Experimental Evaluation and Comparative Analysis of Settlement Models for Predicting Long-Term Behavior of MSW Landfills

In this study, the experimental settlement data monitored from two sets of laboratory-simulated anaerobic bioreactors (AN-1 and AN-2) were used to analyze the time-dependent stabilization behavior of municipal solid waste (MSW). The composition of MSW in landfills often includes hazardous fractions, influencing its settlement characteristics. The experimental biogas data were fitted with different kinetic models, and the modified Gompertz and Logistic models demonstrated the best fit, with a coefficient of determination of R2 ≥ 0.996 and a root mean square error (RMSE) of 6.69–13.61 for both reactors. A comparative study of various settlement models with experimental data was carried out to assess the model's reliability using statistical performance indicators. Significant disparities were observed in predicted settlements, with RMSE values ranging from 6.81 to 14.56 for soil mechanics and biodegradation-incorporated models. These disparities arise from an overestimation of settlement attributed to mechanical compression, particularly on the initial day, which fails to accurately represent the actual settlement behavior. The statistical analysis revealed considerable variability in settlement predictions among published models, with the R2 and RMSE ranging 0.76–0.99 and 2.08–26.27. An independent t-test was performed to statistically validate the differences in settlement predictions, confirming significant variations in model accuracy. It was also observed that predicted data from various settlement models and experimental observations demonstrated that MSW exhibits highly nonlinear compression behavior. To address this limitation, an improved empirical model based on a first-order decay rate was proposed, effectively capturing the nonlinear, time-dependent settlement behavior of MSW. The hazardous waste components, such as contaminated sludge and industrial byproducts, undergo similar decomposition and settlement mechanisms. Therefore, the proposed model can potentially be extended to predict the settlement behavior of landfills containing hazardous waste. This study contributes to optimizing landfill design and postclosure management by enhancing settlement prediction accuracy, which is crucial for the safe containment of municipal and hazardous waste.