Pumped hydro could solve a lot of problems in the electricity market as variable renewable energy sources replace coal-fired generation. EcoGeneration asked a panel of experts how this massive form of storage can reshape the NEM and set us on track for a clean energy future.


Panel participants
  • Andrew Blakers, director of the Australian National University Centre for Sustainable Energy Systems
  • Simon Kidston, executive director, Genex Power
  • Martin Kennedy, head of sales for hydropower Australia, GE Australia
  • Roger Fulton, hydro engineering consultant, GHD
  • Chris Gwynne, program director, Battery of the Nation, Hydro Tasmania
  • Roger Whitby, chief operating officer, Snowy Hydro

Although not much is known about the reliability and emissions settings for the National Energy Guarantee, what role do you see pumped hydro playing if the policy is adopted?

Andrew Blakers, ANU: Regardless of the federal policy adopted, storage will be required to cope with the ever-increasing penetration of variable renewable energy. PV and wind are now the first and second energy technologies in terms of net new generation technology installed worldwide, with coal relegated to third spot. PV and wind constitute virtually all new generation technology installed annually in Australia. The current build rate of PV and wind in Australia is 3GW a year. If this is continued until 2030 then more than 50% of electricity will be renewable.

Simon Kidston, Genex Power: Pumped storage hydro is crucial for the Australian electricity market as it provides reliable, affordable and dispatchable generation. In addition, it stabilises the grid, particularly as the penetration of renewables increases, given renewables by their very nature are intermittent and volatile.

Martin Kennedy, GE: Depending on the final design of the National Energy Guarantee, pumped hydro has the potential fill the dual obligations of the policy by providing dispatchable power to meet the reliability obligation and enabling higher penetrations of renewable energy to meet the emissions obligation. Grid stability requires both short term (fast response) reliability and longer term (load shifting) reliability. Pumped hydro comes into its own when it comes to load shifting, which requires gigawatts of power to be shifted for hours and days at a time. It is the most common and cost-effective way to do this.

Roger Fulton, GHD: Low-cost wind and solar will inevitably replace ageing coal stations as they are retired. Wind and solar will inevitably, and fortunately, drive Australia from its present 18% renewable generation to 80-100% renewable generation, possibly by 2030. As wind and solar as a proportion of the total power generation rises above about 50% of the total, the countrywide system will require storage to provide system stability which was previously an inherent part of large rotating generators in the coal-fired power stations.

Emissions settings will be improved by the clean renewables wind and solar, but wind and solar is subject to changes in supply – the sun goes down each night and the wind strengthens and decreases. This means reliability also requires the storage of energy during times of excess, and generation from this stored energy during times of shortage. Both pumped storage and battery technologies can be constructed for almost whatever size is required, and both technologies currently cost about the same per megawatt. However, pumped hydro has much more energy storage, can last indefinitely and at present prices is very much cheaper per electricity unit by a factor of between 5 and 15.

Chris Gwynne, Hydro Tasmania: Pumped hydro energy storage can play a critical role in medium to long term storage in the national electricity system and also the provision of “boutique” or specialised ancillary services in some instances. These attributes will help enable and optimise the introduction of new renewable energy into the system.

Pumped hydro also has the advantage of being able to positively impact all three elements of the energy trilemma: reliability, affordability and emissions reduction. Other technologies can meet some but not all aspects – for example, they can meet reliability or emissions reduction requirements but come at a higher cost (impacting on affordability).

Roger Whitby, Snowy Hydro: In terms of the National Energy Guarantee pumped hydro meets both the reliability guarantee and emissions reduction requirements, as pumped hydro is dispatchable and flexible from both the generation (supply side) and pumping (load side) operation, and it firms up intermittent renewables and supports their future growth.

Whatever the emissions and reliability settings of the National Energy Guarantee pumped hydro is well suited to contribute to it, says Roger Whitby of Snowy Hydro.

What improvements would need to be made to the design of the National Electricity Market for Australia to get the most out of pumped hydro?

Blakers, ANU: The NEM needs to include adequate reward for the provision of large-scale storage (10-100 hours), spinning inertia, rapid power response (20-200 seconds) and black start capability – all of which is available from PHES systems.

