Smoke Extract Points Locations for Single Domain Atrium

[Buildtech2]

Our Atrium is one domain with extraction points in the ceiling located at 3 different spaces in atrium. The total smoke extract rate is 71 cum/sec. It was assumed that equal amount of extract air will be extracted from 3 different spaces (24 cum/sec) though there is no physical barrier between these 3 spaces. However, we can extract only 50% if air (12 cum/sec) through the ceiling void from one of the space mainly due to restrictions with the beams and expected air (24 cum/sec) cannot come to the fan intake duct but that can be compensated from other spaces.

I am assuming that the smoke movement, regardless of the origin of the fire, will not be affected since the entire atrium is one common domain.
As per the principles of the smoke movement, the smoke will travel from below the ceiling creating an equally distributed smoke exhaust from the atrium.

lease see attached required air flow rates as per CFD in Atrium and possible air flow rates as per shop drawings.

Appreciate, if anybody can tell me if I am assuming correctly.

 

[stookeyfpe]

You haven't provided enough information to offer an opinion. Is the atrium protected by an automatic sprinkler system? Are you exhausting the atrium - if so what was the design fire and heat release rate assumptions? Are you designing this to NFPA 92B or does your building code have a specific criterion for atria?

 

[Buildtech2]

Yes, atrium is protected by an automatic sprinkler system and there is a dedicated exhaust system for the atrium. The design fire heat release rate is 2 Mw. The atrium smoke management system was designed as per NFPA 92 B.

Please see attached calculations

 

[UFT12]

I m not sure I can help much, I just wanted to comment on the heat release rate. Per NFPA 92, A.5.2.1 (by the way, NFPA 92B is obsolete), a suggested size of heat release is around 5000+ kw for atria. I m not saying the 2000 kW you considered is wrong, just to assume that such fire size is coming from engineering judgement.

Also, your assumption that the smoke movement regardless of the origin of the fire will not be affected, may not be entirely true. The location of the plume makes a difference if it is evolving near walls. The smoke layer interface z may be lower than 14 m in this case. Just though it may be worth checking placing the fire origin on a location where walls come nearest possible.

By the way, if you have a CFD model, what are the criteria for the simulation and what results are you you getting? Do they confirm the results of the xls sheet?