• Welcome to the new and improved Building Code Forum. We appreciate you being here and hope that you are getting the information that you need concerning all codes of the building trades. This is a free forum to the public due to the generosity of the Sawhorses, Corporate Supporters and Supporters who have upgraded their accounts. If you would like to have improved access to the forum please upgrade to Sawhorse by first logging in then clicking here: Upgrades

Origin of Egress Calculations

jklancher

Registered User
Joined
Apr 11, 2023
Messages
1
Location
Minnesota
I've been reviewing the egress width requirements for various occupancies in the IBC and NFPA. I see a range of 0.15" per occupant to up to 0,5" per occupant depending on the use of space. Does anyone have access to the logic or rational for how the limits were created. I'm wondering is there were some sort of fluid dynamics type work done related to humans passing through doorways, etc. Any insights or links to background info would be greatly appreciated.

Best regards, Joe
 
My guess is that the numbers are what a committee could agree on. According to a man that was there, the number fifty as a trigger for two exits was just that....there was no empirical evidence to back it up.
 
In the good old days we used units of exit width. One unit of exit width was 22", and an additional 12" could count as a half unit. The number of occupants per unit of exit width depended on the occupancy and whether the building was sprinklered.

At some point, I think in the 80s, exit width calculations were changed to be based on inches per occupant. Stairs require more width than doors or corridors because people move slower on stairways. Sprinklered buildings could have less width per occupant as an inducement to sprinkle buildings and because a fire would spread slower and produce less smoke in a sprinklered building. There probably was an empirical basis for assigning different widths per occupant, but I seriously doubt that anything as sophisticated as fluid dynamics was ever used, except possible for large assembly occupancies such as sports stadiums.
 
How they decided on these factors lies somewhere in the vaults of ICC. Some reasoning had to be submitted (like it is required now) to make a change. Therefore, some science may have been involved. I am sure fluid mechanics plays a part in the study of crowd movement. There have been several studies on the subject, and many computer modeling applications are based on these studies. As artificial intelligence continues to improve, these computer models will be better at replicating real-life scenarios.

Means of egress is like a highway system during the morning rush hour. Everybody is leaving at about the same time from various communities around the city, all converging into the city. As the number of cars increases, the number of lanes increases. When there is an imbalance (too many cars for the number of lanes), things get backed up, and you have your morning traffic jam. The DoT then acquires funding and adds as many new lanes as it can based on available land and cost.

Buildings are much the same way, except the DoT is usually reactive whereas the building department is proactive. The building department will not allow you to increase the occupant load without increasing the egress width, especially in an existing building. Thus, you are required to make the changes to the means of egress before the occupant load is permitted to be increased with a Certificate of Occupancy.

A Brief History of Egress Width Factors (UBC through IBC)

I have no history with NFPA, so I do not know how they achieved their numbers because my experience with building codes started with the UBC. Before the 1991 UBC, egress width was calculated by dividing the total occupant load by 50. This method goes back to 1955 (the oldest UBC copy I have is the 1955 edition).

The 1991 UBC changed the width calculation to 0.3"/occupant for stairs and 0.2"/occupant for everything else. The 1997 UBC provided a table that modified the widths slightly for Group H and Group I-2. This table was carried over into the 2000 IBC but also added reductions for installing a sprinkler system and remained the status quo until the 2009 IBC.

The 2009 IBC eliminated the sprinkler reductions (which pissed off a few people), but the specific requirements for Group H and I-2 were eliminated. However, those areas were already covered in other parts of the IBC by establishing minimum widths greater than those for common corridors.

The 2012 IBC saw the comeback of the reduced width factors for sprinkler systems but also added the emergency voice/alarm communication system as a required condition. This is the situation as it remains today with the 2021 IBC.
 
I believe the 22" came from 95th percentile male shoulder width in post WWII research. (Should raise concerns that it's all based on size of people 75 years ago.) That provided the 0.2" or 0.22" in NFPA as a basis for per occupant. The factors like for stairs - 50% - were probably educated committee guesses.

I'd love to know where the 30" threshold for guards came from.
 
I cannot confirm but I believe these calculations came from troop movment requirements on ships from WW2 with amphibous landing ships, I cannot point to a direct reference, but had a boss (900 years ago) who worked with WW 2 specifications for the Navy for amphibous landing ships for the WW2 Pacific Ocean Amphibous Landing efforts. I think Fred and Barney were on the US Navy Rowing team at the time.
 
The 1991 UBC changed the width calculation to 0.3"/occupant for stairs and 0.2"/occupant for everything else.
Didn't all of the legacy codes move in this direction, I believe as a result of the BCMC 1988 (?) Report on Means of Egress? I caught the very end of those discussions, attending BCMC meetings starting in 1987. Huge impact on assembly seating requirements, among other changes.
 
NFPA 101 Appendix has some background on their numbers. It also includes a warning about the data being old and probably not matching current demographics (aka we're fatter now):

Screenshot 2023-05-02 102323.jpg
Screenshot 2023-05-02 102430.jpg
Screenshot 2023-05-02 102451.jpg
 
Top