30 Second Earthquake Warnings

Bill Kendrick writes "A new network of seismic stations may help give as much as a 30 second warning before a major earthquake, giving time to shut down gas lines, stop public transit, etc. Yahoo! News has the story." There are lots of qualifiers in here ("as much as," "some earthquakes") but any warning is probably better than none.

Comments

Now all we need

[anthony_dipierro]

Is a way to buy earthquake insurance online in 30 seconds.

Lawsuit Potential

[Unknown Poltroon]

What happens when these things are shut off and the quake dosent happen? Or when they dont work? You KNOW some idiot will sue eiter way. Also, what about false alarms? How long before people start ignoring this? Hey, dont get me wrong, i think its fantastic and should be put into use immediately, but these days......

California-Centric Comment...

[Bravo_Two_Zero]

Hey, I live on the other coast, so I'm not likely to be personally impacted, but...

So many fault lines run beneath densely-populated areas, the expensive warning system may be of dubious quality to Californiana. I thought the Northridge quake was pretty much centered under a moderately-populated area? But, for the heavily monitored areas on the other side of the Pacific Rim, it could be a real boon, particularly to quakes that cause tsunamis.

Muhahaha. That is all.

[realgone]

Turn off gas mains? Shut down public trasport? Bah! Let's put this power to some real use -- beeper notification, for instance.

Potential Employer: Well, we've reviewed your resume and references and, if we may be frank, you're entirely unqualified for this position.

*sound of beeper going off*

Me: Did I mention I can destroy you all with the power of my mind?

Potential Employer: Okay, that's just about enough of that. Securit--

*cue 7.1 earthquake*

Me: $400k sound good to start?

Statistics. Security.

[Nyarly]

Ultimately, though, for this to be useful, it would have to be automated from end-to-end. Thirty seconds isn't enough to double check the system, and introducing humans into the mix just adds error. But the automated system has problems:

First, "near 100 percent accuracy" isn't 100% accuracy. There's no information here about false positives, but my guess is that, like many testing and alarm problems, tuning out one way, opens up the other side disproportionately. So you'll always have issues where the gas mains don't cut off in time, but people'll understand. But when the gas cuts out from too many false positives, people'll call for the whole thing to be turned off.

The other side of this is: I hope it's not supposed to be wired to the Internet. How's that for a script kiddie prank. "1 4m 2 733t 4 j00! 1 0wn3d CA! h4h4h4h4h!" Punks. Still, even with a large private network, being able to fake an earthquake isn't a small thing.

Tsunami Warning

[ELCarlsson]

Scientists take a few minutes to make sure that there is actually the threat of a tsunami before they issue an alert. They don't want too many false alarms in which case people would start to ignore them. But if there is only 30 seconds warning before an earthquake may happen that doesn't give them any time to verify the occurance. Too many false alarms and people will start to ignore it.

Re:I can see it now...

[Jonny 290]

30 seconds of pre-quake hysteria

I'll take that over three years of pre-Daikatana hysteria any day.

Any warning better than none?

[junk95]

I agree that this discovery is not something to
laugh at. Short-term prediction of earthquakes
is the ultimate goal of seismologists everywhere and the benefits from accurate short-term prediction are huge. Any progress made at this
field is welcomed.

But 30 seconds? And without even 100% accuracy?
It takes more time than this to leave your house. It even takes more time than this to leave the 90th floor of a sky-scraper (although sky scrapers are generally earthquake-proof).
It even takes considerably more time to issue an alert in a large city.

Sorry guys, but 30 seconds would offer nothing more than panic. More people would be killed by trying to evacuate a building with say, 500 employees (within 30 secs) than any given earthquake could dream off (not to mention heart attacks and such). And false alarms would get to everyone's nerves after the first few months.

So, although the discovery is interesting, it is of little to no practical use.

Re:Any warning better than none?

[Cy Guy]

Actually, if you think of it as 30 seconds of additional warning, I think it would be useful.

Most of the damage caused by an earthquake isn't caused instaneously in the first second of the event. Rather as structures are shaken they eventually crumble, break loose, etc. Having an additional 30 seconds would allow more time for people to get to door jams or outside or just away from book shelves etc.

And maybe there wouldn't be enough time shut gas lines off prior to the quake, but a button could be pushed (or signal sent automaticlly by the early warning system) that would automatically shut off the gas in 2 minutes, unless the instruction is rescinded. While that method might not eliminate all gas explosions caused by the quake, at least the resulting fires would not be fed by additional gas. Also, fireman would have an additional 30 seconds to suit up to fight the fires and be that much more ready to respond when/if they find out a fire/explosion has resulted from the quake.

