Scaling up laboratory column testing results to predict coupled methane generation and biological settlement in full-scale municipal solid waste landfills

Prediction of methane (CH₄) generation and settlement of biodegrading municipal solid waste (MSW) is of primary interest to landfills aiming at biogas recovery for energy generation and MSW stabilization. We investigate these two concurring processes using datasets from 35 laboratory column tests and 8 pilot- and full-scale landfill cells available in the literature. We fit the datasets using three CH₄ generation models, i.e., conventional first-order decay (FOD) model, coupled FOD model, and coupled Gompertz model. The latter two models are proposed in this study which couple CH₄ generation with biological settlement strain (ε_B) instead of elapsed time. Each model requires only four to five input parameters which can be reasonably estimated a priori based on the initial conditions of the MSW and landfills. The performances of the models are compared using jackknife resampling approach and normalized root-mean-square error (NRMSE) values. The coupled Gompertz model results in on average 50% lower NRMSE when predicting the time-dependent CH₄ generation in all the datasets compared to the other two models. Thus, we demonstrate that CH₄ generation from biodegrading MSW in landfills can be better predicted using the corresponding ε_B than the elapsed time.