Liquid Air Energy Storage

Liquid air energy storage ("LAES") is a thermo-mechanical long-duration energy storage technology that stores electricity by cooling ambient air to its cryogenic liquid state (-196°C), storing it in insulated low-pressure tanks, and later re-gasifying and expanding it through turbines to regenerate electricity. The technology builds on over a century of industrial air liquefaction experience and offers a geographically unconstrained alternative to pumped hydro and compressed air energy storage, with energy densities roughly 4–8 times higher than CAES. LAES was first proposed for grid-scale energy storage by Smith in 1977 and has been developed commercially by Highview Power (UK), which demonstrated the technology at pilot scale (300 kW/2.5 MWh and 5 MW/15 MWh) and is now constructing the world's largest commercial-scale LAES facility at Carrington, Manchester (50 MW/300 MWh, operational by 2026), backed by £377 million in total investment including £70 million from Centrica and funding from the UK Infrastructure Bank. Highview's total UK project pipeline exceeds 7 GWh, with two 3.2 GWh facilities at Hunterston and Killingholme advancing through Ofgem's cap-and-floor regulatory regime. China has commenced construction on a 60 MW/600 MWh LAES demonstration in Qinghai, and a 5 MW demonstration in Japan (with Sumitomo Heavy Industries) is planned for 2025. Academic reviews report standalone LAES round-trip efficiencies of 50–60%, rising to 70–90% for hybrid configurations that integrate external waste heat or cold sources. Capital costs are estimated at 900–1,750 EUR/kW. The system uses no lithium, no rare-earth materials, and only ambient air as its working medium, with a 30–50 year design life and components sourced from mature industrial supply chains.

The liquefaction of air has been an established industrial process since the late 19th century. Carl von Linde developed the first practical air liquefaction system in 1895 using a regenerative cooling cycle, and Georges Claude improved the process in 1902 by introducing a cryogenic turbine (expander) alongside the Joule-Thomson valve, achieving significantly higher efficiency. Early attempts to use liquid air as an energy carrier date to 1899, when a liquid air engine was tested for transportation applications. The concept of using liquid air for large-scale grid energy storage was first proposed by Smith in 1977, who described a system using adiabatic compression and expansion processes at 1,048 K and 85 bar, projecting 72% round-trip efficiency. However, the idea remained largely academic for decades due to the dominance of pumped hydro storage and the absence of a commercial need for alternative long-duration storage. Interest revived in the early 2000s in the UK, driven by research at the University of Leeds led by Professor Yulong Ding and colleagues, who investigated cryogenic energy storage as a means of integrating intermittent renewable generation. This academic work led to the founding of Highview Power Storage (later Highview Power), which built and operated the world's first LAES pilot plant — a 300 kW/2.5 MWh system based on a Linde-Hampson liquefaction cycle, where liquid air was stored in a low-pressure cryogenic tank. The technology was subsequently patented. In 2018, Highview Power, in partnership with waste-to-energy company Viridor, developed a larger 5 MW/15 MWh LAES demonstration plant co-located with a source of waste heat, with an expected lifespan of 30–40 years. This second plant validated the integration of thermal energy storage systems (both hot and cold) to recover compression heat and regasification cold, significantly improving round-trip efficiency. Highview announced plans for its first commercial-scale plant in October 2019: a 50 MW/250 MWh facility at Carrington, Greater Manchester. Construction began in November 2020. The project subsequently grew to 50 MW/300 MWh and secured a transformative £300 million funding round in June 2024, including £70 million from Centrica and backing from the UK Infrastructure Bank, bringing total investment to £377 million. Groundbreaking took place in November 2025, with operations planned for 2026. MAN Energy Solutions was selected to provide the turbomachinery train. Highview subsequently announced plans for four larger 2.5 GWh facilities across the UK by 2035, representing £3 billion in total investment and 2 GW of capacity — approximately 20% of the UK's projected energy storage requirements. Two 3.2 GWh facilities at Hunterston and Killingholme are advancing through Ofgem's cap-and-floor regulatory scheme, with final awards expected in summer 2026. The combined portfolio exceeds 7 GWh. Internationally, Sumitomo Heavy Industries and Hiroshima Gas are developing a 5 MW LAES demonstration in Hatsukaichi, Japan, using waste cold from an adjacent LNG terminal, planned for 2025. China commenced construction of a 60 MW/600 MWh LAES demonstration project in Qinghai in 2023 — which upon completion will be the world's largest LAES facility. In 2024, a LAES project at Shijiazhuang Tiedao University in China achieved grid-connected power generation.

