Q. I read with interest your article “Roof Ventilation Retrofit”, which I found through the JLC [Journal of Light Construction) Newsletter. As I have a current roofing issue, I Googled for your contact information and found your blog. I would like to get your independent advice before I talk to a contractor.
The semi-contemporary house has a traditional attic on the eastern part of the house and cathedral ceilings on the west. The original roof (GAF Timberline, 340 lb shingles) was installed in 1981 and performed without any noticeable problems, except for the gradual wear of the shingle grains. In 2005, I had the 24-year old roof replaced as a precaution. It was a clean re-roofing job. The decking was found to be in good condition, except for some loose nails.
Owens Corning 50-year architectural shingles were selected, with added attic ridge vents. (The contractor also installed ridge vent for the cathedral roof.) Ice and water shield were specified at leading edges (6 ft) and 15 lb felt for underlayment for the remaining roof. Chimney flashings were replaced, but not the wall flashings.
Around 2009, during a heavy rain storm, I noticed some water dripping from the window frame in the master bedroom. When I tried to contact the roofing contractor, I found that his numbers were disconnected and that he had moved out of state. I climbed onto the roof to inspect; did not see any obvious problems with shingles or flashings. However, I suspected the problem to be with the flashing at the headwall. This flashing had not been replaced with the re-roofing; the workers simply pushed the new shingles under the original flashing. I caulked the joint from the roof to the vertical wall and for some time the problem seemed to be fixed.
More recently, there was occasionally some water dripping again at the same location at the MBR window frame. On a warm day, I smelled some musty odor in the loft area. (Location 5 on the cross section sketch) Also, the Sheetrock on the head wall at the joint to the roof felt warmer than the rest. I removed a lamp at this wall and traced the smell coming from inside the wall, which could be mildew. I concluded that the head wall flashing needs to be replaced and the headwall opened to investigate. However, before I do this, I would like to have your expert advice and recommendations.
Enclosed is a cross section sketch for insulated roofs and walls and provision for ventilation. All cathedral ceilings have R-30 Fiberglass bat insulation installed between rafters, with a vent channel against the roof deck. Roofs have 2ft overhang and continuous soffit vents. The attic section has wall vents at both ends. Ridge vents were added in 2009 with the re-roofing.
The MBR roof connects to a head wall, which has 4 awning windows. Therefore, even with venting channels in the cathedral ceiling, possible vent air flow through the head wall to the ridge vent is blocked by the awning window framing, except at the center and the two ends of the wall. Some air may find its way through gaps to the ridge. But, this roof section is essentially unvented, did perform OK with the original roof, but is troublesome since the re-roofing.
I enclose a sketch of the concept for installing a new head wall flashing. As a minimum, I expect to remove the siding and Styrofoam and install new underlayment and seal with tape around the windows. I also intend to remove the sheetrock at the inside and remove the insulation to inspect the cavity and remove any mildew which might be present. I wonder if I should seal the area where the roof rafters join with the head wall framing with closed cell foam?, or even fill the head wall, which consists of a number of small cavities, entirely with foam?
A. Thank you for the two sketches, which makes it easier to visualize the situation.
Looking at the cross section of the wall where the leak occurred at the window head in the master bedroom during a heavy rain storm, it is possible that wind-driven rain penetrated the head flashing of the window, if it was not properly installed. Unfortunately, common practice is to install the head flashing over the housewrap or felt paper, when it should be behind it.
More recent leakage under different conditions adds another element to the puzzle. The musty odor you smell on warm days, combined with a warm drywall, tells me that there has been leakage through the ridge vent, which the roofer installed in 2005. If the ridge vent is not externally-baffled (like Shinglevent II), it may have ingested rain and possibly snow over the past eight years, which would drip down the headwall, wet the insulation, negating its R-factor — thus the warm wall in warm weather. The water would be absorbed by the framing at the bottom of the headwall. The musty smell may be a warning of decaying framing.
Your thought that the leakage may be occurring at the joint of the headwall and the cathedral roof of the master bedroom is unlikely, in my opinion. Any leakage taking place in that area would have to run down the entire length of the rafters without wetting the drywall ceiling — an unlikely scenario since you have told me that the vapor retarder of the fiberglass insulation is Kraft paper.
Kraft paper is a relatively ineffective vapor retarder with many joints, which raises the potential for condensation on the sheathing of the cathedral ceiling running down to the joint with the outside wall into which it drips down until stopped by the window headers.
I suggest that you open the area where you smell the musty odor to determine if there is soaking insulation and structural damage that needs immediate attention. I do not think that you need to replace the flashing at the bottom of the headwall unless further investigation points out that there may be a contributing problem there.
As to the cathedral roof ventilation being interrupted by the four windows in the headwall on its way to the ridge vent, there should not be any air flowing through the fiberglass insulation in a wall, as it diminishes its R-factor. The joint of the rafters and the headwall should be tightly sealed. What should have been done by the roofer is to install Air Vent Inc.’s Flash Filter Vent to allow proper venting of cathedral roof, which is not taking place now.
The top ridge vent is also not allowing proper ventilation of the long main roof because it is short-circuited by the soffit vent of the 2-foot overhang on the other side of the peak. Air, like water, seeks its easiest travel paths. The intake at the overhang soffit vent is exhausted through the ridge vent, affecting negatively any ventilation from the soffit vent at the base of the long and relatively shallow roof pitch.
And if the top ridge vent is not externally-baffled, but internally-baffled or not baffled at all, precipitation intrusion is possible — and likely — especially with such short distance, which may encourage weather intrusion through the soffit vent and add to the water that may leak down the headwall.
An externally-baffled ridge vent deflects the wind over the ridge vent, increasing the draw of intake air from the lower soffit vent following the Bernoulli Principle. Internally-baffled vents do not allow this principle to work, as the exhaust feature of the ridge vent is halted. They may lessen the chance of weather penetration, but even that is not absolute.
You had no problem with the original roof without a ridge vent because you had no ridge vent and no air circulation in the headwall. The installation of a non-externally ridge vent changed the dynamics of the functioning of the envelope by allowing airflow — albeit very slow airflow — through the headwall insulation if the joint of the headwall and roof rafters is not effectively sealed.
I also would not apply Styrofoam or any other hydrocarbon rigid insulation in contact with asphalt-impregnated felt paper; the petroleum distillates in the felt will attack the rigid insulation and damage it. Housewrap would be better. Or use polyiso rigid insulation.
If you do any siding removal and repairs, I would suggest that you introduce a rain screen behind the siding, using Benjamin Obdyke’s Home Slicker, which also comes with integral Typar housewrap. Be sure to follow the directions to make sure that the rain screen is properly screened at the top and bottom in order not to allow insect intrusion.
I hope you find this helpful in attacking the problem at its source and finding no need to undo and repair things that should not be needing repairs until proven otherwise. Take it one at a time.
Please let me know what you find as you undertake the repairs, and if my conclusions are correct. It’s always possible to miss something when not on the spot looking at the problem.