Optimal storage capacity with intermittent wind energy
Work in progress
I study the optimal operation and capacity of grid-scale storage in a strongly wind-dominated electricity system with or without fossil backup. Wind generation follows a continuous-time stochastic process calibrated to German wind production data. A social planner with isoelastic preferences chooses charge and discharge rates to maximize intertemporal welfare subject to storage constraints. The model delivers a simple feedback policy and enables efficient numerical solution. Absent any fossil backup, the price elasticity of demand is a primary driver of the marginal value and optimal size of storage.
An optimal wind-plus-batteries system without backup in Germany features more than ten times the 2023 installed wind capacity and storage energy covering hours to days of average German demand, with a levelized cost of more than twice current wholesale electricity prices. By contrast, an optimal wind-plus-hydrogen system features less extreme wind capacities and storage energy covering weeks to months of average demand, despite storage losses. The wind-plus-hydrogen levelized cost is well below gas or coal levels when discounting with 2% annually. A deterministic benchmark confirms the first-order relevance of uncertain wind output: storage needs are driven by the risk of prolonged low-wind spells. Hydrogen's far lower energy capacity cost dominates efficiency losses and power capacity costs in this setting.
Allowing for a fossil backup sharply reduces optimal storage. Marginal abatement costs are initially negative, then spike when near zero emissions: it is very costly to avoid using an already present fossil backup even in the rarest events of prolonged low renewable production.
Links: Coming soon.
Presented at (selection):
2025: EAERE conference (Bergen), FAERE conference (Nantes), ENTER Jamboree (Stockholm), UAB (Barcelona, ENTER exchange seminar), University of Leipzig (Lunchtime seminar).