Kidston, Genex Power: Pumped storage hydro is already favoured as the dominant form of energy storage in Australia, with three large-scale projects running for a number of decades and a further three proposed for development (including the Kidston Pumped Storage Scheme). The market is transitioning from a stable, slow, carbon-based market to a more rapid, intermittent renewable market. As the penetration of renewables increases, pumped storage hydro will fit perfectly into this new marketplace, given the ability to generate almost instantly on demand and provide frequency control and stabilisation services.

Kennedy, GE: The NEG needs be designed in a way which provides clear and unambiguous signals to support ongoing investment in the National Electricity Market. While I couldn’t comment on specific improvements, we are looking forward to further details on the National Energy Guarantee and contributing to the consultation process. Pumped hydro has been operating successfully in Australia since the 1970s, but there are certainly mechanisms that could be considered if we want to drive further investment.

Fulton, GHD: Significant improvement would be valuing – and paying for – long duration storage. Most of the time energy storage will only be required for fairly short durations – 2 or 3 hours, maybe up to 8 hours. As the system or weather perturbations become deeper and more systemic, more hours of storage becomes more valuable in supporting the absence of renewable generation, such as during cyclones or severe storms (wind turbines cannot operate in very strong winds).

Other changes in the NEM would place high value on rotating inertia, spinning reserve and fast instantaneous reserve that balance out load and frequency changes in the system, which pumped storage schemes provide in high quality. Variable speed pumped storage units could be rewarded for being able to provide greater ratio of effective inertia, and hanging in longer during under-frequency events.

Gwynne, Hydro Tasmania: The market needs to better understand the technical requirements of the system of the future and ensure market signals (existing or new) encourage the most cost-efficient and technically effective solutions. The market may need to decide that some system support services could be best provided by non-market (government or regulated) mechanisms. The market would also need to value certain reliability and system security requirements like inertia that have historically been provided “free” (largely) by the coal generators.

Whitby, Snowy Hydro: While not strictly within the scope of the market the regulation of monopoly transmission needs to be improved through longer term planning and an update of the regulatory investment test to ensure that new transmission lines can be built to support renewables development. Early signs of longer term transmission planning can be seen through the work AEMO is undertaking through the Integrated Grid Plan.

Pumped hydro would benefit if the National Electricity Market were adapted so that a value could be placed on long-duration storage, says Roger Fulton of GHD.

How quickly can new pumped hydro facilities be built and what are the requirements for smooth project delivery?

Blakers, ANU: New off-river PHES systems can be built in 2-4 years. Thus, PHES facilities can be readily built at sufficient speed to keep up with high PV and wind deployment rates. The very large number of potential off-river PHES sites allows sites to be avoided if they are problematical from environmental, heritage, geological, hydrological, access or other considerations. Standardization of off-river design, and the avoidance of rivers and remote sites, allows highly parallel construction of the twin reservoirs, the penstock/tunnel and the electromechanical components.

Kidston, Genex Power: The build time of any pumped storage hydro scheme depends on the size of the facility and the location. However, as a general rule of thumb there are about two years of technical feasibility studies and engineering reviews, followed by 3-4 years of construction time. By utilising an abandoned gold mine, Genex was able to greatly reduce construction cost and time.

Kennedy, GE: These are long-lived investments. Assuming permits and such are all in place, construction can take 3-4 years at least, or longer in the case of larger projects with significant tunnelling or civil aspects.

Fulton, GHD: Given that over 20,000 potential pump storage sites have been identified by the ANU study, and probably about 200 are required nationwide, this means that only the very best sites need to be developed. Virtually any new generation project can take a year of planning, permission and consultation before construction starts. But a very good pumped storage site will probably be within easy drive on good roads (as compared with the remote sites that are usually associated with river-based hydro schemes). This makes it easy for the largest deployment of labour and machinery. Two years construction would be a very fast implementation; three years construction reasonably achievable.

Gwynne, Hydro Tasmania: The speed of project delivery is site-dependent, particularly in relation to whether the project is greenfield or utilises existing infrastructure. In Tasmania, utilising existing hydropower infrastructure and scalable sized developments would mean that pumped hydro projects could be brought online within 3-4 years. Effective stakeholder management, early approvals, co-ordination with transmission providers and site selection to minimise issues and risks are all keys to smooth and successful project delivery.

