As a matter of fact I do. I’ll post proper data this evening or tomorrow, but let’s start with a figure of 32 deaths/TWh from a particular plant with no FGD or SCR. It’s very location dependent also.
I grant that coal is not the only, or in some regions not even the most common, alternative to nuclear.
However, of the types of generators commonly used, I think coal and nuclear are the only two that have nigh-inexhaustible fuel supplies.
[ul]
[li]Hydroelectric - there are only so many rivers.[/li][li]Natural gas/oil - I think that supply will cross demand in the next 20 years, resulting in huge price increases. No, I don’t have a cite for that.[/li][li]Nuclear - all the uranium we could want can be mined.[/li][li]Coal - Anthracite will correct me if I’m wrong, but I believe coal reserves are something like 20 to 100 times larger than oil reserves. Maybe more.[/li][li]Solar is limited by how much power the sun dumps on us, and requires vast land resources - though point-generation is going to be more of an option.[/li][li]Wind farms similarly require vast land area.[/li][/ul]
My point is that in the face of unfettered power demand growth, we’ll need either coal or nuclear - no matter what we are using now.
A factual addition, and a couple comments.
Airborne Mercury emissions from coal-fired power plants, while the absolute numbers look rather small, are in fact a significant concern. Coal plants are the largest source of human-generated mercury emissions in the country, and the U.S. EPA has identfied mercury as the toxic of greatest
concern from power plants. [EPA will be coming out by 2004 with regulations on mercury emissions from coal and oil fired plant, to take effect most likely by 2006 or 2007.]
However, to say that nuclear power plants have ‘zero emissions’ is a tad disingenous. Maybe your definition of ‘emission’ doesn’t include spent fuel and the radioactive debris left over after decommissioning the plant, but it’s still a pretty big waste problem. In fact, if you consider size of the problem x number of years you have the problem, the radioactive waste issue is just giant compared to anything coming out of a coal plant. And we have no real solution for it (other than, “bury it, and hope nothing happens in the next few thousand years”)
And (hope this isn’t getting too much of a Great Debate) I think douglips “Coal or Nuclear” choice is a false dilemma (one that the nuclear industry would like us to think is true, of course). Conservation is a huge ‘source’. According to the Rocky Mountain Institute (eco-efficiency freaks, yes, but very solid researchers), “since the Arab oil embargo in 1973, the United States has gotten more than four times as much new energy from savings as from all net expansions of domestic energy supplies put together” Likewise, other sources could be developed much further, especially if we invested a tiny fraction of what we’ve spent subsidizing nuclear power. I’ll cut this off now, but if anyone wants a nuclear debate, let me know.
In fact, in the long run, solar (and the indirect solar of wind, hydro, and biomass) is the only ‘nigh-inexaustible fuel supply’. (well, for the next, what, couple hundred million years, anyway)
Two quick additions.
First, for more info on energy efficiency and alternatives to nuke/coal, one place to start is Rocky Mountain Institute: http://www.rmi.org/sitepages/pid318.asp
Second, on death figures from air pollution. Most studies are epedemiological studies that take daily air pollution records for, say a city, and daily death records for the same area, and look to see if there’s any correlation. Good studies will also look at pollution over a few days at a time (to account for people who took a day or two to die from the pollution), and try and factor out other causes of increased mortality (like high temperatures, pollutants besides the one you’re looking at, time of year, etc.)
The basic finding from most studies is that people die more when there’s a lot of tiny dust in the air (PM2.5 is the technical term). Once you have a value for that correlation, it’s just doing some math to calculate deaths/year from PM pollution. The limitations here are that many of these deaths are people who are already sick or dying, so the real effects might seem exaggerated. On the other hand, it’s very hard to make good estimates of the long-term effects of air pollution, like how damaged are your lungs after 30 years of breathing soot, and does that mean you’re more vulnerable to pneumonia or something, so most mortality estimates don’t include long-term damage, which could be serious.
Personally, I think the biggest tragedy of all is that natural gas, which is present at just about every oil well, is usually just burned off when the oil is pumped b/c it isn’t cost-effective to collect and ship the gas. What a waste that we can’t figure out something better there.
Okay, I’m coming back with some detail as promised. But first, quercus is quite correct about the need to include delayed or chronic illness or mortality, and this was in fact included in the figure “32” I mentioned earlier.
Basically, leaving out global warming, which is as I said very uncertain, the dominant impacts by far are public health impacts from atmospheric emissions at the generation plant.
An example of how this is calculated. First work out how much SO[sub]2[/sub] is emitted from your plant. Use dispersion modelling to find out where it goes. This requires weather data inputs (wind speed and direction, rainfall, mixing height). Using a finite element grid, we can calculate the extra concentration of sulphate particles at each location in the study area.
An exposure-response function is used to work out how many people will suffer a particular illness as a result of the increase in concentration. For example, the exposure-response function relating sulphate concentration to chronic bronchitis in adults may be:
7.8 x 10[sup]-5[/sup] cases per year per person per ug/m[sup]3[/sup]
In the example coal plant it works out as 21 cases/TWh. For the purpose of expressing this impact in monetary terms, this illness has been valued by economists to be “worth” EUR 105,000/case.
Considering mortality, it is probably better to consider “Years of Life Lost” than “Deaths”. There are two reasons for this. First, many people who die as a result of air pollution do not have long to live anyway. Secondly, the air pollution is likely only one of several factors in an individual’s death.
For the sample station, the number of years of life lost works out at around 320/TWh. To get to a number of excess deaths, you may crudely divide by ten to get 32.
Well, hibernicus, I must say I am learning some very interesting things about the health effects of coal power from you. Good show.
Overall, it looks like even per GW*hr that coal power may be responsible for many more deaths than nuclear, which is as I expected. However, the magnitude of the difference may be quite a bit more than I thought. There still, IMO, is not a clear link between particulates and other emissions and proof of deaths, and I don’t think the nuclear waste issue has been considered fully. Still, it really looks like nuclear will end up being much safer overall by most any measure.
This sounds very high to me (32 years of life lost per GWh). I will do a rough, BOTE calculation to demonstrate:
In Austin, TX, the city utility’s peak capacity is about 2500MW (2.5GW). The actual time averaged usage is about half of peak capacity (this is a SWAG). This gives us an annual usage of a little over 20,000 GWh/year. Not all of our power is from coal (we have coal, NG, fuel oil, nuclear, hydro, photovoltaic and wind supplying our utility, with coal & nuclear being the base load workhorses and hydro supplying a small amount of the base load), but if it was that would equal 640,000 life-years lost per year. Since the utility serves somewhere between 541,000 persons (1990 census city population) and 1,000,000 persons (rough estimate of current population served including those outside of city limits), this comes out to somewhere between 0.64 and 1.2 life-years lost per person per year.
Without anything else to kill us, that would tell me that the life expectancy here should be somewhere between 55 (=120/(1+1.2)) and 73 (=120/(1+.64)), assuming a maximum life potential of 120 years. Since there are MANY other things to kill us, the life expectancy should be much lower. The fact that most of us around here aren’t kicking off here at a real young age is what gives me a problem with the number.
Like I said, this is a rough BOTE calculation, not fully considering the generation source mix, how much of the generation (most of it away from the city) actually affects the inhabitants of the city, actual energy usage (as opposed to SWAG based on peak/average power load), etc.
It is high - he said 32 lives lost per TWh, or 320 years of life lost per TWh, but you come up with GWh. This introduces a factor of 1000 error in your calculations. You further confuse the “320 life-years ~= 32 lives” argument and come up with “32 life-years”. But let’s flip the envelope over and see what happens.
