Print Page

Centrifuges and Nuclear Weapon Proliferation 

by Ivan Oelrich and Ivanka Barzashka

nuclear blast
Even moderately technically advanced nations are able to build and operate gas centrifuges for uranium enrichment. This spreading capability is creating dangerous opportunities for nuclear weapon proliferation.  There are two routes to producing a nuclear weapon: one using highly-enriched uranium (HEU) and the other using plutonium.  The simplest possible bomb using highly enriched uranium is easy to fabricate but, in the past, producing the uranium has been an extraordinary technical and industrial challenge.  Plutonium does not occur in nature and must be produced in a nuclear reactor.  While building and operating a reactor and chemically separating the plutonium are not trivial, in practice, it has been easier than producing HEU.  However, the design and construction of a plutonium bomb is much more complex than the simple uranium bomb.  Therefore, potential proliferators have two paths to a bomb, but either option presents a roadblock.  The uranium route allowed simple bomb design but getting the material was difficult. The plutonium route allowed easier material production but manufacture of the bomb was difficult.  Gas centrifuges are changing this equation by making uranium enrichment easier. Thus, the simple bomb design can use the simpler material production, creating an open road to a nuclear weapon.

Natural uranium is composed of two isotopes – atoms with slightly different weights that are essentially identical chemically but that can have very different nuclear properties.  The bulk of natural uranium is uranium-238 (U-238), but it is the uranium-235 (U-235) isotope (composing less than one percent of natural uranium) that is the fissile material that powers nuclear reactors and nuclear bombs.  To be useful, U-235 must be enriched (or increased in concentration) to 3 to 5 percent for a reactor and to about 90 percent for a bomb.
Early in the nuclear age, uranium was enriched by the gaseous diffusion method.  Gaseous diffusion is inherently inefficient; the enrichment achieved in a single step of the process is very small so the material must be passed through thousands of steps.  As a result, gas diffusion plants are enormous and expensive to build.  After each step of the process, the uranium hexafluoride (the gaseous uranium compound used as feed for the diffusion plant) must be recompressed, which makes gas diffusion plants prodigious consumers of electric power.  It is no coincidence that the first gas diffusion plant built during the World War II Manhattan Project was located in the middle of the vast Tennessee Valley Authority hydroelectric project – it was the only place in the U.S. that could provide enough electric power.  Because of the difficulty and cost of building and operating gas diffusion plants, only wealthy technologically advanced countries were able to enrich uranium, whether for nuclear reactors or nuclear weapons. Industrial scale gaseous diffusion plants were operated in all declared nuclear weapons states .

High-speed gas centrifuges are changing this picture.  Gas centrifuges date back to before World War II, but a range of technical problems stopped their industrial application.  However, steady technical advances have gradually improved the performance and reduced the cost of centrifuges.  High-strength materials and precision machine tools have become more widespread, allowing ever more nations to produce centrifuge parts.  Today, centrifuges are the method of choice for uranium enrichment, although some gas diffusion plants are still in operation. In 2007, 72 percent of all enrichment was done by centrifuge plants and only 28 by gaseous diffusion. By 2017 centrifuges are projected to account for 96 percent of all uranium enrichment activity .

Current centrifuge facilities have been built ostensibly to enrich uranium for nuclear reactors.  Centrifuges present a proliferation danger because precisely the same machines that produce low enriched uranium fuel for a nuclear reactor can produce highly enriched uranium suitable for nuclear weapons. Even a modest centrifuge plant, one sized to fuel a single nuclear power plant, can produce enough HEU for about twenty bombs in a year.  The difference between a military and a civilian enrichment plant is just how the machines are piped together in cascades. This means that a fuel plant could quickly be converted into a weapon plant.  Moreover, if the starting material for a bomb program were not natural uranium, but fuel-grade uranium that has been stockpiled, then the time to further enrich the uranium to bomb-grade would be reduced by more than 65 percent. Details of these calculations are provided in the section on Separation Theory.

IAEA Broken Seals, DPRKCentrifuges also make clandestine enrichment for weapons harder to detect and verify.  Because of their size and huge power consumption, gas diffusion plants are difficult to keep hidden from, for example, overhead satellite photography.  A centrifuge enrichment plant is more compact and might use as little as one tenth of the power of a diffusion plant.   There may be some hope of detecting leaking uranium hexafluoride gas accidently released from a gas centrifuge plant but, in general, a clandestine plant would have no characteristics that would distinguish it easily from any large light industry facility.  A centrifuge plant could also more easily be hidden underground or hardened against attack.  For example, satellite photographs  suggest that the Iranian plant at Natanz is covered by thick concrete and a deep layer of earth [2].

