Concentrated solar thermal power — a core climate solution
I promised this week would focus on technology. Other than energy efficiency (see here), I don’t believe any set of technologies will be more important to the climate fight than concentrated solar power (CSP).
I have a long article on CSP in Salon, “The technology that will save humanity: The solar energy you haven’t heard of is the one best suited to generate clean electricity for generations to come.”
OK, maybe “will” should be “may help” [I’m an optimist, sue me!] and CP readers have heard about CSP for a while (see here). But I do think CSP deserves much more attention:
It is the best source of clean energy to replace coal and sustain economic development. I bet that it will deliver more power every year this century than coal with carbon capture and storage — for much less money and with far less environmental damage….
How much less? Many industry experts told me CSP will likely deliver power for well under $.10 per kilowatt hour fully installed in the next decade.
What is its market potential? I think it could be more than two wedges, which is several thouand gigawatts:
It would be straightforward to build CSP systems at whatever rate industry and governments needed, ultimately 50 to 100 gigawatts a year growth or more.
Why is CSP so important?
Because it’s the only form of clean electricity that can meet all the demanding requirements of this century….
The Salon article goes through those requirements and explains why. The article also goes through some of the fascinating three-millenium history behind “One of oldest forms of energy used by humans — sunlight concentrated by mirrors.”
One final point — based on the early comments on the Salon piece, I realize that many people think that a flaw in CSP is that it would require long-distance transmission lines that would lose massive amounts of power. Well, my (terrific) editor cut out a key clause that I probably should’ve put back:
“We will need more transmission in this country,” [especially low-loss long-distance high-voltage DC lines].
I’ll have to do a HVDC lines piece soon, but you can read more than what want to know about HVDC here at Wikipedia, which notes:
Depending on voltage level and construction details, losses are quoted as about 3% per 1000 km [600 miles].
CSP is a technology that all progressive should become knowledgeable about.
Related Posts:
April 14th, 2008 at 9:29 am
This sounds like excellent technology, I’d love to see it take off. But I’m curious, how well does it scale up or down? Instead of putting PV panels on my roof, would it be possible to have a tiny CSP system instead? Or is there some minimum size to make such a system practical?
April 14th, 2008 at 9:49 am
There are smaller scale CSP technologies for generating electricity — see for instance, http://www.stirlingenergy.com/default.asp.
But those aren’t suitable for thermal storage, so I’m not sure they are going to be long-term winners for utility scale generation. And they are too big for homes.
That said, if you want some solar thermal, you can get a water heater on your home (I have one) and in most parts of the country, this will have a good payback — especially if you’re old water heater need for placement or you are doing any new construction.
April 14th, 2008 at 10:38 am
I too am a big fan of CSP and happy to see coverage.
One comment I have about the Salon article… you imply that the water problem is a non-issue by stating that air cooling can be used. While true that air cooling can be used that is also true of any other power generating technology from coal to nuclear. The caveat is that air cooling increases capital costs and/or reduces efficiencies. Therefore the cost estimates for cents/kWh that you give would likely not hold for air cooled systems… I assume. A more complete coverage of this issue would be helpful.
Also, it would seem that the U.S. regulatory system favors local generation over long distance transmission. In addition to mechanisms such as tax credits and loan guarantees, are there regulatory or structural issues that must be addressed to enable the free flow of electricity on a country wide scale?
April 14th, 2008 at 10:55 am
Ausra quoted me figures with air cooling. I’ll get back to you.
The government helped give us the railroad system, highway system, and the Internet. They are going to have to give us a HVDC system that connects to wind in the midwest and CSP in the SW.
April 14th, 2008 at 11:01 am
Joe, I’ve spent a lot of time researching CSP. I like it. It’s a nice technology. But when I compare the “stop-global-warming-effectiveness” of CSP to liquid-fluoride thorium reactors (LFTR) there’s just no comparison. LFTR has the potential to provide all of the energy the planet needs from a material that is far more abundant than uranium and far less difficult to recover. Take a look at this:
http://www.energyfromthorium.com/ppt/thoriumPotential.ppt
especially slide 11.
