By Bob Moser, Americas correspondent

Alternatives for CSP cooling are still few and far between, with wet and dry cooling methods the only feasible options today - both of which have drawbacks. But developers and researchers are exploring different configurations for hybrid cooling systems that could combine the best of both worlds.

Dry cooling has been around for the last 15 years, but research and development haven't evolved the process beyond its faults during that time, particularly in the problem-area of ambient temperature, says Babul Patel, principal consulting engineer in renewable energy for Nexant. With CSP plants focused on dry, sunny areas of the United States, during the summer months when CSP must perform it suffers with dry cooling because of intense solar radiation.

“It's a double-edged sword, in the sense that dry cooling is feasible, but it impacts plant output exactly when CSP has the ability to deliver the maximum output,” Patel said.

Hybrid cooling

At the top of the list for new cooling approaches, at least in terms of feasibility, are varied approaches to a hybrid cooling system. A recent study by NREL found that climate factors play an important role in how a parabolic trough plant's efficiency and cost outputs are affected by hybrid cooling.

The minimal use of water, primarily in an air-cooled condenser, may prove to be the best way to lower the amount of water used while still harnessing its benefits as a cooling agent, says Jordan Macknick, energy and environmental analyst with NREL.

“Options could be misting of the air, or flooding the (cooling) area before pushing the air through,” Macknick said. “We found though that optimizing the size of your air-cooling condenser, making it much larger than what's been considered before, may increase your upfront capital costs but get you a lot more output in the plant. We predict a lower performance penalty and lower cost penalty over time by adjusting the size of your air-cooled condenser”

NREL's study found that dry cooling would generally achieve a 90-92% decline in water consumption, with an increase in cost for generating electricity of 3-8%, depending on the climate.  

“I think that 3-8% range is much smaller than what other people thought the actual impact would be (of dry cooling),” Macknick said. “In past years I had heard talk of 15-20% increases in cost, but that's not what we found.”  

Non-traditional water sources are the other option with the best potential for applicability at low cost, analysts say. Shallow brackish water and recycled wash water collected from mirror washing are the two options furthest along.

Squeegee clean

NREL has been working with researchers at the University of Colorado-Boulder on ways to reduce the amount of water used to wash CSP mirrors. Current practice usually involves spraying the mirrors with a hose, but a high-pressure spraying system with a squeegee involved could reduce washwater by 90% if it's automated, Macknick says.

While washwater reduction is great, in context it's a drop in the bucket compared to the daily water demand for wet cooling. About 20 to 30 gallons of washwater are used for every Mwh generated, compared to 700 to 900 gallons per Mwh used now for wet cooling.

“I think you definitely will see developers try all options, from using recycled wastewater or shallow brackish water, but there will still be issues with using those,” Macknick said. “You may have a performance penalty to pump or treat that water, so I think development will be very site-dependent, and a lot of developers will start trying to locate themselves closer to those sources.”

Non-water liquids could become feasible as a cooling source but would add considerably to capital costs, requiring another stage to cool down that liquid, which couldn't simply be evaporated like water. Some researchers are reportedly testing high-pressured carbon dioxide gas or ionized air.

Smarter siting?

Finding further cooling efficiencies may also drive CSP developers to choose sites more carefully. Basic thermodynamics of a CSP plant are based on the steam cycle, with thermal dynamic efficiency defined by inlet and exhaust temperatures. The ideal CSP locale would have clear, sunny skies but cold ambient temperature.  

“The high desert area could be more attractive, or areas in Colorado that are 6,000 to 7,000 feet in elevation,” Patel said. “But those areas are few and far between, or don't have easy access and grid connection. CSP also likes flat land, and the higher you go the tougher it is to get 200 acres of flat land.”

When compared to water usage at coal-fired energy plants in the American southwest that includes water used during the coal mining, transport and cooling at the plant, CSP's wet-cooling system uses less water. And compared to biofuels based on corn or soybean, CSP wet cooling waters less per acre than those agriculture systems. With those points made, CSP uses more water per MW than a combined-cycle gas-fired plant.

“I think there will be some developments in hybrid-cooled systems,” Macknick said. “There has to be a clever way we can utilize just a small fraction of the amount of water the wet-cool model uses now to achieve similar results. That's where the potential is at.”

To respond to this article, please write to the Editor:

Rikki Stancich: rstancich@csptoday.com