The URL can be used to link to this page

Home My WebLink AboutMethods for Conducting Post-Spill Environmental Studies 1989GENERAL TECHNICAL APPROACH METHODS FOR CONDUCTING POST-SPILL ENVIRONMENTAL STUDIES, HEALTH RISK ASSESSMENT, AND NATURAL RESOURCE DAMAGE ASSESSMENT APRIL 1989 SUBMITTED TO: EXXON CORPORATION VALDEZ, ALASKA General Technical Approach METHODS FOR CONDUCTING POST-SPILL ENVIRONMENTAL STUDIES, HEALTH RISK ASSESSMENT, AND NATURAL RESOURCE DAMAGE ASSESSMENT by Chemtrack Environmental Services, Inc. Tetra Tech, Inc. Research Triangle Institute for EXXON Corporation Valdez, Alaska April 1989 Tetra Tech, Inc. 11820 Northup Way, Suite 100 Bellevue, Washington 98005 LIST LIST CONTENTS OF FIGURES OF TABLES INTRODUCTION FIELD INVESTIGATIONS RISK INTRODUCTION DEVELOPMENT OF PROGRAMMATIC OBJECTIVES REVIEW AVAILABLE DATA DEVELOP FIELD SURVEY OBJECTIVES DEVELOP SAMPLING AND ANALYSIS PLAN DEVELOP QUALITY ASSURANCE AND SAFETY PLANS CONDUCT FIELD INVESTIGATION ANALYZE SAMPLES ANALYZE AND INTERPRET DATA DISSEMINATE RESULTS ASSESSMENT OF CHEMICALLY CONTAMINATED SEAFOOD INTRODUCTION HAZARD IDENTIFICATION DOSE-RESPONSE ASSESSMENT EXPOSURE ASSESSMENT RISK CHARACTERIZATION UNCERTAINTY ANALYSIS ii uo Oo oO NN ODO OT WW mo BO DY SF KF Fe KF Fe PO 0 Oo DOD WwW WwW DP CO NATURAL RESOURCE DAMAGE ASSESSMENT TASK A. TASK B. TASK C. REFERENCES EVALUATE IMPACTS OF OIL ASSESS DAMAGES TO HABITATS AND RESOURCES OF CONCERN DEVELOP RESTORATION PLANS iii 27 27 35 51 54 Number FIGURES Page Tetra Tech's approach to developing and conducting field investigations for Exxon 4 Overview of the risk assessment process 14 Hypothetical dose-response relationships for a carcinogen and a noncarcinogen 19 Tetra Tech's approach to assessing damages to natural resources 28 Framework for classifying measurement methods 43 iv INTRODUCTION Public concern regarding environmental degredation resulting from the EXXON VALDEZ 011 spill will require extensive multidisciplinary environmental studies over both the short-term and the long-term. These studies will educate the public on the degree of environmental degredation, and will also provide Exxon with the data needed to determine its extent of liability and for protection against unreasonable liability claims. The data that are generated on behalf of Exxon must therefore be of the highest technical quality and must be responsive to concerns expressed by the public, state and federal agencies, and Exxon. Chemtrack Environmental Services, Tetra Tech, and Research Triangle Institute have teamed to provide scientists and logistical support personnel capable of providing Exxon with high quality environmental services to determine the extent of environmental degredation attributable to the EXXON VALDEZ oil spill. The general technical approach to carrying out the scientific investigations is provided in this report. Further details will be supplied upon request. Field investigations will provide the foundation for assessing environmental degredation. These investigations will be multidisciplinary, including aspects of water quality, sediment quality, fisheries resources, benthic community structure, kelp bed assessments, and seabird and marine mammal studies. Data provided by field investigations will be used in human health risks assessments to determine potential risks due to direct contact with the spilled oi] and due to ingestion of seafood harvested in the vicinity of the spilled oi]. Lastly, data will be used in a Natural Resource Damage Assessment (NRDA) to determine the degree of injury to natural resources, and the economic value of those lost resources. From the NRDA it may be possible to identify damaged resources for which restoration may be possible. Alternative techniques and associated costs of such 1 restoration will provide Exxon with alternatives for actively mitigating damaged resources. FIELD INVESTIGATIONS INTRODUCTION Field investigations to determine the spatial extent and severity of contamination will be an essential component of Exxon's cleanup program. The data generated by these investigations will assist in the determination of liability for damaged resources for which Exxon will be responsible, and will provide essential information for determining potential risks associated with the consumption of seafood harvested in the vicinity of the oil spill. Investigations conducted in the immediate future will also provide a point of reference for documenting the recovery of habitats and resources over time, and will provide information for the design of long-term monitoring programs. Specific objectives of these investigations May include the following: 2 Determination of the spatial distribution and sediment and water concentrations of spilled oil . Evaluation of impacts to resources (e.g., water quality, zooplankton, intertidal and subtidal invertebrate species, kelp beds, fish populations, seabirds, mammals) . Determination of the bioaccumulation potential of petroleum hydrocarbons 2 Evaluation of the degree of degradation of coastal resources caused by the spilled oil. Tetra Tech will follow a sequential process for the design and implementation of field surveys for Exxon (Figure 1). It is highly recommended that Exxon INPUT FROM EXXON AND AGENCIES INPUT FROM EXXON AND AGENCIES INPUT FROM EXXON AND AGENCIES INPUT FROM EXXON AND AGENCIES DEVELOP PROGRAMMATIC OBJECTIVES REVIEW AVAILABLE DATA + Site Characteristics + Nature and Extent of Contamination * Contaminant Sources DEVELOP FIELD SURVEY OBJECTIVES * Objectives * Testable Hypotheses DEVELOP SAMPLING AND ANALYSIS PLAN + Station Positions + Sampling Frequency + Variables + Statistical Tests DEVELOP QUALITY ASSURANCE AND SAFETY PLANS + Sample Collection + Sample Analysis CONDUCT FIELD INVESTIGATION + Water Quality + Sediment Quality + Biological Resources ANALYZE SAMPLES + Water + Sediment + Biota DATA ANALYSIS/INTERPRETATION * Regression Analyses + ANOVA + Multivariate Analyses DISSEMINATION OF RESULTS + Field Report + Draft and Final Report + Professional Publications Staff INPUT TO RISK ASSESSMENT AND NATURAL RESOURCE DAMAGE ASSESSMENT Tetra Tech's approach to developing and conducting field investigations for Exxon. include representatives from state and federal agencies in several steps within this process. The following steps are included in the process: . Identify programmatic objectives 7 Review available environmental data to characterize the study area . Develop field survey objectives and select sampling variables . Develop sampling, quality assurance, and safety plans a Conduct field investigation 2 Analyze samples 2 Evaluate data and interpret results 7 Disseminate results to Exxon in clearly written reports It is essential that each task is addressed to ensure that the program meets Exxon's objectives and provides legally defensible data. Each step is discussed in greater detail below. DEVELOP PROGRAMMATIC OBJECTIVES Programmatic objectives will be developed in consultation with Exxon, and state and federal agencies as appropriate. These objectives will encompass Exxon's overall goals for conducting the field study, and will provide the initial direction for conducting the review of available data and developing the specific field survey objectives. Development of programmatic objectives will also assist in defining the variables (e.g., sediment and water petroleum concentrations; toxicity; petroleum hydrocarbon concentrations in fish, invertebrates, or marine mammal tissue) that will be sampled and evaluated. REVIEW AVAILABLE DATA Before conducting extensive field investigations, a preliminary review of available data is recommended to facilitate the design and implementation of an appropriate, cost-effective sampling program. Data generated as part of the environmental studies associated with the Valdez terminal are easily available. A data review would include: . General site characteristics (e.g., ecological setting, location of kelp beds and other areas of concern, spawning grounds, water currents, benthic communities, sediment composition) 7 Beneficial uses of the aquatic system (e.g., commercial fisheries, recreation) . Nature and extent of current and historical contamination problems or biological effects . Potential sources of contaminants, other than petroleum hydrocarbons released by the oil spill . Data gaps. Based on the information listed above, recommendations can be made for the collection of additional data to meet the programmatic objectives. Comprehensive data summaries and problem identification studies completed by Tetra Tech for numerous aquatic systems, including those performed under the Puget Sound Estuary Program in Elliott Bay, Everett Harbor, Budd Inlet, and Sinclair/Dyes Inlets, have preceded field investigations and action plans in many of these areas. DEVELOP FIELD SURVEY OBJECTIVES Prior to preparing the sampling plan, specific objectives will be developed to direct the formulation of testable hypotheses. These objectives will account for the information gained during the review of existing data and the compilation of supplementary information. The development of these objectives will include selection of variables required to meet the programmatic assumptions. Objectives and variables will be stated in specific terms and might include the following: . To determine whether sediment concentrations of polycyclic aromatic hydrocarbons have increased as a result of the oil spill . To determine the relationships among petroleum hydrocarbon concentrations and potential impacts to species of concern to NOAA and other natural resource trustee agencies. Field survey objectives will be met through the use of testable null hypotheses (i.e., statements that water and sediment quality conditions are not impacted by the factor of concern, and statements that can be verified or refuted on the basis of field survey results). Hypotheses will be as specific as possible to permit a statistical evaluation of the data. DEVELOP SAMPLING AND ANALYSIS PLAN The actual design of the field survey will be documented in the sampling and analysis plan. This plan will provide the rationale for selection of sampling locations, frequency, and variables. Information gained in the review of available information and the development of field survey objectives will provide the basis for developing the sampling plan. The survey will be designed to collect those data that are currently unavailable but that are necessary to satisfy the programmatic objectives. 