Second floor too hot in summer

[PP_]

Hi All,
I am located in Nashville, TN. Mine is a common problem mostly seen with
a little older homes "second floor is hot and uncomfortable in summer",
even with A/C running in full blast.

My home:
Is a 1 1/2 level home (split level). Almost 1700 sq.ft. AC is 3 ton, with a trunk and branch
duct that connects to 10 registers in the downstair rooms, and three registers to the second floor rooms. This is shown in my first figure, with supply ducts (plain) and the return duct as crossed.

There is only one return grille 18" x 25", with a 16" duct that runs almost 30 feet to
the air handler. In the second figure, I have tried to show you the location of the
return grille. It is on the ceiling of the stairs, almost a central location. There are no return
grilles in any other locations, unfortunately.

My problem:
My second floor rooms are uncomfortably hot.

I thought I would do a little research and gather the options that I have on what could be done.
One thing I realized is that the hot air in the second floor rooms is
stagnant, not able to reach the return grill. The location of the return grill is on the
ceiling of the stairs (see my second drawing), almost 5 1/2 feet below the second floor ceiling.

My question is if I add grills into the rooms on second floor to 8" flex, connect them
to a 12" flex and bring it a feet below the return grill, would this help.
Since it is hot air in the second floor, it wouldn't move down towards the return, correct?
To push the hot air through the duct, could I use a inline fan? If I am
using an inline fan, where would you recommend the position of inline fan? Other options??

Kindly comment on what options I have to resolve my situation. I might have
missed out details, let me know and I could provide them.
Thank you for your time, and for sending in your comments!

 

[Dana]

Any doored-off room with a supply register needs a return path that still works when the door is closed. For most systems a door-cut big enough to move that much air would be egregiously large, but using the stud bay of a partition wall as a "jump duct" is a common solution:

Cut in a grill near the floor on one side of the wall, and near the ceiling on the other, both roughly the same number of square inches as the supply register. Of course you can use flex duct for jump ducts too.

Flex duct that is just flopped in place with large sags and turns has much higher duct impedance than spec, which cuts into flow. Flow can be improved pretty dramatically by stretching it tight for all straight sections, and keeping the turn radius long (at least a couple of duct diameters) where it must turn.

Ducts and duct boots that leak are often diverting 20% or more of the air to somewhere other than intended, which also be improved. Use duct mastic (not tape) on all joints and seams, and seal the duct boots to the ceiling/wall /floor around it's perimeter with polyurethane caulk (which never gives up) &/or can-foam for big gaps. It's sometimes necesssary to reinforce the cualking with a lap of housewrap tape, since the adhesion to just the gypsum (rather than paper facers) isn't reliable long term.

 

[Sponsor]

 

[PP_]

Thanks, Dana.
Thanks for your reply. When you use a stud bay as a "jump duct",
the second floor rooms (the rooms with hot/stagnant air) needs a grille near
the ceiling and the other near the floor, correct? Say, the duct opens to a hall that
meet the stairs, do you think the air would flow through the jump duct?
What makes the hot air in the room flow through the "jump duct"
downwards? The ceiling is ~8 ft high.

Thanks

 

[Dana]

Of course the other half of dealing with the problem would be to reduce the heat gain on the second floor. Comfort problems are usually primarily due to deficiencies in the building envelope, with the mechanical systems being secondary. Both are important, but the more you can do on the building end the less cooling is required.

Upper floors tend to have less shading factor on windows, and if the insulation in the attic isn't fully up to snuff it's worth air-sealing the attic floor and piling on cellulose over whatever pre-existing insulation there might be. Nashville is in DOE climate zone 4A, where the IRC 2015 code minimum for new construction is R49, or R38 if the R38 is fully out over the top plates of the exterior framed wall at the eaves. R49 would be about 13-14" of combined insulation depth, R38 would be 10-11". In older homes, even with 2 x 12 joists the roof deck is often too low to get the required 1" clearance between R38 insulation and the roof deck fully over the top plates of the exterior walls. Most older home have less than R25 in the attic, and it makes a real difference.

