Closing the Gaps
Securing High Enriched Uranium in the
Former Soviet Union and Eastern Europe
by Robert L. Civiak
Download Full Report
(2 MB PDF)
Executive Summary
Proposal 1
Proposal 2
Proposal 3
Appendix A
Appendix B
Acronyms and Abbreviations


Rapid Blend-Down of All Excess Russian Weapons-Origin HEU

A Large HEU Stockpile Remains in Russia Outside of the HEU Deal

According to unofficial estimates,14 the Russian military has about 1,000 metric tons of "weapons-grade equivalent" HEU.15 Five-hundred tons are covered by the current HEU deal, of which 140 tons have been blended down and 360 tons are still in weapons and at storage facilities, waiting to be blended.16 The remaining 500-ton Russian stockpile of separated HEU, which is not covered under the HEU deal,17 is sufficient for 20,000 nuclear weapons.18

President Putin recently pledged to reduce Russia's deployed nuclear weapons stockpile to about 2,000 deployed strategic warheads. At an average of 25 kilograms per warhead, those 2,000 warheads would contain about 50 tons of HEU. If Russia were to retain an additional 100 ton HEU stockpile composed of tactical nuclear warheads, inactive strategic warheads, and an HEU reserve, it would still need only 150 tons of HEU. That would leave at least 350 tons of excess weapons-grade HEU remaining outside of the existing deal. Thus, while the existing HEU deal is a good start, it addresses only half of the excess weapons-grade HEU Russia may soon have from nuclear weapons disassembly (see Figure 1).

There is a pressing need to expand the existing HEU agreement, but negotiating further reductions may take some time. The Russian government may be averse to large additional, unilateral reductions in its HEU stockpile, since that could limit its ability to re-enlarge its nuclear weapons arsenal. Nevertheless, in the interests of security, the US should seek negotiations on expanding and accelerating HEU blend-down as soon as possible. At the current blending rate of 30 tons per year, the HEU remaining under the original agreement will not be eliminated until 2013. If the deal were extended at this rate, the additional excess HEU would not be blended and sold until 2025. That is much too long for this dangerous material to remain available for potential diversion.

while the existing HEU deal is a good start, it addresses only half of the excess weapons-grade HEU Russia may soon have from nuclear weapons disassembly.

One way to speed up the elimination of this material is to simply increase the pace of the existing HEU deal. However, any significant increase under the current arrangements would overwhelm the commercial nuclear fuel markets upon which this deal depends. Nuclear fuel prices might fall significantly, and Russia might not be able to recover the costs of blending the HEU. Increasing sales of Russian LEU could force USEC to close the only operating enrichment plant in the United States. Concern over the loss of American jobs and the national security risk of becoming more dependent on Russia for US nuclear fuel requirements could in turn undercut American support for the HEU deal. Thus, attempting to expand the existing HEU deal under its current arrangements would more likely lead to its collapse than to its expansion.

The solution to this problem is to increase the pace of blending excess Russian HEU into LEU while maintaining the existing level of sales. Two hurdles remain for this to work. First, a way must be found to assure US commercial interests that Russia will not prematurely bring the HEU, which it has blended into LEU, into the marketplace. Second, the cost of the blending operation must be kept low, since it will not be recovered through LEU sales for an extended period. Both of these problems can be solved if the extra HEU is blended to only 19.9-percent U-235 and is kept in Russia under Russian ownership.

Blending the HEU to 19.9-percent U-235 provides nearly the same level of security as lower levels of enrichment. By international convention, uranium enriched to less than 20-percent U-235 is classified as LEU that is not useful in making fission weapons. While this 19.9-percent LEU could be re-enriched to HEU more easily than could natural uranium or the 4.4-percent LEU, which is produced under the existing HEU deal, such enrichment is still well beyond the ability of terrorists and most governments to carry out. Even expert designers would need several hundred kilograms of 19.9-percent LEU to produce a nuclear weapon. Considerably more material would be needed for the simple "gun-barrel" weapon design that terrorists might more readily fabricate.

HEU can be blended to 19.9-percent LEU at a fraction of the cost of blending HEU to 4.4-percent LEU, since a smaller amount of less expensive uranium feed stock is needed for blending to 19.9 percent. This makes such a proposition affordable. Furthermore, because of the additional expense of further blending, the existence of large stockpiles of 19.9-percent LEU is little more of a threat to commercial nuclear fuel markets than is HEU.

Proposal for Rapid Blend-Down of Additional Russian HEU

We recommend that the Administration seek to extend the existing HEU agreement with Russia based on the following elements.

  • Speed up HEU conversion. The United States will pay Russia its costs, plus a modest incentive payment, to blend an additional 30 tons per year of HEU from nuclear weapons to 19.9-percent enriched LEU using natural uranium as a blendstock.
  • Make sure the LEU does not disrupt the international market. The 19.9-percent enriched LEU will remain in Russia, without further blend-down, until at least 500 tons of HEU have been blended into LEU under the terms of the existing HEU agreement.
  • Arrange Future LEU Transactions. The United States will agree to purchase, and Russia will agree to sell, the LEU once it has gone through an additional downblending step from 19.9-percent enriched LEU to fuel grade material. The pricing for this arrangement will be worked out separately from the framework agreement, as is done under the current HEU deal.
FIGURE 1.US and Russian HEU Stockpiles
Click to enlarge...

There is a security benefit to be gained in simply speeding up the blending of the 360 tons of HEU remaining under the existing deal. However, the primary US goal should be obtaining Russian agreement to expand the amount of HEU eliminated. Expanding the deal would be much more beneficial than merely speeding it up, both in reducing the risk of diversion of weapons-useable material and in limiting Russia's ability to re-enlarge its nuclear stockpile. The rapid blend-down could proceed in parallel with US-Russian negotiations to go beyond the initial 500 tons. Even with a doubling of the blend-down rate, it would take six more years to complete the initial goals set out in the original HEU deal. However, the United States should not allow Russia to view the payments it receives for rapid blend-down solely as a means for reducing the cost of blending under the existing agreement. The US should make it clear to Russia from the outset that its interest lies in expanding the HEU deal, not just speeding it up (see Figure 2).

Depending on the attractiveness of the overall deal, Russia might agree to a substantial unilateral reduction of its remaining HEU stockpile. However, Russia will likely require a reciprocal reduction by the United States before it agrees to fully eliminate its excess HEU. The costs of a rapid blend-down program and the incentives for both Russia and the United States to further reduce HEU stockpiles are discussed in the following two sections.

FIGURE 2.Proposal for Rapid Blend-Down of Additional Russian HEU
Click to enlarge...

Cost Estimate for Doubling the Blend-Down Rate of Russian HEU

According to Russian officials, no major new facilities would be needed to double the amount of Russian HEU that gets blended down each year, increasing from 30 to 60 tons the amount of HEU eliminated annually. These officials have told the DOE and non-governmental analysts that there is sufficient capacity at existing facilities to blend 50 tons of HEU to LEU per year. DOE analysts believe Russia could increase the rate to 60 tons per year by installing new equipment costing about $1 million.19 If these reports are accurate, operating expenses will be the primary cost component of the blending operation. Five types of activities contribute operating costs to the blending operation20 (see Figure 3).

FIGURE 3.The Blend-Down Process
Click to enlarge...

1. Initial processing: This involves dismantling the weapons components, shredding the HEU metal into small pieces, and putting it through a series of chemical processes to convert it to uranium hexafluoride. Uranium hexafluoride is the preferred form for blending, because it becomes an easily mixed gas at relatively low temperatures. These initial processing operations are labor intensive, and their cost is difficult to estimate. Based on incomplete information available about processing operations in Russia for the existing HEU deal, it appears that the initial processing for 30 tons of HEU would cost between $15 and $30 million per year. It is the largest cost component of the proposed blending operation.

2. Transport: The HEU must be shipped thousands of kilometers by rail, in heavily secured cars, from the facilities that do the initial processing to other facilities that carry out the blending operation. If 500 to 1,000 kg of HEU is shipped per train, and each trip costs $100,000, the annual transport costs for blending 30 tons of HEU would be $3-6 million.21

3. Blendstock production: To blend directly to nuclear fuel grade, Russian HEU must be mixed with a 1.5-percent enriched blendstock,22 which could cost as much as $40-50 million per 30 tons of HEU. However, it is possible to defer this high cost by postponing the process that requires the expensive blendstock. The Russian HEU can be blended to 19.9-percent LEU using much cheaper natural uranium, at a cost of only $3 million per 30 tons HEU. Subsequent blending of the 19.9-percent enriched LEU to fuel grade LEU would still require the addition of costly blendstock. Ultimately, the total cost of blendstock production for converting weapons-grade HEU to fuel grade LEU would be roughly the same; blending to 19.9-percent in the near-term allows for significant nonproliferation achievements while deferring major costs23 (see Figure 4).

FIGURE 4.Two Options for Blending 30 Metric Tons of HEU to LEU
Click to enlarge...

4. Blending: The actual blending is done by joining the outflow from two separate pipes containing the HEU and the blendstock, both in the form of gaseous uranium hexafluoride. The enrichment level of the product is adjusted by regulating the flow rates from the two pipes. This low-cost operation would add less than $1 million to the total cost of blending 30 tons of HEU.

5. Storage: A substantial number of specially designed containers would be needed to store the 140 tons of 19.9-percent LEU hexafluoride that would be produced each year. Storage containers for 4.4-percent LEU cost $5,000 each, and can hold 2.5 tons of LEU hexafluoride. Assuming that cylinders for storing 19.9-percent LEU hexafluoride cost about the same price, but because of criticality concerns24 hold less material, the annual cost for purchasing containers would be about $2 million. There would be additional costs to store, monitor, and maintain the containers. If, however, they were stored in existing buildings ,25 the total cost for storage, including the cost of the purchasing, monitoring and maintaining the containers for about a decade, would come to about $3 million per year.

In sum, the annual cost for processing and blending 30 metric tons of metallic HEU weapons parts to 19.9-percent LEU hexafluoride, including storage, is estimated at $25-43 million. The price will also have to include any incentive payment that the Russian government may require to perform the blend-down. That price will be subject to negotiation, but if the incentive payment were in the range of $15-20 million26 per year, the program could be implemented for a total cost of about $40-60 million per year.

Some preliminary steps that are needed to bring this proposal into effect, are being initiated. The Nuclear Threat Initiative (NTI)27 has decided to fund Russian and American experts who will refine the estimates for both the operating costs and the capital upgrades that Russia would need to make in order to increase the blending rate by 30 tons per year. NTI is currently in the process of putting the teams together for those activities.

The incentive payment would be only one of several benefits that Russia could obtain from expanding the HEU blend-down program. These benefits, as well as additional benefits to the United States, are discussed in the next section.

Benefits from Expanding the HEU Blend-Down

Expanding the Russian HEU blend-down program could reduce the risk of HEU diversion for malevolent terrorist purposes. Russia would benefit from this reduced risk, as would the United States.

The rapid blend-down would benefit Russia in several other ways. The expanded blending operation would provide employment for a substantial number of Russian workers. These jobs would go primarily to workers at facilities that formerly supported the Russian nuclear weapons program. The social benefit of providing these jobs is of central concern to Russia. There would also be a significant security benefit from stabilizing the economies of cities that previously depended on nuclear weapons production. Such stabilization would further reduce the risk of diversion of fissile material and nuclear weapon-related equipment or information. Russia would also experience a savings in security costs by reducing the number of buildings and the area in which HEU is stored. It is difficult to estimate the extent of the potential security savings, but they would likely measure in the millions of dollars per year.28

In addition, Russia will benefit from the lower costs of future downblending operations. Since the US would cover the cost of the initial blend-down to 19.9-percent enriched LEU, Russia would only have to cover the cost of subsequent blending from 19.9-percent to 4.4-percent LEU. The primary savings would be the $15-30 million per year cost of the initial processing from weapons components into uranium hexafluoride, which will already have been carried out under US funding. As noted above, the cost of feedstock, which would be the single largest cost in subsequent blending to meet nuclear fuel specifications, would be about the same as it is now.

These benefits and a modest additional payment may provide a sufficient incentive for Russia to expand the downblending of HEU by 200 - 300 tons. Such an expansion of the existing HEU deal would be a significant security achievement.

Even greater reductions might be possible. However, larger reductions in Russia's HEU holdings would eventually impinge upon the size of its nuclear weapons stockpile or on reserves of HEU that it plans to hold for potential use in nuclear weapons. As HEU reductions approach the limit of excess Russian HEU, the Russian government is unlikely to continue to downblend its holdings without a reciprocal arrangement from the United States. The United States must address the issue of reciprocity if it wants to obtain the security and arms control benefits of deeper reductions in HEU stockpiles in Russia. Reciprocity should be based on the amount of HEU remaining in the stockpiles of each nation, rather than on how much each side has eliminated. The US stands to benefit from such an agreement because Russia produced substantially more HEU during the Cold War than the United States, and would therefore have to make larger cuts in its stockpiles in order to reach equal footing with the US. An understanding based on reciprocity would be an extension of the principle of parity in deployed warheads, which has been the foundation of bilateral arms control agreements between the US and Russia for some time.

As noted above, information on the size of the Russian HEU stockpile is uncertain. Similarly, the United States government has not released information on its total inventory of HEU. According to unofficial estimates, the United States has produced about 635 tons of weapons-grade equivalent HEU.29 The US has declared 100 tons of that HEU as excess and is blending some of it into LEU. If, as assumed above for Russia, the US were to retain only sufficient HEU for 2,000 deployed strategic warheads and a combination of tactical nuclear warheads, inactive strategic warheads, and an HEU reserve, containing twice as much HEU as the deployed stockpile, it would also need 150 tons of HEU. Therefore, the remaining 385 tons of US HEU could be considered excess.

The Bush Administration recently announced that in addition to 1,700 to 2,200 operationally deployed strategic nuclear weapons, it plans to maintain as many as 8,000 other nuclear warheads, including tactical warheads, a "responsive force," and an "inactive reserve." Furthermore, under current policy the US plans to retain all HEU from retired warheads, except what it has already declared as excess, for potential future use as fuel for US naval vessels.

Storage casks containing the first shipment of Russian LEU to the United States

A full discussion of whether the US needs these large reserves is beyond the scope of this paper. Nevertheless, the United States must decide whether its interests are best served by maintaining a stockpile of at least 10,000 nuclear weapons and additional HEU, or by supporting mutual reductions to much lower levels than exist today. In any event, the US should be willing to provide an accounting of its HEU stockpiles. Before truly deep reductions in the Russian HEU stockpile can be achieved, the United States and Russia must adopt a comprehensive transparency regime that provides each nation with some assurance of the size of the other's HEU stockpile.30 As the numbers of nuclear weapons in US and Russian arsenals fall, such a transparency regime is necessary to provide assurance that the other side is not maintaining large stockpiles of nuclear materials that could assist in a breakout from agreed upon numbers of nuclear warheads. The achievement of such a transparency regime would provide an arms control benefit that might rival the benefit in increased security, which is currently the driving interest in reducing HEU stockpiles.