Variation from BS5839 - baffled!

[kurnal]

An old 7 storey mill with basement was converted to 80 flats in the early  1990s. 10 flats per floor, 4 staircases.  The old timber floors remain, protected on the underside by plasterboard but compartmentation is questionable. I lifted a loose floorboard and found  a 1 metre deep void between floors at all levels and unstopped cavities going in all directions.

No design documents, certificates  or plans are available but I guess for reasons of a lack of compartmentation an L3 detection system was installed in the common areas, after a fashion. Heat detectors in flat lobbies, with sounders, smoke in the corridors. The alarm does nothing  but detect fire and raise an audible alarm everywhere- - no interfaces to ventilation etc.  To me thats not a satisfactory compensating feature but Hey ho.

But heres the rub- there isnt a single manual call point anywhere except one adjacent to the panel for testing the system. I cant fathom why a designer should stray so far from guidance - to me it just doesnt seem to make sense.

But before I recommend the installation of 32 call points can anyone see a reason why they may have done this?

[nearlythere]

An old 7 storey mill with basement was converted to 80 flats in the early  1990s. 10 flats per floor, 4 staircases.  The old timber floors remain, protected on the underside by plasterboard but compartmentation is questionable. I lifted a loose floorboard and found  a 1 metre deep void between floors at all levels and unstopped cavities going in all directions.

No design documents, certificates  or plans are available but I guess for reasons of a lack of compartmentation an L3 detection system was installed in the common areas, after a fashion. Heat detectors in flat lobbies, with sounders, smoke in the corridors. The alarm does nothing  but detect fire and raise an audible alarm everywhere- - no interfaces to ventilation etc.  To me thats not a satisfactory compensating feature but Hey ho.

But heres the rub- there isnt a single manual call point anywhere except one adjacent to the panel for testing the system. I cant fathom why a designer should stray so far from guidance - to me it just doesnt seem to make sense.

But before I recommend the installation of 32 call points can anyone see a reason why they may have done this?

The only reason I can think of is that they were not provided so the system would not be abused and the single point for user testing.

[kurnal]

I suppose you are right NT , it goes back to a bad decision to accept a common areas fire alarm and full evacuation strategy in lieu of proper compartmentation.  But if we accept that full evacuation is required should we not provide some means for people to raise the alarm before it develops to the extent of operating the heat detector in the flat or smoke detection in the corridor?

[BLEVE]

Maybe they misinterpreted the commentary at clause 20.1 relating to malicious activation? regardless, I cant see how it is possible to provide such a system unless voice alarm/microphones are provided instead of the call points?

[nearlythere]

I suppose you are right NT , it goes back to a bad decision to accept a common areas fire alarm and full evacuation strategy in lieu of proper compartmentation.  But if we accept that full evacuation is required should we not provide some means for people to raise the alarm before it develops to the extent of operating the heat detector in the flat or smoke detection in the corridor?

Certainly I think in your case yes, in theory, but it would be appropriate to take into consideration the nature of many of those who would occupy the block and the possibility of abuse of the system.
I don't think the books give much consideration to yobs, social misfits and the likes, do they?

[jokar]

Kurnal, Do they actaully evacuate the premises when the alarm goes off?  How do the FRS fight a fire in such a premises, one would assume that with all the hidden voids no officer in his right mind would commit crews to the building unless they could be assured that the risk was minimal.

[kurnal]

Yes they evacuate - or thats what the plan says. Try to is nearer the mark I guess. Theres a single security concierge on duty. Worse still is that several exits are on mag locks- and theres no button, no green box and no red box. 

They were not aware of the voids till I pointed them out. It came over as a bit of a shock.  I am considering a strategy to deal with those as far as possible- first we need to confirm their full extent by sampling a couple of floors. Plus trying to research the as built plans- but its changed hands a couple of times and have not been passed on. (I know.)

There are 2 firefighting shafts with rising mains. These have good integrity and retain the masonry of the old mill.

