← All terms

Glossary · Energy industry & energy transition

CAES, Compressed Air Energy Storage

CAES stands for Compressed Air Energy Storage (in German: Druckluftspeicher). These plants use electrical energy to compress air and later use the compressed air to generate electricity again. Strictly speaking, a compressed air energy storage plant does not store electricity but an energy potential: the air is held at elevated pressure. When it is expanded, it drives a turbine or another expander.

The principle is simple; the technology is not. During charging, electric motors drive compressors. They force air into a storage space, such as an underground salt cavern, a rock cavern or a pressure vessel. During discharging, the compressed air flows back in a controlled manner, is expanded and drives a machine that generates electricity via a generator. Two parameters must be distinguished. Power indicates how quickly the storage system can absorb or deliver electricity (in kilowatts, megawatts or gigawatts). Energy content indicates how long this power is available (in kilowatt-hours, megawatt-hours or gigawatt-hours). In compressed air storage this separation is particularly clear: compressors, turbines and generators determine the power. Storage volume, pressure range and temperature management determine the energy content. Charging and discharging power can differ greatly.

Compressed air energy storage belongs to the mechanical storage systems. Battery storage, by contrast, stores energy electrochemically. In hydrogen storage, electricity is converted into chemical energy by electrolysis. Technically closely related is pumped storage: it too converts electricity into another physical form of energy and back again later. There, water is pumped into a higher-lying reservoir; in compressed air storage, air is compressed. Both technologies need suitable sites. However, they differ in geography, efficiency, design, permitting and possible storage duration. The Huntorf power plant is therefore occasionally described as the pumped storage plant of northern Germany.

Heat determines efficiency

Air heats up when compressed and cools down sharply when expanded. This heat question is not a side issue. It determines how much of the electrical energy input is later available again as electricity. If the heat of compression is discarded, a considerable part of the energy is lost. If the air is reheated before expansion, more work can be extracted from it.

  • Diabatic compressed air storage: It largely releases the heat of compression to the environment. The cold compressed air therefore has to be reheated before the turbine. In the two classic plants, Huntorf (Germany) and McIntosh (USA), this is done by burning natural gas in a gas turbine. Such plants are hybrid power plants: they store compressed air but additionally consume fuel. For Huntorf, the literature gives an efficiency of about 42 percent, for McIntosh about 54 percent. McIntosh uses a recuperator that preheats the combustion air with turbine exhaust gas. Because two energy carriers are used, electricity and gas, these values depend strongly on the calculation method. They cannot be compared directly with the efficiency of a battery.

  • Adiabatic compressed air storage: It stores the heat of compression in a high-temperature heat storage unit and returns it to the air during discharging. With a suitably designed turbine, no supplementary firing with natural gas is then needed. However, the heat storage unit must cope with high temperatures, many cycles and large amounts of energy. In plans and operator statements, power-to-power efficiencies of around 70 percent are cited.

  • Isothermal or nearly isothermal concepts: The air temperature is meant to remain as constant as possible during compression and expansion. This is technically demanding because heat has to be transferred very quickly. Such concepts are predominantly at the development or demonstration stage.

The designation CAES therefore does not yet reveal which climate impact, efficiency or mode of operation applies. A compressed air storage plant with gas reheating plays a different role in the power system than an adiabatic storage plant without combustion.

Example plants

Huntorf near Elsfleth in Lower Saxony went into operation in December 1978 as the world's first compressed air energy storage plant. Today it is operated by Uniper. The plant forces air into two salt caverns. When needed, it burns the compressed air together with natural gas in a gas turbine. According to Uniper, the turbine reaches its full output of 321 megawatts within twelve minutes. Originally the output was about 290 megawatts. The compressor capacity is only around 60 to 70 megawatts. Charging therefore takes considerably longer than discharging: about eight hours of compression correspond to roughly two to three hours at full load. Not until 1991 did a second plant of this kind follow, McIntosh in the US state of Alabama. It has an output of 110 megawatts and a considerably longer discharge duration.

Meanwhile, China in particular is driving the technology forward. In January 2025, the Nengchu-1 plant in Yingcheng County, Hubei Province, was connected to the grid at full capacity. It operates without supplementary firing and uses two caverns of a former salt mine. According to the operator, it has an output of 300 megawatts, a storage content of around 1,500 megawatt-hours and a conversion efficiency of about 70 percent. These values come from the operator and have not been independently measured.

Site, storage volume and grid connection

Large compressed air storage plants are worthwhile mainly where large underground cavities are available. Salt caverns play an important role here. They are created deliberately by solution mining, are largely gas-tight and withstand high pressures. In Huntorf, the operating pressure lies roughly between 50 and 70 bar, depending on the fill level. However, the same caverns are also candidates for natural gas or hydrogen. This creates competition for sites: whether a suitable cavity is actually available for CAES depends on geology, permitting, grid connection and other usage plans.

