Why the power grid is filling up with batteries

A power grid has one rule it can't break: supply has to match demand every second. For most of the last century, that meant keeping power plants ready to ramp up and down on command. Now a new tool is taking over a big part of that job: the grid battery, a close cousin of the lithium-ion cell in your phone.

What a grid battery actually does

A grid battery is rated with two numbers. Megawatts (MW) measure how fast it can push power out. Megawatt-hours (MWh) measure how much energy it holds. A 100 MW battery that holds 400 MWh can run at full output for four hours. That's its duration.

The International Energy Agency sorts what batteries do into three jobs (IEA, May 2026):

• Energy shifting: charging when power is plentiful and discharging when it's scarce, often soaking up midday solar and releasing it after sunset.
• Ancillary services: fast bursts that keep grid frequency stable and cover sudden plant outages.
• Congestion management: sitting at bottlenecks in the wires to relieve strained lines.

Energy shifting has taken over. It was the main purpose of about 40% of new projects in 2015 and more than 90% in 2025. Batteries are getting longer, too: the average new project in 2025 offered about three hours of storage, up from about two in 2023 (IEA).

Storing power does cost something. The U.S. utility-scale battery fleet had an average round-trip efficiency of 82% in 2019, meaning about 82 of every 100 units of electricity put in came back out (EIA).

How fast it's growing

Worldwide: 108 GW of new battery storage was deployed in 2025, 40% more than in 2024. That's more than the peak year for new gas-fired plants, about 107 GW in 2002. Around 80% was utility-scale, and China accounted for about 60% of the total. Lithium-iron-phosphate (LFP) cells now make up about 90% of deployments (IEA Global Energy Review 2026). The IEA says battery costs fell by more than 90% between 2010 and 2025.

In the U.S.: utility-scale battery capacity reached 43.6 GW at the end of 2025. Operators added another 8.3 GW in the first half of 2026, bringing it to nearly 52 GW. They've reported plans for 14 GW more in the second half of 2026, 26 GW in 2027, and 14 GW in 2028 (EIA, Aug. 7, 2026). At the start of the year, developers planned 24 GW for all of 2026, after a record 15 GW in 2025. More than half was slated for Texas, at 53% (12.9 GW), followed by California at 14% and Arizona at 13% (EIA, Feb. 20, 2026). Plans are plans: 8.3 GW built plus 14 GW still planned comes to a little under that 24 GW target.

Where you can see it working

California is the clearest example. The state now has more than 55 GW of solar, which is more than its peak demand, so on sunny days its "net load" (demand minus wind and solar) can drop close to zero. Its battery fleet grew from under 1 GW in 2019 to more than 17 GW. On the evening of March 29, 2026, batteries covered more than 40% of the state's load. In early 2026, batteries handled more than 60% of California's hour-to-hour ramping. In April, they did more than 40% of ramping in Texas's ERCOT market (IEA).

Why it touches prices

Batteries are a business. The EIA notes that pairing solar with storage lets operators "store power when wholesale electricity prices are low and discharge electricity when prices are high" (EIA). In practice, that means batteries add supply at the hours when the grid is tightest and power is most expensive. How much that shows up on any household bill depends on the market, the utility, and regulators. There's no single answer.

What slows it down

Building a battery is quick. The IEA puts median construction at about 275 days, compared with more than two years for gas and over six years for nuclear. Getting one connected is slower. In the U.S., Europe, and Japan, projects typically take two to two and a half years to start operating, mostly because of permitting, financing, and grid connection. The IEA also notes that safety risks "remain low relative to the scale of deployment," and that public confidence depends on strong safety standards.

The takeaway

Batteries won't replace every power plant, and most hold only a few hours of energy. But they've become one of the main ways grids keep up with rising solar output, and they're being built faster than almost anything else on the grid.

This article is for education only. It is not investment advice.

---

Sources

• U.S. EIA, "Battery storage capacity averaged 70% growth over the last three years" (Aug. 7, 2026). https://www.eia.gov/todayinenergy/detail.php?id=67925
• U.S. EIA, "New U.S. electric generating capacity expected to reach a record high in 2026" (Feb. 20, 2026). https://www.eia.gov/todayinenergy/detail.php?id=67205
• U.S. EIA, "Utility-scale batteries and pumped storage return about 80% of the electricity they store" (Feb. 12, 2021; 2019 data). https://www.eia.gov/todayinenergy/detail.php?id=46756
• IEA, "Battery storage is scaling up and taking on a larger system role" (May 29, 2026), CC BY 4.0. https://www.iea.org/commentaries/battery-storage-is-scaling-up-and-taking-on-a-larger-system-role
• IEA, Global Energy Review 2026, "Technology: Battery storage." https://www.iea.org/reports/global-energy-review-2026/technology-battery-storage

Verification notes

• Verified (primary): every figure above comes directly from the linked EIA or IEA pages, read on 2026-10-05.
• Definitional, not sourced: the MW/MWh explanation and the "100 MW × 4 hours = 400 MWh" example are standard unit definitions, used for illustration.
• Our arithmetic: "8.3 GW built + 14 GW planned = a little under 24 GW" compares EIA's August data with its February plan. EIA didn't say this itself.
• Different sources disagree: the EIA says the U.S. added a record 15 GW of utility-scale batteries in 2025. The IEA, using Benchmark Mineral Intelligence data, says "over 16 GW" of U.S. utility-scale additions and 19 GW counting behind-the-meter. We use EIA figures for U.S. numbers.
• Dated figure: the 82% round-trip efficiency is from 2019 data, the most recent full year in that EIA article. Current fleet efficiency wasn't checked.
• Third-party: the California and ERCOT figures are IEA analysis of CAISO data and IEA's ERCOT figures. We didn't check them against CAISO or ERCOT directly.