CO2 from nukes

Every now and again the question of how much CO2 nuclear power plants produce, compared with other forms of power production, comes up; and whenever it does, I’ve forgotten where I last saw the figures. So now someone (thanks S) sent me some that look good (and are, by repute, biased against nukes if anything), I’ll put them in here.

The original report is but I don’t read german, so I fed it through babelfish which seemed to a pretty good job. I’ll skip all the boring caveats and stuff and just show the end result: this purports to show Total greenhouse gas emissions from electricity generation options (including upstream processes and substances used to manufacture plant)

Electricity from:                                   CO2e
Nukes-Germany(fuel)                                   32
Nukes-France(fuel)                                     8
Nukes-UK(fuel)                                        32 
Nukes-Russian(fuel)                                   65 
U.S. nuclear -                                        62
AKW (uranium after import mix)                        32
AKW (uranium only from Russia)                        65
Import coal power plant                              949
Import coal CHP                                      622
Lignite power plant                                1.153
Lignite CHP                                          729 
Natural gas combined-cycle power plant               428
Gas-CHP                                              148
Natural gas cogeneration plant                        49
Biogas Blockheizkraftwerk                           -409
Onshore wind park                                     24
Offshore wind park                                    23
Water power                                           40
Solar cell (multi)                                   101
Solar-Import (Spain)                                  27
Power Efficiency (medium)                              5

[Update: thanks WW: The unit is grams of CO2 (and equivalents) per kWh of electricity.]

[Update: "Blockheizkraftwerk" means "block heat and power plant", or CHP -W]

The differences among the nukes is just the fuel source, if I’ve read the report properly. The US uses lots of coal electricity to make its nuclear fuel, the French use nuclear, and who knows what the Russians use. So… French nuclear is better than just about anything else, whereas German is beaten by wind. I think there is something broken in that (if you could get your wind turbines made in France they would come out as casuing less CO2 than the same turbines made in Germany). Increased efficiency (assuming I’m interpreting that last line correctly) is bested only by Biogas-CHP, which somehow comes out negative. Not quite sure how they have calculated that (from the comments: I guess the biogas plant is negative because it also generates heat for houses, so it saves fosiil fuels used for heating. It would come up close
to 0 without use of the heat. That would also explain why the NG cogeneration plant has such a low value.

OTOH… table 5 lower in the report says that the cheapest in financial not CO2 terms is… you guessed it, lignite.

22 thoughts on “CO2 from nukes”

  1. An important point to keep in mind is that newer enrichment technologies will be much more efficient–particularly Silex, a form of laser isotope enrichment currently being commercialized by USEC. Even compared to centrifuges, Silex is an enormous improvement in energy expended for a particular amount of usable nuclear fuel.

    Even so, current light-water reactors are notoriously inefficient both in their efficiency in utilizing nuclear fuels and their conversion efficiency. We can do a lot better. Check out this French proposal for a molten-salt thorium breeder reactor started using nuclear waste from current reactors”. This is the thing James Hansen has been talking about lately, and if we can make it work it will make beating global warming and dealing with our nuclear waste problem a lot easier.


  2. I assume that biogas is negative due to unharvested biomass, though that depends heavily on assumptions about farming methods.

    The method they’re using for computing CO2 production for fuel processing seems rather odd, though; it would probably make more sense to just subtract the amount of electricity used to produce the fuel from the amount of electricity produced. Of course, if non-electrical processing is used, that doesn’t apply.

    The ‘natural gas cogeneration plant’ number is also fishy; it probably represents the additional electricity produced by adding a cogeneration plant to an existing power plant.


  3. The claim that biogas is carbon negative is wrong. To be carbon negative requires permanently removing carbon from the active carbon cycle; obviously burning biogas does not do this.


  4. Aha. But burning the biogas to carbon dioxide is preferable to simply letting the methane escape into the air. So via some methane/CO2 tradeoff formula one arrives at an equivalent carbon negative value.



  5. I guess the biogas plant is negative because it also generates heat for houses, so it saves fosiil fuels used for heating. It would come up close to 0 without use of the heat. That would also explain why the NG cogeneration plant has such a low value.

    “Blockheizkraftwerk” means “block heat and power plant”.


  6. Sounds about right. I recall some figures made on the same basis that shows nuclear, hydro and wind as roughly level pegging, solar PV with two or three times the emissions (although this is heavily reliant on the electricity source used to refine the Si metal)gas about ten times that and coal 30 times that PV number.

    Differences in detail, yes, between the two sets, but orders of magnitude are similar.


  7. I remember reading in the mid 90s that the CO2 emissions from the cement used in the concrete to build the nuclear powerstation offset the carbon savings it produced in generation. It was in the ENDS magazine, I think.

    I’ve never been able to get a proper figure for this since nor the (more valid) marginal cement CO2e comparison with a coal or gas powerstation.

    Anyone heard this one before?

    [I’ve certainly heard it before, but never properly sourced. I’m pretty sure thats just scare-mongering by people who don’t have the numbers to hand. But if you ever find their source or numbers, do let me know -W]


  8. Having bought a few hundred thousands of bucks worth of optics for sub Ebay prices when USEC folded the AVLIS program Eli is a bit skeptical about Silex. If anyone wants mirrors for copper vapor laser wavelengths, drop Eli a line, he still has two trunks of the stuff in the basement and Ms. Rabett wants the space.

    USEC folded the program because Soviet uranium from old bombs was very cheap. They ground the narrow band coated optics up in cement mixers because the wavelengths were classified.

    Los Alamos has been pursuing the IR version of the wil o the wisp since the 1970s. The Rabett remembers seeing spectra with the wavelength axis labeled classified back then. (you can google John Lyman talking about Silex being naive, and Eli believes John). If you can’t efficiently produce narrow band 16 micron light you can’t do the job economically and no one has really been able to do that.


