Evaluation of Parameterization Techniques for Unsaturated Hydraulic Conductivity Functions for Municipal Solid Waste

A study was conducted to evaluate methods for parameterizing constitutive functions describing the unsaturated hydraulic conductivity (Kθ) and water retention curve (WRC) for municipal solid waste (MSW). Water retention curves and Kθ were measured using a hanging column apparatus and the multistep outflow (MSO) method on MSW specimens compacted at three different dry densities. Parameters for constitutive functions describing the WRC and Kθ were determined via nonlinear regression on the WRC data, nonlinear regression on the WRC and Kθ data, and numerical inversion conditioning on the water content and transient outflow data. Nonlinear regression of the measured WRC and Kθ and numerical inversion reproduced the measured Kθ with the least error. Nonlinear regression on only the WRC data resulted in the least error between the predicted and measured WRC but the greatest errors between the predicted and measured Kθ. Flow simulations using parameters obtained with any of the three parameter-estimation methods typically resulted in errors in predicted flow corresponding to errors in the specimen water content ranging between 1 % and 5 %, with some simulations returning errors of as much as 10 %. Inclusion of transient flow data (through inversion) or Kθ data results in improved parameterization for constitutive functions for Kθ. However, reasonable reproduction of flow data from the MSO tests was achieved using parameters obtained from regression on only WRC data.