Battery energy storage systems (BESS) are increasingly needed to provide flexibility and balancing services to accommodate rising shares of variable wind and solar generation, and rapid electrification with electric vehicles, heat pumps, and AI data centres. BESS deployment depends on whether market revenues are high enough and reliable enough to justify their capital-intensive investment. Revenue stacking across multiple markets is essential for a viable business model. However, most valuation studies neglect market-based operational constraints, including batteries being ‘skipped’: not being dispatched even when they offer the lowest-cost action. We quantify the value of revenue stacking across Great Britain’s day-ahead wholesale market and the balancing mechanism, and the cost of battery bids being skipped by the system operator. Using three years of half-hourly price data and a co-optimisation framework, we show that participation in the balancing mechanism increases BESS revenues by up to 250% relative to providing wholesale arbitrage alone. Stacking reduces revenue volatility, as diversifying income streams offsets the Balancing Mechanism’s higher price volatility. These gains are highly sensitive though. We show that each 10% increase in battery bids skipped reduces profits by 7%, or around £12,000 per MW per year. Similarly, pay-as-bid settlement means imperfect price capture proportionally erodes returns. These results highlight the need for policy changes that enable the better integration of BESS, as market design and dispatch integration can be as important as technology costs in determining whether batteries will deliver low-carbon flexibility at scale. • Optimisation of battery energy storage operating in both wholesale and balancing markets • Operation in both markets yields 250% higher profits than in the wholesale market alone. • Batteries can be ‘skipped’ in the merit order due to lacking information on their status. • A 50% skip rate reduces BESS profit by 30% when operating in both markets.
Gale et al. (Mon,) studied this question.