On June 1, a battery installation in California’s sun-scorched Kern County quietly rewrote what the American grid can do after dark. The Tumbleweed project became the first major U.S. battery facility capable of discharging power for a full eight hours — twice the industry standard — without burning a drop of gas.
It isn’t the biggest battery ever built. Size was never the point. The grid needs power at 2 a.m., and for years, clean energy had no convincing answer for those final hours before sunrise.
A battery that outlasts the sunset
Tumbleweed went online June 1 in Kern County, delivering 125 megawatts of capacity that can discharge continuously for eight hours — double the four-hour standard that has defined U.S. grid battery design for years. That four-hour benchmark wasn’t arbitrary; it reflected where equipment costs and market opportunities converged. Crossing to eight hours is a meaningful operational and market milestone, not merely a number on a spec sheet.
Scale matters here too. The only previous eight-hour U.S. battery was a 6-megawatt installation on Nantucket in 2019, purpose-built to address a specific island-grid problem. A 50-megawatt, eight-hour project came online in Australia just weeks before Tumbleweed. At 125 megawatts, Tumbleweed is the first to deliver this duration at a size that can genuinely stress-test the concept on a major interconnected grid.
The engineering behind the milestone is less dramatic than the milestone itself. Rev Renewables ran a second construction phase and “literally doubled the number of battery boxes on the site,” according to Cody Hill, who leads storage development for the company. The same contractor, Mortenson, handled both phases. No exotic chemistry, no experimental hardware.
Why eight hours changes California’s clean-energy math
California’s solar window runs roughly from 8 a.m. to 6 p.m. A four-hour battery charged at noon can push power back to the grid until around 10 p.m. — useful, but not sufficient. Tumbleweed can absorb cheap midday solar and discharge it all the way to 2 a.m., closing a gap that has frustrated grid planners for years.
The remaining six overnight hours — 2 a.m. to 8 a.m. — are already low-demand. California currently covers them with a mix of wind, nuclear, geothermal, hydropower, and some fossil gas. Eight-hour storage doesn’t need to solve that window entirely. It only needs to bridge from sunset to the point where overnight clean sources can carry the load, and those two pieces together create something close to a complete clean-energy cycle for a typical day.
California law requires a zero-carbon grid by 2045. Longer-duration storage isn’t a side feature of that plan — it’s a core structural requirement.
Policy that pushed ahead — even when regulators stepped back
California regulators mandated long-duration storage procurement in 2021, directing the state’s investor-owned utilities to begin securing contracts by 2026. The utilities missed that target. The California Public Utilities Commission responded by pushing the deadline to 2031.
Two nonprofit community power providers didn’t wait. California Community Power and Ava Community Energy signed contracts for Tumbleweed in 2022, moving forward while the larger utilities stalled. Ava, which serves customers in Alameda and San Joaquin counties, controls 50 megawatts of the project; California Community Power holds 75 megawatts.
Those anchor contracts were essential to making the project financially viable. Hill was direct: merchant-market revenue alone would not have justified the cost of doubling the battery capacity. The mandate existed precisely to fund infrastructure before the grid urgently needed it. Tumbleweed is now online ahead of any formal requirement — the outcome the original policy intended, even after regulators lost their nerve halfway through.
Lithium-ion’s quiet victory over the long-duration startup wave
For roughly a decade, venture capital poured into companies built on one premise: that lithium-ion batteries could not economically handle storage beyond four hours. Scores of startups raised billions developing alternatives. The results have been, broadly, a wave of bankruptcies and almost no utility-scale deployments.
Tumbleweed uses lithium-iron phosphate cells from BYD — a high-volume, commercially proven technology. Hill said he takes calls from startups pitching novel storage devices but chose a bankable chemistry already in mass production. “It improves along every metric, and it gets cheaper,” he said.
That doesn’t eliminate every alternative technology; at durations of 100 hours or more, lithium-ion material costs become prohibitively expensive, and developers like Form Energy and Noon Energy are working in that space. But the four-to-eight-hour window — where most startup competition was concentrated — now belongs to proven chemistry.
What comes next for long-duration storage
Tumbleweed is operational, but the real test is just beginning. California Community Power and Ava are still working out how to bid their portions into California’s wholesale markets, and the run-time data accumulating over the coming months will be the first meaningful evidence of how eight-hour batteries actually perform in a major grid environment.
Lithium-ion costs continue to fall. Hill expects further duration extensions to follow as the economics improve — whether that means ten hours, twelve, or beyond remains an open question that Tumbleweed’s performance data will help answer.
Grid planners haven’t reached consensus on when abundant renewables make longer-duration bulk storage essential everywhere, and that debate will continue. But if Tumbleweed’s model replicates across California and beyond, the state gains a working template for covering nearly every hour of a typical day with clean electricity — well before the 2045 deadline arrives.
Carlos is an engineer with strong expertise in technical and industrial topics. He previously worked at international companies such as Siemens and speaks Spanish, German, English, and Italian.
