Thanks. Byrnes' write-up says "Since the bodies were contaminated with life-long radio-active isotopes, each were buried in a lead-lined coffin, buried extra-deep, and covered with many feet of concrete as their bodies were radioactive. The graves are to stay undisturbed unless prior approval of the Atomic Energy Commission. After a two-year investigation, it was determined that there would never be one single control rod used to employ in an atomic pile; modern ones have scores of rods. The incident was not determined an accident or an act of sabotage."
So it sounds like all three men received a similar interment.
But now you did it! You sent me down a Memorial Day History Rathole!...
"The incident was not determined an accident or an act of sabotage." -- But what does that leave? Abject stupidity? Poor training? But wouldn't stupidity or lack of training be classified "accident"? Away I go...
Of note:
"After a two-year investigation, it was determined that there would never be one single control rod used to employ in an atomic pile" -- so, tragically, some good came of their deaths. But how could the designers have NOT foreseen that?
"Failure Modes & Effects Analysis (FMEA)" was conceived, developed, and first used by the US Military in the late 1940s. FMEA was one of the first systematic techniques for failure analysis. It was formally established in the United States military via Military Procedure MIL-P-1629, titled "Procedures for Performing a Failure Modes, Effects and Criticality Analysis", dated November 9, 1949. It was developed as a reliability evaluation technique to determine the effect of system and equipment failures, classifying them according to their impact on mission success and personnel, equipment, and safety. The immediate trigger was practical: the US Military developed the technique specifically to reduce sources of variation and failures caused by variation in munition production. But that did not get translated to nuclear physics until much later!
There were two Los Alamos Lab criticality deaths years before the Idaho deaths:
- Daghlian, 1945 — died after dropping a neutron reflector onto the demon core, causing a prompt criticality excursion
- Slotin, 1946 — demonstrating a criticality experiment, lowering a beryllium dome over the core. The screwdriver he was using to maintain a gap slipped, and the dome dropped, fully covering the core for an instant — bouncing too many neutrons back at it.
The Los Alamos scientists knew well the risks of what they were doing — the trick was finding how far you could go before a dangerous reaction was triggered. They even had an informal nickname for it: "tickling the dragon's tail," knowing that if they roused the beast, they would be burned. Crucially, after these accidents, new protocols meant an end to hands-on criticality experiments,
with scientists forced to use remote control methods. So the physics community had already learned — at fatal cost — that a single human action could cause an instantaneous criticality excursion. That lesson was baked into nuclear physics culture by 1946, fifteen years before SL-1.So Why Didn't That Knowledge Protect the SL-1 Crew?
There were multiple interlocking failures that go far beyond FMEA:
- Institutional Silos — the Weapons World vs. the Reactor World. The criticality knowledge lived in the weapons physics community at Los Alamos. The SL-1 was designed by Argonne National Laboratory for the Army's power reactor program — a different institution, a different mission, a different culture. The weapons physicists who viscerally understood prompt criticality were not the engineers designing compact Army reactors for remote radar stations. The knowledge existed, but it did not travel across organizational boundaries in any formalized way.
- The Design Was Driven by Cost and Simplicity, Not Safety Physics. More control rods would effectively have reduced the reactivity worth of any individual rod. Since the addition of more control rods would have increased costs, it is easy to imagine the decision to go with a minimum number based on something other than safety. This is the core of it: the people making the design tradeoff likely did know that concentrating reactivity in one rod was dangerous — but cost and the Army's portability requirements won. That is not a failure of FMEA. It is a failure of engineering ethics and safety governance.
- The Critical Parameter Was Never Even Calculated. Prior to the accident, no one had computed the prompt criticality rod withdrawal distance. The complex and irregular arrangement of burnable boron strips made modeling the SL-1 core particularly difficult, and calculations for predicting criticality were based on greatly over-simplified computations. This is extraordinary: the designers of a nuclear reactor never calculated the distance one of its control rods had to be pulled to cause a catastrophic runaway. The Los Alamos physicists would have been horrified.
- The Operators Were Dangerously Junior and Undertrained. The Army Nuclear Power Program used very junior personnel with minimal training — this previous approach was recognized as inadequate for the complexity of nuclear operations only after the accident. The operators handling the SL-1 were young enlisted men and a Navy Seabee — not nuclear physicists. The institutional memory of "what happens when you go prompt critical" was not transmitted to them in any meaningful way.
- Known Warning Signs Were Suppressed. The SL-1's history of frequent control rod sticking was downplayed and ruled out as a cause of the accident long before the damaged core was closely examined. Management was actively filtering out safety signals.
| What FMEA covered in 1961 | What FMEA missed at SL-1 |
|---|
| Hardware component failures | Human actions as failure modes |
| Individual part malfunctions | System-level emergent behavior |
| Known failure mechanisms | Uncalculated physical limits |
| Mechanical reliability | Design philosophy risks |
| Documented hazards | Institutionally suppressed warnings |
James Reason created the "
Swiss Cheese Model," one of the most elegant and useful mental models ever produced by safety science. We often hear of it today in aviation accident context by "Captain Steeeve, "Hoover," and others. Reason visualized every complex system as having multiple layers of defense — procedures, training, equipment design, supervision, regulatory oversight. Each layer has holes in it (latent failures, bad assumptions, tribal knowledge, design flaws). Most of the time the holes don't align across layers and disasters are stopped or degraded. But
occasionally — through a specific combination of circumstances, chance, timing, and triggering events — the holes line up perfectly and a trajectory of failure passes clean through every defensive layer simultaneously.What makes it so powerful as a framework is what it implies about blame. When the cheese holes line up:
- It wasn't just the operator who pulled the rod too far
- It wasn't just the designer who concentrated 80% of reactivity in one rod
- It wasn't just the manager who suppressed the sticking rod reports
- It wasn't just the regulator who never required the criticality calculation
It was all of them, simultaneously, on that one night. SL-1 is almost a textbook Swiss Cheese diagram:
- Layer 1 — Physics knowledge not transferred from Los Alamos ✅ hole
- Layer 2 — Critical parameter never calculated ✅ hole
- Layer 3 — Single rod design never challenged ✅ hole
- Layer 4 — Sticking rod history suppressed ✅ hole
- Layer 5 — Junior operators undertrained ✅ hole
- Layer 6 — No interlock to prevent over-withdrawal ✅ hole
Every layer failed simultaneously on January 3, 1961.
I spent a lot of time in plants and control rooms, so all that is a personal journey.
OK, that's it. I'm outta here! Time for some sunshine and hiking. Somber Memorial Day!