Whitby, Snowy Hydro: Snowy Hydro is currently preparing for final investment decision for the Snowy 2.0 pumped hydro expansion project. First power from Snowy 2.0 is expected in 2024.

Pumped hydro projects that make use of existing hydropower infrastructure could be brought online in about three to four years, says Chris Gwynne of Hydro Tasmania.

What are some myths about pumped hydro that need to be cleared up to reassure any doubters in Canberra?

Kennedy, GE: Myth 1: Pumped hydro requires a lot of land and water. Fact: While traditional hydro can often require damming rivers and flooding valleys, pumped hydro sites are much more compact and don’t require much water as it is continuously recycled up and down through the system.

Myth 2: Pumped hydro is expensive. Fact: Pumped hydro is by orders of magnitude the cheapest way to store energy at GWh scale. This has been demonstrated around the world and is why countries such as Portugal, China and Switzerland are turning to pumped hydro to stabilize their grids and support growth in renewables.

Myth 3: Batteries will soon overtake pumped hydro on cost. Fact: The prices discussed in the press for the big battery in SA range from $400-600/kWh of installed storage capacity, which is 2-3 times the equivalent price for the Energy Australia-Arup seawater pumped hydro plant in South Australia. And this is without even allowing for the fact that pumped hydro plants last about six times longer than batteries.

Fulton, GHD: That you need existing lakes to start with (off-river storage has many advantages over existing waterways); that you must be beside a river to provide initial filling of the reservoirs (even trucking the water into a pumped storage scheme is estimated to cost only about 1% of total capital cost); that evaporation losses are very large (in most areas annual rainfall exceeds annual evaporation. In dry areas, evaporation control measures can largely mitigate evaporation losses).

Gwynne, Hydro Tasmania: Firstly, validity of the technology: Pumped hydro is a proven global technology – it is just not that common in Australia. It has a very low environmental footprint (small new reservoirs) and has multiple uses in a power system (not just energy storage).

Secondly, efficiency: Pumped hydro cycles typically run at about 80% efficiency. While there are energy losses at an individual site, it is important to look at efficiency on a whole-of-system basis. The system efficiency is improved because the system can use otherwise wasted energy (such as any excess from wind and solar generation), particularly when done in conjunction with existing hydro generation systems.

Whitby, Snowy Hydro: There is often this misconception that pumped-hydro is a “new thing”, but it’s a proven technology. Snowy Hydro has operated pumped hydro from Tumut 3 power station since the 1960s. Pumped-hydro operating in the market is commercially viable.

Martin Kennedy of GE says pumped hydro is the cheapest way to store energy at gigawatt scale “by orders of magnitude”.

Do you favour large remote facilities, smaller ones close to capitals or a bit of both?

Blakers, ANU: A GIS survey by ANU has uncovered about 22,000 potential PHES sites, with storage potential in the range 1-300 GWh (see http://re100.eng.anu.edu.au/). Nearly all of them are off-river. Since only about 450GWh of storage energy (and 20GW of storage power) is required to support a 100% renewable electricity system, only the very best of these sites need to be developed. There are many excellent potential sites close to windfarms, solar farms, major transmission lines and cities. The most robust storage solution will comprise a variety of PHES systems with power in the range 0.1 to 2GW, and be very widely distributed to support local regions in the event of transmission line failure.

Kidston, Genex Power: We saw the Kidston site as ideal, not only for the locational advantages but also for the infrastructure left behind, the pre-existing environmental permits and the absence of community impact. In terms of location, access to a transmission line is of great importance, however given the scale and dynamics of these projects, usually the most ideal sites are located far away from urban communities and therefore may require transmission lines to be extended, as is being done for the Kidston Pumped Storage Project.

Kennedy, GE: One of the features of pumped hydro is that you need to put the plant where the resource is and there is adequate network infrastructure. A recent study by the ANU found 22,000 possible sites across Australia, clustered primarily along the Great Dividing Range. The notion of power generation being separate from load sources is not new, however, as thermal generation sites have often been built well away from Australia’s major cities. While putting pumped hydro plants in the suburbs of the cities may prove challenging, the idea of having a variety of different sized systems dispersed throughout the NEM makes sense from a network stability perspective.

