« ПредыдущаяПродолжить »
firing systems; safety, security, and use-control systems; engineering support for production and dismantlement of nuclear weapons; field support to the military; and surveillance and support of weapons in stockpile. We perform substantial work in programs closely related to nuclear weapons, including intelligence, nonproliferation, and treaty verification technologies. As a multiprogram national laboratory, Sandia also conducts research and development or DOÈ's energy and science offices, as well as work for other national security agencies when our special capabilities can make significant contributions.
My statement addresses the topics specified by the subcommittee, including the efficacy of the Science-Based Stockpile Stewardship Program and its future prospects; the adequacy of our current test readiness posture; the value of exercising the nuclear weapon design process; the stockpile annual certification process; and the condition of the nuclear weapons complex infrastructure. In addition, I will comment on the issue of programmatic planning that was raised in the most recent Foster Panel report." I would also like to refer you to the attached appendix, which summarizes some of Sandia's recent accomplishment in the nuclear weapons program and other national security programs. We state our core purpose as "helping our nation secure a peaceful and free world through technology.” I believe the accomplishments reported here will convince you that we are “on course” in fulfilling that purpose.
EFFICACY OF SCIENCE-BASED STOCKPILE STEWARDSHIP Science-based stockpile stewardship was adopted about ten years ago in response to the curtailment of all major warhead development programs and the increasing likelihood of an indefinite moratorium or even permanent ban on underground nuclear testing. With respect to Sandia's stockpile responsibilities, it is my judgment that the Science-Based Stockpile Stewardship Program has met expectations during this last decade. We have developed and proved an array of diagnostic and design tools that are enabling us to meet the enormous challenge placed on us by the loss of nuclear testing and that permit us to fulfill our responsibilities for the stockpile with high confidence and predictability.
However, it is crucial to note that nearly everything Sandia designs and surveils can be tested and is tested-using nonnuclear processes. We subject our components and subsystems to extensive nonnuclear testing at every stage of development and service life. We have made fair progress during the last decade under the Science-Based Stockpile Stewardship Program in strengthening our testing capabilities, and good progress in advancing our modeling and simulation capabilities to deal with those aspects that cannot be directly tested, such as the effects of hostile radiation fields on our components and stems.
The primary application of nuclear testing for Sandia in the past was to confirm the functionality of nonnuclear weapon components and the warhead as a system when exposed to hostile environments such as the full radiation fluxes and extreme mechanical impulses of fratricide, preemptive strike, or nuclear-armed anti-ballistic missile (ABM) defenses. This is a different purpose than that served by the underground nuclear testing performed by the nuclear design laboratories, Los Alamos and Lawrence Livermore, for development, safety, or confidence testing related to the performance of the nuclear explosive assembly or “physics package.”
Perhaps the best example of the efficacy of the new science-based stockpile stewardship tools is Sandia's development and qualification of the MC4380A neutron generator for the W76 Trident warhead. Neutron generators must be designed for ruggedness against severe environments such as acceleration, vibration, high voltage, radiations, and mechanical impulse. In the past, we relied on an iterative design process involving numerous physical tests and whatever modeling tools were practical at the time.
Relying on data from past underground nuclear tests and aboveground simulations using accelerators and reactors, our scientists and engineers have developed large integrated software models that simulate three-dimensional radiation transport and mechanical response. These models allow our designers to visualize the electrical and physical performance of a neutron generator under many combinations of conditions while it is still “on the drawing board.” A design engineer can change the model and re-visualize its performance on the computer many times before committing to a physical prototype.
The combination of advanced computational simulations such as these and a comprehensive suite of several kinds of nonnuclear tests made it possible for the first time in the history of the program to qualify a neutron generator design for performance reliability and resistance to hostile radiation effects without underground nuclear testing.
I must note here that, although in the past we would always subject such components to hostile radiation in underground nuclear tests to try to directly evaluate the effects of radiations on their function, such "effects testing" could never be made an ideal test. The underground exposure was itself always a compromise to the anticipated levels of stress that components might be subjected to in wartime. This was because the levels of neutrons, x-rays, and gamma rays were in different proportion to what would be experienced in a space or atmospheric encounter and because the fluxes were always different than the anticipated levels. Thus, the confidence by which we have certified the new neutron generator design-without having had an underground “effects” test—is on a different basis, but not necessarily a lesser basis, than was our past practice.
Sandia also has responsibility for the integrated arming, fuzing, and firing (AF&F) system of the W76-1/Mark-4A life extension program. Science-based design tools will permit us to perform the redesign of this complicated and critical assembly at lower cost and with higher quality than was previously possible. The redesigned unit will combine advanced fuzing options, modern nuclear safety improvements, and enhanced reliability. Moreover, we are incorporating surveillance features into the unit so that its "state of health" can be assessed with minimal intrusion.
Qualification of the new W76-1/Mark-4A AF&F will involve both testing and simulation using tools provided by science-based stockpile stewardship. We must conduct a variety of environmental tests in the laboratory to evaluate the unit's performance under various normal and abnormal conditions. We will perform system flight-tests with de-nuclearized payloads to achieve flight environment conditions that cannot be simulated in the laboratory. Radiation tests using aboveground simulators will provide radiation effects testing for most spectra of concern. Parameters derived from all these categories of tests will be incorporated into computational models that can calculate system performance over a broader and more intense range of conditions.
New modeling and simulation capabilities developed under science-based stockpile stewardship are providing powerful capabilities for life extension programs, and they reduce the total number of physical tests needed over the term of a project. However, it is important to understand that physical testing will not be eliminated by computer simulations. Models never achieve perfection, and nature sometimes has surprises in store that become apparent only during physical tests. Moreover, the fidelity of the computational models themselves must be validated with experimental data. For example, the codes used to model radiation effects for the neutron generator program were validated against experiments performed with aboveground simulators and past underground tests.
For these reasons, science-based stockpile stewardship at Sandia is a program of advanced nonnuclear testing as well as computational modeling. Sandia's Test Capabilities Revitalization Project (an FY 2003 construction item) is important in this regard, as it is essential to modernize our field testing and experimental infrastructure to support warhead qualification, development, surveillance, and model validation.
Definitive evidence of the efficacy of science-based stockpile stewardship will be available when we complete our first full-scale life extension program for a major warhead system. We expect the W76-1/Mark-4A life extension project to enter production in 2007. Several large certification hurdles must be surmounted before production can be authorized. However, I expect that we will be able to meet those challenges with the tools that we have developed and are continuing to improve under science-based stockpile stewardship.
In the past thirty years or so, the nuclear weapons program has developed several aboveground experimental facilities in an effort to simulate many of the phenomena of hostile environments as a substitute for underground "nuclear effects testing." The last time Sandia used underground testing for certifying a system against hostile environments was with the W88/Mark 5 program. The six underground radiation tests we conducted for that program were supplemented with more than 1,000 aboveground radiation tests using fifteen different simulators. The available suite of aboveground simulation facilities today, augmented with improved computational models, has allowed us to separately simulate or model most nuclear-weapon radiation spectra.
It must be recognized, however, that our simulation and modeling capabilities have limitations. The extreme radiation fluxes and mechanical impulses of a nuclear detonation cannot be directly simulated. In addition, the physical size of hardware systems that can be tested for complete system response is limited. In the future,
ensuring the fidelity of some aspects of our computational models may not be possible without access to nuclear testing.
However, these limitations may or may not prove to be important in the long run, depending on how international nuclear threat environments evolve. Given today's conditions of threat and technology, science-based validation (as opposed to nucleartesting-based validation) does not, in my opinion, present significant difficulties for Sandia's certification and validation responsibilities. I must caution you, however, that this conclusion applies to Sandia's needs and cannot be directly extrapolated to the role that nuclear testing has played for validating the functioning of the nuclear physics packages, which are designed by either Los Alamos or Lawrence Liver
The question of whether three years, one year, or three months is an adequate lead time for conducting an underground nuclear test may be important as it relates to matters of safety, confidence, or perhaps development associated with the nuclear physics package of a warhead. The time required to carry out a test will also depend critically on whether the problem that led to the need to test is one that could affect a large percentage of the stockpile. However, because of Sandia's success in developing an altemative methodology for hostile effects certification, urgent need for testing will no longer be the crucial issue it once was for ensuring performance in hostile environments for the systems for which we are accountable.
EXERCISING THE WEAPON DESIGN PROCESS Much of the supporting science for stockpile stewardship can be exercised in laboratory investigations, but design skills can
only be proved on real products. System life-extension projects serve two purposes: They modernize older systems that need refurbishment, and they exercise the competence of the weapons engineering skills that we require for the future. However, exploratory work on advanced concepts will also be necessary to ensure that our design skills are sufficiently challenged for evolving needs in the nation's nuclear forces.
The nuclear weapons complex has not been engaged in a new system design since 1992. During the past ten years, we have exercised our competencies with a few modification programs, exploratory projects, and subsystem enhancements. Assuming that a new warhead design will not be authorized for the foreseeable future, full-system life extension programs are the only effective vehicle for exercising the design process.
