Return to List of Contents

Comprehensive Test Ban Treaty

Acronym Report No.9

The first two years, 1994-1995

With its mandate to negotiate a multilateral comprehensive test ban treaty supported unanimously by 156 states in the United Nations General Assembly in December 1993, the NTB Committee of the CD was convened in January 1994, with Mexican Ambassador Miguel MarŹn Bosch as its Chair. MarŹn Bosch attempted to pilot the talks to conclusion before the NPT Review and Extension Conference in April-May 1995, in accordance with policy statements from the G-21 Group of Non-Aligned States, but was thwarted by the nuclear weapon states.2 In particular, France and China had domestic reasons for preventing any possibility of conclusion before 1996: China wished to continue nuclear testing and France wanted to keep the option open pending election of a new President in May 1995.

The first year's negotiations therefore prepared the ground by examining the range of technical and political questions pertinent to a comprehensive test ban. By the close of the CD in September 1994, it had produced a 95-page draft text in the loose form of a treaty.3 It was peppered with over a thousand pairs of brackets enclosing concepts, words and punctuation that had been proposed but were not yet agreed. This document, attached to the 1994 NTB Committee Report, is referred to as the 'rolling text' and has become the basis for the negotiations.

In accordance with the CD principle of rotation among the three groups4, the Ambassador of Poland, Ludwik Dembinski, took the Chair of the NTB Committee in 1995. Negotiations continued apace, in the shadow of the NPT Conference and its surrounding politics.

During 1994 and much of 1995, the nuclear weapon states had sought to protect their nuclear weapon programmes by means of exemptions in the treaty. France and the UK had wanted to keep a provision for safety tests 'in exceptional circumstances'. The Clinton Administration, which had decided against a 1 kt threshold in 1993, had defined 'zero' as anything up to 1.8 kg (4 lb.), the yield permissible for a 'one point safety test'. Since the US Department of Energy had been prepared only to provide a ten year assurance of the reliability of the US arsenals, the United States also proposed a ten year opt out or 'easy exit' clause. There was overwhelming opposition to such a provision, which would have undermined the indefinite and universal character of the treaty. This caused the US to withdraw it at the beginning of 19955. After dropping their proposal for safety tests in April 1995, the UK and France joined the threshold debate among the P-56.

There was general dissatisfaction with the US threshold of 1.8 kg, as it was believed that US technology was sophisticated enough to gain considerable design information at this level through hydronuclear experiments (HNE). Russia wanted up to 10 tons and the UK would accept 40-50 kg, while France reportedly favoured a threshold of 100-300 tons. Officially China opposed any threshold whatsoever, and had submitted language for the treaty scope that would prohibit 'any nuclear weapon test which releases nuclear energy'. However, it was reported that if there was to be a threshold, China would prefer one much higher, in the hundreds of tons.

China's strategy on scope was altogether different. Making a distinction (which many do not accept) between nuclear explosions for military and civilian purposes, China proposed that PNEs should not be prohibited, but rather should be subject to rigorous authorisation and verification procedures. The non-nuclear weapon states had made it clear that they wanted nuclear explosions to be comprehensively banned, with no thresholds or exceptions. However, during the early stage of negotiations, some had indicated that they were prepared to turn a pragmatic blind eye to hydronuclear testing at extremely low levels, providing that all the P-5 could come to an agreement.

On May 15, less than three days after the NPT was indefinitely extended in New York, China conducted a further nuclear explosion. On June 13, the newly elected President Chirac announced the resumption of French testing. The prospect of France breaking its moratorium, imposed in 1992, shook the negotiations and caused an international outcry. Though the upsurge of public opposition targeted French testing in the Pacific in particular, it also conveyed a strong sentiment against all nuclear testing and for nuclear disarmament in general. Also in the aftermath of the NPT Conference, the Pentagon had re-opened the threshold debate in Washington, urging testing up to 500 tons, which the Department of Energy and the Arms Control and Disarmament Agency (ACDA) opposed. Forced on the defensive, France was the first to announce that it was abandoning its push for a threshold ban. Declaring to the CD on August 10 that it envisaged a 'truly comprehensive prohibition' France endorsed the Australian scope language prohibiting 'any nuclear weapon test explosion or any other nuclear explosion.'7

A day later, President Clinton committed the US to 'a true zero yield ban' on all nuclear explosions.8 With this announcement, Clinton resolved the heated threshold debate amongst the government agencies. With the P-5 locked in disagreement over permitted levels, and China adamant that it would not accept the US definition of 'zero = 4 lb', the choice was between going to true zero, as the non-nuclear countries wanted, or taking the level much higher, as desired by the other nuclear weapon states. The problem was: how high? The JASON report, commissioned by the Department of Energy from 14 nuclear and security experts, concluded that sub-kiloton tests would be of marginal utility in ensuring stockpile safety. Combined with the public and international outrage against the French resumption of testing, this convinced Clinton that the wiser choice would be a real zero. Whether the French announcement had meant complete zero or only a dropping of tens or hundreds of tons to a hydronuclear limit of a few kilograms, it would have looked foolish if they had backed down, and so France confirmed the true zero on August 16. The UK followed on September 14, adopting both the zero yield concept and the US and French interpretation relating 'supreme national interests' to stockpile safety and reliability. Although Yeltsin stood beside Clinton when the US President announced on October 23 that both supported a fully comprehensive, true zero yield test ban, Russia has not yet on its own behalf endorsed the concept or the front-runner Australian scope.9

By the end of the CD's 1995 session, Dembinski was able to present an updated rolling text, which covered 97 pages, with over 1200 pairs of brackets around disputed language. The major political breakthroughs were on scope, verification and duration of the treaty. The US dropped its ten year opt out clause first. Then the UK and France withdrew the special provision for safety tests, and France, the US and UK abandoned thresholds and committed to a true zero ban. However, Russia was playing hard to get, and China had not moved on PNEs or general scope language. Substantial agreement was obtained on the international monitoring system, including the location of seismic, radionuclide and hydroacoustic stations. The rolling text had been cleaned of many redundancies and contradictions, but little further progress had been made on the hard political choices concerning on-site inspections, the implementing organisation (particularly the composition of the Executive Council), and entry into force. Little had been done on the preamble, and no further decisions had been taken on China's desire to include articles covering security assurances, no-first-use, the peaceful uses of nuclear energy, and the relationship of this treaty with other agreements (principally intended to establish the le

Structure of Negotiations

Despite an attempt by India to link the NTB Committee, which has to be re-established each year, with the fate of other committees which had been paralysed during the whole of 1995, the test ban negotiations started on time. The NTB Committee was convened on January 23, under its new Chair, Jaap Ramaker of The Netherlands. It re-adopted the mandate which had been negotiated in 1993 by Yoshitomo Tanaka, then Ambassador of Japan.


Under Agenda Item 1 (CD/1238)

"Nuclear Test Ban"

(Adopted at the 666th Plenary meeting of the CD on 25 January 1994)

In the exercise of its responsibilities as the sole multilateral disarmament negotiating forum of the international community, the Conference on Disarmament decides to re-establish an Ad Hoc Committee under item 1 of its agenda entitled "Nuclear Test Ban", and to give priority to its work.

The Conference directs the Ad Hoc Committee to negotiate intensively a universal and multilaterally and effectively verifiable comprehensive nuclear test ban treaty, which would contribute effectively to the prevention of the proliferation of nuclear weapons in all its aspects, to the process of nuclear disarmament and therefore to the enhancement of international peace and security.

Pursuant to its mandate, the Ad Hoc Committee will take into account all existing proposals and future initiatives, as well as the work of the Ad Hoc Group of Scientific Experts to Consider International Co-operative Measures to Detect and Identify Seismic Events. The Conference requests the Ad Hoc Committee to establish the necessary working groups in order to carry forward effectively this negotiating mandate; these should include at least two working groups, one on verification and one on legal and institutional issues, which should be established in the initial stage of the negotiation, and any others which the Committee may subsequently decide upon.

The Ad Hoc Committee will report to the Conference on Disarmament on the progress of its work before the conclusion of the 1994 session.

In accordance with this mandate, two working groups were convened: Working Group 1 on Verification, chaired by Grigori Berdennikov of the Russian Federation, and Working Group 2 on Legal and Institutional Issues, chaired by Mounir Zahran of Egypt. Numerous Friends of the Chair were appointed or re-appointed to continue consultations on some of the more complicated or contentious issues.

