Ben Heard and Geoff Russell collaborated on a post for DecarboniseSA titled Green Nuclear Junk that takes careful aim at an antinuclear meme that is mostly based on a series of false assumptions that include a table of mortality figures made incorrect by dividing by 8.76 instead of multiplying by that same number.
With Ben’s permission, I have decided to publish another copy of his important article in hopes of increasing its exposure and generating a new series of comments. My goal is to add just a little more pressure on Jim Green — the national antinuclear campaigner for Friends of the Earth, Australia — to come clean and admit his mistakes. It would be even better if Green began to make amends for the environmental and economic damage caused by his antinuclear activism, but that is probably a futile hope.
Green Nuclear Junk
By Ben Heard and Geoff Russell
In their determination to attack nuclear power and those who support it, anti-nuclear activism has walked away from the scientific process. As a result, nearly the entire community of environmental organisations in Australia is currently standing behind figures that are completely mathematically incorrect. Will they correct these blatant errors and open their publications to expert external review? Or is correct maths and good science optional when you wear the colour green?
The great scientist Carl Sagan famously said that extraordinary claims need extraordinary evidence. So how is it that Jim Green, an anti-nuclear campaigner with no scientific journal publications, can accuse James Hansen, one of the most extraordinary scientists of the last 50 years, of junk science?
In Green’s recent article “James Hansen’s nuclear junk science” he does precisely what good scientists don’t do. He cherry picks data.
For those who came in late, Pushker Kharecha and James Hansen recently calculated, in the journal Environmental Science and Technology, that the historic deployment of nuclear power had likely prevented 1.84 million air-pollution related deaths, and by mid-century would prevent a further 420,000 – 7.04 million such deaths.
In response, Green has the temerity to call Hansen a “policy flake”. That’s bizarre. For his outspoken policy advocacy on climate change such as the phase out of the coal industry and the cessation of tar sands operations, together with advocacy for fee-and-dividend carbon pricing, Hansen has been rightly lionised by the left and respected almost universally. But there is another platform to his policy position: that nuclear power must be deployed to prevent further climate destabilisation. Same man, same intellect, same background. This should give pause to those on the left who oppose nuclear power.
But according to Green, Hansen is now peddling junk science, claiming that Hansen got the mortality rates from nuclear power wrong. So the guts of Green’s article is a table of numbers giving deaths per gigawatt year of various energy technologies, with nuclear faring just as poorly as coal. He would have us apply this table in place of the figures provided by Kharecha and Hansen. He boldly criticises the authors for their sourcing relating to nuclear mortality, saying this:
They say: “About 25% of these deaths are due to occupational accidents and about 70% are due to air pollution-related effects (presumably fatal cancers from radiation fallout; see Table 2 of ref 16).” Ref 16 is a 2007 article in The Lancet — which makes no effort to explain or justify its figures for nuclear power deaths.
This shows both arrogance and laziness. Firstly, The Lancet is one of the world’s most highly respected medical journals. As far as sourcing goes, that’s a good first step.
Furthermore, Green is just plain wrong. The Lancet says exactly what is contained in the estimate: “…occupational effects (especially from mining), routine radiation during generation, decommissioning, reprocessing, low-level waste disposal, high-level waste disposal, and accidents.” They also clearly cite the estimate as a summary from the ExternE project.
ExternE is a huge actuarial project run between 1998 and 2005 involving an array of experts in Universities across Europe, under the auspices of the European Commission. The table in question is the summation of thousands of pages of methodology, assessment and reporting, all of which is publicly available. ExternE spends 250 pages justifying its nuclear power death estimates.
Kharecha and Hansen apply the expert information in this table to calculate that nuclear power had saved 1.84 million lives since 1971. But rather than relying on this work by experts, Green refers readers to an alternate table of figures, shown below.
Forgive the interrupted prose, but the process by which this table was developed is best explained with bullets.
- As source for this table, Green directs the reader to a page at an anti-nuclear website Choose Nuclear Free. The page is authored by Green.
- The page repeats the table, and references a more detailed paper. This paper is also authored by Green.
- The paper repeats the table again, this time listing the various sources.
For figures on biomass, rooftop solar, and oil, Green draws on information from a non-peer reviewed webpage found through the Lifeboat Project.
- In converting these mortality factors from terawatt hours (TWh) to gigawatt years (GWy), he butchers them by accidentally dividing the figure by 8.76 when it should have been multiplied by 8.76. All of these figures are numerically incorrect. For example, Biomass should show 105 deaths per GWy instead of 1.4. Oil should be 315 instead of 4.5. Green’s coal figures are a from a mish-mash of sources, some of which measure in deaths/GWy and others in deaths/TWh. Ignoring this and just using the Lifeboat figures for simplicity, the range should be more like 131-2,435 deaths per GWy.
- This source (Lifeboat Project) also provides a number for nuclear power (0.04 deaths per TWh which would have, erroneously, converted to 0.005 deaths per GWy by Green’s maths or 0.35 using actual maths) . Green ignored it, presumably because this was too low.
- He instead builds his own nuclear mortality factor by:
- Firstly selecting a 1996 estimate from the IAEA for fatalities from Chernobyl (26,000) and normalising to deaths per GWy. That this 1996 expectation hasn’t eventuated is demonstrated by later assessments of the evidence which Green ignores (see our reference below to the 2007 UNSCEAR assessment).
- Then selecting a single, non-peer reviewed discussion from a US nuclear physicist (Garvin, 2001) for mortality from the rest of the nuclear energy chain. We have serious misgivings about this assessment itself, however that deserves its own investigation;
- Then summing the two figures above; then finally
- Cherry picking a single line from a 424 page 2006 report of BIER as justification to double the upper end of his own mish-mash figure.
Dear readers, we give you junk science. This isn’t so much cherry picking as it is a half-baked cherry pie. Yet this unbelievably bad bit of work is hosted by a site with a tag line “Accurate information about Australia’s energy options”.
But this is especially concerning because Choose Nuclear Free also hosts a joint statement against nuclear energy signed by every major environmental organisation in Australia. This is why peer reviewed science matters. Environmental organisations cannot be permitted to excuse themselves from that process in undertaking their activism and get away with endorsing a non-peer reviewed, mathematically incorrect melange of cherry picked sources that is then leveraged to influence national policy. That’s exactly the practice they object to when it is evident in climate change denial. By insulating themselves from review and indulging in group-think on nuclear, they are just as guilty.
Green’s case falls apart when he tries to claim there have been “countless” accidents in the nuclear energy chain. Wrong. You can count them alright, and that’s what the Energy Related Severe Accident Database does. It tells us that between 1970 and 2005, in the OECD coal incurred 81 severe accidents (defined as 5 fatalities or greater) across the energy chain resulting in 2,123 fatalities. For nuclear, the figures are zero and zero. In non-OECD nations, the figures for coal are 1,507 severe accidents for 29,816 fatalities, and the figures for nuclear are 1 and 31 (being direct fatalities from Chernobyl). ExterneE replicates these findings. This mis-step in logic by Green is extreme.
Green mentions serious hydro accidents in brackets as though they somehow don’t quite count. The ENSAD records two hydro dam failures alone, Banqiao and Shimantan, as responsible for 26,000 fatalities, and ten further failures causing a further 4,000 deaths. These accidents counted for those victims. It is unacceptable to devalue human life when it fails to support the anti-nuclear argument.
The article is padded with discussion of the effects of low-level radiation and Chernobyl. This is an area of ongoing scientific uncertainty, and credible organisations provide differing conclusions. But Green leverages this uncertainty to the most extreme result.
His firstly distances himself from the advice of the peak body while criticising Hansen for attending to it. This is in every way akin to climate denialists dismissing the IPCC. Green remains comfortably distant from the up-to-date advice of the UN who stated, following a major review in December 2012,
“Because of the great uncertainties in risk estimates at very low doses, UNSCEAR does not recommend multiplying very low doses by large numbers of individuals to estimate numbers of radiation-induced health effects within a population exposed to incremental doses at levels equivalent to or lower than natural background levels”.
Green rejects Hansen’s upper limit estimate of 4,900 deaths from nuclear accidents, stating that credible estimates for Chernobyl range from 9,000- 93,000. The link provided to support this statement is to another non-peer reviewed article co-authored by Green himself, which itself provides no source for these figures. But the figures can only be arrived at by doing precisely what UNSCEAR cautions against: multiplying very low doses by large numbers of individuals.
Hansen is correct to quote the figure for known fatalities for Chernobyl of 43 (28 in the event and immediate aftermath plus 15 fatal cases of latent thyroid cancer). He is reasonable in offering an estimate of 4,900 addition deaths for the industry as a whole, being fatal cancers that may have occurred, but at a rate that is too small to distinguish from what is normal. He is well-supported when he reiterates that this is a probable over-estimate due to the wealth of evidence suggesting that low-dose radiation is simply not harmful, including the 2007 observation from UNSCEAR that after 20 years of exhaustive studies of Chernobyl there was “no persuasive evidence of any other health impact in the general population that can be attributed to radiation exposure”. This is hardly going to be the end of the matter for all concerned. But it is certainly not junk science.
Of course, Green glosses over the 64 Gt of greenhouse gas emissions avoided through nuclear power deployment to date. That’s unreal, but sadly typical. It seems anti-nuclear activism would happily cook the globe if it meant no more nuclear power.
[Note: France supplies electricity at around 80g CO2/kWh. Australia’s National Electricity Market produces around 800g CO2/kWh]
The article attacking Hansen is not a critique. It is a deliberately constructed mis-reading of a scientific paper and its reputable sources, aimed at discrediting a celebrated scientist whose work has become problematic for the author. It is clumsy, lazy, shoddy and deceptive, having failed to read and understand the source material and then accusing Hansen and The Lancet of a lack of rigour.
At every turn it seeks to weight the argument against nuclear by steering away from the best sources, the best science, the multi-paper reviews, and the peak bodies. It provides instead a self-referential pastiche, cobbled together from the fruits of non-peer reviewed activism, cherry picking, merging and simply butchering select references in the process.
To top it off, this error-ridden product is currently leveraged by every major environmental organisation in Australia to push their agenda and influence Australia’s policy directions on energy. That’s not science. That’s not a critique worthy of publication in a peer reviewed scientific journal. That’s an attack on the scientific process itself, and a direct analogue to the techniques of climate change denial.
Hansen is but one in a growing group of highly credible voices who have reached the same conclusions: that an effective strategy to address climate change simply must include nuclear power, and that the hazards presented by nuclear power have been greatly exaggerated and pale in comparison to the threat of climate change. If environmentalism wishes to retain any kind of moral high ground in climate change as we push past 400 ppm CO2, it must reject the junk science of non-peer reviewed anti-nuclear activism. It must evolve to a position based on the transparent application credible, expert science.
Ben Heard is Director of ThinkClimate Consulting, a climate change and sustainability advisory firm. Geoff Russell is a mathematician, computer programmer and longstanding member of Animal Liberation SA. Both authors have rescinded previous positions of strong opposition to nuclear power and have become vocal nuclear advocates.