Is it appropriate to add attic-floor insulation after converting the attic with open-cell spray foam?
I am in Northern Virginia and recently completed an integrated attic-insulation and HVAC project on my townhouse. I would appreciate input from people with building-science and residential HVAC-design experience.
The original project included:
-
Removal of the existing attic-floor insulation;
-
Installation of open-cell spray foam at the roofline and other exterior attic boundaries;
-
Aeroseal duct sealing; and
-
Replacement of the HVAC system with a 3-ton variable-capacity hybrid system.
My understanding was that insulating and air-sealing the roofline would bring the attic within, or substantially within, the home’s thermal envelope. The air handler and ductwork serving the top floor are located in the attic.
Immediately after the new HVAC system was installed, it was unable to cool the house adequately during hot weather. On several days when the outdoor temperature exceeded 92°F, the system operated at full capacity for extended periods, but the indoor temperature remained around 79–80°F and could not reach a 75°F setpoint.
The contractor’s HVAC technician observed the problem during installation and returned twice afterward. The technician confirmed that the equipment was operating at full or nearly full cooling capacity, but it still could not bring the house down to the thermostat setting.
A post-project blower-door test later showed that the house was substantially tighter than it had been before the work. The test-out technician did not identify an obvious building-envelope or spray-foam defect that would explain the cooling problem.
The contractor initially discussed replacing the 3-ton outdoor unit and indoor coil with a 4-ton system. They are now recommending that I first add approximately 6 to 10 inches of blown-in insulation on the attic floor. This insulation would be installed in the same general location from which the original attic-floor insulation was removed.
The contractor believes this may reduce the cooling load enough for the existing 3-ton system and has described the work as reversible if it does not solve the problem.
My questions are:
-
Is it considered a sound building-science approach to install substantial insulation at both the roofline and the attic floor?
-
Would this create two competing thermal boundaries, or is that concern overstated?
-
After adding attic-floor insulation, would the attic still reasonably be considered conditioned or semi-conditioned space?
-
Could this configuration create humidity, condensation, mold, or other moisture risks, particularly because the attic has no dedicated dehumidifier or intentional ventilation?
-
How would the added attic-floor insulation affect the air handler, supply and return ducts, and bathroom exhaust ducts located in the attic?
-
Would the attic need dedicated supply air, a return-air pathway, transfer grilles, mechanical ventilation, dehumidification, or another form of temperature and moisture control?
-
Could this approach reduce or negate some of the intended benefit of moving the thermal and air-control layers to the roofline?
-
Before accepting this proposal, what calculations, measurements, or design information should the contractor provide?
I would also appreciate recommendations for objective performance criteria. For example, should the contractor be expected to demonstrate that the system can maintain a specified indoor temperature and relative humidity at an agreed outdoor design condition?
I am not opposed to a technically sound corrective measure, but I am concerned about materially changing the attic assembly without understanding the long-term moisture, durability, and HVAC-performance consequences.
Thank you for any guidance!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
4ton and townhouse usually don't go hand in hand unless you have a wall of glass. Something somewhere is off. How many inches is the OC foam?
Open cell attics need to be conditioned, so at least a supply register is needed.
I would first confirm the air handler is working as intended. You can get a cheap differential pressure gauge (ie 2"wg Dwyer magnahelic) and measure the pressure drop across the air handler. This will tell you if your ducting is sufficient and from the data sheet you should be able to back calculate a rough CFM.
With a precision digital meat thermometer measure the temperature on the return and supply of the air handler.
From these two measurements if you equipment is running. The 3 ton unit on max should run around 1000CFM and should have temp drop around 20F across the unit.
Thank you. I found the contractor’s commissioning screenshots from the day after the installation, so I have some additional data.
The contract specifies an average of approximately 8 inches of open-cell spray foam at the underside of the roof deck and approximately 3.5 inches at the exterior attic walls.
During the contractor’s test:
Outdoor temperature was 88°F (at the outdoor unit?);
The heat pump was operating at 100% cooling capacity;
Compressor speed was 4,200 RPM;
The Connex control displayed 1,248 CFM indoor airflow;
Indoor fan speed was 955 RPM;
Displayed indoor static pressure was 0.34 in. w.c.;
Return air was 76.4°F dry bulb and 65.5°F wet bulb;
Supply air was 58.0°F dry bulb and 54.9°F wet bulb; and
The dry-bulb temperature split was 18.4°F.