Who's who in 1996

Nuclear Test Ban Committee
Chair: Ambassador Jaap Ramaker (Netherlands)

Working Group 1 on Verification
Chair:Ambassador Grigori Berdennikov (Russia)

Technical Verification: Peter Marshall (UK)
On-Site Inspections: Ambassador Mark Moher (Canada)
IMS: Patrick Cole (Australia)
DC: Ralph Alewine (USA)
Associated Measures: Richard Ekwall (Sweden)

Working Group 2 on Legal and Institutional Issues
Chair: Ambassador Mounir Zahran (Egypt)

Executive Council: Ambassador Nacer Benjelloun-Touimi (Morocco)
Entry into Force: Ambassador Antonio de Icaza (Mexico)
Funding: Yukiya Amano (Japan)
Preamble: Marshall Brown (USA)
Preparatory Commission: Don Sinclair (Canada)

The Chair's Working Paper

On the penultimate day of the first part of the 1996 session, Ramaker presented his working paper 'Outline of a draft Comprehensive Nuclear Test Ban Treaty'.10 This is structured as a draft treaty, with a preamble and 17 articles, but has not been cleaned of brackets. A few weeks earlier, China and India had explicitly warned against any attempt to ride roughshod over their concerns and pre-empt decision-making on the options contained in the rolling text. Caught between this and pressure from the US and others to produce a clean draft text by the end of March, Ramaker appears to have adopted a two-stage strategy. The outline working paper is intended to provide 'delegations and their capitals with a structured and manageable product of these negotiations so far, and highlights the major outstanding issues on which political decision-making is required.' It provides both outline and guide to the likely form and content of a final treaty, but without removing any states' cherished proposals from view.

The working paper incorporates articles which are already agreed in the rolling text, such as the standard treaty provisions:

Ramaker highlights six outstanding issues: the preamble; scope; the composition of the Executive Council; some of the functions of the international data centre (IDC), particularly the level of information and analysis it should provide to states parties; on-site inspections; and entry into force.

Where states' individual proposals are hard fought, such as scope, Ramaker has presented the heavily-bracketed rolling text, together with an indication of a clean formulation that has attracted wide support. In other cases, such as the composition of the Executive Council, the working paper provides text developed by a Friend of the Chair after consultations with the delegations. This has not necessarily been bracketed, although it is understood that final agreement may not yet have been achieved.

Four bracketed articles appear at the end, covering China's proposals on the peaceful use of nuclear energy, PNEs, security assurances and the relation of the treaty to other international agreements. Their inclusion in this way indicates the general view that there is no place for them in the treaty, while at the same time meeting China's insistence that its decision on these issues cannot be pre-empted by another representative - even the Chair. In some cases, such as PNEs, China's desire for them is opposed altogether; other articles, such as security assurances, are supported in principle by many, but not within the context of the CTBT.

Attached to the head treaty is a three part protocol covering the IMS, inspections, and associated confidence-building/transparency measures. There are also annexes listing states and detailing the stations chosen for the IMS.

If Ramaker's strategy is to work, the delegations and governments will have to see in his working paper the form of a CTBT that they would be prepared to accept. By indicating preferences, along with the bracketed alternatives that states have not yet been willing to withdraw from the rolling text, he hopes to encourage the relevant political decisions to be taken. This would enable the rolling text to be considerably slimmed down over the next few weeks, which he could then reflect in an unbracketed 'Chair's text' in June. If governments do not take the opportunity to compromise on positions that have commanded little or no support, it will make Ramaker's task of finalising the treaty more confrontational.

Iranian and Australian model

To some extent, the groundwork was laid for a Chair's text in February, when Iran11 and Australia12 each tabled a vision of the possible treaty. Although these differed in the particular solutions they proposed, most notably on scope, there was a conceptual similarity in how the models approached some of the most difficult issues such as entry into force and the composition of the Executive Council.

Presenting Iran's draft text on February 22, Foreign Minister Ali Akbar Velayati stressed that the purpose was to help the CD 'to perceive a middle ground - a packageÓwhich may constitute a compromise amongst the various and, at times, contradictory positions.'13 Michael Costello, the Australian Secretary for Foreign Affairs and Trade said much the same, denying any attempt 'to establish an alternative or parallel negotiating process'. The purpose was to 'demonstrate tangibly that a CTBTÓis indeed within reach.'14

Reactions to the two drafts was mixed, even amongst those states which publicly welcomed them. China, India, Pakistan and even Russia emphasised that they were merely national views on what a text might look like, and should not be allowed to pre-empt the rolling text. India and Pakistan reportedly criticised Iran within the G-21 for giving legitimacy to the Australian draft, which they had been preparing to reject out of hand. However, it is also understood that Iran had consulted several non-aligned delegations several weeks before presenting its draft. Iran had apparently sought co-sponsorship. Though there were no takers, Iran's draft was warmly welcomed by the majority of non-aligned countries, despite disagreement with some of its solutions. Australia reportedly did not consult with other delegations, preferring to go it alone. This may have contributed to the ambivalence with which Western colleagues greeted the Australian initiative. Some feared that the drafts would complicate the solutions that the Chairs and their Friends were trying to hammer out. However, in demonstrating 'the extent to which there is already widespread agreement'15 the Iranian and Australian drafts were received positively. Most importantly they conveyed a sense of urgency, underscoring the need for delegations to let go of 'their' language or proposals and seek compromise for the sake of concluding a CTBT in 1996.


The framework and purpose of the treaty are now being addressed fully. Initial assessment of the Chair's working paper indicates support for his outline framework. While not deleting the proposals on peaceful uses, PNEs, security assurances and other treaties, Ramaker has reflected the majority view that these should not be in the final treaty, providing a strong signal to China that it is now time to move on.

The preamble, which was hastily cobbled together at the end of 1994 and barely considered in 1995, has now become one of the most important sites for negotiating the meaning, spirit and context of the treaty.

Although China withdrew some of its controversial preambular language referring to 'the complete prohibition and thorough destruction of nuclear weapons' in September 1994, it maintained its paragraph on security assurances and no first use. These are opposed by the other nuclear weapon states and regarded by many others as not sufficiently or directly relevant to the test ban treaty's main purpose to be worth fighting for in the preamble. In January, India tabled a working paper on the preamble, with five proposals for new or additional language. One made general reference to the relationship between the cessation of testing and nuclear disarmament, while three proposals referred to the total elimination of nuclear weapons within a 'time-bound framework', 'agreed time-frame', and 'time-bound process'. India also proposed an additional paragraph: 'Emphasising that the principal objective of this Treaty is to end the qualitative improvement and development of nuclear weapon systemsÓ'16

Having made these proposals for the preamble, India also proposed additional language for the articles on review to ensure that the preamble is not neglected. To the paragraph referring to a conference after ten years to 'review the operation of the Treaty',17 India wants added: 'and with a view to ensuring that the objectives, purpose and the provisions of the Preamble to the Treaty are being realised'.18 The US reportedly objects that reviewing the operation of the treaty already includes the preamble, while other states may be more sympathetic to this attempt by India to underscore the importance of the nuclear disarmament. India also linked its text proposals with the political demand (which is non-aligned policy) for an ad hoc committee on nuclear disarmament to be established by the CD. India's strategy appears to be less about the CTBT per se, so much as a desire to establish the legitimacy of nuclear disarmament negotiations in a multilateral forum other than the NPT.

The Iranian draft treaty took most of India's proposals forward, including the verbatim paragraph on qualitative development and another for 'the speediest possible achievement of an agreement for total elimination of all nuclear weapons within a time-bound framework'.19 With at least the three western nuclear weapon states adamantly opposed to any mention of time-bound framework, Australia judged it prudent to leave the term out. However, conscious of the importance to non-aligned states of the nuclear disarmament context, Australia's model treaty says that a CTBT 'will constitute a meaningful step towards the realisation of a systematic process to achieve nuclear disarmament.'

Ramaker's draft did not seek to offer an alternative to the bracketed rolling text on the preamble. Marshall Brown, the Friend of the Chair for the Preamble, is still consulting with delegations and seeking a compromise. Having successfully kept proposals for the elimination of nuclear weapons in a time-bound framework out of the NPT Principles and Objectives decision, it is considered highly unlikely that the nuclear weapon states - or at least the US, UK and France - will agree to it in the CTBT. Indeed, some think that this is the reason India has chosen the time-bound language for its stand. Almost all the rest of the G-21 have indicated that they would not insist on 'time-bound framework', but they want more than the Australian model offered. Believing that part of India's purpose may be to reassert the CD as the forum to negotiate on nuclear disarmament (and lessen the role of the NPT enhanced review process agreed in May 1995), some may insist that the CTBT preamble should at least be as strong as the nuclear disarmament language in the NPT Principles and Objectives decision. Thus the CTBT might be identified as 'a meaningful step in a programme of action to reduce nuclear weapons globally, with the goal (or 'ultimate goal', though this is disliked by G-21 representatives) of eliminating those weaponsÓ' Something along these lines might be acceptable. For India and the non-aligned the 'programme of action' reference implies further steps. Since the P-5 accepted this term to secure indefinite extension of the NPT, a similar acceptance in the CTBT context should not pose an insuperable obstacle.

