Condensation in Wall Assemblies: Can Vapor Diffusion Through Highly Permeable Air Barriers Increase the Risk?

3 1 s t RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 1 0 – 1 5 , 2 0 1 6 C ô t é e t a l . • 9 1
Condensation in Wall Assemblies:
Can Vapor Diffusion Through Highly
Permeable Air Barriers Increase the Risk?
Jean-François Côté, PhD; Hank Staresina;
and Rémi Saucier, TP
Soprema Inc.
1668 J.B. Michaud St., Drummondville, QC J2C 8E9
Phone: 819-478-8166 • Fax: 819-478-8044 • E-mail: soprema@soprema.ca
Abstract
Blindside waterproofing provides challenges that can be addressed using prefabricated
polymer-modified bitumen sheet membranes. Benefits of a fully adhered system, such as
preventing lateral water migration on the interior side of the membrane, will be presented,
along with ways of taking advantage of the concrete curing process for an intimate bond
between the concrete and membrane. Blindside waterproofing systems based on modifiedbituminous
sheets and accessories meet the expectations of designers for soil conditions
and contaminants and gas permeability of membranes, while providing contractors the ease
of installation and detailing. They allow a cost-effective solution to accomplish watertight
structures.
Speakers
Jean-François Côté, PhD — Soprema Inc.
Jean-François Côté joined Soprema in 1999 as a research chemist and has been the
company’s director of strategic development since 2009, coordinating the activities related
to product and systems development. He is co-chair of the ASTM D08-04 Subcommittee and
is an active member of various technical committees in organizations such as ARMA, SPRI,
CSA, and ULC.
Hank Staresina — Soprema Inc.
Hank Staresina is Soprema’s technical specialist for building envelope solutions. He
has been a part of the Canadian construction industry since 1980. Hank has been involved
with a number of high-level manufacturers, providing him with a vast knowledge of industry
practices. He serves on the board of directors of the Building and Concrete Restoration
Association and is past president of the Sealant Waterproofing Association of Ontario.
9 2 • C ô t é e t a l . 3 1 s t RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 1 0 – 1 5 , 2 0 1 6
INTRODUCTION
Vapor diffusion is the process of water
molecules moving through porous materials
(e.g., wood, insulation, plastics, concrete,
etc.) driven by differences in vapor pressure.
It is one of the many mechanisms that move
water through wall assemblies. Vapor diffusion
itself isn’t the cause of moisture problems;
rather it is the moisture deposited by
this and other processes that may create
moisture concerns, such as mold, wood
decay, and corrosion. To minimize these
problems, building codes attempt to provide
some level of vapor diffusion control.
Some materials can block vapor diffusion
and are called vapor-impermeable
or vapor retarders, such as polyethylene
or asphaltic membranes. Other materials
freely allow water diffusion and are called
vapor-permeable, like fiberglass batt insulation.
Most other materials fall somewhere in
between these extremes.
All components of wall assemblies serve
one or more important purposes. Wall membranes
installed behind the cladding are
primarily used to protect the assembly
from wetting during construction and provide
a water-resistive barrier (secondary
plane of water protection). In many modern
wall assemblies, this membrane may also
become part of the air barrier system and
therefore form an integral component to the
vapor control system.
Today’s offering of air barrier membranes
is quite extensive. They may be mechanically
fastened or adhered sheet membranes, or
liquid-applied membranes that are rolled,
sprayed, or brushed onto the sheathing.
There are dozens of products available on
the market, made from a number of different
materials, from organic fibers to synthetic
plastics with material properties featuring an
extensive range of values.
Among these properties, water vapor
permeance is no exception. Air barrier
membranes can be found with permeance
ranging from zero (in the case of air/vapor
barriers) to above 75 U.S. perms (highly
permeable). The appropriate selection (and
positioning) of all components is critical to
ensure proper moisture management in wall
assemblies. However, some misunderstandings
persist within the industry that highly
permeable membranes are superior to less
permeable alternatives. This is not always
the case.
Ideally, all moisture sources should be
controlled, thus eliminating any concerns of
moisture damage. Most building assemblies
include some materials that are moisturesensitive.
These are materials that may lose
their functionality or pose health risks to
the building occupants if subjected to elevated
amounts of moisture. However, eliminating
all moisture sources over the service
life of a building is difficult. Consequently,
walls should be designed to allow for drying
of incidental moisture accumulation or construction
defects.
