Research Priorities to Support Dispersant Use Decisionmaking
The primary response methods for oil spills in the United States consist of the deployment of mechanical on-water containment and recovery systems, such as booms and skimmers. Under the Oil Pollution Act of 1990 (OPA 90), the U.S. Coast Guard (USCG) passed rules for vessel and facility response plans that specified the minimum equipment capabilities for oil containment and recovery for likely and maximum spill volumes. Mechanical recovery is not always effective, thus OPA 90 also called for national and regional response teams to develop guidelines for other on-water response strategies, specifically the use of chemical dispersants and in-situ burning. Regional Response Teams have designated areas and conditions where dispersants and in-situ burning may be considered as appropriate response strategies when mechanical recovery is determined to be insufficient to protect sensitive resources. There are three types of approvals for dispersant use: case-by-case approval, expedited approval, and pre-approval.
In early 2002, the U.S. Coast Guard proposed changes to the rules for oil spill response capabilities to include minimum capabilities for dispersant application in all zones where dispersant use has been pre-approved. Thus, availability of dispersant application assets will no longer be a limiting factor in the decision-making process. Instead, other factors, such as effectiveness and effects, will be the major drivers of the decision whether or not dispersants should be used. Furthermore, the ready availability of dispersants and their application systems is expected to result in an increase in situations where dispersant use may be considered to combat oil spills, even in nearshore settings. Timeliness of the decision-making
process is even more critical for nearshore spills where oil can quickly threaten highly sensitive resources.
Oil spills occur under a very wide range of conditions; thus decisions about appropriate response strategies have to be made for each event. Even for spills that meet pre-approval guidelines for dispersant use, the Federal On-Scene Coordinator (FOSC) has to make a series of decisions to determine whether or not dispersants would be effective and appropriate. Where dispersant use is not pre-approved, many more decisions must be made and the decision-making process becomes that much more difficult. One of the primary objectives of this study was to identify gaps in the information needed to support decisionmaking regarding appropriate use of dispersants. The decision-making framework used most frequently in the United States, and shown in Figure 2-4 in Chapter 2, is used to organize and prioritize the recommendations made in Chapters 3, 4, and 5. Many of the recommendations are relevant to more than one question in the decision-making process, but each recommendation is matched to the question it most strongly supports.
In response to the statement of task, the committee reviewed and evaluated existing information and ongoing research regarding the efficacy and effects of dispersants as an oil spill response technique. The statement of task specifically directed the committee to address “how laboratory and mesoscale experiments could inform potential controlled field trials and what experimental methods are most appropriate for such tests.” All experiments, whether conducted at the bench-top or field scale, represent an attempt to measure or otherwise quantify the contribution of one variable among many that interact to dictate a specific outcome. Depending on the scale of the experiment, many variables may need to be held constant so that the change in outcome can be mapped against variation in a single variable of interest. Unfortunately, this simplification, if not taken under consideration, can reduce the realism of tests to the point that the results are of limited application to real world situations. For example, when bench-top tests are conducted, temperature, salinity, and other natural variables are set and held constant to compare the effectiveness of different dispersants and oil combinations. At the other end of the spectrum, when field tests are carried out, researchers have very limited control over environmental conditions, thus each test result is specific to a narrow set of environmental conditions. If the test is conducted under conditions typical of most spills, some extrapolation is possible; however, spills occur over such a wide range of settings that a relatively large number of tests at a variety of geographic locations would be necessary to identify the range of settings in which the dispersants would be effective. Even if this were possible, many of the most important variables (e.g., concentration of dissolved phase constituents or dispersed oil droplets)
are extremely difficult to measure under field conditions. For example, one of the most challenging aspects of tank tests involves determination of a mass balance. Trying to obtain a mass balance in the field would be, at best, very difficult. Laboratory and wave-tank experiments and monitoring of spills of opportunity have the potential to answer key questions, if properly designed and conducted. Laboratory and mesocosm studies are most appropriate to test specific hypotheses on specific mechanisms or processes. Field tests are expensive, are difficult to control, offer many challenges to achieving good scientific measurements, and represent only a single set of conditions which are not readily controlled.
