The surprising effectiveness of mechanical energy storage

Posted: December 31, 2025

As renewables generate more of our power, we need much more capacity to store that power and release it to the grid when the sun’s not shining or the wind’s not blowing. Luckily, turnkey battery energy storage system (BESS) prices fell by 40% in 2024 alone and the U.S. is expected to have nearly doubled its grid-scale battery storage in 2025. 

Today, we want to dive into the alternatives to batteries for grid-scale energy storage—pumped hydro, compressed air and thermal energy storage—and take stock of the role they continue to play in our rapidly changing grid.

Pumped hydro energy storage (PHS)

Pumped hydro energy storage (PHS) is both the earliest form of grid-scale electrical energy storage and, up until the last year, by far the most prevalent. When demand for electricity is low, PHS stations pump water from a lower reservoir to a higher reservoir. When demand for electricity is high, water flows downhill from the high reservoir back to the lower one, spinning turbines that generate electricity.

As recently as 2019, pumped hydro provided some 94% of grid-scale power storage in the U.S. Since then, electrochemical battery storage has grown exponentially from a single-digit percent of grid storage to exceeding pumped hydro as the largest component of U.S. grid storage in 2024: 26 GW to pumped hydro’s 23 GW.

But PHS has continued growing over the last two years. In 2024, the State Grid Corporation of China (SGCC) completed the largest PHS plant to date at 3.6 GW, and more installations are in the works.

The UK company, RheEnergise, is pioneering a new PHS innovation: using a proprietary mineral-rich fluid 2.5x as dense as water. The denser fluid means the company can build smaller installations that have the same power capacity as hydro plants 2.5x larger. It also means the plants can be built in locations with shallower elevation gradients for less cost. This past summer, the company completed installing the main mechanical works at a demonstration plant in Devon and is targeting commercial sites throughout North America, including mines, data centers and solar and wind installations.

It's easy to see why pumped hydro has long been the first choice for grid-scale storage. It’s highly efficient—discharging about 70%–85% of the energy it stores—and it can discharge power on the order of 1–3 gigawatts for 8–20 hours. Only geothermal energy storage (discussed below) beats those stats. So batteries don’t look likely to replace pumped hydro storage any time soon.

Nonetheless, pumped hydro isn’t always the most viable option: Not every place has easy access to water and uneven terrain.

Compressed air energy storage (CAES)

Compressed air energy storage (CAES) often has the lowest cost per kWh once you take into account maintenance, operations and other overhead. CAES systems use off-peak energy to compress air in large underground caverns. When electricity demand rises, they decompress it to run generator turbines. Some facilities can release energy for up to 25 hours.

The main challenge for CAES systems is that compressing air heats it to around 650°C (1200°F)—so hot that it can destroy the caverns meant to store it. Conversely, decompressing the air cools it to -150°C (-238°F)—so cold that it may damage the turbines it needs to spin. So, CAES systems must cool compressed air and reheat it as it decompresses.

Legacy CAES systems are diabatic: They burn natural gas to reheat the compressed air. Because they use more energy—and fossil fuel—to recover the energy stored in the compressed air, they operate at about 42%–55% efficiency: much less efficient than pumped hydro.

Some new CAES systems are adiabatic: They store the heat extracted from the air as it’s compressed by using it to heat stone or molten salt. The system then uses that stored heat—instead of natural gas—to decompress the air. Companies developing new a-CAES systems estimate them to be about 60%–70% efficient, with efficiency increasing with capacity.

The next step for compressed air storage is liquid air energy storage (LAES). Highview Power is building the world’s first commercial LAES facility in Manchester, UK. It’s slated to start storing energy in 2027. The plant will compress and chill air in insulated storage tanks at temperatures so low that the air liquifies. The heat captured in the chilling process will then be used to evaporate the liquified air to run turbines. Because it’s storing and reusing the waste heat from cooling, the Highview plant is expected to approach 70% efficiency.

Gravity batteries

Instead of pumping water to an elevation and generating power as it falls back to level, some companies are building facilities that lift heavy weights on cables and generate power as they fall back to earth. The weights can be made of concrete, composite material—really anything, including recycled and locally available material—which proponents say makes it perfect for a circular economy.

In late 2023, the Swiss company, Energy Vault, connected the first commercial-scale gravity battery to the grid in Rudong, China. Now it’s partnering with ENEL to build a smaller-scale project in Texas. The Rudong battery releases 25 MW of power over four hours and is expected to achieve an impressive 80% efficiency. Other startups are also looking into installing gravity batteries in disused mine shafts and old oil wells.

Thermal energy storage (TES): Sand, rock, salt

In June, the Finnish company Polar Night Energy completed the world’s largest sand battery for the Finnish district heating network, Loviisan Lämpö. The battery is an insulated cylinder 13 meters tall and 15 meters wide, filled with 2,000 tons of crushed soapstone. When electricity is cheap and abundant, it heats the soapstone to 400°C (752°F).

The town of Pornainen then uses that heat to heat water pipes that feed its district heating system. But other installations could use it to heat air, generate steam or even regenerate electricity. In theory, it could be used to store heat for one of the adiabatic CAES systems mentioned above.

The battery has a relatively low power capacity of just 1 MW, but it can discharge that power for 100 hours. For even larger applications, Polar Night Energy also advertises larger 2 MW and 10 MW models.Other thermal energy installations use different types of rock and sand and even molten salt sourced as industrial byproducts. When used to deliver heat (and not regenerate electricity), the efficiency of thermal energy storage approaches 90%. The low cost of materials and high efficiency make the technology an extremely cost-effective option when storing power for heat is the goal.

Geothermal energy storage

We profiled Sage Geosystems in October, which uses geothermal energy to both store and generate renewable power with its EarthStore system. The company drills deep into the earth, either through abandoned oil wells or at new sites. It then uses electricity to pump water deep into those wells. Both the pressure of the earth and the high underground temperatures pressurize the water so it generates power when released back to the surface through a turbine. Because geological heat adds energy to the water underground, the power it generates can be twice as much as the electricity used to pump it underground, giving it an efficiency of over 200%!

The company expects to complete its first commercial-scale facility next year in Texas, with a capacity of 3 MW. The Texas facility is targeting 6–10-hour storage, though Sage envisions other installations that could store energy over several days, possibly allowing for longer discharge times than either pumped hydro or CAES.

Where do batteries fit in the grid-scale storage mix?

Pumped hydro, compressed air and thermal storage all have one big advantage over booming new battery technologies: They can discharge their power over much longer time scales—over 8 hours to multiple days. Batteries still discharge in less than five hours. These mechanical storage systems can help the grid cope with the weather disruptions that come along with renewable energy.

But, as a new report from Ember notes, batteries are still a key tool that may make 24-hour solar power a reality very soon. In future issues, we’ll take a look at how battery technology is changing and how it stacks up against mechanical and thermal energy storage.