The case for CAES

Storage allows energy production to be decoupled from its supply. By having large-scale electricity storage available, system planners would need to build only sufficient generating capacity to meet average electricity demand, rather than peak demands. In theory, a typical plant could operate with 40 per cent less generating capacity than would otherwise be required. This could achieve considerable financial savings by avoiding the need to invest in peaking and intermediate plants. Other benefits derive from the generators being able to operate more efficiently because they do not need to change output. The result is reduced emissions and capital investment.

The benefits are potentially huge. According to the newly formed Energy Storage Council, such technologies could have "a $175 billion positive economic impact on the US economy" over the next 15 years."

A proven technology

The bagpipes, originating at least 2500 years ago, can be regarded as the first known application of compressed air energy storage. But the first major application in the power industry was the 290 MW Huntorf facility in Germany, installed in 1975-78. The second was a 110 MWe plant, built at McIntosh, Alabama, in 1991. Both of these use salt caverns for their underground air reservoirs. Salt domes or salt strata, as used for natural gas storage, are ideal for CAES and can be found in a large number of regions in the USA, as well in other countries.

Huntorf was originally built to provide black start in support of nuclear/coal units, to cover peaks and to give frequency support, rather than to meet the needs of the electricity market, as envisaged by today's promoters of CAES.

However, experience with Huntorf has served to reinforce the proposition that CAES does indeed serve today's market requirements. For example, the black start function has evolved into a more dynamic role within the total generation resource system, including the accommodation of renewables such as wind.

A 2001 laser survey of the salt solution mined caverns at Huntorf showed that after more than 20 years of operation, there was practically no deviation in contour from the original sonar survey conducted in 1984. Furthermore even at high rates of withdrawal, salt contamination was very low (less than 1 mg salt/kg air). This is confirmed by the absence of hot section corrosion in the turbine.

The benefits of decoupling

Both Huntorf and McIntosh use a single train configuration, with compression and generation on one shaft.

In Huntorf, as in McIntosh, the motor/generator is connected via self-shifting synchronous clutches, ie in the generation mode the compression train is disconnected, and during compression mode the turbine is disconnected by disengaging/engaging the appropriate clutches.

The single shaft compression/generation approach limits the Huntorf facility to a short, four hour, generation cycle, and a long, 16 hour x 60 MW, compression cycle.

This was appropriate at the time Huntorf was designed but today's market requirements are different. For example, energy storage as a commodity (electricity as stored air energy) needs to be delivered when demand increases, to offset volatility in electricity prices, in the same way that natural gas storage has served the gas market.

To meet such requirements, some changes in the machinery arrangement are desirable, especially decoupling of the compression from the main shaft, allowing the flexibility of tailoring the compression (charging or "power absorption") cycle to requirements independently of the generation cycle.

Integrating an advanced gas turbine with air storage has been proposed as one possibility for decoupling compression and expansion, while retaining the gas turbine for simple cycle peaking or emergency use.

In the Alstom multi-shaft CAES concept, the compression is decoupled from the generation (power island) and the compressors are separately motor driven.

With this arrangement:

• the compression can be optimised for standard available frame sizes to suit pressure and volume requirements of the storage facility;

• LP and HP compression can be separated to match motor drives using 4 pole motors of 50MW+ rating, currently in service for compressor drives.

Increased daily generation of 8 hours to 16 hours can be accommodated by adding a duplicate compression train. Multiple smaller compression trains provide "variable" power absorption even during daily generation, to respond to load swings or sudden load rejection etc.

The CAES ET11NM design philosophy is simple: use an available, reliable and proven gas turbine as the main LP expander with its combustion system.

The GT11NM, with more than 500 000 operating hours and 4500 starts, was selected, with an air expander turbine integrated on the same shaft, coupled to either an air cooled or hydrogen cooled standard frame generator in the 300MW class.

The addition of an 85 per cent efficient exhaust heat recuperator with supplemental firing provides additional heat to the air discharged from the cavern storage to precisely match the GT combustor needs after the first expansion in the air turbine.

The standard GT11NM compressor is removed at the turbine housing and replaced with the AT expander and casing. This ensures that the existing turbine operating parameters are maintained, as illustrated in the ET11NM TS diagram.

The Norton project

One of the first potential projects in the USA to benefit from the compression decoupling is the huge planned facility at Norton, Ohio, which is permitted for 2700 MW of capacity, and as a commercial project, will be one of the largest bulk energy storage facilities, including hydro pumped storage schemes, to be built in the USA.

This plant, the Norton Energy Storage project, will consist of 9x300MW nominally rated CAES units, supported by an underground storage cavern volume of 338 million cubic feet 2200 feet below the surface, originally mined in a limestone formation.

The plant will be built modularly with continuous plant construction over a five year period.

A key feature of the project is its ability to help provide reliable full electric power service for midrange and peak hours, extending the capabilities of large, low cost baseload generation.

Figure 9 shows a (nominal) 300MW power island with recuperator, with power delivery stepped up from 21 kV to 138 kV or 345 kV, depending on transmission line requirements.

Using substantially less natural gas fuel than an equivalent sized combined cycle plant the emissions are much lower as well. With Dry Low NOx (DLN) combustors and post combustion catalytic treatment of exhaust gases, emission levels of 3.5 vppm can be achieved for both CAES and combined cycle plants. However, for the same level of emissions, CAES produces 1.73 MW while the CC plant produces 1.0 MW, resulting in lower total emission per MWh.

The Norton limestone cavern can sustain a storage pressure of 1500 psig. The decoupling of compression trains allows for maximum flexibility of energy storage and up to 16 hours of daily generation during the week. The generation/compression cycles are illustrated in Figure 10. A single unit could provide 18 days of continuous generation.

Meeting market needs

The future potential applications of CAES are many and varied, in such areas as: increasing system utilisation, generation deferment, operating reserve and dynamic benefits; regulating output from stochastic renewables, such as wind; load levelling on island supplies and other electrical networks; network reliability improvement and deferment of transmission network reinforcement; reducing network electrical losses and enhancing system stability; distribution network control, voltage and frequency control; supply reliability improvement; and electricity trading.

One particular area demanding more detailed future investigation is the integration of renewables, in particular wind.

Europe is a leader in wind generation, which accounted for 17 000 MWe or about 2 per cent of total electricity produced in 2001. In the USA, installed wind capacity in 2001 almost doubled compared with 2000, rising to 4240 MWe. More large projects are planned in Texas, which alone added 900 MW in 2001, as well in New England and the Northeastern United States.

Regulating the output from such large wind projects is one area where bulk energy storage can really benefit the industry.