What Safety Protocols Changed After The Demon Core Incidents?

When I talk with younger engineers at conferences I often boil the changes down to a few concrete things that came directly from those accidents and subsequent reflections. First, hands-on critical experiments were banned — you don’t use a screwdriver or a manual spacer anymore. Remote handling, shielding, and mechanical rigs became standard. Second, formal governance: written procedures, peer review, criticality safety specialists, and the ‘‘double contingency’’ principle are mandatory now. Third, instrumentation and monitoring improved hugely — continuous dosimetry, area neutron detectors, interlocks, and visible criticality alarms, plus clear emergency response steps.

On top of that the community adopted standards and best practices (think the ANSI/ANS 8-series and later agency rules), routine training, and periodic audits. Computational modeling (MCNP-style neutronics) and pre-test simulations now replace a lot of head-in-the-box experiments, and near-miss reporting and a stronger safety culture discourage theatrical demonstrations. It’s a cleaner, more boring world than those early days, but I’d much rather read a detailed procedure and watch a robot move a piece than hear about somebody trying to ‘‘feel’’ a setup — and that practical caution still feels like the right legacy from those tragic lessons.

I spent a long afternoon once digging through old lab reports and first-person recollections about those Los Alamos incidents, and the more I read the more I could feel how sharply the culture shifted after the two criticality accidents. At the time, experiments were often improvised, done by hand, and treated almost like puzzle-solving games — which sounds romantic until you see the photographs and realize how dangerously close people got to an assembled core. The immediate practical reaction was brutal and direct: hands-on approach to critical assemblies was effectively outlawed. Simple tools and stunts — the infamous screwdriver trick, for example — were swept away and replaced by engineered barriers, remote rigs, and explicit written procedures. People who had been allowed to ‘‘feel’’ the experiment were replaced by remote manipulators and mechanical shims so no human body could be the last line of defense between subcritical and critical geometry.

Beyond those immediate mechanical changes, the institutions themselves rebuilt how they governed work. Formal criticality safety programs were created — complete with specified mass and geometry limits, mandatory peer review for experiments, and the emergence of dedicated safety specialists who signed off on procedures. The ‘‘double contingency’’ mindset became a core design principle: systems had to be safe unless two independent things went wrong at once. That rippled into administrative controls too: stricter access rules, buddy systems or team sign-offs, mandatory dosimetry and area neutron detectors, and fast-acting criticality alarms. Emergency response protocols and medical follow-up were written into the playbook so exposures were tracked and treated systematically rather than ad-hoc.

Over the longer term I’ve loved seeing how technology and culture joined forces. Remote tooling, cameras, and robotic manipulators took over the fiddly parts; computer codes allowed far safer pre-calculation of margins; and formal standards and audits meant lessons were taught across labs instead of being internal folklore. But the human lesson always sticks with me: you can invent all the tools you want, but the single most important change was people agreeing not to re-glorify risky behavior. That cultural shift — from ‘‘brave tinkerer’’ to ‘‘procedural scientist’’ — probably saved more lives than any single gadget. I still think about how those policy changes map to other risky hobbies I have, and I try to bring that same stubborn caution to anything I tinker with at home.