Hawaii home attic considerations and passive cooling with TPO — Controlling temperature and moisture
Hello,
I am looking for help in Hawaii — we have had many challenges with our roof and now trying to get our home cool. We are very sensitive to anything that could cause water damage or mold.
Home: Single Wall, 1961 single story, 2000 square feet
Location: Oahu, Hawaii (climate zone 1)
½ of home has shallow ceiling (max height ~ 24”) with drop ceiling
½ of home is open beam ceiling
Insulation: Attic has faceless fiberglass insulation installed on top of the shallow drop ceiling
Roof: 2/12” Pitch, 3 years old. 5/8” tongue and groove plywood with Peel and Stick Bitumen (CertainTeed Flintlastic)
The home has NO Air Conditioning, NO heat
The windows are open nearly 24/7 all year
We generally have trade winds that blow 12-20 mph
Temperature: Lows 63, Highs 85 Humidity: 50-80% RH
PROBLEM: Home is excessively hot – 100 degrees in summer
We are considering adding a TPO membrane to our roof.
CONCERNS — Ensuring sufficient attic ventilation to:
1. Reduce any moisture problems
2. Release heat absorbed by insulation during day so it does not radiate into home (seems to happen even though heat rises)
3. Help cool the home
QUESTIONS:
1. Could adding a TPO roof cause moisture problems?
2. What do we need to consider about attic ventilation?
Is there a specialist we should consult to ensure this is done properly?
Thank you for any help!
Here is a good link to the current climate:
https://forecast.weather.gov/MapClick.php?w0=t&w1=td&w2=hi&w3=sfcwind&w3u=1&w4=sky&w5=pop&w6=rh&w7=rain&w8=thunder&AheadHour=48&FcstType=graphical&textField1=21.3&textField2=-157.86&site=all&unit=0&dd=&bw=&BackDay.x=44&BackDay.y=6
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
User-7431379,
First of all, can you tell us your name? (I'm Martin.)
What do you mean by a "drop ceiling"? If you are talking about suspended ceiling panels set into a metal grid, that means you don't have a ceiling air barrier. The insulation installed above that type of ceiling is almost worthless -- you can't get any value out of insulation unless the insulation is adjacent to an air barrier.
It isn't clear that you'll be able to lower your indoor air temperature very much unless you are willing to close your windows and install an air conditioner. But if you want to see whether ceiling insulation helps, you should consider:
1. Installing an airtight drywall ceiling (assuming that you currently have a suspended ceiling).
2. Installing a continuous layer of rigid foam insulation above the roof sheathing of the part of your house that has no ceiling insulation.
Another possibility is that you have enough glass to permit solar heat gain through your windows. I'm just guessing, though -- if your windows are all open, that wouldn't be happening.
Martin,
Thank you.
By "drop ceiling," I mean a shallow attic space. The ceiling is Densarmor Plus. The fiberglass batts lay on top of this. There was no attempt to seal the ceiling - it was just coated in drywall mud and prime/paint. I don't know how "airtight" this ends up.
We have considered foam on top of the roof, under TPO. The roofer we are speaking with thinks TPO will alone cool the house sufficiently. My concern is what if it does not... and then I have spent all that money, and we are still very hot inside.
I started getting concerned about moisture after reading some articles on your website.
One concern was that TPO was too cool and would not allow the attic to get sufficiently warm to dry out any moisture accumulation (likley natural in humid climate + rising hot moist air from house?). Other discussions were about the cool roof causing condensation on the inside of the roof deck.
I don't know if these are possible in Hawaii since we are pretty warm all the time.
Windows -- we have lots of them and they are not very good (I assume low E value). Three panel awning type.
>"One concern was that TPO was too cool and would not allow the attic to get sufficiently warm to dry out any moisture accumulation (likley natural in humid climate + rising hot moist air from house?)"
That's REALLY unlikely in a US climate zone 1 location. The exterior surface of the TPO will never drop below the outdoor dew point temperature, and if you're cooling at night with the windows open the indoor dew point won't be appreciably higher than the outdoor dew point as long as you are using reasonable exhaust ventilation while cooking or bathing. During the heat of the day any nighttime moisture accumulation would be purged by the heating of the roof deck.
Installing 5-perm perforated radiant barrier on the underside of the rafters would limit amount of moisture cycling in the roof deck too, raising the average temp of the roof deck (warmer = drier) while lowering the temperature in the attic.
Sorry Martin -- my name is Andrew.
I totally missed that...and, that I was showing up under a user ID!
Thank you for all your help
>"2. Release heat absorbed by insulation during day so it does not radiate into home (seems to happen even though heat rises)"
The thermal mass of fiberglass batts is negligible- they don't absorb or re-radiate very much heat at all. The roof deck and rafters and even the drop have far more thermal mass for storing and re-radiating heat. Fiberglass is somewhat translucent to radiated heat, making it less optimal for insulation under a hot radiating roof deck.
Heat isn't affected by gravity- it doesn't rise. Air is affected by gravity. Hot air is less dense than cool air, and is affected by gravity, causing the cool air to sink, hot air to rise. But radiated heat is always from hot to cold.
Mounting perforated aluminized fabric type radiant barrier on the underside of the rafters will reduce the amount of heat radiated at your fiberglass insulation without creating a moisture trap. It won't be as effective as installing 8" of (IR-opaque) cellulose on your suspended ceiling to bring it up to code minimum, but it would make a noticeable difference, especially if the fiberglass is low-density R19 or R11 goods. A light colored TPO roof will also reduce the roof temperature, reducing the amount of stored heat available for re-radiating to the interior.
Dana,
Thank you for the reply.
What gave me pause is this – the bedrooms covered by attic space with fiberglass insulation warms slowly throughout the day, while the open beam high ceiling area (no attic or insulation) warms very quickly. However, in the late afternoon, the temperatures swap. The open beam room also cools quickly. I assumed that the attic/insulation was absorbing heat, holding it, and then re-radiating it into the living space with overall higher thermal mass?
I have attached a comparison chart (with Govee thermometers). On April 21, the open beam room hit 91 degress at 1317. The bedroom closet (attic + insulation) was 82 degrees, and still 81 degrees at 2248. By that time, the open beam room had cooled to 79 degrees.
Our attic is mostly too small for much access to do anything to now. I thought about E barriers...but, did not do that and the contractor did not recommend it at the time.
The rooms are behaving exactly as one might expect. The insulation reduces the rate at which the room gains heat, but also reduces the rate at which the room loses heat.
I still think that Dana's idea of reducing attic temps with a radiant barrier is all that you need. You have to date most probably been relying on air movement (open windows and maybe efficient ceiling fans?) for thermal comfort and that should continue.
I just pulled up Climate Consultant 6 using a Honolulu weather file; see attached screenshot. You should play with this free software to get a sense of the impact of various strategies on thermal comfort. Pretty interesting: https://www.greenbuildingadvisor.com/article/how-to-use-climate-consultant-4. In the attached it shows dependence on active cooling for about one-quarter of the time, but I bet you can get that number down quite a bit OR expand your sense of comfort beyond ASHRAE Standard 55...
Peter
Regarding the time lag of the insulated vs. open beam rooms, the insulation prevents the roof deck /attic from cooling as quickly as in the open beam room, delaying the peak ceiling temperature. There is heat being stored, but it's stored in the thermal mass of the rafters & roof deck, not the insulation.
The amount of ventilation to the open beam room relative to the other rooms is another factor that can affect temperature lags too. How much ventilation is there in a bedroom closet, really? Compared to a bedroom with an open window under the insulated attic or an open space with an open beam the closet ventilation is pretty much zero. Unless you're sleeping in that closet it's not a particularly useful or important comparison.
