URANIUM ENRICHMENT TECHNOLOGIES: PROLIFERATION IMPLICATIONS

by Sean Tyson

Since the construction of the gaseous diffusion plant at Oak Ridge, Tennessee in 1944, hundreds of methods have been pursued in the search for a more efficient way to enrich uranium. Of these, ten have been developed to the point where they have either been in commercial production or are approaching commercial production. Others possess special characteristics that make them attractive for weapons production regardless of economic cost. This essay discusses the ten developed processes and will also review three other processes which have strong potential for future development. A review of the proliferation-relevant features is presented below in both narrative and tabular form, with a brief description of the status of each process and comments on possible future dissemination of the technology.

I. THE ENRICHMENT PROCESSES

Gaseous diffusion, a relatively simple technology, has been successfully developed for an enrichment program by almost every country that has attempted it, including Argentina, the US, the USSR, France, and the PRC. The development of a diffusion barrier has proved to be a temporary obstacle at most. One of the main attractions of gaseous diffusion is that it is a proven technology. Most costs for the process were capital costs, which the five aforementioned countries have long since recovered. It has therefore been profitable for traditional suppliers to continue using this process, even with the advent of such developments as the gas centrifuge. In addition, gaseous diffusion plants are capable of producing tons of product annually. For countries developing new enrichment programs, however, a number of characteristics of gaseous diffusion make it undesirable as an enrichment process. These include the fact that the separation factor is low, requiring multiple passes of the feed through a cascade, which consumes more energy. Tables 2 and 3 show that while gaseous diffusion is a relatively widespread technology, with common components and materials, the disadvantages outweigh the advantages from the standpoint of economics and in terms of a covert weapons program. More specifically, gaseous diffusion requires a large plant size, and a long equilibrium time.

Gas centrifuge technology has proven to be a popular route for countries interested in uranium enrichment. Table 1 shows the higher separation factor, low energy consumption and common feed which combine to make the process attractive both from economic and weapons program perspectives. Despite the obstacle of export control lists, some countries have shown that with the proper resources, the technology can be acquired (e.g., Pakistan's acquisition of technology from Urenco and Fragema). Centrifuge technology may be the most widely disseminated enrichment technology, and it is being further developed while newer methods of enrichment are appearing. In Table 3, gas centrifuge technology is the only process with positive marks in every category, indicating a very great possibility for future proliferation.

Aerodynamic processes do not appear to be economically competitive with centrifuge processes because of their high energy consumption and low separation factor, nor with diffusion processes due to a slightly greater energy consumption. Using a carrier gas with the feed increases the cost of production. In addition, characteristics of the hydrogen gas in the feed mix make it necessary to take further measures at additional expense to ensure safe handling. If cost is no object, however, an aerodynamic process may be attractive for purposes of a weapons program. The process requires a relatively small plant, the modularization of the cascades simplifies conversion to a high enrichment level, and the time involved to produce the product is comparable to that required in gaseous diffusion. Although South Africa has had some success with the Helikon process, Brazil recently terminated its own jet nozzle (Becker) project at Resende due to cost and engineering difficulties in developing the technology. Its success in developing a centrifuge enrichment plant may also have had an effect on the decision to halt development of the aerodynamic plant. The production capacity of the FRG's jet nozzle plant at Karlsruhe is only a fraction of the production capacity at its centrifuge plant at Gronau.

Chemical-exchange processes are possibly the last route that would be taken by a country wishing to acquire enrichment capabilities for a covert enrichment program. Although this approach uses little energy or maintenance, it involves a long equilibrium time and large facilities. To date, only two chemical exchange methods have been developed; the Asahi Chemical Exchange Process and the French Chemex process. France has offered to sell its Chemex process to countries on the condition that they not pursue other enrichment paths. Details of the Asahi process are not readily available. Even if a catalyst is found, the process is inefficient in enriching uranium above 5-6%. Without a second process to enrich the uranium further (e.g., EMIS), producing large amounts of highly-enriched uranium via a chemical exchange is not feasible. But if a country's program were to utilize two enrichment processes, the chemical exchange process could be used to produce large amounts of reactor-grade fuel, part of which could then be diverted for use as source material for the second process (e.g., a small centrifuge program). This initial enrichment would greatly increase the effectiveness of the second process. However, chemical exchange in itself is not useful for a covert weapons program.

