4 ms·

Out-of-spec doesn't mean it can't work. It just means you have to redo the design using what's effectively a different part than you originally planned on.

jacquesm 4y ago
Yes, but the idea here was to construct a nuclear power plant, not to build a large pot for boiling soup in. You can't take a critical component like that, spec it and then suddenly pretend the spec never mattered in the first place, then you have to admit that you're just winging it. Changing the spec of the reactor vessel essentially translates into a complete redesign of the reactor itself unless you are willing to compromise on other aspects, such as safety, longevity and so on. We're not talking about a bracket here. Or an O-ring. When was the last time something as stupid as an O-ring decided the fate of... oh, never mind.
- otter-rock 4y ago
  It's a requirements change. Those happen all the time in everything. Why would that be completely forbidden or impossible here? Like you said, it might really suck. But the NRC does not allow trading off safety the way you suggest.
  - jacquesm 4y ago
    Requirements changes are not driven by one-off material defects in critical pieces of hardware. That's simply a bending of the rules for economic reasons. And it is one of the main reasons for me to oppose nuclear: the fact that people will be people and that at the root of every one of those disaster and near disasters there was someone who thought they could get away with something. We are ill equipped to deal with this kind of responsibility, especially across a timeframe measures in decades. At the same time I would love to see us solve the climate change problem, and I recognize that we will likely have a nuclear component in there. But it will have to be done by the book or we'll end up regretting it - again. If we're going to start out with the normalization of deviance on the #1 critical component of a reactor then I think we are on the wrong path: https://becht.com/becht-blog/entry/normalization-of-deviance-the-pathway-to-disaster/ https://becht.com/becht-blog/entry/normalization-of-deviance...
    - otter-rock 4y ago
      I see your concerns, and I think you're probably just unaware that they're addressed better than you know, but in a different way than you'd expect. When they accept an expensive out-of-spec part, it's because they can safely redesign the less expensive parts. The only thing that hurts is their long-term profitability because it'll be less efficient than they hoped. Big, expensive parts are always treated as one-of-a-kind articles with unique requirements. It's how bridges and roads get built, for example. With nuclear, they have special teams that do nothing but manage every type of risk you could ever think of for every piece of nuclear material in the country. It's very advanced, and many industries are trying to use those techniques themselves now. Nobody is normalizing or bending anything. Every potential problem is thoroughly addressed on an individual basis. There is no deviance from the safety requirements. Nuclear is decades ahead of every other field in terms of risk management.
      - jacquesm 4y ago
        I'm sorry, I don't believe in magic or appeals to authority, I believe in physics and once you start bending the rules for one part who is to tell where it will stop. Having one leg too short won't be fixed by making the other one a little longer. The materials science requirements on nuclear reactor vessels are such that if the safety margins are so narrow that the specs are going to be even possible to be violated by a manufacturing defect that you have a problem anyway, either in the selection of your manufacturing partners, your design process, your risk assessment, your QA or all of the above. For me that breaks trust at a level that I'm not going to buy the 'we'll fix it by using better parts somewhere else' argument. The stated defect has effects that can only be properly ascertained by destructive testing and since the whole point was that there was only one of these that leaves a large part of the end result in the realm of 'hope' and 'belief', neither of which are part of any materials science course that I'm familiar with. The knock on effect of changing the vessel specs would essentially require a complete rework unless the reactor were de-rated in either safety, generation capacity or lifespan. And because 'safety' is the only one that doesn't show up in the books that's where my worry would be. 'special teams' don't amount to much, we've had special teams all along and a long series of fuck ups. This is exactly the problem.
ncmncm 4y ago
What you are talking about here is called "normalizing deviance". We have had very bad experience with how that process plays out. It is particularly pernicious around safety.
- otter-rock 4y ago
  That's not what it's called. Imagine you write software for rabbit-mq. But then you get a requirements change to use zero-mq for whatever reason. Is that deviance? Have you normalized it? No, and no. Deviance is when you're dropping messages in production when you shouldn't. Normalizing is when you ignore it because "it's usually fine". You're probably thinking that my analogy doesn't work because changing software libraries is different than accepting an out of spec part. But why? If it's out of spec, it's a different part. You can make a spec around that part and design to it all the same. Just because it wasn't your first choice doesn't mean it's dangerous.
  - ncmncm 4y ago
    Point is that it is arbitrarily hard to tell whether a response is adaptive. Maladaptive responses are much cheaper than (unmentioned) alternatives, so are systematically more attractive to management.
    - otter-rock 4y ago
      So nuclear engineers just aren't allowed to do their jobs, and the safety requirements were never good enough to begin with.
      - ncmncm 4y ago
        They certainly were not good enough at Fukushima, in a half-dozen particulars. The next pile of reactors down the coast did not melt down, solely on the strength of a single engineer who succeeded, with enormous difficulty, in upholding standards management objected to, but that he knew were essential. Without him, there would now be two major uninhabitable areas there. He must have known Fukushima would melt down, and must have wanted it held to the same standard as his, but lacked authority to extend his standard so far.

jacquesm 4y ago

Yes, but the idea here was to construct a nuclear power plant, not to build a large pot for boiling soup in. You can't take a critical component like that, spec it and then suddenly pretend the spec never mattered in the first place, then you have to admit that you're just winging it. Changing the spec of the reactor vessel essentially translates into a complete redesign of the reactor itself unless you are willing to compromise on other aspects, such as safety, longevity and so on. We're not talking about a bracket here. Or an O-ring. When was the last time something as stupid as an O-ring decided the fate of... oh, never mind.

otter-rock 4y ago
It's a requirements change. Those happen all the time in everything. Why would that be completely forbidden or impossible here? Like you said, it might really suck. But the NRC does not allow trading off safety the way you suggest.
- jacquesm 4y ago
  Requirements changes are not driven by one-off material defects in critical pieces of hardware. That's simply a bending of the rules for economic reasons. And it is one of the main reasons for me to oppose nuclear: the fact that people will be people and that at the root of every one of those disaster and near disasters there was someone who thought they could get away with something. We are ill equipped to deal with this kind of responsibility, especially across a timeframe measures in decades. At the same time I would love to see us solve the climate change problem, and I recognize that we will likely have a nuclear component in there. But it will have to be done by the book or we'll end up regretting it - again. If we're going to start out with the normalization of deviance on the #1 critical component of a reactor then I think we are on the wrong path: https://becht.com/becht-blog/entry/normalization-of-deviance-the-pathway-to-disaster/ https://becht.com/becht-blog/entry/normalization-of-deviance...
  - otter-rock 4y ago
    I see your concerns, and I think you're probably just unaware that they're addressed better than you know, but in a different way than you'd expect. When they accept an expensive out-of-spec part, it's because they can safely redesign the less expensive parts. The only thing that hurts is their long-term profitability because it'll be less efficient than they hoped. Big, expensive parts are always treated as one-of-a-kind articles with unique requirements. It's how bridges and roads get built, for example. With nuclear, they have special teams that do nothing but manage every type of risk you could ever think of for every piece of nuclear material in the country. It's very advanced, and many industries are trying to use those techniques themselves now. Nobody is normalizing or bending anything. Every potential problem is thoroughly addressed on an individual basis. There is no deviance from the safety requirements. Nuclear is decades ahead of every other field in terms of risk management.
    - jacquesm 4y ago
      I'm sorry, I don't believe in magic or appeals to authority, I believe in physics and once you start bending the rules for one part who is to tell where it will stop. Having one leg too short won't be fixed by making the other one a little longer. The materials science requirements on nuclear reactor vessels are such that if the safety margins are so narrow that the specs are going to be even possible to be violated by a manufacturing defect that you have a problem anyway, either in the selection of your manufacturing partners, your design process, your risk assessment, your QA or all of the above. For me that breaks trust at a level that I'm not going to buy the 'we'll fix it by using better parts somewhere else' argument. The stated defect has effects that can only be properly ascertained by destructive testing and since the whole point was that there was only one of these that leaves a large part of the end result in the realm of 'hope' and 'belief', neither of which are part of any materials science course that I'm familiar with. The knock on effect of changing the vessel specs would essentially require a complete rework unless the reactor were de-rated in either safety, generation capacity or lifespan. And because 'safety' is the only one that doesn't show up in the books that's where my worry would be. 'special teams' don't amount to much, we've had special teams all along and a long series of fuck ups. This is exactly the problem.

ncmncm 4y ago

What you are talking about here is called "normalizing deviance". We have had very bad experience with how that process plays out. It is particularly pernicious around safety.

otter-rock 4y ago
That's not what it's called. Imagine you write software for rabbit-mq. But then you get a requirements change to use zero-mq for whatever reason. Is that deviance? Have you normalized it? No, and no. Deviance is when you're dropping messages in production when you shouldn't. Normalizing is when you ignore it because "it's usually fine". You're probably thinking that my analogy doesn't work because changing software libraries is different than accepting an out of spec part. But why? If it's out of spec, it's a different part. You can make a spec around that part and design to it all the same. Just because it wasn't your first choice doesn't mean it's dangerous.
- ncmncm 4y ago
  Point is that it is arbitrarily hard to tell whether a response is adaptive. Maladaptive responses are much cheaper than (unmentioned) alternatives, so are systematically more attractive to management.
  - otter-rock 4y ago
    So nuclear engineers just aren't allowed to do their jobs, and the safety requirements were never good enough to begin with.
    - ncmncm 4y ago
      They certainly were not good enough at Fukushima, in a half-dozen particulars. The next pile of reactors down the coast did not melt down, solely on the strength of a single engineer who succeeded, with enormous difficulty, in upholding standards management objected to, but that he knew were essential. Without him, there would now be two major uninhabitable areas there. He must have known Fukushima would melt down, and must have wanted it held to the same standard as his, but lacked authority to extend his standard so far.