Hacker News new | past | comments | ask | show | jobs | submit login

I don't want to live next to a nuclear power plant, but I would be fine with living near a dozen of them.

You see, that higher number implies a greater infrastructural and economic investment. Further, the sole large power plant in an area is automatically a military target, whereas if the same capacity were split across many facilities it becomes impractical to attack or control them all.

I'd love to see each municipality in the US above a certain population own and operate its own small reactor, using it to power the municipal utilities. But I do have a bit of a problem with a federal agency operating the only nuclear reactor in a 100-mile radius. It just ends up managed differently, becoming a political power center in addition to an electrical power center.




> You see, that higher number implies a greater infrastructural and economic investment.

A higher number of reactors also means more chances that one of them fails because reactors that don't exist can't fail but those that do most certainly can.

As such the security gains, from infrastructure synergies, would have to be massive to actually be able to offset that.


The reactor that does not exist fails by releasing radioactive fly ash from the coal plant that was never shut down, because nothing was ever built to replace it. Don't discount substitution effects. Your argument would not put safety belts in cars, because the belt that was never installed cannot strangle or entrap its passenger.

Also, do you know of any reasons why the 1000th instance of a design might be less prone to failure than the 1st, or 10th?

Can you think of any reasons why a car door handle might be more reliable (for the same cost) as dirigible door handles? There are many thousands of car door handles in use daily, such that all common failure modes have been seen, and then addressed in later manufactured models. The handle that fails can make the next handle made better able to avoid that specific failure mode.

You want things to fail just a little bit, but not enough to hurt anyone or cost too much money. If something fails, that means it isn't over-engineered for its intended purpose. And the failure point may then be examined to make the next design better, and improve upon existing maintenance strategies.


> The reactor that does not exist fails by releasing radioactive fly ash from the coal plant that was never shut down, because nothing was ever built to replace it. Don't discount substitution effects. Your argument would not put safety belts in cars, because the belt that was never installed cannot strangle or entrap its passenger.

That's a non-sequitur, there are alternatives besides coal just like there are more solutions to the problem than merely increasing energy production.

> Can you think of any reasons why a car door handle might be more reliable (for the same cost) as dirigible door handles?

A car handle is only one piece of a bigger machine, one could argue it's actually rather unimportant because if your car handle fails your car still drives, as such I'm not sure that's actually a good example.

How many iterations did we have on cars, as a whole system, so far? Over a century of designing cars and how close are we to a car that never fails? Which should be a way easier task than trying to make nuclear reactor safe, we had more time for it and even way more need for it, yet we are still nowhere close to having our "perfect cars", as such I just don't see how "perfect nuclear" is anywhere in our reach.




Applications are open for YC Winter 2020

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: