AI-OPTIMIZED SOLAR ORGANIC RANKINE CYCLE WITH THERMAL ENERGY STORAGE FOR DISPATCHABLE POWER GENERATION IN PAKISTAN

Abstract

The intermittency of solar energy remains a major challenge for achieving reliable and dispatchable renewable electricity generation in Pakistan. This study developed an integrated solar Organic Rankine Cycle (ORC) system coupled with thermal energy storage (TES) and AI-based operational optimization to improve energy conversion, storage utilization, and dispatchability. A quantitative simulation-based framework was developed using 8,760 hourly operating observations representing Pakistan-relevant solar conditions. Three configurations were comparatively evaluated: conventional solar-ORC, solar-ORC with TES, and AI-optimized solar-ORC with TES. The  simulation results indicated that annual electricity generation increased from 421 MWh for the conventional configuration to 548 MWh with TES and 612 MWh under AI optimization. Mean thermal efficiency increased from 14.8% to 19.7%, while dispatchable operating hours increased from 4,210 to 6,120 hours annually. TES utilization improved from 71.4% to 84.7% under AI-based control, while LCOE decreased from USD 0.142/kWh to USD 0.109/kWh. The AI model achieved a testing R² of 0.948, indicating strong predictive performance within the simulation environment. Overall, the findings suggest that integrating TES with AI-based dynamic optimization can improve the flexibility, efficiency, and economic performance of solar-ORC systems. However, the reported results are  and require validation using experimental and plant-level data before practical deployment.