Following the zero yield decision, and with more countries coming under pressure to accept the generally worded Australian scope text, non-nuclear weapon states want to ensure that this does not allow those with advanced technology to carry on business as usual. The fact that the US, UK and France had backed the Australian scope before adopting the zero yield decision was not missed. Combined with the US Department of Energy's announcement of its 'sub-critical' test programme for 1996 and 1997, this has fuelled G-21 demands for a clearer statement of intent: if not in the scope, then in the preamble.

Pakistan made the most far-reaching proposal, seeking an additional article on purposes and objectives of the treaty. The intention would be to tie down as treaty text the assurances given by ACDA Director John Holum to the first CD plenary of 1996: 'Without nuclear testing the nuclear weapon states will not be able to pursue confidently such technologies as the nuclear-explosion-pumped X-ray laser, the so-called nuclear shotgun, enhanced electromagnetic pulse weapons, microwave weapons and enhanced-radiation weaponsÓThe true-zero test ban will also place out of reach new 'mini-nuke' and 'micro-nuke' concepts -- technologies designed to use nuclear explosive yields in small amountsÓBy fending off such developments, the CTBT will help make nuclear war less likely, and sustain today's trend toward smaller nuclear arsenals with shrinking roles in national defences.'20

Though Pakistan's proposal of an additional article is unlikely to be taken up, nuclear weapon development will have to be covered in the preamble to get full backing for the Australian scope. India's paragraph on qualitative improvement and development is widely supported among non-aligned states. The US has indicated a willingness to consider some form of preambular language along these lines. The UK objects that this cannot be 'the principle objective', but would be willing to consider language identifying this as a result or consequence of the treaty. The answer may be to make the prevention of new and qualitative developments a 'desire' or 'aspiration'. Though no solution yet commands consensus, there is optimism that an adequate compromise will be found (providing the goal-posts are not moved either by India or by further announcements from the US Department of Energy).

These are the major areas of contention in the preamble. There is also disagreement about including a paragraph on protection of the environment, and whether the reductions in the nuclear weapon arsenals are 'deep' or not. India wants all reference to the NPT to be omitted. Iran dropped all reference to the NPT and left out 'deep'; Australia kept one reference to the NPT, omitted 'deep' and caused surprise by deleting the environment paragraph, despite its national position of support, on grounds that it was opposed by some and 'notÓessential in the context of this treaty.'


Scope is the backbone of the treaty, determining what is to be prohibited or permitted. During 1994 and 1995, as previously discussed, the real debate was conducted among the P-5 and focused on activities not prohibited, with each having a different requirement. That substantially changed when the US, France and the UK adopted the zero yield concept, throwing their weight behind a scope formulation originally proposed by Australia in March 1995:21

  1. Each State Party undertakes not to carry out any nuclear weapon test explosion or any other nuclear explosion, and to prohibit and prevent any such nuclear explosion at any place under its jurisdiction or control.

  2. Each State Party undertakes, furthermore, to refrain from causing, encouraging, or in any way participating in the carrying out of any nuclear weapon test explosion or any other nuclear explosion.

This is reflected in Ramaker's working paper. Although the rolling text on scope, with all its brackets, is faithfully reproduced, Ramaker also identifies the Australian scope with a note that it has attracted wide support.

Despite public appearances to the contrary, Moscow has not yet committed to the majority position, preferring language based on the PTBT. This would add 'underground' to the environments prohibited in that 1963 Treaty. In 1994, China had proposed scope that would ban any nuclear weapon test explosion 'which releases nuclear energy', intended to exclude thresholds and hydronuclear testing but exempt PNEs. On March 28, China accepted that a 'common understanding' had been reached on the phrase 'any nuclear weapon test explosion', so that the CTBT would 'without any threshold, prohibit any nuclear weapon test explosion.'22 With that understanding, China therefore withdrew its additional phrase referring to the release of nuclear energy. However, China refuses to endorse the Australian scope formulation because it cannot accept that 'any other nuclear explosion' should also be banned.

In June 1995, after the indefinite extension of the NPT and Chirac's decision to resume French testing, and as the debate over thresholds raged amongst the P-5 and within the US agencies, Indonesia and India both tabled scope texts that would prevent any continuation of low yield testing by the nuclear weapon states. Indonesia's approach was simple and radical: prohibit all nuclear weapon testing, whether explosive or not, and any nuclear explosion. On February 8 Indonesia withdrew its proposal in favour of the Australian language, with the understanding that this 'will ban all types of nuclear tests in all environments for all time.'

India's proposal was based on (but not identical with) a definition originally proposed by the US based Natural Resources Defense Council (NRDC)23:

  1. Each State Party undertakes to prohibit and to prevent, and not to carry out, any nuclear weapon explosion, or any other nuclear test explosion, or any release of nuclear energy caused by the (rapid) assembly or compression of fissile or fusion material by chemical explosive or other means, at any place under or beyond its jurisdiction or control.

  2. Each State Party undertakes, furthermore, to refrain from causing, encouraging, assisting or in any way participating in the carrying out of any nuclear weapon test explosion or any other nuclear explosion.

The nuclear weapon states object that this definition would prohibit the use of nuclear weapons, place some civilian applications in a grey area and complicate verification. Their opposition makes it certain that this scope cannot win. Despite this, and criticisms from others that this definition is ambiguous and could encompass nuclear operations such as laser ignition, which India itself wishes to develop, the proposal remains in the rolling text. Indeed, India recently inserted the word 'rapid' (in parentheses above) in the hope that this would clarify the intended exemption of civilian operation; but others remain unconvinced. On several occasions in 1996, India's Ambassador, Arundhati Ghose, has emphasised that 'as the PTBT drove testing underground, we do not wish the CTBT to drive testing into the laboratories by those who have the resources to do so.'25 Therefore, India argues that the scope should leave 'no loophole for activity, either explosive based or non-explosive based, aimed at the continued development and refinement of nuclear weapons.'26

Aiming to encourage the CD to come to agreement on scope, Germany and Sweden withdrew their early proposal that the text should explicitly ban 'preparing' to test. However, wanting to ensure that 'apparently imminent non-compliance'27 should fall within the purview of the implementation regime, Germany proposed modified text in the sections on consultation and clarification and the Executive Council.

The Australian model treaty reproduces its own scope formulation, since this is now the front runner. Despite Tarmidzi's withdrawal, the Iranian draft revived the Indonesian formulation on nuclear tests:

  1. Each state party undertakes to prohibit, to prevent and not to carry out, any nuclear weapon test or any other nuclear explosion at any place under its jurisdiction or control.

  2. Each state party undertakes, furthermore, to refrain from causing, encouraging, or in any way participating in the carrying out of any nuclear weapon test or any other nuclear explosion referred to in paragraph 1 of this Article.

The Iranian draft, like Australia's model, provides an example of what the treaty could be; its suggestions are not proposals as such. Though there was speculation that Pakistan might adopt this scope, it did not do so. Since Indonesia has withdrawn its formulation and no-one has taken it up, this demand for an explicit ban on all testing activities, including laboratory and simulation experiments, is now considered dead.


China's desire to exempt so-called 'peaceful' explosions from the CTBT is proving one of the greatest hurdles to agreement. Unlike the US and Russia, China has never conducted a programme of PNEs, a point incorporated in China's rationale for not wanting to close the option off now. At the height of its campaign around 'atoms for peace' in the 1950s, the US developed its 'Plowshares' programme to use nuclear explosions for civilian purposes. Not very successful, the programme was finally abandoned in 1977. The US would have been prepared to give up PNEs by the time the NPT was being negotiated, but this was unacceptable to the Soviet Union, which conducted more than a hundred such explosions. So Article V of the NPT provided a right for non-nuclear weapon states to receive the 'potential benefits from any peaceful applications of nuclear explosions'. Only the five nuclear weapon states, as defined in the NPT, would be able to conduct PNEs 'under appropriate international observation and through appropriate international procedures'. In fact no non-nuclear weapon state has ever requested a nuclear explosion to be conducted on its behalf. The Soviet programme continued into the 1980s, but with growing questions over its environmental safety and economic viability. Apart from the US and Russia, PNEs have not been pursued, although India claimed that the nuclear explosion it conducted in 1974 at Rajasthan was a peaceful explosion.

