Pumped Storage Technology

The basic layout of a pumped storage hydropower plant involves two reservoirs, one above the other, and a turbine/pumping hall capable of both generating power from the stored water in the upper reservoir and pumping water from the lower reservoir back to the upper. For hydropower plants, in general, the energy available from a given volume of water is greater, the greater the head of water. In the case of the pumped storage plant this head is the vertical distance between the upper reservoir and the turbines. The greater this distance, the more energy a given quantity of water can store; put another way, the larger the head, the smaller the volume of water needed for a given amount of energy. However, the pumped storage head will be limited by the type of turbine that can be utilized.

While the highest head available would in theory be best, very high heads require Pelton turbines to exploit them efficiently and these cannot be used as pumps. A very high-head plant would, therefore, require separate pumps and turbines, as was used in the earliest pumped storage facilities. Using separate pumps and turbines is more expensive than using a single-pump/turbine unit. Therefore, most pumped storage plants use Francis or Deriaz turbines, which can be used in both modes. This limits the head that can be used to achieve good efficiency to about 700 m. Modern multistage pump turbines may be capable of extending this to around 1200 m.

A pump turbine may not achieve the efficiency possible when using independently optimized pumps and turbines, but the best combined pump turbines are capable of reaching around 95% efficiency for generation and 90% for pumping, leading to a roundtrip efficiency of 86%. Most plants operate in the 75–80% range.

The Francis turbine is the type most commonly used as a pump turbine. While highly efficient, it has fixed blades so the blade design is generally a compromise designed to optimize both generation and pumping efficiency. The Deriaz turbine is of similar design but with adjustable blades making it possible to optimize for generation and pumping independently. These have been used for pumped storage plants in several parts of the world but will generally be more costly than Francis turbines because of the additional complexity.

Variable-Speed Operation

While most pumped storage hydropower plants have been built using pump turbines that operate at a fixed speed that is synchronized with the grid there are significant advantages to be gained by having the ability to operate at variable speed. For pumping, variable-speed operation allows the pump to function with surplus power at different demand levels and it can take power while the demand level is changing on the grid, allowing for much greater flexibility of operation. In generation mode, variable-speed operation allows the unit to supply varying quantities of power. A fixed-speed turbine can only supply its rated output at grid frequency.

Variable-speed operation requires that the turbine generator be decoupled from the grid through a power electronic interface so that the electricity produced by the turbine at variable frequency can be injected into the grid at the synchronous frequency, or the latter can be adjusted when in pumping mode to allow variable pumping capacity. Therefore, a variable-speed pump generator will be more costly than a fixed-speed unit.

Pumped Storage Sites

The single greatest limitation (aside from economics) faced by pumped storage hydropower is the availability of suitable sites. A plant of this type requires two reservoirs at different heights. This can be difficult to engineer.

In rare cases it is possible to find two existing lakes that can be utilized to create a pumped storage facility. If natural lakes are exploited, it will be necessary to take into account the fact that the water level in both will vary more widely than it would naturally and assessing the environmental impact of these changes. More commonly, a natural lake might form one reservoir while the second is human-made. The third option is for both reservoirs to be human-made. However, this can add significantly to the capital cost of such a plant.

A further, and yet rarely attempted, solution is to use the sea as the second or lower reservoir. This is the layout of the 30 MW Yanbaru seawater pumped storage plant in Japan. One other idea that has never yet been exploited is to bury the lower reservoir underground in a suitable geological formation.

Plant capability depends on both the size of the reservoirs and the head, or vertical, height between them. The volume of the reservoirs will determine the overall capacity of the plant to store and supply energy. The more water, the more energy it can contain. However, for a given storage capacity, the output will depend both on the size of the turbines and the head. A high head can deliver more power from a given flow of water than a small head.

Performance

Pumped storage plants are capable of rapid response to sudden changes in demand. The Dinorwig plant in Wales can go from standby to synchronization at full output of 1320 MW in 12 seconds, a ramp rate of 110 MW/sec. Other, more modern pumped storage plants have been specified to be capable of output ramping when in operation at up to 500 MW/min, which compares favorably with gas turbine peaking plants and modern combined cycle power plants.

With adequate control of evaporation and leakage, the energy stored in a pumped storage reservoir can be retained indefinitely. This is not a significant advantage when most plants cycle daily, but it could prove so under circumstances where energy needs to be stored seasonally, such as solar power for use in winter, or wind power for use in less windy summer months.

