Department of Energy (DOE) Secretary Rick Perry recently announced the launch of the Versatile Test Reactor (VTR) project, flagging it as one of the department’s top priorities. The project, which would be the first new DOE test reactor in decades, would differ from the DOE’s operating test reactors because it would be cooled by liquid sodium instead of water, enabling it to produce large numbers of “fast” neutrons. The DOE says that such a facility is needed to develop new reactors that use fast neutrons to generate electricity. US nuclear plants today are light-water reactors, which use slow (“thermal”) neutrons.
The Union of Concerned Scientists (UCS) questions the need for a dedicated fast neutron test reactor and, more generally, has serious concerns about fast reactor safety and security, detailed in a critique it released last year. Fast reactors pose nuclear proliferation and terrorism risks in part because they commonly use fuels containing plutonium, a nuclear weapon-usable material. Most fast reactor concepts also involve reprocessing of their spent fuel, which separates plutonium in a form that is vulnerable to theft.
Perry reported that the DOE has determined the “mission need” for the reactor, the first milestone for any new large department project, but the mission-need statement fails to make the case that the VTR is needed for fast-reactor development. Regardless, the Nuclear Energy Innovation Capabilities Act, signed into law in September 2018, directed the department to build the VTR by the end of 2025. For its part, the department does not believe it can be completed before the end of 2026.
A Hefty Price Tag
Missing from Secretary Perry’s February 28 announcement was any estimate of the cost. In response to a Freedom of Information Act (FOIA) request, UCS has learned that a “rough order-of-magnitude” estimate for the VTR’s construction and startup is $3.9 billion to $6.0 billion. To build the reactor over the next seven years would require the DOE to spend, on average, $550 million to $850 million annually, which is comparable to the department’s total fiscal year 2019 budget for nuclear technology development of approximately $740 million. The DOE has requested $100 million for the project (which it now refers to as the Versatile Advanced Test Reactor) in fiscal year 2020.
Cheaper, Faster Alternatives
The FOIA documents also reveal that the DOE’s determination of mission need misquotes its own 2017 user needs assessment to justify the new test reactor. In fact, there are ways to simulate the range of neutron speeds typical of a fast reactor in an already existing test reactor, such as the Advanced Test Reactor at Idaho National Laboratory or the High Flux Isotope Reactor at Oak Ridge National Laboratory. This could be accomplished by using neutron filters and possibly a different type of fuel. Going that route would be significantly cheaper: A 2009 DOE assessment suggests that this approach could achieve the minimum requirements necessary and would cost some $100 million to develop (in 2019 dollars), considerably less than the VTR project’s projected price tag.
Equally important, using one of the two currently operating test reactors could likely provide developers with fast neutrons more quickly than the VTR project. The proposed test reactor would not be operational before the end of 2026, according to DOE’s proposed schedule, which it describes as “aggressive.” A recent DOE study estimated that it would take about 10 to 13 years for such a reactor to begin operation. Moreover, after the VTR startup, it would need to operate for some time—perhaps a few years—before it could be reliably used for testing, assuming there will be at least a few unforeseen problems. Thus it could be well over a decade before the VTR would become available. In contrast, the DOE estimated it would take seven years for the alternative system to become available at an operating test reactor.
The VTR mission-need statement also exaggerates the technical capabilities needed by the reactor developers who would use a fast test reactor. One of the main objectives of a test reactor is to bombard fuels and other materials with neutrons to study how they withstand radiation damage. This damage can be measured by a unit called “displacements per atom”—that is, the number of times each atom in a material sample is affected by a neutron. The more displacements per atom that a test reactor can provide per year, the faster a given test can be completed. The mission-need statement claims that reactor developers need a facility that can achieve at least 30 displacements per atom per year, although the reference it cites, a 2017 user needs assessment, only calls for a minimum of 20.
This difference is significant because researchers have shown that using the cheaper filtering approach in the High Flux Isotope Reactor at Oak Ridge National Laboratory could provide about 20 displacements per atom annually, so that the higher rate provided by a new reactor would not be needed. The mission need statement did not assess whether the modest additional capability that the VTR could provide—that is, 30 instead of 20—would be worth the substantial additional cost. If a reactor developer wanted to test a fuel sample up to 200 displacements, it would take 6.7 years in the VTR at a rate of 30 per year, compared to 10 years at the High Flux Isotope Reactor at a rate of 20 per year. But given that it might take more than a decade for the VTR to become available, it is far from clear it would help developers to achieve their goal any sooner than the alternative approach, which could be available in seven years.
A Better Test Reactor
Perhaps the clearest statement casting doubt on the need for the VTR was made by Westinghouse, one of the potential commercial users of fast reactor technology that the DOE surveyed in its user needs assessment. The originator of the pressurized light-water reactor (LWR), Westinghouse is now interested in developing a liquid lead-cooled fast reactor. However, in its response to the DOE survey, it states that any new test reactor should include capabilities for light-water reactor testing because “LWR technology will continue to be the backbone of nuclear energy for decades to come.”
UCS agrees with that statement. If the United States needs a new test reactor—and it may soon, given existing test reactors are many decades old—it would make more sense to build a new thermal neutron test reactor with the capability of generating fast neutrons if necessary, not the other way around. The technology is well-established, and the commercial need for a source of thermal neutrons is far more likely than for fast ones.