At first, China was assumed to be advocating PNEs in order to use it for leverage later in the negotiations, but as its position has hardened, this view has come to be doubted. When China abandoned its requirement for language specifying no nuclear energy, it took the occasion to re-emphasise its opposition to banning PNEs: " an important principle, any disarmament or arms control treaty should not hinder the development and application of science and technology for peaceful purposes. Therefore it would be incorrect if [the] CTBT should ban PNEs... As a populous and developing country with insufficient per capita energy and mineral resources, China cannot abandon forever any promising and potentially useful technology that may be suited to its economic needs. China fully shares the concern over the possible misuse of PNEs. However, we think that the issue can be solved by establishing a strict application and approval procedure and an effective international on-site monitoring mechanism for the whole process of PNEs.'

Although China is in a minority of one in wanting PNEs, its obduracy led to a growing fear that this issue could break the treaty. Some began to seek a way out. Russia had begun to circulate behind the scenes an idea based on the 1977-1980 tripartite talks, in which the treaty would institute a moratorium on PNEs until they could be conducted without military benefit. Iran took this a step further in its draft treaty. While the Iranian scope language prohibited PNEs (and insisting that Iran itself did not favour their inclusion), the Iranian draft gave the Conference of States Parties responsibility for considering 'in exceptional circumstances and in the case that the real benefit of nuclear explosion for the sole purpose of purely peaceful scientific research and civilian applications are demonstratedÓ..a specific request for conducting a peaceful nuclear explosion.' A PNE would only go ahead if four fifths of the Conference of States Parties agreed, and with verification 'to ensure that it will be conducted for purely peaceful purposes.'28 Considering that this condition would never be met, several states expressed initial interest in the proposal, seeing it as a face-saving way to get China on board.

However, such a provision in the treaty could have far-reaching consequences. As it stands, Iran's provision again opens the PNE possibility to all CTBT states parties, a point which India reportedly insisted on. As a treaty subsequent to the NPT, this provision could be interpreted as taking precedence in law, thus making a nonsense of the NPT's injunction against acquisition of nuclear materials for nuclear explosive devices. If restricted to the P-5, as China wants, the provision would provide another point on which India and Pakistan could complain of discrimination. Moreover, it would provide a right (and excuse) for the nuclear weapon states to maintain nuclear design teams, laboratories, and research facilities for nuclear explosions in perpetuity. This could make future nuclear arms control or nuclear disarmament measures even more difficult. While some may argue that the US programme of stockpile stewardship and its equivalent in other nuclear weapon states would do that anyway, there is a big difference between plans set up under the umbrella of 'safety and reliability' and a right in a universal and indefinite treaty. MinAtom and sections of the Russian military which have been supporting PNEs would no doubt join the Chinese in exploring the possibilities. Even if the US did not expect ever to conduct PNEs, the prospect of others putting additional research facilities into the technology would ensure that the laboratories received large sums of government money - at least to track what the opposition might find out! Stockpile stewardship, on the other hand, is being used partly as a crutch to reduce the immediate impact of a CTBT, and would be likely to lose cash and credibility over time.

Though China may have been initially pleased that the Iranian draft was being considered seriously, states are returning to the view that a CTBT should close the door on all nuclear explosions, with no exceptions. Canada immediately declared itself 'allergic' to PNEs. Others have returned to the overwhelming technical evidence that it would not be possible to verify that a PNE is not a weapons-useful test. Some are still anxious about how to get China into the treaty if the PNE issue proves to be a non-negotiable condition. Recalling China's primary argument that the PNE option should not be closed off for the future just because the technology isn't adequate at present to ensure there is no military benefit, Mexico pointed out that the treaty will contain an amendment procedure that would always be available if warranted in the future. Thus 'the absence of any mention of peaceful nuclear explosions in the treaty does not mean banning them forever.'29 Although amending the treaty would be difficult and subject to consensus among states parties attending a specially convened Amendment Conference, would the four-fifths agreement of the Conference of States Parties be more easily obtained?

Opposition to PNEs is almost universal. Ramaker reflected this by sidelining the PNE provision proposed by China. No state has taken up and proposed Iran's suggestion. Yet China remains adamant that there should be no CTBT without permission for PNEs. While the negotiating states tend to focus on the difficulties of verification and the technological problems of ensuring that the country conducting them is not learning anything of military value, public opinion is more worried about the long term health and environmental effects of continued nuclear explosions, whatever the purpose. Neither Russia nor the US have an exemplary record in this. If a PNE provision gives the nuclear weapon research laboratories a new lease of life, it could undermine a central objective of the treaty. Though everyone wants China in the CTBT, a PNE provision may be too high a price to pay.

The Implementing Organisation

As Ramaker's working paper indicates, the treaty will be implemented by a Comprehensive Test Ban Treaty Organisation (CTBTO), likely to comprise a Conference of States Parties, an Executive Council, and a Technical Secretariat. The international data centre (IDC) will be under the auspices of the Technical Secretariat. The Technical Secretariat will be headed by a Director-General, appointed for a renewable term of four years by the Conference. The major political decision concerns the composition of the Executive Council.

There is now virtual agreement that the CTBTO will be in Vienna, co-located with the IAEA, from which it will be structurally independent but with some sharing of facilities and resources. From the beginning, Sweden, backed by Egypt and Brazil, had proposed that the IAEA should be entrusted with full responsibility for implementing the CTBT. The US and China were strongly opposed, with the majority of the states undecided, seeing merit and drawbacks in each option. The IAEA made its case in four submissions before the NTB Committee: in February, May and June of 1994 and in December 1995.

Replying to detailed questions from the Committee on the legal, technical and financial issues relevant to IAEA involvement in implementing the CTBT, the December 1995 paper argued that CTBT verification would naturally dovetail with existing IAEA safeguards agreements with states parties to the NPT, the Treaty of Tlatelolco and the Treaty of Rarotonga. '[T]he IAEA is already verifying in the territory of non-nuclear weapon states parties to comprehensive safeguards agreements that no nuclear material is diverted to any nuclear explosive device and consequently that no nuclear material is diverted to any nuclear test.'30 The IAEA would thus be able to 'extend [the] prohibition of nuclear tests to states without comprehensive safeguards agreements and would also establish an additional verification system to provide additional assurance that no tests are taking place.'31 Admitting that it had no particular expertise in hydroacoustic, infrasound, EMP and satellite monitoring, the Agency argued that it had some basic background in seismology, considerable expertise in radionuclide monitoring and some experience in analysing satellite imagery. Moreover the IAEA had considerable experience in on-site inspections and the infrastructure to train the specialists required in the other fields.

The IAEA put at 13-22 percent the proportion of overheads which it would charge to administer the CTBTO, and indicated that operations and staff costs for the IMS and IDC would require specific funding, whether the CTBTO was part of the Agency or not. Nevertheless, the paper argued that the CTBTO could make savings in administration and management services if it became part of the IAEA. It indicated that its own preference would be for the CTBTO to become a department within the Agency, under the general authority of the IAEA Director General but with a separate budget and a degree of autonomy.

As delegations considered the IAEA's offers, it seemed that the US was prepared to reverse its opposition and back the IAEA proposal to include the CTBTO as an autonomous department. However, there was not enough support for this, and in February, Sweden and Egypt withdrew their proposal instead.

Executive Council

As the Conference would normally meet annually, the main decision-making body for questions of verification and compliance will be the Executive Council. Therefore countries which consider themselves as having a special interest, including the nuclear weapon states, want assured seats. The G-21 countries tend to reject any concept of permanent seats, emphasising the need for regional and political balance, with no state excluded.

Iran's draft drew from the IAEA structure, with a council of 65 members elected from the conference according to their region. Although all states would have the right to be on the council, there is no specific structural mechanism preventing the exclusion of a particular state if its region persistently fails to elect it. Australia looked to the 1992 Chemical Weapons Convention (CWC) precedent, suggesting a smaller Executive Council of 41, regionally designated, but with at least two members per region per session designated by alphabetical rotation. This would guard against any permanent exclusion of any particular state party. Australia sought to balance interests and accountability by specifying that particular priority should be accorded to those states parties: a) having the highest number of IMS stations; b) having sizeable nuclear infrastructure; and c) which have ratified before the treaty enters into force.

Morocco's Ambassador Benjelloun-Touimi, as Friend of the Chair, came up with an option somewhere between the IAEA and UN regional models. As it appears in the Chair's working paper, the EC would consist of 45 members:

8 from Latin America;

6 from Eastern Europe;

9 from North America and Western Europe;

9 from Africa;

6 from the Middle East and South Asia;

7 from South-East Asia, the Pacific and the Far East.

He then defined the selection process to be followed by each geographical region. At least one third to be filled according to the following priorities, as determined by the region itself: nuclear capabilities, number of IMS stations, expertise and experience in monitoring technology, and financial contribution to the CTBTO. One seat would be subject to alphabetical rotation. The remaining seats would be filled by either rotation or elections. Although interest has already been expressed in this idea, it is understood that the African states were particularly disappointed in their under-representation. A slightly modified revision of Benjelloun-Touimi's paper was brought out on March 29, but already Brazil had given notice that robbing Latin America of a seat to give to Africa would be unacceptable. While there are still details and balances to be worked out, it is thought that the final treaty will contain an Executive Council formula very close to the concept proposed by the Friend of the Chair.

Preparatory Commission and Funding

Some preliminary consultations on setting up the Preparatory Commission and on funding have been undertaken by Friends of the Chair from Canada and Japan. If a CTBT is concluded, the Preparatory Commission would be convened under the auspices of the UN within 30 days after the treaty had been signed by a certain number of states (probably 50). This PrepCom would be responsible for overseeing the development of the CTBTO and verification regime for at least two years, in preparation for full implementation when the treaty enters into force. Provisional cost estimates for the PrepCom and verification system during the two year (expected) setting up period have now been estimated as in the region of 74.5 million US dollars.

The most controversial issue is funding. There is acceptance of the principle that the CTBTO and IMS should be funded in accordance with the UN scale of assessments, adjusted to take into account differences in membership between the UN and the CTBTO. To reduce international costs, the US had initially proposed that states should be financially responsible for the running costs of stations on their own territory. This was opposed by many, who feared that it could either lead to an unfair burden on some countries or cause delays in providing stations identified as necessary to the IMS. In August, the US proposed a compromise according to which the CTBTO would finance the capital, operation and maintenance of IMS stations set up after the CTBT is signed. Rather than all payment being made in cash, the US further proposed that parties should have the option of meeting all or part of their annual assessment either by direct payment or by a 'contribution credit'. While negotiators welcomed the US' willingness to compromise, there are concerns about how this would work in practice. States have questioned whether this could cause a 'cash flow' crisis in the Organisation if too many states opted for payment in kind. Some have also raised the fear that this could be used as a precedent for funding other UN-related bodies, with even worse consequences than the present UN financial crisis.

Responding to these concerns, the US has explained that the contribution credit payment option would ensure that the CTBTO becomes operational as quickly and effectively as possible. Contribution credits would be available primarily for the establishment and operation of IMS stations, their operations and maintenance, and the costs of establishing the IDC (including the operations and maintenance of the prototype IDC in Virginia). The US does not propose that administrative and overhead budget items should be offset by contribution credits.

Several states have responded with objections or alternative suggestions. Noting the concerns about liquidity and the possibility of a cash flow crisis arising from this proposal, South Africa suggested that the contribution credits should be limited to the establishment and upgrading of the IMS infrastructure and that there should be a limit (a 'percentage cap') on the amount of credit a state party can receive. Sharing these concerns, Germany advised that the US proposal could also lead to inflation of costs and that states which were not in a position to make contributions in kind could feel discriminated against. Germany indicated a willingness to accept a modified form of the contribution credit proposal, allowing only capital costs associated with establishing and upgrading IMS facilities to be offset. Costs on existing facilities and running costs would not be eligible.

Though the issue is not yet resolved, it is likely that a system involving a more limited incorporation of contribution credits than the US originally intended will be accepted.


The verification regime will aim to provide a cost effective system that imposes a high degree of uncertainty and crippling financial and political costs on anyone contemplating a violation, thereby deterring any state from attempting to cheat. The envisaged verification system will comprise: an international monitoring system, the core of which would be a global seismic network; transparency, consultations and clarification; and on-site inspections. Differences over the conduct, timing and decision-making procedure for inspections is threatening to derail the treaty. Central to an agreement is whether and under what conditions (if any) a state may use information from its own national intelligence sources to supplement IMS data. Additional problems have arisen over the function and products of the international data centre: whilst the US argues that the IDC should only compact and process the data from the IMS stations, leaving the analysis up to states parties, other countries want the IDC to provide a 'user friendly' bulletin with some level of analysis and interpretation of the data.

International Monitoring System (IMS)

By the end of December 1995, an IMS of at least four technologies had been practically agreed. These were:

Three proposals have yet to be resolved: i) Russia's recent request for seismic and monitoring stations at the operational test sites; ii) China and Pakistan's long-standing desire to have international satellite and electromagnetic pulse (EMP) stations; and iii) whether noble gas monitors should be part of the IMS.

monitoring the test sites

In January, Russia caused surprise with a late proposal that four seismic and radionuclide stations should be located at the four operational test sites: Nevada, Lop Nor, Novaya Zemlya and Moruroa (though with the French decision to close the Pacific Test Site, Russia is prepared to forego this). Russia contends that Novaya Zemlya is monitored more closely than Nevada and Lop Nor, arguing for 'identical transparency' at the test sites.33 China objected that Lop Nor was monitored more closely than the global average (while ignoring the fact that the P-5 are more capable of a nuclear explosion than average) and rejected any further enhancement of the detection level as 'excessive and unacceptable'.34 While most states would have been likely to agree the Russian request had it been made before the list of stations for the IMS was finalised, they do not wish to deal with China's threat to re-open the entire IMS agreements. Additionally, some query how much the IMS would actually gain from the changes that Russia proposes. Russia and China both argue it on principle, although the concern is clearly political. When challenged for the lateness of this new demand, Russia has indicated that its perception of IMS requirements had altered with the zero yield decision. The IMS is designed with a 1 kt 'measure' for global coverage, although there will be regional variations, and through synergy the combination of technologies should ensure detection at much lower levels.

The reported target of Russia's concern is the Nevada Test Site, but Russia has refused US offers of bilateral confidence-building measures. With China hostile to increased monitoring at Lop Nor , the question became deadlocked. The majority of the CD would be willing to modify the system if necessary, but are resistant to opening the IMS issue up again at this late date. The P-5 are now attempting to resolve the problem among themselves.

satellites and EMP

An electromagnetic pulse (EMP) is characteristic of nuclear explosions and may be detected from a considerable distance. China, followed by Pakistan, argued very early on that the IMS would be inadequate unless it included networks of satellites and EMP sensors. While not discounting the usefulness of such data, the majority of states argued that it would be neither necessary nor cost effective to provide these networks as part of the IMS, since there are high quality commercial and national monitors already in place. On March 28, China offered the first glimmer of hope when it signalled a review of its position. However, any decision to drop the requirement for specific satellite and EMP networks as part of the IMS would no doubt have to be accompanied by agreement for getting access to such data by other means. This inevitably relates to the question of NTM admissibility.

noble gas monitoring

The release of radioactive isotopes of certain gases, notably argon and xenon, are characteristic of nuclear explosions. Past experience at the test sites show that even the most skilfully conducted underground explosions may vent these gases unpredictably. While argon-37 has a half life of 35 days, xenon-133 has a half life of only 5 days, and xenon-135 a half life of 9 hours, so investigation must be prompt if these are to be detected. Depending on what agreements are reached on conduct and timing of on-site inspections, these may not be quick enough to detect the xenon. In the view of many states, a network of noble gas monitors, co-located with the already agreed sensors for radioactive particulates, could increase the chances of picking up this important evidence. Moreover, given the unpredictability of venting, such a network would greatly increase the uncertainty and risk undertaken by a potential violator. China is among the small number of countries that regard such a network as otiose.

International Data Centre (IDC)

The seismic, hydroacoustic, infrasound and radionuclide monitoring stations identified for the IMS will feed data to an international data centre, which will be incorporated as part of the Technical Secretariat. In its raw form, this data will amount to hundreds of millions of bytes of computerised information or thousands of pages per day. States want the IDC to send them regular bulletins, but differ about what this should cover. All agree that sending out the raw data would be unmanageable. There is agreement that the role of the IDC is to provide states with the information they need to assess compliance with the treaty, and that all states should have equal access to the raw data. There is also agreement that any actual judgement on compliance will require political as well as technical assessment, and should be left to states parties. However, technology levels vary, and few have the capacity to analyse the raw data regularly or effectively. To be able to make such a judgement, or to raise doubts and request an inspection, states with limited resources want the IDC to analyse the data and send them a comprehensible, 'user friendly' bulletin. For many this would include highlighting anomalous events as detected by the IMS, providing probability values and/or providing a preliminary identification of an anomalous data set. Afraid that this could lead to a usurpation of states parties' responsibility to assess compliance, the US argues that the IDC should only process, compact and disseminate data.

