Executive SummaryThe Remedial Action Sampling and Analysis Plan (RASAP) for the Hanford Site's 100 Area soils requires sampling and analyzing the soil material for contaminants of concern (COC). However, the RASAP does not include a protocol for analyzing only a fine-grained fraction (sand, silt, and clay) and then relating the measured COC activities and concentrations in this fraction to those in the predominantly coarse-grained (gravel) whole soils in the 100 Area. Selectively sampling only the finer fractions of a coarse-grained soil would introduce a systematic bias that results in consistently overestimating the actual activities and concentrations of COC in whole soil. Such sampling bias, if uncorrected, would result in over-remediation (remediating below the remedial action goal [RAG] levels) of these soils. Therefore, we conducted a study to assess the scope of this problem by analyzing available particle size-COC activity and concentration data for 100 and 300 Area soils, and determined the effects of sampling bias. Our recommendations include sampling protocols and methods to determine the COC activities and concentrations of the whole soil based on the measurements conducted on a finer size fraction of the whole soil.The important findings of our investigations are that 1. In gravelly soils (100 and 300 Areas), the activity and concentration of any COC per unit mass in any finer fractional soil material was consistently biased higher than those of the same COC in the whole soil (all grain sizes sampled). The degree of sampling bias for each COC could be quantified as a sampling bias factor (SBF) that relates the activity and concentration in any finer size fraction of a soil to that in the whole soil.2 The SBF for any COC in a soil increases with decreasing particle size sampled. SBFs for different COC in a soil can vary significantly (up to an order of magnitude) depending on their chemical properties and affinity for specific mineral surfaces.3. SBF for a single COC in soil samples obtained from different locations from the same waste site can be significantly different (two to three times higher than the lowest value), reflecting spatial variability engendered by the source, the type, and the distribution of contamination.4. The SBF for a COC in soils from different waste sites were different, and these values were affected by the soil texture, the degree, the source, and the distribution of contamination.5. Waste site soils would be over-remediated (remediated below remedial action goal values) without SBF corrections. SBF corrections could be made by using corrected remedial action levels (CRAL) computed from RAG and SBF (for selected particle-size finer) for each COC.6. The soils in the vicinity of the 116-C-1 waste trench would be over-remediated (excavation of ~57% to 280% of excess soil volume) without appropriate SBF corrections for contaminant radionuclides 60 Co and 152 Eu.iv Based on our analyses and conclusions, we recommend that 1. A sampling grid should be established (using a geostat...