Pumped-storage hydroelectricity
Pumped-storage hydroelectricity (PSHE) is the most established utility-scale energy storage technology. This system stores energy in the form of water in an upper reservoir, pumped from another reservoir at a lower elevation. During periods of high electricity demand, power is generated by releasing the stored water through turbines, as with a conventional hydropower station. During periods of low electricity demand, and when cheap or surplus electricity is available to the grid, the upper reservoir is recharged by using electricity to pump the water back to the upper reservoir.
Although some efficiency losses are inevitable, PSHE plants are usually highly efficient, with ‘round-trip’ efficiencies reaching more than 80%.
Reversible pump-turbine/motor-generator assemblies can act as both pumps and turbines, and are fully proven commercially. The materials used to construct the pump/turbines are standard and are not in short supply globally. The operating fluid is water, which is freely available in many parts of the world. Seawater may also be used, but may require more sophisticated, corrosion-resistant materials.
PSHE has the highest capacity of the energy storage technologies known and tested. It can be sized up to around 4GW capacity. There are four PSHE facilities in the UK: Dinorwig, Wales – 1,728MW capacity; Foyers, Scotland – 305MW; Cruachan, Scotland – 440MW; and Ffestiniog, Wales – 360MW. In total, these plants provide 30GWh/day, satisfying a small proportion of electricity demand.
More such facilities may be built. SSE, which owns the Foyers plant in Scotland’s Great Glen, has proposed two new PSHE plants close by, one at Balmacaan and the other at Coire Glas. Each of these facilities would be rated at 600MW and capable of supplying up to 30GWh/day.
Compressed air energy storage
Compressed air energy storage (CAES) plants are similar to pumped hydro power plants in terms of applications, output and storage capacity. However, instead of pumping water from a lower to an upper reservoir during periods of excess power, a CAES plant compresses ambient air and then stores it under pressure, generally in an underground cavern. When electricity is required, the pressurised air is released and expanded in a gas-expansion turbine driving a generator, for power production.
Underground CAES storage systems are the most cost-effective, with capacities of up to 10GWh. Surface units are typically smaller and more expensive, with capacities of 60MWh. There are two operating first-generation CAES systems: one in Germany and one in Alabama.
Owing to the high isentropic exponent of air, the temperature of compression is very high with CAES, resulting in very high discharge temperatures from the compressor. The air is normally stored at a pressure of about 70bar. This heat of compression has to be extracted during the compression process, using inter-coolers and an after-cooler.
Diabatic CAES
In a conventional diabatic system, the loss of heat energy has to be compensated for during the gas expander (power generation) phase. This compensation is achieved by heating the high-pressure air in combustors, generally using natural gas – though a combustion gas turbine exhaust in a recuperator can also be used to heat the air from storage before the expansion cycle.
Alternatively, the heat of compression can be thermally stored before it enters the cavern and used for adiabatic expansion (see below), extracting heat from the thermal storage system. This technique minimises specific gas consumption and reduces the associated carbon dioxide emissions by 40-60%, depending on whether the waste heat is used to warm up the air in a recuperator.
The electrical ‘round-trip’ efficiency of this method is around 42% without, and 55% with, the use of waste heat.
Two facilities – the Huntorf plant in Germany and the McIntosh plant in the US – use single-string turbo-machines where the compressor-motor/generator-gas expander are on the same shaft and are coupled via a gearbox. In future CAES plants, the motor-compressor unit and the turbine-generator unit are likely to be decoupled, for greater plant flexibility.
Adiabatic CAES
A much higher efficiency, up to 70%, can be achieved if the heat of compression is recovered and used to reheat the compressed air during expansion through the turbine. This means that there is no longer any need to burn extra natural gas to preheat the decompressing air.
A consortium led by German company RWE is developing the components and heat storage system for an adiabatic CAES project called ADELE. The pilot plant is scheduled to open in 2018.
