Electric grids have always run on one rule: supply must match demand every second. For a century, engineers met it by adjusting power plants that burned coal, gas or water on command. Wind and solar do not take orders from a control room. As they grow into major sources of electricity, utilities need somewhere to put energy when the sun is high and take it back when the sun goes down. Increasingly, that place is a battery.

A steep drop in cost

The change is driven by price. BloombergNEF's annual survey tracks the cost of lithium-ion battery packs, and it has found them falling from well above $1,000 per kilowatt-hour in 2010 to about $115 in 2024 in inflation-adjusted terms. Solar has followed a similar path. The International Renewable Energy Agency has reported that the cost of electricity from utility-scale solar fell by roughly 90 percent between 2010 and 2023.

Cheaper panels created the problem batteries now help solve. Solar output peaks at midday, when demand is often moderate, and drops quickly at sunset just as people get home and switch on lights and appliances. California's grid operator popularized a picture of this pattern, the "duck curve," in which net demand sags in the afternoon and then climbs steeply in the evening. Without storage, operators must either throw away excess solar at midday, a practice called curtailment, or ramp up gas plants fast at night.

What the first big projects showed

A turning point came in South Australia. In 2017 the Hornsdale Power Reserve, built with Tesla batteries, began operating near Jamestown with a capacity of 100 megawatts, then the largest lithium-ion battery in the world. It was later expanded to 150 megawatts. The state had suffered a statewide blackout in 2016 and was looking for ways to steady a grid heavy in wind power.

Hornsdale's strength was speed. When a large coal unit tripped offline in December 2017, the battery responded within a fraction of a second, injecting power to arrest the fall in grid frequency before slower plants could react. Grid managers later reported that batteries had cut the cost of frequency control services in the state considerably. The result changed how many system operators think about batteries, from an expensive add-on to a flexible tool for stability.

California is now one of the largest users in the world. Battery capacity in the state grew from about 500 megawatts in 2018 to more than 10,000 megawatts in 2024, and on many evenings batteries now supply a large share of the power that used to come from gas plants. Texas, which has plenty of sun and wind, has also added storage quickly.

Where the technology is heading

Most grid batteries today use lithium-ion chemistry, and most are designed to discharge for about four hours. That works well for shifting solar energy into the evening. It does not help with multi-day stretches of cloud and calm, which are rare but real.

Several approaches are being tested for longer durations:

• Iron-air batteries, which use rust-based chemistry and aim to discharge for 100 hours at low cost. Companies such as Form Energy have built early projects.
• Flow batteries, which keep energy in tanks of liquid electrolyte that can be scaled up separately from the power rating.
• Sodium-ion batteries, which avoid lithium and could be cheaper in some settings, with commercial production now beginning in China.
• Pumped hydropower, still the largest form of grid storage worldwide by a wide margin, which moves water uphill and lets it run back down through turbines.

Engineers are also rethinking how batteries connect to the grid. Conventional inverters follow the grid's existing signal. Newer grid-forming inverters can create that signal themselves, helping a power system hold its frequency and voltage even when few spinning turbines are online. That capability matters because the old fleet of large generators provided stability almost as a side effect, and its loss has to be replaced on purpose.

The obstacles

For all the progress, batteries are not a cure-all. Four problems stand out.

The queue. In many countries, getting connected to the grid takes years. A 2024 analysis by Lawrence Berkeley National Laboratory found that nearly 2,600 gigawatts of generation and storage were waiting in US interconnection queues at the end of 2023, more than the total installed capacity of the American power system. Projects that reached operation had typically waited about five years.

Transmission. Even with storage, power must get from where it is made to where it is used. New high-voltage lines can take a decade or more to permit and build, and local opposition is common.

Supply chains. China dominates the production of battery cells and of the processing of key minerals including graphite and lithium. Trade restrictions and tariffs can quickly alter the price of imported systems.

Safety. Lithium-ion batteries can catch fire, and the consequences at large sites can be serious. A fire at the Moss Landing battery facility in California in January 2025 led to evacuations and renewed scrutiny of siting and emergency planning. Industry data suggest the rate of failures per unit of installed capacity has dropped sharply as designs and safety standards improved, but communities are asking for more transparency and stricter rules.

Lessons from stress on the grid

The blackout that struck Spain and Portugal in April 2025 became a reference point for the debate over how to run grids with high shares of solar and wind. Investigators focused on voltage control and the way the system responded as conditions deteriorated. Many engineers said the event underlined the need for more fast-acting resources, including batteries with grid-forming capability, and for clearer rules about which plants must provide voltage support. It did not show that renewables are unreliable, but it did show that a grid changing this quickly needs deliberate design.

Denmark offers a reassuring counterexample. It has produced more than half of its electricity from wind in recent years while maintaining one of the most reliable grids in Europe, helped by strong connections to neighboring countries and flexible demand.

What the numbers mean

Storage will not replace every other source of reliable power in the near term. Grids will still need a mix of tools: transmission, demand management, hydropower, nuclear plants where countries choose them, and some gas capacity for the rarest and longest shortfalls. What batteries have done is make the daily rhythm of solar and wind manageable at a cost few predicted a decade ago.

That shift has quietly changed the central question for power planners. They used to ask whether renewables could supply a reliable grid. The more pressing question now is how fast the wires, permits and factories can keep up with the equipment that already works.