Ralph Alewine, Friend of the Chair on the IDC for most of 1995, and reappointed in 1996, issued a working paper at the end of February, in which he itemised a set of eight 'core' services and related products and presented three options on data presentation and analysis.35 The core services (with products in parentheses) are:

After consulting with delegations about the kind of functions and products they required, Alewine identified three options, with different levels of screening and analysis of data:

  1. raw data provided, but no standardised event screening; filtering and analysis of data on a national basis (or by the IDC using criteria established by individual states parties);

  2. internationally standardised event screening according to criteria established by the CTBTO; additional filtering according to nationally requested criteria; technical assistance;

  3. internationally standardised event analysis; screening according to national criteria; provision of technical expert evaluation for identification of events in the IDC screened event bulletin; technical assistance to establish nationally relevant screening criteria and to set up a system for data analysis at national data centres.

A standard data package would be provided with all three options, but because of the greater level of analysis in option 3, this would be around 5-10 Mbytes per day, as compared with 50-100 Mbytes per day for options 1 and 2. Under all the options a state party would have the right to be provided with all the raw data on request. Option 3 would provide states parties at no extra cost with customised subsets of raw data, as determined by an individual state's particular regional or political concerns; option 2 would not charge extra for 'reasonable efforts'; while this service would incur additional cost under option 1. In effect, option 1 is presented as the cheapest for the CTBTO to provide, while option 3 would be most expensive. However, option 1 would be more costly on an individual basis to those states which wished to avail themselves of the extras, like customised summaries. Constructing and equipping national facilities would be very expensive if each state party had to be responsible for doing the analysis itself.

Option 1 may be said to represent the 'extreme' US position, as put forward last year. None of these options represents the opposite 'extreme' of a bulletin which could only flag up suspect events. Option 3 comes closest to the position favoured by most G-21 countries, which in June 1995 issued a statement calling for 'the Organisation [to] have the capacity and the responsibility of analysing all data'. While it was assumed that western countries would go for option 2, Germany has argued that this does not make the most cost-effective use of IDC expertise. In a follow-up working paper, Germany summarises the Friend of the Chair options thus:

Basic:no filtering of data
Option 1: customised filtering at a national level (states parties responsible)
Option 2:additional standardised automatic filtering
Option 3:additional expert filtering (IDC responsible)

On the basis that the IDC would have the knowledge, expertise and tools to do the most effective job, Germany suggests that it would be absurd for each state party to duplicate the costs and facilities necessary under options 1 and 2. Dismissing the US argument that option 3 would usurp the responsibility of states parties for assessing compliance, Germany argues that all the options are simply different degrees of technical evaluation and screening of data; therefore option 3 is a further technical screening of events not filtered out by the standardised event screening in option 2. Far from straying into judgement, this option - or something closer to option 3 than to option 2 - would, in Germany's opinion, be a more cost effective use of the international verification regime and would ensure greater participation by states with limited technical capabilities of their own.

With Germany wading in on the side of the majority of non-aligned countries, and the US adamant that the IDC's products should be more limited, this issue is far from resolved. Though the question of IDC responsibility is political, this issue is primarily one of finance and participation. It may be sticky, but is not expected to be a treaty breaker.


After nearly 20 years of discussions under the auspices of the CD, the ad hoc Group of Scientific Experts to Consider International Co-operative Measures to Detect and Identify Seismic Events (GSE), chaired by Dr Ola Dahlman of Sweden, began testing GESTT-3 (GSE Technical Test 3) in January 1995. The purpose of GSETT-3 is to evaluate the effectiveness of an international seismic network and data transmission to a single IDC.

Originally set up to provide data and experience on which to base the seismic component of CTBT verification, GSETT-3 was designed to test a primary global network of about 50 seismic stations, and 100-150 auxiliary stations feeding data to a single IDC, using the existing US data centre in Virginia. Involving scientists and specialists in many different countries, GSETT-3 has examined the involvement of national data centres, provided training experience and also tested a specially designed communications system to handle the enormous data exchange.

Since the close of the 1995 session of the CD, the GSE has held meetings from November 27 to December 1, 1995 and from February 12 to 23, 1996. By the end of 1995 the GSETT-3 network comprised 41 primary stations and 76 auxiliary stations, of which 30 primary and 33 auxiliary stations are earmarked for the IMS, as currently envisaged. Although problems of software overload were identified, the GSE judged the concept of a single IDC to be 'proven effective'37. The provisional IDC at Virginia had demonstrated 'the feasibility of providing a wide range of event bulletins, raw data and derived products' to national data centres (some of which would need upgrading to a more reliable level of operational capacity). GSETT-3 was deemed successful at detecting explosions to at least 1 kt, though location capability varied regionally. This was attributed to inadequate and uneven distribution of stations and the need to improve the tuning and calibration of the system being tested.

On-Site Inspections (OSI)

Inspecting the area of a suspected event may be the last resort for the states parties to decide whether there has been a violation of the treaty. Because of national security considerations, this is one of the most sensitive issues to resolve - and perhaps the most neglected. At the beginning of the year, Canada was appointed Friend of the Chair. Attempting to clean the existing text of muddle, Ambassador Mark Moher and Don Sinclair identified five central issues which must be determined by political decision: the basis on which an inspection could be requested; the decision-making process; the role of consultations and clarification; timelines; and the role, if any, for a technical evaluation.

The dominant view is that OSI should be rarely invoked, with penalties for abuse, but that the provisions in the treaty should be robust enough to deter anyone from running the risk of cheating. Negotiating the actual provisions has exposed deep differences over national security, equal access to information, and what constitutes a timely and effective inspection regime. Anxious to protect military and commercial confidentiality, governments want to prevent unnecessary intrusion (and restrict or manage legitimate access). There is no agreement yet on the timing and decision-making process, particularly the 'trigger mechanism' of how an inspection is initiated. Most states agree that only states parties should have the power to request an inspection, with the CTBTO taking responsibility for the conduct of the OSI and presentation of report(s). A few have strongly argued for a technical evaluation of an OSI request from experts outside the Technical Secretariat, which the majority consider unnecessary. There is also disagreement over whether the inspection team report should provide the data only, the data plus a technical statement interpreting them (e.g. as characteristic of a nuclear explosion, something else altogether, or inconclusive); or the data, a statement, and a technical conclusion. The majority prefer the middle option, considering it inappropriate for the inspection team to give a conclusion, but unhelpful merely to get a mass of raw data.

Positions are still polarised over decision-making, specifically, whether an OSI must be authorised by a positive decision of the Executive Council in order to go ahead. The origin and level of evidence required or permitted is central to this question - and to the verification regime as a whole. Specifically, the US (backed by France, Russia, the UK and the Western and Eastern European groups) wants any kind of relevant information permitted, including that which they gather through national technical means. China (backed by India and Pakistan and several G-21 countries) is opposed to any incorporation of NTM. These opposing approaches are again reflected with regard to access. In general, the US, UK, France and most western states emphasise the necessity for quick access and the prompt gathering of time-critical evidence, such as aftershocks and the venting of radioactive gases. All delegations accept that militarily sensitive facilities unconnected with a potential treaty violation should be protected from inspection by means of managed access provisions, which are expected to be developed along the lines of the CWC.

China, India, Israel and Pakistan argue for OSI to be a tool of last resort, used rarely, and only undertaken if a mandatory and stringent process for consultation and clarification has failed to resolve an ambiguity. Russia also advocates a period for consultation, but is more flexible concerning its requirements and timing. Most states are in favour of a consultation and clarification process, but want to ensure that it is not used to delay an inspection so that vital evidence is lost.

For the first two years, Russia was appointed Friend of the Chair on OSI. In December 1994, a Group of Experts reported on the kinds of evidence which might be associated with nuclear explosions in particular environments.38 Focusing on underground explosions, they identified the venting of radioactive isotopes of xenon and aftershocks close to the site of the explosion as important indications which would vanish if not observed within a month. Traces of specially built roads, debris or tailings might also still be discernible to a quick inspection, but erased if the violator was given time.

