From Data to Decision: A Closed-Loop Framework for Co-Optimizing Energy Recovery and SO2 Control in Sludge Gasification

The rapid expansion of urban wastewater treatment has rendered the sustainable management of sludge a global environmental challenge. Gasification technology is emerging as a promising treatment pathway due to its high conversion efficiency and resource recovery potential. However, this process oxidizes the sulfur in the sludge into SO2, introducing an inherent conflict between its energy-saving and emission-reduction objectives. To resolve this dilemma, this study proposes a three-tier data-model-decision framework. First, by integrating material and energy flow analysis (MEFA) with reaction mechanisms, multisource process signals are mapped into the energy recovery efficiency and SO2 emission intensity per unit of sludge. Second, we develop a mechanism-informed feature engineering method and comprehensively compare multiple algorithms to select the optimal surrogate model for process optimization. Finally, jointly considering operational importance and feasibility, this process is systematically optimized under varying emission scenarios using DE, PSO, and NSGA-II algorithms to achieve coordinated emission reduction and energy saving. The results demonstrate that, despite the intrinsic tension between the two objectives, the proposed method can improve energy recovery efficiency by 45.28%-65.34% without additional treatment costs. In particular, the mechanism-informed feature engineering substantially enhances model performance. By bridging data science and industrial ecology, this work systematically identifies the operational bottlenecks and optimization potential of the sludge gasification process, providing actionable decision support for different environmental goals.