Energy Storage News

Isothermal Compressed Air Energy Storage

A number of industrial processes (e.g. steel production, air separation, oil refining, nitric acid production, ethylene oxide production, etc) which produce high pressure gas at high temperature use expansion turbines to extract energy from the gas either in the form of mechanical work or of electrical current. This fact suggests the idea of using compressed gas as an energy storage medium. Obviously the cost of the expansion turbine itself is not prohibitive or this technology would not be utilized in existing industrial processes. However, other economic consideration enter the equation when one considers the deliberate compression and storage of gas rather than utilizing compressed gas that is the natural product of some other industrial process. Among these considerations are:

  1. Typically expansion turbines are not reversible so that a separate compression train is required thus raising capital costs.
  2. A purpose build pressure vessel is required rather than utilizing some structure which already exists as part of some industrial process, again raising capital costs.
  3. If electricity is being used to run the compressors then one must pay for the difference between the electrical energy consumed during the compression process and the electrical energy produced during the expansion process. This electrical efficiency charge is given by: Generation Cost (1-η)/η where η is the round trip efficiency (electrical energy out/electrical energy in). For example if η= 0.7 and you are storing offshore wind energy generated at a cost of $0.10/kWh then the efficiency charge is $0.10(1-0.7)/0.7=$.043/kWh.
  4. If the gas is at ambient temperature, then as it expands it cools down sufficiently to produce water and ice which interferes with proper operation of the expansion turbine. This problem can be overcome by injecting heat into the expanding gas, but, in this case, an additional source of energy other than the compressed gas is needed.

Compressed air can be stored in steel containers at 3000lbs/square inch. To get an idea of what sort in energy density is achievable at this pressure consider the reversible isothermal compression of an ideal gas. If the temperature of the gas is 20C then change in the free energy in this case is given by:

ΔF = nRTln(3000/14.7)

Where n is the number of moles of gas, R is the gas constant, and T is the absolute temperature (i.e. degrees above absolute zero). For one mole of gas ΔF is given by:

ΔF = 8.3145293.155.319=12,964Joules=3.6 Watt hours

In order to translate the expression into a volumetric energy density we need to know the volume occupied by an ideal gas at 3000lbs/square inch at ambient temperature (=20C). In order to determine this number we use the idea gas law:


In the MKS system of units R=8.3145J/mol K, T=293.15K (=20C), P=3000 6895=2.069107 Pa. Therefore the volume in cubic meters is given by

V = 8.3145293.15/2.069107= 1.1810-4 cubic meters = 0.118 liters

The energy density is therefore given by:

Density= 3.6Wh/0.118l=30.5Wh/l

The energy density of a real compressed air storage system will be less than this value, since a pefectly reversible isothermal process can be realized in practice. This energy density is not particularly high (Lead acid batteries have an energy density of 40Wh/l and lithium ion batteries have energy densities over 180Wh/l), but it is not so low that one would a priori reject this method of energy storage as being impractical.

Since the molecular weight of air is approximately 0.029Kg/mol the gravimetric energy density is given by:


One can use the above number to do a simple scale calculation concerning the use of compressed air as an energy storage medium. Even though compressed air energy storage does not use up the earth's atmosphere, it does hold a certain portion of it in storage containers where it is not available for use. The current US electrical energy usage is approximately 12,000kWh per person. Let us suppose that we want to store 1 days worth of this electricity for each of 9 billion people. What percentage of the earth's atmosphere is required? I will assume a practical storage density of 100Wh/kg. In this case the required mass of air is given by:

Mass = [9,000,000,000(12,000,000/365)]/100 =2.961012kg

Three trillion kilograms may sound like a lot of air, but since the mass of the atmosphere is 5E18Kg the percentage stored is extremely small. Even six months worth of storage would represent a tiny fraction of percent of the total atmosphere.

A potentially huge economic advantage of compressed air as an energy storage medium is that, unlike battery electrodes and electrolytes it does not have to be manufactured out of expensive, hard to process material, and it does not wear out with repeated use. Of course one has to pay for the container (the steel tank, the under ground chamber, the underwater airbag, etc.) but the volume of the container is filled with a free, universally available substance which will not wear out with use.

