Attachment 1.
The components of Iraq's clandestine nuclear programme

1. The acquisition of weapons-usable nuclear material

1.1 Procurement and indigenous production of uranium compounds

1.1.1 Material declared and subject to IAEA safeguards
1.1.2 Procurement of yellow cake and uranium dioxide
1.1.3 The Al Qaim uranium recovery facility
1.1.4 The Al Jesira uranium conversion facility
1.1.5 Uranium pilot plant development at Tuwaitha
1.1.6 Summary
Table 1.1 Material balance - Tuwaitha uranium projects

1.2 Development of indigenous uranium enrichment capabilities

1.2.1 Electro-magnetic isotope separation (EMIS)
1.2.2 Gaseous diffusion uranium enrichment
1.2.3 Gas centrifuge uranium enrichment
1.2.4 Chemical and ion-exchange uranium enrichment
1.2.5 Laser isotopic separation
1.2.6 Summary

1.3 The intended diversion of research reactor fuel

1.3.1 The "crash programme"
1.3.2 The recovery of HEU- Project 601/603
1.3.3 The further enrichment of the HEU- Project 521 C
1.3.4 The conversion to metal of HEU - Project 602/602B
1.3.5 Summary
Table 1.3 Iraq's research reactor fuel inventory as verified by the IAEA on 19/20 November 1990

1.4 The production and separation of plutonium

1.4.1 The indigenous reactor - Project 182
1.4.2 The use of the IRT 5000 reactor
1.4.3 The separation of plutonium
1.4.4 Summary

2. Weaponisation

2.1 Background
2.2 Facilities
2.3 Research and development
2.4 Missile delivery system
2.5 Programme documentation
2.6 Summary

1. The acquisition of weapons-usable nuclear material

1.1. Procurement and production of uranium compounds

1.1.1 Material declared and subject to IAEA Safeguards

a. Low enriched uranium

In 1982 Iraq imported from Italy 1,767 kg of uranium enriched to 2.6% in U-235 in the form of UO2 powder. The material has been verified and fully accounted for and remains in Iraq, under the control of the IAEA, at Location C (a storage complex close to Tuwaitha), in the same form as it was received.

b. Natural uranium

In 1979, Iraq imported from Italy 4,006 kg of natural uranium as U02 powder and 508 kg uranium as UO2 in the form of pressed fuel pellets. The UO2 powder and the pellets were used in the Experimental Research Laboratory for Fuel Fabrication (ERLFF) for research and development activities. Of the 4,514 kg uranium received, 4,323 kg uranium have been accounted for, leaving 191 kg not accounted for. This amount is less than the declared accumulation of "material unaccounted for" and measured discards over the period 1982 to 1990 and may be considered to be consistent with the nature of the facility operation. The balance of this material has been verified and fully accounted for and remains in Iraq, under the control of the IAEA, at Location C.

c. Depleted uranium

In 1979, Iraq imported from Italy, 6,005 kg of depleted uranium as UO2 powder. The material has been verified and fully accounted for and remains in Iraq, under the control of the IAEA, at Location C, in the same form as it was received.

d. Highly enriched uranium

Iraq's inventory of research reactor fuel which was imported from Russia and France contained almost 50 kg of highly enriched uranium, based on pre-irradiation values. All of Iraq's inventory of research reactor fuel, as listed in Table 1.3, was fully accounted for and removed from Iraq - the last consignment having been shipped in February 1994.

1.1.2 Procurement of yellowcake and uranium dioxide

In the period 1979 through 1982, Iraq procured yellowcake from both Portugal and Niger and uranium dioxide from Brazil. At that time, neither Niger nor Brazil were party to the NPT, nor had either concluded a comprehensive safeguards agreement, which would have required notification to the Agency of the transfers of such material to Iraq. Portugal, a party to the NPT, but without a comprehensive safeguards agreement at that time, notified the Agency of the transfers to Iraq.

The yellowcake procured from Portugal was supplied in two batches. Batch 1, received on 20 June 1980, consisted of 429 drums containing 138,098 kg of yellowcake and batch two, received as three consignments over the period from 17 May 1982 through 20 June 1982, consisted of 487 drums containing 148,348 kg yellow cake. By letters dated 6 August 1981, 1 June 1982 and 21 July 1982, Iraq notified the IAEA of the receipt of this material, which confirmed the complementary notifications received from Portugal at the time of shipment. Iraq's entire holding of the material of this origin was verified against comprehensive packing lists provided to the IAEA by the Iraqi counterpart, detailing the original production lot number together with weight data for each drum. Verification measures involved weighing, non-destructive assay and sampling and analysis from which it was concluded that all of the yellowcake received from Portugal was fully accounted for and remained intact, as shipped, except for the loss of about 40 kg from a drum damaged during Iraq's salvaging/concealment activities in 1991. This material remains in Iraq, under the control of the IAEA, at Location C, in the same form as it was received.

The yellowcake procured from Niger was also shipped in two batches. Batch one, received on 8 February 1981, consisted of 432 drums containing 137,435 kg of yellowcake and batch two, received on 18 March 1982, consisted of 426 drums containing 139,409 kg yellowcake. By letter dated 6 August 1981 Iraq notified the IAEA of the receipt of the first consignment but did not provide notification of receipt of the second consignment. Iraq's entire holding of material of this origin was verified against comprehensive packing lists for both batches, provided to the IAEA by the Iraqi counterpart, detailing the original production lot number together with weight data for each drum. Verification measures involved weighing, non-destructive assay and sampling and analysis from which it was concluded that all of the yellowcake received from Niger was fully accounted for. This material remains in Iraq, under the control of the IAEA, at Location C, in the same form as it was received.

Iraq did not report to the IAEA the 1981/1982 import of uranium dioxide (UO2) from Brazil and its existence in Iraq was only recognised at the time of Iraq's revised declaration of 7 July 1991. Verification and accountancy of the UO2 procured from Brazil was complicated by the fact that Iraq was unable to provide adequate shipping documents for all of the material and declared that it had used some 4,422 kg out of its estimated total receipts of 27,000 kg UO2. Iraq declared that there had been two receipts of UO2 from Brazil, the first in August 1981, consisting of 7,914 kg UO2 in 120 drums, and a second receipt in the first half of 1982 consisting of 128 drums containing from 17,300 to 19,200 kg UO2. Iraq claimed not to know how much material was in the second shipment asserting that it had arrived without shipping documents and that the material had not been weighed in Iraq. The only available documentation for the two shipments was a list of weights for the first shipment and a list of analytical results for the second. Verification activities carried out during IAEA-12 showed the amount on inventory to be considerably less than declared - thus putting into question the reported consumption. Furthermore, the varied and unusual physical forms of the UO2 raised doubts as to its origin.

An extensive verification effort was subsequently undertaken involving weighing, non-destructive assay, sampling and analysis and microscopic examination of the physical form and properties of a comprehensive series of samples of the material. In this way the range of powders and granules comprising the Brazilian UO2 material were characterised and shown to be distinctly different from the material overtly imported or indigenously produced.

The task was finally completed in July 1994 when, with the co-operation of the Government of Brazil, it was possible to confirm the origin of the UO2 on the basis of the chemical and physical characteristics determined by the IAEA. At this time it was also possible to gain confirmation of the amount of material shipped to Iraq. These data enabled the IAEA to verify and balance Iraq's declared usage against the material remaining on inventory. Of the 24,260 kg UO2 received by Iraq from Brazil, 3,600 kg was used to produce UCl4, UF4, and uranium metal - the rest has been verified and remains in Iraq, under IAEA control, at Location C.

1.1.3 The Al Qaim uranium recovery facility

The phosphate rock deposits of western Iraq contain uranium in the range of 50-80 ppm. A large deposit at Akashat is mined to supply a phosphate fertiliser plant at Al Qaim, some 150 km distant. During the period 1982 to 1984 a plant (Unit 340) for the extraction of uranium from the process phosphoric acid was constructed and commissioned. Operating at design capacity the plant should have produced 103 tonnes of uranium per year - equivalent to 146 tons of yellowcake - assuming 317 operating-days and processing 3,600 m^3 per day of phosphoric acid containing 75 PPM uranium at a recovery efficiency of 93%. Over its six years of declared operation the plant should have produced about 600 tonnes of uranium contained in nearly 900 tonnes of yellowcake. However, Iraq declared a production of only 109 tonnes of uranium in 168 tonnes of yellowcake, i.e., less than 20% of the design capacity of the plant.

The investigation of this apparent inconsistency was greatly facilitated by the presence of a set of operating records - daily production reports - covering the period from 1986 through 1990 and containing day-by-day data on input and output phosphoric acid flows and their respective uranium contents, the relative levels of two key chemical tanks and the number of drums (including drum serial numbers) of yellowcake produced.

An extensive evaluation of these data was undertaken to assess the consistency of the daily operating data with the yellowcake production. On the basis of sampling at the Akashat mine, a relationship was derived between the uranium and phosphorous pentoxide content of the ore, which enabled the calculation of uranium in the input acid stream. On this basis it was possible to derive a theoretical estimate of the plant production which was in very good agreement with the declared production.

This analysis also showed that the poor performance of the plant was due to the low assay of the feed acid (~ 60% of design value), the inability of the acid plant to meet the design flow-rate of 3,600 m^3/day (~ 50% of design flow-rate), failure to sustain the 93% design recovery efficiency (actual values typically 78%) and the fact that the plant on average operated only 214 days per year as opposed to the design operation of 317 days per year.

1.1.4 The Al Jesira uranium conversion facility

The Al Jesira uranium dioxide and uranium tetrachloride (UCl4) production facility, located west of Mosul in northern Iraq combined a UO2 plant of 185 MT/year design capacity, designated as Project 212 and code named the "Wax Plant", and a UCl4 plant of 105 MT/year design capacity, designated Project 244. Both plants sustained considerable damage through aerial bombardment and were thus rendered inoperable in January 1991. Inspection of the facility was complicated by actions taken by Iraq to conceal the true function of the facility which involved the removal of all nuclear material from the facility, the transfer of 2,500 cubic metres of uranium bearing liquid waste to a petroleum storage tank, near Mosul, some 30 km distant from Al Jesira, and the removal and burial of uranium contaminated plant components and waste disposal system pipework at Al Adaya.

a. UO2 production

The UO2 production plant was based on designs provided by a Brazilian company. The plant, which was constructed by Iraq in the period July 1985 to July 1989, was based on the well-proven technology involving the dissolution of the input yellowcake in nitric acid followed by multi-stage solvent extraction, ammonium diursnate precipitation, its filtration and calcination to uranium trioxide, from which the UO2 was produced through hydrogen reduction. Design production capacity was 23.7 kg UO2/hr. The plant began its commissioning phase of operations on 5 July 1989, which continued through to the end of January 1990. This phase was beset with difficulties and the plant operating records show that only 8,879 kg UO2 was produced. The plant went into routine operation in February 1990 and, apart from being shut down during the month of April of that year, continued to operate until 2 December 1990, by which time all of the available Al Qaim yellowcake had been processed. It was necessary to prepare the plant to process either Niger or Portuguese yellowcake and, during December and early January 1991, there was sporadic operation to clean up waste and scrap and to prepare the process for a new feed material of different chemical form.

The Al Jesira UO2 plant produced 420 drums containing 99,457 kg UO2 (86,607 kg uranium). Of these 420 drums, five were used for UCl4 production at Al Jesira, four were used for UCl4 production in the Chemical Engineering laboratory (Tuwaitha Building 85), and two were used for uranium metal production in the Experimental Research Laboratory for Fuel Fabrication (ERLFF - Tuwaitha Bldg 73). The remaining 409 drums are currently stored under IAEA control at Location C.

Al Qaim yellowcake containing 98,512 kg uranium was received at Al Jesira and was converted into UO2 containing 86,607 kg uranium, resulting in a difference of 11,905 kg uranium. This difference has been investigated in detail and it is estimated that 10,140 kg uranium can be accepted to be present in waste products and damaged plant components, leaving 1,765 kg uranium unaccounted for. This figure is deliberately conservative and could be reduced if greater allowance were to be made for losses resulting from accidental dispersal through Iraq's concealment activities, losses to solvent extraction fluids and losses through dispersal resulting from the aerial bombardment.

b. UCl4 production

The UCl4 production plant, Project 244, was constructed at the Al Jesira site based on design and operating experience gained from the UCl4 pilot plant (Project 242) built and operated in Building 85 at Tuwaitha. Construction of the Al Jesira plant started in February 1988 and operations commenced on 1 February 1990. The plant consisted of two parallel production lines with a combined capacity of 105 MT/year of UCl4. Only one line was operational.

The UCl4 plant operation was limited to a period of 72 hours during the month of February 1990, when it was used to produce a total of 1,200 kg UCl4, containing 780 kg uranium from an input feed of 1,036 kg UO2 containing 901 kg uranium and generated waste containing 121 kg uranium. Following this brief period of operation the plant was shutdown for maintenance and repairs and was never again brought back into operation. All of the UCl4 produced at Al Jesira is stored, under IAEA control, at Location C.

