GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

High humidity issues in a spray foam attic

Danny Kelly | Posted in General Questions on

Have an interesting one here for you building science detectives.
Existing two story home built in the 1950s – Charlotte, NC (mixed humid) – approx. 3000SF. Existing mechanicals were a furnace in the crawl space with duct work distributed throughout the house and an air conditioner in the attic serving the second floor. Old batt insulation in the ceiling, wall insulation undetermined (we think none) and batts in the floor.
Owner received an energy audit and implemented some of the recommendations. Removed all batt insulation in the attic, removed both HVAC systems and all duct work, performed air sealing in the crawl space and the attic (at the floor and ceiling) and installed 5 ½” of Demilec open cell spray foam on the roof deck. Existing roof did not have a ridge vent – shingle color is a dark gray. Now that everything is installed, there is very high humidity in the attic and condensation on the unit and condensate drain.

There are already several threads on open cell vs. closed cell, vented roof vs. unvented and blown-in insulation vs. spray foam insulation so please let’s stay with our fixed variables we have here in our discussion.

Here are few readings we took. Interior temp. – 75 degrees and 45% RH. Exterior temp. – 89 degrees and 50% RH (low for here this time of year). Attic temp – 84 degrees and 75% RH.

Now the troubleshooting:
HVAC – we are asking for the manual J and checking the refrigerant charge to be sure the unit is sized correctly and performing properly. Typically in these situations the units are oversized and short cycling which is definitely not the case here – during the hottest part of the day the system is running non-stop and cannot keep up so do not think that the HVAC is the main issue although it could be contributing.
Spray foam – one of the problems with spray foam is there is not really any way to do quality control on it. Have read before that if the mix is not correct there will be odors, discoloring and the foam would be extremely hard. We do not see any signs of these issues. I was expecting to see large gaps where the old soffit vents were but it appears to have a tight air seal at the perimeter. We did inspect the foam insulation with a thermal imaging camera – temperature was 84 degrees for 90% of the area. There were several areas where the depth of the foam was not consistent and we saw about a 5 degree difference within the same rafter bay (Installer said this was typical). A handful of small ½” voids where the temperature got up into the 90s. None of which you would think would allow that much moisture to enter the attic. What else can we look for to determine the quality of a spray foam job?

Big question is how did the humidity get so high up there. I do not think it is coming from the house below – that air is conditioned and relatively low humidity – if that air was getting into the attic it would improve the situation. I do not think it is coming from the exterior either – if there was a hole in the foam/roof, the moisture would be going from the attic to the exterior. So how did it get there? I think two things may have happened – the day they sprayed the foam, the attic conditions were probably pretty similar to the exterior and probably even hotter. Let’s say hypothetically – exterior was 90 degrees and 70% humidity – inside the attic was probably 110 degrees with the same amount of moisture in the air as the exterior. After the foam was installed the attic temp. dropped from 110 to 85 degrees. The amount of moisture in the air remained the same thus the humidity would increase with a temperature drop. Secondly to compound the problem, the blowing agent for the spray foam is water – as the foam cures, offgasses, etc. it released all of the water used to install it so added more moisture into the air.

Since the attic is only passively conditioned through leaks in the duct work and the leaks in the ceiling, there is nothing working to remove the moisture from the attic. The only way for the moisture to leave the attic right now is through diffusion through the roof to the exterior and diffusion through the plaster to the conditioned area of the home. Diffusion is a slow and weak force so it could take weeks for the attic to equalize with the conditioned portion of the home.

Our plan of attack is to temporarily run a dehumidifier in the attic until we get the RH to match that of the conditioned area below. I think once we get the attic acclimated with the rest of the house the levels will somewhat mimic each other.

Any flaws in our logic? Any other things we should look into? Is there a way to calculate the size and number of dehumidifiers based on RH and volume? Any other suggestions on avenues we can explore to solve this problem? Thanks for the help.

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. Danny Kelly | | #1

    oops - forgot to ask my main question. Has anyone else had a similar situation? If my theory was true, you would think there would be stories like this all over the place. I have never heard of anything like this before and I know there are foam companies doing retrofits every day without issues so makes me think we are missing something. Thanks again.

  2. carpeverde | | #2

    Haven't run into this on any of my retrofits yet, but one of my clients is living in a small 900sf house consisting of two similar sized disconnected sections, only one of which is air conditioned. Both are wood frame with open cell foam for insulation and standing seam roofed. It's Central Texas hot & humid. We were concerned about humidity in the attic of the conditioned side. He installed a humidistat and turning the AC on for about a minute between long cycles when the system is off brings the attic relative humidity down fast. He's been in the house for 3 years and doesn't mind this "chore" and enjoys $50 electric bills. My thoughts are that in order to reduce RH in the attic of an unvented attic assembly, would it be preferable to treat the attic volume as a return air plenum and thereby further reducing the delta-T between the attic and the occupied rooms below?

  3. homedesign | | #3

    This is a "good one".
    Some people will likely suggest that the problem is inward solar vapor drive.
    Your theory seems to be moisture trapped during construction.
    I am going to guess that these may be contributing factors but perhaps not the primary source.

    There is a similar "problem" discussed over at JLC
    Too bad the original OP has not returned with more data.

    Have you measured temp & RH in the crawl space?
    Have you taken measurements in all 3 spaces(crawl,living&Attic) early in the morning and late afternoon?

    You say the crawlspace was air-sealed "at the floor"....
    How well was this done?(Especially around plumbing chases)
    Is it near perfect?
    How do you know? did you pressure test between crawlspace and attic?

    How good is the source (bathroom,kitchen) moisture removal?

    How about some photos?.... Dan M. sometimes will attach photos if you request.
    Photo of outside of house, underside of roof deck and ceiling of crawlspace(near plumbing) would be interesting.

    Bill Rose says that most Attic problems can be traced to crawlspace or basement.

    John B (Hot/Mixed Humid)

  4. homedesign | | #4

    Danny ..
    I meant to say ....
    Your theory seems to be moisture trapped during RETROFIT.