Fulton, GHD: I would favour large remote facilities, up to the constraints of the existing or proposed transmission losses. The proposed Snowy 2.0 is the largest of the pump storage projects currently being investigated. It will be able to store 350,000MWh – this is as much as 35 million domestic batteries (10 kWh). However, smaller pumped storage schemes associated with solar farms close to cities can reduce the total reliance on big transmission lines – although we presently rely on transmission lines every day in virtually every part of the country.

Gwynne, Hydro Tasmania: Siting is likely to favour facilities that are co-located with new renewable generation like high potential wind sites or solar sites and proximity to existing transmission systems.

Whitby, Snowy Hydro: There is room in the NEM for a whole range of projects; the ones that stack up will be built. For Snowy Hydro we favour the Snowy 2.0 project. Snowy 2.0 is strategically located between the capital cities of Sydney and Melbourne and the size and scale of the project make it globally significant.

Suitable sites for pumped hydro will often be far from transmission, says Simon Kidston of Genex Power, and developers must weigh up the costs.

Do you have any comments about the environmental impacts of pumped hydro and the hurdle of securing approvals?

Blakers, ANU: The environmental impact of off-river PHES required to support a 100% renewable electricity system is small. The total area of reservoir required is about 3,600 hectares, which is a small fraction of the existing artificial reservoirs. If a particular site is problematical, then an alternative site can readily be found from the list of 22,000 possible sites. Water requirement for a 100% renewable electricity system will be much smaller than for the current fossil fuel system because wind and PV do not require cooling towers.

Kidston, Genex Power: Building a pumped hydro storage facility using an abandoned mine is ideal for permitting and the approval process. The closed loop system minimises environmental impacts that are otherwise present for hydroelectric river-run facilities. In essence, we are turning an abandoned mine that is of little use into a multi-million-dollar asset.

Kennedy, GE: Compared with traditional hydro, the environmental impacts of pumped hydro are limited as the footprint is much smaller and it is not necessary to dam rivers or flood valleys. The ability to use existing dams (such as Snowy) or to repurpose a disused mine site can also help simplify approvals as these approaches can leverage infrastructure and approvals already in place.

Fulton, GHD: Environmental impacts of pumped hydro can be positive with a closed system – just pumping and generating between two off-river reservoirs at different heights.  As with Kidston and other proposed disused mine systems, the pumped hydro will make positive, renewable use of large, unsightly holes in the ground (and will incorporate environmental clean-up). Other STORES (Short-Term off River Energy Storage) pumped storage systems incorporating “turkey nest” dams on elevated unused or agricultural land will have minimal environmental impact, most likely using tunnelling and underground power station, with very little surface works except for the two freshwater lakes.

Gwynne, Hydro Tasmania: The development processes need to be very carefully planned and managed, with sound regional stakeholder engagement and involvement critical to success.

Pumped hydro can avoid many of the traditional environmental difficulties associated with normal hydro generation because damming of rivers is not required for pumped hydro. It utilises off-stream reservoirs for storage.

Whitby, Snowy Hydro: A critical part of any new development should be to consider its possible environmental impacts, including the requirement for new water storages. In the case of Snowy 2.0 the project’s footprint is largely underground and we are using existing reservoirs, so there will be no change to downstream water use or environmental flows. The long-term benefits to the environment come from lowering emissions as we transition from coal-fired generation to renewables. Decarbonisation of the NEM cannot be done in a stable or economic way without the firming capabilities of pumped hydro.

Research by Australian National University professor of engineering Andrew Blakers has identified thousands of “excellent sites” for pumped hydro facilities.

Can you describe some of the latest technology and improvements in pumped hydro. How much extra yield might be gained from upgrading an old facility?

Kennedy, GE: While pumped hydro has been around many years, it is certainly inaccurate to characterise the technology as static. When the majority of Australia’s existing pumped hydro capacity was installed, it was not uncommon to have separate turbines and pumps. Nowadays it is more common to utilise reversible turbines that can spin both ways, reducing the capital cost of pumped hydro plants. Other innovations have included the development of multiple-stage turbine systems, which can significantly improve system efficiency for high-head projects and the development of variable speed technology, which substantially increases the operating range, especially in pumping mode.

