Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
4 Declarations and Dismantlement The nuclear weapons and fissile materials that will become excess as a result of arms reductions are only a part of the world stocks of these items (see Chapter 11. Thus the measures taken to address the urgent problem of manag- ing excess nuclear weapons and fissile materials from dismantlement of weapons through storage and long-term disposition of the resulting fissile materials must be seen not only as ends in themselves, but also as steps toward an overall regime designed to achieve higher standards of security and transparency for the total stocks of weapons and fissile materials in the United States and the former Soviet Union-and, ultimately, worldwide. The com- mittee envisions a reciprocal regime, built in stages, that would include: 1. reciprocal declarations of total stocks of nuclear weapons and fissile materials; 2. cooperative measures to confirm and clarify those declarations; 3. agreed, monitored subtractions from the stocks available for military use, including: · monitored warhead dismantlement, · commitments never again to use agreed quantities of fissile materials for weapons purposes, · safeguarded storage and long-term disposition of excess fissile material stocks, and 87
88 DECLARATIONS AND DISMANTLEMENT 4. agreement on and monitoring of additions to those stocks, including what- ever warhead assembly continues, and a verified cutoff of production of fissile materials for weapons. Such a regime, if agreed between the United States and Russia, would directly serve the three security objectives outlined at the beginning of this report- limiting the risk of theft, limiting the risk of breakout, and strengthen- ing arms reduction and nonproliferation. It would also provide a sound base for building a similar global regime. Although complex and far-reaching, such a regime can be approached incrementally, contributing to confidence at each step while posing little risk. Measures specific to excess weapons and materials, such as monitoring of warhead dismantlement (discussed in this chapter) and secure, safeguarded storage of excess fissile materials (discussed in the next chapter), will be essential building blocks of this larger regime. Virtually none of this broad regime is currently in place. But the end of the Cold War offers an opportunity to begin building it that is both unprecedented and unlikely to be repeated. The Clinton administration, in its nonproliferation initiative of September 27, 1993, has taken the first steps in this direction, announcing that it would propose a global convention to ban production of fissile materials for weapons and that it would voluntarily submit excess U.S. fissile materials to International Atomic Energy Agency (IAEA) safeguards.) In addition, on December 7, 1993, Energy Secretary Hazel O'Leary declassified the amount of weapons-grade plutonium that the United States has produced and the amounts held at several Department of Energy (DOE) sites.2 More remains to be done, however. Weapons and fissile materials in the former Soviet Union are currently of greater concern than those in the United States. Achieving substantial im- provements in the management of these weapons and materials in the former Soviet Union, however, will in many cases require reciprocity from the United States. THE CASE FOR A BROAD REGIME Some more limited objectives can be achieved by efforts focused only on excess weapons and materials such as the highly enriched uranium (LIEU) ~ White House Fact Sheet, "Nonproliferation and Export Control Policy," September 27, 1993. 2 Secretary O'Leary announced that the United States had produced 89 metric tons of weapons-grade plutonium. In addition, the Hanford site produced 13 tons of reactor-grade plutonium. DOE also declassified the total current plutonium inventory at Savannah River, Rocky Flats, Hanford, Los Alamos National Laboratory, Lawrence Livermore National Laboratory, Argonne National Laboratory-West, and Idaho National Engineering Laboratory. But the "total quantity of plutonium at Pantex remains classified due to a proliferation concern that the amount of plutonium in a nuclear weapon could be determined by correlating the number of dismantlements being released to the public, to future increases in die plutonium inventory." The DOE press release also stated that "today's release should be considered only a beginning of a process." DOE Press Release, December 7, 1993.
DECLARATIONS AND DISMANTLEMENT 89 purchase under negotiation and the planned fissile materials storage site to be built in Russia with U.S. assistance. But measures focused only on these excess stocks would leave the size and status of the other stocks unknown. Creating a broader regime covering the total stockpiles of nuclear weapons and fissile materials would make clear that this total stock was of legitimate interest to the world community and would have the following specific benefits: · Strengthening Current Arms Reduction Agreements: Measures to veri- fiably eliminate the warheads to be retired under recent arms agreements, monitor the resulting fissile material, and build confidence that there were not other large, unmonitored stocks of excess weapons and materials available would substantially strengthen the arms reduction regime, complementing the limits on launchers in the Strategic Arms Reduction treaties (START I and START II). Such measures would work synergistically with the measures already agreed, to make rearmament more difficult, costly, time-consuming, and observable and therefore less likely. ·Providing the Basis for Deeper Reductions: Similarly, a regime for agreed, monitored, balanced reductions in the stockpiles of nuclear weapons and fissile materials would lay a foundation for deeper, post-START II nuclear arms reductions. Without a regime designed to build confidence over time in the knowledge of the stockpiles of weapons and fissile materials, concerns about the military advantage that might be gained by retaining large hidden stocks could make the United States, Russia, and other nuclear powers reluctant to agree to reduce to substantially lower levels. · Improving Resistance to Theft: Such a wide-ranging regime would pro- vide the basis for significantly improving security and safeguards for nuclear materials, particularly in the foyer Soviet Union, where the current dramatic political and economic transformation necessitates strengthening these vital functions. In order to make comprehensive improvements, it is essential to have an understanding of how large the stocks of fissile materials are, where they are located, and the like. The requirement to provide declarations would focus each party to the regime on the task of accounting in detail for all the material in its possession and reviewing its own management procedures. Moreover, the declarations and the visits involved in confirming them would provide a more educated basis for U.S. offers of assistance to Russia in improving safeguards and security, allowing discussions on this subject to be more meaningful and comprehensive, and less impeded by secrecy (see Chapter 5~. · Strengthening the Nonproliferation Regime: Including monitored sub- tractions from the total stocks of nuclear weapons and fissile materials in the arms reduction regime would help convince the rest of the world that the nuclear states were seriously pursuing their obligations under Article VI of the Non-Proliferation Treaty (NPT). Indeed, a number of non-nuclear-weapon states have specifically called on the United States and Russia to agree to such measures. In particular, agreement on the first steps toward such a regime
90 DECLARATIONS AND DISMANTLEMENT during preparations for the critical 1995 NET review would help create a favorable atmosphere for an indefinite or long-term extension of the treaty. Similarly, a new openness and willingness to accept international monitoring on the part of the largest nuclear powers would improve the prospects for gaining acceptance of strengthened safeguards elsewhere. Applying strict standards of security and accounting to excess fissile materials resulting from arms reductions could provide the base for setting similar standards for civilian fissile materials worldwide (see below). · Providing Information: Current public knowledge of the stockpiles of nuclear weapons and fissile materials is limited, although as noted above, the United States has recently begun declassifying some of this information. As yet, Russia has not reciprocated. Total inventories of weapons and the size of reserves and "excess" stocks have not been authoritatively disclosed by either country. The uncertainties in U.S. intelligence estimates of Russian stocks amount to thousands of weapons and tens of tons of fissile material.3 Each side's intelligence services have in the past spent billions of dollars attempting to acquire the information that would be exchanged under this regime. Such information provides a basis for defense and arms control planning; for coordi- nating efforts such as the HEU purchase and the planned plutonium storage site in Russia; and as mentioned, for more educated offers of assistance in manag- ing these weapons and materials. · Building Confidence: Establishing such a regime of transparency and re- ductions in nuclear weapons and fissile materials would reflect and deepen the significant Russian-American cooperation in denuclearization. Experience to date with agreements such as the Strategic Arms Limitation Treaties (SALT I and II), the Intermediate-Range Nuclear Forces Treaty (INF), the Conventional Forces in Europe Treaty (CFE), and START I and II suggests that working together to reach agreement on reductions, declarations, and monitoring, and then to implement those agreements, has ~ far-reaching confidence-building effect. Indeed, the exchange of information alone has generally proved helpful in resolving uncertainties and concerns, even in cases where the data initially provided were not immediately accepted as accurate.4 Such a regime could 3 Lawrence Gershwin, National Intelligence Officer for Strategic Programs, told Congress in 1992, that the Central Intelligence Agency estimated that Russia then had 30,000 nuclear weapons. "The uncertainty is plus or minus 5,000, which gives you a sense of how uncertain it is. That uncertainty has not improved . . . because we still don't get direct information on how many weapons are at sites and how many are in inventory." See House Defense Appropriations Subcommittee, Department of Defense Appropriations for 1993, Pt. 5, p. 499, May 6, 1992. Recent statements by Victor Mikhailov, Russia's Minister of Atomic Energy, suggesting that Russia had significantly larger stockpiles of weapons and HEU than previously believed, have highlighted those uncertainties (see Chapter 2). 4 hi the case of He CFE treaty, for example, the requirement to exchange information provided each party with a wealth of hitherto unavailable information. Although initial Soviet data on treaty-limited items in the limitation zone were not accepted, it was the requirement for full reporting itself that provided the basis for questioning the Soviet figures and ultimately resolving the issue; in the end, few would argue that the declarations required by CFE were not a useful, indeed essential, part of the CFE regime. Given the large quantities of nuclear weapons and materials that would have to be accounted for in a regime such as that
DECLARATIONS AND DISMANTLEMENT 91 address some concerns that have undermined confidence: for example, in both the United States and the former Soviet Union, opponents of the ongoing reductions have pointed to the lack of any requirement to venflably dismantle the weapons to be retired as a key justification for continuing suspicion. ·Improving Democratic Management: In democracies the information needed to make decisions on security and related issues should be available to the public. Secrecy in this area has affected not only security debates but environment, safety, and health (ES&H) discussions as well, since the United States and Russia have been reluctant to release ES&H information that might provide details of weapons or fissile material production. The committee shares the view of President Reagan's Blue Ribbon Task Group on Nuclear Weapons Program Management: One of the national security responsibilities of DOE leadership is to make available sufficient information to allow informed public debate on nuclear weapon issues. The Task Group urges that DOE review its classifi cation procedures to ensure that criteria are based upon current requirements rather than historical precedent.S The Secretary of Energy has statutory authority under the Atomic Energy Act to declassify restricted data, and Secretary O'Leary has begun what she has said will be a continuing process with the declassification of some plutonium stockpile data on December 7, 1993. In addition, the 1993 Defense Authonza- tion Act specifically granted authority to declassify stockpile information if the United States and Russia reach agreement on reciprocal release of such data.6 In Russia, some organizations with major responsibilities in these areas (such as the Foreign Ministry and GOSATOMNADZOR, the Russian equiva- lent of the Nuclear Regulatory Commission) appear not to have access to information relating to nuclear stockpiles that is necessary to carry out their duties.7 The need to provide this information to the United States would envisioned in this chapter, and the long time over which they were produced, ambiguities can be expected in this case as well, requiring similar efforts to resolve them. 5 "Report of the President's Blue Ribbon Task Group on Nuclear Weapons Program Management," July 1985, p. 13, cited in Energy Research Foundation and Natural Resources Defense Council, Rethinking Plutonium: A Review of Plutonium Operations in the U.S. Nuclear Weapons Complex (Washington, D.C.: April 1992), pp. 52-53. 6 The primary relevantlanguage from the Atomic Energy Actis Section 142: "(a) The Department [of Energy] shall from time to time determine the data, within the definition of Restricted Data, which can be published without undue risk to the common defense and security and shall thereupon cause such data to be declassified and removed from the category of Restricted Data." In the case of Restricted Data determined to "relate primarily to the militarization of atomic weapons," this determination must be done jointly with the Department of Defense. The recent modification to this language appears in the conference report on the Department of Defense Authorization Act, Report 102-966, October 1, 1992, p. 338. 7 For example, while President Yeltsin has given GOSATOMNADZOR responsibility for regulating safeguards and security over both military and civilian nuclear materials in Russia, GOSATOMNADZOR officials report that they have been denied the access to information and facilities necessary to carry out this responsibility. See Mark Hibbs, "Watchdogs Say MINATOM Withholding Material Theft and Diversion
92 DECLARATIONS AND DISMANTLEMENT inevitably make it available to wider circles in Russia, beneficially broadening participation in decision making itself potentially a major step toward im- proving the management and security of fissile materials in Russia. · Providing Incentives: Some have proposed that the United States provide direct incentives, monetary or otherwise, to the states of the former Soviet Union for steps such as accelerating dismantlement or committing fissile material to peaceful purposes under monitoring. This is part of the idea behind the HEU deal: for example, dismantlement would free HEU that in turn would earn hard currency, providing a direct incentive for the dismantling.8 If such incentives are to be offered, there must be a means to check that the specified goals are being met, which a transparency regime would provide. · Strengthening Management Organizations: In Russia, the Ministry of Atomic Energy (MINATOM) and the military are struggling to meet the challenge of managing the large-scale reductions now in progress in the midst of a drastic weakening of central authority, the disappearance of their tradi- tional Cold War missions, and drastic declines in their former budgets and status. Substantial erosion of these organizations could greatly increase the risks of theft of nuclear weapons and fissile materials. A regime based on cooperation in nuclear reductions, combined with appropriate incentives for accomplishing particular tasks (such as warhead dismantlement and secure management of warheads and fissile materials) could provide these organiza- tions with a new and compelling mission to replace their old tasks and with the resources needed to carry it out. Similarly, if structured to utilize and expand on the capabilities of interna- tional organizations particularly the IAEA the regime the committee proposes could significantly bolster their ability to carry out their global non- proliferation roles. · Addressing Some Ukrainian Concerns: Ukrainian officials have repeatedly expressed concern that if they fulfill their denuclearization pledges and ship the nuclear weapons now on their territory back to Russia for disman- tlement, Russia might add the weapons or the materials in them to its own military stocks. Although the key Ukrainian nuclear concern is the more general security threat it perceives, Russian willingness to permit Ukrainian monitoring of the dismantlement of weapons removed from Ukrainian territory has been an important factor in discussions of this issue. A broader regime would go further in addressing these Ukrainian concerns. Data," NuclearFuel, August 16, 1993; and "Uranium, Plutonium, Pandemonium," The Economist, June 5, 1993. ~ The United States has now agreed, however, that it will not insist on transparency measures to guarantee that the HEU it purchases came from dismanded weapons rather Dan from excess stocks. Since Russia could therefore continue die deal for a number of years without dismantling any additional weapons, this fact may significantly limit the effect of the agreement as a dismantlement incentive.
