Saving energy with spray foam

Building codes require minimum R-values for all buildings, and the codes typically provide a prescriptive table that details the R-values required for specific climate zones. However, the R-values listed are based on laboratory test procedures and do not take into consideration field performance of insulation. Seldom does insulation perform in the laboratory as it does in the field, but advances in building-envelope technology and test procedures have given roofing professionals an opportunity to address many factors that can contribute to the total energy performances of building systems.

One example of this is spray polyure­thane foam (SPF). During the past 30 years, a great deal of scientific research has been conducted to identify factors other than R-value that affect thermal performances of roof and insulation systems. The Spray Polyurethane Foam Alliance (SPFA) has been active in promulgating research to put real numbers behind these concepts. During the past two years, SPFA has conducted thermal performance research in attics and wall assemblies with coordination and input from Oak Ridge National Laboratories (ORNL), Oak Ridge, Tenn.; R&D Services, Cookeville, Tenn.; National Association of Home Builders' Research Center; Syracuse University, Syracuse, N.Y.; and Architectural Testing Inc., York, Pa.

During the summer of 2006, SPFA asked Mark Bomberg, a building scientist at Syracuse University and principal of TI Research, a consulting company in Syracuse, to analyze the relative energy performances of SPF roof systems. The goals of the research were to evaluate the factors affecting energy performances of SPF roof systems in the field, review existing data from building-envelope research that relates to these factors, and develop a matrix that could be used by contractors to provide a more accurate estimate of the effective field performances of SPF roof systems in various climates and building scenarios.

The research also reviewed published scientific papers about the factors that affect thermal performances of roofing materials, including thermal drift (aging of gas-filled foam insulation), thermal bridges created by mechanical fasteners, effects of air movement on energy performance and moisture gains in roofs, and reduction of surface temperature on cool roofs, among others.

The research report, "Energy Performance of SPF Roofs," suggests a new concept for evaluating the relative energy performances of roof systems. According to the report, "since the concept of R-value represents dry materials that are not affected by air or moisture flows and does not include other aspects of energy performance, a new terminology is needed. We believe that focus on air and moisture flows requires equivalent consideration as heat transfer through an assembly. We, therefore, propose to replace R-value with new concepts developed in consultation with the industry and referred hereto as: Energy performance ratio (EPR) and energy performance R-value (R_ep)."

Following is a summary of some of the points within Bomberg's research, as well as existing literature, regarding R-value performance and predictions for SPF roof systems with reflective coatings. I also will discuss examples of R-value predictions for SPF over existing roof systems.

The new concepts

R_ep is thermal resistance measured with a standard test procedure that includes the effects of air infiltration and moisture ingress measured under standardized air pressure, temperature and humidity gradients on thermal resistances of building assemblies.

EPR measures the efficiency of thermal insulating materials. EPR is defined as the ratio between Rep and standard R-value, which does not include the effect of air infiltration and moisture on thermal performance of a building assembly.

The research report includes only information published in public domain and, therefore, the values discussed may not represent some service conditions; however, they identify the magnitude of probable performance reductions.

Additionally, the report explains the reason for better retention of thermal resistance in thick layers of SPF and other foam insulation that use nonwater-blown systems. Although there are no test results for the currently used blowing agent HFC-245fa, taking into consideration the physical properties of this blowing agent, roofing professionals can expect results similar to those obtainable with the recently discontinued blowing agent CFC-11.

The report also shows there already is a sufficient amount of information to allow estimation of the savings related to use of reflective coatings. The research reiterates that surface reflectance of roof surfaces has a greater effect on roofs with low R-values. As the R-value of a roof system increases, the benefit of a reflective surface reduces to 30 percent, 13 percent and 10 percent for roof systems with R-values of 6, 12 and 18, respectively.

However, the research shows the savings strongly depend on surface cleaning and coating maintenance. If you take a worst-case scenario of conditions and use a surface or coating reflectance of 0.6 and compare it with a roof with a reflectance of 0.2, the reflectivity benefit reduces to 9 percent, 6 percent and 5 percent for roof systems with R-values of 12, 18 and 24, respectively. But energy cost savings ranging from 6 percent to 8 percent for an R-24 roof are realistic.

ORNL research is presented in the report and includes experimental verification of values predicted with computer models. The diminishing R-value effects of fasteners and gaps were combined and used as one entry. The percentage of these effects is similar in different climates so the diminishing R-value winter effect of 23.8 percent and summer effect of 17 percent can be used.

The ratio (0.77) between measured (8.6 percent) and calculated (11.2 percent) effect of fasteners in a 4-inch-thick polyisocyanurate roof is used to estimate the R-value results for energy performance calculations, which are 18.3 percent R-value reduction for winter and 13.1 percent R-value reduction for summer.

The report also explains why the effect of moisture is difficult to address without an advanced heat, air and moisture computer model. The report presents the inherent complexity of moisture transport and role of moisture redistribution within building materials. Unless computer modeling is performed for a roof's construction, the report suggests using the following data: For a cold climate region, an average value of 18 percent reduction of labeled R-value is used, and for a mixed climate, 30 percent reduction of labeled R-value appears appropriate for thermal insulation materials permeable for moisture.

The value of SPF

Using the data from the research, consider a 3-inch-thick fiberglass roof with 0.25 percent moisture content. The multipliers caused by the effect of fasteners and moisture are 0.845 and 0.82, respectively (these numbers are calculated from the research). Therefore, the resulting actual thermal resistance is 0.82 x 0.845 x 10.5 = 7.3 (ft² hr F)/(Btu) as compared with the labeled R-value of 10.5.

Let us consider what happens if you install a new SPF roof over a conventional built-up roof system with 1-inch thickness and reflective coating and one-way roof vents (one per 500 square feet) to permit drying of fiberglass insulation. The external insulation reduces the effect of mechanical fasteners and improves the contribution of the old roof to 0.915 effective R-value performance of the existing fiberglass insulation.

The effect of reflective coating on the R-15 roof (SPF plus the fiberglass) is assumed to be 11 percent, so an SPF roof adds 1.11 x 6 (typical aged R-value of SPF) = 6.66 (ft² hr F)/(Btu). The total thermal resistance of the newly installed SPF roof system with 1-inch SPF reroofing becomes 9.7 + 6.7 = 16.4 (ft² hr F)/(Btu).

If 2-inch-thick foam were used for reroofing, the effect of reflective coating would be less (9 percent) but long-term thermal resistance of the foam would be higher. An SPF roof would add 1.09 (10.9 percent) x 2 (inches of SPF) x 6.15 (R-value per inch) = 13.4 (ft² hr F)/(Btu), bringing the total R-value to 22.1 (ft² hr F)/(Btu), which is 13.4 plus the fiberglass R-value of 9.7.

This example highlights that because of synergy in controlling heat, air and moisture, adding 1 inch of foam and coating increases the in-field R-value performance of the existing permeable roof by 8.9 (ft² hr F)/(Btu) and by adding 2 inches of foam and coating increases the in-field R-value by 14.8 (ft² hr F)/(Btu).

More research

As new test procedures are developed and additional data quantified, a more complete evaluation of the energy performances of SPF roof systems can be obtained. The full report is expected to be available during the first quarter of 2007 and can be purchased online from SPFA at www.sprayfoam.org or by calling (800) 523-6154.

Mark Bomberg is a building scientist at Syracuse University, Syracuse, N.Y., and principal of TI Research, a consulting company in Syracuse. Mason Knowles is executive director of the Spray Polyurethane Foam Alliance.