BS9999 staircase widths

[Phoenix]

Like many I have harboured a certain amount of caution about the validity of the smaller exit width and stair width requirements in BS9999.  

Here is a real and current situation:

1.    A building of ground and six upper floors
2.   Ground floor has its own exits
3.   There are two staircases, neither of which are lobbied
4.   In assessing means of escape from the upper floors we have to discount the largest staircase and this leaves a single staircase (the smaller of the two) down which everyone has to escape
5.    The storey exits to this staircase are 810mm wide
6.    The staircase is 940mm wide
7.     The staircase serves floors 1 to 6
8.     The final exit from the staircase is 1040mm wide
9.    An evacuation drill has been carried out to determine if the stairs are adequate for the current population, observations were made at all levels to see how quickly the floors cleared of people
10.    At the time of the drill there were 480 people on floors 1 - 6
11.    The population was reasonably evenly distributed between the floors with the exception of floor 3 which had only about 40 people on it
12.    Floors 1, 2, 4, 5 and 6 each accommodated approximately (480 – 40)/5 = 88 people
13.   The storey exits of 810mm are wide enough for this number of people, according to all guides
14.    The staircase is narrower than the minimum recommended width for new buildings, according to all guides (being only 940mm)
15.    The building has a risk profile of A2
16.    The BS9999 Table 15 figure for A2 over 6 floors is 2.2mm per person
17.    940mm/2.2 = 427 people so this indicates that the safe maximum occupancy on all the upper floors, aggregated together, is 427
18.   In the test there were 480 on the upper floors and this is greater than the BS9999 maximum of 427 [The main objective of the test evacuation was to determine if the over-population (12%) would, in reality, lead to people being unable to leave the fire floor in the standard 2 and a half minutes - we were hoping to find that there was sufficient safety margin in the BS9999 figures for the over-popuation not to be significant (what we found was something different)]
19.    During the test evacation the occupants did not delay in making their escape but moved immediately and swiftly towards the exits
20.     The time it took for people to fully evacuate the 6th floor and enter the staircase was 5 minutes 52 seconds
21.   The time it took for people to fully evacuate the 5th floor and enter the staircase was 6 minutes 20 seconds
22.    The time it took for people to fully evacuate the 4th floor and enter the staircase was 4 minutes 41 seconds
23.    The time it took for people to fully evacuate the 3rd floor and enter the staircase was 2 minutes 52 seconds – There were no more than 40 people on this floor
24.   The time it took for people to fully evacuate the 2nd floor and enter the staircase was 3 minutes 48 seconds
25.   The time it took for people to fully evacuate the 1st floor and enter the staircase was 2 minutes 50 seconds  
26.   The standard target time for the evacuation of people from the floors into the staircase is 2 minutes and 30 seconds
27.   If there had actually been a fire on one of the top floors, instead of it being a test evacuation, and if the location of the fire had meant that everyone had to use the smallest staircase, then there is a real chance that people would have died due to their inability to leave the fire floor in less than around 6 minutes  
28.    This building is overpopulated with 480 people on the upper floors
29.   Reducing the population of the upper floors to the BS9999 compliant (actually, semi-compliant as the minimum 1000mm cannot be provided) figure of 427 would be an overall reduction of 53 which converts to a reduction of approximately 9 per floor
30.    It would seem to me that if we reduced the number of occupants on each floor by about 9 then the time required to clear the floors is unlikely to reduce from the observed time to 2 and a half minutes.
31.    If we could increase the stair width to 1000mm and reduce the population to 1000/2.2 = 454 then the building would fully comply with BS9999 recommendations
32.    It also seems to me that taking the actions referred to in 31. above would be extremely unlikely to reduce the required safe escape time to anywhere near 2 and a half minutes
33.    It therefore seems to me that this observed evacuation of a building significantly undermines the figures in Table 15 of BS9999

I'm not after solutions to the problem in the building, we have that covered.  I'm interested to know if other people are as shocked as I am at this result and, if so, what should be done about it.

