Text Only | Disarmament Diplomacy | Disarmament Documentation | ACRONYM Reports
Back to the Acronym home page
British Policy
South Asia
About Acronym

Disarmament Diplomacy

Issue No. 11, December 1996

The IAEA's Safeguards Programme '93+2': Progress and Challenges
By Suzanna van Moyland


A special nuclear Safeguards Committee will reconvene 20-31 January 1997 in Vienna at the International Atomic Energy Agency (IAEA). It will further consider important proposals put forward as part of the IAEA's Programme to Strengthen the Effectiveness and Improve the Efficiency of Safeguards. With revelations about Iraq's clandestine nuclear-weapon programme and difficulties in implementing safeguards in North Korea, and thus concerns about the damage such cases might do to the credibility of the Non-Proliferation Treaty (NPT), the Programme was initially given enormous support. However, this changed as some States hesitated in the light of burdens that they might be required to undertake.

The Programme's aims have been to give the IAEA greater access to information and locations connected to nuclear activity, enable it to develop better analytical skills, generate more co-operation with States and facility personnel, and capitalise on newer technology. Thus safeguards would become more effective and efficient, and the IAEA more able to detect undeclared nuclear activity at an early stage in States that had made commitments not to develop nuclear weapons.

The learning curve

Iraq is an original signatory to the NPT. As a Non-Nuclear-Weapon State (NNWS) it had also fulfilled its obligations to commit to a full-scope, bilateral safeguards agreement with the IAEA, which is modelled on INFCIRC/153. Yet after the 1991 Gulf War, it was found to have been pursuing multiple paths to develop a nuclear weapon at both declared and undeclared facilities. If undetected, Iraq might have produced a nuclear weapon in late 1993. While North Korea had signed the NPT in 1985, IAEA 'ad hoc' inspections to verify its initial declaration of nuclear activities did not begin until 1992. Evidence emerged swiftly to suggest that all might not be in order. A special inspection request was turned down and North Korea threatened to withdraw from the NPT.

In these cases, two verification tools proved particularly useful: satellites and High-Performance Trace Analysis (HPTA - also called environmental sampling). In North Korea, for example, satellite pictures provided by the United States (US) indicated that a two-storey building had become one-storey. Moreover, results of swipe samples taken from glove boxes did not correspond with the declaration made by North Korea. Furthermore, the case of Iraq had highlighted the need to 'shop' abroad for technology useful for swift development of nuclear weapons.

The IAEA Board of Governors responded to these challenges in 1992 in a number of ways. It reconfirmed the IAEA's right to carry out special inspections (if existing information and access is not considered by the IAEA to be adequate) in States with full-scope safeguards. It also highlighted that early provision of design information (and modifications) was required. Importantly, it acknowledged that the IAEA could use all available sources of information, including that provided by States. Finally, it endorsed a voluntary expanded reporting scheme for States to provide information on exports, imports and production of nuclear material and specified equipment.

This laid the foundations for Programme '93+2'. It also demonstrated the Programme's evolutionary nature by utilising and clarifying tools available to the IAEA, and encouraging other voluntary measures. Recent experiences and newer technology, plus comparative analysis with more recent treaties' verification regimes, called for the framework established by INFCIRC/153 and subsequent practice to be reviewed almost 25 years after it had been drawn up.

The core lesson was that the existing safeguards regime had been demonstrated not to provide other committed parties with confidence of compliance. States could not always be relied upon to make accurate declarations of their nuclear activity. The IAEA would need to adjust further to detecting undeclared activity, as well as more traditional methods of verifying the accuracy of a State's nuclear material accountancy, and they would need to be better equipped and supported to do so. Thus still more was needed to plug the gaps.

Programme '93+2'

The Programme formally began in 1993, when the IAEA Board of Governors requested that the Standing Advisory Group on Safeguards propose ways to tighten the verification regime for States with full-scope safeguards. The Programme's initial deadline was to be the NPT's Review and Extension Conference two years later, in April-May 1995 - hence its nick-name Programme '93+2'. The Conference's Principles and Objectives supported the Programme, stating that "...the Agency's capability to detect undeclared nuclear activities should be increased".

