Fullscreen HTML5Med Res (??MB)HTML4

One of the problems associated with renewable energy sources is that they are intermittent – they stop producing power when the Sun goes down or the wind stops blowing. Proponents of renewables see pumped hydroelectric storage and batteries as being the solution. Are they the answer? PUMPED HYDROELECTRIC STORAGE by Dr David Maddison Tumut 3 image by Colin Henein P umped hydroelectric storage is a method of storing or releasing large amounts of electrical energy to or from a nation’s electricity grid. Typically, it uses surplus electricity to pump water to a higher elevation and then later releases it through a hydroelectric generator back when it is needed. The gravitational potential energy of the water is the stored energy. Pumped storage was used as early as the 1890s in Austria, Italy and Switzerland for better management of water resources but not initially for storage of electrical energy. In the 1930s reversible hydroelectric turbines became available and the first pumped hydroelectric storage scheme was built near New Milford, Connecticut, USA in 1930, although in that case separate pumps were used rather than reversible turbines. Compared to other large scale electrical energy storage schemes, pumped hydroelectric storage is relatively cheap, requires little maintenance and with the right geography, can be implemented on a massive scale. It has the disadvantage of relatively low energy density, so a huge volume of water raised to a suitably high elevation needs to be utilised. The low energy density is the consequence of gravity being the weakest of all the natural fundamental forces. Grid-scale energy storage has traditionally been used for “load balancing”. This enables a power station to run at peak efficiency even though it means that at certain times it will be generating too much power. Rather than reducing the output of the power station, which could result in a loss of efficiency, its excess energy is stored. So the output 16  Silicon Chip The world’s first pumped hydroelectric storage power plant in Connecticut, USA, reproduced from Popular Science magazine, July 1930. Note how the pipeline consists of wooden staves in part, a common technology of the time and which was also used in parts of Victoria’s Rubicon Hydroelectric Scheme (see SILICON CHIP, February 2013). siliconchip.com.au A typical use of pumped hydroelectric storage. During the daytime, water flows downhill through turbines, producing electricity. At night, water is pumped back up to the reservoir using excess electricity, ready for tomorrow’s use. of the power station remains relatively constant against a varying electrical demand. It also means that electrical demands that exceed the total power of the generators can be met, for as long as there is stored water to discharge. This enables a smaller power station to be built than would otherwise be needed to satisfy peak demand. A typical application would be to store surplus energy at night (when demand is lowest) and release it when demand is highest during the day. Today, it is government policy to have a high and increasing amount of intermittent energy sources such as solar and wind generators to supply the grid. This causes very difficult grid management problems and instability issues. Pumped storage is one way to smooth over the constantly varying outputs of these intermittent energy sources in order to stabilise it. Even better, wind and/or solar generators could be used to directly supply power to pump water into a reservoir and thereby have no direct electrical connection to the grid. This approach will be used in some places such as the Espejo de Tarapacá project in Chile which will use solar power to run its pumps. Note that pumped hydroelectric storage is suitable to stabilise only relatively small amounts of intermittent energy and would not be suitable for backing up an entire grid which had substantial inputs in the form of intermittent energy. Huge amounts of storage would be required to do this. In most countries, the lack of suitable sites, the large cost Approximate proportion of grid-scale energy storage around the world. These are the latest available figures (2011) but current estimates suggest over 140,000MW of pumped hydro storage. These figures only specify the deliverable power, not the total time that power could be delivered. Note also that over 99% of grid scale storage is pumped hydro. siliconchip.com.au A typical pumped storage hydro plant. This one happens to be the Raccoon Mountain Pumped-Storage Plant in Tennessee, USA but its features are typical. This one has a natural lake as its lower reservoir, an artificial upper reservoir and it can produce over 1.6GW of power for 22 hours. The upper reservoir takes 28 hours to fill. and the environmental damage of such facilities would make them impractical. Unfortunately traditional forms of pumped storage generators and pump units are not especially well suited to smooth the rapidly varying outputs of solar and wind generators. However, new variable speed pump-generator units are available that are more suited to this application. Alternatives to pumped storage The worldwide installed capacity for grid-scale storage is overwhelmingly pumped hydroelectric storage, being over 99% of installed capacity. However, there are a number of other options for grid scale electrical storage