Hydrocarbon-Resolved Methane Prediction from Diluent Biodegradation in Oil-Sands Tailings

Methane generation from anaerobic biodegradation of fugitive diluent hydrocarbons is an important source of greenhouse gas emissions from oil-sands tailings, yet predictive tools that preserve hydrocarbon-level information remain limited. This study develops a hydrocarbon-resolved methane-prediction model and tests it on a case study involving a twelve-component diluent mixture containing BTEX, normal alkanes, and iso-alkanes. The model integrates stoichiometric methane yields, compound-specific lag times, Monod-type hydrocarbon consumption, logistic activation, and a single methane-conversion factor to simulate cumulative methane production and group-level methane contributions through time. Model performance is evaluated against measured methane and residual hydrocarbon data using normalized mean square error. The model reproduces cumulative methane with improved normalized mean square error relative to the existing stoichiometric benchmarks, while group-resolved outputs and robustness analyses show that predictive performance is governed primarily by conversion efficiency and lag structure. On the other hand, inclusion of an unresolved biodegradable-substrate fraction did not strengthen model agreement. These results indicate that the modeled hydrocarbon set captures the principal methane-generating substrate pool and that the proposed framework provides an accurate and mechanistically interpretable basis for methane prediction in oil-sands tailings.