“It will take all of us working together – government officials, and diplomats, academic experts, and scientists, activists, and organizers – to come up with new and innovative approaches to strengthen transparency and predictability, reduce risk, and forge the next generation of arms control agreements.”
– Wendy Sherman
U.S. Deputy Secretary of State
June 2, 2022
Challenges of Nuclear Stockpile Stewardship Under A Comprehensive Test Ban
Share this

March 1998

By Harold P. Smith, Jr. and Richard S. Soll

The nuclear weapons policy and posture of the United States face unique political and technical challenges as the country balances a requirement to maintain its nuclear stockpile against the obligation and desire to provide strong leadership in arms control and non-proliferation. In a world of uncertain dangers and evolving security needs, nuclear weapons and the robust deterrent deriving from them remain fundamental to U.S. national security, although at a level sharply reduced from that required during the Cold War. As Under Secretary of Defense for Policy Walter B. Slocombe told a congressional panel last year, "nuclear weapons will continue to fulfill an essential role in meeting our deterrence requirements and assuring our non-proliferation objectives" until the conditions for the complete elimination of nuclear weapons in the context of general and complete disarmament are realized.(1)

When President Bill Clinton signed the Comprehensive Test Ban Treaty (CTBT) September 24, 1996, at the UN General Assembly in New York, after nearly three years of negotiations, a new phase of the nuclear age began. The treaty eliminated one of the primary means of maintaining the safety and reliability of nuclear weapons, but the requirement to maintain a high level of confidence in weapons safety and reliability remains. Although the treaty's "early" entry into force may be blocked by certain states whose ratification is necessary (such as India and Pakistan), their possible continued intransigence will have no effect on nuclear testing by the five acknowledged nuclear powers (the United States, Britain, China, France and Russia), which have all signed the treaty. In addition to the principle of international law that signatories should not violate the objectives of a treaty once it has been signed, each nuclear-weapon state has already declared a unilateral moratorium.

One of the key events in the treaty's negotiations came in August 1995, when President Clinton announced that the United States would seek a true "zero-yield" test ban. At the same time, he reaffirmed the vital role of nuclear weapons in national security: "I consider the maintenance of a safe and reliable nuclear stockpile to be a supreme national interest of the United States."(2)

The maintenance of safety, security and confidence in the nuclear deterrent, while the United States continues to provide global leadership in arms control and non-proliferation, will require new and innovative perspectives and processes as the stockpile becomes the oldest in the 50-year history of the nuclear age.

In the absence of underground nuclear test explosions and with no new U.S. nuclear weapons in development, the future arbiter of confidence in the nuclear arsenal will be the Stockpile Stewardship and Management Program (SSMP), conducted jointly by the Department of Energy (DOE) and the Department of Defense (DOD). The program specifies activities to maintain a high level of confidence that the nuclear stockpile will meet DOD requirements. Rather than relying on the empiricism of testing, the program provides a set of initiatives that will promote an understanding of the fundamental sciences of nuclear explosives and the effects of aging on those explosives. The approach is new, parts of it are relatively unfamiliar, and the program will require an estimated 15 years to reach maturity; thus, the United States could incur some risk as it progresses along the learning curve of science-based stewardship. Nonetheless, several SSMP-related initiatives are already underway, including efforts in stockpile surveillance, the evaluation of aging effects and the development of advanced computational technologies.

The SSMP is tied to a new certification procedure. Each year, DOD and DOE must review the stockpile and recertify its safety and reliability. In the event that the secretary of defense and the secretary of energy determine that a high level of confidence in a particular type of nuclear weapon deemed critical to our nuclear deterrent can no longer be certified, the president, in consultation with Congress, would be prepared to withdraw from the CTBT under the "supreme national interest" clause in order to conduct whatever testing would be required. Obviously, the Department of Defense would play a critical role in such a momentous step.

The successful implementation of the new SSMP approach will require an understanding of the issues raised below and more importantly, a sustained commitment by the current and future administrations and Congresses to maintaining the safety and reliability of the U.S. nuclear deterrent while sustaining and promoting a prohibition on nuclear testing. Political skill will be just as important as technical skill, given the U.S. need to balance the disparate capabilities and interests of the recognized nuclear "haves" (the five weapon states); undeclared regional nuclear powers (India, Israel and Pakistan); various aspirants to nuclear weapons status; and the remaining countries. While this latter group embraces the concept of a denuclearized world, its members have an understandable concern regarding their own security in a world of sovereign powers, some of which have or could have nuclear weapons.


