Hacker News Reader

Comment by etrautmann | Hacker News Reader

Somehow they didn't include solar, wind, geothermal, or hyrdo in this analysis? It's still a good point though I'd imagine all of those would have similarly low mortality rates.

View on HN

Solar and wind might surprise you. Anytime you're working on a rooftop or on a tower, your risk of falling to your death is not zero.

Hydro is in a special case too, because a sudden dam collapse could kill thousands of people in the blink of an eye. Johnstown Flood Disaster comes to mind. Whether you blame the collapse on the hydropower is another question.

Call me a crackpot, but I believe geothermal may be worse than the rest put together. By leaching the heat the earth needs to "lubricate" its plate tectonics, I think you might be delaying, amplifying, and shifting the focus of the earthquakes. I think our overdue Hayward earthquake is being influenced by the GWh of heat energy taken out at Salton Trough and the Geysers. Force is building up like in an engine you let the oil out of. I theorize we're about to find out what the Earth-scale equivalent of throwing a rod is.

This is outside of my normal area of expertise, but from what I gather, the total geothermal power output was around 14 GWH in 2017 [1], with an estimated maximum of 2TW possible worldwide if we develop every spot available [2].

It looks like the current estimates of baseline power lost through plate tectonics from the mantle are 35-47 TW, "from secular cooling, internal radioactivity, and core heatflow across its base" [3,4]. To bring these numbers into scale, there is approximately 173,000 TW of heat energy bombarding the earth [5]. I am not able to find estimates for the total thermal energy stored in all of the earth, as there is decent debate about what the inner 'layers' are made of; however, it seems doubtful that we are making a large impact even on localized regions with a max current output of 14 GW when compared to the 35 TW per year lost naturally.

[1] https://www.iea.org/topics/renewables/geothermal/

[2] https://en.wikipedia.org/wiki/Geothermal_power

[3] http://adsabs.harvard.edu/abs/2016AGUFMDI52A..02D

[4] https://www.quora.com/How-much-energy-is-stored-in-the-earth...

[5] https://en.wikipedia.org/wiki/Earth%27s_internal_heat_budget

Did you switch units from wh to w?

And in 1975, Banqiao Dam in China killed 26,000 people immediately from flooding, and another 145,000 in the aftermath, according to China's official estimate.

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

>Anytime you're working on a rooftop or on a tower, your risk of falling to your death is not zero.

Okay, but I'd still be interested in:

a) A measure of death risk to the general public, i.e. people not directly involved in the production of the energy. It's important to distinguish internalized vs externalized risk.

b) A measure that accounts for the potential cost of preventing these deaths. That is, one that could capture insights like "source X and source Y currently have the same death count per kWh, but it would cost half has much to achieve a 10% reduction in deaths from X compared to Y".

The ratio of internal/external deaths would hugely effect the measure in b) -- it's much easier to stack on measures that can prevent deaths if they're employees of the provider than if it's the general public.

And so, on a), I would guess that measure for wind power is really hard to distinguish from zero.

> A measure of death risk to the general public, i.e. people not directly involved in the production of the energy. It's important to distinguish internalized vs externalized risk.

That's kinda creepy. Like we should be ok with Appalachian coal mining's horrible safety record because the general public never dies in mining collapses or gets black lung. Who cares about some faraway people just so long as I can run my crypto farm at under 5¢/kWh?

Very "Hunger Games".

certainly shouldn’t be okay with it, but if arbitrary people in a 100km radius were effected rather than just miners, yes that would be worse, because their “decision” to remain in the area is further removed from the problem than the decision to remain a miner

So what are you thoughts on coastal cities? Should the government do something about them?

My memory is every so often a major hurricane comes along and kills a bunch of people. The people who die probably didn't choose to put the city where it is. The risks are much higher than the nuclear industry.

And some follow up questions, because I always vent on Nuclear (sorry, don't mean to take it out on you, the lack of evidence based policy on environmental issues just gets me every time):

Is dying in a hurricane better than the health impacts of nuclear, which are many years delayed?

Do you think that the evacuation from the region around Fukashima (deaths: 10s of people), which killed more people than the radiation (deaths: 0) is evidence that the standard of "were effected" is too high?

Have you ever exposed yourself to so much solar radiation that you felt heat on your skin? Do you feel that this level of risk is remotely acceptable, given the number of deaths caused by skin cancer each year?

_PS_ I've worked on a coal mine. Deaths mining coal is already a crazy low figure, Australian coal mines are probably safer than supermarkets by the statistics. I'd be shocked if erecting and maintaining solar panels and wind turbines is safer than nuclear. Roofs and spinning blades will kill people.

