Hacker News Reader

Comment by sketerpot | Hacker News Reader

In the case of TMI, the steam formation did slow down the reaction -- just not enough to prevent damage to the fuel rods. Of course, the reactor's pressure vessel acted as a passive heat sink once the fuel melted and easily prevented the fuel from escaping, but it was still costly and scary, and the response of the operators at the time can only be called a ridiculous clusterfuck.

This sort of thing is why I really like pebble bed reactors: you can just shut off the coolant and walk away, and they'll sit tight. (The operators of China's HTR-10 research reactor actually do this.) Everything is designed to withstand the maximum temperatures they could possibly achieve. Light water reactors have an impressive safety record, and the modern versions aren't susceptible to the problems that led to TMI, but inherently self-moderating reactors are just really aesthetically pleasant.

View on HN

>the response of the operators at the time can only be called a ridiculous clusterfuck.

This is unfair to the operators. The accident revealed some reactor and control room design flaws plus some equipment out of service that left the operators in the dark re what state the reactor was in during the accident. They knew the info they were getting was bad and took heroic steps to get better data, including sending men down into radioactive zones to read thermocouples manually with a volt meter, among other things. It is the anti-nukes who perpetuate the myth that the reactor operators freaked out and just mindlessly started throwing switches, closing valves willy-nilly. Unfortunately, the operators did make the situation worse but it was not due to stupidity or incompetence. One problem for the operators was their training was based on some assumptions that were not true for this accident. TMI was a pressured water reactor and one of the cardinal sins taught in training was never let the primary coolant system "go solid", i.e., no steam void in the pressurizer. A solid piping system could easily be burst by even a mild pressure transient which was why they opted to drain more coolant from an already overheating reactor. Tragically, the pressurizer was going solid because a steam void had formed in the core, something their training did not adequately address and they could not infer from the info available at the time. For obvious reasons, they had to make critical decisions within the time constraints and the data actually at hand, not 6 months later in an academic study.

Thank you for clarifying my terse statement; I can see how it could be taken as knocking the operators, which was not what I meant at all. Of course, the operators weren't a bunch of dimwits -- they had seriously flawed information about what was going on in the reactor, and the error-reporting system wasn't designed to cope with the kind of cascading failure that they saw. I wasn't criticizing the operators, but rather their actions, and that wasn't their fault.

(I can't even conceive of a reactor operator who would just start randomly fiddling with valves. That's just so far removed from everything I know about them, it would be like a horse reciting Shakespeare.)

Reply on news.ycombinator.com