Large-scale dynamic optimal power flow problems with energy storage systems

Abstract

In this paper, a novel approach is presented to solve large-scale Dynamic Optimal Power Flow (DOPF) problems associated with power grids containing Energy Storage Systems (ESSs), using a distributed computing architecture, referred to as a Smart Grid Communication Middleware (SGCM) system. The problem is solved by splitting the time horizon into shorter time intervals using the SGCM system, solving one iteration of a Primal Dual Interior Point Method (PDIPM) for each resulting sub-problem in parallel and exchanging boundary variables after each iteration in three different stages. The sub-problems eventually converge to a solution near to an optimal solution of the entire problem without having to solve the original Security Constrained OPF (SCOPF). The methodology is tested on the German transmission grid, where the computational effort is shown to be reduced significantly; thus, with an entire horizon of 96 time steps and a decomposition into 8 sub-problems, the runtime can be decreased from over 2 hours to less than 10 minutes with an overall system cost increase of less than 1%, which makes the DOPF model with inter-temporal ESS constraints suitable for online applications.