Building mobile nuclear power plants will be a challenge, but successfully meeting the challenges could alter the future trajectory of the energy and fuels supply industry.
That is one of the largest and most consequential sectors of our modern, mobile, industrialized economy. There are no guarantees, but compared to many research and development projects, Project Dilithium has both a reasonable prospect of success and an almost unimaginable potential payoff.
Whoever picked Project Dilithium as the name chose a brilliant metaphor for anyone who views Gene Roddenberry’s Star Trek franchise as an inspiration for dreaming about the future. Vessels like the Starship Enterprise rarely worry about fuel; all they need is an infrequent addition of some “dilithium crystals.”
Unsurprisingly, there are skeptics and naysayers who aren’t excited about the prospect of mobile nuclear power plants. The Union of Concerned Scientists’s Dr. Ed Lyman, acting director of UCS’s Nuclear Safety Project , is pessimistic enough about the prospects for success that he has declared that the project “won’t work.” (See Bulletin of Atomic Scientists Feb 22, 2019. “The Pentagon wants to boldly go where no nuclear reactor has gone before. It won’t work.” )
The National Interest chose This Might Be the Military’s Worst Idea Ever as the headline an article by Michael Peck that didn’t match the superlative at all. Instead, it laid out some good reasons why mobile nuclear power plants might be a pretty good idea as long as certain questions can be effectively addressed.
Several other publications produced articles with headlines claiming that “experts” were horrified, critical, or skeptical, but a fair number of those pointed to Ed Lyman as the “expert” and his article in the Bulletin of Atomic Scientists as the place where he documented his concerns.
From my perspective, Dr. Lyman errs by dismissing some of the technical advances that have improved the chances for successful accomplishment of the project’s stated objectives. He has an exaggerated opinion on the risks. He underestimates the potential value of successfully developing the capability to build systems that can meet the stated objectives.
The Army is only a lead customer with the skills and resources required; there are countless potential customers whose interest will be apparent once there is an actual product available for purchase.
A product that comes close to meeting all of the Army’s requirements for Project Dilithium might be especially appealing for settlements in Alaska or The North of Canada. Senator Lisa Murkowski has often expressed the interests her constituents have in small nuclear power systems. They want machines that can operate cleanly for years on small quantities of fuel that eliminate the complex and expensive transportation of diesel fuel.
That problem is essentially the same one that is such a recognized headache for Army planners. Except it’s one that never stops under current available product constraints.
Some proposed options don’t have much of a chance.
I must admit that I think Dr. Lyman is correct in his evaluation of one particular contender in the field. In fact, his opinion of the Los Alamos Special Purpose Reactor (aka Megapower) is a bit less critical than mine. He describes INL’s evaluation of the design as pointing out “several major safety concerns, including vulnerabilities to seismic and flooding events.”
I think he overlooked the part where the evaluators gently told the designers that their beautifully-optimized, computer-designed model cannot be manufactured.
It relies on unobtainable tolerances in drilling ~ 2,000 channels, each 1.5 m long, in a stainless steel monolithic cylinder with 1 mm thick webbing between the channels. Additionally, it envisions the ability to seal about half of those channels at each end to retain fission product gases.
I might be a bit of an outlier, but I believe system designers should make sure their ideas are functionally possible within manufacturing capabilities before they spend too much time determining if they can survive extreme seismic or flooding events.
It turns out that the beautifully optimized but impossible to manufacture design shares one of the perennial problems that have plagued nuclear technology development since the 1940s. It is a substantially scaled up version of a system that has been proven at a smaller scale using highly enriched fissile material. LANL developed Kilopower for a NASA mission to Mars. It worked fine in tests, why not immediately make it 10 times bigger and more powerful?
The prototype power system uses a solid, cast uranium-235 reactor core, about the size of a paper towel roll. Passive sodium heat pipes transfer reactor heat to high-efficiency Stirling engines, which convert the heat to electricity.Demonstration proves nuclear fission system can provide space exploration power
Therefore, this article will continue on the assumption that the only real contenders for near term success in meeting Project Dilithium’s specifications will use a variant of a high temperature gas reactor and a Brayton Cycle heat engine.
Army tested a mobile nuclear power plant in early 1960s. It was a direct Brayton Cycle machine
Here is a quote from an October 2018 study conducted for the Army’s G-4.
ML-1 was a true mobile power plant. Its main advantage was the ability to substitute a single nuclear fuel load to displace and eliminate the need to transport the equivalent of 400,000 gallons of liquid fuel. Unlike the other Army reactors, ML-1 did not use water for coolant, substituting a sealed reactor design with pressurized gas (nitrogen) to drive a closed cycle gas turbine. This design made possible a significant reduction in both size and weight, enabling it to be truck-mobile. The reactor could fit in a standard International Organization for Standardization (ISO) container for ease of shipment by standard military transportation systems.Study on the Use of Mobile Nuclear Power Plants for Ground Operations, Army Deputy Chief of Staff (DCS) G-4 p. 2.2
Though ML-1 worked, it didn’t work well. There were a number of design flaws and unreliable components that led to project schedule delays. By 1966, the Army had lost almost all of its R&D funding as the Johnson Administration attempted to support both an expanding conflict in Vietnam and growing social programs. Programs that weren’t at the top of the performance list were cut.
