Lawrence Berkeley National Laboratory

The security-constrained AC optimal power flow (SCACOPF) is one of the most important problems that the industry has to solve on a daily basis in several electricity markets across the globe within strict time limits. Mathematically, the SCACOPF is formulated as a mixed-integer non-convex program, which is a very challenging class of problems to be solved in practice, and even obtaining feasible solutions can be difficult. In addition, most independent system operators (ISOs) impose a deterministic n-1 security criterion and, consequently, further increase the dimensionality of the problem. In this paper, we present an algorithmic approach that obtains a feasible solution to this problem with a 1-hour time limit. Our approach combines contingency selection, decomposition, and tailored warm start to the non-convex problem, and its performance is illustrated with our results in Challenge 2 of the Grid Optimization (GO) competition created by the U.S. Advanced Research Projects Agency-Energy (ARPA-E).