Introduction

Short-duration battery storage capacity in the U.S. increased from 1,694 MW in January 2021 to 43,419 MW in January 2026 (p.1), a 25x increase in five years. Current forecasts call for installed capacity to double again by end of 2027, reaching almost 90,000 MW (p.1). The success of two- to four-hour lithium-ion batteries has created an opening for longer-duration solutions capable of storing up to four days of power.

Early moves on LDES

Texas now has more than 18,000 MW of installed battery capacity, while California has just over 16,000 MW (p.2). Battery discharge in ERCOT topped 10,000 MW on 13 March, supplying more than 20% of total demand (p.2). In California, supply topped 12,000 MW on 29 March, accounting for 42% of total demand (p.2).

Despite this growth, the absence of market structures to make long-duration storage bankable has been a key roadblock. As of 2024, “no market mechanism[s] exist to address LDES needs” (p.3).

State targets give LDES a big push

Several states have enacted procurement mandates for long-duration energy storage (LDES). California requires 6,000 MW of new generation capacity by 2032, with at least 1,500 MW capable of discharging for eight hours or more (p.4). Virginia’s HB 895 requires Dominion Virginia to install at least 4,000 MW of long-duration storage by 2045, while Appalachian Power must add 520 MW by the same date (p.4). Massachusetts legislation calls for 5,000 MW of energy storage by 2030, including 750 MW of multi-day storage, and the Governor’s executive order targets 5,000 MW of new storage by 2035 (p.4).

Hyperscalers come calling for LDES

Data centre and AI companies are emerging as significant backers of LDES. Xcel and Google announced a deal for 300 MW of long-duration storage to power a new Minnesota data centre, with Google contributing $50 million toward a distributed battery programme (p.5). In April 2026, Meta reserved 1 GW/100 GWh of storage capacity from Noon Energy, with first delivery slated for 2028 (p.5).

LDES potential as a peaker replacement

An analysis of Wagner Unit 3 in Maryland — a 60-year-old, 305 MW facility — illustrates LDES’s potential as a peaker replacement. The plant’s O&M costs in 2024 amounted to more than $309/MWh (p.6), with annual operating costs estimated at roughly $113 million, pushing costs per MWh above $800 (p.6). A 100-hour battery sized at 325 MW would have a discharge capacity of 32,500 MWh, providing a 15% cushion over the unit’s peak generation period (p.6).

Many technologies in the running

Several companies are advancing LDES technologies with durations ranging from 4 to 100+ hours, including iron-air, liquid air, compressed air, CO2 batteries, and carbon-oxygen fuel cells (p.7). No single technology is expected to dominate the market.

Is hydrogen a long-duration storage option?

IEEFA is doubtful that the ACES renewables-to-hydrogen project in Utah will prove economically competitive or environmentally beneficial (p.7). The project will only be able to produce about 63% of the hydrogen needed to meet the planned 30% blending target (p.8). Using hydrogen as a fuel for gas turbines is unlikely to be economically or environmentally effective.

Conclusion

Total U.S. battery storage capacity now tops 51 GW (p.8). IEEFA believes a similar rapid expansion is possible in the long-duration storage market, driven by state mandates, federal tax incentives, and business demand from the AI sector.