This article discusses the advantage of compressed air energy storage (CAES) system. CAES has been proposed as an alternative to pumped hydro storage for large-scale, bulk energy management. CAES systems typically rely on electrically driven air compressors that pump pressurized air into large underground geological formations such as aquifers and caverns for storage. When the power is needed, turboexpanders connected to generators convert the compressed air back into electrical energy. Like pumped hydro, CAES can be scaled to sizes compatible for supplementing large renewable energy facilities. The lifetime costs for a CAES system can make it work as a means for storing cheap off peak electricity and selling it during peak hours, but capital costs and difficulties finding suitable geological structures have limited the technology’s applications. To make CAES more useful for storing wind-powered electricity, the systems have to become less expensive and have greater flexibility in sitting.
Baseload electricity production is pretty straightforward: You generate enough power to meet the instantaneous demand. And that's easy enough to accomplish using a collection of large thermal generating stations powered with fuels such as coal or nuclear power plants.
But electricity producers have been moving away from providing baseload power. Some of this is due to changes in the industry, where deregulation has made peak power more lucrative than baseload, and so many new small plants have been built to supply that market. Some of it, too, is due to the rise of renewable energy such as solar and especially wind power, which have the benefit of producing no greenhouse gases and can in some cases be cheaper to build.
From the standpoint of the grid operator, however, wind and solar energy have a significant drawback: they supply power intermittently in quantities that can be predicted but not assured. Unlike fuel-based generation, wind and solar cannot be dispatched easily or reliably. A lull in the wind or a passing cloud, for instance, can cut production significantly. And in the case of wind in particular, the time of day when the most power is produced does not synchronize with the greatest demand.
For years, engineers have been pursuing various ideas for getting around the problem of renewable energy intermittency, often by storing the power produced off-peak for release in more advantageous times of day. An inexpensive and efficient energy storage technology would have the potential to transform renewables from an opportunistic source of power into something akin to traditional, dispatchable generation assets.
Unfortunately, energy storage technology has yet to fully meet the needs of grid operators. Some of the largest-scale solutions, such as pumped hydroelectric storage or compressed air energy storage, are limited by the available geography or geology. Banks of batteries or flywheels can be sited with more flexibility, but they are too small in scale to provide bulk power management.
The University of Windsor in Ontario, where we work, has partnered with Hydrostor, a Toronto-based company, to develop a new concept—underwater compressed air energy storage—that may be both widely distributed and capable of grid-scale storage. In 2011, we built and demonstrated a UW-CAES pilot project that showed that air compressed in underwater bags could store and release energy. Based on our results, we think this concept merits further, larger-scale investigation.
Energy storage is one of the most fundamental engineering activities; the flexed bow stores energy that can be transferred to an arrow, and the wound spring stores energy that powers the hands of a clock.
Storing electricity has been problematic because it's not easy to store it in bulk in its original state; capacitors, which hold electric charge on conducting plates, have proven good only at holding charge temporarily and supplying bursts of energy. Electricity at the grid scale has more successfully been stored by transforming it into chemical or mechanical potential energy, and then running the process backward to dispatch power back into the power grid when needed. For instance, rechargeable batteries store electricity in the form of reversible electro-chemical reactions. While the most familiar batteries are small scale, capable of powering cell phones and automobiles, some sodium-sulfur units are large enough to store more than 200 MWh. But that isn’t enough to store more than an hour's output from a mid-size wind farm.
More energy still can be stored by converting electricity into mechanical power, and moving material to create potential energy. Pumped hydropower is one attractive form of mechanical storage: excess electricity is used to pump water into a reservoir, which later releases water to run turbines when the power is needed. Pumped hydro can be efficient and can work on a large scale, but the number of locations where such facilities can be sited is limited.
A potentially more flexible system involves compressed air. Compressed air energy storage has been proposed as an alternative to pumped hydro storage for large-scale, bulk energy management. CAES systems typically rely on electrically driven air compressors that pump pressurized air into large underground geological formations such as aquifers and caverns for storage. When the power is needed, turboexpanders connected to generators convert the compressed air back into electrical energy. (Rather than simply expanding the compressed air, CAES systems that have been deployed in Germany and Alabama supplement the stored energy with natural gas combustion.)
Like pumped hydro, CAES can be scaled to sizes compatible for supplementing large renewable energy facilities. The CAES facility in McIntosh, Ala., for instance, can produce 110 MW for up to 24 hours. Both the McIntosh plant and the one in Huntorf, Germany, have achieved admirable rates for reliability and availability— between 90 and 99 percent for both facilities.
The lifetime costs for a CAES system can make it work as a means for storing cheap off-peak electricity and selling it during peak hours, but capital costs and difficulties finding suitable geological structures (generally caverns within salt formations) have limited the technology's applications. In order to make CAES more useful for storing wind-powered electricity, the systems have to become less expensive and have greater flexibility in siting.
A few years ago, wind power developer Cameron Lewis was looking into ways to store electricity generated by wind turbines at night for use during the day. Not satisfied with the limits of traditional CAES, Lewis had an inspired thought while gazing across a lake: Instead of using the solid walls of an underground cavern, why not use the considerable weight of water in a lake or ocean to hold compressed air under high pressure? This insight became the basis of the concept of underwater compressed air energy storage, or UW-CAES.
