About the SA Blog Network



Opinion, arguments & analyses from the editors of Scientific American
Observations HomeAboutContact

World’s first solar power plant that can work at night

The views expressed are those of the author and are not necessarily those of Scientific American.

Email   PrintPrint


How can one use solar energy after the sun sets? Simple: store the sun’s heat in molten salts.

The world’s first solar power plant to employ such technology—a thermal power plant that concentrates the sun’s rays with mirrors on long, thin tubes filled with the molten salt—opened in Syracuse, Sicily, on July 14. Dubbed Archimede—after the famous Syracusan scientist Archimedes who supposedly coined the term "Eureka" for scientific discovery and reputedly repelled a Roman fleet through the use of mirrors to concentrate the sun’s rays and burn the invading ships—the power plant can harvest enough heat to generate five megawatts of electricity, day or night, and can store enough energy to keep producing power even at night or during cloudy daytime hours.

In addition to the benefit of storage, molten salts also operate at a higher temperature (roughly 550 degrees Celsius) enabling them to capture more of the sun’s energy—as well as create the steam for turbines in conventional power plants. Meaning that such solar thermal power plants could be swapped in for fossil fuel-burning ones. This power plant is on the grounds of a natural gas combined-cycle power plant.

The molten salts also don’t burn like the oils used as working fluids in other concentrating solar power plants operating today. If Archimede springs a leak, it will end up with piles of fertilizer (the salts in question are potassium and sodium nitrates, which are literally used as fertilizer in agricultural applications). Of course, the freezing point of such salts is a balmy 220 degrees C so Archimede will have to capture a lot of heat to keep the salts fluid. That’s why the plant’s owner—Italian energy giant Enel—is supplementing the sun’s rays with a little old-fashioned natural gas burning as well.

Of course, it requires 30,000 square meters of special parabolic mirrors and 5,400 meters of high heat-resistant pipe to collect the sun’s rays in the molten salts, even in Syracuse. All that adds up to a building cost of roughly $80 million for just 5 megawatts of electricity. But Italy is hardly alone in pursuing such plants: the Andasol power plants in Spain use more than 28,000 metric tons of such salts to store thermal energy from its otherwise conventional concentrating solar power plants and the U.S. company SolarReserve plans to deploy such molten salt technology in its "power towers" coming soon to the Nevada desert. Eureka!

Image: Courtesy of Archimede Solar Energy

Rights & Permissions

Comments 17 Comments

Add Comment
  1. 1. sethdayal 1:55 pm 08/4/2010

    At $80M producing 9,200 MWh/yr Archimede is 52 cents a kwh in a public utility over 100 cents in a private. New Asian builds of American designed nuclear reactors are coming in at less than 2 cents a kwh (public).

    And it stores power overnight? How about over a rainy weekend or next winter? Not so much. Wonder how much it would cost for that – $2 a kwh?

    But after some stupid dems who are about to lose an election for just that reason, to just slightly more practical repug’s, send some big tax breaks their way, why it’ll be cheaper than coal

    Link to this
  2. 2. Bob Stedi 2:30 pm 08/4/2010

    I’m not sure why this is considered a world first. Andasol 1 in Spain is pretty much the same and has been operating for nearly two years.

    There are score of other Solar Thermal Molten Salt plants, both existing and planned: See

    The South African development is particularly interesting:

    Note that Eskom technology strategy and planning manager Barry MacColl believes that during the life span of this project, which is estimated at about 20 years, the lines will cross where solar becomes more cost efficient than coal.
    And this is from a company that is also operates coal and nuclear plants,

    Link to this
  3. 3. dbiello 2:56 pm 08/4/2010

    Bob – It’s because this is the first to employ the molten salts as a working fluid, not just as a storage option. No synthetic oils coursing through the tubes of this solar thermal power plant, just molten salts.

    Link to this
  4. 4. JamesDavis 7:45 pm 08/4/2010

    "sethdayal", you really are one of those damn dumb anti-environmental republicans, aren’t you. It takes 15 billion dollars (batteries not included) to build one nuclear power plant and 10 years to bring it to production. Compare 80 million to 15 billion and how many liquid salt power plants can you build out in the Nevada desert? …the bigger the holding tank, and the more turbines on a liquid salt plant, the longer it will produce more electricity.