Stu

[colin todd]

All it means stewie is that it would take longer than it took an edinburgh audience to evacuate a theatre in 1911, according to anecdotal evidence from the duty manager and a scenic technician, neither of whom had stop watches.

However, it is quite a long time to evacuate, so one has to wonder what the premovement  time was.

[Phoenix]

Negligable pre-movement time ( a lot of them knew there was going to be a drill and were ready to go!).  The extended times were entirely down to congestion in the staircase and at the storey exits.

Stu

[kurnal]

Useful and worrying information Stu. It confirms our earlier discussions on 9999 and our concerns over the interpolation on exits of less than 2 units width. As you say an ADB solution for a new building would have provided at least a 1300mm wide stair in this case. Shudder  to think how it would have been had the ground floor also used this final exit!

[colin todd]

But much as I loathe BS 9999 with a passion, Big Al, Stewie is assuming that an evacuation time of 2.5 minutes has some mystical significance.  Now I admit it came (like all good things) from Edinburgh, rather than Bathmat Lock, but the point about BS 9999 is that it is not trying to get people out in 2.5 minutes. So, you are simply proving a given by finding that BS 9999 solutions will not get people out in 2.5 minutes.

[kurnal]

Without getting hung up on the minutiae  here is a practical illustration that in this case if the means of escape were taken towards the maximum acceptable variations as set out in BS9999 then it is unlikely that persons would escape before the tenability limits are reached. Indeed BS9999 claims that its recommendations for the widths of stairs are conservative as persons on the fire floor are more likely to respond more quickly than others. Stu's example indicates how this premise cannot be relied on, notwithstanding the effective elimination of pre-movement time in this case.

I agree that setting prescriptive evacuation times of 2 minutes, 2.5 minutes, 3 minutes or w.h.y. is all  fairly arbitrary but 100 years experience since the Edinburgh Palace fire has shown that evacuation times such as these  are a reasonable balance and suitable benchmark for most none engineered buildings.

Though the evacuation time is not prescribed in 9999, we all know that table 15 is the illegitimate son of ADB table 7 but with much interpolation and extrapolation. And Table 7 was itself the second cousin of the Post War Building Studies Fire Grading of Buildings subject to a rounding exercise on metrication. Did I ever tell you I can prove a link in my family tree to Bonnie Prince Charlie even though I don't wear the tartan in public?

[Tom Sutton]

As many of you will know, many moons ago we used to provide a service witnessing evacuation drills in medium to large buildings, and I personally have witnessed many. I found the full evacuation usually took between fifteen and twenty five minutes. So are you suggesting 3 minutes to clear stage one and the remainder of the time to clear stage two. I think this was very unlikely and six to seven minutes to clear stage one in more likely, considering this was under the prescriptive method, using theoretically 2 to 3 mins to clear stage one but practically anybody guess.

[kurnal]

Well I don't know Tom. I have carried out 4 evacuation drills in the last 3 weeks.  3 in multi occ office buildings and in a large warehouse - footprint 100,000 sq m. That is as big as they come.  Each of the office buildings was certified clear in under 3 minutes and the  warehouse had the cause of the alarm verified, all 2000 persons confirmed outside the building and all areas searched by fire wardens in 8 minutes. We had 10 observers in that case.

Two of the office drills were unannounced, whereas the warehouse managers had been briefed.

[colin todd]

Big Al, it was not called the Edinburgh Palace.  Maybe you are thinking of the place the queen hangs out in (until we become the peoples republic of scotland). And I am not sure how you worked out the tenability limit for the floors without any knowldege whatsoever of the premises.  We might as well sack all fire engineers, and just use elderly crystal ball gazers such as yourself to tell us such things.

