An anti-nuke organization recently posted a
deceptive comment that implies that nuclear power is heavily dependent on mining while wind is not. I did the math and found that mining impact of nuclear power and wind is roughly comparable (wind requires considerably more steel and concrete) and both are two orders of magnitude lower than coal. Capital costs for wind and nuclear are initially similar for the same average capacity, but in the long term nuclear generates about three times as much power and displaces three times as much CO2 due to the longer operating lifetime of nuclear power plants. Deaths from wind and nuclear are dominated by electrocution from the resulting electricity, then non-radioactive mining polution and construction, with radiation effects being much smaller than the preceeding causes.
Lets look at the facts, based on an honest attempt to determine them by mining google and doing basic calculations and putting numbers in context. For an equivalent amount of power, wind turbines not only require mining, they amount of mining is in the same ballpark as nuclear. The actual amount of mining ranges from about three times less to three times more for wind compared to nuclear, depending on the grade of ore used - without reprocessing or breeders. And both require less than 1% as much mining as coal.
I am not trying to rag on ground based wind power. I like wind. But we need to be honest about the costs of wind. Wind still looks good compared to coal when you consider the costs. But asserting that wind power eliminates mining compared to nuclear just doesn't hold water.
Before we get into details, lets first look at a graph that shows the power generating capacity as a result of a constant annual expendature (overnight cost) of $52.8 billion a year, the amount needed to build one nuclear power plant per month.
(click on the image to enlarge)
Carbon Dioxide emissions would be rougly proportional to power generated. This graph shows a time period of 187 years. The capital expenditure is constant for 120 years; the remaining years show the residual capacity left over from that expendature. For 60 years, you are building new nuclear power plants, so generating capacity increases. Then during the next 60 years, plants are retiring as quickly as you build new ones so generating capacity holds steady; this phase would be continued indefinitely in the absence of alternatives. The last phase shows the ramping down of generating capacity after you stop investing. This is shown primarily to illustrate the residual power generating capacity of both wind and nuclear power plants. It would be unfair to both sources to cut off the graph when you stop spending money as the delayed return on investment would not be apparent. Wind and nuclear cost about the same amount of money per average gigawatt generated (rather than faceplate capacity). The problem with wind is the shorter lifespan of wind (20 years vs. 60 years). This means that after only 20 years (vs. 60), your money is going to replace old turbines rather than build new ones. Thus, for the same investment, nuclear ultimately produces around 3 times as much power and displaces about 3 times as much greenhouse gas emissions. My advice to wind turbine manufacturers would be to work to extend the useful life of the turbines, if at all possible. This graph does not include the costs of massive upgrades to the electrical grid that would be needed to ship wind power around the country in order average out regional fluctuations to more closely approximate baseload power. This graph does not include decommissioning costs; for nuclear, those are around 10% of the cost of construction - wind would probably be similar. Operating and maintenance costs have not been included; for both wind and nuclear, construction costs dominate. A construction time of 5 years was used for nuclear and 1 year for wind farms.
Ultimately, the nuclear power plants displace 3 times as much CO2 for the same cost even though wind gets a head start of about 4 years. It takes 20 years for nuclear to catch up on total capacity and about 28 years to catch up on total power generated, but after that nuclear rapidly blows the doors off of wind. Given the concern about avoiding a tipping point, this might seem to be a disadvantage for nuclear, but it isn't. If we are in danger of hitting the tipping point in the next 25 years, we will hit it either way (at this level of expendature) - it will just happen a few years later with wind. If the tipping point is more than 25 years away, then nuclear stands a much better chance of avoiding it all together. The way to avoid an earlier tipping point is to spend money faster. Indeed, we should spend money faster than the rate shown in the graph. And we should use a mix of various technologies. But nuclear has the greatest return on global warming for a given level of investment, among the power generation methods examined, and thus should constitute the largest portion of our investment because the task before us is difficult enough, financially, without making it worse than it has to be.
Numbers here have not been cherrypicked. I use reason to form my opinions rather than, as is very common, using it to rationalize pre-existing opinions (prejudices). I was, however, limited by the poor public availability of numbers. Most of the sources are linked at the end of the diary. I also try to put the numbers in context rather than the common practice of citing impressive sounding numbers with a pertinent denominator for the numerator or comparison to alternatives.
Initial Mining Impact without recycling
Nuclear Wind Coal
Fuel Ore 8750 3,324,585
Steel 733
Steel, Tower 13750
Turbine Head 1982
Turbine Blades (composite) 1095
Concrete 8444 6000
------ ------
Total Material 17927 22827
Total Metal 9483 15732
Metal - grams/KWhr 0.981 1.628
coal - grams/KWhr 335
All Numbers in table are in Tons/GWyear, unless otherwise noted.
For comparison, Coal requires 500 grams (1 pound) per kilowat hour. Thus it requires about 500 times as much mining for the same electricity as uranium with no reprocessing or breeder use. Wind also beats the pants off of coal.
Breeder reactors would improve this about 100-500 times. We haven't even considered the coal yield or the construction of the coal power plant.
Recycling can improve both sets of numbers in the long term. Even if we recyle90% of the metal (per life cycle assessment) from the wind turbine tower twice (to extend it to 60 year life of nuclear plant), the amount of metal mining to compensate for the fact that the wind turbine needs replacement sooner, we still have 6093 tons/gigawatt year of metal used in the first 60 years. Nuclear fuel can be reprocessed and used 3 times with outdated reprocessing technology, which lowers the uranium ore usage to something like 2917 tons per gigawatt year. Breeder reactors can not only use that fuel roughly 95 additional times, they can also use the enrichment waste; ultimately, this could reduce uranium ore usage per gigawatt year over the long term to something like 88tons. Maybe a little more if you take radioactive decay into consideration while we wait for breeders. With breeders, we could live off the current waste and waste from the next 60 years for thousands of years. Over 60 years, the wind turbines will use a bit less concrete but that has a lower embodied energy and CO2 per ton and probably less mining impact as well. Because wind turbines have a shorter life than nuclear, the construction materials are available for recycling for the next two 20 year cycles. However, nuclear fuel is eligible for reprocessing a couple times after only one year; but the ultimate benefit from breeder reactors might not be obtained until 60 years when plants built today are replaced.
Now consider what happens when we figure over 60 years with the uranium being recycled three times and the recycled portion of the steel and copper in the wind turbines being recycled 3 times in 3 successive 20 year generations but nothing but fuel is recycled for the nuclear plant:
Initial Mining Impact without recycling
Nuclear Wind Coal
Fuel Ore 8750 3,324,585
Steel 733
Steel, Tower 13750
Turbine Head 1982
Turbine Blades (composite) 1095
Concrete 8444 6000
------ ------
Total Material 17927 22827
Total Metal 9483 15732
Metal - grams/KWhr 0.981 1.628
coal - grams/KWhr 335
All Numbers in table are in Tons/GWyear, unless otherwise noted.
If you continue to explore the costs far into the future, after everyone reading this is dead, mining to support wind goes down further but it is likely that nuclear does as well.
World Nuclear Association says:
* In most respects the environmental aspects of a uranium mine are the same as those of other metalliferous mining.
* Most uranium mines in Australia and Canada have ISO 14001 certification.
* Radioactivity associated with the uranium ore requires some special management in addition to the general environmental controls of any mine.
* The uranium itself has a very low level of radioactivity, comparable with granite. Virtually all the radioactive material from the associated minerals in the ore processed ends up in the tailings dam.
Actually, I find this assessment a little pessimistic as the safety precautions required to contain mildly radioactive uranium ore (mining of other metals and coal also have radiation, but since it isn't related to civilian nuclear industry, the standards aren't as high), will tend to reduce other environmental effects.
Nuclear Reactor:
35 tons fuel/year for 1GW. Reactor doesn't consume significant amounts
of uranium when shut down so capacity factor is irrelevant. 2% of high
grade ore is uranium metal (some mines have ore as high as 70%, lower
grade ore is less) and ore is enriched 5X for use in light water reactors.
So we will multiply by 250 to get 8750tons/GWyear of ore - without
reprocessing. Only around 1% of the energy in the fuel will be used
the first time around, and reprocessing increases this to 3% and about
1% is ultimately unusable leaving around 96% recyclable in future
generations of breeder reactors. Not only that, the breeders don't
require the 5X enrichment step and can probably use much of the waste
from the enrichment process.
"Modern nuclear reactors need less than 40 metric tons [44 tons]
of steel and 190 cubic meters of concrete per megawatt of average
capacity." This would be 506tons/megawatt, 506,000 tons/gigawatt, or
8444 tons/gigawatt year of concrete.
Wind Power
3MW wind turbine with a typical capacity factor of 30% is equivalent to
about 1MW. It takes 1000 of these to equal a 1GW reactor. A 105 foot
tower alone weighs 275 tons. Life span of 20 years. Therefore 275,000
tons of steel are used over 20 years or about 13750 tons per year. 90%
of the steel is estimated to be recycled which gives 1375 tons per gigawat
year for future generations of turbines. These weights don't include the
turbine head or the concrete base. Concrete 450 cubic meters or 1200 tons
each. For 1000 turbines this is 1,200,000 tons or 450,000 cubic meters of
concrete. The life cycle assesment expects that the concrete will be
landfilled rather than recycled. This is 6000 tons of concrete per
gigawatt year. Use of steel and other materials for the turbine head
is another 168 tons of material per turbine or 167,000 tons for 1GW.
The Blades are 21.9 tons of polymer composite per turbine or 21900 tons
for 1GW. Note that I have used a 90% recycling figure but the life
cycle assessment actually says that 80% of the non-concrete material is
recycled; 90% of the tower and the copper wire is recycled but other
parts are not. For example, 21.9 tons (per turbine) of composite blade
material is incenerated.
Note here that we are using actual yields for uranium ore and assuming that steel and other metals are 100% yield. The actual number is probably closer to 66%. Thus, actual metal mining for steel is 50% higher than the numbers given here, which gives a significant error in favor of wind.
Solar also requires mining and has other impacts. I won't look at in detail but will just do a couple quick calculations that show that photovoltaics aren't likely to have any magic advantage here. Cadmium sulfide used in solar cells is of particular concern as it has the potential to result in roghly two orders of magnitude more loss of life in the long term per gigawatt year of power than nuclear if it isn't reclaimed. Photovoltaics at 160W per square meter (with 21% capacity factor) would require 7413 acres of solar cells for the same 1GW. If the support structure for those cells averages out equivalent to a 1/8" thick plate of aluminum, then that is 283,487 tons of aluminum which has to be mined just for the support or 9450 tons/gigawatt year. Lamination complicates recycling. 1/8" glass covering (borosilicate) is another 241489 tons or 8049tons/GWyear (approximately 70% sand which is dredged rather than mined). NNadir has a snark laden diary that looks at some of the environmental costs of solar.
A wind turbine lasts 20 years vs. 60 years. This is both and advantage and a disadvantage. Since with both forms of energy you are essentially paying for the capacity up front, with wind you have to pay for 20 years in advance vs. 60, which can help with capitol in the short term. However, when you consider the cost and capicity factor, wind costs as much upfront as nuclear and only produces power for one third the time before replacement. On the other hand, a given rate of production will continue to grow nuclear capacity for 60 years and after 20 years production will be used up on replacement turbines. 2007 cost of wind is 1,300euros/kilowatt (faceplate) or $1.833 $/watt.
If one must use low grade ore instead of high grade ore, the mining could be ten times higher than the figures given above. That would be 5 times as much as wind. Correcting the 100% yield for iron/steel, however, would reduce that to around 3.333 times. This is still more than two orders of magnitude less mining than coal for the same amount of energy.
Here is an estimate of capital costs:
Initial Mining Impact without recycling
Nuclear Wind Coal
Fuel Ore 8750 3,324,585
Steel 733
Steel, Tower 13750
Turbine Head 1982
Turbine Blades (composite) 1095
Concrete 8444 6000
------ ------
Total Material 17927 22827
Total Metal 9483 15732
Metal - grams/KWhr 0.981 1.628
coal - grams/KWhr 335
Coal produces
2 Pounds (0.907kg) of CO2 per KWh. Thats 1000tons/gigawatt hour or 8.766 million tons/gigawatt year. Coal produces 2.71 tons (ibid) of CO2 per ton of coal. Thus, we can calculate that it requires 0.738pounds of coal per kilowatt hour. Thus a coal plant producing 1GW average consumes 369tons of coal per hour or 3.235 million tons per year. Oak Ridge Laboratory is quoted as
estimating 0.913 pounds/KW, a bit higher.
Loss of life and disability, measured in Disability Adjusted Life Years (DALY), for construction of the wind turbines (per GW equivalnet) is rather shocking, though when compared to coal it would be favorable. These numbers are for 1000 3MW Vestas turbines - equivalent to a single 1GW nuclear power plant.
DALY
Climate change: 1580 85% Electricity
Carcinogens 1250 Electricity, Steel production, copper production
Respiratory inorganics 5910 Concrete
Total: 8740
DALY per GWyear 437
Like nuclear, operation of wind turbines likely results in a net savings of life when compared to using coal. Also, life cycle assesments which use impact of current electricity production tend to penalize carbon free power plants and let carbon generating power plants off easy since they include the impact of past mistakes. LCA calculations wer based on 74% coal use in Denmark. Use of Hydro vs coal reduces carcinogens about 95%, Respiratory inorganics 57%, and climate change about 93%. These ratios can't be directly applied, since 26% of the energy is already not from coal. As an approximation, I divided each DALY by 0.74 before applying the corrections. Climate change: 149, Carcinogens 84, Respiratory inorganics: 3434. Total: 3667. Total DALY per GWyear: 183. This would need to be increased by 5% to 192 since for every 20 wind turbines shipped, one wind turbine would need to be installed at the manufacturing site for the turbine and raw materials. This is still a rough approximation as the impact of electricity was assumed to be entirely from coal and it was assumed that hydro and wind were comparable in impact. checkthemarkets.com says that grand coulee dam used 24 million tons of concrete and 12 tons of steel to produce 6.8GW of electric power. Hydroelectric dams in the US have a load factor of about 0.42. Grand coulie has been operating for 66 years. If we assume a 100 year lifespan, that gives 84033 tons of concrete per gigawatt year. Incidently, 77 people died in its construction (and dam failures result in major catastrophies with high loss of life making hydro one of the most dangerous forms of power in terms of directly attributable deaths). Thus, it uses 14 times as much concrete per gigawatt year, so we will reduce the 3434 DALY from respiratory inorganics to 245 as an approximation giving a DALY of 478 total or 23.9 per GWyear for wind turbine based wind turbine construction.
Although the DALY and Loss of Life Expectancy (LLE) numbers aren't directly comparable, it should be noted that the DALY for wind, in even the most favorable scenario, dwarfs the effects of all radiation related deaths from nuclear power. Likewise, we would expect the LLE or DALY from nuclear plant construction would dwarf all radiation effects. If we assume that for every cancer death (which would cause an individual LLE of around 25years, there are 10 people who suffer from non-terminal cancers with a DALY of 2.5years, then 1 year of LLE would convert to around 2 DALY.
