[NCI Logo]


It has recently been revealed that Iraq had initiated a crash program in August 1990 to build a nuclear weapon within eight months, by recovering highly-enriched uranium (HEU) metal from its inventory of French- and Russian-supplied research reactor fuel. Although the research reactor fuel was under International Atomic Energy Agency (IAEA) safeguards, it was only being inspected twice per year, leaving open the possibility that the conversion of fuel could have been completed in the interval between visits.

It remains an open question whether Iraq would indeed have been successful in obtaining enough HEU in weapons-usable form for at least one nuclear device prior to the IAEA inspection scheduled for April 1991, if the Gulf War had not intervened. This issue depends on three interrelated factors: (1) the minimum amount of HEU required for a weapon of a design compatible with the purpose intended for it, (2) the rate of fuel processing that could be achieved and (3) the level of progress made in other aspects of the weaponization program. In addition to providing historical insight, an assessment of these factors is relevant to the determination of how close Iraq may be to possession of nuclear weapons today, with its existing level of expertise and capability, should it be able to acquire the requisite amount of fissile material.

The inventory of HEU in research reactor fuel possessed by the Iraqis at the time of the Gulf War was as follows (not counting materials at 36% initial enrichment or below):

1. HEU requirements

How much HEU would Iraq have needed for a weapon? One of the most important weapon parameters is the nuclear yield that can be guaranteed with high probability. This is determined by the efficiency of the weapon (the fraction of material fissioned), which depends strongly on the degree of supercriticality (the number of critical masses present) at the time of the initiation of the chain reaction. The desired overall weight of the device places a constraint on the yield obtainable from a given quantity of fissile material.

A good example of a basic, largely declassified implosion design accessible to Iraq is the weapon dropped on Nagasaki, with HEU substituted for plutonium in the core. With a 10.2 centimeter-thick reflector of natural uranium, a material which Iraq had in plentiful supply, about 18.5 kg of HEU enriched to 93.5% U-235 (or its equivalent) would be required.

Utilizing HEU in an implosion device would result in a greater yield than the same amount of material in a gun-type device (such as the weapon dropped on Hiroshima). According to Carson Mark, former head of the theoretical division of Los Alamos National Laboratory an HEU implosion device capable of attaining a two critical-mass assembly could generate a yield of around 10 KT HE (high-explosive) equivalent, on the order of the Hiroshima and Nagasaki weapons. This degree of supercriticality can be attained with one critical mass of material at normal density in a relatively unsophisticated implosion device which would compress the core by a factor of 2 (1.41), which is considerably less than the factor of two achieved in the Nagasaki device.

Such a design would be compatible with the stated goal of the "crash program," which was to produce weapons components from the research reactor fuel by purifying 20 kg of 93%-enriched equivalent HEU metal in six months. This design is also consistent with the the fact that Iraq planned to use a composite reflector/tamper of natural uranium and iron, materials which were readily available.

To meet the goal of the crash program, it would have been necessary to process the 26 kg (23.3 kg of 93%-equivalent) fresh fuel and lightly irradiated fuel, provided the overall losses in processing could be kept to approximately 20% or below. Otherwise, some of the heavily irradiated fuel might also have to have been used. The 3.5 kg of 93-equivalent HEU contained in fuel of unknown burnup may also have been easily accessible.

Using another material as a reflector would have enabled Iraq to reduce the amount of HEU necessary. With a beryllium reflector of 15-cm thickness, such a device would only require 12 kg of 93% equivalent HEU. No evidence has been uncovered that Iraq had been planning to use beryllium as a reflector, however. Furthermore, the critical mass reduction achieved by using a beryllium reflector would not alone result in increased efficiency, without simultaneous improvements in other aspects of design, such as the achievable compression. This is because the lower critical mass of the HEU-beryllium system is a consequence of the increased moderation of neutrons, which would tend to lower the efficiency of the weapon.

2. Processing time

The IAEA has argued that conversion of the reactor fuel would have been so difficult that Iraq could not have completed the task in six months and would have needed from 12-18 months instead. Therefore, the IAEA asserts that the diversion would have been detected by the routine bi-annual inspection. However, this conclusion, which is based in part on information provided by the Iraqis themselves, appears to be based on an overly pessimistic assessment of Iraq's technical capabilities.

The extraction and purification of uranium metal from aluminum in the fresh fuel would be "peanuts ... a good undergraduate chemistry lab exercise," according to experts on research reactor fuel at the U.S. Department of Energy. One series of straightforward (and widely known) chemical steps which could be used are: dissolution in caustic-nitrate solution, precipitation of uranium as ammonium diuranate, drying, denitration, fluorination and bomb reduction with magnesium. Solvent extraction would not be necessary. This applies to fuels based on uranium aluminide (the French-supplied fuel), uranium-aluminum alloys (the Russian-supplied fuel) or oxide dispersion fuel. Batch sizes would have to be limited to less than 1 kg U for criticality reasons, but that would have little impact on the time scale because so little material was required in total. The most difficult step would probably be the reduction of uranium tetrafluoride to metal, since it is an energetic process requiring a sealed ceramic crucible. However, it is known that Iraq already had produced hundreds of kilograms of natural uranium metal using this process.

Furthermore, one pass would likely have been sufficient for the purification of uranium from the unirradiated fuel. Processes for Mo-99 extraction from irradiated isotope production fuel targets exist in which uranium metal, sufficiently pure for refabrication into new fuel, is separated from aluminum and fission products in one pass.

The chemical operations involved here are not associated with large uranium losses. Descriptions of the process dating from the 1960s indicate that losses can be kept below a few percent.

Because the specifics of its initial irradiation have not been publicly disclosed, it is unclear whether the "lightly irradiated" fuel would be substantially harder to reprocess than the fresh fuel. However, the French have said that the fuel was loaded into the Tammuz reactor and brought to power "until it became radioactive." U.S. experts believe the irradiation time was less than one day and also that the reactor was not operating at full power at the time, since it usually takes several days to bring a reactor to full power; they believe that the fuel was irradiated at no greater than 1/10 of full power; according to this information, the average burnup of the fuel would have been about 4 kWD/kgU. Thus the initial radioactivity level of the fuel was quite low, and would have decreased significantly after a couple of years of cooling. Actually, there would have been approximately 0.01 Ci of Cs-137 per kilogram of fuel, a quantity which could be handled in a glovebox with local shielding without causing acute radiation sickness. It is therefore likely that the "slightly irradiated" fuel could be processed without the use of a hot cell, and therefore would not have been appreciably more difficult to handle than the fresh fuel. One solvent extraction cycle, which can reduce residual fission product concentrations by a factor of 100 or greater, would probably have been sufficient. It is known that Iraq was capable of performing PUREX-based solvent extraction, since it previously had separated a small quantity of plutonium from irradiated reactor fuel.

Therefore, 23.3 kg of 93% equivalent HEU would be available with relatively simple chemical processing, meaning that overall process losses of up to 15% could be tolerated to meet the 20 kg goal quantity, and greater losses could be compensated for by increasing the thickness of the reflector in the weapon. U.S. experts have estimated, on the basis of process flowsheets, that at least 1 kg of HEU could be obtained per week based on a single shift (40 hrs/week), which means that the job could be completed in under six months. Increasing the number of shifts and process lines would lead to corresponding reductions in the conversion time.

Even the "heavily irradiated" fuel would not have been inaccessible in the time available. Iraq had some experience with the PUREX process and the separation of HEU from this fuel could have taken place concurrently with the processing of the other material, since that did not require use of a hot cell.