The Atomic Energy Commission wasn’t interested in pursuing suitable reactor design improvements without a strong demand signal from the military service.
Though it has been almost five decades since the DOD funded a program of for designing and deploying mobile nuclear systems, the supporting technology base has made substantial advances.
Initial concept of using gas to move heat from fission reactors to heat engines
Even if ML-1 had achieved its full design potential without technical issues or delays, it still would have been an inefficient machine. The reactor used in the system was capable of minimally adequate temperatures for a Brayton Cycle; 3 MW of heat would only produce about 300 kW of electrical power.
Brayton Cycle turbo machines were also in their infancy in the early 1960s. The theory was well-established, but functional machines had only been available for a couple of decades. Many of the most common applications for the machines were still in high performance aircraft where the engines were only expected to run for a few hundred hours before being replaced. Brayton Cycle machines had barely begun to penetrate land based power generation, or maritime propulsion.
The fundamental idea of matching a high temperature gas cooled reactor with a simple Brayton Cycle heat engine originated during the final years of the Manhattan Project. Farrington Daniels, Director of the Metalurgical Laboratory in Chicago, had pre-war high temperature industrial experience in developing processes to fix nitrogen from the atmosphere. Once the Met Lab had essentially completed its assignment for the Project, Daniels and his scientific colleagues began meeting informally to discuss ways to put fission energy to productive use.
Daniels suggested a “power pile” composed of uranium spheres mixed with beryllium oxide moderator spheres and cooled by helium gas. The Manhattan District assigned Monsanto to lead the project and provided initial funding in December 1945.
In April, 1946, GE, Westinghouse, Allis-Chamers, the US Navy, the US Army Air Corps and the Clinton Laboratories joined the Manhattan District and Monsanto in the project. They supplied engineers and scientists on an in-kind basis. Everyone on the project knew that it was a pilot system that should be constructed quickly so that it could help develop knowledge of material performance at high temperatures.
The Daniels Pile project was defunded almost as soon as the civilian AEC took over responsibilities from the Manhattan District of the Army Corps of Engineers. Before that happened, Daniels and his team of Project scientists and industrial partners had done enough design work and testing to convince themselves that a gas cooled, beryllium-moderated reactor could operate at high enough temperatures to provide useful information for subsequent design efforts.
Daniels spent at least five more years trying to find funding for his idea, but the AEC monopolized all nuclear work. The appointed commissioners at the top of the organization had determined that America had no need for any new power sources. Instead, the commissioners and their political overseers had determined that the primary mission of the U.S. Atomic Energy Commission was to develop weapons and other military applications.
Aside: The tight linkage in public perception between nuclear energy and nuclear weapons started in the earliest days. It’s quite possible that some of the participants in decisions like naming the AEC wanted to keep scary bombs linked to useful power applications. End Aside.
Coated particles invented as a way to retain fission products at high temperatures
After several years of unsuccessful efforts to convince various government agencies to pursue atomic power using high temperature gas cooled reactors, Daniels was recruited by the Rockefeller Foundation to create a solar energy research program.
However, his pebble bed reactor concept simmered and was eventually taken up by Rudolf Schulten in Germany. He devised a way to coat actinide fuel particles with fission product-retaining coatings. He also devised a method of assembling thousands of particles along with graphite moderator into 6 cm diameter spheres that could be “piled” into a cylinder to form a critical mass. The spherical shape of the pebbles results in at least 40% of the cylindrical volume being empty and available for a turbulent flow of cooling gas.
The idea of coating actinide fuel sources with refractory style graphite and silicon carbide coatings was nurtured by a minor, but passionate multinational group of technologists for the next 60 years. Another major fuel form, a prismatic graphite block with drilled coolant channels and drilled voids to contain compacts of particle fuel was developed. Several operating reactors including Dragon, Peach Bottom 1, AVR, Ft. St. Vrain, THTR, HTTR, and HTR-10 have provided an increasingly extensive database of fuel and component performance.
A variety of coatings and coating processes have been tested. Fuel kernels of various compositions including actinide (thorium, uranium and plutonium) oxides, carbides and a mixture of the two have been irradiated and examined. A comprehensive summary of the various testing development programs written by P. A. Demkowicz, B. Liu and J. D. Hunn was recently published as Coated particle fuel: Historical perspectives and current progress.
Since 2003, the US Department of Energy has been engaged in a carefully planned series of irradiations and post irradiation examinations for a high potential system that uses a series of four coatings on a kernel of uranium oxy carbide. That program is nearing its completion with final testing aimed at identifying potential issues that might result from exposure of the high temperature fuels to excessive moisture or chemical contaminants.
For certain postulated configurations, including those that can meet the Project Dilithium requirements, there is no need to wait for the results of those final tests. These systems do not have any pathways that allow water or other contaminants to get into the system. Even the ultimate heat sink is dry atmospheric air.
