Energy storage: CAES ready to go main stream

Ask most experienced electric-generation professionals about the value proposition offered by compressed-air energy storage (CAES) and you’ll likely get a yawn. They probably will recall the lone US plant (there only are two in the world), installed nearly 20 years ago in McIntosh, Ala, to prove the concept’s viability, and then add something like “it works, so what?”

Such a response may have been warranted because CAES had no compelling economic justification until the recent addition to the generation mix of a critical mass of intermittent renewables. Bob Kraft, who has spent his career designing and improving gas turbines and their component parts, told the group that in a world demanding ever smaller carbon footprints CAES can play a significant role by maximizing the value of wind and solar resources at an affordable cost.

Pumped-storage hydro (PSH) can achieve the same result, he acknowledged, but its penetration is limited by environmentally driven siting constraints and significantly higher cost of development. Kraft also noted that the price of a CAES system does not increase with size as quickly as it does for a PSH facility. The example he offered: To double the megawatt-hour storage capacity of a PSH plant you have to double the size of the reservoirs, a major budget item; you also must double the size of the CAES reservoir, but it is only about 15% of project cost.

One of the most important points Kraft made during his presentation was that the faster the electric system can respond to the ups and downs in load given the inherently variable generation characteristic of wind and solar resources, the less capacity you need to back up and smoothly integrate renewables. That piqued the interest of many in attendance who had listened over the last year or so to gas-turbine OEMs touting the need for essentially 1 MW of fast-response GT capacity (or spinning reserve) for every megawatt of wind installed.

Kraft figures CAES can back up intermittent renewables with less than half the capacity (megawatt rating) that would be required if peaking GTs were selected to provide the same service. Plus, much of the CAES capacity could come from underutilized conventional gas turbines converted to energy-storage assets. Energy Storage & Power LLC, the company Kraft manages, owned in part by PSEG Global LLC, is working on several major CAES projects—one involving the retrofit of a GE 7B engine for energy-storage service.

Note that CAES responds very quickly to a change in system demand because the expander turbine operating on compressed air from storage can ramp at rates that could top 25 MW/sec. Ramp rates for gas turbines generally are about 30 MW/min today. For a warm start, the expander turbine can reach full capacity in about three minutes.

First-generation CAES. To illustrate the recent advancements in CAES technology, Kraft showed a diagram of the McIntosh system (Fig 13). The purpose of this facility was to optimize base-load coal and nuclear generating plants and increasing peaking capacity. Important points to remember:

• McIntosh has a single compressor train that puts 1300-psig air in a storage cavern solution-mined from a salt formation 1500 ft under ground. The size of the 22-million ft3 cavern is roughly equivalent to the volume of a 20-story building occupying a typical city block. The cavern is relatively small compared to similar facilities used for storing natural gas.
• In charging operation, the motor/generator is disconnected from the expander turbines by opening the right-hand SSS clutch shown in the diagram and engaging the other clutch. The m/g operates as a motor, consuming 0.81 kWh of grid power for every kilowatt-hour of power generated when air is released to the expanders.
• When fully charged, this cavern can support 26 hours of continuous power-train operation at rated output. Heat rate is 4100 Btu/kWh.

ES&P’s CAES2. Kraft introduced his company’s CAES2 system with Fig 14, showing how a new or existing simple-cycle gas turbine could be transitioned to energy-storage service. Using gas turbines of different sizes, the cycle shown is scalable and adaptable to a range of conditions, with a unit producing from 5 to 450 MW. For applications requiring from 5 to 20 MW of CAES, compressed-air storage is practical in an above-ground vessel; from 20 to 450 MW, storage would be in an underground cavern. Kraft stressed that the CAES2 uses only proven technology and equipment.

Key points of this design include the following:

• Multiple compressors, each equipped with its own intercoolers, arranged in parallel for charging the storage vessel or cavern. This improves flexibility for renewables integration and enables simultaneous charging and discharging of the cavern.
• A degree of power augmentation is achieved by extracting some compressed air after it does work in the first couple of expander stages and injecting this bleed air into the combustion wrapper. Also, keep in mind that the expander exhaust generally is below ambient temperature, usually about 50F. Alternatively, this cold exhaust air can be routed to the GT inlet to maintain nameplate performance levels on hot days.
• Emissions from CAES2 are minimal because the only fuel that must be burned is that in the gas turbine, which is equipped with a DLE (dry, low emissions) combustion system. GT exhaust heat is transferred to air discharged from the cavern in a recuperator before it enters the expander, eliminating the need for the external combustor shown in Fig BK1.

The CAES2 described in the drawing for a Frame 7B-E engine produces 172 MW at a heat rate of 3771 Btu/kWh. It requires only 0.71 kWh to produce a kilowatt-hour of power for the grid. Performance is better than for first-generation CAES primarily because of the recuperator.

Next, Kraft discussed the value proposition for converting to CAES a 2 x 1 F-class combined cycle seeing limited service (Fig 15). Here the 2 x 1 combined cycle is repurposed as a 1 x 1 unit with the CAES expanders coupled to the second 7FA gas turbine. Output of the 1 x 1 portion of the project analyzed would develop 238 MW without duct burners in operation, 268 MW with supplementary firing of the heat-recovery steam generator (HRSG).

Total output of the CAES portion of the plant is 388 MW—a production increase of 150 MW compared to the conventional plant with no duct firing. Kraft estimated the capital cost of the additional power at about $720/kW—including the new expander and compressor trains, conversion of the HRSG to a recuperator, construction of an underground storage cavern, and balance-of-plant requirements.

Kraft closed out the CAES portion of his presentation by comparing the cost of bulk energy storage alternatives today. Were a new 7FA-powered CAES2 plant developed from the ground up rather than converting an existing unit for energy-storage service, it would cost upwards of $950/kW—or about a third more than converting an existing plant. A new PSH facility would cost from about $2500 to $4000/kW.

Lithium-ion batteries were mentioned as an alternative to PSH and CAES2, but considered uneconomic in sizes above about 20 MW and for discharge times of more than one hour at the current stage of development. Cost estimate is $1500/kW.

Power augmentation. Kraft addressed power augmentation in the last part of his presentation—specifically ES&P’s patented humid air injection (HAI) system for combined-cycle plants (access www.ccj-online.com/archives.html, click Spring 2004, click “Water injection a concern? on cover). The system promises cold-day performance on a hot day for a nominal $300/kW depending on the specific powerplant arrangement, ambient environment, etc.

Simply put, HAI involves the injection of a steam/air mixture into the GT compressor discharge just ahead of the combustor. A standard motor-driven compressor is installed to supply the compressed air for this purpose. The extraction point for steam preferred by thermal engineers is the cold reheat line because it is thought to provide the best combination of power augmentation and heat rate.

The first commercial project is planned for a 2 x 1 7FA-powered combined cycle at PSEG Fossil LLC’s Bergen Generating Station. It is expected to deliver a 50-MW net increase from the Bergen unit on a 95F day. Output of each GT increases by about 35 MW, but the steam turbine output drops by about 15 MW because less steam flows through it, and the air compressors increase the parasitic power draw.