Fukushima

Because ju-st stated "I think the engineering and construction [of fission reactors] is not a solved problem" [0]. This is untrue. We have well-tested fission reactor designs and know how to put them together. This is a solved problem.

> Is this the study that took all roof-related deaths as solar-panel-related?

I can't parse this. Would you be so kind as to restate the question? :)

I'm not sure why the fact that a nuclear plant (that's been -as you say- running past its optimal design life) that

* Got hit with the fifth largest earthquake ever recorded

* Then got hit immediately after by a twenty foot high tsunami

* Then exploded

* Then melted down

kills fewer people than a single average coal-fired power plant [1] kills every year as a part of normal operation, fails to significantly increase the cancer risk of the majority of on-site containment workers, and fails to create centuries-long contamination [2] fails to make you question your assumptions about the failure modes of these power generation devices.

[0] https://news.ycombinator.com/item?id=10259794

[1] According to http://www.eia.gov/electricity/annual/html/epa_04_01.html , there were 300 such coal plants in the US in 2013. :)

[2] http://www.huffingtonpost.com/2015/01/04/fukushima-rice_n_64... Four-and-a-half years is far less than 100 or 200 or 500. :)

I don't know what study that is but it doesn't appear to be this one. Scott Alexander's blog post links to [0], which takes a completely reasonable approach to calculating the number of deaths due to solar panel installation. Search for "Falls are the leading cause of fatalities in the construction industry." and read on from there.

> Designing reactors is not the problem. Running them safely and building and decommissioning them efficiently is.

There have been two accidents that caused serious damage of any kind in nuclear power generation's 61 year history, [1] Fukushima and Chernobyl. As I mentioned before, Fukushima was hit by the fifth-largest earthquake ever recorded, and a twenty-foot tsunami, and has managed to only maybe kill two people (those guys died of blood loss, rather than radiation poisoning), and might contribute to ~150 cancer deaths as the years go by. This puts Fukushima's worst-case death rate at ~3.75 deaths per year.

What's more, -as I linked upthread- the area is now producing food that's 100% safe to eat. We're a substantial way down the road to recovering from the second largest disaster in nuclear power's history and it only took four years.

An absurdly pessimistic estimate of all deaths and potential deaths due to increased cancer risk related to fission power plant operation over the past 61 years puts the number at ~4,500. Even though this is a silly and inappropriate metric, that's ~73 deaths per year. If we look at deaths in just the US, choking on one's own food is an order of magnitude more hazardous than that.

> ...am I supposed to exclude nuclear explosions from the list of risks of a nuclear reactor?

Yes.

No power generation reactor has ever undergone a nuclear explosion. No currently operating power generation reactor, nor any such reactor that would be approved for construction can undergo a nuclear explosion. What's more, modern reactor designs (that is, reactor designs a fair bit newer than those at Fukushima) fail-safe in any conditions, short of deliberate sabotage by super-strong, radiation-proof superheroes that are able to enter the reactor vessel while it's in operation and intelligently tamper with the fuel rods contained within.

Not even Chernobyl underwent a nuclear explosion. [2] The explosion at Chernobyl was its steam piping exploding due to overpressure. The Chernobyl reactor was housed in an unhardened structure [4] that was more suited to storing goods and foodstuffs than containing fires and explosions. The steam explosion wrecked the containment structure, allowing the smoke from the flaming irradiated carbon control rods to escape into the surrounding countryside.

> You think I want coal?

I -charitably- think that you want safe power generation. I don't see why you're spending so much time and effort on combating something that has -worst case- over 61 years killed 15% of the the people power generation with coal-fired plants kills just in the US every single year.

Fight the big fights first: Decommission fossil-fuel-powered power plants. Dramatically increase the standards for hydroelectric dam design, construction, and maintenance. Institute programs that ensure that all workers who work at dangerously high elevations are properly secured against falls. Design and deploy helper bots to recognize and deliver aid to those who are choking on their own food.

Once all that's done, move on to shutting down something that's less hazardous than crossing the road or eating one's dinner.

[0] http://nextbigfuture.com/2008/03/deaths-per-twh-for-all-ener...

[1] This is the first commercial power generation fission reactor. It ran from mid 1954 until mid 2002: https://en.wikipedia.org/wiki/Obninsk_Nuclear_Power_Plant

[2] Even though a undergraduate-level physics student can design a bomb capable of undergoing a nuclear explosion, it requires deliberate effort to design and precise engineering to make such a thing. [3] You cannot turn a fission reactor into a nuclear bomb, unless you remove its fuel rods and use those to make a bomb. But then, we're talking about fission reactor safety, not the general safety of fissile materials.

[3] Note that a bomb that undergoes a nuclear explosion is very different from a "dirty bomb". A dirty bomb is made by wrapping conventional explosives with radioactive material. The explosives are used to spread the material around as far as is possible. No nuclear explosion takes place.

[4] This was widely regarded as a bad idea at the time. No power generation reactor in operation today is housed in an unhardened structure.