As of August 2006, Iraq Watch is no longer being updated.
Click here for more information.
IRAQ'S SHOP-TILL-YOU-DROP NUCLEAR PROGRAM
by David Albright and Mark Hibbs
Bulletin of the Atomic Scientists, Vol. 48, No. 3
In a series of articles that began in March 1991, we have tried to separate fact from fiction about Iraq's ability to build nuclear weapons and to produce material to fuel them. After exposing Iraq's efforts to enrich uranium and design an atomic bomb, U.N. and IAEA experts zeroed in on how Iraq put its program together. The basic answer is that along with determination and persistence, Iraq had a great deal of foreign help.
Iraq's "Petrochemical Three," the secret nuclear program conducted under the authority of its Atomic Energy Commission with links to the Defense Ministry and the Ministry of Industry and Military Industrialization, received massive infusions of money and resources. Like the Manhattan Project that built the first atomic bombs in the United States, Iraq's program simultaneously pursued a number of different technical avenues to the bomb. Not knowing which efforts would succeed, Iraq poured billions of dollars into its multifaceted quest.
Providing for these programs required the establishment of elaborate procurement networks in Europe, North America, and Asia. Like the technical quest, the procurement effort was carried out on many fronts at once. Diplomacy and secrecy were required, because few companies would knowingly supply a nuclear weapons program, or even a secret nuclear program that was ostensibly for civil purposes. Iraq showed great ingenuity in hiding its purchases behind such innocuous pursuits as automobile manufacturing, dairy production, and oil refining.
Some of the basic infrastructure Iraq needed for the program-factories, electrical supply, and power equipment-was easy to obtain from abroad. But the more specific the equipment Iraq sought, the more other countries' export controls began to bite. Crucial transfers of components were blocked.
To evade export controls, Iraqi officials divided equipment orders into innocuous subcomponents. They also tried to buy machines to make sensitive components in Iraq. They used middlemen to disguise the destination of equipment and materials they were importing and to assure Western suppliers that the products the suppliers were exporting would be used for peaceful purposes. And the Iraqis seized opportunities to obtain any uncontrolled equipment and material that was available, often in enormous quantities, regardless of whether they were ready to use it. That practice generated the impression that Baghdad was closer to having a nuclear bomb than it in fact was [see "Digging for Gold," below].
Inspections after the Gulf War revealed that in some cases, export controls did not suffice. Iraq's calutron program was successful largely because few of the materials or components were controlled. On the other hand, in its centrifuge program, Iraq blatantly circumvented Western controls on sensitive centrifuge design information and components. The failure to prevent the spread of this know-how-which is highly classified in all Western countries-partly explains why Baghdad's procurement effort remains the most sensitive aspect of its secret nuclear program.
Western companies and governments, particularly Germany, do not want to be embarrassed by public revelations about their involvement with Iraq's nuclear program. Firms fear that commercial secrets will leak out. But exposing the way Iraq tried to buy itself a nuclear weapons program is critical to determining where export controls failed-or worked-and what improvements are needed if the Iraqi experience is not to be repeated.
On January 12, 1992, armed with information from the German government, officials of the International Atomic Energy Agency (IAEA) accused Iraqi Foreign Ministry officials of failing to declare large quantities of materials and components Iraq had obtained for its gas centrifuge program. Maurizio Zifferero, head of the IAEA's ninth on-site inspection mission, told the Iraqis to produce a comprehensive statement about their procurement effort or face a long, slow process in which the agency would match Iraq's declarations, item by item, with Western governments' lists of what certain companies had sold Iraq.
The next day, at a meeting held at the Tuwaitha Nuclear Research Center, technical experts from the Iraqi centrifuge program gave the inspectors the most complete description of the program to date. The Iraqis acknowledged that they had imported German materials and components-and added that they had acquired 100 tons of maraging steel and other raw materials needed to manufacture centrifuge components.
The Iraqis claimed that they had destroyed the materials or had made them unusable. As proof, they took inspectors to the State Establishment for Mechanical Works at Iskandariya, 80 kilometers south of Baghdad, and showed them large amounts of powder, which the Iraqis said were the remains of ferrite (magnetizable) spacers for centrifuge motor stators, and flat ingots left after melting down maraging steel. Heat impairs the special, super-strong properties of maraging steel.
According to the ninth IAEA report, Iraqi authorities told inspectors that because export controls had been tightening, they had made "large procurements as opportunities presented themselves, even though they had no immediate plans for the materials in the quantities ordered." Although the Iraqis had not settled on a centrifuge design and were not ready to make large numbers of centrifuges when the war broke out, according to the report, they knew what they needed to make certain key components.
In the aftermath of the January 12 confrontation, a more complete picture of Iraq's procurement effort is emerging. Nevertheless, the full extent of foreign help and the identities of all important suppliers are not known. Most of the available information covers the centrifuge program. Less is known about the procurement efforts for the parallel calutron program, and for the third major branch of the technical effort, weaponization. Each success in evading export controls considerably shortened Iraq's path to the bomb. Conversely, each failure held the Iraqis back.
