The quantitative assessment of uncertainties in the exposure parameters for the individual exposure pathways provides considerable information about the variability and sensitivity of the calculated results. These results are important because point estimates of these parameters are used to determine the extent of remediation necessary through the Superfund process. The point estimates that are provided as guidance by the U.S. Environmental Protection Agency (EPA) are often conservative and can result in an overestimate of the potential risk. The scope of this work is to perform an uncertainty analysis on the standard Superfund residential scenario risk equations using available statistical information for the uncertain exposure parameters. The results are used to quantify the degree to which the standard default values overestimate the predicted percentiles of exposure (90–95th) that they are intended to estimate and to determine which parameters are responsible for the majority of the variation.
Residential exposure pathways associated with contaminated soil and groundwater are evaluated in this report and include the ingestion, inhalation, and dermal contact pathways. The external exposure pathway from exposure to soil contaminated with radionuclides is also evaluated. Exposure calculations are performed using models present in the EPA's Risk Assessment Guidance for Superfund. In this document, these models are not used to calculate actual risk values. Instead, the exposure parameters generally held to be constant for all sites are used to evaluate the variability in the predictions of the individual models. Uncertainties in the risk estimates that result from site- or contaminant-specific variability are not assessed for this effort. For the exposure parameters, a sensitivity analysis is performed to determine the most sensitive parameters in each model.
The outcome of this work can be used to focus the attention of the risk assessor on the expected variability, range of variation, and therefore the reliability of the point estimates that are used, and on the parameters that caused the variation. In addition, a relative ratio between the point estimates (PE) as set forth in EPA guidance and the predicted percentile risk results is presented for each pathway analyzed. The predicted percentiles of the exposure parameters are referred to as multiplicative exposure factors (MEF). These PE/MEF ratios are used to quantify the degree of conservatism present in the default exposure parameters that are recommended by the EPA for Superfund sites for each residential pathway analyzed. This general method of determining a multiplicative exposure factor that is constant for all sites can be used to derive values for the exposure term that are more representative of the EPA’s stated risk management goal of protecting 90 to 95% of the potentially exposed population for a given land use scenario. Alternatively, these ratios can be used as part of the uncertainty assessment in a baseline risk assessment to estimate the degree of conservatism present in the exposure parameters of each pathway.
The probabilistic distributions of the uncertain exposure parameters used in these models are collected from the most recent sources in the literature. The parameters for which distributions are assigned include exposure frequency, exposure duration, body weight, surface area to body weight, ingestion rates for water and soil, inhalation rate, exposure fraction, adherence of soil-to-skin factor, and the gamma shielding factor. Professional judgment is used to supplement the data and provide a distribution for the few parameters for which no consensus distribution is available.
2. RESIDENTIAL LAND USE RISK EQUATIONS
This section provides the equations and recommended point estimates provided in various EPA sources for the residential pathway. Under residential land use, residents are expected to be in frequent, repeated contact with contaminated media. For carcinogens, the exposure assumptions account for daily exposure over long term and generally result in high potential exposures and risk. For noncarcinogens, the exposure assumptions do not account for accumulations over the life time of the exposed receptor. Risk from groundwater contaminants is assumed to be primarily from direct ingestion, inhalation of volatiles from household water use, and dermal contact while showering. Risk from soil is assumed to be from direct ingestion, inhalation of dust and particulates, dermal exposure from chemicals, and external exposure from radionuclides.