Application of the SmartSampling Methodology to the Evaluation of Contaminated Landscape Soils at Brookhaven National Laboratory

Portions of the SmartSampling{trademark} analysis methodology have been applied to the evaluation of radioactive contaminated landscape soils at Brookhaven National Laboratory. Specifically, the spatial, volumetric distribution of cesium-137 ({sup 137}Cs) contamination within Area of Concern 16E-1 has been modeled probabilistically using a geostatistical methodology, with the purpose of identifying the likelihood of successfully reducing, with respect to a pre-existing, baseline remediation plan, the volume of soil that must be disposed of offsite during clean-up. The principal objective of the analysis was to evaluate the likelihood of successful deployment of the Segmented Gate System (SGS), a novel remediation approach that emphasizes real-time separation of clean from contaminated materials during remediation operations. One primary requirement for successful application of the segmented gate technology investigated is that a variety of contaminant levels exist at the deployment site, which would enable to the SGS to discriminate material above and below a specified remediation threshold value. The results of this analysis indicate that there is potential for significant volume reduction with respect to the baseline remediation plan at a threshold excavation level of 23 pCi/g {sup 137}Cs. A reduction of approximately 50%, from a baseline volume of approximately 1,064.7 yd{sup 3} to less than 550 yd{sup 3}, is possible with acceptance of only a very small level of engineering risk. The vast majority of this volume reduction is obtained by not excavating almost all of levels 3 and 4 (from 12 to 24 inches in depth), which appear to be virtually uncontaminated, based on the available data. Additional volume reductions related to soil materials on levels 1 (depths of 0--6 inches) and 2 (6--12 inches) may be possible, specifically through use of the SGS technology. Level-by-level evaluation of simulation results suggests that as much as 26 percent of level 1 and as much as 65% of level 2 soils may actually be uncontaminated. Additionally, numerical experiments have been conducted to investigate the effects of selective excavation on the volume and average activity of the remediated materials. These numerical experiments indicate that nonselective excavation may result in mixing of contaminated and uncontaminated materials such that the total volume of material above the threshold excavation level of 23 pCi/g may exceed the baseline volume, thus defeating volume-reduction efforts