Although it seems inconsistent that the plant would be shut down after only a few days operation, it should be recalled that the plant had been commissioned well ahead of the need for its contribution to the supply of UCl4 for the EMIS programme. The commissioning of separators at the Tarmiya EMIS facility began in February 1990 and only eight separators were in partial service before operations were interrupted by the aerial bombardment in January 1991. Even in full operation the Tarmiya plant would have required an annual feed of no more than 3,000 kg UCl4, an amount well within the production capacity of Project 242 (Tuwaitha Building 85).

1.1.5 Uranium pilot plant development at Tuwaitha

The principal production and use of uranium compounds at Tuwaitha took place in three locations:

• Chemical laboratories (Building 15B) which processed Brazilian-origin UO2 to produce UF4, uranium metal and UF6.

• Experimental Research Laboratory for Fuel Fabrication (ERLFF - Building 73) which processed Brazilian origin UO2, Al Jesira origin UO2 and Al Qaim yellowcake to produce UO2, U3O8, UO3, UO4, UF4 and uranium metal.

• Chemical Engineering Research laboratories (Building 85) which processed Brazilian origin UO2 and Al Jesira origin UO2 to produce UCl4.

Of particular note is the development of Iraq's capabilities with respect to the production and casting of uranium metal, which originated in Tuwaitha in the middle of 1986. The first phase of this work, which continued through March 1987 was carried out in Building 15 and involved some 30 experiments involving the magnothermic reduction of UF4. The experiments resulted in the production of discs of uranium metal of eight centimetre diameter, having individual weights in the range 600 to 900 grams - 19 such discs remain on inventory at Location C. The experimental work was discontinued in Building 15 and work was not resumed until the beginning of 1988 when facilities in Building 73 were then utilised for the task. The early work in this second phase concentrated on the development of methods to improve the purity of the UF4 feed material and it was not until November 1988 that uranium metal production recommenced. The metal produced in this phase was again in disc form but somewhat thicker - termed "derbies" to distinguish them from the previously produced "discs" - and typically weighed 1.3 kg. Phase three involved continued efforts to improve the purity of the UF4 feed material and a change in the physical form of the produced uranium metal to a solid cylinder of about 5 cm diameter and similar length with a typical weight of 1.5 kg.

By late-1989, this research and development had enabled Iraq to establish its capability to produce uranium metal of high purity with relatively small process losses. On the basis of this capability a larger scale plant was designed and constructed in Building 64 at Tuwaitha with the capacity to produce 20 kg of uranium metal per day. The plant was still under commissioning in January 1991 when Building 64 was heavily damaged in the bombardment of Tuwaitha. Despite the severe damage to the building, much of the equipment, which was general purpose in nature, was salvaged and is currently located at the Al Zahf Al Kabir metallurgical facility in the Taji area, where it is subject to ongoing monitoring and verification.

Some 1,150 kg of natural uranium metal was made in the period 1986 to January 1991, of which 1,000 kg remains in Iraq under IAEA control. About 150 kg was used in a series of metal purification and melting and casting experiments at Tuwaitha and Al Atheer. The most interesting pieces cast were a 5 cm diameter sphere and a small number of 5 cm diameter hemispheres, except for 10 small uranium bullets and 9 cast rods, all castings and machined uranium pieces were unilaterally destroyed by Iraq by dissolution in HNO3 as a concealment measure. Examination of the bullets and bars indicates only rudimentary melting and casting capabilities but, as claimed by Iraq, and supported by PC-3 programme documentation, Iraq expected that considerable improvements would be achieved through utilisation of the more advanced equipment that was soon to be installed at Al Atheer. Much of that equipment was blocked by the export embargo imposed by Security Council resolution 661 (6 August 1990) and all key equipment that was installed at Al Atheer was subsequently destroyed under IAEA supervision.

Iraq's exploration of UF4 and UF6 production technology spanned the period 1981-1985 and, in 1986, led to the design of Project 206. This project was based on a fluidised bed reactor using anhydrous hydrofluoric acid to produce 2 kg/day of either UF4 or UF6. Before construction was completed, Project 206 was modified to produce 1-2 kg UF4/batch and was renamed Project 231. However, according to the Iraqi counterpart, the modified equipment was never operated and attention was focused on rotary kiln technology.

Project 226, based on rotary kiln was technology, was constructed and commenced operation in mid 1986. This project used UO2 of Brazilian origin as the feed material which was reacted with Freon 12 as the fluorinating agent, to produce UF4. Project 226 was operated intermittently until 1991 and produced some 250 kg of UF4. A small quantity of the UF4 produced was used in 1987 to make uranium metal but the stated purpose of Project 226 was to provide a secure supply of UF4 for eventual conversion to UF6 to satisfy the needs of the gas centrifuge development programme. In the event, the material was not required and remains on inventory in Location C.

The lack of success with Project 206 also prompted consideration of the utility of batch processes using boat type reactors and small-scale experiments were carried out in 1985-1986 using both Fluorox as the fluorinating agent as well as direct fluoridation using fluorine gas. On the basis of this work, the direct fluorination method was selected for further development and a larger laboratory-scale boat type reactor unit, with a capacity of 50 g UF6 per batch, was constructed in 1986. This unit operated in Building 15B at Tuwaitha until mid-1987 when it was transferred to Rashdiya. The unit was replicated at Rashdiya and the two units constituted Project 234.

According to the Iraqi counterpart the amount of UF6 produced by the unit operating at Tuwaitha was 3-4 kg and by both units operating at Rashdiya was about 4 kg. In 1988 a third unit (Project 235) was constructed at Rashdiya, based on Project 234 designs, and this unit is reported to have been used to produce a further 500 grams UF6. Several other Projects for UF6 production and purification are documented by the Iraqi counterpart, including Projects 230, 232, 233, 236, 237, 238 and 238A, but were declared not to have proceeded beyond the design stage.

The total recorded production of UF6 is about 8 kg which, according to the Iraqi counterpart was hydrolysed to liquid waste except for 500 grams which is contained in a standard 1S cylinder. The hydrolysed waste and the remaining 500 grams UF6 are on inventory in Location C.

According to the Iraqi counterpart Projects 234 and 235 provided adequate supplies of UF6 to support the development work of the centrifuge programme. The counterpart also declared its confidence in its capability to exploit flame reactor technology, which was the basis of Project 236, to provide sufficient UF6 to support the pre-production development phase. This expressed confidence was based on their declared acquisition of an assembly drawing of a 1970s design flame reactor.

Research and development work on UCl4 production and purification at Tuwaitha is well recorded in IAEC/PC-3 documentation. Initial experiments commenced in 1982 in Buildings 9 and 15 and later, circa 1987, were transferred to Building 85, the Chemical Engineering Research Laboratories where activities continued until January 1991. Fifteen laboratory-scale research projects and pilot-scale production and purification projects were implemented during the nine years period. Many different feed materials, including, UO2, UO3, U3O8 and UO4:2H2O were tried as were different reaction techniques such as fluid bed, static bed (boat type) and rotary reactors with liquid, vapour and gas phase chlorination.

The extensive experimentation culminated in the design and construction of a pilot scale production unit, Project 242, in Building 85, which used UO2 as the feed material and gas phase chlorination. Project 242 which had a production capacity of 20-40 kg UCl4 per day commenced operation in 1988 and continued, on a campaign basis, until the end of 1990. During this period some 5,000 kg UCl4 was produced using Brazilian UO2 and Al Jesira UO2 as feed material. Project 242 was very successful and the chemical and operating experience so gained was used to design the industrial scale UCl4 facility at Al Jesira.

Three projects, 241 B, 245 and 244 were implemented from 1987 to 1990 to establish the capability to meet the purity requirements for EMIS feed material. These projects which were all based on sublimation were used to purify some 1,100 kg of UCl4.

The nuclear material balance for these Tuwaitha locations (Table 1.1) shows a total receipt of 14,789 kg uranium of which 13,117 kg uranium has been verified and remains on inventory at Location C. The resultant inventory difference or "material unaccounted for" (MUF) is 1,672 kg uranium which represents 11.3 % of the total receipts. Some components of this MUF comprise strata which are physically present but difficult to verify, with any certainty, such as the Building 73 waste, plant hold-up, uranium losses to metal slag and others for which Iraq has provided a plausible explanation backed up by documentation, such as the hydrolysis of UF6 and the dissolution of uranium metal. Conservative assessment of these components would reduce the MUF to 1,086 kg uranium or 7.3 % of the receipts. Given that some large inventory strata are inhomogeneous and thus potentially subject to large sampling errors, and accepting that the loss of some material, due to the bombardment and Iraq's salvage and concealment activities, cannot be discounted, the MUF value is not considered to be unreasonable.

