Half a decade ago a scientist, an engineer and a businessman met in a Dublin backyard to conduct an experiment. They heated water in an electric tea maker, then poured it into a bisected segment of pipe. At the bottom of the half-pipe lay a length of wire, and one of the men held a ruler next to it. As the hot water gushed through the pipe, the wire shortened by several centimeters; when they poured cold water over it, it returned to its original length. The trio thought they might be onto something big.

The strange, morphing wire they tested was made of a material called shape memory alloy. Such metals (and some nonmetals) shift into and out of predetermined shapes when subjected to certain temperatures or to pressure or electrical stimuli. Invented 60 years ago, shape memory alloy has been used in fields such as biomedicine and aeronautical engineering. But one of its most stubbornly elusive applications is harvesting energy from hot water. Now those former backyard experimentalists—founders of a company called Exergyn—say they have created an engine that uses morphing wire and hot water left over from industrial processes to generate electricity.

About a third of the energy used by industry in the U.S. is lost as heat, according to some estimates. “A lot of energy is wasted in industrial processes or during heat exchange, when water is used to cool off machinery or power plants,” says Rigoberto Advincula, a professor of macromolecular science at Case Western University and a waste-heat expert who is not connected to Exergyn. The heated water produced as a by-product of industrial applications—including electricity generation—is not quite hot enough to produce steam that can power an engine to run a generator.

Some power plants and industrial users send their hot waste water through secondary engines, which convert a small percentage of energy in that water into electricity using a process called the organic Rankine cycle. This technology, however, requires chemicals to generate power from the heated water. Some of the best chemicals for doing this are dangerous or harmful to the environment, and therefore often banned or restricted. Cleaner chemicals do not draw energy from waste water as efficiently. This adds to the operational costs of running these engines, which means they are not always cost-effective, according to Jonathan Koomey, an Earth systems lecturer at Stanford University's School of Earth, Energy and Environmental Sciences who is not associated with Exergyn.

This is where the morphing wires come in. Shape memory alloys have unique molecular properties—they shift between predetermined shapes depending on their temperatures. That is why this material is the heart of Exergyn's new engine, which uses nitinol, a variation on the original shape memory alloy. “The name represents the fact that it is an alloy of nickel (Ni) and titanium (Ti) and the NOL refers to the [former] Naval Ordnance Laboratory, where nitinol was invented,” says Preston MacDougall, professor of chemistry at Middle Tennessee State University, who is not connected to Exergyn.

At the molecular scale, nitinol is oddly orderly. “Most alloys have no real structure at the molecular level. They are like solutions of metals, as opposed to the orderly arrangement inside something like a salt crystal or a diamond,” MacDougall says. Nitinol's molecules, however, form regular cuboids with 90-degree angles. Under a microscope, MacDougall says, it looks like a bunch of stacked shoe boxes. Heat those molecules and they reorient themselves ever so slightly—the right angles become acute or obtuse—such that the material contracts. Cool the material and the molecules regain their right angles, and the shape memory alloy returns to its original size and shape.

Exergyn's engine exploits this shape-shifting behavior to turn waste-heat water into electricity. Its developers say the engine could be installed in a waste-heat plumbing system, where it would cycle hot water into its cylindrical piston chambers. Each piston is attached to a nitinol wire. “As the hot water comes in, [the wire] contracts a relatively small amount, but does so very powerfully,” says Alan Healy, Exergyn's CEO. Then the engine cycles cold water into the piston chamber. The nitinol expands, and the piston springs out again. On the other side of the piston is a viscous fluid. The moving piston pushes the fluid through a hydraulic transmission that spins a generator, creating electricity. “Using the properties of these materials to generate energy sounds counterintuitive, but it’s not unique,” Advincula says. After all, scientists invented shape memory alloys over half a century ago, and engineers have long recognized these materials had energy-harvesting potential.

In the 1970s a mechanical engineer named Ridgway Banks, working at Lawrence Berkeley National Laboratory, developed a shape memory alloy engine that he believed would save the power industry and other sectors billions in energy costs. Banks patented the engine, but it never revolutionized the energy sector as he hoped. Its technology was complex and required too many chemicals, and the morphing wires wore out too quickly to be practical.

But academics and engineers kept the idea alive in journals and laboratories around the world. In 2010 General Motors partnered with the U.S. Department of Energy’s Advanced Research Projects Agency–Energy to invent an array of waste-heat recovery technologies, including some based on shape memory alloys. So far, however, GM has only used shape memory alloy to replace more complicated mechanical errata on its vehicles, such as a device that makes it easier to close a Corvette's trunk.

Exergyn is not claiming a major technological breakthrough. Healy says the company applied its brainpower to several decades of existing research to create more durable nitinol wires—although exactly how they tweaked the formula is a closely held proprietary information. This advance was needed, according to Healy, because shape memory alloy wears out after a certain amount of morphing. Replacing the wires on a large scale is annoying and costly—major impediments to an engine’s market readiness. “A couple of years ago the best anyone had accomplished was a million cycles,” he says. “We've got over 10 million cycles on our wire.” Healy says he is eyeing the biogas industry for his first customers. “But there’s no shortage of potential applications,” he adds. “Some food companies have waste-heat water from cleaning their machinery or there are data centers using water to cool their servers.”

Still, Exergyn has not yet proved its worth. A company that buys one of Exergyn's engines will want to know the technology will pay for itself after a given period of time, via savings in the company's electricity bill. “One point of reference would be looking at the average cost of electricity for industrial customers,” Koomey says. Exergyn will also compete with the organic Rankine cycle engines that businesses already use to recover energy from waste heat—which Exergyn might have an inherent advantage over, according to Koomey. “Something with lots of moving parts and esoteric fluids [like the organic Rankine cycle engine] will probably be much more expensive than an engine that just uses metal moving back and forth in water,” he notes. “But, the proof is in the device.”

At the moment that proof is still a prototype in Exergyn's Dublin offices. It is capable of pumping out upward of five kilowatts of electricity; 24 hours of that amount would power four U.S. households for a day. The first real field tests will come next year, when Exergyn will devote a serious chunk of the $9 million its raised from investors and grants toward installing 10-kilowatt trial engines at several biogas plants in Dublin. Healy believes the trials will finally show a cost-effective path toward energy efficiency. “Businesses won’t do the right thing,” he says, “unless you give them [an] option that saves them money.”