Proliferation food chain
Lewis Dunn, a U.S. nonproliferation expert, said at a public meeting in mid-1991 that the history of efforts to tighten nuclear export controls is like working down a "proliferation food chain." That is, the more important an item is for producing plutonium or highly enriched uranium or for making a nuclear weapon, the earlier it was covered by export controls. Currently, tight controls cover export of completed nuclear facilities or their basic units, such as gas centrifuges and their major components. Subcomponents and equipment and technology to manufacture major components are less regulated. These often have many uses and are officially referred to as "dual-use" items.
Iraq had only limited success in acquiring tightly controlled items. It was more successful in obtaining dual-use equipment. Iraq's export permit applications almost always listed civilian industrial uses for such equipment.
Some of the exports never made it to Iraq. In July 1990, German and Swiss customs officials stopped a shipment of special computer numerically controlled (CNC) machines for making endcaps and baffles of centrifuges. CNC machines are highly sophisticated, modern machine tools, which can be programmed on-line to perform complicated and precise machining of metal parts. Also confiscated were 888 forgings for endcaps and baffles that would be finished on the CNC machines. Earlier that year, U.S. and British customs officials caught Iraq trying to smuggle detonation capacitors from CSI Technologies in California [January/ February 1992 Bulletin]. These capacitors, which were of military standard and specification, would have been ideal for developing the implosion mechanism for a nuclear weapon.
The embargo imposed on Iraq after it invaded Kuwait in August 1990 kept a shipment of aluminum piping from leaving Switzerland. The material might have served to connect gas centrifuges in a "cascade" or assembly. There was enough piping to connect 5001,000 centrifuges.
But a great deal of technology, equipment, and materials got through before the embargo was imposed. Most of it never raised eyebrows, as business with Iraq was considered legitimate. The West favored Iraq in its bloody, eight-year war with Iran. During the last half of the 1980s, the U.S. government approved the sale to Iraq of $1.5 billion worth of computers, electronic equipment, and machine tools that could be used in its nuclear, chemical, and ballistic missile program. Europe supplied even more. France sold Saddam Hussein $5 billion worth of aircraft, missiles, and armor during the war. A U.S. customs official involved in stopping the sale of the detonation capacitors to Iraq summed up the situation: "The enemy of my enemy is my customer."
In this climate, many Western companies and individuals felt justified in exporting their products to Iraq and rarely questioned Iraqi assertions that the goods were for peaceful use. That changed only after Iraq invaded Kuwait and Iraq's programs to build nuclear weapons and other weapons of mass destruction were exposed after the Gulf War. The exposure of Iraq's military programs has stiffened the resolve of companies that contributed to Saddam's effort to deny that they ever had anything to do with Iraq's clandestine nuclear program. But the specific nature of some of the equipment exported should have aroused the suspicions of companies and export control authorities. According to the eighth IAEA report, in some cases "the presence of application-specific fixtures removes most doubt as to intended use." Some of the companies supplying this equipment might not have known that Iraq's nuclear program was the final customer. But according to the report, the intermediaries who dealt with Iraq "must have known (or could reasonably have inferred) the intended uses." One inspector said that Iraqi officials insisted on successful demonstrations of centrifuge component manufacturing equipment before they accepted shipment. In these cases, one IAEA official asks, "How could they not have known?"
Most of the emerging information has to do with Iraq's gas centrifuge enrichment program. Many different companies and individuals were involved. Some may not have been aware that they figured in Iraq's nuclear ambitions. Others must have known. Some acted as important hubs in Iraq's purchasing networks. Although most of the key links were in Germany, where until recently both export laws and their enforcement were relatively lax, other countries were involved as well.
Interatom. This wholly owned subsidiary of the German firm Siemens AG was folded into its parent company in 1990, because its key nuclear markets, building breeder and high-temperature gas-cooled reactors, dried up during the late 1980s. But one of the last contracts it obtained, in mid-1989, was from Iraq's Industrial Projects Company (IPC) to build a workshop for "tube processing." Western intelligence organizations believed at the time that IPC was a procurement operation for Iraq's clandestine centrifuge program. Iraq might have approached Interatom because it also engineers cascades for Urenco, the European centrifuge enrichment consortium. In 1990, Iraqi officials sought from Interatom know-how on design and production of equipment used in removing enriched uranium gas from the cascade.
According to Interatom documents, the contract involved three parts: "training of Iraqi technicians, planning and construction of a workshop complete with appropriate infrastructure, and supply of processing and testing equipment." Strabag AG, a major German construction firm, was awarded a contract to build the facility.
Germany's export control authority, the Federal Economics Office (known as BAW), approved the first stage of the program, which was carried out later in 1989. But in mid-1990, after BAW belatedly learned of the link between Iraq's centrifuge program and IPC, it cut off the second stage of the program. On August 2, 1990 (the day Iraq invaded Kuwait), BAW reiterated that "in light of the risk that the training and workshop supply contract would pose to the international reputation of the firm," Interatom should "immediately break off contact" with IPC and not supply equipment to the factory under construction in Iraq. BAW reminded the firm of pending amendments to Germany's export laws, which would "extend punishment to [companies] that unintentionally, but carelessly, aid nuclear weapons programs in other countries."
Interatom officials have publicly denied that the company ever contributed any nuclear-related technology or equipment to Iraq's centrifuge program and have expressed doubt that the facility could have been used in a uranium enrichment program. But inspectors determined that the workshop in question, code- named B-01, was intended to be an integral part of Iraq's centrifuge manufacturing program at the Al Furat factory. The ninth inspection report says that upon completion, B-01 would have been used for manufacturing cascade pipework and for assembly and mechanical testing of centrifuges, and that one area was possibly intended for a 100-machine cascade.
