Brave New Climate

Pumped-hydro energy storage – cost estimates for a feasible system


Guest Post by Peter Lang. Peter is a retired geologist and engineer with 40 years experience on a wide range of energy projects throughout the world, including managing energy R&D and providing policy advice for government and opposition. His experience includes: coal, oil, gas, hydro, geothermal, nuclear power plants, nuclear waste disposal, and a wide range of energy end use management projects.

[Ed: Peter is seeking feedback. Some reviewers comments he has already received are given at the end of this document, for reference.]

Pumped-Hydro Energy Storage – Tantangara-Blowering Cost Estimate


Energy storage is essential for intermittent renewable energy generation and is valuable with coal and nuclear generation too. Pumped-hydro is a mature technology and is generally the least cost option for large scale energy storage.

This paper provides a rough cost estimate for a pumped-hydro energy storage facility that would utilise existing dams and reservoirs in the Australian Snowy Mountains Hydro Electric Scheme.

The paper is in response to questions about the cost of pumped hydro energy storage, especially as a component of a fossil-fuel-free electricity generation scenario for Australia. The paper is the fourth in a series of articles (see here for a listing — search this page for articles authored by Peter Lang).

Figure 1: Conceptual diagram of a pumped-hydro energy storage facility.

Project Description

The proposal is to build a 9 GW pumped-hydro facility by connecting two existing reservoirs with tunnels, pipes, generators and pumps.

The two reservoirs are Tantangara (the upper reservoir) and Blowering (the lower reservoir). The difference in elevation between the water levels in the upper and lower reservoirs is 875m (see Appendix). The tunnels to join the two reservoirs would be 53km long (Figure 2).

Figure 2: Proposed Tunnel Line for Tantangara-Blowering Pumped Hydro Scheme

The facility would generate 9GW peak power, for 3 hours per day from 6 hours of pumping at full pumping rate. It could generate for longer at less than 9GW, or if pumping was for longer than 6 hours per day, or if the pumping rate is greater than assumed in this analysis.

The proposal is to bore three tunnels, each of 12.7m diameter, to link the two reservoirs. The reservoirs would be tapped near their down-stream ends to maximise the volume of water storage available. About 3km of the tunnel at the lower end would be steel lined to prevent leakage. There is just about sufficient ground cover along most of the remainder of the tunnel route to avoid the need for waterproof lining or grouting for most of the remaining 50 km of tunnel.

The rocks to be tunnelled through would be mostly granite and similar strong, hard igneous rocks. There may be some hard sedimentary rocks, possibly including minor amounts of limestone. The rock would be fractured and faulted.

The power station is assumed to be similar to Tumut 3 in that it would have six turbines in a power station of similar size to Tumut 3. The differences are that the power output will be six times that of Tumut 3 (because six time the hydraulic head). There would be six pumps rather than the three at Tumut 3. Another difference is that some 3km of steel lining will be required in the tunnels.

Issue – Francis turbines, which are best for pump storage, currently have a hydraulic head limit of about 600m. Tantangara-Blowering is 875m hydraulic head. The manufacturers are working on increasing the upper limit. I understand manyufacturers are working at the moment to double them up.

Cost Estimate

The table below provides a ball park cost estimate for the scheme. The tunnel costs, the main component, were calculated using costs for rock tunnel excavation from world wide experience. The power station, pumps, etc, were estimated by multiplying the original costs (from 1967) for Tumut 3 by 10 (that is approximately how much inflation between 1967 and 2009), and by doubling the cost of this equipment to allow for the six times greater water pressure and, therefore, power output. The greatest uncertainty is the cost of the steel pipes, tunnel lining, power station and ‘Allowance for Other’.

How realistic is this cost estimate?

World experience is that hydro projects cost about US$2,000/kW to US$4,000/kW. The Electricity Storage Association gives a range of costs for Pumped-Hydro of US$500/kW to US$1500/kW. This project is A$744/kW.

So the proposed Tantangara-Blowering facility is towards the low end of the range (if the figures are correct). The main reason for the lower than average cost is that the reservoirs do not have to be built; they already exist. However, this project requires more tunnelling than is normal for pumped-hydro facilities.


The Tantangara-Blowering pumped hydro scheme would be a high capital cost investment for just 3 hours of peak power generation per day.

This is the most economic of the four projects investigated.

Pumped hydro is often economically viable for providing peak power for a system comprising mostly fossil fuel and/or nuclear generation (France’s system is the ideal example). But pumped hydro is not well suited to intermittent, unscheduled generators.


I would like to acknowledge the help of some friends, all of whom have had a life time of experience working on hydro-electric projects around the world (all but two still are). Our work on hydro sites and our friendship go back 39, 33, 33, 29, and 19 years. I would like to thank Rich Humphries and Dr Derek Martin for their invaluable assistance (and for taking time off from catching neutrinos to help me: they are building a large underground cavern deep below Sudbury, Ontario to catch neutrinos that will be beamed through the rock from Chicago); Tim Little (Canada), Andreas Neumaier (currently in Asia) both of whom have a life-time of experience designing and constructing hydro-electric projects; and David Purcell who was the first boss I had who could influence me. He grew up in Cabramurra, the construction camp for the Tumut 1 and Tumut 2 projects in the Snowy Mountains Scheme. On weekends he would go with his father, a civil engineer, around the scaffolding and walk-ways as he checked out the construction work and instrumentation on the Tumut Pond concrete arch dam, Tumut 2 Dam and the Tumut 1 and Tumut 2 underground power stations. What a way to start a life. Certainly, today’s young can’t get that sort of early experience of the real world.

