The Nuclear Safety Smokescreen

By: Hisham Zerriffi

U.S. advocacy of a "Comprehensive Test Ban" (CTB) treaty to end all nuclear explosions is tied to the start-up of a major new initiative called the Science Based Stockpile Stewardship (SBSS) program. The SBSS program will allow the U.S. to retain a large number of nuclear warhead designers for an indefinite period. It is also tied to maintaining the Nevada Test Site in a state of permanent readiness to resume testing should the United States decide to withdraw from the CTB under a "supreme national interest" clause that it wants built into the treaty. The Department of Energy's SBSS program would replace underground nuclear testing with a combination of above-ground experimental facilities and advanced computational abilities. These new facilities are justified by the Department of Energy (DOE) on the basis of ensuring the continued "safety and reliability" of the nuclear weapons arsenal as it ages. However, as IEER discusses in its new report, "The Nuclear Safety Smokescreen," there are two fundamental technical problems with the DOE's justification of the Science Based Stockpile Stewardship program. First, the DOE often does not distinguish between safety and reliability. Second, the DOE implies that aging will have a significant effect on the safety of nuclear weapons. For instance, the DOE has stated that "The effects of aging on weapons components can affect their long-term safety and reliability. Safety may be affected by chemical or structural changes in the HE [high explosives] or detonators, which may lead to altered response to impact or fire." To examine the DOE's claims for the SBSS program, we decided to go back to basics, starting with definitions of safety and reliability. Quite simply, safety is making sure warheads don't blow up when you don't want them to, while reliability is making sure warheads do blow up when you want them to. Moreover, while safety is purely a technical issue, reliability also has political and military (strategic) aspects. For example, the type and level of reliability required for retaliation to a nuclear attack (a deterrence strategy) differs substantially from the reliability required for a first nuclear strike against a heavily armed nuclear adversary. Using DOE documents specially compiled through a Freedom of Information Act request, IEER analyzed data relating to the types of safety and reliability problems that have been experienced with warheads in the past, what types of problems the DOE expects for the future, and how these relate to the types of facilities the DOE plans to construct as part of its Science Based Stockpile Stewardship Program. Safety Problems Two key parts to a nuclear warhead are the "primary" and the "secondary." The primary is the first stage of the nuclear explosive which contains both high explosives and one or both fissile materials, plutonium-239 and highly enriched uranium. The secondary contains both thermonuclear (fusion) and fission components. Though both are important to warhead safety, by far the most crucial component to the nuclear safety of a warhead is the primary. DOE data show very clearly that there have never been aging-related safety problems associated with either the primary or the secondary of a warhead. Defects with these nuclear components, collectively called the "physics package," were design-related -- they were not problems due to aging. (See box below.) Causes of Warhead Safety Problems Caused by Aging Cause other than Aging Nuclear Component Problems: Affecting Primaries: Affecting Secondaries: 0 0 38 2 Non-Nuclear Component Problems: 8 18 While primaries do deteriorate with age, this deterioration affects the reliability of the warhead, not its safety. According to DOE data, aging only affected the safety of non-nuclear components of warheads, such as the parachute system, gas transfer system and radar. (See Centerfold.) These findings suggest that the SBSS facilities, which are designed mainly to investigate parts of the physics package, would have little or no relevance to maintaining the safety of the nuclear arsenal. The non-nuclear components of the warheads that had aging-related safety problems are ones that can be functionally tested apart from the warhead, even when re-designed. This capability is the responsibility of Sandia National Laboratories and would not involve SBSS facilities. Additionally, the National Ignition Facility (NIF), a laser fusion facility to be constructed at Lawrence Livermore National Laboratory in California, is not designed to study non-nuclear components or primaries (except during late stages in the explosion, at which point safety is a moot point). Furthermore, this facility operates at volumes which are much smaller than a nuclear explosion. As a result, all information obtained from this facility must be scaled up, a difficult process when applied to existing warheads. In New Mexico, the Dual Axis Radiographic Hydrodynamic Test (DARHT) facility is being constructed at the Los Alamos National Laboratory. This facility is intended to study the implosion of primaries. However, this may not be necessary, considering the historical data on safety problems with primaries. All warheads in the arsenal are certified to be "one-point safe"1 ( an important measure of safety. Therefore, further testing seems unnecessary as long as the primary undergoes no modifications. Because modification could introduce uncertainty in warhead safety and reliability, a number of experienced analysts have specifically recommended against modifying nuclear components. By focusing on the nuclear components of warheads, the DOE seems to be ignoring this advice. While the utility of these facilities in maintaining the safety of existing nuclear warheads is highly questionable, they could add significantly to the weapons design capabilities of the United States. Facilities similar to the National Ignition Facility and the Dual Axis Radiographic Hydrodynamic Test facility have been used in the past as part of the weapons design program. The tables in the Centerfold describe SBSS facilities and their potential for designing nuclear warheads. A variety of official documents discuss weapons design capabilities for the period following the signing of the CTB. Perhaps the most striking example is testimony by C. Paul Robinson, director of Sandia National Laboratories, before a Senate committee in March of this year. Robinson refers to the CTB and the halt in new weapons production only as a "hiatus" that could be several decades long, noting that children "entering kindergarten this year" will be the future engineers and scientists designing the next generation of weapons systems.2 In this view, new warheads will eventually be designed and built. Reliability Problems DOE has