The wind is a plentiful but unreliable source of energy. As wind power gradually takes over an increasing portion of the load in the grid, it becomes crucial to have a means of storing energy.
By Kent Krøyer
Energy can theoretically be stored in several ways, but very few of these are presently ready for use. One is the storage of river water in large mountain lakes above hydroelectric power plants.
Simply putting the hydroelectric power plant on pause while using wind energy results in the storage of vast amounts of potential energy as the lake is filled with additional water from the river. The rising water level in the lake becomes a charging battery for the grid. This is a wellknown principle, and the technology is mature, but it involves two prerequisites:
A mountain lake and a river.
The Geesthacht pumped storage facility in Schleswig-Holstein, Germany has an efficiency of around 75%. The lake holds a maximum of 3.2 million m³ water. It stores 600 MWh in nine hours with a 96 MW pump capacity, and it can carry a load of 120 MW Islands of energy at sea for five hours. Photo: Vattenfall
Pumped storages are quite efficient
If you are less fortunate and only have a mountain lake but no river, you will need pumped storage. In this case, a large electrical pump lifts water from a lower reservoir to a higher mountain lake whenever energy in the grid is cheap. Much of the technology is similar to that of the aforementioned hydroelectric plant. Examples of pumped storage can be found in Germany with a total effect of 7,000 MW. Almost half of this power comes from eight plants owned by Vattenfall. The largest pumped storage facility in Germany is Pumpspeicherwerk Goldistahl in Thuringia, which was planned by the DDR in 1965 and finally completed in the reunified Germany in 2003. It features two fixed and two variable RPM pumps/turbines, which offer power flexibility in both directions. Energy efficiency is around 80% (i.e., 20% of the energy is lost).
Power lines create markets
Flat countries like the Netherlands have no direct means of using these kinds of energy accumulators.
One solution to the lack of mountain lakes is to build power lines to neighboring countries, including mountainous countries. This provides an opportunity to buy and sell energy between countries – in principle, the same as storing it in colossal batteries. The energy efficiency of powerlines is high, close to 96% for a round trip. An example is Skagerak 4, the forthcoming Danish subsea DC cable to Norway.
“Most of the loss comes from the conversion from AC to DC and back, around 0.8% for each conversion. The loss from the DC undersea cable is quite small,” says Søren Damsgaard Mikkelsen, who is project manager for Skagerak 4 at Energinet.dk, Denmark’s transmission system operator.
Power lines and DC converters are expensive, though. Skagerak 4 will cost 430 million Euros, half of which will be spent on the power conversion equipment alone. However, with sufficient international power lines, a country can become a transit market for energy and profit from regional price differences.
In contrast, having too little power line capacity could mean that wind farms have to be taken off-grid when the wind blows. A solid argument for energy storage capacity as a supplement to power line capacity.
Mountain lakes for flat countries
Another solution to the lack of mountains was invented a few years ago when two Danish engineers went fishing in a small boat. During the trip, they came up with an ingenious substitute for mountain lakes especially for flat countries: The buried pumped storage. The invention’s official name is Energy Membrane, and the patented idea is simply to bury a large, rugged balloon under a massive load of sand or dirt. A pump fills the balloon with water when energy on the grid is cheap, and potential energy is stored as millions of tons of sand are lifted several meters. Energy is returned to the grid when the stored water passes through a turbine that drives an electrical generator. A small demonstration model proved the principle, and a larger 50 × 50 m demonstration model was tested in 2011. It repeatedly lifted a three-meter thick layer of earth one meter into the air.
“We gained some practical experience on how the membrane needs to be laid out and how to weld the edges together. But the test results were as expected. There is an energy loss of almost 1% in the membrane and in the earth. The losses in pump/turbine and motor/generator are not included in this figure as these components are not the final choices. We still expect an overall efficiency of around 80% at full scale, just like other pumped storages,” says the CEO of GODevelopment, Asger Gramkow.
The next milestone is a pilot-project sized 200 × 200 m. The final full-scale storage will cover an area of 500 × 500 m. It will have an effect of 30 MW and a stored capacity of 200 MWh. So far, there have been no technical showstoppers.
50 × 50 m demonstration model being laid out. Pieces of plastic are welded together. Later comes the expensive part – placing a thick layer of earth on top. Photo: GODevelopment
Storing energy as heat is irreversible
Considerable amounts of energy can be stored as heat in district heating plants. It is a means of solving a problem at the times when wind energy supplies a large share of the grid power. Power plants with cogeneration of heat and electricity are generally unable to turn down electricity production when demand for heat becomes higher than demand for electricity. This typically happens during cold winter nights. A consequence can be an export of very cheap electricity.
Furthermore, winter nights often coincide with strong winds. This means that all of the wind farms kick in with a superfluous production of electrical energy at a time when demand for electricity is particularly low. The result is that the price per MWh can go negative. That is when electrical power boilers come to a rescue.
Heating plants can store considerable amounts of energy just by raising the temperature in their large water tanks by a few degrees. An alternative to an electrical power boiler is a heat pump, which is more energy efficient.
Boilers and heat pumps are extremely energy efficient, and the technologies are well known. But once converted to heat, the energy is lost from the grid.
Compressed air: Short term storage
Compressing large volumes of air, CAES (Compressed Air Energy Storage), represents a means of storing energy for the grid. As with water storage, this requires a pump (compressor) to fill a tank. To extract the energy as electricity, you need a turbine with a generator. In contrast to water, however, air is compressible like a spring. Due to thermodynamics, the temperature rises while the spring is being loaded. This creates a problem in relation to the storage of energy since wasting the compression heat represents a considerable loss relative to the accumulated energy, stored as air pressure. The use of the stored energy is also problematic since the compressed air becomes freezing cold when it is decompressed. In order to protect the turbines, it must be reheated when it is time to use it.
The energy efficiency of CAES plants is therefore rather low, around 45-50%, and the technology is not yet mature. A CAES plant requires a monstrously large cavity for storing the compressed air. A subterranean cavern such as an old salt mine or water reservoir with long pipes deep beneath the Earth’s surface is ideal for this purpose. A working example is the CAES plant in Huntorf, Germany, which was the world’s first of its kind. It was built in 1978 and has been supplying reserve capacity for the German grid ever since. The maximum effect is 290 MW, which it can supply for three hours. It uses two underground caverns, totaling 310,000 m³. The maximum air pressure is 70 bar, and the energy efficiency is 42%. The problem of cold decompression is solved by adding and combusting natural gas in the process, which provides extra power to the turbines at an extra cost. A new CAES plant, ADELE, is being built in Germany by RWE Power. This plant will achieve an efficiency of 70% due to a new solution to the temperature problems. The plant capacity will be 1,000 MWh, and pressures will top at 100 bar.
Instead of venting the precious compression heat away, this plant stores it in an isolated regenerator tank on the way down to the subterranean air storage cavern. The working principle of the regenerator is thus not far from the function of the regenerator in a Stirling engine. In the CAES case, it is a large tank filled with special ceramics that will retain much of the heat generated by the compression – up to 600° C. When the stored energy in the compressed air is to be used, it will be decompressed through the regenerator, thus picking up the stored heat on the way up to the turbine. The regenerator must therefore be heavily insulated to minimize energy loss. This CAES plant will therefore provide energy storage for a matter of days rather than of weeks.
Research needed on chemical batteries
Storing large amounts of energy in chemical batteries is a goal for many researchers worldwide. The ubiquitous lithium-ion battery is popular in cellular phones and laptop computers because of its high energy density, but this technology is far too expensive when it comes to the large storage needs of power plants. The equally well-known lead-acid battery technology is less expensive, but it does not take kindly to deep discharges. Lead-acid is, however, a candidate for improvement.
Sodium-sulfur batteries have been used several times for grid energy storage, especially in Japan. Its energy density is high, its efficiency is high (around 90%), and the battery withstands the wear of 2,000-5,000 power cycles. Furthermore, the price tag is acceptable. This battery nevertheless has drawbacks. It requires an operating temperature of 300° C, and the chemistry is extremely corrosive. It also needs to be protected from moisture since the hot sodium will explode if it comes into contact with water. The city of Presidio in Texas installed a large sodium-sulfur battery as power backup in 2010. It can supply the entire city with 4 MW power for eight hours.
Energy density is not an important factor when it comes to grid energy storage, while price and reliability are. Powerful new chemical combinations are on researchers’ drawing boards (such as vanadium flow batteries and lithium-air), but the components for such technologies are not yet ready for use.
The concept of large, centralized energy storage may not, however, be the right choice. If the future involves widespread use of electric cars, then their relatively small batteries may serve collectively as a distributed energy reserve, even if the battery technology is deemed unsuitable for centralized energy storage. The idea is that most cars are parked somewhere most of the time anyway. So, while the owner sleeps or works, the car battery may serve the grid operator to mitigate the day’s peak loads. Basically, this requires a plug for the grid connection and an ingenious control system that allows an easy negotiation of terms between the operator and the owner. The owner needs to be sure that her car is sufficiently charged when she plans to use it. In the meantime, the battery capacity could be sold to the grid operator – leaving the owner as a new micro-dealer at a local energy spot market. This beautiful idea is, however, still just a sweet fantasy.
“It is certainly technically possible to build this kind of infrastructure, but there are important obstacles to consider. First of all, there has to be enough value for the car owner to let the grid operator wear his expensive battery, and that may not necessarily be the case. The same value point of view goes for all parties involved in the structure,” says Per Lund, Chief Engineer at Research and Development, energinet.dk. He emphasizes that the payment structure alone demands an internet-connected clearing house to identify unique car owners, just like account payments for cellular phones. And this is no easy task.
“Society was ready to pay for the cellular structure, and it may also be prepared to do it again. No one knows yet,” he says.
The same car batteries can furthermore be reused for power peak shaving if they are collected by the grid operator once the car owner discards them because they are worn out. Even at half capacity or less, they can be of use as a central energy storage if there are plenty of them and the price is low. This idea would also insure that car batteries are collected for an environmentally sound remining of any precious metals.
Commercial hydrogen energy storage is now available
An old idea, the conversion of electrical energy to hydrogen via electrolysis, is now reaching maturity, and further improvements to this technology are not far off.
The Danish company H2Logic is now building a series of small commercial hydrogen stations, where owners of hydrogen cars can fill their tanks to a pressure of 700 bar. The hydrogen stations are, in fact, small chemical factories, where hydrogen is formed by an electrolytic process that consumes electricity. In principle, the production of hydrogen at the stations can be done at night when electricity is less expensive. At the moment, however, there are too few hydrogen cars on the road to make an appreciable difference. The hydrogen station in the city of Holstebro in Denmark has made an agreement with the grid operator to use only green power from wind turbines.
“The energy efficiency is not high, around 50-60% percent from electricity to hydrogen and again 50-60% from hydrogen to electricity when fuel cell technology is used both ways. Total efficiency for hydrogen storage of electrical energy is thus around 25-30 percent. The lost energy disappears as heat,” says professor Søren Linderoth, Head of DTU Energy Conversion.
Søren Linderoth, however, confirms that this heat loss is useable, for instance if the hydrogen energy storage is located at a combined power/heat plant.
Furthermore, Danish scientists at Risoe DTU and the Haldor Topsoe company have developed a new electrolytic process that has an energy efficiency of close to 100%. The process uses an SOEC (Solid Oxide Electrolysis Cell), which is related to SOFC fuel cells but with the opposite function. It converts electrical energy to hydrogen at 800-1,000° C.
“This technology has great potential as energy storage for wind energy. The companies that want to use this process expect it to be ready approximately five years from now. I am not sure about that, but it certainly is not like putting a man on the moon,” he says.
This high-temperature technology also makes it easy to convert the hydrogen to methane, which can be distributed via gas pipes as with the Danish natural gas transmission system.
“It is also a solution to the problem of switching the transport sector to renewable fuels. In Italy, for example, many vehicles are built to be powered by natural gas,” he says.
Other means of energy storage are still too small Energy can be stored in a multitude of other interesting ways, such as supercapacitors, superconductor rings, and fast-spinning flywheels. So far, these technologies are either unready or unsuited for the heavy storage needs of the transmission system operators except for balancing the grid and for short timed peak shavings.
Report on CAES plant use in Huntorf, Germany
Adele, the adiabatic CAES project
Illustrated video of ADELE
Presentation of ADELE
The sodium-sulfur battery in Presidio, Texas