Nuclear Future Unclear

Whilst pondering this piece, I saw George Monbiot has added his ha’p’orth. I don’t know if he writes his own headlines, but I can’t help noting that “Japan nuclear crisis should not carry weight in atomic energy debate” is a prescriptive statement and the real world is normative (a distinction I probably owe to Nassim Taleb, as I’m currently reading Fooled by Randomness). Lovelock has also apparently used that ill-advised word, “should”. Maybe it should not carry weight, but it will. In fact it already has, since countries from China and India to Germany and France are reviewing their nuclear plans.  At the very least, increased safety compliance costs will be imposed on nuclear energy suppliers.

As I write things look grim. The Japanese Emperor has prepared the nation, the head of the IAEA is flying in, global stock-markets are in freefall and even the UK is politely advising nationals to “consider” leaving the Tokyo area. Russian and EU energy officials have described the current status as “worst-case scenario” and “out of control” respectively.

Maybe the Fukushima plant will by some miracle be pulled back from the brink. But the problem – as discussed in a comment on The Climate Philosopher‘s blog – is that there are now multiple interacting problems. Radiation levels are preventing access to the site and there is contention for the available resources of water, equipment and personnel.  And last-ditch measures – dropping water from helicopters and using water cannon – are being adopted to try to cool parts of the complex. Sometimes a solution presents itself in a crisis, but because there are so many problems at Fukushima, it’ll take more than a calm, clear-headed, courageous individual to save the day, as Thomas Tuohy did at Windscale. The only real hope is that electrical power to the plant is restored allowing the electric pumps that failed in the earthquake and tsunami to be restarted. Assuming they’re still operational.

The Climate Philosopher’s initial post only considered the problem for the case of a single reactor. But as we saw yesterday (Wednesday), one event can force the evacuation of the site, stalling measures to control all 6 reactors and (at least) 2 pools of spent fuel rods, at least temporarily. It seems only a matter of time before the site has to be evacuated for an extended period. If – though I’m starting to suspect when – this happens, the other reactors will release further radiation and the spent rod cooling pools will boil away.

The cooling pools are at least as much a worry as the reactors. Apparently they could even go critical if they dry out, but more likely they’ll simply burn. They now have minimal containment, so the result is likely to be a radioactive plume, presumably not on the scale of Chelyabinsk, though in a much more densely populated region. Hopefully the wind will still be blowing out to sea if such a release does happen. It occurs to me that in some ways a single explosive event, rather than a gradual release, might be preferable, since, if the leak occurs over time, and the wind direction varies, radioactive material will be spread over a wider area. And besides the direct effects, access to the site would be further restricted.

It is now almost irrelevant how serious the situation becomes, since the risks of nuclear power have now been highlighted.  The technology may well be safe if reactors and waste storage facilities are properly designed and managed.  Location is always going to be a problem, though.  We don’t have to worry just about earthquakes and tsunamis.  Monbiot supposes we can blithely:

“…add a fifth [condition for nuclear power], which should have been there all along: no plants should be built in fault zones, on tsunami-prone coasts, on eroding seashores or those likely to be inundated before the plant has been decommissioned or any other places which are geologically unsafe. This should have been so obvious that it didn’t need spelling out.”

Trouble is, George, that rules out practically all of them, since nuclear power needs a constant supply of large quantities of water.  And not only will global warming increase flood risks, it will also lead to sea-level rises which may be significant within the lifetime of reactors being planned today.

Much has been made of the folly of the General Electric (GE) Mark 1 Boiling Water Reactor (BWR) design and so-called Generation III passive-cooled reactors are undoubtedly less risky.

But the public will still be worried.  Let’s assume the wisdom of crowds is at play and the population are not just ignorant proles.  What might the concerns be?

1. Just because we avoid what can now be seen to be obvious problems doesn’t mean we’ll avoid subtle ones in the future.  We don’t know what we don’t know.  Hindsight is a wonderful thing.

An account of the fire at Windscale suggests causal factors were not just design misjudgements that must have made sense at the time – air-cooling and a change of usage mode which meant the thermocouples became suboptimally located – but also unexpected effects, in particular the deterioration of the crystalline structure of graphite under neutron bombardment.  Who would have thought of that?  And even then the precise cause of the initiation of the fire remains obscure.

Even in the best designs there is bound to be scope for human error and unexpected physical processes.

2. We are systematically underestimating risks.

There was inevitably going to be a severe earthquake followed by a tsunami in Japan.  It was only a matter of time.

Not only that, the GE Mark 1 BWR was known to be unsafe.

Why then was this accident allowed to happen?  Even without hindsight, it would have been cheaper to scrap the Fukushima BWRs decades ago and replace them with reactors of a modern design.  The storage pools could have been made safer and separation from the reactors increased.

Here’s a clue.  I read in more than one place that the criterion for design of French reactors is based on protection against a 1000 year flood.  Sont-ils des noix?!  Are they nuts?  Here’s another thought prompted by Taleb.  If there are 1000 reactors in the world vulnerable to a variety of 1000 year risks – and let’s be charitable and assume each reactor is only vulnerable to one risk – then we can expect a nuclear crisis once a year on average!

It seems that the Fukushima complex considered only the risk of a magnitude 7 quake.  Not the worst possible.  And the tsunami was not a second disaster as many media commentators have implied.  The one caused the other.  I was struck after the 2004 Indonesian tsunami that isolated Pacific islanders, who survived by heading for the hills, explained that their ancestors had warned them that: “after the ground shakes, the sea will invade the land”.

The problem, of course is the process of risk assessment.

When I get on a plane the chance of not making it to my destination is less than one in a million.  And the public expects certainty.  Any aircraft accident near-miss results in lengthy enquiries and safety improvements of various kinds.

The possibility of an accident at a given nuclear facility needs to be reduced to at most billions to one.  Multiple levels of containment are needed to allow for the unknown unknowns.

The issue is a thorny one.  Corporate interests are involved.  There is political reluctance to write off significant investments.  Hubris and other admirable human traits play a part.

Nuclear risks are not the only ones we need to be concerned about.  Climate risks are similarly misjudged.  It’s not enough to only have to worry about 1000 year floods or droughts or heatwaves or winters.  Given (say) 100s of possible 1000 year disasters, some will occur every few years!

And then there are risks of various kinds to human health, such as epidemics, and to the economy, such as banking crises.

Dealing with the risk management deficit is a huge task, but here are a couple of recommendations:

1. A formal quantitative approach is needed.  We currently have more rather less tolerance of risks that affect many people, are  disruptive, expensive to address and that require public rather than private expenditure.   Risk levels need to reflect and be seen to reflect what the public would choose.  Otherwise political support will not be forthcoming when it is needed, such as during the nuclear plant planning process.  And, as shown by their attitude to flying, the public has very low risk tolerance.  Standards are lower (mere 1000 year floods indeed!) in domains where the connection between procurement decision and disaster is tenuous. Public involvement in the design of nuclear reactors (or flood defences or epidemic and other disaster planning) is minimal.  The effects of mistakes appear only decades later.  The market cannot be relied on to impose discipline.  Government must step in.

2. Independent agencies may regulate particular industries, but a strong central government function is needed – an Office of Risk Assessment and Mitigation, perhaps, or maybe simply Disaster Avoidance.  At present, it seems the UK Chief Scientist takes on this role.  Maybe responsibility for risks could be made more explicit, but science is not be the only discipline required, so perhaps a new department should be established.  I would not envisage a large staff.  The role would be more akin to audit of government departments, functions and divisions of responsibilities.  And brainstorming scenarios.  A similar office may be needed at other levels – large cities on the one hand and the supranational (so the EU and UN) on the other.  Regardless of the details, teeth are needed.

I was shocked that Japanese towns were so vulnerable to the recent tsunami.  After all, this sort of thing has happened many times before – tsunami is a Japanese word.  I was even more shocked the nuclear plants were so exposed to risk.  It may be somehow rational at an individual level to ignore 1000 year risks – after all a given disaster is unlikely to occur in one’s lifetime, or before an executive has enjoyed his healthy pension – but we are now in an era where we are not only more interconnected than ever before, magnifying and proliferating risks, but also have come to expect a lifetime free from war, plague, pestilence and the other ills brought to previous generations by the ancient Horsemen.