Isothermal CAES
Isothermal CAES is an emerging technology that attempts to overcome some of the limitations of traditional (diabatic or adiabatic) CAES. Traditional CAES uses turbo-machinery to compress air to around 70bar before storage. Without inter-/after-cooling, the air would heat up to 900K, making it impossible – or prohibitively expensive – to process and store the gas. Instead, with the isothermal method, the air undergoes successive stages of compression and heat-exchange to achieve a lower final temperature close to ambient. In advanced-adiabatic CAES, the heat of compression is stored separately and fed back into the compressed gas upon expansion, thereby removing the need to reheat with natural gas.
Isothermal CAES is challenging, since it requires heat to be removed continuously from the air during the compression cycle and added continuously during expansion to maintain an isothermal process. But companies developing this method expect a potential round-trip efficiency of 70–80%.
Storage options for CAES
Very large stores are needed for CAES – regardless of the method selected – because of the low storage density. The most suitable locations are artificially constructed salt caverns in deep salt formations. Such caverns offer several benefits: high flexibility, no pressure losses within the store, no reaction with the oxygen in the air, and a salt host rock.
If no suitable salt formations are present, natural aquifers can be used, although there is a danger of oxygen reacting with rock and with any micro-organisms present. This risk is also a factor with another option that is being investigated – depleted natural gas fields. With the latter location, the mixing of residual hydrocarbons with compressed air also has to be considered.
Cryogenic energy storage
Cryogenic energy storage (CES), sometimes called liquid air energy storage (LAES), uses liquefied air or liquid nitrogen, which can be stored in large volumes at atmospheric pressure. ‘Cryogenic’ refers to a gas in a liquid state at very low temperatures. Uniquely, CES systems can also harness low-grade waste heat from co-located processes, converting it to power.
Although novel at a system level, the components and sub-systems of CES systems are mature technologies available from OEMs. As a whole, the technology draws heavily on established processes from the turbo-machinery, power generation, industrial gas and air separation industries. The turbo-machinery is virtually identical to that used in CAES systems, and requires no rare or unsustainable materials of construction. A 350kW/2.5MWh CES demonstration plant in Slough has been under operational testing by Highview Power Storage since 2011.
CES involves three core processes:
• Charging: the system uses electrical energy (excess or off-peak) to drive a conventional air liquefier. Extracting ambient air from the surrounding environment, the gas is cleaned, compressed and cooled until it undergoes a phase change to a liquid, producing a storage medium that is four to six times more energy-dense than compressed air at 200bar, and around 700 times more dense than atmospheric air.
• Storage: the liquid air is stored in an insulated tank at low pressure, which functions as the energy store. Again, this equipment is widely deployed for the storage of bulk liquefied natural gas, nitrogen or oxygen (2,000t – 200MWh equivalent – to 100,000t tankage, enough for 10GWh).
• Discharging (power recovery): when power is required, liquid air/nitrogen is drawn from the tank and pumped to high pressure. Ambient heat is applied to the liquid air via heat exchangers resulting in a phase change from liquid air to a high pressure gas which is then used to drive an expansion turbine generator.
During the power recovery process, very cold gas is exhausted. This is then recycled back into the liquefaction process, reducing the energy demands for producing liquid air. The introduction of low-grade waste heat (below 120°C) into the power recovery system increases the amount of power that can be extracted. Such waste heat is readily available from traditional thermal power plants, and from many industrial processes such as the manufacturing of steel, cement, and chemicals. Using waste heat of around 115°C, the electrical round-trip efficiency may potentially be as high as 70%.
By integrating low-grade waste heat, CES can penetrate untapped markets, and potentially increase the energy efficiency of manufacturing and existing thermal power generation methods.
Hydrogen energy systems
Hydrogen differs from other candidates for utility-scale energy storage in that it has a high calorific value, so is useful in itself as a fuel rather than being simply an energy vector. However, unlike air and water, hydrogen is not freely available and, once extracted from other substances, is hard to contain.
Nevertheless, hydrogen is a valuable product, with annual production of 55Mt/year. Most hydrogen is used in industrial processes, such as synthetic fertiliser manufacture and fuel desulphurisation.
However, more than 95% of the currently available hydrogen is ‘brown’ hydrogen, derived from fossil hydrocarbons, so cannot be considered truly sustainable. Despite this problem, a growing amount of ‘green’ hydrogen is now being produced, mainly through the electrolysis of water.
Renewable hydrogen
production
Alkaline electrolysis is a mature technology for large systems, whereas proton exchange membrane (PEM) electrolysers are more flexible and can be used for small, decentralised solutions. The conversion efficiency for both methods is about 65-70%.
Another technology under development is high-temperature electrolysers, which could represent an efficient alternative to PEM and alkaline systems, with efficiencies of up to 90%.
Hydrogen storage
As with the technologies described above, excess electricity is converted into hydrogen by electrolysis. The hydrogen is then stored and eventually re-electrified. The ‘round-trip’ efficiency is 30-40% but could reach 50% as more efficient technologies are developed.
Despite this low efficiency, interest in hydrogen energy storage is growing, owing to its much higher storage capacity compared with batteries (small-scale) or pumped hydro and CAES (large-scale).
Small amounts of hydrogen – up to a few MWh – can be stored in pressurised vessels at 100-300bar, or liquefied at -253°C. Alternatively, solid metal hydrides or nanotubes can store hydrogen with a high density.
Large amounts of hydrogen can be stored in man-made underground salt caverns of up to 500,000m3 at 200bar, corresponding to a storage capacity of 167GWh hydrogen (100GWh electricity).
Compressing hydrogen to the pressures required for storage is a complex process. The gas has such a low molar mass that it tends to ‘slip’ back towards the low-pressure end of the compressor, creating more inefficiencies.
Medium pressures require piston-type compressors, while the very high pressures (700bar) that are being developed for transport applications tend to require the diaphragm type. Compressors at these pressures are inevitably expensive, although they do not require rare or unsustainable materials.
Hydrogen re-electrification
Hydrogen can be re-electrified in fuel cells with efficiencies up to 50%. Alternatively, the gas can be burned in combined cycle gas turbine (CCGT) power plants, with efficiencies as high as 60%.
Pumped heat electrical energy storage
In pumped heat electrical energy storage (PHEES), electricity is used to drive a storage engine connected to two large thermal stores containing a mineral particulate – such as gravel. To store energy for electricity, the excess electric power drives a heat pump, which pumps heat from the ‘cold store’ to the ‘hot store’, resulting in the former cooling to around -160°C and the latter warming to around 500°C at 12bar. To recover the energy, the heat pump is reversed to become a heat engine. The engine takes heat from the hot store, delivers waste heat to the cold store, and produces mechanical work. When recovering the stored energy, the heat engine drives a generator to produce electricity.
The requirements for PHEES are two low-cost (usually steel) tanks filled with the mineral particulate and a means of efficiently compressing and expanding gas. A closed circuit filled with the working gas connects the two stores, the compressor and the expander. A monatomic gas, such as argon, is ideal as the working gas, as it heats/cools much more than air for the same pressure increase/drop – significantly reducing the storage cost. The expected electrical ‘round-trip’ efficiency is 75-80%.
The use of gravel as the storage medium makes this storage solution low in cost. The materials used suffer from no potential supply constraints and are all considered sustainable. Plant sizes are expected to be in the range of 2-5MW per unit; aggregated units could provide gigawatt-sized installations. Technology development firm Isentropic has taken this technology to development stage, with commercial systems expected this year.
Flywheel electricity storage systems
Flywheel electricity storage systems (FESSs) can be viewed as kinetic or mechanical batteries. They use excess electricity in a motor that accelerates a rotor (flywheel) to a very high speed, which stores the energy in mechanical/rotational form. This stored energy is converted back by slowing down the flywheel, providing power through a generator. The rotor spins in an almost frictionless enclosure.
A single FESS unit manufactured by Beacon Power can deliver 100kW power and store 25kWh. Such units can be built from modules into large energy storage units, for purposes such as frequency regulation. An example is the first known operating smart energy matrix frequency regulation plant, which comprises 20 such units, with output energy of 5MWh at a power of 20MW.
Flywheels offer rapid response times and large numbers of charge cycles, but must be housed in robust containment and require precision components, so they are relatively high in cost.
Most modern high-speed FESSs comprise a massive rotating cylinder – a rim attached to a shaft – supported on a stator by magnetically levitated bearings. To maintain efficiency, the flywheel system is operated in a vacuum so drag is reduced. The flywheel is connected to a motor/generator that interacts with the utility grid through advanced power electronics.
FESSs offer the advantages of low maintenance, long life – 20 years or tens of thousands of deep cycles – and negligible environmental impact. They can bridge the gap between short-term ride-through power and long-term energy storage, with excellent cyclic and load-following characteristics.
Typically, users must choose between solid steel or carbon composite rims. The choice of material will determine the system cost, weight, size, and performance. Composite rims, while expensive, are both lighter and stronger than steel, so they can achieve much higher rotational speeds. The amount of energy that can be stored in a flywheel is a function of the square of the rotational speed, so higher speeds are desirable.
High-power flywheels are being used in many aerospace and uninterruptible power supply applications, and 2kW/6kWh systems are being used in telecommunications applications. For utility-scale storage, a ‘flywheel farm’ approach can be used to store megawatts of electricity for applications needing minutes of discharge duration. Several flywheel farm facilities are in the planning or construction stages.
Batteries – flow-type (redox flow)
A redox flow battery (RFB) is a device that can accumulate (charging mode) and deliver (discharging mode) energy via reversible reduction-oxidation reactions of electrolytes, either in liquid or gaseous form, that are stored in separated storage tanks. The name ‘redox’ refers to chemical reduction and oxidation reactions employed in the RFB to store energy in liquid electrolyte solutions.
In a RFB, power is decoupled from the energy storage capacity since the power is determined by the number of cells and their size, while the energy capacity is a function of the volume and concentration of electrolyte. Redox flow batteries can operate to high levels of depth of discharge but have lower energy densities. Various redox couples have been tested but only zinc bromine (Zn/Br) and all-vanadium (V/V) redox batteries have reached the market. For example, ZBB has produced an energy storage unit that can deliver 25kW power and store up to 50kWh of energy. Combined into large modules, such units can store 500kWh energy, with the potential to be upscaled even further to at least 6MWh.
A few companies manufacture V/V redox flow cells. A typical energy storage unit with 10kW power and 100kWh energy can be modularly upscaled to deliver 40kW/400kWh of power and energy. Larger systems can be designed to meet higher power and energy requirements. Zn/Br and V/V redox batteries have been developed for applications such as solar energy fuelling stations, telecoms, and remote area utility power.
The characteristic separation of power and energy in RFBs also provides design flexibility in their application. The power capability (stack size) can be directly tailored to the associated load or generating asset while the storage capability (size of storage tanks) can be independently tailored to the energy storage need. So RFBs can economically provide an optimised storage system.
RFBs suit applications with power requirements in the range of tens of kilowatts to tens of megawatts, and with energy storage requirements in the range of 500kWh to hundreds of MWh.
Batteries – lithium-based
The first commercial lithium-ion (Li-ion) battery was released by Sony and Asahi Kasei in 1991. These batteries were used for consumer products and many companies are now developing larger-format cells for use in energy-storage applications. In addition, the emergence of electric vehicles (EVs) powered by Li-ion batteries is expected to create synergies.
The batteries have been deployed in a wide range of energy-storage applications, ranging from energy-type batteries of a few kWh in residential systems, with rooftop photovoltaic arrays, to multi-MWh containerised batteries for the provision of grid ancillary services.
There is a wide range of sub-chemistries within the family of Li-ion cells, each with specific operational, performance and safety characteristics. Li-ion cells may be produced in cylindrical or prismatic (rectangular) format. These cells are then typically built into multi-cell modules in series/parallel arrays, and the modules are connected together to form a battery string at the required voltage, with each string being controlled by a battery management system.