Fiberglass (particularly low density fiberglass from a decade or two ago) is somewhat translucent to infra-red radiation. In a vented attic the temperature of the fiberglass an inch or two below the top of the fiberglass can be several degrees warmer than the attic air, tracking the temperature of the roof deck. It's also not very air retardent, and can underperform during the heating season too. As little as 3" of cellulose blown over the top will fix both problems.

But it's important to find and fix all of the air leaks at the upper floor ceiling/attic floor level before adding any insulation. There are usually fairly large often less obvious air leaks from flue & plumbing stack chases, electric chases to find and fix.

If the windows are single pane, a low-E exterior storm window can cut the solar gain by more than half, and will also make a difference in the winter performance. The hard coat low-E coating on surface #2 (the interior facing side of the storm window) is a cost adder over clear glass storms, but pays back in lower energy use quicker than clear glass despite the upfront cost.

A cheaper option would be heat rejecting window film applied to the interior surfaces of the windows, but that would only help with the cooling load, and would also reject wanted solar gain in the winter. Nashville is still in a heating dominated climate, despite being in the steamy southeast.

If you're the type who REALLY wants to chase down and fix the duct balance and air flow situation, buy a dual port manometer and start measuring things. An Energy Star well balanced duct system isn't fully commissioned until the room-to-room pressure differences are all under 3 pascals (0.012" of water), under all conditions, doors open/closed, at any air handler speed. Most inexpensive (<$100) hand held manometers only have a resolution down to 0.01", and aren't well calibrated, but they're still good enough to find the worst-offenders. When measuring at the bottom of it's absolute range and resolution it's worth measuring in both directions, since zero may not really be zero. But it would tell you right away which rooms need the jump ducts the worst, and which ones are probably OK, assuming the supply duct isn't disconnected or leaking or too high impedance due to a twisty-turny-saggy installation. A room to room pressure difference north of 0.03" is an energy & comfort disaster, using a return duct called " the great outdoors" for some of it's return path, pulling in quantities of humid outdoor air into your air conditioned house. Most homes where the duct system wasn't actually measured & commissioned will have a few of those.

A cheap manometer is also capable of doing the diagnostics at the overall air handler and duct system too. (It's what they're designed for, really.) Air handler cfm are usually specified at a nominal 0.5" water, but some systems with high duct impedances and dirty filters can hit north of 1" by quite a bit. Doubling the static pressure cuts flow by about 15%, tripling the static pressure cuts it by about 20%. It might be worth your time to read through this 5 part primer before diving too far into modifcations of the duct system.

 

[Dana]

Thanks, Dana.
Thanks for your reply. When you use a stud bay as a "jump duct",
the second floor rooms (the rooms with hot/stagnant air) needs a grille near
the ceiling and the other near the floor, correct? Say, the duct opens to a hall that
meet the stairs, do you think the air would flow through the jump duct?
What makes the hot air in the room flow through the "jump duct"
downwards? The ceiling is ~8 ft high.

Thanks

The pressure induced by the supply register moves the air up/down/sideways (including outdoors) by whatever paths are available, delivering orders of magnitude greater pressure than stack effects or convection buoyancy from to temperature gradients.

Supply-register-only situations pressurizes the room, and like making a hole in a balloon or bag, it doesn't matter what side the exit hole is on. A jump duct puts the escape hole where you want the air to move, allowing much higher flow and lowering the overall pressure in the room.

If your goal is to remove hotter air near the ceiling out the jump duct the grill on the room side of the partition wall needs to be near the ceiling. It could be just set up as a transfer grille, the grill on the other side being near the ceiling too, but that cuts into privacy and light-leakage issues. A 6-7'
section of a 2x4 stud bay cuts light transfer almost completely, and cuts sound by at least a few dB.

