Stochastic Power-Gas-Battery Expansion Planning Integrated High Penetration of Green Resources using a Stochastic Bi-Level Model

This paper introduces a groundbreaking bi-level stochastic mixed integer linear programming (MILP) model, uniquely integrating the natural gas network with high renewable energy penetration, specifically wind and solar, into generation, transmission, and battery expansion planning. This innovative model efficiently addresses the complexities and uncertainties associated with renewable energy integration, capturing both upper-level decisions (generation, transmission, battery expansion) and a detailed lower-level natural gas network. Tested on two networks – a 6-bus electricity system with a 4-node gas grid and a 24-bus system with a 10-node gas grid – the model demonstrates significant cost efficiencies. For the 6-bus system, integrating renewable energy reduced the objective function by 28.1%, from $388 million to $279 million. In the 24-bus system, renewables achieved an 11.6% reduction, from $1,190 million to $1,052 million. Additionally, optimal budget allocation resulted in a 20.4% increase in costs when reduced by 20.6% in the 6-bus system, while the 24-bus system showed resilience with only a 1.5% cost increase under similar constraints. These results provide critical insights for policymakers, operators, and stakeholders, emphasizing the importance of renewable energy integration and budget prioritization for cost-effective and sustainable expansion planning.