Tornado Sirens Malfunctioning

[Kevin Rider]

Is it just me or does it seem like there is always issues with tornado sirens? I know people aren't supposed to rely on just the sirens for tornado warnings but in reality many of them do.



Article Link: http://ksn.com/2015/04/09/sedgwick-county-investigates-siren-malfunction/

 

[Jarod Carlisle]

I wondered why it was so long after the warning came out before heard the first sirens.
It's curious to see that they can't test the area - specific warning without an actual warning being in place. I would think that would be something they could easily.

 

[rdale]

Considering there are tens of thousands of sirens in the nation, it wouldn't be surprising to have one or two malfunction now and then. I'm curious who the vendor is but can't quite tell from the video.

EDIT: Found it. ATI Systems. I've been doing sirens for years and have never heard of them. Guess we know why

http://www.atisystem.com/press/news.htm

 

[Kevin Rider]

I don't doubt a few can fail every so often

Considering there are tens of thousands of sirens in the nation, it wouldn't be surprising to have one or two malfunction now and then. I'm curious who the vendor is but can't quite tell from the video.

EDIT: Found it. ATI Systems. I've been doing sirens for years and have never heard of them. Guess we know why

http://www.atisystem.com/press/news.htm

I don't doubt a few will fail every so often but it seems to happen in bulk around the Wichita area a lot. Maybe it is the company and system that you mentioned. It seems like they are constantly malfunctioning whether during severe weather or going off in the middle of a sunny blue sky day outside of the testing time.

 

[Jake Wallentine]

How does ATI not have the ability to do a "quiet" test? That seems like that would be a key feature to siren functionality? The guy in the interview acted like they wouldn't be able to really test to see what went wrong due to it being a live system. I get that but there should be some way to trigger a silent test where it gets turned on except for the actual siren itself.

 

[rdale]

I'm sure they do. What they probably don't have is the ability to quiet test with an actual polygon. My complete guess is that something in the actual TOR warning text / polygon confused their software decoder and that's why it didn't trigger.

 

[James Hilger]

Has happened many times as well as the phone alerts here in NW Ar.

 

[rdale]

Phone alerts are not connected to sirens at all... What sort of problems are you having? Are you referring to some app you have?

 

[Randy Jennings]

Sedgwick County's systems appears to have a feature that many (likely most) systems don't - automatic activation from a NWS warning. Every system that I have personal knowledge of in the DFW area requires a "man-in-the-loop" to active sirens manually. That is an interesting feature, as their have been cases of humans failing to press the button when they should. Having said that, this is likely a very hard feature to implement, as the NWS warning system wasn't designed to activate sirens. The only explanation I see of how this works is in ATI's datasheet about their MassAlert product where it says "web-based National Weather Service interface option for severe-weather activation." I'm guessing the software downloads the CAP/ATOM feed (info at http://products.weather.gov/PDD/cap1.1final.pdf) and looks for the <polygon> tag, or they have an iNWS alert and parse the LAT/LON at the end of the warning. As rdale noted above, the software could have misparsed the warning. Another possible theory is that the county emergency management officials got the warning from another source (NOAA weather radio or NWSChat) first and decided to activate them manually before the software received the CAP/ATOM feed or iNWS alert. The website for CAP/ATOM (https://alerts.weather.gov/) says it is updated every 2-3 minutes, so there could have been a delay.

As for testing the activation based on a NWS warning, that would be very hard to do. The only test message NWS sends out is the Required Weekly/Monthly Test and that message doesn't contain a polygon, just FIPS codes (http://www.nws.noaa.gov/directives/sym/pd01017012curr.pdf).

This is an interesting feature, and I commend Sedgwick county for trying it. However, I don't see this as a popular feature in the EM community. Here in DFW, sirens are city based and every city has its own criteria and process for siren activation. I don't think most EMs would be willing to hand it over to the computers.

 

[rdale]

Today they don't because many EMs are "old school" (i.e. fire service: "200 years of tradition unimpeded by technological progress") but that will change over time. Hopefully We're still working hard on standardizing criteria, look for more from the AMS, IAEM, and possibly FEMA & NWS this year yet.

 

[James Hilger]

Phone alerts are not connected to sirens at all... What sort of problems are you having? Are you referring to some app you have?

Weathercall and some city's do it.