7 Unless otherwise requested, the survey will be designed to facilitate statistical testing of the program hypotheses. Proposed data analyses will also be documented in the sampling and analysis plan. It is essential that the statistical tests needed to test hypotheses be selected before conducting the field survey because the assumptions that underlie those statistical tests may influence station locations and the number of replicates collected. By selecting data analysis procedures during design of the field program, Tetra Tech ensures that the data collected will be limited to what is required to meet the programmatic objectives. DEVELOP QUALITY ASSURANCE AND SAFETY PLANS Tetra Tech routinely prepares quality assurance and safety plans to ensure that all samples are collected and processed correctly, and to ensure that all necessary safety precautions are observed to during sample collection and sample processing. Typically, each field program has a designated quality assurance officer and a safety officer who are otherwise not associated with the program. These individuals are responsible for implementing their respective plans and ensuring compliance by all field personnel. Quality assurance and safety protocols will be strictly observed during all field investigations. Quality assurance measures include station positioning, decontamination of sampling equipment surfaces, checklists to ensure that all types of samples and numbers of replicates are collected, proper sample storage facilities, proper shipping methods, and full chain-of- custody sample tracking procedures. Safety measures will include use of hard hats on vessels, a minimum of two people to handle grab samplers and coring devices, proper chemical handling procedures, availability of respiratory protection, and maximum 10-hour field days. CONDUCT FIELD INVESTIGATION Field investigations may involve both intertidal sampling and remote sampling from a vessel. The type of instrumentation required for sample collection will depend on the selected variables. Water samples will be collected using cylindrical containers such as Niskin or Van Dorn bottles. Zooplankton will be collected using plankton nets. Sediment samples will be collected using grab samplers or coring devices. Organisms for studies of bioaccumulation may be collected from intertidal areas, from subtidal habitats using a trawl or grab, or from unimpacted areas and then caged and moored in potentially contaminated areas. The degree of station positioning accuracy will be determined in consultation with Exxon. At some stations, relocation accuracy to within a few meters may be warranted. In others, much less stringent positioning accuracy may be required. Tetra Tech (1987, 1988) has developed two guidance manuals for the U.S. EPA that discuss station positioning accuracy that may be required to meet various program conditions, and that evaluate the ability of available positioning methods to meet those levels of accuracy. ANALYZE SAMPLES Biological and chemical samples will be analyzed according to protocols approved by &xxon. Biological analyses may include sediment toxicity bioassays; zooplankton, fish, and benthic community analyses; and seabird, marine mammal, and kelp bed censuses. Biological analyses will be conducted in accordance with protocols such as the Puget Sound Protocols (Tetra Tech 1986c). Chemical variables may include analyses of conventional variables (e.g., total organic carbon, sediment grain size, total sulfides), metals, organic compounds, and compounds specific to petroleum hydrocarbon degrada- tion. Chemical analyses will be subcontracted to a laboratory that can demonstrate satisfactory analysis of compounds on, but not limited to, U.S. EPA's Target Compound List (formerly known as Priority Pollutants List) in accordance with U.S. EPA's Priority Pollutant Performance Standards. 9 Analytical methods will include use of precision gas chromatographic/mass spectral (GC/MS) analysis. Tetra Tech routinely oversees the analysis of biological and chemical samples in accordance with the Puget Sound Protocols (Tetra Tech 1986c), or for some chemical samples (i.e., metals and organics) according to U.S. EPA's Contract Laboratory Program (CLP) procedures. Tetra Tech's experience in writing the Puget Sound Protocols and the Clean Water Act Section 301(h) sample collection and analysis protocols (Tetra Tech 1986b) will enable our team to quickly review laboratory methods, select appropriate laboratories to analyze samples, oversee the laboratories, and work with laboratories to resolve problems before they become uncorrectable. Quality assurance/quality control (QA/QC) requirements may differ among projects. For example, this project may involve enforcement actions or litigation, and may need more stringent QA/QC requirements than a project involving routine ambient monitoring. For the proposed study, QA/QC requirements will be identified in consultation with Exxon. ANALYZE AND INTERPRET DATA A variety of analytical techniques are available to evaluate environ- mental data. The quality of the data will influence the selection of an appropriate data analysis approach. Analytical tools include graphical presentations, statistical procedures, models, and water quality and biological indices. Simple data analysis techniques, such as graphical presentations, may be adequate to determine spatial or temporal distributions of oi] concentra- tions. Linear regression of an unreplicated multiyear data set may provide a limited evaluation of temporal trends in contaminant concentrations. Correlation analyses can also be used to determine which variables (e.g., petroleum hydrocarbon concentrations) may be determinants of biological impacts (e.g., changes in benthic community structure, kelp bed size, zooplankton biomass, fish densities). 10 Replicated data will permit the use of statistical analyses [e.g., t-tests, analysis of variance (ANOVA)] to determine the significance of differences between reference and impacted sites. ANOVA requires a minimum of three replicates at each station to estimate the mean value and associated variance. The key species and community approaches to determining the spatial extent and magnitude of impacts can be used to further elucidate possible relationships between aquatic biota and water quality variables. The objectives of the two approaches are to document changes in aquatic biota and to relate those changes to measured water quality variables. Abundances of individual species form the basis of the key species approach, while multivariate methods such as classification and ordination analyses are used in community approaches. A variety of numerical and physical modeling techniques can also be used as data evaluation tools. Tetra Tech has developed and applied water quality models to evaluate pollutant dispersion and advection in surface waters, to predict changes in water quality conditions due to proposed actions, to simulate the effects of acid precipitation on watersheds and lake water quality, and to perform waste load allocation analyses. Results of these modeling efforts provide additional insight into the relationships among the physical, chemical, and biological components of aquatic systems. Data analysis and interpretation will follow the sampling and analysis plan. The statistical tests will have already been selected and the appropriate data will have been collected. During this phase of the Program, preliminary results may redirect the remaining analyses. However, the data collected should be of sufficient quality to be used in other statistical tests that require the same general level of replication and station positioning as those tests outlined in the sampling and analysis plan. 11 DISSEMINATE RESULTS Results of field surveys will be delivered in clear, concisely written reports. Typically, a brief field report is prepared immediately following completion of the field phase of the program. This report documents station locations, types of samples collected at each station, date and time of sampling, general observations, and any problems encountered. At Exxon's discretion, this report may be issued as a stand-alone document or incor- porated into the study report as an appendix. The main body of the study report will include an executive summary; an introduction; a methods section that details field, laboratory, and data analysis methods; a results section; and discussion section. Raw data will be provided in separate appendices. Tetra Tech maintains a highly trained publications staff in all regional offices. These individuals provide editorial and graphical support. All deliverables, including cruise reports, draft reports, and final reports, undergo thorough technical and editorial review prior to release. 12 RISK ASSESSMENT OF CHEMICALLY CONTAMINATED SEAFOOD INTRODUCTION Risk assessment is a scientific procedure to determine the probability of adverse health effects from a specific exposure to a toxic agent. Risk assessment differs from risk management, although both are components of regulatory decision-making (National Research Council 1983). Risk assessment provides the scientific basis for public policy and action. In risk management, risks are interpreted in light of legislative, socioeconomic, technical, and political factors, and appropriate controls are determined. Direct measurement of human health risks is possible in certain limited circumstances. Such circumstances generally involve a single high exposure or repeated moderate exposures to a specific chemical, and a clear cause- effect relationship. For example, direct measurement of cancer risks might be possible in a population of workers exposed to an industrial chemical spill. In contrast, it is virtually impossible to directly measure the health risks of eating seafood harvested from an area such as Prince William Sound during recreational activities. Models that predict health risks are therefore needed. Risk assessment procedures described below focus on predicting health risks from long-term exposure to relatively low levels of contamination. The following sections provide an overview of the steps in risk assessment. These steps would be followed in a risk assessment of seafood harvested within the vicinity of the EXXON VALDEZ oi] spill. The general format for risk assessment and all definitions of terms used herein are consistent with those provided by National Research Council (1983) and U.S. EPA (1986a,b,c,d; 1988). As indicated in Figure 2, risk assessment is a multistep process comprised of the following tasks: 13 RISK ASSESSMENT PROCESS HAZARD IDENTIFICATION (Chemicals of Concern) TOXICITY EXPOSURE ASSESSMENT ASSESSMENT RISK CHARACTERIZATION UNCERTAINTY AND PERSPECTIVE Figure 2. Overview of the risk assessment process. 7 Hazard Assessment -- Selection of chemicals of concern and qualitative evaluation of their potential to cause adverse health effects (e.g., birth defects, cancer) in animals or in humans 7 Dose-Response Assessment -- Quantitative estimation of the relationship between the dose of a substance and the probability of an adverse health effect . Exposure Assessment -- Characterization of the populations exposed to the toxic chemicals of concern; the environmental transport and fate pathways; and the magnitude, frequency, and duration of exposure . Risk Assessment -- Estimation of risk for the health effect of concern, based on information from the dose-response and exposure assessments 7 Uncertainty Analysis -- Qualitative and where possible, quantitative description of the assumptions and limitations inherent in each step of the risk assessment. A summary of the role that each of these steps plays in the risk assessment process is provided below. Although these steps are discussed sequentially, the risk assessment process is highly iterative. Information developed in each step of the risk assessment is useful in subsequent steps and provides feedback to preceding steps. For example, dose-response information may be used in the hazard assessment step to select and characterize chemicals of concern. Toxicity profiles developed in the hazard assessment may affect selection of dose-response variables, and may provide meaningful qualitative information for use in the risk characteri- zation or uncertainty analysis. The following summary is, for the most part, abstracted from several recent U.S. EPA guidance manuals (Tetra Tech 1986a; PTI 1987) on risk 15 assessment of chemically contaminated seafood. Further elaboration on the uses of risk assessment and risk management in evaluation of chemically contaminated seafood is available in the U.S. EPA guidance manuals. HAZARD IDENTIFICATION The first step in the risk assessment process is to define toxicological hazards posed by the individual chemical contaminants in the seafood samples. These hazards are defined by identifying chemicals of concern, and by constructing a toxicity profile for each contaminant of concern. The contaminants of concern for a particular seafood risk assessment should be selected based on the following criteria: a High persistence in the aquatic environment . High bioaccumulation potential 7 High toxicity to humans (or suspected high toxicity to humans based on mammalian bioassays) : Known sources of contamination in areas of interest . High concentrations in previous samples of seafood from areas of interest. Further screening of chemicals is possible based on preliminary risk analysis. For example, some chemicals have relatively low toxicity. Only extremely high concentrations (e.g., >100 ppm) in seafood would cause concern, assuming a high seafood consumption rate for 70 years (for discussion of consumption rates, see Exposure Assessment section). Toxicity profiles consist of a summarization of qualitative and quantitative aspects of chemicals of concern that indicate the types of 16 health injury or disease that may occur under specific exposure conditions. The following kinds of information may be useful in developing toxicity profiles: . Physical-chemical properties (e.g., chemical speciation, vapor pressure, octanol-water partition coefficients) a Metabolic and pharmacokinetic properties (e.g., metabolic degradation products, depuration kinetics) 7 Toxic effects for specific routes of exposure (e.g., target organs, cytotoxicity, carcinogenicity, mutagenicity). Physical-chemical properties are important in determining the environmental behavior of chemicals of concern and in evaluating the relative importance of potential exposure pathways leading to human receptors. Metabolic and pharmacokinetic properties are important in characterizing the behavior of a chemical within the body, including its assimilation, transport, transfor- mation, accumulation, and elimination. For a given chemical of concern, toxic effects may be expressed in variety of ways depending on the magnitude, duration, route of exposure, and sensitivity of exposed individuals. Available toxicity information may be limited to studies of experimental animals or may include a range studies on both experimental animals and humans. The range of information developed for the toxicity profiles is then used to determine whether toxicity observed in experimental animals or humans in one exposure setting could also occur in humans under the specific exposure conditions of concern in the risk assessment. Toxicity profiles may also influence the nature and extent of subsequent steps in risk analysis. For example, the endpoint of concern in the dose-response assessment may be selected based on the most severe adverse effect identified in the toxicity profile. In the absence of quantitative data for other steps in the risk assessment process, the toxicity profile constitutes the best evaluation of risk. Toxicity profiles are available for approximately 17 195 chemicals from the U.S. EPA, Office of Waste Programs Enforcement and Office of Environmental Criteria and Assessment. DOSE-RESPONSE ASSESSMENT After the potential hazard associated with each contaminant of concern is characterized, the relationship between dose of a substance and its biological effect is determined. Dose-response data are used to determine the toxicological potency of a substance, a quantitative measure of its potential to cause a specified biological effect. There may be many dose- response relationships for a substance if it produces different toxic effects under different conditions of exposure. However, risks associated with a substance cannot be ascertained with any degree of confidence unless dose-response relationships are quantified, even if the substance is known to be "toxic." In developing risk assessment methods, U.S. EPA recognized that fundamental differences exist between carcinogenic dose-response variables and noncarcinogenic dose-response variables that could be used to estimate risks. Because of these differences, human health risk characterization is conducted separately for carcinogens and noncarcinogens. Key dose-response variables used in quantitative risk assessment are potency factors for carcinogens and reference dose (RfD) values for noncarcinogens. A generalized illustration of the role of these variables in dose-response relationships for carcinogens and noncarcinogens is shown in Figure 3. The carcinogenic potency factor [expressed in units of (mg/kg/day)~1] is typically determined by the upper 95 percent confidence limit of slope of the linearized multistage model that expresses excess cancer risk as function of dose. The model is based on high-dose to low-dose extrapolation, and also assumes that there is no threshold for the initiation of toxic effects. The RfD (expressed in units of mg/kg/day) is an estimated single daily chemical intake rate that appears to be without risk if ingested over a 18 FREQUENCY OF TUMORS Q = WwW - on” > n uw ° > oO z Ww > So w c uw on - Oo a "a u Ww 2 x< ° e + LOW-DOSE REGION OF CONCERN A~+—— SLOPE - POTENCY FACTOR 7 7 DOSE OF CARCINOGEN RTD NOAEL UF DOSE OF NONCARCINOGEN LEGEND OBSERVED DATA POINTS @ = Chemical A | Chemical B MODELS = — = Low Dose Extrapolation ——— Models Fi Within Observed Data Range RfD Reference Dose UF Uncertainty Factor NOAEL No Observed Adverse Effects Level Figure 3. Hypothetical dose-response relationships for a carcinogen, and a noncarcinogen. 19 lifetime (Vettorazzi 1976, 1980; U.S. EPA 1980, 1987a; Dourson and Stara 1983). It is usually based on the relationship between the dose of a noncarcinogen and the frequency of systemic toxic effects in experimental animals or humans, and assumes that a threshold exists for the initiation of toxic. effects (Dourson and Stara 1983). The threshold of observed effects is divided by an uncertainty factor to derive an RfD that is protective of the most sensitive members of the population. EXPOSURE ASSESSMENT Exposure assessment is the overall process that links sources and distribution mechanisms of chemical contaminants in the environment to human receptors. For risk assessment of chemically contaminated seafood, the exposure assessment is typically conducted in a number of stages: = Characterization of the environmental transport and fate pathways of chemicals of concern . Estimation of average concentrations of chemical of concern in seafood species . Characterization of potentially exposed populations, including fisheries harvest activities and patterns of seafood consumption s Estimation of the magnitude and duration of the dose of each chemical of concern incurred by the human receptor population. Mathematical models of key physical and chemical processes, which may be supplemented with experimental studies, are often relied upon to predict transport and fate pathways. Uncertainties associated with such predictions in the complex natural environment may be enormous, and may be avoided altogether by direct measurement of contaminant concentrations in potential exposure media. Thus, appropriately designed and conducted sampling and 20 analyses programs provide the best means for determining concentrations of chemicals of concern in edible portions of seafood species. For risk assessment of chemically contaminated seafood, the characteri- zation of exposed populations may include an identification of fisheries harvest zones, analyses of fisheries activities within those zones, an evaluation of the distribution of the catch, and characterization of consumption patterns that can be used in dose estimation. Only selected steps may be performed in a given exposure assessment, depending on data availability, study objectives, and funding limitations. Where comprehensive catch and consumption statistics are not available, estimates of seafood consumption rates may be based on standard values for the U.S. population or other assumed values. Dose is the amount of a chemical received by an organism over a specified time interval, and is usually expressed in units of mg chemical/ kg body wt/day. Because the oral route of exposure is the only route considered for consumption of chemically contaminated seafood, dose estimation may be based on the amount of a chemical that is ingested (i.e., an "ingested dose") or on the amount of material that is ingested and assimilated by absorption across the gastrointestinal lining (i.e., an “absorbed dose"). An ingested dose is obtained by multiplying the concen- tration of a chemical in seafood (expressed in units of mg/kg wet wt) by an ingestion rate (expressed in units of kg seafood/day), and dividing this product by an estimated human body weight (expressed in kg). The absorbed dose is obtained by multiplying the ingested dose by an assimilation coefficient. The assimilation coefficient is actually a composite variable that indicates the following: 7 Differences in actual assimilation efficiency between experimental animals that were used to derive dose-response variables and in humans . Age-specific differences in absorption efficiency within the human population 21 . Differences in the bioavailability of a chemical between the exposure medium used in determination of a dose-response relationship (e.g., water) and the exposure medium of concern in the risk assessment (e.g., seafood) . Differences in chemical speciation between the exposure medium used in determination of a dose-response relationship (e.g., inorganic arsenic in drinking water) and the exposure medium of concern in the risk assessment [e.g., total (i.e., inorganic plus organic) arsenic in seafood]. In summary, the dose, its duration and timing, and the characteristics of the population receiving it are the critical measures of exposure for the risk characterization. RISK CHARACTERIZATION In the risk characterization stage, the results of the dose-response assessment and the exposure assessment are combined to estimate the probability and extent of adverse effects associated with consumption of chemically contaminated seafood. Because of fundamental differences between dose-response variables for carcinogens and those for noncarcinogens, risk characterization is conducted separately for carcinogens and noncarcinogens (see Dose-Response section above). In its simplest expression, carcinogenic risks are estimated as the probability of excess lifetime cancer by multiplying the estimated human dose by the carcinogenic potency factor. For noncarcinogens, risks are expressed as a non-probabilistic risk index by dividing the estimated dose by the reference dose. The risk index is compared to a value of 1 (i.e., where the estimated dose equals the reference dose) to evaluate the chemical hazard (Stara et al. 1983; U.S. EPA 1985). Risk index values greater than 1 indicate that the estimated exposure is potentially of concern. Because data on chemical interactions are limited, estimated risks for individual 22 chemicals are usually summed to obtain an approximate estimate of total risk for a chemical mixture. Because technological limitations preclude analyzing seafood samples for all potentially toxic chemicals, risk estimates should not be interpreted as estimates of total risk associated with seafood ingestion. The risk characterization step can be much more complicated than indicated here (Tetra Tech 1986a; PTI 1987). Independent risk estimates can be calculated for each chemical in each exposure medium (e.g., different species of seafood) at various locations defined in the exposure assessment. Depending on the available data, risks can also be partitioned among subsets of the exposed population. The objectives of such an approach are to determine who is at risk and which combination of chemicals and media account for that risk. The results of the risk assessment may be presented in both a tabular and graphic format. In the supporting text of a risk assessment, all final estimates of risk should be rounded to one significant digit, or an order of magnitude if appropriate. Interpretation of noncarcinogenic risk index values is based on direct comparisons with reference dose values and supporting information developed in the toxicity profiles. Risk index values greater than 1 provide only a general indication of concern for potential toxic effects. The severity of such effects must be interpreted in light of information concerning the dose- response relationship for the specific chemical and toxic endpoint in question. Interpretation of carcinogenic risk estimates may be based on the following comparison of health risks for the study area: 7 Health risks for consumption of seafood from a reference area . Health risks for consumption of alternative foods (e.g., charcoal broiled steak) 23 a The range of allowable health risks determined in regulatory decisions. In general, U.S. EPA decisions concerning individual lifetime risk estimates in the range of 10-7 to 10-3 are made on a case-by-case basis and are strongly influenced by considerations of the population risk (Travis et al. 1987; Travis and Haltemer-Frey 1988). Population risk is an estimate of the number of cancers produced within a population of specified size per generation. High individual risk estimates may translate into low population risk (i.e., less than 1) where the size of the exposed population is small. Individual risk levels greater than 10-3 are usually subject to regulatory action, and those less than 1077 are rarely subject to regulatory action. UNCERTAINTY ANALYSIS There are numerous factors associated with each step of the risk assessment process that contribute to uncertainty in the risk characteriza- tion. Quantitative approaches to the characterization of uncertainty are described in the U.S. EPA guidance manuals on risk assessment of chemical contaminants in aquatic organisms (Tetra Tech 1986a; PTI 1987), but are beyond the level of discussion intended for this summary. The following major factors are associated with uncertainty in risk estimates: 1. Uncertainties in the determination of the weight-of-evidence classification for potential carcinogens. 2. Uncertainties in estimating Carcinogenic Potency Factors or RfDs, resulting from: 2 Uncertainties in extrapolating toxicologic data obtained from laboratory animals to humans 5 Limitations in quality of animal study 24 . Uncertainties in high-dose to low-dose extrapolation of bioassay test results, which arise from practical limitations of laboratory experiments and variations in extrapolation models. Variance of site-specific consumption rates and contaminant concentrations. Uncertainties in the selection of assumed values for consumption rates (e.g., 6.5 g/day, 20 g/day, and 165 g/day) when site-specific data are not available. Uncertainties in the efficiency of assimilation (or absorp- tion) of contaminants by the human gastrointestinal system. Variation of exposure factors among individuals, such as: 7 Variation in fishery species composition of the diet among individuals . Variation in food preparation methods and associated changes in chemical composition and concentrations due to cooking. Uncertainties associated with risks from chemical contaminants that were not included in the analysis of seafood tissues, but may be present nonetheless. Adequacy of the study design to provide the numbers of samples needed to describe accurately contaminant concentrations in each species at various locations and times (e.g., seasons) throughout the study area. 25 9. Uncertainties in interpretive aspects of the risk characterization, which arise from such factors as: 2 Lack of data for contaminant concentrations in seafood from relatively clean reference areas, which are needed to describe accurately "background" levels of risk associated with seafood consumption a Lack of reference diet information needed to assess accurately "background" concentrations of chemicals of concern in non-seafood components of the diet, which may be particularly important for comparisons of high exposure populations, whose diet may include a dispropor- tionate amount of seafood, with the general population, whose diet may consist of an average amount of seafood plus a variety of other foods . Lack of additional risk-benefit information concerning the influence of diet on other health endpoints (e.g., heart disease), which, again, may be important for comparisons of high exposure populations, whose diet may include a disproportionate amount of seafood, with the general population, whose diet may consist of an average amount of seafood plus a variety of other foods. In conclusion, uncertainty ranges (e.g., 95 percent confidence intervals) around estimates of mean risk may typically span 3-5 orders of magnitude. The approach taken by U.S. EPA (1980, 1985, 1986a; Tetra Tech 1986a; PTI 1987) is to estimate a plausible upper limit to risk. In this way, it is unlikely that risk will be underestimated substantially. Moreover, the plausible-upper-limit estimate serves as a consistent basis for relative risk comparisons. 26 NATURAL RESOURCE DAMAGE ASSESSMENT Tetra Tech proposes to conduct a natural resource damage assessment for the Port Valdez oi] spill in the three phases that correspond to Tasks A, B and C in Figure 4. The purpose of Task A will be to evaluate the impacts of the oil spill. This will be accomplished by integrating historical baseline data (e.g., National Oceanic and Atmospheric Administration data collected under the Outer Continental Shelf Program), ongoing monitoring data collected by agencies and private parties, data collected during recent and ongoing field investigations (initiated in response to the oil spill), and information in the scientific literature. An economic assessment of damages to natural resources and habitats of concern (Task B) will proceed in parallel to Task A and will use information generated during the execution of Task A. Finally, cost-effective restoration plans will be developed in Task C. These restoration plans will be based on technical information generated during the execution of Task A (e.g., information on the nature and extent of impacts to various biological resources), and on economic information generated during the execution of Task B (e.g., revenue lost to the local economy as a result of impacts to harvested marine life). A report will be prepared and submitted to Exxon at the conclusion of each of these three tasks. TASK A. EVALUATE IMPACTS OF OIL Activities conducted under this task provide the technical basis for determinations of impacts to natural resources under the trusteeship of NOAA and other agencies. Four major subtasks are involved: . Subtask A.1. Evaluate distribution concentrations of contaminants 7 Subtask A.2. Identify contaminant pathways, habitats of concern, and resources of concern 27 Task A: EVALUATE IMPACTS OF OIL + Evaluate Distribution and Concentrations of Contaminants + Identify Contaminant Pathways, Habitats of Concern, and Resources of Concern + Evaluate Impacts of Oil + Prepare Impact Assessment Report Task C: Task B: ASSESS DAMAGES TO HABITATS AND RESOURCES OF CONCERN + Determine Baseline for Natural Resource Services * Determine Types of Natural Resource Services * Quantify Losses in Natural Resource Services + Measure Value of Natural Resource Services + Determine the Market for Resource Services + Determine Damages + Prepare Damage Assessment Report DEVELOP RESTORATION PLANS + Assess Potential For Restoration for Each Habitat and Resource of Concern + Develop Cost-Effective Restoration Plans + Prepare Restoration Report Figure 4. Tetra Tech's approach to assessing damages to natural resources. . Subtask A.3. Evaluate impacts of oil on natural resource receptors 2 Subtask A.4. Prepare Impact Report Each of these four major subtasks is discussed below. Subtask A.1. Evaluate Distribution and Concentrations of Contaminants The first step that Tetra Tech will undertake in evaluating oil concentrations is to determine the quantities and distribution of oil present at the site, the present and historical sources of oil, and the oil fractions that are likely to migrate out of the immediate area of the spill. Information on these topics will be available in the scientific literature (e.g., refereed scientific journals, agency reports, and technical documents prepared for private parties) and from field investigations conducted in response to the oi] spill. After compiling available information, the studies will be reviewed critically to assess completeness of the data, technical quality of the data, and adequacy of the data to substantiate the position Exxon wishes to take regarding damages to natural resources. The conclusions drawn from this critical review will also be used to determine whether the collection of additional data (i.e., a field survey) or the initiation of long-term monitoring is warranted. For each study, important questions that will be considered during the critical review include the following: 2 Are the assumptions of the study appropriate to its objec- tives, and to the environmental characteristics of the site and adjacent areas? . Are the study hypotheses reasonable and relevant to the study objectives? 29 . Will the survey design provide data that are adequate to test the stated hypotheses? . Is a priori knowledge of the receiving environment integrated into the study design? . Were field, laboratory, and analytical procedures appropriate and properly executed? Tetra Tech's experience with numerous regulatory programs for the U.S. EPA has repeatedly shown that the quality of environmental data is often low, and therefore, that the available data are often insufficient to determine whether impacts or potential for impacts exist, and if so, the extent and magnitude of those impacts. Commonly encountered problems with data quality include inadequate sample replication in the field (thereby precluding or limiting statistical testing), poor field sampling techniques (which render data questionable), and inadequate or poorly executed laboratory procedures. Commonly observed problems with chemistry data include high variability among values of laboratory replicates, poor accuracy of values from spiked samples, and high detection limits. Each of these problems has the potential to render data suspect or unusable, and hence, to weaken Exxon's position regarding impacts and damages to natural resources. In many cases, available data will be sufficient to assess impacts to natural resources. In other cases, however, data gaps will be identified in the critical review process. When data gaps are identified, Tetra Tech will, at the request of Exxon, design and execute field investigations to collect essential data. Field surveys will be designed to collect supple- mentary data, or comprehensive data, as warranted. All technical concerns that are relevant to critical reviews of available data will be considered in the design and execution of field surveys, thereby promoting the collection of high quality data that are adequate to substantiate Exxon's position regarding impacts and damages to natural resources. After field 30 surveys have been completed, the newly collected data will themselves be reviewed critically to assess their quality and adequacy. Having defined a comprehensive, high quality data set, Tetra Tech will use those data to estimate the distribution and concentrations of oi] in the receiving environment. Data values will be plotted Manually or electron- ically to determine the spatial extents and gradients of oi] concentrations in three dimensions. Analyses that may be used to determine the spatial extent and magnitude of contamination include univariate and multivariate statistical analyses, regression, correlation, and gradient analyses. The size of the potentially contaminated area will be determined, the extent of contamination, and the potential for migration of oil will be estimated. Spatial and temporal data from the spill site, supplemented by information from other comparable spill sites (if available) and information on the typical behavior of oi] under similar conditions, will often be sufficient to estimate the period and frequency of oil releases from the site. Estimates of the dynamics of oi] distribution and transport will be made to assess the potential for offsite impacts to natural resources. Subtask A.2. Identify Contaminant Pathways, Habitats of Concern, and Resources of Concern Tetra Tech will conduct this major subtask simultaneously with Subtask A.l. The environmental setting largely determines the possible pathways for oi] transport, and the habitats and species of concern. Primary character- istics of the environment that Tetra Tech will examine include the hydro- graphy, bathymetry, and geology of the site and surrounding area. These three factors determine the habitats and natural resources that are present (e.g., kelp beds, anadromous fishes), their spatial and temporal distri- butions, and the mode and rate of transport of oi] within and away from Port Valdez. 31 The location of the oil and its proximity to resources of concern will also be examined in detail by Tetra Tech to determine existing and potential damages to natural resources. In many cases, oil will already have impacted natural resources. In other cases, resources of concern are located some distance from the site of oil contamination, and it will be necessary to determine whether those resources have been affected, or will be affected, as a result of contaminant transport. Available empirical data may be sufficient to document that such transport has occurred, and that natural resources have been damaged. However, transport and subsequent damage to natural resources may not be evident based on empirical data, and it may be necessary to predict whether oil transport and subsequent damage are occurring. The basic approach that Tetra Tech will use in performing such predictions will be to estimate the rate of oi] transport that has been realized since contamination was initiated, and to determine whether that rate has been sufficient for oi] to have been transported into the area where resources of concern are located. When conducting such predictions, Tetra Tech will: . Determine the likely pathways for oil transport (e.g., transport by surface and tidal currents) . Identify the possible transport mechanisms (e.g., volatiliza- tion, sorption, advection, diffusion, sedimentation) a Consider possible chemical transformations of oil under expected environmental conditions (e.g., hydrolysis, photolysis, oxidation-reduction, microbial degradation) . Apply appropriate transport models. Appropriate models may be very simple or very complex. Simple models might provide estimates of the quantities of oil constituents that have been volatilized from the intertidal zone. More complex models might consider 32 the effects of tidal currents on transport of oil at the surface and at depth. Subtask A.3. Evaluate Impacts of Oil Having evaluated the distribution and concentrations of oi] (Subtask A.1), and contaminant pathways and habitats of concern (Subtask A.2), Tetra Tech will evaluate the potential impacts of oi] constituents on resources using information on the environmental partitioning of the oi] and on the vulnerability and effects of oi] constituents on the resident biota that is found in the scientific literature. When information on the environmental partitioning of oi] and on the effects of oil constituents on resident biota is consistent with empirical evidence collected in the receiving environment, it will often be possible to establish with reasonable certainty that the oil constituents are impacting habitats and biota in the receiving environ- ment. Information on the environmental partitioning of oil constituents is important because different species use different parts of the receiving environment. For example, pelagic species are not likely to be affected by oil that is adsorbed to particulates, deposited on the bottom, and incor- porated into the sediment matrix. However, demersal fish that prey on benthic organisms may experience direct and indirect effects of that oil (e.g., direct toxicity, bioaccumulation, impaired reproductive success). Alternatively, pelagic and demersal species of fish may be affected by low molecular weight constituents of the oi] that are dissolved in the water. Environmental partitioning is also important because oi] constituents may be more persistent or available to the resident biota when in one part of the environment than another. Deposition of oi] adsorbed to particulates is likely to result in a reservoir of contaminated sediments that persists through time, continually impacts the resident benthic biota, and is bioturbated, aerated, and released to the overlying waters over extended periods of time that may exceed the time over which oil] is transported into the area. Conversely, oi] in dissolved form may be quickly transported out 33 of the immediate receiving environment, and may be diluted to such low concentrations that impacts are difficult or impossible to detect. Other 0i1 constituents may be degraded by microbes into more toxic or less toxic chemical forms. A considerable body of scientific information now exists that documents the effects of oil constituents on the resident biota. Tetra Tech will use this information, in conjunction with empirical information on the spatial and temporal distribution of oi], and data on the habitats and biota in the receiving environment, to document the extent and magnitude of impacts to habitats and biota under trusteeship to NOAA and other agencies. Direct evidence of impacts to the receiving environment include the identification of oil constituents in the water, sediment, and tissues of the resident biota. Indirect evidence of the impacts of oi] constituents on the resident biota include reduced diversity, abundance, and variability of the resident biota; the presence of different species than would be expected under the extant environmental conditions; the absence of species known to be sensitive to contamination; and altered predator-prey relationships. Such indirect evidence of impacts represents the cumulative effects of oi] constituents on critical life history stages of the resident biota, including impaired reproductive success, abnormal physiology, sublethal toxicity, and elevated disease prevalence. Each of these latter types of impacts is typically inferred from available empirical data because each is extremely difficult to detect under uncontrolled conditions. Tetra Tech will identify the natural resources that could be affected by oil by integrating information collected under this subtask with information from Subtasks A.1 and A.2. directly or indirectly. Direct evidence of contaminant effects will include the presence of habitats (e.g., National Marine Sanctuaries, Estuarine Research Preserves) or species under the trusteeship of NOAA or other agencies within the documented area of impact, and the presence of contaminants in tissues of such species. Habitats of particular importance will include habitats critical to endangered or threatened species, spawning and nursery habitats for commercially and recreationally harvested species, and areas that support 34 commercial and recreational fisheries. Species of particular importance will include all life stages of commercially and recreationally harvested species, all life stages of prey organisms of harvested species, and all life stages of threatened and endangered species. Indirect evidence of impacts will include changes in the habitat or resident biota that are consistent with the documented impacts of the oil, and predictions of possible impacts based on models of oi] transport and fate, as discussed above. Subtask A.4. Prepare Impact Assessment Report Tetra Tech will prepare a complete, accurate technical document that summarizes and interprets all finding from this subtask. This document will form the technical basis for any position Exxon may wish to take regarding documented, suspected, or predicted impacts of the oi] spill on habitats and resources. The technical skills and experience of the Tetra Tech staff, plus the full-time editorial and publication staff available at the Bellevue office of Tetra Tech will ensure that the Impact Assessment Report will be accurate, comprehensive, concise, and clearly written. TASK B: ASSESS DAMAGES TO HABITATS AND RESOURCES OF CONCERN Task A evaluates the nature and extent of physical and biological impacts associated with the oil spill on natural resources under the trusteeship of NOAA and other agencies. Task B uses information from this task, as well as other information, to assess damages to those natural resources. To standardize determinations of damages, Natural Resource Damage Assessment regulations were proposed by the U.S. DOI in the mid 1980s, and revised in May 1988. Not everyone agrees that current NRDA regulations provide proper guidance in every respect (Desvousges et al. 1989). Many NRDA provisions/issues are being contested in lawsuits. Among these are five important economic issues: 35 2 Measuring damages as the lesser of restoration/rep]acement cost or the diminution of use values = Limiting recoverable damages to foregone "public uses" of natural resources . Choosing willingness to Pay as the valuation criterion for estimating natural resource damages, rather than willingness to accept . Preferring market-based valuation methods over nonmarket valuation methods for measuring damages from foregone uses 7 Excluding foregone nonuse values from damages, except when no foregone use values can be determined. Any changes in these important issues may significantly affect the magnitude of recoverable natural resource damages. The rules and Procedures for conducting a formal natural resource damage assessment are still evolving. Nevertheless, the NRDA regulations do provide a frame of reference that Tetra Tech will use in structuring the activities of this task and describing its proposed method of damage determination. Tetra Tech's approach to a full scale determination of natural resource damages largely follows procedures established for Type B NRDA assessments. The approach proceeds in the seven subtasks listed under Task B in Figure 4. Subtask B.1. Determine Baseline for Natural Resource Services-- The difference between with-injury and without-injury (i.e., baseline) service levels is the basis for determining foregone natural resource services. All else being equal, the magnitude of natural resource damages will therefore increase as the difference between with- and without-injury service levels increases. With-injury service levels are usually easier to 36 measure than baseline service levels because they are observed (or are observable). Baseline service levels must be estimated. Even when information on uses of natural resources following an injury exists, determining how many people would have used the resource if it had not been injured is not possible. The difficulty of establishing baseline service levels varies with the two principal types of oi] releases: 7 Single releases . Multiple (or continuous releases). It is usually easiest to estimate baseline service levels for a single release, such as an oil spill. In such cases, natural recovery may occur relatively soon, restoring natural resource services to without-injury levels. Consequently, baseline service levels could be interpolated during the short injury/recovery period using the pre-and post-injury service levels. Multiple (or continuous) releases of oi] complicate the deter- mination of baseline service levels because they must be estimated by extrapolating service levels prior to the releases. Alternatively, baseline service levels can be estimated using "control areas." Using this approach, baseline service levels for a similar, but uninjured, natural resource in a nearby area can be substituted for extrapolated service levels for the injured natural resource, if such control areas exist. This approach was used in the Martinez oi] spill assessment by RTI. Subtask B.2. Determine the Types of Natural Resources Services-- The second step in the Tetra Tech approach to a full-scale assessment is to determine the types of natural resource services that are injured by the hazardous substance releases. Three important economic issues arise in this step: 37 . Definition of eligible uses . Determining private vs public uses . Treatment of nonuse values. The definition of eligible uses is the first critical issue. The U.S. DOI definition of the "uses" of natural resources that can be included in the damage assessment process significantly influences the scope of the regulations. In this case, U.S. DOI has taken a relatively narrow view of what constitutes use by including only "committed" uses [Section 11.14(h); 51 FR 27727]. Under the U.S. DOI's definition, recreational uses of natural resources appear to be the most significant type of use. These uses include both direct and indirect types. Examples of direct uses are boating, fishing, swimming, and hunting. Indirect uses include relaxing in a wildlife refuge, bird watching, or nature photography. Tetra Tech will work closely with Exxon to determine relevant uses. These efforts will include reviewing outdoor recreation plans that cover many of the natural resources within their boundaries, and reviewing historical records that document uses that existed before the releases occurred. The second important aspect in determining the types of natural resource services involves the subtle distinction the regulations make in how natural resources are used. The regulations distinguish between private uses of resources for personal income (i.e., private damages) and private uses that result in no income gain (i.e., public damages; see 51 FR 27680). An example will help to illustrate this distinction. Suppose the release of oil injures the natural biological services from an estuary. Under the regulations the public trustee is allowed to recover any losses from decreased uses of the estuary by any citizens. However, the trustee is not allowed to recover any lost wages or income for people who conduct a 38 business at the estuary (e.g., a marina or boat rental service). These individuals would be required to bring a private action. The basis for this distinction is primarily U.S. DOI's interpretation of the CERCLA legislation. Although the CERCLA definition appears to apply only to public resources, such resources may still have both public and private uses. Using the estuary example, the estuary's values would be based on all the services it generates for society. These services would include both uses and nonuses (i.e., potential or optional uses). The basis for measuring damages would be the change in the estuary's services, with and without the injury. Whether private losses such as reduced income or wages would be included is more complicated. Based on the value of the estuary's services, these would not be included because they are payments to individuals for their financial capital or labor services. Lost income is not reflected in the value of the resource itself. However, the Tetra Tech approach recognizes that it is important to avoid double counting if private claims are sought. The third and final important issue in determining the types of natural resource services to be included in the full-scale assessment is the treatment of nonuse values. The regulations define options value as the dollar amount that people who are not currently using the resource are willing to pay to preserve their option to use that resource in a certain state-of-being in the future. Existence value is the maximum amount people are willing to pay to know that a resource would continue to exist in a certain state-of-being, even though they have no plans to use the resource. The "state-of-being" can be interpreted as a level of quality (e.g., the quality of an estuary in its pre-injury condition). The economics literature has only recently begun to resolve the conceptual and empirical issues involving option value, or nonuse value more generally. Nevertheless, there is a consensus among natural resource economists that nonuse values are a legitimate part of the total value of natural resource services. There is less consensus on how large these values might be. When defined and 39 measured correctly, Tetra Tech's position is that nonuse values should be added to use values to get the total damages. The Tetra Tech team has considerable experience in measuring nonuse values. RTI has pioneered the development of survey-based methods that have received considerable attention in peer-review journals. RTI also is nationally recognized for its survey cap» ‘lities which will provide Exxon with a damage assessment that can withstand professional scrutiny. Subtask B.3. Quantifying the Losses in Natural Resource Services-- In this step, Tetra Tech will determine the extent to which natural resource services have been reduced as a result of the injuries. The types of services that are relevant to full-scale assessment include: . Provision of habitat, food, and other needs of biological resources . Recreation 7 Other products or services used by humans . Flood control . Groundwater recharge 2 Waste assimilation. Methods for quantifying natural resource services will be selected, based on: . The degree to which a particular service is affected = The degree to which a service can be used to represent a broad range of related services 40 7 The consistency of the measurement with requirements of economic theory 2 The technical feasibility of quantifying changes in services at reasonable cost . The preliminary estimates of services at the assessment area and control area based on resource inventory techniques. Tetra Tech's approach to measuring losses in natural resource services will closely follow U.S. DOI's general procedural guidelines. However, Tetra Tech will measure the range of lost services that is consistent with sound economic principles. Subtask B.4. Measure the Value of Natural Resource Services-- There are two important factors to consider when measuring the value of losses in natural resource services that result from an oi] release: the criterion for measuring values, and the methods chosen to measure damages. Each of these factors can have a major effect on how the damage assessment is conducted and on the magnitude of damages. Tetra Tech's approach addresses both factors in this crucial subtask in the full-scale assessment. Valuation Criteria--Economics uses two basic criteria for measuring the value of the reduction in services from an injury to natural resources: . The willingness-to-pay (WTP) criterion: How much would an individual be willing to pay to have avoided the injury? 2 The willingness-to-accept (WTA) criterion: How much compensation would an individual require to be as well off as he was without injury? 41 The WTP criterion obtains an individual's value for an injury by determining how much he would have paid to avoid the loss. The WTP criterion however, is constrained by an individual's ability to pay. The WTA criterion asks how much compensation an individual would require to be at the same level of utility or well-being that he was without the injury. Neither valuation criterion is superior for a damage assessment. Willingness to accept is probably more appropriate from a conceptual point of view: it would be desirable to know the amount of compensation that would offset the values of losses in natural resource services from the release. However, willingness to pay has a much better track record in providing reliable estimates (see Cummings et al. 1986; and Mitchell and Carson 1988). Tetra Tech proposes to include both valuation criteria in a full-scale assessment. This departs from the U.S. DOI regulations which include only willingness to pay. Team experience suggests that using both criteria provides a stronger case for establishing the range of damages. Measurement Method--Choosing a method for measuring damages is the second important part in determining the value of lost resource services. Figure 5 compares the alternative methods by adapting the Smith and Krutilla (1982) framework. As shown in the first two vertical sections of the figure, the classification consists of two majors elements: . The types of linkages between reductions in resource services and their observed effects . The types of assumptions required to use the measurements methods. The measurement alternatives can be grouped into two classes: behavioral and nonbehavioral. The behavioral alternatives use either direct, observed behavior, indirect observations, or expressed preferences of households and firms to link values to resource services. Market price and appraisal methods are designed to directly measure resource services. In the market price method, the resource services are actually traded in the 42 TYPE OF LINKAGES BETWEEN REDUCTION IN RESOURCES SERVICES AND ITS OBSERVED EFFECTS TYPE OF ASSUMPTIONS REQUIRED FOR MEASUREMENT METHODS MEASUREMENT METHOD Market exists for resource services or market exists for comparable resource services. + Market price * Appraisal BEHAVIORAL Market exists for products produced with resource services or restrictions on individual's preferences not technically observed in delivery of resource services. + Factor income + Hedonic property value + Travel cost Resource services can be described and valued in simulated market using expressed preferences. + Contingent valuation Value of services lost at least equal to the cost of replacement or restoration. + Replacement costs + Restoration costs NONBEHAVIORAL Reference: Adapted from Smith and Krutita (1982). Assumes values can be transferred from one resource to another. Figure 5. Framework for classifying measurement methods. * Unit day value market. This method assumes that a market exists for trading these services--an unlikely occurrence for most resource services. The appraisal approach, also a direct behavioral method, resembles the market-price approach. However, instead of using values that are directly observed during market transactions, it uses an appraiser's knowledge of markets for similar resources to estimate market price. As illustrated in Figure 5, the second group of behavior-based valuation alternatives uses indirect linkages to value natural resource services. Using these techniques, a household or firm's observed choices (e.g., a visit to a recreation site or a property choice) are indirectly linked to a resource, with the household behavior revealing an implicit value for them. Using this group of alternatives eliminates the need to assume that resources are directly involved in a market transaction. Instead, these methods (including the factor income, hedonic property value, and travel cost models) require only that the resources of concern be related, or linked in some way to some other good or service traded in a market. Clearly, the assumptions for the indirect behavior-based methods are less restrictive than those required by the direct behavior-based methods. Contingent valuation, the last type of behavior-based valuation method, assumes that individual's behavioral responses to reductions in resource services can be estimated by eliciting individual's expressed preferences for them. In effect, this method assumes that expressed preferences are consistent with the behavior individuals would reveal in a market if it existed. The second class of valuation alternatives, nonbehavioral, is shown in the lower two horizontal sections of Figure 5. These alternatives are considered nonbehavioral because they exclude the use of individual's behavioral responses (e.g., visiting a substitute recreation site) in valuing resource services. The most prominent method in this category, replacement and restoration cost, uses the costs of restoring or replacing the reduced services, thus omitting any role for behavioral changes or preferences. 44 The U.S. DOI regulations also designate the unit-day valuation method as an alternative for valuing reductions in resource services. In this method, values measured for other resources are transferred to injured resources. In effect, the unit-day value method implicitly accepts the assumptions of the methods used to determine the values for other resources (typically, either travel cost, contingent valuation, hedonic techniques, or some combination). This method also assumes that the transferred values are representative of those for the injured resource; either that the resources have identical characteristics and yield the same use values, or that any differences between them have no effect on values. Finally, the unit-day value method also assumes that some "off-the-shelf" values are available. For some recreational activities or services (e.g., fishing, boating, swimming, and to a lesser extent, wildlife-related activities), this assumption is likely to hold: for other less frequently studied resources, it may not. In the U.S. DOI regulations, the required hierarchy of alternative methods is as follows: : Market price, when applicable . Appraisal, when applicable . Any nonmarketed resource methodology which measures willing- ness to pay including, but not limited to, factor income, travel cost, hedonic pricing, contingent valuation, and unit- day value. The U.S. DOI regulations also add that contingent valuations should be used for measuring option and existence values only when there are no use values. The team's experience indicates that this hierarchy is unlikely to affect the choice of a measurement method; nonmarketed resource methodology will typically be used because the market price and appraisal methods are inappropriate for most natural resources. The main choice will be between 45 the types of nonmarket methods. The choices of a method depends on the following four factors: 2 The nature and magnitude of the injuries . The type of resource services affected by the spill . The types of decision makers affected by the spill (e.g., household or businesses) . The time available. Based on the team's experience, no valuation method is unambiguously superior. Each has its own strengths and weaknesses. RTI's research suggests that it is necessary to match the method with the types of resource services that are lost. The following general guidelines can be used in evaluating valuation methods: . The market price or appraisal methods are preferred for any resource services that are traded in markets. a The factor income method is most applicable to agriculture or extractive uses (i.e., if the products are sold in markets). . The hedonic method can be used to value recreation and perhaps agricultural uses. It also may be used for some aesthetic values. . The travel cost method is best suited for recreation uses. . The contingent valuation method is best suited for recreation uses and for aesthetic and existence values. 46 . The unit-day valuation method is best suited for simple recreation uses because recreation has more information available. However, it is poorly suited for recreation sites with unique characteristics or ones involving complex recreation issues. The costs of applying a method depends largely on the specific circumstances involving each resource. It also depends on the cost of analyzing the data. The hedonic approach involves the most complex analysis and consequently, is the most expensive. Contingent valuation and travel cost analyses should cost approximately the same. They are on the same level of complexity and involve many similar tasks. The team has also found that more than one approach or multiple methods can be used. These methods can provide consistency checks on the damage estimates. RTI has considerable expertise with all the NRDA valuation methods. Their expertise includes: 7 Developing new ways of implementing the contingent valuation method 7 Comparing contingent valuation and travel cost methods for single sites . Developing damage assessment models that can be transferred among sites. This experience will be most useful in assisting NOAA in choosing the most effective method for measuring damages. Subtask B.5. Determine the Relevant Market for Resource Services-- Tetra Tech will work closely with NOAA to determine the relevant market for the natural resource services. Possible market areas include: 47 . Local market area--only people in the immediate area are affected 7 Regional market area--people in the immediate areas and several adjacent counties are affected . Multi-state market area--people within several hundred miles are affected. The interaction of three main factors usually determines the relevant market area: . The characteristics of the services provided by the natural resource . The proximity and accessibility of the natural resource to potential users . The availability of substitute natural resources to potential users. Generally, the magnitude of natural resource damages increases as the size of the affected market area increases, other things being equal. Identifying the relevant stakeholders (i.e., the number of people affected by an oil release) for nonuse values is more problematic than for use values. Clearly, the characteristics and "uniqueness" of the injured natural resource are important determinants of nonuse values. However, the role of proximity and accessibility is less clear. Nonusers presumably base their willingness to pay on some knowledge of the natural resource and its attributes. Sutherland and Walsh (1985) hypothesize that this knowledge "is transmitted by the media and by people visiting the site." Knowledge of the natural resource to nonusers will probably decline as distance increases. However, this inverse relationship between distance and nonuse values may be very weak for unique natural resources. 48 In most instances a hazardous substance release will affect natural resource services levels over several years (and perhaps decades), which complicates the identification of relevant stakeholders. Specifically, as changes occur in population levels, income, tastes and preferences, and as new roads are built, the number of people affected by a natural resource injury may change. Therefore, the magnitude of natural resource damage estimates will be sensitive to changes in relevant stakeholders over time. Subtask B.6. Determine Damages-- The sixth step in Tetra Tech's approach to a natural resource damage assessment will use the best valuation method for the relevant market of natural resource services to measure damages on an annual basis. Tetra Tech will develop a total damage estimate by converting annual damages into their present value (through a process known as discounting) and summing them. The dollar value of damages occurring in different years must be discounted before summation to properly account for the time value of money. Specifically, discounting assumes that people prefer consumption in the present to consumption in the future. Thus, when faced with a choice of receiving $100 now or $100 one year from now, people choose the immediate payment. Taking this a step farther, people must be offered some extra compensation before they will choose to wait a year for a particular payout. For example, even without inflation, a person may require a $110 payout one year from now before being indifferent between this payout and the immediate $100 payout. The extra compensation for waiting a year represents that person's implicit tradeoff of future consumption. In other words, the 10 percent ($10/$100) premium