Costs

An energy storage plant such as a pumped storage hydropower plant will depend for its revenue on being able to buy power at low cost and then sell it at higher cost. The income will, therefore, vary depending on a wide range of conditions. From an economic point of view the capital cost of building the plant will be the most important factor in determining its viability. This is likely to be relatively high because, like most hydropower plants, pumped storage is a capital-intensive technology.

At the top end, capital costs are likely to be as high or higher than for a traditional hydropower plant, which, as shown in Chapter 8, are generally in the range $1000–2000/kW. A 500 MW plant proposed for construction in California has an estimated cost of $1.1 billion and a capacity of 500 MW, or around $2200/kW. In contrast, the Tianghuangping pumped storage plant in Zhejiang province, China, cost $1.1 billion for 1800 MW when it came on-line in 2001, around $600/kW. Much of the difference can probably be accounted for by the lower labor costs in China.

Small pumped storage plants are likely to be relatively more expensive than larger installations.

Compressed-Air Energy Storage Principle

A CAES plant requires two principal components: a storage vessel in which compressed air can be stored without loss of pressure, and a compressor/expander to charge the storage vessel and then extract the energy again (Figure 10.1). (The latter might, in fact, be a compressor and a separate expander.) In operation the plant is broadly analogous to the pumped storage hydropower plant. Surplus electricity is used to compress air with the compressor and the higher-pressure air is stored within the storage chamber. This stored energy can then be retrieved by allowing it to escape through the expander, an air turbine that is essentially a compressor operating in reverse. The expanding air drives the air turbine, which turns a generator to provide electrical power.

The compression and expansion functions of the CAES plant can be performed by the two primary components of a standard gas turbine. As seen in Chapter 4, the gas turbine comprises three components: a compressor, combustion chamber, and turbine. If the combustion chamber is removed and the two rotary components separated, then these can alternately use electricity to compress air for storage and extract energy from it again to regenerate the stored power (Figure 10.2).

In practice, a slightly different arrangement is preferred that is closer still to the gas turbine. A rotary gas turbine compressor is used to compress air that is then stored in the storage chamber. When power is required, compressed air is extracted again and fed into a combustion chamber where it is mixed with fuel and ignited, generating a higher-pressure, higher-temperature thermodynamic fluid that is then used to drive the turbine stage of the plant.

Since a plant operated in this way requires natural gas or another fuel, it is not a straightforward energy storage system. However, the economics of this mode of operation appear to be the most attractive because it can generate more electricity than was used to store the compressed air. Additional generation is between 25% and 60% depending on the plant design. A further advantage is that the turbine stage of the plant does not have to drive the compressor as it would in a conventional gas turbine, so it can generate up to three times more power than it would when coupled to a compressor. Therefore, turbines for CAES plants are relatively smaller than for a similar generating capacity gas turbine.

Compressed-Air Storage Facilities

The most important part of a CAES plant is somewhere to store the compressed air. Small-scale CAES plants—with storage capacities up to 100 MWh and outputs up to 20 MW—can use aboveground storage tanks built with steel pressure vessels but large, utility-scale plants need underground caverns in which to store the air. The natural gas industry has used underground storage caverns for years to store gas; these same caverns can provide ideal storage facilities for a CAES plant. However, the demand for such a cavern limits the development of CAES to places where such storage caverns are available.

A number of different types of underground caverns can be exploited. The most expensive is a human-made rock cavern excavated in hard rock or created by expanding existing underground mine workings. Such a site must be located in an impervious rock formation if it is to retain the compressed air without loss so the suitability of underground coal mines and limestone mines will depend on whether they are air-tight.

Salt caverns are another type of storage site, one that has been commonly used for gas storage. These are created within naturally occurring underground salt domes by drilling into the dome and pumping in water to dissolve and remove the salt to create an enclosure. Salt deposits suitable for such caverns occur in many parts of the world.

It is another type of geological structure, however, an underground porous rock formation, which offers the cheapest underground storage facility. Structures of this type suitable for gas storage are found where a layer of porous rock is covered by an impervious rock barrier. Examples can be found in water-bearing aquifers, or in porous underground strata from which oil or gas have been extracted. Aquifers can be particularly attractive as storage media because the compressed air will displace water within the porous rock, setting up a constant-pressure storage system. With rock and salt caverns, in contrast, the pressure of the air will vary as more is added or released.

