全球智库信息集成服务系统

Summary

There is no clear, internationally accepted definition of what activities or technologies constitute a nuclear weapons program. This lack of definition encumbers nuclear energy cooperation and complicates peaceful resolution of proliferation disputes. A “nuclear firewall” could enhance the distinction between nuclear weapons–related activities and other non-weapons uses of nuclear technology. Applying a firewall framework for analyzing nuclear programs could improve international governance of nuclear technology and facilitate peaceful nuclear cooperation and disarmament. It could also expand the time and means available to key states and international bodies, such as the International Atomic Energy Agency and United Nations Security Council, to diplomatically resolve impending proliferation crises.

Defining Nuclear Weapons

The Nuclear Non-Proliferation Treaty (NPT), which establishes the norms and rules that guide the international management of nuclear technology, does not define the term nuclear weapon. Nor does it identify the evidence that would determine whether a state is seeking to manufacture or acquire nuclear weapons.

Such definitional and analytic ambiguity exacerbates the task of distinguishing whether components, equipment, nuclear materials, and facilities are related to nuclear weapons programs or, instead, are for purely peaceful applications of nuclear technology. It also complicates national and international deliberations over the legitimate boundaries for peaceful civil nuclear applications, as well as the handling of proliferation risks and responses.

The Carnegie Endowment for International Peace has developed an analytic approach—a nuclear firewall—to help distinguish between activities and programs that are purely peaceful and those that merit definition as nuclear weapons–related. Like an information-system firewall, the nuclear firewall would

• identify those activities, materials, and equipment that should be inhibited because they are purely or strongly associated with nuclear weapons programs;
• distinguish activities that should be facilitated because they are fully consistent with peaceful applications of nuclear technology and know-how; and
• assess in-between activities depending on transparency and reassurance measures that states would undertake.

The Firewall Framework

To develop the firewall, the Carnegie team worked with leading technical and policy experts from nuclear weapon and non-nuclear-weapon states to identify pertinent indicators and contextual factors that demarcate peaceful from nuclear weapons activities. Tested against historical case studies, the endeavor yielded a framework that can withstand challenges posed by the lack of complete and accurate information about a country’s activities and the possibility of deliberate efforts at deception.

The process predictably revealed that (1) technical parameters alone are rarely sufficient to draw reliable assessments about the overall orientation of a country’s nuclear program; and (2) in some cases, the absence of certain activities, items, or policies can better indicate whether a program’s stated purpose is indeed peaceful.

The resulting multidimensional framework—designed to be country-neutral, transparent, and easily employed—has several features:

• Evaluates the presence and absence of activities, equipment, materials, patterns of behavior, and the broader context. The firewall assesses whether these elements—individually and collectively—are compatible with the purposes states proffer for them. The identity of a country should not prejudice analysis of its nuclear program.
• Provides insights into the nature and direction of nuclear programs and helps assessment of potential proliferation concerns. The firewall can suggest which combinations of particular actions and other indicators should, over time, increase or decrease the sense of assurance or level of concern about a given state’s nuclear-related activities, which could inform discussion of such concerns in national or multilateral settings.
• Augments effective implementation of all three pillars of the NPT, namely nonproliferation, the peaceful use of nuclear energy, and disarmament. By helping users to systematically identify weapons-oriented activities and enhance policy options for redressing them—the firewall can facilitate legitimate applications of peaceful nuclear energy and highlight ways that states can reassure others of the peaceful orientation of their programs. Further, by identifying comprehensive indicators of nuclear weapons programs, it addresses a necessary condition for progress toward disarmament. For without a technologically detailed template for defining how to turn military nuclear programs into purely peaceful ones, nuclear disarmament will not fully enhance security and therefore will not be politically achievable.

While elements of the framework have been vetted with distinguished international experts, governments and civil society should study them and consider which might merit further technical development and which might be ripe for implementation in multiple institutional settings. A firewall application could be employed by individual states assessing proliferation risks and making decisions about strategic trade controls, by international organizations and in multilateral forums, and by academic institutions and the nonproliferation community more broadly. Carnegie stands ready to assist with these efforts.

Introduction

Atomic energy, with its potential for peaceful and military applications, emerged internationally in the 1940s. In subsequent years, states incrementally established institutions and rules for managing this potential, such as the International Atomic Energy Agency (IAEA), created in 1957. In 1968, the Nuclear Non-Proliferation Treaty (NPT) opened for signature, and it came into force in 1970.¹ To motivate states to join the treaty, its principal authors, the United States and the Soviet Union, agreed that the obligations to stem weapons proliferation needed to be matched with obligations to facilitate access to peaceful applications of atomic energy and progress toward nuclear disarmament. These obligations were reflected in Articles I, II, and III of the treaty (nonproliferation), Article IV (peaceful nuclear cooperation), and Article VI (disarmament). By 1995, when the terms of the NPT called for an international conference of parties to decide whether, and under what conditions, to extend the treaty, 178 states had joined. At the 1995 Review and Extension Conference, the parties decided to extend the treaty indefinitely. Today, there are 190 NPT parties.

Inevitably, much has changed in the world since the treaty’s enactment and extension. The circumstances in which the nuclear order must be maintained in the coming decades have changed dramatically and in ways that the NPT and other multilateral instruments could not anticipate.

The NPT and related agreements, instruments, and arrangements—such as the IAEA, the Zangger Committee, and the Nuclear Suppliers Group—remain invaluable foundations of nuclear order. However, innovations and new developments in technology sometimes arise and, inevitably, take time for the existing order to comprehend and manage. Similarly, institutional innovations take time to become widely implemented. For example, while progress has been made over the past two decades to introduce and promote the IAEA Additional Protocol, the goal of making it universal remains elusive. Other efforts to enhance these arrangements, such as the IAEA’s advancement of the State-Level Concept, have proven contentious. In this situation, some states have been tempted to act individually or react to address what they perceived to be challenges and/or opportunities for gain in the nuclear order. Yet such uncoordinated action frustrates efforts to build consensus and may undermine the goals of the NPT.

Leading states in the international system could significantly strengthen the prospects of fulfilling the NPT’s bargains on peaceful nuclear cooperation and disarmament if they cooperated in devising more comprehensive means for assessing and addressing proliferation risks. For such cooperation to be feasible, useful, and internationally accepted, it must be based on solid scientific and engineering knowledge and experience, as well as transparent analytic tools that could be applied consistently and generically to all states. Less transparent and comprehensive assessments of risks, and/or favoritism in addressing them, will over time not only fail to manage emerging challenges but also erode the foundations of the existing order and further exacerbate stresses in the regime.

The Nuclear Firewall Concept

With these premises, the Nuclear Policy Program of the Carnegie Endowment for International Peace engaged external experts in a multiyear, low-profile brainstorming exercise to explore whether we could reliably identify generic activities or programs aimed at the development of nuclear weapons—which is prohibited in principle by the NPT—and offer options for employing such analysis in a menu of policy applications. The effort was inspired by the conviction that a clearer divide between weapons and non-weapons nuclear activities would simultaneously facilitate more effective implementation of the NPT’s three pillars (nonproliferation, the peaceful use of nuclear energy, and disarmament).

This project was stimulated in part by recognition that the NPT does not define what a nuclear weapon is and therefore what constitutes nuclear weapons development activity prohibited under the treaty. The resulting concept of the project—a nuclear firewall—suggests a way to address this omission by helping to draw a line between activities that are exclusively weapons-oriented and those that are peaceful or for dual use. The firewall first seeks to identify activities or combinations of activities that experience suggests are either uniquely useful for the purposes of developing, acquiring, and sustaining an operational nuclear weapon capability, or whose most plausible use is in support of such an aim. Such activities are in the domains of research and development, testing and engineering, acquisition of materials, components and systems, construction of facilities for these purposes, and development of a military infrastructure (both physical and human) to employ nuclear weapons. This focus on weapons-related activities derives from the NPT’s clear requirement that non-nuclear-weapon states refrain from manufacturing or acquiring nuclear weapons. Though individual indicators by themselves are rarely definitive, a program whose multiple activities occurring together reflect a cohesive effort to acquire a nuclear weapon would be in clear violation of this requirement.

For dual-use activities (that could serve military and/or civilian purposes), the firewall seeks to identify observable or otherwise measurable technical and contextual means and trends to distinguish activities that may confidently be regarded as having a nuclear weapons orientation from those that appear to be compatible with their stated non-nuclear-weapons purposes. In cases where the exclusivity of an activity for legitimate (non-nuclear-weapons) purposes is questionable, it identifies lines of inquiry about the alternative scientific, technological, and commercial explanations that could help clarify whether an application other than nuclear weapons development is plausible.

Two Analyses: The Parts and the Whole

These lines of inquiry into alternative explanations address two levels of analysis (see figure 1 on page 10). The first is to assess whether individual activities have rationales that can plausibly justify, for example, their technical and/or commercial parameters, the materials employed, and the quantities involved. The second is to assess whether the program or enterprise to which a specific activity is related has the attributes of a credible program. To illustrate the second level of analysis in the domain of nuclear energy, the firewall project has delineated a standard set of traits, or a template, for a nuclear power program. That is, we identify activities and undertakings that comprise the common practices (technical, legal, institutional, and regulatory) of virtually all genuinely peaceful nuclear programs. This template of activities and behaviors provides a basis for comparison with any particular program. Consistency with this template, or deviation from it, permits assessment of the plausibility of alternative explanations when assessing other countries’ nuclear activities and programs.

