Comparative reliability evaluation of lateritic soils as hydraulic barriers

Abstract

A numerical investigation to determine the existence or otherwise of congruence between laboratory-based and field-based models for predicting hydraulic conductivity of compacted clay liners has been conducted. The comparisons here are based on values of the reliability index estimated with the use of the two models for cases when hydraulic conductivity is assumed to be normally and lognormally distributed. The laboratory reduced Proctor and modified Proctor compactive efforts were assumed to correspond to the lowest and highest compactor weights (165 kN and 338 kN for sheepsfoot rollers only from literature) respectively. The standard Proctor compactive effort was assumed to correspond to the mean compactor weight of 261.4 kN. Reliability estimates from the laboratory-based model were consistently higher than the corresponding values from the field-based model for each of the variables common to the two models. As laboratory samples usually involve small samples which are often unrepresentative of field situations, it is considered that reliability levels estimated with the use of the field-based model provide more realistic assessment of compacted lateritic soils as hydraulic barriers. Compactor weight of 300 kN and above may be needed for effective compaction of lateritic soil liners. Reliability-based designs of lateritic soil liners should incorporate appropriate probability distribution types for variables in the field-based model.