All three types of underground storage structures require sound rock formations to prevent the air from escaping. They also need to be sufficiently deep and strong to withstand the pressures imposed on them. It is important, particularly in porous-rock storage systems, that there are no minerals present that can deplete the oxygen in the air by reacting with it. Otherwise, the ability of the air to react with the fuel during combustion will be affected, reducing the power available during the generation phase of the storage generation cycle.

Underground rock structures capable of storing compressed air are often widely available. For example, a survey in the United States found accessible sites of different types across 80% of the country.

Turbine Technology and CAES Cycles

A CAES plant generally exploits standard gas turbine compressor and turbine technology, but because the two units operate independently, they can be sized differently to match the requirements of the plant. The larger the compressor compared to the turbine, the less time it requires to charge the cavern with a given amount of energy. The Hundorf plant that was built in Germany required 4 hours of compression to provide 1 hour of power generation, whereas the McIntosh plant in Alabama needs only 1.7 hours of compression for 1 hour of generation.

As a consequence of compression and generation being separated, a CAES plant turbine can operate well at part load as well as full load. More complex operation is also possible. The Alabama plant, for example, uses two turbine stages with the exhaust from the last turbine is used to heat air from the cavern before it enters the first turbine. Fuel is not actually burned in the compressed air until it enters a combustion chamber between the first and second chambers.

Many of the refinements used to improve gas turbine performance outlined in Chapter 4 can be used in CAES plants too. For example, the compressor can be divided into two sections with air cooling between the stages to reduce its volume (intercooling) and heat from the turbine exhaust can be recovered and used to heat the compressed air extracted from the storage chamber (recuperation). Reheating, where the turbine is divided into two stages with an additional combustion chamber between stage one and stage two can also be applied versions of these latter two are used in the McIntosh plant, as noted before.

Mechanical components are never 100% efficient so there are consequential energy losses during compression and expansion in a CAES plant. There is also an additional source of energy loss. When air is compressed it generates heat and this heat energy is lost in a conventional CAES plant. A proposed additional refinement to the conventional mode of operation involves capturing this heat and storing it for use to heat the pressured air as it exits the storage chamber, before entering the turbine. This adiabatic cycle could theoretically be used to design a CAES plant that has no need for additional fossil fuel and that could achieve a roundtrip efficiency of 65%.

In principle a CAES plant could be of virtually any size and one proposed project would have had a generating capacity of 2400 MW. However in practice most schemes are likely to be smaller than this, in the tens or hundreds of megawatts range. Startup for the two plants that have operated was around 12 min but both could be brought into service in 5 min if necessary. Round trip efficiency without the use of additional fuel will be low for conventional CAES plants such as the two that have operated but, as noted, refinements could improve this.

Costs

There is little experience with CAES so any cost estimates must be considered tentative. However, it would appear to be an economically attractive option for energy storage. Proposals within the last 10 years or so for conventional CAES plants in the United States have had installed costs of $400–900/kW depending on size and storage type. An adiabatic plant is likely to be much more expensive, with potential costs as high as $1700/kW.

Battery Principle

A battery is a device that can exploit a chemical reaction to produce electricity. The reactions upon which a cell is based will define the particular cell type. In all cases the reaction will occur spontaneously if the reactants are mixed, generating heat in the process. However, in the battery the reactants are separated and only allowed to react in a particular way.

The chemical reaction used in every battery can notionally be divided into two half-reactions and the battery will contain two electrodes, called the anode and cathode; each electrode is associated with one of these half-reactions. The half-reactions involve the creation of charged ions and the capture or release of electrons. Under normal circumstances where the reactants are intimately mixed, these processes occur simultaneously at the same location. However, in a battery the two electrodes are separated by an electrolyte that will allow charged ions to pass from one electrode to the other but will not allow electrons to pass. These can only cross from one electrode to the second to complete the reaction through an external circuit. This is the electrical current that can be used to drive electrical and electronic equipment. It is this separation of processes by the use of a selective filter (in this case, the electrolyte) that allows the cell to generate power.

As mentioned, batteries are generally divided into primary cells and secondary cells. Primary cells contain reactants that will only react once to produce power. After that the cell is spent. A secondary cell can be recharged by applying a reverse polarity to the cell, reversing the chemical cell reaction and regenerating the original cell reactants. It is this type of cell that is of use for energy storage.

