Abstract
This study introduces a novel multi-generation system integrating a combined cycle power plant with a trireforming methanol production unit, uniquely utilizing on-site freshwater and hydrogen to enhance resource efficiency. A comprehensive Artificial Intelligence (AI) framework compared Genetic Programming (GP), Deep Neural Networks (DNN), and XGBoost for process modeling, with XGBoost demonstrating superior predictive accuracy (R~2 > 0.975 for critical outputs). Furthermore, Multi-objective Salp Swarm Algorithm (MSSA) and Nondominated Sorting Genetic Algorithm Ⅲ (NSGA-Ⅲ) were employed to address conflicting thermodynamic, economic, and environmental objectives. The optimized configuration yielded substantial improvements over the base case, achieving up to a 54.50% increase in methanol production, a 32.60% reduction in payback period, and a 38.90% decrease in overall environmental impact. This work bridges the gap between power generation, carbon capture, and chemical production, offering a robust, AI-driven framework for sustainable industrial processes.
NETL
NSTL
Elsevier