DECLARATIONS AND DISMANTLEMENT 93 In 1992, a similar list of objectives led the U.S. Senate to attach the so- called Biden Condition to its resolution of advice and consent to ratification of the START I treaty, citing the risk of "loss of control of nuclear weapons or fissile materials in the former Soviet Union" and requiring the president to seek an arrangement to monitor the total stockpiles of nuclear weapons and fissile materials in the former Soviet Union and the United States, using "reciprocal inspections, data exchanges, and other cooperative measures." This condition is legally binding on the U.S. government, though when it must be carried out was deliberately left ambiguous.9 Indeed, some elements of the regime the committee envisions have been official U.S. proposals in the past. In 1953, President Eisenhower, in his Atoms for Peace speech, called for transfers of specific quantities of fissile materials from military stockpiles to civilian purposes under international safeguards. The idea of cutting off production of such materials and shifting some of the existing stocks to civilian purposes-known as "cutoff and transfer" was a major element of U.S. arms control proposals for many years thereafter. By 1965, this proposal had evolved into a formal U.S. proposal for monitored destruction of thousands of nuclear weapons and transfer of the resulting fissile materials to civilian stockpiles. Building such a regime will not be easy, however, despite the compelling motivations to do so. Far-reaching changes in the way the nuclear weapons complexes in both the United States and the former Soviet Union do business will be required, including the exchange of substantial quantities of information that is currently classified.~° The committee is convinced, however, that declassification of this information would advance U.S. security and nonprolif . ~ . . eratlon objectives. In principle, the most sensitive information related to stocks of weapons and materials would be the numbers and locations of currently deployed strate- gic forces, because of the possibility of an attack on those forces. Yet that information has been exchanged in great detail as part of the START I agree 9 Specifically, the condition requires that the President "seek" such an arrangement "in connection with any furler agreement reducing strategic offensive arms," including START II. Arguing that a requirement to reach such an agreement in parallel with START II could seriously delay that treaty, the Senate Armed Services Committee, under the chairmanship of Senator Sam Nunn, opposed the Biden Condition in its report on the START treaty. But when the subsequent Foreign Relations Committee report specified that the "in connection with" language was not intended to prevent action on START II in the absence of such an arrangement, Senator Nunn withdrew his opposition, and the treaty, with the attached condition, was approved overwhelmingly by the Senate. In May 1993 testimony to the Senate Foreign Relations Committee, Secretary of State Warren Christopher acknowledged that no action had yet been taken to implement the Biden Condition, but indicated that the administration intends to fill this gap. A In particular, a range of information related to the size and location of all parts of the stockpiles of nuclear weapons and fissile materials, and information related to the weapons components-specifically the amount of fissile material they contain could be declassified as part of the regime proposed here. If, in some cases, the amount or isotopic composition of fissile material in particular components was considered sensitive, somewhat more complex monitoring arrangements could be devised that would provide confidence in overall figures without revealing those related to a particular specific device.
94 DECLARATIONS AND DISMANTLEMENT meet. It is difficult to argue that the numbers and locations of weapons that are not deployed, or of materials that are not fabricated into weapons, are more sensitive than those of weapons that are part of the active military force. Thus, it is difficult to justify the current practice of releasing information on all deployed strategic delivery systems, while keeping information about the corresponding weapons and about most aspects of the stocks of fissile materials secret. As already noted, DOE has begun to address this issue with the release of some information about plutonium stocks. Objections that might be raised against a declaratory regime are similar to those introduced when inventory declarations became part of the INF treaty, START I, and START II; yet the parts of those regimes involving declarations, verification, and reductions that have been carried out to date have proven beneficial. A traditional objection is that if the other party underdeclares its holdings and keeps a secret stock, it could gain a significant advantage if drastic reductions were carried out. Clearly, however, that is more an argument against the drastic reductions than against the exchange of information. Moreover, the argument simply reinforces the case that deep post-START II reductions may be impossible to achieve without greater confidence in each party's knowledge of the other's stocks of nuclear weapons and fissile materi- als. Another argument occasionally heard is that ignorance of the total stock- pile itself keeps the opponent guessing and therefore has some deterrent effect. Yet in an age of cooperation, transparency is more stabilizing than ignorance. Such a regime would involve costs for the associated monitoring and coop- erative measures. Monitonng of warhead dismantlement, for example, would probably require a permanent foreign presence at the dismantlement sites. These costs would probably be in the range of tens of millions of dollars per year for each side (not counting the costs of dismantlement itself). Russian Attitudes. Russian officials have expressed differing views con- cerning the different parts of such a regime. On February 12, 1992, Russian Foreign Minister Andrei Kozyrev, in a comprehensive statement to the United Nations Conference on Disarmament in Geneva, called for "a reciprocal ex- change of data between all nuclear powers on the number and types of existing nuclear weapons, the amount of fissionable matenals, and on nuclear weapons production, storage, and elimination facilities." This proposal, however, was never pursued by either side, and officials at MINATOM and other agencies ~ To verify warhead dismantlement at existing facilities, a single perimeter-portal monitoring system would be needed in Me United States (at Pantex), and several in Russia. The existing perimeter-portal monitoring system at Vodcinsk cost $45-$50 million to install, wide an annual operating cost of $10-$20 million. (See U.S. Congress, Congressional Budget Office, U.S. Costs of Verification and Compliance Under Pending Arms Treaties (Washington, D.C.: U.S. Government Printing Office, September 1990).) The L\EA safeguards more Wan 1,000 installations worldwide (see Table 2-1), wide an annual safeguards budget of $60-$70 million, Cough a few of these facilities (particularly enrichment, reprocessing, and plutonium fuel fabrication plants) account for a disproportionate share of We safeguards costs. The proposals outlined in this chapter might require roughly doubling We annual IDEA safeguards budget.
DECLARATIONS AND DISMANTLEMENT 95 have expressed considerable resistance to such broad-ranging transparency. More recently, on August 17, 1993, Russia's representative to the Conference on Disarmament reiterated Russia's willingness to agree to a verified cutoff of production of fissile materials for weapons, to put excess fissile materials under IAEA safeguards, and to exchange information relating to such matenals, on the basis of reciprocity. Agencies other than the Foreign Ministry, however, may continue to have different views. The most consistent theme the committee has heard from Russian officials is that anything more than the most limited transparency measures would require reciprocity from the United States, which U.S. negotiators have so far not been prepared to offer. The committee believes that persistent diplomacy by the United States, coupled with offers of reciprocal openness and continued financial assistance, would stand a good chance of overcoming the obstacles to taking the steps outlined in this chapter. Such a regime must be built in stages. Determining which steps should be pursued in which order is primarily a matter of negotiating tactics, a subject beyond the scope of this report. But urgency is in order. In the current envi- ronment of reasonably cooperative relations between Russia and the United States, a deliberate effort to understate the stocks and maintain substantial secret stockpiles appears unlikely. The more information exchanged while this remains the case, the more difficult it will be to create a secret stockpile in the future. Translating general good will into substantial understanding removes the seeds of suspicion and protects against the worsening of political relations. IMPLEMENTING A BROAD REGIME Fissile materials and nuclear weapons have a complex life cycle including mining, milling, processing, and enrichment of uranium; production of pluto- nium in special reactors; separation of the plutonium from the highly radioac- tive "targets" from those reactors; fabrication of fissile material weapons components; assembly of nuclear weapons from these and other components; deployment of nuclear weapons; retirement and disassembly of nuclear weap- ons; and storage and eventual disposition of fissile matenals.~3 The regime envisioned in this report would apply a variety of measures to different parts of this life cycle. The measures involved should be seen as mutually reinforcing, working together to build confidence that the information exchanged was accurate and that the goals of the effort were being met. i2 For a list of the Russian institutions the committee visited during a visit to Moscow ~ May 1993, see Appendix A. ~3 For a useful short description of this life cycle, see National Research Council, The Nuclear Weapons Complex: Managementfor Health, Safety, and the Environment (Washington, D.C.: National Academy Press, 1989), Appendix B. In both the United States and Russia, the actual deployment and operation of nuclear weapons is the only part of this process controlled by the military. The rest of the process is controlled by the department in charge of nuclear energy, the Department of Energy in the United States and the Ministry of Atomic Energy in Russia.
96 DECLARATIONS AND DISMANTLEMENT Although it is true that technical measures are not available to verify the total stockpiles of nuclear weapons and fissile materials with great accuracy, such a network of measures could build confidence over time much as a bank audit, which never counts all of the money in a bank's possession, builds confidence that the bank's records are basically accurate. Stockpile Declarations and Monitoring The fundamental basis for an overall regime would be a series of declara- tions by each party to the regime, specifying its holdings of nuclear weapons and fissile materials. Consideration would have to be given to how and in what sequence the various categories of weapons and fissile materials should be addressed. In addition, declarations would include locations of stockpiles, as well as descriptions of plutonium production and uranium enrichment plants, facilities for fabricating fissile material weapons components, and nuclear weapons assembly and disassembly facilities. In general, confirming that particular declared facilities held the items de- clared would be relatively straightforward. If it were considered too sensitive to provide full information on the locations of all inventories of weapons and materials at all sites, various sampling techniques might be used.~4 The key advantage of declaring all major sites is that any weapons or materials detected outside those sites would then be clear evidence of a secret stockpile. The more difficult problem will be assessing whether there are significant undeclared stocks at undeclared sites. This problem could be partly addressed through three primary approaches: 1. National intelligence already provides rough estimates of other nations' holdings of nuclear weapons and fissile materials, which could be checked for consistency with declarations- although, as noted, uncertainties in U.S. esti- mates of Russian stocks are currently large. Such national means of intelligence were the sole means of verifying arms agreements such as the SALT treaties, and remain an essential foundation for verification of more recent agreements incorporating on-site inspection. But because nuclear weapons and fissile ]4 For example, each side might tag an the weapons in its possession (a process known as "self-tagging") and provide the other with a list of the tag numbers; various sampling schemes under which one side could demand to see the weapons corresponding to particular tag numbers could then be envisioned, without revealing the locations of the entire stock of weapons. A conceptually similar approach might be implemented without the existence of physical tags: each side might provide the other with a table containing the locations and serial numbers of every weapon in its stockpile-but in encrypted form, so that the table could not be read. (Both sides already rely for their national security on the success of their encryption technologies for transmitting sensitive information.) The table could then be "de-encrypted" one line at a time for the purposes of inspection. For example, inspectors visiting a declared site might demand to see the line in the table representing a particular warhead at that site. A warhead that did not have such a line on the table would then be evidence of violation. For more on this concept, and other means of monitoring warhead and fissile material stockpiles, see S. Drell et al., "Verification of Dismantlement of Nuclear Warheads and Controls on Nuclear Materials," JASON, MITRE Corporation, ISR-92-331, January 1993.
DECLARATIONS AND DISMANTLEMENT 97 materials are much smaller than long-range missiles or bombers, and do not necessarily have comparable operational signatures, national technical means of intelligence will be less effective in monitoring them, and there will inevita- bly be less confidence in the accuracy of declared inventories. A number of techniques have been applied over the years: releases of krypton-85 from reprocessing plants, for example, have provided considerable information on production of separated plutonium. Power consumption at enrichment facilities, heat output from production reactors, and similar data have helped round out the picture of fissile material production. Intelligence has also provided some information on which to base estimates of nuclear weapons production and deployment. 2. In addition to providing baseline estimates against which declarations could be compared, national intelligence might detect stockpile activity outside declared sites, or other information that clearly contradicted the exchanged declarations a possibility that would help deter any party that contemplated maintaining either a secret stockpile or secret production facilities. 3. Exchanges of operating records of major production sites, followed by visits to those sites, could help confirm the information exchanged and reduce the uncertainties in unilateral intelligence. Certain characteristics of reactor buildings, waste from reprocessing, and tailings from enrichment plants can help determine how much material was produced and when, and these findings can be compared to the operating records for consistency. The latter techniques, sometimes known as "nuclear archaeology," are still being developed and cover a broad spectrum. is Physical and radiological examination of the interior of plutonium produc- tion reactors, for example, can provide information about both their design and the power levels at which the reactor has operated over its history. There are important uncertainties involved in this approach, however, including compli- cations introduced by replacement of reactor parts and changes in design over time. Examination of the reprocessing wastes where the plutonium was sepa- rated can also provide some information, though for programs as old, large, and diverse as those of the United States and the former Soviet Union, this information is likely to be limited. Enrichment facility operating records can be checked for consistency with the tailings of depleted uranium that they produce as waste: examination of the various isotopes in these tailings can indicate when the uranium was enriched, and whether it was enriched only to a few percent or to weapons-grade. is For a general description of these concepts, see, for example, Steve Fetter, "Nuclear Archaeology: Verifying Declarations of Fissile-Material Production," Science & Global Security, no. 3' 1992, pp. 237-259. These techniques are also extremely important in a nonproliferation context as the IAEA's current efforts to verify past production in North Korea and South Africa make clear. It might therefore be helpful for the two sides, in parallel with the confirmation effort, to undertake a joint research effort to refine these approaches further.
