Sensitivity of Remediation Goals to Risk Assessment Uncertainty and Implications for Alternatives Analysis

摘要

Background/Objectives. This presentation is about fully utilizing and optimizing what can be learned from the site characterization to inform remedy selection. Along the way it will shed light on the role of risk management in the remedy selection process and address the issue of defining remedy effectiveness or success metrics. The linchpin that ties together the topics of site characterization, remedy selection, risk management and remedy effectiveness is sensitivity and uncertainty analyses. Uncertainties are rarely, if ever, eliminated. However, a careful analysis of uncertainties can provide risk managers with a higher degree of confidence in making protective cleanup decisions. The same is true when a clear understanding of the sensitivity of remedial goals (RGs) is appropriately incorporated into cleanup decisions. Approach/Activities. The baseline risk assessments for Portland Harbor identified PCBs as an important contaminant to consider in developing RGs and sediment management areas (SMAs) in the draft Feasibility Study (FS). Risk management recommendations identified the mink population as an important assessment endpoint for developing RGs and SMAs, and for analyzing alternatives vis-à-vis the NCP protectiveness criterion. The assumptions used in the baseline ecological risk assessment (BERA) and in a site-specific bioaccumulation model yielded a mink-PCB preliminary remediation goal (PRG) in sediment of 31 μg/kg. A preliminary uncertainty analysis presented in the BERA indicated that mink risks probably were overestimated. Therefore, a thorough uncertainty analysis was conducted at the start of the draft FS to quantify the sensitivity of the mink- PCB RG to BERA uncertainties and assumptions. An overview of that uncertainty analysis will be presented, along with key findings. Exposure, dose-response, and organism- to-population extrapolation uncertainties were all considered. The RG was found to be sensitive to these uncertainties. The best estimate of the mink-PCB RG was 256 μg/kg, with a range of 36 to 1,192. Thus, the sensitivity analysis demonstrated that PCB RGs considerably higher than the 31 μg/kg PRG are likely to satisfy the protectiveness criterion. Results/Lessons Learned. Understanding the sensitivity of RGs to uncertainties and assumptions about baseline risks should not hinder cleanup decision-making, or drive risk managers to conservative decisions. Rather, the sensitivity analysis results should make it clear to risk managers that there is a wide range of RGs that could be found to satisfy the protectiveness criterion. From there it becomes easy to understand why all of the NCP criteria are critically important in setting RGs, analyzing FS alternatives, and selecting remedies. Understanding that protectiveness is properly viewed as a range rather than as a single point highlights the importance of the balancing criteria. For example, implementability and cost criteria both become less favorable as the probability of protectiveness