Mechanical Extraction of Dead End Corridor

[GB]

I am currently running a CFD model to look at the removal of smoke from a dead end common corridor of approximately 15m on the 2nd floor of an apartment block.

I conducted a comparative model with conditions faced in a ADB 7.5m dead end corridor with a natural smoke shaft however the AI did not accept that a comparison could be made as they are not like for like conditions due to the extended distance.

I have reviewed the guidance by the Smoke Control Association in conjunction with BRE Report BD2410 as I have been asked to use BS9991 for this element.

The BRE Report used fires between 0.25MW and 2.5MW as the source with a door gap between the apartment of origin to the common corridor / lobby of 0.1m x 2m in height. This required between 160 - 640 air changes per hour to maintain a tenable limit within the corridor when the door to the apartment was closed with the 0.1m x 2m gap being constant.

My questions are: why such a large gap between the apartment of origin to the corridor?

What this a modelling / grid size / time dependant issue or does that size of gap relate to another factor / publication I am unaware of?

Is there an alternative methodology that can be used for a more realistic scenario that doesn't have an apartment door gap of 0.1m x 2m that could be considered?

[kurnal]

if i recall the gap was to represent a door partially ajar. I think it explains the rationale in the  BRE study.

[GB]

Thanks Kurnal,

You are correct in that was the assumption.

It does seem another excessive assumption however with no FD20 door allowed from a room of origin, a fire having sufficient ventilation within a modern apartment to reach 2.5MW with a window breaking providing further ventilation at a later part of the scenario.

It feels that the 2005 report was written with modelling boundaries and restrictions rather than what would be expected at a scenario of this type and would this be modelled differently if it were completed today?

[wee brian]

How else will you get a fire hose through the door?

[GB]

Thanks Wee Brian

The simulation time frame has to allow for the door to the apartment being fully opened to allow for fire fighting (0.78m x 2m) then adding a window breakage and the corridor and stair doors being opened during FF ops.

The 0.1m x 2m is present from ignition to the end of the simulation period i.e after firefighting operations.

It is this pre firefighting operational phase of a 0.1m x 2m gap which results in such a high ach rate which is over engineered in my opinion however I am open to being shown differently.

[John Webb]

Is it possible to reverse the situation and pressurise the corridor to prevent smoke entering through the door that's been left ajar? Has the advantage that the air handling equipment only deals with ambient air temperatures, not heated smoke and gases.

I did some work a long while back (mid-1980s) on the movement of smoke into a corridor for the old Home Office in connection with siting of detectors etc. in hotel and other sleeping accommodation. I can't recall the work being published as such - I think it was fed into BS5839. (Although this was looking at the fit of doors rather than one ajar.)

[AnthonyB]

I know of cases where the dead end corridor has been more like 30 metres, using two smoke shafts and fans for clearance and sprinklers to each flat, so you can get the extended distances, the sprinklers were however a key element in the modelling.

[kurnal]

Where travel distances are extended, it is common to use push and pull arrangements where two fans are used, one at either end of the protected corridor, each is capable of operating in extract or make up mode and addressable to the smoke detectos in the corridor to pull or push whichever would be most advantageous.

Trouble is its all so subjective and as ever Bulls**t baffles brains. The worst case I have seen was a 45m dead end corridor with a push and pull arrangement and when you looked into the fire engineers report the fans were sized based on conventional fire door leakage with the threshold taped. It was approved though.

There is usually  a boost facility to enhance the system performance during fire service intervention.

[GB]

You are correct Tony in that with a reduced fire size i.e 1MW which could be restricted due to the presence of the sprinklers, the extraction rate was around 160ach with the 0.1 x 2m gap.

With the 2.5MW fire, the ach ramped up considerably with the very large gap causing the over engineering.

I am convinced that the BRE report conducted almost 10 years ago would not use the same design parameters with the advance of computation facilities and models as I don't think the permanent gap is justified however as BRE have produced a gospel stating it so, the AI are only allowing preaching from that book of the bible!!

[wee brian]

I seem to remember there were some problems with modelling door gaps back then.

Maybe you should ask the guy who did the work.

[GB]

Thanks Wee Brian, you are right in that that may be the best option.

I will drop BRE an email and see if they respond.

many thanks