  9. “I remember reading in the mid 90s that the CO2 emissions from the cement used in the concrete to build the nuclear powerstation offset the carbon savings it produced in generation.”

    Sounds most unlikely to me. You use more concrete to anchor windmills into the ground than you do to build a nuclear station….for the same power output that is. So if it were indeed true for nuclear it would be even more true for wind.


  10. Well, according to the Silex website, we’ll find out soon enough if the concept is “naive”:

    “The Uranium application of SILEX is currently in the third and final stage of development – called the “Test Loop”. In accordance with the SILEX-GE Agreement, the Test Loop program is being fully funded by GE. The Test Loop, which is being built at GE’s nuclear (Fuel Fabrication) facility in Wilmington, North Carolina, USA, will verify performance and reliability data for full scale (commercial-like) facilities. This key engineering demonstration program is scheduled to be completed in mid 2009.”

    Given current economic circumstances and the amount of enrichment capacity available worldwide, I’m not sure if Silex will really get commercialized, but apparently they convinced GE to pay for a demonstration facility. Will it work? I imagine GE thinks so.


  11. Sovietologist, given the inefficiencies of converting CO2 laser output to the right wavelength by Raman shifting (and somehow you have to have the right medium to do it in) Eli has no doubt it is inefficient. The point is that the rush of Soviet uranium onto the market after 1995 or so forced USEC to shut down AVLIS AND mothball a new centrifuge plant in Ohio somewhere. Now if this pile has been consumed they might be looking for new capacity.

    The issue is the following UF6 has an incredibly complicated IR spectrum even in a beam. The lines are very, very close to each other. The isotope shift between 235 and 238 is small (just take a look at the ratio of masses for the UF6) so you need to find a line of the CO2 laser whose Raman shift in some material that is transparent to the CO2 and the Raman shifted line, is by some miracle right on the 235 UF6 16 micron line. Straight Raman is a loser, maybe you have some sort of four wave mixing process. You then dissociate the excited UF6 with another laser or do something to collect it It makes no sense to shoot out the 238 UF6 leaving the 235, because you need to process a lot more material. Just looking at the process cartoon on the Silex site gives me the feeling they are naive, and I knew John Lyman and he has been pushing this peanut up the hill for a long time and knows where the rocks are. If he says it is naive, he knows.


  12. Well, the era of cheap Soviet uranium is definitely ending. Not only are they running out of spare nukes to scrap, but a lot of Russians think that they got played for chumps selling their uranium at absurd bargain-basement prices. Furthermore, the remaining U235 will mainly be used by the Russians for their subs and as VVER fuel, particularly for the new VVER-1000 reactors Russia is exporting.

    Maybe I’m overly credulous, but I get the impression that Silex has gotten to a point where they probably have some kind of answer to this problem. You (and Lyman) are right that building an laser that’s up to the task is an extremely difficult challenge. But GE-Hitachi has thrown some serious money behind Silex, and I imagine that they probably did their homework first. This doesn’t mean that Silex will turn out to be commercially viable. Heck, it may turn out next year that the test project will fail to live up to expectations and the whole concept will be abandoned. But if Silex does work, it could be considerably more energy efficient than centrifuges even with the colossal inefficiencies in the CO2 laser.


  13. Jacobson’s “analysis” on nuclear is based on some truly bizarre assumptions–namely, that building new nuclear plants in the US will lead inevitably to a nuclear war somewhere, so we should incorporate the potential environmental externalities of nuclear war into the calculus of nuclear power’s environmental impact. It’s amusing as a rhetorical gambit, but it’s totally out of touch with the way these things work in practice, given that most of the countries acquiring nuclear “facilities” are not going to build a complete nuclear fuel cycle, just reactors. Furthermore, the cat’s out of the bag on this one–Article IV of the NPT says that signatories have a right to develop civilian nuclear technologies and this will not change, no matter what the US does. Furthermore, Jacobson used out-of-date studies with maximally pessimistic assumptions (gaseous diffusion powered by coal) in the rest of his assessment of nuclear power. The German study Connolley discusses here is vastly superior in both its methodology and assumptions.


  14. Dear Sovietologist,

    well, maybe the assessment of Jacobson might not be perfect in every detail, but I would not judge it as “analysis”. Clearly, nuclear energy is not silver bullet, it might be used more than today, hardly is it the best sustainable long-term energy joice. I am afraid, we are not gonna live without nuclear anytime soon…



  15. William, I think the table is missing an important point, that electricity, like money is fungible. The CO2 cost in any particular place calculated that way reflects the mix of generating plants when the nuclear plant was built, not necessarily the CO2 cost of the next one. One could play games. For example Germany and France trade electrons across the border, so Germany should use French electrons to build nuclear plants??

    [Yes, I agree, and am not quite sure how you get around that one. Maybe it has to be traced backwards… current french electricity counts as nearly-CO2-free, but was build using CO2, so you should count that in… -W]


  16. Having read the report now, I think the treatment of CHP is troublesome. They make oil heating the reference system and calculate a carbon credit based on that, which gives negative emissions for biogas.

    But, there is no reason that heating must be done with oil, except when a CHP system is installed. The heating can also be done with biogas, when the electricity comes from nuclear or wind.

    Or the heating could be done with a heat pump using electricity, and if that is emissions free electricity, CHP using fossil fuels starts to look quite bad.

    Then, maybe we should include a carbon penalty, rather than credit, for the fact that carbon intensive heat must be co-produced in a CHP system …


  17. To the extent that the difference reflects only energy used in producing the fuel and the nuclear plants, what you get out of it is that the French cost is the baseline. Since nuclear plants are best run at full load, as you build them they will take over more and more of the associated energy costs.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s