IMHO it's a bit of a judgement call on this one Messy. Strictly speaking 7273-4 does not apply as I see it because the locks are not connected to the fire alarm. Whether this is appropriate or not will be determined by the risk assessment but there was some guidance on this in Graham Shiels guide, now obsolete but never fully replaced. The fact is that if 7273-4 did apply the recommendation would be for a weekly test on the door release mechanisms throughout but the box would be treated as an alarm call point, once per year for all devices unless it was necessary to operate some more often due to the set up of alarm zones.
But in this case there will be no weekly test of the release mechanism as there is no link to the fire alarm. To me this, depending on the control set up, represents a higher risk of individual failure but may pose less risk of general failure. If there is centralised control then you may consider treating it as a fire alarm and testing all zones weekly. If it is individual devices operating independently the risk of failure is higher but the consequences may be lower. But this is my greatest concern- is it appropriate to the risk? Who uses it, what will they expect to happen, will they know they have to sound the alarm as well as operating the green BGU? Do the Public use the doors or trained staff? If you cannot test all door releases centrally then more frequent tests of each break unit may be justified but I don't think you will find any specific guidance on frequency.
Just my opinion, probably will be corrected by others, but there appears to be no specific guidance on this.
Thanks for your commenst so far
Kurnal - There are around 60 individual break glass override units (nobody knows the true figure yet) within this building. Just a few are at final exits, however the majority are distributed throughout the premises.
The resp person involved is a low fire risk scientific & medical research establishment. There are high value (monetary and commercial sensitivity) goods, so the premises have various security layers throughout the premises, some staff having restricted access to parts.
There is no central control as these are stand alone units - I have been having a bit of a headache about this, but due to the nature of the business, and the fact that there are absolutely no public & visitors will be accompanied - I believe there is a justification not to link the units to the AFD
It would be very difficult and perhaps disproportionate to test all 60(ish) every week, as assuming a 2 minute test per device (including completing paperwork) and a 3 minute walk between units - that's 5 hours a week - and more if the results need entering on a database! So I am considering three options
A) Testing say 5 per week (at the same time as the fire alarm) which means they all get tested quarterly.
B) Assess the location of each unit and provide it with a high or low risk rating based on the numbers of staff expected to use that route, and the expected fire growth rate of the area they would be evacuating from. The higher risk units would attract a more frequent test than the lower risk units.
C) Test 15 or so per week (on the same day - so 1.5hrs work a week which still sounds a long time to me), so all units are tested monthly
Any comments???