and these include battery banks, large flywheels, or compressed/ liquefied air. Rechargeable batteries can be used to store energy but they are expensive and tend to degrade over time. Flywheels can store energy by being spun up by a motor generator set and then when energy is needed, the generator is used to produced electricity. But flywheels tend to be uneconomic in the sizes required for large scale energy storage. The King Island Hybrid Power Station in Tasmania is an interesting example of a power station that uses several power generation and storage technologies. It comprises a 2.45MW wind farm (nameplate, with a capacity factor of around 29% so the effective output is 710kW), a 100kW solar array and a 6MW diesel generator plant working with a backup battery and a flywheel. It originally used a vanadium redox flow battery which had a storage capacity of 800kWh and an output power of 200kW. However, the system was not robust and was replaced with a lead-acid battery with a capacity of 1.6MWh and a power delivery of 3MW. Another part of the system is a flywheel. This does not store a large amount of energy but is used as part of a “diesel rotary uninterruptible power supply (DRUPS)” whereby a flywheel is kept spinning as an energy reserve and when supply falls it drives a generator to supply power. If after some period of seconds supply does not increase, a diesel generator is started to make up the demand. The whole King Island system requires a $7 million dollar per year subsidy from the Tasmanian Government ($2,500 per person). You can see a real time schematic of January 2017  17 the system in operation, including power flows at www. kingislandrenewableenergy.com.au/ Compressed air energy storage typically utilises an old mine or geological structure such as an excavated salt cavern or depleted gas well to store compressed air at times of excess or cheap energy and then it is released through a turbine to generate electricity at times of peak demand. One company is developing bags of compressed air that are stored underwater. See http://hydrostor.ca/ Compressed air storage can also be used in conjunction with a natural gas turbine to improve its efficiency. A variation of compressed air storage is to liquefy the air and allow it to expand back to its gaseous state to generate electricity via a turbine. Hydroelectric turbines Only certain types of hydroelectric turbines are suitable for pumped storage if a single unit is required rather than utilising separate pumps to send the water back up to its reservoir. The type of turbine used in any particular application is determined by the water head and flow rate available. The three most common types are the Pelton wheel which is best for a large water head and low flow rate (for more on the Pelton wheel see SILICON CHIP “The Historic Rubicon Hydroelectric Scheme” February 2013, page 18); the Kaplan turbine which is best for low water head and high flow rate; and the widely-used Francis turbine which is good for a great variety of conditions, mainly medium head and medium flow rate applications. Unlike the Pelton wheel and the Kaplan turbine, the Francis turbine can also be used as a pump, making it ideal for use in pumped storage schemes. About sixty percent of the installed hydroelectric capacity in the world uses the Francis turbine. The Francis turbine can spin up quickly so changing power requirements can be quickly accommodated and it is available in a wide range of power capacities from a few kilowatts to 800 megawatts. The turbine consists of three main parts: the spiral casing, the guide vanes and the runner blades (or runner). In turbine mode, the spiral case distributes water around the periphery of the turbine inlet, after which it passes over the adjustable guide vanes, which Francis turbine, which can also function as a pump to reverse water flow. (Image courtesy Eternoo Machinery Co.) direct the flow onto the runner blades at the required angle for the present flow rate. The runner blades cause the tangential flow of water to be converted into rotational motion of the main shaft which turns an alternator. Variable speed hydroelectric generators Motor-generator equipment connected to a Francis turbine as used in hydroelectric storage schemes has traditionally only been able to be operated at a single speed and power rating. For example, if a plant had three 100MW generator units to be used for pumping water and there was 270MW of surplus energy to be utilised for pumping, only the first two units could be used to absorb 200MW of this surplus energy. The third 100MW unit could not be used as it would require 100MW to operate and only 70MW would be available so the 70MW would have go to waste. By contrast, a set of three variable-speed motor-generator units could adjust their speed to utilise all available energy for pumping and could each operate at 90MW. In addition, when operating in generator mode a variable speed unit can be adjusted for optimal efficiency of operation when only a partial load is being drawn. In a single speed motor-generator set the stator’s magnetic field and the rotor’s magnetic field are said to be coupled as they both rotate at the same speed. In a variable speed GE’s variable-speed hydro generator can run as either a generator or a pump. www.gerenewableenergy.com/ hydro-power/large-hydropowersolutions/generators/variablespeed.html 18  Silicon Chip siliconchip.com.au Aerial view of Tumut 3 Power Station. The red area contains the penstock (pipes) and power house. The orange area is the Talbingo Dam Reservoir, the upper storage of the scheme. You can explore this in more detail with the ability to zoom in and out at http://globalenergyobservatory.org/ form.php?pid=45928 Cross-section of turbine and pump arrangement at Tumut 3 power station. There are six generators which originally had a capacity of 250MW each but these were all upgraded to 300MW in 2009-11. Three of the generators have underslung pumps to pump water uphill for storage. unit these two magnetic fields are decoupled and either the stator or rotor magnetic field are fed via a frequency converter. A “double-fed induction motor-generator” (also known as a double-fed induction machine, DFIM) is the current standard design for variable speed motor-generators. It may be feasible and economical in some circumstances to convert an older fixed-speed storage plant to a variable speed one. See www.hydroworld.com/articles/print/ volume-21/issue-5/articles/pumped-storage/converting-tovariable-speed-at-a-pumped-storage-plant.html Video: “How does GE’s Hydro Variable Speed Pumped Storage technology work?” https://youtu.be/CDlvjkfpX_o, “GE Hydro Pumped Storage” https://youtu.be/2qZxfnMDrco micro-hydro generators, each of which has a power output of 140kW, were added to the outlets of the six generator cooling systems, which recovered otherwise wasted energy. Then in 2009-11 Tumut was upgraded with new turbine runners and other improvements to each of its six generators, increasing its overall power output from 1500MW to its present capacity of 1800MW (1.8GW), under ideal conditions. Even though the Francis turbines used at Tumut 3 could theoretically be used for pumping (as at other pumped storage facilities), in this case there are separate under-slung pump units for pumping water. Tumut 3 pumps water between its lower reservoir at Jounama Pondage and Talbingo Reservoir as its upper storage. The water head is approximately 155 metres. Snowy Hydro has not published the electrical storage capacity of Tumut 3 or the way it is used in typical operation but we estimate it as follows: there is approximately 160 gigalitres of active water storage. The six turbines (before the upgrade) had a total discharge capacity of 1,133,000 litres per second. This implies that it would take around 39 hours to discharge all active water storage at maximum power. Hence, there is about 70.2GWh of electrical storage. In energy storage mode, the three pumps each have a Australian pumped storage projects Australia has three working pumped hydroelectric projects in operation and one in the planning stage. Tumut 3 in the Snowy Mountains has the greatest power generating capacity with up to 1800MW output, followed by Wivenhoe in Queensland with 500MW and the Shoalhaven scheme in NSW with 250MW maximum output. The Tumut 3 power station of the Snowy Mountains Hydro-electric Scheme was Australia’s first pumped hydroelectric storage scheme, completed in 1973. In 2003 six Pumped storage calculations In calculating the power that can be generated by any hydroelectric project the two main numerical considerations are the water flow that can be directed into the turbine/alternator and the head of the water. These items scale linearly so doubling of either the flow or head will result in doubling of the power that can be produced. The power produced is given by the equation: siliconchip.com.au power (watts) = head (metres) x flow (litres per second) x gravity (9.8 metres per second squared) x efficiency factor Let’s do a real-world calculation for the Tumut 3 power station discussed above. We will consider the power produced from discharging the water and disregard losses from initially pumping it into the upper reservoir. It has a head height of 155m and a flow rate of 1,133,000 litres per second (prior to the upgrade). Without considering the efficiency factor, this yields 1721MW of power generation. Note that before the upgrade it had a quoted power output of 1500MW so this implies an efficiency of 87%. When doing calculations for pumped schemes consider that there is an efficiency loss in both directions. January 2017  19 Wivenhoe Power Station near Brisbane in Queensland. An aerial view can be seen at http://globalenergyobservatory.org/form. php?pid=45950 capacity of 99,000 litres per second (297,000 litres per second total) so the Talbingo Reservoir would take 448 hours to refill, assuming the lower reservoir could store all the water that was discharged. Of course, the storage is unlikely to be fully discharged in normal operation. Wivenhoe Power Station, located near Brisbane, is a 500MW pumped hydroelectric scheme which utilises a lower reservoir created by the Wivenhoe Dam and an upper reservoir created by the Splityard Creek Dam. The lower reservoir is approximately 100m below the upper one and is connected by two pipelines 420m long and between 6.8m and 8.5m in diameter. The power station has two 250MW pump-generator machines, said to be Australia’s largest hydroelectric machines, each having a rotating mass of 1450 tonnes. There is 5000MWh of capacity so, for example, 500MW could be produced for 10 hours. The station is connected to the grid via 275kV transmission lines. Like all hydroelectric schemes Wivenhoe has an exceptionally long expected service life – 100 years – and has been in service since 1984. A generator was added to the outlet of the Wivenhoe Dam in 2003 to provide 4.5MW and this is known as the Wivenhoe Small Hydro. It is not directly associated with the pumped storage scheme. (You may recall that the Wivenhoe Dam was associated with the Brisbane floods of 2011 and subsequent enquiry). The Shoalhaven Scheme is located on the South Coast hinterland of NSW and is used for water supply and up to 240MW of hydroelectric storage power. It has two combined power stations and pumping stations. The lowest one The diagram at left shows the Shoalhaven Scheme, which is a combined pumped hydroelectric system and a water transfer system to supply drinking water to Sydney, about 150km away. The Kangaroo Valley Pumping and Power Station (above) is the middle of three such stations. 20  Silicon Chip siliconchip.com.au The proposed Kidston Hydro Project will use two existing unused mining pits plus a “turkey’s nest” reservoir. is the Bendeela Power Station and has two 40MW combined pump-turbines to provide 80MW. In pump mode it can pump water to the Bendeela Pondage located 127 metres above. The Bendeela pondage is located below the Kangaroo Valley Power Station (1977) and has two 80MW power stations for a total capacity of 160MW. When operating in pumping mode it can pump water 480 metres up to the Fitzroy Falls Reservoir. The Burawang pumping station is not used for pumped storage but to pump water into the Wingecarribee Reservoir from where it can be released into the Warragamba or Nepean Dams. The scheme can produce 240MW of power. The proposed Kidston Hydro Project (about 1300km northwest of Brisbane, Qld) will utilise two mining pits which were formerly part of the now-closed Kidston Gold Mine. In addition, a “turkey’s nest” reservoir will be constructed to provide two storage reservoirs (an upper and lower) and a “balance reservoir” to effect a pumped storage scheme using mostly existing artificial structures. (A “turkey’s nest” reservoir or dam is one constructed above ground by a continuous wall built around the entire circumference of the contained water area. The amount of earthworks required for a turkey’s nest type of reservoir is typically considerably greater than damming a natural structure such as a valley.) There would be a vertical shaft from the upper reservoir and an underground generator station, with the outflow connected to the lower reservoir. According to a feasibility study by Genex Power, the proposer of this scheme, it would be able to continuously produce 250MW of electricity for six hours giving a storage capacity of 1500MWh. It would have two 125MW fixed speed turbines, a head height of between 194m and 230m and able to ramp up to maximum power in 30 seconds. However, according to a report on the Renew Economy website the power output will now be 450MW for five hours for up to 2250MWh of energy. This would involve building an upper reservoir that is 35-40 metres higher than originally planned. An associated solar PV array is also planned for the site. The scheme could be topped up with water if necessary with via a pipeline from the Copperfield Dam 18km away. There is also an existing 132kV transmission line that connects to a substation near Townsville. Plan of Kidston Hydro Project showing the main features of the upper reservoir, the vertical shaft from the upper reservoir, the underground power station and the transfer tunnel from the power station to the lower reservoir. siliconchip.com.au January 2017  21 As well, some sites have been identified as suitable for “turkey’s nest” dams based on elevation differences and horizontal distances between reservoirs but no costing or existing land use considerations were made. These include some on the Eyre Peninsula in South Australia and at Geraldton and Albany in Western Australia. The latter sites would use seawater and the sea as the lower reservoir. A cost estimate quoted for a cliff-top “turkeys nest” site in WA for a system that can produce 700MW to 800MW for six hours is $5 billion. In addition, other sites have been identified in northern Australia as part of a proposed scheme to export renewable energy to nearby Asian countries. Pumped storage projects from around the world As with all pumped hydro storage schemes water would be pumped to the upper reservoir at times of low demand and/or cheap electricity availability, and released during periods of high demand or high electricity prices. A unique feature of this project is that it is the first to propose using disused mines for pumped storage, to minimise costs. In addition, the facility offers a “blackstart” capability. This refers to the ability to start other power generators in the absence of grid power. This is a particular problem with wind turbines because they cannot start producing power unless there is pre-existing grid power available with which to synchronise their AC output. This factor contributed to the recent extended South Australian state-wide blackout. The Government’s Australian Renewable Energy Agency (ARENA) has committed $6.2 million to a feasibility study for this project and Genex Power Limited estimate the cost of building the facility at $282 million. They expect to commence construction this year and have it running in 2019. Proposed Tantangara-Blowering Pumped Hydro Scheme. In 2010 an independent geologist and engineer named Peter Lang proposed an enhancement to the Snowy Mountains Hydro-electric Scheme comprising a pumped storage system that could produce 9GW for three hours per day, after pumping water for six hours. Similarly, a lesser amount of power could be produced for a longer time, eg, 1.5GW for 18 hours. Tantangara would be used as the upper reservoir and Blowering as the lower reservoir, with a difference in elevation of 875 metres. Three 53km long, 12.7 metre diameter tunnels would be bored through to join the two reservoirs. More details about the proposal, discussion, cost and problems can be seen at https://bravenewclimate. com/2010/04/05/pumped-hydro-system-cost/ Now let’s look at some hydroelectric storage projects from around the world. The first utilises a turkey’s nest as the upper reservoir and the sea as the lower reservoir and water supply. It is significant because it requires only an appropriate elevation and no natural structures that can be dammed or a supply of fresh water. The other combines solar generation with a pumped hydro storage scheme. Fluctuations in solar electric production are automatically smoothed as the power is used only to pump water and is not directly fed into the grid. Finally we look at hydraulic rock storage. Okinawa Yanbaru Seawater Pumped Storage Power Station. This pumped hydroelectric storage power station in Japan was the first to utilise a turkey’s nest reservoir in combination with the sea as its lower storage reservoir and water supply. It was built as a pilot plant with a capacity of 30MW and was commissioned in 1999. It utilises a head height of 136m and has a flow rate of 26,000 litres per second from the reservoir which has a capacity of 564 megalitres, suggesting an electrical storage capacity of 180MWh. The system uses a variable speed turbine based upon a gate turn off (GTO) thyristor converter-inverter AC excitation system to provide maximum efficiency for both pumping and generation. As with many such structures the surface of the reservoir in contact with water is covered with an impermeable membrane to prevent water leakage. The Espejo de Tarapacá project in Chile is a 300MW capacity pumped hydroelectric storage project that uses seawater pumped 630 metres up to a natural depression in the Atacama Desert. It utilised three 100MW reversible Francis turbines which pump water uphill at 45,000 litres per second during the day and discharge it at night at 28,000 litres per second. The capacity of the pondage is 52 gigalites. The cost is US$400 million and construction is set to commence this year. It will be combined with a 600MW solar PV array by 2020 and the two plants working in combination will deliver solar energy 24 hours per day, stated to be without subsidies. Video: https://vimeo.com/152150996 Other potential sites in Australia Hydraulic rock storage A number of likely sites have been identified for pumped hydroelectric storage in Australia. One is for a pumped seawater scheme in Portland, Vic, associated with the Portland Wind Farm. Another study used graphical information systems to look for suitable sites in central Tasmania and the Araluen Valley in NSW. Heindl Energy GmbH (www.heindl-energy.com/) has developed a concept they called “gravity storage” or “hydraulic rock storage”. It utilises a large cylinder of rock that has been carved out of the ground. The system is “charged” by having water pumped in beneath the cylinder which raises it above ground level. When energy is to be released the Bird’s-eye view of Peter Lang’s proposal for a TantangaraBlowering Pumped Hydro Scheme. 22  Silicon Chip siliconchip.com.au Okinawa Yanbaru Seawater Pumped Storage Power Station. water is allowed to discharge through generators to create power. The water is forced up into an above ground pond. It has the advantage that large amounts of countryside don’t have to be occupied by dams and ponds. The economics of this concept are as follows: The storage capacity of the system depends on the mass of the rock and the height that it can be raised. If a rock cylinder is made which is the same height as its diameter the mass of the cylinder increases proportional to its radius cubed. For stability, the rock cylinder cannot be pushed out of the ground by more than half its height otherwise it could tilt. Since the height that the cylinder can be raised is the same as the radius and since the energy storage capacity is proportional to the mass times the height the cylinder is raised (the same as the radius) we can see that the energy storage capacity increases according to the radius to the fourth power. If the radius of the cylinder is doubled the storage capacity is increased by sixteen times. The construction of the cylinder involves cutting a circular channel to separate the cylinder from the surrounding rock and then undercutting the rock cylinder to separate it at the bottom. The circumference of the channel and base to be removed will be proportional to construction costs and doubles as the radius is doubled and the area of the base of the cylinder increases by four times as the radius is doubled for an increase of capacity of 16 times. To be conservative we could take construction costs to scale with the more expensive of these two operations, excavating the base of the cylinder which scales with the radius squared. Doubling