In this context it is interesting to compare compressed air with water in elevated reservoirs in pumped hydro storage systems. On liter of water weighs 1kg. Thus in a pumped hydro storage system with a head of 1000 meters the energy density is (1kg/l1000m9.8m/sec2)/(3600Wh/Joule)=2.7Wh/l. Thus the energy density of a pumped hydro storage system is an order of magnitude lower than that of a compressed air storage system. If a larger natural formation with a high head exists the cost of creating a 'container' by building a damn may be relatively cheap in spite of the low energy density. However, if an excavated underground reservoir is used then the higher energy density of compressed air is a potential economic advantage.

So the question arises: If compressed air has potential advantages as a energy storage medium, why hasn't it been tried? The answer is that has been tried twice: A compressed air storage facility went on line in Huntorf, Germany in 1978, and a second facility went on line McIntosh, Alabama in 1991. Both are still in operation. Both of these facilities used solution mined underground salt domes as storage containers. Both of the facilities use air compressors which are far from isothermal. State of the art commercial air compressors typically operate at about 400C. The air is not stored at this temperature but is allow to to cool down to ambient temperature. Such a cooling process is inherently irreversible (i.e. It involves an increase of entropy and a decrease of free energy.), and so cannot in principle approach the efficiency of a reversible isothermal compression.

The problem of water condensation during cooling is solved by combusting the compressed air with natural gas in a gas turbine. Gas turbines normally consume a substantial part of their own energy output compressing air prior to combustion. A pre-existing supply of compressed air thus reduces the required fuel input per kWh of electrical output. Energy Storage And Power the company founder by Dr. Michael Nakhamkin, who was the leader of the McIntosh CAES project, claims that second generation CAES designs utilizing natural gas turbines can obtain 60% to 70% of their output energy from compressed air.

Obviously these gas turbine/compressed air systems do not represent pure energy storage, so that in the long term they would require either bio-mass based or synthetic fuel for their continued operation. In the short term, however, natural gas supplies are still relatively abundant, so why haven't any more of these plants been built in the last twenty years? One would have to guess that, in spite of the operation success of these plants the economics are not attractive. Some combination of low efficiency, high cost, and high risk associated with the creation of the underground storage chambers seems to keep investors away from this technology. Low efficiency, of course, is a potential economic killer, no matter how cheap the capital cost. I have not been able to find any definite information on the round trip efficiency of these plants. In this posting concerning the operation of the Huntorf CAES plant the following numbers are given:

Compressor Operation 60MW for ≤ 12 hours
turbine operation 290MW for ≤ 3 hours

If we take the upper end of the operation time ranges for the compression and expansion phases we find:

Input Energy 60MW12 hours = 720MWh
Output Energy 290MW 3 hours = 870MWh

In order to know the storage efficiency we would need to know what percentage of the output energy comes from compressed air and what percentage comes from natural gas. If we assume that 65% (the middle of the range for second generation CAES plants as estimated by Energy Storage and Power) we find:

Efficiency = 0.65870/720=78%

However, the Huntorf plant is a first generation CAES design so the 65% assumption for the contribution of compressed air is probably optimistic.

The company, Energy Storage and Power, whose director/chief technology officer, Dr. Michael Nakhamkin, was instrumental in the design and building of the McIntosh CAES plant, is marketing second generation CAES plants based on gas turbine designs. Their plan is to design CAES plants using industry standard components (expansion turbines, gas turbines, compressors) rather than using custom build equipment as was done in the case of the McIntosh CAES plant. They are hoping to hold down development costs by this methodology, but they do not appear to have any customers.

Dresser-Rand Corporation who supplied components for the McIntosh CAES plant is also attempting to market turbine based CAES technology under the brand name SmartCAESTM. An engineering company called Chamisa Energy is planning to design and build CAES systems using the Dresser-Rand technology. They have located a site in the Texas Panhandle with appropriate salt dome geology and there is talk of groundbreaking in late 2012 and completion in 2014. However, it is not clear to me that a firm commitment to build this project exists at present.

Considering that three and a half decades have passed since the first gas turbine CAES plant went online the industry response to this technology seems underwhelming.

Recently several companies have started working on a new version of CAES technology which in which the compression process is approximately isothermal and which stores the thermal energy which flows out to the air during compression and returns it to the air during expansion, thus eliminating the need for an external fuel source. Some brief comments about the these companies are given below.

