Last week, the Jasons, a group of distinguished scientists who advise the Department of Defense, released the unclassified summary of their review of the Reliable Replacement Warhead, or RRW. Walter Pincus covered the report in the Washington Post. The summary is posted on the FAS website.
Although almost everything about nuclear weapons and their design is classified, the unclassified summary contains many hints of problems with the RRW that are yet to be overcome.
According to the report, the nuclear components of the RRW design have been fully tested during the time the U.S. was conducting nuclear tests. This confirms a previous report that the primary of the RRW is based on the well-tested but never weaponized SKUA-9 explosive package.
The challenge is to certify a design that will work without ever testing the weapon. That can be done but the Jasons judge that the RRW is not yet at that point. No details are provided but the review committee seems to believe that the computer models that must be used in lieu of actual testing to simulate the performance of the nuclear explosion are not sufficiently developed to give adequate confidence that the new design will work.
There is much left unsaid here. First, what is confident enough? We can never have 100% confidence in any nuclear weapon, even one that is thoroughly tested. I am fairly certain that if you asked the authors of the report whether the current design would work each would express high confidence that the design would work. But if “high” means 90% confidence and the criterion is 99%, then they are not yet confident enough. The problem the RRW faces is that we now have extremely high confidence in the existing warheads, in the 98-99% range. If it is to live up to its name, the Reliable Replacement Warhead has a very high hurtle to leap. Otherwise, why bother?
The report also calls for additional experiments and calculations. These include hydrodynamic tests in which less than critical masses of plutonium are squeezed and shaped by conventional explosives like those used to trigger a nuclear weapon. They may include tests on smaller quantities of plutonium to better understand its behavior under the extreme conditions found in a nuclear bomb. Anne Fitzpatrick and I described these experiments in great detail in a recent FAS Occasional Paper.
The report specifically calls out for the need for better understanding of the “surety features” of the warhead. These are the parts of the warhead that make an accidental explosion impossible even under extreme condition, for example, if the building or vehicle the warhead is in catches fire. Also, surety features make the warhead impossible to explode without special codes or authorization so, if the warhead is stolen, it cannot be exploded by a terrorist. One might think that a stolen warhead would at the very least provide terrorists with a source of plutonium that they could use in their own bomb but when, at a meeting, I asked John Harvey of NNSA about this he said that the warhead even contains features that will make reuse of the plutonium difficult. Harvey couldn’t provide any details on how this could work and I can only speculate. Perhaps some mechanism to automatically contaminate the plutonium with neutron-absorbing material so it will not be useful for a rebuilt bomb? Whatever these features are, they are a major justification for the RRW but the Jason committee does not believe that they have been adequately proven.
Finally, another major justification for the RRW is that it will, in the long run, save money. These savings will accrue by designing the weapon from the beginning to be simpler to manufacture and maintain. The Jason report notes that, while they were not asked to examine costs, the effects of proposed new and cheaper manufacturing techniques on the weapon’s performance are not yet adequately understood. Moreover, the report recommends that, as a hedge, the existing proven manufacturing processes should be maintained until the new procedures are fully understood. This is yet another reason to question the DOE’s as yet unsupported claims that the RRW will save money: The new manufacturing process may indeed be cheaper than the old process but maintaining both the new and the old process cannot be cheaper than the old process by itself. (Keeping in mind that maintaining the process will almost certainly require maintaining the equipment and keeping the people active. DOE consistently argues that it has to keep people doing whatever it is they do to maintain expertise.)
The report does not say that the RRW will not work; certainly it does not say that the warhead is unworkable, but it does state that more research, calculation, and experiment need to be done. The unclassified summary of the report does not put it in these terms but it seems clear to me that the RRW is not yet ready to pass from the design phase (called 2a) into the engineering phase (called phase 3) as DOE wants.
This is further support for California’s Senator Feinstein’s bill (S. 1914) that would put further development of the RRW on hold until a thorough reevaluation of U.S. nuclear doctrine has been completed.
And this leads to the one question the DOE did not ask the Jason committee, because DOE doesn’t even ask itself, is: why must the RRW have the characteristics that it has?
The way the problem is presented, the RRW is what it is and the question is whether it can be certified with adequate confidence given the tools available today. (The committee’s answer is no.) Similarly, the RRW is what it is and the question is what manufacturing capability, called the “complex” by the DOE, is needed. If the DOE’s justifications for the RRW —reduced cost, greater reliability, simplicity of manufacture and maintenance—were taken seriously, it is more natural to turn the question around: Given the evaluation tools we have available today, what sort of warhead could we certify? Given a goal of having the smallest, cheapest complex possible without the need to maintain esoteric skills, what sort of warhead could we build?
Virtually all of the critical potential or feared problems with nuclear warheads, certainly the ones that cannot be resolved without testing, are really plutonium problems. Questions about aging and stability or differences between cast and forged pits are uniquely plutonium problems. Similarly, the hardest and single most expensive part of weapon manufacture is the plutonium pit production. Questions about warhead certification mostly boil down to questions about certifying that the plutonium pit will operate as planned. After more than decade of trying, DOE last week was finally able to certify a new pit for the W88 warhead and that pit was a reproduction of an existing pit developed through nuclear testing.
Why do we need plutonium instead of uranium? Because the powerful high density plutonium core is what ignites the larger, much more powerful secondary, the hydrogen part of a hydrogen or thermonuclear bomb. This also explains DOE’s claims that it needs to maintain design expertise. Design expertise for what? The simplest atomic bombs are quite simple—the challenging design is the design of sophisticated two-stage thermonuclear bombs.
And why do we need thermonuclear bombs? Because only they are capable of explosive force measure in the hundreds of thousands of tons of TNT, an order of magnitude more powerful than the bomb that destroyed Hiroshima.
And why do we need bombs of such immense power? Because of the missions, which are largely holdovers from the Cold War and remain unexamined almost two decades later. These missions include a disarming first strike against Russian nuclear forces, a mission the administration claims no longer exists although nothing else can explain the numbers of nuclear weapons, nor their day-to-day deployment, nor their characteristics, including their immense power.
And why do these missions persist? Because U.S. nuclear doctrine still considers nuclear weapons to be legitimate instruments of state power, indeed, the foundation of our security. This entire discussion about weapons—aging, manufacture, maintenance—hangs on outdated, dangerous, immoral, counterproductive doctrine that has been carried forward from the Cold War by unthinking momentum. The RRW certainly is not the cure, nor is it the disease; it is just a symptom of our doctrine.
It may well be that the U.S. needs a new nuclear weapon. If the country decides that nuclear weapons have any use at all it is highly improbable that the weapons left over from the entirely different conditions of Cold War will just happen to be most appropriate. For example, if the country adopted a nuclear doctrine of minimal deterrence, that is, the one and only mission for nuclear weapons is to survive an attack by nuclear weapons and retaliate to deter that nuclear attack in the first place, then the weapons could be very simple gun-assembled uranium weapons. These would be big and clunky, and delivered by slow aircraft rather than fast missiles and incapable of a first disarming first strike. They could be stored disassembled in deep tunnels. And we would need perhaps a dozen of them. And the result? All the “problems” of the DOE weapon complex would disappear. No need for plutonium pit facilities, no need to train elite bomb design teams, no questions of reliability, no need to ever test a weapon again. If that were what the RRW was going to be, I could imagine signing up for that.