BS9999

[SidM]

Can it have retrospective application?

[CivvyFSO]

If you mean can it be used to risk assess an existing building, yes.

[SidM]

Would it have retrospective effect for the purposes of audit?

[CivvyFSO]

If during an audit a risk assessment had been done to BS9999 I would have no problem providing that it had been done properly. Certain aspects of BS9999 i.e. interpolating exit widths down from 1050 etc are blatantly wrong and I would happily challenge them, but in general the extended travel depending on the premises layout etc is reasonable.

I would not use BS9999 myself to justify someone elses extended travel distance, but the fact remains that given the right circumstances the same values that are in BS9999 could be deemed safe without even referring to the document. i.e. You might have 60m 2 way travel in a warehouse, so depending on the nature of the premises I may agree that this is actually reasonable using the very same thinking that BS9999 used; high roof, alarm system, management etc.

It is either safe or it isn't.

[Phoenix]

interpolating exit widths down from 1050 etc are blatantly wrong and I would happily challenge them

I am inclined to agree with you in general, Civvy, but I think you'd have to hesitate before formally challenging the guidance in 9999.  The application of 'x'mm per person for widths below 1050mm is supported by some (not me) and there is some weight behind the recommendations of 9999.

Intuitively I agree that following 9999 can lead to very large numbers of people depending on what seems to be very little exit width capacity.  But then, I don't know - I only know what I've been brought up with.  My intuition might just be the product of a career long exposure to ADB and traditional fire safety principles.

   
Stu

[CivvyFSO]

I don't think that any suitable research has been done since the research that is stated in Post War Building Studies No 29 was considered, to show that any more people can get through an exit until the width allows people to exit side by side. Once we can physically get two people side by side through an exit there is going to be a sudden jump in the exit capacity around the 1000-1100mm mark. (Some research into widths etc has actually been done, but it tends to point towards less p.p.m than the standard 40, so fire engineers etc are not that quick to quote it. Pauls, Nelson & Mowrer are names that jump to mind regarding this.)

I am not saying that 110 persons should be a fixed limit, it should be subject to variations just like anything else is, but consider the following:

A room with 2 x 1045mm (just to be awkward) doors, one at each end
ADB allows 110, BS9999 allows about 433 in the room with the best case, i.e. absolute minimum width per person
This exit will generally only allow single file escape
Taking 40ppm as standard flow gives about 11 minutes evacuation time if we lose one exit.

Now imagine a little box that you can fit these 433 people in, with a roof high enough to warrant the door width that is to be applied. Using the standard equations, light a growing fire in it and see the conditions at the 11 minute mark. I don't think it would be classed as 'conducive to the long-term health goals' of any persons involved.

Now remember that this has not taken into account that once a considerable queue forms at the exit this 'flow' can break down. (There is plenty of case history of insufficient exit capacity, or too many people arriving at the same exit.) There has been quite a bit of research on this, and Fruin's levels of service gives a good indication of the possibilities. The research is often based towards the movement of crowds, and not specifically escape for fire.

BS9999 regularly creates these queues, without (IMO) considering the actual impact that these queues can have on exit capacity.

[Phoenix]

Good points Civvy, have you (or has anyone else) any idea where the revised figures came from in 9999?

I, like many, tend to have a flexible skepticism for innovations in fire safety but also tend to feel that it is safe to place a certain amount of trust in high profile publications such as 9999.  I know that there are a thousand things wrong with it but I would very much like to believe that the important parts such as exit widths have been published after convincing evidence has been provided that support this published guidance.

Here's another thing about 9999.  And some people will recognise this after a recent discussion...

It recommends reducing travel distances by 25% for places where alcohol is consumed but it makes no similar recommendation in respect of increasing exit widths by 25%.  

This goes very much against the understanding of how places undergo evacuation.  Places where alcohol is served are often quite crowded.  Crowded premises have evacuation times that are heavily dependent upon the queueing times at exits.  Queueing times are dependent upon exit widths.  

Travel distances are important in more sparsely populated premises.  In such premises evacuation times are dependent on how long it actually takes to get to the exit and, as there are no queues at exits, flow rates through exits are irrelevant.  So exit widths lose significance.  

In short:

Crowded premises - Exit widths are most important.

Uncrowded premises - Travel distances are most important.

Travel distances become almost irrelevant where people have to join the back of a queue to wait to exit, so the 25% reduction has little significance.  Exit widths are important yet they are not adjusted where there is alcohol consumption.

It doesn't make sense to me.

I really think the contributors to this forum would have made a much better job of the guidance document than whoever was responsible for it.

Stu

[Mr. P]

I have been imagining your example Civvy. My scene on it is that some of the people are the fuel load. Sardines and tins come to mind.
Having been told that 9999 is based on RA it seems you can use it to accompany a variety of scenarios/risk levels for new or existing premises.

to quote Phoenix -{Travel distances are important in more sparsely populated premises.  In such premises evacuation times are dependent on how long it actually takes to get to the exit and, as there are no queues at exits, flow rates through exits are irrelevant.  So exit widths lose significance.}

Are you saying if there are fewer people they can travel further? -Because they don't have to queue? With larger numbers  of people, if the exit routes and door widths are adequate to take that number before people become further burdened by the fire/smoke/heat etc. or can move into another compartment/protected area, the TD may also not be quite so critical?

Sorry, this is just a cat & pigeon statement ( not our Cat mind you)...

[Phoenix]

Mr. P,

I only put that bit in as an example to support Civvy's little attack on 9999, not to alter the course of this thread.

However, when you ask:

Are you saying if there are fewer people they can travel further? -Because they don't have to queue?

I would respond by saying that that is not what I said, no.  I said, where there are more people, the significance of travel distances diminishes but the significance of exit widths increases.  Therefore 9999 should maybe have thought about increasing exit widths where alcohol is consumed (by 25%?).

[Another issue that has a bearing on this argument is that of the response times of differing centiles of the population - see PD7974-6.]

Stu

[Mr. P]

Phoenix, as I said cat's & pigeons. For those carrying out the RA's, it leaves alot of responsibility on them which the quick buck maker seems to care little of... they apply (not RA) and walk away and the 'punter' does not often realise they as RP could have come back/re-dress of the assessor.

There are 2 ways to sleep at night - you don't care - or - you know you have done your job well.

If the situation comes up for me re: the widths and numbers game for premises where alcohol is consumed, I'll case it and put it up for them to increase the widths. Most places I get are existant/change of use, not too many purpose builds.