All the tools of nuclear alchemy are on display at Los Alamos, from reactors, lasers, particle accelerators and ion beam facilities to a cryogenics lab and a giant foundry. Look closer, however, and it soon becomes apparent that Los Alamos is reeling from profound changes. Once the epicenter of a nuclear weapons complex designed to create thousands of new weapons a year, Los Alamos today is heralding the end of the nuclear era.
The United States has not designed or produced a new nuclear weapon in more than a decade, and is mothballing or retiring much of the capability to do so. The last nuclear explosion in the United States was triggered at the Nevada Test Site in 1992 (See "Beyond Ground Zero,"), shortly before the Bush administration declared, and Congress wrote into law, a moratorium on nuclear testing. This year the Clinton administration hopes to win Senate passage of the Comprehensive Nuclear Test Ban Treaty (CTBT) banning future nuclear tests for 144 signatories, including the declared nuclear states China, Great Britain, France, Russia and the United States. (Although the treaty has not been ratified, all the signatories are acting as if it had been.)
"Any one of these changes would have required a large adjustment, but taken together they've sent us into completely uncharted territory," says Joseph Martz, a nuclear weapons program leader at Los Alamos. Martz's job today is to monitor an aging stockpile for signs of deterioration, knowing that in 10 years most of the people who designed and tested the weapons, and the manufacturing base that produced them, will be gone.
"That means we might literally need as much as a decade's warning of a problem in the stockpile that would require remanufacturing," says Martz. "That's a pretty awesome technological challenge."
Not everyone is comfortable with the rapid mothballing of the nuclear weapons complex. Senate Foreign Relations Committee Chairman Jesse Helms, R-N.C., has indicated early opposition to the CTBT, for instance, stating in a letter to the President that the treaty is "very low" on the committee's agenda.
"I don't think Sen. Helms will bring the CTBT up this year, which is a good thing," says Sen. Jon Kyl, R-Ariz., a staunch opponent of the treaty. "Here we have an aging stockpile of the most awesome and complex weapons ever devised, and we're not testing them for reliability and safety like you would any other military or commercial product."
But Sen. Jeff Bingaman, D-N.M., a member of the Armed Services Committee and a proponent for both Los Alamos and Sandia National Laboratories in New Mexico, believes the test ban enjoys strong public and congressional support. "The conclusion I've drawn from hearings with the directors of the national laboratories is that we can retain confidence in our stockpile without testing, but it's going to require significant effort and bipartisan congressional support," he says.
At the core of the debate over the CTBT and recent changes in the nuclear weapons complex is a fundamental question: Can the United States really maintain a credible and safe nuclear deterrent without the benefit of new designs, weapons production or nuclear testing? Energy Department officials insist they can, with an ambitious program called Science-Based Stockpile Stewardship. Others both within and outside the nuclear weapons complex aren't so sure.
The first shock to the government's nuclear facilities came in 1988, with President Bush's decision to halt all new weapons development and production. That directive ensured a steadily aging stockpile, and threw into question the future of the nation's small band of nuclear weapons designers. Since that time, all nuclear facilities-design and development laboratories, massive production facilities and the Nevada Test Site-have slowed to a virtual halt or recast themselves for new missions.
Of the seven major production facilities that constituted the vast nuclear production complex, only the Pantex Plant in Texas (where U.S. nuclear weapons are being dismantled under the START treaties) maintains anything close to its Cold War pace of operations. The total U.S. nuclear weapons complex will shrink from 29.1 million feet of floor space in 1985 to a projected 6.4 million in 2005. Because the other declared nuclear states that have endorsed CTBT continue to manufacture new nuclear weapons, a number of scientists consider the mothballing of the U.S. production complex to be troublesome.
"At some point we're going to have to remanufacture if not new weapons, then major components using new, modern processes," says Michael Anastasio, a senior weapons designer at Lawrence Livermore National Laboratory in California. "Unfortunately, many of the problems we've identified in the past with testing occurred in that transition from the design lab to the manufacturing and assembly line. So all of these little changes are driving us out of our comfort zone into the red zone."
The 1992 halt to all nuclear tests compounded the uncertainties created by the halt of weapons designs and production. Nuclear testing was viewed as the essential experiment that gave nuclear scientists confidence in their calculations. It was also used as the "pass/fail" exam for identifying and advancing the most promising weapons designers.
"Careers were made and broken at the test site. Weapons designers, including myself, were promoted based on our ability to conduct successful tests," says George H. Miller, director of national security programs at Lawrence Livermore, which, along with Los Alamos and Sandia, shares the stewardship of the nuclear weapons stockpile. "So the organization we evolved to efficiently design, produce, test and field new weapons, along with our entire reward system, was all swept away by the end of the Cold War. And that has been a really traumatic transition."
