Global Risk

JASON and Replacement Warheads

11.20.09 | 7 min read | Text by Hans Kristensen

Claims that nuclear weapons need to be as safe as a coffee table might drive warhead replacement

By Hans M. Kristensen and Ivan Oelrich

The latest study from the JASON panel is an unambiguous rejection of claims made by the National Nuclear Security Administration (NNSA), the nuclear weapon labs, defense secretary Robert Gates, and U.S. Strategic Command (STRATCOM) that some or all U.S. nuclear weapons should be replaced to ensure the future reliability of the arsenal.

The executive summary of the study, Lifetime Extension Program (LEP), finds “no evidence that accumulation of changes incurred from aging and LEPs have increased risk to certification of today’s deployed nuclear warheads.”  The study concludes that the lifetime of today’s nuclear warheads “could be extended for decades, with no anticipated loss in confidence, by using approaches similar to those employed in LEPs today.”  [Emphasis added.]

The JASON appears to have prevented a wasteful and counterproductive nuclear warhead replacement program. Even so, we expect parts of the report’s conclusions to be used by proponents of nuclear warhead replacements in the months and years ahead.

The Surety Argument

With JASON’s rejection in 2006 that pit aging is a reason to build replacement warheads, and its latest conclusion that stockpile reliability is achievable with the existing Live Extension Programs, only two of the core justifications used by proponents of RRW remains: surety and training.  (There is no agreed definition of the terms safety, security, and surety.  The DOD defines that nuclear safety reduces the probability of accidental explosion of the warhead;  security reduces the possibility of unintentional or unauthorized intentional use of the warhead;  surety combines these aspects.)

The report leaves the door open for replacement warheads by concluding that addition of nuclear surety or use-control features to the nuclear explosive package of reentry vehicles on ballistic missiles (W76, W78, W87, and W88) “would require reuse or replacement LEP options.”  Note that replacement warhead would not necessarily be a new design, but could be new components in an existing design.

Additional use-control, not warhead reliability, has thus become the main technical justification for building replacement warheads and we expect to see a sudden emphasis on surety by those who want to build new warheads.

This begs the questions:  how much surety is enough, who sets the bar, and what is it worth?

All U.S. warheads contain one or several surety features to prevent unauthorized use and accidents (see Table).  The last time the United States went through a stockpile-wide safety and security related upgrade was in the early 1990s.  Back then several weapons were phased out because they didn’t meet new safety and security standards, and new features were added to others. Not all nuclear weapons were created equal, however, and those that were seen as too important to retire were allowed to remain in the inventory even thought they did not meet the standard. But they will be gone soon.

U.S. Nuclear Warhead Surety Features

All U.S. nuclear warheads have surety features but details vary greatly due to history and deployment. Click for table.


After 9/11, the administration began arguing for raising the standard again. National Security Presidential Directive 28 (NSPD-28) issued in June 20, 2003, ordered the “incorporation of enhanced surety features independent of any threat scenario,” a capability-based safety philosophy based on technology rather than threats.  Under the headline “Urgency of RRW,” NNSA Director Thomas D’Agostino told Congress in February 2008 that “after 9/11 we realized that the security threat to our nuclear warheads had fundamentally changed.”

We have repeatedly probed officials about this alleged change, and they say it has to do with fear that terrorists will do anything to steal and use a nuclear weapon. The theory was that terrorists would go to greater length to steal U.S. nuclear weapons than the Soviet Union. Existing security features and well-protected storage sites are no longer sufficient; a nuclear weapon must be as inherently safe against unauthorized use as a coffee table, as one senior official recently put it.

Who can be against safety of nuclear weapons?  But if the price is several billion dollars then it is appropriate to ask what the surety and safety requirements are for U.S. nuclear weapons, how they have been set, by whom, and for what purpose.  Another way to pose the question is:  How will we know when we are done?

Since 9/11 the government has already spent huge sums to improve the physical security of nuclear weapons at bases and storage sites, and has upgraded use control features of some weapons.  The safety of US nuclear weapons is probably better today than it has ever been.  The greatest weakness is almost certainly administrative, as when military personnel loose track of the weapons, which happened in August 2007 at Minot Air Force Base.  We also note that suggestions to increase surety by changing the deployment and readiness of nuclear weapons, for example, taking weapons off alert or removing warheads from missiles and storing them separately, are dismissed out of hand.  So there are some actions that could improve surety that are clearly out of bounds.

