IRAQI INSPECTIONS: LESSONS LEARNED
by David KayDavid Kay is the Secretary General of the Uranium Institute, an international trade association. Mr. Kay was the team leader for three IAEA inspections in Iraq. The following is adapted from a transcript of a talk given for the Program of Nonproliferation Studies at the Monterey Institute of International Studies on February 10, 1993.
What are the lessons learned from Iraq? I've been surprised to find out how many people do not understand what the Iraqi challenge was. To address this question, it's important to look at what the Iraqis were doing, so you will understand the need to try to draw lessons from the Iraqi experience.
Background: The Clandestine Iraqi ProgramIn June, 1981, Iraq's Osirak reactor, which was being constructed with French assistance, was attacked by the Israelis. From information acquired during the inspections of Iraq, we know that after the attack, the Iraqis had a full inter-governmental task force review what their policy should be with regard to their clandestine nuclear program. One of the questions before that task force was what Iraq's attitude should be about continued membership in the Nonproliferation Treaty. It essentially broke down into two camps. The diplomats said Iraq should stay in the NPT, and go ahead with the clandestine program. They said it would just draw too much attention to the program if Iraq got out. The scientists, being far more upstanding, said no, get out of the NPT; it didn't protect us anyway when the Israelis attacked our reactor, and let's go ahead with our clandestine program as is our right.
At the final meeting, Saddam Hussein turned to the individual who eventually became the scientific head of the Iraqi program, Jaffar Dhia Jaffar, and said, "Dr. Jaffar, if we stay in the NPT, will it in any way hinder the clandestine nuclear program?" Jaffar says his answer was an immediate and unequivocal no; he said it would have absolutely no effect upon Iraq's program. That is one of the most important lessons to learn. If anything has to make tomorrow different from today, and all other days previous, it has to be that governments that are trying to make a similar decision (that is, if we stay in the NPT, will it make a difference) have to not be able to reach such an absolute conclusion with such assurance.
The Iraqis did reach this conclusion, and Iraq went ahead and developed an enormous program. The current estimate is that it involved between seven and twelve billion dollars of expenditures. It involved physical facilities all over Iraq from about 60 kilometers to the south of Bagdad, all the way out to the Syrian border, all the way up north of Mosul. These are large facilities: the facility at Tarmiya, the civil construction for which was done by a Yugoslav construction firm, is about 3.5 kilometers on a side, and there's a similar one at Ash-Sharqat. The Iraqis also had in place three major uranium enrichment programs: a calutron program (the original method used by the US, which was also used by the Soviet Union, Britain, Japan and the PRC); a gas centrifuge program, which acquired or built enough parts for at least 10,000 gas centrifuges, which is a huge number; and a chemical enrichment program. They also did a significant amount of research on gaseous diffusion, laser enrichment and jet nozzle enrichment as well.
At the time of the war, they were already in at least the fifth iteration of an implosion design device, with a calculated yield of 20 kilotons, which is roughly the size of the weapon used at Hiroshima. Equally important is the fact that they cut the weight of the device in half. This in fact is the way you really measure efficiency in a weapon design program, particularly if you have to put the device on the end of a missile. Scientists initially develop things that are essentially delivered by 747s, and then the engineers scale it back. At the unclassified weapons museum at Sandia National Laboratories in Albuquerque, they have the initial US H-bomb, the one that was in the inventory for about two years. It's huge. It's just unbelievable, particularly when you consider it had to be - and was - dropped from a B-52. If you look at the models of how it was scaled down, you can get an idea of what the Iraqis did. They were also looking at more advanced technologies. They had work progressing on the production of lithium 6, which is necessary for tritium for thermonuclear boosting.
The Iraqis had already validated their design work by testing various weapons components. Here again is something technical that is important to understand, and, for a lot of people, very difficult to understand. As long as you are not interested in developing the latest cutting edge multi- stage fusion device, it is no longer necessary to test weapons by taking a bomb out and setting it off. Weapons are tested at the component level, with inert material, and with computers. There has been tremendous progress in this area, so much that it misleads people. For example, the computer that I carry around, a Macintosh 140 Powerbook, with 8 megabytes of memory, etc., has more computer power than was available to US scientists doing weapons design work in 1967, when our major nuclear program went ahead. We managed to obtain the Iraqis' hydrodynamic codes, which are the simulation codes of how material behaves under extreme pressure. A team of modern weapons designers that looked at them remarked on how primitive they were, but we had the people who had worked in the US and British programs in the early days of the programs look at them as well. We asked them to baseline the Iraqi codes against codes that were used in the 1950s and 1960s. The codes that the Iraqis used are freely available - they run on PCs - and they are much, much better than the codes that were available to weapons scientists then.
