In the closing days of the Bush administration, we see nuclear advocates laying down markers for the debate about nuclear weapons that is expected early in Senator Obama’s presidency. In a recent speech at Carnegie and in an article in Foreign Affairs, Robert Gates, slated to stay on as Secretary of Defense, has called for developing a new nuclear bomb — the so-called Reliable Replacement Warhead (RRW) (“so-called” because it is no more reliable than current warheads and it will not replace them, but it is otherwise appropriately named). General Kevin Chilton, head of Strategic Command, has made similar statements.
Those of us who are interested in working toward a world free of nuclear weapons realize that progress will involve many steps, some large, some small. One important step will be ratification of the Comprehensive Test Ban Treaty (CTBT). Some CTBT supporters suspect that the outlines of a deal are coalescing: those who want the RRW will try to make the CTBT and the RRW a package deal, arguing that we will be able to maintain a reliable, safe nuclear deterrent without testing, as the CTBT would require, only if the weapon labs are allowed to proceed with weapon modernization. The Congressional Strategic Posture Commission interim report appears to be at least sympathetic to this view. This artificial link is based on both faulty logic and a long list of unstated and unsupportable assumptions.
The assertion that our nuclear weapons need any modernizing implies, usually implicitly, that current weapons are antiques that are not quite up to snuff. Chilton, in the article cited above, specifically links U.S. modernization to Russian and Chinese nuclear weapons. This superficially makes sense: after all, we don’t send our military out to fight with World War II vintage tanks, ships, and airplanes. Certainly the United States should be armed with the latest and best nuclear weapons; at the very least, our weapons have to be at least as modern as any possible competitors, right? The simple analogy to conventional weapons doesn’t hold because of the types of tasks assigned to nuclear weapons and some confusion about just what a “nuclear weapon” is.
A nuclear bomb is an immensely powerful explosive, whose job is to blow things up. Our tanks fight other tanks and our ships fight other ships in an intrinsically competitive interaction. If our potential enemies improve their tanks or ships, then we must improve ours. But nuclear bombs do not fight other nuclear bombs, they blow things up. They might be expected to blow up airbases, ports, or oil refineries; whatever they are expected to blow up, if they can do that, then they accomplish their mission. There are certain targets that even the most powerful nuclear bombs cannot blow up, for example, targets buried in tunnels in hard rock or targets whose location is unknown or potential targets that we do not even know exist. Whether a target can be destroyed or not depends on how big a bang the bomb makes (and how accurately it is delivered), not how “modern” it is. Even when one of the main missions of U.S. nuclear weapons was to attack and destroy the Soviet Union’s nuclear-armed missiles in their silos, we were pitting our nuclear weapons, not against Soviet nuclear weapons, but against Soviet concrete. The world’s most modern nuclear weapon would be easily destroyed if a 1945-vintage Hiroshima-style bomb blew up next to it.
A conventional analogy makes the point: The Joint Direct Attack Munition, or JDAM, is a family of highly accurate conventional bombs. The bombs are dropped from the world’s most sophisticated, multi-million dollar aircraft, including the B-2, the radar-evading stealth bomber, which costs over a billion dollars. Once in free-fall, the guidance system on the JDAM picks up signals broadcast to Earth from a multi-billion dollar constellation of satellites that make up the Global Positioning System. These satellites broadcast time signals with an accuracy of a few billionths of a second, allowing tiny computers on the JDAM to calculate the bomb’s trajectory to within a few feet. Some JDAMs are outfitted with smart fuzes. When a bomb crashes through a building, it will decelerate when it goes through the floors of the building but not in the airspace between the floors. A smart fuze will detect these decelerations and count the number of floors that it has passed through, allowing the bomb to be exploded on a particular floor. Obviously, this bomb is extremely sophisticated—a marvel of modern military technology. So what is the explosive used in the bomb? Something called tritonal, which is a mixture of TNT and powdered aluminum. TNT, the shorthand for trinitrotoluene, was discovered in 1863. Millions of pounds of it were used in the First World War. The trick of mixing TNT with powdered aluminum to increase its power is a more modern innovation, only going back to the Second World War. Clearly, this astonishingly sophisticated bomb does not depend on some modern sophisticated explosive to make it effective. The explosive is the easiest part of the job. If you can use a highly accurate guidance and fuzing system to place a few hundred pounds of TNT into, say, a communications headquarters, the fact that the bomb uses an explosive that is a century-old technology is not going to save the world’s most modern electronics.
Do not misunderstand: nuclear weapons are very complex, precision devices. And they can have specialized performance, for example, when the Nixon-era ballistic missile defense system used nuclear-armed interceptor missiles, the nuclear warheads for use in space were tailored to produce extra x-rays while those for use in the atmosphere were tailored to produce extra neutrons. A nuclear weapon intended to produce an electromagnetic pulse should be designed to maximize the second derivative of the gamma ray flux (you will just have to trust me on that one). I am simply saying that nuclear bombs do not have to be complex and their required complexity depends on the missions assigned to them, and those missions should be dramatically different now that the Cold War is over.
