Moisture in the Roofs of Cold Storage Buildings
WAYNE TOBIASSON AND ALAN GREATOREX
In a prior study where we used an infrared scan-
INTRODUCTION
ner to find wet insulation in a ballasted roof over
Several kinds of roofing systems of cold stor-
refrigerated spaces (Tobiasson and Greatorex
age buildings (i.e., freezers and coolers) were in-
1994), numerous core samples revealed that the
spected and sampled for entrapped moisture. The
infrared scanner missed large areas of wet insula-
information collected was used to develop main-
tion deep in the roof. Indoor-outdoor temperature
differentials were slightly less than the 27F (15C)
tenance, repair, and replacement recommenda-
tions for these roofs. That information was also
minimum, then specified in ASTM C1153 (ASTM
used to develop recommendations for improving
1990). As a result, we recommended increasing
that minimum to 32F (18C). The current version
the design of freezer and cooler roofs.
The following tasks were performed:
of C1153 contains that recommendation.
Nighttime on-the-roof infrared roof mois-
The indooroutdoor temperature differentials
ture surveys (Tobiasson and Korhonen 1985)
for these Texas roofs were much greater than ei-
(Fig. 1).
ther of these minimums. We probably found most
Daytime indoor infrared roof moisture
of the wet areas in these freezer roofs, but some
surveys.
questions remain and additional core samples are
Visual inspections of the membranes and
needed to verify some of our findings.
their flashings.
The large indooroutdoor temperature differ-
Core sampling of membranes and insula-
ences present during these surveys tended to cause
tions and subsequent gravimetric measure-
the top surface of the roof to be cooler (i.e., darker
ment of the moisture content of those cores.
in the thermal image) where wet insulation was
Removal of 12- 12-in. ( 30- 30-cm) speci-
present. However, when the wet insulation was
mens of the various insulations for labora-
located near the top of the roof, a significant
tory determination of the following proper-
amount of solar energy was stored in it during the
ties:
day. At night this "hot water bottle" of energy kept
1. Dry density
the surface of the roof there hotter (i.e., brighter in
2. Moisture content
the thermogram) than at areas containing dry in-
3. Thermal resistance--as obtained
sulation. Thus, the nature of the thermal anomaly
(i.e., hot/bright or cold/dark) changed with the
4. Thermal resistance--after drying.
type of insulation present, the amount of mois-
ture it contained, and how deep in the roof the
The 2-in.- (5-cm-) diam. cores were analyzed at
CRREL. The 12- 12-in. (30- 30-cm) specimens
moisture was located.
The infrared surveys conducted within the
were analyzed at Owens Corning.
freezers always detected bright (i.e., hot) thermal
anomalies where wet insulation was present in the
roof, since both effects (i.e., loss of insulating abil-
INFRARED SURVEYS
ity upon wetting and enhanced storage of solar
Temperatures varied from daytime highs of
energy in wet insulation) tend to warm that side
88F (31C) to nighttime lows of 62F (17C). The
of the roof.
humidity was high, and a few showers were en-
countered that required us to halt coring and cut-
ting operations and nighttime infrared surveys.
THE ROOFS
However, not enough rain fell to prevent the sur-
A plan view of the seven roofs in Dallas (roofs
faces of these roofs from drying out during the
D1D7) is shown in Figure 2. The temperatures
heat of the day.
Previous Page
