Building a Better Power Plant
A group of academics is testing a radical new way of cooling thermoelectric plants that could reduce the power grid’s dependence on freshwater.

_Young Cho

Cho is a professor in the Mechanical Engineering and Mechanics Department of the College of Engineering.

_Matthew Mccarthy

McCarthy is an assistant professor in the Mechanical Engineering and Mechanics Department of the College of Engineering.

_Ying Sun

Sun is an associate professor in the Mechanical Engineering and Mechanics Department of the College of Engineering.

As searing heat swept across the United States in the summer of 2012, it wasn’t just the air temperature that shot up. Water in rivers and lakes, too, was unusually warm.

This was a problem for some thermoelectric power plants, almost all of which rely on water from lakes or rivers for cooling.

In August of that year, Dominion Resources’ Millstone Nuclear Power Station in Connecticut shut down one reactor for 11 days because seawater temperatures in Long Island Sound, from which the station draws water, were the highest since operations began and exceeded the technical specifications of the reactor. The shutdown reportedly resulted in lost energy output worth several million dollars.

A month earlier, four coal-fired power plants and four nuclear power plants in Illinois requested permission from the state environmental protection agency to exceed their permitted water temperature discharge levels, which are regulated to prevent ecological damage. At one plant, Braidwood in northeastern Illinois, waters in a 2,500-acre cooling lake reached 102 degrees.

This was hardly the first time heat or drought put power plants’ water demands at odds with marine life and the nation’s power needs, and it won’t be the last. As increased development — particularly in the west and southwest —gobbles up precious water resources and warming trends raise temperatures above historical norms, the country’s reliance on water to generate power is increasingly risky.

One solution is dry cooling technology, which uses air instead of water to cool steam condensers inside power plants. No water is wasted, and no fish are threatened.

Although relatively common overseas, dry cooling is used in less than 1 percent of U.S. thermoelectric plants. That’s because dry-cooled power generation is 10 percent less efficient on hot days and is up to five times more expensive to build than more common designs, explains Jessica Shi, senior technical leader/manager, technology innovation research for water conservation at the Electric Power Research Institute (EPRI), an energy industry research outfit.

But what if there were a better design?

A dry run

That was the question posed by the National Science Foundation and EPRI in 2013, when they teamed up to offer a grand challenge to inspire researchers to propose new, transformative cooling concepts.

In Drexel’s College of Engineering, three mechanical engineering and mechanics professors were intrigued by the challenge. Associate Professor Ying Sun, Assistant Professor Matthew McCarthy and Professor Young Cho chalked up some ideas.

“We weren’t allowed to use water, the design had to be at least three times more efficient at heat transfer, and it had to be cost effective,” says Sun, the project’s principal investigator.

After a month, they had about 10 ideas. Most had some fundamental flaw — the systems would have to be too big, or too unreliable, or too expensive.

But one seemed promising, and it was fairly simple from an engineering standpoint. Instead of water, use paraffin. Instead of large metal condensers to cool the paraffin, spray it into the air.

Paraffin is a colorless, soft wax derived from petroleum. It’s similar to candle wax but has a lower melting temperature, which makes it ideal as a phase-change material (a material capable of storing and releasing heat as it changes form, as water does so perfectly) because it easily converts from liquid to solid, and back again.

It’s also nontoxic, readily available and cheap.

“Cost is really important because water is 1 cent per ton — it’s basically free apart from the cost of pumping it — and alternative heat transfer materials are very expensive,” says Sun. “Paraffin is about $1,000 a ton, so it’s affordable, and it’s a classic, stable, proven material.”

The team — which also includes Assistant Research Professor Philipp Boettcher from mechanical engineering as well as a post-doctoral researcher and graduate and undergraduate students — then tackled the problem of how to design a condenser that would convert heated paraffin back to solid form. There were obvious disadvantages to adopting the fin-tube system common in the industry.

“Cost is really important because water is 1 cent per ton — it’s basically free apart from the cost of pumping it.”

“Existing dry cooling technology uses metal fins to convert steam back into water — it’s like the radiator in your car, it has a lot of fins to create a large surface area. However, the effectiveness of the heat transfer of the fins is very low,” says Sun. “It’s also expensive. The dry cooling condenser for a typical 500 megawatt plant costs $100 million because you have to use many ribbons of sheet metal in the fins.”

Instead, they looked to the cooling tower concept already used in some plants, which sprays heated water from the condenser into the tops of open towers. Whatever doesn’t evaporate collects at the bottom as cooled water ready to be discharged. Unlike water, which partially evaporates when it’s sprayed into the air, the paraffin will solidify and can then be collected and used again for another cycle.

“We’re taking the primary steam coming off of the power plant and we’re running a series of tubes through a big bath containing a liquid version of this wax, then we’re taking this wax and spraying it up in the air,” McCarthy explains. “We have these little tiny droplets of wax that have very high surface area. You have millions of droplets. When you take a certain volume of something and chop it up into a million small units, you create a huge amount of surface area.”

The droplet approach improves the effectiveness of heat transfer by a factor of four, well within the range required of the grand challenge, says Sun.

For inspiration, the team looked to simple ideas already in existence. To understand how wax behaves as a spray, for instance, they watched YouTube videos of apples being treated with protective fruit wax coatings before heading to supermarkets. And on a summer vacation to the amusement park Dutch Wonderland in Lancaster, Pennsylvania, Sun bought her young son a Dippin’ Dots Ice Cream treat — a snack created by flash-freezing ice cream droplets.

Their wax-and-spray-cool concept won over the National Science Foundation and EPRI, which jointly awarded the professors and their team a $1 million grant last year to continue the research.

