Why solar towers and storage plants will reshape energy markets

Why solar towers and storage plants will reshape energy markets

Print Friendly, PDF & Email

More details emerge of the contract for the Port Augusta solar tower and storage project, and why it means huge change for energy markets, and a shift of focus from “base-load” fossil fuel to clean, flexible capacity built around “base-cost” renewables.

Print Friendly, PDF & Email


Crescent dunes tonopah
The 150MW solar tower and molten salt storage plant to be built in Port Augusta has been made possible by a ground-breaking pricing and contract structure that could help completely reshape Australian power markets, including the end of “baseload” power as we know it.

The South Australian government announced last week that it had signed a deal with US company SolarReserve to build the 150MW solar tower with molten salt storage project – to be known as Aurora – just a few kilometres from the now closed Northern coal-fired power station.

The output of Aurora will be around 500GWh – roughly the same as the annual consumption of the state government and the various assets it owns.

But it is the unique nature of the contract that explains the difference between what the government will pay SolarReserve ($75-$78/MWh), and what SolarReserve will receive, and will likely serve as a template for more “dispatchable” renewable energy projects in the future.

Essentially, Solar Reserve will provide the S.A. government with some of its needs from other sources in the market when demand and the price is low. Aurora will cover the government for energy and prices when the government’s demand is at its highest, around the middle of the day.

But because the government can and will get some cheap power elsewhere, Aurora will be able store its solar power in its molten salt storage tanks so it can sell into the market at the system peaks, in late afternoon and early evening, when the market prices are highest, boosting its revenue.

For those in the market, it is a little like a power purchase agreement where an off-taker takes part of the output, while the rest of the plant goes “merchant”. But this contract has a lot more flexibility and is more flexible than plants that use traditional cap contracts.

Danny Price, from Frontier Economics, who helped the government design the contract, says it was struck with a mixture of a capacity purchase agreement (CPA) and a hedge, with the CPA designed to break down the fixed costs and the variable costs.

“We had to have a contract for wide range of technologies,” Price tells RenewEconomy, and the contract structure was designed to level the playing field between dispatchable renewables, which may struggle following loads MW for MW.

Although the fine details of the contract are not being released, at least not yet, the significance should not be understated.

South Australian energy minister Tom Koutsantonis boasted last week that it would send “shivers” through the spine of the coal industry – not that there is one left in his state, but there is no reason why this cannot apply in other states.

There has been great debate about how traditional energy markets, which rewarded baseload power with expensive peaking capacity, could evolve to include renewables with zero fuel costs such as wind and solar.

Price says rather than redesigning the whole system, this contract structure plays within the current design – and the ultimate goal is to provide more stability to markets. It also avoids the pitfalls of conventional capacity markets, which have been shown in Western Australia to be disastrous if badly designed.

Aurora will have no incentive to do as wind and solar PV farms do, and bid in negative prices to guarantee dispatch, and the fact that it is competing in peak times means it reduces the ability of the thermal players to game the market up to pricing peaks of $14,000/MWh.

What’s more, it’s a cheaper technology than gas or new coal. “You couldn’t even buy the gas for that price – let alone the capacity,” says Price of the $75/MWh contract price. And new coal, he says, cannot be built for less than $100/MWh.

“Who would build a gas or coal-fired generator if you can get a dispatchable renewable energy technology like solar thermal for $70/MWh? Why would you do it?

“It is a terrific outcome for the market and for renewable energy. It shows how you can construct a contract to encourage dispatchable renewables.

“It means we can get a renewable energy technology up that is cheaper than thermal plant. So while the federal government stuffs around on carbon pricing, technology is just taking them over.”

Price said one of the keys to the project win was the learnings of SolarReserve from its Crescent Dune facility in Nevada, which has been operating for just over a year -– apart from a few months off-line to deal with a leak or leaks in the molten salt storage tank.

Price said the key to the technology was the receiver, and it was “a very good technology.” SolarReserve had learned a lot from its first plant, capital costs had fallen 40 per cent, and costs were continuing to come down.

“The technology is not really a risk. We get the benefit of the learning so far, and we are second cab off the rank.”

Please read this piece by Simon Holmes á Court, which looks at some of the other interesting aspects of this development, particularly his observation that solar thermal technology is already nearing the price predicted by Alan Finkel for 2050, just a few months after the Finkel report was released. 

There are a few things that could still go wrong.

The current contract price is dependent on the $110 million equity  partnership promised by the federal government both before the election and when it was negotiating a tax package with Senator Nick Xenophon, whose dual-citizenship means his immediate political future is now subject to a High Court ruling.

It is clear, and SolarReserve CEO Kevin Smith said as much in a radio interview on Monday, that the pricing would have to be revisited if that component did not come through.

He did, however, say that financing has been lined up. It is not thought that renewable energy certificates will contribute largely to the deal.

While they are trading round $80/MWh now, they are expected to have fallen dramatically by the time this plant is up and running in 2020, and could have negligible value before they expire in 2030.

Solar towers with molten salt storage, because of their high storage levels, offer the flexibility to offer either continuous power 24/7 (baseload), higher outputs over shorter periods to meet specific demand (such as powering Las Vegas between noon and midnight), or acting as a “peaking” power station.

This new South Australian plant appears to be designed to include the last two components. Its nominal capacity of 150MW will likely be reduced to 135MW during the day, where it uses some of its own power to operate and control the sun-following field or mirrors.

But by being dispatchable in such a way, the solar tower will beat both coal and gas at their own game.

Coal’s big weakness, apart from the fact that it is now dirty and expensive, is that it can’t readily be turned off, which is why loads such as hot water systems are switched on only at night – to give the coal-fired power plants something to do during the night.

The arrival of “base-cost” renewables, however, means that much of the power supply can be delivered by wind and solar at cheap prices.

And the emergence of storage in its various forms – batteries, pumped hydro and now solar towers – means that output can now be dispatchable, or can be balanced, and can mean manufacturers may no longer have to operate only at night when the power cost is low. Cheap power may be a day-time event.

Aurora will not need to operate all the way through the night, although it could. If the government needs power, it will likely buy from other sources at a cheaper price, Aurora will deliver to the government during the day, meet and cover its peak demand, and sell into the afternoon and evening system peaks.

Of course, Aurora is not the only solar thermal project happening in the world right now. SolarReserve is building another plant in South Africa, and other companies are building plant in Morocco, Chile and China.


Keith Lovegrove, director of the Australian Solar Thermal Energy Association, recently returned from a conference in China, where he says the main news is that of 20 announced projects of between 50 and 100MW,  three are well under construction and a further five will begin soon.

“The attached pic is a visit to the Shouhang 10MW molten salt tower system there in Dunhuang,” he says.

“It looked good (from the outside at least). Right next door that company is well into building the 100MW salt tower system that is one of the three mentioned above. CSP is alive and well in China and maybe has finally a good chance of life in Australia.”

Please listen to our podcast with Keith Lovegrove and David Leitch, and this article’s author to discuss the solar tower project, its economics, the implications for the market, and what may follow.

(Note: This article has been updated to make it clear that SolarReserve will be buying power on market, and at cheaper prices when demand is low, not the S.A. government. The purchase will be made by SolarReserve as part of its contract to meet the government’s power needs).

Print Friendly, PDF & Email

  1. George Michaelson 3 years ago

    I don’t entirely understand ‘…will not operate into the night…’ I was led to believe the thermal mass was capable of being kept at or near temperature for extended periods if you wanted: the decision to run it and cool down will be economic, not a physical limitation as I understand it. If the peak is early evening, it makes sense to run it then and secure income. But if it has a 25% loading obligation to be available as energy reserve, surely they can do that any time in the next 24-48h period if required?

    • Giles 3 years ago

      it will operate, just won’t be sending power to the grid all night. no need.

      • George Michaelson 3 years ago

        looks like you edited for clarity (or I misread) it makes sense now.

