Green break glass over rides for magnetic locks

[kurnal]

We always recommend that every magnetic and electromechanical lock on an exit door should comply with the relevant BS, be interfaced to the fire alarm system and have a green break glass double pole over ride on the side of the door required for exit purposes.  The double pole switch is there because "in the event of an earth fault, the fire alarm interface and normal exit button may fail to release the lock".

Now a question for the technical guys- Please Gentlemen from your knowledge how likely is such an earth fault  to happen, how many simultaneous faults and errors would be needed and is the risk higher with a electromechanical lock rather than an electromagnetic lock? Because it seems to my simple mind to be rather over the top, especially when you see most green break glass units wired in telephone style pvc multicore cable in a wooden door frame with not an electrical earth in sight.

[Big_Fella]

Indeed....

We are finding more and more of these systems now not calling for the green break glass for this very reason, as it is very unlikely that various faults will occur all at the same time, and given it's wired in telephone cable etc and not monitored, one seems to think why install it.

Although, I suppose it's not impossible and is just a further method of release should such instance occur.

[Galeon]

Is the punter likely to hit a green manual release anyway ? I am not so sure ,even if trained to do so when it all kicks off , it would be interesting if anyone has an statistics of green or in that case red manual call points being operated in actual fire scenarios.

[David Rooney]

I think the "double pole" bit originated from the LFEDA ... there old advice sheet (was it 64 or 65 ??) stipulated it. I don't know where they got their technincal advice from (!?) but the odds of it all happening and not releasing a door are probably the same as Bruce Willis landing on an earthbound asteroid and saving mankind as we know it..... but only if he's done his, method statement, working at height RA, Coshh Statement and confined space training first ofcourse.....!

[Wiz]

We always recommend that every magnetic and electromechanical lock on an exit door should comply with the relevant BS, be interfaced to the fire alarm system and have a green break glass double pole over ride on the side of the door required for exit purposes.  The double pole switch is there because "in the event of an earth fault, the fire alarm interface and normal exit button may fail to release the lock".

Now a question for the technical guys- Please Gentlemen from your knowledge how likely is such an earth fault  to happen, how many simultaneous faults and errors would be needed and is the risk higher with a electromechanical lock rather than an electromagnetic lock? Because it seems to my simple mind to be rather over the top, especially when you see most green break glass units wired in telephone style pvc multicore cable in a wooden door frame with not an electrical earth in sight.

The reason for the double-pole recommendation was based on a fault that occured some years ago when the insulation of both connecting wires to a single pole EDR switch both wore out and shorted on the at the cableway drilled into a metal frame. Therefore when the normally-closed switch of the EDR opened on operation, the metal frame provided a short as if the switch was still made.

With the double-pole EDR switch, each pole must take a different voltage potential i.e +ve and -ve so that if a similar short to the above was to occur the power supply output fuse should blow and the lock therefore fail-safe open.

The risk of such a short is minimal, can be lessened with proper installation practices and the type of cable used in a 'fail-safe' system is fairly irrelevant because it could be argued that less robust cables are just as good as any other type of cable in a 'fail-safe' system.

I have never been impressed with the strong arguments espoused by some that every part of a fail-safe door release system must be engineered to the nth degree because if you show me a circuit that includes a myriad number of elements that fail-safe since they rely on something being 'turned off' and I will show you that every such circuit always relies on something being 'turned on' at the beginning of the operation sequence. Therefore it can never be truly fail-safe.

I have previously asked on this forum if anyone has any figures that show how many lives have been lost due to the failure of a simple fail-safe circuit where a more complicated fail-safe circuit could have prevented such a loss, and no-one came up with anything.

[Dinnertime Dave]

The fire station where I work has a key pad entry with a push button on the exit side. The system fails numerous times each year in the locked position. Fortunately it has a green break glass. Oh and we also leave the key in the inside lock.

[colin todd]

By definition, risk is the combination of liklihood and consequences. While earth faults are of low liklihood, it is slightly embarrassing if, having locked an exit door, it does not open when required to do so in the event of fire, all for the price of a double pole switch instead of single pole. That's why BS 7273-4 calls for double pole. (The LFEPA had to get something right sometime!)

[kurnal]

Yes Colin but in addition sometimes  for security reasons clients would prefer us to omit the green break glass emergency release altogether.

In such a case the possibility of an earth fault / card failure in the panel/ cable breakdown/ availability of staff to operate the keypad/ resilience of the keypad etc  all need to be considered. And probably 101 other things besides.

[messy]

(The LFEPA had to get something right sometime!)

Praise for the LFB???

Colin, are you feeling well?

[colin todd]

Omitting the break glass is a different issue from making sure that if one is provided it works. Omission is a variation that BS 7273-4 recognizes might be necessary in certain high security situations. However, it was not envisaged that these would include someone stealing a bottle of Talisker from the local supermarket.