It's pretty common to use ceiling grilles on both sides of the wall and a short section of flex as jump duct, but you don't want to be making more leak paths through the insulation layer in the attic, adding to parasitic load, but it's also sub optimal if the supply registers are all at the ceiling, since it reduces mixing with the room air.

For best results the jump duct would be placed far from the supply duct, to guarantee that there isn't a short-circuit making most of the flow directly from the supply register to the jump duct register. If the supply registers are at the ceiling, put the room-side jump duct grill near the floor. That forces the cool dry air coming out of the supply duct to cross through and mix with the room air, rather than concentrating the flow near the ceiling.

 

[WorthFlorida]

My last home was two floors and had two AC units, one for each floor. Very common in Florida. After the roof was replaced the second floor was always much warmer because the ridge vent on the old roof allow more attic ventilation than the new roof. You could actually feel the heat radiate down from the attic space and that was with 12 inches of blown fiberglass insulation. Home built in 1989. One solution was to get more ventilation in the attic to expel as much heat as possible. But do take Dana's recommendations.

 

[PP_]

Thank you, Dana.
Quite informative... A lot to go through and digest slowly

A little bit on the locations of my supply registers.
As shown in my attached picture, the supply registers are represented almost
in positions as per the drawing (A, B and C second floor registers are towards the outside walls,
close to the floor). ALL the first floor registers are from the ceiling.
In the kitchen/dining/living/entry/bath (first floor) the register locations are
almost 2/3rd from the inside walls. In the den/bed1 (first floor) it is very close to the
inside walls, almost within a feet or less from the supply trunk.

A, B and C second floor registers are towards the outside walls, near the floor,
and there is a hallway connecting the rooms. I would upload a drawing later this
week...
Thank you!

 

[PP_]

My last home was two floors and had two AC units, one for each floor. Very common in Florida. After the roof was replaced the second floor was always much warmer because the ridge vent on the old roof allow more attic ventilation than the new roof. You could actually feel the heat radiate down from the attic space and that was with 12 inches of blown fiberglass insulation. Home built in 1989. One solution was to get more ventilation in the attic to expel as much heat as possible. But do take Dana's recommendations.

So, your new roof is without a ridge vent? You agreed to modify the existing one?
Thanks!

 

[SteveW]

Maybe I missed it, but do you have a basement, and thus your furnace/indoor AC unit is there?
If so, that adds an 8' column of heavy, cold air to push up to the 2nd floor.

This is an issue where I live. Getting cold air from the basement up 2 floors -- not that easy. We have to rebalance our supply ducts every spring and fall.

 

[PP_]

Maybe I missed it, but do you have a basement, and thus your furnace/indoor AC unit is there?
If so, that adds an 8' column of heavy, cold air to push up to the 2nd floor.

This is an issue where I live. Getting cold air from the basement up 2 floors -- not that easy. We have to rebalance our supply ducts every spring and fall.

No, I do not have a basement. My heater/airhandler is in the garage (next to kitchen/living),
and my AC unit outside. It is a 1 1/2 level or split level home.
Thanks!

 

[Reach4]

You also have the option of adding a minisplit to the hot area.

 

[Dana]

Thank you, Dana.
Quite informative... A lot to go through and digest slowly

A little bit on the locations of my supply registers.
As shown in my attached picture, the supply registers are represented almost
in positions as per the drawing (A, B and C second floor registers are towards the outside walls,
close to the floor). ALL the first floor registers are from the ceiling.
In the kitchen/dining/living/entry/bath (first floor) the register locations are
almost 2/3rd from the inside walls. In the den/bed1 (first floor) it is very close to the
inside walls, almost within a feet or less from the supply trunk.

A, B and C second floor registers are towards the outside walls, near the floor,
and there is a hallway connecting the rooms. I would upload a drawing later this
week...
Thank you!

In doored-off rooms with supply registers near the floor, on the room side put the jump duct register or transfer grille near the ceiling, as far away from the supply register as is practical.

In doored-off rooms with supply registers at ceiling level, the jump duct register or transfer grille goes near the floor, as far away from the supply as practical.

