Overcoming Mythology – Real Analysis Regarding Carbon Intensity of Uranium Fuel Cycle
It might surprise many people who are not actively involved in discussions about our energy future, but there is a persistent myth that comes up with depressing regularity regarding the carbon dioxide emissions associated with various energy sources. The myth is that nuclear power plants – which produce electricity by converting uranium fission heat into mechanical energy through a conventional steam cycle system – are somehow prodigious producers of carbon dioxide.
Not only do I know how to do the chemical and nuclear balanced equations to see that fission does not release any CO2, but I used to spend months at a time sealed up inside a submarine with an operating reactor that had no smokestack or means of discharging any gases. Theory and practice both support my gut feeling that nuclear energy is an emission free source of power.
However, in internet conversations with advocates of other types of energy who are often viscerally opposed to the use of fission, I have continually run into people who make rather vague assertions about all of the fossil fuel that is used in mining for uranium, enriching the fuel, fabricating the fuel rods, and transporting the fuel. They also talk about the energy required to manufacture the concrete, the steel, and the other raw materials associated with constructing the plant. Then they claim that there is an uncounted amount of energy associated with the process of decommissioning the plant after it has served its useful life.
After a number of engagements with people who were unable to provide numbers to support their assertions, I began to realize that part of the argument was quite emotional. Many of the people who actively promote non nuclear alternatives to fossil fuel find it very inconvenient to acknowledge that one of their main selling points – lack of emissions – is shared by a far more capable and reliable non fossil fuel alternative.
The organized forces that oppose nuclear energy apparently recognized that vague arguments did not provide them much cover, so they have worked hard to find and promote studies that “proved” or documented substantial quantities of CO2 in various portions of the nuclear fission power lifecycle. People like Jan Willem Storm van Leeuwen and Benjamin Sovacool have produced frequently referenced papers that attempt to provide hard numbers – but whose methodologies have been frequently called into question by capable and qualified observers. (Hat tip to Charles Barton at Nuclear Green for this link.)
Often, the arguments have devolved into something akin to “he said, she said”, though people with reasonably good technical backgrounds and critical reading skills can readily determine which side of the argument is real and which side is propaganda. Unfortunately, hard-nosed analysis that is not a survey of other people’s assertions is often hard to come by. It exists, but it has often been done by sources that are dismissed out of hand. For some reason, the environmental reports of corporations like Vattenfall is not seen as a credible source of information, even when certified by an outside regulator.
Yesterday, however, I received a document titled Measures of the Environmental Footprint of the Front End of the Nuclear Fuel Cycle, dated August 23, 2010. The document was prepared for the U. S. Department of Energy and has an identification number of FCRD-SYSA-2010-000104. I have not yet been able to locate the document in a publicly accessible archive, but I will keep trying. In the meantime, I thought it would be worth while to provide some information from the executive summary of the 115 page, meticulously documented study.
Front End Fuel Cycle CO2 Emissions, 2050 Scenarios kg CO2/MWh(e)
| Current | 2050 Low | 2050 High | |
| Uranium extraction | 0.78 | 0.35 | 2.17 |
| Conversion | 1.45 | 1.30 | 1.39 |
| Enrichment (centrifuge) | 0.41 | 0.01 | 0.26 |
| Fuel Fab (UOX) | 0.09 | 0.02 | 0.06 |
| DU Management | 0.035 | 0.01 | 0.02 |
| Transportation | 0.011 | 0.01 | 0.01 |
| Total | ~2.8 | ~1.7 | ~3.9 |
Source: Schneider, Carlsen, Tavrides, Measures of the Environmental Footprint of the Front End of the Nuclear Fuel Cycle, dated August 23, 2010, Idaho National Laboratory
Though this study focuses on just the fuel cycle, it provides valuable support for the assertion that atomic fission is virtually emission free. It also provides a substantial amount of support for the assertion that nuclear fission is sustainable and has a very high Energy Return on Energy Invested (EROI) According to one of the tables in the study, the fuel cycle requires the energy investment of just 0.0085 GJ(e)/MWh(e) plus 0.029 GJ(t)/MWh(e).
Here is a quote from the executive summary:
With the exception of water use, these impacts are very favorable relative to other competing technologies for large-scale energy production. For example, front-end processes have been estimated to account for 38% of the carbon footprint associated with production of electricity from nuclear energy. Scaling the above estimate for FEFC emissions accordingly, one estimates 7.4 kg CO2/MWh(e) for nuclear electricity production . For comparison, current average U.S. emissions from natural gas and coal-fired electricity production are 410 and 979 kg CO2/MWh(e), respectively.