There are a number of battery types that can be used for grid and utility applications. The most widely used secondary cell is the lead-acid battery, similar to the type commonly used in vehicles (Figure 10.3). Lead-acid batteries account for roughly half of all secondary cell sales, globally, and are widely available. Another type, nickel-cadmium (NiCad) batteries, are less common. NiCad batteries were used for portable computers until they were superseded by alternatives that were better suited to the duty cycle of such devices. They have also been used for automotive applications, particularly under low-temperature conditions when they behave more reliably than lead-acid cells. Other cell types developed specifically for portable electronic devices include nickel-metal hydride cells and lithium-ion cells. Both are potentially useful for utility storage. A high-temperature battery, the sodium-sulfur battery, has also been popular for utility applications.

In addition to these conventional secondary cells, a variety of devices known as flow cells or flow batteries have also been tested for large-scale energy storage applications. These include zinc-bromide, vanadium-redox, and polysufide-bromide flow cells. None has so far found widespread commercial application but new types are being developed. They are considered attractive because they have much longer lifetimes than conventional secondary batteries.

Traditional electrochemical storage systems boast a best-case cycle conversion efficiency (electricity to cell storage and back into electricity) of 90%, but a more typical figure would be 70%. Most batteries also suffer from leakage of power over time. Left for too long, the cell discharges itself. This means that battery systems can only be used for relatively short-term storage. Flow cells do not suffer from this problem because the reactants are not stored together and this helps reduce long-term energy losses.

An additional problem with traditional secondary batteries is their tendency to age. After a certain number of cycles, the cell stops holding its charge effectively, or the amount of charge it can hold declines. Much development work has been aimed at extending the lifetime of electrochemical cells but this remains a problem. Again flow cells, because of their design, can avoid this problem.

Against this, batteries are able to supply their output extremely quickly—under 5 msec for a conventional battery and less than 100 msec for a flow battery. Some are also capable of very high-power outputs and discharge rates.

Lead-Acid Batteries

Lead-acid batteries were among the first secondary cells to be developed and were used for load leveling in very early power distribution systems. The cell is based on a reaction between lead-oxide and sulfuric acid. Efficiencies of lead-acid batteries vary depending on factors such as the temperature and duty cycle but are typically between 75% and 85% for DC–DC cycling. However, cells discharge themselves over time so they cannot be used for very long-term power storage. If cycled carefully, cells for utility applications can have lifetimes of 15–30 years.

The cells have a water-based liquid electrolyte and operate at ambient temperature. Both high and low temperatures can reduce their performance. They are also relatively heavy and have a poor energy density, although neither of these factors are a handicap for stationary applications. In addition, they are cheap and easily recycled.

Several very large energy storage facilities based on lead-acid batteries have been built. These include and 8.5 MW unit constructed in West Berlin in 1986 while the city was still divided into East and West, and a 20 MW unit built in Puerto Rico in 1994. While the former operated successfully for several years, cell degradation led to the latter closing after only five years. Lead-acid cells have been very popular for renewable applications such as small wind or solar installations where they are used to store intermittently generated power to make it continuously available.

Nickel-Cadmium Batteries

The nickel-cadmium battery is one of a family of nickel batteries that includes nickel–metal hydride, nickel-iron, and nickel-zinc batteries. There is also a nickel-hydrogen battery in which one cell reactant is gaseous hydrogen. All have a nickel electrode coated with a reactive and spongy nickel hydroxide while the cell electrolyte is almost always potassium hydroxide. Cell reactions vary depending on the second component.

The only nickel-based cell that has been exploited for utility applications is the nickel-cadmium cell. Nickel-cadmium batteries have higher energy densities and are lighter than lead-acid batteries. They also operate better at low temperatures. However, they tend to be more expensive. This type of battery was used widely in portable computers and phones but has now been superseded by lithium-ion batteries.

Efficiencies of nickel-cadmium cells are typically around 70% although some have claimed up to 85%. Lifetimes tend to be rated at around 10–15 years though some have lasted longer. These cells discharge themselves more rapidly than lead-acid cells and can lose 5% of their charge in a month. There can also be a problem with disposal since cadmium is highly toxic.

The largest nickel-cadmium battery ever built is a 40 MW unit in Alaska that was completed in 2003. It occupies a building the size of a football field and comprises 13,760 individual cells.

