per·form’ance n. 1 the act of performing 2 functional effectiveness 3 deed or feat 4 a)a formal exhibition or presentation, as a playb)one’s part in this
-Webster’s New World Dictionary Second Edition
The roofing industry has attempted to define roofing performance since at least 1974, when Robert G. Mathey and William C. Cullen collaborated to create the National Bureau of Standards (NBS) Building Science Series (BSS) No. 55 Preliminary Performance Criteria for Bituminous Membrane Roofing.
The objective is straightforward: If you can define the needs of a specific building under design, then, in theory, the roof system could be any system that meets these requirements. Today, that could be multi-layer hot built-up roofing, single-layer single-ply systems, structural or architectural metal, sprayed-in-place polyurethane foam, an all-liquid-applied system, polymer-modified bitumen, or something that hasn’t even been invented yet. Attachment could be fully adhered, partially fixed, loose-laid and ballasted, or even a tension fabric system.
Once the list has been culled by eliminating the systems that don’t meet the stated criteria, the specifier could move on to establish aesthetic, functional, and cost considerations.
It hasn’t worked out that way. In BSS No. 55 back in 1974, hot multiple-ply built-up roofing was the dominant low-slope roof system. NBS proposed 20 criteria for these systems, which it hoped would establish a level playing field. These included:
- Tensile strength.
- Thermal expansion.
- Flexural strength.
- Tensile fatigue strength.
- Flexural fatigue strength.
- Abrasion resistance.
- Tear resistance.
- Shear strength.
- Impact resistance.
- Notch tensile strength.
- Moisture effects on strength.
- Moisture expansion.
- Weather resistance.
- Wind-uplift resistance.
- Fire resistance.
- Fungus-attack resistance.
- Ply adhesion.
Of these 20 items, tensile strength at low temperatures (-30 degrees F./-34 degrees C.) was the most frequently cited parameter. For multiple-ply built-up roofing, this was a minimum of 200 pounds of force per inch of width. In theory, any bituminous system, regardless of the number of plies and type of bitumen used, would perform adequately. Of course, this meant the roof was fully adhered because that was all that was done in those days (on nailable decks, the base sheet might first be nailed, but all ply felts were bonded to one another). Today, for polymer-modified bitumen, “strain-energy” and “fatigue resistance” are more meaningful, but are, unfortunately, not universally accepted.
Consider the systems available now. There are elastic systems that can stretch 400 percent at -40 degrees F., whereas the BUR cannot stretch even 2 percent. There are metal systems that cope with thermal change not by stretching at all, but by using floating clips and flexible transitions that permit contraction and expansion to occur. There is spray-in-place polyurethane foam (SPF) that adheres tenaciously, but with a high thermal coefficient of expansion. SPF works well when using an appropriate elastomeric coating, but fails miserably when coated with inelastic asphalt-aluminum roof coatings. Modified bitumens with a variety of polymer modifiers and reinforcements seem to work fine even though they have half the plies of the conventional built-up roof.
In short, we are still using prescriptive specifications that seem to work only if we stick within a generic type of roofing. (As Cullen and Mathey put it, “… roofs have been traditionally described by prescriptive types of specifications for the individual materials [that] comprise the roofing membrane.”) This approach really means that the specified material is tested to make sure that it’s “normal” rather than ensuring it will work under a defined set of conditions.
Who would have predicted that loose-laid ballasted roofs could work? Or that mechanically attached systems that flutter in the slightest breeze could endure?
This is not to say that performance specifications cannot be developed. They can be, and have been. For example, with the leadership of Underwriters Laboratories and FM Global, there are meaningful fire and wind tests. These are systems tests – everything from the deck up is tested in a manner reflecting actual field installation.
There are also hail criteria, which are obviously important in regions subject to hail. We now have criteria for solar reflectivity, such as ENERGY STAR®, where peak air-conditioning is a design criterion. (We need new tests and criteria for roofs that are substantially vegetated or crowded with photovoltaic panels.)
While there continue to be changes in materials and systems, they are, for the most part, evolutionary, not revolutionary.
Rubber roofing is a good example: It was apparent that natural rubber didn’t have the durability for long-term roofing use, but synthetics did. Roofs appeared as early as the 1960s based upon butyl, Hypalon®, chlorobutyl, polyisobutylene, and neoprene polymers. Unreinforced liquid neoprene Hypalon systems were promoted when high flexibility was needed. In the case of the design of the dramatically different Dulles Intl. Airport, sheet neoprene rubber was applied, followed by a weather-resisting liquid Hypalon coating.
The development of ethylene propylene diene monomer (EPDM) rubber revealed that this material was especially weather resistant. It’s produced as a vulcanized sheet material, which means that it was necessary to develop new ways of holding the sheets down and sealing laps in the field. New flashing materials and techniques had to be introduced. None of these needs could be carried over from the performance criteria established for bituminous roofs.
Laps of early EPDM systems were sealed using solvent-based adhesives. Sheets could be adhered to the substrate with cold adhesives as well. It became apparent that the adhesives were the weak spot and needed improvement. This was ultimately accomplished by using cleaning agents (to remove talc release agents), primers, and butyl-based tapes. The liquid-applied seam adhesives are more vulnerable to workmanship errors, such as pressing too hard with the application brush. Pressure-sensitive tapes are better, but it took the development of ASTM D6383 Standard Practice for Time-to-Failure (Creep-Rupture) of Adhesive Joints Fabricated from EPDM Roof Membrane Materialand exhaustive tests by the National Institute of Standards and Technology (NIST) (formerly NBS) to prove that the tapes solved this critical junction.
Flashings for elastomeric systems also evolved from uncured neoprene with poor heat and UV resistance to sheet EPDM. Term-bars replaced the base- and counter-flashings used in bituminous systems. Penetrations evolved to flexible pre-molded boots and pourable sealers from the troublesome pitch-pans of bituminous systems. None of these steps was predictable using the prescriptive specifications of bituminous roofs. Instead, the laboratory was real-world application and exposure.
Even today, EPDM roofing is not really defined by performance requirements. There is ASTM D4637, which is the current specification for EPDM sheet. It defines three types of sheets (non-reinforced, internal scrim reinforcement, and fabric backed), but does not suggest which is best for a specific project. It prescribes a minimum thickness of 40 mils (1.02 mm), but the industry currently offers 45, 60, 75, and 90 mil (0.090-inch) materials, again without recommending which is best for what conditions. There are requirements for tensile strength, which are apropos to rubber, but certainly not for steel roofing, urethane foam, or even PVC roof systems. Heat aging is conducted at 240 degrees F. for 670 hours for black rubber, but only 166 hours for non-black sheets. (What’s implied is that white isn’t as durable as black – not which is suitable for desired roof life.)
Since there is no single set of criteria that will universally define what will work as a roofing (or waterproofing) system, in the next several columns, I will take each generic category, one at a time, and give an update on existing material standards, recommended practices, and test methods.
I will rely heavily on ASTM standards, most of which are found in Volume 04.04 Roofing and Waterproofing (available from West Conshohocken, PA-based ASTM Intl.). Documents from UL, FM Global, SPRI, NRCA, ARMA, ERA, ANSI, SPFA, MBMA, SDI, RCI, PIMA, and other organizations will be included as well.
Single Ply Roofing Industry
Asphalt Roofing Manufacturers Association
National Roofing Contractors Association
Steel Deck Institute
National Roof Deck Contractors Association
EPDM Roofing Association
Cool Roof Rating Council
National Institute of Standards and Technology
Metal Building Manufacturers Association
Council for Performance Criteria for Constructed Roof Systems