4 ms·
They should just license this tech from Japan, who has already solved this problem and put it in production since 2007.
by bschwindHN 4y ago
They should just license this tech from Japan, who has already solved this problem and put it in production since 2007.
- sholladay 4y agoSo it was operating for four years already when the Tōhoku earthquake and tsunami hit Fukushima? That adds some color to the story that I haven’t seen reported. Edit: Yep, indeed. Wikipedia says it underestimated the quake. https://en.wikipedia.org/wiki/Earthquake_Early_Warning_(Japan) https://en.wikipedia.org/wiki/Earthquake_Early_Warning_(Japa...
- Kenji 4y agoIt warns a few seconds ahead of time. The tsunami appears a few minutes later. What are you gonna do, build a concrete tsunami wall in a couple of minutes?
- sholladay 4y agoThe tsunami hit an hour later. That seems like plenty of time to get fire trucks, backup generators, and other emergency equipment prepped. But of course hindsight is 20/20. The predictions were off and the emergency preparedness wasn’t what it should be. I still think it’s interesting that they did have plenty of advanced warning, albeit underestimated. Most of the reporting I’ve seen made it sound like what you said, that it happened suddenly.
- mistrial9 4y agoI watched the news streams at 2am US Pacific time while that Tsunami hit.. the earthquake was well in advance of the power plant event, and yes they knew. Also the Fk power plant had a tidal wall built around it already, but fatefully not large enough.
- sholladay 4y agoWith the EEW having been around for a while at that point, it seems obvious that there should have been an automated system to begin a graceful shutdown of the reactor when a nearby quake was detected. I guess that may not have helped in this case, given that the EEW underestimated the quake. Although, hopefully such a system would have a sensitivity threshold that accounts for that. Edit: The reactors did shut down automatically at some point, although it’s not clear to me if it happened immediately as a result of the EEW or after damage was detected. At any rate, lack of electricity for cooling was the real problem.
- ilyt 4y ago>With the EEW having been around for a while at that point, it seems obvious that there should have been an automated system to begin a graceful shutdown of the reactor when a nearby quake was detected. Uh, just no. First, it was most powerful registered earthquake in Japan which means that reactor survived every previous one just fine Second, it did shut down when it detected earthquake. The disaster didn't happen because it didn't start to shut down The failure was because emergency power supply needed to cool the reactor during shutdown also failed due to tsunami. Like, fuck, at least read wikipedia page before spewing bollocks
- markoman 4y agoThat's correct about Japan already having this technology. It was quite necessary given their use of some of the highest-speed trains in the world. These trains will stop immediately upon getting the signal from the early warning system. Japanese researchers found that there are 'P-waves' that precede the more damaging waves, which they used to create this early warning system, made up of thousands of seismometers placed throughout the country. Chile hopes to use this same technology for their own public safety. https://en.wikipedia.org/wiki/Earthquake_Early_Warning_(Japan) https://en.wikipedia.org/wiki/Earthquake_Early_Warning_(Japa...
- heynowheynow 4y agoWe've know about P-waves for at least 60+ years. The "early warning" ability is dependent on the speed of light exceeding the propagation delay of physical motion. In case anyone were unclear, there is no magic rock whispering AI that can predict an earthquake. Corralitos "anomaly" and characteristic foreshocks are wishful thinking. There are only alarms for when they are happening.