Lightsail Energy

Lightsail Energy is developing an energy storage system in which air is compressed/expanded using a piston/cylinder arrangement. A spray of water is injected into the cylinder during the compression stroke to remove heat from the air and keep it temperature approximately constant. The removed heat passes through a heat exchanger into some kind of thermal storage. When power production is needed the same piston/cylinder system is run in reverse using the compressed air as the driving force. Heat passes back through the heat exchanger from the thermal storage into water which is again sprayed into the cylinder this time delivering heat to the expanding air.

Some further interesting facts about their design are listed below:

Final Temperature Difference <10°C
Piston Compressor/Expander RPM 1200
Round Trip Efficiency 70%

The round trip efficiency of 70% is not outstanding compared to the most efficient kinds of batteries, but is similar to the efficiency of flow batteries which are often touted as potentially cheap form of electricity storage. The 70% the storage efficiency implies a penalty of 43% of the cost of electrical generation. That 1.43kWh of electricity must be generated in order to deliver 1kWh from your storage system.

Lightsail Energy emphasizes the RMP's (1200) achieved by their design because the power to weight ratio is an important parameter. Most energy storage system are also power delivery systems, and the cost per kW of power is important in addition to the cost per kWh of storage. If mountains of machined metal are required to deliver a decent level of power, then even high round trip efficiency and low storage cost may not be sufficient to deliver economic competitiveness. The fact the same machine can be used for both compression and expansion is a potential advantage of this system over turbine style gas expanders which are not reversible. However, it is overall cost that matters, and since this a completely new power generation system we can have no a-priori knowledge of what the ultimate costs are likely to be.

Lightsail is planning to build storage systems using steel pipe as the storage vessel. These storage vessels will not be as cheap as the solution mined salt dome storage used for the Huntorf and McIntosh CAES plants but they will allow maximum flexibility in siting. Even with this form of storage Lightsail believes that they can compete in the market for diesel backup generators which is one of the most expensive forms of fossil fuel generation. They also hope to eventually compete against gas peaker plants with more advanced versions of their design. Under ground storage is an option if appropriate locations are found.

Their projected costs for a second generation design with underground storage appears to be (I read the number off a graph) about $0.13kW/h. I am not sure whether this number includes the efficiency penalty.

This technology is still in the development phase, and how close Lightsail may be to real world installations is not clear.


SustainX is developing a system of approximately isothermal compression and expansion which is conceptually the same as that of Lightsail Energy. Air is compressed/expanded using pistons in pneumatic cylinders into which water is sprayed to remove/add energy in order to maintain the air at an approximately constant temperature. The same cylinder/piston system is used for both compression and expansion. They specifically describe this process as a mixed phase (water in air) heat transfer process. They do not give any numbers describing their machine's performance. SustainX's storage systems are also in the development phase.

General Compression

General Compression claims to have made patent pending advancements in the field of isothermal compression and expansion and are planning to use them to develop utility scale energy storage systems. They give no details on the nature of these advancements, and they have not yet built any demonstration storage systems.

Oscomp Systems

Oscomp is not an energy storage company, but they claim to have made a significant advance in the field of isothermal compression of gases as described in this paper. They describe their technology as rotary multi-phase near isothermal cooling. They also use the terminology 'liquid injection cooling with fast heat transfer rates'. The fact that they are using a rotary compressor is something new, but the key innovation of liquid injection cooling sounds similar to the technology of Lightsail and SustainX.

Although this posting mentions compressed air energy storage as a possible application of the Oscomp's compressor, their primary business strategy is to get into the market for natural gas compressors. They claim that they can achieve a 60:1 compression ratio in a single stage compared to a 10:1 ratio for the best current commercial compressors. They believe that they can cut the costs of natural gas compression in half and enable continued production from wells which are not currently economic.

If you are worried about global warming, a technology which helps extend supplies of natural gas is probably not the best news in the world. However, an existing market could lead to rapid development of this technology which could then find used in other applications.

The fact that three different groups (and possibly four) have developed mixed-phased isothermal compression at about the same time is interesting. Perhaps this is a technology whose time has come. Whether or not it can penetrate significantly into the energy storage market remains to be seen.

May 13, 2012

Energy Storage News

rogerkb [at] energystoragenews [dot] com