Indsigt: "Atom-fængsel" i Grønland

"Atom-fængsel" i Grønland - RAND rapporten

Her er den originale (engelsksprogede) rapport fra RAND, hvori der blandt andet peges på Grønland som et muligt sted en sikker deponering af materiale til fremstilling af kernevåben. De afsnit der specielt har interesse for os her i Grønland er fremhævet med grønt.

Mandag d. 10. marts 1997
RAND
Emnekreds: Thule Air Base.

Indholdsfortegnelse:
A Concept for Strategic Material Accelerated Removal Talks (SMART): PREFACE; SUMMARY; I. THE CONCEPT OF SMART; II. WHY SMART?; III. OPERATIONAL ASPECTS; IV. COMPARISON OF ALTERNATIVES; APPENDIX: SOURCES AND ASSUMPTIONS OF WEAPON-USABLE MATERIAL CHARTS

Læs den i PDF-format: A Concept for Strategic Material Accelerated Removal Talks

Her følger den originale rapport fra RAND, hvori der blandt andet peges på Grønland som et muligt sted en sikker deponering af materiale til fremstilling af kernevåben.
I denne udgave er fjernet de mange fodnoter med henvisninger til andre rapporter, ligesom fodnoter med enkelte uddybende kommentarer er fjernet. De afsnit der specielt har interesse for os her i Grønland er fremhævet med grønt.

A Concept for Strategic Material Accelerated Removal Talks (SMART)

Brian G. Chow Richard H. Speier Gregory S. Jones DRU-1338-DOE April 1996 Prepared for the Department of Energy

Preface Summary I. THE CONCEPT OF SMART Recent Proposals for Material Removal Fast Material Removal by SMART Relationship to Other Options II. WHY SMART? III. OPERATIONAL ASPECTS Material Involved Unilateral, Bilateral, and Multilateral Approaches Decision Rule for Return of Material Coverage of Costs Incentives Location and Facilities(tryk her for dansk oversættelse ) Physical Security Force Transportation Relationship to Existing Laws, Negotiations and Other Activities Sequencing of Events IV. COMPARISON OP ALTERNATIVES Appendix: SOURCES AND ASSUMPTIONS OF WEAPON-USABLE MATERIAL CHARTS

PREFACE

U.S. officials say that, after Russia ratifies the second Strategic Arms Reduction Treaty (START II), the focus of nuclear arms control between the United States and Russia may shift from delivery vehicles to nuclear warheads. This study proposes a new approach for dealing with the problem of nuclear warheads. We believe that quickly removing weapon-usable material from both countries' territories should receive serious consideration by strategic policymakers, military planners, arms negotiators, and planners of the civilian nuclear fuel cycle.

This study was sponsored by the U.S. Department of Energy. The views expressed in this paper are those of the authors and do not necessarily reflect the opinions or policies of their affiliated organization or of their research sponsor. This paper can be cited or quoted.

SUMMARY

This concept paper explores a new direction in strategic arms control. Under the concept, the U.S., Russia, end ultimately other nations would move beyond the current efforts to protect weapon- usable material from theft or other sub-national diversion. They would take steps so that within a decade excess weapon-usable material will not be subject to national diversion, i.e., breakout from strategic and nonproliferation agreements to make more nuclear warheads.

The quickest way to achieve this goal is to transport excess weapon-usable material out of the countries to a foreign storage site (or sites) where unauthorized removal would be prevented by engineered features and by an appropriately constituted physical security force

There are an infinite number of variations on this concept. But all of the variations depend on one first step: talks.

Our concept is that Strategic Material Accelerated Removal Talks (SMART) would be initiated first between the U.S. and Russia. The talks would establish whether there were any feasibility whatsoever to the concept, what incentives might interest the Russians, and what variations should be explored. If the talks showed promise, they could be expanded into planning and action.

The action would involve far greater quantities of material and greater speed of handling than under present efforts. At present, we are engaged in a multi-year process to improve accounting, transparency and control of Russian weapon-usable material and to construct a secure storage facility for such material at Mayak. In addition, we are removing 500 metric tons of HEU (=beriget uran) from Russian control over a 20 year period. However, Russia may have as much as 700 metric tons of additional HEU. As to excess plutonium in the U.S. and Russia, we are hoping to dispose of it over a period of 20 to 90 years. This leaves a vast amount of excess material in the U.S. and Russia - enough to make well over 100,000 nuclear weapons even after 10 more years of present efforts.

By contrast, under the concept of SMART, excess weapon-usable material would be transferred out of national access in a decade. In this time, the number of potential nuclear warheads in both the U.S. and Russia would be reduced by more than 90 per cent.

SMART is a natural progression of nuclear arms control between the United States and the Soviet Union (now, the Russian Federation). From the Strategic Arms Limitation Talks I & II in the 1960s and 1970s to the Strategic Arms Reduction Talks I & II in the 1980s and early 1990s, the nuclear disarmament process went from limitation to reduction. There is already a presidential agreement that, within 30 days of Russian ratification of START II, new talks will begin. There is also a growing consensus that the new talks will focus on making the reductions irreversible. SMART is a quick way to secure irreversibility.

Some suggest that SMART should wait until accounting data on material is exchanged between the United States and Russia, the many facilities involving weapon-usable material are better monitored, and the in-country storage site is operational. These activities are necessary for SMART. However, discussions on SMART Can proceed in parallel to these programs

The changing political situation in Russia has created the risk of a reversal of arms control progress as well as the risk of the loss of control over substantial quantities of Russian weapon-usable material. The sooner that SMART is pursued, the better the chance to limit these risks. The same changing political situation may limit the windows of opportunity to remove large amounts of weapon-usable material from Russia. Pursuing SMART now will maximize our readiness to seize these opportunities.

SMART would keep open a number of options No decision on final plutonium or HEU disposition would be involved. So a foreign storage site would not necessarily be a strategic material graveyard On the other hand, the storage site proposed here should not be considered a strategic material bank, where the contents are a normal part of everyday commerce, and deposits and withdrawals can be made freely.

Rather, a storage site should be considered a strategic material prison, where the consents are dangerous, escape is physically prevented, and release is subject to considered procedures. Subject to decision rules and controls discussed below, many variants of SMART are possible under which material might gradually be returned to national custody in the future. However, any return of material would be carefully constrained with respect to purpose, amount, and frequency.

Because material could be withdrawn from storage, SMART would be consistent with a number of options for plutonium disposition. The material could be buried. Also, the option would be open for the plutonium to be released slowly back to Russia to allow for Russia's presently-envisaged 20-90 year program of burning in Russian reactors Another option is a 30 year program to burn plutonium in CANDU reactors in Canada. SMART could seamlessly be integrated with this option if U.S. and Russian excess material were stored in Canada. Yet another option is to store the material in a third country and then slowly release the material for MOX fabrication at the German Hanau fabrication facility. In this option SMART could be combined with a 25-40 year program to burn Russian oz U.S. and Russian plutonium in European reactors. In all cases, HEU stored in a third country could be released at an appropriate rate for blending into LEU reactor fuel

This study describes the SMART concept. the rationale for it, various operational considerations, and issues raised by critics of the concept. It also compares SMART to other alternatives currently being pursued or explored and identifies common actions that would be taken under SMART and the alternatives.

We have yet to estimate the implementation costs and the budgetary savings of SMART. The following preliminary numbers, however, could indicate the tradeoff. SMART could cost on the order of $1 billion for interim and long-term storage facilities in a third country, $1 billion for incentives for Russia, and hundreds of millions of dollars per year in packaging, transportation. and physical security costs. These costs should be compared to the nuclear budgets being expended to preserve a "hedge" against a breakout from arms control agreements. The U.S. (DoD and DoE) nuclear forces budget is about $15 billion per year, and the "hedge" might be a significant fraction of it. Russia could have a "hedge" strategy as well. and SMART might also provide it some savings.