2.5 GW * .5 fudge factor * 24 hours/day * 365 days/year = about 11000 GWh/year. Correct?
Using the real numbers from above, I get:
11000 GWh/year * 320 years lost/TWh * .001TWh/GWh = 3500 life-years lost per year.
I get 1.2-2.4 life-days lost per person per year. If you lived there 100 years you’d lose less than 9 months of life on average, which leaves plenty of life to be reduced by car accidents, smoking, 16 oz. steaks, armadillo attacks, and other hazards of Texas life.
I hereby apply Leibniz’s Chainsaw (related to Occam’s Razor) to the rest of your calculations.
Does that sound more reasonable?
Now that made me laugh out loud. Mind if I borrow it?
I should have made one thing clearer. These deaths and illnesses were calculated for the whole of Europe, over a grid of some 3,000 by 2,000 km. The number of life-days lost in Austin because of electricity used in Austin will be very small (maybe 1% of your total; imagine if local power plant really did shorten the average life by 9 months!). The remainder will occur downwind of the plant.
Note also the following:
exposure-response functions are often higher for the US than for Europe, and for western than for eastern Europe. Very crudely put, this may be because members of less healthy populations are more likely to die of something else before pollution gets them.
The results for a particular power plant depend strongly on location, in particular on how many people live downwind of it. If for example you had a power station where the prevailing wind took the plume out over the open ocean, the number of excess deaths and illnesses would be greatly reduced. Also there is huge potential to reduce emissions and hence health impacts. Combine the two (clean coal, good location) and you might have a very different result.
Thanks, Anthracite. I actually will wear this one around the place for a while. My credibility could do with a boost!
By the way, for some reason I assumed you were British. Now I’m even more impressed how much you know about power stations in Britain and Ireland. Learn to spell “sulphur” ;)and you could do my job better than me!
No clear link, no. Just epidemiological studies. As quercus said, you look for statistical correlations, and factor out confounding factors. We can never say “This person died because of coal-burning power plant”.
Not British, but recently I certainly wish I could spend much more time in the UK than I currently do.
“sulphur”. Heh.
I am curious, as a momentary hijack - can you tell me in general terms (I do not ask you to reveal your IRL identity) what it is that you do?
And don’t get cocky. 
I still am Queen of Coal. But even the Queen must be humble enough to learn from others, so she can be more powerful and terrible in the future. And while I know an awful lot about coal and coal power, I have never really looked hard at health effects per se, such as shown in the links you provided. But Hell, there are but 19 hours in the day for me to work, and but 7 days a week…
OK, now that we have a figure of 32 deaths/TWh for coal power, let’s do some calculations for deaths per TWh for nuclear, making some of the above mentioned assumptions.
Let’s start with the assumption of total deaths due to nuclear power = 30,000. This includes 25,000 due to Chernobyl (already dead or still to die), and any other unfortunates along the way (e.g. those poor folks in Japan who brought a barrel of UF[sub]6[/sub] (or something) critical in the last year or so - they get lumped in here.) This should be a good back of the envelope number. Hopefully I won’t end up with one of TXLonghorn’s surrealistic envelopes.
Anyway, from This report on Nuclear Power from the DOE:
Some interesting calculations along the way:
Nuclear Power Capacity in 1998: 349 GW.
Energy capacity (i.e. energy generated 3,057 TWh
if all nuclear plants ran at 100% 24/7)
Generated energy in 1998: 2,291 TWh
(did I get the decimal in the right place?)
World's nuclear capacity factor: 75%
Groovy. That’s a boatload of energy.
Coal's estimated lives/TWh: 32
If Coal had generated that power instead
of nuclear, number of additional deaths: 73312
Is it really that simple? Holy Mother of Uranium. Bituminous Saints Preserve Us! In one year, Coal generated power in the amount of our world nuclear generation would have killed twice as many people as nuclear energy ever has?