Lithium Batteries

Lithium batteries, including both lithium-hydride and lithium-ion batteries, have become popular for consumer electronic devices because of their low weight, high energy density, and relatively long lifetimes. Lithium is extremely reactive and can burst into flames if exposed to water, but modern lithium cells use lithium bound chemically so that it cannot react easily. As with nickel, there are a number of lithium cell variants but the most popular today is the lithium-ion cell. These are designed so that there is no free lithium present at any stage during the charging or discharging cycle.

The use of lithium batteries in grid and utility applications is beginning to grow with units being tested in a number of locations. One large installation, due to start operating in 2013, is a 2 MW lithium-ion facility for the Orkney Islands off the northwestern coast of Scotland. The future development of lithium batteries may benefit from interest by automotive manufacturers in their use in hybrid and electric vehicles.

Sodium-Sulfur Batteries

The sodium-sulfur battery is a high-temperature battery (Figure 10.4). It operates at 300 °C and utilizes a solid electrolyte, making it unique among the common secondary cells. One electrode is molten sodium and the other molten sulfur, and it is the reaction between these two that is the basis for the cell reaction. Although the reactants, and particularly sodium, can behave explosively, modern cells are generally reliable. However, a fire was reported in 2012 at a sodium-sulfur battery installation in Japan.

Early work on the sodium-sulfur battery took place at the Ford Motor Co. in the 1960s but modern sodium-sulfur technology was developed in Japan by the Tokyo Electric Power Co., in collaboration with NGK insulators, and it is these two companies that have commercialized the technology. Typical units have a rated power output of 50 kW and 400 kWh. Lifetime is claimed to be 15 years or 4500 cycles and the efficiency is around 85%. Sodium-sulfur batteries have one of the fastest response times, with a startup speed of 1 msec.

Flow Batteries

A flowing-electrolyte battery, or flow battery, is a cross between a conventional battery and a fuel cell. It has electrodes like a conventional battery, where the electrochemical reaction responsible for electricity generation or storage takes place, and an electrolyte. However, the chemical reactants responsible for the electrochemical reaction and the product of that reaction are stored in tanks separate from the cell and pumped to and from the electrodes as required, much like a fuel cell (Figure 10.5). The processes occurring here are usually somewhat different to the simple electrochemical process described earlier but the principle is similar.

A number of flow batteries have been tested including the zinc-bromide flow battery, the polysulfide-bromide battery, and the vanadium redox battery. A number of newer designs are also in the research stage. Response times for flow batteries are longer than for conventional secondary batteries but they should be able to supply full power within 100 msec. However, flow batteries have not been extensively tested commercially so their overall performance has yet to be established.

Costs

The costs of battery systems vary widely depending on type and size. Large batteries are generally cheaper than similar small installations. Battery costs are normally based on the cost per kilowatt-hour rather than the cost per kilowatt. The cheapest of the conventional cells is the sodium-sulfur battery with a cost in the range of $250–900/kWh. Lead-acid battery costs are in the range of $400–2300/kWh and nickel-cadmium of $1800–2400/kW. Flow batteries are potentially cheaper with costs in the range of $150–800/kWh depending on the type.

Costs

Although a few commercial SMES devices are available, costs have been difficult to establish. In general, the cost is relatively low per unit size (MW) but high in terms of storage capacity (MWh). For short-term grid stability duty, they appear to be competitive with other types of storage such as batteries, flywheels, and capacitors.

Flywheel Principle

A flywheel depends on a rotating mass to store energy that is then held in the kinetic energy of rotation of the rotor. The stored energy is proportional to the moment of inertia of the rotor about its axis (directly related to its mass) and its rotational speed.

Conventional flywheels such as those used for piston engines are fabricated from heavy metal disks made of iron or steel. However, these disks are only capable of rotating at low speeds. For power applications, new lighter composite materials have been developed, capable of rotating at 10,000–100,000 rpm without fracturing under the immense centrifugal force they experience. These are made from carbon fiber or glass fiber composite materials that, while less dense than the metals they replace, can store more energy at the very high rotational speeds at which they operate. See Figure 10.7.

Friction losses can be significant at high rotational speeds and energy storage flywheels generally use magnetic bearings to minimize such losses. These are normally supplemented by conventional bearings in case the magnetic bearing fails. Friction losses from air drag can also be significant at the speeds at which modern flywheels rotate, so they will normally be housed in a vacuum chamber, evacuated to reduce this source of loss too. The whole device must be enclosed within an exceedingly strong container that will prevent the pieces of the flywheel scattering like shrapnel in the event of a catastrophic failure.