The first level of analysis—assessing how much concern exists about a certain item or activity in connection with a nuclear weapons application—is routinely conducted by the five nuclear weapon states under the NPT (as well as some other states), usually on a national basis and privately. This concern may actually be indicative of a clandestine weapons program. The second level of analysis—assessing the peaceful credibility of an entire program—however, is an innovation of the nuclear firewall concept and could be developed further as a standard practice for proliferation assessment.

A central, related innovation is the identification of categories of rationales that might be offered to justify (but also provide cover for) certain activities indispensable for a nuclear weapons program. In addition to the template of a peaceful nuclear energy program, the firewall concept highlights the utility of further expanding analysis in the areas of nuclear energy; ballistic missiles; and nuclear weapons research, development, and deployment. This effort, which requires the expertise and resources of states, would include the mapping of plausible legitimate explanations for activities in each domain and the identification of the requirements and conditions for such applications to be deemed credible. The most immediate benefit of developing this analysis is diagnostic. When and where indicators of concern emerge, the provision of rationales could either generate reassurance that the activities are indeed legitimate and deserving of support, or conversely, reinforce concerns that they are inconsistent with legitimate applications and thus merit further investigation. The aim is to suggest an internationally acceptable basis for engaging in this sort of analysis and inquiry in both intragovernmental settings and intergovernmental dialogues between the IAEA and its member states, within the Zangger Committee and the Nuclear Suppliers Group, or between states and nongovernmental organizations (NGOs) in other institutions and contexts.

This analytical approach of assessing not only a state’s activities but also possible omissions in its technical development, activities, or practices (compared with the templates of legitimate applications of nuclear technology) is a significant value of the firewall. The complementary lines of inquiry described above permit a more comprehensive analysis of nuclear programs than traditional proliferation assessment techniques.

The Five Cs Approach

Consistent with the IAEA’s obligations pursuant to comprehensive safeguards agreements, the IAEA assesses the completeness and correctness of declarations regarding a country’s nuclear activities. The firewall project seeks to provide an additional analytic framework to complement the IAEA’s safeguards work by increasing the breadth and depth of proliferation analysis from two to five Cs (hereafter, the “5Cs”). We add considerations of compatibility, cohesion, and consistency. Compatibility refers to the degree to which individual activities undertaken are indeed commensurate with their stated peaceful purposes, including the sequencing and scale of activities and economic rationale underlying them. Activities that are incompatible register higher concern and may indicate weapons orientation. Cohesion refers to the extent that diverse, individual activities undertaken by a state are interconnected and reflect most of the known elements and pathways of a nuclear weapons program. Finally, consistency refers to the degree to which attributes of a nuclear program are consistent with the established hallmarks of credible civil or nonexplosive applications of nuclear technology (for example, in safety, security, environment, and liability).

The 5Cs describe the firewall project’s holistic approach to assessing nuclear programs. Such an approach can help establish a legitimate basis for evaluating concerns regarding any given nuclear program, as well as provide reassurances about its nature. The 5C approach could aid discussion on potential cases of proliferation and help develop policy options to assuage concerns that might arise from looking at individual activities. Furthermore, it provides an opportunity to do so as a complement to national or multinational proliferation assessments.

Potential Implications and Applications

The project’s comprehensive approach to proliferation analysis has several important implications for efforts to strengthen nuclear governance. By clarifying the interdependence of Articles II and IV, this approach could advance the objectives of all three pillars of the NPT. For example, we considered whether a program that closely conforms to the patterns, scope, synchronicity, and scale of established peaceful programs and exhibits no indicators of nuclear weapons could attract increased international cooperation (for as long as this characterization holds). On the other hand, a state conducting activities that depart from standard practices of peaceful programs—including, especially, activities associated with weaponization—would be more readily assessed as transgressing or at a minimum encroaching on its obligation not to acquire or manufacture nuclear weapons. In such cases, states undertaking more sensitive activities might voluntarily reassure the international community of their commitment to peaceful obligations. The insights developed through this iterative analysis were subsequently tested using historical and contemporary case studies of proliferation and peaceful nuclear development in order to refine their conclusions.

Naturally, extensive discussions were required to establish how the firewall could be used to assess both indicators of proliferation warning and of peaceful orientation. The resulting analytic framework accounts for and weighs such indicators statically and over time to establish whether any given country is moving closer to or further away from nuclear weapons development and acquisition. Whether utilized in institutional or diplomatic contexts, this framework could facilitate a process of discussion or clarification.

