Case study: Atrium requires smoke curtains
A 4-story atrium required some creative smoke control options.
While many of the projects are sufficiently complicated to warrant a Fire Dynamics Simulator (FDS) model, some projects with an atrium are able to make use of the NFPA 92: Standard for Smoke Control Systems calculations. On one project, the atrium included 4 stories where most of the space was open. Balconies were present on the upper floors, mostly for circulation space, but also in some significantly large waiting areas. The atrium closely resembled a “communicating space” as defined in NFPA 101: Life Safety Code, except that it connected one too many stories.
Using “back-of-the-napkin” math, initial estimates for needed exhaust rates were in the ballpark of 150,000+ cfm of exhaust. As anyone would expect for an atrium with such a small footprint, there was resistance to providing a system capable of providing this much exhaust and everything that comes with it. While there was concern about fan sizes and ductwork, there was more concern about the sheer number of louvers that would need to be provided for the make-up air, especially because they must be located below the smoke layer interface and couldn’t be hidden high up in the ceiling.
There was also resistance to significantly altering the design of the atrium, so smoke curtains were recommended to lower the smoke layer interface. By providing curtains on the 3rd and 4th floors, these areas were still considered part of the atrium, but were separated from each other and the lower 2 floors by construction that resisted the passage of smoke. This was an acceptable means for lowering the smoke layer interface from 6 ft above the 4th floor to 6 ft above the 2nd floor in both the eyes of the authority having jurisdiction (AHJ) and the engineer. This reduced the calculated exhaust rate to approximately 50,000 cfm, which was much more palatable. The only significant hindrance to this was the cost of the smoke curtains.
As the concern was only providing a separation from the perspective of smoke, not fire, and occupants did not need to egress through the curtains, the project was able to save some cost. However, due to the large perimeter of the opening, the curtains were going to cost several hundred thousand dollars. It was still a cost savings when paired with the reduction in exhaust rate, but expensive nonetheless.
Moreover, actual curtain selection could not wait until sending the design package out for bid. The project was on a tight budget, and actual quotes had to be procured because not every curtain is capable of doing the job. Portions of the curtains needed to have angles in them, other areas needed to have columns put in place to account for the tracks, and curtains with actual tracks at the edges had to be selected. A surprising number of “smoke-rated” curtains do not have tracks on the sides, which can result in gaps between the curtain and the wall/column of several inches. If an engineer does not vet this particular characteristic, this might go unnoticed until commissioning or even inspection by the AHJ.
With the curtain selection complete, the atrium proper was only connected to 2 stories and the calculations would be simple enough. Naturally, an axisymmetric plume calculation for the basic exhaust rate was conducted. While overhangs from the 2nd floor over the 1st did exist, these were deemed to not be significant. This was based on the fact that the overhangs were approximately 5 ft wide and the areas under the overhangs were solely used for circulation, which was regularly expected to be in use.
Furthermore, with the relatively low (<10 ft) ceiling height on the 1st floor, sprinkler activation (using the Beyler correlation) could be counted on to suppress any significant fires under the overhangs. Therefore, for this project, a balcony spill plume was not conducted, with the approval of the AHJ.
Will Clay is a senior engineer at WSP USA, with 8 years of experience in the fire protection engineering and life safety consulting industry.