We also have certain expectations about what a weapons program involves, and I think this is what fundamentally misled the intelligence communities in a number of countries. When the inspection team first went into a place called al-Atheer, which was the major weaponization site of the Iraqis, one of the inspectors came up to me after a few hours. He said, "You know, you guys are absolutely wrong, this could not be what you think it is." To say the least, I was interested, so I asked him why. He said, "I can tell you, you cannot work with radioactive material in the spaces they have designed and meet the latest OSHA standards." (For those non-Americans in the room, I should explain that OSHA standards are the Occupational Health and Safety Standards that govern US industry.) I swear, he was serious. Now this was someone who was about 35, who worked at one of the US weapons labs, and was not a weapons designer so much as a bureaucrat. His major working life had been in that period when, in fact, we had stopped designing weapons and started worrying about OSHA and EPA regulations. But there is a lot of that. The first time we found a major weapons test bunker, a lot of the scientists said, "They're not making use of all the data. This couldn't be the real site, because there are not enough data collectors and data streams coming off the site." There again, we took them back and had them look at old US weapons testing sites.
If you look in Dialogue, or any of the other databases today, you can find vast amounts of information about weapons designs that were abandoned or not pursued, or are now so outdated that they are now in open literature. A classic case of this is the Scud. The Scud is nothing but, with the exception of a small change in the guidence system, a V-2 rocket. Now you can buy a full set of V-2 plans in Washington or at the Imperial War Museum in London. It's available. Peter Zimmerman has called this "bronze medal technology," as an analogy to Olympic medals. And there is a huge amount of that. Last September, the British home office was fielding applications for four hundred Russian salesmen to go to the major British airshow to sell military parts. Among other things, they were offering to sell, was satellite photography with 10 meters resolution. And, if you wanted to pay a little more, you could get it at half that resolution. There is a tremendous amount of information out there about nuclear, missile, and chemical weapons technology that does not have to be reinvented and tested. The Iraqis understood that.
Components vs. SystemsWhat is the lesson we can draw from this? Certainly, it is that nuclear proliferation is now possible by more routes, more states and, most importantly, with less warning than has recently been thought probable.
Why is this true? One reason is that there is a rising level of technological wealth and managerial skill. Let me stress the last part, managerial skill. The Iraqis defeated the export control regimes, not because of the ineptness of the regimes or because of corruption, but because they learned that they did not have to buy a final assembled instrument with everything they wanted. If they could buy it in parts from three different countries and had the proper project management skills, they could put it together themselves. In fact, if you want to look at scientific talent in developing countries, the most important talent for a nuclear program is not physicists. In fact, the Iraqi program would have gone a lot better is they had locked the physicists up (because they spent too much time thinking about better ways to do things). It is the mechanical engineers, electrical engineers, chemical engineers, project management types that make a difference. Even in countries that we look at as being poor and disorganized, these types of engineers are there in large numbers, and there may be some sectors that are very, very advanced.
For example, when the invasion of Kuwait took place, General Motors was about to establish a very large truck self-assembly plant in Iraq to serve all of the Middle East. General Motors tried to sell the Iraqis a technology enhancement project, in which they would train welders, lathe operators, and other technicians. The Iraqis told GM to send 12 of its best production engineers and they would let them see what they were doing, just to see whether Iraq needed this package. The engineers' assessment of the current state of Iraqi engineering essentially said that the Iraqis were making three-axis computer numerically controlled machine tools of their own, and doing electron beam welding. Incidentally, these are all nuclear- applicable technologies. The GM report listed all of the technical skills that Iraq had and concluded that Iraq did not need GM's help. Other countries in the developing world have the same type of management cadre that Iraq has used so well.
Information AccessInformation access is global. The Iraqis managed to access every data bank. They did not send graduate students exclusively to one country. Instead, they placed graduate students in the same field in three and four different countries. That is extremely smart, because we don't teach alike. When I taught, I tended to use a certain set of references. If you go to a German classroom, or a French classroom, a Russian classroom, you will discover that everyone presents a slightly different view of reality. So, if someone in hydrodynamics has been trained in the UK, France, or the US, they have parallel views of reality which are slightly different because of access to different information. The Iraqis did that, and they did it consistently, in every one of their fields. This does tell you that information is global. There is no hope of controlling information, particularly if what you are after are not the most recent scientific developments, but those that are expedient. For example, there is so much in the open literature today about neutron initiation, one of the keys of setting up an explosion, or X-ray focusing, which is necessary for advanced nuclear weapons. What's not in the open literature can be deduced by anyone who can read and understand what is there. This is true in many fields.
