A roofing professional shares his views about various roofing applications
by Dick Baxter
Editor's note: Following are the author's opinions about various roofing industry issues. Views expressed are not necessarily those of NRCA.
The roofing industry continues its struggle to refine systems and find appropriate (or inappropriate) uses for various products. There have been improvements in some sectors and challenges in others. It seems the challenges outnumber the improvements, but perhaps planned obsolescence and challenges aren't so badat least for those of us in the reroofing/retrofitting business. There usually is a silver lining somewhere.
In the improvement category, G-P Gypsum Corp., Atlanta, has done some work on its Dens Deck product line to make the boards more "user friendly" in some applications. Dens Deck, as we have known it, remains available with all its strengths and limitations. It may be important to know that G-P has eliminated the silicone in the product originally used to provide moisture resistance to the boards in favor of another proprietary additive.
The discontinuation of using silicone and the term "siliconized" will make the product more acceptable to some end users who fear silicone contamination in their facilities (though the chance of silicone being released from gypsum to contaminate anything is about as remote as asbestos fibers leaping from asphalt plastic roof cement to become a "hazard").
Dens Deck Prime from G-P is a significant improvement over the original product for torch-applied and adhesive applications directly to Dens Deck surfaces. Prime comes with a plastic-like coating on one surface, which eliminates surface dust and release of glass fibers common to original Dens Deck. Asphalt-based cold adhesives can be squeegeed over the Prime surface without "rolling up" in front of the squeegee, eliminating the need for priming the Dens Deck surface. Solvent- and water-based adhesives used in adhered sheet membrane applications readily bond to the Prime surface, which creates a more positive shear-plane at the substrate-membrane interface.
The shear-resistance properties of Dens Deck are good as long as the product remains dry. With the primed surface positively bonded to the gypsum and positive bond of the adhesive to primed surface assured, the performance of adhered roof membranes installed directly over Dens Deck Prime should be markedly improved from the perspective of wind-uplift resistance.
The potential calcination problem common to all gypsum products has not been eliminated with the Prime product. Some consideration should be given to potential substrate degradation problems that may be attributable to roof surface temperatures generated by dark or black roof membrane surfaces when a roof membrane is installed directly over Dens Deck boards.
G-P also has retested all available thicknesses of Dens Deck at FM Research for mechanical fastener density and patterns. The new fastening densities are significantly different from those contained in the current FM Global Approval Guide, and significant material and labor savings are available to users who inquire or request updated fastening information from G-P.
For instance, the 5/8- by 48- by 96-inch (16- by 1219- by 2438-mm) Dens Deck product originally was tested using 16 various brands and types of fasteners for an FM Research Wind Uplift Classification 1-90. After subsequent testing, only eight FM Research-approved fasteners per 5/8- by 48- by 96-inch (16- by 1219- by 2438-mm) board are required to provide an FM Research Wind Uplift Classification 1-90. Check the manufacturer's new test results and consult FM Global's Property Loss Prevention Data Sheet 1-29, "Above-deck Roof Components," for proper placement of the required number of fasteners per board.
G-P commissioned some independent testing of its Dens Deck product, which produced some guidelines/caveats for installations using "hot" methods (torches and hot asphalt/coal-tar pitch) directly over Dens Deck. Some caution is in order for contractors because some of the provided guidelines are neither practical nor "doable" in most applications.
When torching polymer-modified bitumen sheets directly to a Dens Deck surface (regular or Prime), do not apply heat directly to the Dens Deck surface and understand that the old "J" pattern for preheating the lap area may cause irreparable damage to a Dens Deck board. Instead, apply torch heat only to a roll's surface. I seriously recommend the Prime version of Dens Deck for such applications because the molten polymer-modified bitumen flows uniformly onto the Prime surface and bonds easily to the Prime board surface.