the radius of the rock cylinder increases the capacity by sixteen times but the construction cost by only about four times. The capacity of a system with a 200m diameter cylinder would be 3GWh. This would provide less than 2kW continuously, for 75,000 people, for a period of 24 hours. It would contain 2,380,000 cubic metres of water at a pressure of 67 atmospheres. The efficiency of the system would be about the same as for pumped storage, 80% or so. Such a system would rise or sink 100 metres, at around 1mm per second. This system has a much higher energy density than a traditional pumped storage system and uses about one quarter the amount of water and much less land. It is expected to be long lived from an investment point of view, with a minimum asset life of 60 years and with low maintenance requirements. Heindl Energy is currently planning a prototype and ways to excavate the sidewalls, the base and a sealing mechanism on the sidewalls have been conceptually determined. The pilot project has a delivery date of around 2020. Videos on the topic: “Hydraulic Hydro Storage for 1600GWh of energy” https://youtu.be/zwVMl_4QRk8 This video shows an earlier implementation of the sealing ring system required to keep water contained. “TEDx Talk Hydraulic Hydro Storage” https://youtu.be/ m3p_daUDvI8 “Comparison of different storage technologies” https:// youtu.be/IZqUut5rNaY The Gravity Power Module This concept from Gravity Power (www.gravitypower. net/) is similar to Heindl’s hydraulic rock storage however in this case the piston does not rise above ground level. Rather than water being pumped between a ground level reservoir and beneath a rock piston as in Heindl’s scheme, in this scheme water is transferred to and from beneath the piston and the area above it. The cost of building the enormous shaft in the ground is claimed to be “surprisingly low”. A Francis turbine would be used for pumping and generation. A proposed design to provide 40MW for four hours would require a 500 metre deep main shaft of around 32.5 metres in diameter, with a 250 metre tall piston of natural rock Artist’s impression of the Espejo de Tarapacá project in Chile. siliconchip.com.au January January2017  23 2017  23 Economics As is the case for all energy storage systems, pumped hydroelectric storage is not 100% efficient. This means that the electricity generated from release of water is less than that required to pump the water into its upper reservoir in the first place. In some implementations of pumped storage, cheap electricity generated during off-peak times is released during peak times when the electricity price is higher. The higher price that the electricity can be sold for during peak times more than offsets the typical 20% loss of energy involved in pumping the water to its upper reservoir as well as taking into account capital costs and running costs of the storage system. excavated from the lower 250 metres of the shaft. The adjacent power house shaft would be 10 metres in diameter. Similar capacity and scaling considerations to the Heindl’s hydraulic rock storage apply to the Gravity Power Module. Heindl Energy’s gravitational storage concept showing rock cylinder, seal (purple), water beneath rock cylinder (blue), underground pump and generator chamber and above ground pond for water. Once that storage was discharged it would take many more days to recharge the storage as noted previously and one would hope the wind would return after 70 hours and stay for a long period. According to the Australian Energy Market Regulator there is currently an installed electrical generation capacity of 48,116MW. If 50% of that was replaced with wind or solar, we would need 48 Tumut 3 systems as backup, even to allow for just a few days without wind or sun! Various storage issues have been considered for wind power and are considered at https://stopthesethings. com/2016/08/31/bulk-battery-storage-of-wind-power-amyth/ and http://euanmearns.com/estimating-storage-reSC quirements-at-high-levels-of-wind-penetration/ Pumped storage and wind power While pumped storage systems do have their advantages, they do not solve the problems of intermittent energy such as solar or wind. To see why, we must consider the enormous amount of energy required by modern society and the low density of energy production of intermittent sources such as solar and wind. Take for example the replacement of a modest 1GW fossil fuel or nuclear plant with wind turbines. As wind turbines typically operate only one third of the time or less, you would have to have three times as many windmills as their nameplate capacity would suggest. So 3000 1MW windmills would be required to generate the same amount of energy as the fossil or nuclear plant working continuously, even before we consider how to store energy for later use when the wind is not blowing. In the Australian context the only existing storage facility that could deliver that much power would be Tumut 3 with a presumed capacity of 70.2GWh. This could provide backup for around for a 1GW system for around 70 hours or just under three days, to account for a condition of no wind. (Left): the Gravity Power Module showing water flow and position for both generating and storage modes of operation . 24  Silicon Chip (Right): detail of generator portion of Gravity Power Module which is located beside the storage shaft. siliconchip.com.au