Clean Power Crescent Dune's 1.1 GW-hour storage capability is almost 40 times the size of the largest battery storage project in construction or built to date.

Published on February 22nd, 2016 | by Susan Kraemer


Crescent Dunes 24-Hour Solar Tower Is Online

February 22nd, 2016 by  

Crescent Dunes

SolarReserve’s breakthrough Crescent Dunes solar tower with thermal storage has successfully generated electricity at its full 110 MW capacity.

With this milestone achieved, the CSP (Concentrated Solar Power) project has now passed the necessary test to begin full commercial operation under its 25-year Power Purchase Agreement (PPA) with NV Energy to supply power well into the night for Las Vegas and other parts of Nevada.

The molten salt receiver actually exceeded design expectations, SolarReserve CEO Kevin Smith told CleanTechnica. The heat transfer efficiency of the receiver is the key performance validation of SolarReserve’s patented solar thermal storage technology.

“We are meeting 100% of our requirements so far. The technology has been fully proven,” said Smith. “There’s additional testing that our EPC; ACS Cobra has to do on the balance of the plant and once that is complete in the next few weeks, we start to ramp up our annual output.”

Under the rollout plan with NV Energy, Crescent Dunes will now begin its official ramp-up over the coming year, with generation increasing gradually each month.

“The PPA gives us a year to ramp up. There is that flexibility because it recognizes that this is a completely new technology at this scale. However, based on the successful test results to date, we believe the ramp-up period will take less than the official timeframe.”

Crescent Dunes completed synchronization with the grid in October. Then, fittingly, the first 24-hr solar tower in the US generated its test electricity at 11 o’clock at night.

Crescent Dunes generated its first solar electricity at night

Crescent Dunes generated its first solar electricity at night.

Oddly, for CSP, the EIA posts a tally not just of output in MWh, but also of solar “fuel” input in BTUs (British Thermal Units).

“They care about BTUs because historically fossil fuel costs are the biggest part of a conventional power plant,” Smith explained, while pointing out the silliness of counting up BTUs in a free fuel like solar.

With his 20 years of engineering experience with conventional thermal power, he finds complaints about solar’s supposed low efficiency amusing. If efficiency is generation output divided by fuel burned, solar is the winner with “infinite” efficiency, as its generation has a fuel input of zero.

The EIA database shows Crescent Dunes first generated 1,831 MWh in November. This represents only several days generation at the end of November, Smith explained. Then the contractor took the plant down again to make some modifications.

ACS Cobra, a global leader in the engineering and construction of thermal solar plants, is a  investor as well as EPC contractor. The Spanish firm has taken an “unhurried approach” to getting the project right.

However, since SolarReserve’s contract doesn’t kick in till Cobra hands it over, SolarReserve has not been harmed by Cobra’s pace. “Most of the delay has not been due to technical issues, but to the remote location of the project and the fact that this was Cobra’s first major power project in the US,” he said.

Smith had confided in October that, in fact, they had actually been more or less continuously operating the salt system and receiver since January, when they devised the algorithm change to reduce standby solar flux, reducing avian mortality. At that time, he noted a near-nil avian mortality rate over that 11-month period.

Crescent Dunes is online

Avoiding the Naysayer Pile-On

A disincentive for premature operation announcements would have to be the astonishingly antagonistic coverage that dogged the first power tower project in the US. Ivanpah grew up in the spotlight, with the media obsessively focused on every possible drawback, including even the idea that completely novel technology needs time to ramp up.

“It’s unbelievable what that project has had to endure and continues to,” said then-NRG spokesman Jeff Holland. “It’s like a lightning rod for news. Everything else is so quiet.”

Ivanpah was the first concentrated solar power tower to operate on direct steam, heating water to steam in the receiver and running the turbines on the steam, a simple and pure concept, as the heir to Luz.

The downside is you really can’t store steam, so the first power tower had no cushion in the form of storage. When that turned out to be difficult even for some daytime operation, BrightSource needed to increase the natural gas use, to “prop up” the system during fluctuating solar resource.

1.1 GWh of Molten Salt Storage Makes Crescent Dunes Different

Crescent Dunes technology is completely different. Molten salt circulates throughout an integrated energy storage system. The salt “cools” to 500 degrees F (hardly cool: this is twice the boiling point of water) before being circulated back through the receiver again to be solar heated to 1,050°F again.

