National Fuel Stockpiles: An Alternative to a Proliferation of National Enrichment Plants?

Frank N. von Hippel

Iran's national uranium-enrichment program has provoked international concern, evidenced by UN Security Council sanctions and even threats of possible military attack, because it provides the means to make highly enriched uranium and thereby a fast route to nuclear weapons. Iran's program is also the current focus of the continuing international debate over the spread of national enrichment and reprocessing capabilities.

If it were possible to persuade Iran to give up its national enrichment program, the movement to satisfy the world's needs for enrichment services without constructing new national facilities would be strengthened.

Iran is not the only non-nuclear-weapon state with an enrichment capability. Argentina, Brazil, Germany, Japan, the Netherlands, and South Africa also have enrichment capabilities[1] that currently are of less proliferation concern for various reasons. Germany and the Netherlands are integrated into Europe and deeply opposed to nuclear weapons. Japan has had huge quantities of separated plutonium as well as enrichment capabilities for decades but has made no move to develop nuclear weapons. Argentina, Brazil, and South Africa all renounced their nuclear-weapon ambitions in 1991.

Iran, by contrast, is a nation with a long history both as an independent regional power and a target for outside attack and intrigue. Even though Iran's leadership has convincingly argued that its acquisition of nuclear weapons would generate more dangers than security, there probably are, as in other countries, voices arguing that a nuclear deterrent would bolster the nation's independence.

Iranian government officials make a number of arguments for their country's uranium-enrichment program that have nothing to do with nuclear weapons. Among them:

  • Iran needs nuclear power to free more oil and natural gas for export and to prepare for the day when its fossil fuel resources are depleted.
  • Peaceful use of nuclear technology is an inalienable right of non-nuclear-weapon states-parties to the nuclear Nonproliferation Treaty (NPT) and, if Iran allows itself to be pressured into giving up that right, the United States will simply go to the next item on its list of demands for policy changes that cumulatively amount to regime change.
  • The cutting-edge technical demands of enrichment technology are good for Iran's high-technology sector.
  • Keeping and developing that technology further has become a national cause for the Iranian people.

There are, of course, rebuttals to each of these arguments.

  • The capital costs of natural gas-fired power plants are much less than those of nuclear power plants, Iranian reserves of natural gas are huge, and, as of 2003, Iran was still wastefully flaring by-product gas from its oil production.[2]
  • The next issues on the U.S. list of demands, Iran's support of Hezbollah and Hamas, do not carry the same weight as the widely shared concern that Iran is developing at least a nuclear weapons option, a perception that may be pushing some of its neighbors to seek their own nuclear weapon options.
  • The technical demands of a centrifuge-enrichment program are relatively specialized and, as illustrated by a comparison of Pakistan with South Korea, nowhere near as economy transforming as other technologies, such as automobile or semiconductor production.
  • Iran's government has promoted and orchestrated the broad domestic support for its insistence on Iran's inalienable right to enrichment technology. If it concludes that Iran has been offered something better in exchange, specifically, a guarantee against attack and an end to the broad technology, investment, and financial-service embargoes to which Iran is currently subject, then it could probably similarly persuade the Iranian public to embrace such a deal as a victory.

The Economic Argument for a National Enrichment Plant

What remains, therefore, is the original and fundamental economic argument for Iran's need for enrichment technology, that it is going to build a large nuclear power capacity (20,000 megawatts electric [MWe], the rough equivalent of 20 large nuclear-power plants, in the next 20 years)[3] and that it has learned that it cannot depend on other countries for its nuclear fuel supply.[4] This same argument has been made for the Brazilian and Japanese enrichment programs and could be made by other countries, including some of the 20 or so that recently have expressed an interest in acquiring their first nuclear-power plants.

Given the huge capital costs of nuclear power plants, it is reasonable to want to insure against a lack of fuel to operate. If Iran's nuclear fuel supply were cut off, it would lose a return on its investment of at least $200 million per year for each idled reactor.[5]

Alternatives to National Enrichment Plants

Thus far, the focus of international efforts has been to try to develop a system of credible guarantees and fuel banks to assure Iran and other states that do not yet have fully functioning enrichment plants that they will have a reliable fuel supply, conditioned only on International Atomic Energy Agency (IAEA) certification that they are in compliance with the NPT. Such guarantees, however, can never be entirely convincing to countries that are not strongly allied politically to the national host of at least one enrichment supplier. Iranian officials complain, for example, that Iran has been subject to endless investigations by the IAEA for what they consider minor infractions and, more recently, what they decry as fabricated documents in the case of alleged nuclear weapons design activities.[6] Tehran could easily argue that such controversies could delay indefinitely the IAEA's delivery of the requisite clean bill of health and subsequent guaranteed supply of low-enriched uranium (LEU).

An alternative approach, originally suggested by Richard Garwin,[7] would be to allow Iran to stockpile several years or more of imported fresh fuel. According to current estimates, Iran has very limited uranium resources. Its "reasonably assured" and "inferred" resources of uranium amount only to about 2,000 tons, enough to provide about 10 annual reloads for a 1,000 MWe reactor. Less certain "prognosticated" and "speculative" resources are estimated at about 16,000 tons, enough to fuel Iran's planned 20,000 MWe of capacity for only five years.[8] If Iran follows through on its nuclear energy expansion plans, it would in any case have to import natural uranium to supply an indigenous enrichment plant.

Cost of a Stockpile of Fresh Fuel

Stockpiling fresh fuel costs more money than buying it on a just-in-time basis. The extra cost is associated with having money tied up in the fuel instead of earning interest or having to pay interest on a loan used to buy the fuel in advance. U.S. utilities typically do not stockpile more than one year of fresh fuel. Switzerland's nuclear utilities, however, have less trust in the reliability of the market and have a general policy of stockpiling three to five years of fresh fuel.[9]

A one-year supply of fuel for a 1,000 MWe reactor costs about 1.5 percent of the capital cost of the reactor, or about $60 million (see box). A five-year fuel supply would cost $300 million, or about 8 percent of the capital cost of the reactor. Assuming a 5 percent interest rate, the extra interest cost would be $15 million per year. This is about 0.2 cents per kilowatt-hour or about 2 percent of the generation cost of electricity from a new nuclear power plant.[10]

As shown below, within large uncertainties, this cost is comparable to the extra cost of building a small national enrichment plant instead of buying enrichment services on the international market.

Cost of a Small National Enrichment Plant

Domestic enrichment of uranium for a few reactors, as proposed by Brazil and Iran, can also be more costly than buying fuel on the international market because small enrichment plants do not fully exploit economies of scale. For example, the same staff could operate a much larger enrichment plant capable of fueling tens of large power reactors.[11]

The cost disadvantage of a small plant is even larger if it does not have access to advanced centrifuges of proven reliability. Today, only Russia and Urenco have such centrifuges. Consequently, France has decided to base its enrichment programs on Urenco centrifuges, as have two of the three companies that are establishing new centrifuge enrichment plants in the United States. Similarly, China decided to use Russian centrifuges rather than its own in its enrichment plants.

Brazil has made public its past and planned investments in commercial enrichment, and Cabrera-Palmer and Rothwell have used those numbers to estimate the production costs for Brazil's enrichment plant.[12] Brazil expects to have spent about $304 million by the time that the capacity of its enrichment plant has reached about 0.203 million separative work units (SWUs) per year in 2015, enough to supply about 1,500 MWe of nuclear capacity. That capital cost of $1,500 per SWU per year production capacity, however, will still be three times the average capital cost of the large, new, multimillion-SWU plants based on Urenco technology that are being built in France and the United States.

Cabrera-Palmer and Rothwell estimate that Brazil's production cost after the expansion would be $147 per SWU, approximately today's market price. If this price does not fall, Brazil would expect to break even while the new U.S. and French enrichment plants, with their lower costs, should earn a handsome return on their investments.[13] If the international price of enrichment falls back to historic levels, Brazil will be losing money on its national enrichment program.

Brazil's enrichment program might be a money loser for another reason as well. In 1998, at a time when the world price of enrichment was about $85 per SWU, the cost of SWUs from Japan's new 1.05 million SWU enrichment plant were reportedly $150 per SWU.[14] Since that time, however, nearly all Japan's centrifuges have failed.[15] As a result, their capital cost will have to be written off over approximately 10 years instead of the 30 years assumed in the analysis of the Brazilian program. Assuming that most of the investment cost is in centrifuges, if this happened to Brazil, its costs would be increased by $10-20 million per year.

Fuel Stockpiles Versus National Enrichment Plants

Thus, within large uncertainties, the stockpile strategy of Switzerland's utilities and Brazil's national enrichment plant have comparable costs. The incremental cost of a fabricated fuel stockpile would likely be lower for Iran than for Switzerland if Iran would have to stockpile natural uranium for its national enrichment program in any case. The cost of Iran's enrichment program is likely higher than Brazil's because Iran has had to harden its enrichment program against the possibility of bombing. Iranian planners could argue, however, that they are making an investment for the future when their nation will have a large nuclear power program where economies of scale can be realized.

From a nonproliferation perspective, a national stockpile is far superior to a national enrichment plant. In any case, a 2 percent increase in the cost of nuclear power from a multiyear fuel stockpile pales into insignificance in comparison to even a modest probability that Iran's enrichment program will provoke another trillion-dollar war in the Persian Gulf region.

U.S. Policy Toward an Iranian Fuel Stockpile

Currently, it appears to be a U.S. policy objective to block Iran from acquiring a stockpile of LEU in any form. At the United States' request, Russia's fuel supply contract with Iran reportedly provides for just-in-time deliveries of fresh fuel.[16] This arrangement probably reflects a U.S. concern that Iran could use some of the LEU in a fuel stockpile to provide feedstock for its enrichment plant. Fed with LEU instead of natural uranium as feed, the potential production rate of weapon-grade uranium from an enrichment plant could be increased more than threefold.[17]

The U.S. November 2007 National Intelligence Estimate, Iran: Nuclear Intentions and Capabilities, concluded, however, that "[w]e judge with moderate confidence that Iran probably would use covert facilities-rather than its declared nuclear sites-for the production of highly enriched uranium for a weapon." This conclusion is plausible because conversion of Iran's overt Natanz enrichment facility to the production of highly enriched uranium would be immediately detected by the IAEA, which would be obliged to inform the UN Security Council, with military action being a possible result. The same would be true if Iran suddenly withdrew fresh fuel from IAEA safeguards.

If one accepts this logic, the key objective should be to establish arrangements to minimize the chance of Iran developing a covert enrichment program. Such arrangements would include, as a high priority, persuading Iran to ratify its additional protocol, which would give the IAEA authority to visit all of Iran's nuclear-related sites.

The standard pre-additional protocol safeguard agreement that Iran accepted when it joined the NPT only requires IAEA access to sites where nuclear material is present. It therefore does not include inspections of the sites where Iran fabricates centrifuges or their key components. If subject only to these safeguards, Iran could fabricate centrifuges for a covert site as well as for Natanz without much risk of detection.

Iran voluntarily complied with the additional protocol from mid-July 2003 until February 2006. The IAEA used this access to take environmental samples and successfully detected activities that Iran had tried to conceal, such as its enrichment experiments at the Kalaye Electric Company.[18] It also monitored Iran's production and storage of centrifuges and components.[19] Iran suspended this access, however, after the IAEA Board of Governors referred the issue of Iran's enrichment activities to the UN Security Council.

In addition, it would be critical for Iran to allow the IAEA to interview personnel the IAEA believed are involved in enrichment-related activities. Such access was provided during the joint IAEA-Iran effort during 2007 to clarify the history of Iran's enrichment activities.

Not a Silver Bullet

Allowing Iran a fuel stockpile is unlikely to be the silver bullet that will solve the current impasse over Iran's enrichment program. It is difficult to believe that some important political faction in Iran is not insisting on keeping the enrichment program because the nuclear weapons option it provides acts as a virtual nuclear deterrent to U.S. attack. The deterrent threat is that, unless the U.S. military is willing to occupy Iran indefinitely, which is unthinkable after the debacle in Iraq, Iran could respond to U.S. or Israeli bombing of its known enrichment infrastructure by quickly launching a covert nuclear weapons program.[20]

Even if the United States reverses its policy and agrees to allow Iran to stockpile fresh fuel, therefore, it most likely will have to accept at least a small continuing Iranian enrichment program until its relationship with Iran improves to the point where a U.S. attack became unthinkable. With a national fuel stockpile in the mix, however, Iran could opt more easily for a smaller enrichment program. More importantly, if the United States accepted a small Iranian enrichment program, it might be possible to get transparency arrangements with Iran that would reduce concerns about the possibility of a parallel clandestine enrichment program.[21]

Other countries that do not feel the need for even a virtual nuclear deterrent but, like Switzerland's utilities, do not fully trust the international nuclear fuel market may, however, find a national fuel stockpile an adequate alternative to a national enrichment plant.

The Cost of Nuclear Fuel

The cost of a kilogram of low-enriched uranium (LEU) in fabricated fuel is the sum of the costs of the original natural uranium, the separative work required to enrich the chain-reacting isotope uranium-235 (U-235), the cost of conversion from uranium oxide to uranium hexafluoride (UF6) for enrichment and back, and, finally, the cost of fuel fabrication.

Uranium. Natural uranium contains about 0.7 percent U-235. Of this, about 0.45 percent typically is separated into LEU, leaving behind 0.25 percent U-235 in the depleted uranium that is the waste product from enrichment. About 10 kilograms of natural uranium would therefore be required to yield enough U-235 for one kilogram of LEU containing about 4.5 percent U-235. The cost of natural uranium has been quite volatile of late.[1] Assuming a cost of $100-200 per kilogram, this would contribute $1,000-2,000 per kilogram to the price of producing low-enriched fuel.

Enrichment work. Enrichment work is purchased in separative work units (SWUs). If the amount of U-235 in the depleted uranium is fixed at 0.25 percent, about seven SWU are required to produce a kilogram of 4.5 percent enriched uranium.[2] The cost of SWUs has almost doubled since 2000, to $150 per SWU.[3] It may fall again as more enrichment is brought online. Assuming enrichment prices of $100-150 per SWU, enrichment would contribute another $700-1,400 to the price of producing a kilogram of low-enriched fuel.

Conversion. Uranium is first extracted from ore as "yellowcake" (U3O8). To be enriched, it must be converted to uranium hexafluoride, which becomes a gas at the low pressures inside centrifuges. When it is discharged from an enrichment plant in the form of LEU, it must be converted again to uranium oxide, this time UO2 powder, which is subsequently fabricated into the cylindrical ceramic pellets that are put into zirconium tubes to make up the fuel rods used in water-cooled reactors. The cost of conversion is currently about $11 per kilogram of uranium.[4] The contribution of the conversion of 10 kilograms of unenriched uranium to UF6 and the conversion of the kilogram of enriched uranium back to UO2 therefore add about $120 to the cost of a kilogram of fuel.

Fabrication. As of April 2006, the cost of fuel fabrication contributed about $240 to the cost of a kilogram of LEU fuel.[5] An assumed cost of $300 per kilogram is used here.

With these assumptions, the cost of fabricated fuel containing a kilogram of LEU would be $2,120-3,820. A 1,000-megawatt nuclear power plant requires about 20,000 kilograms of LEU fuel per year. At about $3,000 per kilogram, the cost of this fuel would be about $60 million, or about 1.5 percent the cost of the nuclear-power plant.


1. The average price paid by U.S. utilities between 1990 and 2005 ranged between $26-37 per kilogram of uranium. In 2007 the average increased to $85 per kilogram and spot prices increased to $230 per kilogram, "Uranium Purchased by Owners and Operators of U.S. Civilian Nuclear Power Plants," Energy Information Administration, U.S. Department of Energy, May 19, 2008. In 2008, the spot market fell to $160 per kilogram as of August 11, 2008, Uranium Intelligence Weekly, "Uranium Price Panel: $62.70/lb U3O8," August 11, 2008, p. 1.

2. If the price of uranium increased, it would make sense to pull more U-235 out of the natural uranium, but this would require more enrichment capacity. Going down to 0.2 percent U-235 in the depleted uranium, for example, would reduce requirements for natural uranium by 10 percent but increase requirements for SWUs by about 10 percent per kilogram of LEU.

3. The "restricted" SWU price shown by the Ux Consulting Company in early 2008 was $150 per SWU. See The price charged by Russia was approximately $20 less per SWU until 2006, after which it is no longer shown.

4. The price doubled since 2002. See

5. World Nuclear Association, "Nuclear Fuel Cycle," April 2006,

Frank N. von Hippel is a professor of public and international affairs at Princeton University's Program on Science and Global Security.


1. Brazil and Japan have national enrichment programs based on gas centrifuge technology. Germany and the Netherlands host Urenco enrichment facilities. Argentina is considering development for commercial application gas-diffusion enrichment technology that was part of its nuclear weapons program. South Africa has decommissioned plants that used a vortex-tube enrichment process to produce highly enriched uranium for its weapons program and low-enriched uranium for its nuclear power reactors.