While designing and building centrifuges has become easier, it is still not yet easy.  A nation could secretly develop centrifuge capability by importing key components and assembling centrifuges domestically. Manufacture requires access to materials, parts and precision machine tools that are not universally available and are unusual enough that monitoring their trade internationally is feasible.  Therefore, it is possible to limit the proliferation and diversion of sensitive technology and curb the development of this potentially dangerous technical capability. To this objective, the International Atomic Energy Agency (IAEA) has defined an export “trigger list”, which is a comprehensive directory of suspect imports [1]. Below are listed a few key items that could be controlled to thwart development of a centrifuge program.
IAEA electronic device for the identification of nuclear material
Centrifuge performance depends critically on the speed of rotation of the rotor, which depends, in turn, on the strength of the rotor material (or, more precisely, on its strength to density ratio).  Most efforts to build centrifuges begin with 700-series  aluminum-zinc alloys as rotor material.  Such alloys are produced by only a few manufacturers and their international trade is monitored under IAEA regulations.  Maraging steel has an even higher strength-to-density ratio.  Maraging steel is a complex material whose properties depend not simply on its composition, but on the careful control of its annealing. Both of aluminum and maraging steel must be shaped with precision machine tools that are becoming more common but are still rare enough that they are monitored under IAEA agreements.  Today’s highest performance centrifuges use fiberglass or carbon fiber rotors with extremely high strength-to-density ratios.  While these materials may be easier to obtain, their precision manufacture is challenging and also requires special equipment.

Other key components of centrifuges are the high-speed motors and the variable-frequency power supplies that drive them.  The power supplies have limited industrial applications and allow some opportunity for monitoring and control.  Ancillary and test equipment is also important, including vibrational test systems, machine tools, multi-plane balancing instruments, mass spectrometers, vacuum pumps, and certain chemicals.

NPT SigningLike diffusion plants, centrifuges use a gaseous form of uranium, uranium hexafluoride or UF6.  The UF6 is produced by chemical conversion of yellow cake (predominantly U308) obtained by milling uranium ore. Uranium exports are monitored by the IAEA; transfer of both source material (natural or depleted uranium) and special fissionable material (Pu-239, U-233 or uranium enriched in U-235 or U-233) above a certain quantity should be reported.  Although countries are obliged to report transfer of uranium abroad, domestic uranium mining and milling does not fall under IAEA safeguards.  The Safeguards Agreement supplemental to the Nonproliferation Treaty prescribes physical IAEA monitoring of both source and fissionable material in all peaceful nuclear activities inside a country, except for mining and milling, which means that verifications start at the conversion plants. Unlike gas centrifuge components, the equipment and feed chemicals needed for producing UF6 are rather difficult to monitor since many of the key items are common to the chemical industry as a whole. Even though equipment has to be especially ordered, ordinary components can be assembled to into a system designed especially for uranium conversion.

Centrifuge plants do not have to be clandestine to be a proliferation concern.  Because only minor changes are required to convert a civilian enrichment plant into a nuclear weapons material plant, every fuel production centrifuge plant is a potential source of nuclear weapons material.  Iran, for example, claims that its centrifuge program at Natantz is entirely for peaceful purposes, specifically to provide fuel for a nuclear reactor.  While the international community is not entirely confident of Iran’s claim, and IAEA inspections confirm that currently Iran is not producing highly-enriched uranium but the same centrifuges could be reconfigured to for weapons production.

The potential dual function of all centrifuge plants calls for continuous international inspection to insure that ostensibly commercial plants are not used to produce nuclear weapon material.  While technical advances are making centrifuge manufacture and operation ever easier, it is still possible to control key technologies and slow the diffusion of enrichment capability.

[1] INFCIRC/254/Rev.8/Part 1a. (2006, March 20). In International Atomic Energy Agency (IAEA). Retrieved March 19, 2009, from
[2] Natanz - Iran Special Weapons Facilities. (n.d.). In . Retrieved March 19, 2009, from