The Fieldpiece application calculated approximately 38,980 BTU/h of air-side capacity using 1,250 CFM and the measured return/supply air conditions.
However, it appears that the 1,250 CFM value may have been entered from the Connex display rather than independently measured. The Fieldpiece differential-pressure probes were not connected, and Delta P was blank. I also do not have separate measurements for return static, filter pressure drop, coil pressure drop, supply static, or independently measured total external static pressure.
The contractor Manual J used 90°F outdoor and 75°F indoor design conditions and calculated:
27,500 BTU/h total cooling load;
21,000 BTU/h sensible load;
6,600 BTU/h latent load; and
985 CFM cooling airflow.
Based on these data, would you agree that the current 3-ton system did not appear to have a major airflow restriction at approximately 1,250 CFM?
Would you still recommend independently measuring airflow and a complete static-pressure profile before drawing conclusions about the duct system or whether it could support a larger variable-capacity unit?
I am also trying to understand why the equipment was already operating at 100% at 88°F outdoor temperature. If the Manual J load is approximately 27,500 BTU/h at a 90°F design condition, while the Fieldpiece calculation indicates approximately 39,000 BTU/h of delivered capacity, I would expect the system to be able to maintain 75°F once the house had stabilized.
Does that discrepancy make you more suspicious of the Manual J assumptions, the calculated air-side capacity, air-distribution issues, or the condition of the encapsulated attic?
There is currently no intentional supply register, return pathway, ventilation, or dedicated dehumidification serving the attic, even though the Manual J identifies the duct location as an “encapsulated attic.”
First, answers to your questions:
1- NO, it's not. You should insulate the roofline OR the attic floor, but generally not both. .
2- NO, not really. The attic being inside or outside the building envelope is determined by the location of the building envelope, and that "building envelope" is an air barrier. You can insulate between locations that are still within the same building envelope, and that's actually commonly done, but for sound isolation instead of thermal (but it usually uses some of the same materials either way). There SHOULD be air "communication" between the attic and the rest of the house though, which means there should be some airflow from the HVAC system through the attic. Note I said "through", not "from" or "to". You need supply and return air pathways.
3- YES, it would still be conditioned space, but likely at some intermediate temperature.
4- Once you condition the attic, it should be conditioned by the building HVAC system, which should keep humidity levels under control. You should not be creating a sort of hybrid "enclosed but not conditioned" attic space. Vent it, or don't vent it. If you don't vent it, it should be sealed off from the outdoors and ventiallated by the HVAC system as part of the house.
5- You'd insulate the ducts. That might help the HVAC system if the attic space is crazy hot, but it's a bandaid, and doesn't solve the real problem.
6- YES, and usually that's provided by the same HVAC system that serves the rest of the home.
7 - Potentially, it depends on other factors. I would not insulate the floor AND the roof though.
8- I wouldn't use the contractor here, I'd use a third party HVAC engineering form or HERS rater. You have a somewhat unsual situation here, and you're likely to spend a lot of money without necassarily solving anything with many mechanical contractors.
You need to know what's going on in that attic. I suspect it's pretty hot. You have about 8" of open cell spray foam under the roof sheathing. Closed cell would be much better, and is what really should have been used, but too late for that. Open cell is around R3.5 or so per inch, so you have about R28. Code minimum these days in most areas is R49, so you're *very* light on insulation here. Chances are the spray foam guys sold you on some marketing about how spray foam R value is somehow "better" than other R value, but that's BS. R value is R value, period. You're far under code minimum, even in olden times when it was R30, and then R38 (which was the standard for quite some time).
Did the previous HVAC system cool the home sufficiently? If so, what size was it? Did the ducting change with the new system? 4 tons is a pretty good size system. For some perspective, 5 tons is about the largest standard residential system, and can handle a large home. Anything over 5 tons is usually a zoned system made up from multiple smaller units. I wouldn't just put in a larger system here, I think something else is going on.
