E
Analysis of the Sensitivity of Dispersed Oil Behavior to Various Processes

There are many complex mechanisms interacting to control oiltransport and fate. They include oil surface spreading, evaporation, entrainment, emulsification, horizontal/vertical advection and natural diffusion (affected by current, wind, wave), sedimentation, oil droplet rising (due to buoyancy), biodegradation, dissolution, and chemical dispersion and associated oil droplet-size changes.

It is very difficult to quantify the importance of these mechanisms on oil concentrations changing with space and time. Decisionmakers contemplating potential use of dispersants must know where and how fast spilled oil is migrating. It is, therefore, important to evaluate how these mechanisms affect oil transport. As discussed in Chapter 4, oil transport and fate models integrate major controlling physical, chemical, and biological processes into one system to identify cause-effect relationships and to evaluate effects of these mechanisms on oil concentrations. Therefore, it was decided to conduct a sensitivity analysis as part of the committee’s work on evaluating the existing literature.

The objective of the sensitivity analysis conducted here was to understand the role that various processes have in the transport and fate of spilled oil, both with and without the use of chemical dispersants, and to determine the sensitivity of this behavior to these processes. While some sensitivity was discussed analytically in earlier parts of Chapter 4, an integrated sensitivity analysis can only be performed with a comprehensive computer model that includes all of the relevant processes. The sensitivity modeling analysis discussed here was conducted for 14 cases with various oil types (crude oil and light and heavy refined oil), environmental



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects E Analysis of the Sensitivity of Dispersed Oil Behavior to Various Processes There are many complex mechanisms interacting to control oiltransport and fate. They include oil surface spreading, evaporation, entrainment, emulsification, horizontal/vertical advection and natural diffusion (affected by current, wind, wave), sedimentation, oil droplet rising (due to buoyancy), biodegradation, dissolution, and chemical dispersion and associated oil droplet-size changes. It is very difficult to quantify the importance of these mechanisms on oil concentrations changing with space and time. Decisionmakers contemplating potential use of dispersants must know where and how fast spilled oil is migrating. It is, therefore, important to evaluate how these mechanisms affect oil transport. As discussed in Chapter 4, oil transport and fate models integrate major controlling physical, chemical, and biological processes into one system to identify cause-effect relationships and to evaluate effects of these mechanisms on oil concentrations. Therefore, it was decided to conduct a sensitivity analysis as part of the committee’s work on evaluating the existing literature. The objective of the sensitivity analysis conducted here was to understand the role that various processes have in the transport and fate of spilled oil, both with and without the use of chemical dispersants, and to determine the sensitivity of this behavior to these processes. While some sensitivity was discussed analytically in earlier parts of Chapter 4, an integrated sensitivity analysis can only be performed with a comprehensive computer model that includes all of the relevant processes. The sensitivity modeling analysis discussed here was conducted for 14 cases with various oil types (crude oil and light and heavy refined oil), environmental

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects conditions (wind speed, waves, diffusion), chemical dispersion effectiveness, and oil droplet sizes in nearshore and offshore Florida Coast. Biodegradation was not modeled in the sensitivity analysis, because models were run for only one or two simulation days. Although the sensitivity results are somewhat dependent on the test conditions and simulation codes selected for this evaluation, they are good indicators of the importance of the model parameters tested. The sensitivity analysis helps identify knowledge gaps, future research needs, and a new approach (potential of using models to assist on-scene decisionmakers for the possible dispersant use) to assess dispersant use. The analysis also can enhance our knowledge and understanding of the combined effects of these mechanisms on oil transport and fate, and beneficial and adverse impacts of using chemical dispersants. MODEL SETUP Code Selection The two codes that were considered are SIMAP (French-McCay, 2003, 2004), and the combined use of ADIOS2 (Lehr et al., 2002) and Lagrangian 3-D GNOME (Simecek-Beatty et al., 2002); both simulate most mechanisms needed to assess dispersant use, except dispersant effectiveness itself and changes in the oil droplet sizes (which are user inputs). The latter codes are used by NOAA for their real-time response to an actual oil spill. The flow field is not simulated by 3-D GNOME, but it is supplied to the code as a model input. Although the code can accept a three-dimensional velocity distribution, NOAA usually uses a simpler two-dimensional flow field, balancing pressure forces, bottom friction, Coriolis force, and water density variation, adjusted by tide and wind for a real-time emergency response. Their capability to reflect a three-dimensional flow field in a real-time emergency response needs to be improved, especially in complex nearshore areas. Oil trajectory predictions during a spill are constantly adjusted to match observed trajectories by re-adjusting the model input, including the velocity and wind fields. Because of the availability of the codes within the public domain, their ease of use, limited requirements to operate the models, and their potential use for determining dispersant use during an oil spill event, ADIOS2 and 3-D GNOME were used for the sensitivity analysis. However, this investigation was not intended to evaluate these codes, but rather to use their simulation results as indicators for sensitivity of oil concentrations to various transport and fate processes. Because the sensitivity results may reflect specifics of these two codes, additional sensitivity analysis is recommended with other codes.

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects Water Body There are many different types of water bodies: nearshore, offshore, semi-confined water, estuaries, lakes, and rivers. Because nearshore environments may impose more difficulties for decisionmakers regarding the appropriateness of dispersant use (Reed et al. 2004), this evaluation focused more on nearshore waters with few additional offshore model runs. However, other water bodies should also be evaluated. A somewhat schematic representation of the southern Florida coast was selected to represent a relatively simple nearshore region. A coastal flow moving northeast along the Florida Keys from Key West toward Key Largo was imposed as the geostrophic flow without accounting for wind and tide. The oil spill location (25°01’N, 80°23’W) was selected to be in 10-m-deep water about 9 km offshore. For some sensitivity test cases, the oil was assumed to be spilled further offshore in 200-m-deep water (25°01′N, 80°11′W). Oil Types Three types of oil were selected: Alaska North Slope crude oil, Intermediate Fuel Oil 300 (IFO 300), and marine diesel primarily used in the southern United States for commercial marine operations. It is useful to evaluate other crude oils, including heavy California crudes, light and heavy crudes imported from Africa, Mexico, Venezuela, and elsewhere. But the Alaskan Crude, IFO 300, and diesel were selected, in part, due to the availability of data on their physical and chemical properties. Their physical properties built into the ADIOS2 code (Lehr et al., 2002) are shown in Table E-1. These oils also represent widely varying chemical components, as indicated in their distillation cuts (also built into the code) TABLE E-1 Densities and Viscosities of Three Oils Characteristics of Oil Alaska North Slope Crude Oil IFO 300 Diesel Oil Density, g/mL 0.8936 at 15°C 0.9859 at 15°C 0.8362 at 15°C Viscosity, cP 23.0 at 15°C 14,470 at 15°C 4.0 at 15°C Surface Tension, Oil-water 26.1 at 0°C — 24.9 at 0°C dynes/cm Oil–seawater 23.8 at 0°C 37.3 at 15°C 24.6 at 0°C Percent Oil Evaporation to Initiate Emulsification 18 100 100