A sketch of the outside of B-01 prepared by the IAEA after several inspection visits is nearly identical to a sketch of a building on a map given to Siemens personnel to help them find the site [see opposite page]. A senior IAEA official said he had no doubt that the Siemens map gave directions to Al Furat.
Inwako. In early 1989 Western intelligence learned that Inwako GmbH, a firm directed by the German arms dealer Simon Heiner, had masterminded the supply of 125 British-made ring magnets to Iraq for use in the top bearings of centrifuges. According to a German intelligence report, Bonn first learned of the Inwako export in April 1990.
Britain learned of Inwako's attempt to help Iraq in 1989, when a British firm solicited by the Germans to supply the magnets alerted export control officials. Britain then tracked Inwako's effort to get the magnets from a second British firm, Endshire Export Marketing, which was willing to supply them. Suspecting that the magnets were for Iraq's nuclear program, Britain let the shipment proceed. Because not all magnets were alike, experts suspected that a second order to the British firm would be forthcoming for the thousands of magnets needed for serial production of centrifuges. That order would have been denied, but not before British experts learned what design Iraq had chosen for the top bearing of its centrifuges. On the eve of the Gulf War, Iraq had not made a decision about the top bearing, and accordingly did not place an order for the magnets.
Inwako is under investigation in Germany for other procurement deals involving Iraq. Correspondence between Inwako and the British firm Technology Development Group (TDG, page 34) shows that TDG had approached the Germans to supply precursor chemicals useful for the production of poison gas.
H&H Metalform. This German producer of vertical flow-forming machines, which cold-press thick metal cylinders into thin-walled tubes, also denies any connection with the Iraqi nuclear program. But H&H supplied several flow-forming machines to Iraq between 1987 and 1989-and according to the eighth IAEA inspection report, an H&H machine was used to make a small number of maraging steel rotor tubes for centrifuges. In a company report from April 1991, H&H claimed that the flow-forming machines it sent to Iraq could not be used or altered to make centrifuge rotor tubes. In an interview in February 1992, Dietrich Hinze, technical director of H&H, expressed "surprise" at the inspectors' findings.
The inspectors found seven other H&H flow-forming machines at the Nassr State Establishment for Mechanical Industries, an im- portant support site for the nuclear program, and two more at a subsidiary establishment at Al Schaula. The Iraqis say these were used to make rocket bodies. The ninth report concluded that with appropriate fittings, all of them would have been suitable for making centrifuge rotors.
Sets of specialized mandrels, metal cylinders that support the maraging steel while it is being flow-formed, are important accessories for flow-forming machines used to make centrifuge rotors. These are not difficult to manufacture, but the finishing mandrel must be precisely engineered to the exact inside diameter of the rotor. According to one IAEA expert, H&H is believed to have supplied some of the mandrels, but H&H denies this. According to a German federal prosecutor investigating H&H, even had H&H supplied centrifuge-specific mandrels, that would not have constituted a violation of German export laws at the time. However, three H&H officials, including Hinze, have been arrested for illegal exports of missile-specific fittings for flow-forming machines.
Recent information suggests that H&H, either inadvertently or deliberately, funneled related technology and equipment to Iraq's nuclear program. H&H obtained for Iraq CNC machine tools made by another German firm, Neue Magdeburger Werkzeugmachinen GmbH, which Hinze said H&H represented. According to the eighth inspection report, two of the Magdeburger CNC machine tools found in the centrifuge program were equipped with fixtures designed specifically for machining centrifuge components.
Technology Development Group (TDG). Beginning in 1987 this British firm, which was co-directed by Safa Al-Habobi, an Iraqi active in the procurement process, was involved in all the business H&H transacted with Iraq. According to Hinze, TDG received a commission on all the equipment H&H sent to Iraq. The arrangement allowed H&H to "get involved in projects that were out of our line," Hinze said in an interview. But he said that his company's business contacts with TDG were aboveboard and that "centrifuges were never discussed."
A German government report from late 1990 asserted that H&H was "50 percent owned by parties in Iraq." H&H has denied this, claiming that funds transferred to H&H by TDG and Iraqi organizations were advance payments for engineering work.
British investigators believe that under Al-Habobi, TDG may have served the Iraqi nuclear program. They believe that Al-Habobi, who was also in the top management of the metalworking firm Matrix-Churchill, Coventry, organized the procurement of a wide array of dual- use machine tools, spare parts, and materials for Iraqi military programs during the late 1980s. Last year, the Bank of England and the U.S. Treasury Office of Foreign Asset Control froze the assets of TDG and Matrix-Churchill pending conclusion of an investigation. In the United States, investigators are still combing through a list of over 1,000 pieces of equipment, subcomponents, and materials Matrix-Churchill obtained for Iraq.
