Hacker News Reader

Comment by xg15 | Hacker News Reader

Can anyone explain to me why "deaths/tWh" is even a meaningful measure?

Of course nuclear energy has one of the highest Wh outputs, no-one is disputing that. However, what does that have to do with the risk of use? That seems like a measure very skewed to make arguments in favour of nuclear power.

I might as well argue that car drivers are safer than pedestrians because the average deaths/horse power is vastly lower.

Also, why did they leave away hydro, water and wind power in those "deaths per x" charts?

View on HN

How is it skewed? It's a cost-benefit analysis. tWh is the benefit, deaths are the safety cost. In other words, if your country needs X amount of power, and you know the deaths/tWh for each energy source, then you can figure out how many people would die if your country used that energy source to produce its power.

> I might as well argue that car drivers are safer than pedestrians because the average deaths/horse power is vastly lower.

This is a bad analogy. A good analogy would be: if you had to travel 1000 miles, it would be safer to do it by car than on foot because car is safer per mile travelled.

> Of course nuclear energy has one of the highest Wh outputs, no-one is disputing that. However, what does that have to do with the risk of use? That seems like a measure very skewed to make arguments in favour of nuclear power.

Erm, wtf? The world's energy requirements are a more-or-less fixed number of tWh, no? Like, if 1 nuclear reactor produces as much energy as 10000 wind turbines, comparing the deaths that would be caused by 1 reactor to the deaths that would be caused by 10000 turbines seems like the obviously correct comparison to make.

Wikipedia has some numbers. Hydro has caused the largest single-event catastrophe.

https://en.wikipedia.org/wiki/Energy_accidents

Relevantly, some nuclear plants are paired with hydro schemes, in order to make the flat output of nuclear plants economically match the bumpy daytime peak of consumer demand.

The reason why many nuclear plants are colocated with hydroelectric dams is that the dam is good source for cooling water in 3rd cooling loop and that it is good backup power source for black start of the nuclear plant. Ability of nuclear plant to scale its output power according to demand is usually significantly higher than of hydro.

It's meaningful because it provides a rational basis for comparing the risk of various power sources. It's the same as comparing deaths/mile to evaluate self-driving cars.

People are overwhelmingly bad at risk assessment, which is why one can be anti-nuclear and afraid of terrorist attacks (both lower risk), while happily driving to work at a coal mine (both higher risk).

Terawatt-hours are a direct measurement of what we want from an energy source, namely, energy. The equivalent for transportation would be deaths/mile, which is indeed a commonly used metric of transportation safety. Deaths/TWh is very meaningful.

It's not a perfect measurement, as Nuclear is similar to solar in that it provides power to it's own schedule. Normally this will mean gas or coal being throttled below their capacity. It's possible that this affects their deaths/TWh negatively, though perhaps not as maybe the sourcing of the fuel is the big problem with those plants, but e.g. if it's 5 people who die building the nuclear and gas plant, and then the gas plant throttles down overnight because the nuclear plant can't, then the nuclear plant wins unfairly.

It's the absolute opposite of skewed. It's fair and representative. If a country needs XX TWh/year, this is exactly how many deaths will be caused by the various production methods.

You car analogy is on the other hand completely wrong: the death/HP is a meaningless indicator. How about deaths per mile travelled? Check out the values at https://en.m.wikipedia.org/wiki/Micromort which has a section for traveling (named Additional)

I'm only aware of two deaths specifically caused by wind power, when two engineers were caught on top of one when it caught fire. It's likely such a negligible number that it wouldn't even show up on the graph.

Given that the point of the graph is to show nuclear at the bottom, wouldn't that especially be a reason to include wind power in the graph?

What about falls? They don't happen often but it seems clear enough that they are attributable to wind power.

According to this, there are multiple fatal accidents in the wind power industry each year: http://www.caithnesswindfarms.co.uk/accidents.pdf

Wikipedia's stats on energy production fatalities have wind being pretty decent, at 150 deaths per PWh (rooftop solar is 440, US hydro is 5, US nuclear is 0.01, and fossil fuels are in the thousands), but if the number is negligible it's only because wind power itself is negligible: https://en.wikipedia.org/wiki/Energy_accidents#Fatalities

Why do you think it's not meaningful? What would be the alternative?

Deaths per capita within 100 miles of generation.

Do you think coal mining deaths should not be counted if they occur 200 miles away from the power station the mine supplies? What about if the deaths occurred a continent away?

Here are a few measurable statictics that Nuclear would win hands down:

* Square miles of uninhabitable land produced/tWh

* Fishing industries destroyed/tWh

* Agricultural land destroyed/tWh

* Peoples displaced/tWh

* Lethal toxic waste produced/tWh

* Clean up cost/tWh

I'm pretty sure hydroelectric would actually win statistics 1, 3, and 4 "hands down". It'd also be a serious contender for 2 and 6.

This is of course excluding coal, which I'm 92% sure makes even hydroelectric (let alone nuclear) look like a peaceful meadow full of fairies and butterflies by comparison.

All I can say is I'd rather be spending time catching fish on a lake created by a hydroelectric system, then cooking those fish at night in the camp site by the lake, than doing the same anywhere inside the Chernobyl 30 km Exclusion Zone or the Fukushima 20 km Exclusion Zone.

If it's fishing you want, the water reservoirs for nuclear power plants make for good spots. I used to live over by the Rancho Seco plant, and its backup water reservoir (and the surrounding land) is still maintained as a public park even decades after the plant shut down (and was maintained as such even when the plant was operational, from what I understand). Good trout over there.

The Chernobyl 30 km Exclusion Zone is actually not that bad, apart from the slight radioactivity. Since there's no human interference, it's essentially a sanctuary for local wildlife.

When you say "win," do you mean be worse? Because I think the others mean it as being better, and I'm getting confused.

"Win" was (I'm guessing) used in a negative context to mean "worse", so I used it in a similar way. Kind of like how Hitler, Stalin, and Pol Pot tend to "win" competitions for murdering the most people.

You think so? I thought they were using "win" to say that nuclear was good, since fossil fuels are much worse on most if not all of those measures.

Isn't cancer one of the main risks of radioactive pollution? These deaths/X metrics don't account for disease.

Well because in the end the amount of power is what matters? I'm not sure I understand your point.

If you replace nuclear with renewable (or the other way around) you'll have to produce the same amount of power, all other things being equal.

Yes, however this assumes that number of deaths of any technology would linearly scale with its energy output.

e.g. if I built n times more wind turbines to match the energy output of nuclear energy then, according to this measure, I would also get n times the number of wind-power-related deaths.

I see no basis for that assumption, especially as nuclear energy has known risks that e.g. wind power doesn't.

> I see no basis for that assumption, especially as nuclear energy has known risks that e.g. wind power doesn't.

But why do the risks matter? One would expect that two turbines or nuclear power plants would kill twice as many people, on average…

Reply on news.ycombinator.com