 

[rdale]

What I'm saying is that they aren't "connected". If sirens fail it doesn't affect Weathercall and vice versa. What sort of problems have you had with Weathercall? I know many people there and can pass that through to the higher-ups direct.

 

[James Hilger]

What I'm saying is that they aren't "connected". If sirens fail it doesn't affect Weathercall and vice versa. What sort of problems have you had with Weathercall? I know many people there and can pass that through to the higher-ups direct.

I obviously have no need for it. Just seen it reported quite abit about all the failures. Seems to be every year one of the options fail to go through or goes out of service right before it's needed around here. Kinda crazy in today's age. It is made by humans though.

 

[Matthew Gaylor]

If I remember correctly, their old system used to sound sirens throughout the entire county during a tornado warning, even if it was just the very far edge of the county. In 2012, when they did this switch, they finally were going to just do it for the cities in the warning. Unfortunately, it looks like that didn't work during this storm.

 

[Mike Johnston]

Yes, the sirens did sound for the whole county in the old days. IIRC, weren't the polygon warnings started back in 2006? (Or maybe it was just an experiment for parts of the central region.) At any rate, if a place like Sedgwick County, Kansas doesn't have its siren system integrated with the polygon warnings yet, it makes you wonder how many other jurisdictions don't either.

 

[rdale]

It is integrated - they just chose a bad vendor

 

[Matthew Gaylor]

Yes, the sirens did sound for the whole county in the old days. IIRC, weren't the polygon warnings started back in 2006? (Or maybe it was just an experiment for parts of the central region.) At any rate, if a place like Sedgwick County, Kansas doesn't have its siren system integrated with the polygon warnings yet, it makes you wonder how many other jurisdictions don't either.

They recently integrated Sedgwick county (2012), which is why they said this is the 1st time they've actually tested the new system because there hasn't been a tornado warning for any part of Sedgwick county until then. And yes, they chose a bad vendor.

 

[Matthew Gaylor]

Actually, now that I think of it, they switched in 2012 to this "new system." On May 19 2013, Sedgwick county was under a tornado warning, they actually got placed under 2, the 1st one included western half of Wichita, the 2nd one turned into a tornado emergency for the city of Wichita. So my question would be, did the new siren system work then and if it didn't why didn't they fix the problem then? In reality, it's like Dale said, they just chose a bad vendor.

 

[Mike Johnston]

Yes, I remember that May 2013 storm and the sirens did activate, at least for the tornado emergency. I was on the phone with my sister who lives in east Wichita when that happened.

 

[Lee Smith-Whaley]

Considering there are tens of thousands of sirens in the nation, it wouldn't be surprising to have one or two malfunction now and then. I'm curious who the vendor is but can't quite tell from the video.

EDIT: Found it. ATI Systems. I've been doing sirens for years and have never heard of them. Guess we know why

http://www.atisystem.com/press/news.htm

I know I'm a little late, but I can say from personal experience that ATI can do a silent test on individual sectors (normally predefined), and that at least what my base uses is junk.

 

[Eian Greene]

I was following similar threads over at The Siren Board not too long ago, and I came across at least 2 cities that had/have problems with ATI's products.

Nashville, TN had 72 ATI sirens installed back in 2002. Metro Nashville citizens complained to Metro OEM often about the 1,000Hz 'censor bleep' sound they used, and how they couldn't hear it very well, even outdoors. Metro Council moved in late 2012 to approve a full replacement of the system with different gear, under direction of Mayor Karl Dean.
By April of 2013, the city had a contractor install 90+ Federal Signal Eclipse and 508 sirens, with the last ATI siren coming down in mid April that year. The bulk of the system in Nashville is Federal Signal's Eclipse sirens. In less dense parts of Davidson County, they've installed 508s. The Eclipse is rated at 115dB/100ft, making it ideal for urban installation. The model 508 is higher powered at 128dB/100ft. Both produce sound at 500Hz, a much lower and far-reaching tone than the old ATI HPSS-16 and HPSS-32 units they used to have.

Another case study to look at is in San Fransisco, CA. They have problems with the units blowing speakers because the tone used, a recording of one of their old Federal STL-10s, is actually too low a frequency for them to handle, so they readily blow the drivers. The lowest frequency the STL-10 produces, is around 170-180Hz. This is well below the lower limit of the compression drivers used.

That said, I wouldn't be surprised if in a few years if Sedgewick County and others with ATI products begin considering other vendors for outdoor sirens.