1.1.6 Summary

1. Iraq's failure to provide complete notification to the IAEA of its importation of UO2 (from Brazil) and yellowcake (from Niger) was in contravention of its safeguards agreement with the IAEA

2. None of the imported yellowcake had been used by Iraq and was fully accounted for through IAEA safeguards verification measures. This material remains under IAEA control at Location C and is routinely verified by the IAEA.

3. An amount of 3,600 kg of the natural uranium dioxide imported from Brazil was used for the production of uranium tetrachloride, uranium tetrafluoride, uranium hexafluoride and uranium metal and has been accounted for in those converted forms. The remainder of the UO2 material of this origin has been unambiguously identified and fully accounted for. This material remains under IAEA control at Location C and is routinely verified by the IAEA.

4. As a result of its extensive audit, the IAEA is satisfied that Iraq's declared production of yellowcake at the Al Qaim facility, although well below the full design capacity of the plant, is consistent with the plant's mode of operation and is in good agreement with the plant operating records.

5. Taking into account the losses due to plant damage resulting from the bombardment and measures taken by Iraq to attempt to conceal the function of the plants, the amounts of uranium dioxide and uranium tetrachloride declared by Iraq to have been produced by the Al Jesira facilities are consistent with the plant input.

6. Again, taking into account the losses due to building damage resulting from the bombardment and measures taken by Iraq to attempt to conceal the function of the buildings, the amounts of uranium compounds and uranium met al declared by Iraq to have been produced at Tuwaitha are consistent with the amounts of feed material consumed.

7. The total amount of material unaccounted for, potentially arising from normal process losses taken together with the circumstantial losses referred to above, is determined to be just less than 3,000 kg natural uranium which is equivalent to 1.5% of the non-static inventory.

Table 1.1
Material balance - Tuwaitha uranium projects

Iraq's former holdings of research reactor fuel are listed in Table 3.1.

1.2. Development of indigenous uranium enrichment capabilities

As stated in the FFCD, Iraq's strategy for the acquisition of weapons-usable nuclear material, established at the end of 1981, was to use electromagnetic isotope separation (EMIS) as the primary technology. The strategy foresaw the development of industrial-scale plants with production capacities of 15 kg/year of highly enriched uranium (HEU - 93%), based initially on natural uranium feed. Gaseous diffusion was chosen as a subsidiary technology with the declared objective of building a plant to produce 5 tonnes/year of low enriched uranium (LEU) containing 4% U-235 to be used as the feed material for the EMIS plants. Assuming that the EMIS plants could have been optimised to use LEU feed material, the combination of the two technologies could have more than tripled the capacity of each EMIS plant.

Other technologies such as gas centrifuge enrichment and laser isotopic separation (LIS) were not included in the initial strategy because of their greater technical complexity and dependency on equipment subject to export controls. Nonetheless, LIS and chemical and ion-exchange uranium enrichment processes were explored, although, according to the Iraqi counterpart, only centrifuge technology was taken beyond laboratory-scale exploitation.

In 1987, faced with what Iraq considered to be overwhelming difficulties in the further development of gaseous diffusion technology, reduced priority was given to this programme and the released resources were assigned to the development of gas centrifuge enrichment.

1.2.1 Electromagnetic isotope separation (EMIS)

According to the Iraqi counterpart and substantiated by PC-3 documentation, the EMIS development programme was organised into three phases with the first phase concentrating on research and development activities using "R40" magnet/separation chambers. These units which were designed to have ion-beam paths of radius 40 centimetres, were 1:2.5 scale versions of the anticipated production-scale units. Phase one was established in Tuwaitha and continued over the period 1982 through 1987. It involved the construction and operation of an electromagnet (Project 101) and two different magnet/separator systems (Projects 102 and 103) all of which were in operation in Building 85 from the beginning of 1985.

The second phase, which overlapped phase one, commenced in 1983 and reached an experimental stage in 1987. Phase two was devoted to development of R50 and R100 pre-production-scale units (Project 104), as well as 1:5 scale model units (Project 105) which were used to investigate multi-magnet series operation as an analytical tool for the production phase configuration. Starting from 1985, a total of one R50 and three R100 magnet/separator systems were built and installed in Building 80 at Tuwaitha and were operated until 1991. According to programme progress reports obtained by IAEA-6, none of these separators achieved more than 20% of their design capacity. This performance is in keeping with Iraq's declaration that the total production of enriched uranium from the development separators at Tuwaitha was only 640 grams with an average enrichment of 7.2%.

