The influence of building design on wood decay

IRG/WP 00-10339

W W Wilcox

The cheapest and most effective way to prevent wood decay is to KEEP THE WOOD DRY! You and I know that, but we often forget and attempt a preservative solution to a problem better solved with good design and construction. It's also the answer to a frequently missed exam question in my Architecture class. Just as caulking is a poor substitute for proper design, so preservative treatment should be the last resort in an exposure problem, because, what if the treatment fails? Are we in a position to guarantee that every stick we treat will give "life-of-the structure" performance? "All wood rots in time," and "I won't use treated wood because the incisions are ugly," are just two statements I often hear from designers which indicate the uphill battle we have to get designers and builders to use wood properly. Substituting treatment for better design and then providing ineffective treatment have created a lack of trust of treated products among designers and lack of understanding of wood has prevented the creation of good design fixes. "Why should I pay more for treated wood, it doesn't last any longer than the untreated" a Hawaiian contractor told me; he bought thick slabs of Douglas-fir, treated green with CCA, cut them to length in the field, and placed them directly on the ground as a stairway. He was right; Formosan termites gobbled them up in a matter of months. Wrong material, wrong treatment, wrong application! Other case studies will be presented which illustrate situations where use of treated wood was perhaps not the best solution to exposure problems in structures. Prosecuted for murder for allowing the deck on your rental apartment building to decay?--you will see that one, too. (The following text is a transcription of a lecture presented from slides) I am honored to have been asked to address this meeting of the IRG. I was particularly gratified when asked to present a talk on the general subject covered by my title, which I interpreted as dealing with decay situations where pressure treatment may not be the solution of first choice. As most of you know, preservative treatment is not my field; my field is the diagnosis and evaluation of early stages of decay and the improvement of the performance of wood in structures by proper use and design. I get involved, after the fact, in trying to determine why buildings, or parts of them, have failed. But I am not a designer. I am not in a position to show you drawings which tell you how to do it right, although I'm afraid that is what the organizers really wanted from me. Also, no one asks me out to structures that are performing well; I never get to see the good ones! So, I will start by trying to discuss general principles of design which may control performance of wood in structures, but will move very quickly into performance issues which I know more about. The goal of exterior skin design for wooden buildings can be summarized by adapting a borrowed phrase with the anagram of KIDS--keep it dry, stupid! Dry wood can never decay! With the possible exception of particulate wood products, most building materials intended for exterior exposure perform well enough and are assembled with enough redundancy that how they are put together usually is the most important criterion in their performance. The controlling factors usually are minute details. An example is this roof-wall intersection, where the vertical membrane ran underneath the roof membrane, allowing rain water to gather underneath the roof covering. This, probably, was a result of a simple error in scheduling the building trades--the workers installing the wall membranes came on site before the roofers. All exterior design should be focused on shingling of intersecting materials so that water is constantly being shed outward, away from the structure. In the U.S. the major uses of treated products are in utility poles and railroad ties (sleepers to the Commonwealth). These products are purchased by sophisticated buyers who know what they want, what the treating industry can produce (and even sit with them on the committees which establish the treatment standards), and they get what they want--a long-lived product. In my view, the performance problems are primarily in the consumer area where the typical user knows very little about what they want and how to use it, and can't tell the difference between a good product and a bad one. In this area quality control and certification become important, and I'll talk more about that area toward the end. I do, however, have two examples from the consumer/structural arena where pressure-treated wood is being asked to perform unusual duties and is expected to perform well. It's hard to think of a structure being asked to perform under conditions of greater decay hazard than a houseboat. Here the architect used pressure-treated redwood siding to try to weather the extreme conditions of the Sausalito, California waterfront, and then liked the patina of the Chemonite treatment so much that he had the large structural and decorative members painted to match. This is a quite attractive effect which is, to the best of my knowledge, performing well. A "work in progress" is the restoration of a historic structure, the San Francisco Conservatory of Flowers. This structure is scheduled to be demolished, piece by piece, with all of the thousands of pieces of glass retained and the thousands of pieces of old-growth redwood holding them up replaced with exact duplicates made of young-growth redwood through-treated with Chemonite. I have every confidence that this solution, the only reasonable solution we could come up with for a historic structure, will perform superbly, while maintaining the historic fabric and composition of this unique structure. Now on to the area where I have more experience! I don't know about other regions, but wooden deck failures are becoming increasingly important in the Western U.S. Over the last several decades, our drive for quality outdoor living has converted decks from small, temporary attachments to our homes to large, important portions of the structure, expected to have the same performance life as the rest of the structure. Decks are, by definition, outside the protective building envelope and serve in much more decay-conducive environments than most of the rest of the structure, with the possible exception of the roof which is specifically designed to shed the elements. Decks are not. With decks going higher into the air, their failure increasingly involves human injury. The most absurd of such an instance of which I am aware was a deck in San Francisco on a Victorian structure which was being used as rental apartments. This truly was a tragic incident because failure of the deck killed a woman, but it was made more tragic by an overzealous, headline-seeking, District Attorney who decided to prosecute the owner of this building for murder! Now that would certainly do something for quality control in the treating industry if a product failure could lead to criminal prosecution for murder. But this was not treated wood. Should it have been? This is a three-story, Victorian building in a major, high-class, residential area of San Francisco. With this sort of view, you can see why this is a prized place to live, and why decks are popular. This building used to have a third-floor rear deck, attached to the structure at the U-shaped marking beneath the sliding glass door. From talking to the renter of that apartment I learned that he had invited a wedding party to his apartment for the reception. It was a nice day so most everyone was out on the deck. At one point he gathered everyone together for a group picture. Everyone lined up against the outboard handrail, with a spectacular view in the background, for the group picture. The apartment renter, who was near the sliding glass door with the camera, reported hearing a loud snap and, feeling the back of the deck rising, jumped through the open sliding glass door into the apartment. He looked back in horror to see the deck tip forward, hit the second floor deck and bounce, causing it to rotate 90 degrees and flip upside-down on the concrete covered back yard. A number of the guests received serious injuries when they hit the concrete, but one unfortunate woman, most tragically the new bride, ended up underneath the deck and was killed instantly. The headlines the next day touted the negligence of the building owner for allowing his deck to rot, thereby causing this accident. I was called to the site, a seriously unnerving experience, by the attorneys representing the bride's estate in civil court to document the role of decay in causing the failure. But I could not, making me persona non grata because I disagreed with the conventional wisdom carrying forward the case and was never asked to do anything more on it. Let's look at how I came to my unpopular conclusion that decay was not the cause of this accident. This was a 40 or 45-year-old, untreated Douglas-fir deck, not at all an uncommon situation in the San Francisco Bay Area. Certainly, some decay was present, as expected, and there was evidence that decayed members had been replaced from time to time in the life of the structure. There was a good deal of decay in the outboard ends of the stringers, to which the handrail had been attached. If failure of the handrail had been the cause of the injuries, I would have agreed that decay was a culprit. But it was not--the handrail broke on the way down. At the back edge of the deck, where it attached to the wall, the stringers appeared to butt directly to V-rustic siding while sitting on a 2" ledger, with blocking toe-nailed into the stringers and face-nailed into the siding. Portions of the blocking had been replaced with plywood, apparently as a repair for decayed wood. Also visible in this area are relatively new, as yet unpainted, standard joist hangers. These would have been face-nailed into the blocking and the stringers. Now lets focus on the SECOND floor deck, of similar construction. The stringers rest on a ledger, too, but here the joist hangers have so-called "hurricane tabs" which firmly attach the rear of the deck to the ledger. Looking again at the ledger for the third floor deck, we see that it, too, used to have hurricane tabs, but they were removed sometime since the building was given its current coat of paint. Also notice that the failed deck no longer has these tie-downs. This configuration made the third floor deck very vulnerable to pulling away from the wall, which the second floor deck could not do, unless the ledger itself pulled off the building. The third floor deck was supported by a full-width beam placed a considerable distance back from the handrail edge. Again, looking beneath the SECOND floor deck, we see that it, too, had that construction, but that someone had moved the beam outward, toward the handrail, since the last painting, making it more stable to excessive weight applied at the outboard edge. My reconstruction of the failure was that the heavy load of the wedding party having their picture taken against the handrail caused the handrail edge to bend downward, applying an upward force to the back end of the deck, which no longer was tied to the ledger, causing it to pull the face nails and lift off the ledger. Now unstable and rotating about the single support beam, the deck tilted forward, the rear hitting the second floor deck causing it to bounce and flip over. It probably hit at an angle, causing the 90 degree rotation, because the guests were collected at one corner of the handrail for a better background view. The cause of this tragedy was not the fact that 40-year-old untreated decks have decay in them, it was the result of someone, during a repair, substituting standard joist hangers for the original ones with tie-downs at the deck's connection to the building, and the failure to move the support beam further toward the outboard edge of the deck, as had been done for the second floor deck. I don't like personal injury cases to begin with, and I can tell you I hope I never have another failure case where the pieces are adorned with flowers. But the point I want to make in sharing this with you is that the solution to this problem is NOT the use of pressure-treated wood. Parenthetically, it also is not murder! Another deck problem I have encountered involves a design feature called "mill construction." This involves Douglas-fir 2x6 or 2x8s on edge, nail-laminated together. The architects tell me that this system was developed on the then-timber-rich West Coast of the U.S. during WWII, when the Federal Government had sequestered all construction steel. This approach allowed the construction of long-span industrial floors of wood, without the need for steel to stiffen them. There is no reason to use mill construction today; in fact, because it is so wasteful of wood, there is good reason NOT to use it. It is totally inappropriate in residential construction, where such massive strength is unnecessary, and, because of all the capillary spaces between laminates, should never be placed in exterior exposure. In this case, a deck was created by simply making the laminates in this area about six feet longer. Since there was no roof overhang, the only protection for this assembly was a coating of "miracle goop." These all fail eventually leading to a severe decay problem in this mill construction deck. Because of the capillary spaces, the decay progressed back into the building and down into the glulam beam holding up the downhill side of the building. Again, the solution to this problem is NOT pressure-treated wood. This, simply, is a stupid design which should never be used in residential construction, especially in exterior exposure. Now, here's a deck that's a challenge! We've got a lot of steep, very valuable, land overlooking the Pacific Ocean in California, which tempts people to do things like this. Does anyone want to take on responsibility for the treatment of the wood holding up that deck, especially in light of the criminal prosecution in San Francisco? Which brings up the issue of quality control for treated wood entering the consumer market. These data are old, now, but they were produced from a sampling of retail-end lumber which contained a brand by a third-party agency guaranteeing that this lumber met their specifications. Unfortunately, commercial politics destroyed even this agency in the U.S., so there is no organized, even partially reliable, third-party inspection quality assurance available to retail consumers in the U.S. We can not expect builders to rely on treated products if we can't even meet our own minimum standards a majority, let alone approaching 100%, of the time. This is a foundation piling, driven into muck soil to support a structure. The non-green patches are where bark was still present during treating and has sloughed off after drying in place, leaving untreated wood. Even where it is green, the penetration was less than 1/8-inch. The tragedy here is that there are actually over 400 such piles, holding up an 80-unit condominium structure, with no way to replace them. The solution to this problem would have been properly pressure-treated pilings, but, without quality control, that didn't happen. Here is a stack of 6x14, Douglas-fir beams, treated with Chemonite. This shows a pretty good representation of the difficulty of treating sawn Douglas-fir. The problem here is not the quality of treatment (that's actually pretty good for sawn Douglas-fir), the problem is that the designer ordered the beams in the wrong length, requiring that they all be cut to length in the field, exposing the untreated wood you expect at the center. Here, treated wood was the recognized solution, but the end-user messed it up. The developer of this project in Kailua, on Oahu, told me at a seminar in the early eighties that there was no point in paying extra for treated wood because it didn't last any longer than untreated wood. Granted, at that time Hawaii was treating green, sawn Douglas-fir with CCA and calling it "treated" under a "Hawaii use only" u32 ?stamp. Still, I jumped at his invitation to see his experience, and this is what I found. The roof of the garages was sodded--a challenge for wood anywhere, but particularly in Formosan termite country. And for the stairways, he was using 3 or 4-inch-thick slabs of sawn Douglas-fir, cut to length in the field, and placed directly on the ground. He was already replacing rotten and termite-eaten treads less than two years into their service life. Maybe you could get away with this with adequately treated wood, but why try? And you certainly won't be successful when the treated wood you're using has penetration like this! Glulams in exterior exposure present a special problem. Most made-up members are too large to treat and, with woods like Douglas-fir where penetration only to the depth of the incisions can be expected, this is clearly not a solution. Treating the laminates also is problematic, not just because of possible effects on gluing, but, in California, the shavings were declared by the State regulators to be a Class I toxic waste. The laminators got rid of their treated wood in a hurry! This project used glulam for nearly everything on the outside of the buildings--through-members protruded beyond the building envelope, blocking and rim-joists were glulam with the sides exposed, they even designed glulam "stubs" to make it look like the beams crossed over the columns. The solution to this problem is NOT pressure treatment. This, again, is simply a matter of stupid design! In fact, because of their innate nature, glulams simply are not members that should be in exterior exposure. Examination of this end, with the various grain orientations of the different laminates, shows that, if this beam were to undergo any significant moisture content cycling, it would develop up-facing checks deep down into the beam, directing water into the center of the member where it becomes trapped and facilitates internal decay. These are solid wood "flying" beams, penetrating the walls on both sides and impossible to flash or caulk successfully because of all of the potential for movement, therefore decaying and leading the decay into each structure. Pressure treatment is not the solution to this problem, this is simply another dumb use of wood. I'm not sure that our fire codes allow this anymore, anyway. I have heard it said by a number of architects that, usually, caulking is simply a dsigner's poor excuse for proper design. Even if it doesn't leak, it creates a lasting, constant maintenance problem for the building maintainer. These examples led to wood decay, but, again, treatment was not the solution. No amount of treatment could overcome the problems caused by this misapplication of grooved plywood siding and scarf-jointed battens. Grooved plywood should never be applied with the grooves horizontal, because it causes up-facing cracks into the center of the panel at the edge of each groove and the groove acts like a gutter directing water to the joint between panels. A properly assembled scarf joint sheds water back to the surface of the two joined pieces, while these reverse scarfs draw water in to the exact location the battens were put on to protect. This is the only case I've ever seen of mushrooms growing directly out of building siding, but treatment would not have helped this either. The siding leaked water into a wall filled with insulation which was, essentially, crushed newspaper, which provided both the moisture and carbon source to support extensive decay. Playground equipment! I've lost track of the number of calls I've gotten from parents asking for a "non-toxic" preservative for the playground equipment their children play on. No matter how you start trying to answer that question, you lose them quickly. This is the equipment my kids played on and I always took the position that I would rather risk their eating a little chemical than running the risk of having one of these huge members decay and come down on their head. Here treatment should be the solution but, because of chemophobic parents, it will only be a workable solution if we can develop deep penetration with that elusive "non-toxic" preservative. Actually, that is what happened with this playground equipment (unfortunately, the "come down on their head" part). The parents club got a number of old 12x12s--former excavation shoring, which had been stored solid-piled on the ground for a number of years. They buried one end in the ground, creating a Stonehenge-like circle of uprights. They fastened eye-bolts to the top of these, now, vertical cantilever members, attached chains which were threaded through old tires, producing a creative, undulating, play surface. Alertness during the construction process would have revealed that these members were badly decayed, as the stack of washers penetrating into the member in an attempt to tighten the nut against a firm surface, showed. Unfortunately, this was another personal injury case. This timber broke at the ground line, hitting a little boy in the head, killing him. This case is just a series of mistakes by well-meaing, but uninformed, people, for which treatment, again, would not have been the solution. But the school's solution--replacing the wooden equipment with steel--also is ill-advised. If my kids were to fall from a piece of playground equipment, I'd much rather they hit their head on wood than a steel 6x6. Wooden roofs should always have a pitch to them, for natural, passive, drainage. All that's needed here is a single pinhole through the roofing material and this roof is a goner. I guess you could get away with this by using adequately treated wood for everything below the water, but why bother? Change the design! Finally, on the West Coast of the U.S., we are building with green lumber which, in many cases, already is infected with decay fungi--the same ones responsible for most above-ground decay. Since most ordinary decay fungi go dormant between wettings, we are building our buildings with the fungi which will decay them already provided. All that's needed is water. The solution to this problem is not treatment, it's kiln-drying. Although I've shown you a number of examples of inadequate wood performance in structures for which I've suggested that preservative treatment would not be the appropriate solution, there are a number of things we should do, I believe, to position preservative treatment in a way to make it available as a solution if other methods prove unsuccessful. Some challenges to preservative research which I recognize would include the following. We need to improve penetration, especially in refractory species like Douglas-fir. Adequate drying prior to treatment, even in large members would help. New approaches, such as vapor metal and cold plasma treatments may hold promise for their ability to increase penetration. We need to find a substitute for visible incisions and, at the same time, achieve deeper incising. I have frequently heard from architects, "I won't use treated wood because those little knife marks are ugly." We need to improve quality control; is the ability to guarantee treatment that meets certain specifications in 100% of pieces treated really not attainable? In our emerging global economy, perhaps quality assurance needs to be taken on by one, or more, international organizations, as in the ISO9002 systems for other commodities. Finally, is it completely impossible to give those parents a "non-toxic" treatment for their playground equipment? The solutions to these issues will require cooperation. Going it alone is an outmoded concept, which, in this era of secrecy and non-cooperation in academia, probably means that the treating industry is going to have to get together and do these things, if they are going to get done. I do, however, have the answer for one of the oldest philosophical questions about being alone. Alan asked me to be a little outrageous and present things that would cause attendees to talk about them in the halls the rest of the week. I think I've done that, Alan. If I went a little overboard, I apologize.


Conference: 00-05-14/19 Kona, Hawaii, USA

Download document (143 kb)
free for the members of IRG.

Order document from secretariat