Pumped hydro currently provides most of the energy storage for the electricity industry, offering large-scale, low-cost, off-the-shelf energy storage in unlimited quantities.
This article was originally published in the August 2024 issue of create with the headline “Under the hydro pump”.
Australia’s future solar-dominated energy system will need long-duration storage capable of riding through night-time and a “wet windless week in winter”: 10-100 hours or more.
This is the realm of pumped hydro, with its very low energy storage cost and its operational lifetime of a century or more. The long-duration energy storage requirements in the 2030s will be much larger than current energy storage needs.
Pumped hydro energy storage
Pumped hydro energy storage (PHES) constitutes most current energy storage for the global electricity industry.
PHES typically entails two reservoirs, separated by an altitude difference of 100-1600 m, spaced several kilometres apart and connected by a pipe or tunnel containing a pump turbine. Water is pumped uphill on sunny and windy days, and returns downhill through the turbine to recover the stored energy when required.
The water can shuttle uphill and downhill in a closed-loop cycle for many decades. The round-trip efficiency of pumped hydro energy storage is typically 80 per cent.
Most existing PHES is located on rivers, usually in conjunction with hydroelectric systems. There is often resistance to construction of new dams on rivers. However, most potential PHES sites are located away from rivers (“off-river”) because most of the global landscape is located away from rivers. For off-river PHES, the two reservoirs are typically located in hilly country, and unconnected to any river, and have combined flooded area in the range 1-10 km2.
The Australian National University produced the Global Pumped Hydro Energy Storage Atlas, which lists about one million PHES sites around the world that do not require new dams on rivers. Energy storage volumes shown in the atlas are 2, 5, 15, 50, 150, 500, 1500 and 5000 GWh. About 100 times more sites are shown in the atlas than would be required to support a fully decarbonised global economy of 10 billion affluent people.
Users can pan, zoom and rotate to a scale of 30 m and download 26 items of technical information for each site. They can select the 3D terrain map setting to visualise individual sites.
Pumped hydro storage comprises both an energy cost ($/GWh, reservoirs) and a power cost ($/GW, tunnel and powerhouse) that can be sized independently. The marginal cost of increasing the storage volume (hours of storage) is low – simply increase the dam wall height, while the tunnel and powerhouse remain unchanged.
Good sites in the atlas are marked with red dots (Class A or Class B), while premium sites are marked with triangles (Class AA) or stars (Class AAA). Premium sites are characterised by large-scale (0.5-5 GW of power for dozens of hours); large head (400-1600 m); large slope in the range 5-25 per cent (head divided by horizontal separation); and large water-rock ratio in the range 5-25 (ratio of the volume of stored water to the volume of rock needed to construct the reservoir walls).
Premium sites in the Global Pumped Hydro Atlas have an indicative cost as little as one tenth that of Class E sites.
Greenfield sites in the atlas require two new reservoirs, while bluefield and brownfield sites use existing lakes and old mining sites respectively. Ocean sites use the ocean as the lower reservoir. Seasonal storage sites slowly (over weeks or months) draw from or discharge water to large nearby rivers.
Pumped hydro in Australia
Australia has extraordinarily good long-duration pumped hydro energy storage sites. It has approximately 5000 good PHES sites and only needs about a dozen (depending on size); energy planners can afford to be very choosy. Storage volume in Australian sites ranges up to 5000 GWh – equivalent to 5000 big batteries (one GWh each) or 50-100 million EV batteries.
Australia has three existing pumped hydro systems (Tumut 3, Kangaroo Valley, Wivenhoe); two under construction (Kidston, Snowy 2.0); and dozens of proposed projects including Pioneer Burdekin, Borumba and Battery of the Nation.
Snowy 2.0 has two GW of storage power and 360 GWh of storage energy (larger than all the utility batteries in the world combined) at a cost of $12 billion, which equates to $33 per kWh for a system that will still be operational in a century. This is far below the cost of equivalent batteries. However, batteries are better for storage up to a few hours. Both PHES and batteries are required for an optimum energy storage system.