Pumped hydro is an excellent match for nuclear and for dirty brown coal fired power stations (because they run at full power all the time and produce very cheap power). Consider the 2007 NEM demand: Peak = 33GW, average = 25GW, base-load = 18GW. Option 1 - nuclear only. 33 GW @ $4b/GW = $132b. Option 2 - nuclear + pumped hydro. 25 GW nuclear @ $4b/GW = $100b + 8GW pumped hydro for $15b = $115b. Conclusion: Option 2 (nuclear + pumped hydro) is the lower cost option.

Reviewers’ Comments

Reviewer 1

I had a quick look at your idea of a pumped storage scheme between the Tantangara and Blowering resevoirs in the Snowys. I have only spent a couple of hours looking at your numbers for the assessment of the technical and economic viability of such a scheme. It is my personal view as an individual and not that of the company I am working for.

Ok, here comes the technical side:

1. One would have to assume that the available head is between the minimum operating level at Tangagara, MOL = 1,207 and the full supply level at Blowering, FSL = 380 because any operator would have to guarantee 95% reliability for his peaking power. Thus, the gross head for power generation is MOL – FSL = 827 m

2. Assuming 3 No.s tunnels of diameter D = 12.7 m gives a cross sectional area of Atot = 380 m2.

3. Assuming a maximum flow velocity of v = 3 m/s to keep friction losses within a reasonable range, the total discharge Q = v * Atot = 1,140 m3/s.

4. With that and assuming a local and friction loss factor of η = 0.85, the energy P computes to be P = 7,860 MW using the formula

P = (9.81*η*Q*H)/1000

5. Assuming an operating time of T = 3 hours/day the available energy is E = P * T//1000 = 8.6 GWh/year

6. The discharge volume V = Q * T * 3,600 = 12.3 106 m3.

7. With a reservoir area of Ares = 2,118 ha and assuming vertical reservoir walls, the drawdown over the 3 hour operating period will be Δh = V/Ares = 0.6 m which is acceptable from an environmental and safety point of view, I think.

8. Assuming that the operator can sell peaking power at 4 times the rate as he has to buy the power for pumping during off-peak times, say $0.60 generating vs $0.15 pumping costs, the annual revenue from the scheme would be R = ($0.60 – $0.15) * 8.6 106 = $3.87 billion per year.

9.I am not a mechanical engineer but I believe that at a head of H = 827 m you are pushing the envelope of what turbines can take. I also think that you may need more than 3 units, say 6 or 9 to keep the size of the turbines and pumps and the width of the underground cavern to a reasonable size.

Now comes the project cost sides:

1. The tunnel is by far the most expensive single item so I have focused on that. I do not have a price for a 53 km long, 12, 7 m diameter tunnel excavated by TBM so have used a 8 km long, 3.5 m diameter tunnel and scaled it up to give an approximate figure for the bigger tunnel. Assuming that the tunnel will be mostly unlined and constructed in essentially sound rock, the costs for excavation, support and care of water during construction is $ 7.56 109, which more than double your figure.

2. You have scaled up the 1967 costs of T3 to arrive at the other major cost items. If my tunnel cost figure is correct, I fear that the other cost items may also be higher than what you have estimated.

3. If this is the case, the construction costs may be closer to $15 billion than $7 billion as you have estimated, which will bring the cost per installed kW back into the range of $2,000/kW which is about what pumped storage schemes cost these days.

I do not mean to discourage you but the capital expenditure for a pumped storage scheme between Tantangara and Blowering seems prohibitive because of the scale of the investment, the high up-front costs and the long period for investors to recover their money. Unfortunately, politicians and banks take a much shorter view of life when it comes to political or financial gains and it seems to me that your idea, as much as I like hydro, seems to be condemned to the ‘not economical’ basket.

Reviewer 2

I took a quick look at your costs and determined that it would take more time than I have to review those adequately.  We have had a lot of challenges with cost escalation over the last several years (until the last year or so) and I would not want to give you a misleading opinion.  That said, I do have some quick comments on your proposal:

— The proposed capacity of 9GW (9000 MW) is huge.  Compare to Itaipu (12,600 MW), previously the world’s largest hydro plant and now Three Gorges (ultimately about 22,000 MW) or Revelstoke (2000 MW, soon to be 2500 MW when Unit 5 is completed).

The proposed number of turbines is 6, which translates to 1500 MW each.  I believe the world’s largest turbine capacity to date is about 750MW. (e.g. the turbines at the Sayano-Shushenskaya dam in Siberia which recently experienced the world’s worst hydro plant disaster to date – check it out on Google if you are not familiar already).

The head is very high as you note.  I believe Francis turbines have been built for heads as high as about 800m, but probably not for large capacity.

Your overall cost looks quite low at less than $800/kW; as you note, world experience is typically several times that today.

Your stated uncertainty of 25% is too optimistic.  When we do a conceptual level estimate by a qualified estimator, we state that the uncertainty on the upper end is as high as 100%. The uncertainty goes down as more design is done, and we would not claim 25% uncertainty until we were into final design.

Steel lining is one of the more critical cost factors.  For many high head projects, the cost of steel lining is the factor that determines whether the project is economic or not. You need to know in situ stresses etc to do a proper analysis.