A World Without Testing

Articles in the press would have one believe that the debate within the executive branch on a test ban was long, bitter and divisive, with DOD single-mindedly arguing vociferously that low-yield testing equivalent to a few hundred tons of high explosives was essential to the security of the country. In fact, DOD's recommendation was to pursue a comprehensive ban that included the clause in the draft treaty text regarding a party's right to withdraw from the accord if it were to decide that extraordinary events related to the subject matter of the treaty jeopardized its "supreme interests," a right that is included in most arms control agreements. The president had no difficulty accepting the recommendation for a variety of reasons, not the least of which was the thorough briefing by DOE and DOD of all the key decision-makers on the technical advantages and disadvantages of low-yield testing. Rather than bitter and divisive, the road to the decision was an excellent example of good government: collegial, thorough and timely, and the decision itself was sound.

Make no mistake; if the only goal were a technical one—to ensure a reliable stockpile—testing offers the best path for any acceptable level of confidence. However, other political and national security goals, whose importance ranks with reliability, can only be met by not testing. Therefore, the correct question is: Can the United States maintain an acceptable level of confidence at acceptable expense without recourse to nuclear testing? There is good reason to conclude that it can.

First of all, some of the country's best and brightest scientists have been working on nuclear weapons for 50 years. The weapons have been extensively tested and, as a result, there is a comprehensive database that was optimized during the Cold War with the presumption of continued testing and with a new system always on the drawing boards. Now, we are able to draw upon that data, experience and talent. Most importantly, the United States has no requirement to develop advanced new designs of nuclear weapons to increase performance. While the role of nuclear test explosions in developing new, increased-performance designs is essential, it is less important in maintaining the status quo, which includes refurbishing, rebuilding or remanufacturing existing weapons as necessary, or making necessary modifications to improve their safety, reliability and effectiveness. Whether nuclear test explosions can be dismissed altogether for maintaining the enduring stockpile is the challenge that we face.

Britain's experience during its self-imposed testing moratorium from 1965 to 1974 offers a critical counter-example of what can happen even to a very competent party. During the moratorium, the British relied primarily on computer simulation to provide the information they needed concerning their arsenal. However, the advent of ballistic missile defense (BMD) using nuclear weapons on interceptors changed the requirements; the British feared that their warheads would be vulnerable to such nuclear explosions, thus forcing them to design new warheads to meet the new and more demanding environment. Under the 1958 U.S.-British nuclear cooperation agreement, the British revealed to the United States their initial concepts for a new warhead design that was derived solely on the basis of computer simulation. However, U.S. weapons designers, who had already had the benefit of testing warheads under conditions simulating the environment of nuclear BMD forces, concluded that there were performance deficiencies in the British waread design. The British database lacked the new and critical insight that American testing had provided during the British moratorium.

Most experts agree that new designs to increase performance require testing, but is it also true that the United States can maintain an enduring stockpile without benefit of nuclear test explosions? That is what the Stockpile Stewardship and Management Program will address.


The Role of DOD

The Department of Defense must be both a faithful partner and a smart customer of DOE. Of the first, there can be no doubt; the latter remains to be determined. In hearing after hearing before various congressional committees, DOD officials have consistently and effectively supported DOE and its budgets to implement the stockpile stewardship program.

With regard to being a smart customer, it is interesting to note what has changed insofar as the nuclear stockpile and nuclear test explosions are concerned. In the past, acceptance of DOE's certification that the weapon was reliable was based on the very evident, indeed awesome, results of a nuclear explosion. Now, DOD will have to become a smart customer, one possessing the expertise to examine and question DOE reports of simulations, comparisons with archival data and non-nuclear experiments in much the same manner as a prime contractor must judge the performance of highly specialized subcontractors. The country cannot and should not provide duplicate capability in each department, but DOD must be able to accept responsibly or, if necessary, reject authoritatively a DOE certification of a particular nuclear weapon. A failed nuclear test would have been more than sufficient in the past. It will not be available in the future.

The nuclear weapon is the only system DOD procures for which it does not already possess the requisite "in-house" technological expertise. That situation will have to change, but it is an easier change today than previously. Veteran designers from the weapons laboratories will be available, as they are today within DOE, to lend their expertise to DOD's portion of the certification process. In the longer term, DOD must ensure the cultivation and availability of a new generation of nuclear weapon engineers trained in warhead design, that is, engineers capable of providing oversight of suitability of the warhead provided by DOE.