Above-ground pressure tanks are possible. For large amounts of energy, however, materials and costs quickly become a problem. Underground caverns offer large volume but tie the plant to one location. A battery can be built close to a substation, a solar park or a bottleneck in the distribution grid. A large compressed air storage plant, by contrast, follows the geology and must be connected to the power grid, and where applicable also to gas or heat systems. A storage plant in northern Germany can absorb surplus wind power if grid connection and market rules allow it. However, it does not remove a local bottleneck in a distant distribution grid.

Distinction from batteries, hydrogen and long-duration storage

CAES is often described as long-duration storage. This can be true but should be stated more precisely. Long-duration storage can mean hours, days, weeks or entire seasons. Existing compressed air storage plants typically discharge over several hours up to about one day. Longer durations are possible but depend on storage size, pressure level, costs and design. For very fast responses and many cycles, battery storage is often better suited. For seasonal amounts of energy, hydrogen is usually discussed: chemical storage can hold very large amounts of energy for a long time, but loses a lot of energy in conversion.

Efficiency alone is not sufficient for drawing the distinction. Batteries usually achieve higher electrical efficiencies and respond very quickly. They are suitable for frequency maintenance, short-term shifting of solar power and applications close to the grid. Compressed air storage can be cheaper per unit of stored energy at large storage volumes. However, it needs suitable sites and more complex plant technology. Which technology makes sense depends on the task: reserve capacity, daily shifting, prolonged low-wind and low-sun periods (Dunkelflaute), congestion management or firm capacity.

A compressed air storage plant offers flexibility if charging and discharging can be controlled in a way that serves the system. This flexibility has limits: minimum load, start-up time, pressure and temperature state and free storage capacity restrict its use. An empty storage plant cannot deliver power, and a full one can no longer absorb surplus. Storage plants are therefore not an abstract reserve but facilities with a fill level, a location and operating costs.

Economic role and regulation

A compressed air storage plant earns money when the price difference between charging and discharging covers losses, operating and capital costs and levies. For diabatic plants, the fuel and CO2 costs of gas reheating are added. With more wind and solar power, there are more frequent hours with low or negative prices and others with high prices. For capital-intensive plants, however, pure energy trading is often not enough if price peaks are rare or hard to predict. Further revenue sources can be balancing energy, redispatch, capacity mechanisms, grid reserve or contracts with large consumers. Each application follows its own rules, and not every remuneration can be combined with every mode of operation. What is useful for the economy as a whole and what is profitable for a private business do not automatically coincide.

Grid fees, levies and the question of whether a storage plant is classified as a final consumer or as a generator have a major influence. In Germany, § 118 (6) EnWG exempts new facilities for storing electrical energy that enter operation by 4 August 2029 from grid fees for the stored electricity for 20 years. This applies only to the extent that the electricity is fed back into the same grid with a time delay, and only to grid fees in the narrow sense. Surcharges and concession fees are not covered. The rule applies regardless of the technology, but in practice it is discussed mainly for battery storage. Whether it fits a specific compressed air plant, especially with gas supplementary firing, must be examined case by case. For commissioning after this cut-off date, the Federal Network Agency (Bundesnetzagentur) is developing a new grid fee system. As of November 2026, this had not yet been finalized.

Typical misunderstandings

Misunderstanding 1: Compressed air storage can absorb electricity surpluses almost without limit. In fact, compressor capacity, fill level, pressure range and site limit the intake. In Huntorf, the charging power is only about one fifth of the discharge power. Misunderstanding 2: Every stored kilowatt-hour is worth the same. A kilowatt-hour from a windy night that is delivered the next evening has a different effect on the system than one that is needed only after several days without wind. The time of availability is part of the value.

Misunderstanding 3: Storage capacity means security of supply. A compressed air storage plant contributes to it if it can be called upon in critical hours and has been charged beforehand. However, it replaces neither the planning of firm capacity nor grid expansion or load management. For the residual load, that is, electricity demand after deducting wind and solar feed-in, CAES can be an important balancing option. How strong its effect is depends on how high and how long the residual load peaks are.

CAES is one possible answer to the growing need for shifting energy over time, but not a universal solution. The term denotes a family of storage technologies. What benefit a plant actually provides emerges only from its heat management, site, storage duration, market rules and grid function.

Sources

Related glossary entries

Edited by eCMS on 11.11.2026 16:42

© 2026 Erik Schmidt. All rights reserved. Quotations with a link to the source are welcome.