Hi Kent, educate me?
Since air doesn't always move up through the ceiling plane, stack can be zero or reverse, what is the sprinkler relying on to trigger a response?
The 2 flush cover pendant style on the right side of this picture are the units we are struggling with and most often found in residential applications...
The trigger would be the heated air these units are relying on... the airway openings allow the temperature to change quickly enough to release the pendant cover and activate the sprinkler heads. Thus some air passage is necessary on these "hidden" sprinkler heads. The pendant covers also have a temperature release mechanism which would separate the cover from the head unit. Basically they are reliant on stack effect and temperature.
The other heads on left side of picture generally (some still need air flow) rely mainly on temperature - they are fully exposed, and the typical home owner does not like the looks of them.
So actually the temperature difference they need to move the air, call it convection or stack, is from the fire. That makes sense and would easily override the lesser temperature differences of house stack effect.
I have had some dealings with my local and state fire marshal and found the line of communications very open and beneficial. After all, they are the final word and they would certainly be aware of the problem.
Thanks for the reply.
There are two easy solutions - move the air barrier location or placement of the head & line as shown in this piece... http://blog.sls-construction.com/2013/bs4d-the-fire-sprinklers-air-...
Very nice Sean! Thanks!
Sealing around the bottom edge of the edge of the cover plate violates the (UL & FM) listing of the product, and it specifically violates NFPA 13, NFPA 13R, NFPA 13D, NFPA 25 and the soon to be out 2015 IECC. The sprinkler standards (NFPA) do not permit paint or caulk on sprinklers and their escutcheons or cover plates. The problem with sealing the bottom edge is that it will delay the activation of the sprinkler. Residential sprinklers must activate quick, before the room temperature becomes untenable to the occupants. By caulking the bottom edge, the air cannot enter and melt the solder connection of the cover plate to the cup and allow the cover plate to fall off, exposing the sprinkler deflector and bulb. Caulk may not only retard the activation times, but it may deflect the spray when the sprinkler is activated.
We understand that, and it makes perfect sense. That is not the issue. We are not talking about sealing the cover plate. It's the housing where it meets the sheetrock - not the escutcheon or the cover plate. On a rare occasion the sheetrock installers can make it a tight fit and everyone is ok with that - nobody complains. The issue is when the sheetrock installers do not get a tight fit (most of the time) - on a tight house, it can greatly increase the air loss to the house especially when there is 30-40 heads in the house.
We completely understand that the unit needs the airflow around the cover and thru the interior of the unit. The cover plate would still have the intended gap, the openings for air flow in the unit itself would not be sealed - just the perimeter of the housing where it meets the sheetrock...
I understand now. You are speaking of the annular space between the cup (housing) and the sheetrock. In the perfect world, there is tight fit but we both know how that works 99.9% of the time. Can this area be filled with caulk or compound? I would still have to say no, because in the strict sense of the listing, this cup is part of the listed assembly and the sprinkler standards prohibit this as explained above. It may, I stress may, be permitted by the manufacturer to insert a foam backer rod in this annular space to seal this space. The foam backer rod would not have the physical attachment to the cup as paint, caulk, compound would have, and there would be little chance of the backer rod interfering with the operation of the cover plate and sprinkler.
I am in Sea/Tac area (Fife) all next week teaching. If there is an opportunity to discuss this further or to see some air testing I'm game.
This is not the only problem with sprinklers. Any pipe in the attic with water in it must be protected from freezing, and the standard practice in the mid-Atlantic has come to be called "tenting". The insulator will cover the pipe with a 24" wide R-30 batt draped evenly over the pipe, then blown material is installed over the rest of the ceiling and up onto this batt.
On a cold day, these pipes can be seen with IR from inside the building, indicating a confused thermal envelope exacerbated by the air leak at the head.
We have solved this problem in retrofits by raking back the blown material, removing the "tent", bending a piece of 1/4" styrofoam fan folded material over the pipe, and sealing at the drywall with single part foam. Insulation is then installed over the completed assembly. Done properly with extra attention at the heads, you can eliminate the majority of the air leakage. This is very expensive and I don't advocate doing it in all cases.
The only real solution to the 2 problems - the air space and the air leakage - is to have the entire system installed inside the house with wall heads. Nothing allowed in the attic. The installer never makes the argument that the warm air can't get to the head if it is on the wall. We have yet to find an installer who will do this without kicking and screaming.