APT NEWS Archive

(Earlier news articles may be accessed in the News Archives)

Accelerator Production of Tritium --- That Could Mean 40 Years of Work

LANL is designing an accelerator that would be built at Savannah River; but LANL scientists would be advisors on the technology for years to come.

Los Alamos Monitor
Mar 1, 1998
By Charmian Schaller, Monitor Editor

Los Alamos National Laboratory is playing a key role in the design of one possible approach to future production of tritium. The work is a major project that applies old and new research in accelerator physics to current needs in the US weapons program. The design role has pumped about 580 employees and millions of dollars into the economy of northern New Mexico already. And scientists at LANL could well be playing an advisory role in tritium production technology for the next 40 years. What's more, there could be interesting spinoffs in fields as diverse as transmutation of radioactive waste and the production of medical radioisotopes.

The Importance of Tritium
In the post-Cold War era, the United States has stopped producing new nuclear weapons and is reducing the number it has on the shelf. But in an uncertain world, Congress still considers it vital that the weapons retained in the arsenal be effective should their use ever become necessary. Tritium, an isotope of hydrogen, is essential to this effectiveness. It increases the explosive power of nuclear weapons. The catch is that tritium has a half-life of 12.5 years. It decays at the rate of about 5.5% per year. What this means is that a key component of nuclear weapons tends to "wear out" in a relatively short period of time. At one time, tritium was produced in Department of Energy reactors, but as the public's safety concerns about reactors increased, the last tritium-production reactor, the K Reactor at Savannah River Site near Aiken, SC, was shut down in 1988. Since that time, the US has been renewing the tritium in its weapons by recycling tritium gas taken from nuclear weapons dismantled as part of the US stockpile reduction program. But the DOE says this process isn't sufficient. It has ruled that a new tritium-production capability must come on- line shortly after the turn of the century --- perhaps as early as 2005.

There are two known ways to produce tritium, a substance that is extremely rare in nature. One uses a nuclear reactor. The other uses an accelerator. LANL has been operating an accelerator for more than 40 years, first as the key instrument at the Los Alamos Meson Physics Facility, and now as the prime source of particles for the Los Alamos Neutron Science Center. On Oct. 10, 1995, then-Energy Secretary Hazel O'Leary announced that LANL would lead in the development of accelerator production of tritium (APT), a technology that could help assure a new tritium supply for US nuclear weapons. She said the DOE would pursue a dual track toward production of tritium. It would explore either construction of an accelerator-based system at Savannah River or use of a commercial reactor for tritium production. LANL was to pursue the development and testing of the APT approach. Subsequently, Burns and Roe was chosen as the contractor on the project. Funding going to laboratory and contractor work on the project this year totals about $100 million. DOE is scheduled to choose one of the two approaches by the end of 1998. The approach that isn't chosen will serve as a backup.

How Tritium Is Made in an Accelerator
Tritium has a nucleus composed of two neutrons and a proton. Helium-3 has a nucleus containing two protons and a neutron. In APT (accelerator production of tritium), a stream of protons is speeded up to nearly the speed of light and slammed into a tungsten and lead target. Neutrons are knocked loose from the target (in a process known as "spallation"). Forty-seven neutrons are produced for every proton that strikes the target. The resulting neutrons are slowed down in water, and then used to bombard helium-3, replacing protons in the nuclei with neutrons to transform helium-3 into tritium. Think of it as analogous to what happens on a pool table. A red ball (a neutron) and two green balls (the protons) lie close together in a pattern (the nucleus of helium-3). An expert takes his shot. A red ball (another neutron) rushes toward the cluster and strikes. One of the green balls goes flying out, and a new pattern is created. Now there are two red balls and one green one (and you have a tritium nucleus instead of a helium-3 nucleus). The tritium is then extracted.

In an reactor, by contrast, 12-foot rods of lithium-6 are bombarded with neutrons to produce tritium and helium-4.

LANL has substantial expertise to offer in APT. The laboratory has been working successfully with tungsten neutron sources and spallation for 20 years.

But the DOE's goal isn't to use the accelerator at LANL for production of tritium. The accelerator that would be built at Savannah River for production of tritium would have 400 times more power than the LANL accelerator (about 400 megawatts compared to 1 megawatt here at LANL). In addition, the new accelerator would require far more water for cooling than is used here. LANL's assignment is to certify the design of such an accelerator in several important areas. The lab's own pamphlet on APT lists its tasks as follows: fabricating and operating a full-power, low-energy prototype of the plant accelerator; prototyping the high-energy accelerator equipment; verifying target/blanket performance; and determining the lifetime and properties of target/blanket components under prototypical operating conditions.

The goal is to design an accelerator that will do the job as cheaply and safely as possible with minimum impact on the environment --- no later than the year 2007. APT must be designed to produce 3 kilograms of tritium per year to maintain US nuclear weapons at the level prescribed by START I (the Strategic Arms Reduction Treaty now in effect). It might be possible to reduce that amount by half at a great saving if START II were to win approval soon in Russia, triggering more reduction in arms and less need for tritium.