2. T.W. Wood et al., "The Economics of Energy Independence for Iran," Nonproliferation Review, Vol. 14, No. 1 (March 2007), p. 89.

3. "It is important to note that the Parliament of the Islamic Republic of Iran approved an Act which obligates the government to install and commission 20,000 MWe of nuclear power plant capacity over the next 20 years." Organization for Economic Cooperation and Development (OECD) Nuclear Energy Agency (NEA) and the International Atomic Energy Agency (IAEA), Uranium 2007: Resources, Production and Demand (OECD, 2008), p. 221.

4. Always cited is the shah's 1974 loan of $1 billion to the French-led Eurodif uranium-enrichment consortium in exchange for the right to purchase 10 percent of the LEU produced and France's refusal to deliver after the shah was overthrown in 1979.

5. Assumptions include a "real" (after correction for inflation) rate of return of 5 percent on the government investment in the reactor. Belkis Cabrera-Palmer and Geoffrey Rothwell, "Why Is Brazil Enriching Uranium?" Energy Policy (2008). This payback would be a component in the nuclear utility's price for the electric power that it produces. The current average capital cost for a new 1,000 MWe reactor is about $4 billion. Jim Harding, "Economics of Nuclear Power and Proliferation Risks in a Carbon-Constrained World," The Electricity Journal, Vol. 20, No. 10 (December 2007), p. 65.

6. See "Implementation of the NPT Safeguards Agreement and Relevant Provisions of Security Council Resolutions 1737 (2006) and 1747 (2007) in the Islamic Republic of Iran: Report by the Director General to the IAEA Board of Governors," February 22, 2008, paras. 35-42.

7. Richard Garwin, communication with author, February 24, 2005.

8. OECD NEA and IAEA, Uranium 2007, pp. 221-223. Fueling a 1,000 MWe light-water reactor requires the enrichment of approximately 200 metric tons of natural uranium per year. "Reasonably assured" resources are in known deposits that have been well explored and can be recovered at less than a specified cost, in Iran's case, less than $130 per kilogram. "Inferred" resources are extensions of such deposits whose existence has been established but whose exact characteristics are not sufficiently delineated to classify as reasonably assured. "Prognosticated" resources are expected to occur in well-defined geological trends with known deposits but have not been explored. "Speculative" resources are expected somewhere within a given region or geological trend.

9. Kevin Alldred, communications with author, August 2008.

10. Nuclear Energy Institute, "The Cost of New Generating Capacity in Perspective," August 2008.

11. Pat Upson, Enrichment Technology Company Ltd, personal communication with author, October 31, 2005.

12. Cabrera-Palmer and Rothwell, "Why Is Brazil Enriching Uranium?"

13. In 2007, Urenco earned a profit of 34 percent before taxes (23% after taxes), Urenco, "Urenco Group-Full Year 2007 Audited Financial Results," April 2, 2008.

14. Mark Hibbs, "Utilities Mulling JNFL's Installation of Centrifuge Rated Around 50 SWU/Y," Nuclearfuel, May 4, 1998.

15. Citizen's Nuclear Information Center, "Rokkasho Uranium Enrichment Plant Down to a Single Line," Nuke Info Tokyo, March/April 2008.

16. Miles Pomper, "Bush Sends Russia Nuclear Energy Pact to Hill," Arms Control Today, June 2008, pp. 32-34.

17. Alex Glaser, "Characteristics of the Gas Centrifuge for Uranium Enrichment and Their Relevance for Nuclear Weapon Proliferation," Science & Global Security (forthcoming).

18. "Implementation of the NPT Safeguards Agreement in the Islamic Republic of Iran: Report of the Director General to the IAEA Board of Governors," November 10, 2003, Annex 1, para. 21.

19. "Implementation of the NPT Safeguards Agreement in the Islamic Republic of Iran: Report of the Director General to the IAEA Board of Governors," February 24, 2004, paras. 37-42, 61-62; November 15, 2004, para. 134.

20. David Albright, Paul Brannan, and Jacqueline Shire, "Can Military Strikes Destroy Iran's Gas Centrifuge Program? Probably Not," Institute for Science and International Security, August 7, 2008.

21. For a comprehensive discussion of possible measures that could be used to strengthen international confidence that Iran is not producing HEU for nuclear weapons, see James Acton and Joanna Little, The Use of Voluntary Safeguards to Build Trust in States' Nuclear Programmes: The Case of Iran (London: Vertic) 2007.