«This page left intentionally blank. United States Environmental Protection Agency EPA-540-R-05-012 Office of Solid Waste and Emergency Response OSWER ...»
2-2 Chapter 2: Remedial Investigation Considerations 3. Identify inputs to the decision. Examples: What are the appropriate fish species, receptor groups, and consumption rates to evaluate? What existing data are available and what must be collected? What is the toxicity of the contaminants to all receptor groups?
5. Develop a decision rule. Example: If exposure at the upper 95 percent confidence limit for fish consumption of the recreational fisher population to the mean contaminant concentration of any one of the three most popular fish species exceeds a cancer risk range of 10-6 to 10-4 or a Hazard Index of 1, risk will be considered unacceptable.
6. Specify limits on decision errors. Example: What levels of uncertainty are acceptable for this decision, considering both false positive and false negative errors?
Similar hypotheses could be established for evaluating each remedial alternative being considered for the site, and for evaluating the effectiveness of the selected alternative. The way in which the process is followed may vary depending on the decision to be made, from a thought process to a rigorous statistical analysis. Additional guidance provided in EPA Requirements for Quality Assurance Project Plans [(QAPPs), U.S. EPA 2001e) describes how DQOs are incorporated into QAPPs.
2.1.2 Types of Data The types of data the project manager should collect are determined mostly by the following
information needed to:
Highlight 2-1 lists some general types of physical, chemical, and biological data that a project manager should consider collecting when characterizing a sediment site. The project manager should
understand the importance of historical changes in some of these characteristics (e.g., water body bathymetry or contaminant distributions in surface and subsurface sediment, water, and biota). It may also be important to understand how characteristics change seasonally, and under various flow and temperature conditions. The relative importance of these types of data variabilities is dependent on the site. It is frequently important to understand the properties affecting the mixing zone or biologically active zone of sediment. Contaminants in the biologically active layer of the surface sediment at a site often drive exposure, and reduction of surface sediment concentrations may be necessary to achieve risk reduction. While sediment sites typically demand more types of data for effective characterization than other types of sites, the type and quantity of data required should be geared to the complexity of the site and the weight of the decision. In addition, the data acquisition process should not prevent early action to reduce risk when appropriate.
Site characterization should include collection of sufficient baseline data to be used to compare to monitoring data collected during and following implementation of the remedy in a statistically defensible manner. Additional sampling could be needed during remedial design, however, to establish reliable baseline data for the monitoring program. Chapter 8, Remedial Action and Long-Term Monitoring, provides a discussion of effective monitoring programs, much of which is also useful during the remedial investigation.
At this time, polychlorinated biphenyls (PCBs) are among the most common contaminants of concern at contaminated sediment sites. The term “PCB” refers to a group of 209 different chemicals, called PCB congeners, sharing a similar structure. Aroclors are commercial mixtures of PCB congeners and weathering of an Aroclor after release into the environment results in a change in its congener composition (National Research Council, (NRC 2001). EPA’s Office of Water Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories, Volume 1, Fish Sampling and Analysis, Third Edition (U.S. EPA 2000b), notes that individual PCB congeners may be preferentially enhanced in environmental media and in biota.
Characterizing PCB risk on a congener-specific basis allows for an accounting of the differences in physiochemical, biochemical, and toxicological behavior of the different congeners in type and magnitude of effects and, therefore, in risk calculations. Although Aroclor analysis can be useful for initial assessment of PCB concentrations, for risk assessment purposes, NRC recommends that PCB sites be characterized on the basis of specific PCB congeners and the total mixture of congeners found at each site (NRC 2001). EPA currently provides congener-specific analyses through its Non-Routine Program under the Contract Laboratory Program (CLP), but it may, in the future, be available through its CLP routine analytical services. However, to the extent that PCB congener-specific data are determined useful at a site, the project manager should not assume this necessarily needs to be done for all samples collected. At times, only a subset of samples or sampling events may need congener analysis. Deciding how best to characterize a PCB site is a complex issue due in part to issues related to dioxin-like PCBs, the lack of congener-specific toxicological data, the need for comparing present and previously collected data, and the cost of congener-specific analyses. The decision about what method or methods to use for PCB analysis should be made on a site-specific basis.