[Graeme]

someone stealing a bottle of Talisker from the local supermarket.

a thief with taste.  Suprised it was not Buckfast

[Wiz]

By definition, risk is the combination of liklihood and consequences. While earth faults are of low liklihood, it is slightly embarrassing if, having locked an exit door, it does not open when required to do so in the event of fire, all for the price of a double pole switch instead of single pole. That's why BS 7273-4 calls for double pole. (The LFEPA had to get something right sometime!)

The risk is people could be trapped. The consequences are they could die by being trapped. I would suggest the liklihood of a 'short' affecting the operation of the switch is low (I've never seen it in over 25 years). So how much of a risk is it?

The circumstances of someone being trapped in a 'typical' set up would need to involve two/three failures;- failure of the fire alarm system and failure of the normal exit method (if it includes such) and failure of the green BGU. After all of this, there would have to be the circumstances of the person not finding another escape route or of being rescued before the 'worst case scenario' happened.

I've previously asked a number of times if anyone has any knowledge of any incident when someone was trapped in the circumstances of a short to a single pole EDR switch and was injured because of it. I've not had an answer.

Despite all of the above, I agree that using just a double-pole instead of single-pole switch shouldn't be too much of a problem for anyone to include in their design, although I am stunned that a double-pole switch from KAC costs more than twice the price of a single pole switch!

What concerns me is how far the 'belt, braces and superglue' method will be extended, and will it affect every facet of a system until we get a system that costs twice as much money for reducing the risk factor by only 1%? Nothing is a 100% fail-safe. So why is 99% so much better than 98%, if it costs twice the cost?

The underlying principle of fail-safe is, that as much as possible 'turns off' instead of 'turns on'. However every circuit, no matter how many' turn-off' stages it has, depends on at least one stage 'turning on'

I've always said 'you show me your fail-safe system' and I'll show you where it doesn't 'fail-safe'!

[Chris Houston]

The circumstances of someone being trapped in a 'typical' set up would need to involve two/three failures;- failure of the fire alarm system and failure of the normal exit method (if it includes such) and failure of the green BGU. After all of this, there would have to be the circumstances of the person not finding another escape route or of being rescued before the 'worst case scenario' happened.

On the first one "failure of the fire alarm system":

People will tend to detect a fire before and alarm system and it is perfectly logical to assume they will try to evacuate before the system has activated.  

If there is less than perfect detection or less than full detection it might react slowly or not at all.  I've not yet seen a system that incorporates detection in every single area of a building and has never malfunctioned once.  

Someone has already talked about the green BGU at their fire station which fails several times a year.  Even so, I don't think inviting anecdotes on a discussion forum is a scientific method and just because no one has personal experience of it means it does not happen.  Otherwise I could hypothesise that noone has ever won the national lottery jackpot as noone on here has any personal experience of it.

So the fire alarm system would not be expected to detect a fire before a human.  The BGU might then fail.  That is the scenario, it only needs 1 failure.  The "not getting rescused" and the "not finding another way out" is the consequences.  The risk of not being able to use the exit only needs the failure of the BGU and this is not a failure of multiple independent systems.

[Wiz]

The circumstances of someone being trapped in a 'typical' set up would need to involve two/three failures;- failure of the fire alarm system and failure of the normal exit method (if it includes such) and failure of the green BGU. After all of this, there would have to be the circumstances of the person not finding another escape route or of being rescued before the 'worst case scenario' happened.

......Someone has already talked about the green BGU at their fire station which fails several times a year.  Even so, I don't think inviting anecdotes on a discussion forum is a scientific method and just because no one has personal experience of it means it does not happen.  Otherwise I could hypothesise that noone has ever won the national lottery jackpot as noone on here has any personal experience of it........

.

Chris your quote' green BGU at their fire station which fails several times a year' is misquoted! it is the access system that was quoted as failing and it is thanks to the green BGU that Dinnertime Dave can get out at dinnertime.

Following your claim 'just because no-one has personal experience of it means it does not happen' I think you meant to say 'just because no-one has personal experience of it does not mean it does not happen'. I then suggest that all fire alarm system cie and power supplies are housed in locked fire and bomb resistant rooms just in case they are damaged in a fire or terrorist attack. Just because no-one has personal experience of it does not mean it can't happen! S*d the cost and aggravation, lets cover every possible scenario!

[Chris Houston]

We need to get the balance right.  But as I think Colin T has pointed out - if someone is going to lock a fire exit then it is reasonable to ask for an override system that does not have 1 point of failure.

The law/guidance does not ask for much - just for fire exits that are easy to open.  If you think that double pole green BGUs are too much cost, don't lock the fire exit shut.