However, just before, in March 1995, when the IAEA Board reviewed the recommendations put forward, they made the important decision to divide the proposed measures into two categories: those for which legal authority was considered to be contained within the existing full-scope safeguards agreement (Part I measures); and those considered to require additional bilateral legal authority (Part II).

Part I Measures already being implemented

In June 1995, the Board approved the implementation of Part I measures. These include:

1) Information on a State's System of Accounting and Control (SSAC), past nuclear activities (including records of material production and decommissioned facilities), and all activities involving "significant amounts" of nuclear material. Prompt, thorough provision of reactor design information and modifications is also required.

2) Complementary access at sites of facilities to verify design information, initial reports and changes to them (known as ad hoc inspections), for material accountancy, or to containment and surveillance measures.

3) No-notice inspections at "strategic points" at a facility or other location where nuclear material is located - a tool not well-exercised prior to '93+2'.

4) Environmental sampling at places inspectors are granted access.

5) Optimisation of improved technologies, such as remote cameras, computerised log sheets for inspectors, and automatic transmission of encrypted data. Also the IAEA is developing its information analysis techniques, using databases and computer programmes to aid cross-referencing and spotting inconsistencies.

6) Deeper co-operation with SSAC to establish where procedures can facilitate inspections, joint support programmes and joint inspections, and sharing facilities such as laboratories.

7) Deeper co-operation with States, including using available direct, secure communications systems between inspectors and the IAEA headquarters, and issuing IAEA inspectors with multiple- or long-term visas, or allowing visaless entry - essential for no-notice inspections.

Implementation of these measures is underway. For instance, the IAEA is establishing computerised proliferation models and secure databases to analyse and cross-reference the increased amounts of information available to it. New remote-monitoring technology and means of relaying encrypted data back to headquarters in Vienna are being established. No-notice inspections of centrifuge plants are being conducted.

Environmental sampling has been a significant Part I undertaking. Trials of this technique, which can identify the isotopic signatures of minuscule radioactive traces with specific nuclear operations, have taken place and a new IAEA 'clean' laboratory at Seibersdorf has been completed in order to handle samples without contamination risks. Procedures have been established to ensure that the samples are analysed 'blind', both in the bulk handling processes now available at Seibersdorf, or, if they are sent abroad to more specialist laboratories for examination, in particulate analysis. Iran, for example, has raised concerns over confidentiality when samples are analysed. Procedures for taking baseline samples have been established. Confusion in late 1995 over taking samples from 'hot cells' at a Canadian facility demonstrated the importance of communication between the IAEA, governments and facility operators about '93+2' undertakings.

Part II Measures under consideration by the Safeguards Committee

Measures under Part II of '93+2' are currently under consideration. These will now require a protocol additional to the original full-scope safeguards agreement. IAEA Member States not currently on the Board of Governors also wished to be a part of this decision-making process, so it was agreed that a Safeguards Committee, Chaired by then Board Chair, Ambassador Johan van Ebbenhorst Tengbergen of the Netherlands, be formed to negotiate by consensus the additional protocol. This Committee is open to any State that has entered into an agreement with the IAEA - not only NNWSs party to the NPT. 65 States attended the first meeting in July 1996, with the European Atomic Energy Agency (EURATOM) and the Brazilian-Argentine Agency for Accounting and Control of nuclear materials (ABACC) having observer status. It met again in October 1996, while discussions continued informally inbetween.

Part II measures considered by the Safeguards Committee include:

1) Information about all buildings on sites of nuclear facilities, nuclear fuel cycle research and development (R&D), uranium and thorium deposits and mines, locations relevant to nuclear activities (e.g. heavy water plants, workshops, nuclear waste stores and stores of nuclear-related equipment), and domestic manufacturers, imports and exports of major items of nuclear equipment.