Wet materials in a wall assembly will
naturally dry by diffusion in the directions
of lower water vapor pressure. When a
polyethylene vapor retarder is used behind
the interior drywall, the greatest amount
of drying occurs to the exterior, and so the
drying rates of the wall assembly benefit
most from highly permeable exterior membranes.
While a very permeable membrane
will permit the greatest amount of drying,
there are diminishing returns. A plot of
normalized drying times, based on Fick’s
Law of diffusion, is shown in Figure 1. The
normalized drying rate is the percent efficiency
of a membrane to an infinitely permeable
membrane. For instance, a 10 U.S.
perm membrane is roughly 90% effective,
whereas a 50-perm membrane is 98% effective.
Consequently, changing a 10-perm
membrane to a 50-perm membrane only
improves the drying times by about 8%;
increasing a 50-perm membrane to a 100-
perm membrane only improves the drying
times by about 1%.
In this study, water vapor permeance
measurements were performed using two
different methods on five self-adhered air
barrier sheet membranes. Typical permeance
values obtained in this study were
used in hygrothermal modeling in order to
quantify the level of risk associated with
condensation, water accumulation, and
mold growth in typical wall assemblies.
EXPERIMENTAL
Water vapor permeance testing was
conducted on all membranes in accordance
with ASTM E96 – 2010, Standard Test
Condensation in Wall Assemblies:
Can Vapor Diffusion Through Highly
Permeable Air Barriers Increase the Risk?
3 1 s t RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 1 0 – 1 5 , 2 0 1 6 C ô t é e t a l . • 9 3
Figure 1 – Relationship between normalized drying rate and permeance.
Methods for Water Vapor Transmission of
Materials.
The determination of water vapor permeance
of materials (in this case, the selfadhered
membranes) by ASTM E96 uses an
apparatus called a cup. Both methods of
E96 were used in this study:
• Desiccant Method, or “dry cup,”
where desiccant is placed in the
cups, leaving an air space at 0% RH
• Water Method, or “wet cup,” where
distilled water is placed in the cups,
leaving an air space at 100% RH
In both methods, the material being
tested was used to seal the opening of
the cup, like a lid, such that it separates
the interior of each cup from its surroundings.
The cups were then placed in
a climate chamber under controlled temperature
and relative humidity conditions
at 23±0.2°C (73.4±0.4°F) and 50±2% RH.
These conditions are commonly referred to
as “Procedure A” for the Desiccant Method
and “Procedure B” for the Water Method
in ASTM E96. The vapor pressure gradient
created between the water/desiccant in
the cups and climate chamber conditions
(1,400 Pa or 0.203 psi) results in water
vapor either entering the cups (when using
the Desiccant Method) or leaving the cups
(when using the Water Method) by diffusion
through the test material.
The filled and sealed cups are then
weighed at regular intervals using a laboratory
scale. Water vapor diffusion (leaving or
entering the cup) can be calculated when
successive weight measurements, plotted
on a graph, indicate that the rate of weight
gain or loss has attained a steady state.
Permeance is then expressed as the weight
of water vapor diffusing through the material
as a function of time, per unit area and
vapor pressure difference. Both methods
were completed in triplicate.
Membranes were also tested for water
vapor transmission rate using another test
method, ASTM F1249-13, Standard Test
Method for Water Vapor Transmission Rate
Through Plastic Film and Sheeting Using a
Modulated Infrared Sensor.
Testing was conducted at 23°C (73.4°F)
using the PERMATRAN-W 3/33 Water
Vapor Permeability Instrument by MOCON.
This instrument uses two diffusion cells
for simultaneous duplicate testing. Each
diffusion cell has two halves separated by
the material to be tested. In one half, a
test gas containing water vapor flowed (the
test gas used was moisture-saturated air,
100% RH). In the second half, a carrier
gas flowed (dry nitrogen, 0% RH), creating
a water vapor pressure difference of 2,800
Pa or 0.406 psi across the tested material.
The carrier gas swept constantly across the
“dry” side of the material and collected all
water vapor diffusing through it.
Carrier gas containing the permeating
water vapor was brought to a pressuremodulated
infrared (IR) detector at a rate of
100 cc/min. This detector measures infrared
energy absorbed by water vapor and
produces an electrical signal. The amplitude
of the signal is proportional to the
water vapor concentration. The amplitude
is then compared to the signal produced
by the measurement of a calibration film
of known water vapor transmission rate.
