Longest time a siren has been stuck on?

[SirenSteveUK]

According to a UK police officer, around 1982 the air raid siren on top of Marble Arch in London went off unexpectedly - some fault at the actual unit as the manual override at the police station failed to silence it. The 'damned thing was screaming for about an hour' before a man from the City Engineers could arrive. Apparently not a single person queried the noise by ringing the police.

On the subject of accidental activations, in 1986 a Carrier Control Unit malfunction in Edinburgh, Scotland, caused the nuclear attack sirens to sound in the city at 07:30 in the morning. This coincided with a visit by the Queen and the then Prime Minister Margaret Thatcher to watch the Commonwealth Games. Again little panic was caused as most people switched on their radios to check there was no real emergency.

[Robert Gift]

Was the 1986 "malfunction" an intentional joke by someone?

[SirenMadness]

If the activation was due to a malfunction, I would not call it intentional. I have, however, seen an "accidental" activation from our E57 sometime after it had done a fire-signal, though I am sceptical as to it being just a growl, because the building never tested the growl signal.

[AllSafe]

Some sirens, such as the Allertor, do have continuously rated motors. I wonder if that is so with the T-135.

[SirenSteveUK]

British Telecom (BT) was blamed for the sirens sounding in Edinburgh in 1986, although BT would not say at the time whether their company accepted responsibility for the mistake. 'The Home Office ruled out computer error as a reason for the mistake, and seems convinced that human error or sabotage were not involved either. This is despite the fact that no similar mistakes have been recorded in the past 12 years, and that the incident happened at the height of a controversial visit to Scotland by the Prime Minister'.

A spokeswoman at the Home Office, which had the responsibility for civil defence in the UK at the time, said that 'BT was checking a carrier control unit in Edinburgh. This is believed to have malfunctioned causing the alarm to go off'.

[Robert Gift]

A spokeswoman at the Home Office said that 'BT was checking a carrier control unit in Edinburgh. This is believed to have malfunctioned causing the alarm to go off'.

Ha! Sounds like a clever prank.

All my similar "accidents" are made to appear as though they have another cause.

[SirenMadness]

What sometimes makes me ponder is the fact that some states and counties only issue an actual alert for one minute, to three at around the most, when the sirens are capable of an average fifteen-minute duty. If the sirens are known to be unserviceable after the danger, there is no reason for them not using their sirens to the longest time frame possible.

[SirenSteveUK]

On more than one occasion in the UK, with many sirens being on top of police stations, etc., and with localised control equipment with every siren (often within the building itself), some police officer would 'set off' a single siren by mistake (not knowing what the equipment was, etc.), and if there were any phone calls from the public - they would claim they were 'just testing'.

[Robert Gift]

Good point, Peter.

I'd give a warning at least 5 minutes, unless an actual tornado is observed and is approaching the warned area.
Then, I would keep the siren sounding for up to it's duty cycle unless the tornado dissipated.

A lot of warnings result in no tornado, or one not even close to the warned area.

[SirenMadness]

You know, they are right in the theory of a false warning wasting their sirens, though the debate really depends on what is the most efficient for you.