One reason the venting system at the plant, which was built by General Electric, did not work is that it relied on the same sources of electricity as the rest of the plant: backup generators that were in basements at the plant and vulnerable to tsunamis. But the earthquake may also have damaged the valves that are part of the venting system, preventing them from working even when operators tried to manually open them, Tokyo Electric officials said.
I am quite amazed that this isn't receiving more 'nuclear panic' notice. I guess people don't realize the concept of seismic fragility, since you must combine it with seismic hazard to get seismic risk. As well, all of this is expressed in terms of probability, and nobody does that well.
Release and relief valves have always had problems with earthquakes. The same with electrical systems. Short pipes between heavy masses get sheared off. All of this happens at 10 cm/s, and we can expect all this to happen with a US BWR at relatively high probabilities. The recent M6 Saguenay earthquake buggered up some valves on a plant, and could have caused deaths if the wind had been perverse.
So, you can have ridiculously fragile nuclear plants if you believe there is virtually no chance of an earthquake somewhere in the eastern US. So little scientific work has been done by the plants in the last 30 years, it becomes easy to make this statement, if you ignore the uncertainties, as these guys are prone to do.