At wits end with upstairs/downstairs heat transfer
![](https://secure.gravatar.com/avatar/bb1d5de3eccf59853ccb564ffc4dbc05?s=52&d=https%3A%2F%2Fimages.greenbuildingadvisor.com%2Fapp%2Favatars%2F200x200%2Fa.png&r=g)
If this is the wrong category, please move to the correct one.
For 1.5 years, I have struggled with what I think is the most a$$inine design of a heating/cooling system ever devised. Not only has it been the cause of much financial misery, but also vexing frustration. And since I am wits end, I am about to institute some sort of jury-rigged contraption to stem the tide.
We live in the foothills of South Carolina (winter lows in the teens and summer highs near 100, USDA Hardiness Zone 8a-7b) and we rent a 2 story 2100 sq ft 4 br 3 ba on-slab house with the 2 air handlers in the attic, two 14×20 intakes, two units on the outside of the house and two thermostats inside (downstairs centrally located and upstairs centrally located right at the stair landing).
The root cause of the problem that allows the transfer of heat in the winter and cool air in the summer is the open floor plan around the stairway which probably measures 10 wide by 18 long, thus allowing unrestricted airflow between the two floors. As if that was not bad enough, two of the registers are too close to the unrestricted up-down airflow corridor. One is in the ceiling 3 feet away (with no archway between it and the up-down corridor. The other…and this is where it becomes a$$inine… is on the vertical wall of the up-down corridor about 2 feet above the lower ceiling height.
In the summer, we struggle with the upper unit running non-stop as lower level heat rises and upper cooling sinks. I have to put a sweatshirt on in the morning to go downstairs and get breakfast while upstairs is mildly uncomfortably warm.
In the winter (as in last night (21 deg F)), we struggle with the lower unit running non-stop, with cold upper level air sinking and newly generated heat on the first floor zipping right upstairs through the corridor preventing the upper thermostat from ever turning on. Unfortunately, the door archways of the upper bedrooms and bathrooms prevent some of this downstairs generated heat from warming the bedrooms. With the upstairs thermostat not stimulated to turn on ALL night, the bedrooms become quite cold.
In the end, we have increased utility bills, frustration with imbalanced temperatures, and either unit getting run no stop for 3 months at a time in their due season.
A few weeks ago, I had closed off the 2 offending registers (which were basically only broadcasting newly generated heat right into the up-down corridor of air). That helped a little, even though they still leak a good bit of airflow. But the real problem is still with 4 other registers pumping hot air at first floor ceiling level and basically a 180 sqft hole leading upstairs and so the problem still persists.
And now, I feel it is time for desperate measures and my wife is not happy about my proposed plan. My plan is at the top of the stairwell landing (which is easier to do than the bottom) is to install a curtain track in the ceiling and have a retractable curtain that can be slid around the upper stairwell landing, similar to those curtains in emergency rooms to separate patients. With a thick enough curtain, I think a good amount of airflow could be prevented, with some loss around the bottom and especially the top near the curtain rings.
Before I go down this path, any other thoughts? FYI, moving is not an option this time. I want to say in advance that I appreciate the help of those with the knowledge to help me fix this.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
![](https://www.greenbuildingadvisor.com/app/themes/greenbuildingadvisor/dist/img/modules/detail-library/thumb1.png)
![](https://www.greenbuildingadvisor.com/app/themes/greenbuildingadvisor/dist/img/modules/detail-library/thumb2.png)
Replies
Part of the problem is stack effect air flow. So better air sealing of the house will help.
Better mixing (just after air leaves the registers) also helps.
I agree with Jon that air sealing would help. You can scout around downstairs for where the cold air is coming in from outside, perhaps assisted by a thermal camera (you can rent a great one from some Home Depots, or you can buy one that attaches to your phone), but best is to hire someone to do a "blower door test" to find air leaks. It's hard to get motivated to do that in a rental but it's worth considering. Some utilities will even do it for free.
You want to block leaks where cold air comes in downstairs and also leaks where warm air escapes upstairs. Typical culprits are recessed lights and unsealed attic hatches.