The design work for the third phase, the production phase, which proceeded concurrently with the other two phases, was finalised in 1987 and foresaw two identically equipped industrial scale plants, Al Tarmiya and Al Sharqat, each with 70 R120 separators for the production of uranium enriched to about 20% and with 20 R60 separators for the production of HEU (93%). The design production of each facility was 15 kg HEU per year, based on natural uranium feed, with the potential of a more than three-fold increase in that production by using LEU as the feed material.

A foreign civil engineering contractor was employed to construct many of the principal buildings at Al Tarmiya but according to Iraq, there was no foreign involvement in the construction of Al Sharqat.

Iraqi records show that installation and commissioning of R120 separators at Al Tarmiya commenced at the beginning of 1990 and that, by the time of the Gulf War, a total of eight R120 separators were in limited operation. Preparations had begun for the second group of seventeen R120 separators to be installed but nothing was accomplished, Iraq's declaration of the total enriched uranium produced at Al Tarmiya as some 685 grams at an average enrichment of 3% is equivalent to only about 20% of design, both in terms of mass and enrichment, but is not inconsistent with the reduced performance that might be expected during commissioning.

Iraq states that it had interrupted operations on the 15 December 1990 and that the damage caused by the bombardment prevented re-commencement.

Construction of the sister facility at Al Sharqat was about 80% complete at the end of 1990. There are no indications to suggest that any EMIS process equipment was ever installed.

1.2.2 Gaseous diffusion uranium enrichment.

a. Background

Iraq declared the existence of a programme to develop the gaseous diffusion process for uranium enrichment to IAEA-3, which arrived in Iraq coincident with the issue of Iraq's 7 July 1991 declaration, which did not include this information. Iraq stated that exploratory work on gaseous diffusion technology had commenced in 1982 with the intention of developing the capability either to directly produce highly enriched uranium or to produce low enriched uranium for use as feed material for the EMIS process. The Iraqi counterpart explained that work had initially concentrated on the development of suitable porous barrier material, on obtaining a theoretical understanding of flow through porous tubes and on diffusion plant cascade design. By 1985 some progress had been achieved in producing barrier material, therefore effort was also placed on compressor, diffuser and heat exchanger design. It rapidly became apparent that a very large industrial infrastructure would be required to manufacture these items and that this infrastructure was beyond the national capabilities at that time.

It was further explained that a decision had been made in 1987 to revise the mission of the team assigned to this task (Group One) such that priority was to be given to the exploitation of gas centrifuge technology for uranium enrichment. Some work on the gaseous diffusion process did continue, although it was limited to research and development on the barrier material and on carrying out practical tests on some compressors that had been procured. Iraq stated that its attempts to reverse engineer a screw compressor procured from the UK were unsuccessful.

b. Research and development

Work had commenced in 1982 with literature surveys of data on separation barriers, followed by experiments on porous tube manufacture and on the characterisation of porous materials. A number of materials, in various forms and deposited by various methods, were investigated during the following three years with little success, due to excessive pore size and unsatisfactory flow characteristics. Iraq claims that a suitable barrier material was developed in 1988 which overcame these adverse properties, but that the barrier tube was still found to be mechanically weak in industrial-scale handling.

In parallel with the above, a survey of compressors judged to be suitable for transporting the process gas was made and specifications were obtained from potential suppliers. Procurement action was taken to purchase compressors from companies in the USA, Germany, France and the UK and attempts were made to locally manufacture a compressor casing, but these were not successful. In 1987 design drawings of a screw compressor were made by reverse-engineering a screw compressor that had been procured from the UK. However, it was soon realised that reproduction of its components was beyond the capacity of the existing national engineering resources and, although some attempts were made to secure foreign assistance, nothing materialised. Concurrent with these activities, a facility for testing compressors was built at Rashdiya but, according to the Iraqi counterpart, was never commissioned due to the change in emphasis of the programme in favour of the centrifuge enrichment process.

Theoretical work on diffusion cascade behaviour and calculation of the performance of a total cascade made up of different sized stages acting in "square" cascade array were carried out. Those calculations were for various cascade sizes ranging from 16 stages in series to 72 stages in series. Theoretical calculations aimed at optimising the geometry and flow parameters of the diffuser were also made.

