International Treaties - CTBT


The U.S. Stockpile Stewardship Program is Far Larger Than Needed to Maintain Safety and Reliability

  • Nuclear weapons are made of parts, and the task of maintaining an arsenal indefinitely without nuclear testing is tantamount to maintaining those parts or replacing them as needed with others at least as functional.  Nuclear weapons do not age, their parts age. 
  • More than 95% of each weapon's parts lie outside the nuclear explosive itself (i.e. outside the "physics package"), and all these non-nuclear parts can be fully tested without nuclear testing (both individually and together as systems).  They can also be replaced as needed (or desired) with even more functional, fully-tested components and systems.  This is the job of Sandia National Laboratory, and it is a job essentially unaffected by the Comprehensive Test Ban Treaty (CTBT).

  • Some weaponization improvements may improve security; others could change or broaden the "uses" of the weapon, as in the creation of the B61-11 earth-penetrating bomb (completed) or the B61 precision-guided glide bomb (under development).  Neither of these ill-advised new weapons, nor others to follow, require nuclear testing to certify and deploy.

  • Within the physics package, all the components except the fission "primary" and the thermonuclear "secondary" can be fully tested and even changed ("improved") if desired without nuclear explosive testing.  The problem of maintaining weapons under a test ban boils down to maintaining or replacing the components of the primary and secondary as needed or when desired.

  • Nuclear weapons safety is concerned with preventing a) nuclear detonation and b) plutonium dispersal in an accident.  All U.S. weapons have less than one chance in a million of measurable nuclear yield in an accident.  All but submarine-launched warheads use chemical explosives that cannot accidentally detonate.

  • Nuclear weapons safety issues arise solely in relation to the primary, as the secondary contains no explosives and in at least some cases no plutonium.  The safety of primaries is an intrinsic property of the design, and there are no mechanisms or processes which produce age-related safety problems in primaries.  The CTBT will not and indeed cannot degrade the safety of existing nuclear designs.

  • The safety of the arsenal could, however, be degraded by the attempt -- already begun -- to design and certify new primaries that cannot be positively tested for their safety characteristics before deployment.  It could also be degraded by inadequate surveillance.  By far the greatest dangers posed by nuclear weapons are inherent in their very existence and to the posture of alert in which they are kept.  Attempts to improve nuclear weapon "safety" by changing the design of weapons are misguided and could in fact decrease safety.  The reliability of a nuclear weapon is the probability of detonation within a very precise yield range, not whether it will explode with large yield.  U.S. weapons are kept extremely reliable.

  • Any reliability problems encountered with secondaries can be fixed by remanufacture.  Only a very few problems capable of degrading the reliability of secondaries have ever occurred.  In these cases, corrective measures have been straightforward.  The U.S. maintains a large factory for secondary components in Oak Ridge, Tennessee (the "Y-12" facility).

  • Test-ban-related reliability concerns are confined to the primary stage only.  All components of the primary can be replaced with new, fully-tested components, should they be found to have aged or to otherwise be defective, except the central plutonium-containing "pit."

  • No aging problems in pits have been detected.  The responsible officials at Los Alamos have repeatedly said in private that there are no mechanisms that will degrade the performance of pits for the first several decades of aging.  There are long-term aging mechanisms, but they are not significant in the near term -- i.e. in the next two decades or more.

  • The nation has a stockpile of more than 10,000 extra pits.  Some, if not many, of these can be used in stockpile weapons, as already proven for one design by two nuclear tests.

  • Present facilities at Los Alamos National Laboratory are adequate to manufacture about 20 pits per year.  Los Alamos plans to begin an unnecessary site-wide upgrade next year to its existing facilities to increase production capacity to four times this rate, if not more, creating a sort of Rocky Flats II. Cost:  about $1.3 billion.

  • Contrary to misleading testimony by the lab directors, the "stockpile stewardship" program has resources, personnel, and capabilities far beyond those actually needed to maintain existing U.S. nuclear forces indefinitely.  Much of this program consists of projects and facilities only marginally related to maintaining the stockpile, such as the National Ignition Facility (NIF), and pulsed power facilities.  The program also includes expensive duplication (e.g. in dynamic radiography and computing), and expensive provisions to prepare for treaty breakout (e.g. tritium supply and NTS).  Current plans include the construction, over the next 13 years, of a new and upgraded complex of nuclear facilities worth more than $16 billion (see below).  The program is essentially a protection racket run by the nuclear labs.

Stockpile Stewardship Capital Costs, 1997-2010
pulsed power facilities (at least eleven) $   505. million
dynamic radiography facilities (seven   1,683.
inertial confinement fusion facilities (eight) >1,200. 
computing (ASCI and buildings alone   3,078. (omits other large computing $)
new and upgraded buildings      973. 
miscellaneous other stockpile stewardship facilities      134.
    subtotal, stewardship capital costs >7,573. 

Stockpile Management Capital Costs, 1997-2010
tritium production (for non-deployed weapons!)     6,529. (accelerator option only; 
                 option not yet chosen)
new buildings and upgrades     1,987.
miscellaneous other stockpile management upgrades        486.
    subtotal, management capital costs >$9,002. 