      • solarguy 3 years ago

        Well I’m sure that at times, sending power to the grid would be required, well into the night when wind is low and other storage options are sub optimal. On the 15MWh needed to run the mirrors, mate surely that’s a typo or a miss understanding.

  2. danielspencer 3 years ago

    Great article. Gamechanging.

  3. bedlam bay 3 years ago

    Giles, can you rebut the biased Weekend Australian article of 19/8/17.

    • Giles 3 years ago

      can you provide headline so i can search it. unless it in the cafe, i don’t get to see it. and i wasn’t in a cafe this saturday!

  4. Alastair Leith 3 years ago

    Giles could you go into a bit of detail on the two tenders this project won in your next article. I’ve lost track of the details (was it 25% and 75% of govt load breakdown?).

    • Alastair Leith 3 years ago

      Found this in one of the links:

      The SA Government went to open tender looking for two power supply contracts to meet its long term power needs:
      75% open to any dispatchable technology, to bring new generation into the market.
      25% constrained to dispatchable renewable energy technologies, to “drive innovation in storage and other technologies”.

    • Giles 3 years ago

      what you mean? it won both tenders, or the combined tender, so providing 100%.

      • Alastair Leith 3 years ago

        the details of what each tender was for. found it now. I thought they changed one of the tender conditions from “renewables only” to “low carbon” to allow for gas (even though they aren’t low carbon when fugitive methane emissions included) but presumably not coal.

  5. Alastair Leith 3 years ago

    “Who would build a gas or coal-fired generator if you can get a dispatchable renewable energy technology like solar thermal for $70/MWh? Why would you do it?

    Well, SA govt might be getting it for $75-75/MWh but it’s only because of the extra capacity bidding into a very artificially high peak market (c.f. the island SWIS grid that never goes above $300/MWh) and flexibility in the PPA with Govt (essentially arbitrage) that they can afford that contracted price.

    It would be a mistake to pencil $70/MWh down as the LCOE of CST with 8-10hr storage in Australia just yet. But I agree with the sentiment of who would build more coal or gas given this outcome. Carbon pricing is coming, Liberal Party or not.

    • Tom 3 years ago

      Agree with both parts of your comment.

      Regarding carbon pricing – it’s similar to the minimum wage. Businesses claim they don’t like it, but they really do.

      Without the minimum wage (or carbon price), there are no baseline figures from which to build your business model. You make an estimation, but then the bank doesn’t agree with your estimation, and you don’t get the finance. Or otherwise you and the bank agree on the estimation, and you are both wrong – yo can’t find the workers (or produce the pollution), and you are making losses. Bugger.

  6. john 3 years ago

    Frankly build another plant to ensure that the price spike does not happen or if another one is not enough build another one.
    So have at least 4 plants to ensure that the cost of energy is always going to be low and better than that when they sell into the market they can lower the cost of energy to the state.
    Does this make sense?

    • Gary Rowbottom 3 years ago

      A small network of these is hugely advantageous – dispatch can be staggered as needed to cover gaps much better – there are stated ambitions of 6 CST plants in SA (and similar in Qld). Fust things fust though – lets get No. 1 cracking.

  7. Mike Westerman 3 years ago

    This works I surmise because the SA government demand is mostly during the day. As solar because more significant, you would think it cheaper much of the time for the SA government to buy spot, rather than take from SolarReserve. It leaves SolarReserve with a fair bit of risk if there are persistent cloudy days. PV plus pumped hydro avoids that risk and is able to buy wind and other PV power to store as needed, at a combined cost comparable with CSP without the big Federal loan.

    • solarguy 3 years ago

      Mike, In this case it’s called CST- concentrated solar THERMAL!

      • Alastair Leith 3 years ago

        the terms are interchangeable far as I’m aware. CSP is more common in USA.

        • solarguy 3 years ago

          Although it can be used, it isn’t definitive. CSP can mean CPV too. Perhaps I’m to pedantic, but you know for sure what technology someone is talking about.