98 DECLARATIONS AND DISMANTLEMENT Unfortunately, weapons assembly and disassembly facilities are not sus- ceptible to such checks of the consistency of the physical status of the facility with operating records. Numbers of nuclear weapons can to a limited degree be checked for consistency with numbers of weapon systems with which they are associated, though the existence of spares, reserves, testing units, and reloads would complicate that approach considerably. As noted above, numbers at declared sites can be checked by routine inspection, and some provision for limited challenge inspections (comparable to those that have been worked out in other recent agreements) would also be useful. Ultimately, while no combination of intelligence and examination of books and facilities could ever prove that declarations were complete, it could go a long way in building confidence and any effort to hide a large stockpile that was not very carefully prepared, so that all the false information provided matched the physical state of existing facilities in a consistent way, would stand a substantial probability of detection. There is also the possibility of secret facilities producing fissile materials or weapons. In general, however, over the period of time necessary to build a strategically significant illegal stock, unilateral intelligence should be capable of detecting covert production on a scale large enough to be of military signifi- cance. Such secret facilities are of more concern in a nonproliferation context, in which the problem is to detect production of one or a few nuclear weapons. As noted earlier, if declarations were made in the current climate of coop- eration, the probability of Russia's deliberately hiding a significant part of its stockpile of weapons or materials seems low and once a good faith declara- tion was made, it would be more difficult for future governments to generate such a secret stockpile. Moreover, in the new environment in the former Soviet Union, with nascent democracies, strong political disagreements, and a newly open press, it is more difficult to keep such secrets than it once was. The possibility of "whistleblowers" would provide an additional deterrent to large- scale violation. The recent case of a Russian chemist revealing a continuing Russian chemical weapons development effort and then being arrested indicates both the potential and the limits of this source of arms control verifi- cation information. These problems of "initialization" that is, determining whether the declared initial inventory of items to be limited by arms control is correct- have also been addressed in the INF, START, and CFE treaties, as well as the Chemical Weapons Convention (CWC). In the end, despite some problems, the agreed data bases and accompanying confirmation measures in all of these regimes have been considered beneficial. The same should be true of a declara- tory regime for nuclear weapons and fissile materials. ~6 See, for example, Frank von Hippel, "Russian Whisdeblower Faces Jail," The Bulletin of the Atomic Scientists, March 1993; and Will Englund, "Ex-Soviet Scientist Says Gorbachev's Regime Created New Nerve Gas in '91,"Baltimore Sun, September 16, 1992.
DECLARATIONS AND DISMANTLEMENT 99 The committee concludes that U.S. knowledge of the stockpiles of weapons and fissile materials in Russia could be substantially improved through decla- rations and cooperative measures to help confirm them, for which a variety of potentially promising approaches are available. Such measures, however, could not eliminate potentially significant uncertainties regarding the possibility of covert stocks. Monitoring Dismantlement Dismantlement the fundamental means by which nuclear weapons are subtracted from the total stock is one of the three stages in managing excess plutonium resulting from arms reductions; it is treated in a separate section below. Monitored Storage of Excess Fissile Material Similarly, the commitment of excess fissile materials to peaceful use or disposal and the monitoring of storage and disposition to confirm that com- mitment the means by which fissile materials are subtracted from the stocks available for military use are fundamental to the subject of this report and are discussed in Chapter 5. Monitoring Assembly Keeping track of dismantlement is intended to provide an understanding of subtractions from the declared stockpile of nuclear weapons. In order to know the net subtraction from the stockpile, it would be necessary to monitor assem- bly as well, though assembly might be subject to somewhat different procedures because of its possibly greater sensitivity. This issue is addressed below, in the discussion of dismantlement. A Fissile Material Production Cutoff Similarly, if agreed transfers of fissile materials to civilian stocks are to be a useful arms control measure, it is important to ensure that new military fissile materials are not produced to replace them. Combined with the monitored transfer of large quantities of existing materials to peaceful purposes, cutting off production of fissile material for weapons would provide a demonstrable sign of progress in arms reduction, capping and reducing the total potential size of the nuclear arsenals that could be produced. The United States has already stopped producing all fissile materials for weapons and has recently proposed a global convention ending such production. Russia has stopped producing HEU but continues to produce weapons-grade plutonium, and has expressed willing- ness to agree to a formal production cutoff.
100 DECLARATIONS AND DISMANTLEMENT Ironically, at present Russia is continuing to produce weapons plutonium while requesting assistance to address the shortage of available space to store it (see Chapter 5~. The Russian government has indicated that the remaining three plutonium production reactors provide necessary heat and power to the areas surrounding them, and that their spent fuel must be reprocessed for logistical and safety reasons, but it has announced plans to end production of weapons plutonium by the year 2000. The U.S. government has begun discus- sions with the Russian government concerning possible assistance in convert- ing these reactors or providing alternate sources of power so that weapons- grade plutonium production can be cut off in the near term. Particularly as a cutoff of production plays a central part in the Clinton administration's September 1993 nonproliferation initiative, the committee believes it is essen- tial, for both substantive and symbolic reasons, that this continuing Russian production of weapons plutonium be ended expeditiously. The politics of other issues, such as the future of nuclear power and nuclear safety in Russia, should not be allowed to interfere with assistance in shutting down this production as soon as possible. Technical means are available to achieve this goal.~7 The committee is convinced that a cutoff of fissile material production could be monitored with relative ease by using a combination of national technical means of intelligence and inspections of fissile material facilities. Such facilities could be placed under IAEA safeguards comparable to those in place in non-nuclear-weapon states; this would allow a global cutoff agreement to be nondiscriminating. If the cutoff were limited to the United States and Russia, less intrusive transparency measures would probably suffice, since the goal would be to detect militarily significant production in states already possessing substantial stockpiles of nuclear weapons. i7 The Russian plutonium production reactors use aluminum~ad fuel, which the Russian government argues must be reprocessed because it cannot be stored safely. There are some questions about this argument: while some U.S. aluminum-clad fuel has been reprocessed for similar reasons (and problems have arisen with storage of some fuel), such fuel has been stored safely in water at the Massachusetts Institute of Technology, for example, for two decades. Even if this argument is accepted, however, two main options are available for cutting off plutonium production. The first is converting the reactors to use fuels that would not require reprocessing. The comparable U.S. N-reactor production facility at the Hanford reservation, for example, used zirconium-clad fuel similar to that used in commercial reactors, which can be stored safely, and can be used in die reactor for several times as long, producing much less spent fuel that requires storage. The second is shutting the reactors and providing alternative sources of power. This would require either new transmission capacity to carry power from elsewhere or the construction of new power plants (and possibly new gas pipelines). This latter option might be more expensive and time-consuming than the former, but would eventually have to be pursued in any case. In 1992, in a letter cosigned by representatives of MINATOM and the Kurchatov Institute, He Russian government formally requested assistance from the United States in converting these reactors. After some internal discussion, the United States agreed to send a team to discuss the practicality of converting the reactors. Delays have been encountered on both sides, but as of late 1993, the pace of efforts in this regard appeared to be . . Increasing.