In place of the venerable design-manufacture-test cycle that long powered the nuclear weapons complex, the Energy Department and weapons laboratories have substituted the ambitious Stockpile Stewardship program. The idea is to deconstruct the extraordinarily complex phenomena of a nuclear chain reaction-high-explosive induced implosion, nuclear fission, tritium boosting and thermonuclear fusion-into its component parts.
In order to shed light on each component, DOE is funding construction of a series of expensive experimental facilities, which taken together represent the largest scientific project in the world. Eventually the data gathered at the facilities will be fed into a massively parallel supercomputer capable of performing 100 trillion mathematical calculations per second (100 times faster than the most advanced computer), theoretically leading to an accurate computer simulation of a nuclear explosion.
In terms of scientific complexity and challenge, many experts believe there are only two projects comparable to Stockpile Stewardship in the 20th century: the Manhattan Project and NASA's Apollo moon shot program. And like both of those efforts, there's not a man or woman involved in Stockpile Stewardship willing to say with certainty whether the goal of the program is even attainable.
"I think this is pretty close to the moon shot in terms of difficulty, because we're requiring increases in computing speed which have never been seen since the invention of the microprocessor," says David Cooper, associate director for computation at Lawrence Livermore and an Apollo veteran with 30 years of supercomputing experience at NASA. "I'm an optimist, so I think we can pull this off. But when you compare trying to simulate an aging nuclear stockpile on a computer to some of the computing problems I worked on at NASA, they were a slam-dunk. This is a half-court shot."
Donald Wolkerstorfer, a program manager and weapons designer at Los Alamos, calls Stockpile Stewardship a scientific "leap of faith." The nation won't realize until around 2005 whether the labs have reached the other side or fallen perilously short.
"The fundamental premise of science dictates that you test your hypothesis with experiments, and that means testing," says Wolkerstorfer. "Instead, we're going to conduct a series of smaller component tests, and rely on the ability of computers to integrate them into an accurate simulation. If you'll recall, that's the basic method we used in deploying the Hubble Telescope."
No Sure Thing
The critical decision to entrust the nuclear arsenal indefinitely to Stockpile Stewardship came during a summer 1995 meeting in Washington. Among the participants were Vic Reis, DOE's assistant secretary for defense programs; Anthony Lake, former national security adviser; Gen. John Shalikashvili, former chairman of the Joint Chiefs of Staff; and the heads of the nuclear weapons laboratories (which work for DOE under contracts managed by the University of California in the case of Los Alamos and Livermore, and Lockheed Martin for Sandia). According to participants, each of the key players represented different, and in some cases competing, agendas.
With the proliferation of weapons of mass destruction considered the greatest single threat to national security, White House officials viewed a continuation of the moratorium on nuclear testing as imperative. Specifically, the commitment of the Clinton administration and the other declared nuclear states to CTBT proved essential in winning international approval for an extension of the Nuclear Non-Proliferation Treaty in 1995. Under the treaty, 180 non-nuclear nations agreed not to pursue nuclear weapons programs. "There's no question that CTBT underpinned our nuclear non-proliferation diplomacy," says a White House source.
Military leaders remained the most skeptical of the moratorium. Privately, a number of senior officers feared the weapons laboratories were being coerced into endorsing a risky program by the promise of an ambitious new mission, significantly increased budgets and new "big science" facilities.
As part of Stockpile Stewardship, the weapons labs will construct a $1.2 billion National Ignition Facility at Lawrence Livermore. The facility will include:
- One of the world's most powerful lasers.
- An $81 million giant X-ray for photographing high-explosive phenomena called the Dual-Axis Radiographic Hydrodynamic Test facility at Los Alamos.
- A $1 billion supercomputer project for all three labs called the Accelerated Strategic Computing Initiative.
- Potentially, a $3 billion particle accelerator for producing tritium, an essential ingredient of modern nuclear weapons with a relatively short half-life of 12 years.
Military officers were concerned that while the United States could still conduct underground "subcritical" tests at the Nevada Test Site, provisions of the CTBT precluded any test that triggered a nuclear reaction, no matter how small. Uniformed leaders were also skeptical of the open-ended nature of the treaty.
"Gen. Shalikashvili understood that the problem is not today's stockpile, which is in good shape and well-tested, but rather the stockpile 10, 20 or 30 years from now," says Reis. "[Stockpile Stewardship] is a very different kind of program. It has a lot more scientific content than the military is traditionally comfortable with."
Despite the obvious allure of the stewardship program, the directors of the national labs had their own concerns. They feared the political will to support such an ambitious effort would wane over time. A Congressional Budget Office analysis, for instance, had already recommended forgoing the National Ignition Facility and consolidating nuclear weapons work at one of the labs.
"Our greatest fear continues to be that political support will eventually diminish and Washington will start chiseling away at the edges of a program with little margin for it," says Lawrence Livermore Director C. Bruce Tartar. The labs' original estimate for running Stockpile Stewardship, officials say, was roughly $6 billion annually, significantly above the slated budget of $4.5 billion a year.
On the other hand, none of the lab directors believed a return to robust nuclear testing was in the political cards. Even if the Senate defeats CTBT, critics and proponents agree that in the absence of a crisis, Congress is unlikely to overturn the current moratorium on testing.
"In our own calculus, we lab directors could not envision a political debate whose outcome would be a return to frequent testing," says Tartar. "My final summary to Gen. Shalikashvili, who was the toughest questioner I've faced since my Ph.D. exam, was . . . that we have a better than even chance of making this program work. But the nature of the science is too complicated to suggest it's even close to a sure thing."
To assuage Pentagon concerns and maintain confidence in the stockpile, the laboratories pledged to retain the capability to resume regular testing within three years of notification. The presidential directive establishing the Stockpile Stewardship program also called for a new, annual certification procedure for the nuclear stockpile. Each year the Secretaries of Defense and Energy must certify that the stockpile is safe and reliable. If that certification is not given, the President can invoke a "supreme national interest" clause and resume nuclear testing.
Frozen in Time
Even with such safeguards, the Stockpile Stewardship program is clearly in a race against time. The nuclear stockpile is already the oldest it has ever been. Because of the high rate of turnover, the average age of the stockpile during the Cold War hovered at around 13 years. Many of the weapons are now much older than that, and well beyond their original design life.
More important, no one in the surprisingly small fraternity of U.S. nuclear weapons designers-there are fewer than 100 today-has actually designed a weapon in the past decade, nor staked his or her reputation on a real-world test since 1992. Now the Stockpile Stewardship program is attempting to freeze in time the reliability and safety of a stockpile of weapons and the competence of a generation of nuclear scientists. In reality, both are perishable.
To better understand the effects of an aging stockpile, the labs have instituted an "enhanced surveillance" regimen that involves dismantling representative samples of the stockpile each year. While no major problems have been identified to date, experts say the inspections have already led to modifications of some weapons.
Experts compare an aging nuclear weapon, with its various radioactive materials, plastic high explosives and metal components, to a car that sits in the sun for years. "Over time, the glue on the windshield will haze, the upholstery will become more brittle and the dashboard will crack," says Wolkerstorfer. "A similar phenomenon occurs inside a weapon, and we really don't know in what time frame that becomes a problem."
In an effort to buy time, the weapons labs began an aggressive archiving program in 1994. Information about more than 1,000 nuclear tests going back four decades is being updated and entered into computer databases for easy retrieval and cross-checking. Archivists are also conducting extensive videotaped interviews in an attempt to preserve the expertise of a dwindling generation of weapons designers and engineers who had numerous new designs and nuclear tests under their belts.
"While we always recognized that much of the expertise in this business resided in people's heads, we've been surprised at how many of the details existed almost in the realm of folklore that was passed along from one generation to the next," says Charlie Miller, chief of archiving at Los Alamos. Within 10 years, he notes, most of the weapons designers with experience in nuclear testing will have retired. "So we're trying to lay down the information that will smooth the path if we ever have to go back to testing."
Lab directors are perhaps most concerned with recruiting and training the next generation of nuclear scientists to ensure they will have the confidence-and the competence-to make the politically unpopular decision to rekindle a long-dormant nuclear flame if necessary. In that sense, the construction of a series of advanced experimental facilities is viewed as virtually a quid pro quo for the labs' commitment to Stockpile Stewardship.
"The way we validated people in the past has disappeared, and we need these facilities to stress our people and their judgments in an arena in which they can fail," says Lawrence Livermore's Tartar. "Even with these new facilities, the question remains whether we can keep from fooling ourselves about how good we are and how much we really know. I think we can. Without these facilities, my own judgment is there's not a chance in hell we can."
At the national weapons laboratories, the talk these days frequently centers on the dangers of fooling yourself. A few of the scientists are old enough to remember a three-year testing moratorium in the early 1960s, and their surprise when renewed testing revealed that defective weapons had been mistakenly put into the stockpile. Others note the inherent fallibility of calculations designed to do things like extrapolate temperatures on the surface of the sun from the microscopic point of an experimental laser. Then again, perhaps it is inevitable that a generation of scientists who devoted their lives to unleashing and then harnessing the nuclear genie sense hubris in plans to confine it to a laboratory flask.
James Kitfield is a staff correspondent at National Journal.