Our team at the University of Windsor has been working to flesh the concept out into a full-fledge technology. Like conventional CAES systems, UW-CAES comprises an air compressor, an air storage reservoir, and a turbo-expander. The distinctive feature which sets UW-CAES apart from its conventional equivalents is in the air storage reservoir. UW-CAES uses distensible underwater air accumulators—think of them as large, pleated balloons. Anchored to the bottom of a lakebed or the seafloor, these accumulators would expand or contract in response to the amount of compressed air stored in them. Hydrostatic pressure from the water pressing in from the sides and top would maintain a constant pressure within the accumulators, regardless of how much air is inside. Thanks to this, UW-CAES releases its stored energy at a constant rate, even as the amount of air remaining dwindles.
There is an obvious downside: Since traditional CAES uses rigid geological formations to store the compressed air, it can operate at higher pressures and thus store more energy per unit volume. The pressure for UW-CAES is a simple function of hydrostatic forces, which is determined by depth. Fortunately, suitably deep water (80 m ought to be sufficient) is much more widely available than the salt domes and saline aquifers capable of housing a CAES facility. What's more, many of the world's largest cities are situated on coasts, meaning that UW-CAES systems could be built close to the sources of electricity demand.
We have modeled the system using a three-stage polytropic compression/expansion process with intercooling, using ambient air as the working fluid. When the system is charged—that is, energy is being stored—an electric motor runs the multi-stage air compressor unit to pump pressurized air into the submerged air accumulators. Of course, it's a basic law of physics that part of the energy used to compress air raises its temperature, and if that heat dissipates, it is energy that's lost. To capture that energy, after each compression stage, heat generated by the compression process is extracted from the air and stored in a thermal recovery unit consisting of heat exchangers and a storage reservoir. During system discharge—when electricity is generated—air flows back through the thermal recovery unit to be reheated prior to each expansion stage to prevent freezing and other challenges related to expanding gases. The heated air is then sent through and expands through several turbines, which drive a generator to produce electricity.
An important performance measure of energy storage systems is its round-trip efficiency. Unlike conventional CAES, no outside energy is used to heat the air as it's being expanded. By relying on thermal recovery to heat air in the expansion stage, the UW-CAES can be expected to achieve a round-trip efficiency of greater than 70 percent, according to thermodynamic analyses.
That compares favorably to other large- and grid-scale energy storage systems. Depending on the chemistry, rechargeable batteries have a round-trip efficiency of between 60 and 90 percent. Pumped hydro has a similar, though narrower, range—between 65 and 80 percent. There's no reason why CAES using underground caverns couldn’t have a similar efficiency as the UW-CAES system, as long as they used thermal recovery. In practice, however, underground CAES systems such as the McIntosh facility operate with a round-trip efficiency of around 53 percent.
UW-CAES should not only be just as efficient as existing energy storage technologies, if not more so, but also should have a comparable price. The University of Windsor team has been working mostly on proving the technical aspects of the system, but our industry partner, Hydrostor, has estimated that the capital cost for a full-scale UW-CAES system should be about $350 per kWh.
Although the individual components of the UW-CAES system have long been technologically established, the integrated system being developed by Hydrostor and the University of Windsor is the first of its kind. As such, the combination and adaptive optimization for the UW-CAES system configuration has required extensive field testing. A pilot project was needed to prove that UW-CAES is feasible.
Last year, we set up an experiment in Toronto Harbor to test the concept. We deployed a 1:5 scale, 10-ton lift bag anchored to two nine-ton concrete cylinders at a depth of 30 m below water surface. That's not a bad approximation of the accumulators in an operating UW-CAES system; the concept borrows from underwater lift and salvage balloons.
Because we were mostly interested in studying the charge and discharge cycles, the experimental set-up was kept simple. A compressor on the shore fed compressed air through a 25 mm line to an inlet at the top of the balloon. We ran compressed air into the balloon for about three minutes, raising the pressure from 182 kPa to 195 kPa. We then released the air more slowly, though we experienced a pressure loss due to friction and the pipe layout.
Minor issues aside, the pilot project successfully demonstrated that a system was feasible. Aside from creating a working concept of a UW-CAES system, the pilot project also proved to be a unique exercise in determining system equipment selections with real world constraints. In addition, the project identified topics that have yet to be thoroughly studied, such as fluid dynamic interaction with submerged inflatable structures.
After successfully completing the pilot project, Hydrostor has begun development of a 4 MWh full-scale demonstration facility. This facility is expected to be located approximately 7 kilometers from the shore of Toronto at a depth of 80 m below surface on Lake Ontario. Hydrostor and the University of Windsor are working to optimize the system efficiency and address some of the hydrodynamics challenges already identified.
UW-CAES has a promising future with the flexibility, reliability, and predictability the technology brings, especially in a continuously evolving electrical grid. Energy storage will be an integral part of the future electrical infrastructure, and we think UW-CAES has the potential to be a leading technology.