For people in skyscrapers, the 30 seconds would at least be sufficient for the elevator to stop at the next floor, so it doesn't get caught in between floors. And it would increase the likelihood that subways could either get to the next station or not a leave they are already at.

Re:Any warning better than none?

[leucadiadude]

Most useful I would think would be to give firemen the time to move their highly useful emergency apparatus out of a building that might fall down, or at least open big heavy doors that might get jammed. Several fire stations in San Fran had their outer doors get jammed in the last sizeable quake, rendering that euipment useless just when it is most needed.

What could you do with 30 seconds?

[Spamalamadingdong]

It might take 30 seconds to leave your house, but it takes a lot less than that to go stand in a doorway. A list off the top of my head:

• Electronic-ignition gas appliances could shut themselves off until the "all clear" sounded.
• Trains, buses and other public transit could come to a halt before the rails or roadway ruptured.
• Traffic signals could go 4-way red for half a minute.
• Electrical systems could prepare for shutdown to prevent fires and other damage from short circuits (kill the circuit breakers at the first heavy shaking), and hospitals and the like could automatically start their emergency generators to prepare for cutover.

None of these things would have large costs from false alarms, but might prevent a lot of death and destruction if a real tremblor hit. Just being able to shut down the electrical grid would prevent transformers from being damaged by shorts and make it much easier to bring the system back up, and killing most open flames from water heaters and furnaces would prevent a lot of fires. This has real potential.

Re:What could you do with 30 seconds?

[leucadiadude]

"Electrical systems could prepare for shutdown to prevent fires and other damage from short circuits (kill the circuit breakers at the first heavy shaking), and hospitals and the like could automatically start their emergency generators to prepare for cutover."

Since I work in the power generation industry (in CA actually) I believe the first part of your comment above is not really applicable. Most of the damage to the grid in the past has come from the ground shaking uprooting and breaking stuff such as foundation bolts hold heavy stuff in place, or ceramic insulators on transformers or large power circuit breakers and transmission towers. Then this damage can cause an electrical short, which is then almost instantaneously isolated. The systems that are already in place can isolate shorts, grounds and opens in fractions of a second, on the order of a few cycles of a waveform of ac. I.e., the existing fault isolation systems and relays are already so fast a thirty second warning would be superfluous.

The second part of the above comment is useful however, because most blackstart emergency diesel or combustion turbine generators can take anywhere from 10 seconds to 2 minutes to get up to to speed and voltage and assume load. With a thirty second headstart you could get them up to speed and have a near seamless transfer of load when and if the shaking knocks out your circuit on the main grid.

Re:What could you do with 30 seconds?

[Amazing Quantum Man]

In CA, all new construction and any sale of an existing home requires that the property have an automatic cutoff valve attached to the gas line.

Mine cost about $200. We'll see how good it is when the next big one hits.

Not very useful

[Hoi Polloi]

30 seconds? Just enough time to kiss my ass goodbye. Now if they install sirens they can make sure you are awake when you die.

The equipment might not be the softest point

[Spamalamadingdong]

Most of the damage to the grid in the past has come from the ground shaking uprooting and breaking stuff such as foundation bolts hold heavy stuff in place, or ceramic insulators on transformers or large power circuit breakers and transmission towers. Then this damage can cause an electrical short, which is then almost instantaneously isolated. The systems that are already in place can isolate shorts, grounds and opens in fractions of a second, on the order of a few cycles of a waveform of ac. I.e., the existing fault isolation systems and relays are already so fast a thirty second warning would be superfluous.

It depends what you are trying to do. If you are trying to prevent shorts between lines from setting off brush fires (making demands on a fire-fighting system which probably has its hands full already), killing the power before the shorts occur might still be a good idea; it would be saving things in general, rather than the grid in particular.

Re:The equipment might not be the softest point

[leucadiadude]

Your reply about killing the power before shorts occur doesn't fully comprehend the magnitude of the complexity of operating a large grid electrical distribution system. Only as a last resort do you drop circuits. Things get too hard to control, balance load etc when you start leaving the normal load practices. No power company is going to advocate dumping circuits in case of a short. Such a strategy will be worse than the cure. And would likely result in the whole grid possibly going down, just when you need to keep as much load up as possible. It is complicated and time consuming to restore a large grid if it goes down, not a trivial thing. I hear your comment about the fire prevention, but there are actually more important considerations system wide militating against a pre-emptive dumping of load on an earthquake warning.

Re:The equipment might not be the softest point

[Spamalamadingdong]

I wasn't proposing to dump the load on warning (please read what I wrote). I proposed to prepare to dump load on warning, but not dump it until detection of a tremblor which was strong enough to actually produce damage. The warning system would not be superfluous; it would be used to vet inputs from local detectors so that a passing gravel hauler hitting a pothole doesn't cause the transmisson line going over the road to shut down.