Fulton, GHD: Hydro-electric generation is an old technology but has enjoyed the advantages of modern technology for greater output and efficiency. Most pumped hydro installations have a single reversible pump turbine which operates as a turbine powering the generator in one direction, and when power flow is reversed the generator becomes a motor which drives the turbine in the reverse direction, pumping the water back up to the upper reservoir. Optimal speeds and conditions for pumping and turbining are different, so usually a fixed-speed reversible pump turbine is a compromise between the two actions pumping and turbining, with both actions operating away from their best efficiency point.

Modern, powerful solid-state technology has given us the ability to provide variable-speed motor generators, so the pumping speed can be different from the generating speed, allowing both actions to operate at their best efficiency point. This can provide a significant efficiency and increase in both directions (2-3%, which doesn’t sound much but is worth millions of dollars).

Also, modern computational analysis has allowed optimisation of the pump and turbine design. In the old days, the design was an art, using model turbines and trial and error to get the best design. And then, the turbine rotor (runner) would be cast steel, with small variations between differing blades which reduced efficiency. Computer design, computer manufacture and modern materials now provide the highest possible efficiency, and allow machines that can work at much greater flow and pressure than ever before.

Most old hydro stations have had their turbines upgraded with modern design runners. It is not uncommon to get 10-15% more power, with efficiency gain of 3-5%. There aren’t a lot of old pump storage stations around Australia, but they can benefit from the same upgrade results (if not already upgraded). But putting a Ferrari engine into an old Holden can only go so far.

Gwynne, Hydro Tasmania: A key technology improvement is variable-speed pumped hydro machines which can support system security and reliability when pumping as well as generating. Traditional pumped hydro machines can only provide that support while they are generating.

Noting that there is very little pumped hydro currently in Australia, upgrading pumped hydro facilities could yield a capacity gain of 3-5%, primarily due to a more efficient design.

Whitby, Snowy Hydro: Snowy Hydro’s largest power station Tumut 3 also has pumping capabilities. The station was recently upgraded by fitting new computer-designed turbine runners that lifted the capacity of the station from 1,500MW to 1,800MW. One of the latest developments in high-head pumped storage is the use of variable-speed reversible Francis turbines, and the Snowy 2.0 project will utilise this technology. The improvements in technology allow for more flexibility and responsiveness when operating. These capabilities allow operations to better match the increasingly volatile supply and demand patterns in the market that are occurring within the increasing levels of intermittent generation.

Describe in a few words your most-favoured type of pumped hydro facility.

Blakers, ANU: Virtually all good PHES sites in Australia are off-river. There are thousands of excellent sites characterized by large head (400-700m), large storage (5-50GWh), steep slopes between upper and lower reservoirs and close proximity to cities, PV and wind farms and transmission lines.

Kidston, Genex Power: Genex is of course biased towards the Kidston Pumped Hydro Project … however in terms of operating projects, the Dinorwig Power Station in North Wales has a nameplate capacity of 1,728MW and was constructed into a cliff face on the coast. It’s been operating since 1984 and is the main form of grid stabilisation in the European energy market given its enormous scale.

Kennedy, GE: It is tough to say, as no two customers or project sites are the same and each pumped hydro system is optimized to its particular circumstances. Globally, most projects store at least 200MW of energy for 6-8 hours, at which scale pumped hydro is highly cost-effective. In Australia, most projects under consideration seem to be in the 200-250MW range (such as Kidston and Cultana), however as Snowy 2.0 shows, much larger facilities are also possible.

Fulton, GHD: I’ll not describe my most favoured type but my most favoured pumped hydro facility – the Kangaroo Valley pumped hydro station, part of the NSW Shoalhaven scheme. Situated in one of the most truly beautiful little valleys in Australia, this scheme is being well maintained and operated by Origin Energy, who are working the electricity market by pumping water during times of low-priced tariff and generating to support the system (and make money) when the price of electricity is high. The scheme also has strategic value in being able to use the pumps to supplement Sydney water supply in times of critical water shortage. And finally, it is awesome to stand in the top of the station and looked down 100 feet of shafts and generator and turbine and pump and large valves. Maybe not the most modern, maybe not the most efficient, but my most favoured. Moreover, it certainly has potential for upgrade and expansion.

Whitby, Snowy Hydro: Snowy 2.0 of course, closely followed by Tumut 3.