Rooms with supply ducts that have no doors blocking the path to the main return do not need jump ducts or other return path enhancements.

Partition wall cavity jump ducts and other enhanced return path possibilities are laid out fairly clearly in this document. The following images are all from that site.

Even if the return path grille can't be placed optimally, having at least SOME return path still improves flow and mixing.

I can't stress enough how important it is to make flex duct runs as tightly stretched as possible. HVAC contractors swear by them for the ease of installation and the ability to get around obstacles, but just letting it flop in place or make tight turns can cut the flow by a large fraction due to the increased friction/turbulence of the less-than-smooth interior surface of unstretched flex. There is SO much poorly installed flex duct in residential construction that some are almost ready to ban it.

 

[Dana]

For the record, are the same ducts used for both heating and cooling?

If yes, optimizing the system for room to room, floor by floor temperature balance for cooling almost always results in some amount of temperature imbalance for heating and conversely. If there are already balancing vanes for tweaking the flows they might be currently optimized for heating, resulting in cooling season comfort issues. Top floors of 1.5 story houses usually have more problems with this than full 2-story houses, but all multi-story houses buildings will have this issue to a greater or lesser degree.

 

[PP_]

Thanks, Dana.
The same ducts are used for heating and cooling. I will check whether there are dampers/vanes
in the supply trunk. The registers have dampers, do you recommend tweaking them?
For the dual port manometers, are there brands/models that are better than the $100 ones.
I though I could look for a used one that is a bit more sensitive. Have you ever used
Magnehelic Pressure Gauge Max Pressure 15 PSIG? [This is from the link you have
provided - energyvanguard]. Would this work for my measurements?

You mentioned that the flex ducts needs to be as stretched as possible; this is with the available
insulation, correct. I have not bought any ducts, I would gather as much info as possible
before starting my work. (I am managing with ceiling fans and portable AC unit for now).

Thank you very much for sharing your knowledge )

 

[Dana]

Recommissioning/commissioning the equipment that's already in place is always the cheapest and first step, rather than "replace and hope".

Start by stretching your ducts tight, easing any sharp turns or kinks, and sealing all the duct connections to air handlers & register boots, sealing the seams of the boots with duct mastic, and sealing the boots to the sheet rock, etc.

Rooms that have no obvious return paths other than the door either need to keep the door cracked a couple of inches (a simple eye-hook latch can be handy), or create a return path with a jump duct or transfer grille, etc.

With the obvious stuff like that fixed you can then start fine tuning with a manometer. I can't really recommend one low-end manometer over another, but ANY manometer will be better than just shooting in the dark. I haven't used the Magnehelic Pressure Gauge, but would trust the Energy Vanguard guys to tell you what it's best suited for, or what can/can't be done with it,

 

[Jadnashua]

If you have access to the attic, add some more insulation. One thing I did that had an immediate comfort level increase was to add a radiant barrier to the underside of the roof rafters. Note that that can increase the heat of the actual roof deck some, as it would then be getting heated from both the top and from the bottom by reflecting that heat back verses into the attic space. Prior to doing this, you could feel heat radiating from the attic into the ceiling. Immediately after stapling the stuff up, the ceiling became the same temperature as the interior walls, decreasing the heat load on the structure. The attic temperature dropped over 20-degrees on a sunny day (I haven't measured it recently).

I also blew in additional insulation, but that was after the radiant barrier by a few years. The evidence is obvious in my row of townhouses...my roof will have snow on it in the winter for nearly a week longer than those next to it. In the winter, it helps keep any heat radiated up into the attic from reaching the roof deck (as does the insulation).

If you're in the market to reroof, consider using a metal roof with a profile so that there's an air gap underneath it. One of those that looks sort of like a shake will have nearly 3/4" of an air gap underneath. The metal will reflect over 90% of the incoming IR energy, keeping the roof deck considerably cooler, and thus, the attic as well. A side benefit is that most of them have a free, life-time warranty, some for both labor and materials.

