Almost every form of energy conversion creates heat, making it one of the most prevalent forms of energy. When used for mechanical work, thermal energy is most efficient when it can be moved, stored, and converted at its highest possible temperature. But most of today’s pumps and compressors are made from superalloys and ceramics and can’t handle that extreme heat. A team from Georgia Tech has developed a ceramic pump they and others expect to spur a new generation of highly efficient, low-cost systems for storing, transporting, and converting surplus thermal energy produced by renewables like solar and wind.
For two long days and nights, Professor Asegun Henry’s team of researchers at Georgia Tech’s Woodruff School of Mechanical Engineering anxiously watched their newly developed ceramic pump move molten tin heated to record-high temperatures.
At any moment... they expected something to go wrong. The extreme heat could shatter a shaft, break a bearing, melt a seal, as it did during countless tests conducted throughout the pump’s four-year development cycle. The students were forced to monitor the machine in case it started leaking dangerously hot metal. They didn’t want to burn down the lab.
Henry and his students hoped this test would be different. Over the years, the team developed a number of innovations and improvements to perfect a pump they and others expect to spur a new generation of highly efficient, low-cost systems for storing, transporting, and converting surplus thermal energy produced by renewables like solar and wind. It could serve as a new, cheaper type of battery that could generate electricity when needed, a big challenge in the advancement of renewable energy. The nuclear and chemical processing industries, among others, could also use the system for advanced applications.
“It’s a big undertaking because you have to change the whole infrastructure of the system,” Henry said.
To handle those needs, the pump had to withstand temperatures that could exceed 1,326 °C, making it the first of its kind. As testing moved into its third day, the new pump held. Based on past failures, though, the team expected the worst.
Almost every form of energy conversion creates heat, making it one of the most prevalent forms of energy. When used for mechanical work, thermal energy is most efficient when it can be moved, stored, and converted at its highest possible temperature. But most of today’s pumps and compressors are made from superalloys and ceramics and can’t handle that extreme heat. The components break, corrode, weaken, or wear when exposed to high-temperature liquids for an extended period. Instead, they must be stored in a cooler part of the system, limiting the system’s overall efficiency.
The need for a stable heat storage and transfer system is especially important for advanced and emerging nuclear energy applications. The system, for example, could serve as a less expensive, more efficient, and faster way to cool down a reactor in case of an emergency. Finding that solution, Henry said, has been a “grand challenge” in the nuclear energy industry and one of the main reasons the Advanced Research Projects Agency-Energy (ARPA-E) is funding his work.
“The scary thing with nuclear is the runaway effect, where the reactor heats up faster than you can cool it,” Henry said. “This type of pump can move heat faster, and the liquid will stay warmer before it boils. That buys you a whole lot of safety.”
To create the pump, Henry had to first choose the material for its components and the heat transfer media that would flow through it.
For the media, Henry knew he had to use a liquid. Liquids are easy to move and control without expending a lot of energy, making them the most efficient and effective way to transfer heat. “The higher the temperature, the more useful the liquid is,” Henry said. “The problem is, what liquid to use.”
Fluids commonly used in heat transfer systems include salt, glass, and oil. But those materials have too many disadvantages—heat limitations, high viscosity, corrosiveness, low thermal conductivity and density—for the types of extreme-heat applications Henry had in mind. Molten metal, however, especially tin, performs much better in all of those areas.
“It is low viscosity; as soon as it melts, it flows like water. And it’s electrically conductive. That’s huge because that capability doesn’t change and the electrons can move freely; the liquid connectivity is almost the same as a solid connective,” Henry said. “With metal you end up in a different regime of heat transfer. Heat transfer in liquid metal is extremely effective. You get a whole lot more heat transfer given the energy needed to move it.”
Molten metals, however, can eat away at metal pipes and other components. “It’s like trying to make piping for sugar water with pipes made from sugar,” Henry said.
That last piece helped Henry decide to build all of the components of the pump—a gear pump, in this case, for simplicity and reliability—from ceramic. Ceramics, though, are brittle. The team feared ceramic gears and other components used to pump and circulate extremely hot metal would crack and fail. But new breeds of ceramics can retain their mechanical stiffness at temperatures over 1,000 °C. After some research, Henry chose Shapal Hi-M Soft, introduced in 2012 and known for its increased ma-chinability, mechanical strength, and thermal conductivity.
Despite those attributes, extreme heat caused one version of the pump to expand and vertically misalign by 1 mm. As a solution, the team intentionally misaligned the pump by connecting a vertically angled, flexible tungsten sleeve on the insulating shaft connecting the pump to the motor.
Another major design problem remained. Most pump seals are made from polymers, which vaporize under extreme heat. After about three years of trials and testing, the team finally settled on pure graphite to seal the pump, pipes, and valves. “They’re real squishy and are good up to about 3,000 °C,” Henry said.
But graphite oxidizes. To prevent that, the team built a simple vacuum chamber around the pump and filled it with a “cover gas” of nitrogen. It also allowed the team to use tungsten on the outside of the seals to strengthen areas where tensile forces were applied. The chamber doesn’t greatly impact the cost of the system; it can be made from thin metal and double as a protective covering.
“When you take oxidation off the table, it opens up a range of materials you can use,” Henry said.
Not What They Expected
By the morning of the third day of testing, Henry’s students were begging to go home. Henry made them wait. He had to be sure that this version solved past problems. After 72 hours running at an average of 1,200 °C, Henry pulled the plug on the pump and declared the new design a success.
“It’s like trying to make piping for sugar water with pipes made from sugar.” –Asegun Henry, Georgia Tech
“I said shut it down because it wasn’t going to break,” he said. “The students had given up before, so everyone was excited. It was a big triumph.” Even though the pump worked, Henry is still working on future iterations. The teeth of the drive gear wore slightly, a problem Henry could solve by replacing the Shapal with a harder material such as the more expensive silicon carbide. He’s now looking for a cost-effective source. Henry will also try replacing the gear pump with a centrifugal pump to increase its speed. Then the new rounds of testing will begin. He expects to have a commercial version ready within two to three years.
“When you’re doing engineering and technical development there are different types of barriers to success,” he said. “But we had tested enough aspects to know that it would work. We knew it could be done. You can’t let the challenge or the effort be a barrier to trying again and again. If you have good ideas, you shouldn’t give up.”