Optimized Two-Stage Planning Model for Integrating Compressed Air Energy Storage With Uncertain Correlated Wind Farms in Power Systems

Abstract

As energy demands grow, integrating renewable energy sources (RESs) and energy storage systems (ESSs) has become essential for reducing carbon emissions and reliance on conventional power plants while maintaining grid security. Among ESS technologies, Compressed Air Energy Storage (CAES) stands out as a promising solution but remains underexplored in grid applications. This study introduces a two-stage mixed-integer non-linear programming (MINLP) model for optimizing CAES integration with hybrid power systems. The first stage addresses planning variables, including site selection, capacity, and power ratings, minimizing total costs and voltage deviations. The second stage optimizes the operation of CAES units and conventional generators, focusing on reducing operational costs, energy losses, carbon emissions, and voltage deviations. The model employs probabilistic approaches with the five-point estimation method (5-PEM) and Clayton copulas to handle wind power variability. Optimization is achieved using MOPSO-TOPSIS and a hybrid NSGAII-MOPSO-TOPSIS, with Tabu Search Algorithm (TSA) for local refinement. The model’s effectiveness is validated on a modified IEEE-57 bus system with wind farms. Results demonstrate significant economic, environmental, and technical benefits of CAES integration. The hybrid approach (Case 3) outperforms MOPSO-TOPSIS (Case 2), achieving a 3.68% reduction in daily fuel costs and a 6.05% reduction in carbon emissions, compared to 2.92% and 4.23% in Case 2. Case 3 also shows superior reductions in power losses (5.04% vs. 4.84%) and voltage deviations (0.81% vs. 0.60%).

Keywords

• Compressed air energy storage system (CAES)
• Hybrid optimization algorithms
• Integrated energy system
• Energy storage
• Wind farm modeling
• Sustainable energy