DECLARATIONS AND DISMANTLEMENT 101 Adding It Up As noted, such measures would work together synergistically. To undertake a militarily significant "breakout," a potential violator would have to deliber- ately leave large quantities of weapons or materials out of the initial declara- tions (or successfully produce both later in secret plants without detection); successfully falsify decades of operating records in a way consistent with the state of all the existing facilities; provide delivery vehicles to launch the weapons, in the context of the overlapping START verification regime; and so on. Each of these hurdles, while not insurmountable in itself, provides an additional risk of detection. The combination of measures would make the possibility of successful evasion acceptably remote. INTERNATIONALIZING THE REGIME Most if not all of the regime described above can and should be extended worldwide. The standards set in managing U.S. and Russian excess weapons and fissile materials can provide the base for improving management of these items throughout the world, and the opportunity to do so should be taken. As the Clinton administration's September 27, 1993, statement on nonproliferation policy put it, world stocks of fissile materials should be "subject to the highest standards of safety, security, and international accountability." Declarations of weapons holdings should be made by all the declared nuclear-weapon states, while declarations of fissile material holdings should ultimately include all states.~9 Such universal reporting of stocks of fissile material, which should include information on all imports and exports of fissile materials, would complement the information that the non-nuclear-weapon parties to the NET are already required to give to the IAEA, providing a sub- stantially firmer base for planning international fissile material management policy, which will remain an essential aspect of nonproliferation. Similarly, as additional states come to participate in nuclear arms reduc- tions, arrangements comparable to those described in this chapter for monitor- ing subtractions from their stockpiles and committing excess fissile materials to non-weapons use or disposal should be put in place. Making a cutoff of production of fissile materials for weapons a global accord, as recently proposed, rather than solely a U.S.-Russian pact, would have particular significance, marking a major step forward in nonproliferation efforts. A global cutoff would establish the fundamental principle that it was no 8 White House Fact Sheet, "Nonproliferation and Export Control Policy," September 27, 1993. ~9 At present it is probably not realistic to expect states that have not formally declared their nuclear weapons capability, such as Israel, India, and Pakistan, to declare the number of nuclear weapons available to them; hence declarations of nuclear weapons holdings would apply only to acknowledged nuclear-weapon states. These "threshold" counmes are also likely to be reluctant to declare their holdings of fissile materials, but He regime should encourage them to do so.
102 DECLARATIONS AND DISMANTLEMENT longer legitimate for any state to produce the essential ingredients of nuclear weapons, except for peaceful purposes under safeguards. If states such as Israel, Pakistan, and India could be convinced to accept such an agreement, it would cap their undeclared arsenals without requiring them to either acknowledge or roll back those arsenals immediately. Such a first step would go a long way toward limiting the potential for a nuclear arms race on the South Asian subcontinent. At the same time, the stringent standards of security and accounting that should be set for storage and processing of excess fissile materials from weapons (see Chapter 5) should be extended to all civilian weapons-usable fissile materials worldwide. Such a step would significantly reduce the risks of diversion or theft of nuclear materials from civilian fuel cycles. The IAEA secretariat and organizations in several countries are now working on concepts for such universal reporting and safeguarding of fissile matenals. MANAGING AND MONITORING DISMANTLEMENT Current Practices Dismantlement in the United States In the United States, nuclear weapons are being dismantled at the Pantex plant in Amarillo, Texas, at a rate that has varied over the last several years, reaching 1,600 warheads in 1991 (see Table 4-1~.2° DOE is striving to increase this rate to roughly 2,000 per year. The United States plans to dismantle a large fraction of both its tactical and its strategic arsenals, though decisions on the number of weapons to be retained as inactive reserves remain to be made-up The U.S. dismantlement rate is limited by the size of the available infra- structure and by a set of practical considerations, most of them related to the need to maintain applicable standards of protection for environment, safety, 20 One type of weapon, the W-33, which did not include plutonium components, was dismantled at the Y-12 plant in Oak Ridge, Tennessee, rather than at Pantex. See U.S. Congress, General Accounting Office, Nuclear Weapons: Safety, Technical, and Manpower Issues Slow DOE's Disassembly Efforts, GAO/RCED-94-9 (Washington, D.C.: U.S. Government Printing Office, October 1993). 2} Recent Defense Department statements suggest that the inactive reserve-nuclear weapons that remain assembled, but are not among the 5,100 weapons slated to be in the future active U.S. nuclear force-may take on greater significance than it has had in the past. Undersecretary of Defense John Deutch, for example, recently told Congress that because problems with the weapons complex and the end of nuclear testing would leave the U.S. ability to produce new warheads "severely constrained," some warheads would be kept in the inactive reserve "to replace active weapons if necessary." Deutch argued Mat the inactive reserve "holds the Nation's only capacity for augmenting our significantly reduced active nuclear forces in response to a reversal in current geopolitical trends or the emergence of a new strategic threat." (See U.S. Congress, House Appropriations Subcommittee on Energy and Water Development, Energy and Water Development Appropriations for 1994, Part 6, p. 1311.) The appropriate size and operational posture of the U.S. nuclear arsenal is being reexamined in new studies by die Defense Deparanent and tile National Security Council.
DECLARA TI ONS AND DISMANTLEMENT 103 TABLE 4-1 Warhead Dismantlement in the United States Fiscal Year Numbers Retired and Disassembled 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 535 1,416 1,360 960 860 927 574 1,068 510 1,134 1,056 1,546 1,274 SOURCE: U.S. Department of Energy, cited in U.S. Congress, Office of Technology Assessment, Dismantling the Bomb and Managing the Nuclear Materials, OTA-0-572 (Washington, D.C.: U.S. Government Printing Office, September 1993), p. 24. and health. Dismantlement is conducted under carefully designed, preapproved, step-by-step procedures, which are time-consuming. The existing facilities and personnel are working close to capacity. Thus, significantly speeding the pace of dismantlement would require either hiring and training a substantial number of extra personnel, in order to add an additional shift at existing facilities, or building new facilities. Even hiring and preparing workers for an additional shift would take several years because there are extensive screening and train- ing processes for personnel who are to handle nuclear weapons. Since, at the currently scheduled rate, planned U.S. dismantlements would be largely com- plete by the year 2000 in any case, such steps would not drastically shorten the remaining time to completion. The weapons components resulting from dismantlement are either stored, destroyed, disposed of as waste, or processed to recover valuable materials. The plutonium weapons components, known as pits, are currently being placed in intermediate storage at Pantex, while the highly enriched uranium components are being shipped to the Y-12 plant at Oak Ridge, Tennessee, where they can be stored or processed for use as nuclear fuel in naval or civilian reactors. High- explosive components are being burned in the open at Pantex, but environ
104 DECLARATIONS AND DISMANTLEMENT mental and public acceptance considerations may malce it difficult to continue that practice.22 Dismantlement in Russia In Russia there are four sites where weapons are assembled and can also be disassembled: Arzamas, Penza, Zlatoust, and Nizhnaya Tura. Information about Russian dismantlement rates is uncertain. Russian officials responsible for these programs have indicated that their dismantlement rate is somewhat greater than that of the United States. In public testimony, the Department of Defense has estimated that the current dismantlement rate in Russia is approximately 2,000 per year, comparable to the U.S. rate.23 In both official and private discussions, Russian officials have indicated that rates as much as twice or even three times that of the United States could be attained (for exam- ple, the Central Intelligence Agency reports that the Russians have indicated a capability to dismantle 4,000-5,000 weapons per year, which the agency says it has no reason to doubt).24 Why Russian dismantlement is not currently proceeding at the maximum attainable rate is uncertain. Russian spokesmen have claimed that dismantle- ment rates are severely limited by available storage capacity for the fissile components of nuclear weapons. However, making existing storage sites available, possibly with modifications, in both the nuclear weapons complex and the military complex would provide adequate space (see Chapter 5~. The economic and budgetary turmoil in Russia appears to be one source of significant problems for dismantlement. Workers at some key nuclear sites, including those involved in dismantlement, have gone unpaid for months at a time and have threatened strikes. To the extent possible, the U.S. government should attempt to be helpful in ensuring that sufficient resources are available to accomplish critical tasks such as dismantlement. The planned HEU deal should be a step toward that objective, and additional options for providing financial or other incentives for dismantlement should be pursued. Unlike the United States, Russia has assumed some formal obligations to dismantle nuclear weapons, which might be seen as seeds from which the broader transparency regime might grow. Commitments to dismantle weapons removed from certain states of the former Soviet Union are contained in some 22 For a more detailed account of current U.S. dismantlement practices, see U.S. Congress, Office of Technology Assessment, Dismantling the Bomb and Managing the Nuclear Materials, OTA-0-572 (Washington, D.C.: U.S. Government Printing Office, September 1993). 23 Ashton B. Carter, Assistant Secretary for National Security and Counterproliferation, Department of Defense, testifying at a hearing of the House Foreign Affairs Committee, "U.S. Aid to the Republics of the Former Soviet Union," September 21, 1993 (transcript, Federal News Service). In Russia, unlike the United States, limited production of new nuclear weapons is also believed to continue. 24 Testimony of Lawrence Gershwin, National Intelligence Officer for Strategic Programs, Senate Government Whirs Committee, February 24, 1993 (transcript, Federal News Service).
DECLARATIONS AND DISMANTLEMENT 105 of the accords reached in the framework of the Commonwealth of Independent States, in the negotiated letters accompanying the Lisbon Protocol, and in an agreement between Russia and Ukraine reached in April 1992, which contains detailed provisions for Ukrainian monitoring of the dismantlement of weapons removed from Ukrainian territory. Russian officials report that the lader agreement is currently being implemented and that, as of the spring of 1993, half of the tactical nuclear weapons removed from Ukrainian territory had been dismantled, with Ukrainian monitoring.25 In addition, as currently conceived, the arrangements for the HEU purchase now in the final stages of negotiation and for U.S. funding of a fissile material storage site in Russia will both specify that the material involved must come from dismantled weapons (though meas- ures to verify this will be limited or nonexistent). This creates an obligation to dismantle enough nuclear weapons during the 20-year period of the agreement to provide 500 metric tons of HEU. Monitoring Dismantlement With the exception of the monitoring called for under the Russian- Ukrainian agreement (and limited openness to the public at Pantex), no measures are in place that are specifically designed to increase the transparency of the dismantlement process. Such measures would increase the confidence of the parties to the current reductions accords, as well as the international com- munity, that dismantlement is in fact taking place and that the denuclearization process is being securely managed.26 Increased transparency for weapons dismantlement has thus far been resisted within the U.S. government and some sectors of the Russian government, for three reasons: (1) the need to protect sensitive weapons design information, (2) the urgency of proceeding with dismantlement, and (3) the costs of monitoring. These objections have some merit; yet the process of introducing increased transparency measures need not significantly slow down the process of disman- tlement, unduly compromise sensitive information, or break the bank. More- over, as described above, there are compelling motivations for increasing transparency. Although the uncertainties concerning dismantlement rates are greater for Russia than for the United States, monitoring of dismantlement 25 Interview win General Sergei Zelentsov (retired) and Colonel-General Vitali Yakovlev, former commander and current deputy commander, respectively, of the Russian military's 12~ Main Directorate, in charge of nuclear weapons, Moscow, May 1993. Ukraine has insisted that all weapons removed from its territory be dismantled, and Kazakhstan appears to take a similar position. Some weapons being withdrawn from Belays, however, in particular the modern SS-25 mobile intercontinental ballistic missiles, may well be incorporated into Russian strategic forces. Some Ukrainian officials continue to claim Mat the monitoring provisions of the April 1992 agreement have not been implemented, though others indicate that they are fully satisfied with these verification arrangements. 26 As part of Me broad regime outlined here, such increased transparency could begin before dismantlement as well, including, for example, monitored storage of excess weapons awaiting dismantlement.
