Distributed robust coordinated multiple electricity substitution planning under network constraints and uncertainty

To achieve sustainable social development, China is promoting an electricity substitution strategy as part of its clean development initiative. This paper proposes a distributed robust coordinated optimization approach to address multiple electricity substitution planning challenges under network constraints and demand uncertainty. A linear decision rule-based robust model is developed to handle demand uncertainty when sizing electric boilers, electric heat pumps, and electric flue-curing tobacco machines, enabling adjustable conservatism while minimizing network impacts. Substitution necessity constraints are introduced to quantify cost gaps between electricity and fossil fuel devices, revealing that device capacities and costs increase with the threshold, and planning schemes become infeasible once the threshold reaches 0.3. A bi-level distributed optimization framework coordinating multiple electricity substitution planning subproblems at the lower level and enforcing power flow constraints at the upper level, is solved by the Augmented Lagrangian Alternating Direction Inexact Newton algorithm. By exchanging sensitivity matrices, the algorithm improves convergence, thus achieving the same accuracy as the Alternating Direction Method of Multipliers with roughly half the iterations. Extensive case studies demonstrate that network constraints reduce planned capacities by 7.03%–11.84% depending on node location, with the largest impacts occurring in summer and winter, verifying the proposed method’s efficiency and robustness.