Published with some editorial changes in
Arms Control Today, January/February 2005, pages
25-9.

Introduction

Negotiation of a fissile material cut-off treaty (FMCT) has
been blocked for several years by the failure of the Conference
on Disarmament (CD) to agree on its program of work. At
this stage therefore we can only speculate about the likely
substantive provisions of an FMCT. The negotiating
mandate for the CD, drawn from a 1993 UN General Assembly
Resolution[1],
has the following elements:

• the treaty is to ban "the production of fissile
  material for nuclear weapons and other nuclear explosive
  devices";
• the treaty is to be "non-discriminatory,
  multilateral and internationally and effectively
  verifiable".

The US has recently concluded that effective international
verification of an FMCT is not realistically achievable.
The US supports the early conclusion of an FMCT to establish
"cut-off" as an international norm, but is
concerned that negotiation of a verification regime will
seriously delay the treaty.

This raises two issues: the practicability of effective
verification; and whether the detailed verification system must
be specified in the principal treaty instrument.

Verification Considerations

Is there anything in the FMCT concept that makes it
inherently incapable of effective verification? This
depends on the objectives of the FMCT, as reflected in its
substantive provisions, yet to be negotiated. We must
take care that questions of verifiability are not
confused with differences over objectives, or issues of
treaty architecture.

For example, the generally held FMCT concept does not
proscribe production of additional nuclear weapons from
unsafeguarded stocks of fissile material existing prior to the
FMCT's entry-into-force (EIF). Rather, the
objective is to ensure these stocks are not added to.
Some may question whether the existence of large stocks of
fissile material outside the FMCT reduces the treaty's
usefulness - but this is an issue relating to the
appropriateness of the FMCT's objectives, not its
verifiability.

The maxim underlying nuclear "peaceful use"
treaties - the principal one being the NPT - is "trust
but verify". The subject matter of the NPT is
of such fundamental importance to national and international
security that no state would be prepared to rely on trust
alone. The existence of a credible verification mechanism
- in the form of IAEA safeguards - is essential to maintaining
confidence in the effectiveness of the NPT and reinforcing the
commitment of treaty parties.

Most states approach the FMCT with a similar perspective -
the FMCT would not be considered credible without a
verification mechanism. Here, we are not starting with a
clean slate - there is already a highly developed verification
regime for nuclear material and activities. We have over
40 years experience with IAEA safeguards. In addition,
there is considerable experience with bilateral verification
and confidence-building mechanisms (CBMs), some directly
applicable to nuclear weapons and sensitive materials.

Treaty Architecture

In the area of multilateral verification treaties, there are
two alternative precedents. One is for a single treaty
containing both the basic treaty objectives and commitments and
the details of the verification system – the approach
taken with the Chemical Weapons Convention (CWC). The
disadvantages of this approach include the time required to
negotiate the treaty – a US concern in the case of the
FMCT – and the degree of inflexibility in the
verification system. Updating the verification system
would be a major political exercise.

The alternative approach – demonstrated very
successfully by the NPT – is to have the basic political
commitments in a principal treaty, and to set out the
verification system in a secondary agreement (or series of
agreements – in the NPT's case each party concludes
a safeguards agreement with the IAEA based on the model in IAEA
document INFCIRC/153). This approach separates largely
political from largely technical subject matters, and allows
for an adaptable verification system.

The NPT was concluded in 1968 and entered into force in
1970. The model safeguards agreement, INFCIRC/153, was
not concluded until 1972. The negotiation of the model
Additional Protocol (INFCIRC/540), agreed in 1997, illustrates
how this basic approach allows flexibility for major updates of
the verification system. Another advantage of having
separate negotiations on the technical details is that such
negotiations can proceed quite expeditiously – despite
their complexity, INFCIRC/153 and INFCIRC/540 each took only
about 18 months to conclude.

Given the commonalities in subject matter between IAEA
safeguards and FMCT verification, it would seem sensible to
follow the NPT route, thus enabling the political commitment
– the international norm – to be established
quickly, and the details of the verification system to
follow.

Objectives and Scope of FMCT

The basic objective of FMCT will be to proscribe production
of fissile material for nuclear weapons or other nuclear
explosive devices. Accordingly, upon EIF each party would
undertake:

• not to produce fissile material for nuclear weapons;
• to accept international verification on relevant
  facilities and nuclear material to verify this
  commitment;
• not to use any fissile material subject to verification
  under FMCT for nuclear weapons - i.e. the principle of irreversibility would apply, "subject
  material" could not be withdrawn for weapons use.