Additional Sources
Parliamentary Office of Science and Technology Postnote October 2006 Number 268 – Carbon Footprint of Electricity Generation
Rod, good article. In the last paragraph, you have an interesting comparison with coal and natural gas CO2 emissions. Do you have a similar comparison for the life-cycle of Wind and Solar? One thing I never hear the pro-wind/solar people talk much about is the fact that wind towers need substantial concrete foundations, and I believe they use a LOT of steel in the fabrication of the tower and the turbines (I’m not sure what the actual ‘blades’ are made out – they might be some sort of lightweight non-metallic material for all I know, but when I say ‘turbines’, I mean the actual mechanical electrical generator unit which the ‘axle’ from the ‘blades’ connects to, which generates the electricity [I’m sorry I don’t know the correct terms, but I hope you understand what I mean with the terms I’m using]).
I know very little about how commercial-size solar plants are created, but for solar thermal, I would presume that all those acres of mirrors which are used to concentrate the sunlight on the boiler must have concrete foundations also, and probably steel supports?
Anyhow, my point is, it looks like people are looking at the entire lifecycle of nuclear with a magnifying glass. . . are they applying the same level of scrutiny and analysis to wind and solar?
Jeff Schmidt,
IPCC AR4 assessment of life cycle emissions for various technologies (incl wind and PV but not CSP) : http://www.ipcc.ch/graphics/ar4-wg3/jpg/fig-4-18.jpg
I’m wondering if I can get a little explanation of that graphic? In particular, it looks like for everything, there are two lines, a “high” line and a “low” line – am I correct in interpreting that as giving a minimum and maximum range for that technology? (That is, there could be a potential for both nuclear and wind to have virtually zero emissions (the ‘low’ bar which almost doesn’t show up at all in the graphic), or small, but large enough to be meaningful emissions at the high end?
In that range question. . . I’ve been wondering something related: When you look at the life-cycle emissions for Nuclear power, could it be *improved*, hypothetically, by using electrically powered equipment where today, perhaps we use equipment which is NOT electrically powered (maybe powered by gas, diesel, nat. gas, etc) but COULD BE, then supplying that electric power with electricity from a nuclear plant? That is, can nuclear power be used to reduce the carbon emissions of the nuclear power life-cycle, but perhaps it isn’t being done today?
I know that most Steel is refined in blast furnaces which burn Coke and either CO2 or CO (don’t remember now) along with iron ore, which causes chemical reactions which cause some of the carbon to bond with the iron to make steel, and some of the carbon gets released as CO2, I believe? Are there any large, commercial scale processes for refining Iron ore into Steel using electricity for most of the heat, and only enough carbon to bond with the iron to make steel, so that no carbon is released from the process? (That would of course benefit *anything* made from Steel, including Wind Towers, as well as Nuclear). What about Concrete? I’ve heard that Concrete production currently produces a lot of carbon emissions – is that another place where you might be able to use nuclear power to reduce the carbon to almost zero?
I ask, because it seems to me that, maybe, there is opportunity to *truly* make nuclear power be almost zero carbon emitting, if we can find ways to remove carbon from the other parts of the lifecycle. . . In addition to materials, what’s the possibility of using grid power (from nuclear plants) to do the actual construction of the nuclear plants, instead of diesel powered cranes, bulldozers, etc? A nuclear power plant needs high-power transmission lines, right? Well, if you build the transmission lines before the rest of the plant, you could maybe use those lines to provide lots of power to the site during construction, then later, of course, the plant would provide lots of power out-bound from the site when it is operating?
Concrete can be manufactured to produce very little CO2 (while marketing claims zero-carbon – it is probably safer to say low carbon footprint).
http://www.cenin.co.uk/low-carbon-cement.php
Westinghouse has a hard enough time laying cement that can stop a terrorist airplane attack and withstand an earthquake at the same time. I’d expect new nuclear plants will be one of the last in line to switch to zero-carbon concrete.
Some cement can actually be CO2 negative, so I think it’s find to say it can be zero-carbon.
Recycled steel can be re-processed in an electric arc furnace for lower CO2 emissions. You can also use feedstock from a blast furnace or direct-reduced iron to begin the process for new steel (but this increases costs). I would expect an EAF to be more efficient than a blast furnace (which would also contribute to reduced emissions). If there are benefits to this process for nuclear, it seems they would also apply to other technologies heavily reliant on steel (such as wind).
Ohhh. . . sorry for a double post. . . but I just had another thought.
I know that anti-nuke groups like to oppose the re-licensing of old nuclear plants to extend their operational life. . .
But, if you are worried about CO2 emissions, isn’t there a good argument that most of the ‘lifecycle’ GHG emissions attributable to nuclear are up-front in the construction of the nuclear plant, and that by extending the life of a nuclear plant, you further reduce the ‘average’ kg CO2/MWh?
I admint I’m not completely clear about the above point: the executive summary you cited said, “For example, front-end processes have been estimated to account for 38% of the carbon footprint associated with production of electricity from nuclear energy.” It’s not clear to me if, by ‘front-end processes’ they are *including* construction of the plant, or just the uranium fuel mining/enrichment/fabrication processes?
The remain approx. 62% of emissions – are those emissions which will continue to happen if the plant is allowed to continue operation, or are they, as I’m thinking, mostly emissions which have already occurred by the time a plant is seeking relicensing, and continued operation would not imply further emissions?
Seems like a crude model that looks at the operating cost of nuclear, including fuel costs, could be used as a “smell test” for some of the assertions of CO2/MWe for nuclear fission. Of course some of the assertions are very easy to discredit, especially those that have to attribute to nuclear power the CO2 from nuclear weapon induced fire storms (Mark Z. Jacobson and Mark A. Delucchi).
Rod – again thanks for this post (and all your posts). There’s one question I have for studies of nuclear energy’s carbon footprint – does the study assume that the energy required for mining, extraction and fuel creation comes from fossil fuel or (as far as possible) from nuclear generated energy? It seems obvious to me that nuclear energy should be used to create more nuclear energy. Studies would have to make this clear.
The executive summary you quote makes it clear that the study is for the current generation of enriched solid fuel, low burnup (ie non-breeder) reactors. (The title of the study should have made that clear, that it’s only one of the many possible nuclear fuel cycles. And pretty much the worst one, IMO, as well.) For me, a serious study would also include the numbers for high-burnup (breeder) reactors, especially molten salt reactors. My feeling (without even going to the back of an envelope) is that assuming high burnup would drop the CO2 emissions by a factor of 30 to 100, depending on the assumptions made.
It would be great to see a much more complete study.
Sane estimates show that CO2 emissions from nuclear, wind and solar are in the same ballpark. People who make a large number of pessimistic assumptions might find that nuclear has 2-3x the emissions of solar & wind but that’s still a fraction of you get from fossil fuels… Not worth worrying about, since scientists figure our climate would stabilize if we cut our CO2 emissions to 20% of current levels, and a nuclear economy comes in way under that.
One area where people can pick pessimistic assumptions is in the area of U enrichment: energy consumption is several times less for gas centrifuges than for the old gas diffusion method, so people can make today’s nuclear industry look worse than it is.
Another issue in nuclear economics/ecology is the question of how long the plant lasts, since the construction of the plant is expensive and requires materials and energy. We’ve learned that plants built in the 1970s have lasted longer than originally planned, which improves the $ economics and the ecological accounting. If you want to make nuclear power look good or bad in the future, the assumption of how long plants last is a key variable to mess with; I’ve seen estimates that future LMFBRs will have a 70-year lifespan. Is that true? We’ll have to wait 70 years to really answer that one.
I got interested in environmental accounting around 2001, but I discovered pretty quickly it gets mixed up in politics. Promoters of all sorts of energy technologies will search the literature and pick the most favorable estimates to put into their calculations. Opponents do the same thing, picking the worst numbers. Of course, if you’ve got a lot of factors, the difference can really snowball.
My most important takeaway is that “if it’s expensive, it’s not green;” ultimately money gives somebody a license to consume resources, so if the guy who works at a solar panel factory gets the bucks to drive a huge pickup truck and the guy who owns it builds a huge house and flies around to conferences and sales meetings, that all has to go on the ecological “bottom line”.
It was the projection of declining uranium ore grades (and the high levels of energy intensity required to mine unconventional reserves) that worried LCA researchers and energy analysts most. Just like oil, you have proven reserves, unproven reserves, and unconventional reserves. What does the industry look like after a massive build out, and large scale development of unconventional reserves? We are currently in a production deficit of 1:3 for every pound of uranium mined (we make it up with decommissioned weapons stock).
As usual for this commenter, he simply repeats antinuclear propaganda without having checked it using other sources.
Decommissioned weapons stock from Russia, have pushed the price of new uranium down, making it unprofitable to produce more. When this program is over, both the price, and the market will rebound. There is no shortage of high-grade uranium ores currently unexploited. The charge that we are running out of easy uranium, is as hollow as the accusation that it nuclear power’s carbon burden is high.
If the nuclear industry continues to run at about the same size it runs at now, most people don’t see any supply problems for 100 years or so.
On the other hand, if the nuclear industry grows enough to make a real dent in global warming (a factor of 5 or 10) , the picture becomes more murky, and the price of uranium starts to be an issue.
Now, it’s certainly possible to get better fuel utilization in converter reactors, with and without reprocessing, but a transition to breeders of some sort may become necessary in the 2050 timeframe.
Put breeders in the mix, and you’ll never get people to agree about economic, never mind ecological, calculations. The issue is that breeders are slow: particularly, fast breeders have a high fissile inventory requirement, so it could take anything from 100-1000 years to consume a mass of U238 that is put into the system depending on the nature of the reactor and the reprocessing cycle. (Thorium-cycle reactors might use Th more quickly, but would be starting with a small fissile inventory)
So, if the ‘payoff’ for mining uranium comes over a very long interval, you’ve got to think seven or more generation ahead, and the main tool that’s used to reason about this is the ‘discount rate’ which assumes that the economy grows at a constant percentage rate, so that a dollar today is worth more than a dollar ten years from now.
The trouble is that nobody knows what the ‘discount rate’ really is. It’s positive for a growing economy, but the nature of compound interest is that a small difference in the rate has a huge impact over a human lifetime. The soviets certainly made bad decisions because they used a zero rate in their planning in the 20th century, but if you’re a doomer, say you think pollution from fossil fuels is going to trash the environment and hurt the economy, the discount rate could be negative at times in the future, which would have our children be quite delighted that we had the foresight to supply them with lots of fuel.
@EL – like your worries about the health effects of low level radiation, worrying about uranium supplies is also unfounded. Comparing them to fossil fuel is a bit silly. First of all, we have been consuming fossil fuel at an increasing rate for more than 150 years. During some periods of growth, the rate of consumption increase was incredibly high – take a look at the decade of the 1950s, for example.
Secondly, the fuel has an incredible energy density difference, a factor of 2-5 million times as much energy per unit mass as fossil fuel.
Thirdly, because it is so dense, it is incredibly cheap per unit of heat, so cheap that we currently “waste” about 99% of the potential energy. I put the waste in quotes because unlike the fossil fuel that we wasted in the early days of inefficient steam engines, we have not really thrown away the useful raw material. Instead, we have carefully stored it away for the use of future generations. Leaving used nuclear fuel behind is not leaving a burden – it is bequeathing a resource!
Finally, the fact that we have a large inventory of useful, already mined materials left over from previous years should not be seen as a negative for nuclear. Wouldn’t it be great if we had several decades worth of fossil fuel just sitting around in tanks left over from preparing for a war that thankfully never happened. Of course, such a thing would be impossible because of the rate at which we consume fossil fuel. Our entire national strategic reserve of oil would only last a few months if called upon because we just do not have any place to store more.
Rod, the academic literature doesn’t support the Vattenfall numbers. I’ve gone through this over and over again, in part for my efforts on Wikipedia.
Jan Willem Storm van Leeuwen and Benjamin Sovacool are irrelevant, they got ~ 600 g/kWh in most cases and in any case find nuclear to be only marginally energetically permissible. No one agrees with them. They are not the issue. The issue is the dozen or so papers that have produced ~ 60 g/kWh
I would like to know the number is ~ 3 g/kWh as Vattenfall shows, or 7.4 as your recent reference finds, but I know these are not tenable and I don’t ever offer them hoping to convince someone. The 60 or so number is solid and won’t be debunked anytime soon. If you want to be optimistic, this can be somewhat of an upper bound.
We really need to focus on what the number says. If, for sake of argument, nuclear had an EROEI of 10:1 and you assumed that the 1 part investment came from coal then you would find the indirect carbon emissions of nuclear to be 1/10th of coal. This is shocking close to the real scenarios being presented in the academic literature and the methodology they use. There is no point in arguing over the EROEI. We should focus on the fact that most of the energy consumed is electricity, with structural materials like concrete constituting the next largest fraction.
Concrete can not be substituted with nuclear energy, and as such, there is no avoiding the necessity of concrete for the productive use of nuclear power.
Electricity, however, is the very thing produced by nuclear power and counting electricity against nuclear in lifecycle analysis is shockingly disingenuous. Electricity is the major player in all of mining, conversion, enrichment, and even fabrication.
The number in g/kWh glosses over these details and focusing on the number instead of the details is folly. I will, however, be very happy to obtain a copy of the new report you mention.
The EU’s ExternE project evaluated a number of external effects of various power sources, including nuclear. Their result for greenhouse gases was that nuclear at 5g/kWh is lower than anything except hydro at 4g/kWh. Wind is at 10g/kWh+ and Solar PV at 34g/kWh+.
http://www.externe.info/expolwp6.pdf p17
Note that wind uses more concrete and steel than nuclear. Dedicated solar plants use more steel.
The values of 60g/kWh I have seen rely indirectly (via general review) on the discredited Storm van Leeuwen papers, but I would be interested to read other studies if you have some links.
