They do it with Mirrors: Concentrating Solar Power

Note 5/3/09 Some more recent CSP articles are here.

I’ve just spent some time reviewing a pile of reports on concentrating solar power (CSP) technologies for Ratepayers United Colorado (RUC), so here is a summary of the various types.


Technology descriptions

        Concentrating Photovoltaic

o       This technology uses mirrors or lenses to focus sunlight on high-efficiency photovoltaic chips.  The extra sunlight makes it worth the expense of making a more efficient higher complexity chip because each chip can convert more sunlight to electricity, with conversion efficiencies often twice as high as the efficiencies of conventional soar panels.  This also has the advantage of saving silicon (which is in short supply) for making chips.  Problems are that they do not work as well as conventional panels in diffuse light through clouds (because the light cannot be focused) so they are only appropriate for areas with very little cloud cover, and the extra light heats the chips more, which lowers their efficiency, and so may require some sort of additional cooling loop.

o       Article links:



        Parabolic trough

o       The oldest CSP technology, parabolic trough plants, known as SEGS 1 through 9, have been operating reliably in theParabolic solar trough technology  -  such as this SEGS plant - will be reborn in Red Rocks, California.
California and Nevada deserts since the 1970.  Parabolic trough plants work by focusing sunlight on pipes by means of parabolic mirrors.  These pipes contain a working fluid (several have been used, from water and superheated steam to molten nitrate salts.)

o       Parabolic trough technology is currently experiencing a revival, with several new plants being built.  The using of Organinc Rankine Cycle generators allows solar trough plants as small as 1 MW to be built (such as the new Saguaro plant north of
Tucson, AZ.

o       Parabolic trough technology allows energy to be stored as heat, which is much less expensive than storing electricity.  This allows the energy from these plants to be available at times of peak demand, making the electricity much more valuable.

o       Steve Raabe recently wrote an article for the Denver Post providing a good overview of the prospects of this technology in
Colorado.  The only point that he missed is the potential for hybridization with existing coal and gas plants.  By preheating steam for an existing fossil fuel fired turbine, CSP can make an old power plant operate much more efficiently.  Arnold Leitner of SkyFuel ( tells me “Our preliminary engineering estimates, satellite imagery of the locations and solar data show that SkyFuel could supply 50-100 MW-electric solar steam to the Comanche power plants generating an estimated 65,700-131,400 MWh of pure solar power at the facility via the existing steam turbine. SkyFuel could deliver this solar-generated steam to the power plant at an effective fuel cost commensurate to the fuel cost of burning natural gas at a modern combined cycle power plant at fuel price of 7-8 cents/mmBtu. In other words, through a FuelSaverTM at a coal-fired power plant SkyFuel could provide solar energy at the price of natural gas generation.”

o       Considering that solar power is available during peak demand, gas ifered generation is the appropriate cost comparison (as opposed to wind power, which does not deserve (or need) a price premium due to its unpredictable timing.

o       A variant on this called Concentrating Linear Fresnel Reflector (CLFR) uses many thin mirror strips instead of parabolic

troughs to concentrate sunlight from a large field onto just two tubes of working fluid.  This has the advantage that flat mirrors are much cheaper to produce that parabolic mirrors, and also allows for a greater density of reflectors in the array, allowing more of the sunlight to be used.

        Power Tower

o       Power Tower technology is similar to solar trough technology in that it uses mirrors to concentrate sunlight on a working fluid which is then used to superheat steam to run a turbine.  The difference is that the mirrors concentrate all the sunlight onto a single receiver at the top of a tower.  This allows for higher temperatures, but leads to engineering problems because of the high temperature at the receiver.

o       Lower price per watt is theoretically possible compared to trough technology because of the higher temperatures.

o       So far, only two pilot plants have been built and operated, called Solar One and Solar Two (actually the same facility as Solar One, but converted to use nitrate salts as the working fluid rather than superheated steam).    Both Solar One and Two incorporated thermal energy storage.  Due to the success in demonstrating the technology of Solar Two, a commercial 15MW plant Solar Tres is in the planning stations.  This station will incorporate enough thermal storage in molten salt tanks for 24h operation.

        Solar Chimney

o       A Solar chimney consists of a large greenhouse (multiple square miles of area covered by a transparent roof) which is sloped gently up to a central hollow tower or chimney.  The sun heats the air in the greenhouse which then rises up the chimney driving an air turbine (similar to the hydroelectric turbines used to generate power at dams) in the chimney as it rises.  Water filled tubes on the floor of the greenhouse serve as heat storage which allows the chimney to operate even at night and on cloudy days.  The amount of water in the tubes can be changed to alter the profile of power production and match it closely to the power demand the chimney serves.

o       The edges of the greenhouse can actually be used for agribusiness to grow plants, so not all the space taken up is solely devoted to electric production. 

o       The beauty of solar chimneys is that they are extremely low tech, and can be built without heavy equipment using simple materials.  The only exception to this is the turbine, and even that is much less complex than turbines used to generate power from wind, because the wind in a solar chimney is much more regular than naturally occurring winds and storms that wind turbines have to deal with. 

o       The first solar chimney was built in
Manzanares, Spain and ran continuously for 32 months in the late 1980s with 95% availability (considerably better than most coal and nuclear plants.)   See a video tour of this chimney I ran across on EcoGeek.

o       A 200MW chimney is planned by EnviroMission of Melbourne Austrailia for the Austrailian Outback.

o       It may be possible to build solar chimneys on south-facing slopes or simply as an extra layer of glazing on tall buildings with a turbine at the top which would make them even cheaper by avoiding the necessity of building the tall chimney (my idea).


        Dish Stirling

o       A Dish Stirling system is a parabolic mirror which focuses heat directly on a Stirling engine, a simple closed-cycle engine which operates simply using any heat source.  Sometimes hybridized with a fossil fuel source to provide heat when the sun is not shining. 

o       Dish
Sterling systems have the advantage of small size and scalability, because each individual mirror-engine system produces only around 25kW, but many can by linked together.

o       Because the suns rays are focused directly on the engine, there is little opportunity for thermal storage, a great advantage of several other thermal concentrating technologies.

o       Stirling Energy Systems currently has a few demonstration systems in operation.  They have signed purchase agreements with two
California utilities to build a total of around 1 GW of electric generation, but both projects are still in early testing phases.

o       According to Sandia National labs, this is the most efficient technology for converting sunlight into electricity.

CSP Technology comparisons

Technology Scale Levelized cost per MWh Pros/Cons Complexity/ deployment

Concentrating PV



No storage option; does not work well on cloudy days

High.  Beginning to be deployed in last couple years.


25 kW per dish


High efficiency, modular.  No thermal storage.

Just beginning to be deploy

Parabolic Trough

Most: >50 MW for economies of scale;

Organic Rankine Cycle 1MW+

$8-$18 current

$6 potential

Can hybridize with existing fossil plants.

Storage, well understood technology, needs water for efficient cooling.

Plants operating consistently for 30 years in CA.

Solar Tower

>30 MW

$18+ current,

$5 potential,

can hybridize with existing fossil plants.

Storage.  Potentially cheaper than solar trough. Needs water for efficient cooling.

2 pilot plants with operational history.  First commercial plant now operational.

Solar Chimney


Not yet known.

Baseload power, low maintenance.

Low complexity, great potential for 3rd world.


Comments (31)