    Link to this
  5. 5. dwbd 10:20 pm 08/4/2010

    Chuck Devore, a state assemblyman from California said this about solar thermal power:

    "…And you need to honestly take a look at what’s happening out in the desert. I’ve visited the solar fields out at Cramer Junction. They produce as I recall about 135 MW of power. They are solar thermal fields that have been up probably for about 20 years now. In other words not very much power. And they cover about 1000 acres or so. And what I was struck by when I went out to visit them was the fact that there was no plant or animal life whatsoever underneath those panels. The entire 1000 acre area was a dead zone, because the plants are a fire hazard for the parabolic trough mirrors. And they spray, like an herbicide to keep the plants from growing and to keep the dust down because the dust and sand abrade the mirrors and increases their maintenance costs. And so what I found somewhat interesting was that in that entire area it was devoid of any life. Now people who don’t live in the desert may think that’s all well and good, there’s no life out there anyway, but as a guy who has spent a lot of time out in the eastern High Sierra I can tell you there’s a lot of life out there…. How much of the desert are we willing to cover? Because the energy density of solar power is so minimal compared to the energy density of something like nuclear power where you can produce hundreds of times the amount of electricity on a much smaller imprint on the ground…."

    If you really want to learn the truth about Solar Thermal Power, there is an excellent analysis going on right now here:

    I notice David Biello has confirmed the absolute impracticality of Desert Solar Thermal power by his inability to challenge any of the logic of their analysis – after I challenged him to do so.

    It’s not too late, David, if any of their analysis is faulty, they will be more than welcome to be corrected by you.

    Link to this
  6. 6. sethdayal 2:42 am 08/5/2010

    You know James, you need to learn to read and stop reciting greenwash bullshit from the denier shop down at Big Oil.

    AP1000 build $1.2B/Gw 2007, 1.3 cents a kwh

    That’s Real cost of American nuclear power built by American engineers in five years or less overseas for public power companies

    Project started 2009 and is on track for completion in 2013 – 4.5 years. .

    The Japanese with a much better regulatory system than our own and higher wages build American designed reactors for $1.4B/Gw one and half cents a kwh.

    Budget cost South carolina AP1000 build Summer 2 3 $4.5/Gw 4.7 cents a kwh if built by public power – bonneville or TVA

    You could built a few CSP plants for those prices but the power out of them would be negligible.

    Link to this
  7. 7. dbiello 8:47 am 08/5/2010

    What I find mystifying is why it’s always absolutes? Why can’t it be solar AND nuclear? Or wind farms AND gas turbines? Heck, even modular reactors on the sites of former coal-fired power plants? It’s the absolutism and fervor of the proponents of any given energy technology that I find bizarre. I understand that funding is ultimately a zero-sum game but this absolutism pushes advocates to ridiculous extremes, typically in denigrating other potential technologies.

    That’s an opinion based on years of observation. Here’s a fact: I don’t have a position on Archimede, I’m just reporting its existence.

    Link to this
  8. 8. frgough 11:29 am 08/5/2010

    It takes $15 billion and 10 years because of greens and their jackboots on our necks.

    When energy is expensive and delayed, turn over the nearest rock and you’ll find an environmentalist.

    Link to this
  9. 9. darthwilliam1118 1:54 pm 08/5/2010

    From sethdayals own linked article: "The cost of building a kilowatt of AP1000 capacity is between $1,000 and $1,200, " the cost of the CSP plant in the article is $16,000 per kilowatt. So a brand new technology is more expensive than a 60 year old government subsidized one? Well gee that’s a surprise. Also for the nuke plant that is just cost of construction, doesn’t include operations, fuel, insurance, waste disposal, plus those pesky uranium mines, shipping, processing ,etc. A nuke projects tend to go way overbudget, like 100-200% so the real cost may be more like $4,000 per kw.

    The big benefit of CSP is that the plant cost is basically it -there is no fuel cost, no mining, no waste, insurance is less (no meltdown risk). Also USA has only 6% of world’s uranium supply, although Australia has 23%. Sunlight is unlimited, free and no supply chain issues other than clouds.