For the purposes of the FMCT, it is expected that
"fissile material" would encompass separated (unirradiated) plutonium, uranium-233, and highly enriched
uranium (HEU). It would probably also need to include
separated neptunium, which is recognised by the IAEA as a
material of potential proliferation significance.

Non-proscribed activities Production of
fissile material for civil purposes, and for non-explosive
military purposes such as naval propulsion, would be permitted,
but only under verification to ensure fissile material is not
diverted to weapons. Recycle (or clean-up) of plutonium
from weapons - an established stockpile stewardship practice
– which does not involve new production of fissile
material, would also be permitted.

States affected The FMCT would apply to
three groups of states:

• non-nuclear weapon states (NNWS) party to the NPT;
• the five NPT nuclear weapon states (NWS);
• the three "nuclear capable" states outside
  the NPT - India, Israel and Pakistan[2].

NNWS party to the NPT are already committed not to produce
or use nuclear material for weapons purposes, and to accept
IAEA safeguards on all their nuclear material and activities to
verify this commitment (comprehensive safeguards[3]). In principle
therefore the FMCT should not involve any additional
commitments from NNWS that are implementing both an NPT
safeguards agreement and an Additional Protocol. The
principal effect of the FMCT - and its verification task -
relate mainly to the NWS and the non-NPT states.

Scope of the FMCT A major issue to be
resolved in the negotiations is the scope of the FMCT - the
facilities and material to which verification would
apply. In broad terms the basic options are:

• a treaty of wide scope, covering all
  nuclear facilities and nuclear material - other than
  non-proscribed military activities, i.e. naval propulsion,
  and subject to the issue of stocks (discussed below); or
• a treaty of focused scope, concentrating on
  the most proliferation-sensitive fissile material production
  facilities - i.e. reprocessing and enrichment facilities -
  and relevant production from those facilities.

There would be substantial problems in trying to extend the comprehensive safeguards system to the NWS and non-NPT
states:

• truly comprehensive safeguards covering all nuclear
  material cannot apply in the NWS and non-NPT states while
  they have nuclear weapons and therefore will retain, outside
  verification, nuclear material existing at the FMCT's
  EIF;
• the cost of verification on the comprehensive model in
  the NWS would be very high.

Taking these and related considerations into account,
Australia has concluded that a separate, distinct FMCT
verification regime would be required for the NWS and non-NPT
states. The FMCT commitment to be verified would be the
same as that verified by IAEA safeguards in the NNWS - i.e.
that fissile material is not being produced for nuclear weapons
purposes – but the verification approach would be more
appropriate to the circumstances of the NWS and non-NPT
states.

Australia has proposed a focused approach, involving:

• the monitoring of facilities that can produce fissile
  material (production facilities) – i.e.
  enrichment and reprocessing plants;
• verification of fissile material subject to the FMCT
  (subject material) - separated plutonium, U-233, HEU
  and separated neptunium produced after EIF. This would
  require verification measures at downstream facilities
  handling these materials.

In addition, the FMCT verification regime will need to
include measures aimed at detection of possible undeclared production facilities - see below.

Stocks

A major issue - fundamental both to the substance of the
FMCT and to the prospects of successful negotiation - is how to
treat past production of fissile material, i.e. stocks existing
at EIF.

The issue of stocks does not arise in the case of states
which have accepted comprehensive safeguards under the
NPT. All holdings of nuclear material in these
states are subject to IAEA safeguards. The issue relates
to those states which are outside comprehensive safeguards and
have the capability to produce fissile material, i.e. the NWS
and the non-NPT states.

The term "stocks" can have a wide meaning,
ranging from fissile material in weapons, to bulk material
declared surplus to defence needs, and to civil stocks of
fissile material. Essentially, the FMCT could not apply
to all pre-existing stocks held by the NWS and the three
non-NPT states, as this would amount to instant
disarmament – clearly an unrealistic objective.
The FMCT will cap future production, but it must be recognised
that past production in the NWS and non-NPT states would be
outside verification.

If the parties so wished, however, the FMCT could contain a
mechanism by which parties could place surplus and civil stocks
under the treaty in accordance with the principle of
irreversibility. Some parties may wish to do this as a
mutual CBM.