Third country interim and long-term storage facilities for strategic material could be sited in a variety of locations offering physical security, reasonable logistics, environmental acceptability, and an equitable location with respect to the U.S. and Russia. Depending on Russian criteria for a politically acceptable site, some candidates include Canada, Greenland, Australia. Iceland, Northern Scandinavia and Switzerland. Transportation of material to these sites could be undertaken in a consolidated manner to minimize the physical security risks. Once the material was at the sites, it would be protected by a physical security force and ultimately also by engineered features to protect it from seizure by a sub-national or a national government.

Our experience in discussing SMART with others is that some may want to rule out the idea immediately because it cannot deal with one particular principle or another that they hold dear. This principle could be to deal with current programs before exploring new ones, to avoid considering new expenditures, to keep weapon-usable material in-country until it is fabricated into reactor fuel, or to minimize shipments of nuclear material

This paper addresses these concerns -- how they can be resolved or Whether they are unfounded. SMART has many flexible and accommodating variants that still preserve the key objective of preventing massive nuclear reconstitution. If such concerns are used to shape, as opposed to kill, SMART, we may arrive at a great contribution to making nuclear disarmament irreversible.

The political opportunity rapidly to remove large amounts of weapon-usable material from Russia could appear and disappear in a fleeting moment. We recommend that the concept for Strategic Material Accelerated Removal Talks between the U.S. and Russia be evaluated by both countries seriously and quickly.

I. THE CONCEPT OF SMART

Under Strategic Material Accelerated Removal Talks (SMART), the U.S., Russia. and ultimately other nations would move beyond, the current efforts to protect weapon-usable material from theft or other sub-national diversion.' They would take steps to that within a decade, not in decades as currently planned, excess weapon-usable material will not be subject to national diversion, i.e., breakout from strategic and nonproliferation agreements to make more nuclear warheads The quickest way to achieve this goal is to have excess weapon-usable material transported out of the countries and to a foreign storage site (or sites) where unauthorized removal would be prevented by engineered features and by an appropriately constituted force.

RECENT PROPOSALS FOR MATERIAL REMOVAL

The idea of removing excess weapon-usable material from the U.S. and Russia has been proposed recently by people associated with various organizations. In 1993, a RAND study' recommended that The United States should make FSR (former Soviet Republic) weapon-grade plutonium as inaccessible to FSRs as possible so that it is difficult to make nuclear weapons with it. The best way to accomplish this is to remove plutonium from FSRs. The study further proposed that United States' own material might have to be subject to similar treatment or arrangement in order to make the agreement symmetrical and equal to both sides.

In 1994, the National Academy of Sciences' suggested with respect to Russian plutonium Management, control, or outright ownership of the stores and the material in them might be transferred to other parties, such as an international consortium formed for that purpose. The material might even be physically relocated to some other country, possibly in return for cash, as in the case of the HEU deal.. .The concerns motivating such proposals apply primarily to Russian plutonium, but if reciprocity or parallelism is desirable for political reasons, similar steps could be taken with U.S. plutonium. In 1995 an advisor to the Joint Chiefs of Staff, whose views were identified as "his own"', argued in favor of the rapid removal of non-strategic nuclear weapons and excess fissile material to a secure repository in a third country--monitored either bilaterally or by an international presence and guarded by their own military units -- until both countries are confident the final disposal can be accomplished in a fully secure environment. A 1996 study by four authors from Harvard University's John F. Kennedy School of Government' recommended removing excess KEU and plutonium from Russia as quickly as possible -- by purchase. ...even if the HEU deal goes as planned, it will take two decades to implement, which does not solve the immediate pressing need to secure Russia's HEU. Thus, we argue for a drastically accelerated timetable ...The United States should aim to purchase as much Russian HEU as possible, as quickly as possible, given capabilities for secure and safe transportation....Because of the proliferation risk posed by Russia's stockpile of excess weapons-grade plutonium, the United States should offer to buy it --or at least some of it.... Our concept is that talks would be initiated first between the U S and Russia for the purpose of placing their excess strategic material in a third country or countries so as to prevent national diversion by both U S. and .Russia. These Strategic Material Accelerated Removal Talks (SMART) would establish whether there were any feasibility whatsoever to the concept, what incentives might interest the Russians, and what variations should be explored. There are an infinite number of variations on this concept. But all of the variations depend on one first step talks,

At a later phase, the same concept can be applied to the out-of- country storage of nuclear material from undeclared nuclear-weapons states (India, Pakistan, Israel and North Korea), other declared weapons states (China, France and U.K.), and nonnuclear weapons states (such as Germany, Japan. Belgium, Switzerland, South Africa, and Brazil). The benefit of a bilateral U.S /Russia agreement would still be significant, even if the later phase of multilateral talks failed.

FAST MATERIAL REMOVAL BY SMART

It is important to emphasize that there are two types of diversion of nuclear assets -- subnational and national, because measures aimed at preventing one type of diversion might do little to prevent the other. We need to make sure that neither one is ignored, and that measures, not necessarily the same for both, are devised to prevent them. Subnational diversion would be undertaken by a small group of people, and the diverted material or weapons would go to rogue military units, terrorist groups or nuclear- aspiring countries. In contrast, national diversion would be undertaken by the host government owning or storing the nuclear assets, and the diverted assets could be used for nuclear weapons with little warning.

In the last several years, the United States and Russia have made considerable efforts to account for, better protect, and reduce the vast amount of weapon-usable material produced during the Cold War The HEU Purchase Agreement signed in February 1993 allows 500 metric tons of HEU to be removed from Russia over a 20 year period. Both sides also recognize that accounting and transparency are the prerequisites for measures that protect materials from subnational and national diversion. In the January 1994 Summit, the United States and Russia agreed to establish a joint working group to consider...steps to ensure the transparency and irreversibility of the process of reduction of nuclear weapons, including the possibility of putting a portion of fissionable material under IAEA safeguards. Particular attention would be given to materials released in the process of nuclear disarmament and steps to ensure that these materials would not be used again for nuclear weapons. In November 1994 under Project Sapphire, the U.S. purchased and removed 625 kg of HEU (about 89% enriched) from the Republic of Kazakhstan.

In March 1995, President Clinton declared a surplus of 200 tons of fissile material from the U.S. weapons stockpile. The United States is encouraging the Russia to do the same. The United States and Russia are engaged in a multi-year process to improve the control of weapon-usable material in Russia and other republics. Modern surveillance and monitoring equipment will be developed and installed at about 40 to 50 sites, where sensitive nuclear material is stored.' A secure storage facility for such material is also being constructed at Mayak. The United States and Russia are also evaluating options to dispose of excess plutonium.

The measures taken thus far are either prerequisites or stepping stones for more comprehensive programs. Our concern is that, even after 10 more years of present efforts. there will still remain a vast amount of excess material in the U.S and Russia -- enough to make over 100,000 nuclear weapons. It may be 20 to 90 years before excess nuclear material in the two countries is eliminated or removed to third countries. In the meantime, a government might reverse its course; and the material could quickly be used to reconstitute its massive nuclear arsenal.

In addition to the 500 metric tons of HEU to be blended down, Russia may have as much as 700 metric tons more in its operational warheads, inactive warheads, nuclear pits and material stockpiles (Table 1)." The United States currently has about 800 metric tons of HEU. As to plutonium in the U.S. and Russia, they together have about 300 metric tons. Table 1:
Weapon-Usable Material in Russia and the U.S. in 1995 (in metric tons) HEU Pu Russia 1200 200 U.S. 800 100 Total 2000 300 START I & II, as well as earlier agreements and initiatives, will drastically reduce the number of active or operational warheads. As recently as 1989, the United States had 22,500 active tactical and strategic warheads. The number dropped to 9,000 in 1995 and is expect to decline to 5,000 by the year 2003 (Figure 1). The reduction is similarly drastic in Russia's nuclear arsenal, from 40,000 in 1989 to 12,000 in 1995 and 6,000 by 2003.