Serious consideration should be given to determining the true value and potential contribution of field testing. The body of work completed to date has provided important, but still limited, understanding of many aspects of the efficacy of dispersants in the field and behavior and toxicity of dispersed oil. Developing a robust understanding of these key processes and mechanisms to support decisionmaking in nearshore environments will require taking dispersant research to the next level. Many factors will need to be systematically varied in settings where accurate measurements can be taken. It is difficult to envision the appropriate role of field testing in a research area that has yet to reach consensus on standard protocols for mesocosm testing. The greater complexities (and costs) of carrying out meaningful field experiments suggest that greater effort be placed, at least initially, on designing and implementing a thorough and well-coordinated bench-scale and mesocosm research program. Such work should lead to more robust information about many aspects of dispersed oil behavior and effects. When coupled with information obtained through more vigorous monitoring of actual spills (regardless of whether dispersants are used effectively in response), this experimental work should provide far greater understanding than is currently available. Upon completion of the work discussed below, the value of further field-scale experiments may become obvious. In any case, such field-scale work would certainly be better and more effectively designed and executed than is currently possible. Future field-scale work, if deemed necessary, should be based on the systematic, coordinated bench-scale and wave-tank testing discussed throughout this report and summarized in the remainder of this chapter.
The committee also realizes that funding for oil spill research is very limited and this situation is not expected to change significantly in the future. Therefore, it is important to prioritize the research recommendations according to how they will improve the ability to make informed and science-based decisions on whether dispersants are the appropriate response tool for a specific spill. As stated previously, the committee’s efforts have focused on questions arising from the potential use of dispersants in shallow, nearshore settings. Each of the questions in Figure 2-4 is
listed below, followed by the research priorities needed to support the decisions.
D.1 WILL MECHANICAL RESPONSE BE SUFFICIENT?
Given that the use of dispersants does not remove oil from the environment, dispersant use is only considered when the answer to this question is “no,” which then leads decisionmakers to evaluate the appropriateness of dispersant use. Although a subset of factors that may limit the effectiveness of mechanical response are readily evaluated (e.g., spill occurs too far from shore to allow safe and effective mechanical spill response), other factors require an ability to forecast environmental conditions at the spill site or along the projected path of the surface slick. Thus efforts to support real-time, tactical decisions regarding mechanical response will indirectly and directly support real-time decisionmaking regarding dispersant use.
D.2 IS THE SPILLED OIL OR REFINED PRODUCT KNOWN TO BE DISPERSIBLE?
As discussed in Chapter 3, there are two parts to this question of chemical effectiveness: (1) Is the freshly spilled oil dispersible? and (2) If the oil is initially dispersible, how long before it becomes non-dispersible? Currently, responders use “rules of thumb,” experimental results for specific oils, and past experience to determine if an oil will be dispersible. Generally, it is possible for experienced and knowledgeable responders to predict whether a specific oil is initially dispersible. However, the more difficult determination is the length of time remaining until the oil weathers to the point that it is no longer dispersible. The state of the practice is to use visual observations to estimate the degree of emulsification or to conduct a first application of dispersant and see if the surface slick disperses. Such observations provide a qualitative approach to whether an oil has dispersed; however, they will not begin to answer the quantitative question as to the effectiveness of dispersant application. There is significant confusion among decisionmakers on how to interpret existing data on dispersant effectiveness. Results from experiments designed to answer one type of question are inappropriately used to answer a different question or to predict behavior under very different conditions (e.g., laboratory results are inappropriately used to estimate field effectiveness). The key areas of research include studies that will allow better predictions, using simple models, of the weathering processes that limit dispersant effectiveness for different oil types and environmental conditions.
NOAA, the Environmental Protection Agency (EPA), the Department of the Interior (including MMS and USGS), USCG, relevant state agencies, industry, and appropriate international partners should work together to develop and fund a research program to identify the mechanisms and rates of weathering processes that control the chemical effectiveness of dispersants. The research program should include both bench-scale tests and wave-tank experiments. Because of the limited funds and high costs of wave-tank experiments, it is essential that wavetank studies be well coordinated. Agencies and industry should work together to establish an integrated research plan that focuses on collecting information about key aspects of dispersant use in a scientifically robust, but environmentally meaningful context. This new work will require systematic analysis using rigorous experimental design and execution, making use of standard chemical and other measurement techniques carried out by trained, certified personnel. Specific recommendations for these experiments are listed below.
Research should be conducted in laboratory and wave-tank systems to investigate those parameters that control oil dispersability, including oil rheology and chemistry, dispersant rheology and dispersant chemistry, and dispersant-oil ratio. Past research on these topics often did not include measurement of important system characteristics (e.g., energy input to experimental systems) and response variables (e.g., oil droplet-size distributions), and the results are often contradictory. Future studies should conform to accepted standards of experimental design (discussed in Chapters 3, 4, and 5) that support statistical analysis of the data.
Experimental bench-scale tests should be used to characterize the energy dissipation rates that prevail over a range of operating conditions to determine the functional relationship between variables for a range of oil viscosities and weathering states. Furthermore, evaluation of chemical effectiveness should always include measurement of the droplet-size distribution of the dispersed oil.
The relationships between energy dissipation rates and chemical effectiveness should be determined for a variety of oil-dispersant combinations, including a range of oil viscosities and weathering states. Oil dispersant chemical efficiency tests should be designed to collect data that can be used in fate and transport modeling.
Wave-tank studies should be designed to specifically address the chemical treatment of weathered emulsions of water-in-oil. Oil mass balances should be reported. In addition, the droplet-size distribution of the dispersed oil should be measured and reported.
D.3 ARE SUFFICIENT CHEMICAL RESPONSE ASSETS (I.E., DISPERSANT, EQUIPMENT, AND TRAINED PERSONNEL) AVAILABLE TO TREAT THE SPILL?
Under the proposed U.S. Coast Guard rulemaking, assets to treat spills in U.S. waters with chemical dispersant will be available within 12 hours after the spill in areas that have pre-approval plans. This increased availability of chemical response assets will likely result in more frequent consideration of dispersants as a response option for all spills, including those closer to shore and in shallow waters. If dispersant application becomes a required capability, it will be necessary to implement methods and procedures to ensure the readiness of response equipment and supplies for dispersant use, similar to the requirements for mechanical response equipment.
D.4 ARE THE ENVIRONMENTAL CONDITIONS CONDUCIVE TO THE SUCCESSFUL APPLICATION OF DISPERSANT AND ITS EFFECTIVENESS?
This question addresses environmental and operational effectiveness. Currently, it is not possible to predict the overall field effectiveness of dispersants for a spill event, a critical aspect of the trade-off analysis. Resource trustees need to be able to evaluate the benefits of reduced loadings of oil on shoreline habitats and smaller slicks that threaten water-surface resources compared to increased risks from dispersed oil plumes on water-column and benthic resources. Both potential risks and potential benefits depend on the effectiveness of dispersant application, particularly in nearshore settings. Better information is needed to determine the window of opportunity and percent effectiveness of dispersant application for different oil types and environmental conditions. Currently, dispersant effectiveness is a user input to fate and transport models, but potential effectiveness should be estimated by a physical-chemical efficiency model that integrates all of the complex processes controlling oil weathering and oil entrainment into the water column. Furthermore, there is no standard definition of field effectiveness and how it should be reported.
Relevant state and federal agencies, industry, and appropriate international partners should develop and fund a research program that provides the data necessary to predict, through modeling of the chemical, environmental, and operational conditions, the overall effectiveness of a dispersant application, specifically including conditions representative of nearshore physical settings. Two general types of modeling efforts and products should be recognized: (1) output intended to support
decisionmaking during preplanning efforts; and (2) output intended to support emergency response to provide “rough-cut” outputs in hours. Detailed and specific recommendations are discussed at length in Chapters 3 and 4. The research program should consider the following issues:
Energy-dissipation rates should be determined for wave tanks over the range of operating conditions that will be used in dispersant effectiveness tests. The wave conditions used in dispersant effectiveness tests should represent a specific environment of interest. It may be necessary to conduct experiments over a range of energy dissipation rates to adequately represent the spill environment.
The design of wave-tank dispersant-effectiveness studies should test specific hypotheses regarding factors that may influence operational effectiveness. These factors include oil properties expected to prevail under spill-response conditions such as water-in-oil emulsification and the potential for heterogeneity in the rheological properties of the floating oil (e.g., formation of a “skin” that resists dispersant penetration).
Tanks test studies should be conducted to determine the ability of mechanical recovery methods to retrieve oil that has been ineffectively treated with dispersant and re-floated oil. A more complete understanding of how dispersant use may subsequently limit mechanical recovery, if the dispersant is ineffective, could greatly reduce concern about the reliance on operational testing of dispersant effectiveness during early phases of spill response.
Experiments should be designed to provide data on the rate and consequences of surfactant wash-out for both dispersed oil droplets that re-coalesce and surface slicks that were treated under calm conditions. Coalescence and resurfacing of dispersed oil droplets should be studied in flumes or wave tanks with high water-to-oil ratios (to promote leaching of surfactant into the water column) as a function of mixing time and energy dissipation rates.
Evaluation of dispersant effectiveness in wave-tank tests should include measurement of oil concentrations on the water surface, in the water column, lost to the atmosphere, and on wave-tank surfaces. Oil mass balances should be reported. In addition, the droplet-size distribution of the dispersed oil should be measured and reported.
Currently, protocols for monitoring effectiveness of dispersants (e.g., Specialized Monitoring of Advanced Response Technologies [SMART]) are for guidance only and reflect the concern that use of dispersants not be unnecessarily postponed until monitoring assets can be put in place. Often, monitoring resources cannot be mobilized within the timeframe of emergency dispersant applications. SMART protocols have not been up-
dated since the first design. However, it is very important to document actual oil concentrations under dispersed oil slicks, to validate dispersed oil model predictions and document actual exposures to sensitive resources. Monitoring would also support evaluation of the effectiveness of dispersant applications. NOAA and USCG should develop updated SMART protocols and consider adding a detailed standard operating procedure (including instrument calibrations and data quality objectives) for each sampling and analytical module.
D.5 WILL THE EFFECTIVE USE OF DISPERSANTS REDUCE THE IMPACTS OF THE SPILL TO SHORELINE AND WATER SURFACE RESOURCES WITHOUT SIGNIFICANTLY INCREASING IMPACTS TO WATER-COLUMN AND BENTHIC RESOURCES?
This trade-off analysis is the most difficult step of the process because of the lack of quantitative tools to predict the fate and effects of the dispersed oil plume and the benefits associated with less surface oil. It is also one of the most critical questions needing answer for adequate and appropriate decisionmaking. As discussed in Chapter 2, current ecological risk assessment (ERA) workshops on dispersant decisionmaking use a very qualitative approach that is difficult to apply to nearshore conditions where the potential impacts are not easily characterized. For instance, in offshore settings, one might reasonably assume that there is only a small likelihood that organisms on the seafloor may be exposed to significant concentrations of dispersed oil. Such assumptions may not be reasonable in some nearshore settings. Resource trustees need better information on the likely exposure regime, the mechanisms of toxicity of dispersed oil, and appropriate endpoints.
Oil trajectory models for dispersed oil plumes could be valuable tools to predict exposure, but they are incomplete in terms of their representation of the natural physical process involved, verification of the codes, and validation of the output from these models in an experimental setting or during an actual spill. As discussed in Chapter 4, the ability of models to predict the concentrations of dispersed oil and dissolved aromatic hydrocarbons in the water column with sufficient accuracy to aid in spill decisionmaking has yet to be fully determined.
As discussed in Chapter 4, one of the most significant weaknesses in correlating laboratory-scale and mesoscale experiments with conditions in the field results from a lack of understanding of the turbulence regime in all three systems. Likewise, one of the biggest uncertainties in computer modeling of oil spill behavior (with and without dispersant addition) comes from specifying horizontal and vertical diffusivities. It is very difficult to integrate all interacting transport and fate processes and oil
properties to predict how much oil will be found in specific areas during an actual oil spill without the use of models.
Oil trajectory and fate models used to predict the behavior of dispersed oil should be improved, verified, and then validated in an appropriately designed experimental setting or during an actual spill. These models should meet the needs of both planning and real-time decisionmaking in complex nearshore settings. Key elements of the model improvements include:
Studies should be conducted to quantify horizontal and vertical diffusivities and the rate of energy dissipation in the field (shear in vertical dimension, variations in the vertical diffusivity as a function of depth, sea-surface turbulence, etc.) under a variety of sea states that can be used as inputs into models (to improve the physical components of dispersed oil behavior), as well as to better design laboratory and mesocosm systems that may be suitable for estimating dispersant performance at sea.