    You mentioned that the open cell at roof deck was only 5-1/2"
    I am guessing that the foam does not encapsulate the roof rafters?
    Are the bottom faces of the roof rafters exposed?
    Judging by your extra hot attic ... I guess exposed rafter bottom.

    You mentioned a new AC unit in the attic.
    Is there a unit in the crawlspace?

    John B (Hot/Mixed Humid)

  5. homedesign | | #5

    Perhaps of interest is a new Article (JLC June 2010)
    About REPLACING a rotting roof
    This was a cold climate problem that was related to spray foam at the roof deck.
    Not that the spray foam created the problem ... but It may have CONTRIBUTED


  6. adkjac | | #6

    IMHO... this is easy... you have 2 air seals. I did a similar job here except we didn/t seal the attic from the liviing space. So... you should condition the attic space now that you sprayed the foam to the rafters. That is the point of putting the foam on the rafters,,,, to add the attic to the conditioned space. And as to your foam... it sounds like it is installed fine...

    yes... your dehumidifer will dehumidify... AC can do the same thing too.

  7. Danny Kelly | | #7

    Thanks for the replies - some good ideas.
    John - 2x6 rathers so the bottoms are barely covered with a flash coat of foam, you can see them through the foam though - very thin layer of foam. And yes there is a seperate system in the crawl for the first floor.
    Have not taken readings in the crawl space will do that and some zonal diagnostics with the blower door is a good idea, can see if the attic is truly seperated from the outside. Had not really thought of the vapor drive from the exterior, was more thinking % RH from outside to attic was lower but we have been getting thunderstorms each afternoon on that hot roof - that could be contributing for sure.

    I guess we could add a return up there or a permanent dehumidifier but pretty typical around here to not completely condition the attic, it is typically a "semi-conditioned" attic so the owner would prefer that it performs like others he has seen that did not need additional measures. I guess installing a humidistat in the attic isn't a bad idea as long as we can be sure it does not need to run much.
    Thanks for the suggestions.

  8. Armando Cobo | | #8

    - I would start by having a pressure test on your house and a thermal imaging study. Your temperature and humidity between house and attic are way too high to be “normal” if you really did a good job re-insulating and sealing. More than likely you have to many holes and gaps in the envelope and the interior walls/chases/pipes/wires, etc.
    - Other possibilities could be your AC is not sized properly to dehumidify the house. A manual J is as good as the person who runs it… bad info in equals bad info out; and I’ve found in my many years of experience that most Manual Js I see are way wrong. Find someone that truly knows how to do it; usually is a mechanical engineer (NEVER, EVER LET THE HVAC CONTRACTOR THAT INSTALLS THE SYSTEM DO YOUR MANUALS J, D, S & T). Remember, bigger is not better. Also, was your system tested for proper air delivery? How do you know you are getting the correct amount of cfm where it needs to go?
    - Variable speed units do not dehumidify well at low speeds (in case you are using it). In areas of high humidity, you must use a dehumidifier as part of your hvac system as well.
    - Charlotte is in CZ3 but on the border with CZ4, which means that 5 ½ “ open-cell is NOT ENOUGH INSULATION. Either you install 10” on the inside or you need to install rigid insulation on the outside. Just wait until the winter time; you could see huge condensation under the roof decking and/or your 2x6 due to thermal bridging.
    - All conditioned attics and crwal spaces NEED ventilation, make sure you have it.

  9. God | | #9

    Armondo... not the same zone where I just built for others... open cell sprayed to roofdeck, walls, floor above cellar... HVAC in cellar, none in attic. System... house perform super. attic in summer is fine... just a bit warmer than rooms, no moisture problems... winter, just a bit cooler than rooms, no moisture. We did put one small register in cellar, ducts are mastic in cellar.

    I think this home's problems will be solved with tiny registers in attic for supply and return to further condition space. And as to built in moisture... time will take care of that... to the point of closing registers in attic in a year.

    also... when you say ventilate... I hope you mean add registers to condition... not ventilate to the outside which has been found to be a problem moisture wise.

  10. homedesign | | #10

    You were not kiddin
    You do have a lot of nicknames.

  11. Armando Cobo | | #11

    Dear God,
    I really do not want to disagree with you, I don’t want to go way down south when my time is up; however, you must be getting OLD and not reading correctly, ;-))) …there is a second AC & ductwork in the attic plus a temporary dehumidifier.
    There is no mentioned of the wall assembly, except that there MAY BE NOT be insulation. So lets see, no telling what cladding and its condition, windows and their condition, no moisture or air barrier, poor or no insulation, I doubt there is ANY sealing around openings or thermal bypass mitigation on the walls… and you all wonder where the moisture is coming from? On top of that, you have and AC and dehumidifier sucking moisture not only from the inside air but from all the neighbors in the block! Do WE think here on earth warm/moist air rises? Check your bathroom ceiling after a shower.
    By ventilation on conditioned attics or crawl space, I meant supply air.

  12. adkjac | | #12

    God can't understand Armondo...but loves all his children regardless.

    aj... God and all... stand by our posts...

  13. homedesign | | #13

    Quote from original post:

    Owner received an energy audit and implemented some of the recommendations.

    Danny, it would be interesting to hear what recommendations the owner chose Not to do.

  14. Danny Kelly | | #14

    Hi John - we were not involved in the initial audit - were brought in more on the back side so not familiar with all the details only what we discussed with the owner. Seems like the audit suggested good, better, best scenarios.
    HVAC - recommended everything from tune up and duct sealing to complete replacement - went with complete replacement including all ductwork
    Crawl space - add vapor barrier, air sealing, closed crawl - they did vapor barrier and air sealing but left the crawl as a vented crawl
    Attic - air seal, add insulation, attic tent (keep thermal boundary at ceiling level) - radiant barrier and foam roof deck. Went with air sealing, removing existing insulation and foam roof deck.
    I'm sure there were some recommendations for the exterior walls, doors and windows - no work in this area. May have been some ventilation recommendations - none were implemented - still have very leaky exterior walls. I'm sure there were recommendations on appliances, lighting, water saving measures, etc - they were concentrating on the HVAC and thermal envelope at this time so none of those were implemented at this time.

    Armando and God (must have been a good question to get god in on the mix). I do agree and typically explain to my clients that the spray foam does not meet the required R-Value when doing roof decks but there are literally hundreds of homes around here getting very good results with 5" on spray foam. Adding ventilation and/or supply/return grills may be our ultimate solution but again there are plenty of homes performing well without those measures so trying to figure out why this one is acting so different.

    Good news is took another reading in the attic and the humidity is dropping with the help of the dehumidifier - down about 10%. Inside temperature of the house is holding good at 74. Going to try to get by on monday for blower door and some zonal diagnostics.

  15. homedesign | | #15

    Crawl space - add vapor barrier, air sealing, closed crawl - they did vapor barrier and air sealing but left the crawl as a vented crawl

    I assume the vapor barrier was put OVER the ground, Eh?
    What method did they use to air seal the underside of the LivingSPACE?
    I would think that air sealing above an existing crawlSPACE would be very challenging even with spray foam.....(Especially around plumbing stacks and mechanical boots)

    There is also the issue of the mechanical in the crawlSPACE.
    How perfect was the air sealing of all the new ductwork that is NOT inside the LivingSPACE?
    Did they duct-blaster test?
    What are the conditions (Temp&RH) in the crawlSPACE?
    How(method) did they air seal between LivingSPACE and AtticSPACE?

    remember to take photos

  16. adkjac | | #16

    names... whatever... focus on the problem..... really seems like a mess. you have not tightened up the house very well at all... but you did tighten up the one place with the porblem too much. you have the attic sealed at the roof and and the ceiling if I read this right. And you have an AC unit in there that is cold when operating.. And you don't have any way for the air to change right? None? And you expect it to change? So you put in a dehumidifier, That seems like a good idea.... why not add the registers like I said.. a very simple fix and be done.

    Without fixing up the rest of the home including dealing with the vented crawlspace... you are goiing nowhere toward completing the rehab. If money is the issure.... then the plan was no good to start with. What did the assessment cost? Green rehabbing is a money pit to me.

    The biggest challenge that home has now is the job is one thiird done.


  17. user-659915 | | #17

    The crawl space left fully vented is a major red flag. As JB points out it is very difficult to properly seal the floor over a crawl space from below. It's possible therefore that the house has been left with a major entry point for humid air at the bottom and no exit at the top, and with crawlspace and attic possibly even directly connected by the remains of an old heating chase. I'd recommend very strongly that the crawl be sealed ASAP.

    By the way, I'm still not clear about the mechanical systems. Danny, how is the 2nd floor heated? You indicate a separate A/C system for the second floor but the only mention of a furnace is in the crawl space. Is the 2nd floor system actually a heat pump? If the upper floor is cooled by independent A/C but shares its heating with the first floor I don't understand how either system could be properly balanced. This would be a recipe for multiple points of failure.

  18. homedesign | | #18

    You Know
    It looks like the owner paid for an Energy Audit....
    Got a list of recomendations and then randomly executed SOME and ignored OTHERS.

    Air Conditioning and Dehumidifying the attic does not sound like a solution.

    You say the "5 inch foam attic deck retrofit" is working well for many other homes in your climate.
    How many had a CrawlSPACE? How Many had a VENTILATED CrawlSPACE?

  19. adkjac | | #19

    For once a chorus... Happy Father's day to all the Fathers out there

  20. Danny Kelly | | #20

    Thanks for all the great feedback and questions - will try to answer all of them (how do you guys get those grey boxes to appear?)

    Yes - our vapor barrier covers the ground. We also roll it up the foundation when the exterior grade is greater than the crawl until we are above grade on the exterior. Yes it is difficult to air seal a crawl and we will never get it 100% sealed but do the best the can. The biggest issue is all of our older houses have 1x6 floor decking, roof sheathing and exterior wall sheathing so lots of opportunities for leaks. We concentrate on sealing around all plumbing and electrical penetrations, around all HVAC boots, around fireplaces, tub drains and any thermal bypassess or chases to the attic. We have not measured the Temp/RH of crawl yet. The temperature and RH on the first floor of the home are fine - I think the crawl space is pretty well seperated from the thermal envelope and there is no evidence of crawl moisture getting into the home. Have tried to find chases where crawl is connected to attic and have found none.

    We have not performed a duct blaster test but feel pretty good that the system is tight - the HVAC company used has done hundreds of energy star homes. We have had some issues running duct blasters on existing homes - some of them are so leaky we have a hard time pressurizing the house to 25 Pa. Why it is harder to pressurize than depressurize is beyond me. Also - I do not think duck leakage is the issue here - if anything, more duct leakage (on both supply and return sides) would actually help our problem by actively conditioning the attic. There is a gas furnace and AC in the crawl serving the 1st floor and a heat pump and AC in the attic serving the 2nd floor.

    Air sealing from the living space to the attic space is pretty easy. An old plaster ceiling with very few penetrations. Foamed around HVAC boots, top plates, wires, plumbing penetrations, etc. All insulation was removed so very easy to access everything. There are no chases, thermal bypasses or duct work connecting the crawl to the attic.

    I did not invent this semi-conditioned attid idea - like I mentioned before there are hundreds of attics around here that have 5" of open cell foam on the roof deck with no supply or return registers. These homes are performing very well. Most of them are on homes with vented crawl spaces, which are pretty common here. I do think the problem is there is no way for the air to exchange - this concept I believ relies on duct leakage and communication through the ceiling plane. Most of the homes that have this sytem have a sheetrock ceiling, I wonder if the fact we have a plaster ceiling makes it harder for the attic to communicate with the conditioned second floor. This may be the difference from our house to the others that have this same system. I think we probably could add a supply and return in the attic and solve the issue - before we accpet that as the fix however, we would lke to know why others do not need this and why ours does. The other thing we would need to look at if we are going to actively condition this area is to be sure our HVAC unit is sized to handle the additional volume.

    As for the cost - the HVAC system was around $17K, I think the foam was between $2K - $3K. The audit was $400 and the vapor barrier and air sealing of the crawl and attic was $1500. Removing the existing insulation in the attic was $500.

    Thanks for all the feedback - happy father's day to all.