G-P's recommended mopping temperatures for asphalt to be applied over Dens Deck (regular or Prime) are unrealistic because Type III asphalt at the recommended temperature cannot be mopped nor will it readily flow from a dispenser. It might be possible to apply coal-tar pitch at the recommended temperature, but that probably would be risky (in that the person mopping probably will prefer a higher temperature) and costly because of the amount of extra coal-tar pitch that would be applied. Simply know that generic gypsum is susceptible to degradation from calcination when exposed to temperatures common to torching and/or normal asphalt application temperatures, and compensate for the potential heat exposure.
A logical combination for Dens Deck and a hot-applied built-up roof (BUR) membrane might be the inclusion of a perforated base sheet in the BUR membrane system installed directly over Dens Deck.
Asphalt could be applied to a base sheet's top surface at normal temperatures (375 F to 425 F [191 C to 218 C]), and the area of heat-affected Dens Deck would be minimized. Adhesion would be positive between the roof membrane and substrate, and moisture released from the gypsum would be dissipated into nonbonded areas between the base sheet and substrate surface. The Prime product would be preferred in such an installation. Recognize that gypsum is inherently temperature-sensitive, and protect the board from prolonged exposure to high temperatures during application of the hot-applied roof membrane system.
White mineral-surfaced cap sheets or light-colored surfacing aggregate should be installed on roof membranes directly applied to Dens Deck surfaces to minimize anticipated roof membrane surface temperatures, especially in high sun-load environments.
The purveyors of TPO roof membrane systems continue to struggle through the age-old process of providing a suitable material at a competitive price. The number of entrants into the TPO business and amount of TPO manufacturing capacity in the United States gives new meaning to the word "competitive." Because there basically are no outward product differentiations among available TPO roof membranes, the only variable in the equation is price. Is this not déjà vu? Will one be better than the other? Worse? How long is "long enough" for expected service life?
Whatever the answers are to these questions, those of us who test roofing materials have ascertained that "TPO is not TPO is not TPO"chemical formulations and basic raw materials may be different in the same product during any given time period. And with the experimentation with new fire-resistant additives and other variations in chemical composition, current TPOs have no substantial performance history and are essentially new, untested products.
The major concern is that the potential of TPO as a roof membrane will be permanently damaged in the experimentation process. There still are those who think black TPO membranes appeal to some market segments. Keep in mind that TPO polymers are inherently heat-sensitive, and adjust your thinking accordingly. Black may be a great color, but its heat-absorptive properties are significant and have the potential to adversely affect a TPO membrane's long-term performance. Simply put, all other factors being equal, a black TPO roof membrane inherently does not have the same potential for weathering as a white or light-colored membrane.
Remember, too, that TPO polymers are terribly ultraviolet- (UV-) sensitive, and without appropriate UV stabilizers in the basic TPO compound, the material is not suitable for use on a roof. There are manufacturers who can become more "competitive" by eliminating UV stabilizers from the TPO compound and relying on aggregate ballast to provide UV protection for the TPO membrane. Anyone who has ever set foot on a ballasted roof more than once should be able to figure out the fallacy of this approach. If wind does not displace the ballast, heating, ventilating and air-conditioning repair/maintenance or roof repair crew members won't put ballast back when finished with their work. A TPO roof membrane absent of appropriate UV stabilizers exposed for even a short time period will provide a relatively short service life.
In addition, independent testing indicates that the fire-resistance properties of some TPO roof membranes are not as advertised. Experimentation with different fire-resistant additives by some manufacturers has resulted in some relatively unpredictable fire-resistant properties in some TPO product lines.
NRCA continues to monitor the chemical composition of available TPO roof membranes to determine whether there have been any significant changes in the manufacture of TPO roof membranes. Routine, regular evaluations of available products may provide "baseline" data for follow-up performance evaluations on "in situ" roof membranes selected at random by the Midwest Roofing Contractors Association and NRCA.