  8. Peter F 3 years ago

    Seems that they will get the value of the RECs. Even though they will be worth nothing for 60% of the plants life they will get useful value for the early years. That combined with the ability to sell into the peak market will more than double their gross margin and because it is front loaded makes a significant difference to the Net Present Value of the project

    • Mike Westerman 3 years ago

      Knowing SolarReserve I would be staggered if they haven’t gotten their claws into 75% of “all current OR FUTURE rights arising from a legislated scheme imbuing value to reductions in emissions of greenhouse gases”!

  9. Peter F 3 years ago

    Giles where do you get the capacity factors for various power plants. Is it available in monthly and annual forms

  10. Malthus Anderson 3 years ago

    Solar thermal is a big step forward but how do we prevent hundreds of birds, especially swallows, from being killed by the intense heat as it is directed from hundreds or thousands of mirrors to the central tower?

    • Ren Stimpy 3 years ago

      Quite easily, but how do we prevent precious butterflies like you from being gobschmacken? You poor thing!

    • mick 3 years ago

      if they are that stupid they would probably want to evolve a bit more

    • Thijs 3 years ago

      Are those the same birds that commit suicide in the coal plant chimneys?

    • Mike Westerman 3 years ago

      Birds are not as dumb as some – they avoid objects that are so intensely bright – the mirrors are only focused at a point on the tower while elsewhere they are reflecting a single sun.

    • solarguy 3 years ago

      Simple, they focus the mirrors from the ground up instead of sky down to the receiver, that reduces a lot of the problem. And keep in mind birds don’t like bright shiny things they will avoid it, just like using old head light in your vegie patch to scare of birds.

    • Chris Fraser 3 years ago

      A feral cat is much more dangerous than a CST tower.

    • BushAxe 3 years ago

      The receiver tower is nearby 200m tall so I can’t see many birds reaching that!

    • Joe 3 years ago

      No….not another scare campaign about the Birds…to try to smear RE. First it was wind turbines chopping our little feathered friends to pieces. Now it is ‘death rays’ zapping our feathered friends out of the sky. With all this RE our little feathered friends are now an endangered species. Malti, if you are seriously concerned about our Birdo friends there are plenty of other threats far worse than RE, like the domestic kitty cats, land clearing, tree felling, aeroplane strikes, motor vehicle strikes, ingestion of our plastic pollution junk, etc…I could go on but you get the picture that it is us Humans and our way of living that is the threat to little Birdo.

    • Craig Allen 3 years ago

      The problem became apparent during initial testing at the Ivanpah facility in Nevada. It happened in standby mode when all the mirrors were focused at a spot above the tower. They solved it by re-configuring the system so that the mirrors instead focus the light in a giant ‘pancake’ shape with no one spot getting more than 4x sun intensity.

      So now what shall we do about the billions of tons of toxic coal ash spread across thousands of square kilometers of coal ash ponds?

      And what about the toxic filth being released into waterways such as Wollongambi River in the Blue Mountains – killing fish, waterbirds, crayfish, invetegrates etc.? http://www.abc.net.au/news/2017-02-26/blue-mountains-pollution-mine-wollangambe-river/8303644

      • Caffined 3 years ago

        Sorry, a little late on this but it warrants a comment.
        How do they prevent birds , and insects etc from going near the huge circulat hot zone when the tower is working.?
        Knowing that the heat concentration must extend some distance away from the focus point
        Note, im sure you will have seen the many videos of birds being incinerated whilst the tower is operational.

        • Mike Westerman 3 years ago

          I don’t think you can “prevent” wild creatures from coming into fatal contact with human activities. Mitigation of risks by the operators and adaption by the birds seems to have helped but in any case, compared to the birds killed by habitat destruction, cars, phone towers and cats, solar towers aren’t the first target in avian preservation efforts.

          • Caffined 3 years ago

            I agree Mike,
            I was really questioning Craig’s comment that “they solved” the problem. That is clearly not the case.
            But, it might not disapear from the future discussion if it follows rhe wind turbine/bird senario where some areas of the USA have effectively prevented Wind Farm developments in prime locations due to bird protection .
            Is there a “Save the Galah ” movement in SA ?

          • Mike Westerman 3 years ago

            Fortunately the yellow bellied parrots stay down south and the lizards up at Whyalla don’t fly!

  11. Tom 3 years ago

    I’ve got a question if anybody knows the answer:

    How long does it take the actual steam generator in a CST plant to ramp up to full capacity once called upon? What is the “ramp-up”?

    Is it comparable to OCGT? Or is it more like a CCGT?

    • Mike Westerman 3 years ago

      Tom it depends on how you operate it – all STs have a ramp up if they are allowed to go cold because blades need to heat up without distorting, shafts need to extend etc. Most industrial STs will go from dead cold to full steam over an hour or two, whereas an ST that hs had steam bled in or is already warm will run up over a few minutes (ie similar to aeroderivative OCGT). So much faster than a CCGT, way slower than a hydro.

      • Alastair Leith 3 years ago

        presumably they will warm up turbines prior to the evening shoulder or any projection of high demand? the steam can be flashed almost instantly on call using thermal transfer from the molten salt conduit presumably?

      • Tom 3 years ago

        Interesting. Thanks.

      • Hamish Barker 3 years ago

        Steam turbine can be kept hot at small cost. The limit on ccgt ramp rate can be the steam drum of the boiler. It is a thick component, thermal changes cause stress, too much temp change ( which goes with pressure change ) leads to fatigue life consumption of the drum wall, which in the end limits number of start cycles.

        But a solar thermal plant might need fewer evaporator tubes due to better conductivity of salt to water/steam heat transfer rather than conventional plant’s hot furnace gas to water/steam. Whether this can translate to a smaller steam drum, where separation of steam from water occurs, i don’t know. If a smaller drum can be achieved (possible also to use once through steam generator hx which eliminates drum albeit at the cost of different, maybe tighter of boiler wayer chemistry control requirement ), the wall thickness can be less, and thus the maximum load ramp rate can be increased.

    • BushAxe 3 years ago

      The ST will be running continuously to provide power for the plant equipment (pumps, motors, etc) so it will always be ready for dispatch. Ramp-up (rate of output change) would be 20-30MW/min maybe faster depending on the ability to increase steam production.

      • Tom 3 years ago

        5 minutes to full production. That’s OCGT speed.


  12. brucelee 3 years ago

    With the forecast of rooftop PV to cover 100% of daytime load, it basically HAD to be designed as a peaker right?

    • BushAxe 3 years ago

      SA will become a significant exporter of power eventually, we need to build the interconnector with NSW.

    • david H 3 years ago

      It looks that way to me, The plant is designed with a maximum output of 150MW(e)? but is only expected to produce 500GWh/yr, which if averaged over the year equates to c. 57MWe of continuous output.

  13. Joe 3 years ago

    It just warms my blood reading about Premier Jay and his Solar Thermal, just the latest knockout punch on FF. But the push back just keeps coming. We had Glencore’s CEO a week or so again going in hard to defend Coal…but that was him singing from his company’s song sheet as you would expect. Yesterday it was the turn of Bluescope’s CEO to go harder again…the love affair with Coal just gets stronger all the time….with the idea that taxpayers should “subsidise” Coalers ( old and build new ones ) to keep on burning that little black wonder rock.

  14. bedlam bay 3 years ago

    I think the article questioning the $540 M financing of Port Augusta may have been in the finance section of “The Australian” on 17/8/17. I will try to get a copy.
    Tony Simons

  15. bedlam bay 3 years ago
  16. Gary Rowbottom 3 years ago

    Hooray, ecstatic to see this happen at last. Thrilled the SA Government has recognised the job value in this project, as well as the obvious direct benefits to the electrical supply system. To be fair, that was recognised and promoted by various state and Federal political parties too, but it was only the SA Government that had the opportunity to make it happen – I am delighted they took that opportunity. Much work to do for Solar Reserve, but inside 12 months I will be excited to see the tower begin to arise from the bushland north of my home town. Thanks to all supporters of this project.

Comments are closed.

Get up to 3 quotes from pre-vetted solar (and battery) installers.