The Problem of AnalysisAnother reason why proliferation is possible by more routes with less warning has to do with the intelligence services themselves, in two different ways. One is what I call "mirror imaging." Has anyone ever wondered why the US Polaris submarine and the equivalent Russian ballistic missile submarine have exactly the same number of tubes? God did not create a rule that said that there can only be X number of tubes on a missile submarine. This happened as a result of one side looking at the other, and drawing conclusions about what is necessary. We tend to look at the way we have done things and assume that everyone will do it that way. There was a lot of mirror imaging in the intelligence communities about what was going on. This process is continuing as the US national laboratories are trying to rapidly become nonproliferation institutes, and a lot of strategic Soviet analysts are trying to retool themselves as proliferation specialists.
People are taking the skills that they acquired during a lifetime looking at a Soviet program, or an American program, and trying to apply them. There is a tendency to think that what you knew was necessary there, will be necessary here. Watch out.
The second, and far more dangerous, intelligence problem is that we know from the Iraqi program that they had learned to defeat national technical means (NTM), i.e., signal and satellite intelligence. There had been two major routes. There was a major leakage from the US intelligence program on key satellite data. It had been passed to the Soviets by Christopher Boyce. For one reason or another, the information had been passed to the Iraqis, and they had learned with great sophistication how to conceal their activities from US satellite detection. Also, during the Iran-Iraq War, the US shared strategic intelligence information with the Iraqis. The Iraqis were able to look at US data, and they could tell from the data how it was obtained. They were smart enough, with the help of East Germans and others, to devise ways of beating the system. The intelligence community has invested billions of dollars in NTM and doesn't like to hear this. Unfortunately, though, it really is true that NTM is relatively easy to defeat if you can spend the money and you have the skills. At the time the war broke out, there was very low-grade human intelligence that there was a possible nuclear site at Tarmiya, which in fact was a major site where Iraq's calutron program was going to be assembled. The photo interpreters had looked at it and didn't believe that it was a major site because it had no security fences around it, no anti-aircraft around it, and no electrical power going into it. It looked like another one of the industrial complexes you find a lot around Baghdad. The site was put on secondary target list. After it got hit by a pilot who had some leftover ordnance, it was 48 hours before they did a photo post-strike recon of the site. The pictures were just amazing. There were five huge cranes on the site and something that looked like giant frisbees being pulled out of the damage and over 100 identifiable Iraqis on the site. Another good rule of the US military is when something goes on during combat that you don't understand, wipe it out because then you can worry about it afterwards. So they did, and then they sent the inspectors in after the war. Among other things, it turned out the Iraqis had run 150 megawatts of power 25 km underground, because someone had told them that one of the rules a photo interpreter uses is power. There was no guard fence there because there had been a military exclusion zone 50 km around the site. The lesson that we can draw from this is that we should not count on intelligence by itself to give us warning. This is one of the most disturbing lessons of the Iraqi experience.
Another reason proliferation is possible by more means is that a significant leakage of expertise and material across borders is a reality. Centrifuge technology, very advanced centrifuge technology, came out because someone was willing to pay money to someone who was willing to take money. Before we start pointing fingers at Germans, or even worse, at Russians, or Kazakhs, or whoever, let's remember we are in the midst of downsizing our own nuclear weapons program in the US. People do not have to get on a plane and fly to point A. All they have to do is plug a Macintosh in with a modem and they can work any place in the world. A lot of the data exchange on the centrifuge program into Iraq did not come from people going to Iraq. It was done in other places. This also makes it much harder to track and identify, since people aren't necessarily flying into countries where they shouldn't be.
There is another problem with material. One of the critical components for the Iraqi centrifuge program was maraging steel. I often tried to convince the Iraqis to tell us who their supplier was. Once I tried the argument - "Don't tell me you can't rat on people who have been your friends. They charge you, you pay a risk premium, and everyone knows that you pay a risk premium because there's a risk that the supplier will be exposed. Go ahead and tell us." Instead, Jaffar pulled out the invoice (without the supplier's name on it) and, sure enough, they had only paid a couple of cents per pound over the world market price for maraging steel. That tells you that for a lot of commodities there is, in fact, not much of a risk. And, maraging steel is one of the more strictly controlled. That is something you have really got to worry about. Material does get out and will continue to get out.
