Fukushima: soteigai or zatzusei
Monday the independent panel appointed to investigate the Fukushima nuclear accident released a 507 page interim report. Most of the document focuses on specific operational decisions and tactical choices.
Several specific failures are highlighted: insufficient planning, poor regulation and oversight, inadequate training and exercising, a breakdown in communications within the government and between the government and the operator of the nuclear power plant.
The previous paragraph could be quickly edited to apply to nearly every serious industrial accident: Bhopal, TMI, Deepwater Horizon, various large-scale blackouts and others. The same failures are referenced in most after-actions for events large and small.
Also typical has been most of the media coverage focusing on personal failures by political, regulatory and corporate leaders.
But toward the end of the report — and the 22 page English-language executive summary — are several atypical bits of analysis worth much more attention than given so far.
It is not easy to admit an absolute safety never exists and to learn to live with risks. But it is necessary to make effort toward realizing a society where risk information is shared and people are allowed to make reasonable choices.
A quarter century ago I made some extra Yen editing Japanese-to-English translations. This time I will mostly leave the first draft as it is. There is a kind of clarity in the slightly awkward but more literal rendering.
Even for an accident of low probabilities so long as extremely large scale damages are anticipated once it occurs… due consideration should be given to the risks involved and precautionary measures should be taken.
It was a major shortcoming for the safety of both nuclear power plants and surrounding communities that a nuclear accident had not been assumed to occur as a complex disaster. Disaster prevention programs should be formulated by assuming complex disasters, which will be the major point in reviewing nuclear power plant safety for the future.
It cannot be denied that the viewpoint of looking at a whole picture of an accident was not adequately reflected in nuclear disaster prevention programs in the past.
The nuclear disaster prevention program had serious shortfalls. It cannot be excused that nuclear accidents could not be managed because of an extraordinary situation that… exceeded the assumption.
The Investigation Committee is convinced of the need of paradigm shift in the basic principles of disaster prevention programs for such a huge system, which may result in serious damage once it has an accident.
Whatever to plan, design and execute, nothing can be done without setting assumptions. At the same time, however, it must be recognized that things beyond assumptions may take place. The accidents this time present us crucial lessons on how we should be prepared for such incidents beyond assumptions.
Low probability, high consequence events deserve our sustained attention.
Reasonable assumptions will be exceeded.
The chairman of the investigation panel, Yotaro Hatamura, has been especially critical of the tendency to blame the crisis on soteigai. This is often translated as “unforseeable events,” but is probably closer to “unimaginable events.” (Echoes of a “failure of imagination” in the 911 Commission report.)
Hatamura is an engineer. His best known work is probably Learning from Design Failures in which he examines more than 100 cases to “uncover the root cause, reveal the scenario that led to the unwanted event, describe what happened so readers can clearly repeat the steps in their mind, and propose ways to avoid those mistakes in the future.” It is a very detailed, case-by-case, engineering oriented approach to disciplined thinking. He is a solution-oriented guy.
But Hatamura has also become an advocate for clearly distinguishing between complexity and non-complexity and what can — and, even more important, cannot — be done to manage complexity. With a little effort we can foresee complex events. We have a much more difficult time imagining how our strategy for the complex must differ from our strategy for the merely complicated or novel or known.
The Japanese for complexity (see above) includes kanji a classically minded literalist might read as “a surprising recurrence of miscellaneous elephants.” If you can imagine how you would manage that, you are on your way to being able to manage the cascade of a complex event.
The final report is expected in June.