Faced with water scarcity, the CSP industry looks to dry cooling as a possible solution. Promising it may be, but the stakes remain high.

By Toby Price

When you consider that CSP plants use approximately 800 gallons of water per MWh generated it becomes evident that the industry is sometimes at odds with its sustainability agenda. Historically, CSP plants have used evaporative water-cooling to reject heat produced in the steam cycle, but more recently the industry has been exploring ways to reduce its expensive H₂O dependency.

The cooling process described above accounts for approximately 90 percent of a parabolic trough power plant's water consumption. Consequently, CSP plant owners face huge water bills, especially when precious supplies have to be piped in from distant sources.

Regions where CSP plants are most productive often have relatively little water. Consequently, they often face tight water restrictions and tough environmental regulations. “A lack of water is often the common denominator of CSP plants,” explains Torresol Energy’s CEO, Álvaro Lorente. It is no wonder then that reducing water dependency figures among the CSP industry’s list of priorities.

Air or dry cooling offers a viable alternative to traditional wet cooling and dramatically curbs water consumption because steam-cycle heat is rejected directly to the air without the need for water.

Dry cooling technology is already being adopted by companies in the sector such as BrightSource Energy, which has fitted its Ivanpah Complex in California’s Mojave Desert, a project that will nearly double current solar thermal electricity output in the US. “Our commitment to the environment includes our decision to use air instead of water to cool our plants. By air cooling, our plants reduce water usage by more than 90 percent,” BrightSource Energy announces on its website.

Performance subject to climate

Although the water-saving potential of dry cooling is impressive, it does have its drawbacks. Energy consultancy, WorleyParsons, reports that a dry-cooled plant in the Mojave Desert would provide 5 percent less electricity per annum than a wet-cooled plant.

Location significantly influences the performance of dry-cooling technology. “Cooler climates are more suited for dry cooling. In dry, hot desert areas, the loss of efficiency and capacity are too severe,” explains Babul Patel, Senior Consultant at Nexant Inc. Air cooling at a site in New Mexico for example, would have less impact on electricity costs than in the Mojave Desert because daytime temperatures are considerably lower.

The performance penalty of using dry cooling also varies by technology. A separate study by WorleyParsons revealed that the annual electric output of a trough plant fell by 4.6 percent following the introduction of dry cooling, whereas that of a power tower fell by only 1.3 percent.

The relative costs of using air instead of water also need to be considered. “Air-cooled condensers can be up to four times [the cost of] conventional condensers and wet cooling towers. In addition, air-cooled condensers require higher auxiliary fan power compared to wet cooling towers. This loss of MW will impact on O&M costs,” says Patel.

Bruce Kelly, a technical specialist at Abengoa Solar, also reports that dry cooling imposes a nominal 8 to 9 percent penalty on the levelised energy cost (the cost of an energy-generating system including all costs over its lifetime), although he does emphasise that the expected penalty could be reduced to 7 percent if further improvements to the dry cooling process are made, such as reducing the cost of air-cooled condensers by optimising the tube and fin geometry in conjunction with the design air velocity and fan power demand.

WorleyParsons has also demonstrated that the net present value of an air-cooled CSP plant can be improved by using a larger collector field, which offsets lower steam cycle efficiency, resulting in higher power output during peak summer hours.

Hybrid an option

Another way in which water consumption can be cut at a lower cost than dry cooling is to employ a hybrid wet-dry system. “This approach maintains output near 100 percent even at high ambient temperatures, with annual water consumption about 10 percent that of a wet cooling tower,” says Patel. He does stress, however, that this process involves higher capital costs as it requires both a conventional wet cooling tower and a dry cooling surface condenser.

In a study performed on behalf of the US National Renewable Energy Laboratory, Bruce Kelly calculated that the LEC would increase 2-10% if hybrid cooling is employed instead of water cooling, although he also reported that hybrid systems reduce the energy cost penalty to below that of dry cooling.

Another alternative is to use a Heller indirect air-cooling system, invented in the 1950s by Hungarian thermodynamics professor Laszlo Heller, which is claimed to provide a reduction in LEC compared to a dry-cooled plant. Manufactured among others by GEA Power Cooling, Heller systems are widely used, although they have not been adopted by the CSP sector because costs are somewhat uncertain at present. Nevertheless, the Electric Power Research Institute has kicked off a comparative study of indirect air-cooling (due to be completed in mid 2010), on the theory that it is the most economic dry cooling solution for large-scale thermal applications.

New technologies, such as advanced high-temperature heat transfer fluids, could also improve efficiencies while reducing water consumption, while the use of gas instead of steam turbines could also eliminate the need for cooling water.

Dry cooling clearly penalises a CSP plant’s bottom line, but with water prices on an upward trajectory the cost of energy from a plant with dry cooling may soon equal that of a wet-cooled plant, making the technology viable in the near future.

In this case, should we deduce that dry cooling will become the industry norm in coming years? As Lorente points out: “It is difficult to predict whether these systems will become the industry standard in the future, especially when the current trend is towards major CSP complexes (of several hundred MW), which could justify the construction of infrastructures to transport water from further a field”.