Let’s go a bit further, and calculate deaths/TWh for nuclear power. I’ll use the above DOE figures for generation capacity, but because I don’t have historical figures and I’m too lazy to click on any but the topmost link on google, I’m going to use the forward projections and come up with a fudge factor for how much nuclear energy has been generated in the past twenty years, by taking the lowest figure for next 20 year capacity and reducing it by 20%, then multiplying by our capacity factor calculated above.
DOE capacity projections for nuclear: 303-368 GW
(for the next twenty years)
Pretend last twenty year average cap. is: 240 GW
Generated energy per year: 1577 TWh
(at 75% capacity factor)
Total energy generated in 20 years: 31,000 TWh
Oh, this is working out to be pretty easy to calculate.
Deaths by nuclear power: 30,000
Deaths per TWh by nuclear power: 1
OK, so I guess I’ve found the answer I was looking for, at least in a back of the envelope kind of way. Coal is roughly 30 times more deadly than nuclear per generated unit of energy. I’ve deliberatly tried to underestimate generated nuclear power in the past to make nuclear look worse. If you were to just apply 1/20 of the nuclear deaths to 1998, you’d end up with 1500 deaths vs. the proposed 73000 for coal you’d get a factor closer to 50.
This doesn’t address the ‘what if’ questions about nuclear waste disposal. I suspect that if there ever is a problem with nuclear waste leakage, it will be a localized problem such as contaminating groundwater. Exploding nuclear plants a la Chernobyl are far more devastating than a leaking radioactive waste dump, at least at first glance. These leaks will then start causing deaths in the affected areas, but at what rate? I think that type of question is probably destined for Great Debates, so I’ll leave it rhetorical for now.
Now, the question is, what about other forms of power that have been mentioned such as Natural Gas, or even the chemical waste stream from a solar cell plant? Hibernicus, do you have any other off-the-cuff numbers for deaths/TWh by any other power source?
This has been a simply fascinating thread, BTW. Good job for starting it douglips. Up to this point, except for wondering about the tie between particulates and other emissions and deaths still, I have to agree that the figures seem reasonable. This is what GQ threads are supposed to be about - I would say it is more GD, but we were searching together for an answer, rather than debating a point. Fun!
Una
douglips, I just want to point out that you’re comparing public health impacts from coal with major accident impacts from nuclear.
The main impacts from nuclear are identified as emissions of radon from abandoned mill tailings, atmospheric emissions of long-lived radionuclides like C-14 and I-129 from the reprocessing stage and also occupational accidents and emissions from the use of fossil fuels in the cycle.
The number of deaths from major accidents is only one measure and fails to take into account larger, if less spectacular, ongoing impacts.
Also, the major accidents figure results almost entirely from a single incident (Chernobyl), so we cannot really conclude anything about the safety of nuclear power in general. It is statistically meaningless.
Unfortunately I know nothing about nuclear power, so I really do recommend you go to the ExternE site if you are serious about looking for numbers. Download the Belgian nuclear report, read the summaries, and you’ll know more about the environmental impacts of nuclear power than I could ever tell you.
Hail, King and Queen of Coal:
Do environmental health effects depend on the “kind” of coal that is burned? Anthracite, what did you find in Polish plants? It seems that this debate has been largely based on efficient, well-run plants in the U.S. and U.K. However, in Central and Eastern Europe, it’s my understanding that they burn “brown” coal for heating energy, with disastrous consequences.
I can only give you anecdotal evidence. I lived in Czech Repbulic from 1994-1995. In winter, you could smell, taste, see, and feel the pollution from the brown coal in the city of Prague. I am a relatively healthy person, but when I lived there I had severe respiratory problems that lasted the entire winter. Every time I stepped outside, I could feel the pollution in my lungs, experiencing discomfort with every breath. I left the city every weekend to get away from it - and as soon as I was outside an urban area, I felt better. As soon as “heating season” ended, I recovered.
