I’m going to break the format of this blog a bit, such as it is, to make a few journalistic observations about the US response to the earliest days of Fukushima.
Firstly, the US NRC seems to have been shell-shocked by the event. It should have been obvious from the first day that Unit 1 was in meltdown. The commission knew, for a fact, that there was no cooling in the core; this is well established. And it does not take a degree in nuclear physics to know what happens to nuclear cores without cooling. Yet it was three days before anyone was willing to admit the obvious, even to themselves. By all accounts the commissioners just watched the clock as over forty hours-without-cooling ticked by, and continued to assure the world that Japan had it under control. It is as if a lifetime of fervently espousing the impossibility of core-melts had made them unable to see what was right in front of their eyes.
This is cognitive dissonance on an epic scale.
Secondly, the Defence Department seems to have been much quicker to grasp the nettle. It was the Navy’s nuclear reactor division that appears to have been the real driver for belated US action. (The voluntary evacuation order, for instance, and a strongly worded message to the Prime Minister suggesting ‘heroic action’ — ie: suicide squads — when it looked like TEPCO were pulling out.)
Thirdly, the limited US evacuation recommendation (50 miles) was far too limited from a safety perspective. Within 72 hours it was clear to everyone that the citizens of Tokyo were in serious jeopardy. This is to say there was substantial evidence of a zirconium fire in the unit 4 spent-fuel pool (not least the fact that the building exploded, despite there being no fuel in the reactor core) — an event that would have released a tremendous amount of radioactive fallout, and would almost certainly have led to the loss of a much larger pool nearby. As it happened, the fuel had begun to melt but the Japanese were able to bring the pool back from the brink; but nobody at the time took this for granted.
When questioned about this, authoritative experts repeatedly stress that the NRC simulations showed no threat to Tokyo, even from a spent fuel fire. What they don’t say, however, is that the NRC simulations (which use a system called RASCAL) were only capable to modeling effects up to 50 miles out, and so they could not have shown a threat to Tokyo under any circumstances. The Germans had already advised their citizens to evacuate the capital, no doubt at some diplomatic cost. So had the French, the Russians and the Chinese — all countries with heavy nuclear investments and none of them exactly wilting violets when it comes to radiological hazards.
Tokyo was in serious trouble, and the US must have known.
This leads to my final observation: that the US decision not to order an evacuation beyond 50 miles — quickly and obediently echoed by the UK — was a political decision not an evidence-based calculation. It was a decision to put US citizens in Tokyo at risk as a favor to the Japanese government, who were desperately worried that a mass international evacuation would have led to uncontrollable panic in their capital.
It was a gamble that paid-off. Other countries looked to the US for cues and echoed their recommendations, stemming the tide of evacuation recommendations. Mass panic was averted. TEPCO — with ‘heroic efforts’ and no small amount of heroic luck — were able to save the spent-fuel pool, and with it the residents of Tokyo. (Although it remains in jeopardy).
In other words, Japan owes the US a solid, and they’re acutely aware of the fact. Sore misgivings about US bases on Japanese soil have all but disappeared since the crisis.
The town on Plymouth, on England’s south coast, can trace its history back to the bronze age. A natural harbor, it has a proud maritime tradition. The pilgrim fathers left from its port in 1620 to settle the New World. It is currently home to the largest operational naval base in Western Europe.
On 29 July, last year, Plymouth almost became a radioactive ghost-town – the victim of a reactor meltdown due to a technical mishap. The UK’s own Pripyat or Fukushima. Or so recent disclosures suggest.
Plymouth isn’t even home to a nuclear reactor. Not a permanent one anyway. Its naval base services many of the UK’s nuclear submarines. The near catastrophe was was caused when a series of “unidentified defects” [echoes of Normal Accidents here – ed] led to the loss of both primary and secondary power to the onsite subs’ coolant systems for more than 90 minutes.
A subsequent investigation pinned the failure on a “defect in the central nuclear switchboard.” A failure that would almost certainly have been deemed “impossible” by anyone studying the system’s blueprints in advance.
Who knows how close Plymouth came that night. The incident certainly seems serious enough, but the secrecy around such events invariably prohibits proper scrutiny for decades. I highlight the scare mainly as a rejoinder to those who dismiss revelations about near-misses in the past with the argument that technologies have changed since then, making past failures irrelevant.
Technologies change, yes, but the fact that complex systems fail for unexpected reasons that defy our risk analyses remains. It is as constant as our willingness to wager, over and over, that this time, this time, the systems are safe and the assurances are accurate.
As someone with a longstanding interest in the media coverage of Fukushima and how it seems to exculpate the nuclear industry, I was struck last night by a passage from Eric Schlosser’s fascinating new book “Command and Control“. In it he outlines the recommendations of a top secret 1959 RAND report — authored by one-time MIT professor and Undersecretary of Defense, Fred Iklé — concerning how the US should manage the publicity fallout from an accidental nuclear bomb detonation. Schlosser quotes directly from the report:
“If such an accident occurred in a remote area, so that leakage to the press could be prevented, no information ought to be made public. […] If the accident has been compromised and public statements become necessary, they should depict the accident as an occurrence which has no bearing on the safety of other weapons.”
The report further recommended that the crisis be drained of its immediacy by establishing an authoritative “board of inquiry” that would take several months to reach its conclusions. Schlosser quotes again:
“During the delaying period the public information program should provide the news media with all possible news about rehabilitation and relief. There is always a strong and continued interest in such news after a disaster. Within a relatively short time the interest in rehabilitation tends to crowd out reports about destruction and casualties.”
Any parallels with Fukushima are, I am sure, entirely coincidental.
(Page 195 on my kindle, for anyone interested to read further.)
Fukushima has been back in the headlines of late, with the ‘surprise’ discovery that hundreds of tons of Strontium-laced radioactive water has been leaking into the Pacific, where the processes of bioaccumulation and biomagnification will ensure that it eventually winds up in our sushi. I put scare-quotes around ‘surprise’ because it really wasn’t a much of a shock to people intimately involved with the disaster. Individual leaks might be difficult to spot, but they know how much water they are pumping into the ruined reactor buildings, and they know how much water they are storing. It isn’t a very complicated math problem to figure out how much water is being lost. Especially, one might imagine, to the nearest hundred tons. Seriously, where did people think that water was going exactly?
With all the focus on the leaks though, I worry we have been missing the much (much) bigger dangers the continuing disaster poses. There is still the spent fuel-pool sitting atop unit 4, which is badly damaged and increasingly unstable. Even a relatively modest earthquake (the normal and entirely expected kind) could bring it down. If that happens, which it very well might, it would inevitably lead to a radioactive fire that would dwarf Chernobyl’s by several orders of magnitude. It is difficult to make a case that it would not (at minimum) be the end of Tokyo as a habitable city. (The implications of this alone are difficult to comprehend. It would mean relocating over 30 million people, and who knows what for the global financial system.) Less bad for the environment but almost as bad for Tokyo would be if the reactor cores melts its way down to the city’s aquifer, which runs under the plant. Again, a very real possibility.
Such fears, should they be realized, would come as ‘surprises’ to the vast majority of the media-reading public, but they really shouldn’t. The mainstream media might be ignoring these issues for now, but there are plenty of knowledgeable people trying to make themselves heard about the dangers. This article, for instance, does a pretty good job of summarizing some of the issues.
In the time I have been exploring the debates around Chernobyl and Fukushima, I have often been surprised by the extent to which authoritative sources seem to frame their evidence in ways that minimize the appearance of danger. Occasionally, however, one cannot help but be a little awed by the artfulness involved.
Take, for example, the following map, which depicts the ‘Deposition of 137Cs [Caesium 137] throughout Europe as a result of the Chernobyl accident’:
This map – originally created by the Russian Institute of Global Climate and Ecology (IGCE) but faithfully and prominently reproduced by the IAEA in their seminal (2005, also 2006a) 20-year report on “Chernobyl’s legacy” – is interesting and important in many ways.