[Tom Sutton]

Well Kurnal things have improved dramatically because it was certainly not the case in my experience. My premises were multi storey office blocks (6+7 storey) department stores and the teaching block of Liverpool teaching hospital and none of them came near to your experiences.

[Midland Retty]

It all depends on the size / use / occupancy type of the building being evacuated, how well trained the occupants are on knowing what to do if an alarm is raised, their mobility etc etc etc.

Take a care home. Has anyone ever witnessed an evac drill in a res-care home where staff evacuated the effected compartment (let alone the building) in either 2, 2.5 or 3 minutes? Would it be unreasonable to expect them to do so?.

Conversely I have witnessed evacuations of large buildings which took place in about three minutes. Yet an evacuation I witnessed two weeks aago t a secure mental institution the "ward" took about 10 minutes to be evacuated. Is that acceptable? 

How realistic are those 2, 2.5 and 3 minute timescales? What are they trying to achieve? How accurate are they? Are they decent benchmarks? To me a sensible approach has to be taken based on the building you are assessing.

I was once told that the quoted evacuation times were handy because by the time the fire service arrive the person co-ordinating the evac will know if anyone is left in the building!!! Really??? Does anyone else factor that in when they decide on evacuation times?

Alot of Responsible Persons ask me "If you want people out of the building in three minutes or under why do we need 30 min F/R doors?"

[colin todd]

Retters, you witnessed the evacuation of a secure mental institution?  Did the chief  fire officer evacuate too, or did he remain in his office.

[Midland Retty]

Oh now Sir Col really even for you that's scandalous.Behave yourself !

[Phoenix]

Of course the two and a half minute target time is plucked out of the air but I suspect that there are no numbers in any fire safety guides that are not so derived.

All of fire safety is about probabilities and we try to set standards and make recommendations for buildings that will put them at the low probability end of the risk spectrum.

The ceiling heights in my office building are not high and smoke and heat from a growing fire could be expected to move horizontally fairly rapidly.  I do not think that there is any way that we can view a period of around 6 minutes to evacuate what might potentially be the fire floor as being at the low probability end of the risk spectrum.  

This is just one result and we do, of course, need more than a single result before we commit to reconsidering the efficacy of the figures in BS9999 but I did want to flag this result up so that others who see similar results might have something to build on.

Most evacuations I oversee do tend, more or less, to meet the notional target time of around two and a half minutes for evacuating the floors and this does tend to put them at the lower end of the risk spectrum.  This office result is way off and I would love to hear of other similar results when the limits of BS9999 are pushed.

I found the full evacuation usually took between fifteen and twenty five minutes ....... six to seven minutes to clear stage one is more likely.

Tom, if you have seen these long evacuation times regularly then I think something is wrong and some improvements are required.

..... an evacuation I witnessed two weeks ago a secure mental institution the "ward" took about 10 minutes to be evacuated. Is that acceptable?  

Maybe, maybe not - we do take steps to limit the rate of fire growth in such situations.  We know that evacuation times are likely to be relatively long so reducing the potential for rapid fire growth is very important.  My office building had a (possibly generous) A2 risk profile which accepts a medium fire growth rate.

It all depends on the size / use / occupancy type of the building being evacuated, how well trained the occupants are on knowing what to do if an alarm is raised, their mobility etc etc etc.

Exactly.  But in my office building, I should point out, the staff were all well trained, they all knew what to do when the fire alarm sounded and there were a negligable number with mobility problems.

A lot of Responsible Persons ask me "If you want people out of the building in three minutes or under why do we need 30 min F/R doors?"

This just illustrates how little your average RP knows about fire safety and how important our jobs are.  No one and no guide has ever stated that we "want people out of the building in three minutes or under."  I hope everyone on this forum has an immediate logical and rigorous answer to this question.  

Stu

[colin todd]

My job isnt important. More people die of cold related illness in the London boroughs every winter than die from fire in the whole of the UK all year.  Hypothermia risk assessors are the important ones.