Fundamental Principle

LAES stores electrical energy by converting it into the thermodynamic potential of cryogenically liquefied air. Air is cooled to -196°C, at which point it transitions from gas to liquid, reducing its volume by a factor of approximately 700. This liquid air is stored in insulated low-pressure tanks at near-atmospheric pressure. When electricity is needed, the liquid air is pumped to high pressure, re-gasified (warmed back to gas), and expanded through turbines to drive generators. The system operates as a closed thermodynamic cycle with three distinct phases: charging (liquefaction), storage, and discharging (power recovery).

The energy quality of cryogenic storage is noteworthy: research shows that cryogenic energy storage technologies may have higher energy quality (defined as the ratio of exergy change to enthalpy change) than high-temperature thermal energy storage technologies operating at equivalent distances from ambient conditions. This means that the cold stored at -196°C represents a higher proportion of recoverable useful work per unit of energy stored compared to heat stored at an equivalent temperature above ambient.

Charging Phase: Air Liquefaction

During charging, surplus electricity powers an industrial air liquefaction plant. The process follows established cryogenic engineering principles, with three main liquefaction cycle architectures used in LAES:

Linde-Hampson Cycle: The simplest configuration. Air is compressed to high pressure (typically up to 4–5 MPa), cooled in a heat exchanger by the cold return stream, and then expanded through a Joule-Thomson ("J-T") valve. The J-T expansion cools the air further; if the air is sufficiently pre-cooled, it partially liquefies. The liquid fraction is separated and stored, while the cold gaseous fraction is recycled through the heat exchanger to pre-cool incoming air. This cycle is simple but has the lowest efficiency because the J-T valve expansion produces no work.

Claude Cycle: Proposed by Georges Claude in 1902, this cycle adds a cryogenic turbine (expander) in parallel with the J-T valve. A large fraction of the compressed air is diverted through the expander, which extracts work while cooling the air more efficiently than a J-T valve alone. The Claude cycle achieves substantially higher liquefaction efficiency and is the basis for most modern LAES designs. Studies show the Claude cycle has a superior cost-benefit ratio compared to Linde-Hampson and Collins cycles. A three-turbine Claude cycle (replacing a single cold turbine with three operating at different temperature stages) increased round-trip efficiency from 47% to 57%.

Collins Cycle: A more complex variant with multiple expansion stages, offering higher efficiency but at greater capital cost and operational complexity.

During liquefaction, the compression process generates significant heat. In adiabatic LAES configurations, this heat is captured in a high-grade warm storage ("HGWS") system — typically hot water tanks, packed beds of gravel or ceramic, or other sensible heat storage media — for later use during the discharge phase.

Storage Phase

The liquid air is stored at near-atmospheric pressure in vacuum-insulated cryogenic tanks. These are the same type of vessels used industrially for storing liquid nitrogen, liquid oxygen, and LNG — proven, commercially available equipment. Because liquid air is stored at low pressure, the tanks do not require thick-walled pressure vessel construction, reducing cost. Storage duration is limited only by the thermal insulation quality of the tanks (which determines the rate of boil-off) and can extend from hours to days or longer with minimal losses. The energy is effectively "frozen" in the liquid air until needed.

Discharging Phase: Power Recovery

When electricity is required, the stored liquid air is pumped to high pressure by a cryogenic pump (typically 6–15 MPa). It is then re-gasified by passing through heat exchangers. The high-grade cold released during regasification is captured in a high-grade cold storage ("HGCS") system — a cryogenic packed-bed regenerator — for reuse during the next liquefaction cycle. This cold recovery is critical: recycling the regasification cold to assist the next charge cycle is the single most important factor in achieving acceptable round-trip efficiency.

After initial regasification, the now-gaseous high-pressure air is further heated using the stored compression heat from the HGWS. The hot, high-pressure air is then expanded through a multi-stage turbine train (typically four stages with inter-stage reheating) to generate electricity. The quasi-isothermal expansion — where the air is reheated between each expansion stage using stored thermal energy — maximizes the work output from each stage. The expanded air is vented to the atmosphere at near-ambient conditions.

Thermal Energy Storage Systems

The round-trip efficiency of a LAES system is fundamentally determined by how effectively two thermal stores operate:

High-Grade Cold Storage (HGCS): Stores the extreme cold (~-196°C to -100°C) released during regasification of liquid air in the discharge phase. This cold is recycled to assist the next liquefaction cycle, dramatically reducing the electrical energy required for charging. Packed beds of quartzite pebbles, alumina spheres, or phase-change materials at cryogenic temperatures are used. The HGCS is widely identified as the performance bottleneck of the LAES system.