Energy is fed into a flywheel using a motor to rotate it up to its maximum speed. Energy is extracted from the rotor through a generator driven by the flywheel shaft. Extracting energy from the flywheel will lead to its rotational speed lowering. The makes grid synchronization virtually impossible so most flywheels use electronic AC–DC–AC converters to eliminate this problem and permit variable-speed operation. Flywheels are generally maintenance free.

Flywheel Performance Characteristics

For a flywheel to operate effectively as an electrical energy storage system it must be kept rotating at its full operating speed. This requires continuous energy feed to compensate for friction losses. Roundtrip efficiency estimates vary with some manufacturers claiming up to 90%. Other sources such as the U.S. Department of Energy suggest 70–80% is more typical.

The energy stored in a flywheel is a function of the mass of its rotor. The larger the rotor, the more it can hold. However, the power it can deliver will depend on the size of the generator used for energy extraction. Most flywheels are designed for backup power or grid support functions during which they are expected to deliver high power for a short period of perhaps a few seconds at most. Such units can have small rotors but relatively large power extraction systems. Other flywheels have been designed for long-term energy delivery. These will have rotors that are capable of storing a large amount of energy compared to the size of the energy extraction system.

Flywheels can usually respond extremely quickly. In grid backup systems they should be capable of reaching full power within half a cycle, 25 msec, and some are quoted with response times of 5 msec. Such units will probably be able to supply their full output between 5 seconds and 15 seconds.

Commercial flywheels are available with power ratings between 2 kW and 2 MW and with storage capacities between 1 kWh and 100 kWh. One of the largest commercial systems is a unit with ten 100 kW flywheels used by the New York Transit System to support its electric traction power network. This system can supply 1 MW of power for 6 seconds. However, the largest flywheel so far constructed is one used in Japan for fusion research. This system can supply 340 MW for 30 seconds.

Costs

The commercial cost of a flywheel system is around $2000/kW, although costs have been put as low as $500/kW. The cost per unit of stored energy is around $1 million/kWh for small commercial flywheel systems though it may be lower for large flywheels. These costs reflect the fact that most flywheel systems store a small amount of energy but can deliver it in a quick burst of high power.

Energy Storage Capacitor Principles

A simple electrostatic capacitor comprises two plates with an air gap between them. When a voltage is applied to the plates, charge builds up on them to neutralize the voltage by creating an equal and opposite static-charge voltage across the plates. The charge will continue to build as the voltage is increased until it is high enough to cause air to breakdown and start conducting electricity.

The amount of charge the capacitor will hold can be increased by placing a dielectric material between the plates that reduces the strength of the field between the plates, allowing more charge to build up before breakdown occurs. An electrochemical capacitor is similar to this in that it has a dielectric material between the capacitor plates, but in this case the dielectric is a liquid electrolyte such as sulfuric acid or potassium hydroxide (Figure 10.8). The capacitor plates themselves are inert materials that will not react with these reagents.

When a charge is applied to the capacitor it causes the usual charge buildup. However, in this case the charge on each plate is neutralized by an opposing layer of charged ions from the electrolyte. This creates a double layer of charge at each plate, effectively leading to two capacitive charged layers, which massively increases the amount of charge the unit can hold.

The simplest electrochemical capacitors use a water-based electrolyte, which restricts the voltage that each can support to around 0.9 V before electrolysis of the water begins. This limits the amount of charge the capacitor can hold but otherwise these capacitors have excellent characteristics. Basic capacitors of this type (known as symmetrical capacitors) have two identical electrodes made from carbon. By varying the construction of one electrode, it is possible to increase the amount of energy the device can hold. Devices of this type are called asymmetrical capacitors. A second type of electrochemical capacitor uses an organic electrolyte, which allows it to support a voltage up to 2.7 V. Both symmetrical and asymmetrical versions of these capacitors are available too. Since single capacitors of either type can only support a relatively low voltage (in grid terms), electrochemical capacitors are usually stacked in series to allow higher voltages to be exploited.

Performance Characteristics

Electrochemical capacitors can be cycled for tens of thousands of times without degradation provided the voltage across them is kept below the maximum so that no internal reaction takes place. However, once they are charged they do lose charge slowly through leakage in the same way as a battery. The leakage levels in water-based electrochemical cells are similar to those of a lead-acid battery. Leakage levels are lower with organic-based electrolytes.

