170 5 189.37 00.886 C4 naphthalenes 185 5.55 267.97 00.386 Biphenyls 154 3.9 005 00.247 Acenaphthylene 152 4.07 000 00 Acenaphthene 154 3.92 010.66 00.504 Dibenzofuran 168 4.31 005.42 00.111 Fluorene 166 4.18 013.52 00.366 C1 fluorenes 181 4.97 057.39 00.286 C2 fluorenes 196 5.2 184.39 00.562 C3 fluorenes 211 5.5 272.13 00.436 Dibenzothiophene 184 4.49 028.19 00.393 C1 dibenzothiophene 199 4.86 133.96 00.846 C2 dibenzothiophene 214 5.5 345.13 00.553 C3 dibenzothiophene 228 5.73 692.83 00.679 Phenanthrene 178 4.57 067.78 00.796 Anthracene 178 4.54 000 00 C1 phenanthrenes-anthracenes 192 5.14 143.84 00.499 C2 phenanthrenes-anthracenes 207 5.25 366.41 01.003 C3 phenanthrenes-anthracenes 222 6 459.43 00.252 C4 phenanthrenes-anthracenes 237 6.51 241.73 00.0446 Fluoranthene 202.3 5.22 000 00 Pyrene 202.3 5.18 000 00 Sum (Kow <5.6) 2,286 14.964 SOURCE: Stout, 1999. NOTE: Concentrations are calculated from the PAH composition of whole Orimulsion data from Stout (1999) and equilibrium partitioning (data provided by D. French McCay, A.S.A. Narragansett, RI) to estimate dissolved PAH composition (PAH concentrations listed as zero if log Kow >5.6). tion of Orimulsion-400 after 24 hours’ exposure in benchtop 4-liter stirred-bea- ker experiments. Figure 2.5 presents a histogram plot showing the relative abun- dance of high-molecular-weight PAH in fresh Orimulsion and the dissolved con- stituents in the water generated by gently stirring a 1,700-mg/L dispersion of Orimulsion in fresh water for 24 hours. The PAH dissolved in the water are enriched in the more soluble low-molecular-weight two-ring PAH (naphthalene and its alkyl-substituted homologues), which were the predominant dissolved- phase components in the near-equilibrium 30 percent aqueous phase in the origi- nal Orimulsion fuel mixture (e.g., see Table 2.4). As discussed further in the section titled “Water Column Processes - Dissolution,” the high-molecular-weight

BEHAVIOR AND FATES 25 PAH (e.g., phenanthrenes, dibenzothiophenes, and chrysenes) in the dispersed bitumen phase (Figure 2.5) can continue to dissolve to an appreciable extent given the extremely high surface-area-to-volume ratio of the dispersed bitumen droplets. However, this dissolution process would be masked at the dilution volume used in the experiment. In addition, the coalescence of bitumen droplets into larger agglomerates would tend to limit such behavior in the experimental apparatus used, while it would not be a factor in an open-ocean environment. WATER COLUMN PROCESSES To date, there have been only a limited number of smaller experimental spills and no accidental spills of Orimulsion. As a result, much of our knowledge regarding the behavior and fate of spilled Orimulsion is based on laboratory-scale studies and models. In preparing a “real-world” Orimulsion budget or mass balance based on predictions from benchtop or flume experiments, it is critical to caveat the results because of the difficulty in simulating and scaling “real-world” conditions. Factors that cannot be simulated in laboratory-scale experiments include water turbulence and density profiles, unconfined water volumes, continual dilution of bitumen, and the absence of container walls (Stout, 1999). Moreover, because Orimulsion is not a homogeneous fluid, there may be special peculiarities in its behavior and the break- down of the emulsion that are specific to the nature of the spill event. Dispersion Orimulsion is expected to disperse (spread apart) rapidly when spilled into an open-water environment. In fact, for spill response purposes, fresh Orimulsion can almost be considered predispersed oil. The near-equilibrium dissolved-phase PAH in the aqueous phase of the product would behave as a neutrally buoyant, or near neutrally buoyant, contaminant. Dilution of this dissolved-phase PAH solu- tion would depend on the mixing characteristics of the surrounding water. As an example, two small spills of Orimulsion were intentionally released into the Caribbean Sea in July 1996 (French et al., 1997). Using the reported diffusion coefficients for this experiment, neglecting boundary conditions and buoyancy forces, and assuming Fickian diffusion, the maximum dissolved-phase PAH con- centration of the experimental spills would fall to less than 1 percent of the initial concentration in less than two minutes. However, eddy diffusion coefficients are functions of the flow field and vary in both time and space. For this reason measured concentrations can differ by orders of magnitude (Okubo, 1971). Ac- tual dilution will therefore depend on the specific nature of the spill and the concurrent environmental conditions. Also, additional dissolution of PAH from the dispersed bitumen droplets will further complicate the picture as described in the following section. In general, more open and energetic conditions will result in more rapid mixing and lower concentrations.

26 1.0 0.9 Fresh Orimulsion 0.8 Dissolved in Water 0.7 (Trial 6') 0.6 0.5 0.4 Rel. Abund. 0.3 0.2 0.1 0.0 e s s e an s s e s s s s e s s e s s e e e e en n e n e en r n e n e en n e ne ne ene en ene en e e n ene ene en en en en h r y s s s al ale ale hy l o f u o r e o re t h r ace ace h e t z n r r th y r y ry ry t h py r py r ryl ) ) e th th th lu lu op op an /p s/p hr an a p h ph ph ap h ben 1 -f 3 -f en a nt h n t h t h i t h i uo r es a a i a a l n ne C - ch - ch o r o ( a c,d ,i )p 2 4 - N -n D C C P h s/ s/ n z o zo F e he C lu n z ,3 g, h - n ce n ne ne e en hr t C k )f C2 C4 A o( B e 1,2 zo ( ( re h re di b dib - an t ran z n th t 3- o r l uo en n o Be n an n an C1 C f lu 3 - f B de he h e 1- C In -p 4 -p C C2 C FIGURE 2.5 Histograms comparing the distribution of PAH in fresh Orimulsion and a representative distribution of PAH dissolved in water SPILLS OF EMULSIFIED FUELS after 24 hours. NOTE: Because of space limitations, compound names are given only for every other PAH. SOURCE: Stout, 1999.

BEHAVIOR AND FATES 27 In these open-water experimental spills, the bitumen concentration quickly fell below the critical concentrations reported by Febres et al. (1995) and Brown et al. (1995), where there is a transition from an emulsion to a cloud of suspended bitumen droplets. In such a situation, the behavior of these droplets depends upon the initial concentration, the mixing energy of the system, and the salinity and temperature of the water. The initial size bitumen droplets in an Orimulsion spill are nearly all below the size limit that spill responders use to define dis- persed oil (Lunel, 1993). According to Li and Garrett (1998), fluids with a high viscosity such as bitumen resist being broken by normal turbulence, so it is unlikely that the droplets would be further reduced in size. Small droplets will tend to remain suspended in the water column pending coalescence and the formation of larger agglomerates. Coalescence is discussed further in “Coales- cence and Bitumen-SPM Interaction” later in this chapter. If there were no coalescence of the bitumen or adherence to suspended sediments, the droplets would be expected to behave in a similar manner as dust particles in the air. While dust particles are heavier than air, atmospheric turbulence keeps them suspended. Similarly, even minimal water turbulence would keep the bitumen droplets from either sinking (in fresh water) or floating (in seawater) as would be expected, based solely on density considerations. Coalescence, which would increase droplet size, could change this result because the vertical velocity of the bitumen droplet due to buoyancy increases with the square of the droplet diam- eter (small particles) or linearly with the droplet diameter (large particles). In an area of steady currents, vertical shear diffusion would cause the bitu- men plume to become elongated and progressively more dilute in the direction of the current. Shorelines or other boundaries could restrict the dispersion of the subsurface droplets but, to the extent that the relatively weak buoyancy forces can be neglected, would not cause any increased concentration of the bitumen. The exception would be any bitumen that has resurfaced or sunk due to coalescence. Dissolution An Orimulsion spill into a water body results in input of dispersed bitumen droplets and the dissolved constituents in the aqueous phase (BTEX, dissolved PAH at near-equilibrium levels, and surfactants). The initial and final concentra- tions of dissolved constituents will be a function of the volume of Orimulsion spilled, the volume of the receiving water body, and the turbulence. Laboratory studies by Brown et al. (1995) have confirmed that the transfer of initially dissolved components from Orimulsion into water is essentially instanta- neous, usually occurring within 30 minutes. This process is reflected in Figure 2.6, which shows the dissolution behavior observed when Orimulsion was intro- duced into seawater at varying salinities in controlled equilibrium batch studies. These curves document the behavior expected from simple dilution of the near- equilibrium dissolved PAH (see also Figure 2.5) and the surfactants in the origi-

28 SPILLS OF EMULSIFIED FUELS 1 Concentration (mg/L) 0.75 0.5 Concentrated Tampa Bay Water (36.6 ppt) 0.25 0 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000 5,500 6,000 Reaction time (min.) 1 Concentration (mg/L) 0.75 0.5 Tampa Bay Water (23.0 ppt) 0.25 0 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000 5,500 6,000 Reaction time (min.) 1 Concentration (mg/L) 0.75 0.5 Diluted Tampa Bay Water 0.25 (9.8 ppt) 0 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000 5,500 6,000 Reaction time (min.) FIGURE 2.6 Mean values of total dissolved aromatic hydrocarbons from Orimulsion as a function of salinity and batch-mode equilibrium exposure time. SOURCE: Brown et al., 1995. Used with permission of the University of Miami. nal Orimulsion aqueous phase. Interestingly, the laboratory data suggest that total dissolved aromatic hydrocarbon concentrations are actually higher at high salinity (36 ppt) compared to brackish water salinities (9.8 ppt). This behavior seems counterintuitive at first. However, Brown et al. (1995) believed that the