2-4 Chapter 2: Remedial Investigation Considerations
Currently, metals are also among the most common contaminants of concern at Superfund sediment sites. Concentrations of bulk (total dry weight basis) metals in sediment alone are typically not good measures of metal toxicity. However, in addition to direct measurement of toxicity, EPA has developed a recommended approach for estimating metal toxicity based on the bioavailable metal fraction, which can be measured in pore water and/or predicted based on the relative sediment concentrations of acid volatile sulfide (AVS), simultaneously extracted metals (SEM), and total organic carbon (TOC) (U.S. EPA 2005c). Both AVS and TOC are capable of sequestering and immobilizing a range of metals in sediment.
2.1.3 Background Data
Where site contaminants may also have natural or anthropogenic (man-made) non-site-related sources, it may be important to establish background or reference data for a site. When doing so, project managers should consult EPA’s Role of Background in the CERCLA Cleanup Program (U.S. EPA 2002b), the EPA ECO Update - The Role of Screening-Level Risk Assessments and Refining Contaminants of Concern in Baseline Ecological Risk Assessments (U.S. EPA 2001f), and Guidance for Comparing Background and Chemical Concentrations in Soil for CERCLA Sites (U.S. EPA 2002c).
Although the latter is written specifically for soil, many of the concepts may be applicable to contaminant data for sediment and biota. It should be noted that a comprehensive investigation of all background substances found in the environment usually will not be necessary at CERCLA sites. For example, radon background samples would not be normally collected at a chemically contaminated site unless radon, or its precursor was part of the CERCLA release.
Where applicable, project managers should consider continuing atmospheric and other background contributions to sites to adequately understand contaminant sources and establish realistic risk reduction goals (U.S. EPA 2002b). For baseline risk assessments, EPA recommends an approach that generally includes the evaluation of the contaminants that exceed protective risk-based screening concentrations, including contaminants that may have natural or anthropogenic sources on and around the Superfund site under evaluation. When site-specific information demonstrates that a substance with elevated concentrations above screening levels originated solely from natural causes (i.e., is a naturally occurring substance and not release-related), these contaminant normally does not need to be carried through the quantitative analysis. However, these contaminants should be generally discussed in the risk characterization summary so that the public is aware of its existence. The presence of naturally occurring substances above screening levels may indicate a potential environmental or health risk, and that information should be discussed at least qualitatively in the document. If data are available, the contribution of background to site conditions should be distinguished (U.S. EPA 2002b). This approach is designed to ensure a thorough characterization of risks associated with hazardous substances, pollutants, and contaminants at sites (U.S. EPA 2002b).
For risk management purposes, understanding whether background concentrations are high relative to the concentrations of released hazardous substances, pollutants, and contaminants may help risk managers make decisions concerning appropriate remedial actions (U.S. EPA 2002b). Generally, under CERCLA, cleanup levels are not set at concentrations below natural or anthropogenic background levels (U.S. EPA 1996a, 1997c, 2000c). If a risk-based remediation goal is below background concentrations, the cleanup level for that chemical may be established based on background concentrations.
2-6 Chapter 2: Remedial Investigation Considerations In cases where area-wide contamination may pose risks, but these risks are not appropriate to address under CERCLA, EPA may be able to help identify other programs or regulatory authorities that are able to address the sources of area-wide contamination, particularly anthropogenic sources (U.S. EPA 1996a, 1997c, 2000c). In some cases, as part of a response to address CERCLA releases of hazardous substances, pollutants, and contaminants, EPA may also address some of the background contamination that is present on a site due to area-wide contamination.
2.2 CONCEPTUAL SITE MODELS
A conceptual site model (CSM) generally is a representation of the environmental system and the physical, chemical, and biological processes that determine the transport of contaminants from sources to receptors. For sediment sites, perhaps even more so than for other types of sites, the CSM can be an important element for evaluating risk and risk reduction approaches. The initial CSM typically is a set of hypotheses derived from existing site data and knowledge gained from other sites. Natural resource trustee agencies and other stakeholders may have information about the ecosystem that is important in developing the conceptual site model and it is recommended that they have input at this stage of the site investigation. This initial model can provide the project team with a simple understanding of the site based on available data. Information gaps may be discovered in development of the CSM that support collection of new data.
Essential elements of a CSM generally include information about contaminant sources, transport pathways, exposure pathways, and receptors. Summarizing this information in one place usually helps in testing assumptions and identifying data gaps and areas of critical uncertainty for additional investigation.
The site investigation is, in essence, a group of studies conducted to test the hypotheses forming the conceptual site model and turning qualitative descriptions into quantitative descriptions. The initial conceptual model should be modified to document additional source, pathway, and contaminant information that is collected throughout the site investigation. Project managers should also be aware of the spatial and temporal dimensions to the processes depicted in a CSM. Although these are difficult to represent in static graphical form, it is important to consider the relevance and role of these dimensions when using the CSM and developing hypotheses or inferences from them.
A good CSM can be a valuable tool in evaluating the potential effectiveness of remedial alternatives. As noted in the following section on risk assessment, the CSM should capture in one place the pathways remedial actions are designed to interdict to reduce exposure of human and ecological receptors to contaminants. Typical elements of a CSM for a sediment site are listed in Highlight 2-2.
Project managers may find it useful to develop several conceptual site models that highlight
different aspects of the site. At complex sediment sites, often three conceptual site models are developed:
1) sources, release and media, 2)human health, and 3) ecological receptors. For sites with more than one contaminant that are driving the risks, especially if they behave differently in the environment (e.g., PCBs vs. metals), it is often useful to develop a separate CSM for different contaminants or groups of contaminants. Highlight 2-3, Highlight 2-4, and Highlight 2-5 present examples that focus on ecological and human health threats.
2.3 RISK ASSESSMENT Consistent with the National Oil and Hazardous Substances Pollution Contingency Plan (NCP), a human health risk assessment and an ecological risk assessment should be performed at all contaminated sediment sites. In addition to assessing risks due to contaminated sediment, in many cases, risks from soil, surface water, ground water and air pathways may need to be evaluated as well. One of the outputs from the risk assessment should be an understanding of the relative importance or contribution of the pathways depicted in the conceptual site model to actual risk. This understanding is generally key to making informed decisions about which remedial alternative to implement at a site.
Generally, the human health risk assessment should consider the cancer risks and non-cancer health hazards associated with ingestion of fish and other biota inherent to the site (e.g., shellfish, ducks);
dermal contact with and incidental ingestion of contaminated sediment; inhalation of volatilized contaminants; swimming; and possible ingestion of river water if it is used as a drinking water supply.
Separate analyses should also consider risks from exposure to floodplain soils and may include direct contact, ingestion, and exposures to homegrown crops, beef, and dairy products where appropriate. The relevance and importance of each pathway to actual risks will vary with different contaminants or contaminant classes at a site. In addition, the risk assessment should include an analysis of the risks that may be introduced due to implementation of remedial alternatives (see Section 2.3.3, Risks from Remedial Alternatives). As with all remedial investigation (RI) and feasibility study (FS) data collection efforts, the scope of the assessments should be tailored to the complexity of the site and how much information is needed to reach and support a risk management decision. It is important to involve the risk 2-8 Chapter 2: Remedial Investigation Considerations assessors early in the process to ensure that the information collected is appropriate for use in the risk assessment.
Screening and baseline risk assessments are designed to evaluate the potential threat to human health and the environment in the absence of any remedial action. Generally, they provide the basis for determining whether remedial action is necessary as well as the framework for developing risk-based remediation goals. Risk assessments should also provide information to evaluate risks associated with implementing various remedial alternatives that may be considered for the site. Detailed guidance on performing human health risk assessments is provided in a number of documents, available through EPA’s Superfund Risk Assessment Web site at http://www.epa.gov/oswer/riskassessment/ risk_superfund.htm. The Risk Assessment Guidance for Superfund (U.S. EPA 1989, also referred to as “RAGS”), provides a basic plan for developing human health risk assessments. Specific guidance on the standardized planning, reporting, and review of risk assessments is available at http://www.epa.gov/ oswer/riskassessment/ragsd/index.htm.