Their next step was to evaluate how well their concept would work in a real system. They used some of the money from their initial grant to work collaboratively with Advanced Cooling Technologies, a thermal management engineering company in Lancaster, Pennsylvania, to construct a 5-kilowatt prototype (a plant is usually at least 500 megawatts) and filed a patent on the idea.

They’re now conducting tests to better pinpoint how energy-efficient the design could be.

“To design the final system, we need to more accurately understand the solidifying and melting processes that these waxes are experiencing,” says McCarthy. “Liquid wax droplets are solidifying as we blow cold air over them, and solid wax particles are melting in a slurry moving past the hot steam tubes. These are the two phase-change processes that we need to characterize.”

The team has built a series of experiments for the prototype that will allow them to measure the heat transfer coefficient as solid wax particles fall into liquid wax and float through a series of heated tubes that simulate the steam in a typical thermoelectric system. The experiments will tell the team how long it takes the particles to flow through the system and at what rate they’re melting.

“We’ll come up with correlations to predict their performance, and we can use those models to design a full system,” says McCarthy. Meanwhile, a consulting company, Worley Parsons in Reading, Pennsylvania, is conducting economic feasibility studies on their design.

A group of academics are testing a radical new way of cooling thermoelectric plants that could reduce the power grid’s dependence on freshwater.


By 2025, lack of water availability is projected to constrain thermoelectric power plants in many areas of the country.

Getting Real

Skepticism about the team’s system, which in a perfect world wouldn’t even be able to be implemented on a full scale for at least five years, abounds among experts, McCarthy admits.

“There are plenty of people in the power plant industry who think this is a pretty radical idea,” he says. “But academics kind of blur the line between science and science fiction. We come up with crazy ideas and have the ability to think outside the box because we’re not beholden to any existing technology. We don’t have anything stopping us.”

For some members of the team, this is the first time they’ve designed a project that had to make sense in the real world.

“I learned to consider many practical problems,” recalls Han Hu, a PhD candidate and the student leader on the team. “Is the design cost-effective? How long will the structure and material sustain before fatigue? Is the design safe to run 24-7-365?”

In May, Sun and her team received word that they had been awarded a larger, $3 million grant to manufacture a second, more evolved prototype.

This new grant comes from ARPA-E, a government agency formed in 2009 to fund out-of-the-box energy technology innovations.

This time, the team’s previous grant funder is their partner on the project — a validation of sorts. The new proposal uses the same fundamental concept of moving phase-changing material from the steam side to the air side, but with a different circulation technique. Rather than pumping and spraying the phase-change material, it encapsulates it in a highly porous mesh-type structure.

The new concept addresses the potential for contamination that was inherent in the original prototype.

“You have these particles being sprayed around, and if they get dirty that can clog your system and that can hurt performance over time,” McCarthy says. “This new idea has none of those concerns, but it has some other concerns. We’ve shifted the drawbacks from contamination to performance. The new idea we’re proposing is arguably more reliable, but the performance might be a little bit less and the cost might be a little bit more.”

If a practical design emerges from the academics’ blackboard, it will be a step toward protecting the country’s energy security. In 2009, a U.S. Department of Energy report imagined a scenario in which at least 25 percent of the nation’s thermoelectric capacity used dry cooling technology and the rest was converted to the wet recirculating type. If such a situation became reality, it would reduce the industry’s need for freshwater by more than 25 billion gallons a year — one of the better scenarios modeled by the DoE.

“Our goal is usually to publish a sexy, high-impact paper to our community, but this is quite different,” Sun says. “This is real-life challenges. You are using some of your fundamental knowledge and turning it into practice with the hope of changing the world with groundbreaking technology.”

The Great Thermoelectric Thirst

Each day, the nation’s thermoelectric power plants withdraw an estimated 161 billions of gallons of water for cooling — enough to supply 160 cities the size of New York with drinking water. Thermoelectric power plants are responsible for 41 percent of all freshwater withdrawals in the United States in a given year, more than any other sector, and supply 90 percent of America’s energy. Much of the water they withdraw is recirculated, rather than consumed; nonetheless, the overwhelming majority of plants are based on designs that require a steady, abundant source of cool water.

Thermoelectric power plants create electricity by converting steam heat to mechanical energy. They boil water to create steam, which then spins turbines to generate electricity, according to an explanation from the Union of Concerned Scientists.

The heat used to boil water can come from burning a fuel, from nuclear reactions, or directly from the sun or geothermal heat sources underground.

Once steam has passed through a turbine, it is condensed back into a liquid so it can be reused to produce more electricity. Typical power plants condense this steam by cooling it using water drawn from a nearby lake or river.
There are three main types of cooling systems used by thermoelectric power plants.

Once-through withdraws cool water from a nearby source (a lake, river, ocean, etc.) and circulates it through condensers that are exposed to the steam leaving the turbines. The now warmer water is discharged back to the source. This is an older, simple design that works well in areas that have abundant supplies of cool water. However, when discharge water exceeds temperature limits, fish life is threatened.

Since the 1970s, more power plants have been built or converted to recirculating or closed-loop cooling systems. These systems reuse the discharge water in a continuous cycle rather than return it to the source. In order to cool the water down so that it can be used again, most plants use cooling towers that spray water through the air; others use outside cooling ponds. These systems withdraw less water overall than once-through systems, however, more water is lost to the environment due to evaporation.

Dry-cooling systems require no water; instead they use lots of metal to create a large surface area to cool the steam leaving the turbine using air. Though they conserve water, they cost more to build and are less efficient than other methods. They account for only about 1 percent of U.S. power plants.