NFPA 13 and Sprinklering Wood Trusses in Concealed Spaces

[Brainhurts]

Hi All,
I have been diligently researching a seemingly straightforward issue related to NFPA 13 and the sprinklering of wood trusses within concealed spaces. During my search, I came across several resources and articles, none of which got to the point.
I’ve got a 1-hour rated floor assembly using wood trusses, very similar to UL L521: https://www.woodworks.org/cad-revit-details/light-frame/1-hour-truss-floor-assembly/

Despite my efforts, I am still grappling with a key question: Can fire-stopped wood trusses (utilizing the 160 cubic foot rule) be employed to avoid sprinklering within concealed spaces? I have encountered hints suggesting that specific exceptions apply to trusses but NFPA 13 sections that address this are vague. If this is the case, I’m puzzled as to why TJI construction appears to be favored over fire-stopped wood trusses. 2x wood trusses are known to better withstand fire than TJI trusses with OSB webs.

If you could provide any guidance or direct me to additional resources, I would greatly appreciate it. Thank you for your time and assistance.

 

[mtlogcabin]

2x wood trusses are known to better withstand fire than TJI trusses with OSB webs.

However the heat deforms the metal press plate connecters very quickly resulting in a truss failure.

The disadvantages of a metal gusset plate’s behavior while being exposed to extreme heat conditions from a fire include th fact that the galvanized metal gusset plate’s teeth do not penetrate the wood all that deep. The wood the plates are connecting, have a charring rate of 1/40 inch per minute. The truss plate’s teeth have a minimum depth of 3/8 inch. So during a typical fire, the wood will lose 3/8 inch of its body in approximately 15 minutes leaving the teeth of the metal truss plate completely exposed. This will then cause the metal truss plate to fall off allowing that joint to lose its integrity. Once a joint on a truss is compromised, the weight it previously carried will be transferred to the next closest joints. These joints in return can not carry the additional load while being exposed to the fire causing them to lose their integrity resulting in a complete failure of the truss assembly.

Fire Load Performance Characteristics Of Metal - 1362 Words | Bartleby

Free Essay: 5 White Paper: Fire Load Performance Characteristics of Metal Gusset Plate Trusses vs. Finger Joint Glued Trusses St White Paper: Fire Load...

www.bartleby.com

Would it be cheaper to fill the space with insulation?

 

[steveray]

The short answer is that you need to follow one of the numerous exceptions in NFPA 13 exactly.....With any big span, you would have to do every truss if I recall correctly....MT has a good solution, but watch out for the ceiling listing that may require something extra when you add insulation....

 

[Brainhurts]

This is really confusing - because in researching fire rated assemblies, all the I-joist options need a two layer gyp ceiling and all the truss joist options need only a one layer gyp ceiling. Ie. the I-joists don't cut it with a single-layer of gypsum in a real UL fire test.

Also SBCA refutes the plate issue:

Metal Connector Plates In Fires | Structural Building Components Association

sbcindustry.com

However it looks like it's still an active discussion since 1989???? this is an article from 2022?:

How They Fail: Gusset Plates - Building Construction for Firefighters

Sal Ancona examines the behavior of gusset plates during a recent fire and the lessons learned they provide about the tenuous structural integrity of some buildings.

www.fireengineering.com

30+ years, but we still make them the same way....? seems like throwing a few nails in there would help.... or intumescent paint on the plates?? thicker plates??? surely some innovation must be happening....

 

[mtlogcabin]

Also SBCA refutes the plate issue:

No they did not

"The fire will not cause the plate to pull or curl away from the joint. It is the load on the wood members that would cause this action."

Every report I have read indicates the failure of a metal plate truss is a result of the loading on the truss. Charring accrues and within 15 to 20 minutes it reduces the holding power of the metal plate once one fails those loads are transferred and added to other connections which may not be able to handle the additional loads resulting in a catastrophic failure.

 

[steveray]

Most of the truss plates I see these days have nails in them as well these days...Not sure if it is a fire thing or other.....

 

[Inspector Gadget]

I just dealt with this with a building using NFPA 13. Now, I'm using Canadian Codes for what a fire block can be - usually its things like solid lumber, plywood, osb, regular gypsum, etc. In this particular case the client was intending on using a ULC-tested insulation product (due to the requirement for sound attenuation), which under our codes has met the CSA standard required to serve as a fire block. All they had to do was run the insulation from the ceiling to the underside of the floor deck at periodic intervals (as in, run a batt vertically next to a web joist, with the understanding that the batts in the remaining part of the cavity would hold , or cut down a batt here and there to close the space.



It's worth noting, however, that the language says to measure "from the top of the batt insulation." The architect told the client to fireblock at every fourth joist because they had calculated the volume without factoring in the insulation. When I factored in the insulation, the spaces were vastly larger. (Think 12-13" tall web joists with two layers of batt insulation.

But wait, there's more.

Also worth noting in NFPA 13 is the following:


Note that this clause does not require insulation to go from ceiling to the underside of the roof deck; just the underside of the top chord. A standard 2x3 or 2x4 top chord is 1.5" thick, so if the insulation comes to 1.5" of the floor deck, everything is fine.

 

[steveray]

I just dealt with this with a building using NFPA 13. Now, I'm using Canadian Codes for what a fire block can be - usually its things like solid lumber, plywood, osb, regular gypsum, etc. In this particular case the client was intending on using a ULC-tested insulation product (due to the requirement for sound attenuation), which under our codes has met the CSA standard required to serve as a fire block. All they had to do was run the insulation from the ceiling to the underside of the floor deck at periodic intervals (as in, run a batt vertically next to a web joist, with the understanding that the batts in the remaining part of the cavity would hold , or cut down a batt here and there to close the space.

View attachment 13851

It's worth noting, however, that the language says to measure "from the top of the batt insulation." The architect told the client to fireblock at every fourth joist because they had calculated the volume without factoring in the insulation. When I factored in the insulation, the spaces were vastly larger. (Think 12-13" tall web joists with two layers of batt insulation.

But wait, there's more.

Also worth noting in NFPA 13 is the following:
View attachment 13850

Note that this clause does not require insulation to go from ceiling to the underside of the roof deck; just the underside of the top chord. A standard 2x3 or 2x4 top chord is 1.5" thick, so if the insulation comes to 1.5" of the floor deck, everything is fine.

We have this in our US UL:

17. Blanket Insulation

Unless specifically described in the individual designs, the addition of insulation in the concealed space between the ceiling membrane and the floor or roof structure may reduce the hourly rating of an assembly by causing premature disruption of the ceiling membrane and/or higher temperatures on structural components under fire-exposure conditions.

Insulation in G500, L500, M500 and P500 Series Designs — For 1-hour-rated G500, L500, M500 and P500 Series assemblies, fiberglass insulation, either loose-fill, or faced or unfaced batts or blankets may be added to the plenum or joist space above the gypsum board, provided an additional layer of gypsum board is added to the assembly. The gypsum board should be of the same type as shown in the individual designs. The base layer of gypsum board should be attached with the fastener type and spacing as described in the design. It is not necessary to tape the joints of the base layer. The finish layer of gypsum board should also be attached with the fastener type and spacing as described in the individual design. The length of the fasteners should be increased by a minimum of the gypsum board thickness of the additional layer. The joints in the finish layer should be finished as described in the design.

Other methods of adding insulation in the plenum or joist cavity are not permitted unless indicated in the individual designs.

 

[Inspector Gadget]

We have this in our US UL:

17. Blanket Insulation

Unless specifically described in the individual designs, the addition of insulation in the concealed space between the ceiling membrane and the floor or roof structure may reduce the hourly rating of an assembly by causing premature disruption of the ceiling membrane and/or higher temperatures on structural components under fire-exposure conditions.

Noted ....

FYI: This is a building where everything has to be constructed to a ULC-listed assembly (or on the Canadian "component additive" method), both of which take insulation is taken into account. Insulation was *required* for sound attenuation.

But it's a good reminder that if the assembly didn't include insulation, that it would have had to have been specified differently by the designer.

 

[steveray]

Noted ....

FYI: This is a building where everything has to be constructed to a ULC-listed assembly (or on the Canadian "component additive" method), both of which take insulation is taken into account. Insulation was *required* for sound attenuation.

But it's a good reminder that if the assembly didn't include insulation, that it would have had to have been specified differently by the designer.

I looked at the Canadian version and it seemed scarily sparse....

 

[Inspector Gadget]

I looked at the Canadian version and it seemed scarily sparse....

There are a lot of UL assemblies that are also ULC tested. Plus we also have what's called a "component additive" system that can provide a fire-resistance rating based on providing a time to each bit of an assembly.

Further, for modest construction, there are umpty-gazillion accepted generic assemblies that are essentially tables in the Code.

 

[steveray]

There are a lot of UL assemblies that are also ULC tested. Plus we also have what's called a "component additive" system that can provide a fire-resistance rating based on providing a time to each bit of an assembly.

Further, for modest construction, there are umpty-gazillion accepted generic assemblies that are essentially tables in the Code.

I was speaking mainly about the general notes we see in the beginning I didn't see this there:

 

[tmurray]

I was speaking mainly about the general notes we see in the beginning I didn't see this there:

View attachment 13988

A lot of that is subject to the code or ULC-S101 for us.