The report detailed other, less time-critical manifestations: surface cratering; evidence of an underground cavity or rubble zone; surface changes from ground spallation; residual underground radioactivity; the presence of radioactive argon-37 gas; changes in ground water level; and fractured geology, which might be revealed by anomalies of heat, pressure and gas flow. There is general agreement on the types of evidence sought and the appropriate technologies to gather them, with minor exceptions. It is also accepted that the initial phase of an OSI should include overflights, but little work has been done on the conditions and limitations. Overflights can be invaluable for reconnaissance, initial inspection, radionuclide detection and to narrow down the area for intrusive inspection, but care would have to be taken to reassure states concerned about espionage.

Red light, green

The competing requirements of access and confidentiality must be balanced. Those who want to ensure that the risk of inspections - and their likelihood of unearthing convincing evidence - is strong enough to deter a violator, want a presumption of quick access. In general these states would support a simplified decision-making process. Countries which are more sensitive of the opportunities for inspections to be used abusively, including those with a pronounced fear of espionage, want an inspection request to be rigorously considered and subject to a positive decision by the Executive Council. Taking a leaf out of the CWC negotiations, these alternative decision-making procedures are referred to as 'red light' (meaning that an OSI request would be automatically carried out unless the Executive Council countermands it) and 'green light', in which no inspection can go ahead unless specifically authorised by a majority decision of the Executive Council. There is also disagreement whether the decision should be by a simple majority, or by two-thirds or even three-quarters, as Pakistan, India and China prefer.

Last year, the US proposed that an OSI could be conducted in two stages or phases. Its intention was to separate the access requirements for collecting time-critical evidence from the longer more intrusive inspection that would be expected to gather detailed information. Recognising that it was unlikely that a full OSI would be acceptable without some kind of decision by states parties, the US intention was to ensure that at least the first stage would take place automatically, unless there was a red light, while the second phase could be either red or green. France supported this idea, but linked it with the question of evidence, by suggesting that an OSI request based solely on national means should be subjected to a green light process, while a request using international data should go ahead automatically unless stopped by a red light decision. The UK and Russia remained sceptical about the two-phase concept. China embraced the idea, but wanted each phase to be subject to a separate green light decision, which rather undermined the US' original purpose! Pakistan argued against the two-phase process on grounds that the evidence from aftershocks and noble gases was neither as time-critical nor as vital as suggested by the US and others, and could be masked in a well camouflaged test. However, others consider that the unpredictability of gas venting means that a quick-access inspection would make a cheat think twice before risking discovery. The UK argues that an OSI should be triggered automatically, providing the Technical Secretariat deems there is enough evidence to require investigation.

Those which oppose the two-phase approach, such as the UK, see an inspection as a continuum, developing from a less intrusive stage to more intrusive investigation, including drilling. In the UK view, no additional Executive Council decisions should be required if the Technical Secretariat deems it necessary to proceed to a higher level or intrusion during an inspection. If at any time an inspection comes up with evidence that is sufficiently compelling to clear or condemn, it could be terminated. China has proposed a detailed time-line to ensure that inspection teams do not remain on a state's territory for long periods. Russia favours a single inspection decision, but wants it to be made by the Executive Council (green light). Iran and Australia both suggested a two-phase OSI, with presumption of access (red light) for the first, if triggered by IMS data. It now looks certain that agreement on the decision-making process will be bound up with resolution of the question on national technical means.


The term 'national technical means' (NTM) implies different things to different delegations, which is part of the problem. To some it covers only technology such as satellites, bhangmeters, and seismic, infrasound and radionuclide monitoring stations that might not be directly linked into the IMS, but could be categorised as 'IMS-type' data. These could provide a useful supplement to the IMS, and if appropriate, might be accessed directly from commercial or national sources by the Technical Secretariat. At a more controversial level, NTM can include intelligence gathering technology, satellites, signals and communications intercepts. Though the means may be militarily sensitive, the information could be of relevance for the detection and identification of a suspected event. Then there are human and 'illegal' intelligence sources, including whistleblowers, leaks and spies. These might provide an alert or help to pinpoint the area for investigation, but may not be verifiable. Even if such information were permitted, which many states including Russia oppose, it may be necessary to get corroborative evidence from accepted IMS, national or commercial resources. It is conceivable that a state may use its human intelligence to narrow down the search, even if it is not formally accepted. Because of its ambiguity the term 'NTM' is unlikely to appear in the treaty, but the question of what information (if any) can be used to supplement the IMS data from the four basic networks will have to be resolved. Although the permissibility of NTM will affect the treaty in several ways, procedurally the question is likely to arise in only two situations: an OSI request and the judgement on whether a violation has been committed.

Whilst backing the IMS as a cost effective basis for verifying the CTBT, the US, Russia and Western and Eastern European states argue that it would be too easy for a cheat to calculate the chance of detection and design its clandestine explosion accordingly. The uncertainty introduced if NTM evidence were also allowed would greatly enhance the treaty's deterrence capability. But China and several non-aligned countries are more worried about NTM being available to only a few states, and used in exclusive or discriminatory ways. Together with their demand for dedicated international satellite and EMP monitoring networks as part of the IMS, China and Pakistan have repeatedly argued against any incorporation of NTM. Referring to Pakistan's 'historical experience of undue harassment', Foreign Minister Sardar Aseff Ahmad Ali said that the use of NTM as a trigger for OSI would 'allow the use of information which could be subjective, selective and unreliable...'39 China contrasted the 'relatively objective and just' IMS with NTM which 'are controlled and used by individuals or small groups of states parties.' Thus, Ambassador Sha concluded '[i]f NTM are incorporated into the international verification or used for triggering OSI, that would inevitably put most of the states parties...into an extremely unequal position...This is obviously unreasonable and unjustifiable...'40

China and Pakistan are not the only countries sensitive to abuse of national intelligence means. In the past few months, US intelligence has been reported as the basis for accusing India of preparing a nuclear test, Pakistan of preparing a test in response, China of exporting ring magnets (for uranium enrichment) to Pakistan in violation of its NPT commitments, and Russia of possibly having conducted a low level nuclear explosion at Novaya Zemlya. These allegations have been denied by the governments concerned, but have sparked furious international debates. For India, Pakistan and China, these incidents confirm their opposition to NTM. Assuming their validity, US and western representatives have tended to argue that without such intelligence sources, the wrongdoers would have got away with it. The accused states have objected to the arbitrary use of the information and the failure of the US to let others interpret the evidence independently. In June 1995, the G-21 stated the view that 'the judgement by the Organization [regarding OSI] should be based on data received from the IMS'41 Though this is taken to mean opposition to NTM, the position of most non-aligned countries is not so clear. Recognising that NTM can provide cost-effective enhancement of the deterrent and detection capabilities of the verification regime, they are more interested in restricting the potential for abuse than in banning NTM altogether. Several delegations have indicated that NTM data might be more acceptable if it could be verified or corroborated by IMS information. Similarly, those which are against allowing NTM as a basis for an OSI request are more open to its use to supplement or reinforce IMS data in a request or evaluation. Even Pakistan accepts that NTM should be permissible for use by a challenged state party in responding to an OSI request against it.

According to Iran's draft text, an OSI should be based solely on IMS data, but supplementary information could be supplied to the Technical Secretariat. Australia's model treaty allows an OSI request to be based on data from the IMS 'and/or by other elements of the treaty verification regime... [including] any relevant supplementary data or information.' To even things up, Australia also proposes associated measures promoting 'access by all states parties to other technical information and data relevant to the verification of the basic obligations of the treaty.'42

South Africa has been prominent in seeking ways of bridging the gap between the non-aligned and western states over NTM. Acknowledging the concerns of fellow non-aligned states that NTM may be open to abuse, the South African Ambassador J. S. Selebi argued that denying the validity of intelligence data in the treaty would not prevent the surveillance from continuing. Therefore, since supplementary information would benefit all by enhancing the verification regime, it would be better to regulate its use and provide conditions for fairer access to the information, where relevant.43 South Africa followed up with a working paper on OSI which developed the French proposal allowing both IMS and NTM data, but according each a different decision-making weight. South Africa proposed that if a state party wished to rely on NTM data in making an OSI request, it should provide the 'unprocessed information' to the Technical Secretariat for evaluation. This would get round the objection that NTM would be biased because they were not open to alternative interpretation, an accusation frequently levelled at US intelligence.

The purpose of the South African approach is to bridge the perceived gap between NTM and the international verification regime, by providing the CTBTO with a role in evaluating NTM. Although the US and Russia may object to providing raw data and may argue that NTM information is too sensitive to be entrusted to the international organisation, this may be the only way out of the current impasse. Certainly data from national monitoring stations not included in the IMS and from satellites and similar technical means could be provided to the Technical Secretariat in an acceptable form. If a state wished to rely solely on human or signals intelligence that it regarded as too sensitive to be revealed, and if there were no corroborating evidence from the IMS, it is likely that its case would be too weak to get Executive Council approval for an OSI or judgement of non-compliance in any event.