The 2005 IAEA report has become the ‘go-to’ document for authoritative information about the accident’s health and environmental impacts. And although Caesium 137 was not the only radiological pollutant put out by the meltdown, it was the most significant. The map thus represents a central part of a narrative that is deeply implicated in our current energy policies and risk calculations.
It is not the integrity of the data portrayed in this map that I will comment on here (although the data is interestingly contested,) but rather the presentation of that data: the map’s semiotics, the logic of its construction, and the relationship of both to the report that reproduced it.
Let’s begin with impressions. All images are subjective, of course, but at first glance the map might seem disturbing. Ominous red ink, with its connotations of danger, spreads deep into Western Europe. A band across the Scandinavia reflects the prevailing winds at the time of the accident, and a second ‘hotspot’ hovers over Austria. To me at least, the map suggests an ‘archipelago of hazard’ that links Ukraine and Belarus together with Austria and the Nordic states as partners in radiological hazard.
When understood in its proper context, however, the map’s connotations are far more reassuring.
To see why a map of radioactive fallout can be viewed as reassuring, it is necessary to look at the argument the IAEA report is making, and the charges to which it is responding. In 1995, with the 20-year anniversary of Chernobyl fast approaching, the accident was undoubtedly going to be in the news. This could have been a disaster for the nuclear industry. The IAEA’s official toll of ‘56 deaths’ simply wasn’t credible by then, and it was highly likely that the media would be seduced by the growing literature attesting to regional horrors and attributing mass deaths to the accident.
The IAEA report was an effort to get ahead of the story. Its (now widely circulated and highly trusted) figure of “no more than 4000 deaths due to Chernobyl” does not deny a legitimate ‘mortality crisis’ in the region but, unlike the critical literature, it resolutely refuses to attribute almost any of that crisis to the meltdown. Instead it attributes the well-documented plummet in the region’s life-expectancy to various exogenous factors. Many, the report suggests, are attributable to the socio-economic upheaval that followed the collapse of the Soviet Union and affected the entire post-Soviet block. The more extreme mortality around the plant itself, meanwhile, it attributes to an irrational ‘nuclear fear’ driven by an alarmist media and an infrastructure of compensation. (It argues that when the international community designated the affected population as “victims” rather than “survivors,” people in the region were led “…to perceive themselves as helpless, weak and lacking control over their future” which, in turn “…led either to over cautious behavior and exaggerated health concerns, or to reckless conduct […] and unprotected promiscuous sexual activity” [IAEA 2005]).
As far the IAEA is concerned, in other words, the principal thing we have to fear from Chernobyl’s fallout is fear itself. The slightly bizarre (and subsequently ridiculed) charge of death-by-promiscuity was removed from the second edition of the report, but the rest still stands today. Chernobyl didn’t kill people directly, we are told, the Soviet Union killed them by collapsing, the media killed them with alarmist coverage, politicians killed them with compensation, and people killed themselves with their own irrationality.
The map is reassuring, at least from the nuclear industry’s perspective, because it appears to support this narrative. It suggests that Norway, Sweden, Finland and Austria were all subject to Chernobyl’s radiological fallout along with Belarus and Ukraine, yet we know that none of the Western European countries suffered a significant mortality crisis after the disaster. Their people faced some restrictions and economic hardships, to be sure. Certain ruminants such as reindeer (in Finland) and wild boar (in Germany and Austria) were kept from the food chain for many years, but there were no deaths or decreases in life-expectancy that were even remotely comparable to those in the East. In other words, the map suggests that the many deaths that some critics attribute to Chernobyl seem to align much more closely with social and political geography than they do with the geography of Chernobyl’s fallout. Its subtext, in essence, is that Chernobyl’s pollution was not so bad as to justify more than 4000 deaths.
But let’s look again; this time at the map’s key:
The numbers on either side represent different units of radiation: Becquerels (or kilo-becquerels per meter squared) on the left, and curies (per square kilometer) on the right. Both sides are equivalent so I will speak to the left as it is easier to follow.
Ignore the units for a moment, however, and instead look closely at the scale: the different segments, and the colors that represent them. Notice first that anything under 2 kBq/m2 is yellow, (nowhere escapes the palest hint of radioactive contamination). See next that the scale now goes from 2 to 10: a span of eight units, represented by a slightly deeper yellow. Then it goes from from 10 to 40: a span of 30 units, and a still deeper yellow. Pause here to note that there is now a change in ‘color’ rather than ‘hue,’ and we move from yellows to the more ominous shades of red. So the next segment, 40 to 185 kBq/m2 – now a span of 145 units – is light red. The penultimate segment, 185 to 1480 – a span of 1295 units (!) – is darker red. And the final segment, represented by the deepest red, begins at 1480 and has no upper boundary whatsoever.
Readers of the report can only speculate what the highest caesium concentrations might be. (Although a different IAEA report [IAEA 2006b] – one conducted concurrently with the ‘Chernobyl’s Legacy’ report but not formally considered in its conclusions – found that concentrations of 90Sr [Strontium, which is generally considered to be a less significant pollutant than caesium] “exceed 4000 kBq/m2 over large areas” around the site of the accident [IAEA 2006b: 3]).
And besides, what scale goes from 0 to 2, to 40, to 185, to 1480, to… infinity? It is not linear, clearly, but nor is it logarithmic. I have stared at this scale for a long time without discerning its scientific logic. My only conclusion is that its guiding logic is guided less by science than by an intention to deceive. Look again at the map. Notice that all the ‘red’ areas outside of Ukraine and Belarus fall into the 40-185 range: at least an entire order-of-magnitude lower than some of the levels in the areas around Chernobyl itself, even though the map’s coloring inclines us to equate them. The scale and its coloring, without necessarily lying, artfully occludes the fact that the levels of caesium pollution around Chernobyl were far higher than in other areas of Europe.
Now couple this realization (that pollution was far higher around the disaster) with the fact that other countries – including the UK, (which doesn’t even make it ‘into the red’ pollution wise, and is certainly no nuclear critic) – felt compelled to test and restrict the sale of livestock, at great expense, for decades after the accident to protect the health of its citizens. Understand also that Chernobyl sits at the heart of the Dnieper reservoir system: an agricultural floodplain – sometimes described as “the breadbasket of Russia” – where the former Soviet Union grew (grows) the vast majority of its grain (the distribution of which maps clearly onto political boundaries). And suddenly the stories of mass deaths and enormous health costs from Chernobyl become all the more plausible.
The IAEA is often viewed as an impartial voice on matters nuclear, but it is important to remember that the organization’s core mandate is the promotion of nuclear power. Some industries have paid lobbyists to defend their interests, others have entire divisions of the United Nations.
IAEA (2005) “The Chernobyl Forum 2003-2005: Chernobyl’s Legacy: Health, Environmental and Socio-economic Impacts. And Recommendations to the Governments of Belarus, the Russian Federation and Ukraine” Second revised version. Available online: http://www.iaea.org/Publications/Booklets/Chernobyl/chernobyl.pdf
IAEA (2006a) “Environmental Consequences of the Chernobyl Accident and their Remediation: Twenty Years of Experience: Report of the Chernobyl Forum Expert Group ‘Environment’”
IAEA (2006b) “Radiological Conditions in the Dnieper River Basin: Assessment by an international expert team and recommendations for an action plan” Radiological Assessment Report Series; International Atomic Energy Agency. Vienna.
The high-consequence / low-probability nature of atomic risks inevitably puts nuclear skeptics at a disadvantage.
Imagine, for instance, a hypothetical engineer who, in the early 1970s, studies the risk analyses of the Fukushima Daichi nuclear plant and correctly identifies the various shortcomings that will one day bring it down. She understands that the risks of earthquakes and tsunamis are higher than envisaged, and that the plant is ill-designed to accommodate them. This leads her to presciently imagine a natural disaster that causes a chain of meltdowns that, at their peak, legitimately threaten the future of Tokyo: home to 35 million people and a cornerstone of the global economy.
Now imagine her future. It is comforting to believe that her insight would have led to a reconsideration of the plant – a slew of changes that would have undone Fukushima’s fate and protected Japan in April 2011. But, alas, this narrative isn’t very credible. It neglects the fact that all through the development of that plant – and, indeed, all nuclear plants – there were credible critics who made such warnings and were ignored.
