Scalable Multi-Value Joint Transmission and Generation Expansion Planning via Implicit Differentiation

Current electricity grid planning methods typically address transmission or generation expansion separately and only optimize for cost. As such, these methods do not adequately capture the realities of an evolving wholesale market. We address this shortcoming by introducing the multi-value expansion planning (MEP) problem in which transmission, generation, and storage can be jointly optimized for multiple objectives, such as cost, emissions, and consumer surplus. To solve the MEP problem, we develop a stochastic implicit gradient-based algorithm that scales well with the number of scenarios and problem dimension. Our numerical results suggest that investing in more transmission can significantly increase consumer surplus and reduce emissions.