Moisture Damage in a Panelized Wood Roof - Two Examples

"The pioneers are the ones with the arrows in their backs", goes the old expression. The building industry in general heeds this warning. We employ time-tested techniques and standard assemblies to assure our clients that the building they purchase today will last for years to come. But one of our standard assemblies may be susceptable to hidden, long-term deterioration; leading to sudden and unexpected failure. This page covers two projects where a kraft paper layer stapled to the bottom of a panelized wood roof facilitated the collection and retention of water vapor in the between-joist cavities. The accumulation of water and saturation of the wood coupled with the high ambient temperature led to wide-spread and severe corrosion of the prefabricated metal hangers.

I first encountered this problem in August of 1989. At that time I thought the particular project conditions to be so unique I would never see this problem again; that is, until I was asked to look at a similar facility in February, 1995. I am not certain this problem can/does exist in all areas of the country. I leave it to my industry fellows to make their own determinations in the field.

Background - August 1989

I was called to consult on a large furniture warehouse structure across the bay in Hayward, California (about 30 miles from San Francisco) by Mr. John Goveia of Technical Roof Services, Inc. The building owner was puzzled by the seemingly spontaneous failure of a small number of 2 x 4 roof joists. Here is what I found:

Project #1 - Hayward, California
Facility Use Furniture Warehouse (recently vacant).
Deck Construction Cap sheet roof over 1/2" plywood over 2 x 4's to 4 x sawn purlins at eight foot centers.
Insulation None. Kraft paper stapled to bottom of joists.
Wall Type Tilt-Up Concrete. One whole wall line facing the Bay is filled with roll-up doors for truck loading/unloading.
Approximate Age 25 years.
Plan Area 252,000 square feet.
Environment Located approximately one half mile from the San Francisco Bay shoreline.
HVAC None.
Ventilation Six large powered exhaust fans. At the time of my investigation, three fans had been purposely disconnected.


Using a scissor lift, I was able to view the joist hangers up close. The majority appeared to have corroded, some extensively. I pulled some of the kraft paper off of the joists to clear the area for photos and was surprised with the amount of liquid water trapped in the joist space, resting on top of the kraft paper. In one instance I found a quarter of a cup!

Of the fallen joists, the common characteristics were the lack of any nail holes on the narrow face (from the sheathing) and the absence of the top flange of a panelized joist hanger (due to corrosion) on one end. Without the diaphragm nailing, these joists were free to fall to the warehouse floor. Given the general state of deterioration of the metal hangers throughout, it could be reasoned that a significant number of additional joists would have fallen were it not for the edge/field diaphragm nailing.

Mr. Goveia was first to advance the idea that the corrosion could have spread to the diaphragm nailing. We sampled the diaphragm nails and purlin hanger nails at random locations. Each group showed significant corrosion due to moisture. Some of the plywood nails, though they looked good from a surface view, had experienced corrosive necking - a loss in diameter within the region of the plywood layer. Some nails had lost fifty percent of their original diameter. In certain locations, the nail heads on the top flange of the purlin hangers seemed to be slowly melting away.


In an effort to reduce the inside temperature (and to increase ventilation), the employees had left several of the large dock doors open during the daytime to permit the cool bay breeze to blow into this warm building. That cool breeze also carried with it a great deal of moisture. When that moist air entered, it tended to rise...up to the level of the roof, through the gaps between the staples in the kraft paper, into the even warmer joist space...until it encountered the water-tight roof membrane; end of travel. The wood members absorbed moisture to the point of saturation and the interface of saturated wood and metal hanger fostered corrosion.

As day turned to night, any free vapor condensed on the upper surface of the kraft paper (the cold side of that boundary). The moisture in the wood members would attempt to equilibrate through their thickness, leaving a less than saturated exterior surface. With the next roof warming, the trapped water moved into the surface of the wood, recreating the saturated condition and assuring that the corrosion could continue. This process went unnoticed for years until the hangers which supported the occasional joist were so corroded they failed.