2) Complementary access to anywhere on a site of a facility and other locations during inspections (these could be no-notice), to decommissioned and closed-down facilities, and where material containing uranium or thorium is located. Also access is sought to nuclear-related R&D locations and other locations identified in the Expanded Declaration as related to nuclear activity, equipment etc. - these proposals have been controversial.

3) Environmental sampling at locations where inspectors have complementary access and to anywhere on a State's territory.

4) Co-operation with States to establish simplified procedures for designating inspectors to States, and acceptance of direct communication systems between inspector and Headquarters when not provided for by the State.

These proposals have met with some resistance, most notably from Germany and Japan, although Belgium, Spain and more recently South Korea have also been reported to have reservations. Information about, and inspector access to, locations where no nuclear material is declared to be located have proved most controversial. Industry has been more sceptical about '93+2' and positions have been to some extent dependent upon which government agency is more influential in the negotiations.

A particular stumbling block has been commercially sensitive R&D, and manufacture, import and export of nuclear-related technology. Misappropriation of information in a competitive industry is one concern. Germany has also held that such access would run contrary to its constitutional laws on searches of private property. It was for this reason that an earlier proposal for access to buildings near facilities was dropped. Notably, however, under the Chemical Weapons Convention (CWC), which will enter into force in April, challenge inspections can be conducted anywhere, including these very locations. Moreover, it is hard to imagine that customs and excise, health and safety inspectors, and tax inspectors do not have intrusive inspections powers.

Various ways have been sought to accommodate these reservations, such as managed access (e.g. shrouding sensitive technology, no cameras etc), prior notice at R&D locations and mutually working out ways in which the IAEA's requirements can be satisfied. It should therefore be possible to find middle ground that would not detract from the core objective of detecting clandestine nuclear-weapon activity - remembering that R&D is the starting point for such programmes.

A more worrying concession would be for the IAEA to have to identify specific reasons, such as problems or inconsistencies, before inspectors are given access to locations identified in Part II's expanded declaration but as having no nuclear material. This would become self-defeating as it would make those inspections akin to special inspections, which the IAEA already has the authority to conduct but which has been seen in the case of North Korea to increase sharply the political tension and might now be interpreted more as a signal to the United Nations Security Council. Furthermore, one could easily imagine the possible delays and contentions as to whether the IAEA did indeed have good reason to conduct such an inspection. (Notably, in agreements such as the confidence-building Conventional Forces in Europe Treaty, no explanation is necessary in order to request an inspection.) The more restricted the IAEA are in inspecting such locations, the more important smooth access anywhere for taking environmental samples becomes. However, it is possible that this could be made similarly complicated to achieve.

The central question with any compromises is whether they may provide loopholes that could be exploited in the future. These kinds of inspections aim to generate a norm whereby such inspections would be routine and not finger-pointing - hence the fact that they must be conducted in all NNWSs, however good its non-proliferation record has been. That inspections of locations where no nuclear material is located would be routine, however, would not mean that they would be frequent because the IAEA's budget could not accommodate that, so industry should not be over-burdened.


A second dimension to '93+2' is to make IAEA safeguards procedures more cost-effective. Japan has been a particular proponent of this. Indeed the Programme is a good opportunity to review this issue. At the same time, though, the IAEA's budget has been at zero growth since 1985, while its safeguards responsibilities have increased significantly. A central goal of '93+2' is to decrease burdens for both the IAEA and industry. Under existing procedures, numbers of routine inspections are calculated in relation to the quantity of nuclear material handled. Thus, for instance, Germany, Japan and Canada are heavily inspected. A pay-off is envisaged, whereby fewer routine inspections at large facilities would be conducted because the probability of detection and thus deterrence is enhanced through more no-notice inspections and additional access at nuclear-related locations. This pay-off, however, cannot be implemented unless a full and integrated package of Part I and II measures is in place.


'93+2' was designed for States with full-scope safeguards. However, the multi-faceted issue of universality and of unequal safeguards commitments impacts upon negotiations.