The battery management system, comprising electronic subsystems, is important. The cells should be charged and discharged within controlled parameters, since they lack the capacity of aqueous technologies to dissipate overcharge energy. Safety characteristics are ultimately determined by the attributes of system design, including mechanical and thermal characteristics, electronics and communications, and control algorithms, regardless of electrochemistry.
One US company has manufactured Li-based electrical energy storage units with 2MW of power and 500kWh of energy and has deployed more than 20MW of grid-connected energy storage units since 2008.
Another US company has developed a unit providing 1MW power and 250kWh energy – these storage units can be assembled into larger systems. One 1MW/250kWh unit has been operating since 2008.
AES, a US independent power producer, has deployed large-scale battery storage, with project sizes in the range 16–32MW. SSE is trialling a 2MW/500kWh Li-ion battery in Orkney. The project, owned and operated by a contractor, was commissioned in 2013. The battery is used to store wind energy when there is insufficient export capacity or demand on the islands, thereby maintaining wind generation.
Batteries – metal-air
Metal-air batteries are the most compact and, potentially, the least expensive batteries available. They are also environmentally benign. However, electrical recharging of these batteries is difficult and inefficient. Many manufacturers offer refuellable units, where the consumed metal is mechanically replaced and processed separately, but few developers offer an electrically rechargeable battery. Rechargeable metal-air batteries typically have a life of only a few hundred cycles and an efficiency of about 50%.
The anodes in these batteries are commonly available metals with high energy density, such as aluminium or zinc, that release electrons when oxidised. The cathodes or air electrodes are often made of a porous carbon structure or a metal mesh covered with proper catalysts. The electrolyte is often a good hydroxide-ion conductor, such as potassium hydroxide. The electrolyte may be in liquid form or a solid polymer membrane saturated with KOH.
While the high energy density and low cost may make them ideal for many primary battery applications, the performance of secondary batteries needs to be confirmed before they can compete with other rechargeable battery technologies.
high-temperature batteries – sodium-sulphur/sodium-nickel halide
Sodium sulphur (NaS) batteries were developed by Ford in the 1960s but it was not until the 1990s that commercialisation was successful, when the technology was adopted by NGK Insulators and the Tokyo Electric Power Corporation in Japan.
The batteries use electrodes of molten sodium and sulphur, with a ceramic separator as the conductive electrolyte. Battery temperature is held in the range 280-350°C, which can be a problem for intermittent operation. The energy storage module based on the NaS battery can provide 50kW power and 360kWh energy.
There are more than 300 NaS energy facilities worldwide, with the largest being a 34MW/245MWh unit for wind energy stabilisation in Japan. More than 270MW of stored energy, suitable for six hours of daily peak shaving, has been installed. Demand for the batteries is expanding, as a way to stabilise renewable energy output, and to provide ancillary services. US utilities have deployed several MW systems for meeting peak demand (peak shaving), back-up power, firming wind capacity, and other applications.
This technology has a good reputation for cycle life. A fire at a NaS battery facility in 2011, which led to the release of hydrogen sulphide, raised safety concerns, although these appear to have been overcome by design changes.
The ZEBRA, or sodium-nickel chloride (NaNiCl) battery, uses a molten sodium anode, a molten sodium aluminium chloride electrolyte and a nickel cathode. As with NaS, it needs an elevated temperature to operate: 260-360°C. Initially developed for electric vehicles, NaNiCl batteries offer a potentially attractive solution for stationary energy storage. They offer low levels of self-discharge and good lifetimes, though there are concerns that energy is required when the battery is not in use to maintain its operating temperature.
There are limited grid applications to date. A 400kW unit is under development in the US. More recently, GE has begun manufacturing this technology, renaming it the sodium metal halide battery.
batteries – nickel-based
The nickel-cadmium (Ni-Cd) battery, in commercial production since the 1910s, has seen periodic advances in electrode technology and packaging to remain viable. While not excelling in typical measures such as energy density or first cost, Ni-Cd batteries remain relevant by providing simple implementation without complex management systems, while providing long life and reliable service.