I think your curtain idea would help, but it would be no means provide complete air sealing, and if it makes your wife unhappy that's no good.
Stack effect is orders of magnitude smaller issue here than the poor layout of the registers. Any chance of moving some of them?
Changing the thermostat location's may get you half way there.
I agree that moving thermostats to the critical comfort locations is a good idea.
I think you may be dismissing the stack effect a little too quickly. We don't know how leaky this building is, but it's not unheard of for infiltration to be half the heat load in a leaky building on a cold day, if not averaged over the season. Maybe not the best estimate, but in the absence of any data, I'll choose a simple number. When driven by stack effect, that heat load all occurs on the first floor. If the other heat loads are distributed between the first and second floor, that means 3/4 of the heat load is on the first floor and 1/4 on the second. I don't think that should be ignored, especially given that it interacts with the poor distribution system layout to amplify the problem that layout causes.
Stack effect pressures in 2 story houses are miniscule compared to air handler pressures. Whether the house is leaky or not, the duct design is far more significant than the stack effect.
Absolutely, stack effect pressures are miniscule compared to air handler pressures. But that's comparing apples to oranges. The stack effect pressure drives air leakage, which is part of the heat load. The air handle pressure drives air distribution which is part of the heat delivery. Both heat delivery and load are important for temperature distribution but there's no reason 1 Pa pressure in one should have the same effect on temperature as 1 Pa pressure in the other.
Perhaps you thought someone was arguing that the stack effect would effect the heat supply. That's not at all what I'm arguing. Its effect on the load is what I'm talking about.
I second Dana...
Moving the upstairs thermostat, or replacing it with one like the Ecobee that can use additional wireless sensors (placed appropriately in the bedrooms), may go a long way to improving things. I would tackle the upstairs thermostat placement first as it could be a relatively cheap and effective improvement.
Are the returns both on the attic floor/upstairs ceiling? Or is the lower unit return ducted down to the first floor? For heating it's nice to have a return at floor level to draw the cold air off the floor where it settles.
You don't want to put your curtain between the return and the space served by the air handler.
IMHO, closing registers rarely improves things and can cause a slew of problems.
You may want to look into balancing your ductwork assuming it has adjustable dampers. There may be some low hanging fruit in easily fixable flaws that are hindering air flow. Poorly installed flex and leaky duct work are common problems.
Do you have some pictures of your attic (with the ahu's and ductwork?)
Here are some photos of the open air corridor. You can see the evil register (which belongs to the bottom unit). While I am sure the house is not as airtight as it could be, I think the cross section of this top-down air corridor (it is actually 10x20 on the 2nd floor) is magnitudes higher than the cumulative total of the air leaks.
It might take me a day or two to get up in the attic for photos.
View 1
View 2 - if you click on photo, you can see the evil register towards the left edge of photo
Proximity of thermostat to put-down air corridor.
And I second Josh, especially his comments on the returns. Where is the downstairs unit return located? Are the filters located at the returns or at the air handlers? Is the downstairs return filter new and serviceable? Has someone put a restrictive 1" media flow-stopper in there? If the filter is anything but a 99 cent see through fiberglass filter, buy one and test to see if there is some improvement. If you happen to get someone with a thermal imager and blower door, maybe you could go for the trifecta and find someone with a flow hood. They could get you flow rates of all the registers and returns. If that downstairs return isn't low in the building and if it's weak in the cfm it's returning, that is really hurting you. Good hunting!
GaryinKC, the downstairs filter is located in the first floor ceiling about 6 feet from the up-down air corridor. It is not see through filter. But, the problem, at least in the winter, is that the lower unit is pumping out air fine (even against two closed registers which still leak a lot of air). And all this hot air goes upstairs and prevents upper floor thermostat from ever turning on.
I've seen large differences in comfort by increasing return flow rates. If you are supplying 600 cfm to the lower level and returning 400 at the return grille--could be drawing the missing 200 cfm through leaks higher in that looong return chase-then these conditions could be aiding mother nature's desire for the warm air to rise. Might not change much but I'd still spend the buck for a non-restrictive filter to find out. I'll echo the advice to seal sources of infiltration down low and sources of exfiltration up high. If it's not too hard to get into the attic and you promise to move about safely, you might poke around the air handlers and other penetrations for ways conditioned air might be streaming into the attic.