Matrix-Churchill offered a lucrative, long-term contract for a tool shop for automobile parts manufacturing to the Swiss metal-working firm, Schmiedemeccanica SA, Biasca, Switzerland, on the condition that a financial company, Durand Properties, buy up to 18 percent of Schmiedemeccanica's stock. Soon after, Durand Properties became wholly owned by Iraqis. Meanwhile, the Iraqi procurement firm IPC placed an order with the company for 1,000 metal forgings that was intended as the first part of an order for 4,000. The forgings, which look like transmission gears, matched closely the dimensions of forgings that would be inserted in a CNC machine and turned into centrifuge endcaps and baffles. These were confiscated by customs officials at the Frankfurt airport in July 1990. The forgings are believed to have been intended for use in CNC machines ordered by Iraq but made by Schäublin AG, another Swiss firm. Five of these machines were seized in 1990.
Bruno Stemmler. Another apparent link between H&H and Iraq's centrifuge effort was Bruno Stemmler, a former employee of MAN-Technologie AG and Dornier GmbH, two companies which have been important partners of Urenco, the European enrichment consortium. Stemmler, a former centrifuge engineer, said in a series of interviews that he visited Iraq in the late 1980s in the company of H&H management and of Walter Busse, another former MAN employee.
Stemmler said that H&H wanted to create new markets for its flow-forming machines, primarily in the areas of irrigation and solar energy. Stemmler hoped for a solar energy contract with the Iraqis. But he said the Iraqis were also having problems with the corrosion of maraging steel after flow-forming and wanted more help from H&H.
Stemmler said he did not intend to help the Iraqis with centrifuges, but they asked him about his former career as a centrifuge expert and then about centrifuges. On his trips, Stemmler met with up to 15 Iraqi centrifuge design engineers; H&H personnel, he said, were not present during those discussions. He visited a site where he saw what he believed was a centrifuge mechanical test stand, and he advised the Iraqis on a problem in the vacuum system. He said Iraq had a good team of centrifuge specialists but not enough hands-on technicians. Stemmler received a small amount of money from the Iraqis for his visits, via H&H. Stemmler offered Iraqi officials his patented knowledge about oxidizing maraging steel to prevent corrosion by uranium hexafluoride. A high quality oxide layer is important in developing centrifuges with a long lifetime. Stemmler said he wanted H&H to sell his oxidation patents, but Hinze asserts that he never knew Stemmler had any such patents.
Stemmler also says he sent a hand-written note through H&H to Iraqi officials containing supplementary information to his patent on obtaining a good oxide layer on a metallic surface. He said he gave Iraqi experts enough information to start oxidizing maraging steel rotors. But when shown pictures of Iraqi oxidized rotors recently, Stemmler said they had oxidized them incorrectly. He said he knew at the time of his visits that Iraq would buy an oxidation furnace, which it did not have at the time. Iraqi officials told him that they needed one not just for centrifuge parts but for other maraging steel items as well.
Stemmler might have provided more assistance to the Iraqi centrifuge program, but his collaboration was stopped by a 1989 German investigation of his activities in Iraq as well as those of Busse and H&H. When he halted his activities, he had begun preliminary design work of a bottom bearing, a crucial centrifuge component. Stemmler said this work was done without a contract from Iraq, and he believes the Iraqis gave him this task to test his knowledge as well as to determine which subjects he could advise them about. He never concluded an agreement with the Iraqis on the sale of his oxidation patents.
The complex links among companies and individuals are only part of the puzzle. A piece-by-piece examination of the centrifuge procurement effort shows that Western export controls made it difficult for Iraq to obtain large numbers of finished components-but they contained enough loopholes to allow raw materials, know-how, subcomponents, and manufacturing equipment to slip through. Information from IAEA inspection reports and other sources reveals how Iraq exploited these loopholes.
Centrifuge design. "The Iraqi centrifuge design conforms substantially to early West European designs," according to the ninth inspection report. "However, no component is identical in design; all showed evidence of intelligent adaptation and development based on sound principles." Inspectors said the changes improved the machines. But the report concluded that although the Iraqi centrifuge experts were capable, "it is unlikely that they were able to make the observed design modifications without outside help."
Who supplied that help remains a major unanswered question. According to several sources, the West European designs at issue are for early Urenco machines built in the 1970s at Almelo in the Netherlands and Gronau in Germany. The Iraqi design incorporates principles similar to those developed for the G1, an early Urenco production model designed in Germany, but it also has features from later designs known as G2 and G3.
Stemmler said that the Iraqis showed him general, not detailed, blueprints for a G1-type centrifuge with some modifications. A Urenco official has said there is no indication that Iraq obtained Urenco blueprints for specific centrifuges, which are carefully controlled and catalogued. Nor have IAEA inspectors substantiated suspicions that Iraq obtained such blueprints.
Nevertheless, suspicions center on Germany as the source of design information. The German centrifuge component operation has been "too decentralized" compared to Urenco operations in Britain and the Netherlands, said one Western enrichment official. In Germany, many subcontractors have made components for Urenco centrifuges and enrichment facilities and have obtained sensitive design information from Urenco to do so.
After the confrontation with IAEA inspectors during the ninth inspection, Iraqi authorities acknowledged they had received "advice from abroad" but said they had not wanted to rely on foreign assistance. They said that outsiders had accelerated the program, but that Iraq had invested great resources in developing its own troubleshooters because the outsiders had been unable to solve all the technical problems. One inspector said this might be true to a point, but "the imprint of Germany is there."
Iraq acquired many of the subcomponents and some key parts of centrifuges but in widely varying quantities-evidence of its "buy-when-available" policy. Many of these components were obtained through intermediaries. Iraq was preparing to manufacture the most important parts on its own and was in widely varying states of readiness for these tasks. [See "The Gas Centrifuge, Piece by Piece," page 32.]
