The lifespan of the uranium dioxide that fuels the country’s commercial light-water nuclear reactors is less than five years, so refueling occurs frequently. And, every 18 months, a typical LWR produces 30 tons of solid waste that must be buried forever. The waste volume is low enough that most plants are able to store it on-site. The nuclear industry is looking for better solutions.
Nuclear fuel in a next-generation reactor design called the Traveling-Wave Reactor, or TWR®, could last for decades and produce 80% less radioactive waste, according to TerraPower® LLC, the Bill Gates-founded company behind the TWR reactor. Instead of water, liquid sodium will cool the core, enabling higher operating temperatures and, therefore, greater efficiency.
The tandem of unparalleled fuel longevity and minimal waste is possible because of the reactor’s “breed-and-burn” design and its distinctive core configuration. While most reactors use uranium-235 as their power source, the TWR reactor’s primary fuel would be depleted uranium, the waste byproduct of enriched uranium production consisting largely of uranium-238. U-238 also happens to be the most common uranium isotope found in nature. But the isotope is not fissile: It cannot easily sustain a chain reaction. To fission, it needs a “jump start,” and in a TWR, that would come from high-energy neutrons elicited by a small amount of uranium-235.
Two parallel “fission waves” then move slowly through the core. The first wave transmutes the U-238 into highly fissile plutonium-239 (the “breed”). The second wave consumes the plutonium and produces heat (the “burn”). Under the right conditions, the reaction could be sustained for 40 years, perhaps longer, without refueling.
“On paper, the TWR reactor works wonderfully,” said Wade Marcum, Henry W. and Janice J. Schuette Professor in Nuclear Science and Engineering at Oregon State University. “But then comes the engineering. How does TerraPower turn this idea into reality?”
In 2015, the company asked Marcum’s research group to solve a particularly vexing problem: Measure the displacement under load — the stress — sustained by TWR fuel pins caused by flow-induced vibration. FIV arises wherever a fluid moves past a flexible structure. Marcum likens it to a flag waving in the wind that gradually wears out from interacting with itself and the air. To some extent, vibrations occur in every reactor core as the coolant flows up and around the long, cylindrical fuel pins. Where the fuel pins contact adjoining structures, the surfaces chafe against one another, which leads to fretting wear, material fatigue, and, potentially, failure of the cladding that encases the pins.
“It’s a subtle phenomenon that starts as microscopic wear points that become exacerbated over a long period of time,” Marcum said. “Eventually, the cladding may breach and expose the nuclear fuel.”
According to Mathieu Martin, a senior thermal hydraulic engineer with TerraPower, FIV is among the most common causes of component failure in the operating reactor fleet. Undetected and unchecked, the harm it causes can require the replacement of damaged fuel, which increases plant downtime and operational costs.
Typically, new fuel pins are subjected to the same coolant flow rates found in a nuclear core for 1,000 consecutive hours, then examined for damage before being put into service.
“There’s a level of credibility in extrapolating those results out four or five years, but it doesn’t really work when we need the pins to last 40 years,” Marcum said.
“We needed to develop an entirely new measurement technique,” Martin added.
With guidance from TerraPower engineers, Marcum’s team set out to develop instrumentation that could detect and monitor fuel pin wear and tear under conditions similar to those of a TWR core. Whatever they came up with had to be small and unobtrusive: Pins in the TWR design are organized in compact hexagonal arrays, and each pin is wrapped helically from top to bottom with a wire spacer. The configuration prevents the pins themselves from touching one another and encourages coolant flow. It’s a tight squeeze with scant leftover space.
Their technique entailed wrapping a single fiberoptic sensor along the entire length of the cylinder. Differences in the backscatter of laser light indicate the level of strain every few millimeters. After two years of focused work, the researchers tested the device, which successfully measured the structural response of a slender cylinder during hydraulic flow.
Next, the team built a small-scale test loop that replicated flow conditions in the core. The device housed a hexagonal array containing 19 stainless steel fuel pins, each 1.25 meters long. (Full-size pins measure 4.7 meters long, and full-size arrays can contain nearly 300 pins.) Tungsten pellets stood in for fuel slugs, and flow tests were conducted with water instead of liquid sodium.
To accommodate the fiber-optic sensors, each pin was engraved with a 200-micron-deep channel running parallel to its wire spacer. “We instrumented every single one of those pins, and we exposed the assembly to all relevant conditions that would ever be experienced in the reactor — and even slightly beyond that,” Marcum said. “We were able to measure, at any instant in time, the intensity of flow-induced vibration, and especially what happens at the physical interactions where the wire spacers contact adjacent pins. It worked. Then we broke out the Champagne.”
They weren’t done. In the final phase of the project, Marcum’s group tested the instrumentation in a full-length test loop containing 61 full-size pins. The results, obtained last September, are awaiting publication, but Martin and Marcum characterized the experiment as a complete success.
“Now we have a way to measure flow-induced vibration and isolate areas of higher risk that are more prone to damage,” Martin said. “The information is quite important for TerraPower to move forward with the Traveling Wave Reactor.”
Questions: editor@engr.oregonstate.edu