The proximity of the Bay, favorable orientation of the dock doors, disconnection of the exhaust fans and the kraft paper all seemed to add up to an atypical condition where the moisture was actually pumped into and collected in the building. Corrective work included adding nails to supplement the existing diaphragm nailing near the center of the building, the addition of a seventy foot wide plywood overlay around the 483 ft. x 522 ft. perimeter and numerous wood blocks and angles required to support the existing framing. The costs for this work were borne by the tenant per the lease agreement.

Background - February 1995

The problem cited by the head of facilities was drops of discolored water from the roof structure staining some of the merchandise. I informed them of my findings at the above project and made arrangements for a site visit. Here is what I found:

Project #2 - Ontario, California
Facility Use Clothing Distribution Warehouse (occupied and running).
Deck Construction Cap sheet roof over 5/8" plywood over 2 x 4's to 3 1/8" glulam purlins at eight foot centers.
Insulation Triple foil layer plus kraft paper stapled to the bottom of the joists.
Wall Type Prefabricated aluminum panels supported by the perimeter braced steel frames. Two opposite wall lines filled with truck doors.
Approximate Age 15 years
Plan Area A partial second floor of 93,600 square feet sits over a basic plan area of 295,000 square feet.
Environment Located approx. half way between Los Angeles and San Bernardino. The Pacific Ocean is more than 35 miles away.
HVAC Numerous large package units on the roof.
Ventilation Smoke exhaust fans - only make-up air from the package units.


With the assistance of a facilities worker and a scissor lift, I was able to view the areas of greatest concern and find additional locations where the same problem was in an earlier stage. The pattern that emerged is best seen in the following photos.  You may click on the thumbnails to access the full size photos.


Photo #1 - Water vapor penetrates the kraft paper boundary and collects on top of the paper and in the adjacent wood fibers. Of note is the fact that the hot vapor tends to collect on the up-slope (high) end of any joist bay leading to pillowing of the kraft paper at those locations. Near the roof ridges, pillowing can be seen at each end of the joists bays.


Photo #2 - The indications of moisture damage can be minimal, such as in this case of a stained purlin hanger. See an example of this type of hanger exposed in Photo #4.


Photo #3 - Over time, the water vapor collects and wets the surface of the wood sufficiently to initiate rusting of the joist hangers. While you will not see this process by looking at the kraft paper, you will, in advanced stages, note the presence of a dark brown or black trail down the face of the purlin below the joist location.


Photo #4 - When you peel back the kraft paper and foil insulation at such locations, you can see both the discoloration of the wood and the extensive corrosion.


Photo #5 - Deterioration of the joist hangers starts where the moist wood makes contact with the metal joist hanger.  A whitish substance collects on the surface of the hanger, indicating the factory finish has been compromised.


Photo #6 - The deterioration continues until the side is fully corroded.  Note the presence of the telltale dark streaks on the face of the purlin.


Photo #7 - Eventually the whole hanger is involved.  Rust particles begin to move with the vapor along the top surface of the kraft paper resulting in lateral staining.


Photo #8 - Were it not for the diaphragm nailing, this joist would be ready to fall.  The hanger is fully corroded.  Note the weathering of the bottom surface of the roof plywood.


The source of the moisture was the real stumbling point. In the hot, dry climate of Ontario, California it seemed improbable that the HVAC make up air carried moisture into this building. There has never been a problem with roof leaks at this facility. The stock is shipped dry in plastic sacks or cartons, so moisture does not enter by this method. The dock doors all have seals so there is no significant transfer of air during the loading and unloading process.

In my opinion, the moisture came from the center's employees. A loose calculation of water gain over the life of the structure follows:

  1. Each person loses (conservatively) approximately one pound of water through sweat or exhalation every work day.
  2. There are more than 150 employees working in this facility daily.
  3. Assume employees are present on 90 percent of the days during the year.
  4. The building has been in use since January 1981.
  5. Total Water = (365 days)(0.90)(1 lb.)(150 persons)(14 years) = 689,850 pounds = 344.9 tons of water.

In this case, the air conditioning system forces the warm, moist air up towards the roof framing and above the level of the return air grills - a distinct boundary between temperatures and humidities is noticeable when working on a scissor lift. The moisture works on the edges of the kraft paper, weakens them, then enters the dry, warm joist cavity. Once inside, the corrosion mechanism can begin.

Of note is where the path of the water vapor was otherwise blocked by a steel and concrete floor structure, the vapor traveled until it found an opening and moved up to the second floor roof area. Increased moisture damage was noted directly above stairs and conveyor openings in the second floor.

A program of cutting back the kraft paper near all metal hangers was completed. Each location was reviewed for degree of corrosion and a judgement made about the need for supplemental supports below each joist. As part of the re-roofing process, areas of severe deterioration were examined for corrosive necking of the diaphragm nails. An overview of the repair follows.

The Repair


Photo #9 -As part of the below-roof work, areas showing significant hanger deterioration were marked by pushing gardener's flags up through the roof sheathing.  If almost every hanger showed signs of rust, the flags were placed only at each end of the roof purlins.


Photo #10 - An area of roof with the roof membrane freshly removed.  The dark color of the plywood comes from the trapped moisture. The clumps you see are what remains of the heads of the nails once they have corroded.  No nails, no structural diaphragm, no lateral resistance.


Photo #11 - Here are some samples taken from the jobsite.  They are 10d (pennyweight) shorts, 2 1/4" long .  Note that some of the nails have lost material from their shanks in the region between the nail head and where the nail was embedded into the roof framing.


Photo #12 - A close-up of some of those nails.  The middle two have experienced "corrosive necking" - a thinning of the shank due to corrosion.


Photo #13 - Moisture vapor geysers.  Created by the original roofing crew when they nailed their base sheet to the plywood.  The end of the nail was exposed to the same corrosive conditions as the joist hangers.  As it corroded, the loss in diameter created a path for moisture.


Photo #14 - Moisture was pushed up into the space between the roofing and the sheathing. Downslope from this location, the roof sheathing was removed and replaced due to dry rot.


Photo #15 - If we cannot depend on the original nails, then new fasteners must take their place.  Through testing, we determined that renailing the diaphragm was not feasible due to potential splitting of the wood.  Each new fastener location was subdrilled to eliminate this possibility.


Photo #16 - #8 x 2 1/4-inch long wood screws were used to replace the damaged nails.  Our selection was based on shank diameter relative to 10d nails - especially at the contact between the plywood and framing member. Screws were installed between each existing fastener.

What does this mean for you?

1.First and foremost, this should alert you to what to look for in your facility: Telltale rust staining and buckling kraft paper. If the problem exists, have your contractor or maintenance persons cut the kraft paper away from the face of the purlins by 2" to 3" to encourage air circulation and drying. In the absence of a sealed, moist environment, the corrosion process will be stopped. Also, visual access to the hangers will permit you to assess the degree of damage and take appropriate steps. You will experience a small drop in "R" value of your roof system, but the structural integrity of your building should be your first concern.

2.The importance of proper ventilation cannot be underestimated. I know that the general trend is to make buildings more energy efficient (i.e. they have no "holes" for the passage of air), but the retention of moist air from kitchens, or bathrooms or process machinery or even people might contribute to a real problem in the long term. The opinions of our mechanical engineering colleagues must be sought to prevent this type of problem in the future.

3.If you are responsible for plant facilities, you might want to look into the type of cushioned dock bumpers which seal the end of your trucks tight against the door. This will maintain the balance of your air systems in your building. Also, the dock doors in unoccupied bays should be kept closed at all times.

Scott McVicker, S.E.
McVicker Associates, Inc.
Structural Engineers
507 Purissima Street
Half Moon Bay, CA. 94019
Phone: (650) 726-9494
FAX: (650) 726-9498

Back One Page]