Every aspect of Iraq's nuclear weapons program shows signs of a patient, no-expense-spared search for help abroad.
Calutrons. Far less is known about Iraq's procurement of calutron components than about the centrifuge program, although the calutron program, too, depended heavily on foreign help. The difference was that export controls rarely hindered Iraq's attempts to buy calutron components. Iraq's success in hiding its calutron pro-gram from Western intelligence agencies can also be traced in part to a lack of export controls and the monitoring that accompanies them. Until last year, calutron technology was thought to be too expensive, cumbersome, and outdated for any would-be nuclear weapons state to pursue.
One of the few cases about which some details are known involves the massive iron pieces used in calutron magnets and in connecting several magnets into a "racetrack" (see September 1991 Bulletin). Iraq contracted with foreign foundries to cast and rough-machine the large iron components for calutrons being installed at Tarmiya. Final machining was done in Iraq at the Al Radwan facility.
Voest-Alpine AG, an Austrian foundry, has been identified by Western officials as one of several foundries where calutron parts were cast, although the officials do not believe the company knew that the parts were for calutrons. The first order to this foundry came directly from the Iraqi State Electric Establishment. A second order for 28 pieces, six of them identical to the first order, came from a German company. According to the eighth inspection report, the foundry's management has said that the company would have received a larger order from an intermediary if it could have met a required deadline. IAEA officials believe that other contracts must have been placed with other foundries, since the number of pieces found in Iraq is larger than the number the identified foundry supplied.
Much of the infrastructure at Tarmiya, the main calutron site, was built by foreign companies. A Yugoslavian firm, Federal Directorate of Supply and Procurement (FDSP), may have been the main contractor for the project. Electrical installation was provided by a Yugoslavian company, EMO. One inspector said that Siemens AG of Germany provided some of the electrical distribution equipment that connected transformers to ion sources at Tarmiya.
Making a weapon. Tightening export controls did slow Iraq's effort to develop a deliverable nuclear bomb, as a 1990 Iraqi progress report made clear (January/February 1992 Bulletin). Nevertheless, Iraq tried to obtain components illegally and exploited loopholes to acquire raw materials and manufacturing equipment when components were blocked. The main site at Al Atheer was being outfitted mainly as a materials research facility. Iraq was therefore purchasing a great variety of state-of-the-art equipment for this facility.
Iraq imported hundreds of tons of HMX high explosives, the most desirable conventional explosive for nuclear weapons. Western intelligence believes that Chilean explosives expert Carlos Cardoen exported most of the HMX to Iraq. Cardoen, who set up a cluster bomb factory near Baghdad, is under investigation by the U.S. Justice Department. Who made the explosive and how it got to Iraq is unknown, although Eastern Europe is suspected.
Streak cameras are capable of taking high-speed photographs useful in a nuclear weapons development program. A Japanese firm, Hamamatsu Photonics KK, legally exported two streak video cameras and support equipment to Iraq's Ministry of Industry and Minerals in September 1989. The declared use was for experimental automobile engine research at the University of Technology, Baghdad. According to the eighth inspection report, the cameras had a speed and resolution suitable for weapons development.
Iraq acquired from the Swiss firm Asea Brown Boveri a large cold isostatic press that could be used to shape explosive charges, but the press was found in a location without adequate safety equipment for high explosives work. It also obtained a vacuum induction furnace (a very high-temperature furnace) manufactured by the German firm Arthur Pfeiffer Vakuum Tecknik GmbH, which might have been useful in making uranium metal components.
Money. An effort of the magnitude of Iraq's nuclear weapons program requires the ability to transfer large amounts of money internationally. Iraq used bank letters of credit to purchase what it needed, and few questions were asked by the European and American banks-including the scandal-ridden Bank of Credit and Commerce International-that arranged Iraq's commercial transactions.
An Atlanta-based branch of Italy's Banco Nationale da Lavore (BNL) is now under investigation by the U.S. Department of Justice for conspiracy, wire fraud, and mail fraud related to financing Iraq's global procurement effort. The U.S. government recently obtained evidence that BNL provided nearly $1.6 billion to Iraq's Ministry of Industry and Military Industrialization to buy petroleum interests, the investment proceeds from which were allegedly to be used for the procurement effort. These interests were to be managed by the brother and the son of Jaffar Dhia Jaffar-a deputy minister of the ministry as well as vice chairman of the Iraqi Atomic Energy Commission. Who headed Iraq's nuclear weapons program is still unknown, but Jaffar is a leading suspect.
The IAEA inspection effort in Iraq provides a snapshot of one country's success in attempting to procure equipment and know-how for a nuclear weapons program. Iraq spent 10 years and billions of dollars, but the evidence shows that Iraq got as far as it did because of political circumstances, suppliers' ignorance, commercial deception and greed, and even scientific curiosity.
Western governments that backed Iraq against Iran looked the other way when firms sent dual-use items to Baghdad, and companies have justified their behavior on the basis that Saddam Hussein was battling Teheran's bearded mullahs. But there is no reason to believe these firms would not have supplied Iran had their governments made another choice.