>Call me a crackpot, but I believe geothermal may be worse than the rest put together. By leaching the heat the earth needs to "lubricate" its plate tectonics, I think you might be delaying, amplifying, and shifting the focus of the earthquakes.

I cannot imagine that we are anywhere close to technologically capable of having even a miniscule impact on plate tectonics.

Interesting. Trade “the planet’s climate” for “plate tectonics” and you’ll find a lot of like-minded Americans, particularly in Congress.

The Earth's crust has something like a million times the mass of its atmosphere.

Actually, TIL (a google search revealed) it's only 5 times more massive.

That is not correct either.

Earth's crust: 2.77 * 10^22 kg

http://adsabs.harvard.edu/abs/2007AGUFM.V33A1161P

Earth's atmosphere: 5.15 * 10^18 kg

https://en.wikipedia.org/wiki/Atmosphere_of_Earth#Density_an...

The crust is about five thousand times more massive.

The factor of 1000 difference is what you get when you accidentally forget to convert between kg and tonnes.

5000 it is.

The mass of the atmosphere is same as mass of 10m layer of water (because they both create 1 atm hydrostatic pressure). Obviously, the crust is much thicker than 10m, also much denser than 1000kg/m^3.

That's similar to what the fracking advocates said before they started causing earthquakes.

Causing earthquakes is easy (I lived close to a mine, and earthquakes happened). Affecting plate tectonics ... I feel like that is something entirely different.

I do agree it seems like a bad idea to leach heat from the inner surface of earth in large quantities. I actually feel the same way about wind as well (a bad idea to take enormous amounts of energy out of the weather system before understanding it).

Technically you don’t take it out. You just move the energy somewhere else. It all ends up back as heat eventually. We already move vast quantities of energy around with our food production and distribution systems.

What about solar? Change an area of land that previously reflected most of the light back into space to absorb it instead.

Lubricating these faults often has the reverse effect; it allows the plates to constantly slip and having more frequent, smaller earthquakes. That being said, the geysers quake a few years back hit the theoretical maximum for that portion of the fault so who really knows. See below for info on the topic as well as the Parkfield Experiment.

https://earthquake.usgs.gov/research/parkfield/fluids.php

Perhaps not!

There's a famous blogpost that claims that rooftop solar panels are about ten times more deadly than nuclear power stations, because people might fall off the roof when installing them: https://www.nextbigfuture.com/2008/03/deaths-per-twh-for-all...

The majority of American PV generation is from ground mounted solar farms, not rooftop systems. I have looked many times for statistics about worker mortality in solar farm construction, sadly to no avail. I believe that deaths-per-TWh is significantly lower for solar farms. Dangerous falls are much less common during the construction of large farms and panels in farms tend to be more productive than those mounted on rooftops.

In California, for example, distributed rooftop solar installations generated 10.7 TWh in 2017 while solar farms generated 23.1 TWh.

https://en.wikipedia.org/wiki/Solar_power_in_California#Gene...

I wonder how the figures change if you require solar in the first place?

Presumably, if solar was required by planning regulations (where sane), for new-build or major roof repairs, the installation risk could then be nearly eliminated as an extra factor, given you are going to have the problem of putting a new roof on anyway.

It would still take extra time or extra people to install solar alongside other work on the roof, and thus create extra risk.

Depends on how you are doing it. If the solar is also your roof surface it isn't an extra job, it is just a materials choice.

edit - also, given how much risk there is in the prep and teardown for scaffolding, combining roof jobs will generally reduce the overall risk a lot, even if you can't combine tasks any further than that.

With the difference that one person dropping dead from a roof hurts one family, a nuclear plant going off spoils a huge area for generations.

Which would make sense if one person fell off a roof for each exploded plant, but the ratio is a tiny bit bigger than that.

People don't risk falling off scaffoldings when building and maintaining nuclear power plants?

OSHA is much more vigilant in power plants than on random roofs.

So we need 4400 persons falling off a roof to be par with nuclear power plants. I still consider this ratio a much better outcome than a power plant going off anywhere near (radius to be defined in Mm depending on wind direction).

How? There has been a handful of nuclear plant accidents ever, but hundreds of thousands of people work on roofs every year worldwide to install solar panels.