To accomplish this, DOD personnel will participate in the forensics, by which weapons are taken from the stockpile and dissected in order to determine whether they are aging predictably and/or acceptably. Under a process called "dual revalidation," DOD personnel will work closely with their DOE colleagues on two independent teams, each led by personnel from one of the two weapon design laboratories (Lawrence Livermore National Laboratory and Los Alamos National Laboratory) to ensure mutual, thorough and independent understanding of each nuclear weapon system. Thanks to the good sense of former Secretary of Energy Hazel O'Leary, who set aside a potentially damaging recommendation of the Galvin Committee,(3) there will continue to be two weapon design laboratories to ensure independence and to provide directly to DOD two unabridged reports on each weapon system.

Hopefully, from this kind of peer review, with its two independent reports to an educated and cooperative customer, we will understand whether our weapons will continue to be safe, secure and reliable. Peer review may not be a substitute for testing, but given the political importance and advanages of banning nuclear test explosions, it is the next best thing. The United States has successfully encouraged its allies to do the same.

The Department of Defense is depending on DOE to develop and maintain the forensics to identify and evaluate potential safety and reliability problems in U.S. weapons, and to build and maintain the facilities and capabilities to refurbish, rebuild or remanufacture the weapons as required. These are not easy tasks: some materials and manufacturing techniques will not be acceptable or will be unavailable, and original vendors will have disappeared. Nonetheless, despite tight budgets and possible complications incurred in remanufacture, the two departments must find a way to provide the necessary replacements and must ensure that the exercise of the supreme national interest clause, under which the president can decide to forgo the terms of the CTB Treaty and resume testing, will not be necessary.

To carry out this mandate, DOD will support whatever DOE budgets are required, but DOD should not provide the funding. Maintenance of the nuclear arsenal, for sound and long-standing reasons, has never been the province of the military. It should not become so today.


Deterrence Under a CTB

Deterrence, unlike art, is not solely in the eye of the beholder; it is also in the eye of the holder. Deterrence is as much dependent on the will and confidence of the deterrer as on the fear of the deterree. If the government were to lose confidence in the reliability and safety of its stockpile and if such loss of confidence were to become known, deterrence could vanish. Therefore, one of the principal vulnerabilities of dual revalidation, strangely enough, is the would-be whistle blower who claims that a given weapon is unreliable or unsafe. Unlike Cassandra, who was correct but not credible, one worries about the malcontent or the misinformed who is credible but not correct in claiming that a given weapon will not work or is not safe. Such news will inevitably get into the media and justifiably cause public concern. Eventually, candidates for office may play on public anxiety by using these allegations in much the same way that the purported but non-existent "missile gap" distorted the presidential election of 1960.

At the same time, the United States will not have the ability to test and thus to show whether the whistle blower is a malcontent or a Cassandra. The ensuing debate could degenerate into the world of the "schoolmen" of the Middle Ages where, this time, the debates will not be over angels dancing on heads of pins but over endless computations, and the nuclear weapons experts, upon whom the government must rely, will lose the confidence of the public and the national leadership. While the concern is real, there remains, nonetheless, no other way than to maintain a steadfast commitment to competent, independent review by peers and knowledgeable evaluation of those reviews by the customer. To be forewarned of the problem is, perhaps, to be forearmed.


Debunking Misconceptions

There are two misconceptions clouding the CTB debate concerning U.S. relations with the other nuclear-weapon states. The first is that the French would not have had to conduct their most recent series of tests (during 1995 and 1996) in the South Pacific if the United States had been willing to provide them its computer simulations. The second is the assumption that because the United States has superior computational capability, the CTBT will give the United States a clear advantage over Russia in an era without testing.

Allegedly, the French would not have needed to test nuclear weapons in the South Pacific if the United States had provided them its impressive computational models, which, it is claimed, have become the archive of all the information, all the knowledge, that one needs to develop and maintain nuclear weapons. The assertion is simply not true. The models or computer codes are only as valuable as the databases upon which they are based. Application of U.S. codes to the French database would be of little value to the French, nor would their codes be of any significant benefit to the United States.

