Adaptive two-stage stochastic energy infrastructure expansion planning

This paper introduces an adaptive two-stage stochastic optimization model for energy infrastructure expansion planning under demand uncertainty. Unlike traditional two-stage models, which can be too rigid, or multi-stage models, which can be overly flexible, our model seeks a balance between commitment and flexibility. It allows each investment decision to adapt at one or two designated adaptation times to the unfolding uncertainty while maintaining a static policy before and after these times, where the adaptation times are determined within the proposed optimization model. The model’s performance is evaluated using two metrics and compared with conventional approaches. To enhance the model’s practicality, five strategies are presented for sharing adaptation times among related investments, further reducing the number of plan revisions. A case study on Rwanda’s electrification plan demonstrates that this approach can save up to 3.44% compared to the two-stage model, while requiring significantly fewer adaptations than the multi-stage model. Additionally, the model also provides actionable guidance for policy makers including the optimal adaptation frequency and timing of policy revisions and the ideal planning horizon length. • An adaptive two-stage stochastic energy planning model is proposed. • The balance between commitment and flexibility in investment is studied. • Five approaches for sharing adaptation times are developed. • Case study shows substantial cost savings while preserving flexibility. • Model results offer valuable insights to enhance the planning process.