What went wrong? There will be a series of reasons, both proximate (sticking valves, operator errors, slow response) and ultimate (siting, design, and such).
What can we say at this point about the failure to design for an earthquake-tsunami one-two punch?
1. The safety systems detected the earthquake and, as designed or subsequently modified, shut down the chain reaction to below criticality.
2. The earthquake might have damaged some systems, e.g., might have drained some of the 10m-deep pool at Number 4 holding removed fuel rods.
3. The tsunami overtopped the sea walls built to protect the installation. That constitutes an obvious siting and design insufficiency, given Fukushima was near a subduction zone that generates tsunamis. These were errors made forty years ago by Tokyo Electric Power and the Japanese regulators.
4. Because basement electrical control rooms were flooded by the tsunami, backup power failed and batteries were only good for four hours. Some sensor data and control functions were lost. Another obvious design insufficiency, easily avoided by locating this room on a higher floor.
5. The Fukushima boiling-water reactors were designed by G.E. in the 1960s to be cheaper than the dominant design, the pressurized water reactor. But its slimmed-down containment was heavily criticized in the early 1970s for safety reasons that are likely to be relevant to the Fukushima disasters.
6. About one-fourth of the U.S. reactors in operation today are of this cheaper G.E. Mark I design. The Chernobyl reactor that melted down in 1986 was a much cheaper Soviet design; not only did it not have a containment but the reactor was surrounded by very flammable graphite, what caused the building to burn for a month while lofting radioactive particles up to 30,000 feet to create fallout over a far wider area than anything likely from the multiple Daiichi disasters.
7. As Matt Wald notes, “One simple improvement, in use now in most plants, is to keep some spent fuel in “dry casks” — steel cylinders filled with inert gas, sitting in small concrete silos. These have no moving parts and are unlikely to be bothered by earthquakes or tsunamis.” It sounds as if Tokyo Electric Power and the Japanese regulators failed to update their fuel-handling.
8. Most of the four hundred reactors operating in the world today are of the more robust pressurized water design and most are not threatened by tsunamis, so many of the lessons of Fukushima may not prove relevant to them.
9. Most of the reactors currently being constructed have forty additional years of operating experience and safety research to guide them. (The G.E. Mark I represented only ten years of commercial nuclear power experience.)
While the use of coal for electricity generation appears cheap, that is only because we take so little account of its hidden costs (climate change), pollution (coal-fired plants put more uranium into the air each year than all of the uranium consumed in nuclear reactors), and miner deaths. In China alone, there are an estimated 5,000 miner deaths each year. For hydroelectric, there are dam failures that wipe out downstream towns. Per megawatt generated, the hydro fatality rate around the world is a hundred times higher than for nuclear electricity.
For commercial nuclear power, the death toll was about 50 in the first fifty years, all from Chernobyl. One death per year, on average. While Daiichi has already added perhaps a dozen to that total, it is clear that nuclear is still, by far, our safest method of generating electricity.
The nuclear contamination downwind of a damaged reactor must be compared to those of other industrial plants, say those for agricultural chemicals.
The 1984 Bhopal disaster was the world's worst industrial catastrophe with perhaps 15,000 deaths and a half-million injuries from the leak of methyl isocyanate gas in to the air, which went undetected for a week.
Such facilities get little of the design and regulatory oversight that characterizes the nuclear power industry. Nuclear power is not inherently safer so much as it has been made mostly safe by foresight.
Were not nuclear power associated in peoples’ minds with something far scarier, nuclear bombs, we would have treated it about like we treat a pesticide plant. But I wouldn’t downgrade the nuclear power precautions; instead I would upgrade the safety requirements and disaster plans for many manufacturing operations that can generate fallout from a fire or leak.
---
My qualifications in this area, besides a Ph.D. in physiology and biophysics atop a physics-engineering background, come largely from a three-day course on reactor safety (and past accidents) given by the NRC. I reviewed the relative safety of various power sources in Chapter 19 of my 2008 book, GLOBAL FEVER: How to Treat Climate Change (University of Chicago Press). I am a professor at the University of Washington School of Medicine in Seattle.