Integrated electricity-gas system planning under cross-vector uncertainty: A scalable multi-stage stochastic framework

The growing interdependence between integrated electricity and gas systems (IEGS) calls for planning methods that capture their coupled long-term uncertainties and operational interactions. This paper develops a multi-stage stochastic framework for assessing the interplay of integrated electricity and gas systems (IEGS) in expansion planning under uncertainty, leveraging a scalable Column Generation and Sharing (CG-S) solution strategy to tackle computational challenges. The model jointly optimises investments in electricity transmission and generation, as well as natural gas (NG) processing facilities, under uncertainty in NG availability and prices, electricity demand growth, and capital costs. It integrates detailed hourly power system operations and daily gas flows, preserving realistic temporal granularity. Case studies on the Australian East Coast Energy System demonstrate that the CG-S algorithm reduces convergence times by up to 45% while maintaining stable memory use. The proposed planning framework enables the identification of a coherent pathway for strategic, coordinated investments across integrated electricity and gas systems, while uncoordinated deterministic practices risk oversizing early generation investments by up to 55% due to the inability to leverage cross-system synergies.