Early Ni-Cd cells used pocket-plate technology, a design still in production. Sintered plates entered production in the mid-20th century, to be followed by fibre plates, plastic-bonded electrodes and foam plates. Cells with pocket and fibre plates generally use the same electrode design for both the nickel positive and cadmium negative, while sintered and foam positives are now more commonly used with plastic-bonded negatives.
All industrial Ni-Cd designs are vented types, allowing gases formed on overcharge to be dissipated but requiring water replenishment to compensate. This requirement has led to the implementation of separator designs that allow varying levels of recombination, with some products designed for telecoms or off-grid renewable energy applications, achieving near maintenance-free operation with respect to the electrolyte.
Ni-Cd batteries found use in earlier energy-storage applications, including the Golden Valley Electric Association BESS near Fairbanks, Arkansas, sized for 27MW for 15 minutes, commissioned in 2003. Ni-Cd has also been used for stabilising wind-energy systems, with a 3MW system connected to a wind/diesel hybrid system on the former Netherlands Antilles island of Bonaire, commissioned in 2010. This was part of a project to make the island the first community with 100% of its power derived from sustainable sources.
Batteries – lead-acid
Lead-acid chemistry is the most mature rechargeable battery technology. It is a low-cost and popular storage choice for applications such as power quality, automotive, uninterruptible power supply and telecoms. Though inferior in power density to Li-ion, lead-acid batteries are still used for large-scale energy storage. But application for energy management has been limited owing to short cycle life.
A 1MW/1.5MWh lead-acid system by GNB Industrial Power and Exide has been operating for 12 years and was replaced in 2008. Another 1MW/1.2MWh system has been operated by Stadtwerke Herne, Germany, since 1996. Other lead-acid systems have been deployed in sizes of 10 to 20MW. The largest was a 40MWh system in Chino, California, built in 1988, which operated for several years.
Advanced lead-acid batteries, which include carbon-doping of the electrodes, are being commercialised, with improved cycle life and durability. Storage systems using these batteries are expected to start testing shortly. However, the scarcity of lead poses questions for the future.
Superconducting magnetic energy storage
A superconducting magnetic energy storage system (SMES) stores energy in a magnetic field created by the flow of electric current in a superconducting inductor. The inductor must be cryogenically cooled below its superconducting critical temperature. Energy is added or extracted from the magnetic field by increasing or decreasing the current in the inductor.
At steady state, the superconducting inductor does not dissipate energy, so the energy may be stored almost indefinitely. Several systems were deployed in the 1990s, but the technology was not adopted widely. New companies are developing SMES devices. A 24kV SMES magnet has been tested as a research system at Florida State University, and much work is being done on this technology in Germany. SMES offers high cycle life and rapid response, but has a relatively low energy density and high cost, and requires energy to cool the magnet.
Super-capacitors
Electrochemical double-layer capacitors (EDLC), or supercapacitors, store energy in the form of separated charges at porous electrodes divided by an electrolytic solution. The high power density but relatively low energy density of EDLCs make them suitable for voltage and frequency stabilisation. EDLC storage is slowly being deployed. A pilot project of a 300kWh/100kW uninterruptible power supply system, using electrochemical capacitors for bridging power, was carried out by EPRI Power Electronics Application Center in the US in 2003.
This technology offers high cycle life and rapid response, but has a relatively low energy density and high cost, and suffers from a high rate of self-discharge compared with other electrochemical energy storage technologies.
Graphene super-capacitors
This super-capacitor is an early-stage technology using graphene, which has high conductivity, mechanical strength and the potential to be produced in extremely thin layers. Work by General Electric has achieved high energy densities and fast charge capacity at a small scale.
The material is said to be sustainable, but, although graphene is basically carbon, it is made with mineral graphite – reserves of which are depleted. To be truly sustainable, graphene would have to be synthesised from other carbon sources.