Metal roof systems are outwardly "foolproof." It is easy to see that water cannot get through a preformed metal roof panel. It's also easy to visualize that metal roof systems drain readily and their finishes should provide long-term corrosion protection for base steel. After all, metal roof systems have been in place since before the beginning of the past century in some cases. So why is so much forensic attention necessary for new metal roof systems?
Reference to the Architectural Sheet Metal Manual from the Sheet Metal and Air Conditioning Contractors' National Association and comparison of some of the details from current metal roof system suppliers will provide most of the answer to this question. There will be none of the time-proven methods for making metal "work" in the current metal roof system suppliers' literature. The old "tinners" figured out how to make metal roof panels perform relatively flawlessly by making provisions for usual anticipated movement of metal, minimum penetration of metal roof panels by fasteners and positive drainage of metal roof surfaces. It really is a relatively simple concept.
But in the "good old days," roofs were small by today's standards; drainage was easier to ensure; and compensation for movement of metal sections was not quite as critical as with some of the massive metal roof systems installed today. Some suppliers of metal roof systems have given a great deal of thought as to how to make their systems functional and relatively foolproof; others have a long way to go.
The most significant problem with the new breed of metal roof systems is that they are designed to be installed by "erectors," not "tinners" or sheet-metal mechanics. Indeed, "erectors" rarely possess even the most rudimentary sheet-metal tools but are rather armed with screw and caulk guns for the typical metal roof system installation. A talented "erector" can assemble a typical, well-conceived preformed metal roof system with caulk/sealant, baffles and screws to provide satisfactory long-term performance. Less than talented "erectors"even with "good" metal roof systems-provide job security for roof consultants. If manufacturers would adopt some basic sheet-metal working techniques that have proved to be effective for many years, some problems with "screwed and glued" metal roof systems could be avoided. But then they would have to find some good field talent, which, as we all know, is hard to do.
Basic roof panel alignment during the installation phase seems to elude most metal roof system "erectors." Fundamental alignment techniques such as the use of gauges, chalk lines and measuring tapes are not particularly well-understood. Misalignment of metal roof system panels does not necessarily affect the water-tight integrity of the panels but makes completion of the roof system at rakes, eaves and ridges much more complex (and less foolproof) than necessary. The usual result is deformed accessory/trim metal caused by the 5-pound (2.25-kg) hammer used to provide a "force-fit" and eventual failure of massive quantities of sealant used to maintain a watertight condition for the duration of a contractor's one-year warranty.
Many "leak" problems associated with metal roof systems are the direct results of minimum to no positive slope. On relatively narrow buildings where roof panels can be installed in one length (no endlaps), minimum slopes are not particularly troublesome except, perhaps, at the ridge cap. On wide buildings necessitating the joining of several roof panels to cover the expanse, the treatment and securement of endlaps becomes critical. If sealants are not 100 percent complete through the entire metal panel juncture, wind-driven water will be forced into laps to cause leaks as the water is forced through the lap and over the end of the overlapped metal roof panel.
Although metal roof system manufacturers/brokers continue to contend that a 1/4-in-12 (1.2-degree) slope is sufficient for their roof systems, the reality is that at that slope, everything in the construction process must be close to perfecta condition not widely experienced in the construction business. Some systems are much better conceived than others; some common sense in comparing different metal systems will go a long way in determining which metal roof systems have the potential to provide long-term satisfactory service. The better ones will cost a little more.
Low slopes also generally result in localized water ponding on metal roof panel surfaces either in areas immediately above critical endlaps or where some minor metal roof panel deformation occurs because of minor misalignment of purlins/structural steel or construction damage.
Designers continue to attempt to use interior gutters at either parapet perimeters or in valley areas, and manufacturers/brokers of metal roof systems continue to promulgate the myth that such gutter systems can be made effective. Perhaps such gutter systems can be most effective in Phoenix; Tucson, Ariz.; or Los Angeles (except when heavy rains occur about two times per year), but they are invariably and inevitably troublesome in areas that have frequent, heavy wind-driven rains. The guideline here is that drainage from a metal roof system never should be impeded by vertical barriers. The cause of leakage invariably will be "splash-back" into the roof-wall or roof-gutter junctures or separation of gutter joints in interior gutters caused by typical metal movement.