With a known test film area and the measured
water vapor, the transmission rate
for the test material can be determined and
expressed in units of weight per unit area
per day. This number was then converted
to permeance by using the vapor pressure
difference between the cell halves.
PERMEANCE RESULTS
Results of the experimental testing performed
on the five membranes are shown
(as averages) in Table 1. Membrane E could
not be measured using ASTM F1249, as the
vapor transmission rate exceeded the range
of the test equipment. The same applied to
a sixth membrane (A, not shown here) for
which ASTM E96 results will be known later.
Results are generally in agreement with
manufacturer-published data, except for
Membranes E and F (higher permeance).
Variability around the average was low for
all membranes in all test methods, except
Membrane D, for which variance was significantly
higher, exceeding 47% in the Water
Method, for example.
Because the ASTM E96 methods are
performed with air at different RH values
in the cup, and ASTM F1249 is performed
with nitrogen at 100% RH, it is not shocking
to observe different permeance results for a
given air barrier membrane, depending on
the method used. The general trend, however,
is respected for all test methods.
Building codes have grouped materials
into classes depending on their water vapor
permeance:
• Class I vapor retarder: <0.1 U.S.
perm
• Class II vapor retarder: 0.1 to 1.0
U.S. perms
• Class III vapor retarder: 1.0 to 10
U.S. perms (semipermeable)
• V apor permeable: >10 U.S. perms
All air barrier materials tested in this
study exhibited permeance results (as per
ASTM E96 Water Method) in the “high
permeance” category (from 10 to 76 U.S.
perms).
HYGROTHERMAL MODELING
The WUFI 5.2 Pro (WUFI) computer
model was used to simulate the hygrothermal
performance of wall assemblies in
various configurations and climates. The
simulations were repeated for four different
climate zones. Major Canadian cities
in each climate zone were selected from
the WUFI climate file database representing
10th percentile cold or hot years over a
30-year period:
• Temperate Marine (Climate Zone
4C): Vancouver, British Columbia
• Cold-Humid (Climate Zone 5):
Toronto, Ontario
• Cold-Dry (Climate Zone 6): Québec
City, Québec
• V ery Cold-Dry (Climate Zone 7):
Edmonton, Alberta
9 4 • C ô t é e t a l . 3 1 s t RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 1 0 – 1 5 , 2 0 1 6
Water Vapor Permeance (U.S. perms)
Test Method ASTM E96 ASTM F1249
Membrane Water Method Dessicant Method
B 10.1 6.2 20.8
C 23.1 19.0 44.9
D 30.1 18.0 41.0
E 71.4 37.7 –
F 75.8 39.9 52.7
Table 1 – Permeance results using three test methods at 23°C (73.4°F).
Interior boundary conditions were established
by assuming a high indoor moisture
load representative of a residential building
with low outdoor ventilation rates and high
moisture production (EN 15026 approach).
This model generated relatively high indoor
humidity values and therefore established
conservative estimates on drying rates and
wall behaviors.
The rain load was determined using
ASHRAE Standard 160P, with
a rain exposure factor of 1.0
and rain deposition factor
of 1.0. This established an
upper limit to the degree of
wetting of the cladding. To
maximize solar radiation, the
wall orientation was directed
due south, with a short-wave
absorptivity of 0.4 and longwave
radiative emissivity of
0.9. These values are approximate
for a light-grey cladding,
such as stucco or fiber
cement siding. Explicit radiation
balance was also included,
with the default settings,
to provide better resolution of
the long-wave counter radiation
with the environment.
The parametric study was
based on a typical 2-x-6-in
wood-framed wall with ½-in.
plywood sheathing, with an
interior polyethylene sheet
vapor retarder and ½-in. gypsum
wallboard. The wall was
insulated with R-19 fiberglass
batt insulation. The parameters
and their values are
provided in Table 2.
The simulations considered the impact
of nonstoring cladding (materials that
absorb or store relatively small amounts of
water when exposed to rain such as vinyl
siding or fiber cement panels) and reservoir,
or moisture-storing cladding (materials
such as brick and stucco that can
hold significantly larger amounts of water).
The respective water-storage capacities were
used from internal WUFI property data. Also
considered was the impact of additional
exterior insulation (either
vapor-permeable or impermeable)
from R 0 to R 18. Air barrier membrane
vapor-permeance variables
were selected at three levels: 1, 10,
and 30 U.S. perms.