My reference:
Valuing the greenhouse gas emissions from nuclear power: A critical survey
BK Sovacool – Energy Policy, 2008
Paper hosted by anti-nuclear site:
http://www.nirs.org/climate/background/sovacool_nuclear_ghg.pdf
Google Scholar Record:
http://scholar.google.com/scholar?cluster=2585748549109587992
This paper makes reference to the paper your link to (Dones) over and over and over again. If we’re asking who has analyzed the issue more, Sovacool has. Dones did not include either operation or decommissioning, and in Sovacool’s study, every part of the cycle matters. The “average” of 66 g/kWh is taken by only averaging the values given for each part of the cycle. In other words, although Dones gives a number of 5-10 g/kWh, Sovacool’s interpretation of Dones’s number is higher than this.
Aside from the fact that Dones number is off-the-bat averaged in as higher than the reported number, the methodology is beaten one way and other, and in the end, yes, this paper claims that Dones uses numbers that are not good enough, and that better numbers get closer to 60.
You suggest that 60 g/kWh came from Storm’s methodology, but WHICH methodology? They changed their approach time and time again and found limits to how much the numbers could be fudged upwards and now their number of around 100 is only as high as other commentators would allow them to claim in good faith. Note this almost completely a product of the EROEI.
Believe me, I want to refute this. But I could write 10 pages on the subject and still not have clarity on what numbers are justified and what are not.
IMO…
If you want my real opinion/suspicion, I believe that precisely because nuclear got so much attention in this area, lots of studies cropped up that were able to nudge the numbers up higher. I doubt the same attention and body of literature has been applied to wind power, in fact, I know it hasn’t. The best I’ve seen is a 2 MW turbine that is already out of date, although more modern than the Per Peterson numbers that were thrown around everywhere in the blogosphere recklessly. Nuclear and wind started out at the same numbers and then nuclear was studied further and its number increased x10. If the same effort was directed to wind, the same thing would be observed, and we would have both nuclear and wind at around 60 g/kWh (my suspicion). BUT, the generational changes in power plant characteristics complicates the issue drastically. New nuclear plants are slimmer, but wind gets a larger advantage by counting recent advances.
Sovacool’s description of the limits of Dones’ study is incorrect (I’m not entirely sure that Sovacool references this specific report but it may be a case of referencing a common basis report). The following quotes from Dones 2005 explain what is included:
p5:
The life cycle of all stages of the considered energy systems has been systematically and consistently considered. An energy system or energy chain includes: energy resources extraction and processing, production of infrastructure, and fuels, transport, conversion to electricity or heat or mechanical energy, and waste management.
p10:
The modelled nuclear cycle includes mining, milling, conversion, enrichment, fuel fabrication, power plant, reprocessing, spent fuel conditioning, interim storage of radioactive waste, and final geological repositories of highly and intermediate level radioactive waste.
Operation and decommissioning are definitely part of the Dones numbers.
Sovacool had to mention Dones in order to attempt to discredit him; but “averaging in” one reasonable assessment with a slew of ridiculously high numbers doesn’t make Sovacool’s numbers better. He also mentions Dones’ objections to the SvL numbers but doesn’t do anything about them. He just incorporates 3 SvL studies straight into the numbers along with Barnaby and other fantasists.
Really Sovacool would have done better not to try to aggregate the studies. It doesn’t do anything useful to add such disparate numbers, and he even fails to separate out the diffusion and centrifuge options from his magic all-up number, despite talking about the difference it makes.
@theanphibian:
If you want my real opinion/suspicion, I believe that the studies that cropped up to push nuclear’s numbers higher were specifically designed to do that. Plenty of people have a financial motive in trying to tie nuclear down as much as possible. Nuclear fission threatens the massive profits available from selling gas lifted out of the ground for $0.50 per million BTU in Qatar and selling it for $9.00 per million BTU as LNG in Japan. It also demonstrates that you do not need to sacrifice reliability for cleanliness. No one should be able to sell solar panels simply because they supposedly have a lower carbon footprint because they definitely do not have a lower carbon footprint than a well designed, built, and operated fission based generator.
With regard to Savocool’s study, you may recognize that Jan Willem Storm van Leeuwen’s work is irrelevant, but Savocool does not. Here is a quote from his Acknowledgement section in the very paper that you quote:
Mark A. Delucchi from the University of California Davis, Paul Denholm from the National Renewable Energy Laboratory, Roberto Dones from the Swiss Laboratory for Energy Systems Analysis, V.M. Fthenakis from Brookhaven National Laboratory, Paul J. Meier from the University of Wisconsin-Madison, and Jan Willem Storm van Leeuwen provided invaluable and outstanding comments and suggestions in the revision of the manuscript.
