RMI's 30% battery forecast is a 2024 number in 2026 clothes

RMI's new battery report arrives with a headline that belongs to an earlier policy era: the 30% annual growth in U.S. deployment the think tank projects over the next five years comes from a 2024 Wood Mackenzie forecast, made when the Inflation Reduction Act was still in full effect. The same consultancy cut its five-year utility-scale number to 16% earlier this year, and the forecast it released alongside RMI's report has utility-scale deployment flat in 2026 and growing 8% through 2031.

The headwinds are easy to list and hard to argue with: import tariffs, Treasury rules that make systems reliant on China-linked parts expensive to own, and the August executive order restricting deployment of foreign-produced bulk power components, which analysts say stacks a further constraint on the others. The 30% figure, then, reads as what batteries would be worth building if the policy stack had held still, and it has not.

The operating record holds up regardless of the growth rate. RMI cites an ICF analysis of the 2022 heat wave, when 3.4 GW of grid batteries met 6% of California's demand and materially strengthened the system's ability to maintain reserve margins; two years earlier, with less than 1 GW connected, the state saw rolling blackouts at a demand peak roughly 10% lower.

The report's most bankable material is its ledger of what storage pays the places that host it. Across project agreements in California, Massachusetts, Nevada and Texas, RMI puts average annual gross tax contribution near $6,600 per MW, with some installations above $10,000. A 250 MW/500 MWh battery that came online in February will return about $46 million to Medway, Massachusetts over 20 years, while Purple Sage, a 400-MW project in Nevada, will pay local governments $3.4 million a year even after partial state property tax abatements, and Libra, a solar-and-storage installation in the same state, will pay $170 million in property and sales taxes over its life. RMI also calculates that Suffolk County, New York could avoid $5.3 million in annual health costs by transitioning its 500-MW Port Jefferson peaker.

Those payments, not the growth rate, are what decide which projects get built: they convert a county board from a risk into a counterparty and arrive years before the interconnection queue moves. Consent is the product, and storage has the easier version of that sale because the health math argues back against a capital stack that, as we noted when the carbon-capture mandate fell, has lately been gas's. The queue is where the forecast fight resolves; if utility-scale deployment goes flat in 2026 as Wood Mackenzie now expects, the withdrawals will surface there, and the projects left standing will be the ones whose payment schedule a town has already signed.