Following is a more expanded version of this column, under the same title, which was published in the May 2003 issue of Buildings magazine.
In Canada on Jul. 1, 2002, and in the United States on Jan. 1, 2003, polyisocyanurate roof insulation manufacturers adopted a new test method for predicting long-term thermal resistance values, now commonly referred to as LTTR-values or just LTTR. This new test method, CAN/ULC-S770, was developed in Canada and is based on ASTM C 1303-95. It became a national standard in Canada at the close of 2000 and is now included in ASTM C 1289-02, the ASTM standard specification for faced polyiso insulation.
LTTR is defined as the thermal resistance value of a closed-cell foam insulation product measured after storage for five years under prescribed laboratory conditions. According to the text of S770, the five-year aged value is equivalent to the time-weighted average thermal resistance value over 15 years and should be used as the design value. Research cited in the test standard shows that the thermal resistance values of boards stored in a laboratory for a given period of time are consistent with those of boards removed from roof systems after the same period of time. Therefore, the test method prescribes accelerated aging in the laboratory to measure the change in thermal resistance, which is attributable to changes in cell gas composition caused by diffusion of air into and blowing agent out of the foam cells. This phenomenon is commonly called aging. S770 applies to any foam plastic insulation that uses a blowing agent other than air to enhance thermal performance and, as a result, ages. Foam plastic insulation types subject to this method include polyiso, extruded polystyrene, and sprayed polyurethane, although so far only the polyiso roof insulation industry has formally introduced a set of long-term R-values (LTTR) determined according to the new test method.
S770 may be viewed as the culmination of evolving thermal test method research and development over the past 20 years or more. In the late 1970s and into the very early '80s, R-values were reported on the same piece of product literature in two columns, one for Time of Manufacture (TOM) and one for Aged, presumably for 180 days. The users of this information were then free to choose which value best served their purposes. Of course, the higher values at the time of manufacture (before aging occurred) were usually chosen because these values represented less expensive options, although they were also less representative of actual in-service thermal performance. At the same time, other methods calling for heat aging at varying temperatures were also used. Each method yielded a different R-value for the same product in the same thickness. Accurate product comparison, therefore, was difficult if not impossible, and a certain level of confusion naturally prevailed.
In 1981, an attempt to address this confusion was undertaken by the Roof Insulation Committee of the Thermal Insulation Manufacturers Association. This industry group developed a six-month conditioning practice to establish a standardized method for determining and reporting R-values. The committee chose the six-month period because it approximated the completion of the first phase of diffusion (aging), after which changes in cell gas composition would occur extremely slowly over many years. It was also a practical method that required a relatively short time period before R-values could be determined, an important consideration given the dynamic nature of foam insulation development. This method achieved widespread recognition throughout the roofing industry and was only recently replaced by S770.
In the late 1980s, a joint initiative involving private industry and the federal government led to research at Oak Ridge National Laboratory (ORNL) to develop a test method using very thin slices of foam to accelerate aging. This work resulted in ASTM C 1303-95, which introduced the thin slicing technique, but many viewed it as somewhat difficult to perform, preventing it from gaining wide acceptance. However, it was based on a very important scientific fact: the rate of gas diffusion (aging) is inversely proportional to the square of the thickness of the insulation. For example, a two-inch thick insulation board ages four times as fast as a four-inch-thick board. In other words, very thin slices age very quickly and thermal values of the slices can be used to predict the aged thermal values of full-thickness boards.
S770 is based on ASTM C 1303-95 but includes more prescriptive language, making the method easier and reportedly less expensive to conduct. According to S770, thin slices (six to 12 mm) are cut from the surface and the core of the product, parallel to the facers. The testing periods in days are then determined from the scaling equation provided in the test method. The following table illustrates the number of days required to age a 10 mm slice to the same state of aging that a full-thickness board would reach after five years, clearly revealing the relationship between thickness and aging, the scientific principle underlying the thin slice method.
Because of the aging process, the previously recognized R-values determined after six months of conditioning are higher than the average thermal values for a 15-year period (LTTR). Consequently, if a specification called for R-20 polyiso roof insulation before 1/1/03, 2.7-inch-thick polyiso would have been used in the roof system. Today, because of the change in test methods, a 3.3-inch-thick board would be required, a difference of 0.6 inch. Increased insulation thickness could affect certain details of roof construction, since it may require slightly longer mechanical fasteners and corresponding adjustments to wood nailer thickness.
As a result of this evolution in test method development, the roofing industry now has an accepted, scientifically based, consensus developed test method to predict with confidence the long-term (aged) thermal resistance value of certain foam plastic insulations. Understanding this important development in test methods is essential for correctly specifying polyiso roof insulation in today's low- and steep-slope roofing market.
Richard Roe, RRC, CCPR (firstname.lastname@example.org) is director of Technical Services at Atlas Roofing Corporation in Atlanta.