I happen to be a writer and an analyst. I know a little about how to slant work to shade meaning, even when you are writing about a subject where there are a lot of numbers. There are whole books on the topic of How to Lie With Statistics. Savocool had assistance in editing his paper from a man who has been famously trying to do that to nuclear energy for decades.
@Joffan and @theanphibian: Do any of these figures take into account typical capacity factors for the energy sources? I would argue that with Wind having, according the figures I’ve seen, about a 30 percent capacity factor, and Nuclear having about a 90 percent capacity factory, that you need 3X as many wind turbines to really be equivalent to a given amount of nuclear power (and that’s not even accounting for the grid-scale energy storage you would need to pair up with wind or solar generation in order to ‘smooth out’ the energy so that you actually do get just as much energy from wind or solar as nuclear). So, 1gW of nuclear is about equivalent to 3gW of Wind. When you take that into account, doesn’t wind start to look much worse than nuclear if you are trying to reduce carbon emissions? As for solar PV, I believe Solar PV capacity factors are more like 15 percent, aren’t they? So, with solar, you are looking at needing something like 6X as many watts of solar as nuclear, plus 2X as much grid energy storage as with wind.
Am I missing something important in the analysis, or is that fair? If we are going to use those numbers, I’d say that you should triple the value for wind, so Wind becomes 30g/kWh, and Solar PV becomes 204g/kWh (unless the capacity factors are already figured into the above numbers).
@Jeff Schmidt Capacity factors are taken into account in all of these studies. I am not aware of a study that calculates indirect lifecycle emissions without the capacity factor, and if it did exist it would be laughable. Generally they are extremely sophisticated and take into account much more minute detail than capacity factor – and I mean this for both the anti-nuclear sources and pro-nuclear sources. The units of g/kWh (grams of CO2 gas) in themselves indicate an analysis that looks at inputs versus output to grid (kWh) in a realistic deployment. Vattenfall even included effects of transmission line loss for different sources, nuclear has a major advantage there because centralized plants have better access to higher voltage transmission. Generally there is no need to consider the need for x number of nuclear plants vs. y number of wind turbines since everything is counted in a per-unit basis.
Suffice it to say that 204 g/kWh is not a correct interpretation of a study’s results. There is, however, room to bicker over capacity factors between 13% vs. 18%, for instance. Just like some people will argue 70-some % CF for nuclear by sampling the French fleet that has to throttle on a regular basis to meet power demand. The case of NG and coal completely blows up in complexity because a majority (or near-majority) of their power availability is not used due to lack of demand for it (they are load matching units), but for these units the direct Carbon emissions dominate so the over-investment in terms of concrete and steel for construction, coupled with the lower capital cost, leads to a situation where the assumed CF does not make a major difference. You should be on the lookout for danger, however, in the case of something like wood or biofuel because comparing those to either renewables & nuclear or fossil fuels are comparing apples to oranges. Such comparisons can be effectively made, however, will respectable and diligent investigators, just like the IEA and EIA does compare (fairly to some extent) energy production in toe units.
Arguing that we can’t use nuclear power to help alleviate our carbon emission intensive energy profile because we will have to use those sources to build the plants and manufacture the fuel is just silly. Anyone that makes this argument is not interested in an intelligent discussion. What is the CO2 number for a plant built and fueled in France? Maybe we should promise to only run enrichment plants when the wind is blowing.
Recently I came across a discussion of the ecological footprint of nuclear power. Here are some excerpts from the Global Footprint Network’s Ecological Footprint Atlas 2008, pp 30-31:
Beginning with the initial 1997 edition of the National Footprint Account, a new energy component, nuclear land, was included in the Ecological Footprint along with the carbon land component. … It was assumed that the Footprint of generating a unit of electricity in a nuclear power plant was the same as that for generating a unit of electricity by a power plant using a world-average mix of fossil fuels.
In 2007 the National Footprint Accounts Committee concluded that this emissions proxy approach was not a scientifically justifiable method for calculating the Footprint of nuclear electricity. This decision, which was preceded by numerous meetings and two public comment periods, and had an 82 percent approval rating in Global Footprint Network
Rod
I want to add that Sovacool is clearly anti-nuclear. There is really no difficulty in making this statement, a quick look at his publication list reveals this. Or have a look at the Wikipedia article, dedicated entirely to him, which speaks to the reliance on his published work on this subject. The 66 g/kWh is also debunk-able. Note that if a study didn’t include emissions due to decommissioning, it is probably because they perceived there to be negligible emissions in the process. But even though their opinion was 0, Sovacool omitted it from the average. There are a lot of factors at work in these studies that I openly admit to not be sufficiently educated on. The researchers up to this point have, indeed, done impressive work.