  21. Mark Tyrol | | #21

    Attic tent does not meet Code - no insulation.

  22. wjrobinson | | #22

    Danny.,... You so agree exactly with what I see and I am hundreds of miles away. Yeah... I think you have nailed it and should not be concerned with additional HVAC load. Try simply adding a tiny bit of conditioning... if it works... move on. You may be on to something as to the plaster...

    As to my install... we did NOT seal ceiling plane... I think that was a mistake... Either you are conditioning the attic or you are not.

    Thanks for the fathers day wishes... my is going great... on to the next ho down...

  23. Danny Kelly | | #23

    Yes Mark - I understand the attic tent will not meet the 2009 IRC Energy Code for new construction but here in NC that will not be adopted until 2012 - I am hoping by then there will be some more options to choose from. Currently to pass inspection the generally accepted practice is to stuff a fiberglass batt between the stairs and the plywood panel of the pull down stairs which we all know does no good at all. [I am not arguing that this is acceptable, just explaining current reality]. Currently, 95% (unscientific poll conducted by me) of the houses we go in the PDS have no insulation or poorly insulated as mentioned above. The attic tent is a cheap fix that will greatly reduce air infiltration, as we all know, most of our heat loss (and gain) is through air leakage so this is one of the low cost high return products that is a no-brainer to install. If there are better products out there, I would love to hear about them. I have seen many similair products but the only one I have seen with the proper R-value is a "coffin" which seems to be pretty costly. (Maybe a good idea for a future product highlight article on GBA. ) Anyway - we did not install one in this case since we were moving the thermal envelope from the ceiling plane to the roof sheathing.

    Another problem we see often is when there is only a central return in the upstairs - for some reason these always seem to be right next to the pull down stairs. The PDS rarely have a good seal around them so the return air is sucking the hot air out of the attic rather than air from the conditioned space and homeowners say they can't keep their second floor cool. Simply installing the attic tent alone has solved this problem in several houses for us. Although not a perfect product it does have its upsides and can typically greatly improve current conditions.

  24. adkjac | | #24

    Danny... You keep mentioning... exactly what you did that worked which is what I suggest and exactly what did not... Break the seal to the ceiling. That is what you know works and what I also think will work. The leaky pull down stairs helps you since you want that to leak being that the knew air barrier is the roof.

    I hope you go forth soon and do... then get back and tell us about the success.

  25. adkjac | | #25

    I swear I did not type knew for new... what happened?

  26. Danny Kelly | | #26

    A quick update for everyone - thanks for all the feedback.
    Went by and took some more measurements today and ran the blower door - a few stats:
    Interior of house - Temp 78 degrees and RH 47%
    Exterior - T - 98, RH - 45%
    Attic - T - 86, RH - 54%
    Crawl - T - 73, RH - 80%

    Ran the blower door with attic hatch closed - 4600 CFM50 (pre-retrofit blower door was 5800)
    Opened the hatch - went up to 4700 CFM50
    Pressure from interior of house with reference to attic: -14 (Put hose thorugh PDS)
    Pressure from attic with reference to exterior with PDS open: -46Pa
    Pressure from attic to exterior with PDS closed: -25Pa

    Not sure what some of these necessarily tell us but I think the good news is:
    1. RH in attic fell from 85% to 54% in about 5 days - starting to feel pretty good about our theory of trapping moisture in the attic during the install.
    2. Feel like the foam did a good job of sealing the attic from the outdoors A. blower door only jumped from 4600 to 4700 when opening the PDS. B. When measuring the attic with reference to outdoors got a -46Pa which shows it is not very connected.

    Not quite sure what to think about the -25 Pa when the PDS is closed. I think all that is telling me is that the size of the hole from the attic to the outside is the same size as the size of the hole from the attic to the conditioned part of the house. Could be 10 square inches or 10 square feet.

    I think we are going to continue to run the dehumidifier until the attic reaches equilibrium with the house. Once there - will turn off the dehumidifier and see if it maintains the same RH as the house. Either way, I think we decided it makes sense to add the registers in the attic - both a supply and return. Cannot see the down side of doing this and not sure why this is not done in all semi-conditioned attics. We may install dampers to close these registers in the winter to try to prevent possible issues with water condensing on the roof deck. Can anyone think of a reason NOT to install the supply and return. Was thinking just adding a small one of each - balanced - to not increase the load on the system.

  27. homedesign | | #27

    Not sure what some of these necessarily tell us but I think the good news is:
    1. RH in attic fell from 85% to 54% in about 5 days - starting to feel pretty good about our theory of trapping moisture in the attic during the install.

    Still too early to tell.... lots of variables.
    There is a thread at JLC about another out of control attic(spray foamed at the roof deck)
    The owner reported larger RH swings (40% to 76% same day!) depending on the time of day and Sun conditions.
    Also is it possible that that the ATTIC RH is going down BECAUSE you are running a dehumidifier?

    I think that with all the framing wood and spray foam in the attic that there can be large quantity of water storage that can charge and discharge.

  28. Danny Kelly | | #28

    Yes John - perhaps we are a little too optimistic. I did read the JLC thread - that seemed even more unusual that ours. With the huge variations he is experiencing sounds like he has some connections the exterior and a lot of solar vapor drive going on - especually since he mentioned he typically had the higher RH on days that it rains. Our situation does seem pretty similar although we have not experienced the up and down that he has, have been getting a steady decline. Although you are correct, probably mostly due to running the dehue. We have also been leaving the pull down stairs down each morning to allow the HVAC system to dehumidify some as well, so we do have several forces currently working in our favor. I would like to hear from the foam contractor how they think the solar vapor drive is handled. I guess time will tell - real test will be when we turn off the dehumidifier and leave the stairs closed. May be up to a weel before we are ready for that.

  29. adkjac | | #29

    The house I have to check has no issues. The air in the attic is slightly higher or lower temp depending on season but I have never noticed moisture issues... vapor drive etc. I still think something is different... And so far if the home is built right, the differences are you sealed the ceiling up and I did not, and you did a retrofit, where I did new construction. I would run the humidifier via sensor to turn it on. And I would leave it. If you do have encapsulated moisture you are protecting the home from high RH and in a year at most all will be dried out. After a year... you may see no need. Could happen quicker but I don't have a clue. The spray foam manufacturer should be able to help you I would think too.