The second misconception presumes that the absence of testing will give the United States a significant advantage over Russia. This seems highly unlikely, however, because the challenge of maintaining a nuclear deterrent and adhering to the CTB Treaty is essentially the same for Washington as it is for Moscow. The United States may have begun development of nuclear weapons a few years before the Russians, but more than 50 years have now passed with the result that any U.S. lead has disappeared. Russia's best and brightest scientists have been working on its weapons for the same half century—more than enough time to overcome the U.S. head start, which itself was essentially nullified by effective espionage. The Russians have extensively tested their weapons, with the result that they have a comparable database. The allegation that the United States has a lead in computation is probably true, but computation is by no means everything. Cleverness still counts, as do mathematical and physical insight, and the United States has no monopoly on such insights.

It is important to understand, further, that computational capability is only one of the tools available to nuclear-weapon states for maintaining a high level of confidence in their respective nuclear weapon stockpiles. Other tools include rigorous programs of weapon inspections, laboratory tests and experiments, simulators, and other theoretical and experimental research. In addition, the capability to produce, refurbish and remanufacture nuclear weapons has been and will continue to be, under the CTBT, indispensable to maintaining confidence in the stockpile. This entails the active participation of creative individuals with a thorough understanding of the design and development of nuclear weapons. Russia has outstanding capabilities in all of these areas.

This brings us to the fundamental question: Will the CTBT impact equally U.S. and Russian capabilities to maintain their respective stockpiles? Historically, U.S. practice had been to design, develop and deploy a new warhead for each new weapon system. Given the evolving national security requirements of the time, such replacements occurred well before the end of the useful stockpile lifetimes of the replaced warheads, thus allowing the United States to avoid, for the most part, age-related stockpile problems. In addition, the United States resolved some serious stockpile problems by developing new and improved replacement nuclear warheads which, of course, required testing, sometimes of an extensive nature. Testing was also used to answer definitively questions regarding weapon performance that could not be resolved by other means.

Whether the Russians depended on nuclear testing to maintain confidence in their stockpile to the same extent as the United States is difficult to say. There is reason to believe that they did ot. We think that the Russians ensured stockpile reliability through conservative weapon designs that included lavish use of fissile material and high explosives and by remanufacturing nuclear weapons at periodic intervals before age-related problems appeared. In summary, one can conclude that over the years the Russians (as well as the other nuclear-weapon states) have developed effective tools that work for them to ensure the level of confidence they require in their nuclear weapons under a CTBT.

The United States must find ways to discourage all parties from building the massive arsenals that were held by the United States and the Soviet Union during the Cold War. The world does not need another half century of anxiety about the triggering of Armageddon, either on a global level or regionally.

The CTB Treaty, as a corollary to the nuclear Non-Proliferation Treaty (NPT), is a required step. The United States could not have argued so strongly during the NPT extension conference to make the treaty permanent if it had not committed itself to no further testing. The United States was placed in that position by moral and pragmatic considerations, and the president's decision on a test ban was the correct one.

Using the various tools which our scientists have developed and continue to develop, the United States will be able to ensure through the Stockpile Stewardship and Management Program that its stockpile is eminently safe, secure and reliable. The United States should do everything feasible—as this administration is doing—to persuade the other four declared nuclear powers and the other countries of the world that the CTB Treaty can enhance the security of us all. Adherence to the treaty is in the best interests of all sovereign powers, even though the nuclear terrorist may remain undeterred no matter how reliable or large our arsenal. By never losing sight of our vital national security interests and the global requirement for nuclear stability, the United States can make the treaty a reality while, at the same time, making the maintenance of a robust deterrent as certain as science will permit.


1. Statement before the Senate Governmental Affairs Subcommittee on International Security, Proliferation and Federal Services, Hearing on Nuclear Weapons and Deterrence, February 12, 1997. [Back to (1)]

2. Statement by President Bill Clinton released by the White House, Office of the Press Secretary, August 11, 1995. [Back to (2)]

3. Robert Galvin, Secretary of Energy Advisory Board, "Alternative Futures for the Department of Energy National Laboratories," February 1995. [Back to (3)]


Harold P. Smith, Jr. was assistant to the secretary of defense for nuclear, chemical and biological defense programs from June 1993 to January 1998, where he was responsible for nuclear matters, including stockpile management and implementation of the CTB Treaty. Richard S. Soll is senior scientist and head of special projects at Science Applications International Corportation's Center for Verification Research in Springfield, VA.