I’m also eager to see the Schneider, Carlsen, Tavrides study, because as I’ve pointed out, it is thinkable for the current declared “consensus” to be toppled. A 2010 installment with the thoroughness of several national lab researchers would be a gem.
When wind and solar start adding the carbon burden of their backup from coal and gas their numbers don’t look so rosy. Without this added their numbers are simply lies. When you can show that IN TOTAL a non-hydro, renewable system, supplying reliable, dispachable power, at the same level of reliability as a nuclear plant AND show a lower CO2 output end-over-end, we’ll talk.
@EL – First of all, the study does contain a range of estimates. I only posted on tiny table out of a 115 page long study that contains a number of additional references. Even that table shows that the work includes a high and a low estimate for 2050 (and for other years in the study.)
Sovacool did a meta analysis. He also threw out a whole bunch of studies for various reasons. Some of the studies that remained in his survey population use ancient technology or silly assumptions that are not born out by real world experience in processing lower grade ores. Surveys of other people’s work might count for originality and may lead to an inkling of the truth in archeology. In physical science it is kind of like doing a study of people evaluating the shape of the Earth and determining that it must be some kind of statistical version of somewhere in between flat and round. The answer would simply be WRONG.
Using the results of that “study” to predict the emissions from nuclear fission into the future provides a BAD estimate of the amount of CO2 that would be released. Based on the real world analysis provided by the detailed technologically based study that I am quoting, Sovacool would overestimate the emission level of a fleet of nuclear plants by a factor of at least 9. The study done for the INL also offers specific suggestions for improving the already impressive numbers by increasing uranium utilization from its meager value of less than 1% today as we develop more capable fuel cycles.
If your real mission is a cleaner planet, stop holding your hands over your ears and saying “I can’t hear you” while babbling continuously. Your comments in a previous thread were one of the reasons that I posted this.
You asked for a copy of the study. I can share it via email. My address is at the bottom of the blog home page and at the bottom of every single article page.
On the other hand, if your real mission is to sell weak and unreliable alternative energy sources that consume a VAST quantity of precious physical and human resources for every kilowatt-hour produced, you are free to continue trying. Don’t expect any compromises or interest from me.
I guess most people here know this summary of analyses:
http://www.world-nuclear.org/info/inf11.html
Note the huge difference for nuclear between centrifugal and diffusion enrichment (see table 2).
Does anyone here at Atomic Insights want to have a (friendly) discussion about the relative carbon footprint of Thorium Fuel Cycle versus Uranium Fuel Cycle (Thorium requiring only one part in 200 the amount of ore to be removed from the ground and no enrichment step)?
Just say front end emissions are 0, buys you 30% (but depends on the study).
Then you need to talk about plant construction, operation, and decommissioning. If these are more efficient in material and energy inputs then the number will be lower. These variables, however, can be relatively well correlated to overall cost of the activities. So take whatever assumptions you have about these processes and scale the number accordingly.
theanphibian – the 30% only applies to diffusion enrichment, which is in the process of dying out. All future considerations for uranium fission should use an assumption of centrifuge or more efficient technology, which basically reduces normal uranium-fission fuel front-end energy by a factor of ten.
Molten salt reactors are a different kettle of fish in any case, whether using uranium or thorium. They show enormous promise for modularity/deployability (low pressure reactor), utility (eg. high temperature) and “safety” (as if the current reactors aren’t safe enough, different rant). Other arguments advanced eg. anti-proliferation appear less well-based. The uncertainties around operating the plant with the processing/extraction of radioactive fluids is still an area that seems full of unknowns to me.
@theanphian – I do not agree that the variables of energy consumption in plant construction, operation and decommissioning have any relationship at all to the cost of the plant. To an overwhelming degree, the cost drivers for the ultimate cost of a nuclear power plant are related to the cost of people and financing. Under current rules and processes, it is possible for a valve that costs $2,000 when installed a coal plant to cost $50,000 or $100,000 by the time it is installed in a nuclear plant.
That nuclear grade valve did not require 25 or 50 times as much input energy as the conventional grade valve. It just required a whole lot more human time for QA, planning, and retest.
@Bob – I am always willing to have a friendly discussion, but I hate fighting with my brothers, especially when it is over difficult claims.
It is fine and dandy to produce pretty slides showing that 100% of thorium will eventually produce heat while only 0.5% of uranium currently does produce heat. However, no one is actually producing heat by consuming thorium today, so all of the heat being produced by uranium fission must be happening because all of us existing nukes are too blind to see the advantage. Either that, or we are still hypnotized by Shaw.
