Toward Scalable Stochastic Unit Commitment - Part 2: Assessing Solver Performance

In this second portion of a two-part analysis of a computational approach to scalable stochastic unit commitment, we transition our focus from approximating accurate stochastic process models of load to solving the resulting stochastic optimization models in tractable run-times. Our solution technique is based on Rockafellar and Wets’ progressive hedging algorithm, a scenario-based decomposition strategy for solving stochastic programs. To achieve high-quality solutions in tractable runtimes, we describe key customizations of the progressive hedging algorithm for stochastic unit commitment. Using a variant of the WECC-240 test case, we demonstrate the ability of our approach to solve moderate-scale stochastic unit commitment problems with reasonable numbers of scenarios in less than 30 minutes of wall clock time on a commodity hardware. Further, we demonstrate that the resulting solutions are high-quality, with cost typically within 1-2% of optimal. Our optimization model and associated test cases are publicly available, serving as a basis for evaluating the relative effectiveness of stochastic unit commitment solvers.