There is, however, little reason to doubt that many firms were victims of Iraqi deception regarding the end uses of their products. That is particularly the case for many common items, like small iron magnets, aluminum motor housings, and bolts of copper wire. Records seized by the IAEA testify that most of the items Iraq ordered from abroad were ordinary.
But not all. Some items used to manufacture centrifuge rotor assemblies-flow-forming machines, CNC-machine tools, and specialized welding gear-were critical to the success of the program. Lax export controls, particularly in Germany and Switzerland, gave Iraq easy access to this equipment.
Iraq's access to Western centrifuge enrichment know-how was even more critical. Key components in Iraqi centrifuges clearly show the fingerprints of Urenco designs. While the suspicion that Iraq obtained classified Urenco blueprints has not been confirmed, design details of Iraqi components have since prompted an even more worrisome proliferation scenario: On the basis of early Urenco centrifuges, yet-unknown freelance experts have been improving them. Unless the IAEA and the U.N. Special Commission find the source of this assistance, the roots of Iraq's clandestine nuclear program may still bear fruit.
Digging for gold
Inspectors' finds in the Iraqi deserts have become a bone of contention between the International Atomic Energy Agency (IAEA) in Vienna and the U.N. Special Commission in New York. Under U.N. Security Council Resolution 687, the U.N. Special Commission is empowered to eliminate Iraq's weapons of mass destruction, and the IAEA is responsible for on-site inspections of Iraq's declared and undeclared nuclear facilities and materials. Using intelligence information, the Special Commission designates sites for the IAEA to inspect. The IAEA organizes and carries out the inspections, and reports its findings back to the Special Commission.
In theory this arrangement is straightforward, but in practice several disagreements over what to inspect and when to release information to the public have occurred. In mid-January, when the U.S. media was leaked a preliminary report while inspectors were still in the field, the transatlantic tiff flared up.
The IAEA inspection team, following up on information provided by the German government, confirmed that a German company had supplied equipment and parts for centrifuge motor stators, including 240,000 small ferrite spacers, which become magnetized during operation. (Some contemporary news reports referred to the spacers as "magnets.") Since Iraq's design called for 24 spacers per motor, the spacers would have equipped about 10,000 centrifuges.
On the basis of this piece of information, the Washington Post reported on January 14: "Had all the centrifuges been assembled and operating, diplomats said, they may have been capable of producing enough enriched uranium to fuel three or four nuclear bombs a year." Other major U.S. newspapers followed with similar reports, and on January 15, London's Financial Times declared in a headline, "Iraq Might Have Four Bombs per Year."
The spacer find, along with other discoveries, confirmed long-standing suspicions that Iraq had planned eventually to go beyond building a few hundred centrifuges to several thousand. But the toffee-colored objects, each the size of sugar cube, are simple to manufacture and are not controlled by export regulations.
"The spacers were available on the market, so Iraq bought them," one IAEA official said. The purchase, officials believe, follows the pattern of other Iraqi activities, which was to obtain whatever possible.
The conclusion in press reports that the spacers would have enabled Iraq to build 10,000 centrifuges was "analogous to the assertion that if Iraq got 40,000 hubcaps, it could make 10,000 automobiles," the official said.
The spacer story was leaked by U.N. officials in New York "against the instructions of the IAEA" after a member of the inspection team, Special Commission Assistant Director Robert Gallucci, sent the report to his boss, Rolf Ekeus. After the leak, IAEA Director General Hans Blix fired off a letter of complaint to Ekeus. (Days after the magnet story, Gallucci took a top job in the U.S. State Department.)
IAEA officials say that poorly timed press leaks accompanied by rumors that Iraq was close to producing bomb quantities of highly enriched uranium have hindered their attempts to gain the confidence of Iraqi scientists and engineers who still hold important pieces of Iraq's nuclear puzzle. In the background, however, are more important differences about how to get to the bottom of Iraq's nuclear program.
Throughout the inspection effort, some U.N. officials have taken a hard line. "Only if we push Iraq to the brink will we ever find out what is going on," one U.N. official said. These experts are impatient with the slow-but-steady approach favored at the IAEA. Some U.N. officials have tried to transfer responsibility for long-term monitoring of Iraq's nuclear program from the IAEA to the U.N. Special Commission.
The differences between the two agencies blew up again when IAEA inspectors, under last-minute orders from the Special Commission, searched fruitlessly in Iraq in mid-February for evidence of a plutonium production reactor allegedly operating underground, and of a heavy water production plant. Ironically, the Iraqi News Agency reported the reactor story first, sure that the search was a wild-goose chase. Quoting a European diplomatic source the next day, February 13, the New York Times reported that "all the specialists believed" that Iraq had the reactor. The rumors that launched the search had been flying since before the Gulf War, with no hard evidence to support them.
U.N. and IAEA officials are now working to patch up their differences. The Special Commission has agreed to include the IAEA in its deliberations on picking sites to be inspected, and the Special Commission will play a greater role in the inspections.
The gas centrifuge, piece by piece
A gas centrifuge consists of thin-walled cylinders held in a vacuum and spun at very high peripheral speeds. To achieve these speeds, rotating components must be manufactured from specialized materials and machined very precisely to minimize imbalances.
Non-rotating components. The top magnetic suspension bearing of a gas centrifuge contains a pair of ring magnets: The upper one is suspended in a housing containing oil that is resistant to the highly corrosive uranium hexafluoride gas fed through centrifuges (the oil serves as a damper). The other is fitted to the top cap of the rotor assembly. Iraq had not adopted a final design for the top bearing but was exploring one design using aluminum-nickel-cobalt magnets and two designs using more advanced cobalt-samarium magnets. Iraqi officials said they had acquired about 125 of the former type and a few dozen of the latter type for research.
Inwako GmbH, a German arms dealer located in Bonn, supplied aluminum-nickel-cobalt magnets, which were made by a British firm and procured by another, Endshire Export Marketing, then shipped to Germany before going to Iraq. The supplier of the cobalt-samarium magnets is unknown, although an IAEA inspector said that they might have been made in Britain.
The bottom bearing is a pivot or cup assembly mounted on an oil damper, combined with a needle attached to the bottom cap of the rotor assembly. The relatively sophisticated design Iraq was using includes on the tip of the needle a ball etched with spiral grooves, which pump the oil up between the ball and cup, minimizing friction on the ball. Experts consider the development and etching of the pattern difficult.
Evidence that Iraq possessed spiral-grooved bottom bearings was one of the surprises of the IAEA inspections. The ball in question corresponds in size to the ball of a G2 centrifuge but the pattern is not thought to be a Urenco pattern, according to an IAEA official. Iraqi officials have said they had photo-etching equipment to put the pattern on the ball, but that it was destroyed during the bombing. Inspectors have not seen it. A U.S. official said that in 1987 Iraq obtained a laser cutting system, which might have been used to laser-etch the pattern after Iraq received an electron-beam welding system [see main article].
A molecular pump, which has internally machined grooves, keeps any uranium hexafluoride leaking through the top bearing from entering the vacuum area between the rotor tube and the outer casing. Iraq acquired 84 tons of aluminum alloy tubes from a German firm, enough to manufacture 6,000 pumps. One of the CNC (computer numerically controlled) machines it acquired from Neue Magdeburger Werkzeugmachinen GmbH had special attachments to manufacture molecular pumps.
The rotor spins in a strong, vacuum-tight outer casing to minimize drag and to contain debris if the rotor breaks. Iraq received from a German company about 300 tons of thick aluminum alloy tubes to manufacture outer casings-enough for about 2,500. An order for 310 tons more was stopped by the 1990 embargo. Iraq declared that it had also received enough aluminum forgings to make several thousand top and bottom flanges (caps) for the casings, as well as a CNC machine from Neue Magdeburger to machine the casings.
Ring-shaped stators (hysteresis motors) designed to operate in a vacuum spin the rotor at high speeds- expected to be about 60,000 revolutions per minute for the Iraqi machines. Iraq developed or acquired a design for an axial hysteresis motor. Through the German company Rhein-Bayern Fahrzeugbau GmbH, Iraq ordered 240,000 magnetizable ferrite spacers-24 are needed for each centrifuge stator-and 10,000 soft iron ring band cores, one per stator. Rhein-Bayern acted as a middleman for the Iraqi purchase and had design specifications for these "magnets." The spacers and rings were supplied in several shipments to the State Electrical Industries Establishment in Baghdad between January and May 1990, according to the ninth inspection report. From the same company Iraq acquired a die-casting machine to manufacture coil rings for stators. These subcomponents were enough to assemble 10,000 stators for centrifuge motors-and the source of inaccurate extrapolations that Iraq had been ready to manufacture 10,000 complete centrifuges [See "Digging for Gold," above]. Iraq also declared it had acquired enough aluminum forgings to manufacture several thousand bottom flanges, which hold the stator.
Inverters, also known as frequency changers or converters, power the motors and are easy to obtain for single-machine test stands. But inverters suitable for cascades are not. The Iraqi Electromagnetic Equipment Establishment in Falluja ordered five high-frequency inverters, rated at 15 kilovolts and 25 amperes, from Acomel SA in Neuchâtel, Switzerland, in early 1989; they were shipped in May of that year. Iraqi officials said at the time that they were for manufacturing ball bearings and spindles. An enrichment expert said that each inverter could have powered 100 centrifuges.
Scoops are small tubes at the ends of the central post that extract enriched and depleted uranium hexafluoride. They are relatively easy to make, and Iraq had procured some scoops and some tubing to make more.
Sealed valves are manufactured from material resistant to uranium hexafluoride; they prevent leakage of air into the centrifuge, which would react with the uranium hexafluoride and destroy it. Iraq declared it had acquired 700 uranium hexafluoride-resistant bellows valves from Swiss companies Balzer AG and VAT AG, and the U.S. company, Nupro. An official at VAT said the company had not sent the high-quality valves the company makes for Urenco, although the valves sent could have been used for gas centrifuges.
The rotor assembly. Centrifuge rotor components-a thin cylinder or a number of connected cylinders, fitted with internal baffles and disc-shaped endcaps-are made from high-strength materials such as high-tensile-strength maraging steel, aluminum alloys, or specially wound carbon fibers held together by special epoxy. Iraq used carbon-fiber rotors in its research and development program, but it would have produced rotors made of maraging steel.
Iraq admits to purchasing (but has refused to reveal the supplier of) 100 tons of maraging steel, sold as both preformed cylinders ("preforms") and in sheets; Western governments are now trying to track down the supplier in Europe. The material is believed to have come from Germany or Britain. About three tons of the steel had been used in Iraq and Europe to manufacture centrifuge components. The steel was believed to be "grade-350," an ideal quality for the type of centrifuge Iraq was developing. It was melted down, impairing its maraging properties, before inspections started.