Storage in molten salt can stay hot for months, according to Smith. Normally, it isn’t left there, of course, but cycled daily as needed, tapped by night for generating electricity, and replenished by day by the sun.

Due to its innovative molten salt solar heat storage, SolarReserve’s US-designed technology can generate dispatchable solar electricity 24 hours a day, or on demand when a utility requests it during peak demand periods.

Crescent Dune's 1.1 GW-hour storage capability is almost 40 times the size of the largest battery storage project in construction or built to date.

Crescent Dune’s 1.1 GWh storage capability is more than twice the size of the largest battery storage project in construction or built to date (which is 400 MWh).

The cost to include the 1.1 GWh of thermal energy storage at Crescent Dunes is very low because it is integral to the firm’s patented molten salt receiver and storage design.

Batteries have appropriate uses — to run an EV or take a solar roof off the grid, for example. But adding a battery to an equivalent utility-scale PV project cannot come close in lifecycles or cost.

The US Department of Energy has estimated that, by 2020, battery prices will be $125 per kWh. But that is about 10 times the cost — even back in 2010 — to build thermal energy storage at only $30 per kWh.

Compared to a lifetime of at least 30 years for a molten salt storage system (and no replacement costs), batteries aren’t close to competitive. With 10 hours daily of storage, the 110 MW Crescent Dunes project stores 1,100 MWh/day. SolarCity’s recent utility-scale battery in Hawaii stores only 50 MWh/day.

PV is the cheapest solar for small-scale daytime use. But CSP with molten salt storage is the cheapest solar that provides bulk energy that is dispatchable-on-demand round the clock, whether that is after sunset or before sunrise.

“It is shortsighted to only look at the immediate cost, and not really on the value that you can give to the energy system, the role that CSP can play to backup, and in combination with PV and wind add up to a really very high share of renewables, because it can provide power on demand when other fluctuating renewables can’t do it,” said Christof Richter, Executive Secretary of SolarPACES.

To fully de-carbonize the grid, we need PV for days and CSP for nights.

CSP Can Displace Natural Gas Plants Internationally

For now, CSP is cheaper than natural gas only internationally, where the fuel costs up to $10 per mmBtu or higher, not $2 like in the US.

Next, for example, in South Africa, SolarReserve is breaking ground on Redstone this year, and bidding three additional 150 MW projects in the current Round 4.5, and has a permitted 260 MW CSP/PV combo in Chile, Copiapo.

But the recent solar ITC extension was gained in return for allowing the fossil fuel industry to export oil and gas. And shipping natural gas overseas will likely raise its price back home, helping to make the case for CSP instead.

Smith confirmed that CSP with storage is technically able to run as a peaker plant, since it can start up, on demand, in about the same time — about 20 minutes or less. As spinning reserves increasingly take over the first seconds to half-hour required, “peaker plants are utilized both as day ahead and within day,” he said. “You get some day ahead warning, but within a day you might only get three hours’ notice or less.”

He is not enthusiastic about operating CSP as a peaker plant, however, as peakers might only be deployed a few hours a day, or a few days a month; dramatically lowering their capacity factor.

“And then you have a revenue problem. You don’t want to just wait for somebody to call you. After all the sun is shining all day.”

So, a typical workday for Crescent Dunes will be a solid 12-hour shift, working for NV Energy from noon to midnight.

Complete our 2017 CleanTechnica Reader Survey — have your opinions, preferences, and deepest wishes heard.

Check out our 93-page EV report, based on over 2,000 surveys collected from EV drivers in 49 of 50 US states, 26 European countries, and 9 Canadian provinces.

Tags: , , , ,

About the Author

writes at CleanTechnica, CSP-Today and Renewable Energy World.  She has also been published at Wind Energy Update, Solar Plaza, Earthtechling PV-Insider , and GreenProphet, Ecoseed, NRDC OnEarth, MatterNetwork, Celsius, EnergyNow, and Scientific American. As a former serial entrepreneur in product design, Susan brings an innovator's perspective on inventing a carbon-constrained civilization: If necessity is the mother of invention, solving climate change is the mother of all necessities! As a lover of history and sci-fi, she enjoys chronicling the strange future we are creating in these interesting times.    Follow Susan on Twitter @dotcommodity.

  • I mistakenly left out a sentence from this article, here inserted between these two sentences:
    The US Department of Energy has estimated that, by 2020, battery prices will be $125 per kWh.

    “Currently battery prices are expected to come down to $300 per kWh to build.”

    But that is about 10 times the cost — even back in 2010 — to build thermal energy storage at only $30 per kWh.

  • DPL

    The acronym “PV” is not defined anywhere in the article. Fail.

    (Yes, I know what it means due to the magic of Google. That’s not the point.)

  • J.H.

    Whats ironic is that all this energy is going to the COCH BROTHERS

  • J.H.

    I wounder how ( scalable) this molten salt technology can be, micro? I’ve always have been a csp via parabolic trough (George Plhak) diy, kind a guy. I’d like to see the patent. Storage is what it is all about. But I like the hibrid PV CSP systems

  • Musi

    thanks for the article. any ideas what the estimated annual capacity factor is and how it would compare to other 24/7 CSP plants like Gemasolar ?

    • Bill_Woods

      Estimated annual production is 500,000 MW-h, or 57 MW-yr.
      Capacity is 110 MW, so capacity factor is 52%.

  • Stan Hlegeris

    “The salt “cools” to 500 degrees F (hardly cool: this is twice the boiling point of water)”

    Good article, but pay attention to accuracy. It makes no sense to say 500F is “twice” the boiling point of water. 212F is a meaningless number resulting from Fahrenheit’s goofy placement of the freezing point of water at 32F and the arbitrary units he scratched on his thermometer.

    We can all just agree that 500F is hot, as is 260C or 533K. Water at that temperature contains twice as much thermal energy as water at some lower temperature, but the silly numbers assigned by Fahrenheit contain no information about this.

    • Ulenspiegel

      Yep, it would help to give temperatures in K, a simple estimate of efficiency is then possible by using the carnot cycle.

    • Matt

      Oh the USA was so close to meters, kg, and C when I was doing my engineering degree. Then the politician with small minds and smaller balls, stop it for fear of getting their grand mother upset. If only we could charge them for all the mix up that have happen since. The Mar missions come to mind.

  • John Ragozzino

    Let’s wait for 3 month’s data before we shut down the natural gas generators…..

  • Brian

    This could replace natural gas in some areas, and pv plants could replace natural gas in others. With the fracking, and earthquakes, natural gas is unacceptable. Biogas from landfills, and food waste could be acceptable, but not natural gas. We don’t need it. One solar power plant 100 miles square in Nevada could supply the nation with clean electricity. Of course we would never do this because too much would be lost in transmission, but a massive network of decentralized PV solar power plants, CSP solar power plants, rooftop PV, and wind farms could power the nation, without dangerous natural gas. We don’t need polluting natural gas. We can shut all our dirty coal plants by dramatically scaling up clean solar and wind power.

  • eveee

    In the caption, the claim that Crescent Dunes 1.1GWhr storage is 40 times the size of the largest battery storage project in construction or built to date is probably outdated.
    The largest battery storage project under construction today in the US, is a 100MW facility, for SCE and built by AES, that operates for four hours, for 400MWhr.

    “The 100 MW project being developed by AES Energy is a lithium-ion battery system that has a four-hour duration (400 MWh)”
    Thats just under 3x smaller, not 40x. Battery storage is coming along.

    • Thanks. I updated that caption.

    • “In construction or operation” is correct. I’ve learned the hard way in covering renewables sadly that “in development” doesn’t necessarily mean something will get made.

      “In development” just means something like (depending on the technology) like – that lab tests look promising, or permits are under way or have been obtained, or bids have been made or land secured.

      But there’s many a slip, especially for big jumps like going to 400 MWh battery storage for utility scale application when currently the largest is 54 (Tecachapi).

  • Philip W

    Interesting article, I like CSP with molten salt storage more and more.

    There is also a great video showing the Ivanpah solar farm. I know it’s an ad for a camera-drone, but the pictures are still amazing though.

    • Very sci-fi!

    • We’re lucky to have Susan’s expert articles on the topic here. 😀

      Very cool video. 😀 That plant is beautiful from the air.

  • Harry Johnson

    Hallelujah! Naysayers to CSP take note, PV isn’t the only game in town. The simple fact is that steam driven turbines work and when storage is added, it works 24/7. Some geeks like the gee whiz technology of batteries but the cost difference is just too much and we’ll need those batteries for a billion vehicles.

    • Haha – so true! Especially in California and the other states that have pledged 100% EVs by 2050.