Water and land requirements for pumped hydro are small. Averaging across 150 Australian Class AA greenfield sites reveals water and land requirements below one GL and 10 hectares per GWh respectively.
The water is recycled indefinitely between upper and lower reservoirs. For comparison, 8000 GL of irrigation water was applied to two million hectares of crops in Australia in 2022.
If Australia’s entire future energy storage requirement of about 1000 GWh were to be met using off-river PHES, the water requirement for the initial fill would be about 1000 GL spread over a construction period of about 20 years. This amounts to about five litres per person per day (equivalent to half a minute of a morning shower). The land requirement would be 100 km2 which is equivalent to 3.3 m2 per person – about the area of a king-sized bed.
All other things being equal, a site with double the head compared with another site would have double the energy storage and double the storage power, but considerably less than double the cost. Heads in Australia range up to 1300 m.
Large head is an important attribute. It is striking how many pumped hydro proposals in Australia utilise heads below 300 m when much larger heads are readily available. For example, a sample of 150 Australian Class AA sites with energy storage volume of 50, 500 and 1500 GWh reveals a typical head of 600 m.
The Lithgow site is at a strategic location just west of the Blue Mountains. New high-power transmission into Sydney to accommodate the “electrification of everything” is constrained by national parks. Placing a large buffer storage inland from these national parks to absorb the daily output of solar and wind farms allows maximum use to be made of the existing transmission.
The site is close to the 500 kV powerline that runs from the Hunter Valley south to Goulburn, which allows the storage to service existing transmission entering Sydney from the north, west and south-west.
The sites near Coffs Harbour are typical of the Great Dividing Range, comprising upper and lower reservoirs in generally cleared land with steep forested land in between. A tunnel can connect the upper and lower reservoirs with low disturbance to the land in between. Pioneer Burdekin has such a configuration. In the case of Coffs Harbour, heads of up to 1300 m are available.
There is a Class AA bluefield site (150-500 GWh) located near the Snowy 2.0 tailrace, on the west side of Talbingo Reservoir. Once the Snowy 2.0 tunnel is finished, it may be possible for the tunnel boring machine to tunnel west beneath Talbingo Reservoir and then build the required four kilometre-long tunnel for a “Snowy 3.0”.
This would allow continuity for the Snowy 2.0 workforce, tunnel boring machine, tunnel-lining concrete plant, work camps, transmission easement and access roads.
Pumped hydro in Tasmania can buffer Tasmanian wind generation to ensure maximum load factor in existing and new Bass Strait power cables. In South Australia, heads of 300-600 m are available in 15-150 GWh sites near Port Augusta. Development of such sites would provide South Australia with a large energy storage buffer in case of failure of transmission connection to the east. It would allow the excellent solar and wind resources in this region to make maximum use of the existing transmission into Victoria and NSW.
In Queensland, the Pioneer Burdekin and Borumba pumped hydro proposals offer storage energy and storage power of 170 GWh and seven GW respectively. This is a large fraction of the ultimate storage requirements for Queensland.
Victoria has many greenfield and bluefield site options northeast of Melbourne from which to choose. There are also interesting brownfield options that utilise the Yallourn, Hazelwood and Loy Yang coal mining sites. Perth has relatively few good pumped hydro options because maximum available heads are about 200 m. Interestingly, the Pilbara has several large sites that could support the mining industry.
Professor Andrew Blakers is a world-renowned leader in renewable energy development and Professor of Engineering at the Australian National University.
This article is an edited extract from Engineers Australia’s Accelerating the Energy Transition series. Read the full article alongside Alan Finkel’s view on the critical barriers to transition and Neil Greet’s take on energy security.
Pumped hydro is not a source of power, it takes power from the grid off peak at a cheap price then stores it to be sold back at a higher price The response time for hydro to load demand is very slow so most times you are taking power then generating just in case power to compensate for expected surges and demands. There is a far better way no one wants to discuss