The IAEA has also said that the Iraqis planned to build a 50-centrifuge cascade to increase the enrichment of the 80%-enriched material to 93%. Since Iraq did not have a single centrifuge assembled yet for this purpose, this undertaking would have taken "at least one year...," according to the IAEA. Thus even if the Iraqis could have converted the research reactor fuel relatively quickly, this enrichment step would have clearly caused a substantial delay.

However, this enrichment plan would appear to be completely illogical in the context of a crash program seeking a single weapon, since the Iraqis already had more than enough material for a weapon of the type described above. For a given core-reflector combination, the critical mass of 93%-enriched HEU is only 20% less than that for 80%-enriched material. The IAEA claims that the Iraqis were so concerned about process losses that they could not afford to neglect this potential reduction in critical mass.

It strains credulity to believe that the Iraqi nuclear establishment would have been so foolish as to embark on this strategy. Re-enrichment of all the 80%-enriched material in the fresh fuel would increase the amount of 93%-equivalent HEU by 0.78 kg (7%), unrealistically assuming one-hundred percent recovery (0.8/0.93 x 13.7 kg - 11.0 kg). such optimal conditions would have increased the amount of 93%-equivalent HEU from the fresh and lightly irradiated fuel by a mere 3%. The losses associated with the several additional processing steps necessary for enrichment would in all likelihood have erased or even overwhelmed this gain.

3. Status of weaponization program

The ability of Iraq to use the limited amount of HEU available to it from the research reactor fuel depended on its successful development of the implosion package to the specifications outlined above. A recent report states that their implosion program had made "remarkable progress" in HE testing, "but no evidence has come forth indicating that compression levels reaching approaching a factor of two" had been reached. However, this report does not clarify whether the Iraqis had aquired the capability for a compression of 1.4, which some experts assert is not a particularly challenging achievement. Furthermore, an HE testing program would be relatively asy to conceal, even today. In light of the considerable amount of information which Iraq has withheld in the past, it is far from obvious that the full extent of its nuclear weapons program has even now been revealed.

Given the fact that no nuclear-weapon components from Iraq's original program have ever been recovered, and that Iraq maintains the same nuclear expertise that it had before the Gulf War, Iraq today might lack only the fissile material needed to rapidly assemble nuclear weapons.

Dr. Edwin S. Lyman
Scientific Director
November 30, 1995


R. Blanco and C. Watson, "Head-End Processes for Solid Fuels," in the Reactor Handbook, Volume II: Fuel Reprocessing, 2nd Ed., Interscience Publishers, New York, 1961.

R. Lewis and F. Martens, "Technical and Economic Assessment of the Use of Highly-Enriched Uranium In Critical Experiment Facilities, Research and Test Reactors, and Power-Reactor Prototype Facilities," RSS-TM-3, Argonne National Laboratory, June 1977.

M. Hibbs, "Experts say Iraq Could Not Meet Bomb Deadline Even With Diversion," Nucleonics Week, August 30, 1995.

International Atomic Energy Agency (IAEA), Statement to Security Council Briefing by Ambassador Ekeus on 1995-08-25, August 1995a.

IAEA, Statement to the 1995 IAEA General Conference by Director General Blix, September 1995b.

J. Carson Mark, "Some Remarks on Iraq's Possible Nuclear Weapon Capability in Light of Some of the Known Facts Concerning Nuclear Weapons," Nuclear Control Institute, Washington, D.C., May 1991.

H. Paxton and N. Pruvost, "Critical Dimensions of Systems Containing 235U, 239Pu and 233U," 1986 Revision, LA-10860-MS, UC-46, Los Alamos National Laboratory, July 1987.

UNSCOM, "Annex to the First (sic) Report on the On-Site Inspection In Iraq Under Security Council Resolution 687 (1991), 22-30 September 1991: Al-Atheer Progress Report."

Please send comments to Dr. Edwin Lyman at nci@access.digex.net.

[Saddam & the Bomb Page]] Saddam & the Bomb Page[What's New Page] What's New Page[Home Page] NCI Home Page