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects shown in Table E-2. Intermediate fuel oil is heavy and viscous, and marine diesel is light and much less viscous. Alaska North Slope crude falls between them in density and viscosity. IFO 300 and this particular marine diesel have more higher-molecular-weight components than Alaska North Slope crude oil, as indicated by higher distillation temperatures assigned in ADIOS2. For the sensitivity analysis, 10,000 barrels (roughly 1,500 tonnes of crude; 1,600 tonnes of IFO 300; 1,400 tonnes of diesel) of oil were assumed to be released to the surface of this water body over one hour. Environmental Conditions Selected environmental variables were wind speed, wave height, horizontal diffusion, and vertical diffusion. These variables can either be entered independently, or the latter three can be computed by the model as a function of the wind speed and current, using relationships such as Morales et al. (1997) described in Lehr et al. (2002) and Simecek-Beatty et al. (2002). The latter option was used here, resulting in the following values: TABLE E-2 Distillation Cuts of the Three Oils Used in the Modeling Sensitivity Analysis Oil Cut Number Alaska North Slope Crude Oil Intermediate Fuel Oil 300 Diesel Fuel Oil Weight Fraction, wt percent Temperature, °C Weight Fraction, wt percent Temperature, °C Weight Fraction, wt percent Temperature, °C 1 1.0 42 1.1 180 1.1 120 2 4.0 98 1.1 200 1.1 140 3 5.0 127 6.4 250 1.1 160 4 5.0 147 9.4 300 3.2 180 5 5.0 172 7.2 350 5.2 200 6 10.0 216 8.1 400 20.4 250 7 10.0 238 6.0 450 31.9 300 8 5.0 247 3.0 500 25.5 350 9 5.0 258 4.9 550 9.7 400 10 5.0 265 9.8 600 1.0 450 11 5.0 272 14.7 650 — — 12 10.0 282 10.7 700 — — 13 30.0 >282 17.4 >700 — —   SOURCE: Data from Environment Canada, 2005.

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects 2, 10, and 25 m/s for the speed of south-southwesterly wind 0, 0.9, and 5.5 m for the breaking wave height 160, 350, and 660 cm2/s for the horizontal diffusion coefficient 23, 51, and 97 cm2/s for the vertical diffusion coefficient. The mixing depth was selected to be 10 m for the oil releases on the 10-m-deep water, while it was assigned to be 200 m for the oil releases in deep water. The mixing depth can represent the maximum depth of surface mixing (e.g., by Langmuir circulation) or a diffusion floor imposed by a pycnocline or thermocline, though neither Langmuir circulation nor density stratification was simulated in the modeling. The vertical diffusion coefficient through and below the mixing depths was assigned as 0.11 cm2/s. The diffusion coefficients calculated by ADIOS2 and 3-D GNOME above the mixing depth may be greater than realistic values, especially in offshore deep water. Moreover they are temporally and spatially constant. While it is possible to estimate surface diffusivities using real-time surface current data (e.g., Ojo and Bonner, 2002), for simplicity and comparative purposes, the above values were used for both the nearshore and offshore modeling cases. Dispersant Application The sensitivity analysis was performed with and without dispersant applications. It was assumed that dispersant application began six hours after the release and it took six hours to treat the entire slick, even though in a real spill case, there may be some operational limitations that prevent complete areal coverage. Because the dispersant effectiveness is the measure of how much oil on the water surface is entrained into the water column below, the oil concentration in the water column is almost directly proportional to the dispersant effectiveness. Thus, the sensitivity analysis did not vary the dispersant effectiveness. It was arbitrarily assumed that 50 percent of the oil on the water surface would be “dispersed” (entrained into the water column from the water surface), and that the oil droplet diameters would be the same as, or reduced by a factor of five from the baseline sizes, resulting in slower rise toward the water surface. The baseline oil droplet sizes were assigned to be between 10 and 70 µm. Although the dispersant does not increase the oil droplet sizes, two additional distributions of larger droplet sizes (two and four times the baseline sizes) were also selected to evaluate an effect of oil droplet sizes on the vertical oil distribution. These four oil droplet-size distributions, each including seven droplet sizes, are shown in Figure E-1. Figure E-2 shows the representative rising velocity of an oil droplet of Alaska North Slope crude oil in 20° C (roughly 68° F) sea water (kinematic viscosity

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects FIGURE E-1 Oil droplet-size distributions. of 1.1 × 10−6 m2/s). NOAA’s ADIOS2 code includes oil and sediment interaction. For this modeling the suspended sediment concentration was assigned to be 5 mg/L. However, these models do not include oil-particle interaction, which may be important in estuaries and high energy coastal waters with varying salinity and high concentrations of suspended organic and inorganic matter. Modeling Results and Evaluation Sensitivity analyses were performed by combining the three oil types, three wind speeds, three wave heights, three horizontal diffusion coefficients, and three vertical diffusion coefficients with and without dispersant application, as well as four oil droplet-size distributions with dispersant applications. Table E-3 shows the 14 test cases for the sensitivity analysis. Cases 1 and 2 are baseline cases for the crude oil without and with a dispersant. Note that oil droplet sizes in Case 2 were assigned to be 20 percent of those of Case 1. Cases 3 through 6 are for environmental changes without and with a dispersant application. Cases 7 though 10 are for three oil types without and with a dispersant. Cases 1 through 10 are those whose oil releases are in 10-m-deep water (nearshore), while Cases 11 through 14 have 200 m of water depth (offshore) at the release point with four sets of oil droplet-size distributions.

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects FIGURE E-2 Oil droplet rising velocity. Note: Oil viscosity in a surface slick changes as the oil weathers due mainly to emulsification. Thus, no single value for oil viscosity was applied to calculate the rising velocity. The rising velocities shown here were estimated with the viscosity (1.07 × 10−6 m2/s) of sea water at 20°C. The sensitivity modeling was performed for 24 simulation hours for the nearshore cases and 48 hours for the offshore cases. Simulation results for evaporation, natural dispersion, chemical dispersion, amount floating, and amount beached at 14, 24, and 48 hours are summarized in Table E-4. Oil volume percent may be converted to weight percent using oil density. However, as will be discussed later, oil density changes with time due to selective evaporation of lighter-molecular-weight components and emulsification. These sensitivity analysis results are discussed below under four categories: (i) Main characteristics of oil transport and fate with and without a dispersant (Cases 1 and 2), (ii) Environmental conditions (Cases 3-6), (iii) Oil types (Cases 7-10), and (iv) Oil droplet sizes (Cases 11-14). Model results of Case 1 (without a dispersant) and Case 2 (with a dispersant) are shown in Figure E-3, as an example of predicted oil distributions shown in Table E-4.

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects TABLE E-3 Model Input for Fourteen Sensitivity Test Cases Case No. Disp. Use Oil Type Wind Speed (m/s) Wave Height (m) Hori. Diff. Coeff. (cm2/s) Vert. Diff. Coeff. (cm2/s) Amount Evapor. to Start Emulsif. (percent) Oil Dropl. Size as Percent of Baseline Sizes (percent) Water Depth at Release Point (m) 1 No Crude 10 0.9 350 51 18 100 10 2 Yes Crude 10 0.9 350 51 18 20 10 3 No Crude 2 0 160 23 18 100 10 4 No Crude 25 5.5 660 97 18 100 10 5 Yes Crude 2 0 160 23 18 20 10 6 Yes Crude 25 5.5 660 97 18 20 10 7 No IFO300 10 0.9 350 51 100 100 10 8 No Diesel 10 0.9 350 51 100 100 10 9 Yes IFO300 10 0.9 350 51 100 20 10 10 Yes Diesel 10 0.9 350 51 100 20 10 11 Yes Crude 10 0.9 350 51 18 20 200 12 Yes Crude 10 0.9 350 51 18 100 200 13 Yes Crude 10 0.9 350 51 18 200 200 14 Yes Crude 10 0.9 350 51 18 400 200 Main Characteristics of Oil Transport and Fate With and Without a Dispersant Cases 1 and 2 have a wind speed of 10 m/s with corresponding wave and diffusion coefficients (see Table E-3). The 3-D GNOME code prescribes that the oil slick is advected with a speed equal to the underlying current velocity plus 1 ~ 4 percent of the wind speed in the direction of the wind, while oil in the subsurface water is carried by the underlying current. For Case 1 without a dispersant application, Figure E-4 shows the location of the predicted oil plume 24 hours after the spill and the oil spill location marked by “+.” Black spots represent oil floating on the water surface, that traveled about 50 km from the spill site over 24 hours. The colored areas show different ranges of oil concentrations in the top 1 m of the water column; note that oil plume in the water column is following a different trajectory than the oil on the surface. The figure also indicates the area of the top 1 m of water column containing oil to be roughly 25 km2. Evaporation accounted for the loss of 34 volume percent of 10,000 barrels of the spilled Alaska North Slope crude oil over 24 hours, while only 3 volume percent was naturally entrained (dispersed) into the water column, as shown in Figure E-3. The remaining 63 percent was floating on

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects TABLE E-4 Summary of the Simulation Results at 14, 24, or 48 Hours after the Oil Spill Case No. Oil Distribution Evaporation (volume percent) Natural Dispersion (volume percent) Chemical Dispersion (volume percent) Floating (volume percent) On Beach (volume percent) 1a 34 3 0 63 0 2a 31 3 37 29 0 3a 36 0 0 64 0 4a 30 31 0 38 1 5a 30 0 42 28 0 6a 28 31 24 17 0 7a 10 0 0 90 0 8b 18 73 0 9 0 9a 8 0 49 43 0 10b 21 44 34 1 0 11c 35 3 37 25 0 12c 35 3 37 25 0 13c 35 3 37 25 0 14c 35 3 37 25 0 aValues are at 24th simulation hour. bValues are at 14th simulation hour because no floating oil existed after 16 hours. cValues are at 48th simulation hour. FIGURE E-3 Predicted oil distributions at 24 hours after the release of Alaskan North Slope crude oil with (Case 2) and without a dispersant (Case 1).

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects FIGURE E-4 Predicted oil movement at 24 hours after the release at point + for Case 1. the water. Natural dispersion caused by wind, waves, and current is the only mechanism in Case 1 to disperse oil into the water column. One aspect of complexity comes from the fact that oil consists of a wide range of hydrocarbons (see Table E-2). Although oil toxicity comes from the cumulative impacts of multiple hydrocarbon components, low

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects and intermediate-molecular-weight components such as BTEX and PAH tend to cause more acute risks to aquatic biota, as discussed in Chapter 5. These components usually evaporate faster and to a greater extent than higher-molecular-weight components such as wax, resins, and asphaltenes. The latter are contributing components in the formation of mousse, which makes it more difficult for a dispersant to work effectively (see Chapter 3). Figure E-5 shows the predicted composition (a relative volume fraction of each distillation cut) of the Alaska North Slope crude oil floating on the water surface at different times after the spill for Case 1. Because of evaporation, the oil composition after the release is changed from its initial composition. This figure shows that cuts with lower distillation temperatures evaporate faster and more completely than those with higher distillation temperatures, thus increasing the relative percentage of the latter components with time. For example, the heavy distillation cut #13 increased its relative volume fraction from the initial 30 percent of the oil to 46 percent under 10 m/s wind speed (Case 1) after 24 hours on the water surface. At that time, 63 volume percent of the oil was floating on the surface, thus potentially available to reach shorelines. As shown in Figure E-5, all of the first three cuts and most of cut #4 (those distilled at 147° C [roughly 296° F] or lower) evaporated within six hours. Compounds present in cuts 1–3 (e.g., below 127° C [roughly 260° F]) would include alkanes with fewer than 8 carbons and the monocyclic aromatics, benzene and toluene. Additional compounds present in cuts 1–5 (e.g., be- FIGURE E-5 Predicted composition of floating Alaskan North Slope crude oil with a dispersant under 10-m/s wind at various times (Case 1).

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects The predicted volume fraction of each distillation cut and the total volume of the floating Alaska North Slope crude oil over the first 24 hours after the oil spill for Case 2 were very similar to the undispersed oil slick, as shown in Figure E-7. As discussed above, most of the first five cuts (those distilled at 172° C [roughly 341° F] or lower—including alkanes with <10 carbons plus benzene and toluene, and most of the C2- and C3-substituted benzenes) evaporated within six hours, before the dispersant was applied. Thus, the dispersant applied from 6 to 12 hours after the oil spill does not introduce these chemicals to the water column. The remaining oil in both Cases 1 and 2 is mostly composed of heavier molecular-weight cuts. Environmental Conditions Wind and currents affect waves and diffusion in horizontal and vertical directions. As the wind becomes stronger, more oil is entrained into the water column. Consequently, less oil floats on the water surface and less oil is available for evaporation, although it is somewhat counterintuitive for less evaporation with stronger wind. Table E-4 and Figure E-8 show the strong effect of wind on oil entrainment, with the percentage entrainment varying from 0 percent at 2 m/s (Case 3), to 3 percent at 10 m/s (Case 1) to 31 percent at 25 m/s (Case 4). Oil concentrations in the water column vary depending on the amount of oil naturally dispersed (entrained), but they also reflect the diffusivity (which increases at higher FIGURE E-7 Predicted composition of floating Alaskan North Slope crude oil with a dispersant under 10-m/s wind at various times (Case 2).

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects FIGURE E-8 Predicted oil distributions 24 hours after the release of Alaskan North Slope Crude (no dispersant applied) under 2-, 10-, and 25-m/s wind. There is no oil dispersed by a chemical dispersant for these three cases. wind speed) in the water column. As discussed in Chapter 4, there is significant uncertainty in the structure of vertical diffusion, and this is manifest in uncertainty in the subsurface concentration of dispersed oil. This example reveals complexity of wind and currents controlling waves and diffusion in horizontal and vertical directions, affecting oil movement. Similar to Case 1 (10-m/s wind), Case 4 (25-m/s wind) caused all or most of the first five cuts to be evaporated, as shown in Figure E-9, which presents the predicted composition (relative volume fraction of each distillation cut) of the Alaska crude oil floating on the water surface. The combined effects of wind (thus wave energy and diffusion) and the use of dispersants are examined by comparing results of Cases 2, 5 and 6 (wind speeds of 10, 2, and 25 m/s with dispersants) with Cases of 1, 3, and 4 (without chemical dispersants). An interesting aspect of these cases is that as the wind increases from 2 m/s to 10 m/s to 25 m/s, more natural dispersion occurs, resulting in less available oil on the water surface to be dispersed by the chemical dispersant (see Table E-4). However, these model results may be artifacts of the dispersion modeling, because the dispersant effectiveness is the measure of how much oil on the water surface is entrained into the water column below and the same 50 percent dispersant efficiencies were imposed in these three cases regardless of the wind conditions. In reality, stronger wind tends to increase dispersant effectiveness, at least up to a certain wind speed. As stated previously, the

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects FIGURE E-9 Predicted composition of floating Alaskan North Slope crude oil without a dispersant under 25-m/s wind at various times (Case 4). need to specify dispersant effectiveness as a model input is the weakest part of the dispersant application assessment, yet unfortunately the dispersant effectiveness is probably the most important parameter. Effect of Oil Type Cases 7 and 8 have the same conditions as Case 1 except for the oil type: Alaska North Slope crude oil in Case 1, heavier IFO 300 in Case 7, and lighter diesel in Case 8. The Alaska North Slope crude oil started to emulsify when 18 percent of the oil was evaporated. On the other hand, the IFO 300 and diesel are not expected to emulsify. The IFO 300 and diesel evaporated 10 volume percent and 18 volume percent, respectively, over 24 and 14 hours (see Figures E-10 and E-11 with and without a dispersant), which was less than the 34 volume percent for Alaska North Slope crude oil. This may be expected from Table E-2, which shows that these refined oil products have a low percentage of low-temperature distillation cuts. Moreover, the IFO 300 did not disperse into water due to its high viscosity (~ 15,000 cP). Diesel with very low viscosity (~4 cP), on the other hand, was greatly dispersed (73 percent), and after 16 simulation hours, there was no oil floating on the surface. Note that oil viscosity varies as oil weathers. The simulation results without dispersant application indicate that the viscosity of Alaskan North Slope crude oil floating on the water surface (Case 1) changed from its original (un-weathered) value of about 20 cP to about 200,000 cP (and still increasing) 24 hours after the spill. For the IFO 300 (Case 7), the floating oil viscosity changed from over

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects FIGURE E-10 Predicted oil distributions at 24th hour after the release of Intermediate Fuel Oil 300 with (Case 9) and without a dispersant (Case 7). 10,000 cP initially to 40,000 cP over 24 hours. For diesel (Case 8), it changed only from 4 cP to 8 cP over 16 hours. These oil viscosity changes have a significant effect on how oil spreads and on how effective a dispersant would be in dispersing it. FIGURE E-11 Predicted oil distributions at 14th hour after the release of diesel oil with (Case 10) and without a dispersant (Case 8).

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects When it was assumed that the Alaska North Slope crude oil did not form an emulsion, the model predicted that 18 percent of the oil was naturally dispersed into the water column, compared to 3 percent when an emulsion was allowed to form. On the other hand, when mousse was assumed to be formed after 1 percent evaporation, only 1 percent of the oil dispersed into the water column. These results reveal the importance of oil type, oil properties, and emulsification on oil dispersion. Cases 2, 9, and 10 with dispersant use correspond to Cases 1, 7, and 8 without dispersant use (Cases 1 and 2 are with Alaska North Slope crude oil, Cases 7 and 9 are with IFO 300, and Cases 8 and 10 are with marine diesel). In spite of the very high viscosity, the models predict that 49 percent of the IFO 300 would be in the water column within 24 hours as a result of dispersant, due to the assigned 50 percent dispersant effectiveness as model input (see Figure E-10 with and without a dispersant application). By contrast, Figure E-11 shows that 78 percent of the diesel was dispersed into the water by a combination of natural and chemical dispersal. This amount is essentially the same without the use of the chemical dispersant (Case 8), implying that if this diesel is spilled, there is no need to use the dispersant under a 10-m/s wind. Again, because these results are based on using dispersant effectiveness as a model input, the fate of real oil may be somewhat different. Figures E-12 and E-13 present the predicted volume fraction of each distillation cut of floating IFO 300 and marine diesel, respectively, after the spill for Cases 9 and 10. As shown in Figure E-12, only the first cut and FIGURE E-12 Predicted composition of floating Intermediate Fuel Oil 300 with a dispersant at various times (Case 9).

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects FIGURE E-13 Predicted composition of floating diesel oil with a dispersant at various times (Case 10). some of the second cut (those distilled at 200° C [roughly 392° F] or lower) of the IFO 300 were evaporated within 6 hours; thus application of a dispersant after 6 hours would not entrain much of these components into the water column. For diesel, all or most of the first four cuts (those distilled at 180°C [roughly 356° F] or lower) were evaporated before the dispersant was applied (Figure E-13). For both refined oils, these cuts consist of only small portions (2 and 6 percent) of the refined fuels, and the majority of the oils are mostly composed of heavier-molecular-weight cuts (containing parent- and alkyl-substituted PAH). These results indicate that without models it is very difficult to integrate all interacting and sometimes competing transport and fate processes, oil types/properties, and dispersant use to predict how much oil will be found in specific areas during an actual oil spill. Thus, transport and fate models should be used to assist decisionmakers to take appropriate remedial actions during an actual oil spill. Effect of Oil Droplet Sizes A dispersant application is expected to result in entrained oil composed of many more small droplets, which rise more slowly, as discussed previously (see Figure E-2). To evaluate this effect, four oil droplet size distributions (see Figure E-1) were simulated. Case 12 has the baseline droplet sizes (diameters varying from 10 µm to 70 µm), whereas Case 11 has an 80 percent reduction in droplet size due to the dispersant applica-