    • But what about the potential of used EV batteries at a big discount?…

      Though, I agree, CSP + storage is great and important as well. Has had a hard few years…

  • Interesting, first time I see the price per kWh comparison between molten salt storage and lithium ion batteries. I wonder what the overall PPA price is for energy coming from this plant and if the price comparison in storage costs includes O&M differences. I would think that a molten salt storage system requires more maintenance than a battery storage system.

    • neroden

      It does require more maintenance. It probably still is cheaper for nighttime energy delivery.

      It’s interesting because CSP really is competing with *storage*, not with power generation. It’s more expensive than wind and it’s more expensive than daytime solar and it’s more expensive than hydro, but it’s cheaper than batteries, cheaper than compressed air, and competitive with pumped-storage hydro.

      CSP is one of several nighttime technologies which will be used. The heavy deployment of batteries for other reasons will probably cover the short-term “sunset” usage peak. I expect that the “noon to midnight” service cycle of this CSP will be replaced eventually with a “sunset to sunrise” service cycle.

      • GCO

        Agreed that CSP seems very well suited for storage.

        Now, on what do you base your “more maintenance” claim compared to batteries?

        Keeping a battery system going for 30+ years will require some pretty heavy work I reckon, likely at least one complete swap or rebuild. Wouldn’t that end up costing more than maintaining pumps etc (or whatever wears down in a CSP plant)?

    • The PPA was for 13.5 cents a kWh, regardless of whether it is generated day or (from storage) by night. That is comparable to the 14.5 cents SolarCity gets for its PV+battery combo in Hawaii this year.

  • Thanks, Susan. Learned much about CSP again from you (& molten salt storage). Quite interesting, and am curious to see where this tech goes.

    • Ross

      CSP has been getting little love but this article is really positive about it, particularly on the storage costs and its use as a nighttime power source.

  • LA Fran

    As a hybrid approach, could a molten salt storage solution be used with PV, if the electricity was used to heat the salt. Actually I suppose the question is not can it, but rather how would it compare cost wise to using Lithium batteries instead.

    • joshua

      With a plant that concentrates the sun 5000 times, the efficiency can get up to ~80% (even higher for higher levels of concentration).

      If you instead did PV->electric heaters->molten salt, then you would be limited to the efficiency of the PV, ~25% at best.

      The simplest explanation of why this would not be as good is that for CSP, we are able to use the entire solar spectrum to heat up the working fluid. In the case of PV, we only capture the part of the spectrum that the panel is tuned for.

      But, if you were talking an off-grid installation, where you wanted to use thermal storage to provide heating, then this could definitely work (since then you are comparing it to PV->battery->heaters).

    • vensonata

      Yes, I asked the same question a few years back, and got a similar answer to what Joshua says below. But simple thermal storage for domestic hot water and space heating is as cheap as molten salt. It can use PV and need be no higher than 150 F. Most people have the storage container already in the form of a hot water tank. A half decent insulated house (and many American houses are not even ‘half decent’) is also a storage vessel. In that case instead of turning down the thermostat in the day when you are out, you turn up by 4 degrees and then in the night you let it fall to 65 degrees. Since space heating and water heating are the major portion of energy demand in houses this will be an important form of storage for PV electricity….as well, of course, as the electric car.

    • LA Fran, previously, I’ve interviewed a couple of DOE award winners who do believe that molten salt storage can indeed work alone: what they said is at

      • Frank

        Nice link, and nice article. I think some of that “drama” with Ivanpah is like the BS drama (often in the form of a problem or conflct) that they insert into a lot of “reality shows”. It’s just there to try to make the story more appealing or interesting. It’s certainly not there to make it more accurate. Of course it’s very annoying when that is treated as a reason to be against it.

    • SolarReserve’s combination PV/CSP tower in Chile is bidding at “well under 10 cents per kWh” The PV is not used to electrically heat the salts, (more inefficient) but rather to supply the daytime generation, while pretty much all the CSP is soaked up in the storage and then used at night or dispatched when needed.

      • I like that approach… curious to see how quickly that takes off.

    • Burnerjack

      Or, as an alternative, using flywheel/motor pairs like those from Beacon Power? It all comes down to storage space, capital and operating costs.
      Lithium batteries, like most things, have fairly finite lifespan. Flywheels, on the other hand, have near infinite lifespans. Replace the bearings, inspect for cracks, etc. and rock on.

Back to Top ↑