Dead End Corridors

Updated: An Approach to Mitigating Dead End Corridors

8/19/2024

For years, architect clients have asked me for a code solution to solve the challenge of corridors that exceed the dead end limit. As a quick reference, dead ends are limited to 20 feet, except in Groups B, E, F, I-1, M, R-1, R-2, R-4, S and U with a full NFPA 13 sprinkler system, where the limit is increased to 50 feet (2021 IBC 1020.5).

A dead end corridor in a residential building.

Up until recently, my normal suggestion has been to place a cross-corridor door to break up the length of any one corridor segment to less than 50 (or 20) feet. This has been accepted in numerous jurisdictions and also validated by an ICC staff opinion that I requested on the topic. In many cases, I have seen the cross corridor door on magnetic hold opens, to release upon activation of the fire alarm system

Cross-corridor door on magnetic hold opens.

But on a recent project, a local AHJ has taken the stance that a cross corridor door does not mitigate a dead end condition. I scheduled a meeting with the AHJ to explain how this has been done on other projects and to share the ICC interpretation, but they were reluctant to budge. After some negotiations, the AHJ allowed the door, but insisted that it not be held open and also required a large "Not an Exit" sign on the door itself.

I have drawn up a similar situation in the image below. The corridor is serving a Group R-2 occupancy, so the dead end limit is 50 feet.

Sketch showing the dead end condition with added cross-corridor doors.

New Development: An ICC Committee Interpretation

A couple of years after obtaining an ICC Staff Opinion on this matter, I finally decided to request an ICC Committee Interpretation. A Committee Interpretation is developed and approved by the ICC Interpretations Committee and generally carries a little more weight than a lone staff opinion. Fortunately, the Committee Interpretation came back favorably. Check out the language below (or you can view the entire document here).

Yes. The installation of an intervening door within a corridor may be utilized to meet the dead-end corridor maximum length requirements of Section 1020.5. The strategic location of an intervening door, which does not swing in the direction of egress travel and is not provided with exit signage, significantly increases the probability that the occupant will turn around instead of continuing beyond the door into the last section of the dead-end corridor looking for an exit. The placement of the intervening door must be such that each corridor segment on either side of the door is within the dead-end corridor length requirements of Section 1020.5.

Dead End Corridors: What Do You Think?

Do cross-corridor doors mitigate the 20 foot dead end limit? Do you think they should be permitted to be held open?
Let me know your thoughts in the comment box!

25 Comments

Matt Benfer

4/16/2020 09:45:54 am

Sounds like a valid approach. I think the "Not an Exit" sign is a must have on this type of a door as the whole purpose is to prevent the rest of the normally dead end corridor as being attempted to be used for egress. Whether the door is held open or not doesn't seem to be a huge issue to me, except that I would think that a door which is not magnetically held open is probably more likely to be propped open by other means, this negating the door.

Chris Campbell

4/16/2020 09:58:50 am

Thanks for the reply, Matt. I like the point about the door being more likely to be propped open if the magnetic hold opens are not provided...will have to bring that up next time I chat with this particular AHJ!

Mike Gartner link

8/9/2023 09:50:59 am

Because the length of a dead end corridor is measured from the end point of the corridor to the location where occupants have a choice of two directions leading to separate exits, it would be up to the AHJ for acceptance. In the strictest interpretation I would say no, but with less restrictive interpretation, possibly with limits. I would think that the hall door must be operable by persons with a key only making second corridor occupants subject only to Common path requirements (assuming hall is less than 20 feet). Either way less than ideal and should be avoided where reasonably possible.

Ben Dobben

10/1/2024 07:32:20 pm

I concur, in the past I have had this very issue, and I have had ridiculously long entrance halls to mitigate this problem.
In the past I haven't, but now include the building department as the drawing progress.
They don't mind a few questions, but really tire of this quickly.
We had to change the exit porch due to recent code changes, as the existing building was too close to street, and yet existing happened 100 or more years ago. What was typical on this street made house was not on the official site plans was that it wasn't showing, and so we argued that it may be that at the 1950s town survey may not have included none heated spaces.
I suggested to the contractor is just do a repair (except exit stairs) a little at a time, spacing different phases months apart, stretching what should have been a two week project to a 12 month project.

Brian Lukus

4/16/2020 06:14:46 pm

I have seen this situation approved with a locking hardware set on the cross-corridor door in addition to the door holder and signage. Some other things to consider:

1) How many times can you add cross corridor doors? Twice? three times? If this were a B occupancy, are you proposing a door can be added to get another 50ft of dead end?
2) The intent of the dead end length is to limit the path of travel such that the occupant(s) have to turnaround to find an exit. If the door is on a hold open, the fire could be elsewhere in the building, (doors only required to release on the smoke next to the door per NFPA 72), so what is the held open door preventing in a fire emergency that has not spilled out into the corridor? The emergent building evacuation is still occurring and the occupant(s) could be mentally stressed from the fire incident and miss the exit.
3) I know Codes are written around fire emergencies, however they are also accepted for other emergent building evacuation. This held open cross corridor door does nothing to prevent the dead end condition in non-fire emergencies. Yes not a fire Code mandate to design for non-fire emergencies but as life safety professionals, is it prudent design to think about this topic especially in assembly occupancies where these other emergencies have occurred?

4/17/2020 07:57:56 am

Hi Brian,

Thanks so much for sharing your thoughts!

1) I think the common path of travel limit prevents the cross-corridor door from being applied multiple times. For a Group B, this is typically 100 feet, so you would likely only be able to provide one cross-corridor door and still meet the 100 foot common path.

2) Good point. I do appreciate the previous comment above that if the door is not held open by the fire alarm system, it is more likely to be propped open with the door stop.

3) I think this is a fair point to consider. Fire is just one type of emergency - there are clearly many others. I would note that in a non-fire emergency, you would not have a smoke-filled corridor, so the chance of missing the exit in the first place is much less.

Kelly Kidwell

5/9/2020 11:49:44 am

2) You can require all doors on magnetic holds to close upon alarm regardless of alarm initiation point. It's all in the wiring and programming, but that could mitigate concerns regarding doors being held open when no smoke is in that corridor.

Bruce Verhei

4/16/2020 09:57:25 pm

Good evening.

Exit distances were extended in almost every use or occupancy except H shortly before creation of the I-code series. The IBC exiting chapter most closely followed the legacy UBC. Scissor stairs, less restrictive second way out of boiler rooms, and some other minor items came from the other two codes. More or less Chapter 10 of the UBC was cut and pasted in creation of the new code.

That Chapter 10 was result of recent complete rewrite. The consultant Boeing hired to do the rewrite and seek successful adoption of this replacement was able to convince many of the obvious: We use travel distance in part as a surrogate for exiting time. By exiting time we almost always are considering needed time to leave before smoke, meaning sight obscuring, toxic, maybe hot, and choking gases are thick enough to impede travel to persons not wearing SCBA.

A door across a dead end does nothing to substitute for travel time to point of having a choice of more smoky or less smoky pathway. If concept is that you are providing a location to shelter in place considerably more would be needed.

Requiring rating, and having every door from a space opening into the exit access to be smoke rated and be self-closing with mag-hold open might be an acceptable alternative where not all ready required.

Pressurization of the exit pathways is the gold standard.

I do not see adding a door as mitigating the original problem.

I would favor a code change that extended dead end distance where sprinklers and pressurization are provided.

Best.

Bruce Verhei
Olympia, WA

4/17/2020 07:59:13 am

Thanks for sharing your thoughts Bruce!

Ryan Indorf

4/27/2020 11:54:07 pm

If an AHJ is hesitant to approve a cross-corridor door to prevent continued travel in the opposite direction of an exit, I would be curious to hear their thoughts on interruption gates within exit stairs.

In regards to an automatic closing door in this scenario, I agree with the above comment that all emergency egress events should be considered and not solely fire. However, it may also be worth considering what might happen over time if the door is NOT held open. A wedge in the door completely defeats the purpose of the door, regardless of the emergency.

5/9/2020 11:57:09 am

I think in a lot of ways, dead ends are bad design. Bad for egress and life safety, but also bad for circulation under normal circumstances. My first line of defense is trying to get the architects to improve their design and eliminate long dead ends altogether. Normally a long dead end is an opportunity for another through-way for circulation or an opportunity to add some closets.

In my mind, a case where you need several doors to eliminate a dead end is unacceptable. There needs to be a rethinking of the space--or why not an additional exit? At that point you're probably about to hit the limit for travel distance anyway.

A door is the last resort when nothing else can work. And to me it seems less risky in a fully sprinklered building.

5/18/2020 11:30:05 am

Thanks for the thoughts Kelly, hope all is well with you!

David Dillon link

10/18/2021 07:31:25 am

Hello,

Great article, I find myself dealing with my first dead-end issue here at work in NC. The corridor is a connection between 2 town owned buildings to prevent people having to go outside. The 2 building are a community center and a senior center. The corridor is 30' long and sprinklered, a set of doors on one side that the owner wants to lock to prevent after hours entry into the senior center.

I would like to email the drawing to the author for comment.

10/19/2021 08:53:39 am

Hi David,

Thanks for the comment. Please feel free to email me at Chris (at) buildingcode (dot) blog.

Regards,

Chris

James Samuelson link

2/1/2022 01:35:46 pm

"This has been accepted in numerous jurisdictions and also validated by an ICC staff opinion that I requested on the topic."

Can you share the ICC Staff opinion, I am running thru the same discussion here in Ohio.
Thanks!

Steve Bergeron

4/7/2022 02:03:51 pm

Is the consensus here that a cross-corridor wall and door would be treated like any other corridor wall/ door from a fire-rating standpoint?

For example, in a 3-story R2 apartment building, corridor walls are permitted to be 1/2 hr and doors are 20 min. Would there be any reason for the cross-corridor wall/door to be a higher rating?

4/7/2022 03:33:42 pm

I'm not aware of anything that would require a higher rating.

restorerz link

1/19/2023 08:59:35 am

Fire safety is important in order to: Reduce the risk of injury to employees and customers. Reduce damage to facility/building. Protect against possible fines.

Hamed Galal link

12/23/2023 04:25:44 am

I believe adding door in the corridor in order to duplicate the dead end limitation is not accepted, as this eliminate the concept of dead end entirely
also this would violate the common path limits
and definitely will violate the single exit rule for this dead end pocket that is being mitigated by adding such door
I totally disagree with this solution, as if it is accepted, then why it is not added clearly to the explanatory sketches at codes handbook for example

Jeff Evans

4/11/2024 07:33:36 pm

Finally. . .

https://www3.iccsafe.org/cs/committeeArea/pdf_file/BU_21_68_23.pdf

Ajay V Prasad

5/28/2024 11:38:32 am

Thanks Jeff, I was going to the same place. I'm still getting push back on your FI for the exit access door vs. egress path remoteness.

Kevin H Scott link

8/20/2024 11:12:17 am

I disagree with the requirement for the NOT AN EXIT sign. The dead-end corridor requirement only applies when more than one exit is required, which is also the same threshold for EXIT signs. So, the only time the NOT AN EXIT sign would be needed in a corridor is when EXIT signs are required. If exit signs are provided, then the occupant should be able to follow the EXIT signs and not need the NOT AN EXIT sign. The only time the EXIT signs would not be visible is when the corridor is filled with smoke, and if that is case, the NOT AN EXIT sign is not visible either. Therefore, the point of the cross corridor door is when the occupant is following the wall in the smoke-filled corridor and reaches the door at the end of the dead-end corridor, he/she is forced to turn around and follow the corridor back. They were not able to read the sign back at the point where they made the wrong turn because the corridor was filled with smoke. In my mind, yes the sign is correct, but it will never provide any value because the assumed need is flawed.

8/20/2024 11:36:16 am

Kevin - thanks for sharing. The "Not An Exit" sign was something required by that specific AHJ, but it is not part of the ICC committee interp. I wouldn't suggest it unless you have an AHJ specifically requiring it.

Warren Ducharme

8/20/2024 01:31:25 pm

FROM RI Staff Architect.
This solution is used to subvert a Specific (dimensionally specific not performance based) code requirement and is totally unacceptable.
WHY Two Short Dead END corridors are still a DEAD END.
Possible Compliant method the additional 24+ feet would be worthy of consideration IF the first segment END of Corridor served only the end space. (Relocate the first space door to the compliant corridor and remove the end of second corridor door and the ell wall segment to make the last END of Corridor door the entry into and from the last space) ALL Other common Path and travel Distances must of course comply. COMPLIANCE IS THE ALTERNATIVE. Resistance is futile.

8/20/2024 01:43:38 pm

Warren - acceptance of any design is ultimately up the AHJ. As a general concept though, the ICC Interpretation indicates that this approach meets the intent of the code for dead end corridors.

The Building Code Blog

Updated: An Approach to Mitigating Dead End Corridors

New Development: An ICC Committee Interpretation

Dead End Corridors: What Do You Think?

Leave a Reply.

Categories

Sign up to receive Building Code Blog Updates

Archives

Home

About

Blog

Contact

PRIVACY POLICY