The Long-Term Thermal Resistance (LTTR) of roof insulation can be different from the laboratory performance due to many factors that can be mitigated based on the type of insulation used. Photo courtesy of ROXUL, Inc.

R-Value Comparison: Stone Wool and Polyisocyanurate

With a general framework established for looking at R-values in insulation, let's look more closely at two specific types of insulation that are used for roofing applications, namely stone wool board and polyisocyanurate (polyiso) board insulation.

Insulation Characteristics

Polyisocyanurate is a commonly used insulation board, particularly for roofing. The trade association known as the Polyisocyanurate Insulation Manufacturers Association (PIMA) defines it as a closed-cell, rigid foam board insulation consisting of a foam core sandwiched between two facers. The facers are composed of various organic and inorganic materials, usually paper and fiberglass. It is widely used in residential and commercial markets for both roof and side wall applications. Among the characteristics that have contributed to its popularity is a comparatively high R-value per inch of thickness. In fact, it can currently claim to have the highest tested R-value per inch of any other commercially available insulation on the market. The foam core is moisture resistant, the board can achieve an FM Class 1 fire rating, it is widely available, and competes well on price.

By contrast, stone wool roofing insulation does not contain any foam plastics or blowing agents. Rather, it is manufactured out of volcanic basalt rock mixed with some recycled slag in a furnace. The melted rock is then literally spun into wool with minor amounts of organic binder and process additives to create the desired density of the end product as a rigid board. The stone wool fibers are typically non-directional which is important for achieving the multiple performance characteristics of the final product. In terms of R-value, stone wool provides very good performance, but most notably, that performance has been shown to remain constant over time. Since there are no blowing agents, there is nothing to out gas, meaning that the thermal performance does not change. It also has superior qualities in regard to fire resistance, dimensional stability, sound attenuation, and water repellency.

When looking to select or specify roofing insulation, architects and designers may first look at the published R-values of these two insulation products based on lab test results. Stone wool insulation, when tested under ASTM protocols at a mean temperature of 75°F (i.e. 50°F outside, 100°F inside) will yield an R-value of approximately R-4.0 per inch. However when tested at lower mean temperatures, such as 25°F (0°F outside and 50°F inside) that R-value has been shown to increase up to approximately R-4.4 per inch. Conversely, at higher mean temperatures, such as 110°F (135°F outside and 85°F inside) it tests at about R-3.7. These values may vary based on different products from different manufacturers, but the trends have been clearly established and are characteristic of the material.

R-values for polyisocyanurate insulation are similarly tested per ASTM standards at a mean temperature of 75°F. They are further tested per the LTTR method for a weighted average aged R-value. Most manufacturers now publish an R-value of R-5.7 per inch as a result of that testing. However, as far back as 2003, the National Roofing Contractors Association (NRCA) has performed independent testing of polyisocyanurate insulation, through a third-party laboratory, by using actual samples from different manufacturers. These tests were performed according to ASTM protocols and used the 75°F mean temperature as the basis. Additionally, the ASTM protocol also specifies, but does not require reporting, of testing samples at mean temperatures of 25, 40, and 110°F which the NRCA performed to see if there were any differences in R-value.

Comparison of R-values for faced polyisocyanurate insulation boards prior to JAN. 1, 2014. Chart courtesy of ROXUL, Inc., with permission from NRCA, PROFESSIONAL ROOFING magazine.

This work has been carried out and reported in numerous articles and reports in Professional Roofing magazine and elsewhere. The NRCA testing concluded that there is indeed a difference based on temperature and there is a notable range in performance between manufactured products. Of note, it determined that as the mean temperature drops below 75°F, the R-value also drops down to an average low of about 4.75 at a 25°F mean temperature. Further, when the outdoor temperature gets above the 75°F mean, the R-value also drops below the R-6 value, which was claimed by polyiso manufacturers prior to January 1, 2014. These findings became the basis for the NRCA to recommend using a more representative value of R-5.6 per inch when designing for warm climates and R-5.0 per inch when designing for cold climates.

Compared to polyisocyanurate insulation, stone wool insulation thermal performance improves at cold temperatures and performs consistently at all temperatures. Graph courtesy of ROXUL, Inc.

Comparison of R-values for faced polyisocyanurate insulation boards after JAN. 1, 2014. Chart courtesy of ROXUL, Inc., with permission from NRCA, PROFESSIONAL ROOFING magazine.

Independent Laboratory Testing

Separate from the NRCA studies, an independent organization known as Building Science Corporation (BSC) has conducted a series of similar tests on different thicknesses of polyisocyanurate insulation at different test temperatures that more closely replicate anticipated environmental conditions. Mean temperatures in the real world aren't governed by a 50°F temperature difference between inside and outside, so they conducted tests with a constant 72°F inside temperature and a variable outside temperature to generate the relevant mean temperatures.

On this basis, BSC first tested samples of 2-inch-thick polyisocyanurate insulation which had a commonly published LTTR aged R-value of R-12.1 or just slightly better than the R-6 per inch for a 1-inch sample. The NRCA doesn't differentiate between solid or stacked boards, so their 2 inch recommended R-value is simply twice the R-5.6 or R-11.2. All of the samples tested showed a decrease in R-values at winter outdoor temperatures below 32°F or summer solar heated roof temperatures above 113°F. Further, they determined that the change in performance is not linear across temperature changes meaning that it is very difficult to predict actual performance between tested temperatures. They also found that there was a notable variation in performance between manufacturers. In essence, this test agreed with the NRCA findings indicating that the design R-values should be reduced from the published values for more accurate calculations.