Delivering Reliable Reserve Commitments from Intermittent Electricity Resources

Abstract

Renewable Energy Producers (REPs) in electricity supply are increasing the need for flexible reserve facilities to manage their short-term intermittencies. Thus, reserve markets have become increasingly profitable. REPs may be attracted by this profitability but reserve markets require reliable commitments. It is an open question whether reliability is aligned with the profitable operations of REPs who use storage assets and/or financial hedging to manage their output reliability risks. Analyzing this question at an operational level requires a detailed analysis of the market microstructure and trading arrangements. We find that the outcomes depend on intraday market arrangements, either discrete (D-ID) auctions or continuous (C-ID) trading. We model a profit-maximizing REP's operation, with and without storage, providing reserve services via a multi-stage stochastic mixed-integer program with decision-dependent probabilities, supported by DID or C-ID trading. Using Benders decomposition and stochastic-dual-dynamic programming (SDDP), we analyze German data, which includes both C-ID and DID trading opportunities. Our findings indicate REPs could most effectively use C-ID trading for profitable and reliable reserve market participation, wherein batteries neither enhance participation nor improve system reliability. Conversely, in DID trading, REPS mostly rely on batteries for reliable, profitable participation. Accordingly, we also provide some regulatory recommendations.

Keywords: Renewable energy, electricity, reserves, storage, intraday markets, multi-stage, stochastic optimization, non-convex master problem