Maybe this should be a completely separate thread, but I wanted to ask what you know about “brown” coal and what the environmental consequences might be for a population exposed to emissions over the long term, in terms of air-pollution related illnesses such as athsma. What is the difference between “brown” coal and what we burn in the U.S.?
A coal virgin, easily blinded with science,
M
P.S. And no, let’s not debate the safety of East European nuclear plants vs. East European coal plants. Shudder.
Yes. I should have been more clear. My reasoning has been that, well, there are no major accident impacts from coal. About the worst thing I can think of happening is a coal mine collapsing and killing 100 miners. Tragic, but not on the scale of thousands of deaths like a nuclear catastrophe. Further, it was my WAG that besides major accidents, the other health impacts from nuclear were minor. However, in light of the sheer volume of power generated, minor effects add up (as is demonstrated by the coal numbers.) This may not have been a valid assumption.
Another assumption I was making was that any of these emissions would be comparable in some way to the emissions in the pre-coal process. For example, fossil fuels used in the cycle of fuel processing should be roughly canceled by fossil fuels used in the mining process for coal. Were your death figures taking into account all the pre-coal process? I was assuming that that was deaths/life lost due solely to the coal plant operation, and did not take into account all the pre-processing (mining, transportation, radon emitted from the coal mine itself, etc.) This was another reason I only wanted to include nuclear generation accidents/emissions, and not mining or other process effects - I didn’t think you counted it. Mea culpa if you did.
To me, this assumption (coal pre-process == nuclear pre-process) seemed rather reasonable, because while mine tailings from a uranium mine will be a source of radon, it is my further assumption that much much more mining must happen to get coal out of the ground per unit energy, and the sheer scale will balance it out. This may also not be a valid assumption - since uranium must be enriched, perhaps we have to dig up 800 tons of ore to get 1 kg of fuel - I don’t know for sure.
This gives another interesting side to the question - how does mining impact for the two compare, per unit energy yield?
True. I was using Chernobyl in an attempt to get nuclear power to look as bad as possible. I have heard nuclear engineers on the radio talk about how the Chernobyl design and other Soviet design is so inherently unsafe that you’d never get a license for it in the US or Europe. This implies to me that we will never have a disaster like Chernobyl except in the former USSR. Hopefully we’ll never have such an accident again, but by taking the highest estimate possible for the event, and adding more deaths on top of it, I was hoping for decent approximation.
I’ll have to check it out - perhaps I’ll have some time this weekend to go over some numbers.
In the meantime, let me sum up some questions for you and for further discussion:
[ul]
[li]Was the figure you quoted of 32 deaths/TWh just for coal generation emissions, or for the entire coal process (mining, transportation accidents, transportation emissions, person falls into coal pulveriser, etc.)?[/li][li]How much mining must be done per unit energy yield for coal vs. nuclear? (mining impact - say, tons of ore per TWh)[/li][li]What are the effects of the coal pre-process, if not covered in your 32 deaths/TWh? (mining, transport, pulverisation)[/li][li]Are there any effects from the coal post-process? (gathering ash that needs to be landfilled, transportation of said stuff, etc. Ignore saleable products such as fly-ash sold for industrial use, as I don’t want to degenerate into a discussion of how many mobsters get buried in concrete made from coal fly-ash…)[/li][li]What are the effects of the nuclear pre-process and post-process? (refining ore, enriching fuel, processing spent fuel, transportation to waste sites, etc.)[/li][li]Finally (and perhaps a Great Debate, but perhaps we can get some solid data in this thread), what are the risks in the long-term storage of spent nuclear fuel? Best case scenario/worst case scenario?[/ul][/li]
I don’t know if we’ll be able to cover it all, but this thread has been pretty cool so far and I don’t want to blow it now.
If you can help me out with some of the coal questions, I’ll try to dig up some nuclear numbers.
Well, I do hope you’re not referring to me. But it’s nice to know we’re not just talking among the three of us at this stage. I thought everyone else might have wandered off.
They certainly do. At the station I know best, we’ve moved from coal with 2.2% sulphur content to coal with 0.8% S. That’s a 64% reduction in SO[sub]2[/sub] straight away. Also fuel nitrogen and fuel ratios will vary.
But much more important is whether the exhaust is cleaned up before being emitted to atmosphere. Your Prague experiences were a result of high levels of particulates emitted from low chimneys and forming a thick smog at street level in cold, still conditions. The “heat island” effect is at work here.
Well, I’ll let Anthracite take the substantive question. But from an emissions point of view, the thing about lignite is it has such a low heating value that you have to burn a lot of it. So, even though it’s a very “clean” fuel, by the time you’ve burned enough to get the amount of energy you want, you’ve actually produced more pollution than if you had been burning bituminous coal.
I don’t know about uranium mining, but the figures we had for coal mining (USA) were:
Accidents 0.145 fatal + 14 major non-fatal /Mt coal
Lung Cancer 0.33 deaths /Mt coal
pneumoconiosis 0.35 deaths /Mt coal
These figures represent a huge human cost (and I’m aware many Dopers have lost relatives in this way), but are much lower than corresponding figures in developing countries.
Yes and no. The figures for every stage of the fuel cycle were calculated, including mining, transport, etc. It turned out that the generation stage was very dominant. I could add in the relatively small numbers from the other stages, but because they are small and the uncertainties are so great, it didn’t make sense to do so.
[quote]
**
In the meantime, let me sum up some questions for you and for further discussion:
[ul]
[li]Was the figure you quoted of 32 deaths/TWh just for coal generation emissions, or for the entire coal process (mining, transportation accidents, transportation emissions, person falls into coal pulveriser, etc.)?**[/li][/quote]
In theory everything was included (see below). In fact, the results for those other stages are smaller than the margin of error for the figure for generation emissions alone.
French uranium mines yield 0.15% U on average. It takes 22 t of mined uranium to make 1 TWh in a PWR. So, you have to mine 15,000 t of uranium ore to make 1 TWh.
It takes 360,000 t of coal to make 1 TWh.
From mining and preparation:
[list]Death and injury to mine-workers as a result of industrial accidents
Occupational diseases such as pneumoconiosis and cancer among mine-workers
Damage to (or removal of) surface habitats
Subsidence at the surface due to collapse of mine tunnels
Disruption of the water table
Noise and dust blow
Accumulation of solid waste
Leaching of contaminant materials (acid and trace elements) from coal waste dumps
Emission of global warming and pollutant gases from the exposed coal seam and from mining equipment[/ul]
From transportation:
[ul]Pollutant and global warming gas emissions from trains, ships, vehicles
Noise and fugitive coal dust generated during loading/unloading and in transit
Water pollution due to acid water drainage from coal bunkers
Occupational and public injury due to road/rail accidents, accidents at sea or loading/unloading
Visual impact and land use of storage facilities[/ul]
But if you’re talking deaths, the answer is safely less than 1/TWh, which is why I didn’t add it.
Transportation, land-use, leaching of heavy metals into groundwater, uptake of heavy metals into vegetation…
This depends completely on the coal, and this number has a huge amount of potential error in it.
Taking a plant with an annual average Net Plant Heat Rate of 9500 Btu/kWh, and looking at 4 types of “representative” coals, we get:
Texas Lignite - 6500 Btu/lbm - 730,770 tons to make 1 TWh
Black Thunder PRB - 8800 Btu/lbm - 539,772 tons to make 1TWh
Illinois Basin - 11,000 Btu/lbm - 431,818 tons to make 1 TWh
Central Appalacian - 13,000 Btu/lbm - 365,384 tons to make 1 TWh
So this number could be well more than double the previous number. These calcs also assumed a constant plant heat rate, which is false, since with the Texas Lignite and PRB the heat rate would be worse, and thus one might expect (roughly) 750,000 and 550,000 tons per TWh respectively.