As far as I can tell, the original pro-nuclear "deaths per TWh" calculation, and follow-up work, neglects mundane industrial accidents that occur during the construction and operation of nuclear power plants. Only incidents involving unplanned radiation exposure are counted. But these same kinds of mundane accidents do get counted in the numbers for renewable energy. Further, as I note in another comment, the solar PV numbers neglect that more TWh come from solar farms than from rooftop solar installations, and that constructing solar farms is safer (lower risk of fatal falls). The methodological difference isn't significant if you just want to highlight how dangerous the fossil fuel status quo is compared to non-combustion options. But if you want to compare deaths-from-renewables vs. deaths-from-nuclear, it doesn't make sense to omit the mundane accidents from just one side of the balance sheet.

As an example of "mundane" nuclear deaths, 6 workers have been killed at the Surry Nuclear Plant. 2 were fatally scalded by an accidental steam release in 1972. 4 were killed by a steam explosion in 1986.

https://en.wikipedia.org/wiki/Surry_Nuclear_Power_Plant#Even...

5 Japanese workers were killed by a 2004 steam accident at Mihama-3.

https://en.wikipedia.org/wiki/Mihama_Nuclear_Power_Plant#200...

Falling equipment killed a worker at Arkansas One in 2013.

https://en.wikipedia.org/wiki/Arkansas_Nuclear_One#March_201...

solar power produces more waste than nuclear.

https://www.energycentral.com/c/ec/are-we-headed-solar-waste...

https://www.foreignaffairs.com/reviews/review-essay/2017-08-...

I agree a comparison with those would be interesting however the article does explicitly state which sources are included in the analysis:

  Here we limit our comparison to the dominant energy sources—brown coal, coal, oil, gas, biomass and nuclear energy; in 2014 these sources accounted for about 96% of global energy production.1

The Banqiao dam failure alone is enough to drag hydro quite far up the list. Even excluding it, other disasters add up to worse outcomes than nuclear. Hydro plants apart from being an environmental disaster suffers from creating all kinds of hazards downstream if not maintained regularly. When they are, they're mostly ok, until they're not - consider e g.the Mosul dam in Iraq where lack of maintenance has significantly increased the risk to a couple of million people.

Some analyses have shown that nuclear beats out all those, too: https://www.iaea.org/sites/default/files/publications/magazi...

That article is from 1979. Wow.

Probably best taken with a grain of salt given how much solar in particular has ramped up.

Some of those numbers are ridiculous even for 1979. Looking at Figure 2, the metric tons of materials numbers for fossil fueled plants exclude the fuel, which is by far the largest material input to fossil fueled power.

Example: that figure shows material requirements of about 10 tons per megawatt-year for electricity from coal. A megawatt-year is 1000 * 24 * 365 = 8,760,000 kWh.

Kentucky Coal Education says that it takes about 1 pound (0.45 kg) of coal to produce 1 kWh of electricity. The actual material requirements for a megawatt-year of electricity from coal are upwards of 3,900 metric tons. Even with the low-efficiency PV modules of 1979, PV electricity required less mined material than coal electricity.

http://www.coaleducation.org/lessons/twe/ctele.htm

This discrepancy is ironic in a document that states in the introduction "If you calculate the risk of only part of a system and compare it with the corresponding part of another, by judiciously choosing the component you could prove that any energy system is riskier (or safer) than any other system. You would obviously be proving precisely nothing."

This method of proving precisely nothing is employed by the author in Figure 2, only regarding materials requirements instead of risk.

Here is a 2012 piece from the same author - 'The Impossible Dream? Why Renewables Won't Reduce CO2 Emissions by Much' - https://www.energycentral.com/c/pip/impossible-dream-why-ren...

Maybe it is just me, but I suspect he might just have a tiny little bit of bias.

Are those "major"?

In the USA, renewables provide two-thirds as much power as nuclear. If you want to claim nuclear is major, you have to accept renewables are too.

But it's not fair to group a bunch of different power sources all under "renewable." Nuclear is at 20%, the biggest renewable is hydro at 7.4%, followed by wind at 6.3%. I might call all of those reasonably major.

But solar PV falls way behind at 1.2%, geothermal is 0.4%, and solar thermal only 0.1%. I'd call those minor.

https://www.eia.gov/tools/faqs/faq.php?id=427&t=3

what's not fair is not the "grouping", but the consideration of "major"... to assess the merits of each approach to set their fractions you compare their intensive properties not their extensive ones.

What percentage of that renewable power generation is not backed by traditional fossil or nuclear based generator? I guess 0%

This why nuclear power should be THE answer to global warming. But it isn't ... Nuke industry is dead and it is still disparaged by large swaths of environmentalists.

That's not the whole truth. That two-thirds is often backed by often stand-by NatGas generators, complicating the issue on if those additional plants should be counted separately.

Depends on the country.

If you are looking at making an argument on safety for going forward, you should probably look at them.

Reply on news.ycombinator.com