Welcome to the age of diminishing returns

Thursday, March 17, 2011

Fukushima: the nuclear martingale


The "martingale"(*) is an old scheme for betting on the roulette game. It is simple, seductive, promising, and a recipe for disaster. The Fukushima catastrophe may be the result of the same way of thinking: a nuclear martingale.


It is unbelievable how many people today still think they can  make money out of a roulette game by using the old "double-up" method, also called "martingale". It works like this: you bet on a 50% event, red/black or odd/even, and you double the wager every time you lose. Eventually, when you win, you recoup your losses and make a small profit. Or, at least, this is the theory.

In practice, the martingale method is a recipe for disaster. At best, with it you will not be better off than with playing numbers at random. But it is worse than that: the martingale gives you a good chance to go broke by seeking to double "one last time." Fortunately for gamers, most casinos have betting limits; casino managers want their customers to lose money but not so much that they will go broke and kill themselves.

The martingale strategy is related to what Nassim Taleb has termed "Black Swan", an improbable but catastrophic event. A black swan event is not just a stroke of bad luck. It is something that you create by a series of wrong choices made with the best of intentions. It is the idea of the martingale: a sort of game of chicken played against the laws of probability. In a sense, you try to scare reality by raising the stakes - such as when you double the wager at the roulette game. But whatever small success you obtain in this way, reality is not easily scared and it comes back with a vengeance in the form of black swan: the bigger and the more catastrophic the more you had tried to avoid it.

Martingale-like schemes are typical, for instance, of the financial world. The subprime mortgage crash that started in 2007 is a good example of this strategy as noted by Nassim Taleb. Many financial schemes may be based on similar ideas. And, in these cases, there are no casino managers who stop people from falling into the martingale trap and go broke.

The Fukushima nuclear disaster may be telling us that there is a similar mechanism at play with technology in general. When we design machinery that is dangerous and prone to failure we try to reduce risks by tight regulations, specifics, and centralised control. Of course, these strategies are expensive and therefore are best implemented over large scales. So, we are raising the stakes by building bigger and more expensive systems in order to hedge the risk of failure. In the case of nuclear energy, the result is the concentration of power production in large plants. That strategy seems to work, within limits: on the average, the safety record of the nuclear industry is not bad. But when something goes wrong with a nuclear plant, it tends to go wrong in a big way, such as with Chernobyl and Fukushima.

So, are we protecting ourselves against small failures at the cost of risking large ones? In such case, we would be playing the martingale on a truly gigantic scale. The problem is not specific with nuclear technology. We tend to hedge risks with all kinds of technologies at the cost of risking catastrophes.

Think of coal as an example. We know that burning coal in power plants carries risks. In addition to local pollution, coal may be a major factor in overheating the whole planet because of the greenhouse effect associated with carbon dioxide (CO2) generated by combustion. Against this risk, we are presently planning a major effort in terms of "carbon capture and sequestration" (CCS) - a technology usually referred to as "Clean Coal." The idea of clean coal is that CO2 can be stored underground in geological reservoirs and therefore prevented from reaching the atmosphere.

We may well be playing the martingale with this idea. Suppose that the carbon capture technology will be used on a large scale and that we end up relying on "clean coal" for a major fraction of the worldwide power production. Then, we will have hedged the risk of climate change by raising the stakes: invested money and resources on a specific technology. Likely, we will have reduced local pollution and the amount of CO2 emitted in the atmosphere. But do we know enough about the physics of the sequestration process to guarantee that the stored CO2 will stay there? ? How can we rule out that we'll get that CO2 back in the atmosphere all together and much sooner than expected?

As a black swan, this one borders the unimaginable. Maybe it is an improbable event; sure, but it would be much more improbable if we were just to stop burning coal.

But we just don't seem to be able to reason that way. We tend to go always for the bigger and the more sophisticated technological solution and that carries enormous risks - maybe in terms of unlikely events, but not impossible ones. We are addicted to technology (as noted by George Mobus) and we don't seem to be able to realise that at some point technology starts showing diminishing returns (as Joseph Tainter has noted).

Maybe this is exactly what we are doing with civilisation: playing the martingale. We are hedging small risks by developing technologies, regulations, laws, and controls, all in order to keep society together. But the risk is the improbable, but eventually unavoidable, total collapse. The biggest black swan of all.


___________________________

(*)  In probability theory, the term "Martingale" refers to processes akin to random walks. Its origin is obscure, but a discussion on its meaning can be found here

9 comments:

  1. It's time to get the nuclear industry off taxpayer-funded welfare. No more loan guarantees and liability limitations for that vicious, corrupt and polluting industry. If they can't afford to pay their own insurance in the much-loved free market, then they can bloody well fail.