The divide between the five declared Nuclear-Weapon States (NWS), which are not committed under the NPT to be safeguarded and the some 177 NNWSs parties to the NPT, is likely to be a prominent feature of the January Safeguards Committee meeting. At a political/military level, NWSs cannot continue to drive for tightened non-proliferation without solid moves to disarm, as they are committed to do under the NPT. This keeps cropping up, for example at the Comprehensive Test Ban negotiations and the NPT Review and Extension Conference, and now with '93+2'. The other side of the coin, however, is that NNWSs must be aware that a solid and credible safeguards system for NNWSs will be a necessary condition for far-reaching disarmament by the NWSs. Break-out would do nothing for nuclear disarmament. At an economic level, NNWSs may be reluctant to commit to more safeguards when competitors in industry in NWSs do not. Additional paper work and inspections may be necessary in the initial phase, until everyone settles into the revised regime.

Consequently, NWSs will be expected to commit to as many '93+2' measures as they possibly can. Measures NWSs can undertake include: placing an increasing number of facilities under safeguards; agreeing to Part II measures at safeguarded facilities; transferring military fissile material stockpiles to the safeguarded civil sector; and improve transparency at their own fuel-cycle related R&D locations. The extent to which NWSs participate in reporting exports and particularly imports of nuclear-related technologies, however, goes to the heart of difficulties and disagreements relating to the NPT. Just how binding the reporting system for such technology will be is likely to be a prominent issue at January's Committee.

This touches on another criticism in relation to the NPT: that despite the NPT's bargain of foregoing the nuclear-weapon option in return for transfers of peaceful nuclear technology, discrimination exists in the form of other export controls. Strengthening safeguards and introducing a binding reporting system for NNWSs would help bridge this gap felt between more and less industrialised countries.

India, Israel and Pakistan are outside the NPT and have no treaty obligations to accept full-scope safeguards. They have only site-specific safeguards agreements with the IAEA - modelled on INFCIRC/66. These nuclear-capable States signalled at an early stage that they did not consider '93+2' to be relevant to them. While adoption of aspects of the Programme would be very positive confidence-building measures, attempts to lever them into adopting '93+2' measures at this stage could stall the whole process.

Among States with full-scope safeguards, the issue of equal treatment is also complex. (Though not party to the NPT, Brazil is also committed to using nuclear materials and facilities for only peaceful purposes under the Treaty for the Prohibition of Nuclear Weapons in Latin America and this commitment is verified by the quadripartite agreement with ABACC and the IAEA. Cuba, also a member of the IAEA but not the NPT, has signed but not ratified this regional Treaty.) On the one hand, the IAEA cannot single out a few NNWSs and, moreover, no one knows who may be a proliferator in the future. On the other hand, not all NNWSs may agree to additional protocol measures immediately.

Adopting the additional protocol

Under discussion is an additional protocol to a safeguards agreement and not a treaty. Moreover, States with full-scope safeguards have treaty commitments to enable the IAEA to verify that they are not developing nuclear weapons. Thus, discussion of entry into force in the traditional sense might not be appropriate.

The best way forward would be for Part II measures to be agreed and applied on a bilateral basis between the State and the IAEA. Arrangements for application of Part II measures in regional organisations ABACC and EURATOM may well add a further, more lenghty loop. If adjustments to domestic and regional legislation are needed, provisional application would send positive signals. States may be anxious about a two-tier safeguards system if some States hesitate initially to sign the additional protocol. However, agenda-setting by leading industrialised countries would create the solid foundation necessary to encourage others to join the strengthened safeguards regime.


An agreed additional protocol for Part II measures would provide an excellent backdrop for the NPT Preparatory Committee in April within the enhanced review process. More important than time-frames, however, is substance. Programme '93+2' presents an opportunity to strengthen safeguards and it may be difficult to revisit these issues in the near future.

Suzanna van Moyland is Arms Control and Disarmament Researcher at the Verification Technology Information Centre (VERTIC).

© 1999 The Acronym Institute.

Return to top of page

Return to List of Contents

Return to Acronym Main Page