Verification Methods

It is envisaged that verification under the FMCT would
comprise three basic elements: routine verification activities
for declared facilities and material; verification activities
aimed at detection of possible undeclared fissile material
production; and complementary measures aimed at transparency
and confidence-building.

Declared activities In concept this
aspect would be very similar to IAEA safeguards. FMCT
parties would be required to declare all relevant facilities
– enrichment and reprocessing facilities, and downstream
facilities handling subject material.

Declared facilities would be monitored – through
inspections, containment and surveillance, and other measures
– to verify there is no undeclared production of fissile
material, and that declared fissile material is not diverted to
nuclear weapons (or purposes unknown).

For enrichment plants, verification would be applied to all facilities, including those producing low enriched
uranium (LEU), to ensure there is no undeclared production of
HEU. In principle verification would not be applied to
LEU – but in view of the advantages of LEU as a feedstock
for undeclared HEU production, some verification measures for
LEU may need to be considered, particularly in the case of
states with smaller arsenals.

For reprocessing plants, verification would be applied to
verify throughput. Verification would be applied to separated plutonium, and to facilities in which
separated plutonium is present, sufficient to ensure the
plutonium remains under treaty commitments. Verification
would cease to apply once plutonium has been returned to a
reactor as fuel and irradiated, since irradiated plutonium is
of no further strategic value until it has been
reprocessed.

Undeclared activities Here, two broad
forms of verification activity are envisaged: routine
activities aimed at evaluating the completeness and correctness
of FMCT declarations; and inspections based on suspicion of a
breach of FMCT commitments.

A major challenge for IAEA safeguards is the detection of
undeclared nuclear facilities, particularly - because of the
limited observable indicators - centrifuge enrichment
facilities. In the past, for IAEA safeguards, techniques
for the detection of undeclared facilities were limited, and
detection of undeclared nuclear material was seen as the major
indicator of the existence of undeclared facilities. Now,
the increasing availability and capability of techniques for
detection of undeclared facilities has led to a revolutionary
change in safeguards. In the ongoing program to
strengthen IAEA safeguards, emphasis is being given to
detection of undeclared facilities.

A whole suite of new measures is being established,
including: more effective information collection and analysis
(including sharing of intelligence information); satellite
imagery; and - through the Additional Protocol - wide-ranging complementary access to apply verification measures such
as environmental sampling. Another technique under study
is wide-area environmental monitoring – there is
practical experience with this technique from UNSCOM/UNMOVIC
activities in Iraq. It can be expected that methods
similar to these would apply under the FMCT.

An issue raised for the FMCT is, how to recognise
non-compliance. Since the NWS and non-NPT states have
"undeclared material" - fissile material
pre-existing EIF – how can it be determined whether
particular material is the result of undeclared production
post-EIF? There are techniques to resolve this question
– nuclear material can be dated – but, as is
increasingly the case with IAEA safeguards, the principal focus
of FMCT verification to counter undeclared activities will be
the detection of undeclared enrichment and reprocessing
facilities.

As with IAEA safeguards, detection of undeclared enrichment
and reprocessing facilities will also be a challenge for FMCT
verification – but in some respects the problem is more
manageable for the FMCT, because of extensive intelligence
information held by the relevant states on each other, and the
limited motivation to cheat for states that have nuclear
arsenals they consider adequate – and in some cases are
actively reducing. The more difficult cases might be
states with small arsenals - India and Pakistan - where
bilateral CBMs could have an important role in complementing
international verification.

Similar to the Additional Protocol, under FMCT the
verification agency would have the right to request access to
locations to resolve questions and inconsistencies arising from
information analysis. In addition to access initiated by
the verification agency, FMCT parties may require a
verification mechanism they can initiate directly, along the
lines of the CWC's challenge inspection mechanism.

National security aspects The NWS and
non-NPT states are concerned to protect national security and
proliferation-sensitive information relating to past nuclear
weapons programs and ongoing stockpile stewardship
activities. Hence, verification methods and procedures
will need to be carefully defined. Managed access provisions will be essential, probably elaborated in greater
detail under the FMCT than in the Additional Protocol.

Other verification methods, and transparency/confidence
building measures In addition to measures drawn
from IAEA safeguards experience, other approaches that could be
relevant to FMCT include "open skies" (flight
corridors available for verification purposes), and bilateral
or regional access arrangements. Such measures could have
an important confidence-building role, complementing
international verification, e.g. between India and Pakistan, or
in the Middle East.

Verification of HEU produced for naval
propulsion This is often raised as a problem for
the FMCT, because the design of naval fuel is highly
classified. Appropriate verification is required to
ensure this does not become a route for diversion. This
is complex but not insurmountable – the Trilateral
Initiative between the US, Russia and the IAEA demonstrates
the practicability of innovative approaches to verifying
fissile material of sensitive composition, shape and mass.

Verification Intensity

Although some of the methods of IAEA safeguards are readily
applicable to FMCT verification, the treaty objectives are very
different. This would be reflected in the design of the
FMCT verification regime.

Comprehensive safeguards are designed to provide assurance
against horizontal proliferation, i.e. the acquisition
of one or more nuclear weapons by a NNWS. The acquisition
of just one or two nuclear weapons would be a dramatic change
in the strategic status of such a state. The sensitivity
of IAEA safeguards, reflected in technical parameters such as
goal quantities (e.g. the significant quantity of
8 kg plutonium), detection probability, timeliness goals, and
inspection frequency, has been set accordingly.

On the other hand, with states outside comprehensive
safeguards, the concern is essentially vertical proliferation, i.e. additions to existing arsenals. The
FMCT would address this by providing assurance that stocks of
fissile material held outside international verification will
not increase. For states that already hold nuclear
weapons, their concern will be with treaty violations that
could substantially alter strategic relativities: for states
with thousands of weapons, a strategically significant
violation might involve hundreds of weapons. Thus
the design of verification approaches to deter vertical proliferation could be qualitatively different to those
directed at horizontal proliferation.

A major development in IAEA safeguards implementation is a
move away from uniformity to safeguards based on a
"state-level approach", taking into account
appropriate state-specific considerations. This
development remains a work in progress[4] – but the concept is readily
applicable in the FMCT context. Applying a risk-informed
approach, decisions on verification intensity could take
account of considerations such as the following.

Since the NWS will retain military stocks outside the FMCT,
it is reasonable to assume they have concluded, before joining
the treaty, that those stocks are sufficient for their
foreseeable needs. Hence they should have little
incentive to cheat. This is especially the case for the
US and Russia, which are dramatically reducing nuclear weapon
numbers and are committed to irreversibly transferring
substantial quantities of fissile material out of weapons
programs. In these circumstances rigorous verification is
not required. On the other hand, for states with small
arsenals, verification intensity will need to reflect the fact
that small-scale violations could have a serious effect on
strategic relativities.

Conclusions

The US, Russia, France and UK have announced cessation of
fissile material production for nuclear weapons, and it is
understood China has also ceased such production. The
FMCT would make an important contribution to nuclear
non-proliferation, by formalising this situation and making it
irreversible. A further important contribution would be
capping the fissile material available in the non-NPT states,
which are otherwise under no restraint. The FMCT would
also help establish conditions under which further nuclear
disarmament, involving all relevant states, would be possible
– a significant consideration in the context of the NPT
review process.

To achieve these benefits will require the treaty to be
effective, and this requires a credible verification
regime. There is a substantial foundation to build on,
drawing on experience from IAEA safeguards and bilateral
verification arrangements.

Undue delay in concluding a normative treaty can be avoided
by separating negotiation of the principal treaty from
negotiation of the verification system. The latter would
be a largely technical negotiation, which can commence in
parallel with the principal negotiation but be concluded
subsequently.

Whether a particular verification regime provides the degree
of assurance required by the parties - hence is considered
"effective" - is a matter for judgment, based on
many factors: the verification objectives; the verification
methods and standards; related CBMs; other information
(including intelligence) available to the parties;
incentives/deterrents reinforcing compliance; and so on.
Only when we have defined the objectives and main
features of FMCT verification will it be possible to design the
verification system and to judge whether it will be
sufficiently effective – but there seems no in-principle
reason why this should not be the case.

---

John Carlson is Director General, Australian Safeguards and
Non-Proliferation Office. He is also Chairman of SAGSI,
the IAEA's Standing Advisory Group on Safeguards
Implementation. This paper reflects the personal views of
the author and does not necessarily represent the views of the
Australian Government, nor of SAGSI.

---