Multi-Objective Capacity Configuration Optimization for Zero-Carbon Industrial Parks with CSP-Based Green Power Direct Connection

Zero-carbon industrial parks (ZCIPs) must balance renewable supply, carbon emissions, and multi-energy reliability. This study examines capacity planning for a ZCIP connected to an external concentrated solar power (CSP) plant through a green power direct-connection contract. The contract gives the park bounded scheduling rights over its offtake and allows thermal energy storage (TES) charging or discharging within an agreed operating envelope; the CSP operator retains physical control of the plant. A tri-objective model minimizes annual cost, net accounted carbon emissions, and a representative-day supply adequacy deficit for electricity, heat, and cooling. Four seasonal typical days are used, and the problem is solved by an improved MOPSO. Under equal evaluation budgets, the improved algorithm achieves the highest mean HV and lowest mean IGD. CSP direct connection raises the hypervolume from 5.871 × 109 to 7.838 × 109; allowing electricity sales raises it from 4.563 × 109 to 8.508 × 109. Two stress cases, the maximum electricity-load day and maximum residual-load day, are evaluated using the fixed compromise configuration under a normal grid-connected boundary. The results support ZCIP planning and green power direct-connection policy design.