SITE INFORMATION AND CORRESPONDENCE

Home Browse Search

David Stavarek, R.G. June 29, 1998 Page 7 hg�"separated" by current mathematical techniques to produce specific output PDFs of contaminant concentrations. As a consequence, I cannot reject the proposed method of using discrete adjustment factor values for estimating cleanup levels from a PDF of overall risk. The proposed approach is similar to the one used traditionally for calculating cleanup levels. However, I would like to recommend that if clean up levels are necessary to be estimated,the authors should show a verification of the procedure in the risk assessment report. Specifically,I am recommending to re-insert the estimated cleanup values into a new spreadsheet with the exposure model, run the Monte Carlo simulation analysis, and produce a new output PDF of risk. If the area under the curve of JUi latest outvut PDF does not exceed more than 5% of the target risk level (whether a 1x10 6 or 1x10 ), then the proposed approach may be considered acceptable to OEHHA. Specific Comments Regarding the Probabilistic Approach to Risk Assessment 1. Guidance documents. At present,the State of California does not have official documents describing the principles and guidelines for the preparation of probabilistic health risk assessments. It is suggested that the following documents be used as reference: • USEPA(1997). Guiding Principles for Monte Carlo Analysis. Risk Assessment Forum, Washington, DC. EPA/630/R-97/001. • USEPA(1997). Policy for Use of Probabilistic Analysis in Risk Assessment at the U.S. Environmental Protection Agency. Office of Research and Development. May 15, 1997. • USEPA Region 8 (1995). Use of Monte Carlo Simulation in Performing Risk Assessments. Region 8 Superfund Technical Guidance. Denver, CO. • USEPA Region 3 (1994). Use of Monte Carlo Simulation in Risk Assessments. Region III Technical Guidance Manual Risk Assessment. Hazardous Waste Management Division, Office of Superfund Programs. Philadelphia, PA. EPA903-F-94-001. 2. Dealing with correlations in algorithms for exposure dose. The Workplan does not mention how correlations or covariance between exposure factors will be dealt with in the final risk assessment. This is a topic that is often criticized in the Monte Carlo analysis,but for which there is no consensus on a procedure to be used. A procedure that I have used and which I have formally proposed to the ASTM for inclusion in the probabilistic risk assessment guidelines, is shown below. The method is based on the strong association of exposure factors (BR BW EF ED) with age. The exposure dose