University of Wollongong compares the radiation exposure to the public from coal vs. nuclear for the power plant alone.
[National Council on Radiation Prodection and Measurements (NCRP)] also found that the radiation exposure from an average 1000 MW power plant comes to 4.9 person-sieverts a year for coal-fired power plants and 0.048 person-sieverts a year for nuclear-fired power plants.
When the entire cycle from mining to disposal is consider, the nuclear power plant exposure is 1.36 person-sieverts a year. Even using the linear no-threshold model, which overestimates the effects of low-level radiation, that would be 0.034 cancer fatalities per gigawatt year; assuming the average life is reduced by 25 years from cancer, this would give a loss of life expectancy of 1.258 per reactor year of operation which would probably be equivalent to a DALY/GW year of about 2.5. Non-radiation induced cancers would probably be ten times higher as a result of arsenic and other chemical carcinogens produced by mining which gives a DALY/GWyear for nuclear of 25, which would be comparable to wind turbines produced entirely using clean wind or nuclear power rather than fossil fuels. Thus, while there are a lot of limitations in this comparison of the impact on human life, it would appear that the effects would be roughly comparable for wind and nuclear.
The life cycle assesment of the Vestas wind turbine is very flawed in its reporting. It mostly gives relative risks for the recycled vs. no recycling scenarios and the coal vs hydro scenarios but doesn't report many absolute numbers that can be compared to other power sources. It also uses hydro as an alternate power source but not wind itself, though wind might be problematic due to capacity factor limiting production time.
The most dangerous product of nuclear power, in terms of actual lives lost, isn't radiation, vented radioactive gases, radioactive waste, or accidents - it is electricity. 411 people died from electocution in 2001. This is equivalent to a chernobyl every ten years even with the inflated estimates for Chernobyl (assumes linear-no threshold is accurate at low does - it is not). Respiratory deaths alone from coal, estimated at 24000-30000/year, are equivalent to 6-7.5 Chernobyls per year.
The construction fatality rate for wind turbines is around 0.15/TWh or about 1.314 per gigawatt year. Construction fatalities for nuclear are probably even lower. Coal mining in the US has 0.3 accident deaths per million tons which suggests that mining accidents are not a signicant cause of mortality - if mining accident death rates are similar for other forms of mining, mining accident deaths would be of order one death per thousand gigawatt years for nuclear and wind. Mining accidents in China are two orders of magnitude worse and other industrialized contries about one order of magnitude worse.
Ground based wind power deployment may be practical up to the faceplate capacity of hydroelectric power as the hydroelectric plants can hold back water when the wind is blowing and release it when there is inadequate wind. That at least deals with the short term fluctuations. Seasonal variations are more problematic. That would increase the actual power generated by the hydro/wind combination over hydro by itself as the generating capacity of a hydro plant exceeds the capicity of the water source that feeds it. Massive improvements in the grid could average out regional fluctuations. Prematurely retired natural gas plants could be kept on standby to provide backup power for a few decades.
Radiation is a giant red herring used by anti-nuke advocates. Ultimately, radiation from mining, processing, operation, disposal, and acidents is a small part of the impact on human life from operating a reactor and the impact is similar for wind or nuclear and much lower than for fossil fuels.
Flying Wind Power could possibly be a significant improvement over ground based wind.
In order to reduce our dependency on petroleum, I would use the first 50GW of power, or so, for rail electrification. This would help prevent rising fuel prices, and the trade deficit which results, from undermining our economy so badly that we can't upgrade our power generating capacity.
References:
http://www.phyast.pitt.edu/...
http://www.infra.kth.se/...
http://www.world-nuclear.org/...
http://www.futurepundit.com/...
http://www.world-nuclear.org/...
http://www.phyast.pitt.edu/...
http://www.blm.gov/...
http://en.wikipedia.org/...
http://en.wikipedia.org/...
http://www.uow.edu.au/...