Electromagnetic processes, with the notable exception of the calutron, are in such an experimental stage of development that they are unlikely options for a national enrichment program, although they are theoretically as efficient and economic as laser processes. The three electromagnetic processes (Electromagnetic Isotope Separation (EMIS), ion cyclotron resonance, and plasma centrifuge) are similar in that they require frequent maintenance to remove feed from the collection chamber for recycling, though EMIS does appear more wasteful in this regard. They may also be useful for the separation of plutonium isotopes, a process otherwise practical only through laser methods. In addition, the three need little space for an enrichment facility, and could easily handle high-level rather than low- level enrichment. As can be seen in Table 2, all three methods are capable of high enrichment in a single stage, and the two modern methods (ion cyclotron resonance and plasma centrifuge) use less energy than gaseous diffusion or aerodynamic methods. EMIS, however, does require vast amounts of energy.

The components for the two advanced methods are not common. In contrast, those used for EMIS are commonplace, and, until the discovery of Iraq's program, these components were not subject to export controls.

Like the electromagnetic processes, the different laser-related processes (Atomic Vapor Laser Isotope Separation (AVLIS), Molecular Laser Isotope Separation (MLIS) and Laser-Assisted Processes (LAP)) are currently under development and are therefore unlikely candidates for enrichment programs. Although their capabilities seem impressive, to date, all tests have been conducted on a laboratory scale only, so their performance on a commercial scale is questionable at present. As shown in Table 1, MLIS and LAP are potentially the most efficient types of enrichment technology available because of their high separative ability, low energy use and ability to use common feed forms (UF6 gas). Although much of the technology associated with the processes is no longer subject to export controls, the level of technology is advanced and the difficulty in integrating different technologies for this enrichment process is great.

Table 2 shows a large number of countries that are to some degree involved in laser-related research, but it should be noted that the most recent developments have all been made in technologically advanced countries such as the US and Japan. Even if a developing country acquires the laser equipment and technology, as Pakistan (AVLIS) and South Africa (MLIS) recently have, it will have to do extensive research to develop the process into a usable system.

Table 3 shows that the AVLIS and MLIS processes share several characteristics that make them attractive from a prospective proliferator's point of view: their enrichment plants would require little area, conversion from low to high grade enrichment is not difficult, energy use is low, and they have a short equilibrium time. Add to that the fact that they are particularly suited to separating isotopes of plutonium, and the two may be the most desirable methods in existence. The AVLIS enrichment process is in an advanced state of development in the US; if a more desirable feed is developed, AVLIS will likely become a common enrichment method in the coming decades. Although the AVLIS process as conceived by the US employs metallic uranium fabricated from UF6 as feed (a routine but costly process), British Nuclear Fuels, Ltd., among others, is working to produce the feed directly from natural uranium, avoiding the complication of producing UF6 and its associated costs. AVLIS's subsequent dissemination among advanced countries will increase opportunities for proliferation among those countries now considering enrichment alternatives. MLIS has not been developed in the US to the same extent as AVLIS, but research is continuing elsewhere, and it may also be employed on a commercial scale in the future. The CRISLA process has been researched in the PRC as well as the US and Japan. Some promising results have been reported but no independent tests of its potential have been made.

Thermal enrichment processes have not been employed on a large scale since World War II, and considering that the U.S. dropped it as soon as its gaseous diffusion plant at Oak Ridge began operation, it may not be economically competitive with gaseous diffusion, and may not have a comparable throughput. But considering the relatively simple technology and low maintenance of the process, it, like EMIS, is a real alternative to other advanced techniques. In fact, like the chemical-exchange method, thermal diffusion could be used in tandem with another enrichment process (e.g., EMIS) to produce highly-enriched uranium. This was the route the US took in World War II.

II. FUTURE ALTERNATIVES

The Iraqi nuclear program has challenged the nuclear nonproliferation regime by demonstrating that common materials could be employed to pursue "an obsolete method" of enriching uranium for a nuclear weapons program. By employing technology 40 years old, the Iraqi program successfully evaded the controls enforced by the traditional suppliers. This glaring gap in the nonproliferation regime's controls has made it necessary to reevaluate not only current technologies, but all older technologies that are to any degree effective. While the control lists have concentrated on the most common modern enrichment technologies, the possible role of both the old and the new must be considered in a future nonproliferation regime.

In summary, one can note the following points about alternative enrichment technologies:

1. Gaseous diffusion is economically unattractive for most countries and is undesirable for a covert weapons program.
2. The gas centrifuge approach remains relatively economical, and improvements to the process are likely to increase its capability in the future. Considering the number of countries that now possess or are developing the technology, it will probably continue to be a viable option as long as the more modern technologies are unavailable.

Aerodynamic processes have not proved to be economically competitive with the gas centrifuge or most technologically advanced methods, although they have produced more enriched uranium than any of the developing technologies. In Brazil's case the process also has been beset by engineering difficulties and high development costs. Nevertheless, aerodynamic processes may produce a given amount of highly-enriched uranium more efficiently, and possibly faster than the chemical-exchange methods, and it is not difficult to increase the enrichment level for a cascade or a facility. Because of the small area required, it is also relatively easy to conceal an aerodynamic enrichment plant, despite its high energy demand. A prospective proliferator, therefore, might be interested in the approach.

Chemical-exchange processes are economically attractive, but, by themselves are useful mainly for producing low-enriched uranium. Countries are unlikely to develop chemical enrichment processes indigenously, since France has offered its technology, albeit with restrictions. However, in conjunction with a secret enrichment program (e.g., EMIS), a chemical plant could provide the feed for a process that would generate the highly-enriched uranium necessary to build a nuclear weapon. A close watch would certainly have to be kept on the production of such a plant and the distribution of its product.

The state of the two modern electromagnetic processes (ion cyclotron resonance and plasma centrifuge) is such that a judgment is not possible regarding their future use. EMIS is another matter. Even with newly- enacted export controls, knowledge that the technology works, coupled with the fact that the necessary components are widely available to any country makes it a logical choice for the few countries that have great resources and covet nuclear weapons.

At a theoretical level, and in the laboratory, various laser methods appear to be ideal enrichment processes, as they proved to be capable of high levels of enrichment with relatively little energy use, and would require small plants. Should they prove to be economical in practice, it is likely that they will become common within the next two decades. As more countries acquire the technology, some of the technology will inevitably spread to countries without safeguards.

The liquid thermal process has not been used since World War II. However, like EMIS, this process is simple and the components used in it are not difficult to acquire. Although the plant size is greater than for other methods, a small plant could be used with a second process providing added enrichment, and thus could produce the material for a country's covert nuclear weapons program.

III. THE TABLES

Table 1 shows characteristics of the enrichment processes that may affect a country's interest in acquiring the technology, and the availability of the materials or techniques needed: The separation factor is basically the proportion of product to the tails, or waste that is extracted from the feed, which contains two isotopes, Uranium-235 (235U) and Uranium-238 (238U). The greater the separation achieved in one pass, the fewer passes needed to achieve a desired level of enrichment. This will directly affect both the number of units or stages needed and, subsequently, the size of an enrichment facility and the energy required by the process. Energy use is expressed in kilowatt hours per Separative Work Unit (SWU). A low energy requirement is attractive economically, and is of use in concealing an enrichment facility, since the additional energy requirements of a covert enrichment facility would be easier to disguise. Feed shows the form in which the uranium compound is introduced into the enrichment process. Uranium hexafloride (UF6), though highly corrosive and requiring certain specialized equipment, is the most common form used in today's enrichment processes; its characteristics are well understood, and because it is used in many enrichment technologies it is cheap to produce and it is relatively easy to acquire. A problem with the AVLIS enrichment system developed by the US is that it is currently planned to use as feed metallic uranium produced from UF6. This conversion entails additional costs, and to produce amounts that would ensure competitive prices for the AVLIS product, the uranium industry would need to be assured of a reliable demand for the product, in an amount sufficient to cover the additional cost in producing the feed. Finally, Critical requirements show those components and skills that are absolutely essential when developing an enrichment technology. Most of the required technologies are subject to export control, and as a result the would-be enricher would have to utilize some form of subterfuge in acquiring any of them. Also, even if the actual components are available experience and extensive training is needed to employ the equipment.

Table 2 shows the degree to which an enrichment process has been disseminated internationally, and the highest state of development reached. Not all countries have the same capabilities, and some of the technologies have not been employed for uranium enrichment for decades. For example, under electromagnetic enrichment processes, countries listed as possessing EMIS enrichment technology are the United States, Iraq and the USSR; the US used over 1000 calutrons in World War II for its enrichment program, and subsequently abandoned the technology when gaseous diffusion enrichment became available. Recently, Iraq built functioning EMIS separators ("magnetrons"), but its program had barely progressed beyond the research and development stage. However, the table does give an indication of how widespread a technology is, which may increase the potential suppliers for a country interested in the technology. And if a process has reached a high state of development, then it will be easier to acquire or replace components related to that technology.

Table 3 describes the different processes in terms of positive and negative proliferation factors. The values given are relative, based on comparisons between the processes. The chart is not meant to be comprehensive, only to assist in a brief analysis of the technology in terms of its advantages or disadvantages for future proliferation. Technology proliferation is a relative judgment as to how widely available a technology is; all else being equal, the more widely disseminated the technology is, the easier it will be to acquire. Component availability is an indication of both the difficulty in producing the critical components, and of the difficulty in importing them. This takes account the effect of export control lists such as that of the London Suppliers Group. The training level is more of an indication of the experience in and knowledge of a process than of the actual education required. For example, the chemical processes by and large employ techniques widely understood throughout the chemical and petrochemical industries, whereas it is unusual to find the ability to integrate the various disciplines and skills required in AVLIS in national enrichment development programs. The plant size required for an enrichment process may render a covert enrichment program easy to detect or impossible to find. A gaseous diffusion plant may cover hundreds of acres whereas a laser enrichment plant requires very little space. It is also the case, however, that a plant dedicated to small-scale production of weapons-grade material (e.g., 100kg of 90% enriched 235U/year, enough for four warheads) need not be a very large facility, regardless of the technology involved. Conversion ease shows the possibility of converting a legitimate enrichment program into one for weapons production. Processes requiring a small inventory and possessing a high enrichment capability will be preferable. A process using a common feed will be easier to utilize, and cheaper, than one employing nonstandard feed. Whether a process is economically profitable may determine its suitability to an enrichment program--Iraq's program clearly demonstrated that this is not necessarily a deciding factor, however. Low maintenance refers to the amount of direct attention required for operation of the equipment as well as how prone the equipment may be to frequent failures. As was mentioned above, a low energy consumption is desirable in any process. Finally, a short equilibrium time, such as that in a electromagnetic or laser process, is preferable to a long equilibrium time, such as that in a chemical process.

Table 2: Technology Holders and Status of Development

Table 3: Proliferation Sensitivity

(Table 2 and 3 are not listed in this file, due to their lengths.)

SOURCES, Tables 1 and 2

1. Albright, David, "Reprocessing and Enrichment Technologies," April 1991, from Cochran, Thomas B., et al, "Nuclear Materials Production Technologies and Processes,"1991.
2. "China Defends Nuclear Sales to Iran," San Francisco Chronicle, 11/5/91, p. A9.
3. Cochran, Thomas B., et al, "Nuclear Materials Production Technologies and Processes," Nuclear Weapons Databook, Volume II; US Nuclear Warhead Production (Cambridge, MA: Ballinger Publishing Co., 1987).
4. CRS Report For Congress, Non-Proliferation: A Compilation of Basic Documents on the International, U.S. Statutory and U.S. Executive Branch Conponents of Non-Proliferation Policy, (Washington, D.C.: Congressional Research Service, 1990).
5. Eerkins, Jeff W., "Uranium Enrichment With the CRISLA Process," June 1989; paper prepared for USCEA International Enrichment Conference, 6/18- 21/89, Monterey Plaza Hotel, Monterey, CA.
6. Emerging Nuclear Suppliers and Proliferation Project, DOC 54, DOC 109, DOC 1185, DOC 2387, DOC 3569, DOC 4980, DOC 5224.
7. Fujii, Yasuhiko, "Recent Trends and Technical Development of Isotope Separation and Chemical Exchange Uranium Enrichment," Atoms In Japan (November 1990).
8. International Nuclear Fuel Cycle Evaluation, Vol. 2, Enrichment Availability (Vienna: IAEA, 1980).
9. Kerst, Ronald A., "Proliferation of Uranium Enrichment Technology," 1991; unclassified document, Z Division, Lawrence Livermore National Laboratory.
10. Kidwai, Azim, "Nuclear Breakthroughs, Laser Export Reported," Nuclear Developments, 8/21/91.
11. Krass, Allan S., et al, Uranium Enrichment and Nuclear Weapon Proliferation (London: SIPRI, 1983).
12. Lewis, Paul, "U.N. Suspects Iraq Has A Fourth A-Plant," New York Times, 10/3/91.
13. Rhodes, Richard, The Making of the Atomic Bomb (New York: Simon and Schuster, 1988).
14. U.S. Dept. of Energy, "Nuclear Proliferation and Civilian Nuclear Power, Volume II: Proliferation Resistance," December 1979.
15. "The World's Nuclear Fuel Cycle Facilities," Nuclear Engineering International, 12/87.