Pumped hydro slowly starts to put its head up

Print Friendly, PDF & Email

NSW is getting excited about pumped hydro, but many projects will only get up if supported by analysts such as Rosie Scenario and Phil Wright.

share
Kidson Pumped Hydro Project
Print Friendly, PDF & Email

Pass the valium, the NSW Government takes the initiative

About two weeks ago, the NSW government announced that it was holding a briefing day looking for EOI [Expressions of interest] to take WaterNSW existing assets and see how some pumped hydro could be added on.

As Amy Kean’s email stated:

“The EOI process is investigating the opportunity for WaterNSW’s assets, such as 42 dams across the state, to offer private developers the key ingredient required to develop new energy investments, such as pumped hydro, hydro generation, floating solar and other technologies.”

I wasn’t able to make the briefing. The great advantage of the existing infrastructure is that there are likely to be far less issues around environmental approvals and water licenses. In addition brownfields developments can be far cheaper than greenfields developments.

Origin’s Shoalhaven expansion half the cost of Snowy 2.0 in terms of power, more expensive for energy

The same was true of Origin Energy’s announcement of its co sponsored ARENA study for a doubling of power at its existing Shoalhaven pumped hydro scheme. This plan would see Origin spend $240 million for a 235MW increase in output.

This is, of course, only half of the cost of building Snowy 2.0 – which on current estimates is around $4 billion for 2000MW increase in power. That is, $2 million a MW.

Origin’s water licence won’t change so it will only be able to produce for 14 hours whereas Snowy could produce for a couple of weeks. So Origin’s plant is far cheaper on a power basis, but more expensive in terms of energy.

Likely that the EOI process will produce new options that have lower power cost than Snowy 2.0

It will be surprising if there isn’t one technically viable project emerging from a study of 42 dams.

And we would expect that for daily arbitrage, that is 4-8 hours of daily output a brownfields plant will be significantly cheaper on a power basis than Snowy 2.0. Still, we will have to wait and see.

Pumped hydro economics work @ $1m a MW, but not at $2m

There are three potential revenue sources for pumped hydro.

Price arbitrage, ie selling electricity at a higher price than cost of electricity used to pump the water back to the top of the hill. This cost typically has to allow for round trip efficiency of about 75%. That is, pumped hydro is a net consumer of energy.

Cap or insurance products: The standard product in the existing market is a $300 cap. This guarantees the buyer of the cap won’t pay more than $300/MWh over a given period for a given number of MW.

Ancilliary services. It could be argued that this is code for “the project doesn’t make enough money.” Ancilliary services as we increasingly hear are black start, inertia, frequency control.

There is a capital cost for pumped hydro to provide these services and we think that batteries are inherently better suited to many of them. In any case, there just isn’t that much revenue.

Shoalhaven 2 will need to do better than Shoalhaven 1

In the 12 months ended May 8, 2018 the existing Shoalhaven power station produced 85 GWh and earned $9 m of pool revenue.

It should be obvious that even if there was zero cost of pumping, zero fixed costs that $9 million of revenue isn’t going to justify $240 m of capex.

This brings us to the point that we previously made looking at Wivenhoe pumped storage in Queensland (Read: Wivenhoe, the litle pumped hdyro plant that didn’t.

The owner of that pumped hydro station is also the owner of a coal fired power station, and so there is very little incentive to operate the pumped hydro. It’s cheaper to burn coal and the price is higher than if you have to pay to pump the electricity.

In addition, the pool revenue won’t reflect the internal hedge that Shoalhaven provides the wider Origin Gentailer Group.

Limited arbitrage signal in NSW

We used the excellent NEM Review data service to get the 17,500 sequential half hourly prices for the 12 months ended May 8, 2018.

We then used Excel to sort these into a 365 row, 48 column table and then used the excel large and small functions to sort the prices each day into the low prices and the high prices.

From this we were able to establish the average 2, 4, and 6 hour price arbitrage for each day and get the yearly average.

That is, for the 4 hour arbitrage we used the average of the highest 8 half  hours on the day and subtracted the average of the lowest 10 half hours. Using 10 half hours corrects price for round trip efficiency but we also need to adjust volume.

Surprisingly a 5% gross return was available assuming 6 hours generation and a brown fields capital cost of $1 m/MW

Figure 1 Perfect foresight daily arbitrage gross profit NSW 2017/18. Source: NEM Review, ITK

Of course this assumes perfect foresight of daily prices and that your plant has zero impact on either or both the buying and selling price. So it’s very much produced by Ross Gittins’s Italian analyst friend “Rosie Scenario”.

All we need is for her English assistant “Phil Wright” to complete the spreadsheet and it’s off to the Board for approval.

What about $300 caps? Well in NSW for the year to June 30, 2020 these are selling for a time weighted average of $8.40 a MWh. No post Liddell closure cap prices are yet quoted by the ASX.

As it happens there were only 2 half hours where NSW price exceeded $300 MWh in the past 12 months. Very different to the year before.

Each MW of caps @ $8 MWh adds around $70,000 of revenue per year, but may require some arbitrage revenue to be foregone and if prices stay above $300 for more than 14 consecutive hours, would expose Shoalhaven to risk.

In the present market that seems unlikely and in the context of a bigger portfolio is easily covered.

Even so the combination of caps and arbitrage would likely justify Shoalhaven 2, but probably still wouldn’t justify Snowy 2.

In short, more price volatility is needed for these projects to really get a sign off. Its probably appropriate to be developing some of these projects but only as a secondary priority.

The highest priority in NSW is surely getting common agreement on how to deal with the forthcoming closures of the coal generation and specifically making sure there is enough transmission infrastructure to encourage wind & PV plants to get out of the computer and onto the ground.

David Leitch is principal of ITK. He was formerly a Utility Analyst for leading investment banks over the past 30 years. The views expressed are his own. Please note our new section, Energy Markets, which will include analysis from Leitch on the energy markets and broader energy issues. And also note our live generation widget, and the APVI solar contribution.

David Leitch

David Leitch is a regular contributor to RenewEconomy.com.au. He is principal at ITK, specialising in analysis of electricity, gas and decarbonisation drawn from 33 years experience in stockbroking research & analysis for UBS, JPMorgan and predecessor firms.

Print Friendly, PDF & Email

11 Comments
  1. Mike Westerman 7 months ago

    Interesting analysis David. It seems tho’ that for an investor contemplating a solar farm, and knowing the daytime price curve is declining over time where penetration of solar is high, while evening peak pricing is rising, would find bundling a deal with PHES attractive.

  2. Malcolm M 7 months ago

    The more immediate need for such pumped storage capacity is in the Vic market. Over the last few weeks with Basslink out of action we have seen a foretaste of what the Vic market will be like when the wind farms under development start supplying the market. Night-time spot prices have often dropped to $7/MWh compared with peaks of over $100/MWh. Many of our existing hydro stations that have lower and upper storages should be capable of adding pumps. The hydro operators would then be able to offer floor price contracts to renewable developers, and would then have more water in their upper pools to use at times of high power prices. Hydro stations that would be suitable include
    – Murray (the lower pool of Khancoban has about the same capacity as the upper Geehi storage)
    – Dartmouth
    – Eildon

  3. Jon 7 months ago

    Thanks David
    The price volatility will come as mor FF plants close down and more RE generation comes online.
    Storage systems (battery and Pumped hydro) have two cost comparison points generation MW and storage MWh, would be interesting to see bith

  4. ashentegra 7 months ago

    Excellent analysis and thank you David. I suspect the arbitrage opportunity has narrowed significantly with the introduction of the Hornsdale battery and more battery installations announced – today by Infigen for its Lake Bonney SA windfarm.

    If this kit profitably removes the most extreme prices and encourages imitation through more installations, as is already happening, the benefits to pumped hydro shrink dramatically. I would hate to be committed to a half-decade engineering construction project and see the price arbitrage opportunities shrink and shrink and shrink each year. I think batteries will eat their lunch.

    Adding a battery to any windfarm is a no-brainer. The frequency oscillation discount is removed, production above contracted can be saved and even on still days when wind cannot offer into the market, a zephyr that intermittently turns blades a little and pumps out a few volts has a use, a destination.

    The widest price differential is between late night production and afternoon consumption. Wind, coal and run-of-river hydro all stand to benefit from batteries, much more than, say, solar or gas. However, coal power businesses are universally being run for cash and cannot justify battery capex when the coal plant will clearly be closed by cost in the near future. They can of course free ride on the wind/battery combos of other players in the meantime, as their withdrawal from bidding into the midnight market increases coal’s returns.

    Disclosure: IFN.ASX

  5. BushAxe 7 months ago

    You have to look at the long term future of thermal generation as it looks like there’s going to be requirement for alot more peaking generation. So I think we’ll see a balance between storage and peakers, $240m is cheap for 235MW of generation and alot cheaper to run than OC generation.

  6. Nick Abbott 7 months ago

    This analysis is wrong because it does not appreciate the increase in volatility that renewables are bringing.
    In this week just gone in Victoria there were 16 hrs where the price per MWh averaged just $15.83. There were also 9 hrs where the price averaged $211.66. Big difference.

    The amount of generation from wind and solar will double in the next couple of years given the projects currently under construction. The volatility we see in Victoria and SA will happen in NSW. And storage will be needed to flatten that out.

  7. Chris Drongers 7 months ago

    Pumped storage is almost certain to increase mixing of warm/cold, fresh/salty layers in the existing water storages. The warmer, seasonally less salty discharges will benefit ecosystems further down the river.

    • Mike Westerman 7 months ago

      Unlikely in the Shoalhaven scheme because water used for the PHES will be taken from the surface water well upstream of the deepest sections of Lake Yarrunga. Mixing can be a significant problem in stratified lakes, or lakes with difference source waters (as is the case in Snowy). Hence the attraction of avoiding these issues by developing schemes on “dry” topography, rather than existing reservoirs. Shoalhaven is a special case in that Phase 1 was constructed so as to facilitate Phase 2.

      • Chris Drongers 7 months ago

        What are the problems with mixing in stratified lakes if those lakes are artificial impoundments anyway? I was thinking of the problems with Murray River Cod and Yellowbelly breeding caused by discharges of cold bottom water from the Hume Weir and other storages. In inland areas, early and late dam inflows can be more saline than mid-season flows; mixing would allow a greater fraction of the saline flows to be used by ‘blending’ with the fresher than required mid-season flows.

        • Mike Westerman 7 months ago

          One problem is that the deoxygenated bottom water becomes acidic and releases iron from the soil, which when in contact with oxygen precipitates out, reducing sulphates to sulphides. Both the H2S and iron cause problems from corrosion to mortality of a range of species (including humans on occasion). If the artificial lakes are designed as shallow ponds of not much more than the required cycling volume, these problems tend not to occur.

          • Chris Drongers 7 months ago

            Thanks

Comments are closed.