The firewall project consciously refrained from prescribing how states individually or collectively should respond to concerns that may arise from this analysis. The linkage between assessment and policy prescription is left open for further discussion in the context of how the firewall could be applied in the future. However, to be effective, the analysis must be timely, recognized as technically valid, country-neutral, reliable (it would yield broadly similar results based on the same data set or information regardless of who used it), and robust (it employs all the indicators necessary to reach a valid conclusion). The analysis should also have built-in resiliency to defeat efforts to game it by determined proliferators, as well as to protect against biased or undependable results. In a complementary way, this analysis could inform states’ efforts to reassure others that their nuclear programs are peaceful.

Governments, NGOs, academic institutions, and even private enterprises could use the firewall in a range of applications. Some potential analytic and diagnostic applications include informing multilateral deliberations on issues of proliferation concern and helping to guide national and multilateral strategic trade implementation and outreach. Some potential constructive or normative applications could include informing future actions that would strengthen the delineation between nuclear weapons activities and the rest.

Such applications, all of which would require further explication of the requirements, as well as development and testing, are discussed later in the paper. To enhance the political and technical acceptability of these applications, an innovation such as the firewall should have the following attributes:

• Scope: cover all three pillars of the NPT.
• Application: consistently and uniformly apply to all NPT non-nuclear-weapon states, with an explicit effort to enhance the legitimacy and political acceptance of the firewall concept and tool.
• Country neutrality: utilize to the maximum possible degree established, generic criteria, indicators, procedures, and institutions without prejudice or favor to the state involved.
• Clarity: explicitly delineate the boundaries between rights and obligations, and in practice, impose limitations and conditions on certain types of activities.
• Transparency: make the general principles public, even if some sensitive technical details and applications are withheld exclusively for official governmental use.
• Robustness: include all relevant known indicators to make the firewall as resilient to the evolution of technology as possible, as well as to prevent efforts to game it by potential proliferators. This will require periodic reassessment and refinement.
• Parsimony and user-friendliness: employ only those parameters necessary to generate valid conclusions and provide user manuals that enable them to be easily and effectively employed.
• Modularity: allow for the application(s) to be employed at different levels of resolution, with varying degrees of access to information.

The Firewall Framework

The firewall distinguishes activities and patterns of behavior in non-nuclear-weapon states that raise degrees of concern about nuclear proliferation, as compared with activities and patterns that are unambiguously peaceful. The former would obviously warrant closer monitoring, possibly alongside further steps by the state to reassure the international community about the peaceful purpose of its activities.

The firewall looks at both the presence and absence of activities and their evolution over time as indicators of purpose. It systematically identifies and weighs indicators and patterns of behavior. This approach diminishes challenges associated with the availability of comprehensive and valid information. It provides a holistic picture of a weapons program (as opposed to, for example, looking only at fissile material production). This approach also avoids the pitfall of technological determinism, which may arise in using methodologies focused on critical path analysis. Historically, many proliferators have pursued multiple parallel approaches to nuclear weapons development, thus underscoring the need to assess a broad range of activities over time. Augmenting this descriptive analysis, the firewall framework also utilizes a template of a credible nuclear power program to assess the compatibility and consistency of a state’s nuclear activities. (As discussed later as an area of potential future development, a thorough assessment of all plausible non-nuclear-weapons applications of relevant technologies and activities and the incorporation of such templates would greatly enhance this framework.)

The results of a firewall analysis as it is applied to each case can be represented along a color spectrum ranging from green to yellow to red. This coloring scheme can be used in two ways: to delineate and describe the character of particular activities, either singularly or in combination; and to illustrate the sum output of all of the indicators and patterns, which provides an overall assessment of the state of a nuclear program at any given time. An activity or capability that registers green is peaceful. Dual-use activities are generally considered yellow, and a program whose capabilities and activities could be utilized for peaceful and/or military purposes is coded yellow. Finally, an activity that has exclusively weapons-oriented applications is coded red. A program registers as red when its collective capabilities and patterns of activities are uniquely, or strongly, associated with a nuclear weapons program. When looked upon comparatively over time, this coloring scheme portrays not merely the status of any given nuclear program but also its trajectory toward or away from nuclear weapons. Pictographically, clusters of activities and other patterns of behavior as viewed through the firewall are captured in a pyramid (see figure 2).

The green-yellow-red representation of the output signifies three of the firewall’s major policy contributions: (1) to provide a basis for determining that particular capabilities and activities—alone or in combination—are inconsistent with the obligation not to seek to acquire nuclear weapons; (2) to provide warning that a nuclear program is trending away from legitimate peaceful purposes toward illegitimate military ones—that is, from green to yellow or red; and (3) to identify patterns of behavior that could be modified to enhance confidence in the peaceful intentions underlying any given nuclear program.

The purpose of the firewall is not simply to assess where states fall on this color spectrum, but also to reveal concerning trends and patterns in a state’s activities well in advance of reaching the red zone. To this end, a proliferation warning and reassurance application of the firewall could be built