While some states have expressed interest in a solution along the lines South Africa has proposed, they argue that since evaluation of data is involved, the Executive Council should be required to authorise an OSI. It now seems likely that resolving the opposing positions on access and incorporation of NTM may lie in a compromise involving both. In return for permitting NTM subject to evaluation by the implementing organisation or corroboration by IDC/IMS data, the non-aligned states are almost certain to insist on a green light decision-making process. In that event, the two-phase approach requiring two distinct decisions would have little to recommend it to those, such as the US and western states, who want a presumption of access. They may then back the Russian proposal for a single green light decision to initiate an OSI, but push for a simple majority rather than the more stringent two-thirds or three-quarters favoured by Pakistan and India. Even if an OSI were initiated by a single decision, it may be conducted in stages, with the Technical Secretariat being responsible for moving it to a more intrusive stage or ending the inspection, as determined by the evidence. In that event, China and others may insist on time-lines to ensure that the inspection would not be open-ended and interminable.

If compromise is not possible, and if China, Pakistan and India continue to oppose any incorporation of NTM, the US has warned that it may as well walk away from the CTBT, since it would not get Congress to ratify. On several occasions US representatives have underlined that this is a bottom line issue, a 'treaty breaker'. However, compromise by the US will also be necessary if this issue is to be resolved.

Consultation and Clarification

Several states, including China, Israel and Pakistan, want a mandatory process of consultation and clarification, to give an accused state the chance to clear its name before its territory is intruded on by an OSI team. According to China's proposed time-line, this could delay an OSI for three weeks, which others regard as unnecessarily long. Whilst many advocate a period for consultation, they argue that its duration and requirements should be flexible, not obligatory. China and Israel seem to want the consultation period to take place prior to an OSI. The majority consider that it should be initiated as soon as a request has been submitted, but that an inspection should not have to wait for full conclusion of the clarification process; rather the two can proceed in parallel, and could complement each other.


Transparency and confidence-building measures have also been considered. Several states, including Algeria, Iran, India, Indonesia, Nigeria and Pakistan have advocated closure of the test sites, of which only three will be left if France carries out its pledge to close the Pacific sites at Moruroa and Fangataufa. The US and Russia have argued that closure would not be feasible because the sites are also research laboratories. While not wishing to abandon their goals of closure, clean-up and restoration of the land to their former inhabitants, some delegations have begun to consider interim confidence-building measures. The main purpose would be to ensure that no-one continues with clandestine low yield or hydronuclear testing contrary to the zero yield decision. In this regard, Iran proposed that states parties 'shall cease all activities related to nuclear testing and close those parts related to testing in the site...[and] shall also ensure that specifically designed equipment for testing shall be destroyed.'44 Iran also proposed that detailed information on past tests, including dates and yields, should be provided to the CTBTO.

While this may be opposed by the nuclear weapon states, other transparency measures involving notification of large chemical explosions, large scale construction and so on, are likely to be agreed. The details have not yet been fully worked out to everyone's satisfaction, but are unlikely to become a blockage.

Entry into Force (EIF)

Although preparations can be made, the verification regime and decision-making apparatus cannot be fully inaugurated until the treaty enters into force. Although it is customary to abide by the provisions of a treaty once it is signed, it becomes legally binding only on entry into force. With regard to this treaty the distinction is academic for the vast majority, since all but the P-5 and the three nuclear capable states which are not parties to the NPT have either committed themselves not to acquire nuclear explosive devices or lack any kind of nuclear infrastructure. The CTBT will materially affect the military ambitions of only eight countries: China, France, Russia, the United Kingdom and the United States, plus India, Israel and Pakistan. Entry into force provisions are thus a balancing act between the requirement that these eight be on board and the desire to have the treaty come into force as soon as possible. Characterised by Canada as a payoff between politically effective and operationally viable, EIF conditions can become central in some countries' debates over ratification. They must be credible in terms of national security, but not so stringent that they turn the treaty into a dead letter.

From the beginning there have been a myriad of options. These boil down to a simple number, a list of particular states and/or a mechanism for avoiding deadlock if one or a few states fail to ratify. Since there are political difficulties with making entry into force conditional upon the accession of a small number of named states, various formulae have been advanced to camouflage this central intention. Of these, the most popular is a list of 68 states identified by the IAEA as having (or having had) nuclear research or production facilities. This was originally proposed by Russia and now has the backing of China, France and the UK. The US, which fears that this would give too many states the power to hold the treaty hostage, prefers limiting the specific requirement to the five nuclear weapon states plus a number such as 40 or 65. To get round this, Iran proposed that the treaty should enter into force if 65 of these 68 states have ratified, which is a modification of the early US proposal for 90 or 95 percent of a list. This is open to the inevitable objection of what would happen if the remaining three were nuclear weapon or threshold states.

With the same purpose of including the test-capable states without identifying them, Australia proposed ratification by the 75 states which are either CD members or observers (while making clear that its national preference was still for a simple number). Until recently, the expanded CD had been the preference of China, France and the UK, but with the expansion question still unresolved, these countries had switched to the IAEA list. Like Iran, Australia sought a veto-avoiding mechanism, proposing that a 'waiver conference' be held two years after signature if not all 75 had yet acceded. The conference, which would be open to all who had ratified, could decide whether to waive the conditions and let the treaty enter into force with a smaller number of states. The idea of a waiver conference was also originated by the US, with the proviso that the P-5 would have to be on board before the conference could take place, a condition which other states regarded as discriminatory. By the time the Australian draft suggested its modification, the US was ready to drop the proposal, although other countries have indicated that they may wish to keep it alive as an option.

Pakistan was among those countries that opposed the waiver conference, because it insisted that India must accede in order for Pakistan to do so. As it stands, the Australian formula would allow a majority of ratifying states to decide on entry into force. A country which has security concerns about an opposing state which has not ratified may then be vulnerable if it is forced by the majority decision either to withdraw or to adhere while the other remains free to do as it likes. Another drawback of the Australian draft is its provision for only one waiver conference. The purpose of the waiver conference is two-fold: to exert pressure on the awkward states and to give the rest the opportunity to bypass them. If the states which have not ratified are not test-capable, waiving the list conditions may be a very useful way to prevent them holding the treaty hostage to other issues. If one or more of the eight have not acceded, and do not respond to pressure to do so, the conference participants might wish to decide against entry into force at that time, without losing the chance of revisiting the decision with a future waiver conference.

Pakistan recently proposed a formula (originated by the UK delegation, but disowned by the UK itself) which covers only the eight nuclear and threshold states. This would allow Pakistan to sign and ratify, secure in the knowledge that the treaty could not be implemented unless India had also acceded. However India, Israel and China consider that such a formula is discriminatory, while South Africa objects that it confers special status on the eight, setting an unacceptable precedent. In addition to the criticism that this would introduce an undesirable discrimination, it would also place undue power of veto in the hands of the targeted states.

In its March 28 statement, Austria drew attention to its proposal on provisional application, originally tabled in June 1995, and updated in February 1996. With a slightly different approach to the waiver mechanism, Austria proposes that if the treaty has not been implemented two years after the date of deposit of the first instrument of ratification, a Conference of all states which have ratified may decide by a majority (to be determined) to let the treaty - or parts of it - be applied provisionally. This would enable the verification system and CTBTO to be inaugurated, with special financial arrangements agreed between the states concerned. If full entry into force were delayed due to the political circumstances or games of one or a few states, such provisional application could ensure that the treaty could be practically implemented and verified in the interim.

In January India added in another variable, with a proposal relating entry into force to nuclear disarmament: '...this Treaty shall enter into force only after all states parties have committed themselves to the attainment of the goal of total elimination of all nuclear weapons within a well-defined framework (of ten years).'45 The bracket around the ten year target date was intended to show that this was negotiable, but had the effect of causing the P-5 to suffer apoplexy. Such a linkage with entry into force is opposed by almost everyone, and was seen to be India's holding position to ensure more genuine negotiations on its preambular and review proposals.

Long regarded as an issue for the endgame, entry into force may not be agreed until the last moment. However, two trends are emerging: a simple numerical condition, relying on political means to ensure adherence by the key states; or the IAEA list, with perhaps a waiver mechanism to bypass if necessary a destructive use of veto power by one or a few states.

© 1998 The Acronym Institute.

Return to top of page

Return to List of Contents

Return to Acronym Institute main Page