Simulating differential settlement of landfill foundations using random fields.

The long-term functionality of composite landfill final cover systems and composite floor liner systems depends on the distortions imposed on these systems by differential settlement. The evaluation of differential settlement is particularly challenging for final cover systems, because of the heterogeneity of waste, and for floor liners over highly variable subgrades (e.g. existing waste, mine spoil, etc.). The present ability to forecast these distortions is limited by a scarcity of long-term settlement measurements at modern landfills and difficulty characterizing waste and subgrade properties. Furthermore, deterministic approaches to landfill settlement prediction are not able to capture the spatial variability in the waste mass and subgrade properties which control this behavior. This paper considers a probabilistic approach to the long-term estimates of differential settlement beneath landfill final cover and floor liner systems using random fields to model the waste and subgrade properties. Finite element analysis of two dimensional random fields is used to calibrate appropriate spatial correlation parameters for use in simpler, one-dimensional analyses to estimate differential settlement.