BEHAVIOR AND FATES 29 increased concentrations in the dissolved phase at higher salinities might be caused by the preferential formation of micelles by free surfactants under high ionic strengths, with an apparent solubility increment resulting from the incorpo- ration of petroleum hydrocarbons into the surfactant-based micelles. Although this hypothesis seems reasonable, additional studies are needed to fully evaluate and confirm it. More importantly, however, a number of researchers interpreted these data to conclude that continued PAH dissolution from the bitumen droplets would not be environmentally significant. What these studies failed to consider was the limited dilution volume of the receiving water, the influence of bitumen droplet coalescence on droplet size distribution, and the importance of equilibrium parti- tioning theory. Equilibrium partitioning theory predicts that additional PAH should dissolve from the bitumen droplets (especially given their high surface- area-to-volume ratios) when the water phase is sufficiently diluted. These topics are considered in more detail below; but first, by way of comparison, Figure 2.7 shows the dissolution behavior of PAH components when No. 6 fuel oil was introduced into the controlled batch-mode equilibrium studies completed by Brown et al. (1995). Unlike the Orimulsion tests, No. 6 fuel oil continues to release dissolved-phase PAH into the water over time even under the rather limited dilution conditions of the laboratory tests. The observed increased con- centration of the dissolved fraction with decreasing salinity is expected given the solubility behavior for hydrophobic organic compounds in seawater. Once diluted into the receiving body of water, dissolved-phase PAH, AE, and monoethanolamine would be subject to loss by bacterial degradation at rates controlled by bacterial population densities, nutrient levels, temperature, and other factors. Adsorption of dissolved-phase PAH by SPM is not expected to be significant compared to direct bitumen droplet-SPM interactions. Stout (1999) concluded, like Brown et al. (1995), that the PAH components associated with the bitumen droplets would be essentially inert, with very little additional abiotic (evaporation-dissolution) or bacterial weathering expected. In more recent theo- retical and modeling evaluations of Orimulsion behavior, French (2000) con- cluded that significant PAH dissolution weathering from the bitumen droplets can occur over time, particularly given the high surface-area-to-volume ratios of the 1- to 80-µm-diameter dispersed bitumen droplets (Stout, 1999). French, using equilibrium partitioning theory, and the initial PAH concentra- tions in neat Orimulsion from two different sources determined that the effective initial concentration of dissolved PAH in the neat fuel would be in the range of 15 µg/L (calculated as shown in Table 2.4 based on the 2,286 ppm of PAH in neat Orimulsion reported by Stout (1999), and 30 ppb (calculated in a similar manner but using the 3,050 ppm of PAH in neat Orimulsion as reported in Table 2.2 by Environment Canada (2001). This represents the maximum theoretical concen- tration that could be seen by exposed organisms. Concentrations (µg/L) of dis- solved PAH and monoaromatic hydrocarbons (MAH) from Orimulsion diluted

30 SPILLS OF EMULSIFIED FUELS 1 Concentration (mg/L) 0.75 Concentrated Tampa Bay Water 0.5 (36.6 ppt) 0.25 0 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000 5,500 6,000 Reaction time (min.) 1 Concentration (mg/L) 0.75 0.5 Tampa Bay Water (23.0 ppt) 0.25 0 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000 5,500 6,000 Reaction time (min.) 1 Concentration (mg/L) 0.75 0.5 Diluted Tampa Bay Water (9.8 ppt) 0.25 0 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000 5,500 6,000 Reaction time (min.) FIGURE 2.7 Mean values of total dissolved aromatic hydrocarbons from No. 6 fuel oil as a function of salinity and batch-mode equilibrium exposure time. SOURCE: Brown et al., 1995. Used with permission of the University of Miami. by the indicated dilution factors were calculated using the equations of French and are shown in Table 2.5. As shown by the data in the table, PAH are predicted to continue to dissolve from the dispersed bitumen droplets as they are diluted into the receiving water body, but the dissolved PAH will never exceed the initial

BEHAVIOR AND FATES 31 TABLE 2.5 Concentration of Dissolved Aromatics (µg/L) from Orimulsion Diluted by the Indicated Factors, as Computed from an Equilibrium Partitioning Model Dilution MAH + PAH MAH PAH 1 51.06 32.05 19.02 10 50.64 31.64 19.01 70 48.25 29.29 18.95 1,00 47.24 28.31 18.93 1,000 34.52 16.30 18.22 10,000 18.43 3.93 14.49 100,000 7.84 0.47 7.37 1,000,000 2.01 0.05 1.96 SOURCE: Data provided by D. French McCay, A.S.A., Narragansett, RI. concentrations in the aqueous phase of the emulsion. Increased dissolution with exposure to noncontaminated water can keep up with dilution initially (up to approximately a thousandfold dilution); however, after that, calculated dissolved- phase concentrations decline due to dilution. The concentration at 1:70 dilution agrees with the lowest salinity result by Brown et al. (1995), as one would expect because the partition coefficients used here are for measurements in fresh water. The more dilute the fuel, the higher is the percentage dissolved (Table 2.6), such that the dissolved concentrations do not represent simple dilutions of the originally dissolved fraction in the neat fuel. The Brown et al. (1995) results indicate that dissolution is very rapid and is complete in less than one hour (at least at 1:70 dilution). TABLE 2.6 Percentage of MAH and PAH Dissolved from Orimulsion Diluted by the Indicated Factors, as Computed from an Equilibrium Partitioning Model Dilution MAH + PAH MAH PAH 1 0.00 0.07 0.0010 10 0.02 0.66 0.010 70 0.11 4.24 0.07 100 0.16 5.86 0.10 1,000 1.16 33.75 0.91 10,000 6.22 81.39 7.17 100,000 26.46 97.70 34.62 1,000,000 67.76 99.76 80.00 SOURCE: Data provided by D. French McCay, A.S.A., Narragansett, RI.

32 SPILLS OF EMULSIFIED FUELS Coalescence and Bitumen-Suspended Particulate Matter Interaction Because there have been no large accidental releases of emulsified bitumen, one can only speculate about the behavior of such a spill. The open-ocean turbulent energy spectrum and lack of boundaries are not easily reproduced in the laboratory. There could be other phenomena that appear only under full-scale conditions and that do not occur in laboratory studies or small field experiments. This uncertainty must be considered in describing expected bitumen coalescence and other Orimulsion spill issues. While the surfactant remains effective, there will be little coalescence of the bitumen droplets, in either fresh or salt water. However, salinity as low as 6 ppt can collapse and deactivate the surfactant. Specifically, dissolved salt dehydrates the EO sheath and compresses the electrostatic double layer around the bitumen droplets (Crosbie and Lewis, 1998a). The proportion of bitumen that disperses or coalesces will depend on the spill volume and conditions that prevail at the time of an Orimulsion spill. Closed- system flume studies (Ostazeski et al., 1998a, 1998b), using seawater with break- ing waves, produced a rapid increase in the mean bitumen droplet size. These larger droplets scavenged other bitumen droplets as they rose to the water sur- face, forming floating tar patties. Minimal coalescence occurred in similar tests using fresh water, presumably due to the higher effectiveness of the surfactant in fresh water. In laboratory studies in both brackish and salt water, the majority of the bitumen ended up floating on or near the surface. In salt water, turbulence contributes to two competing phenomena that affect the dispersed bitumen. On the one hand, the turbulence increases dispersion of the bitumen droplets. On the other hand, it also contributes to particle-particle collisions and coalescence leading to the formation of floating tar mats on the water surface. Stout (1999) predicted that the propensity for droplet coalescence is probably greater under lower-energy conditions. Anecdotal reports from ob- servers at small, open-water experimental spills indicate that as much as one- third of the bitumen will coalesce and surface. Higher-energy situations will lead to small tarballs, whereas lower-energy settings are more likely to produce larger tar patties or ropes. Spill volume and release rate will also be key factors, because larger instantaneous spills will have higher droplet concentrations and rates of interaction. This is an important area for future studies because the percentage of bitumen that surfaces or remains in the water column affects the method of cleanup and expected environmental impacts of the spill. For purposes of this report, an assumption has been made that coalescence and surfacing of the bitumen in salt water and brackish water will be significant. Interactions with Suspended Particulate Material Numerous batch-mode and flume studies have been undertaken to evaluate the interaction of dispersed bitumen droplets and suspended sedimentary material

BEHAVIOR AND FATES 33 (Bitumenes Orinoco, S,A., a; Brown et al., 1995; Johnson et al., 1998; Stout, 1999). Brown et al. (1995) examined the interaction of Orimulsion bitumen with fine and coarse sediment in saline water and reported that adhesion equilibrium reached up to 1 and 2.5 mg of bitumen per gram of coarse sediment at low and high sediment loads, respectively. Adhesion approached maximum values of 2,000 and 6,000 mg of Orimulsion bitumen per gram of fine sediment in batch equilibrium experiments. Orimulsion interactions with SPM are physical adhe- sion processes that seem to be favored by high salinity, sediment surface area, and/or high organic carbon content (Brown et al., 1995). In comparing the behavior of Fuel Oil No. 6 and Orimulsion, Brown et al. (1995) reported that the adhesion of No. 6 fuel oil to Tampa Bay sediments was negligible compared to Orimulsion. For 15-minute contact experiments, the fine sediment fraction showed approximately 1,400-mg/g loadings for Orimulsion and only 6 mg/g for No. 6 fuel oil. The coarse fraction loadings were 3.3 mg/g and 0.03 mg/g for Orimulsion and No. 6 fuel oil, respectively. Under brackish or full-strength seawater salinities, dispersed bitumen-SPM interactions can lead to formation of bitumen-SPM agglomerates that can be transported to the bottom (Brown et al., 1995; Stout, 1999). In flume studies conducted under extremely high-energy and full-strength seawater salinity conditions, up to 11 percent of the total bitumen was observed to be transferred eventually to the bottom in water containing high (45 mg/L) suspended loads of kaolinite (Stout, 1999). Similar experiments con- ducted in fresh water showed no transport of bitumen-kaolinite agglomerates to the bottom, with the majority of the dispersed bitumen remaining suspended in the water column, presumably because of the effectiveness of the surfactant in fresh water. Long-Term Fate and Microbial Degradation of Sedimented Bitumen Droplets It appears that Orimulsion is capable of being biodegraded although the rates are believed to be extremely slow. Brown et al. (1995) observed overgrowth of microbial life in Orimulsion weathering reactors containing Tampa Bay seawa- ter, and Lapham et al. (1999) confirmed the slow aerobic biodegradation of Orimulsion bitumen, with 1-3 percent degraded over a 21- to 60-day period. Compared to degradation rates for crude oil, where 10 percent of the alkanes and 50 percent of the PAH were degraded over similar time scales, it is clear that the rates for bitumen degradation are extremely low. However, the authors con- cluded that it was remarkable that any microbial respiration of the highly de- graded bitumen droplets occurred at all. They also observed that bitumen degra- dation was greater in the presence of the AE surfactant, which was believed either to act as a co-metabolite providing a readily available carbon and energy source to the bacteria or merely to increase bitumen reactivity by promoting its dissolu- tion in water. In a subsequent study, Proctor et al. (2001) found that the addition of seagrass and pinfish to sediment microcosms stimulated the in situ degradation

34 SPILLS OF EMULSIFIED FUELS of bitumen by as much as two- to ninefold. This suggests that bioremediation augmented by the addition of natural marine carbon substrates may be a viable option for responding to spilled Orimulsion in the marine environment. Their studies also demonstrated that the bacteria were not growing on Orimulsion but were simply respiring it as a co-metabolite. SHORELINE PROCESSES Orimulsion Spills on Land and Shorelines Fresh Orimulsion spilled directly onshore initially behaves similarly to heavy oils of like viscosity. It will tend to penetrate deeper into beaches that are wet and have gravel substrates. However, the mobility of Orimulsion decreases rapidly as it weathers and comes in contact with dry substrates, which could reduce the contamination threat to groundwater (AEA Technology, 1996). When spilled onto sand, the fresh emulsion is filtered, breaking the emulsion. The particles can penetrate approximately 4-5 cm before they occlude the available pore space, resulting in a maximum loading of approximately 55-60 percent bitumen in the surface layer on the contaminated sand (Harper and Kory, 1997). The water phase is available for percolation into the substrate, and the potential for ground- water contamination will be a function of local geology and spill size. When floating bitumen strands on shorelines and dries, it becomes stickier and will not resuspend with the tide. Weathered bitumen, which is significantly stickier than fresh Orimulsion, will not penetrate sand as readily and is expected to initially penetrate only the coarsest beach sediments (cobble-boulder). With surface warming of the bitumen-coated substrates by solar radiation, however, the weathered bitumen may become fluid enough to percolate deeper into sedi- ments (Harper and Kory, 1997). Stranding of Weathered Bitumen As long as bitumen that is dispersed in the water remains wet within the sediments, it will penetrate freely into pebble beaches but will remain on or near the surface of sand beaches (Harper and Kory, 1997). Unlike weathered bitumen patties, such dispersed bitumen is more mobile within the coarse-grained beach substrate than a typical heavy fuel oil. Dispersed bitumen can be flushed from sediments, but normal tidal flushing may not provide sufficient energy to remove it completely. Once the bitumen droplets have been exposed to air, they become much stickier and can form a tenacious coating on the surface of the beach sediments. This coating is difficult to remobilize, even if there is subsequent re- wetting of the bitumen. Bitumen that has already formed weathered patties or ropes on the water surface before reaching the shoreline is highly viscous and sticky, with many of

BEHAVIOR AND FATES 35 the same characteristics as weathered fuel oils. It is unlikely to penetrate into sediments finer than pebbles (Harper and Kory, 1997). Sticky bitumen patties can mix with sand suspended by waves. Mixing with more than 1 percent suspended sand can cause the droplets to become heavier than typical seawater and sink. USE OF MODELING AND SCENARIOS TO UNDERSTAND THE BEHAVIOR OF ORIMULSION SPILLS A widely used oil spill model has been modified to simulate possible Orimulsion behavior (French and Mendelsohn, 1995). This model has been used to evaluate the effects of hypothetical Orimulsion spills in Tampa Bay and Dela- ware Bay and River (French McCay and Galagan, 2001) and as an analysis tool in a series of Orimulsion test spills in the Caribbean (French et al., 1997). Although the model incorporates standard spill model approximations and assumptions, certain parameters specific to Orimulsion were estimated from laboratory or small field experimental data. One such key parameter is the rate of bitumen coalescence, which was estimated by doing an empirical curve fit to bitumen concentration (French, 2000). Since this curve fit uses confined sample data, it is most likely too large. Although the algorithms of the model have been published (French and Mendelsohn, 1995; French et al., 1996), the committee was unable to verify that the numerical code in the model accurately represents these algo- rithms. Other spill models have likewise been modified to simulate Orimulsion spill behavior using somewhat different approximations (VKI, 1999) and would pre- sumably provide different predictions, although no direct comparison of the dif- ferent models has been done. Barring an actual spill event, such model compari- sons provide guidance on the sensitivity of selected environmental and computational parameters in model forecasting and may be an area of future research. Because each spill is unique, even an actual spill would only validate a model involving those particular circumstances. Nevertheless, modeling an emul- sified bitumen fuel spill in different scenarios can be useful in sensitivity analysis and in determining where additional data are required. French and Mendelsohn (1995) used their model to assess the sensitivity of the model predictions to key factors such as bitumen coalescence. They also performed comparative studies of hypothetical Orimulsion and Fuel Oil No. 6 spills. They concluded that a large variety of scenarios are necessary to span the range of possible effects of emulsi- fied bitumen spills, both in an absolute sense and in a relative sense for compari- son with spills of other oil products. The committee selected six distinct spill scenarios to assess the expected behavior and fate of an Orimulsion spill. The scenarios developed by the committee were (1) marine—open water, (2) marine—nearshore, (3) estuarine (brackish water), (4) nontidal river, (5) fresh water—quiescent, and (6) on land near water. The emphasis of these scenarios

36 SPILLS OF EMULSIFIED FUELS is on the behavior and fate of the bitumen component because of its unique properties and behavior once spilled into water. The dissolved constituents are also important, but the processes of mixing and dilution of water-soluble com- pounds is well understood. Marine—Open Water Figure 2.8 presents the schematic fate diagram for open-water marine spills of Orimulsion. Away from the immediate vicinity of an open-ocean spill, the surfactant will be diluted or degraded rapidly so that the emulsified fuel will become an independent cloud of bitumen droplets. The dissolved components in the water phase will mix quickly into the surrounding water. In open-water settings, concentrations should rapidly decrease due to mixing and turbulence. Some of the bitumen will coalesce and make its way to the surface as either tarballs or tar patties. The remaining bitumen will generally disperse due to water turbulence. The surface tarballs or patties will be transported by winds and surface currents, in a fashion similar to tarballs from any weathered fuel oil slick. Being only slightly buoyant, they will be subject to overwash, making their observation by spill response personnel difficult. Tarball fields can be spread over long distances and subsequently become reconstructed in convergence zones. Thus, a line of stranded tarballs from the spill could appear on beaches far away from the spill site. Upon exposure to the air, the surfaced bitumen will greatly increase its adhesion properties and could attach itself to any floating material it encounters. Mesocosm roof-top experiments have shown that prolonged expo- sure to sunlight can eventually lead to sinking of surface bitumen patties or Clumps - Surveillance / Recovery of Floating Oil No Re-float Bitumen Inc Long Term Cloud Dec reasing Fate of reas S ing C preadi Suspended oale ng Particles - scen ce Fecal Deposition FIGURE 2.8 Schematic representation of the distribution and fate of Orimulsion-400 spilled into an idealized marine open water environment.

BEHAVIOR AND FATES 37 Standard / Innovative Shoreline Cleanup is Feasible Stranded Clumps - Spill Bitumen Surveillance / Recovery Bitumen / Sand Rollers Long-term Fate: Suspended Some Sediment / Suspended Bitumen Interaction (?%) FIGURE 2.9 Schematic representation of the distribution and fate of Orimulsion-400 spilled into an idealized marine nearshore environment. tarballs (Brown et al., 1995). Suspended bitumen and tar particles may also be subject to zooplankton grazing and, ultimately, sedimentation in fecal material. Bitumen droplets are of a size range similar to single-celled algae (i.e., phytoplankton) and natural silt particles, which are fed on by filter feeders. How- ever, most filter feeders studied to date have proven to be quite selective in their choice of food, and they can reject mineral and other particles. Copepods have been known to ingest droplets of Bunker C fuel oil following an oil spill (Conover, 1971), and up to 10 percent of the hydrocarbons in the water column were associated with the plankton and their feces. Laboratory studies with crustacean zooplankton, primarily copepods, demonstrated uptake of a variety of aromatic and paraffinic hydrocarbons from oil-contaminated food or water (Corner et al., 1976a, b; Lee, 1975); however, there are no data to show if similar behavior might be expected from ingested bitumen droplets. Marine—Nearshore Figure 2.9 presents the schematic fate diagram for nearshore coastal marine spills of Orimulsion, which are expected to behave differently than open-water spills. Because of the land and bottom boundary conditions, vertical diffusive mixing could be reduced and pockets of higher near-surface concentrations of bitumen and dissolved components could persist longer than would occur in open water. This would encourage a greater rate of coalescence and the formation of surface slicks. If the bitumen does resurface in quiescent water, it will tend to form a thin film of less than 0.1 mm (Sommerville, 1999) and can easily be resuspended if sufficient energy becomes available. Surface slicks of weathered

38 SPILLS OF EMULSIFIED FUELS bitumen can also become stranded at the high-water mark on the shoreline during a falling tide and onshore winds. Once deposited, they would be expected to weather as described in the previous section on shoreline interactions. If large patties or ropes of the weathered bitumen mix with sand suspended in the surf zone, they may form large “sand rollers,” which are mixtures of tar and sand that can roll down and populate the nearshore bottom. Once deposited, they may be subject to burial in offshore bars. They could also become buried in the intertidal zone during depositional phases on beaches. Suspended bitumen droplets would probably not strand on the beach but would be transported along the shoreline by the alongshore current. They could, however, be scavenged by suspended particulate matter as described earlier. In any segment of coastline that is of lower energy (or in calmer offshore waters), these bitumen/ suspended sediment combinations could sink. Suspended or dispersed bitumen droplets that are carried offshore would be subject to the same long-term fate as bitumen droplets produced from an open-ocean Orimulsion spill. Estuarine (Brackish Water) Figure 2.10 presents the schematic fate diagram for an Orimulsion spill in estuarine or brackish water environments. Although the surfactant in Orimulsion will degrade in water with salinity greater than 5-7 ppt, the dispersed bitumen droplets will typically not become buoyant until the salinity of the surrounding water is 15 ppt or greater (Deis et al., 1997). Thus, it is possible for the bitumen droplets to begin to coalesce and then either float to the surface, sink to the bottom, or remain neutrally buoyant, depending on the encompassing water den- Spill Surveillance / Minor Clumps in Water Intake Closures Convergence Zones River Some Sedimentation (? %) FIGURE 2.10 Schematic representation of the distribution and fate of Orimulsion-400 spilled into an idealized estuarine (brackish water) environment.

BEHAVIOR AND FATES 39 sity. If the water is stratified and quiescent, it is possible that increased concen- trations of bitumen could form at the boundary between the less dense and more dense water. A likely scenario for an Orimulsion spill in a brackish, low-energy estuary of varying salinity would be that a small amount of the bitumen would coalesce and float to the surface, some would be scavenged by suspended particu- lates, and most would either sink very slowly to the bottom or remain in the water column, gradually being flushed by the normal water exchange in the estuary. Dissolved components (PAH, surfactants) are expected to decrease at rates deter- mined by the degree of tidal mixing and flushing. If the water is sufficiently brackish to cause some of the bitumen to float, this weathered floating bitumen could adhere to estuarine vegetation and impact shore- lines. Bitumen/particle agglomerations could settle into the bottom sediment in quiescent areas where fine sediment accumulation occurs. If there is insufficient energy to cause coalescence, any buoyant bitumen may float to the surface but be subject to resuspension if disturbed. Suspended and dispersed bitumen droplets that are carried out of the estuary to coastal or offshore waters would be subject to the same long-term fate as bitumen derived from an open ocean Orimulsion spill. Nontidal River Figure 2.11 presents the schematic fate diagram for an Orimulsion spill in a nontidal energetic riverine environment. A river spill of Orimulsion would typi- cally result in rapid mixing of the bitumen and dissolved components throughout the water column with subsequent rapid dilution. The surfactant would usually remain effective for a sufficient period of time to allow the bitumen droplets to Surfactant Attached Spill Surveillance / to Bitumen Water Intake Closures Debris Piles Scour Pit Minor Sedimentation FIGURE 2.11 Schematic representation of the distribution and fate of Orimulsion-400 spilled into an idealized nontidal river environment.

40 SPILLS OF EMULSIFIED FUELS become effectively dispersed to low enough particle densities to discourage coa- lescence. If there is a heavy sediment load in the river, there could be a minor amount of scavenging of the bitumen, although laboratory flume experiments have indicated that bitumen-SPM interactions are insignificant in fresh water conditions. There would most likely be little sinking of the pure bitumen outside of any previously formed scour pits or other naturally quiescent pools, where the droplets might temporarily collect but could easily be resuspended. Water in- takes would have to be closed until the cloud of droplets passed. Flood condi- tions, a not uncommon situation in spill accidents, could strand bitumen-laden water on inundated floodplains. As the water drained off, the bitumen could remain on the top sediment layer. The long-term fate under normal conditions may be deposition in calm deltaic areas where other fine sedimentary material accumulates, although there are no data to support this hypothesis. Fresh Water—Quiescent Figure 2.12 presents the schematic fate diagram for an Orimulsion spill into a freshwater region with low or near-zero currents. A spill of Orimulsion into a quiescent freshwater pool would be similar to a spill in a river with less mixing and essentially no bitumen-SPM interaction. As in the river spill, the surfactant would probably remain effective long enough to allow the bitumen droplets to diffuse to a low enough concentration to inhibit coalescence, as long as the receiving water was of sufficient volume to allow breakdown of the emulsion. However, the low mixing energy would delay this process. The PAH fraction in the carrier aqueous phase would be subject to limited dilution if the volume Cleanup Options are Viable e to ang r xch n Wate E e Op - Settles in Days Innovative - Not Stickier Recovery of as Long as Wet Suspended Bitumen FIGURE 2.12 Schematic representation of the distribution and fate of Orimulsion-400 spilled into a quiescent fresh water environment.

BEHAVIOR AND FATES 41 Sticky Where Dry Standard & Innovative Response Option and in Contact with Sediment / Vegetation Emulsion Skin Diluted River Tank Liquid Penetration Particles Penetrate - Spread Like Typical Crude Permeable Sediments - Some Settling of Large Particles FIGURE 2.13 Schematic representation of the distribution and fate of Orimulsion-400 spilled into an idealized land-to-water (through wetlands) environment. fraction of the Orimulsion were significant compared to the freshwater pool volume. Given a breakdown of the emulsion, some of the bitumen could settle temporarily to the bottom, but it could also be resuspended easily. If there were any turbulent energy, the bitumen droplet concentration would be nonzero throughout the water column, with the settling motion of the bitumen offset by random particle motion. For a spill on the surface of a pond, the steady-state vertical concentration of droplets could be an exponentially decaying function upward from the bottom of the pond (Hemond and Fechner, 1994). On Land Near Water Figure 2.13 presents the schematic fate diagram for an Orimulsion spill on land with subsequent migration through a wetland into a riverine environment. If Orimulsion is released on land, it would be expected to behave initially like a fuel oil with a similar viscosity. Penetration of fresh Orimulsion into soil and sandy sediments would be limited to the upper few centimeters, as described in the discussion of Orimulsion spills on land and shorelines. The suspended bitumen droplets are filtered out of the spilled material by the sand or soil, whereas the dissolved-phase surfactants and low-molecular-weight PAH constituents may percolate deeper into the soil and could eventually interact with groundwater. With prolonged exposure to air, the bitumen droplets remaining in the upper soil layers would dry, increasing their stickiness and viscosity, and any pooled Orimulsion would eventually turn into tar mats. If sufficient Orimulsion were spilled to eventually reach a freshwater body, some of the bitumen would be subject to dispersion in the receiving water body. However, under calm condi-

42 SPILLS OF EMULSIFIED FUELS tions, it would most likely behave like a highly viscous, high-density oil and settle toward the bottom in tar patties, tarballs, or tar mats, depending on the spill and environmental circumstances. The dissolved PAH fraction in the aqueous phase would be subject to limited dilution, and it could remain in the receiving waters at higher concentrations than in any of the other scenarios considered. SUMMARY OF THE BEHAVIOR AND FATE OF SPILLED ORIMULSION All of our understanding of the behavior of spilled Orimulsion is based on small-scale laboratory studies, flume tests, small (2-10 barrels) experimental spills in harbors and open water, and computer models developed from these data. Therefore, the ability to predict what happens in a real, large spill event remains limited, and responders should be prepared for a wide range of possibilities for response and cleanup. To provide better prediction tools during spills, the different Orimulsion models should be compared for the same spill scenarios and the strengths and weaknesses of each model should be evaluated. Orimulsion behavior varies significantly when spilled into fresh water or into salt water due to the denaturing of the surfactant when the salinity of the receiving water is greater than 5-7 ppt. As a result, there is a greater tendency for dispersed bitumen droplets to coalesce and surface in brackish or salt water. The competing processes of coalescence (with possible surfacing or sinking) and dispersion into the water column dictate the behavior of bitumen droplets from an Orimulsion spill in marine or brackish water. Therefore, further research must be done to quantify the processes of coalescence and dispersion of the bitu- men droplets. In particular, the role of turbulent energy (magnitude and structure), salinity, spill volume, and spill rate should be evaluated. For spills into open water, Orimulsion would quickly behave as a cloud of dispersed bitumen droplets that are separated from the dissolved PAH and surfac- tants in the 30 percent water phase, which quickly mixes into the receiving water body. The bitumen droplets chemically resemble the residue that would be found in a heavy fuel oil spill at the end of short-term weathering processes. The concentrations of dispersed bitumen droplets and water-soluble constituents de- crease significantly due to dispersion into the surrounding water column. Predic- tions of the dispersion are dependent on knowledge of the vertical and horizontal diffusion parameters that are key factors used in computer models. To improve the accuracy of model predictions in support of local spill response plans, site-specific studies are needed to define diffusion coefficients (energy dissi- pation rates) for areas where Orimulsion shipment and/or loading and offloading operations are planned. The PAH concentration and chemical composition of bitumen are similar to the end product or heavily weathered residues from most crude oil spills. Signifi- cant quantities of the water-soluble PAH constituents have already leached from

BEHAVIOR AND FATES 43 the bitumen droplets. Long-term leaching of PAH can occur at appreciable rates given the high surface-area-to-volume ratios of these 1- to 80-µm-diameter bitu- men droplets. However, given the relatively low concentration of the high- molecular-weight PAH in the bitumen itself, this continued leaching is not ex- pected to be environmentally significant. Limited data suggest that bitumen droplets may interact with SPM to a greater extent than No. 6 fuel oil (which does not break up into small droplets and remains suspended in the water column). Because the bitumen droplets are recalcitrant and remain in the environment for a long time, their interactions with SPM and retention in nearshore, estuarine, and riverine sediments may be important. Additional dispersed bitumen-SPM stud- ies using a wider variety of suspended particulate material types (including organic-rich substrates) at different salinities are recommended. The ultimate fate of spilled bitumen droplets is sedimentation, where contin- ued biodegradation of the bitumen droplets is expected to be extremely slow. Based on the highly weathered nature of bitumen, the bioavailability of PAH is likely to be low. However, given the high surface-area-to-volume ratio of the very fine bitumen droplets, the potential for PAH uptake exists and thus should be evaluated. For Orimulsion spills on land, the water phase can separate from the bitumen and infiltrate the underlying substrate. Because low- and intermediate-molecu- lar-weight PAH and surfactants reside in the aqueous phase, they may reach groundwater. The potential for groundwater contamination is very site specific and outside the scope of this study. Recent literature shows that the surfactant AE are subject to a variety of microbial breakdown pathways to highly oxygenated water-soluble intermedi- ates. Identified breakdown products include fatty acids, alcohols, carboxylated polyethylene glycols, and (for branched AE) intermediate compounds with car- boxylic acid groups on one or more alkyl groups (side chains) as well as the terminal end of the polyethylene oxide moiety. Branched AE such as those found in Orimulsion degrade slowly, and intermediate products can still be isolated (at low concentrations) from laboratory experiments after weeks to months. There- fore, AE and their intermediate degradation compounds can persist for weeks to months after a release. Federal and state agencies should consider developing information on the ambient concentration of these compounds and their degradation products in the environment.

Next: 3 Ecological Effects: Summary and Evaluation of Available Information »
Spills of Emulsified Fuels: Risks and Response Get This Book
×
Buy Paperback | $45.00 Buy Ebook | $36.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Spills of Emulsified Fuels: Risks and Response is part of an evolving body of work conducted by the National Research Council (NRC) to help inform debate and decision-making regarding the ecological consequences of releases associated with the widespread use of fossil fuels. Like earlier NRC reports, it attempts to understand the chemical, physical, and biological behavior of a complex mix of compounds that make up various petroleum hydrocarbon-based fuels. The specific risk factors presented by emulsified fuels are difficult to characterize, mainly because there have been no spills of emulsified fuels to date, and thus there is little practical experience with these products.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!