The claim that the weapons themselves have to be made even more secure came later.  It was not a prominent component of the 2001 Nuclear Posture Review, and a concurrent review of “all activities involved in maintaining the highest standards of nuclear weapons safety, security, control, and reliability” did not result in a list of new warhead surety features in the subsequent Stockpile Stewardship Plans.  Indeed, the 2004-2008 plan instead declared: “The physical protection and security of nuclear weapons…remains [sic] strong….”

But after the Bush administration in 2004-2005 began lobbying Congress for authorization to begin industrial-scale production of new warheads to replace existing ones, the claim that additional warhead surety features are necessary has become a key justification.  For example, STRATCOM has recently proposed consolidating four versions of the B61 bomb into one based on the need for additional surety features (see Figure).

New Bombs For Surety

STRATCOM has recently used hypothetical needs for additional surety features as justification for building a new version of the B61 nuclear bomb. Click to download.


The slide includes some interesting assertions and assumptions underlined by a quote by Osama bin Laden saying it is a religious duty to acquire nuclear weapons to defend Muslims.  The main assertion is that current U.S. nuclear weapons are “not designed to address potential for nuclear terrorism.”  That certainly depends on what the “potential” is.  If it means terrorists trying to force their way into a storage facility, steal a weapon, and detonate it somewhere at their choosing, the claim is almost certainly wrong.  If on the other hand “potential” refers to the most worst-case scenario, where all U.S. safety and defensive efforts fail, then everything is of course possible.  But worst-case scenarios are not interesting when assessing what is necessary; realistic scenarios are.

The statement that only a small percentage of the stockpile has “internal disablement features” to prevent unauthorized use probably refers to bombs and cruise missiles that happen to make up a smaller portion of the stockpile than reentry vehicles for ballistic missiles. But as the Jason report concludes, “All proposed surety features for today’s air-carried systems could be implemented through reuse LEP options.”  Reentry vehicle warheads do not have these surety features, which might be a problem, but adding some does not necessarily requirement replacements, according to JASON.

The claim that all weapons lack modern surety features “to further reduce” the possibility that an accident could trigger a nuclear yield is of course true because one can always add more security features to further reduce the possibility.  The sky is the limit. The issue is, however, how much is needed.  In fact, once the remaining W62 warheads are retired (DOD missed the October 1, 2009, deadline), the entire stockpile will contain surety features that reduce the chance of warhead detonation due to accidents or terrorist attacks to less than one in a million.

The Skills Argument

Proponents of the RRW frequently have argued that it is necessary to build replacement warheads to keep a cadre of scientists, designers, and builders well trained and at the ready so that if sometime in the future we need new weapons we will be able to produce them. The JASON recommends improving the surveillance programs, but the language that “Continued success of the stockpile stewardship is threatened by lack of program stability, placing any LEP strategy at risk,” seems to criticize the NNSA and labs for being so fixated on building new bombs that the surveillance program has suffered.

The training argument depends on a combination of assumptions:  (1) the country will eventually need new nuclear weapons and these will need to be sophisticated weapons requiring high levels of expertise, (2) the expertise needed for continuing stockpile maintenance is not adequate to maintain the expertise needed to design and build new weapons, and (3) the knowledge and skills needed to build new weapons cannot be written down and can only be preserved over the next two or three decades by keeping it alive in people.  The truth of all of these assumptions depends in large part on choices we make about the future missions and requirements for nuclear weapons.  None of the assumptions is of necessity true.


The quest for new weapons is not dead yet.  Now that one main justification, reliability, has been deflated, those who want to continue to build new warheads are more likely to retreat to the second line of defense rather than surrender.  We expect the NNSA, the nuclear laboratories, and the military to focus their efforts on surety technology and laboratory expertise.  We need Congress (and perhaps JASON) to study the need for new surety features, and determine what level is sufficient for real-world threat levels.

This publication was made possible by a grant from Carnegie Corporation of New York and Ploughshares Fund. The statements made and views expressed are solely the responsibility of the author.

See all publications