Unfortunately, the total amount of material inside and outside of active warheads will change little. What appears to be happening is that the material is being kept in forms that will facilitate the reconstitution of the weapons. Some weapons are being kept intact -- which is permitted because current arms control agreements focus on delivery means rather than on the actual weapons themselves. Other weapons are being disassembled, but the nuclear pits are being stored intact. One source estimates that in 2003 that the U.S. will have 2,500 non-operational warheads that have been kept intact and will have stored 12,000 nuclear pits. The Russians may well match the U.S. and have 10,000 to 20,000 non-operational warheads and/or pits in 2003.

Our proposal is meant to address this problem as well as the problem posed by the stockpiles of plutonium and HEU in various countries. Under the concept of SMART, excess weapon-usable material could be transferred out of national access in a decade. Thus, the number of potential nuclear warheads in both the U.S. and Russia would be reduced by 90 per cent. Figure 2 illustrates this massive difference. Time is of the essence because the Russian political situation is not settled, and reconstituting a massive nuclear arsenal remains a distinct possibility.

/////Figure 1: Amount of Weapon-Usable Material in U.S. and Russia Will Remain Huge

/////Figure 2: Drawdown of Weapon-Usable Material Is Much Faster With SMART

RELATIONSHIP TO OTHER OPTIONS

SMART would keep open a number of options. No decision on final plutonium or HEU disposition would be involved. So a multinational storage site would not necessarily be a strategic material graveyard. On the other hand, the storage site proposed here should not be considered a strategic material bank, where the contents are a normal part of everyday commerce, and deposits and withdrawals can be made freely.

Rather, a storage site should be considered a strategic material prison, where the contents are dangerous, escape is physically prevented, and release is subject to considered procedures. Subject to decision rules and controls discussed below, many variants of SMART are possible under which material might gradually be returned to national custody in the future. However, any return of material would be carefully constrained with respect to purpose, amount, and frequency.

Because material could be withdrawn from storage, SMART would be consistent with a number of options for plutonium disposition. These include burying of plutonium at agreed sites, burning of Russian plutonium in Russia, and burning of U.S, and/or Russian plutonium in the U.S , Canada or Europe. We compare a number of options in Section IV.

However, SMART should not be confused with the International Plutonium Storage (IPS) concept studied by the IAEA in the early 1980's and sometimes suggested for Russia today. The IPS concept has many variants, but moat of them allow the plutonium to remain in-country and readily to be withdrawn from storage. The essence of all IPS variants is that IAEA safeguards are applied to the plutonium - providing only after-the-fact notification of national diversion. There are some strong IPS variants, similar to SMART, with out-of-country storage and tightly constrained procedures for releasing material from storage. But these strong variants received little support in the IAEA study.

SMART would be a supplement to, not a substitute for, current efforts to improve the protection of excess Russian, as well as U.S.. material from sub-national threats. Weapons will still need to be dismantled, and the material remaining in Russia will still need secure storage. So why not be content for the time being with the present program against sub-national diversion? Why mount an additional effort?

II. WHY SMART?

In the 1960s and 1970s, the superpowers negotiated. signed and ratified the first and second Strategic Arms Limitation Talks (SALT I & II) treaties to limit and freeze the number of strategic nuclear delivery vehicles. In the 1980s and early 1990s, they negotiated and signed the first and second Strategic Arms Reduction Talks (START I & II) treaties to reduce the numbers of such delivery vehicles. There is already a presidential agreement that, within 30 days of ratification of START II by Russian Duma, new talks will begin.

We believe that the Strategic Material Accelerated Removal Talks (SMART) should be considered a logical next step. Although SALT and START's constraints on nuclear delivery vehicles affect the number of active or operational warheads, these treaties set little restraint on the number of inactive warheads and nuclear pits or the amount of weapon-usable material that either side can keep. Thus, the existing agreements do not prevent either side from undertaking a quick and massive nuclear reconstitution. SMART is designed to do that.

There are current efforts focused on the disposition of fissile materials from dismantled nuclear weapons as well as inventories already on hand. The two classes of alternatives under consideration are burning the materials in nuclear reactors and burying them. We are, however, concerned that current U.S. and Russian discussions on disposal would at best lead to the elimination of excess strategic material in 20-90 years -- a period much too long for comfort, considering the political uncertainties in Russia. Note, for example, that in twenty years Russia moved from its first nuclear test to strategic forces comparable to those of the U.S.

Secured in-country storage is being actively pursued by the U.S and Russia. While this is always necessary for preventing sub-national diversion, in-country storage does little to prevent national diversion. If one wants to remove the danger of massive nuclear diversion by Russia (and have the U.S. engage in comparable actions) within the next decade, physically removing the materials from these two countries is the only hope.

It is sensible to ask why the U.S. should worry about the possibility of massive Russian nuclear reconstitution. After all, if START II is fully implemented by 2003, the U.S. is still expected to have about 5,000 nuclear strategic and tactical warheads. That would be much more than adequate to destroy much of Russian society under MAD (Mutual Assured Destruction), even if Russia reconstituted to its old posture of 45,000 warheads. Thus, following MAD, one does not have to worry about a Russian breakout and has no need to remove the strategic material from Russia, because the U.S. already has enough warheads to deter its nuclear adversaries whether their arsenals are expanded or not.

But MAD has never been the U S. nuclear deterrence policy This is apparent from the rough equivalence of both sides' nuclear arsenals during the Cold War and now. Instead, the U.S. policy since the 1950s has been flexible response or a variation of it. If a nuclear adversary chose to attack only our military targets as opposed to civilian targets, the U.S. has wanted to maintain enough warheads to have the option to respond in kind or proportionally. The U.S. would not want to respond by hitting the nuclear initiator's cities and thus would hope to leave enough incentives for both sides to spare the cities in subsequent attacks.

If our adversary has more nuclear warheads, we would want to have more warheads ourselves. First, with more warheads, our adversary might attack more targets. Then, we would need more warheads for a proportional response. Second, an adversary's nuclear forces are themselves some of our key targets. More warheads in the adversary's arsenal might lead to more delivery vehicles and sites, and more nuclear targets for us to hit. Thus, the number of an adversary's nuclear warheads matters in our flexible response strategy.

Moreover, the quicker our adversary can break out and increase its nuclear arsenal, the more difficult and costly it is for us to prepare ourselves for timely responses. It may cost billions of dollars annually for the U S. to maintain a capability for quick and massive nuclear reconstitution

The possibility of massive Russian reconstitution has not escaped from the minds of the U.S. nuclear planners. The 1994 Nuclear Posture Review says that the U.S. "must be prepared for possible emergence of hostile government in Russia or failure of arms control process in the FSU." The Review adopts a "hedge" strategy which preserves the options for quickly "uploading" additional re-entry vehicles (RVs) to our strategic missiles and reconstituting our nuclear forces.

While preserving the capability to upload RVs may not be costly, other components of a "hedge" strategy can be. For example, the United States may have to go beyond the currently planned capability of producing enough tritium for 5,000 warheads. It may need more tritium in the event that thousands of additional warheads have to be made. The United States also plans to maintain 2,500 inactive warheads in the "hedge" stockpile even after the year 2003." In total, the future active and "hedge" stockpile will have nine types of warheads." Some can be eliminated if no "hedge" stockpile is needed. The more types of warheads there are to maintain, the more expensive will be their stewardship, which involves laboratory and maintenance personnel for warhead upkeep. repair and modification.

Although there is no estimate on what portion of the current annual nuclear forces budget (DoD and DoE) of about $15 billion is attributable to the hedge strategy, it can be considerable.'º The more strategic material we can remove from Russia, the less money we need to spend on maintaining a responsive nuclear posture. Russia could have a similar "hedge" strategy, and SMART can save money for it as well.

Because SMART makes massive nuclear rearmament much more difficult and costly, SMART would reduce the likelihood of such reconstitution. Some, however, might argue that Russia could not reconstitute its nuclear forces quickly even with weapon material on hand, because it would take time and money to acquire the delivery vehicles. If this were true, there would be little need for our proposal. But we do not. consider this argument valid.

By the year 2003, Russia may reduce its SS-19s from 6 to 1 RV and SS-N-20a from 10 to 6 RVs as a means of meeting START II provisions. However, in a breakout, the Russians can quickly reverse the process and upload more RVs onto missiles. No additional delivery vehicles need to be acquired for uploading the strategic missiles. Moreover, many tactical missiles and aircraft are dual-use platforms. The conventional warheads in these delivery vehicles can quickly be replaced by nuclear ones. And the non-nuclear components of nuclear warheads can be manufactured in advance in covert facilities. The confidence of detecting many of these facilities is not high. Thus, if Russia has massive amounts of nuclear material, we have to worry that it may reconstitute a fully deliverable nuclear arsenal quickly.

We have not yet estimated the implementation costs and the budgetary savings of SMART. The following preliminary numbers, however, could indicate the tradeoff. SMART could cost on the order of $1 billion for interim and long-term storage facilities in a third country, $1 billion for incentives for Russia, and hundreds of millions of dollars per year in packaging. transportation, and physical security costs. These costs should be compared to the nuclear budgets being expended to preserve a "hedge" against a breakout from arms control agreements. The U.S. nuclear forces budget is about $15 billion per year, and the "hedge" might account for a significant fraction of it. Russia could ha~e a "hedge" strategy as well, and SMART might also provide it some savings.

The U.S. is currently evaluating alternatives for disposing of U.S. plutonium, with burning and burying as two key options. Some argue that the U.S. should take the lead and set an example for Russia to follow." There are two potential pitfalls. First, Russia may not follow suit in disposing of its plutonium. By acting unilaterally first, the U S. will diminish its bargaining power to induce the Russians to dispose their own plutonium. Second, even if Russia would follow suit, it is likely to take a long period, 20 to 90 years, to dispose of its plutonium.

It is much more urgent to agree on removing the U.S. and Russian material from their own territories than on the method of disposition. The latter can wait. Indeed, waiting may improve the quality of the decisions on disposition. In any case, U S. material is one of the bargaining chips for negotiating a solution that actually reduces the nuclear threat and the "hedge" cost to both sides The U.S. should not unilaterally dispose of the bulk of its strategic material.

SMART is a logical complement to President Clinton's initiatives announced in September 27, 1993. Zn his Nonproliferation and Export Control Policy, he proposed "a multilateral convention prohibiting the production of highly-enriched uranium or plutonium for nuclear explosives purposes or outside of international safeguards." Dealing with nuclear inventories already on hand will supplement the convention to achieve its goal - reducing the access to weapon- usable material. Moreover, President Clinton "seeks to eliminate where possible the accumulation of stockpiles of highly-enriched uranium or plutonium, and to ensure that where these materials already exist they are subject to the highest standards of safety, security and international accountability." Short of actually reducing the size of stockpiles, making them inaccessible is the next beat option.

Because the political opportunity rapidly to remove large amounts of weapon-usable material from Russia could appear and disappear in a fleeting moment, contingency planning for such an operation would be appropriate under any circumstances. But, we emphasize, the feasibility and modalities of such an operation can only be established in one way -- by formal and informal talks, initially between the U.S. and Russia.

What are the modalities to be explored?

III. OPERATIONAL ASPECTS

The basic idea behind SMART is third country storage of weapon material to avoid ready access by the United States and Russia. When we discussed the idea with analysts and planners, many have raised questions about the necessity of SMART, its acceptability by the Russians, and its implementation. In Sections I and II, we have discussed why SMART is needed. We have further suggested that, by discussing the idea with the Russians, we would find out whether SMART is attractive to them.

If the Russians are attracted by SMART, it would probably be for two reasons: First, the same reason for which they accepted SALTs and STARTs -- that SMART would enhance their security. Second, financial incentives.

With respect to the implementation of SMART, there are many questions. Our experience in discussing SMART with others is that some may want to rule out the idea immediately because it cannot deal with one particular principle or another that they hold dear. This principle could be to deal with current programs before exploring new ones, to avoid considering new expenditures, to keep weapon-usable material in-country until it is fabricated into reactor fuel, or to minimize shipments of nuclear material. In this section. we discuss such matters. Because SMART is based on a straightforward concept, it is capable of accommodating many concerns. Through in-depth exchange and careful design of various operational aspects, many issues can be resolved.

MATERIAL INVOLVED

We estimate that, in 2003, the United States will have about 80 metric tons of plutonium and 670 metric tons of HEU outside of active warheads. This material will be in inactive warheads, nuclear pits and material inventory. Russia will have about 190 metric tons of plutonium and 970 metric tons of HEU outside active warheads by the year 2003. The two countries' combined total will be 270 metric tons of plutonium and 1640 metric tons of HEU, Table 2 summarizes these figures. Table 2: Weapon-Usable Material Outside of (and Remaining in) Active Warheads in 2003 (in metric tons) HEU Pu Russia 970

(90) 190

(20) U.S. 670

(80) 80

(20) Total 1,640

(170) 270

(40) This material is enough to make about 100,000 nuclear warheads. Storing this material outside of both countries would form a serious impediment to the reconstitution of these weapons.

Non-operational weapons and pits are "classified shapes" that are not suitable for full international inspection. Because the process of disassembling these shapes to generic forms could, take some time and because our concept seeks rapid removal of the material from national access, it is desirable to consider leaving the weapon- usable material in these forms for the time being -- even though this may prevent full international inspection of the material. The subsection on Relationship to Existing Laws, Negotiations and Other Activities will discuss an arrangement where immediate access to the material would be limited to U.S. and Russian custodians (with inspection as agreed) but where an appropriate force would prevent removal of the material from the storage site(s).

There is also the question whether non-operational weapons should remain in the U.S. and Russia pending, perhaps accelerated, dismantlement or whether intact weapons "physics packages" (including both high explosive and nuclear material. but not necessarily including arming, fusing and other electronic devices) should be removed to the third country storage site(s). This would be a legitimate question for discussion. The decision on removing intact "physics packages" would depend in part on the schedule for weapon dismantlement versus the schedule for material removal to third country site(s).

Regardless of that decision, there is a strong case for pits to be removed to third countries as soon as possible without delaying shipment in order to disassemble them. Pits are excellent forms in which to store nuclear material because pits have been designed to remain intact for long periods. Given the levels of physical security under which the pits must be stored in a third country, early disassembly or conversion to oxides may not be necessary If such disassembly required prolonged storage of the pits in Russia, it might not even be desirable.

The total amount of strategic material to be removed depends a great deal on the agreed meaning of "excess". The 200 metric tons that the U.S. has declared "surplus" is only the beginning of the material that could eventually be deemed "excess". Additional amounts could be declared "excess" as the mutual removal of this material reduces the amount that must be kept in national custody for a "hedge". Ultimately, the agreed meaning of "excess" could be the quantities of material over and above those needed to maintain the weapon levels allowed by START II (or whatever strategic levels were currently agreed) plus agreed amounts for limited tactical nuclear forces, naval reactors, peaceful uses, and working inventories. If the U S. and Russia end up with comparable quantities of material in national custody, then the "excess" to be removed from Russia will be far larger than that from the U.S.

UNILATERAL, BILATERAL, AND MULTILATERAL APPROACHES

The U.S., perhaps with the aid of its allies, could continue to make unilateral purchases of excess Russian plutonium and HEU stocks. However, material other than HEU blended to non-weapon- usable enrichments would be transferred to the foreign storage site(s).

To avoid large and early expenditures, the bulk of excess weapon-usable material in the U.S. and Russia could be transferred under bilateral schedule to the foreign storage site(s). A schedule of deposits into the site(s) would be agreed so that, within a few years of the agreement, the excess weapon-usable material would be removed from each nation's sole control.

The arrangement could most usefully be designed to permit multilateral deposits into the Storage site(s). Other nuclear weapons states -- declared and undeclared --would be desirable participants, as would nonnuclear weapon states. Germany, which faces political problems with plutonium disposition, might be an early participant. Germany's example might encourage other European states with excess plutonium stocks, e.g., Belgium and Switzerland, to join. Japan, which now has a stated policy of avoiding excess plutonium stocks but continues to have excess stocks, could be encouraged to participate. Israel, India, and Pakistan's participation would address the "stockpiles" issue that bedevils the Geneva negotiation of a fissile cutoff convention. Nations such as South Africa and Brazil, with limited supplies of weapon-usable material, could be encouraged to participate in the longer term.

DECISION RULE FOR RETURN OF MATERIAL

The most desirable rule to prevent misuse of strategic material would be to permit removal of the material from the storage site(s) only upon unanimous agreement of the participants in the site(s) and only in quantities, at rates, and under conditions appropriate to approved uses. This would allow withdrawals for safeguarded HEU blending and safeguarded civil plutonium use, and even for naval reactors if LEU designs were not available until existing HEU stocks were exhausted. But in all cases, these uses would be subject to international approval.

If these conditions proved unacceptable because of the loss of national sovereignty over, for example, civil plutonium use, a fallback would be a review conference in. say, 25 years to agree by consensus on any changes in the decision rule for return of material.

If this provision still appeared too restrictive with respect to future breeder programs, the agreement could provide for the removal of material for approved uses at an appropriate but relatively low rate. e.g., 10 metric tons per year, starting at a future date, e.g., 20 years hence. Moreover, the release rate could match the consumption rate so that in-country inventory would not build up. This is not to say that breeders have little proliferation risk. We only mean that, even in a future in which strategic material continues to be used in civilian activities, out-of-country storage could still reduce the risk of massive national nuclear reconstitution. A similar arrangement can also be made with respect to current or future use of plutonium in thermal reactors.

It would be desirable to require that LEU, blended from the HEU in storage, be supplied in lieu of plutonium returns. Even when HEU supplies are consumed, one can continue the LEU substitution by using LEU enriched from natural uranium.

COVERAGE OF COSTS

With Russia in dire financial shape, Western nations would need to pay most of the cost of constructing storage facilities, packaging of the material. and transportation to the facilities. We suggest that many Western governments could share this cost because the security benefits of SMART would be widespread.

The strategic material storage facility now planned for construction in Russia is expected to coat at least $300 million Removing excess material from Russia would leave less material in Russia and could permit the redesign to a smaller in-country storage facility. However, we do not expect that the economies from a smaller storage facility would be significant; and we cannot expect that such savings could defray meaningfully the cost of the new concept

The operating cost of the foreign facility(s) would be another matter. This cost might be shared more equitably, especially since a large fraction of this cost would be for personnel from the participating nations.

If and when material were returned from the facility(s), the coverage of costs could be still another matter. In order to avoid exploitation of the facility(s) as cost-Cree plutonium storage, the owner of the material could be charged full pro-rated costs for all capital, interest, packaging, two-way transportation. and operating expenses upon the return of the material.

INCENTIVES

Given the precedent of the 500 metric ton HEU purchase and the Russian claims for planned civilian use for plutonium, Russia may be reluctant to give up access to its strategic material without some "incentives," The "incentives" that will appeal to Russia are likely to be cash and national security.

Because plutonium has no economic value, there is no particular basis on which to assign a price for the purchase or removal of Russian plutonium. It ultimately will amount to what the United States ia willing to pay and Russia is willing to accept.

The HEU deal may shed some light as to the magnitude of the incentive that the United States is willing to pay. The HEU deal will cost the U.S about $12 billion (undiscounted) over 20 years. Although the United States expects to recoup the money by selling the blended-down uranium, it is likely that The United States might suffer a loss of about $2.2 billion (undiscounted) or a present value of $ 0.8 billion (discounted at 10%). The loss mainly results from the fact that the United States will pay Russia $84 per separative work unit for the enrichment value in the blended-down uranium but will save only $60 in variable costs by running the U S. enrichment plants at a lower capacity.

A rule of thumb is that each Russian or U.S. nuclear weapon contains 15 kg of HEU and 3.5 kg of plutonium, On this ratio, the number of weapons that use 500 metric tons of HEU would use about 120 metric cons of plutonium. If the United States is willing to pay $0.8 billion for the Russians to blend down 500 metric tons of HEU, it might be willing to pay proportionally about $1 2 billion for removing as much as 180 metric tons of plutonium from Russia.

The logic of this estimate is that the United States might be willing to pay about the same amount for each bomb equivalent of material, whether HEU or plutonium. Following this logic, about $1 billion (present value) might be used as an "incentive" to the powerful Russian nuclear establishment MINATOM. If such a payment were offered, it could be structured on the basis of tons of material shipped out of country in order to give Russia every incentive to move quickly. Additional cash flows to Russia could result from its participation in the construction of the out-of-country storage facility(s).

It would remain to be negotiated what fraction of the removed plutonium would be owned by the United States and what fraction still would still belong to Russia.

While MINATOM may be most interested in cash, the Russian military is certain to be concerned with national security. A crude proxy for national security in Russian minds has often been equality of capabilities. Thus, the arrangements for strategic material removal -- beyond levels of material that MINATOM is authorized to sell -- will almost certainly require the removal of U.S. material on some equitable basis that will lead to comparable capabilities remaining in national custody.

The incentives that appeal to Russia may change with time -- perhaps changing rapidly and fleetingly. That is one reason why a contingency plan for rapid material removal would make sense.

In the next section we shall discuss incentives for the nation(s) that might host the storage site(s). Those incentives are likely to be cash (or jobs) and a contribution to world peace adequate to outweigh environmental qualms.

LOCATION AND FACILITIES

We should consider the feasibility of locating the storage site(s) in a remote place so that the physical security forces and any emergency response forces (see below) could defend the material from seizure by (a) sub-national groups, (b) the host government of the territory upon which the site was located, or (c) any other government. An alternative to a remote place is a country(s ) whose government is so politically stable and whose nonproliferation credentials are so impeccable that the possibility of host government seizure is practically nil. In any case, the site(s) will also need to be at some location(s) deemed equitable by both the U.S. and Russia. One important variation is that the U.S and Russia can each select a site(s) for its own material that is acceptable to the other government. Finally, the site(s) will need to be environmentally and logistically appropriate and acceptable to the host government(s). Clearly, to narrow down the choices, the United States and Russia need to discuss the matter with potential third countries.

Our mapping of these criteria against candidate sites yields the following preliminary alternatives (in no particular order): Canada, Greenland, Australia, Iceland, Northern Scandinavia, Switzerland, Antarctica, a Pacific island. Again, it is possible that more than one nation could be used for sites, for example, with one providing interim and another providing long-term storage.

Canada is particularly interesting because of its excellent international reputation and the offer by Canadian utilities to burn up U.S. and Russian plutonium in CANDU reactors. In section IV we discuss the CANDU option along with other alternatives. In spite of the attractions of Canada. its strong security relationship with the U.S. and the difficulty of establishing a physical protection force large enough to protect weapon-usable material from seizure by the Canadian or the U.S. government, may make a Canadian site unacceptable to Russia. As noted above, Russian attitudes -- including attitudes toward a Canadian storage site -- could change suddenly. But in order to explore the operational questions associated with a foreign storage site(s), we will examine a location that might be more . . . Greenland. ( Tekstlinien er faldet ud af faxen. Red. )

We shall discuss Greenland as an example of sites for both interim and long-term storage. Greenland is a self-governing administrative division of Denmark. Although, through Denmark, Greenland is integrated into NATO security arrangements, Greenland was granted home rule by Denmark in 1979 and was released from its obligations to the European Union in 1982.

Greenland is some 2500 kilometers long from North to South and some 1000 kilometers wide from East to West. The small population (58,000: 85% Inuit and 15% Danish) is concentrated on the coasts. The shortest distances from Greenland territory to the U.S. and to Russia are both about 2000 kilometers.

Greenland is in economic difficulties. Since the closing of the Black Angel lead and zinc mine in 1989, it has suffered from rising unemployment. Its main industry, fishing, does not offer long-term promise. The public sector employs 2/3 of Greenland's 23,000 person labor force, and Denmark supplies half of its $667 million annual revenues. The additional employment and income resulting from nuclear storage facilities in Greenland could be a significant incentive for the acceptance of such facilities. The contribution of such an operation to nuclear disarmament and world peace could be an incentive to environmentalists in Greenland and Denmark.

In order to remove material rapidly from national control, it could be stored on an interim basis at sites associated with nuclear weapon storage bunkers (which we presume exist even if they are not currently used) at Thule Air Force Base on The coast in the far Northwest. Over a few years, a long-term storage facility could be constructed inland, where access and egress could be even more tightly controlled.

Thule AFB is isolated. The nearest "municipality" (population 800) is 200 kilometers away. The next nearest "municipality" is 600 kilometers distant. Thule AFB is as near to the North Pole (1500 kilometers) as it is to the capital of Greenland. From the point of view of a location that is equitable with respect to the U.S. and Russia, Thule AFB is well situated. It is roughly 3000 kilometers from both U.S. and Russian territory, and it: is somewhat more than 4000 kilometers from Moscow and Washington, D.C.

As in all remote areas, construction and operations in Greenland would be difficult. Apart from the logistic difficulties resulting from weather and remoteness, there are construction difficulties. Eighty-five percent of the island is covered with ice, and the northern two-thirds of the island consists of permafrost. Moreover, one-third of the island, in the Northeast, (Thule is in the Northwest) has been designated a national park, the world's largest. If a long-term storage site were to be on national park land, this would raise political concerns. But if such concerns and difficulties can be overcome, the storage site(s) could be placed in a location(s) with strong physical security.

In concept, the interim site would have physical security at least meeting Physical Security Convention requirements and at least as stringent as that now provided for U.S, or Russian strategic material. It would be more manpower intensive than the long-term site to compensate for its relatively improvised nature. Also, the ratio of national to multinational personnel would probably be higher than that at the long-term site, to allow time for the buildup of the appropriate physical security force for long-term operations.

Once the long-term site was in full operation, its physical security would be superior to that of the interim site. The long-term site would have engineered choke points to retard the removal of material at a rate greater than some agreed figure. It would also have engineered features to facilitate the total interdiction of material removal under emergency conditions.

PHYSICAL SECURITY FORCE

Nations storing material at the facility would be entitled to participate ln a multinational force providing physical security. An on-site force would be capable of being supplemented within a short time by emergency response forces from the participating nations. Both the interim and long-term sites would have redundant communications to ensure reliable links to these emergency response forces.

The multinational force and the facility(s) would be designed to prevent any one nation. including the host nation (if protection from the host nation were deemed necessary), from removing material Without authorization and to prevent unauthorized sub-national access to the material. In the case of Greenland or other such remote areas, the rarefied logistics would slow the introduction of hostile forces and facilitate the interdiction of attempts to remove the strategic material. Interdiction could depend, first, on engineered chokepoints described in the subsection on transportation below; second, on emergency response forces on the ground; and, third, on non-nuclear air strikes against retarded hostile forces.

Although it is essential to design the facility and the physical security force to resist seizure of the material, the likelihood of military action being visited on the host nation would be exceedingly small. Sub-national attacks will be strongly discouraged by the deterrent effect of the physical security arrangements. National attacks -- by the U.S., Russia, or the host government -- will be discouraged by the transparency of each party's intentions not to seize the material as well as by the physical difficulty of doing so.

TRANSPORTATION

The international transportation of Japanese civil plutonium has raised wide controversy Originally, the difficulties resulted from a debate over the stringency of physical security. Later concerns were based on environmental grounds -- fears of the consequences of accidents and spillage. U.S. legislation now limits the air transportation of civil plutonium over U.S. territory, and IAEA "regulations" -- soon to be issued in revised form -may raise pressure for similar international limits on civil plutonium air shipments.

At the same time, land. sea, and air transportation of nuclear weapons and their nuclear components has been conducted securely by the nuclear weapons states with little public notice. An exception, and one of the greatest nuclear transportation traumas, was the 1968 crash at Thule AFB of an American B-52 carrying nuclear weapons. Continuing "revelations" wich respect to a hitherto secret U.S.- Danish 1957 nuclear defense agreement are still appearing in the press.

It is possible that the rapid transport of hundreds of tons of nuclear material would be marred by one or more such accidents. The spread of nuclear material in an accident would be greatly reduced if the material were not configured as nuclear "physics packages" that contained high explosives or if the material were transported by ship rather than air. Worse than an accident would be an attack on a shipment and its possible seizure. The likelihood of this can be minimized by avoiding a large number of ill-guarded and highly publicized shipments, i.e., by consolidating shipments into a relatively small number of highly-guarded operations.

The implication for the SMART concept is that there is likely to be controversy with respect to transportation. But stringent physical and environmental security with respect to the transportation and the overall contribution of the operation to world peace should mitigate the controversy. With careful preparation of the case for SMART as an effort of historical international importance, support for the operation should not be undermined by limited mishaps.

The transportation load -- and the chances of mishaps would grow if material were expected to be moved back and forth frequently This is one reason for caution against regarding the storage site as a nuclear material bank. Even if the operational arrangements avoided frequent returns of material for civilian use, undismantled "physics packages" might be returned for dismantlement -- with yet another shipment of pits from these dismantled weapons back to the storage site(s). And the ultimate disposition of the nuclear material would probably require removal of the material from the storage site(s).

These later shipments from -- and possibly back to --the storage site(s) will have one advantage. Unlike the original removal of material from the U.S. and Russia, there will be little time pressure. Consequently, this later transportation can be optimized with respect to consolidation of shipments, timing, and secrecy.

Depending on the site(s), land, sea, and air transportation may all be feasible. Thule AFB as an interim site would be accessible by air and, for some of the year, by sea. Later transportation to and from a long-term inland site in Greenland could be by air or by specially constructed (and interdictable) land access.

The long-term site would be designed to allow relatively rapid emplacement of strategic material but to retard its rapid removal. This retardation could be accomplished through a variety of means. Examples are overpacks that could not fit through narrow tunnels and that could not easily be transported long distances, backfilling of high throughput tunnels and dismantling of railroad tracks, and removal of handling equipment for high volume operations. The logistics for lower throughput operations would remain in place.

The long-term site would have additional features that could be activated in an emergency to block the removal of material altogether. Foams, contaminants, and the collapsing of tunnels or other logistical chokepoints could block the removal of material for days, weeks, or months. Of course, this engineered interdiction would be supplemented during the entire period by the actions of the response forces.

RELATIONSHIP TO EXISTING LAWS, NEGOTIATIONS AND OTHER ACTIVITIES

The U.S. Atomic Energy Act and the Nuclear Non-Proliferation Act restrict exports of special nuclear material. These acts might be amended in the name of nuclear nonproliferation and disarmament to permit the coordinated export of U.S and Russian strategic material to a multi-nationally controlled storage site(s).

An alternative would be for the U S. and Russia (and other participants) to each maintain internal control of storage bays containing their respective strategic material -- and to each retain title to that material. Under this arrangement, the transfer of material to the storage site(s) might arguably not be considered an export The bays would operate under agreed inspection arrangements -- balancing the need to verify the quantities of nuclear material against the need for preventing unauthorized access co classified weapon shapes. The multinational force Would dominate the perimeter(e) beyond these bays in order to prevent unauthorized. access or egress.

A slow-moving negotiation is currently underway in Geneva to conclude an international convention to prohibit the production of fissile material for weapons or outside of international safeguards. The convention allows the continued production of material for "peaceful" purposes.

It would be inconsistent with the concept of removing weapon- usable material from national access if the same type of material continued to be produced by that nation. However, the concept would not be entirely undercut if -- in the case of Russia -- some 1,000 metric tons of material were removed and only 1 or 2 metric tons of material were produced annually and consumed (or transferred to third country storage) at a commensurate rate.

Even without the fissile material cutoff convention, all nuclear weapon states have said that they no longer produce fissile material for weapons. It would be desirable -- but not absolutely necessary in the immediate term -- to extend this cutoff to strategic material for civil purposes -- or, at least, to the production of strategic material in excess of current civil consumption. It is essential to understand that SMART would be undercut if sizable buildups of weapon-usable inventories were allowed to occur in the U.S and Russia for any purpose, military or civilian.

The U.S. is preparing to issue a "record of decision" on plutonium disposition by the end of 1996. It is important that no decision be made that would undercut: our leverage with Russia in negotiating the removal of such material from both our nations.

The G-7 and Russia are meeting at a "nuclear summit" in Moscow April 19-20, 1996, to discuss nuclear safety, fissile material control, and plutonium disposition There will be subsequent discussions among these countries. The realization of the concept outlined here would be delayed if it were negotiated by all eight parties before essential agreement was reached between the U.S. and Russia. However, because the G-7 consists of the closest U.S. allies, it would be essential to keep them informed of U.S. intentions and U.S.- Russian progress.

Our concept calls initially for talks, with later steps taken in sequence. We now discuss this sequence.

SEQUENCING OF EVENTS

The concept is to permit rapid removal of excess weapon-usable material from national control in order to retard any reversal of nuclear disarmament and nonproliferation. For that reason, we place emphasis on early and interim activities -- with longer term arrangements analyzed in parallel but executed later.

The following is an illustrative sequence of events to implement the concept: planning and analysis in the U.S. of the concept and the operational aspects of talks, - preliminary consultations with Russia and the rest of the G- 7 on the principle of talks, with simultaneous analysis of interim storage sites and interim operations; - if agreed, announcement and endorsement of the talks; - conduct of the talks aiming at rapid interim storage and covering what constitutes excess material, sites and interim facility designs, interim decision rules for material return, interim plans for cost coverage, consolidation points on national territory, shipping schedules, and transportation modes, interim physical security, interim cutoff of all production of weapon- usable material (or, at least, avoidance of sizable inventory buildups), and simultaneous full engineering design of the interim site, and full operational design for packaging, transportation, and physical security as well as preliminary analysis of the long-term site, the physical security force, and long term operations; negotiation with the host government(s) for the interim site(s) to secure their agreement, and simultaneous initiation of consolidation and packaging of strategic materials at U.S. and Russian shipping Sites; construction of interim facilities, low volume shipping and interim operations, and simultaneous full engineering design of a long-term site and negotiation of a long-term agreement covering the same points as the interim agenda outlined above (with other participants in multinational activities brought into the negotiations); full interim operations, formation of a physical security force, beginning of movement of materials from other nations into the interim facility, construction of the long-term facility; long-term packaging of material and storage at the lang-term facility. reconfiguration of the long-term facility to prevent high rates of material removal. How long would all this take? The major elements of time to get the material out of the U.S. and Russia are (1) negotiations, (2) preparation of an interim storage facility and packaging of material for shipment, and (3) transportation.

Negotiations could take as long as arms control negotiations usually take except for two Unique features. First, there is a sense of urgency, which is the rationale of the entire SMART concept. Second, because SMART does not require any final decisions about the fate of the weapon-usable material, the talks can bypass some of the most intractable issues.

The HEU agreement was negotiated without any particular sense of urgency (judging by the 20 year time frame for its execution) and made the final decision on the fate of the HEU, yet The agreement took less than 10 months to initial, six more months to negotiate, and 11 months to develop implementing contracts." Given a high priority, SMART could take a comparable time.

Preparation of the interim storage site and packaging of the material need not be complete before shipments begin, Only enough storage and enough packaging needs to be in place to accommodate each shipment. with concurrent preparations, shipments might begin shortly after agreement on interim arrangements.

The key time period is that to ship tens of thousands of pits, hundreds of tons of additional weapon-usable material, and possibly thousands or tens of thousands of intact "physics packages". If weapon dismantlement continues while pits and other material are shipped, the shipments of intact "physics packages" can be minimized. This is why current programs -- and even accelerated programs -- of weapon dismantlement complement and do not compete with SMART. Our judgment is that, if the shipments to at least an interim storage site are scheduled to be completed within a decade. some intact physics packages will need to be considered for shipment. A more accelerated shipping schedule could entail even more shipments of intact physics packages. However, if the situation were urgent, such an accelerated schedule might be the wisest course.

The construction of a long-term storage facility and the organization of a long-term physical security force need not govern the pace of removal of material to an interim facility. By analogy with the Mayak facility we would expect such a long-term facility to be constructed and receiving material before the end of a decade.

IV. COMPARISON OF ALTERNATIVES

Three options worth comparing to SMART are The Mayak option, the present program of storing material from dismantled Russian weapons and other weapon-usable material at a new $300 million high-security facility in Russia, presumably blending the remaining HEU over a schedule yet to be determined, and disposing of the plutonium in some fashion -- with burning of plutonium in Russian breeder reactors favored by MINATOM and used as the basis of this discussion. The CANDU option, the program proposed by AECL of Canada to burn U.S. and Russian plutonium in the existing Bruce CANDU reactors in Canada, with fabrication of MOX fuel for the CANDU's to take place in the U.S. and Russia. The Hanau option, in which Russian plutonium is fabricated into MOX fuel in the existing German fabrication facility at Hanau; and the MOX is then burned in European light water reactors. In all these options, there is no defined policy for handling the excess HEU beyond the 500 metric tons covered by the 20 year U. S. - Russian deal. And in all these options, we assume that there are as much as 80 metric tons of excess plutonium in the United States and 180 metric tons in Russia.

The options can be compared in terms of the length of time during which material is vulnerable to national seizure and the extent to which one option leaves open the possibility of later moving to another option. The comparison results in the following assessment:

Mayak option. The storage of material from dismantled weapons and other sources at Mayak is appropriate for all options. The SMART option gives the Mayak facility a focus -- to package excess material for shipment out of country. The existing Mayak option is seen by MINATOM as a precursor to the burning of plutonium in Russian breeder reactors over a 20-90 year period. The 20 year disposal period depends on the dubious assumption that Russia will bring four BN-800 breeder reactors on line by the year 2004; the 90 year disposal period depends on the assumption that one such reactor will become available for plutonium burning by 1998. Realistically, the Mayak option will leave excess weapon-usable material in Russia and accessible to national diversion for a period that may be longer than that of any other option.

CANDU option. U S. and Russian plutonium could be burned in Canada's 8 Bruce A+B reactors over the next 30 years. An important question is where the plutonium would remain until it was burned, The present concept seems to be that it would remain in the U.S. and Russia, where it would be fabricated into MOX This would not address the problem of national diversion in the near term. Indeed, there is some Russian interest in acquiring a CANDU reactor" If the CANDU burning of Russian plutonium occurred in Russia, the accessibility of the plutonium to Russian national seizure would be extended in time.

An interesting variant on the CANDU option would be the SMART CANDU option. This would have the plutonium shipped to Canada as soon as possible as pits, other metallic forms, or oxide -- whatever form was available for early shipment. The disassembly of pits to generic shapes, the conversion of metal to oxide, and the fabrication of MOX fuel might or might not take place in Canada depending on how much additional transportation is acceptable. If Canada were acceptable to Russia as an interim plutonium storage site pending burning in reactors, the removal of the material from the threat of Russian (or U. S.) national diversion could take place as quickly as under the "pure" SMART option.

