ACP and Wood Mackenzie project 27% growth in commercial energy storage

American Clean Power and Wood Mackenzie project the market for commercial energy storage will grow 27% over the next five years, a forecast whose arithmetic leans on hours delivered rather than on a rising count of installations. Their quarterly count has commercial-property installations at 48 megawatts in the second quarter, down from an unusually high first quarter that the report credits to California, where facilities and households installed ahead of the state's top court resolving questions about reduced tax incentives that took effect in 2023.

Megawatt-hours still rose even as the count of installations dipped, and the report points to duration as the reason: the average U.S. storage system now runs 3.5 hours, up from 2.8, so a smaller pipeline of projects puts more capacity into a market that pays for it.

Chemistry is where duration comes from, and lithium-ion still supplies most of it: about 90% of commercial battery systems are lithium-ion on 2025 data from the International Energy Agency. One alternative now has a large order behind it, at a Google data center where the company closed a deal earlier this year with Form Energy and utility Xcel Energy to install 30 GWh of iron-air batteries that Form says can release power for up to 100 hours. Amanda Peterson Corio, Google's head of data center energy, said when the deal was announced that pioneering long-duration storage helps build a more resilient system.

Policy pulls storage into the load queue from the other direction, with states and municipalities requiring data centers and other large load users to bring their own power, which in many cases includes a storage component, while public entities at the local, state and federal levels back virtual power plants, the programs in which utilities call on excess capacity in commercial buildings and homes to support the grid. The report describes distributed storage's value as broadening beyond customer benefits to grid services, as the industry tries to position VPPs as an enabler of load growth.

That distinction matters to how the assets get financed. A battery earning behind-the-meter savings for its host is a customer contract; one earning grid-service revenue through a utility program is closer to infrastructure, and the report's language suggests the second bucket is where the growth is being underwritten, a pattern Google's own procurement shows from the buy side, from the virtual-plant funding with PG&E this publication covered in September to the geothermal contract with Fervo Energy.

This segment's numbers will keep arriving lumpy and policy-timed, as California's incentive scramble showed, and the 27% projection will show up in hours before it shows up in counts of new sites. The next ACP-Wood Mackenzie update is worth checking for installation count and average duration, because the report's own second-quarter data has them moving in opposite directions.