Contrary to the description Dr. Lyman offered in his dismissive article about the potential for mobile nuclear generators, results from DOE’s Triso fuel development and testing program have been anything but “inconsistent.”
The program has produced consistently impressive data on tens of thousands of irradiated particles. Even with fuel burn ups in the range of 19%, the coatings have provided excellent fission product retention even when heated to 1800 C for extended periods of time.
That temperature offers more than 500 C in margin compared to the highest envisioned fuel operating temperatures and 200 C in margin for the highest possible fuel temperature in the most limiting postulated accident.
None of the operated high temperature reactors using coated particle fuels have used a direct cycle gas turbine. Several aborted design efforts have made enough progress to determine that using helium as a coolant is not too hard, but using it as a working fluid in a Brayton Cycle gas turbine leads to some difficult material and mechanical challenges.
The authors of the G-4 study on the use of mobile nuclear power for ground operations noted the possibility of “carbon dioxide, argon, nitrogen” as options to helium. It also reminded readers that the ML-1, the only direct Brayton Cycle nuclear power system that has been operated in the U.S. used pressurized nitrogen as a working fluid and reactor coolant.
Companies that might be in the running and customers that might be following
There are a handful of relatively new start-up companies that have been focusing their design efforts on very small manufactured reactors that might be suitable for powering remote mines, villages, or forward operating bases.
Holos, Ultra Safe Nuclear, StarCore, and X-Energy all come to mind as companies that are working on designs that might be quickly adapted to meet the requirements specified for Project Dilithium. None of them have been working on designs that are directly applicable because none of them had previously identified an appropriate early adopter customer.
There are other potential customers that are interested in reliable electrical power in places that are not readily reachable by wired transmission systems. They can be in remote areas separated by long distances from the existing grid, on islands separated by expanses of water, on offshore exploration platforms, or on ships whose movements prevent wires from being a possible solution.
Because they cannot be reached by wires, they must either rely on diesel or gas turbine generators or they must do without power. In virtually all cases, the unconnected customers survive with a combination of very expensive power from burning refined petroleum in relatively small machines and doing without many of the electrically powered support systems that many of us take for granted.
Those with the fewest resources lean more heavily on doing without power. That often makes their lives a struggle without hope of improvement through industry, education or increased productivity.
Many of these potential customers could benefit from the availability of power from machines with the same characteristics that a military forward operating base needs, so potential suppliers have always identified the military as a potential customer.
Because the DOD’s need to reduce logistics vulnerabilities can be more directly measured in lives lost and dollars spent, it could be an early adopter that provides an accelerating market pull.
Small, manufactured nuclear systems are going to be available fairly soon even without the DOD, but its more pressing needs might provide more resources that enable design completions and manufacturing cost reductions. These will help customers with fewer current resources to afford the power systems that they could already put to good use if they were on the shelf now.
My primary concern about Project Dilithium is tied to the US military’s notorious belief that the technologies it uses are somehow so unique that they must be surrounded by a dense thicket of secrecy, even if they are merely ruggedized versions of power generators or ship propulsion engines that could be used in a wide variety of applications.
Those secrecy rules not only bind up initial development, but they cause enormous cost increases by requiring unique support systems provided by contractors that fiercely protect their special relationships with military leaders, congressional appropriators, and even presidential candidates/office holders.
What about vulnerabilities that Lyman described?
In his BAS article on mobile nuclear plants, Dr. Lyman expressed concerns about the vulnerability of mobile nuclear plants in a hostile environment. One feature of nuclear power systems that many critics forget is the fact that they require robust shielding systems.
The same layers of dense metals and hydrogen containing materials like concrete, plastic and water that protect people from radiation would do a very good job of protecting reactors from penetrating projectiles.
Compared to the diesel and kerosene storage containers needed for current power generators, mobile nuclear should require fewer protective resources.
Even in the unlikely situation that an explosive can be delivered into a properly shielded reactor core, it’s probable that most radioactive material will be effectively retained. Coated particles might be dispersed a short distance, but they seem too dense to be carried very far. It’s probably worthwhile to perform testing to verify this theory.
One of the specific statements made by Dr. Lyman is a myth that continues to be propagated in even the most informed and supportive circles of people with nuclear expertise. He called graphite “the combustible material that brought the world an 11-day radioactive fire after the April 1986 Chernobyl explosion”.
Nuclear grade graphite might be pure carbon, but it isn’t combustible. It’s structure is far too ordered to offer any available sites for the kind of rapid oxidation required for combustion. At Chernobyl, graphite heated to temperatures high enough to produce a red glow was dispersed and landed on combustible materials in the building structure. Those are the materials that actually burned. (Source: Prerelease of a forthcoming IAEA publication on graphite in nuclear reactors.)
Bottom line is that Project Dilithium is a fascinating reach forward that is worth watching. I encourage all decision makers involved to be as transparent as possible, while paying attention to the need to protect certain kinds of details.