Statistical techniques for characterizing residual waste in single-shell and double-shell tanks

Abstract

A primary objective of the Hanford Tank Initiative (HTI) project is to develop methods to estimate the inventory of residual waste in single-shell and double-shell tanks. A second objective is to develop methods to determine the boundaries of waste that may be in the waste plume in the vadose zone. This document presents statistical sampling plans that can be used to estimate the inventory of analytes within the residual waste within a tank. Sampling plans for estimating the inventory of analytes within the waste plume in the vadose zone are also presented. Inventory estimates can be used to classify the residual waste with respect to chemical and radiological hazards. Based on these estimates, it will be possible to make decisions regarding the final disposition of the residual waste. Four sampling plans for the residual waste in a tank are presented. The first plan is based on the assumption that, based on some physical characteristic, the residual waste can be divided into disjoint strata, and waste samples obtained from randomly selected locations within each stratum. The second plan is that waste samples are obtained from randomly selected locations within the waste. The third and fourth plans are similar to the first two, except that composite samples are formed from multiple samples. Common to the four plans is that, in the laboratory, replicate analytical measurements are obtained from homogenized waste samples. The statistical sampling plans for the residual waste are similar to the statistical sampling plans developed for the tank waste characterization program. In that program, the statistical sampling plans required multiple core samples of waste, and replicate analytical measurements from homogenized core segments. A statistical analysis of the analytical data, obtained from use of the statistical sampling plans developed for the characterization program or from the HTI project, provide estimates of mean analyte concentrations and confidence intervals on the mean. In addition, the statistical analysis provides estimates of spatial and measurement variabilities. The magnitude of these sources of variability are used to determine how well the inventory of the analytes in the waste have been estimated. This document provides statistical sampling plans that can be used to estimate the inventory of the analytes in the residual waste in single-shell and double-shell tanks and in the waste plume in the vadose zone.