I suspect you have one or more of several problems here. My first thoughts are "the attic is hot due to insufficient insulation, and is robbing a lot of cooling capacity from the system", possibly also "the attic isn't completely air sealed from the outdoors, so your HVAC system is trying to aircondition the outdoors", and possibly "something is restricting the airflow in the ductwork, or the ductwork is in some way not moving the air to the right locations". That last one would probably result in noticeable hot and cold spots within the home though.
With the data provided, you have basically one of two things going on here as follows:
1- The home is bringing in far more BTU than expected, which could be one or both of insufficient insulation or poor air sealing from outdoors. Your blower door test probably rules out the air sealing issue BUT maybe the home is air sealed from the attic, and the attic is leaking? That might not have shown up with the blower door test.
2- The HVAC system is underperforming for some reason. If it's running at 100%, maybe the airflow is not going where you expect? Makeup air coming in from outdoors? Do you have a way to check if the system is either pressurizing or depressurizing the home? The manometer Akos mentioned can check for this if you use it in differential mode between the indoors and outdoors. Ideally, the indoor air pressure should not change when the HVAC system is running compared to when it's not. If you see a significant change in either direction, then the ductwork is leaking to the outdoors somewhere.
Bill
Thanks, Bill. This has been very helpful.
I pulled together temperature and relative-humidity data from the attic and several rooms and attached four plots covering the last 60 days. I used day counts to make the project milestones easier to follow:
Days 18–19: open-cell spray-foam installation and removal of old attic floor insulation
Days 44–45: HVAC replacement
Days 44–47: period during which the cooling problem was observed
Days 47–49: portable air conditioners were operating
Days 57–59: house unoccupied and thermostat in Away mode
I also added the Washington Dulles outdoor daily high and daily mean temperatures to the two temperature plots. The contractor’s Manual J used the Sterling weather station, so I thought Dulles was the most consistent outdoor reference.
To answer your earlier questions, the previous system was also a 3-ton unit, but it was single-stage. The duct layout was not redesigned or replaced as part of this project. The contractor reused the existing supply and return ductwork, Aerosealed it, and installed a new 4-inch filter cabinet.
The old system struggled significantly during summer weather before the spray-foam work. However, its performance improved noticeably after the roofline was insulated and air-sealed.
Looking back at both the outdoor-weather records and my indoor sensor data, the early-July heat wave was actually more severe than the weather during most of the HVAC-replacement week.
During Days 31–35, while the old 3-ton single-stage system was still operating after the spray-foam work, the outdoor daily highs were approximately:
Day 31: 97°F
Day 32: 98°F
Day 33: 98°F
Day 34: 99°F
Day 35: 94°F
Over those five days, the sensor data show approximately:
Top-floor average: 77.2°F
Office average: 78.0°F
Main-floor average: 76.2°F
Attic average: 82.1°F
I also do not remember the house feeling unusually hot or requiring supplemental cooling during that period, despite the more severe outdoor temperatures.
By comparison, Days 44–47, during and immediately after the new HVAC installation. The outdoor daily highs were approximately:
Day 44: 90°F
Day 45: 95°F
Day 46: 96°F
Day 47: 90°F
Day 44 and 45 are not clean performance-comparison days because the HVAC replacement was still underway (mostly during Day 44). Day 46 was the first complete hot day after installation. On that day:
The office daily average was approximately 79.2°F;
The combined top-floor average was approximately 77.9°F;
The main-floor average was approximately 76.9°F; and
The attic average was approximately 82.2°F.
Daily averages also understate the afternoon comfort issue. The office reached approximately 80.8°F on Day 46 and the system could not bring the house down to the 75°F thermostat setting even while the contractor observed it operating at full or nearly full capacity.
The house felt extremely uncomfortable during the two or three days immediately following the HVAC replacement, even though the replacement-week weather as a whole was not as severe as the early-July (Day 31-35) heat wave.
On Day 47, the contractor installed two portable air conditioners. Together with one portable unit I already owned, there were three portable AC units operating in the house to make the indoor conditions tolerable.
Those portable units operated for only approximately two or three days. During the following week, the outdoor weather became milder, so I stopped using them because the central system was then able to maintain more acceptable temperatures.
This creates two important limitations when interpreting the post-installation data:
The indoor-temperature decline beginning on Day 47 cannot be attributed solely to the new central HVAC system because three portable AC units were also contributing cooling.
The improved performance during the following week coincided with substantially milder outdoor conditions. It therefore does not demonstrate that the new system’s peak-weather performance problem was corrected.
I recognize that this is not a controlled comparison under identical conditions. However, the available data and my experience suggest that the old 3-ton single-stage system, after the spray-foam work, handled the early-July heat wave at least as well as—and possibly better than—the new 3-ton variable-capacity system handled the less-severe replacement-week conditions.
At a minimum, the new system has not demonstrated a meaningful improvement in peak-weather comfort compared with the old system after the spray-foam work.
The plots also show that the spray foam substantially reduced the attic temperature. Before the work, the attic daily average frequently reached the upper 80s and 90s, with some days near 100°F. After the work, the attic was generally in the upper 70s to low 80s, including during very hot outdoor weather.
Therefore, the attic does not appear to remain “crazy hot” in the same way that it did before the spray-foam work.
However, there is currently no intentional supply register, return-air pathway, dedicated ventilation, or dehumidification serving the attic. I am still concerned about whether the attic is properly communicating with the conditioned house and whether it is functioning as a correctly designed conditioned or semi-conditioned encapsulated attic.
I agree that the underlying issue could involve:
The actual house load under peak conditions;
The assumptions used in the Manual J;
Actual equipment performance;
Airflow or air-distribution problems;
The existing supply and return duct capacity;
The attic-conditioning strategy; or
Some combination of these factors.
At this point, however, the contractor is effectively offering me only two corrective options:
1. Add approximately 6–10 inches of insulation back onto the attic floor, keep the existing 3-ton system, and wait to see whether reducing the load on the occupied floors solves the problem; or
2. Replace the outdoor unit and indoor coil with a 4-ton variable-capacity system. They have warned that this could result in high static pressure, return-air noise, increased energy consumption, or other operating issues. My understanding is that they may not accept responsibility for correcting those consequences if they occur.
I understand that the technically correct answer may be not to accept either option until an independent third party evaluates the attic, Manual J, airflow, static pressure, air distribution, and actual delivered system capacity.
However, if these are the only two practical options currently available, which would you consider the less risky engineering choice?
My concern with the attic-floor option is that it may reduce the cooling load on the occupied floors but leave the attic as a more isolated intermediate-temperature space, with the air handler and ductwork still located there and no clearly defined conditioning or moisture-control strategy.
My concern with the 4-ton option is that it may provide the additional peak capacity that appears to be missing, but it could be the wrong correction if the actual problem is equipment performance, duct capacity, air distribution, the Manual J assumptions, or the attic design rather than nominal equipment size.
If you had to choose between those two paths, which would you consider preferable? What minimum testing, design verification, and written protections would you require before allowing either option to proceed?
Thanks again. Your comments have been very helpful.
Based on the readings from the thermostat, it looks like your air handler is running properly, no need for measruments.
Where is the thermostat located and is the air handler mainting the temperature there?
If you look at attic temps post SPF is about 82F on a very hot day. Uninsulated ceiling is about R3, say 1000sqft of ceiling. Your attic is adding
(82f-75f)*1000sqft/R3=2300BTU, basically squat. Adding insulation reduces that squat to less than squat but won't change your heat load appreciably. Not worth the effort at all. That 8" of spray foam is a bit on the skimp side, but again won't change numbers a whole lot.
I have a feeling is more of an airflow distribution or controls problem. If the thermostat is in a location that gets good cooling flow, it might be shutting down the air handler allowing the upper floor to get extra toasty. Can you get runtime data from the thermostat?
How big was the previous AC system?
Thanks, Akos. That ceiling-load calculation is very helpful and is consistent with something I noticed in the contractor’s Manual J.
The Manual J assigns approximately 2,629 BTU/h of cooling load to the third-floor ceiling below the encapsulated attic, which is very close to your estimate. Based on that, it is difficult to understand how adding 6–10 inches of insulation back onto the attic floor would materially change the total calculated cooling load of approximately 27,400 BTU/h.
