Pricing fast-responding electric storage assets in the presence of negative prices and price spikes: a simulation-and-regression approach

Challenged by weather-dependent and intermittent outputs of renewables, modern electricity markets experience frequent price spikes and the occurrence of negative prices. Electric storage can smooth the imbalance between power supply and demand, and thus helps fully utilize renewables. To encourage the use of electric storage in the grids, an accurate valuation of storage technology is of great importance. In this paper, we focus on the use of fast-responding electric storage for real- time power trading and propose a novel valuation model that considers the presence of price spikes and negative prices within a simulation-and-regression framework. By calibrating our model against California ISO data, we assess the economic feasibility of four real-market battery technologies: sodium sulfur (Na-S) molten salt, zinc bromide flow, lithium iron phosphate and lithium nickel cobalt aluminum batteries. Although none of these batteries reaches breakeven of its merchant value, Na-S molten salt and zinc bromide flow require relatively fewer cost reductions. Further, we find that real-time storage optimization is more valuable than day-ahead optimization by a factor of 3.12–16.64, which suggests that fast-responding electric storage technologies are potentially valuable if their capital costs become sufficiently low in the future.