Compensation for metal movement in currently designed metal roof systems typically is accomplished by using "floating" clips or fastening metal roof panels to "Z" purlins that are intended to "roll" with panels' thermal movement. When metal roof panel's longitudinal dimensions are (generally) less than 40 feet (12 m), most manufacturers recommend "fixed" clips for nonpenetrating metal roof panel installation.
Unfortunately, misunderstanding (or lack of understanding) by the "erector" about how the system was designed to function results in "spot" through-fastening of the metal roof panels to "hold them in place" until they can be positively secured with clips and seaming. Unless these "temporary" fasteners are removed and the panels repaired at the point of penetration, the concept of free panel movement is defeated in the construction process. In some instances, metal roof panels are secured on longitudinal laps with "floating" clips only to be through-fastened to rigid structural members at both ends, again defeating the intended compensation for metal movement.
"Seaming" of metal roof panels most commonly is completed using a mechanical roller that rolls and crimps the vertical, longitudinal seams joining metal roof panels together. If a mechanical seamer is not properly maintained, adjusted and/or serviced, the result is poorly formed/crimped seams that, instead of being waterproof, become funnels for water entry at the marginally crimped seam. Rollers on mechanical seamers should be lubricated frequently to minimize wear on rollers and bearings. Double-rolled seams put a tremendous strain on mechanical seam-forming equipment. For large jobs, it would not be uncommon to exchange seaming equipment several times. Some metal roof system manufacturers have regular maintenance regimens for seaming equipment where equipment is exchanged following each application or when it becomes ineffective during a large project.
When problems occur because of some of these construction errors, the most common "fix" is more sealant and/or screws, neither of which address the root of the problem and generally provide only temporary relief from leakage.
Pre-engineered metal buildings offer some initial construction cost advantages, and, in most cases, provide satisfactory service for their intended uses. Metal roof systems should not be installed in areas where relatively high interior humidity can be expected because there is no way to construct an effective vapor retarder in a metal roof system. The vinyl-covered insulation most commonly used in metal roof systems cannot be effectively sealed at laps, allowing moisture vapor to pass through the insulation to condense on the bottom side of metal roof panels. The eventual result is corrosion at the bottom of roof panels and/or securement devices used to hold the metal roof panels in place and seeping condensate at low edges of roof panels.
Time-proven methods of securing metal roof panels at valleys, hips and ridges often are ignored in construction details for new metal roof systems. Gutter accessories are designed more for ease of installation than utility and long-term performance. In most applications, there is no adequate provision for sealing eave ends of metal roof panels. Foam baffles supplied with some metal roof panels must be secured with sealants and/or adhesives to prevent displacement. Foam baffles are not equal to metal wind baffles at hips, ridge caps and valley areas. As with most roof systems, redundancy in waterproofing always should be considered. Critical junctures should be waterproofed using a continuous waterproofing membrane before installation of metal caps/closures. These precautions are virtually never included in manufacturers' details but help to ensure the satisfactory, long-term performance of metal roof systems.
Remember, metal roof systems are watershed roof systems and cannot be made to be waterproof for any significant length of time. Watershed roof systems depend on positive slopes for drainage; securement that allows for thermal expansion and contraction of metal roof system components while providing a watertight assembly; and minimizing penetrations (fastener and equipment) through metal panel roof systems. All sealants should be concealed in joints and laps to provide long-term watertightness; superficial applications of sealants typically are ineffective.
There obviously are more issues to be mentioned, but this should be enough to stimulate thought about general roof system performance. Understanding available options in roofing and waterproofing is an important step toward selection and installation of successful roof systems. Unfortunately, manufacturers' recommendations may not provide the "best" options.
Dick Baxter is president of CRS Inc., Monroe, N.C.