[nearlythere]

Yes they evacuate - or thats what the plan says. Try to is nearer the mark I guess. Theres a single security concierge on duty. Worse still is that several exits are on mag locks- and theres no button, no green box and no red box. 

They were not aware of the voids till I pointed them out. It came over as a bit of a shock.  I am considering a strategy to deal with those as far as possible- first we need to confirm their full extent by sampling a couple of floors. Plus trying to research the as built plans- but its changed hands a couple of times and have not been passed on. (I know.)

There are 2 firefighting shafts with rising mains. These have good integrity and retain the masonry of the old mill.

Why would Building Control not be a point of contact for plans K?

[kurnal]

Yes- we are working on it and also the alarm system installer. But have asked the client to chase up all these contacts.  Its better that way.

[Galeon]

'But heres the rub- there isnt a single manual call point anywhere except one adjacent to the panel for testing the system. I cant fathom why a designer should stray so far from guidance - to me it just doesnt seem to make sense'

And why does that not surprise me that there is no certification ?

[CivvyFSO]

Also, why would you have a call point just for testing the system? The testing of call points is to simply test the call points function isn't it? The sounders could be easily tested from the panel, surely?

[nearlythere]

Also, why would you have a call point just for testing the system? The testing of call points is to simply test the call points function isn't it? The sounders could be easily tested from the panel, surely?

As I understand it the fire alarm user test is simply to check periodically that the system sounds from a call point. If it was solely a call point test, in a very building, the user test regime could mean a call point not being operated for some considerable time.
Haven't looked but does or did the BS not say that a detection system must have a manual call point?

[Wiz]

As a simple fire alarm engineer, I may be missing something here, but my first response is that it is obvious the designer was concerned about malicious activation of mcp's in public areas of the building. I've seen many apartment buildings where supposedly the system designer obtained agreement from the relevant parties for such a variation.

In respect of CivvyFSO's point, if you imagine that the full operation of a fire alarm system has three general phases/steps;
Step 1) route between detecting device and control equipment. Step 2) control equipment response, Step 3) Route between control equipment and alarm warning devices.

Testing the system by operating the Evacuate button (for example) is only testing step 3) and only part of step 2) of the above.  

The only way of testing all three steps of the system is from step 1 other wise you don't know if the 'path' from detection to alarm warning is working correctly. I always believe that testing from mcp's/detectors rather than the cie Evacuate button is a better way (unless you are only checking that all alarm warning devices operate).

[CivvyFSO]

As I understand it the fire alarm user test is simply to check periodically that the system sounds from a call point. If it was solely a call point test, in a very building, the user test regime could mean a call point not being operated for some considerable time.

Yes, part 1 systems should have call points as Kurnal points out. My point is that if they have not wanted to bother putting call points in, then why put one in if it is "just for testing". You seem to be creating something to test just for the sake of it. I would be more inclined to think along the same lines of Wiz, where the main number of call points have been left out to avoid malicious activation. Even though Kurnal doesn't agree with it, it is a subjective matter that is really down to opinion.

I agree that a detection system doesn't automatically make up for a lack of passive protection, but if malicious activation is a problem then it is surely better not to have the call points than to have an alarm that people end up ignoring in a building that is lacking in passive protection?

[Midland Retty]

We would all agree that where compartmentation is inadequate the provision of AFD or enhanced levels of AFD (where detection is already installed) is going to be the most popular solution.

Furthermore where there is the possibility of malicious activations, whacking in manual call points everywhere is undesirable. Perhaps, as others have already said, this was the rationale behind the design of the system.

My course of action would be:-

1) Investigate feasibility / possibility of providsing void detection to protect the gaping voids between floors.

2) Provide MCPs at all final exits, and selected high risk areas as appropriate, complete with alarmed security covers, rather than have 32 call points across the building.

3) Whilst all flats have heat detection, could the householders be encouraged to provide their own stand alone detectors? (home fire safety checks for example) The idea being that the smoke detector should alert the householder quicker than the  heat detector, and if the householder reacts quicker they should hopefully raise the alarm and call for help quicker.