The delay due to the insulation will still be there even after the TPO roof goes up, but the peak won't be as high. Ventilating the attic with soffit to ridge venting to allow convection cooling of the attic would shorten the delay and flatten it even further. Higher pitch roofs work better for convection cooling (both at the exterior roof and attic ventilation) than low-angle roofs.
A spray-on aluminized low-E paint isn't as effective as a radiant barrier, but might still do the trick if the TPO roofing isn't quite effective enough.
Great ideas and thoughts!
Near term, it sounds like concerns of moisture in the attic are not warranted – thank you!
It really sounds like I need some sort of reflective coating on top of the roof, and am leaning toward TPO.
TPO or Insulation + TPO? Any ideas on how I can know if TPO alone will cool us down “sufficiently?”
Or, do we really need insulation under the TPO?
Sounds like plan would be for 2-3” of Polyiso (~ R12-18).
Could this go on top of the Bitumen layer?
Radiant barrier – I wish I joined this group sooner! I don’t see a way to install it now that the ceilings are in place since the attic access is so restricted. Also, it won’t work on the open beam section.
The ceiling in the open beam area is 117+ -- like standing under a heat lamp in summer.
Ceiling Fans – yes, in every room and help a lot
Climate Consultant 4 - looks like a great tool, thank you. I agree for a “normal” home… we
need to cool our roof first.
Venting the Attic – does this need to be extreme like a doghouse… or just increase the eve vents (can increase the number of these) to allow more input and output through soffit (not much room to add) vents? I do have a powered fan in on soffit on the downwind side of the house in hopes of reducing pressure and allowing better flow out. However, after reading in these forums some…I am doubtful that will help.
Closet Measurement – just a convenient place to put the Govee at the moment. Performance inside this walk-in is basically the same in the main bedrooms.
Ventilation in the Open Beam Room – It is ok, but there is no ventilation up high. Biggest cross ventilation comes through front door and out back of home (glass sliding door). I have thought about an exhaust fan near the peak on the side of the house…but, not sure how effective this would be.
Without knowing the thickness of the fiberglass it's hard to make a call on whether roof deck insulation is warranted, but given your current roofing I suspect it isn't going to be necessary. Even the white version of Certainteed Flintastic has an aged SRI of 24. (https://www.certainteed.com/commercial-roofing/products/flintlastic-sa-cap/ ) If your Flintastic isn't the white version it's SRI is even lower, for even higher heat gain.
So, what is the 3-year SRI of the TPO product you're considering?
Look it up here:
https://coolroofs.org/directory
Playing around with the CRRC product parameter search tools, Viking Products Group's off-white Sealtite White TPO clocks in with a 3 year SRI of 85, which would be taking up less than 1/3 of the heat of even the white Flintastic. Carlisle's bright white Spectro-Weld TPO White has a 3 year SRI of 93. There are several other light colored TPO roofing in those ranges.
Anything >SRI 50 will make a very measurable difference on this problem- twice the reflectivity = half the gain, 3x the reflectivy= 1/3 the gain. It doesn't necessarily have to be bright white TPO to fix this problem.
Dana,
Thank you for the eductiona and webiste - a great resource. I talked to the company and the product is Weather Bond (Carlisle residential version) Sure-weld TPO White, SRI = 99/85.
With a 3 year SRI of 85 on that product you should see a dramatic difference in heat gain and peak room temperatures.
Rejecting the heat at the roofing layer is really the "right" way to do it- it's more effective than interior-side radiant barrier.
Getting to the current IRC 2018 code-min on a U-factor basis for the open beam section of roof would require adding 4- 4.5" of foil faced polyiso above the roof deck. Is that a possibility? Alternatively, 4" of HFO blown closed cell spray foam or 7" of half pound open cell foam on the underside of the roof deck, covered with ceiling gypsum (required for fire protection, not merely aesthetics) would also get you there. Even at <1/3 the solar gain the ceiling temps & room temps are still going to peak fairly high in that part of the house if left uninsulated. It's not too pretty to my mind's eye, but the silvery aluminum radiant barrier paints on the ceiling would still make a difference if there is no insulation above or below the roof deck.
Thank you Dana,
We have explored insulating the underside of the open beam ceiling (between beams), but are concerned about the look and also ensuring we are not trapping moisture (raised by a engineer at a Polyiso company).
The bid I am waiting on is for 3.5" Polyiso + TPO.
I do not know if foil faced or not -- does that make a big difference?
Does the foil perform a radiant barrier effect, even without an air gap?
I think thicker than 3.5" is a concern here due to wind lift and hurricane considerations.
"Even at <1/3 the solar gain the ceiling temps & room temps are still going to peak fairly high in that part of the house if left uninsulated" -- that is my concern.
Insulating even to R15-R23 on the under side of the roof deck with 3.5" - 5.5" rock wool, held in place with gypsum board or half-inch t & g plank or something semi-permeable/permeable will make a big difference in the heat gain without creating a moisture trap. It's only a potential moisture trap if using foil or plastic faced foam board on the interior side. (Even 5.5" of closed cell polyurethane would not be a true moisture trap, but the drying rates would be glacially slow.)
If 3.5" polyiso above the roof deck works with your wind lift numbers, that too will make a huge difference.
For a building with no space conditioning, the code minimums do not apply, though you might want to provide for the possibility of adding space conditioning later.
3.5" of polyiso will significantly reduce the rate of heat transfer through the roof. The white TPO will reflect the majority of the heat and then insulation will reduce the rate of transfer of what heat does get through. Together, they will make a dramatic improvement on your indoor comfort. Whether the polyiso is worth the investment or not is difficult to answer. With no space conditioning, it won't save you any money, except maybe a little bit on fan energy. The improvement in comfort is hard to put $$ on. About the best I can do is to say that you would notice the difference between the roof with and without polyiso. If you can afford it and want to do the best you can without A/C, the polyiso will help you get there.
>"For a building with no space conditioning, the code minimums do not apply..."
Really?!!
If it's an occupied dwelling codes kinda do apply, so let's see what's in Hawaiian code.
Hawaii's residential building code is based on IRC 2012 (with amendments), but the state has also adopted the IECC 2015 efficiency standards.
https://up.codes/codes/hawaii
https://up.codes/viewer/hawaii/iecc-2015/chapter/RE_4/re-residential-energy-efficiency#R402.1
In zone 1 the IECC 2015 calls out R30 min for ceilings (or U0.035 max for the entire assembly.)
Among the amendments find:
----------------------------------------------
Section R402 Building Thermal Envelope
R402.1 General (Prescriptive)
The building thermal envelope shall meet the requirements of Sections R402.1.1 through R402.1.5.
Exception: Low-energy use buildings. The following low-energy buildings, or portions thereof, separated from the remainder of the building by building thermal envelope assemblies complying with this section shall be exempt from the building thermal envelope provisions of Section R402.
1. Those with a peak design rate of energy usage less than 3.4 Btu/h • ft2 (10.7 W/m2) or 1.0 watt per square foot (10.7 W/m2) of floor area for space conditioning purposes.
2. Unconditioned space that does not contain habitable space.
------------------------------------------------
Interpreting the absence of space conditioning systems means it automatically has "....a peak design rate of energy usage less than..." is not necessarily well founded. Even if that somehow were deemed acceptible, Section R407 is still in play:
https://up.codes/viewer/hawaii/iecc-2015/chapter/RE_4/re-residential-energy-efficiency#new_R407.1
Note that under R407.1 you get "Tropical Home Points" for:
---------------
R-19 Roof Insulation +Cool roof membrane1 or
Radiant Barrier3
----------------
also:
---------------
No Air Conditioning Installed
------------------
The TPO roof selected and lack of air condition meets that standard in footnote 1, so the 3.5" of polyiso option works IF the house meets the definition of "Tropical Home" in Hawaiian code as opposed to "Standard Home", which may require a more careful reading of the code.
Section 1 seems to be where that is defined:
https://up.codes/viewer/hawaii/iecc-2015/chapter/RE_4/re-residential-energy-efficiency#R401.2.1
It appears to mees the definition of Tropical Home, in which case it can make it on an R407 basis even without the TPO + R19, simply with the 2 points for no air conditioning- it already meets that part of the code. There are no points awarded for the TPO on it's own (without the R19), even though it's going to make a difference.
Thank you for the continued thoughts. I am waiting on the quote for 3.5" Polyiso on top of the roof deck, below TPO layer -- I will have to evaluate how painful the extra expense is. Trying to retrofit these old homes seems challenging.
If this route is afforadable, I hope to pursue it and evaluate suitability. Follow on options could include conditioning or interior (under roof deck) insulation.
Andrew
Hello all -- I wanted to give an update.
I consulted two architectural firms in Hawaii and both said they do NOT spec TPO ever. They both like PVC, but this is hard to find for residential roofs. They both state HIGH occurence of failure in the Hawaii high-UV environment. The rep for Weatherbond 115-mil fleeceback TPO says to expect ~ 20 years. At that time, you can coat it with silicone.
However, the architects are stating 8-10 is more reasonable. As it fails, small holes can develop which allow water to penetrate. And, these failure points may be hard to detect?
I am leaning away from this roof design because there is no place for water to go IF it gets in.... and, by the time we find the water intrusion extensive water damage could occur. It seems like a vented roof or exit path for the water needs to be designed in ... and, this would drive cost up considerably 1: to design it (architect fees $8k+) and 2: labor and material. I have seen some neat designs with nail boards and purlins. This looks like great stuff if we had designed it into the origonal design :((((
Andrew,
All roof will fail eventually, and most of them will eventually develop small holes and leaks that are hard to locate and/or detect. And yes, damage can occur if leaks are left uncorrected. But what damage is that, actually?
If you are still planning on installing the foam+TPO on top of the 3-year old mod.bit. roof, then any water leaking through the TPO will stop at the mod. bit. roof. No harm done. The foam will get wet and polyiso does soak up water, but it dries out again. This sort of roof sandwich is done all the time on commercial buildings with good success. If you budget to reroof or recoat after 15 years or so, you should be in good shape. Or, go with PVC for longer lifespan.
You also have the advantage of a 2:12 pitch on the roof. This is a lot of slope for a membrane roof, and water won't pond inside the sandwich - it will slowly migrate to the eaves and leak out there. Water leaking out of the eaves more than a couple of hours after a rain would be your sign that the roof is leaking.
Peter,
I had similar thoughts, but I supposed that water would be in fact held in this sandwich and the mod Bit/glues would eventually fail. In the interim, mold could likely develop.
One discussion with an architect discussed making a path for the water — I think up to maybe even cutting the mod Bitumen and plywood to make a channel for water to escape under the eves.
I like your thoughts on water leaking out as sign of trouble to prompt investigation, but wonder if this type membrane would actually leak.
I like the 2:12 portion, but I have a very low slope back section — used to be a covered lanai/patio — now enclosed. This is an area where we had extensive water damage already because water sat under a white acrylic coating...acted like a blanket to hold the water. It did not drip out at the eve to indicate a problem.
That is the source of concern.
Thank you,
Andrew
Peter, all,
As I think more about this — It sounds like Bitumen (or other vapor/water barrier) + Foam (Polyiso or other) + TPO is common on FLAT roofs (usually commercial).
How do any of these systems plan for the “eventual failure” and subsequent water intrusion?
CUTTING THE BITUMEN/PLYWOOD (underlaying the Polyiso) to form a slot parallel to the eve:
- Has anybody heard of this technique to allow any water that enters a way to get out?
Thank you,
Andrew