I’ll agree with Jon and Charlie that air sealing is a good part of the problem. A home with mechanicals in the attic might have poorly sealed penatrations at the ceiling vents, a loose piece of plywood over the attic hatch and a few leaky recessed lights and bath fans, and these leaks can have a good effect on air flow through the house. Doing what you can with gaskets and caulking might help.
I am sure leaks may be a problem. But considering that there is a 10 ft x 20 ft (cross section) short circuit for air between the upstairs and downstairs, I would think that would be orders of magnitude higher than lots of little leaks. Also, in the summer, my lower floor is an icebox without the lower thermostat turning on at all. It is all, I suspect from the cold air being generated upstairs falling right away down the up-down air corridor.
Have you done a blower door test?
I agree that the large indoor convection column isn't helping, and that stack effect small leaks aren't an issue but AIR HANDLER driven leakage can be quite large through even small leaks if the ducts aren't balanced or the ducts are leaking into/out of the attic.
Odds are pretty good the air handlers are ridiculously oversized for their zones too, and can move a LOT of air. Any duct or air handler leakage into a vented attic will create pressure differences between indoors & outdoors, driving massive amounts of air even through leaks that would be negligible from a stack effect point of view.
An Energy Star duct system would need to have less than 3 pascals pressure difference (0.012" water column) between rooms under all operation conditions, doors open/closed, at all air handler speeds. Unless the duct system was designed for that and well executed it's almost certain to fail. The inexpensive dual port manometers used for testing static pressures on duct systems don't have that kind of resolution, but many/most go down to 0.01", not that they're very accurate on that end. Using one of those to test the room to room pressure differences means anything other than a 0.00" or 0.01" reading would be fail, but be sure to switch ports, measuring it in both directions to get a sense of whether it's just reading noise. A measurement of 0.03" or higher is pretty much guaranteed to fail Energy Star.
Having a supply duct blowing into the convection-zone stairwell is making the stratification issues worse. Placement of thermostat where it's affected by those convection currents is also pretty lousy, but probably easier to fix than the register placement.
Checking pressure differences between rooms is a good diagnosis step. Checking the indoor/outdoor pressure difference with and without the fan running would also be useful to compare the problem described in your first two paragraphs to stack effect, which would produced pressures on the order of 0.02".
Sorry for your troubles, and thanks for sharing this interesting story of poor design of the HVAC system(and building shell, and floor plan).
In the winter, portable electric heaters in the bedrooms may be helpful to improve comfort. The higher costs of electric heating may be offset somewhat by lowering the thermostat and operating cost of the central heating system.
A well insulated and airtight home will have very little temperature difference between floors. Couple this with adequate sub slab and foundation wall insulation if applicable and the lower level should be as comfortable as the upper levels. A well thought out heating and distribution system is a must for winter comfort.
OP, you could temporarily test your curtain idea / see how much the air transfer between floors is a factor using something like the dust containment walls contractors use - plastic + spring-loaded poles to seal off the opening. ZipWall & e-z up are two brands. If you're handy, you could make cut-to-length spring sticks out of 1x2s or similar. Plus, the poles would be handy for the next remodeling project! :o)
Installing a temp. curtain would not only tell you about the heated/cooled air moving between floors, but a big "pop" when one air handler comes on might point to leaks in the ducting too.
I also have a not-s0-well insulated or air sealed house with a large opening between the first and second floors. I have a single air handler system with the ducts inside the envelope and the return and thermostat on the first floor. On a cloudy day with the HVAC running regularly, the temperature differential between floors is maybe two degrees. My theory is that the tendency for the second floor to be warmer is counteracted by the air handler circulating the volume of the air in the house every 15 minutes or so.
Another simple experiment would be to turn on the fan of the unit that isn't running and see what difference the increased circulation makes.
Good luck!