Last December UCS, in collaboration with the American Association for the Advancement of Science (AAAS), held a one-day workshop on the safety and security of U.S. nuclear weapons. About 20 people participated, including active and retired scientists and engineers from Los Alamos, Lawrence Livermore, and Sandia National labs; representatives from NNSA, DOD and the State Department; independent scientists who are members of the JASON group that advises the government on nuclear weapons and other security issues; and experts from academia and nongovernmental organizations.
To allow for an open discussion, the meeting was held under the Chatham House rule, so the summary report does not identify participants. The discussion covered a range of issues, from improving the safety (risk of accidental detonation or dispersal of plutonium) and security (risk of unauthorized access or use) of weapons through modifications in the warheads themselves, to changes in delivery systems or operational procedures that could achieve improvements. It also covered cyber-security threats, especially the risk that such threats could present to use control.
While participants presented many differing opinions throughout a day of lively discussion, the general consensus was that security is a greater concern than safety. Safety improvements could benefit workers at sites that deal with nuclear weapons, particularly the Pantex plant, where the weapons are assembled and disassembled, but the NNSA’s frequently stated goal of converting all weapons in the stockpile to use insensitive high explosive did not seem to be a high priority for most participants.
On the question of security, participants were generally in agreement that it is a serious concern, but their thinking about how to address it varied. Some favored adding more intrinsic security features (those within the nuclear explosive package itself), saying that new understanding of how weapons work combined with experience from previous testing means that this could be accomplished without reducing confidence in the weapons’ reliability. Others felt that adding such features to all weapons would take decades and could still have the unintended effect of reducing confidence in the reliability of the stockpile. They argued for paying more attention to measures outside the nuclear explosive package instead.
Regardless of their preferred approach, participants widely agreed that any decision to add safety or security features to warheads should be based on sound cost-benefit and risk-benefit analyses that take into account both cost and reliability questions and consider the relative benefits of changes to the nuclear explosive package vs. other types of changes. Participants had differing opinions on whether NNSA and DOD are currently doing this, however.