Facilities were constructed initially at Tuwaitha, then later at Rashdiya, to test the theoretical models of the barrier design and the diffuser. These test facilities included capabilities to check barrier porosity, permeability, robustness and gas flow dynamics for tests with inert gas and with hydrogen fluoride (HF), fluorine (F) and the process gas (UF6). Iraq states that, although a number of facilities to test barrier performance in UF6 were planned, none were completed.

Barrier manufacturing facilities were commissioned to investigate the various proposed manufacturing processes, culminating in a laboratory scale production facility capable of making 18 test barrier tubes per day - several hundred were produced during its operating life-time. In 1986 Iraq proceeded with the plans to test a single barrier tube with UF6. The tests were stated to have been carried out at Rashdiya in 1988, within Project 365, where one barrier was exposed to UF6 for about four months and Iraq claims that promising results were obtained.

Iraq further planned to measure the separation factor of a complete single stage unit, initially using a mixture of two freons having very different molecular weights. A separation-test facility was constructed at Tuwaitha but severe difficulties were experienced in assembly due to the lack of robustness of the barrier tubes. Many were broken before an engineering solution was achieved. However, before the facility was commissioned, the entire project was moved to Rashdiya. The facility was dismantled and transferred to the new site and according to the Iraqi counterpart was never rebuilt.

In 1988 a barrier tube suitable for operation in UF6 was successfully manufactured. The separation performance of a single unit (or stage) was theoretically determined and planning commenced on Project 366 through which to assess the barrier efficiency of 24 stages operating in series. The Iraqi counterpart states that this plan was never completed and that the project was cancelled in 1989. Two further facilities to measure the separation factor in UF6 gas of a single diffuser stage unit and of 48 diffusers acting in series were also planned. The design of the former was completed but, due to the revised programme priorities established in 1987, was never constructed. According to the Iraqi counterpart, the design of the latter was never completed and the project was stopped when still at the basic design stage.

1.2.3 Gas centrifuge uranium enrichment

a. Background

As described by the Iraqi counterpart the team responsible for the development of gaseous diffusion technology (Group One) became independent from PC-3 in August 1987 and was renamed the Engineering Design Directorate - eventually to become the Engineering Design Centre (EDC). At the same time it relocated from Tuwaitha to premises (Rashdiya) in the north-western outskirts of Baghdad, which had formerly been a Ministry of Irrigation research and development establishment. The relocation was coincident with Iraq's recognition that the establishment of the engineering infrastructure that would be necessary to exploit gaseous diffusion on an industrial scale was beyond Iraq's current capabilities. Consequently, it was decided to focus the resources of EDC on the development of gas centrifuge enrichment technology with the aim of establishing a production capacity of 10 kg of highly enriched uranium (93% - HEU) per year by 1994. The facilities on the new site were rapidly expanded and modifications to existing buildings and new building construction continued until early 1991, as work on the centrifuge enrichment process gathered momentum.

Very little technical documentation is available to support Iraq's description of its work on gas centrifuge enrichment technology. There are very few technical reports and not one single example of an official programme report coded in accordance with the system described in the FFCD. However, Iraq has made available to the IAEA a large number of technical drawings from which it has been possible to understand the progression of the design of the various types of centrifuge machines considered in Iraq's development programme.

b. Research and Development

Work commenced in August 1987 with an attempt to develop the oil-bearing (Beams type) gas centrifuge for which extensive design information was available in open US literature. EDC's technical capabilities developed rapidly and, by late 1987, the first oil centrifuge (GS-1) was built and subjected to laboratory trials. Rotational speeds greater than 30,000 rpm could not be achieved due to vibration, high power consumption and vacuum difficulties.

In the face of these difficulties in the summer of 1988 EDC sought foreign assistance through H&H, a German company already involved in the supply of specialist machine tools to Iraq's armaments industry. H&H introduced two foreign nationals who had previously been employed by MAN - a German company that had, in the 1970's and early 1980's, been involved in the design, development and supply of centrifuges to URENCO, the European centrifuge enrichment company which produces low enriched uranium (LEU) for nuclear power station fuel. During the next 2 years the difficulties with unbalance and vacuum were gradually overcome as rotor dynamics and bearing know-how was learnt, with guidance from the ex-MAN employees, and by the import of high quality balancing machines and drive units. By mid-1989 a speed of 50,000 rpm was achieved in vacuum. These mechanical trials were followed by separation tests using a mixture of freon and carbon dioxide gas to simulate uranium hexafluoride (UF6) gas, the medium used