During the Cold War, there was a great incentive to build huge numbers of nuclear bombs. The United States was the first to put some of these bombs on missiles that were small enough to fit on submarines. In the interests of efficiency, both the United States and the Soviets put several bombs on a single missile. Both sides wanted the bombs to be able to attack missile silos and other very hard targets. The Soviets tended to build bigger missiles than the United States. The United States took the more sophisticated course of building smaller missiles and put a great deal of effort into getting nuclear weapons to be as small as possible, which required sophisticated designs and precision manufacturing.
But nuclear bombs are, in principle, pretty simple. The complexity of current weapons comes about because of the extreme performance required, not because of any necessary and intrinsic characteristic of nuclear weapons. The first bomb used in war, the one that destroyed Hiroshima, was such a simple uranium design that Manhattan Project scientists had enough confidence to use it without testing it beforehand. So the first thing to note is that any statement about the need for weapons modernization or the needed sophistication of nuclear bombs rests on a hidden assumption about what the mission for that nuclear bomb will be and how it will be delivered to its target. Since the fundamental missions of nuclear weapons—the the question of what we are asking them to do, what they are for—are now being reevaluated and remain unsettled, it is premature to say just how sophisticated nuclear weapons have to be and whether they need any “modernizing” at all.
We must also be careful not to be confused by what a nuclear weapon is. There is a part that simply goes boom. That can be simple or complex. But the bomb sits atop a missile or in a bomber or submarine that is a truly complex piece of machinery. The bomb will be guided by advanced electronic systems, perhaps including sensors. All together, this makes up a “weapon,” almost all of the components of which are non-nuclear and can be tested fully under a Comprehensive Test Ban Treaty, which only prohibits tests that produce a nuclear reaction. The job of the nuclear part of the weapon could in principle be very simple indeed: to explode and blow something up.
Paying attention to what is the nuclear part of the nuclear bomb is also relevant to DOE’s claim that we need a new weapon because of safety. Most of the safety attributes are in the electronics that is outside the nuclear package and can be developed and tested without nuclear testing. A few safety features are intimately connected to the nuclear package but before the RRW was even thought up DOE had plans for retroactively inserting new safety features, some quite major, into existing warheads.
I have written here about how the characteristics of the RRW imply, without debate or even apparent conscious thought, certain missions for nuclear weapons, missions that are at best of dubious relevance and most likely clearly outmoded and contrary to the security interests of the United States and the world. The most insidious part of this RRW-CTBT “deal” is the implied assertion that the new nuclear bomb we need is exactly the one that the Department of Energy (DOE) and weapon labs are trying to foist on us, along with all of its implied missions.
The United States does not need a new nuclear bomb but let’s assume for a moment that under a CTBT it does. What should that bomb be? The DOE presents many arguments. Some are based on assertion, such as the RRW will save money in the long term, even though the DOE has not done a sufficiently detailed cost study to support that statement (and historically DOE cost estimates are mind-bogglingly inaccurate). Another argument is that the labs have to physically build new bombs simply to continue to be able to build new bombs. That is, the labs need to maintain the expertise required to design and build modern, sophisticated nuclear bombs and, to do that, they need modern, sophisticated bombs to practice on. But without Cold War mission requirements, why do they need to be particularly sophisticated?
If some sort of machine, whether a computer or a toaster, has some manufacturing step that is complex and difficult, any engineer will be proud to develop some clever process to more simply and cheaply carry out that step. But the engineer will admit that that is the second best solution; the ideal solution is to redesign the machine so it still does what it is supposed to do but it does not even include that step in its manufacture. In other words, the ideal solution is not to solve the problem but to design the problem out of the system.
If maintaining design expertise is a challenge, the labs should be designing bombs, and procedures for designing future bombs, that do not require such sophisticated design expertise. If manufacturing plutonium pits is so difficult, then design bombs that do not use plutonium. (The reply will come back that we need plutonium pits because we need triggers for two-stage thermonuclear bombs with yields of tens or hundreds of kilotons of TNT explosive force. But what future missions require such huge yields? During the Cold War, the answer was attacking Soviet missile silos in a disarming first strike. If that is the future mission they have in mind, we should debate that first and only then build the new weapon to carry out that mission.) If maintaining a nuclear weapons industrial complex and the corresponding sophisticated work force is such a challenge, then DOE should design new weapons that do not require either. DOE will respond that that is precisely what they are trying to do with the RRW, but that claim is contradicted by the scale of their multi-billion dollar budget requests for the future nuclear weapons complex. Even if that is DOE’s direction, it does not go anywhere far enough.
Simple uranium bombs with high reliability and yields of twenty kilotons (or the power of the bomb that destroyed Hiroshima) or more would be easy to manufacture. We could design such a weapon, perhaps build one or two, and put the plans on the shelf in case we ever needed it. I can’t help but imagine those language-free schematic assembly instructions that come along with unassembled Ikea furniture, describing how to put a bookshelf together without special skills or complex tools. We should design the Ikea Bomb. The DOE’s arguments for a new nuclear bomb design would be a lot more convincing if DOE were eagerly trying to design themselves out of a job rather than looking at a future that has them building nuclear weapons forever.