Models should include advective transport of entrained oil droplets, and the model codes and results should be validated in flume/tank studies and open-ocean (spill-of-opportunity) tests.
The results of studies to better understand the processes that lead to formation of water-in-oil emulsions should be reflected in the improved models.
Model output should include the concentrations of dissolved and dispersed oil, expressed as specific components or pseudo-components that can be used to support toxicity analysis.
Once the improved models are available, sensitivity analyses should be conducted based on three-dimensional, oil-component, transport, and fate models, and the necessary databases developed (evaporation, dissolution, degradation, etc.) so that oil concentrations and fate can be used in decisionmaking.
One of the major concerns with use of dispersants in nearshore settings is the dispersed oil interaction with suspended particulate matter and the ultimate fate of the droplets.
Relevant state and federal agencies, industry, and appropriate international partners should develop and fund a focused series of experiments to quantify the weathering rates and final fate of chemically dispersed oil droplets compared to undispersed oil. Results from these experiments could be integrated with results from biological exposures comparing uptake of dissolved and particulate oil to provide a comprehensive model of the fate of dispersed oil in aquatic systems.
factors, including seasonal variation in the presence, absence, and number of organisms at the spill site or along a projected path of a surface slick or dispersed oil plume; the sensitivity of the species to various toxic components in crude oil or refined products; and some estimation of the time needed for a given population to recover from acute exposure.
Relevant state and federal agencies, industry, and appropriate international partners should conduct a series of transport and fate modeling and associated biological assessments with and without dispersants, and develop operational envelopes of the dispersant use (e.g., for what oil types and volumes; when, where, and what type of water bodies) for planning prior to actual oil spills. Dissemination of these modeling efforts also provides scientific knowledge and intuition to make rational decisions for dispersant use.
Models are envisioned as a key tool to support decisionmaking on the appropriateness of dispersant applications. These models can be improved with the results of laboratory and wave-tank studies, but they need to be validated by comparing model results with actual field data. Relevant state and federal agencies and industry should develop and implement detailed plans (including preposition of sufficient equipment and human resources) for rapid deployment of a well-designed monitoring effort for actual dispersant applications in the United States. The plans should include measurement of total petroleum hydrocarbon (TPH) and polynuclear aromatic hydrocarbon (PAH) concentrations in both dissolved phase and particulate/oil-droplet phase for comparison to TPH and PAH concentrations predicted by computer models for oil spill fate and behavior. The spill-of-opportunity monitoring planning for dispersant application in the Gulf of Mexico (Aurand et al., 2004) should be finalized and implemented at the appropriate time.
Relevant state and federal agencies, industry, and appropriate international partners should develop and implement a series of focused toxicity studies to: (1) provide data that can be used to parameterize models to predict photo-enhanced toxicity; (2) estimate the relative contribution of dissolved and particulate oil phases to toxicity with representative species; and (3) include an evaluation of delayed effects. Detailed chemical analyses should accompany these tests, including characterization of dissolved and particulate oil composition and concentrations, as well as bioaccumulation. Increased understanding of these variables, and effective incorporation of them into decision-making tools, such as fate and effects models and risk rankings, will enhance the ability of decisionmakers to estimate the impacts of dispersants on aquatic and benthic organisms. To this end, every effort should be taken to ensure that the spill response research community continues to monitor developments in the broad field of ecotoxicology, as various applications of
increased understanding of toxicological effects, on various time scales, at the population and community-level may be of significant value to dispersant decisionmaking (see Chapter 5 for more details).
In addition, consideration should be given to long-term monitoring of sensitive habitats and species (e.g., mangroves, corals, sea grasses) after dispersant application to assess chronic effects and long-term recovery. These data will be valuable in validating the assumptions associated with environmental trade-offs of using dispersants.
The 1989 report recommended that studies be undertaken “to assess the ability of fur and feathers to maintain the water-repellency critical for thermal insulation under dispersed oil exposure conditions comparable to those expected in the field.” This committee re-affirms this recommendation because of the importance of this assumption in evaluating the environmental trade-offs associated with the use of oil dispersants in nearshore and estuarine systems and because it has not been adequately addressed.