Leakage will reduce long-term storage life, but with fast cycling, roundtrip efficiency can be 95% or higher. In addition, capacitors can normally discharge their energy very rapidly without damage so long as excessive internal heating is avoided. Energy density is moderate between 1 Wh/kg and 5 Wh/kg for symmetrical capacitors and up to 20 Wh/kg for asymmetrical capacitors. In comparison, a lead-acid battery has an energy density up to 42 Wh/kg. Response time is fast at 5 msec.

Units can be designed with the capability to deliver power levels between 1 kW and 5 MW, but actual energy storage capacity is generally relatively low at 1–10 kWh. When power is withdrawn from a capacitor the voltage falls, so sophisticated DC–AC conversion systems are needed to maintain a constant voltage output.

Applications

Electrochemical capacitors have been used both for energy storage and for braking energy recovery systems in automotive applications. For grid use they are best suited to backup or fast reaction grid support, offering a similar performance to flywheels. Although capacitors are not yet widely deployed for grid support they have been tested in a number of configurations. These include adding rapid-response storage to small distributed generation grids or micro-grids where they can provide fast-reacting grid support when the output from intermittent renewable resources suddenly falls and before a backup engine–based system can take over. Capacitors are also being tested for high-voltage grid support services.

Costs

The cost of capacitor storage is likely to be similar to that for flywheels at around $2000/kW. Based on cost per unit of energy storage, the price is again expected to be similar to that of flywheels with costs from $7000/kWh to perhaps $1million/kWh depending on the size of the unit. As with flywheels, this reflects the use of this in low storage capacity, high power output configurations for rapid, short-duration delivery of power.

Hydrogen Energy Essentials

Hydrogen, as a fuel, can be generated by the electrolysis of water using an electrical voltage. This has been carried out industrially for many years with the main system being an alkaline electrolyzer, exploited most successfully by the Norwegian utility Norsk Hydro. Large-scale electrolyzers have been built that are capable of handling inputs of 100 MW for hydrogen generation and the product is around 99.8% pure. Conversion efficiency is 90%. Alternatives to the alkaline electrolyzer are proton exchange membrane electrolyzers that are currently being developed and could potentially achieve 94% efficiency, but with the need for a platinum catalyst. High-temperature ceramic electrolyzers are also under development.

Once produced, hydrogen must be stored, for instance, as shown in Figure 10.9. While the gas has a high-mass energy density, it is very light and has a low-volume energy density so it must be compressed or stored in a concentrated state. For power generation applications, storage under pressure in steel or composite tanks is probably the favored method. The gas can be liquefied but only by using cryogenic equipment, making the process costly. There have also been attempts to store hydrogen in the solid state within certain alloys that will absorb it in large quantities. This may eventually offer the best storage method for automotive applications, although the weight of the storage system is currently a major problem.

Once manufactured and stored, hydrogen can be converted back into electricity in a number of ways. It can be burned like natural gas, although the combustion temperature is generally higher than for the former. However, most gas turbines, piston engines, and gas-fired boilers can easily be adapted for its use. When burned in this way it produces only water as a by-product. Alternatively, and possibly most efficiently, hydrogen can provide the fuel for a fuel cell system as illustrated in Figure 10.9. For a future energy economy based on hydrogen, this offers one of the most promising solutions with 60% efficiency achievable in a simple fuel cell and perhaps 70–75% with a hybrid system.

Performance Characteristics

When used as an energy storage medium within an electricity system, hydrogen will generally be used as fuel in a conventional or fuel cell power plant and generally cannot be brought into service as rapidly as some of the fast-acting storage systems discussed above. It should, therefore, be considered as a system of energy arbitrage rather than for grid support.

The main drawback today of hydrogen storage is the round cycle efficiency. With an electrolyzer operating at 90% efficiency and a power plant converting it back into electricity with perhaps 60% efficiency, the best roundtrip efficiency that can be expected is 54%, much lower than other storage systems discussed earlier. Such low efficiency may be tolerable in a renewable energy storage system, such as a wind–hydrogen storage unit where the wind energy must otherwise be shed. It is unlikely to be considered sufficiently efficient for generation from off-peak grid power in most other circumstances if there is an alternative available.

Costs

Since hydrogen energy storage as an electrical energy storage medium has yet to be tested, there are no realistic costs available for practical systems. If it is to be of use it would need to be able to compete with the high storage capacity technologies such as pumped storage hydropower, CAES, or large battery storage.