I think that nuclear power generation has been relatively safe over almost a century exactly due to systems engineering, incorporating procedures and the potential for human error.
I also think that you are right that some of the systems surrounding nuclear energy are suboptimally managed in places (and are often intransparent for the public, which I think is one of the factors why people are still afraid). And I think it's because nuclear power generation is a social commitment on a very long time scale - and humans aren't very good at dealing with cause and effect over long time scales.
While there is, just like in aeronautics, a paper trail connecting to each engineer and expert working on a system, it does not matter whether society can identify a person that is responsible for a decision made half a century ago. Think about leaky permanent[0] waste storage close to densely populated areas (like in Germany). Analyzing what went wrong is suddenly a task for historians, not for engineers. And history is of course necessarily colored in perspective - politicans can point to an expert who was wrong half a century ago.
In engineering and safety culture, iteration is necessary to approach an optimal system - the problem is here: How can we learn from what happened decades ago when the cause wasn't a part, a design or a process, something that can be simulated with a high degree of confidence? But a decision necessarily based on lots of unknowns? What can we do to be able to learn from it next time?
[0] officially temporary but due to how it was done effectively permanent
In engineering and safety culture, iteration is necessary to approach an optimal system - the problem is here: How can we learn from what happened decades ago when the cause wasn't a part, a design or a process, something that can be simulated with a high degree of confidence? But a decision necessarily based on lots of unknowns? What can we do to be able to learn from it next time?
The answer is that you design systems with multiple levels of redundancy, and then monitor them.
That way in order for something bad to happen, four independent things have to go wrong at once. Then you monitor if any of them ever fails. If one does, the other three prevent any catastrophe while you correct it. Then you have a new data point about how to design things better going forward.
The thing is that we have a decent idea of how to engineer things well. That's easy. Throw money and factors of 2 at it until the cows come home. That's not my concern. How do we know they'll actually do the proper engineering? How do we not have another Boeing 737 situation but for a nuclear reactor?
I know as engineers we want to turn everything into an engineering problem but sometimes there are just humans who don't understand the severity of their actions to cut costs or defer maintenance or something and I don't know if it's possible to ever fully engineer these people away or engineer away their influence or engineer away their incentive to do the wrong thing... as much as I'd love to do that.
Maybe I'm just being a naysayer. The thing is that I still believe nuclear power is the best interim solution in existence and that it's necessary to build more nuclear plants. I just don't buy the "totally safe" argument. It's like chemo or amphotericin or something. It'd be great to not have to use it... buuuuut....
Now you're asking a completely different question. It's not a matter of not knowing what to do, it's a matter of not doing it.
But nothing about nuclear is unique in that regard. If you have incompetence then we get Deepwater Horizon, Bophal, companies dumping cadmium, lead, mercury, cyanide, nitrogen, phosphorous or acid in to rivers and groundwater, mining runoff, destroying rainforest, dam failures, basically the entire list of Superfund sites in the US. Those kinds of things can be significantly worse than even a full scale nuclear meltdown. And on top of that, chemical pollutants have no half-life -- they're forever.
There is a 3700 acre coal mine fire that has been burning in Pennsylvania since 1962, causing the town to be abandoned:
It's the same problem with anything -- even solar is silicon. Take a look a the environmental history of Silicon Valley for a morbid lesson in how wrong that can go.
That is a very interesting point - would I rather live next to a DuPont chemical plant or a nuclear power plant? I think I would choose the nuclear power plant.
Edit: With the nuclear plant I at least have feasible means to monitor any danger by myself.
Comments
I think that nuclear power generation has been relatively safe over almost a century exactly due to systems engineering, incorporating procedures and the potential for human error.
I also think that you are right that some of the systems surrounding nuclear energy are suboptimally managed in places (and are often intransparent for the public, which I think is one of the factors why people are still afraid). And I think it's because nuclear power generation is a social commitment on a very long time scale - and humans aren't very good at dealing with cause and effect over long time scales.
While there is, just like in aeronautics, a paper trail connecting to each engineer and expert working on a system, it does not matter whether society can identify a person that is responsible for a decision made half a century ago. Think about leaky permanent[0] waste storage close to densely populated areas (like in Germany). Analyzing what went wrong is suddenly a task for historians, not for engineers. And history is of course necessarily colored in perspective - politicans can point to an expert who was wrong half a century ago.
In engineering and safety culture, iteration is necessary to approach an optimal system - the problem is here: How can we learn from what happened decades ago when the cause wasn't a part, a design or a process, something that can be simulated with a high degree of confidence? But a decision necessarily based on lots of unknowns? What can we do to be able to learn from it next time?
[0] officially temporary but due to how it was done effectively permanent
Edit: Markup
The answer is that you design systems with multiple levels of redundancy, and then monitor them.
That way in order for something bad to happen, four independent things have to go wrong at once. Then you monitor if any of them ever fails. If one does, the other three prevent any catastrophe while you correct it. Then you have a new data point about how to design things better going forward.
The thing is that we have a decent idea of how to engineer things well. That's easy. Throw money and factors of 2 at it until the cows come home. That's not my concern. How do we know they'll actually do the proper engineering? How do we not have another Boeing 737 situation but for a nuclear reactor?
I know as engineers we want to turn everything into an engineering problem but sometimes there are just humans who don't understand the severity of their actions to cut costs or defer maintenance or something and I don't know if it's possible to ever fully engineer these people away or engineer away their influence or engineer away their incentive to do the wrong thing... as much as I'd love to do that.
Maybe I'm just being a naysayer. The thing is that I still believe nuclear power is the best interim solution in existence and that it's necessary to build more nuclear plants. I just don't buy the "totally safe" argument. It's like chemo or amphotericin or something. It'd be great to not have to use it... buuuuut....
Now you're asking a completely different question. It's not a matter of not knowing what to do, it's a matter of not doing it.
But nothing about nuclear is unique in that regard. If you have incompetence then we get Deepwater Horizon, Bophal, companies dumping cadmium, lead, mercury, cyanide, nitrogen, phosphorous or acid in to rivers and groundwater, mining runoff, destroying rainforest, dam failures, basically the entire list of Superfund sites in the US. Those kinds of things can be significantly worse than even a full scale nuclear meltdown. And on top of that, chemical pollutants have no half-life -- they're forever.
There is a 3700 acre coal mine fire that has been burning in Pennsylvania since 1962, causing the town to be abandoned:
https://en.wikipedia.org/wiki/Centralia_mine_fire
It's the same problem with anything -- even solar is silicon. Take a look a the environmental history of Silicon Valley for a morbid lesson in how wrong that can go.
That is a very interesting point - would I rather live next to a DuPont chemical plant or a nuclear power plant? I think I would choose the nuclear power plant.
Edit: With the nuclear plant I at least have feasible means to monitor any danger by myself.