As usual, the Onion nails it: Nuclear Energy Advocates Insist U.S. Reactors Completely Safe Unless Something Bad Happens

Alright, I can’t resist saying more. I grew up in Europe and remember Chernobyl vividly. The nuclear fallout reached as far as northern Britain, and for months we were reminded just how risky nuclear power is. Many years later, I studied systems failures in the aerospace industry, and learned just how hard it is for humans to prevent catastrophic failure in complex socio-technical systems. No, worse: I learned that human organisations often conspire to make complex systems less safe. Or, as Nancy Leveson puts it in her book, Safeware (a must read for anyone interested in complex systems failure):

“In most of the major accidents of the past 25 years, technical information on how to prevent the accident was known, and often even implemented. But in each case… [this was] negated by organisational or managerial flaws.”

The problem isn’t a question of whether small radiation leaks are dangerous or not. And it isn’t even a question of whether, on average, coal-fired power plants emit more radiation than nuclear plants (although this infographic from xkcd helps to put it into perspective). The problem is that human organizations are just inherently too flawed to manage the safe operation of something as inherently dangerous as a nuclear fission reactor. The failings of TEPCO aren’t really news: this is how normal accidents occur.

In the last few years, I gradually became convinced to ignore my objections to nuclear power, on the basis that if we’re going to switch from fossil fuels quick enough, we have to consider everything else. The risks from nuclear power pale into insignificance compared to the risks from climate change on our current fossil-fuel-intensive path. That’s not to say the risks aren’t there. But it’s a trade off between what’s essentially a local risk to people living within a hundred km or so of a nuclear power plant (which I do), versus a global risk to pretty much everyone on the planet. But just as we don’t have a suitable global governance structure for taking appropriate action on climate change, so we also don’t have a suitable global governance structure that’s able to weigh up the risks to, and represent the interests of, different groups of communities in considering whether to build more nuclear power plants to reduce the risk of climate change.

The bottom line is that nuclear power poses the same kinds of inter-generational ethical problems that climate change does: we build nuclear power plants now to enjoy our profligate energy-intensive lifestyle, and leave future generations to cope with the costs and challenges of decommisioning, site clean up, and long term storage of waste. We don’t know how to solve these problems right now, so do we really have the right to leave them to future generations to solve?

It seems to be me that nuclear power can only be part of the solution to climate change once we demonstrate that we’ve fully pursued every other avenue for clean energy and energy conservation, and once we’ve come up with an adequate governance structure for balancing the risks to different communities. Some people make an even stronger argument, based on the economics – e.g. Amory Lovins crunches the numbers and points out that nuclear power is both unsafe and uneconomic.

And, as the Onion reminds us, it’s only safe when nothing unforeseen happens.


  1. Have you ever come across thinkers who nicely characterize our obligations to the future? For example, I can easily see my obligation to my son, and his children.

    But beyond that? I’m not so sure I need to worry. Look at the mess your distant ancestors left in Zimbabwe, India/Pakistan, etc., or mine left in Belgium, and so on.

    In that context, I think nuclear waste storage is already solved. The Eloi can figure it out.

  2. Allow Eli to disagree. As bad as Chernobyl was, the actual loss of life was less over any interval than that from coal burning, and over the past 30 years, orders of magnitude less. What we do is over estimate acute issues and underestimate ones that erode over long periods.

    WRT nuclear, there are ways to limit the maximum possible accident.


    “the calculation of premature deaths per terawatt hour of energy production comes to this conclusion: for coal, 161; for oil, 36; for biofuels, 12; for natural gas, four; for nuclear, 0.04.”

  4. Eli – I think that’s consistent with what I’m saying. The number of deaths attributed to Chernobyl is relatively small (for an major industrial accident). But the ongoing health problems of those living in the region is significant, even 25 years later, and the full extent is still somewhat unknown. My point was there is a difficult ethical question of weighing the risks to people in the local area of a nuclear power plant against the risks to the global population from climate change. A strict utilitarian view would say to sacrifice the smaller number for the larger. But who gets to give themselves this right?

    I think coal vs. nuclear is a false dichotomy. It should really be dirty energy (in which I would include both coal and nuclear, for different reasons) vs. clean energy.

    On your last point, I agree only to the extent that the maximum possible accident at a nuclear power station is limited by the total amount of radioactive material onsite, and the proximity of nearby populated areas over which it can be dispersed. Some designs are less risky than others, but human stupidity can always find a way to maximize the damage. The worst case scenario is dramatically worse (on a per plant basis) than for any other form of energy generation.

  5. BTW George Monbiot looks at Fukushima and draws the opposite conclusion to mine:
    His argument is that because throwing a major earthquake and tsunami at a nuclear power station did nothing worse than a minor radiation leak, then nuclear power is much safer than he previously thought. I could apply the same logic elsewhere. For example, if I encounter someone who ran across the freeway, was hit by a truck, and only broke a couple of bones, then I can conclude that running across the freeway is much less dangerous than I thought….?

  6. @Steve

    I think that someone running across the freeway and getting hit by a truck but being relatively uninjured DOES constitute evidence that should make you adjust running across the freeway towards safety. However, we have strong prior beliefs about the danger of such an action and it will take a lot of data to convince us otherwise. A single incident will not outweigh our strong prior.

    My confidence in the safety of nuclear fission has certainly increased because of these events. Fission reactors are getting safer and safer and even these ancient reactors are still pretty safe.

  7. I’ve spent some of the last week trying to get people to stop panicking about Fukushima, and focus instead on the tsunami damage (today’s numbers: 9079 dead, 12782 missing, Stop buying iodine tablets, and send food and blankets instead. However, I do partially agree. Any nuclear installation which requires continuous cooling or other activity to prevent a serious radiation leak is not truly fail-safe.

    That essentially rules out all actively-cooled plant: certainly all BWRs and actively-cooled spent fuel facilities. I am very optimistic that other nuclear designs may be much safer, but we have to sceptically assess the statements of the nuclear industry, and of governments, who have often tried to play down the risks.

  8. Steve, this pretty much sums up my take on nuclear. Thanks. I think you are right that humans just cannot be trusted to manage so complex systems. Accidents will happen (and has happened) far more frequently than they should according to design criteria. I simply do not believe the optimism when some people argue that such accidents cannot happen on our new plants. However, my main concern is not accidents, but that we have no clues what to do with the waste. What will it cost in the long term? I am usually not very comfortable with the idea of commiting myself to a deal if I don’t know the terms. On the other hand we know there will be a dear cost to pay from sticking to fossils.

  9. quoting Steve E. “My point was there is a difficult ethical question of weighing the risks to people in the local area of a nuclear power plant against the risks to the global population from climate change. A strict utilitarian view would say to sacrifice the smaller number for the larger. But who gets to give themselves this right?”

    Not trying to make a cheap point (but I will understand if I am accused of such): Who gets to give themselves the right to discount coal miner’s lives at a fraction of a “regular” citizen?

    [ One could respond that the coal miners themselves are doing the discounting, because they chose that career. ]

    Your post’s point still stands and is very well put: “we also don’t have a suitable global governance structure that’s able to weigh up the risks to, and represent the interests of, different groups of communities in considering whether to build more nuclear power plants to reduce the risk of climate change.”

  10. I hate blog entries without sources, but I’m hungry and it’s time to eat. I do have open, so it will have to do as a citation.

    1) The maximum atmospheric CO2-content compatible with a human-livable climate is 350 ppm. (Other calculations result in 300 ppm, but let’s not be alarmist.)
    2) The current level is 390 ppm. The atmospheric CO2-content is not only increasing but the rate of increase is increasing.
    3) If we continue producing CO2, there is a significant probability of a mass-extinction event such as an anoxic oceanic event occuring.
    4) From 1, 2, and 3. We must immediately stop producing CO2.
    4) We (whoever we-all may be) have already produced the CO2 involved in building the existing atomic power plants and extracting the existing fuel.
    5) Therefore we should
    a) continue using the already-existing atomic power plants until the already-existing fuel is used up,
    b) shut down CO2-producing (i.e., fossil-fuel fueled) power plants equivalent to the amount of power produced by these atomic power plants,
    c) keep the atomic power plants on-line until all CO2-producing power plants have been replaced by renewable power sources,
    d) shut down the atomic power plants one-by-one as renewable sources come on line.

    C. Mather

    Note: I am ruining an acquaintanceship with the following calculation.

    Let us assume the world-wide probability of a severe nuclear accident for the next 90 years is 10% and this accident will kill 9 million people. Then, (0,1 X 9.000.000)/90= 10.000 deaths/year.

    Within the next 90 years, climate change will kill 100% of the world’s population with a probability of 100% Then, (1 X = 77.777.778 deaths/year.

    [your reasoning from 1-5 is fine, and is pretty much in agreement with my original post. Your numbers in the last part don’t seem to make any sense though. 100% chance of killing 100% of the population???? – Steve]

  11. @C. Mather
    “… 100% chance of killing 100% …”
    for example (but not only)
    Hansen, J., et. al. “Target atmospheric CO2: Where should humanity aim?”
    as discussed in “Why it’s urgent we act now on climate change” and “Climate Emergency: Time to Slam on the Brakes”

    [Neither source mentions 100% chance of killing 100% of the population. The climate crisis is bad enough; this kind of exaggeration is unnecessary – Steve]

Leave a Reply

Your email address will not be published. Required fields are marked *