High-Grade Warm Storage (HGWS): Stores the heat of compression (~100–200°C) generated during the charging liquefaction process. This heat is returned to the air during the discharge phase to boost turbine inlet temperatures and power output. Hot water, thermal oil, or packed-bed stores are used.

Effective thermal integration — minimizing temperature differences across heat exchangers, optimizing the number of compression and expansion stages, and matching the thermal storage capacity to the cycle demands — is the primary lever for improving LAES efficiency.

Hybrid and Integrated Configurations

LAES performance can be substantially improved by integrating external heat or cold sources:

Waste Heat Integration: Co-locating LAES with industrial facilities, power plants, or data centres allows waste heat (which would otherwise be rejected to the atmosphere) to supplement the HGWS, increasing turbine inlet temperature and power output without additional electrical input.

LNG Cold Recovery: Co-locating LAES with an LNG regasification terminal provides a source of extreme cold that can replace or supplement the HGCS. Since LNG is regasified at temperatures comparable to liquid air (-162°C vs. -196°C), the synergy is very strong. Sumitomo's Japan demonstration is designed around this integration.

Organic Rankine Cycle (ORC) Integration: The cold from LAES regasification can drive an ORC using a low-boiling-point working fluid, generating additional electricity from what would otherwise be a waste cold stream.

Performance Characteristics

Academic reviews report the following performance ranges for LAES systems: standalone round-trip efficiency of 50–60% (with some optimized configurations reaching 70%), hybrid round-trip efficiency of 50–90% depending on the quality and availability of external heat/cold sources, capital costs of 900–1,750 EUR/kW, 30–50 year design life, and energy density approximately 4–8 times that of CAES. The Carrington facility is designed for 50 MW output for 6 hours (300 MWh), providing energy for approximately 480,000 homes.

Like CO₂ batteries and CAES, LAES has the property that energy storage capacity (determined by the volume of liquid air tanks) and power output (determined by the turbomachinery size) are partially decoupled. Adding more cryogenic storage tanks is relatively inexpensive, making LAES economically favourable at longer durations (6–12+ hours) compared to lithium-ion, where cost scales nearly linearly with duration.

LAES offers several structural advantages over incumbent energy storage technologies: pumped hydro storage, compressed air energy storage, and lithium-ion batteries.

Geographic Independence

Pumped hydro — which represents approximately 95% of global installed energy storage — requires specific topographical features (elevation differential between two reservoirs) and large water resources, severely limiting deployment sites. First-generation CAES (Huntorf, Germany; McIntosh, USA) requires underground salt caverns or similar geological formations for air storage. LAES has no geographical constraints: liquid air is stored in above-ground cryogenic insulated tanks that can be placed anywhere. This makes LAES deployable at any grid-connected site, including brownfield industrial land, retired power station sites, and urban fringes near demand centres.

Higher Energy Density Than CAES

Liquid air at -196°C and atmospheric pressure has a volume approximately 1/700th of ambient air. This extreme density increase means that a LAES facility stores 4–8 times more energy per unit volume than advanced adiabatic CAES. A study comparing LAES and CAES directly found that while LAES has similar round-trip efficiency, its volumetric energy density is substantially higher. This translates to a smaller physical footprint for equivalent storage capacity, which is particularly valuable for deployment near urban load centres where land is expensive.

Long Duration Without Degradation

Like other thermo-mechanical storage systems, LAES does not suffer from electrochemical degradation. The working medium (air) is free, abundant, and does not degrade with cycling. The rotating equipment (compressors, turbines, heat exchangers) is standard industrial machinery with well-understood maintenance profiles and lifetimes of 30–50 years — far exceeding the 10–15 year useful life of lithium-ion battery systems before significant capacity fade.

No Critical Minerals or Fire Risk

LAES uses only air, steel, and standard industrial materials. It contains no lithium, cobalt, nickel, or rare-earth elements. There is no risk of thermal runaway, no toxic electrolyte, and no fire hazard — significant advantages over lithium-ion battery storage, where fire safety, end-of-life recycling, and supply chain security are growing concerns.

Multi-Service Capability

Beyond simple load shifting (storing off-peak electricity for peak dispatch), LAES can provide frequency regulation, black start capability (restarting the grid after a blackout), synchronous inertia, dynamic voltage support, and short-circuit strength. The Carrington facility is specifically designed to provide both long-duration energy storage and grid stability services from a single zero-carbon asset. LAES can also serve as a source of industrial-grade liquid nitrogen or oxygen as a co-product, and the cryogenic cold from regasification can be used for cold-chain applications (refrigeration, frozen food logistics).

