Multi-objective optimization of energy efficiency, carbon efficiency and economic benefits under the multi-energy synergy of electrolytic aluminum production process

Aluminum electrolysis is a highly energy-consuming and emission-intensive industrial process, and achieving simultaneous improvements in energy efficiency, carbon efficiency, and economic performance has become a critical challenge for modern smelters facing increasingly stringent decarbonization and cost-competitiveness requirements. However, most existing studies still emphasize single-objective optimization, while comprehensive three-objective research that simultaneously optimizes energy, carbon emissions, and economic goals to cope with high energy consumption, high carbon emissions, and limited economic benefits is still lacking. This study develops a multi-objective optimization framework for a 300 kA industrial aluminum electrolysis cell based on 36 months of plant operating data. Firstly, an AC/DC-coupled multi-energy model is established to explicitly represent grid electricity import, on-site PV self-consumption, natural gas use, and major auxiliary demands, together with key process-consistency constraints. Secondly, NSGA-III is employed to generate a well-distributed Pareto set, revealing the trade-offs among specific energy consumption (SEC), carbon intensity (CI), and unit production cost (ECI). Additionally, combined with entropy-weighted TOPSIS and robustness screening, a variety of implementable schemes are identified to achieve factory-oriented selection for different operating preferences. Based on the actual situation of the research object, this study selected a stable and feasible optimization scheme. Under the premise of not violating the process constraints, compared with the benchmark scheme, SEC decreased by 7.7%, CI decreased by 15.8%, and ECI decreased by 4.0%. This framework provides transferable decision support and flow-based diagnosis for the coordinated operation of aluminum electrolysis systems in modern smelters.