I'm renovating my home, and in the course of hunting for a leak in the roof (now found) I removed a strip of drywall that is 8" wide and 16' long, parallel to the 6x12 ridge beam. See image below for layout.

The house is in the San Francisco Bay Area, and has seen sunny and dry conditions for the last 4 months (not foggy here). On many days, around 4pm there is considerable water on the face of the beam, on the side where the drywall was removed.

enter image description here

I'm trying to figure out where the moisture is coming from, to ensure that I do not need to make any repairs (or do any further investigation) before I replace the drywall. Note that there was no evidence of moisture near the beam prior to the drywall being removed, other than that caused by the (now found) leak in one shingle.

I installed a couple of temp/humidity sensors, at the locations shown in red - here is the upper one.

enter image description here

The upper sensor shows that at 4pm the top of the joist bay is 138F and 27% relative humidity. IIRC, each 20F reduction roughly doubles the relative humidity, so were this air to cool to 98F is would hit >100% relative humidity, and therefore condense. The bottom of the beam measured 77F at 4pm, so one would expect such warm high humid air to condense upon contact.

The lower sensor (hung from the bottom of the beam) shows that at 4pm the air near the beam is 82F and 71% relative humidity - see below. Note the relative humidity climbs steeply around 2pm, even as the temperature climbs.

enter image description here

Here is a photo of the beam, lower sensor, and water drops.

enter image description here

Some notes:

  1. AFAIK, there is no barrier between the shingles the 1/2 plywood beneath
  2. There is fiberglass batt insulation between the joists, which end a few inches short of the beam for most bays
  3. Condensation does not appear on the the other side of the beam (where there is no gap in the drywall)
  4. The beam extends to the exterior, with ~18" exterior at each end of the building

Question: where is the moisture coming from? (I'd like to work this out before I replace the drywall, to ensure that I do not need to make any repairs or do any further investigations).

The possibilities that I have considered:

  • A) From the air throughout the building, rising towards the beam. Issue with this: in that case the condensation would form on the other side of the beam too (same temperature), but it does not

  • B) From moisture left in the roof by the heavy rain we had (over winter), that is being expelled as vapor from the joist bay where the drywall is missing. This was my primary theory back in April. Issue with this: we've had 4 months of sunny dry weather, and the moisture content of the 1/2 plywood (where I have access to it) is <10%

  • C) From a leak (e.g. leaking pipe) somewhere in the roof assembly, that is being expelled as vapor from the joist bay where the drywall is missing. Issue with this: there is no evidence of this anywhere else in the roof assembly - e.g. no wet patches in the drywall ceiling. Also, I know of no pipe in the roof assembly.

  • D) Moisture is absorbed from the exterior air (e.g. morning dew) by the asphalt shingles (which look old), then when they get hot it vaporizes and some of it passes downwards through the 1/2 plywood into the joist bays, then through the gap in the drywall and on to the beam. Issue with this: this seems to be the opposite of what the roof's waterproof barrier is supposed to do, and typically warm air rises, rather than sinking.

So I have 4 explanations, none of which quite seems to fit the facts and my understanding. Where is the moisture coming from?

(apologies for the long long question)

  • What you are referring to as "joists" are more properly called rafters. Commented Sep 9, 2023 at 9:34
  • Given the climate in the SF Bay area moisture control rather than energy efficiency is the main concern. Is this house air conditioned? How is it heated? Are state or local building codes mandating a certain level of energy efficency for your renovation? Commented Sep 9, 2023 at 10:06
  • If you have no roof underlay, then the moisture may be water wicking under the shingles.
    – Huesmann
    Commented Sep 9, 2023 at 13:45
  • @Huesmann when you say "moisture may be water wicking under the shingles", what is the source of the moisture that you're referring to? Vapor in the exterior air? Dew on the shingles? Thanks
    – tom
    Commented Sep 9, 2023 at 15:48
  • @JimStewart Good question. The house is currently empty. We recently installed a mini-split system, which is not yet being used, but which I expect will help solve this issue (one unit is at 10', near one end of the beam). The building permit did not require any energy calcs. and thanks for the reminder on "rafters"
    – tom
    Commented Sep 9, 2023 at 15:50

2 Answers 2


Since you have a hole in the ceiling and a large gap in the vapor barrier with the insulation ending there, moisture from the house is being accumulated in the fiberglass (like a sponge) and the damp air from there is condensing on the side of the beam exposed to the moisture-filled fiberglass.

Given the timing, the roof getting hot is evaporating the stored moisture in the afternoon, so it's "exhaling" right onto the cooler beam, on that side.

At night, or anytime the roof surface is cooler than the dewpoint inside, water from the air is condensing in the roof space, and accumulating until the roof heats up and it revaporizes and flows out the hole (thus the humidity goes up "even as the temperature does.")