Waste Heat Integration

LAES efficiency improves dramatically when integrated with external sources of waste heat (from industrial processes, power plants, or data centres) or waste cold (from LNG regasification terminals). A hybrid LAES system co-located with an LNG terminal can use the cold energy released during LNG vaporization to assist air liquefaction, substantially reducing electrical input. Sumitomo's Japan LAES demonstration is located adjacent to an LNG terminal for precisely this reason. Academic studies show that hybrid LAES configurations can achieve round-trip efficiencies of 50–90%, compared to 50–60% for standalone systems.

LAES faces several interconnected deployment challenges spanning efficiency, cost, technology maturity, and market competition.

Round-Trip Efficiency

The most significant technical limitation of standalone LAES is its round-trip efficiency ("RTE"). Air liquefaction is an inherently energy-intensive process: producing one kilogram of liquid air requires approximately 0.6–0.75 kWh of electrical energy. Standalone LAES systems achieve RTEs of approximately 50–60%, which is substantially lower than pumped hydro (75–85%) and lithium-ion batteries (85–95%). The primary efficiency losses occur in the liquefaction process (compression, cooling, and Joule-Thomson expansion), where a significant fraction of the electrical input is converted to heat that must be either stored and recovered or rejected. Recovering the cold energy released during regasification (via high-grade cold storage) and the compression heat (via high-grade warm storage) are critical for achieving acceptable efficiencies, and imperfect thermal recovery is the main source of RTE loss. Hybrid configurations that import external waste heat or cold can push RTE to 70–90%, but this introduces dependency on co-located industrial processes.

Cold Thermal Energy Storage Performance

The performance of the high-grade cold storage ("HGCS") system — which stores the extreme cold released during liquid air regasification for reuse during the next liquefaction cycle — is widely identified as the bottleneck limiting standalone LAES efficiency and cost-effectiveness. Cryogenic packed-bed thermal stores must maintain very low temperatures (-196°C to -100°C) with minimal thermal losses over the storage period. The choice of storage material (sensible heat materials like quartzite, alumina, or latent heat materials with cryogenic phase-change) significantly affects performance, and optimizing cold storage design remains an active area of research.

Commercial Maturity and FOAK Risk

Although the fundamental components of LAES (compressors, cryogenic turbines, heat exchangers, insulated tanks) are individually mature industrial technologies, their integration into a complete LAES system optimized for grid-scale energy storage is novel. Highview Power's two demonstration plants (300 kW and 5 MW) validated the concept, but the Carrington facility (50 MW/300 MWh) will be the first full-scale commercial plant when it begins operations in 2026. No LAES plant has yet accumulated years of continuous commercial operation at this scale. The transition from demonstration to bankable commercial technology requires proving reliability, availability, and cost performance at scale.

Cost Competitiveness

Capital costs for LAES are estimated at 900–1,750 EUR/kW, which is substantially higher than lithium-ion battery storage on a per-kW basis. However, LAES economics improve at longer durations (6–10+ hours) because the incremental cost of adding storage capacity (more cryogenic tanks) is relatively low compared to the incremental cost of adding lithium-ion battery modules. A 2024 critical review found payback periods of approximately 20 years for standalone LAES, improving to 3–10 years for hybrid systems with waste heat integration. LAES economics are highly dependent on the spread between off-peak and peak electricity prices, the availability and cost of co-located heat/cold sources, and the regulatory frameworks governing grid services.

Competitive LDES Landscape

LAES competes with a growing field of long-duration energy storage technologies: CO₂ batteries (Energy Dome), iron-air batteries (Form Energy), advanced CAES, gravity storage, flow batteries, and green hydrogen. Each has different efficiency, cost, duration, and siting characteristics. The LDES market is still nascent, and market share will depend on which technologies first achieve bankable, repeatable deployment at competitive costs. Highview's significant head start in engineering scale-up and its UK pipeline (7+ GWh) provide a competitive advantage, but the technology must prove its cost and performance claims through successful Carrington operations.

Scale-Up and Supply Chain

Scaling from the Carrington facility to the multi-GWh Hunterston and Killingholme projects (each 3.2 GWh) requires significant manufacturing and construction capacity. Highview has partnered with MAN Energy Solutions for turbomachinery and secured Centrica as a strategic partner, but building the specialized supply chain for cryogenic tanks, cold stores, and integrated LAES plants at the pace required by the UK's Clean Power 2030 plan remains challenging.