defined "reliability" very narrowly, so that even a small chance of a slight decrease in performance, either with respect to yield or target accuracy, is considered a reliability defect. This strict definition only seems relevant if the purpose of the nuclear arsenal is to issue a first strike to destroy an adversary's nuclear arsenal. In such a strategy, high accuracy and yield at or above the rated value may be necessary to destroy strategic missiles stored in "hardened" silos. But such a strict definition of reliability is not relevant to a deterrence strategy based on retaliation in response to a nuclear attack. Nuclear weapons are so devastating that the possibility of a small decrease in yield or accuracy below design values would not affect the decision of an aggressor to launch a nuclear attack. So far as IEER has been able to discern, only 12 of 186 problem types identified in the data may be relevant to a strategy based on retaliation to a nuclear strike. The vast majority of reliability concerns seem to connect to a first strike strategy. Yet neither the DOE nor the Pentagon have put their decision regarding the SBSS program in this context. However, this conclusion is very tentative and is only indicated by the data on reliability defects. We cannot arrive at a definitive conclusion because the requisite data have not yet been made public. DOE Omits Other Options The inherent technical design capability of the SBSS program is only one of the many indications that the DOE intends to go beyond the maintenance of the current weapons stockpile. In its Draft Programmatic Environmental Impact Statement (PEIS)3 on Stockpile Stewardship and Management, the DOE explicitly ruled out re-manufacturing or maintenance as alternatives to the costly SBSS program, though it had not given these options due consideration. We briefly describe these options here: Re-manufacturing: The DOE states that precise replication is not always possible and therefore re-manufacturing is not a reasonable alternative. Additionally, the PEIS states that the emphasis of the SBSS program is on "nuclear components which can no longer be functionally evaluated by nuclear tests." The elimination of this option fails to take into account several key points: A number of experts, including former nuclear weapons designers, such as Ray Kidder, J. Carson Mark, and Richard Garwin, have stated that re-manufacture is a reasonable method to maintain the nuclear arsenal after a Comprehensive Test Ban.4 The PEIS does not consider the implications of a recommendation that "fixes" to the primary should be avoided, even if meant as "improvements." Kidder, Mark, and Garwin, among others, have stated that re-manufacturing is preferable to changes in the "physics package" of warheads. The failure to consider systematic ways to deal with safety and reliability issues arising from non-nuclear components is an egregious omission because such problems could be solved by re-manufacturing. Maintenance: This approach is the most similar to the proposed SBSS program. The major difference seems to be that new experimental facilities would not be constructed, while surveillance of weapons would be enhanced. Existing experimental facilities would continue to be used. Eliminating this option makes it clear that this PEIS is biased towards one outcome: the construction of new facilities that would expand design capability. DOE has also not considered an intensification of the Stockpile Evaluation Program, even though it has relied mainly on this program to discover warhead safety and reliability problems. This program is distinct from DOE laboratory capabilities and has been in place since 1958. It was designed to monitor warheads during production and after deployment and to fix any problems that were found. The program withdraws new components and complete warheads from the production line and from deployment. The samples varied over the years both in number and also according to whether the warhead was in production or was deployed. Currently, the program consists of withdrawing approximately eleven warheads of each type from the stockpile every year. Ten of them are system tested, reassembled, and returned to the stockpile. The conventional explosive in the eleventh is destructively tested. Over the decades, the DOE has created a database of problems found with the nuclear arsenal. A majority of these problems (75%) were discovered during the Stockpile Evaluation Program, while the rest were discovered during research and development, through underground testing, and through a variety of other methods. DOE could intensify this program by withdrawing a larger number of warheads for inspection. If new safety-related problems arise in the future due to the aging of warheads beyond their design lives, as postulated by the DOE, then an intensified Stockpile Evaluation Program would seem to be more appropriate than the SBSS program. Yet DOE has not considered this alternative, much less done an analysis showing that the SBSS program would better serve to maintain the safety and reliability of the existing arsenal. Conclusion By eliminating reasonable alternatives from its programmatic environmental assessment, such as the ones discussed above, the DOE has indicated its determination to build new facilities regardless of their relevance to safety and reliability. At the most basic level, the DOE has simply failed to show why it needs to operate its existing design facilities or to construct new experimental facilities for maintenance of the existing arsenal. The DOE already operates a number of facilities of a similar nature, albeit not necessarily as advanced as the new facilities. If the DOE wishes to pursue the SBSS program it has to show a need for these facilities based on specific problems it expects. Principal Findings of "The Nuclear Safety Smokescreen" The Department of Energy's analysis of the need for an SBSS program confuses safety and reliability ( issues which are technically distinct. Consequently, DOE has not justified the need for SBSS facilities as they relate to safety issues, separately from their relevance to reliability issues. The DOE has not related the types and levels of reliability required of the arsenal to overall U.S. military strategy. DOE data show that the nuclear detonators in nuclear warheads (the primaries) have never had safety problems linked to aging. The data clearly indicate that SBSS facilities are not needed for safety of the nuclear package. The SBSS program will give the U.S. powerful capabilities for designing new warheads (as mandated by present nuclear weapons policy). Copies of the full 49-page report, The Nuclear Safety Smokescreen, are available from IEER for $10.00, including postage. See Publications Page for more information on how to order.