Total stewardship and management
capital costs $16,575. million

To these capital costs must be added the circa $44 billion in program costs expected for these years.  The annual cost of this program ($4.5 billion plus $1.5 billion for waste management)greatly exceeds the Cold War average of $3.7 billion (in today's dollars; includes waste management). 

In sum, there are no safety or reliability concerns that prevent the U.S. from promptly ratifying the CTBT.  It is the oversized stewardship program itself that is poised to create these problems by gratuitously creating untested designs.

For further reference, see:  Richard Garwin, "Stewardship: Don't Claim Too Much or Too Little," Bulletin of the Atomic Scientists, May/June 1997; Ray E. Kidder, "Problems with Stockpile Stewardship," Nature, April 17, 1997 and "Maintaining the U.S. Stockpile of Nuclear Weapons During a Low-Threshold or Comprehensive Test Ban" Lawrence Livermore National Laboratory UCRL-53820, October 1987; Greg Mello, "Nuclear Weapons Safety: No Design Changes Are Warranted," Tri-Valley CAREs, Livermore, July 1995; Christopher Paine and Matthew McKinzie, "End Run: The U.S. Government's Plan for Designing Nuclear Weapons and Simulating Nuclear Explosions under the Comprehensive Test Ban Treaty," Natural Resources Defense Council, Washington DC, August 1997; Stephen Schwartz et. al., "Four Trillion Dollars and Counting," Bulletin of the Atomic Scientists, November/December 1995, pp. 32-52.

The stockpile stewardship program has serious dangers.  The underlying justification for most of these new experimental facilities is to maintain, exercise, and improve nuclear weapons expertise in the absence of nuclear testing.  That is, the central problem of stewardship, as perceived by the DOE and the labs, is how to maintain scientists, not weapons.  "Stockpile stewardship" is designed to assure that the stockpile will maintain the stewards, not vice-versa.  Its rising program budgets reflect political negotiations, rather than any underlying scientific or engineering need as regards maintaining the existing stockpile.

Instead, the phrase "maintaining the safety and reliability of the stockpile" has ironically come to mean its opposite:  a program of continued modification and improvement of weapons, including the modification and replacement of primaries and secondaries with new designs, a program that could easily compromise safety and definitely will compromise reliability.

These pending modifications would provide a powerful tool by which nuclear testing advocates will argue that the CTBT must be at least temporarily abrogated, or in the alternative that a second or third generation of advanced "surrogate" testing facilities be built.  The dangers of the first outcome are plain.  The second outcome will further undermine the trust of the non-nuclear weapon states in the regime established by the NPT and CTBT and will at the same time lead to strong pressures to share equipment and data with our erstwhile nuclear "enemies" to facilitate their ratification and subsequent compliance with the CTBT.  This data-sharing and equipment procurement has already begun.

The stockpile stewardship program is a potent vehicle for the internationalization of nuclear weapons technology, both between weapon states and reflecting a growing coincidence of interests in the weapons infrastructures in the weapon states, as well as in the academic computer computation physics community worldwide for new facilities and programs.  The problem of maintaining scientists and engineers can be greatly simplified by the observation that, overall, we need far fewer of them.

What is the alternative?  A conservative program of stockpile maintenance, or "curatorship," based on these principles:
  • If it ain't broke, don't fix it.

  • Don't introduce untestable "improvements" into the stockpile as a way to "exercise" the labs.

  • Provide for the replacement of components which age, with an surveillance program to detect unwanted changes.  This is already in place.

What does an alternative to the present stockpile stewardship program look like?

The U.S. has about 12,000 nuclear weapons of 10 types.  At least 90% of this arsenal was created and is being maintained as a deterrent to Russian nuclear weapons.  Russia, however, is broke, and cannot keep its bombers airworthy or its submarines at sea.  The mutual danger and expense of these huge arsenals could result in further bilateral arms control initiatives in the coming decades, or even this decade.  Thus the "stockpile" in "stockpile stewardship," is hardly fixed, and it would be imprudent to make decisions based on a hypothetical "need" to maintain this arsenal for many decades -- especially when investments and actions based on that "need" may themselves have serious national security costs.  These costs include: Damage to prospects for entry-into-force of the Comprehensive Test Ban Treaty (CTBT)

The U.S. is formally, although not actually, committed to nuclear disarmament by Article VI of the Nuclear Nonproliferation Treaty.  In addition, the International Court of Justice has affirmed that the use or threat of nuclear weapons is generally contrary to international law.  The failure of the U.S. to abide by laws regarding nuclear weapons has affected, and will continue to affect, the stature and entry-into-force of the Comprehensive Test Ban Treaty (CTBT), as well as U.S. nonproliferation objectives.  These facts, which may become more pointed in the next few years, could inspire a reconsideration of discussions of the U.S. nuclear weapons stewardship program.

November 15, 1997

^ back to top 2901 Summit Place NE Albuquerque, NM 87106, Phone: 505-265-1200, Fax: 505-265-1207