This site uses cookies. By continuing to use this site you agree to our use of cookies. To find out more, see our Privacy and Cookies policy.
Skip to the content

IOP A community website from IOP Publishing

environmentalresearchweb blog

« Jack Frost’s underground workshop | Main | The snowline and the climate »

How carbon-free is nuclear?

How much carbon dioxide is produced from nuclear generation? Certainly the power plants do not generate carbon dioxide directly. But there are indirect carbon implications – including from uranium mining and fuel fabrication, which arevery energy intensive activities.

In a 2001 study, Jan-Willem Storm van Leeuwen and Philip Smith commented that: ‘The use of nuclear power causes, at the end of the road and under the most favourable conditions, approximately one-third as much CO2-emission as gas-fired electricity production. The rich uranium ores required to achieve this reduction are, however, so limited that if the entire present world electricity demand were to be provided by nuclear power, these ores would be exhausted within three years. Use of the remaining poorer ores in nuclear reactors would produce more CO2 emission than burning fossil fuels directly’. They developed this analysis is subsequent studies:

This analysis was strongly challenged by the World Nuclear Association which disputed some of the figures and assumptions and the nuclear industry has also pointed to the use of in situ leaching techniques that are claimed to reduce the energy costs of uranium ore extraction.

However this claim, and the WNA figures, as well as the estimates by Smith/Leeuwen, have been challenged by Prof. Danny Harvey. In a new textbook on Carbon Free Energy (Earthscan), he estimates the ‘energy return over energy invested’ (EROEI) ratio for nuclear power production as being 19.5 for uranium ore grades of 1%, down to 17–19 for the current world average grade of 0.2–0.3%. For an ore grade of 0.01%, the EROEI ratio drops to 5.6 for underground mining and to 3.2% for open pit mining, but could be as low as 2 or as high as 10 for in situ leaching (‘ISL’) techniques. However he suggests that ISL involves ‘significant and irreversible chemical and radioactive contamination of underground aquifers.’ And of the 5.4 million tonnes of identified uranium resources, he says only 0.6 mT are amenable to ISL.

Although there can be debates about assumptions and methodology, the basic issue seems clear: as lower and lower grades of uranium ore have to be used, increasing amounts of energy are needed to make the fuel, so that, since the bulk of this energy will for the present come from fossil-fueled plants, the emissions they produce will undermine the advantage of the zero-emission nuclear plants, and ultimately could make the whole exercise pointless – you would be producing more CO2 than if you just used the electricity from the fossil-fueled plants directly as normal.

The assessment of when the so called ‘point of futility’ is reached, when the energy used (and carbon produced), to mine and process the fuel is more than the carbon-free energy produced by the reactor, depends on a variety of complex factors, including the energy efficiency of the fuel fabrication and enrichment processes, and how this energy is provided. Centrifuge methods are much less energy intensive than the diffusion processes so far mostly used for enrichment, but it’s hard to see how improvements in fabrication efficiency could continually compensate when lower and lower quality ores have to be used. The high-grade ores currently used contain around 2% of uranium (20,000 parts per million), the lower grade ores only 0.1% (1000ppm). Granite contains just 4ppm and seawater – 0.0003 ppm. If we had unlimited cheap carbon free energy, then maybe we could extract some of this, but then we wouldn’t need to!

Other analysts have focused on the energy balance issue- and compared nuclear with other options. A study by Gagnon from Hydro Quebec looking at energy outputs to energy inputs (‘energy payback ratios’) over the complete life cycle, indicated that, at present, nuclear plants (PWRs) only generate up to 14–16 times as much energy as is required to build them and produce their fuel. By comparison, on-land wind turbines could produce up to 34 times as much energy as in needed for their construction (they of course don’t need any fuel for operation). Moreover, this figure is likely to be improved as new technologies emerge (an earlier paper by Gagnon had wind ranging up to 79), while as we have seen, the figure for nuclear is likely to fall as lower grade ores have to be used. (Gagnon. L Civilisation and energy payback Energy Policy 36 2008, 3317–3322).

The study of energy balances by Harvey mentioned earlier came to similar conclusions: while as we have seen he claimed that the energy returns over energy invested (EROEI) ratio for nuclear was below 20 and possibly as low as 2, he found that the EROEI ratios for most major renewables were much better than this, even for nuclear using current grades of uranium, and ‘will be decidedly better at lower grades’. Solar PV, which is one of the more energy intensive renewables, had a EROEI ratio of 10–20 at present and this is expected to rise to over 20 given new technical developments. While the EROEI ratio for wind was, he calculated, already up to 50.

Perhaps the last word should go to Benjamin Sovacool from the National University of Singapore, who has produced a paper trying to resolve the differences in views on this issue. It assessed 103 life-cycle studies of the nuclear fuel cycle. He says that the quality of most life-cycle estimates is very poor, with a majority obscuring their assumptions (sometimes intentionally) and relying on poor and/or non-transparent data; but when one selects only the most methodologically rigorous studies, typical life-cycle emissions from nuclear plants appear to be about 66 g CO2e/kWh. Although that is less than the estimate of 112–166 g CO2/kWh produced by Storm van Leeuwen and Smith, it is more than most renewables and 10 times greater than the industry often claim for nuclear power – he says they typically put the life-cycle emissions from nuclear plants, including ancillary fuel fabrication and (in some studies) waste disposal, at 1–3 grams of CO2e/kWh.

Sovacool. B Valuing the greenhouse gas emissions from nuclear power: A critical survey, Energy Policy 36, 2008 pp2940–2953.

All of this may not matter if we are just talking about a few extra nuclear plants, but if larger programmes are envisaged here and elsewhere, then it begins to be important. Certainly if we are thinking in terms of very major UK expansion along the lines of the 146 GW by 2050 seen as possible, if very ambitious, in the new DECC 2050 Pathways report, or even perhaps in the case of the UK ‘nuclear renaissance’ programme envisaged by Robin Grimes and Bill Nuttall in their recent Science review paper ( and by the IMEch in their recent report, which seemed to back the earlier suggestion by the Malcolm Wicks MP, that nuclear should provide 35–40% of UK electricity ‘beyond 2030’:


TrackBack URL for this entry: