03/15/2011 - 13:59

Guarding against disaster: As Japan's tragedy becomes more serious, so does the need to learn from it

Joshua Pollack

Joshua Pollack

Pollack is a consultant to the US government. He has conducted studies in several areas, including arms control, verification technologies, proliferation, deterrence, intelligence, homeland...

More

Even as it worsens, Japan's nuclear accident still pales in comparison to Japan's natural disaster. The estimated death toll from the earthquake and tsunami, now climbing toward five figures, numbs the imagination. It seems almost tasteless to dwell on what amounts to a minor, if sensational, feature of an incomprehensible tragedy.

We should not assume that the present disaster is just a statistical freak, never to be repeated. It's quite early in the nuclear era for a once-in-1,142-years event to roll along and deliver the first serious simultaneous reactor accidents.

Viewed in isolation, though, the evacuation of hundreds of thousands of people from their homes and (as of this writing) the possibility of one or more full-scale meltdowns can hardly be considered a mere detail. In the worst possible case, spent fuel stored onsite could melt and spew radioactivity over a wide area. Even in the best case, it appears that radioactive steam will vent from the ruined cores of the Fukushima Daiichi Nuclear Power Station for some time to come.

So what can we learn from this calamity, which is both the first major nuclear accident resulting from a natural disaster and the first serious simultaneous failure of multiple reactors? For the anti-nuclear camp, it will surely mean that we ought to throw in the towel on nuclear power completely -- nothing less. For the fervent advocates of nuclear power, it will mean instead that we just ought to accept that "stuff happens" sometimes. That might seem like a tough sell, but as noted above, the non-nuclear destruction and death wreaked across northeast Japan overshadow almost any imaginable outcome at Fukushima.

The contest between these perspectives will unfold differently from country to country, but the overall outcome is not in much doubt. As hard as it may be to accept at this moment, humanity can expect to live with this technology for many decades, perhaps centuries to come; the needs of growing populations and economies simply cannot be denied, especially in the shadow of the global-warming threat. But by the same token, the buildup of people and infrastructure worldwide means more potential over time for Fukushima-type events in places threatened by seismic or meteorological disasters.

The public will, at a minimum, insist on safety improvements. It certainly ought to. A number of constructive ideas are likely to surface concerning how to strengthen design standards for future reactors and how to retrofit existing facilities. Whether the results will satisfy anyone is a different question. Accidents on some scale will almost certainly continue to happen every few decades; anyone who failed to recognize this truth before will be hard-pressed to deny it now.

Two sets of questions are now worth asking. The first is how and why the failures happened. The second is whether there are any historical antecedents for this type of problem, how the responsible parties responded, and whether this record offers any hints of an answer to the present dilemma.

Weeks or months may pass before events are fully understood, but based on early accounts from the International Atomic Energy Agency and Japan's Federation of Electric Power Companies (FEPC), Fukushima appears to be a cascading natural-technological disaster. The earthquake triggered automatic shut-downs at multiple reactors. But it also knocked out the power grid, forcing operators to use backup generators to keep coolant flowing into the hot reactor cores. When the tsunami swept onshore, it knocked out the generators, throwing several reactors into crisis. The details of this story may change as more becomes known, but the general outlines are quite plausible.

The "reference event" scenarios for the design of the reactors do seem to have included an earthquake-tsunami combination, but as the FEPC statement says, "the force of the tsunami exceeded the assumed range." That seems understandable. At least one expert is calling the 8.9 (or 9.0) magnitude quake the most severe seismic event in Japan since the year 869. From that perspective, the reactor operators could be congratulated for a relatively successful outcome, so far.

Yet we should not assume that the present disaster is just a statistical freak, never to be repeated. It's quite early in the nuclear era for a once-in-1,142-years event to roll along and deliver the first serious simultaneous reactor accidents. Globally over the span of a few decades, a number of potential hurricanes, earthquakes, or floods might answer to a similar description. Reactors under construction today may well be operating when another part of the world experiences a millennial event of its own.

If that doesn't sound reassuring, it shouldn't be. But the developed world has faced and overcome similar challenges before. During the 19th century, when the United States was growing rapidly, construction and firefighting technology proved inadequate to stop devastating, citywide fires that would occur somewhere once or twice in a generation. Each time the heart of a major commercial center like New York, Chicago, or Boston burned, it would take dozens of overtaxed insurance companies down with it, leaving property owners nearly empty-handed. While pressing for technological improvements, insurers also developed survival strategies based on the principle of geographic diversification. By reducing their exposure to any single location, they could make payments and continue doing business even after a massive conflagration.

An analogous strategy for nuclear safety would also be twofold. On one hand, stricter standards for reactor design (just like improvements in construction and firefighting) would reduce the overall frequency of disasters. The placement of backup generators in reactors near coastlines might be a good place to start the conversation. On the other hand, a systematic effort to reduce the consequences of accidents (just like geographic diversification) could do a great deal to prevent harm when such events -- almost inevitably -- do happen.

One obvious consequence-reduction strategy is to remove the most dangerous substances from the scene. The time is long past to start moving highly radioactive spent fuel away from reactor sites and into long-term, secure storage. As of this writing, the condition of the spent-fuel ponds at Fukushima appears to be deteriorating -- a wake-up call in the best case, and a nightmare in the worst case. Especially compared to current practices, deep borehole disposal looks highly attractive.

Similarly, plans for the adoption of mixed-oxide (MOX) fuel, such as the 32 MOX assemblies reportedly in the core of Fukushima Daiichi unit 3, should be curtailed. MOX fuel, with questionable economic benefits, is more toxic and radioactive than standard uranium-oxide fuel, making a dispersal event more harmful.

Another harm-reduction strategy echoes geographic diversification even more directly. The practice of locating numerous reactors together might seem attractive from an economic point of view, and perhaps also from a social-acceptability perspective. But for future reactors, this practice ought to be revisited. The Fukushima disaster demonstrates the special risks of collocated units, which stand exposed to the same external hazards.

Japan is still grappling with a national crisis and has only begun to recover its dead. Hopefully, it's not too soon to seek valuable lessons for the entire world in the present tragedy.