106 DECLARATIONS AND DISMANTLEMENT should be reciprocal, both because of the benefits of a reciprocal regime and because reciprocity would Neatly improve the Dolitical acceptability of moni torlng measures. . cat ~ The best available means to monitor dismantlement without significantly compromising sensitive design information would be a variant of the perimeter- portal monitoring (PPM) system now in place to verify that missiles banned by the INF treaty are not being produced.27 Under such an arrangement the disassembly facility would be securely fenced, with the exception of monitored entry and exit points. At the entry point, technical equipment could be used to verify that an entering object is a nuclear weapon. A variety of technical means to do so exist that could be used in a mutually supportive manner. The leading technique is x-ray radiography, which could be constrained (to the satisfaction of both the inspecting and the inspected parties) to ensure that the resolution of the image provided was good enough to verify that the entering object was a weapon, but not good enough to reveal the most sensitive design details. Addi- tional methods include passive detectors to observe the radiation emitted by nuclear weapons and active detectors to observe the radiation emitted in response to interrogation by a particle beam, among others.28 At the exit point of the facility, the material going out could be assayed for fissile material content (by methods external to the canisters containing the fissile components, to avoid inspection of the detailed dimensions of the components, which itself is classified information). Although the committee is persuaded that monitored dismantlement using such PPM methods can be made without compromising vital design information, it will be necessary to declassify some limited infor- mation that is now considered restricted data, such as radiation spectra from weapons. Currently, assembly and disassembly of nuclear weapons take place in the same group of structures at Pantex. Similarly, it is the committee's understand- ing, based on discussions with Russian officials, that in Russia, each specific weapon type is assembled and disassembled in the same facility, although within those facilities, assembly and disassembly are segregated. In principle, perimeter-portal monitoring could simply be imposed on such joint assembly and disassembly facilities, thereby monitoring both dismantlement and assem- bly at the same time. Both incoming and outgoing weapons could be counted, with the difference being credited as disassembled weapons; the fissile material content of the weapons components leaving the PPM enclosure and going into safeguarded storage could be assayed after exit; and nonfissile components could be brought into and out of the facility in opaque containers, with the 27 Under the INF treaty, the United States has a PPM installation at the Russian Votkinsk missile production facility, to ensure dlat prohibited SS-20 INF missiles are not produced there. The Russians have similar monitoring opportunities at a U.S. missile facility at Magna, Utah. 28 For more details on pow monitors for identifying nuclear weapons, see DreH et al., op. cit.; and David Albright, "Portal Monitoring for Detecting Fissile Materials and Chemical Explosives," in Frank von Hippel and Roald Z. Sagdeev, eds., Reversing the Arms Race (New York: Gordon and Breach Science Publishers, 1990).
DECLARATIONS AND DISMANTLEMENT 107 monitors learning nothing about their contents except the basic fact that they did not contain substantial quantities of fissile material. Alternatively, assembly and disassembly of nuclear weapons could be seg- regated (as they already are in Russia) in order to exclude assembly operations from monitoring or to impose different types of monitoring and levels of intrusiveness on the two operations. At the Pantex plant, some of the essentially identical structures used for assembly and disassembly could be devoted exclu- sively to dismantlement, and others exclusively to assembly. This would not require significant modifications of existing facilities, although it would probably come at some modest cost in operational efficiency. If, as it appears, segregating assembly from disassembly would not impose substantial costs or delays, this would probably be the preferable approach, to ease the task of designing monitoring arrangements most appropriate to the degree of sensitivity of the activity being monitored. Information concerning the design of weapons types to be retained in active service, for example, may be more sensitive than the design of weapons being retired. The problem of protecting sensitive information related to the nonfissile components flowing into the assembly operation would be reduced. Similarly, if the sides agreed to tag particular weapons to be dismantled under an arms agreement, it would be considerably easier to determine that these specific weapons had been disman- tled if intact weapons were not leaving the same facility. As noted above, however, if a regime is to be built that monitors the net subtraction of nuclear weapons from each side's arsenal, both assembly and disassembly will have to be subject to some form of transparency. The specifics of how the monitoring of dismantlement should be implemented are matters that must be subject to further internal consideration by each party and to bilateral negotiation. MANAGING DISMANTLEMENT FOR ENVIRONMENT, SAFETY, AND HEALTH Protection of the environment, safety, and health must be a critical part of the dismantlement effort. The most obvious and compelling safety issue is ensuring against the possibility of a nuclear explosion. Addressing this problem requires great care, including disabling warheads prior to disassembly, to prevent a nuclear yield. A possibility of conventional explosions that might cause plutonium contamination remains, however. The "Gravel Gerties" used for dismantlement are designed to contain such explosions, limiting damage and contamination to the interior of the particular dismantlement module itself. Nevertheless, precautions must be taken to ensure against such explosions; none has occurred in the decades of operation at Pantex. Other ES&H issues involved in dismantlement include worker exposures to radioactive and toxic materials; transport of hazardous materials to and from the facility; disposal of hazardous materials on-site (including open burning of explosives from disassembled weapons at Pantex, which has been the focus of
108 DECLARATIONS AND DISMANTLEMENT particular public concern); criticality safety; and possible safety issues asso- ciated with storage of weapons and fissile materials (such as the possibility of an aircraft crash on the storage facility). All these issues are being addressed, to varying degrees, but dismantlement will never be a risk-free endeavor. The ES&H dangers involved in dismantlement, however, are far less severe than those of many other U.S. (and Russian) nuclear weapons complex activities, particularly since there is no actual processing of plutonium or other radioac- tive materials. Considering the methods used for dismantlement in the United States, it is the committee's judgment that there is little doubt that dismantling weapons and storing or disposing of the resulting materials is safer overall than storing the assembled weapons indefinitely. Public support for weapons complex operations, however, can be secured only by providing greater openness and public participation in decision making. In the new environment in which DOE finds itself, such participation is re- quired if dismantlement is to continue at projected rates. DOE is making progress in setting up mechanisms to meet these needs. Nevertheless, public involvement is currently embryonic and in need of further developments RECOMMENDATIONS The committee has deliberately included consideration of both dismantle ment and declarations in a single chapter, since both are critical to the creation of a meaningful future control regime encompassing all nuclear weapons and weapons-usable fissile materials. The committee recommends that: · The United States and Russia should make formal commitments that specific quantities of fissile material from dismantled weapons (representing a very large fraction of those materials) will be declared excess and committed to non-weapons use or disposal. Storage and disposition of these materials should be subject to agreed standards of accountability, transparency, and security. The standards for accountability and security should approximate as closely as possible the stringent standards applied to stored nuclear weapons. · The United States should negotiate with Russia to create, through a step- by-step process, a broad regime under which each side's stocks of nuclear weapons and fissile materials would be declared and monitored, and the size of both stocks would be verifiably reduced over time in line with current reduc- tions in deployed delivery systems. This regime would include, in addition to the fissile material steps mentioned in the previous recommendation: 1. a system of mutual declarations of total inventories of nuclear weapons and of fissile materials in civilian and military inventories; 29 See U.S. Congress, Office of Technology Assessment, op. at.
DECLARATIONS AND DISMANTLEMENT 109 2. measures designed to increase confidence in the accuracy of the declarations, and the transparency of each side's nuclear weapons production complexes, including physical access to production facilities and production records for fissile materials; 3. a monitored cutoff of production of HEU and plutonium for weapons. If necessary, the United States should be willing to provide limited funding to assist Russia in the measures necessary to cut off plutonium production; and 4. an agreement providing for perimeter-portal monitoring of dismantlement facilities, counting warheads entering these facilities and assaying the fissile material that leaves. If the net subtractions from each side's stockpile are to be confirmed, some monitoring of warhead assembly will be required as well. · Information concerning the total stockpiles of weapons and fissile materials, and those weapons characteristics necessary for external monitoring, should be declassified as part of this transparency regime. Appropriate reviews to prepare for such declassification should be initiated promptly. · Russia and the United States should dismantle their retired warheads as expeditiously as is practical, consistent with protection for the environment, safety, and health, and cost-effectiveness.