Multi-criteria kinetic modeling of kerogen hydrothermal conversion: assessing network complexity and model selection

The kerogen-rich oil shale deposits constitute a strategically vital unconventional resource base. However, suitable kinetic studies that represent the reactions for developing appropriate technologies are scarce. This study systematically evaluates the predictive accuracy of kinetic models for kerogen conversion under hydrothermal conditions through the interplay among reaction network complexity, objective function selection, and stoichiometric constraints. These simulation choices, while often treated independently, have a cumulative impact on model fidelity. By benchmarking a three-pathway parallel reaction scheme against a single global reaction model across both catalytic (nickel tallate) and thermal systems, the investigation reveals that considering the complexity of kerogen reduces prediction errors by 16–45%. It also highlights a key choice between two optimization methods: average absolute error focuses on overall accuracy, while the sum of squared errors emphasizes precision for the most significant components. Lastly, the study emphasizes the importance of adhering to stoichiometric constraints to ensure physical consistency and prevent overfitting. Concurrently, nickel tallate catalysis nearly doubled kerogen conversion while selectively channeling products toward light oils, underscoring the synergistic potential of coupling optimized catalytic strategies with rigorously formulated kinetic models.