We depend on engineers and scientist who are knowledgeable, experienced, and seasoned in their judgment for making stockpile stewardship succeed. Our confidence in their ability to perform their responsibilities is gained through seeing them succeed with large, complicated, and challenging projects that require them to think through the integration of the many elements of a system into a demonstrable product. Therefore, it is important that the NNSA laboratories continue to offer weapon design work in the form of life extension projects or similar programs on a permanent basis.
I must emphasize that the nuclear weapons program requires an intimate relationship between the laboratories, where designs are created, and the production plants that manufacture the designs. Sandia design engineers work closely with production engineers at the NNSA production agencies and contractors where components are manufactured and war-heads are assembled or disassembled. The laboratories are also the appropriate authorities for certifying production plant processes.
The new generation of engineers and scientists who will perform design and production engineering in the decades ahead will not have had the benefit of experience on full-scale weapon development programs. We must find other ways to qualify those people in the future. The life extension projects approved by the Nuclear Weapons Council for the W76 Trident warhead, the B61 bomb, and the W80 cruise missile warhead are important major projects for exercising the design process and the designers.
STOCKPILE ANNUAL CERTIFICATION PROCESS A major effort of the Science-Based Stockpile Stewardship Program is directed to the annual assessment of the certification basis for nuclear weapons in the stockpile. To perform the assessments that support this annual process, the laboratories conduct reliability and safety investigations and prepare a report for each weapon type in the stockpile. The laboratory directors individually submit their "certification letters” to the Secretaries of Energy and Defense, who in turn integrate the information and formally report the condition of the stockpile to the President.
Assessment Activities at Sandia Sandia's responsibility for stockpile annual certification comprises the nonnuclear subsystems that control the operation of nuclear warhead. Our confidence in the stockpile has always been anchored in the community of experienced engineers and scientists who are expert in the disciplines of stockpile stewardship. Confidence is also maintained through exhaustive non nuclear tests, a long history of fielded weapons and their data, a careful preventive maintenance and replacement program, chemical analyses, computer modeling, and joint or independent reviews of our work.
We test and model the performance of nonnuclear components and systems in a variety of normal, abnormal, and hostile operational environments. We certify weapon performance under normal operating environments, and we verify that components and systems will retain adequate functionality after exposure to hostile environments. Most normal operational environments of concern for nonnuclear components and systems can be simulated without nuclear explosive tests.
Under the Defense Programs Enhanced Surveillance Campaign, we develop tools and models to measure, qualify, and predict the effects of aging on weapon materials and components and to understand how those effects impact weapon safety and reliability. One enhanced surveillance project uncovered unexpected behavior in desiccants designed to maintain a noncorrosive internal atmosphere in the warhead. Our new understanding of desiccant behavior is guiding the formulation of new desiccants for weapons refurbished under stockpile life extension programs. Another surveillance project discovered problems with newly procured material for replacement o-rings, which we were able to intercept.
Two years ago Sandia introduced non destructive, acoustic laboratory testing of strong-links, a major safety component of nuclear warheads, into the core surveillance program. Last year we added a second development from the Enhanced Surveillance Campaign into this core surveillance test equipment that allows us to evaluate the electrical current-carrying capacity of these safety devices. Both of these new tests have allowed us to better predict the useful lifetime of this critical component and enhance our replacement planning strategy.
DoD and DOE/NNSA annually conduct joint flight tests on warheads of each type in the enduring stockpile. Historically, flight tests have uncovered approximately 22 percent of the defects discovered in surveillance databases.
I would like to address the two reports issued by the DOE Inspector General this past year on the surveillance program-one on the testing backlog for flight and laboratory tests, and the other on the significant findings investigation process. The backlog situation was noted by the Foste Panel, and I have referred to this problem in previous years' statements to this committee. While I do not believe that the situation is as dire as some might have suggested, action was necessary on the part of NNSA and the laboratories to improve performance. We are working with the Navy and Air Force to ensure the availability of samples and flight-test vehicles to complete the desired levels of testing. After a hiatus in Air Force cruise missile testing due to missile problems and infrastructure renewal, I am pleased to report that we have begun flight testing again with two successful advanced cruise missile tests, although it will take us several years to catch up with our desired level of testing.
The Foster Report emphasizes that surveillance, assessment, and certification processes for the stockpile should be as rigorous and probing as possible. I am in full agreement that the laboratories should be challenged to improve their processes and adopt the most advanced tools and effective assessment methodologies available. Complacency in this mission space would be inexcusable.
The Foster Report recommends the use of “red teams” within each laboratory and strongly endorses the inter-laboratory peer review function.4 Sandia has practiced red teaming and peer review successfully for decades. Our Surety Assessment Center is a full-time red team that is organizationally independent of the weapon design groups and which reports directly to the laboratory's executive management. In addition, we engage an independent advisory panel with external members to oversee the activities of the Surety Assessment Center and make recommendations directly to executive management on a semiannual basis. Thus, not only do we have a red team, but we also have a red team for the red team!
Peer review at Sandia follows the same model as the Livermore/Los Alamos competitive arrangement. We have a laboratory in New Mexico that supports development programs assigned to Los Alamos, and we have another laboratory in California adjacent to Lawrence Livermore that supports Livermore weapon programs. The example of peer review with the B61 described in the Foster Report 6 is not a new concept, but is basically how the arrangement works in practice. The California designers peer-review the work of the designers in New Mexico, and vice versa. I assure you, it is not a collegial interaction. It is a formal process that is often quite contentious. Within the past few weeks, I have received peer review reports from groups in California and New Mexico that confirm again to me that Sandia's peer review process is vigorous and robust.
Comments on Section 3144
Regarding Red Teams and Peer Review The Defense Authorization Bill at Section 3144 would mandate laboratory "red teams” to challenge internal laboratory assessments and to perform inter-laboratory peer reviews. Actually, red teaming and peer reviews have been standard practices between the nuclear weapon laboratories since at least 1956, when the concept of two competing laboratory clusters was fully implemented.
However, the current language of Section 3144 is faulty in many respects. For example, it would require Sandia to peer-review the assessments of the nuclear design laboratories and vice versa. This fails the first requirement of “peer review”—that the participants indeed be peers! Sandia is not competent to peer-review the nuclear explosive systems of Livermore and Los Alamos; and conversely, Livermore and Los Alamos do not have the competence to peer-review the technologies nor the complexities of Sandia's nonnuclear components. But the longstanding arrangement whereby the California design cluster and the New Mexico design cluster peer-review each other avoids that problem, and has proved to be an effective practice.
I am also troubled by the provision requiring that the President and Congress receive each certification letter and report from each laboratory director and the commander of Strategic Command, including the findings and recommendations of all their red teams. Currently, the laboratories' Annual Assessment Reports and directors' letters are included as background information in the package accompanying the joint certification memorandum from the Secretaries of Energy and Defense. But I do not believe that it would be appropriate, as a routine practice, to forward all red-team findings to the President and Congress. Red-team issues are usually very arcane and highly technical. In the vast majority of cases they can-and shouldbe resolved at the level of the laboratory director.
However, I have always maintained that a minority report from a laboratory director regarding the certification of any warhead should be communicated to Congress and the President as part of any safeguards process associated with a nuclear test ban or moratorium. The Nuclear Weapons Council (a very senior council of Defense and Energy that was created by Congress to oversee nuclear stockpile issues) requires that the laboratories' Annual Assessment Reports be “forwarded unaltered to the Secretaries,” so I do not see this as a worrisome issue.
The Secretaries of Energy and Defense have a responsibility to integrate the laboratory directors' findings and provide the President with the “bottom line,” and I believe that any president would require that of them. Currently, the Nuclear Weapons Council is tasked to perform that integration function and prepares the Nuclear Stockpile Certification Memorandum (to the President) for signature by the two secretaries.
It is surprising that Section 3144 makes no mention of any role for the Nuclear Weapons Council. Under current law 6 the Nuclear Weapons Council has broad responsibility for oversight of stockpile programs. Some of the requirements that Section 3144 would place on the laboratory directors (i.e., in their reports accompanying certification) are already assigned to the Nuclear Weapons Council by statute. would be uncomfortable, for example, evaluating the relative merits of various nuclear weapons for a particular military mission, as would be required of me in my annual certification report as currently outlined in Section 3144. However, this (and other responsibilities) are adequately and appropriately discharged by the Nuclear Weapons Council.
I credit the Foster Panel for focusing attention on the importance of the annual certification process, which was originally established by President Clinton in 1995 by directive. The process was also spelled out in the Resolution of Ratification that accompanied the Comprehensive Nuclear Test-Ban Treaty to the Senate. However, because that resolution failed in the Senate vote, that document today has no formal status. I might add that with the failure of the CTBT ratification, there also is no procedural certainty by which the need for a nuclear test would be communicated within either the Executive or the Legislative branches of the U.S. government. Thus, it may perhaps be time to establish annual certification as a statutory requirement with responsibilities carefully defined in law. However, by moving too quickly with the proposed Section 3144 at this time, we may create, at best, a partial fix that will introduce some unintended consequences.