    You are on the right track... and IMHO doing a very good service for your customer.

  30. Greg Pruitt | | #30

    I have been in the foam business for 6 years. Retrofit installations is over 50 % of our business. Have fnever seen a major problem with humidity in the attic after a few weeks. I suspect that part of the described problem is the plaster ceiling. Sheetrock leaksabout 20% of it's air. The plaster is probably not allowing any escape. a sealed attice depends on leakage from the duct work, can lights and sheetrock to help stabilize the temperatures in the attic with semi-conditioned air. I encourage air leaks. I wouls recommend a air duct and a return. this will solve the problem if you have a complete seal on the roof deck including the rafters. We spray at least 1.5" of foam on the rafters and 5.5" betwee them. We see other contractors spraying 3.5 to 5" and not coating the rafters. You get what youo pay for, and the problem is so many new contractors that really don't care. Our houses seem to out preform most others. On new construction wit full fill the sutds, and caulk every joint including the top and bottom of the studs.

  31. GBA Editor
    Martin Holladay | | #31

    I have to strongly disagree with your statements.

    You wrote, "A sealed attic depends on leakage from the duct work, can lights and sheetrock to help stabilize the temperatures in the attic with semi-conditioned air. I encourage air leaks."

    Duct work shouldn't leak; nor should a drywall ceiling -- even if the attic is conditioned. Discussion of these issues requires "house as a system" thinking. I shiver when I hear that you tell homeowners that you encourage air leaks, leaky ductwork, and can lights.

  32. wjrobinson | | #32

    Martin, Greg has been doing related work for 6 years... I definitely feel something is going on with the fact that the ceiling plane was sealed so well. Experience is everything sometimes. The one home I have done does not have the ceiling specially sealed. And all is well. Until that attic is doing better it has to be conditioned somehow. Refer to my post above. And cut Greg some slack for having successful experience.

  33. MICHAEL CHANDLER | | #33

    I also do all I can to minimize air exchange from the sealed attic to the house. Where we have a pull-down stair or attic storage access we generally build a storage room with 1/2" drywall between the living space PDS or door and the spray foam, flat taped and air sealed to minimize air exchange between the living space and the foamed attic assuming the PDS is not an air barrier and that a door that can be left open will be. If there is mechanical equipment in the attic it is not located in the storage cube.

    So my advice would be to improve on the attic tent to make a more durable seal there but not to open a supply and return into the attic to "condition" it. there will be plenty of heat transfer through the uninsulated drywall ceiling, bulk air (and moisture) transfer is not in anyone's interest from my perspective.

    Once you have the humidity stabilized and are assured that it is not due to a roof moisture or flashing leakage issue then I would suggest that you get a humidistat switch ( and connect it to to turn on a warning light near the parking area or wherever seems noticeable but inconspicuous to signal when the attic humidity exceeds 60%. (Works well to monitor humidity in sealed crawls also and you can drive by to check without bothering the occupants)

  34. Riversong | | #34

    AJ says "Cut Gregg some slack".

    Greg doesn't understand the most important element of a healthy and efficient house, and he suffers under the illusion that "Sheetrock leaksabout 20% of it's air" (whatever that means). In fact the airtightness of ordinary drywall is the standard used by the Air Barrier Association of America against which to measure all other air barrier materials.

    "Experience is everything sometimes"? Not if it's based on ignorance of building science and repetition of poor practice.

    Greg is a perfect example of the quality of spray foam "professionals", and the reason there are so many problems with that industry.

  35. MICHAEL CHANDLER | | #35

    It is really amazing how many things we can do wrong without causing major building failures.

  36. homedesign | | #36

    Michael Chandler:

    Once you have the humidity stabilized and are assured that it is not due to a roof moisture or flashing leakage issue then I would suggest that you get a humidistat switch ( and connect it to to turn on a warning light near the parking area or wherever seems noticeable but inconspicuous to signal when the attic humidity exceeds 60%. (Works well to monitor humidity in sealed crawls also and you can drive by to check without bothering the occupants)

    How often do you drive by?
    What happens when you retire?

  37. homedesign | | #37

    Do you know why conditioned attics are cooler than traditional attics during the summer?

    Because they are Air Conditioned.

  38. Danny Kelly | | #38

    Since this thread has resurfaced, thought I would give an update:
    After running a dehumidifier for about a week, we did get the attic to maintain reasonable levels without running it at all. We tracked the levels in the attic for two weeks afterwards with average temp being 84 degrees and average Relative Humidity at 75%. High and low temps – 90,73 and high/low RH – 72% and 58%. Though the RH was a little on the high side but probably within reasonable levels for the most part. I think the plaster ceiling and the fact that it was a retrofit and not new construction was our main problem and just took the attic awhile to acclimate from being vented to non-vented/semi-conditioned. One interesting result I noticed: I have tracked my own vented attic before and typically during the day as the temperature gets really high, the RH goes down. Makes sense, the amount of moisture is the same as what is outdoors but the higher temp. can hold more air. But in the unvented attic, as the temperature went up, so did the RH while the temp and RH in the conditioned portion of the house remained pretty constant. I wonder how an unvented attic should perform

    After this and reading several other threads – I am beginning to question the benefit of an unvented attic. The fact that the entire premise of the theory is that is relies on leaks in the ductwork and leaks in the ceiling is questionable. If this is true – how much leakage is required and how much is too much? If you ask a foam contractor about air sealing at ceiling level and sealing duct leaks – he will tell you not to do it, that it doesn’t matter since the attic is conditioned. But the attic is not conditioned – it is only semi-conditioned – the attic will generally stay with 8-15 degrees of the house. If you ask a weatherization/building science expert, he will tell you to still air seal at attic and seal all duct leaks. This makes more sense to me – you want the air to go where it is designed, and when you have a 10-15 degree delta T between living area and semi-conditioned attic, you certainly do not want your conditioned air to exchange with the semi-conditioned easily. So brings me full circle – what is the benefit of a semi-conditioned attic? I guess I’ve always heard – stops air flow/stack effect and keeps your HVAC unit and ductwork inside conditioned space.

    If your HVAC unit and ductwork are not in your attic – I do not think there is much of a benefit of the unvented attic – more risks with moisture and not really much of a reward. Just be sure you have good air sealing at the ceiling levels and high R-Value insulation. I can appreciate the benefit of having your HVAC unit and ductwork running in 80 degree air vs. 120 degree air – but do those savings really offset having to condition all of that additional volume? According to my HERS rater – the answer is no. Would be curious to see more information on this. Secondly, the majority of the heat in our attics is radiant heat – with an R-8 foil faced duct with air running at 80CFM – how much heat can really be gained by the cold air in these ducts. Seems to me you can get similar results with a light colored roof and a radiant barrier in conjunction with good air sealing and insulation.

  39. MICHAEL CHANDLER | | #39

    for years I've been using those cheap, inaccurate radio shack thermometer / hygrometers in crawls to check for humidity issues as necessary but recently I had a design customer ask about a way to cheaply monitor humidity in sealed crawls in his rental houses. He doesn't necessarily want to interact with his tenants to check the crawls and the tenants (and my customers) don't want to have to look under their houses to check the humidity (and I don't want them obsessing over every bump in the humidity which another one of my clients does with a remote sensing hygrometer in her kitchen which monitors the crawlspace humidity) so we came up with the idea of setting a light up in a place where it would be obvious during the daily life of the occupant and also obvious to the casual drive-by of the landlord when there was a serious moisture event in the crawl. Locating it in the parking area makes it plenty obvious to the occupant but not as alarming as a buzzer or read-out.

    I've had a fair amount of malfunction with larger fan loads on motion sensor switches and don't want to put the load of a dehumidifier on the hydrostatic switch for fear of similar malfunctions. I do keep a dehumidifier in my steel-lined document storage cube here that has an internal humidistat switch but somehow I don't feel like dedicating a dehumidifier to a system that when operating correctly doesn't need one and when subject to a plumbing leak or gutter overflow needs more than dehumidification. I'd rather get a call from the customer saying that the crawlspace warning light is on.

    When I retire I'm sure those calls will continue to come to me and I'll hopefully refer them to someone else to go look in the crawl for the leak. But then I'm not really planning to retire either.

  40. Riversong | | #40


    Now that your attic has been dried and acclimated, it appears to be closely tracking outside conditions.

    With the reported highs and lows, if the low RH coincides with the low temp (72°/58%) and the high RH coincides with the higher temp (90°/72%), then the attic is closely tracking indoor absolute humidity when it's cooler and rises higher than outdoor absolute humidity when it gets hot (which is almost certainly when the sun shines).

    At 75°/45% indoor conditions, the mixing ratio (lbs water/lbs air) is 0.0082.
    At 85°/75% outdoor conditions, the mixing ratio is 0.0185.
    At attic low temperature, the mixing ratio is 0.0099 (slightly above room conditions).
    But at attic high temperature, the mixing ratio is 0.0215 (16% higher than outdoors).

    This shows that solar radiant drive on a dark roof is moving moisture through the open-cell foam into the attic and that vapor pressure differentials are insufficient to move that moisture back out. Solar drive wins and the attic remains humid as long as the sun shines.

    Install a very light colored roof and the attic will equilibrate at a humidity level closely tracking indoor conditions.

    Moral: dark roofs, unvented attics, and open-cell foam don't make a good partnership.

  41. Danny Kelly | | #41

    Thanks Robert - I kind of suspected that the vapor drive had something to do with it, nice to have some math to back it up and help understand it. I was having a hard time understanding how there could be a vapor drive with lower humidity outside than there was inside. A typical days reading was attic temp - 75, RH - 63% and outside temp 75, RH 79% in the morning and around noon going to - attic temp 88, RH - 70% - outside temp 95, RH 40%. When I get a chance to review your ratios more closely, may help me decipher this. I guess when considering vapor drive, we should be looking at the temperature of the shingles in lieu of the outdoor conditions too.

    If vapor is passing through the foam so easily, I wonder if we are going to have bigger problems in the winter when the "conditioned" air is hotter and more humid than the outside air. Sounds like we could have trouble.

  42. homedesign | | #42

    Danny & Robert,
    Bear with me and/or correct me
    I am trying to learn/visualize this stuff

    Have the conditions in the attic really improved after the "Drying"?

    In Terms of Absolute Humidity (Grams per Cubic Meter)
    The Afternoon Attic Absolute Humidity seems higher now

    I visualize a cubic meter of air as a cube about 39 inches tall
    I visualize a gram of water to be about the size of a sugar cube

    from Danny's original post
    afternoon attic condition = 84degrees/75%= 22 g/m3

    from Danny's recent post
    afternoon attic condition = 88degrees/70%= 23 g/m3 = higher dewpoint

  43. homedesign | | #43

    I still think the main problem here is not ONLY solar vapor drive.......
    the problem was a less than wholistic approach
    the homeowner did his own pick and choose from the recommended Recipe instead of following the Recipe.

  44. Danny Kelly | | #44

    Hey John - I am with you - still trying to figure this out and wouldn;t say I have many answers. I do think we are still occasionally on the high side with our RH but the fact that it came down for a few days and the condensation went away (was all over the boots and HVAC unit) we figured we were making progress. Was consistently very high initially. How are you guys calculating water to air ratio?

    John - what parts do you think are missing from the recommendations. We were not involved in the initial audit nor the foam - all we did was the air sealing and vapor barrier in the crawl space but all that he was doing seemed to make sense to me as far as prioritizing recommendations. The audit probably had 40 recommendations everything from changing HVAC unit to chaging light bulbs. The owner did the majority of the recommendations to the tune of $25K+. The only recommendations not followed from a thermal envelope standpoint was insulating exterior walls, closed crawl and windows - all we thought could wait until a phase II. I initially thought the no insulation (we think) in the exterior walls was probably our problem but the interior conditioned space seems to be performing well.

  45. Danny Kelly | | #45

    Oops - I think I caused some confusion - had some inaccurate numbers:
    Average attic temp - 80 degrees
    Average Attic RH - 63%

    I'll be happy to email you the entire spreadsheet if you are interested - hit me at - dkelly at kellymcardle dot com

  46. Riversong | | #46


    The formulas are quite complex, but you can get a good approximation using a standard psychrometric chart: start at the air temperature on the bottom axis, go up to the curved RH line and then follow the horizontal line to the vertical axis on the right for humidity ratio (or the slanted line to the left for the dew point temperature).

    Your latest temp/RH readings show this:
    AM: mixing ratios of 0.0115 in and 0.0145 out - 0.14" HG (0.07 PSI) vapor drive inward
    PM: mixing ratios of 0.0196 in and 0.0139 out - 0.27" HG (0.13 PSI) vapor drive outward

    This shows that the outside absolute humidity is approximately the same morning and noon, but the inside absolute humidity increases 70% from morning to noon, even though there is a strong outward vapor pressure drive.

    Yes, I suspect you'll have wintertime condensation under the roof deck from outward vapor pressure drive and insufficient solar drive to reverse the flow and dry the deck. Open-cell foam should be sprayed only under a vented roof. The venting also eliminates the inward radiant vapor drive and allows the roof deck to dry.

  47. homedesign | | #47

    I used this online calculator:
    For me the Absolute Humidity (grams per cubic meter) is easier to visualize.
    The calculator gives results in mixing ratio as well.

    I was thinking that the homeowner should have closed up the crawlspace
    especially since there is an air handler and ducts in the crawlspace.
    AND attempted some airtightening at the walls.... The house still sounds very leaky.
    AND since he chose to foam the roof deck... I think he should have gone for more R-value and at least covered the rafter bottoms some to cut down on thermal bridging.
    Those attic temperatures and temperature swings sound really high and I think it is related to the low r-value and exposed rafter bottoms.

  48. Danny Kelly | | #48

    Good points John - will look into these for him - I guess he could still add a few inches of foam to what he has there now.

    As for the crawl, was in surprising good condition for a vented crawl and he didn't even have a vapor barrier on the ground. We did crawl air sealing and installed a vapor barrier and installed a datalogger to monitor conditions to see if further work is required. Should have results on that in a few more weeks.

    Thanks for all the thoughts everyone.

  49. BobHr | | #49

    My initial thought is that there is not enough insulation. 2 x 6 rafter with open cell is about an R20-22. I did not reread the post but was the attic sealed off from the living area. If the air barrier is at the roof level then the attic should be coneected to the interior. Is anything being done for thermal bridging through the rafters. No mentioned of drywall covering the insualtion.

    How are the bathrooms vented? Do the bathroms have fans and are they vented they vented to the exterior. Are they being used.

    How is the drainage on the site. The crawl space floor was sealed but how well. How about moisture trough the crawl space foundation wall. If the humidity level is high can some of the be making it to the attic via gaps around the plumbing stack, flue, etc.

    Was a blower door test done after the improvements to see how air tight the house is.

    The attic temp was between the indoor and outdoor temps but is relative humidty is much higher than both. To me it sounds as if the moisture is getting into the attic but can not get out. Again I think the attic needs to have air cirulation between it and the living area.

  50. Bryce | | #50

    Jumping in late on this issue and have not followed the thread completely, but from what i've read, my guess is that crawlspace and household moisture are rising to the top....warm air rises carrying h20 with it. Lots of discussion about chases and by-passes, but no discussion if the crawlspace or attic were pressure tested WRT to the building envelope? Using a second manometer, check the pressure difference between the interior of the home to the attic and then home to the crawlspace. If the pressures are less than the house to the exterior, then the two spaces are connected, proving there are still chases/by-passes left unsealed. The blower door is a great tool, just don't stop at a simple test to the exterior.

  51. Riversong | | #51

    For me the Absolute Humidity (grams per cubic meter) is easier to visualize.

    Water vapor is invisible - you can't visualize it, even if it's absolute ;-)

    I visualize a cubic meter of air as a cube about 39 inches tall
    I visualize a gram of water to be about the size of a sugar cube

    Well, the cubic meter of air would be 39" tall only if it was also that wide and long, but keeping air in a perfect cube is like herding cats, only moreso ;-)

    And a cubic centimeter of water would be about 3/8" on a side (to a cat herder), a bit smaller than a sugar cube (which is a teaspoon, or a cube approximately 11/16" on a side). Perhaps you're thinking of a gram of hashish ;-)

  52. homedesign | | #52

    Just remember I am from Texas

    I know my approximations are "rough" kinda like "Joe-math"
    But I think it is useful to try and put things in perspective.
    How much water vapor is in the house, attic and crawlspace?
    I think I picked up the sugar cube approx. from Bill Rose Book.
    please bear with my visualization fetish(I am an Architect)
    warning there may be math errors to follow!!

    I realize that Temperature and RH are averages of many different conditions in a region.
    Some packets of air in some parts of the region can be warmer,wetter,drier or cooler

    Getting back to Absolute humidity
    I do not know the dimensions of this attic... but if the 3,000 sf house is full 2-story then the footprint is about 1,500 sf or 140 square meters
    assume average attic height about 1.5 meters then attic volume roughly 210 cubic meters
    assume crawl space 1 meter high ....volume approx 140 cubic meters
    living area 2 meters high x 140 square meters x 2 floors = 560 cubic meters

    Living area seems to stay around 10 grams per cubic meter = 5,600 grams of water
    1,000 grams of water per liter ... then Living area contains about 5.6 litres
    I visualize a litre to be about the size of a litre of Dr Pepper ;-)

    Attic in the morning = approx 14 g/m3 x 210 m3 = 2.94 litres
    Attic in afternoon = approx 23 g/m3 x 210 g/m3 = 4.83 litres

    Attic is fluctuating between 3 and 5 litres

    So... did 2 litres diffuse though the roof assembly via solar vapor drive?
    I thought diffusion was a slow process?
    Or ... did some of the water vapor migrate from the crawlspace and the living area?

    I am thinking that the living area is not showing a high water content because the AC is running full out.

    But couldn't some of the more energetic "warmer and wetter" packets of air migrate or "float" up to the attic?

    Is it also possible that the wood and foam in the attic are absorbing/adsorbing moisture at night and then releasing the moisture when the poorly insulated roof heats up during the day?

  53. homedesign | | #53

    meant to say
    Attic in afternoon = approx 23 g/m3 x 210 m3 = 4.83 litres

  54. BobHr | | #54

    I finished reading page 2 of the responses. I think RH over 50% could lead to mold problems in the long run. You have some choice on contrilling the humidty. Get better air circulation be attic and living areas or run a humidifier.

    I would say that your attic is under insualted. Because of that you fear providing more air flow between the living space and the attic.

    I have heard the argument that open cell will show the roof leaks. I take the view that most roofs are going to leak and that many of the leaks start out very small. To me the insulation may slow drying and lead to greater damage. Damaged roof decking is going to be more diffiult to replace as the insualtion will be attached.

    It is my personal opinion but I do not like the idea of insulating the inside of the roof deck unless the outside is also insulated and a more weather resistent roof is installed. There is movement to going to rain screen walls as this will allow drying of clading and draininage of bulk water that gets behind it. I think a similar approach should be taken with a roof. The building science guys have shown in thier paper "mind the gap" that small gaps can accomlish a lot. Double layer of house wraps, light scoring of the back of rigid foam etc can provide both vapor release and small amount of bulk water.

    Using this philosophy I think the top of the roof deck gets 2 layers of a suitable vapor barrier. Then rigid insulation and a roof that has at least a 50 year life that is back vented and drained.

    For this homeowner they would not have the vapor drive if there was 2 layers of rigid insulation on the outside. His R value would be in the R30 to R 60 range for his area. Currently he has an R19.25 (3.5 x 5.5inches) This is well under the R30 minimum and only 1/3 of the R60 maximim.

    Adding 4 inches of rigid at 6 per inch would add an R24 plus the R19 for a total of an R36. Hardly an extreme level of insulation. What it would do is cut down on vapor drive and thermal bridging. The multiple layers woud allow the deck to dry in both direction and it would give a path for bulk water drainage.

  55. Riversong | | #55


    Was that you narrating the cat herding video?

    You must have a different Joe-Math down there - not the same as Joe (Pi) Bartok, who can heard cats with the best of them:

    And, I didn't think you had square meters in Texas. I thought it's all big pointy feet and tall hats.

    Vapor diffusion is slow in Vermont, but real fast under the Texas sun. Actually, it's slow when driven by vapor pressure, but pretty speedy when solar powered since it's an energy-dependent process. And you're right that anything hygroscopic will suck up moisture and then release it to the air when heated.

  56. Riversong | | #56


    I've never seen any code or engineering standard for "maximum insulation". It's more a cost-benefit issue. The ideal insulation would be of infinite R-value.

    Sandwiching any wooden or cellulosic material between to layers of relatively impermeable insulation or other materials (like a roof-top vapor barrier) is asking for trouble, not preventing it.

    And we've had "rainscreen" roofs for decades - they're called vented roofs. Most roofing materials, except wood shakes or thatch (or any other natural material) do not need an air gap under them - they are designed to be water-proof or water-resistant. A vented roof, however, allows drying of moisture penetrating from either above or below and is the most reliable method of creating a durable roof assembly.

  57. homedesign | | #57

    When I read about solar vapor drive with Brick veneer.. the examples are always after a rain event.
    How about Solar Vapor Drive thru roofs?
    Is the extra moisture from "morning roof" alone enough to contribute to Significant Vapor Drive?
    Or does it take a rain event for significant vapor drive?

  58. Riversong | | #58

    Oh, yeah...

    I visualize a litre to be about the size of a litre of Dr Pepper

    A liter is actually about the size of a liter of Coke or Pepsi or Sprite, not Dr. Pepper, Sarsparilla or Moxie.

  59. Riversong | | #59


    Since most bulk water from a rain event rolls off a roof (that's what it's designed for, after all), I would think that a morning dew would wet a shingled roof just as much as a rainstorm. In fact, the saturated morning air would penetrate the layered shingles more than a surface rain.

  60. Riversong | | #60

    Robert Hronek:

    I think RH over 50% could lead to mold problems in the long run.

    Indoor humidity levels during the heating season greater than 40% can cause sufficient vapor diffusion and air-borne moisture transport to create a potential for condensation within the thermal/structural envelope and consequent mold or rot.

    But it requires a constant 80% RH to create surface mold conditions, and 90%-95% RH to sustain conditions conducive to decay organism growth.

  61. homedesign | | #61

    Hi Danny,
    How about an update?

    and @ post #9
    did you slip up and reveal your secret identity?

  62. MJL Spray Foam Insulation | | #62

    Any Updates Danny? I have been meaning to ask.


  63. Lander4 | | #63

    Interesting stream. Would love an update. Solar drive through open cell makes sense. If I had to bet, I would say that vented crawlspace is a contributor.
    @MartinHolladay - I don't understand the adamance about sealing encapsulated attic from living space. If encapsulation is meant to bring the attic inside the envelop, why create a trapped air buffer? (Particularly if there is HVAC equipment in the space.) Encapsulated crawlspaces are frequently encouraged to have conditioned air (since it is now inside the envelop), so why not the attic?

  64. GBA Editor
    Martin Holladay | | #64

    1. This thread is 4 years old, so of course I have forgotten all about it.

    2. Even though it is 4 years old, I went back and looked at what I wrote. I may have missed it, but I didn't find that I had posted the comments that you attribute to me. I don't think that I wrote that an unvented conditioned attic needs to be separated from living space below by an air barrier at the attic floor. So I don't know why you got that impression.

  65. Lander4 | | #65

    Hi, Martin.

    I read old threads if they are a topic/issue of interest.

    Sorry if I misinterpreted. In comment 31 you state, "...Duct work shouldn't leak; nor should a drywall ceiling -- even if the attic is conditioned…"

    and then Michael elaborates in 33.

    best regards, Lander

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |