Laserfiche WebLink
`tea <br /> r <br /> Sampling <br /> �-- The data collected during 1994 are subject to uncertainty associated with sampling and <br /> analysis. These data are presented in other parts of this health risk assessment. In the <br /> risk assessment, it was assumed that samples collected were representative of conditions <br /> Jo which various populations may be exposed. However, the collected samples may not <br /> be completely representative due to biases in sampling and to random variability of <br /> samples. In general, sampling was biased toward areas of known and suspected elevated <br /> chemical concentrations, which will lead to an overestimation of risk when these results <br /> '~ are assumed to represent a larger area. <br /> The placement of soil borings and monitoring wells was, in part, purposely biased to <br /> detect and characterize potential hot spots of soil and potential groundwater <br /> ;—. contamination based on historical site use. This type of sampling approach is likely to <br /> overestimate the chemical concentrations to which a receptor would be exposed and the <br /> potential health impact to the receptors evaluated. <br /> Analysis <br /> Samples were analyzed using California State Certified Laboratory procedures and were <br /> subjected to limited review, to obtain data suitable for decision-making. However, it <br /> should be understood that sample analysis is subject to uncertainties associated with <br /> precision, accuracy, and detection of chemicals at low concentrations. Analytical <br /> '�- precision and accuracy are evaluated through laboratory QA programs. Uncertainties <br /> associated with precision and accuracy of analysis are generally random errors which may <br /> lead to over- or underestimation of risks. These errors are typically of low magnitude <br /> (well below an order of magnitude) compared to other sources of uncertainty in the risk <br /> assessment. <br /> Due to the limits of analytical methodologies and the complexity of matrixes for <br /> environmental samples, some chemicals present in low concentrations in samples may <br /> not be detected, leading to a possible underestimation of risk. This potential source of <br /> uncertainty is addressed during data evaluation, as described below. <br /> S:\LDC\YELLO.RPT May 4,1995 7-2 <br /> +.r <br />