Environmental Radiological Hazard Mitigation and Remediation

To establish a maximum project extent that could win approval by public stakeholders and regulators, this project employed one of the foremost authorities on kriging^* to develop the probability (to a 90 percent confidence factor) that an unsampled area would be at or greater than the standard 50 picocuries (pCi) per gram action level. For evaluating required remediation of contaminated surface soil (depth of surface soil to be removed), a depth to clean soil was assigned to each grid cell based on its percent of confidence, and a total cubic yard (CY) calculation was performed.

For verification and secondary CY calculations, a simple method was employed to reach the goal, which was to merely compare the locations of “hot” samples to the excavation plan. Because excavators and track hoes need spaces larger than a 90-foot linear grid cell in which to operate, and because it was not being used to evaluate needed remedial actions, this method was deemed operable. Toward this end, inverse distance weighting was chosen as the best method to perform a weighted interpolation on a point dataset. Finally, confirmation sampling was used to verify that soil readings were below the 50 pCi/gram action level.

*Kriging and confidence grids generated by Jeffrey C. Meyers, author of Geostatistical Error
   Management: Quantifying Uncertainty for Environmental Sampling and Mapping.