required for waiting a year for the payout reflects the person's willingness to substitute future consumption for present consumption. Up to this point, the focus has been on discounting from an individual's perspective. A social discount rate is needed when aggregating natural resource damages over time. Over the last 30 years, economists have extensively debated the determination of the "proper" social discount rate. 49 There is a growing consensus among economists that the social discount rate should reflect the social rate of time preference, which is the rate at which people are indifferent to substituting consumption in the present for consumption in the future (Lind 1982). Taking into account inflation, taxes, and rates of return available to investors, Lind (1982) estimates the social rate of time preference as about five percent. The U.S. DOI regulations specify a 10 percent social discount rate for aggregating natural resource damages over time in accordance with the Office of Management and Budget. In general, a high discount rate will lead to a larger present-value damage estimate for damages in the past than a low discount rate. Alternatively, a high discount rate will produce a smaller present-value damage estimate for future damages than a low discount rate. The impact of the discount rate on the magnitude of natural resource damages when there are both past and future damages depends on the specific mix of these damages over time. For this sixth subtask, Tetra Tech proposes to estimate natural resource damages using several discount rates in order to evaluate the sensitivity of total damages to the discount rate. Based on this sensitivity analysis, Tetra Tech will recommend to Exxon the most appropriate discount rate. Subtask B.7. Prepare Damage Assessment Report Tetra Tech will prepare a comprehensive damage assessment report that explains in detail the assumptions, approaches, analyses, and conclusions derived from the execution of each of the above subtasks. The damage assessment report will consist of a characterization of site conditions as identified in the Task A report, an overview of the methods used to assess damages, a description of the assessment of those damages (Subtasks B.1-B.5 above), and an itemization of damage estimates (Subtask B.5 above) and costs associated with the damage assessment. The report will provide possible justifications and bases for any judicial or administrative action that may be taken against Exxon to compensate for damages to 50 resources. Consequently, the report will be prepared in consultation with Exxon and with attorneys who specialize in environmental law. TASK C. DEVELOP RESTORATION PLANS Having evaluated extant and potential impacts of the oil, Tetra Tech will develop restoration plans for habitats and resources of concern. Three subtasks will be conducted, including an assessment of the potential for restoration, the development of restoration plans, and preparation of a restoration report (Figure 4). Subtask C.1. Assess Potential for Restoration for Each Habitat and Resource of Concern The results of Task A will form the technical basis for determining which impacted resources have the potential for restoration, and the degree of restoration that may be achieved. The review of the scientific literature that will be conducted as part of Task A will be invaluable for determining the potential for restoration of various habitats and resources. It will identify habitats and resources that have been successfully restored in full or in part, as well as habitats and resources that have not been restored. It will also identify and suggest potential methods for effecting such restoration. This information, in conjunction with information specific to the Port Valdez spill and the professional experience of Tetra Tech scientists, will be sufficient to identify habitats and resources that have high, moderate, and low potentials for restoration. Results of the damage assessment will assist in evaluating the realized and potential damages, and in prioritizing restoration efforts. Developing Cost-Effective Restoration Plans The evaluation of remedial alternatives and the development of restoration plans is a critical part of Tetra Tech's proposal to assist Exxon with the Port Valdez oi] spill. 51 Tetra Tech has extensive experience recommending remedial alternatives, including the recommendation of preferred alternatives for disposal of contaminated dredged materials from Commencement Bay (for the Washington Department of Ecology), and the recommendation of preferred oil spill cleanup methods in different coastal habitats (for the American Petroleum Institute). The approach that Tetra Tech will use to recommend preferred remedies and identify conditions that will mitigate resource impacts will consist of the following steps: . Identify and screen possible technologies for remediation based on technical feasibility and ability to effectively remediate the contaminants of concern . Assemble viable remedial technologies into a series of remedial alternatives for evaluation based on short- and long-term impacts, long-term protectiveness, potential for recovery, implementability, institutional feasibility, and cost . Rank the alternatives from most preferred to least preferred based on the above criteria and ability to meet risk/resource- based remediation goals (including goals for protection of human health, as appropriate), the environmental setting, and the habitats and biota at risk. This basic approach requires a thorough understanding of the potentially affected site (including habitat characteristics and resident species). Understanding of the potentially affected site will have been achieved through the execution of Task A. This approach also requires highly trained and experienced personnel such as Tetra Tech's, who are familiar with current remedial technologies and their applications and limitations, and the realized or potential impacts of the contaminants of concern on the habitat and resident biota. 52 Subtask C.3. Prepare Restoration Report Recommendations for remediation of the oil spill will be presented in a comprehensive, concise, and clearly written report. The report will identify those resources that have been, or may be damaged by the oil spill. Among those identified resources, recommendations will be made regarding those that have the potential to be restored, and the expected degree of restoration that is possible given existing technology. Resources that have the potential to be restored will be ranked based on their ecological and economic importance. Finally, detailed restoration plans will be developed for those resource for which restoration is possible and desirable. Decisions regarding which resources should be restored will be made in consultation with Exxon and appropriate agency representatives. 53 REFERENCES Dourson, M.L., and J.F. Stara. 1983. Regulatory history and experimental support of uncertainty (safety) factors. Regul. Toxicol. and Pharmacol. 3:224-238. National Research Council. 1983. Risk assessment in the federal government: managing the process. The Committee on the Institution of Means for the Assessment of Crisis to Public Health, Washington, DC. PTI. 1987. Guidance manual for assessing human health risks from chemically contaminated fish and shellfish. Draft report to U.S. Environmental Protection Agency, Office of Marine and Estuarine Protection, Washington, DC. PTI, Inc., Bellevue, WA. 76 pp. + appendices. Stara, J.F., B.E., R.C. Hertzberg, R.J.F. Bruins, M.L. Dourson, P.R. Durkin, L.S. Erdreich, and W.E. Pepelko. 1983. Approaches to risk assessment of chemical mixtures. Report presented at the Second International Conference on Safety Evaluation and Regulation, Cambridge, MA. 23 pp. Tetra Tech. 1986a. Guidance manual for health risk assessment of chemically contaminated seafood. Prepared for U.S. Environmental Protection Agency. Region X, Office of Puget Sound, Seattle, WA. Tetra Tech, Inc., Bellevue, WA. 69 pp. + appendices. Tetra Tech. 1986b. Quality assurance/quality control (QA/QC) for 301(h) monitoring programs: guidance on field and laboratory methods. Prepared for the U.S. Environmental Protection Agency, Office of Marine and Estuarine eee Marine Operations Division, Washington, DC. Tetra Tech, Inc., Bellevue, WA. Tetra Tech. 1986c. Recommended protocols for measuring selected environ- mental variables in Puget Sound. Prepared for the Puget Sound Estuary Program, U.S. Environmental Protection Agency, Region X, Seattle, Washington. Tetra Tech, Inc., Bellevue, WA. Tetra Tech. 1987. Evaluation of survey positioning methods for nearshore and estuarine waters. EPA-430/9-86-003. Final report prepared for Marine Operations Division, Office of Marine and Estuarine Protection, U.S. Environmental Protection Agency. Tetra Tech, Inc., Bellevue, WA. 54 pp. + appendices. Tetra Tech. 1988. Evaluation of differential LORAN-C for positioning in nearshore marine and estuarine waters. Draft report prepared for Marine Operations Division, Office of Marine and Estuarine Protection, U.S. Environmental Protection Agency. EPA Contract No. 68-C8-0001. Tetra Tech, Inc., Bellevue, WA. 54 Travis, C.C., and H.A. Haltemer-Frey. 1988. Determining an acceptable level of risk. Environ. Sci. Technol. 22:873-876. Travis, C.C., S.A. Richter, E.A.C. Crouch, R. Wilson, E.D. Klema. 1987. Cancer risk management: a review of 132 federal regulatory decisions. Environ. Sci. Technol. 21:415-420. U.S. Environmental Protection Agency. 1980. Water quality criteria documents; availability. U.S. EPA, Washington, DC. Federal Register Vol. 45, No. 231, Part V. pp. 79318-79379. U.S. Environmental Protection Agency. 1985. National primary drinking water regulations; synthetic organic chemicals, inorganic chemicals and microorganisms; proposed rule. U.S. Environmental Protection Agency, Washington, DC. Federal Register, Vol. 50, No. 219, pp. 46936-47022. U.S. Environmental Protection Agency. 1986a. Guidelines for carcinogen risk assessment. Federal Register Vol. 51, No. 185, pp. 33992-34003. U.S. Environmental Protection Agency. 1986b. Guidelines for exposure assessment. Federal Register Vol. 51, No. 185, pp. 34042-34054. U.S. Environmental Protection Agency. 1986c. Guidelines for the health risk assessment of chemical mixtures. Federal Register Vol. 51, No. 185, pp. 34014-34025. U.S. Environmental Protection Agency. 1986d. Superfund public health evaluation manual. EPA 540/1-86/060. U.S. EPA, Office of Emergency and Remedial Response, Washington, DC. U.S. Environmental Protection Agency. 1987a. Health advisories for Legionella and seven inorganics. National Technical Information Service PB87-235586. U.S. EPA, Office of Drinking Water, Washington, DC. 126 pp. U.S. Environmental Protection Agency. 1988. Superfund exposure assessment Manual. EPA/540/1-88/001. U.S. EPA, Office of Remedial Response, Washing- ton, DC. 157 pp. Vettorazzi, G. 1976. Safety factors and their application in the toxi- cological evaluation. pp. 207-223. In: The Evaluation of Toxicological Data for the Protection of Public Health. Pergamon Press, Oxford, England. Vettorazzi, G. 1980. Handbook of international food regulatory toxicology. Vol. I: Evaluations. Spectrum Publications, New York, NY. pp. 66-88. 55