    And since no insurer will write a policy for a nuclear plant in the USA without both loan guarantees and liability limitations - the funding for said having been filched from the pockets of taxpayers by your ever-loving government - there will be no more nuclear plants in the USA.

    Turns out Japan's nuclear industry is on taxpayer welfare too. A taxpayer-funded liability limitation of $1.2 billion, in particular.

    The people in Japan pay thrice: once for the industry's too-hazardous-to-insure nature, next for the actual power, and third when a reactor blows - the last with their lives.

    The USA's taxpayer pays four times: an additional theft of money for loan guarantees, which it seems the Japanese taxpayer doesn't have to fork over for.

    ReplyDelete
  2. thanks for this interesting post. these theories make me think, it is true that the man will destroy the earth.

    ReplyDelete
  3. Not necessarily, Alessandra, but there are good chances that we'll destroy ourselves.

    ReplyDelete
  4. Please don't lump Chernobyl in with Fukushima and Three Mile Island. Chernobyl was an older design, and was being operated irresponsibly. Fukushima will not blow radiation all over Japan like Chernobyl did to Russia - that's what containment structures are for. Containment has never been breached at any nuclear facility in the world, and despite extreme events at this plant, there's little danger of that happening there.

    Calm down, everybody...

    DK

    ReplyDelete
  5. @dooberheim,

    According to the NISA, IAEA, and ASN, Fukushima now rates as strictly worse than the partial core meltdown at TMI-2 (Three Mile Island, Unit 2). In fact, according to the ASN, Fukushima now rates as only one level better than Chernobyl. Is this reassuring? Chernobyl garnered the worst rating available. Certainly, these accidents are comparably bad.

    According to the authoritative timeline for Fukushima Daiichi at Wikipedia, containment integrity for Units 2 and 3 is not known but suspected damaged. In other words, containment has probably already been breached. The ominous smoke emitted from these same units yesterday afternoon is not a reassuring sign, nor the fact TEPCO measured a 126,700% increase in iodine-131 in randomly sampled seawater adjacent to the plant.

    A 126,700% increase. I wish that was a crude joke, naturally. It's not.

    References:

    * http://en.wikipedia.org/wiki/Timeline_of_the_Fukushima_nuclear_accidents
    * http://www.presstv.ir/detail/171118.html

    Have a warm week.

    Humbly yours,
    Sess

    ReplyDelete
  6. Fukushima reactor was built in 1971,
    Chernobyl was built in 1977.
    Which one is newer?

    ReplyDelete
  7. Very different kinds of reactors. Chernobyl was a graphite-moderated military reactor designed for the production of plutonium. Fukushima was an LWR civilian reactor; built to produce electricity. It doesn't matter so much - anyway - both are reactors in the past tense by now.

    ReplyDelete
  8. Comrade RutherfordApril 4, 2011 at 8:08 AM

    "Please don't lump Chernobyl in with Fukushima and Three Mile Island. Chernobyl was an older design, and was being operated irresponsibly. Fukushima will not blow radiation all over Japan like Chernobyl did to Russia - that's what containment structures are for. Containment has never been breached at any nuclear facility in the world, and despite extreme events at this plant, there's little danger of that happening there.

    Calm down, everybody..."

    Your hubris is showing.

    "Fukushima will not blow radiation all over Japan like Chernobyl did to Russia"

    It already has:

    Fukushima's radioactive plume reaches the UK
    http://www.newscientist.com/blogs/shortsharpscience/2011/03/for-the-first-time-since.html


    "Containment has never been breached at any nuclear facility in the world, and despite extreme events at this plant, there's little danger of that happening there."

    It already has:
    "A containment vessel at the Fukushima Daiichi nuclear reactor cracked under the pressure of the rising heat and steam inside it yesterday (March 15), and another vessel has possibly ruptured today. Just a few days ago, experts believed the vessels to be fail-safe."
    http://sg.news.yahoo.com/containment-breach-fukushima-implications-20110316-103604-762.html

    Please, Mr. Nuclear Power Is Safe Clean And Too Cheap To Meter, stop. Everything you say can't happen comes true! Think carefully before you speak, as it will come true too.

    This disaster is not over yet, and will take months to just 'control' the situation. Like Chernobyl, this will have devastating effects for decades to come.

    Now we find out that #1 may be spontaneously fissioning all by itself!

    For a more reasoned and unvarnished look at Fukishima, go to Fairewinds Associates website and watch the videos. This guy knows that he's talking about.

    http://fairewinds.com/

    ReplyDelete
  9. Comrades, thanks for these comments but this blog is not dedicated to discussing radioactive contamination. It is all right to comment once, but please don't start a flame. Thanks!

    ReplyDelete