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects tion. Because of uncertainty in oil droplet sizes, simulations with droplet sizes larger than base case (Case 13 with twofold increase in diameter, and Case 14 with 4-fold increase in diameter) were also evaluated. To isolate the effect of droplet-size distribution (uninfluenced by bathymetry, and uncertainty in vertical diffusivity), these four cases were run with oil spilled on the surface of 200-m deep water, further offshore along the Florida coast, and constant diffusivities were assigned over a mixing depth of 200 m. Figure E-14 presents predicted oil migration on the water surface and in the top 1-m water column 48 hours after the oil spill for Case 11. This figure also indicates the oil spill location by “+.” Unlike the shallow water applications (Cases 1 through 10), oil in this case traveled through deep water, ranging in depth from 200 m to over 350 m. Predicted time-varying average and maximum oil concentrations in the top 1-m water column are shown in Figure E-15, indicating the increase of oil concentrations during the first 6 and 12 hours after dispersant application. At 24 hours, the average and maximum concentrations were 0.7 and 3.4 mg/L, respectively, while at 48 hours, they were reduced to 0.4 and 1.7 mg/L. These concentrations are much lower than those appearing in the nearshore case (Case 2) due to unrestricted vertical diffusion in the offshore case, and because the same diffusion coefficients were used over the entire 200-m mixing depth as were used in the shallow water. This latter assumption was used for simplicity and comparative purposes; in reality diffusion in the deeper water is expected to be less than that in the shallow water. As indicated in Figure E-2, the rise velocity of oil droplets ranges from about 2.5 × 10−7 m/s for a diameter of 2 µm to 4.3 × 10−3 m/s for a diameter of 260 µm. Droplets moving at 2.5 × 10−7 m/s will rise only 0.001 m and 0.02 m, over periods of 1 hour and 24 hours, while over the same periods, droplets rising at 4.3 × 10−3 m/s will rise 15 m and 370 m. Meanwhile, a vertical diffusivity of 51 cm2/s will spread oil droplets (both upward and downward) about 6 m and 30 m over the same periods. Thus, the smallest oil droplets behave as if they are neutrally buoyant—transported only by diffusion—while the largest droplets are advected mainly by their buoyancy. Predicted vertical distributions of oil at 48 hours are shown in Figure E-16 for four size distribution cases (Cases 11 through 14). Each vertical distribution is plotted at the horizontal location where the oil concentration within the top 1-m layer is the highest over the area contaminated by oil after 48 hours. Although four cases are plotted in the same figure, they are at slightly different locations. As expected, when the oil droplet sizes increase, more oil is found near the water surface. Thus Case 14 (droplet diameters of 40 to 280 µm) has the highest concentration of 18 mg/L in the top 1-m water column, while Case 11 (diameters of 2 to 14 µm) has a

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects FIGURE E-14 Predicted oil movement at 48 hours after release of Alaskan North Slope crude at point + for Case 11 (80 percent reduction in oil droplet size due to dispersant application). corresponding concentration of 1.7 µg/L. Currently the effect of chemical dispersant application on oil droplet sizes is a model input, but the ability to predict droplet size should be developed and incorporated into oil transport and fate codes. These results again indicate that it is very difficult to integrate all of the interacting and sometimes competing transport and fate processes and

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects FIGURE E-15 Predicted time-varying oil concentrations (average and maximum concentration following plume versus time) in top 1-m water column for Case 11. Note: There was an 80 percent reduction in oil droplet size due to dispersant application. oil types/properties to predict when and how much oil will move to specific areas with and without dispersant application during an actual oil spill event. Thus, transport and fate models should be used to assist decisionmakers in choosing appropriate remedial actions during an oil spill by providing quantitative estimates of oil distributions that change with time and space. This is especially important in nearshore areas, which might experience the greatest environmental sensitivity. Yet these same areas are likely to have the most complex flow fields. Limitations on computer speed and human resources will likely limit, for some time, the accuracy of numerical models to simulate advection and diffusion in near real time, especially considering that spill locations are unpredictable, and multiple “what if” scenarios runs must be run. A consensus regarding “how good is good enough” needs to be developed among decisionmakers and model developers, and used to guide the future development of models and to optimize their use in real time.

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects FIGURE E-16 Predicted vertical distribution of oil concentrations at the horizontal location where the oil concentration within the top 1-m layer is the highest over the area contaminated by oil after 48 hours for Cases 11 through 14. In the meantime, efforts should be made to improve and validate models. This effort should include undertaking research at laboratory and mesoscales to define parameters that control oil dispersibility. The improved models should be used to assist on-scene decisionmakers to determine whether to use dispersants during an actual spill, and feedback should be sought from these decisionmakers as to the utility of the models in this regard. The ADIOS2 and 3-D GNOME codes, and possibly other codes, may support emergency response to provide “rough-cut” predictions within hours of oil spills.

OCR for page 355
Oil Spill Dispersants: Efficacy and Effects This page intentionally left blank.