Managing ChangeThings are changing in a way that we have to worry about because the changes are making the situation worse. One problem is the break-up of the former Soviet Union. In talking to one of Gorbachev's advisers, who is now a senior academician, I had gone through the reasons I am concerned about the break-up of the former Soviet Union, which include the lack of export controls and problems with expertise. We are talking about dismantling 15,000 weapons, which is a process that at the rate of 1,000 to 2,000 weapons a year will take at least 15-20 years. It isn't really the scientists you have to worry about, it's the guy doing guard duty at night who is getting paid virtually nothing. I went through all the standard sets of arguments. The official came up later and he said "You know, it is really worse than you thought it was. I thought I would never bemoan the passing of the control of the KGB. But crime and corruption is much worse today than ever before. I've got plutonium in my lab that I'm not sure will be there when I return." I took that with a grain of salt, because there might be other reasons for those statements. The fact of the matter is that we have a long-term issue there.
The material presents another risk. Somewhere in the order of 4 kg to 12 kg of plutonium is sufficient for a nuclear device. That is an amount that is smaller than an American softball, if it were in a single source. If you look at how drugs are moved around the world, you realize the problem of preventing the transfer of small amounts of material. There are ways to smuggle this material that are not technically difficult.
One of the failings of the last administration, which the current administration better get on quick, is that for over six months the South Africans have been willing to sell their entire HEU supply to the US. We should have bought it. The US ought to get that HEU out of there, every last gram - and, as a purchaser of the entire supply, you can demand rights of inspection to be sure you have every last gram. It is going to be very irresponsible if we go many more months without having gotten it because we may lose the opportunity to buy the entire supply. I should emphasize that the South African Government has been completely forthcoming on this matter. The delay has been entirely on the side of the US government.
In addition to material, there is also the problem of the transfer of technological skill and expertise. In many ways, because of the nature of the Soviet and American programs, you don't have to worry about the top scientists. They've been working on the most advanced weapons design. If you asked the Edward Tellers of these programs, the Livermore senior weapons designers, to do a gun assembly design, which is an easy design that almost anyone could do with a little thought and reading, they wouldn't have designed such a weapon in years, if ever. The technicians, on the other hand - who know that when you are shaping plutonium there are certain properties that are not described entirely accurately in the open literature - are the people you do have to worry about. This is where the expertise is leaking out.
Another problem is the political status given to those possessing nuclear weapons. Every time I hear a British politician saying that they have to keep their nuclear weapons, not because they have any reason for them, but in order to maintain their seat on the UN Security Council, I cringe at the implications such a statement can have. The same thing is true of Secretary of State James Baker's rush to offer the Russians large amounts of money without making a similar offer to Ukraine; this is part of the problem we are now having with the Ukrainians. By giving status, or appearing to give status, to countries that have nuclear weapons for that reason alone, we are in great danger of leading others to conclude that they should either not give up the nuclear weapons that they have or think of acquiring them.
During most of the regime of security known as insecurity, i.e., the "Cold War," we had a system that was extremely intolerant of movements across borders. With the exception of a handful of crises like the Cuban Missile Crisis, the Berlin Wall Crisis in 1959-60, it was a regime of remarkable stability. We are seeing that regime being disassembled at a rate that is without parallel, at least in the last 50 years. If Iraq had not invaded Kuwait when it did in 1990, but if it had waited until 1994, not only would Iraq have had a nuclear weapon, but we would not have had the troops or the ordnance that came to the aid of Kuwait deployed forward in Europe. Those forces are being pulled back.
States today see the US disassembling its military might at a rate that is unparalleled since the end of World War II. They see a lack of an articulated US defense policy that explains why it is developing the force structure that it is, and they ask, where are we? What sort of world are we moving toward? And they start to raise questions. There was a parallel to this in the '50s, up through the early '60s. For example, Sweden and Switzerland had nuclear research programs, and the Swedish one was more than just a group of academics sitting around. I'm afraid that unless steps are taken in 1995 at the NPT Review Conference, and unless other steps are taken soon, we will see a number of states start exploring the nuclear weapon option again, and this time it is going to be different. They can do all the non-nuclear work essentially out in the open. And they only have to worry about the acquisition of special nuclear material at the very last stage and that, for a series of technical reasons, is not going to be very difficult to do. Those things working together, in addition to what Iraq already taught us, mean that for proliferation we are entering a period far more dangerous than anything we have seen before, at least in the last thirty years. And that is really an optimistic view to conclude all of this.
Questions and Answers: