The first CAES merchant21 September 2001
According to its developers, the 2700 MWe compressed air energy storage (CAES) project planned for an abandoned limestone mine in Norton, Ohio, will bring a range of potential benefits.
Capacity margins in the East Central Area Reliability Council Region (ECAR) are projected to decline to very low levels over the next few years, down to less than 2 per cent in 2008, even allowing for the 6.5 GW of new capacity currently planned or under construction. Following the passage of the Federal Energy Policy Act of 1992 creating “exempt wholesale generators”, opening up of transmission lines to competitors and the advent of restructuring and competition, Ohio public utilities, like those elsewhere, have ceased constructing new power plants and divested existing generation capacity. The resulting capacity void is being filled by merchant facilities, and the proposed Norton CAES plant can be regarded as one of these.
The primary purpose of the Norton Energy Storage project is to generate electricity for peak and intermediate periods, usually daytime hours, for sale into the wholesale electricity market in Ohio and the ECAR market. But, as well as helping to contribute to the expected capacity shortage within ECAR, the project will also be able to provide a range of services to the bulk power market. These include commercial services that will be in demand as the unregulated marketplace for electricity develops, as well as conventional services for system control and stability, including: short-term electricity storage; spinning reserve; “ten-minute” reserve capacity; load following; frequency regulation; reactive power (VARs); and periodic electricity storage (daily, weekly etc).
A major impetus for development of CAES is the general movement towards deregulation or at least towards cost transparency in pricing. Under most regulatory structures retail price is the same around the clock and throughout the year. This masks the fact that the cost to generate electricity varies significantly with the load, the kind and number of units in operation, the fuels used, the availability of hydropower, the skill of the various operators.
This means that electricity should cost less to generate at night if a reasonable cost-based unit dispatch is done, but this is not reflected in regulated prices charged to consumers. By allowing consumers to “see” the costs of electricity, many jurisdictions are encouraging the consumers to adapt their consumption behaviours to use greater quantities of cheaper off-peak electricity that, by definition, did not exist in the price structures of the regulated world. Of course, this shifted usage reduces peak period load, which reduces costs during peak periods. The ultimate result is a more even load, at lower demand levels and lower average costs to the benefit of all consumers – at least until growth catches up with prior demand levels.
CAES can influence these dynamics profoundly for a given generator, especially one who is long on base load capacity like coal and nuclear and short on swing capacity like combustion turbine units.
Cycling large coal units can be very expensive because of effects on fuel rates, emissions and capital maintenance requirements. An operator using coal units to serve widely fluctuating units may operate them between full and 50 per cent loading. By storing off-peak output instead of turning down the units, the operator may enjoy lower operating cost, lower total emissions, fewer forced outages, better system control and regulation; he may also be able to idle some units or place them on standby reserve or retired reserve rather than forcing them into service for which they were not designed. The logic was conceived to enable nuclear units to avoid turndown by storing short-term surplus electricity in pumped hydro storage units. The same theories can be applied to using CAES to support a base load fleet, but CAES can offer lower capital cost, greater operating flexibility, less destructive use of the land (or use of much less land), shorter construction time, and a very valuable energy-cost management tool. So CAES can be seen as a tool to manage the timing value of electricity within diurnal and weekly cycles.
Merchant storage project
The Norton plant will be the first compressed air storage facility developed in the USA as a merchant facility, and only the third such compressed air storage facility in the world. There is a 110 MWe compressed air energy plant in McIntosh, Alabama, and another at Huntorf in Germany, rated at 290 MWe. Both these existing facilities use salt caverns. The proposed Norton plant will be the first to use a limestone mine.
The Ohio project is being developed by Norton Energy Storage LLC (NES), which is also responsible for design, construction and operation. NES, a Delaware limited liability company, is a wholly owned subsidiary of CAES Development Company LLC. CAES Development Company, whose mission is to develop air storage based systems to bring system wide value and flexible services to bulk power markets, is in turn owned by its management team and by two limited partnerships formed by Haddington Ventures LLC, a provider of private equity capital to the energy industry and an affiliate of JP Morgan Partners. The principal limited partners in the two LPs include the venture capital affiliates of major banking, insurance and utility companies.
NES signed an agreement with the city of Norton in July 2000 to co-operate on building the plant. In March 2001 the Ohio Power Siting Board issued a report recommending approval of authorisation to build the plant. In August 2001 Norton Energy Storage received its air emissions permit.Other permitting is underway.
Cost estimates for the scheme have not been divulged but the following ranges have been given for capital costs: EPC, $700-1500 million; interconnection, $15-75 million; project development, $25-75 million; total, $750-1650 million. This suggests a cost per kW in the range $275-600. The expected non-fuel O&M cost per year will be around $20-30 million. The units will have all the elements of a combined cycle generating system except the steam and water cycle, plus the storage cavern. Costs are expected to be higher than those for a simple cycle combustion turbine, but lower than those for a new high-efficiency combined cycle unit.
NES will operate by accepting electricity from clients and by means of large electric-motor-driven compressors will store energy in the form of compressed air in the mine. The compression is done during periods of off-peak power demand. As part of the compression process, the air is cooled prior to injection to make the best possible use of the storage space available.
To return electricity to the customers, the compressed air is released through combustion turbines, essentially substituting for the compressed air that is produced by the compressor that is normally integral to a combustion turbine.
In the Norton facility the pre-compressed air is first heated in recuperators by waste heat from the combustion turbine exhaust and some duct firing, expanded across an air turbine and then mixed with fuel and burnt in the combustion chamber of the combustion turbines.
The two turbine bodies, air turbine and combustion turbine, are on a single shaft and both drive a single generator mounted on the same shaft. Unlike conventional combustion turbines, the turbines do not drive compressors. Therefore the turbine work that conventionally would drive the compressor in a combustion turbine is directed to the generator, increasing the output. In fact the power output from each shaft will be nearly three times the capacity that a comparable conventional simple cycle combustion turbine could produce.
One incidental advantage of not having an integral compressor is that the vibrations combustion turbines experience from their compressors is damped completely by the cavern, which effectively serves as a huge surge tank.
The emissions from a CAES turbine will be about the same as they would be for the same combustion turbine in conventional service, assuming both are equipped similarly with selective catalytic reduction. However, the power output in MW from each CAES train is substantially greater because the compression is accomplished when the turbine is not operating. Therefore, the rate of air emissions at full load by the proposed 2700 MWe of capacity will be comparable to emissions from 900 MWe of simple cycle combustion turbine capacity, or around 1300 MWe of combined cycle capacity. In the Norton CAES facility the heat rate is expected to be about 4320 Btu/kWh (HHV at full load) and the electric motors driving the compressors will use about 0.7 kWh per kWh deliverable to market.
The amount of heat lost in cooling the air is significant, but it would be prohibitively expensive to recover and use that low-grade heat. When the energy as electricity used for compression is added to the fuel energy, the cycle efficiency is comparable to that of a combined cycle plant using a combustion turbine of similar design and vintage.
In the envisaged scheme, about 80 per cent of the plant’s output is generated by the combustion turbine. An air-turbine-only option could be considered but the pressure/temperature conditions of the expanding air would be problematic. If the pressure was high enough to achieve a commercially meaningful output, then the temperature of the air at discharge is likely to by far lower than the materials and connections will be able to tolerate; this would result in embrittlement of materials, seal failure, and perhaps leakage from differential expansion. At the other extreme, it may be possible to design a small air turbine that would be serviceable, but it would probably require expensive, exotic materials, innovative engineering and research that, together, would be very costly. Instead, the approach adopted in the Norton project has been to use proven equipment and technology in the interests of cost, efficiency, risk minimisation, known performance, and easier regulatory approvals.
The main components
The project will use nine modified Alstom ET-11NM combustion turbines, each driving a generator nominally rated at 300 MWe. To serve the peak period market with electricity for 16 hours per day, five days per week, each Alstom gas turbine unit will require about 200 MW of compressor driver capacity.
Each of the nine combustion turbine expander units will include an air turbine unit, combustion turbine section, acoustic enclosures, recuperator, exhaust system including fuel stack, fuel system, starting system, lubrication and hydraulic systems, fire protection system, and associated instrumentation & control. The recuperator will include selective catalytic reduction (SCR) technology to control NOx emissions. The combination of dry-low NOx combustors with the SCR system should reduce NOx emissions to 3.5 ppmvd at 15 per cent oxygen.
The inlet of each air turbine section will be connected to the storage mine injection/withdrawal wells and compressors via high pressure piping.
Nine main power step up transformers will be installed at the Norton CAES facility, one for each combustion turbine expander and generator set. The transformers will step up the generator voltage from 21 kV at the generator to either 138 kV or 345 kV, as required, to allow the facility to interconnect with the FirstEnergy transmission system. Existing 138 kV transmission lines traverse the south and east side of the proposed site, which is in Summit County, while First Energy’s 345 kV Star substation is about three miles to the west.
Each of the nine expander-generator units will be served by two compressor trains, giving a total of eighteen trains. Each compressor train has two compressors, each with its own synchronous motor rated at 56 MVA with a 0.95 power factor, giving a load of 100 MWe per train. One compressor in each train is high pressure, the other low and intermediate pressure. Each train includes an inlet air system with filter, a low and intermediate pressure stage with motor a separate high pressure compressor stage with motor, intercoolers, aftercoolers, lubrication and hydraulic systems and instrumentation and control systems.
Each of the nine units will have four step down transformers, from 21 kV to motor voltage (one for each compressor motor, 36 in total). Two solid state variable frequency drives will be provided for starting the compressor motors.
Mining operations began at Norton in the early 1940s. It was operated by the Pittsburgh Plate Glass Company as a source of high grade calcium carbonate (calcite) ore for the production of synthetic soda ash, which is used in glass manufacturing. The rock was later used for cement production. Mining ceased in 1976. The former owner maintained access to the mine from 1976 to 1999. Norton Energy Storage currently owns and maintains access to the mine.
The mine was developed by the systematic removal of limestone at a depth of about 2200 ft below the surface. The removal of ore left a cavity with a volume of about 338 million ft3 spread over about 540 acres. Two shafts, constructed in 1942, were used to access the mine. One for production and one for personnel and utility services. At the surface, an area of about 92 acres had served the mining operations.
The mine was developed laterally from the shaft areas using a system of rooms and pillars, resulting in a very stable structure. Despite being well below the water table, the mine is virtually dry.
Sandia National Laboratories and The Hydrodynamics Group LLC were contracted to characterise the mine as an air storage vessel and evaluate its expected performance. A Sandia team spent six months, November 1999 to April 2000, studying the geology of the mine. It was found to consist of a very dense rock with low permeability – stiff, strong, with few, if any, natural fractures.
The assessment has confirmed the feasibility of using the mine for the compressed air storage power plant and indicates that the stability of the mine will not be compromised by the storage cycles envisaged. Working pressures in the mine are expected to be in the range 800-1600 psi.
The mine will be converted to air storage by constructing conventional natural-gas-type wells into the mine for air injection and withdrawal. Each well will be fully cased and each casing will be cemented in place. Each well will consist of a gas wellhead unit, and associated valves, pressure monitoring equipment and piping. The two existing mine shafts will be sealed.
Mechanical draft wet cooling towers will be used to cool water for plant equipment cooling systems for the first two units, and portions of the cooling load for the remaining seven units. Fin fan coolers, using air instead of water as the cooling medium, will be used for most of the cooling for the last seven units.
Twelve existing concrete silos on the site (capacity over 10 million gallons) will be converted for use in storing make-up cooling water and fire system water.
5 x 16 operation
The project is anticipated to generate power on a “5x16 schedule”, ie 5 days a week for 16 hours a day (80 hours maximum per week for 2000 to 4160 hours annually), resulting in an annual capacity factor of 23 to 48 per cent. The stand-alone electric powered compressors will recharge the storage mine in the compression mode for the intervening off-peak 8 hours plus weekends (generally up to 88 hours per week).
The facility will be designed to operate on natural gas only; no fuel oil will be used. The fuel will be supplied by Dominion East Ohio Gas. Fuel gas compressors will be provided to stabilise the supply pressure to the level required for introduction into the combustor section.
The plan is to construct the Norton CAES facility in increments of 300 MWe, with the first 300 MWe block entering commercial operation in mid 2003, in time to meet the summer peak load.