The previous AC was also a 3-ton system, but it was single-stage.
The thermostat is located on the main floor, between the kitchen and living-room sensors.
During the cooling problem immediately after installation, the contractor observed the new system operating at 100% cooling demand, with the compressor at approximately 4,200 RPM. The thermostat reported approximately 1,248 CFM of indoor airflow and 0.34 in. w.c. static pressure.
The thermostat itself was also above the 75°F setpoint and was not being satisfied. The system continued operating at full or nearly full capacity but still could not bring the thermostat location down to setpoint. Therefore, it does not appear that the problem was caused by the thermostat receiving too much cooling and shutting the system down prematurely.
The temperature data also suggest a possible distribution issue. Comparing the old system during the July 1–5 heat wave with the new system on July 16, the office and main living areas were slightly warmer with the new system, while the basement was noticeably colder. That makes me wonder whether a greater share of the delivered airflow is now going to the basement or lower portions of the house rather than the upper floor.
I am not sure whether the Bryant Connex thermostat provides access to detailed historical runtime, capacity-stage, compressor-speed, or airflow data. Do you know whether the homeowner can retrieve that information directly from the thermostat or app, or whether it requires dealer access?
Based on the relatively small calculated ceiling load, would you agree that adding attic-floor insulation is unlikely to address the primary problem? Would your next priority be checking room-by-room airflow and pressure balance, or reviewing the thermostat and control configuration?
Thanks again.
Have you been tracking humidity levels inside the home? If the new system is doing a poor job of dehumidification, the resulting higher indoor humidity will make the indoor temperatures feel even higher, which could be compounding your problem.
I would have a HERS rater look at this. I think you really need someone to come out and evaluate the system that is NOT tied in with the installation contractor. That doesn't necassarily mean your contractor is a problem, just that they might be missing something.
Bill
Thanks, Bill. Yes, I have been tracking humidity in the attic and several rooms. The living areas have generally been around the low-to-mid 50% RH range, while the basement and attic have sometimes been closer to 60% or slightly higher. So humidity may be contributing to the comfort issue, although the upper-floor humidity was not unusually high during the worst overheating period.
I agree that an independent evaluation is the right next step. Thanks again for the recommendation.
If the system is running continuously and your humidity levels are still that high, you either have a lot of moist air leaking in from outdoors, or the system isn't running at a low enough coil temperature to dehumidify well. Nothing exactly jumps out at me from what you've posted so far, but I'm thinking something isn't right with the HVAC system and it's coil is running too warm. There can be a number of reasons for that, one of which is high fan speed (but be careful with that -- too-low fan speed can cause the coil to freeze up).
Make sure to mention all this to whomever checks things out for you. I again recommend you try a HERS rater and not just a mechanical contractor.
Bill
Thanks, Bill. That makes sense. I will make sure the independent evaluator looks specifically at moisture infiltration, actual airflow, coil temperature, refrigerant performance, and whether the blower setting is too high for proper dehumidification.
The thermostat was reporting approximately 1,248 CFM at full cooling, but I understand that is not a substitute for an independent airflow measurement.
Thanks again for pointing me in that direction.
I think your air handler is running mostly fine. The CFM/ton is too high for good dehumidification but should still not struggle with cooling. Usually this is adjustable in the installer menu. You want about 300-320CFM/ton for humidity control.
I would start with checking register flow rates and go from there. It would also give you a quick verification of the CFM reading from your thermostat. If your ductwork did not change, why is the basement so much colder? Something is off somewhere.
“A post-project blower-door test later showed that the house was substantially tighter than it had been before the work.”
Did they give you your before and after ACH-50 numbers? If you have them, please post them.
“• Removal of the existing attic-floor insulation;
• Installation of open-cell spray foam at the roofline and other exterior attic boundaries;”
What R value was removed and what R value was installed?
My opinion is that changing from a vented attic full of HVAC equipment to a conditioned attic full of HVAC equipment is a marginal improvement at best assuming the duct leakage is low. Your new thermal envelope has a much larger surface area to lose energy and it seems likely your new spray foam has a lower R value than whatever you removed. Assuming you ACH-50 is under about 2.5, I think you will need to throw larger equipment at this problem.
Walta