Probabilistic Design of Gas Collection Systems for a Prototype Bioreactor

Abstract

Prototype bioreactor studies are becoming an increasingly common approach to assessing municipal solid waste degradation in landfills, owing to their greater potential for monitoring and control. Though reports of several bioreactor studies have appeared in the literature, the performance of a laboratory-scale bioreactor in the presence of a gas collection system (GCS) has not been investigated anywhere. Because the properties of waste in a landfill or a bioreactor can vary temporally and spatially, it is essential to account for these variabilities while designing a GCS. In this study, a GCS for a prototype bioreactor was designed using kriging surrogate models to account for the variability in the waste properties. The numerical models required to calibrate the kriging models were implemented through TOUGH3 EOS7CA (three-dimensional simulations). The radius of influence (ROI) of a gas extraction well (a critical parameter that controls the design of a GCS) was estimated by selecting methane generation rate (MGR), suction pressure (Sp), absolute permeability (k), and depth of the extraction well (D) as the input parameters. The ROI and gas pressure distributions inside the bioreactor were determined accounting for the variabilities in the input parameters through Monte Carlo simulations on the kriging model. Gas pressures inside the prototype bioreactor with a single gas extraction well (GEW) system indicated that the bioreactor was unsafe—values far higher than atmospheric pressures were observed. A sensitivity analysis was conducted to achieve efficiency and optimize the design, revealing that Sp and MGR were the critical parameters that controlled the ROI. Hence, the GCS was designed with two GEWs, and the safety of the bioreactor was assessed by varying only the critical parameters with another developed kriging model. Observance of gas pressures in the range of atmospheric pressure indicated that the design was safe. A relation was developed between Sp and MGR, which will help enable a 90% methane recovery at the GEWs. The probability of failure of the prototype bioreactor in the presence of the developed relation was also estimated. The low probability of failure (1.23×10−4) indicated that the design was safe.