 

[Dana]

I

I also blew in additional insulation, but that was after the radiant barrier by a few years. The evidence is obvious in my row of townhouses...my roof will have snow on it in the winter for nearly a week longer than those next to it. In the winter, it helps keep any heat radiated up into the attic from reaching the roof deck (as does the insulation).

Doing it in the reverse order makes more sense. Radiant barrier is only economic if the insulation on the attic floor is very low (6" of fluff or less), and the ducts aren't insulated. Current IRC code min in TN is R49 (about a foot of fluff) and R-8 for ducts ( a couple inches). But even R30 attic insulation and R4 flex is enough to render radiant barrier uneconomic.

And spending the money on air sealing the ducts comes first followed by air leaks into basement or crawlspace (any big leaks at the the bottom of the house) and the attic floor comes WAY ahead of any insulation upgrades.

In heating dominated climates radiant barriers increase heating energy use, due to the loss of the passive solar gain through the roof. A warmer attic in the fall and spring when there's still a heat load is a good thing.

 

[Jadnashua]

The air temp in the attic, year round, is more moderate now. In the winter, any heat lost from the house gets reflected back, but the solar gain would be greater. It's easier to keep the first and second floors close to the same temperature now verses before the radiant barrier. The additional insulation helped a little bit more. Originally, the attic had about 6" of blown-in cellulose. Long time ago, I'd put 6" fiberglass on top of it. THen, later, added the radiant barrier (cheap, quick, clean to install). As the prices of energy kept going up, I then added more cellulose. But, keep in mind, insulation only slows the transmission of heat...a radiant barrier does a pretty good job of literally stopping it if it has a proper air gap. That heat does get to the roof deck from both the top and bottom, but if you compare my roof to those townhouses adjacent, there's no obvious difference in condition to at least the shingles over the 20-years that roof has been installed.

 

[PP_]

Recommissioning/commissioning the equipment that's already in place is always the cheapest and first step, rather than "replace and hope".

Start by stretching your ducts tight, easing any sharp turns or kinks, and sealing all the duct connections to air handlers & register boots, sealing the seams of the boots with duct mastic, and sealing the boots to the sheet rock, etc.

Rooms that have no obvious return paths other than the door either need to keep the door cracked a couple of inches (a simple eye-hook latch can be handy), or create a return path with a jump duct or transfer grille, etc.

With the obvious stuff like that fixed you can then start fine tuning with a manometer. I can't really recommend one low-end manometer over another, but ANY manometer will be better than just shooting in the dark. I haven't used the Magnehelic Pressure Gauge, but would trust the Energy Vanguard guys to tell you what it's best suited for, or what can/can't be done with it,

Hi Dana,
It has been a while, got busy during the last few years.

I got started with fixing the air leaks at the lower floor ceiling.
Started sealing off with duct mastic where there are connections, plumbing lines,
electric fixtures and so on. The way I did is to get cellulose insulation scooped out,
did vacuum to clean up and apply mastic. Once complete, I put back the cellulose
insulation back in place. Cellulose has been compacted (?) to some extend due to
age/gravity(?).

For electric fans and fixtures, do you recommend metal enclosures, or could I do
a plywood box with a 1-2 inch room on all sides of the fixture? My other question
is when it is wood or plywood, does mastic tape adheres to them well? Or do you
recommend any other sealants when it comes to wood/plywood?
Thanks!

 

[Fitter30]

Had a 1850 sq ft two story 3 bed 2.5 bath. Beds all 2 fl in st.louis. One 3.5 ton unit. Installed a 2 ton air hander in attic return in the hall ceiling single grill in two bed and 2 bathrooms and 2 in the master. Didn't need heat upstairs just cooling for 6* difference. First month electric bill dropped $50 in 1988. Run the 3.5 during the day, night the 2 ton. Upstairs was oversized but 40 years ago that was the smallest split. 5.5 ton was great for entertaining 90*+ twenty five people no problem.