IBM applies supercomputer cooling to solar collector for 80% efficiency


April 25, 2013

IBM is developing an affordable High Concentration PhotoVoltaic Thermal (HCPVT) system that uses cooling technology from supercomputers to harvest solar energy and produce drinkable water

IBM is developing an affordable High Concentration PhotoVoltaic Thermal (HCPVT) system that uses cooling technology from supercomputers to harvest solar energy and produce drinkable water

Image Gallery (15 images)

Solar power may provide a clean, abundant source of energy, but we know the sun's rays are capable of much, much more. Aside from generating electricity, we've seen solar energy harnessed to produce drinkable water as well, so why not combine the two processes into one system? That's what IBM and its collaborators are hoping to do with an affordable High Concentration Photovoltaic Thermal (HCPVT) system that uses cooling technology from supercomputers to harvest solar energy more efficiently, and produce purified water at the same time.

The current prototype consists of a large parabolic dish made up of several mirrors, connected to a sun-tracking system. The majority of the sunlight hitting the dish is reflected and focused onto hundreds of triple-junction photovoltaic chips, all fitted to microchannel-liquid cooled receivers. Individually, each chip measures just 1 cm x 1 cm and can generates an average of 200–250 watts over an eight-hour period on a sunny day, at an efficiency of about 30 percent.

Thus far, this roughly matches the electrical power output and basic design of other concentrated solar systems in existence, but the cooling system is what really makes the HCPVT system stand apart. IBM adapted the cooling technology it developed for supercomputers like Aquasar and SuperMUC for use with photovoltaics to create a system that continually pumps water just a few micrometers away from each chip through micro-structured layers.

IBM says this method is 10 times more effective than using air-cooling, and maintains a stable temperature over the chips to prevent them from melting. The cooling system would allow the chips to remain operational at 2,000 times the intensity of the sun's rays, but IBM claims it can still provide a safe temperature up to 5,000 times concentration.

In those IBM's supercomputers, the heat absorbed by the liquid coolant has been used to heat the physical buildings sheltering the system itself, but the HCPVT system developers have a different idea. Instead, the heated waste water could be diverted to a desalination system, where it vaporizes and purifies salt water. The researchers estimate this could produce 30-40 liters of clean water per square meter of the receiver area in a day. Alternatively, the scientists would like to direct the heated water to an adsorption chiller, which could produce air conditioning for a nearby area.

By combining electrical and thermal collection units into a single setup, the research team predicts the HCPVT system would be able to convert 80 percent of the captured solar energy into a usable form.

Another advantage to the HCPVT system is that it would cost considerably less than comparable solar energy systems, while still running more efficiently. Much of the system would be comprised of lightweight concrete and metal foils, as opposed to most other solar collectors that are constructed out of pricier glass and steel, leaving only the small, high-tech components to be produced in Switzerland. This also has the added benefits of lowering the costs for assembly and maintenance, which could potentially expand the number of regions that could implement it.

The research team claims this design would cost less than US$250 for each square meter of aperture area and produce energy at a price of under US$0.10 per kilowatt-hour (kWh). According to IBM, this puts the system on par or lower with energy costs for coal power stations.

Currently, the prototype HCPVT system is being tested at an IBM research lab in Zurich, with additional prototypes planned for Biasca and Rüschlikon, Switzerland in the future. The researchers also hope to begin constructing larger versions of the system in remote locations at some point, but we'll have to see if they can also build the necessary infrastructure needed for such an undertaking.

Scientists at IBM Research, Airlight Energy, ETH Zurich, and Interstate University of Applied Sciences Buchs NTB began collaborating on the project after receiving a three-year grant for US$2.4 million from the Swiss Commission for Technology and Innovation.

IBM Research's manager of advanced thermal packaging, Bruno Michel, explains the HPVCT system further in the video below.

Source: IBM

About the Author
Jonathan Fincher Jonathan grew up in Norway, China, and Trinidad before graduating film school and becoming an online writer covering green technology, history and design, as well as contributing to video game news sites like Filefront and 1Up. He currently resides in Texas, where his passions include video games, comics, and boring people who don't want to talk about either of those things. All articles by Jonathan Fincher

This is awesome!!!!!!!!! I can also be a tester in Maine where we have harsh winters.

Artem Down

Isn't it fascinating what relatively small grants to research organizations can yield, compared to subsidies for Big Oil and Coal? Also, the per watt cost comparison to coal plants is hog wash as it counts none of the many "externalities" conveniently not counted for fossil fuels which are paid none the less by the general public.


I need one of those for my home ! Does IBM need to test this device in central america ? I can do that !


when do you think this might be marketed at the consumer level. would be nice to have this producing water and power on my property.

Grant Baker

@Grant Baker One's thing's for sure, they'll have to downsize that collector unless more than one household can use the system...

Everett Cox

Absorption chillers and vapor cycle desalination need to dump waste heat as well so it is possible to do both with the same heat.

keeping as much shadow as possible as possible off the collection area as possible makes for more energy collection. I prefer an asymmetric reflector with the focal point off to the side.



I realize concrete has come a long way... but it's far too heavy to make a sun-tracking dish out of, right?

"Much of the system would be comprised of lightweight concrete and metal foils, as opposed to most other solar collectors that are constructed out of pricier glass and steel".

If you're looking to be cheaper than steel, there are quite a few options lighter than concrete. Fiberglass construction with metal foil would be my choice.

Although if we could find a way to use carbon fiber on this, I say, go for it.

William Carr

re; William Carr

Heavy does not mean immobile.


Bravo on the use of concrete and foil as an alternative material used for this renewable power.

Gary Richardson

The idea of a concentrated photovoltaic system with water cooling is nothing new, IBM is several years late to this party. zenithsolar sells this exact setup since 2009 but their production system is 70% efficient. Each ZenithSolar Z20 unit generates 5 KWh Electricity plus an additional 11 KWh hot water (up to 100 degrees C) I wanted 1 for my house several years ago but it was too expensive. We will see if IBM costs are any better.


Could the steam be used to power a steam engine or turbine before it is condensed to water? Seems like that would be a little more efficient.

Mike Kling

@Mike Kling, A setup like this would produce very low quality steam. The efficiency of a steam turbine maxes out maybe 50% in the best cases but this would require steam temperatures in excess of 500C. The hotter the inlet temperature and the cooler the exhaust temperature the higher the efficiency. Using the heat from this setup to directly produce electricity would be massively inefficient and drop the overall efficiency from 80 to well under 40%.


Ah, LordFlashHeart, so still more efficient than coal-fired power, which has an efficiency of little more than 30%. Also, there have been recent breakthroughs in the production of super-critical steam, using solar energy, which already make your claims seem outdated.

Marcus Hicks
Post a Comment

Login with your Gizmag account:

Related Articles
Looking for something? Search our articles