  • Thank you for this answer. It's certainly a more plausible explanation than any of the other 4 that I had. Seeing as the underside of the plywood does get to ~15F lower than the conditioned space (at the coldest time of day). It hadnt occurred to met that the fiberglass could hold more moisture than an empty cavity, but it has a massively higher surface area, so that seems possible. Do you know whether all insulation acts that way?
    – tom
    Commented Sep 9, 2023 at 15:57
  • Most insulation does absorb moisture. Closed cell foam does not, or not much. Rockwool is supposed to be hydrophobic, but in this sort of situation I could still see it getting wet, unless it had somewhere to drain to... I've seen open-ended fiberglass rip itself loose from collecting so much water the face wouldn't hold it...
    – Ecnerwal
    Commented Sep 9, 2023 at 16:11
  • Thanks. I have re-installed the drywall, but not yet mudded it. I will leave it 1 week, with a temperature/humidity sensor/logger above the drywall, so that I can monitor the moisture level up there before I commit by mudding it. Any thoughts on whether I should use a vapor barrier primer on the drywall, or just regular PVA?
    – tom
    Commented Sep 10, 2023 at 2:00
  • You should tape the joints (with e.g. masking tape temporarily at minimum, if you "don't want to commit" yet) for air leakage (via smaller holes) reasons. Ideally a ridge vent installed at the peak would allow better drying to the exterior, but if the framing does not look like it was rotting with the situation you had before the leak led you to open it up, it should be similar to that once you close it up again, once it does dry out. On the vapor barrier paint - I guess technically it would mean the area of drywall with foil above would have 2 VBs, which is generally considered bad.
    – Ecnerwal
    Commented Sep 10, 2023 at 2:18
  • Joe Lstiburek talks a lot about getting moisture out at the ridge so as @Ecnerwal says you could put in a ridge vent. For this to work you need vents around the eaves for air to come in, flow through the rafter bays to the ridge vent and out. Do you have any eave vents in the soffits? Or are the rafter cavities closed to mass flow of air? Moreover, Lstiburek says the eave vents need to have a greater capacity for flow into the cavity than the ridge vent has outflow otherwise there would be a pulling of air from inside the house. You want all the air out the ridge vent coming from the eaves. Commented Sep 10, 2023 at 19:04

Building science moisture guru Joseph Lstiburek claims water molecules inside a house move upward because the density of air varies inversely with water vapor content. Moist air is less dense than dryer air according to Joe Lstiburek and so bouyancy transports water molecules to the highest point in a building. He maintains that there must be provision for drying at the high point or accumulation will lead to rot of the roof structure.

If I understand his opinions correctly, an ordinary shingled roof has enough leakage through the overlapping courses of underlayment to allow drying, and even an impervious stick on underlayment is OK if there is an attic to allow drying from the inside. But you do not have an attic!

Lstiburek is evidentally of the opinion that the new "conditioned attics" must not be sealed, but must have some recirculation with the conditioned living space to prevent accumulation of moisture. And if I understand him correctly these will very likely require active dehumidification in the attic. See video of Matt Risinger with Lstiburek. https://youtu.be/zhkGcklWB_Q?feature=shared

You do not have an attic but simply insulated spaces between rafters. Perhaps you can find in Lstiburek's books or other writings or his videos how to deal with this in your climate. He seems to think that a rising and falling moisture content is OK for traditional sawn lumber as long as there is some path to allow drying. But if I had the roof structure you do, I would be wondering if I could pump dehumidified air into the space around the ridge beam to extract moisture.

But the moisture laden air would have to be exhausted. Where? And where would the air going into the dehumidifier come from? Inside or outside? The only thing I can think of is a closed loop for air passing through a dehumidifier, e.g., dehumidified air injected into each rafter bay at the eave and moisture laden air pulled out at the ridge beam going into the dehumidifier.

I would not replace this long strip of drywall you have removed. I would put up metal grills so that there could be air from the living space circulating around this ridge beam. This would allow you to inspect the ridge beam and see how dry it stays. You might consider putting short grills in the interior ceiling at the low end of each rafter cavity. Air could circulate. Of course this might introduce contaminants into the indoor air.


The idea of putting grills in the ceiling at the eaves is probably not necessary and just too radical at least right now. But I think putting a line of metal grills in place of the strip of drywall would be a possible way to allow drying and allow inspection. If the rafters are 2 ft on centers, I would think you could get 2 ft long by 8 in wide white metal grills of the type used for a/c returns.

  • Thanks Jim. Did you see the other answer from @ecnerwal? It roughly recommends the opposite - that re-installing the drywall will stop the moisture migrating into the rafter-bays (at other times of day), which would make sense, as this problem didnt seem to exists before i removed the drywall. I'd be interested to know what you think
    – tom
    Commented Sep 9, 2023 at 15:53
  • @Ecnerwal is a real expert with long and deep experience. I am trying to answer by applying what I understand are the ideas of Joe Lstiburek. It stands to reason that warm moist air from inside would condense on a cold beam and cold decking. It is my understanding that water vapor will penetrate drywall unless it is painted with a vapor impermeable primer or has polyethylene sheeting behind it. I was tryig to figure out how to dry wetted structure. Ecnerwal was trying to keep the water away from the structure in the first place. Commented Sep 9, 2023 at 19:13
  • This kind of roof must be common in the SF Bay area, ask around how people insulate without rotting the roof structure. Exactly where in the Bay Area is this? I have spent some time in North Oakland and in Berkeley and I would like an idea of the setting. Commented Sep 9, 2023 at 19:27
  • Thanks Jim. Agreed that is the difference in approaches in the two answers. The house is in Burlingame, on the flat (~100ft above the bay)
    – tom
    Commented Sep 10, 2023 at 1:54
  • I expect you will replace the drywall for a standard ceiling, and in that case I think you could insert one or several inspection ports to allow convenient checking of the moisture levels on and around the beam. These could be designed to be no more permeable to moisture than the ceiling drywall. Right now you presumably do not know what kind of primer and paint has been used on the ceiling. Specifically, you don't know if mositure (vapor) blocking primer has been used. Is blocking primer generally specified (and used) on drywall ceilings attached to the undersides of the roof trusses? Commented Sep 10, 2023 at 16:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.