Cross sections of one typical
nonstoring and one typical storing/
reservoir wall used in simulations
are shown in Figure 2 and Figure 3,
respectively.
Drying rates were calculated by
setting the wood sheathing moisture
content (MC) to 28%, the moisture
content at which fungal deterioration
can occur, and evaluating
the time required for the sheathing to dry
below 20% MC, the lower limit for fungal
growth. The simulations were started at the
beginning of February to mimic wintertime
wetting, and drying times were measured
for one year. Drying times in excess of a
year were deemed to be at extremely high
risk of moisture damage and decay.
Hygrothermal simulations assessing the
time for saturated sheathing (i.e. 28% MC)
3 1 s t RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 1 0 – 1 5 , 2 0 1 6 C ô t é e t a l . • 9 5
Figure 2 – Nonstoring cladding, no exterior insulation wall assembly.
Figure 3 – Storing cladding, exterior insulation wall assembly.
Table 2 – Hygrothermal study parameters and values.
Variable V alue
Cladding Moisture Storage Capacity Nonstoring: 5.2 kg/m2
Storing/Reservoir: 34.0 kg/m2
Drainage Cavity Ventilation Rate Nonstoring: 100 ACH*
Storing: 5 ACH
Exterior Insulation Vapor Permeance Permeable: 178 ng/Pa·s·m2 (3.1 U.S. perm)
Impermeable: 1.15 ng/Pa·s·m2 (0.2 U.S. perm)
Exterior Insulation Thermal Resistance Thermal Resistance (IP)
(R-value) None/R-6.5/R-18
Air Barrier Membrane Vapor Permeance 57.2/572/1716 ng/Pa·s·m2
(1.0/10/30 U.S. perm)
*ACH: Air changes per hour
to dry below the safe level (i.e., 20% MC)
are presented. Figure 4 shows simulation
results for the Vancouver (Zone 4C) climate,
and Figure 5 does the same for the Québec
City (Zone 6) climate.
Wall assemblies—not including exterior
insulation or including vapor permeable
exterior insulation—had measurable differences
in drying rates, correlated to the
vapor permeance of the air barrier membrane.
However, results show that although
membrane permeance has a large impact
on the drying rates between 1 and 10 U.S.
perm, improvements achieved by increasing
membrane permeance from 10 to 30
U.S. perm are small. The biggest differences
were observed when no exterior insulation
was present. In some assemblies, the faster
drying time is as little as two extra days. In
contrast, it can sometimes take weeks for
mold growth to occur under ideal conditions.
Functionally, the difference between
selecting a high-permeability or “very high”-
permeability membrane has a lesser impact
on the moisture performance of the wall
than other design considerations, such as
the type of cladding or use of exterior insulation.
Walls with impermeable exterior insulation
(not shown) are double vapor retarders;
they do not dry within a one-year period,
regardless of the membrane permeance.
SEASONAL MOISTURE
VARIATIONS
The long-term moisture performance
was assessed by running the simulations
for several years until annual equilibrium
was met; that is, repeatability between successive
seasons was achieved. The metric
for comparison was the number of hours
with high risk of fungal growth or decay at
the sheathing level, and these hours were
called “critical hours for biodeterioration.”
These critical conditions were temperatures
between 4°C (39.2°F) and 35°C (95°F), and
at sheathing moisture contents exceeding
20%. It should be noted that these simulations
do not include incidental moisture
leaks or air leakage. Figure 6 and Figure 7
show the impact of air barrier membrane
permeance and exterior insulation on the
number of critical hours for biodeterioration
in a year, as obtained from simulations
for various assemblies with brick cladding.
Assemblies with fiber cement cladding (not
shown) had very low numbers of critical
hours of biodeterioration.
Of the modeled assemblies, the only
wall types that are at risk of decay are those
with 1.5 in. or no exterior insulation behind
a reservoir cladding. Wall types without
reservoir claddings are not at appreciable
risk to decay regardless of membrane permeance,
although simulations indicate that
using a 10-U.S.-perm membrane is the
9 6 • C ô t é e t a l . 3 1 s t RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 1 0 – 1 5 , 2 0 1 6
Figure 4 – Number of days required to dry plywood sheathing to 20% MC for various
wall assemblies (following hygrothermal simulations in Vancouver – Zone 4C).
Figure 6 – Number of critical hours for biodeterioration in a year for brick
cladding (following hygrothermal simulations in Vancouver and Toronto).
Figure 5 – Number of days required to dry plywood sheathing for various wall
assemblies (following hygrothermal simulations in Québec City – Zone 6).
least susceptible to cause risks (less susceptible
than when using a 30-U.S.-perm
membrane). Wall assemblies with 1.5 in.
or no exterior insulation behind a reservoir
cladding have a direct relationship between
membrane permeance and hours of biodegradation.
In these cases, using a “very high”
permeance air barrier membrane should be
avoided.
To better examine the yearly impacts
on the sheathing performance in reservoirclad
wall systems, the annual sheathing
moisture contents were plotted for
R-6.5-permeable exterior insulation and
compared to assemblies with no exterior
insulation. Figure 8 and Figure 9 present
the annual evolution of sheathing moisture
content for the Vancouver (Zone 4C) climate
and the Québec City (Zone 6) climate,
respectively.
Analysis of the annual plywood moisture
content reveals that highly permeable
membranes perform best when used in
conjunction with permeable exterior insulation.
However, the improvements compared
to a 10-perm membrane are minimal.
Overall, addition of exterior insulation has
a more significant impact on sheathing
performance than any variations in membrane
permeance. In the absence of exterior
insulation, it was found that mediumpermeance
air barrier membranes perform
slightly better than either highly permeable
or impermeable membranes in all of the
simulated climate zones. In all instances,
the use of a 10-perm membrane does not
appreciably increase the risk compared to a
30-perm membrane.
CONCLUSION
Prevention of moisture problems is the
first and most important step in ensuring
the long-term performance of all wall
assemblies. However, if leaks do occur, an
assembly that can dry will invariably perform
better than one that does not.
The main causes of moisture problems,
in order of significance, are bulk water
leaks, air leakage condensation, construction
moisture, and lastly, water vapor diffusion.
Proper rainwater management strategies
and detailing of the water-resistive barrier
are fundamental to minimize bulk water
leaks, whereas continuous air barriers and
exterior insulation are keys to managing
condensation resulting from air leakage.
Construction moisture and vapor diffusion
are managed by the proper placement
and selection of vapor control layers
and careful use of impermeable materials.
Proper installation following good construction
practices will also greatly contribute.
With regard to drying, a more permeable
membrane will enable more drying
than a less permeable one but will also
allow more water vapor to enter the wall
assembly through inward vapor drive when
3 1 s t RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 1 0 – 1 5 , 2 0 1 6 C ô t é e t a l . • 9 7
Figure 7 – Number of critical hours for biodeterioration in a year for brick
cladding (following hygrothermal simulations in Québec City and Edmonton).
Figure 8 – Annual evolution of sheathing moisture content using various wall
configurations (following hygrothermal simulations in Vancouver – Zone 4C).
Figure 9 – Annual evolution of sheathing moisture content using various wall
configurations (following hygrothermal simulations in Québec City – Zone 6).
a reservoir cladding is used. Furthermore,
the membrane vapor permeance must be
considered in conjunction with the adjacent
layers in the wall assembly. A highly
permeable membrane is not as effective if
the vapor diffusion is already restricted by
other layers in the assembly. In addition,
highly permeable membranes are subject
to diminishing returns, whereby increasing
the permeability yields smaller and smaller
benefits to drying.
The wetting and drying characteristics
of wall assemblies are complex, and there
is no universal solution. Resorting to high
permeance membranes may not be the right
approach in all instances; similarly, lowpermeance
membranes are not suitable for
all applications. In many cases, the vapor
permeance of the air barrier membrane has
little or no influence on the performance of
the wall assembly.
As demonstrated, the thickness and
type of exterior insulation and other materials,
including the cladding and interior
vapor control layer also have an impact on
the selection of the air barrier membrane
vapor permeance; therefore, widespread
industry recommendations cannot be made
as to whether a lower or higher vapor permeance
is better. Consequently, specifying
and positioning the vapor control layer must
be done holistically with the design of the
enclosure.
9 8 • C ô t é e t a l . 3 1 s t RC I I n t e r n a t i o n a l C o n v e n t i o n a n d T r a d e S h ow • Ma rc h 1 0 – 1 5 , 2 0 1 6