PUMPED STORAGE HYDROPOWER
Pumped storage hydropower is both the simplest and most widely used technique for storing electrical energy today. It was first deployed in Switzerland around 1904,3 and there is probably around 130 GW of capacity in use, though estimates vary. These plants vary in size from a few megawatts to over 1000 MW, with the largest close to 3000 MW in capacity. Plants can be found in Australia and China across Europe and in Russia, but the largest aggregate capacities are in Japan and the United States. Many are used in conjunction with nuclear power plants so that the latter can operate at full power irrespective of demand. However, some smaller plants are also used for peak shaving and load management duties independent of the availability of nuclear power.
The reservoir-based pumped storage plant is an adaptation of the conventional hydropower plant to enable it to operate reversibly (Figure 10.1). In a conventional hydropower plant with a reservoir water collects in the reservoir and is then released through the plant’s turbines as power is required. This confers an element of energy storage but the water in the reservoir can only be used once.
In a pumped storage plant there is a second reservoir below the turbine hall. Water that has been released from the first reservoir and used to generate electricity is collected again here and stored. The power station is also equipped with pumps—in most cases its power turbines can be operated in reverse as pumps—and during periods when excess power is available on the grid these pumps are used to pump the water that has been collected in the reservoir below the turbine hall back into the higher reservoir above the turbine hall. The water is cycled between the two reservoirs to provide either power or energy storage as needed.
This type of plant is extremely robust and though roundtrip efficiency is lower than for some other technologies, long-term energy losses are low. Leak- age and evaporation are the main sources of loss and, if these are managed well, water loss can be kept small. Today this is the only technology available for very large-scale energy storage.
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 genera- tor 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.