Hanau option. The fabrication of Russian plutonium into MOX and its burning in European reactors could take at least 25 years, If U.S plutonium were treated identically, the conversion of both U.S. and Russian plutonium into spent fuel could take about 40 years. It is not clear that the Russians would agree to store 180 metric tons of their plutonium in Germany at one time, so a "SMART Hanau option" might not be realistic. Indeed, the Russian interest in Hanau may focus on the possibility of having a Hanau-type MOX fabrication facility built in Russia, extending the accessibility of the material to Russian national seizure.

SMART. SMART would be consistent with later burying or burning of plutonium in Russia, the U.S., Canada, or Europe. But it would also allow plutonium to be permanently disposed of in ways that did not entail the proliferation risks of shipment to diverse commercial reactors SMART, at any rate, would protect the material from national diversion while a considered decision was made on the question of how to conduct permanent disposition Moreover, SMART would address the problem of the vulnerability of HEU during the decades before it was blended and released on the market as LEU. Our concept of SMART -- removing material from national access in a decade -- would reduce by more than 90 per cent the nuclear weapon capabilities of the U.S. and Russia far sooner than any other option.

APPENDIX: SOURCES AND ASSUMPTIONS OF WEAPON-USABLE MATERIAL CHARTS

What follows are the sources and assumptions used to produce Tables 1 and 2 and Figures 1 and 2 in the text. While we have used the best available unclassified sources and have made what we consider reasonable assumptions, we must emphasize that the data should be considered illustrative. We do not use uncertainty bands around the curves so as not to obscure the overall trends, but there is substantial uncertainty surrounding the exact values of these curves.

TABLE 1 AND FIGURE 1

The number of operational warheads for both the U.S. and Russia are based on estimates by the Natural Resources Defense Council They have estimated the number of operational warheads in 1989, 1990, and 1995. They have also projected what the number of operational warheads will be in 2003. we used linear interpolation to find the values for the other years between 1989 and 2003.

Our curves for the total amount of material is based on 1995 estimates for both U.S. and Russia. For the U.S. it is estimated that in 1995 there was 100 metric tons of plutonium and ?75 metric tons (rounded to 800 in Table 1) of HEU. We assume that the 100 metric tons of plutonium is constant from 1989 to 2003. For HEU, we assume that there has been a steady consumption of 4 metric tons per year due to naval reactor use. For the years from 1991 to 2003, the 1995 amount of HEU is adjusted to reflect this consumption. For the years 1989 and 1990, the U S. enrichment plants were still producing HEU. We assume that this production compensated for the naval reactor use so that the U S. total HEU in 1989 and 1990 was the same as in 1991

For Russia, it is estimated that in 1995 there were 200 metric tons of plutonium and 1,200 metric tons of HEU. Russian plutonium production is continuing at the rate of about 2 metric tons per year. From 1992 to 2000, the plutonium is assumed to increase 2 metric tons per year. After 2000, the production is assumed to stop and the amount of plutonium is constant at 210 metric tons. In 1990 and 1991, the plutonium production was estimated to have been 3 metric tons per year and in 1989 four metric tons per year. The amount of HEU is assumed to be constant from 1989 to 1995. After 1995, the HEU decreases due to the blend down agreement with the U.S. It decreases 10 metric tos per year from 1995 to 2000, and 30 metric tons per year from 2000 to 2003.

TABLE 2

In 1986, the number of Soviet nuclear operational warheads (both strategic and tactical) peaked at 45,000. Since that time the number of Soviet (now Russian) nuclear warheads has declined sharply. In 1995, the number of operational Russian warheads is estimated to be only 12,000 and may decline to only 6,000 by the year 2003 in order to meet the Start II ceiling.

ln l986, the U.S. had 23,000 operational warheads. This number had been about the same throughout the 1980s, after declining from a peak of about 33,000 in 1967. This number has also sharply declined. It is estimated to be around 9,000 in 1995 and may fall to about 5,000 in 2003.

Together, the number of operational or active warheads will have declined by about 57,000 warheads from 1986 to 2003 Already, 47,000 warheads have been removed.

Based on the planning factors described in this Appendix, we estimate that, in 2003, the United States will have about 80 metric tons of plutonium and 670 metric tons of HEU outside of active warheads. This material will be in inactive warheads, nuclear pits and material inventory. Russia will have about 190 metric tons of plutonium and 970 metric tons of HEU outside active warheads by the year 2003. The two countries' combined total will be 270 metric tons of plutonium and 1640 metric tons of HEU. This material is enough to make about 100,000 nuclear warheads.

FIGURE 2

The curves showing the material in operational warheads are the same as the curves on Figure 1 for the years 1989 to 2003. Prom 2003 on, the number of operational warheads is assumed not to change.

The SMART drawdowns are assumed to take five years to negotiate, so both the U.S. and Russian drawdowns start in 2001 and their starting points are the total amount of material curves from the Figure 1. For both curves, the excess material is removed in the five year period from 2001 to 2006. The removal rate is linear over the five year period To determine how much material is excess. it is necessary to determine how much material will remain in 2006.

For the U.S., we assume that in 2006 it will have not only the material for its 5.000 operational weapons but also a reserve of 43.5 metric tons of HEU and 10.5 metric tons of plutonium which could be used to make nearly 3,000 additional weapons if needed We also assume that there is a reserve of five years worth of consumption of naval reactor fuel (At 4 metric tons per year of HEU, this reserve is 20 metric tons). This leaves 664.5 metric tons of excess material to be removed over the period 2001-2006. In addition, 20 metric tons of HEU is consumed during this period by naval reactors.

For Russia, we assume that in 2006 it will have not only the material for its 6,000 operational weapons but also a reserve of 15 metric tons of plutonium for possible civilian reactor use and two years worth of HEU to be blended down (at 30 metric tons per year) to fulfill its contract with the U.S. This contract lasts until 2015. Therefore, the last 7 years worth of blend down HEU will be shipped out of Russia and then shipped back when it is needed to be blended down. This results in 994 metric tons of material being shipped out of Russia during the period 2001-2006. An additional 150 metric tons of HEU (30 metric tons per year) is blended down and shipped to the U S. during this time.

Our non-SMART drawdown curves are a best guess of how the stockpile might be reduced without SMART -- weighted on the side of optimism that plutonium will be disposed of rapidly. For the U.S. the curve from 1989 to 2003 is just the total material curve from Figure 1. From 2003 to 2006, the material is only reduced by the 4 metric tons per year of naval reactor HEU consumption. Starting in 2006, we assume that excess plutonium is disposed of (by either burning in reactors or by mixing with waste and vitrifying) over a 20 year period, The total amount of plutonium disposed of is 82.5 metric tons. In 2015, we assume that the U.S. begins to blend down and dispose of its excess HEU at the rate of 30 metric tons per year. We wait until 2015 to be sure that the uranium market can absorb this material. As will be seen, we assume that the Russians continue to blend down HEU during this time. The total U.S. HEU blended down is 516 metric tons. At 30 metric tens per year, this takes until 2033. During all of this time, we assume that naval reactors continue to consume 4 metric tons of HEU per year. Even though all excess U.S. material is consumed by 2033, we continue the curve to 2036 (with continuing naval reactor consumption) to match the Russian curve. At 2036 the U.S. has only the material for its 5,000 operational weapons and a reserve of five years worth of naval reactor consumption (20 metric tons of HEU).

For Russia, the curve from 1989 to 2003 is, as in the U.S. case, just the total material curve from Figure 1. From 2003 to 2006, the curve reflects the blending down and removal of 30 metric tons per year of HEU. In 2006, we assume that Russia 's excess plutonium is disposed of over o 20 year period. This involves a total of 174 metric tons of plutonium. During this time, the current HEU blend down contract continues at a rate of 30 metric tons per year until the contract expires in 2015. Starting in 2015, we assume that Russia disposes of its remaining HEU by continuing to blend down the material at a 30 metric tons per year rate. A total of 610 metric tons of HEU is blended down after 2015, and it takes until 2036 before all of the material is disposed of. In 2036, Russia only has the material for its 6,000 operational weapons and a reserve of 15 metric tons of plutonium for possible use in civilian reactors.