    Link to this
  10. 10. violet 10:21 pm 08/5/2010

    though it’s costly now,after long-term practices,these solar power plants will surely play a significant role in collecting sun’s rays efficiently.

    Link to this
  11. 11. sethdayal 3:40 am 08/6/2010

    Darth my friend you need to start learning about energy. The Archimedes plant $80M producing 9,200 MWh/yr is $80000 a kw average compared to $1200 for the Chinese nuke.

    The day to day cost of running the current ancient low efficiency vacuum tube era nukes including lotsa NRC fees, is 1.86 cents a kwh in the US which includes all ongoing costs you mention and a bunch more.

    On of the latest $2B/Gw nuclear builds in China was a 4 year build of twin Candu 6′s at Qinshan in 2004 on time and on budget. Google it.

    It costs a fortune in maintenance and water to keep the mirrors clean and the moving parts moving on CSP desert installations.

    Nukes can used reprocessed nuke waste as fuel (MOX) for about the same cost as new uranium.

    Link to this
  12. 12. j.quasimodo 5:10 am 08/6/2010

    Storing energy as sensible heat in molten salt is no doubt a useful way to level out interruptions by passing clouds, but it would take an incredible amount of salt to keep a big plant running overnight. And there’s an efficiency problem: to draw down the energy in the salt involves a steadily falling temperature, which will lower the steam pressure for the turbines, which makes them less efficient. So storing energy as sensible heat is of limited application.

    Molten salt has been used for decades as a heat transfer medium in certain chemical processes and it is not benign. While it doesn’t catch fire when it leaks, as oils do, it will start fires. It’s also corrosive, so the materials of construction are expensive. And because it does not take advantage of a phase change (like steam or certain organic media) you have to circulate a large volume to move a useful amount of heat — big expensive pumps, high energy consumption.

    Link to this
  13. 13. j.quasimodo 5:36 am 08/6/2010

    Amen to the comment about polarizing the question of choosing winners. The hand of government should be light and deft — otherwise you get foolish policies like corn-derived ethanol and electric (i.e., coal-fired) cars that develop a voting constituency and thus are hard to get rid of.

    That is not to say that market forces will efficiently solve all problems. Our air conditioners, furnaces and cars use drastically less energy than 20 years ago, and that would not have happened without standards. Market forces in the presence of standards will identify the best way(s) to get there, and any subsidies should performance -based, not technology-specific.

    Link to this
  14. 14. jerryd 2:33 pm 08/6/2010

    Sethdayal rigges his solar number then rigges nukes in the other cway.

    While for these first yr the solar is $.52/kw but it should run for 50+ yrs. Now when he does nukes first he lies about nuke costs by a factor of 5-10. They just cost out 4 new nukes in Fla, AP100, and they were quoted at $8.5k/kw.

    The SC are old quotes that no longer apply and recently asked almost the amount he said just for prefinancing cost!! He is also saying the building new nukes cost less that a NG plant would by 2x’s.

    Back to the article, both solar 1+2 did this 10-15 yrs ago.

    Link to this
  15. 15. sethdayal 4:36 pm 08/6/2010

    Jerry my friend Florida is not the only place on earth. in fact between your extremely inefficient and power company and the vast troves of corrupt business men, bureaucrats, and politicians whose palms always need greasing , everything in Florida costs double.

    The SCANA costing is less than 6 months old. Jerry is a great tinkerer but he hasn’t learned to read.

    American engineers used to build nukes here for $.7B/GW ($2010) before Greenpeace attorneys got control of the Nuclear Rejection Commission and forced an order of magnitude increase in cost with no safety benefit.

    Vogtle and south texas projects built in the USA by American attorneys with some American engineers assisting are down to $4.5B/Gw. Vogtle twins in China built by Chinese and American engineers are at $1.2B a Gw.

    Link to this
  16. 16. afoladris 9:36 am 08/8/2010

    Guess this will be the being of NEVADA 2

    Link to this
  17. 17. Bob Stedi 11:42 am 08/12/2010

    The EPR project in Finland confirms darthwilliam’s comment on how Nukes tend to go overbudget:

    "Areva has acknowledged that the cost of a new reactor today would be as much as 6 billion euros, or $8 billion, double the price offered to the Finns."

    That’s $5,000 per kw. — just for construction.

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Email this Article