April 14th, 2008 at 12:07 pm
Kirk — that will take me forever to download. The problem with LFTR, I think, is that it isn’t a commercial technology today — and it it can take a long timefor a non-commercial technology to be able to deliver large quantities of power at prices below $.10 a kilowatt hour.
CSP is commercial now, and will beat $.10 a kilowatt hour within 5 years.
April 14th, 2008 at 12:13 pm
Joe, it’s only 11MB, it won’t take you that long to download.
The “nuclear” option is often dismissed on sites like these because they are fixed on the light-water reactor paradigm. So they get in the mode of saying things like “we can’t build them fast enough” or “we’ll have to build 10 Yucca Mountains” or “we’ll run out of uranium in 50 years anyway”. Those statements might be true for light-water reactors, but not for liquid-fluoride thorium reactors.
This reactor type went from an idea to a demonstration in 3 years in the 1950s, when we had practically no nuclear modeling capability like we do now. It doesn’t require large forged pressure vessels or huge containments. It get much better efficiency than LWRs and uses a tiny fraction of the resources. It can be built fast and super-safe, because the safety features are inherent to the design, not engineered in later like LWRs. Engineered safety can fail, inherent safety is immutable. This reactor CAN stop global warming AND lift the world’s standard of living at the same time. That’s worth working on.
April 14th, 2008 at 12:46 pm
Geothermal could also be very important. We could also tap old oil wells for geothermal energy.
http://peswiki.com/ index.php/ Directory:Geothermal_Oil_Wells
April 14th, 2008 at 1:35 pm
HVDC losses of only 3% per 1000 km seem quite, quite low to me. I would suggest checking more definitive references if the actual number actually matters.
April 14th, 2008 at 2:47 pm
Why does it seem low?
Seems right according to everyone I’ve talked to.
April 14th, 2008 at 3:27 pm
PhilD, air cooling is trivial for some CSP technologies, such as Stirling dishes, since the cooling happens at each dish. Being distributed in this way is not an option for fossil and nuclear plants. However, this is not an option for CSP with Thermal Energy Storage, such as Ausra and BrightSource. They use conventional steam turbines, which require traditional cooling, which can be via water, air, or a hybrid.
Traditional air cooling technology involves fans to drive sufficient air flow as an alternative to cooling towers. The primary problem with this is that it becomes inefficient on very hot days. This has led to what is sometimes called “hybrid cooling”: using air cooling most of the time, and wet cooling when air cooling becomes too inefficient. You can find out more in the EPRI Journal.
April 14th, 2008 at 4:05 pm
I’m a big supporter of CSP and am trying to get the public in the SW interested in this resource at www.solarsouthwest.org. However I want to be realistic about the challenges facing the industry.
To get to paying perhaps $.10/kWh we have to start deploying this technology at a somewhat higher cost. The $.10 kWh for a fossil fuel replacement plant with storage is with 98% probability not going to just pop out of a laboratory but will come from economies of scale in manufacturing and construction. If Americans want to get a leg up on this they should be willing to pay somewhat more now to start catching up to the Spanish. Abengoa in its literature puts the crossover point for cost of CSP over natural gas plants at about 2018. But if we sit on our hands at higher per/kWH costs, we won’t get these deployed in time or remain dependent on countries (Spain primarily) with better renewable energy support structures to do the heavy lifting.
April 14th, 2008 at 5:15 pm
“For average distances of 5,000 km, HVDC transmission losses would be about 25%.”
source:
cohesion.rice.edu/CentersAndInst/CNST/emplibrary/Hartley%2004May03%20NanoTechConf.ppt
see also
http://www.abb.com/ cawp/ GAD02181/ C1256D71001E0037C125683200658E0C.aspx
April 14th, 2008 at 7:10 pm
Some VCs I know are keen on CSP since:
a) They’re clearly well beyond Research into development and Deployment, which is when something can get big enough to generate big $$.
b) They don’t depend on anything exotic in terms of materials or manufacturing. Although there’s plenty of sand on the planet, we’ve seen temporary impediments like insufficient supplies of solar-grade silicon hold PV back. But, it takes a while to build the factories to create that silicon. Still, CSP & windmills don’t need any of that, and I suspect there are a bunch of automobile-industry military-industry supplies who have useful capacity. one wonders, for example:
1 Abrams M1 tank & fuel it uses over its life = ?? CSP or windmills?
c) I’ve looked for home-size CSP with no luck, definitely hard to scale down, like wind turbines, although I’ve seen several companies trying that. We have too many trees around us for the latter, for sure.
d) For solar & home, besides solar hot water and solar PV (where that works), and just doing better insulation [we just put in some more thermal blinds over the big plate glass windows, and they really help], if I had a wish it would be for really smart window systems:
windows that are solar cells when then can be, reflective when they need to be, maybe with electrically-controlled blinds if needed, run by wireless sensor nets (temperature, light) that talk to programmable thermostats. Most of these pieces exist in some form or other, although I haven’t seen them all together.
Anyway: Joe, have you any thoughts on windows in the built environment? and how much effieicny can be gained? [We’re working on local codes, so this is of broader interest.]
April 14th, 2008 at 7:33 pm
The CSP that is being deployed today should be compared to peaking power. In that comparison, it is quite cost competitive already, as peaking power can cost as much as 25 cents per kWh. The deployment of CSP will bring reductions in cost from the standard “learning curve”. While wind is currently cheaper than CSP, CSP is more valuable because it time of day correlates well with load.
Also note that Stirling Energy Systems’ webpage already says the cost is less than 10 cents per kWh.
NREL’s estimates are that we should see 7 cents per kWh around 2010 and 5 cents per kWh around 2020. These estimates pre-date the rapid rise in commodity prices, but CSP is pretty low-tech, and may be less affected (Ausra likens their materials to Farm equipment).
April 14th, 2008 at 8:35 pm
Sopogy makes smaller scale CSP:
http://www.sopogy.com/
Re: costs…don’t believe all you read. I’ll leave it at that.
April 14th, 2008 at 10:28 pm
Michael: thanks, I’ve seen Sopogy. Unfortunately, I don’t really need 100kw on my roof.
Earl: I think “more valuable” is a little imprecise, i.e., it’s certainly true in CA right now by $, i.e., daylight power costs more than night power
But, I think windpower might be “more valuable” in the long term, in the sense that:
a) At least in CA and the Southwest, there’s enough sunlight for all the daylight power we need, given big CSP farms + PV for micro-scale help. Certainly some of the CSPs can do storage, and if someone invents really great batteries, we’d be done.
b) For baseload, though it seems like we could use a lot of windpower, given that big distributed windfarms can do that. See Prof. Jacobson’s page at Stanford (a lot of good stuff):
http://www.stanford.edu/group/efmh/jacobson/
and specifically, under I. b:
Cristina L. Archer and Mark Z. Jacobson, “Supplying baseload power and reducing transmission requirements by interconnectiing windfarms.” 2007.
http://www.stanford.edu/ group/ efmh/ winds/ aj07_jamc.pdf
Windmills seem like a bigger part of the solution outside the sunny Southwest, i.e., where they “lack” nice empty deserts. I especially like putting windmills on farmland, a nice dual-use.
I also recommend Mark’s testimony on April 9 to the US House of Representatives:
http://www.stanford.edu/ group/ efmh/ jacobson/ 040908_testimony.htm
“I will discuss the scientific findings on the effects of carbon dioxide, emitted during fossil-fuel combustion in California, the U.S. and the world, on air pollution and health in California relative to the U.S.”
See “Results from the studies and analyses are as follows:”
basically, more CO2 hurts CA health worse than it hurts most states.
April 14th, 2008 at 10:47 pm
Re: costs above — not referring to Sopogy
April 15th, 2008 at 11:31 am
I came across this article on Concentrated Photovoltaics (CPV).
http://www.popsci.com/ environment/ article/ 2008-02/ shocker-worlds-largest-solar-plant-use-solar-panels
The article said something about the need to keep the CPV at or below 60 degrees C at least for the metal holding the thing together.
Maybe they could make a hybrid out of these things.
First heating: CPV for the first heating of the water to get the water temp to 60 C.
Second heating: The rest of the heating with the Ausra cheaper system.
Third heating: With the more expensive system that Lutz and the Dept. of Energy developed in the 1980’s.
The thing I was wondering about with the Ausra system was the low water/steam temp they were using. Any increase in operating temps helps in the Carnot cycle.
If they could get electrical power out of the CPV, maybe that would help with their efficiencies and costs.
April 15th, 2008 at 1:20 pm
If I did the calculations correctly, then about 6–7% transmission losses per 1000 km for typical HVDC lines.
I am sure that additional advances will be made, but only 3% still seems a futuristic projection to me.
April 15th, 2008 at 6:56 pm
John Mashey, thank you for the nice wind references. I wrote “more valuable” as meaning providing a better cost of alternative minus CSP cost, i.e. a narrow sense. I think we would both agree that wind and CSP are both valuable in a more general sense. Indeed I think we would both say that diversity in methods of production is important, so we want to see a mix of wind, CSP, and eventually geothermal, ocean energy, etc.
For land use, a wind farm takes approximately 8 times the land area as a CSP farm of the same energy output, but since only 5% of the wind farm area is occupied by the turbines, that means the land displaced is about 60% less than a CSP farm.
April 15th, 2008 at 10:08 pm
Earl:
http://www.actewagl.com.au/Faqs/WindFarm.aspx
says 2%.
http://www.wel.co.nz/index.asp?pageID=2145843299 says 2-3%.
I think this is the difference between the footprint of the blades and the (smaller) footprint of the towers, plus roads, transformers, etc.
http://www.ucsusa.org/ clean_energy/ renewable_energy_basics/ farming-the-wind-wind-power-and-agriculture.html
has a good picture , with crops planted right up (close to) the base, since the baldes are normally high enough for tractors to pass underneath with no trouble.
April 16th, 2008 at 4:20 pm
John, I am using the American Wind Energy Association’s FAQ number which include access roads (the rest of this is cut and paste from their FAQ entry):
How much land is needed for a utility-scale wind plant?
In open, flat terrain, a utility-scale wind plant will require about 60 acres per megawatt of installed capacity. However, only 5% (3 acres) or less of this area is actually occupied by turbines, access roads, and other equipment–95% remains free for other compatible uses such as farming or ranching. In California, Minnesota, Texas, and elsewhere, wind energy provides rural landowners and farmers with a supplementary source of income through leasing and royalty arrangements with wind power developers.
A wind plant located on a ridgeline in hilly terrain will require much less space, as little as two acres per megawatt.
April 16th, 2008 at 9:51 pm
Earl: thanks. From the various pictures I’ve seen, I think the 2X variation comes from their assumptions on access roads and rotor spacing. BTW, an interesting new design:
http://www.flodesignwindturbine.org/cms/
It remains to be seen, but this is a spinoff of some serious aeronautical engineering folks with a good customer list. Of course, one must always apply the cautions in:
http://www.wind-works.org/ articles/ FantasyWindTurbines.html
Anyway, both CSP and wind turbines have a lot of merit.
April 29th, 2008 at 8:47 pm
More informations about Concentrated Solar Power :
www.desertec.org
www.trec-uk.org.uk
www.gezen.nl
www.trec-france.org
May 6th, 2008 at 5:22 pm
Concentrated solar thermal is great in some areas, but not all. It takes a lot of room and has yet to be determined what happens when you basically change the ecosystem that is under the ring of a geothermal tower for it to work correctly. The CSP Trough or even the stirling dishes seem to be a better choices for this type of job
www.solarpower-home.com