Ductwork in Exterior Wall
I know, don’t do it. Here is my problem. I am renovating my daughter’s house. She wanted a bearing wall removed and I knew that two supply ducts would be an issue. They are 6″ ducts running from the basement to two second floor bedrooms. Furnace is near the center (front to rear) of the basement at the end of the basement. Bedrooms are above this end of the house.
The 6″ round come off the main plenum and within a couple feet convert to 6″ oval (which I think gives airflow more like 5: round). This goes up the center 2×4 wall and then turns and runs 12′ to the front and to the rear of the house where is meets baseboard type vents. The velocity at the vents seems fine, but the daughter just bought the house and has not lived in it yet, so I don’t know how heat or AC is.
I proposed enclosing 6″ round ducts in the front and rear corners, but she did not like that option. Because of floor framing clearance at the foundation and other issues this would also require use of a number of elbows and significant framing changes. Besides the work involved, I think the elbows, etc, would more than offset any improvement gained by using 6″ round ducts.
My present thought is to use wall stack duct, The exterior walls have Polyiso wall sheathing (I’m guessing at least 1/2″, but the wall seems a bit thick, so maybe it is more. I am thinking about using 2-14″ wall stack so that I do not have to hack into the band joists and so that I can place about 1″ of foam insulation behind it. I will also foam at cavity joints and adjacent to the ductwork. I have not found much information in 2-1/4″ wall stack ducts, but I’m thinking that 12″ to 14″ wide ductwork will probably give me at least equal and probably better airflow than the oval ductwork. I would also be eliminating 1 oval elbow. At the second floor I need to jog the duct over 3-1/2″ to meet the vent. This is about a 15 degree angle and then the duct terminates directly at the vent so essentially I eliminate another elbow.
Sorry for the long post. Any thoughts on airflow or efficiency in the exterior wall?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Climate zone or ZIP code = ???
Are the second floor rooms on a separate zone?
Sorry, suburban Philadelphia (Climate zone 4). Entire house is one zone, approximately 2100 s.f. 3 ton AC and 90+ % gas furnace. 1985 house with R-13 fiberglass in the walls plus the Polyso. I am air sealing any wall cavities plud band joists at the kitchen soffits that are open during renovation (which may only be about 10%) plus will be significantly improving air sealing of return air and supply ductwork. Based on a quick look in the attic it looks like someone added a foot or more of cellulose insulation on top of what had been probably R-19 original insulation (I did not dig though the cellulose, but I'm guessing the original was fiberglass).
Heat loss out the walls from 120F air in heating supply ducts would be considerably higher than from an average stud bay, even if you added more foam to the exterior side, and when the system isn't running you will have a convective loop drive through those ducts due to that heat loss.
If there is a gas-fired water heater you could also run hydronic micro-zoned heating to the bedrooms, and figure out something else for cooling (maybe a single 1-ton mini-split, or a pair of half-ton window shakers- depends on layout and actual loads.) It's a lot easier to find room to run heating plumbing completely inside of conditioned space to the second floor than ducts.
For the amount of effort it takes might make even more sense to install a 1.5 ton 2-head multi-split or a pair of 1/2- 3/4 ton PTHPs micro-zoning the bedrooms and letting the existing mechanical serve first-floor only. The average overnight low temp is still at or above 25F in the Philadelphia area even in January, which is the temp below which most PTHPs switch over to resistance heating. A 2-head multi-split or a pair of 1/2-3/4 ton mini-splits would be more efficient, but would also be a lot more money. If you went this route you may have to add a bypass duct to keep the flow rates high enough on the cooling coils to keep them from icing up, but maybe not, depending on the total duct design.
If the stud bay is 12 ft^2, you have a delta T to outdoors of 100 degrees when the furnace is running, and the insulation outside of the duct is R=4, then you would be losing 12*100/4 = 300 btu/hr while the furnace is running. With two such ducts, you would be losing 600 btu/hr. If you have a 60K furnace, that would be the equivalent of 1% efficiency loss. The penalty would be much worse if not for the presence of at least some exterior insulation.
So if it's a 30K furnace its a 2% efficiency hit, and we should worry about it more?
Hopefully they don't have a 60K furnace for a 2100' house, which has a likely heating load under 30K, maybe under 20K. But they could easily have a 120K furnace, given how things were often sized in 1985. That would make it a 0.5% hit, so we should care even less?
The lower-R wall & the duct convection when the furnace is not running adds up too. The generic problem is that it's moving the building envelope efficiency in the wrong direction.
The subtler point is that the heating & cooling loads of the top floor don't often track well with those of the first floor, making single zone operation both a comfort and efficiency problem. Micro-zoning it with right-sized equipment can help the overall efficiency beyond what a mere nameplate AFUE analysis would imply. It'll be more expensive than banging out more sheet metal, but it's an opportunity moment for making it a more comfortable, livable, and more efficient house.
Dana, you brought up increased heat loss as a problem with the proposed ductwork. One cause that you mentioned is increased delta T when the furnace is running. Quantifying that problem seems appropriate before advocating a more expensive alternative. A 30K furnace would run roughly twice as much as a 60K furnace, so the problem would be worse. I agree with you that, although the problem may be at its worst when the furnace is running, the problem doesn't go away the rest of the time. The original poster did not mention having a problem with temperature differences between floors.
I have been so busy with the renovation work I have not even looked at the furnace size. Around here about 75,000 BTUH would be common. I know that would be oversized, but this is a tract house and therefore not the most efficient construction.
Separate systems for the two bedrooms is not in the budget.
I don't know about temperature difference on each floor, since no one has lived in the house since the purchase. It is a shady site so I don't expect it to bee too bad. Virtually every house in this area is single zone, usually until well over 3000 s.f.
Between new foam insulation and the existing Polyiso sheathing I should end up with about R-8 behind the ducts. I think other ductwork improvements (i.e. sealing) and tightening the building envelope a bit during the renovations will more than offset the inefficiency of these two ducts. My bigger concern is whether the 2-1/4" wall stack will get at least equal and hopefully better flow to these rooms. I'm not saying that they need more flow, but 6" oval is more like 5" round in performance. That is common for most bedrooms around here, but it may be a contributing cause of the typical temperature difference between floors.
Mark,
There are a lot of issues here. As you correctly pointed out, wallstack adds more resistance (added static pressure) to the duct compared to ordinary round duct. The other issue is the static pressure added by the duct fittings at the bottom of the wallstack and the top of the wallstack. (For more on this issue, see How to Move Air Quietly Through a Duct System.)
Finally, there's the fact that you don't have enough room in the stud bay to insulate this hot duct from outdoors. That's a lot of compromises.
I understand that this is a renovation project with a tight budget. Only you (or your daughter) can make this decision, but the system you describe isn't elegant or satisfying.