Engineers like to imagine that truth ‘shines by its own lights’ but this profoundly misrepresents technical discourse. Every technological finding is actually an ‘argument’ with contestable uncertainties and interpretations. And where powerful interests are at stake – as they have been throughout the history of nuclear technologies – unwelcome technical arguments are always vigorously contested. ‘Truth’ is always negotiated, because although veracity does shine to a degree, it rarely shines brightly enough to be distinguishable on its own.
So it is that our engineer’s prescient arguments are met with counterclaims and doubts. ‘The evidence is incomplete,’ she is told. ‘Your calculation is built on an uncertain variable’. ‘Our studies reach different conclusions.’ ‘The plant is safe.’
In most engineering circumstances she could look forward to her vindication, terrible though it might be. Eventually her predictions would come to pass, and her critics would be proven wrong. Many technical and scientific reputations have been built on unpopular claims that are subsequently borne-out: journeys in the wilderness endured for their eventual glories. Karl Popper saw this dynamic as exemplary of the scientific method.
The nature of nuclear risks makes vindication unrealistic, however. We know today that our engineers fears will be realized in time, but it would have been unrealistic for her to expect this or to build a career around it. Nuclear risks have timeframes of hundreds and sometimes thousands of years (millions, in the case of waste storage). These timeframes are vital, given the gravity of the hazards involved, but they fit awkwardly onto human discourse. This is to say that even if Fukushima’s actual risks were orders of magnitude higher than the industry’s calculations promised, any critic who understood the real risk and argued for it could expect to go an entire lifetime without seeing their claims validated. (Even if our engineer had seen the future, she might still have been deterred. Construction on Fukushima began in July 1967; the plant did not fail until almost 44 years later – the length of an entire career.)
With no realistic hope of empirical vindication, our engineer is forced to slug it out in an exhausting crossfire of claims and counterclaims. It is a losing battle. Every year that passes without the disaster she foresees subtly undermines her credibility. This is unfair as, statistically-speaking, even decades of safe operation prove very little about a reactor that is expected to melt down no more than once in every 100,000 years (or more). But humans are not statistical beings. Our imaginations and intuitions are honed for numbers at a human scale. Forty years without a catastrophe intuitively feels like compelling evidence of a reactor’s safety, just as fifty years without an atomic holocaust seems like compelling evidence that “deterrence works” (as I outline in a previous post).
Our engineer’s antagonists fuel and leverage this misperception of safety, and wield enormous resources through which to do so. Billions of dollars are invested in the notion that nuclear is safe, as are an array of expert careers and widespread notions of national security: a pervasive network of economic, military and professional interests. So it is that she comes to be marginalized by a vast ‘nucleocracy’ that controls the official publication channels, the professional bodies, and, through them, the public perception of ‘orthodox’ knowledge.
The inevitable result is a tragic career. For like Cassandra, the mythical Greek beauty who angered the gods, she is cursed with terrible foresights that nobody will heed. She struggles to publish. Her work is vilified by aggressive and well-funded lobbyists. She is appointed to few, if any, official committees. In short, she receives few of the accolades through which modern societies confer expert legitimacy: the ‘right to speak’ on the esoteric nuclear matters that frame our lives and livelihoods depend. She publishes in the ‘alternative’ media – mostly online – whichs offer a platform from which to speak, even as it further drains her credibility.
Such is the fate of the nuclear naysayer.
This vignette is hyperbole to some extent, of course, but there is more far more truth to it than most people realize. For a compelling glimpse into the fate of even well-established scientists who defy the ‘nucleocracy’ see Gayle Greene’s fascinating (2011) article: “Richard Doll and Alice Stewart: reputation and the shaping of scientific ‘truth’” in Perspectives in Biology and Medicine. 54 (4): 504–31. Or look instead to the various critical studies of Fukushima’s safety that were marginalized before the accident and rediscovered afterwards. Also, for an account of why our engineer’s ‘vindication’ is likely to be hollow even when it comes, see this working paper by yours truly.