There are many reasons why uranium still has a lot of life left in it as a heat source and why pushing thorium as an incredibly better source is not a useful tactic. The physical reality is that it is just as possible to extract nearly all of the energy from uranium as it is to extract nearly all of the energy from thorium. They also both have approximately the same energy content per unit mass.
@Rod – At the onset I would like to concede that both Uranium and Thorium fuel cycles have a surpassingly small carbon footprint on the basis of GW-year of energy produced.
While the difference between the two cycles may not be “decimal dust”, we are talking about two forms of power generation that both deserve to be labeled low carbon footprint power generation.
I would tend to agree that it tends not to help to belabor the perceived differences between fuel cycles when equivalent quality of data is not available for both cycles.
Good and reliable numbers are largely available for the Uranium Fuel Cycle case and while there is some operationally verified data for Thorium in a Fluoride Reactor (ORNL MSRE), that research reactor did not include many improvements (particularly in the supporting chemical plant that is proposed by Thorium advocates to permit more efficient operation and closing the Thorium fuel cycle) that Thorium advocates rely on when making their estimates. Until a modern Thorium LFTR prototype is operating there will always be suspicions that someone in the Thorium camp is just a little too aggressive in making their claims (or is perhaps dealing from the bottom of the deck).
My thanks to EL, Joffan, and theanphibian for their kind offerings and observations.
@Bob – You bring up one aspect of the fuel form discussion that does not get enough attention. The MSR/thorium advocacy generally comes from people who are excited about chemical engineering and see no real issue in having a chemical processing plant at every power plant. Wigner, for example, was a chemical engineer.
Most of the rest of us nukes are not all that excited about chemistry and chemical engineering. We think of it as a necessary evil that should be engineered out as much as possible. We work hard to keep our reactor fuel in solid form that is fully contained inside redundant layers of corrosion resistant material. We try to centralize the chemical processing part of the fuel cycle and keep it separated from the power plants.
I am willing to concede that Kirk and his buddies could be on to something big. However, I am humble enough to report that the ONLY time in my academic career that I actually struggled and earned a ‘D’ on a test was in chemistry. I managed, though a lot of extra work, to pull out a B on the final exam and for the course, but I am lazy. Chemistry is hard. In my book, that makes it expensive, so I think there is a lot of overselling going on with relationship to MSRs and thorium.
@Rod
@Robert – though I am not allowed to talk much about the details, it is public knowledge that the Virginia class submarines have no refueling capability because their initial cores are designed to last for the life of the hull. The fuel form that is being proposed for the NGNP has already been tested to a burn-up that is about 4 times as high as the burn-up in conventional light water reactors and it will be tested to a greater number in the relatively near future. Those tests take a lot of time; you can only produce so many neutrons in the ATR and you have to do physical tests to back up the codes.
The effects that you mention regarding cladding damage can be overcome with known techniques.
You wrote the following:
The whole notion of fluid reactors is kind of a
Is it possible that special interests play a role in the development of uranium v. thorium technologies. I heard a LFTR advocate claim that most of the profits in the nuclear industry are made on the nuclear front end (and less by utilities running power plants). He offered this as one explanation for why thorium has not been more widely developed (despite its parallel path of development over last 60 years).
Barry Brook provides some estimates and comparisons of the materials and land requirements for solar thermal, wind and nuclear (AP1000) here: http://bravenewclimate.com/2009/10/18/tcase4/
The comparison overwhelmingly favors nuclear.
So far as I can tell … Barry Brook limits his analysis of nuclear to energy production in the power plant (steel, concrete, land use, and other factors in a AP1000 reactor). But what about the many other production sites involved in the nuclear fuel cycle: mining and tailings impacts, milling and conversion plants, facilities for heavy water enrichment for some plants, uranium enrichment by gaseous diffusion or centrifuge, fuel fabrication facilities, and storage and transportation sites for wastes, materials, and fuels. You have to include these too in the material inventory for nuclear, shouldn’t this be the case?
Capacity factor must matter for example if Open Cycle Gas Turbines (OCGT) are used for backup for wind energy because they have the ability to ramp up and down quickly. In the context of this example the alternative to a wind farm would be Combined Cycle Gas Turbines (CCGT
Capacity factor must matter for example if Open Cycle Gas Turbines (OCGT) are used for backup for wind energy because they have the ability to ramp up and down quickly. In the context of this example the alternative to a wind farm would be Combined Cycle Gas Turbines (CCGT
Sovacool has grudgingly acknowledged that his results do not apply to new Gen III reactors in the comments following this article
http://www.scitizen.com/screens/blogPage/viewBlog/sw_viewBlog.php?idTheme=14&idContribution=2136