Advantages of APT
Paul Lisowski, National Project Director for Accelerator Production of Tritium, said in a recent interview that he sees "a lot of advantages to APT over using a nuclear reactor." Perhaps the most important advantage is improved safety. In an accelerator, neutrons are produced by spallation (one particle strikes another, knocking it out of place). In a reactor, they are produced by nuclear fission (the splitting of the nucleus of the atom). The APT process produces virtually no radioactive waste and presents no chance of a criticality accident. The reactor process, in contrast, produces considerable radioactive waste, and involves a finite risk of a criticality accident plus some nuclear proliferation concerns. In APT, an immediate shutdown is possible. In a reactor, shutdown is slow. Lisowski said that in APT, "The amount of tritium available for release is very, very small ... Nothing is left over..." The reactor process would require shipment of "very hot" materials to Savannah River, and much waste would have to be stored. In summary, Lisowski said, the risk of APT is much lower.

There would be an additional advantage in that the helium-3, a poison from dismantled weapons, already is going to Savannah River for recycling. On the other hand, there would be public policy questions involved in production of tritium in one of some 100 existing commercial reactors. In the past, the US has, as a matter of policy, been careful to separate military and civilian uses of reactors. If the Hanford Site, a DOE facility in Richland, Wash., were to fire up its reactor to produce tritium, it would need MOX fuel (a mixture of uranium and plutonium oxides), Lisowski said. The source of such fuel would have to be weapons-grade plutonium, and such a use of plutonium might raise questions about an existing agreement with the Russians.

While LANL is working on cost and efficiency questions about APT, the DOE's team studying the reactor option must, among other things, clarify the public policy issues by the end of 1998 when DOE must make its decision. Lisowski contends that APT "is less expensive than a dedicated nuclear reactor would be to make tritium --- much less." The cost of APT would range from $1.783 billion to $2.272 billion depending, on whether the project used a design based on START I (3 kilograms of tritium) or a modular design with the flexibility to allow for START II (1.5 kilograms of tritium), he said. Yes, Lisowski acknowledged, that sounds like a lot of money --- until you compare it to other defense costs. That amount is, he said, "about the cost of something like an aircraft carrier." And for that amount of money, you get all the tritium you need for 40 years.

Coups, Breakthroughs, and Special Efforts
Burns and Roe and LANL are hard at work on their APT tasks, and Lisowski made it clear that he believes things are going well. The development of the modular plan --- the design that allows for late design changes should START II win approval --- looks simple in a diagram, but was an inspired idea. It involves building the target area at the mid-point in the potential length of the accelerator. Should START II be approved before construction reaches beyond the target area, it won't be necessary to build the portion of the accelerator beyond the target area. If START I remains the only treaty in place and the portion of the accelerator beyond the target area must be built, then particles would be accelerated in a "U" back to the target. The result: potential savings of as much as $800 million and a great gain in flexibility. The project would have until 2002 before it must make final construction decisions dealing with the START I / START II issue.


Mockup of the Radiofrequency Quadrupole (RFQ) in the accelerator tunnel of the
Low Energy Demonstration Accelerator (LEDA) building in Los Alamos---Photo by Al Cabral
On Jan. 30, team members working on the injector in the Low Energy Demonstration Accelerator accelerated 79 milliamps of protons to 1.25 MeV, setting a world record. Lisowski and all of his team members are working hard to maintain unusual integration and cooperation from the top down in an organization that starts with the DOE Program Manager, steps down to the Project Director (Lisowski), and includes three major subdivisions beyond that level: Technology (conceptual design, engineering development and demonstration, a process led by LANL); the Plant Project (plant design, construction, and commissioning, led by Burns and Roe); and Operations (led by Westinghouse Savannah River Company). Lisowski emphasized the cooperation involving government, scientists from several laboratories, and industry. He said he paid close attention to a General Accounting Office report on big projects that showed only 13 of 82 finished. He wants to finish.

Lisowski has made a point to maintain an external review committee. It's led by Bill Herrmansfeldt of Stanford University. And Lisowski and other team members already are dreaming of spinoffs from APT. The new high-power accelerator technology could have advantages in producing medical radioisotopes, destroying radioactive wastes, producing energy without long-lived nuclear waste, making material for space-power applications, studying radiation damage, and producing neutrons for basic and applied research. People from the project already are working on some of these spinoffs. A conference is scheduled in July to explore what physicians would like in the way of new radioisotopes.

And there have already been some results. One is a major, positive impact on the economy of northern New Mexico, A small example: Burns and Roe has been able to hire some staff members who lost their jobs when other contractors downsized. Another example is the $1 million set aside for involvement of universities in New Mexico and in the Southeast in the project. Students are writing short papers on proposals, and the best of those might be funded.


APT Home | Highlights | News | News Archives | Project Offices