Wetting behaviors of common insulation products were assessed by standard two-hour water immersion. Mineral wool slabs absorbed 8 to 38 times more water than foil-faced polyisocyanurate (PIR) and 4 to 19 times more water than coated glass-faced PIR. Drying within vented benchtop assemblies required 2 to 6 days longer for mineral wool as compared to PIR. Rewetting of mineral wool specimens increased water absorption by 130% to 190% and extended dry times by an additional four days. In comparison, sorption behaviors of PIR remained unchanged. Repeated wetting of mineral wool revealed dynamic holding capacities that varied on the basis of pore structure and slab macrostructure.
New practices in wall design favor exterior insulation outboard of the water-resistive barrier. The exterior insulation layer is now located within a highly variable environment prone to episodic wetting. Under these conditions, effective performance of common insulation materials may not align with design intent as assumed sorption characteristics for mineral fiber and cellular products reflect vastly different properties and test methodologies. These discrepancies, together with varied and unknown exposure scenarios, lead to high uncertainty regarding actual performance in response to liquid water.
Sorption properties of insulation materials are largely influenced by pore structure. Materials having voids that are accessible to adjacent pores and their external environments are referred to as ‘open pore’ (e.g. mineral fibers). Conversely, ‘closed pore’ structures have void spaces that are compartmentalized or closed to adjacent pores and their external environments (e.g. closed cell foam).
Because pore structure is so integral to water absorption, it has an influence in standard sorption testing. For example, test methods for closed pore foam employ full immersion in water (ASTM D2842, ASTM C272 or ASTM C209 (2 hour immersion). In contrast, sorption potentials for mineral wool are evaluated in accordance with ASTM C1104, which utilizes water vapor (95 ± 3% RH), not liquid water as the wetting medium.
In light of non-uniform testing methodologies, a direct comparison of sorption attributes for fibrous and cellular insulation is not possible. Therefore, professionals lack even a conceptual understanding of potential risks for current design practices.
The selected insulation materials represent products intended for exterior facades and rainscreen applications. Test specimens consisted of new, 2-inch thick, 1 ft x 1 ft panels free of visible defects or inconsistencies. Polyisocyanurate panels were bi-faced with either a tri-laminate foil or coated glass.
Sorption testing was performed in general accordance with ASTM C209 which includes two-hour full immersion under one inch of standing water. In accordance with ASTM C209, specimens were allowed to drain for 10 minutes prior to initial weighing.
Vented benchtop assemblies provided an estimate of drying characteristics within a mock wall assembly. Components included cladding, vented air space, wetted insulation specimen, and substrate (Fig. 1). Plastic angles served as closure backing at the top and sides to prevent excess convection. Assemblies remained open at the base and held off from the benchtop surface to allow venting and unobstructed drainage.
Water content was determined by weighing insulation slabs at 24-hour periods. Specimens were reported as effectively dry when two of the three replicates achieved a minimum of moisture content of 0.5% (weight basis).
Phase II rewetting entailed three independent studies, each involving three replicates and seven cycles. Wetting was performed by two-hour full immersion as previously described. After each wetting cycle, slabs were weighed and then oven-dried at 125 F to 150 F. This drying regimen reflects the upper temperature range for typical rainscreens and enclosure systems.
Water Absorption and Drying
Percent water absorption and corresponding drying rates are summarized in Figs. 2 and 3. As expected, sorption behaviors differed on the basis of facer type and pore structure. For example, fibrous mineral wool absorbed 8 to 38 times more water than foil-faced PIR and 4 to 19 times more water than coated glass-faced PIR. Drying times for the two PIR products remained virtually identical. High drying potentials for PIR were attributed to low water absorption coupled with the ability to release water vapor at facer-substrate interfaces, regardless of facer type.
Contrasting outcomes for the two mineral wool products were unexpected. Product declarations report similar densities, binder type, and binder fractions. Factors such as variations in binder distribution, binder curing, hydrophobic additives, fiber orientation, and layer crimping are therefore implicated. It should also be noted that the range of sorption values varied significantly between individual panels of a given product.
Drying times ranged from 1 to 7 days as a function of insulation type. Both PIR products were effectively dry at 24 hour whereas mineral wool slabs required an additional two to six days to achieve the same endpoint of 0.5%. Regardless of the employed methods, these findings defy common claims regarding wetting and drying behaviors of mineral wool.
The Effects of Initial Rewetting
Rewetting of PIR specimens showed no appreciable influence on water absorption. Rewetting of mineral wool specimens increased water absorption by 132% to 195%. Corresponding drying times within vented benchtop assemblies required an additional four days.
Prior research has shown that dry mineral fibers previously subjected to weathering or wetting exhibit increased moisture sorption. The presumed causes are linked to separation of fibers from binder resins or actual loss of binders and hydrophobic additives.
However, a more plausible scenario involves changes in void geometries resulting from bulk water transport during wetting and draining. These changes are expected to occur throughout the three-dimensional matrix where some void volumes increase while others decrease.
This study compared the behaviors of mineral wool and polyisocyanurate (PIR) in response to partial wetting with liquid water. The following conclusions may be drawn from these findings:
- Water absorption is inherently linked to pore structure.
- The reported sorption values for mineral wool are orders of magnitude greater than those derived from standard methods where high humidity, not liquid water, is used as the wetting medium.
- These results are aligned with prior accounts showing similar sorption attributes for mineral wool with corresponding effects to thermal performance. Notable reductions in claimed R-values are therefore expected as even partial wetting negates thermal performance to near negligible levels.
- Drying times are a function of holding capacity, which is particularly relevant for mineral wool prone to partial saturation.
- Claims regarding water repellency and matrix drainage must be balanced against the fundamental realities of pore structure.
- Mineral wool products vary considerably in their wetting and drying behaviors. Even greater differences are observed for slabs of a given product where absorption may vary by an order of magnitude.
This research was funded by a consortium of polyisocyanurate manufacturers.
Owens Corning: LIVE Webinar Series
Webinar Series Schedule
HLW developed a concept to bring healthcare and other experts to the "ground zero" of a disaster and mitigate it ...
Midwest Build Week Online
Midwest Build Week Online is the largest virtual/online event covering all the latest projects and opportunities for the construction industry in Chicago and the Midwest.
WHY DO WE FEEL BETTER WITH WOOD?
This commARCH White Paper focuses on how Wood shows psychological and physiological benefits, according to research.