Temperature-Dependent Thermal Performance of Closed-Cell Foam Insulation Board in Roof Constructions

Abstract

Thermal performance of an insulation material is influenced by the in-service temperature condition. Unlike most other insulation materials, thermal resistance (R-value) of polyisocyanurate (polyiso or PIR) foam insulation with “captive blowing agent” varies nonlinearly with temperature. Building designers consider the constant R-value of different insulating materials for building design and energy calculations, and hygrothermal simulation software tools, such as WUFI, consider the linear temperature-dependent R-value, even for polyiso. However, neither the linear temperature-dependent thermal resistance nor the constant thermal resistance value of polyiso represents the actual thermal performance. This paper aims to quantify the impact of in-service boundary temperature conditions in Canadian climates on the thermal performance of polyiso foam insulation board used in EPDM and PVC roof constructions. Hygrothermal simulations were performed using WUFI Pro, which considers real climate data and hygrothermal properties of constituent roof components for evaluating moisture and temperature conditions in roof constructions. Based on heating degree days (HDD), ten different cities were selected between climate Zone 4 (HDD < 3,000) and Zone 8 (HDD ≥ 7,000). The thermal resistance measurements were conducted using heat flow meter apparatus on four polyiso insulation boards (two new and two aged) of different sizes [thickness, new: 1 in. (25 mm) and 2 in. (51 mm); aged: 2 in. (51 mm) and 3 in. (76 mm)] at five mean temperatures of −4°C (25°F), 4.5°C (40°F), 10°C (50°F), 24°C (75°F), 43°C (110°F), and at a temperature differential of 28°C (50°F). The measured thermal resistance data of the four samples at different mean temperatures were normalized with calculated thermal resistance of each sample at 22°C (72°F). The normalized R-value variation was calculated using in-service boundary temperature conditions determined from hygrothermal simulations and considering linearly varied thermal resistance with temperature, for the selected ten Canadian cities. The normalized R-value data of four polyiso samples revealed 17.4% decrease in R-value under winter conditions of Zone 8.