SITE INFORMATION AND CORRESPONDENCE

Home Browse Search

Philip S. Isorena • April 20, 2000 Page 3 Soil concentration Depth to top of Soil Incremental cancer of benzene contamination risk 120 µg/kg 8.3 ft(100 in) Sandy clay loam 7.6 x 10"° 120 µg/kg 10 in Sandy clay loam 1.6 x 10_' 120 µg/kg 8.3 ft (100 in) Clay 2.8 x 101 120 µg/kg 10 in Clay 1200 µg/kg 8.3 ft Sandy clay loam 7.6 x 10"5 1200 µg/kg 10 in Sandy clay loam 1.6 x 10 1200 µg/kg 8.3 ft (100 in) Clay 1200 µg/kg 10 in Clay 3.5 x 10 Consequently,the lack of sufficient environmental samples in this document provides no confidence in the accuracy of the excess lifetime cancer risk from inhalation of VOCs in indoor air from soil emissions for an adult RME exposure (p.23 and p 27) calculated in the document as 6.8 x 10"$ for GP-3 Area and 5.3 x 10"$ for the Pipeline Area. In addition, U.S. EPA Region 9 and OEHHA have concluded that the Johnson& Ettinger model more accurately estimates the risk. Since volatilization to indoor air is the primary exposure pathway from the type and number of specific analyses that were performed, little faith can be placed in the final cumulative risk estimations by all four exposure pathways. The BTEX concentrations in Soil in Appendix C, Table C-2,provided data for calculating a 95% upper confidence limit for benzene concentrations in the Pipeline Area. However, the calculation of a 95% UCL apparently used the t-value for p=0.1 instead of p=0.05. Furthermore, the algorithm for calculating the 95% UCL was the one for normally distributed data with a large number of samples, rather than lognormally distributed data with a small number of samples. Errors in calculation of the 95% UCL will be transferred through each pathway to calculate the final risk. The authors are referred to Supplemental Guidance to RAGS: Calculating the Concentration Term, US EPA Publication 9285.7—081, May, 1992, and Statistical Methods for Environmental Pollution Monitoring, Richard O. Gilbert, 1987. Additional soil samples in the Pipeline Area and GP-3 area were analyzed for PAHs, and the data is combined in Table C-3. However, this reviewer cannot use this information to arrive at the data presented in Appendix G, as Representative Concentrations—GP-3 Area and Representative Concentrations—Pipeline Area. Most of the data were nondetect, but there were several detection limits reported, and those for G136-15.5 and GB7-4 were higher than detected concentrations in other borings. Consequently, the risk was estimated based on one or two samples with detectable concentrations and a number of samples with nondetects. Since the detections were an order or two of magnitude greater than most of the nondetects, there may be hotspots which were not investigated.