This amount of material would have been enough for 5,000 centrifuge rotors, endcaps, and baffles-only about half of them good enough to use, according to Iraqi estimates, which inspectors found reasonable. Still, production of usable rotors might have been far less, since no maraging steel rotors adequate for uranium enrichment have been found in Iraq.
No evidence has surfaced that Iraq could have made its own maraging steel. It wanted to develop such a capability at Nassr State Establishment, according to a U.S. official, but the idea "never got off the ground."
The rotor tubes are shaped on a flow-forming machine, which cold-presses the preforms into thin-walled cylinders. When they are the correct thickness, the rotors must be cut to the right length, heat-treated to strengthen the metal, and oxidized to make them resistant to uranium hexafluoride. Iraq was setting up a manufacturing operation to make hundreds of these tubes a year but had not yet produced adequate rotors.
Inspectors said that at least one of the flow-forming machines Iraq had obtained from H&H Metalform had been used to make maraging steel rotor tubes. It had been outfitted with rollers, an expanding mandrel, and a finishing mandrel. Inspection of the wear on the mandrel's collar showed that the machine had been used for only about 13 hours. One of the CNC machines acquired from Neue Magdeburger had specialized fixtures to cut the rotors to a precise length.
IAEA officials said the oxidation process Iraq was pursuing-unsuccessfully, according to Bruno Stemmler-was consistent with Stemmler's patent. [See page 35.] Inspectors believe that the Iraqis would have succeeded eventually. Sometime after Stemmler's visits, Iraq had acquired a small and a large oxidation furnace. The large one, with special temperature controls, was manufactured by Degussa AG, a German firm. It had not yet been used and was probably intended for serial production of centrifuge rotors. The small one had been used.
Iraq has stated that it obtained 20 carbon fiber rotors for its research and development program from a still-unknown European "dealer." The inspectors have seen 12 of them; the Iraqis claimed that the other eight were broken during attempts to install endcaps. The inspectors found the story plausible because the endcaps, also made outside Iraq, were slightly too large for the rotors.
Experts believe the Japanese company Toray manufactured the carbon fibers used in the rotors, although they do not believe that Toray had anything to do with making the rotors. The ninth inspection report notes that Toray had only one customer for these types of carbon fibers, which are used specifically in centrifuges. That customer is Urenco.
The windings of at least two of the rotors, according to the ninth report, were different from the rest, and also distinct from windings in Europe. Iraqi authorities claimed to be unaware of the differences and said they had "nothing to do with material and construction specifications." There is no evidence that Iraq was able to make carbon fiber rotors; inspectors found three winding machines at the Dhu Al Fiqar factory near Falluja, but these were apparently intended for missile manufacturing and apparently had not been used.
Top and bottom endcaps are welded to a maraging steel rotor tube to contain the uranium hexafluoride. The top cap would support part of the upper bearing; the bottom cap would carry the rotating part of the bottom bearing. A baffle isolates the chamber of the rotor from which the slightly enriched or depleted uranium is extracted. Unlike many designs, the Iraqi model had two baffles, each slightly different.
A government report to the German parliament notes that endcaps and baffles in Iraqi centrifuges "show design details that are identical to those in German gas centrifuges." The Iraqis have admitted, according to the ninth inspection report, to having ordered 15 CNC machines for making endcaps, baffles, and other components used in the centrifuges' uranium hexafluoride feed and extraction systems. To manufacture and program these machines, Iraq provided the manufacturer detailed plans for these components. Iraq had received 10 machines, and five more were being manufactured, but before delivery the Iraqis wanted the machines' capabilities demonstrated. For this purpose, they provided the manufacturer maraging steel shipped from Germany. The five CNC machines were seized by customs officials following the confiscation at the Frankfurt airport of the forgings made by Schmeidemeccanica. The shipment of maraging steel forgings consisted of 250 top caps, 250 bottom caps, 250 baffles, and 138 unidentified pieces, weighing a total of 1.6 tons.
The ninth report did not identify the manufacturer of the CNC machines, although other sources have said it was Schäublin in Switzerland.
Iraq would have needed all 15 CNC machines to produce these components efficiently, because each would have served a unique purpose. But some of the 10 machines it possessed could have been reprogrammed to do the functions of the undelivered machines. One IAEA official said that Iraq had obtained all the necessary computer programs-and that the Iraqis could not have developed the programs themselves.
Other equipment that Iraq acquired included a number of machines that were to be used to assemble centrifuges. Among the more important were an electron beam welder for maraging steel components-with a special fixture for holding rotors while endcaps are being attached-provided by the German firm Leybold Heraeus; centrifuge balancing machines, obtained from Reutlinger und Söhne KG, another German firm; and metal inert gas welding equipment, also with a specially designed holding device for welding aluminum centrifuge components.
Reprinted by permission of The Bulletin of the Atomic Scientists, copyright 2001 by the Educational Foundation for Nuclear Science, 6042 South Kimbark, Chicago, Illinois 60637, USA. A one year subscription is $28.
As of August 2006, Iraq Watch is no longer being updated. Click here for more information.
Copyright © 2000-2006