Evaluation of Vapor Retarders in Roof Decks in High Humidity Environments by Hygrothermal Simulation

IN COLD WEATHER, high-humidity indoor
environments such as indoor pools pose risks of
moisture condensation in building enclosures,
which in turn can cause mold, decay, buckling,
corrosion, and eventually structural issues of
durability. Most swimming pools are treated
with chlorine, resulting in chloramines that
off-gas into the pool’s air space, irritating skin
and eyes and creating respiratory hazards for
swimmers and spectators. The chloramines also
corrode building materials, especially metals.
A good ventilation system1 with effective air
distribution and sufficient outdoor and exhaust
air to remove toxic and corrosive chloramines is
essential for the health and safety of swimmers,
spectators, and other building occupants.
Acoustics and ambient noise are significant
issues in natatoriums, as these spaces act as
giant echo chambers. Most large community
pools are considered learning spaces.
Therefore, under the Acoustical Society of
America (ASA)/American National Standards
Institute (ANSI) acoustic standards for physical
education teaching environments,2 these
types of natatorium spaces are required to
meet certain minimum reverberation times.
To control indoor acoustics, pools require
sound-absorbing materials that are widely
and uniformly distributed throughout the
natatorium. Both wall treatments and roofs
must be sound absorbers. Moisture loads in
a natatorium can be nearly two to three times
per unit volume the loads in a typical building.1
Therefore, moisture transport driven by vapor
pressure differentials enhanced by temperature
gradients across the enclosure is an important
factor to consider in the proper design of
the roofs.
Current literature provides few guidelines
for designing natatoriums to mitigate the
Feature
Evaluation of Vapor Retarders
in Roof Decks in High
Humidity Environments by
Hygrothermal Simulation
By Gourish Sirdeshpande, PhD
This paper was presented at the 2023 IIBEC
Building Enclosure Symposium.
Interface articles may cite trade, brand,
or product names to specify or describe
adequately materials, experimental
procedures, and/or equipment. In no
case does such identification imply
recommendation or endorsement by
the International Institute of Building
Enclosure Consultants (IIBEC).
risk of moisture condensation in roof decks.
To avoid any risk of moisture condensation,
design professionals frequently recommend
a vapor retarder on the warm side of the roof
assembly, presuming that a vapor retarder will
safeguard the roof deck from condensation and
moisture buildup.
Most roof installations in pools, irrespective
of climate conditions, are perforated metal decks
made of perforated metal, acoustic material,
vapor retarder, insulation, roof decking material,
and a roofing membrane (in that order, from
inside to outside), as shown in Fig. 1. The
entire assembly is anchored by screws that
hold the components together. This type of
anchorage typically results in pinholes and tears
in the retarder. Even when a vapor retarder is
used, corrosion is a commonly occurring issue
with metal roof decks in natatoriums. Such
occurrences question the role of a vapor retarder
in moisture buildup and condensation in these
roof decks.
Although building enclosure consultants and
roofing professionals may assume that there is
minimal risk of condensation with the use of a
vapor retarder, this assumption may be incorrect.
The study presented herein used hygrothermal
modeling to investigate moisture migration
and condensation in composite roof decks with
© 2024 International Institute of Building Enclosure C 18 • IIBEC Interface onsultants (IIBEC) March 2024
and without vapor retarders. In particular, the
effectiveness of a vapor retarder to prevent
condensation and moisture accumulation was
highlighted in this investigation.
STEADY-STATE DEW POINT
METHOD FOR ESTIMATING
CONDENSATION
Most building enclosure professionals use
the best-known simple steady-state design
tool, the dew point method,3 to evaluate
condensation in roofs. This method assumes
that steady-state heat conduction (under worst
boundary conditions) and moisture diffusion
govern heat flow and water vapor flow based
on inside and outside temperatures and water
vapor permeability of the assembly. Partial water
vapor pressures in the roof deck are compared
with saturation water vapor pressures based
on calculated steady-state temperatures in the
enclosure. If the calculated partial water vapor
pressure is greater than saturation (estimated
dew point), it is assumed that water vapor will
condense in all parts of the roof assembly where
the partial pressure is higher than the saturation
pressure. The condensation rate is estimated
based on the water vapor permeability of
the assembly.4
The applicability of this method is severely
limited by its assumptions about steady-state heat
conduction and vapor diffusion. In reality, boundary
conditions of temperature, humidity, wind, and
radiation are in constant flux. Furthermore, water
vapor permeances and thermal conductivity may
vary with the relative humidity, temperature,
and moisture content of the roof components.
Nevertheless, many designers rely on the dew
point method to select vapor retarders. Typically,
a Class I vapor retarder with less than 0.1 perm is
used in all metal roof decks.
TRANSIENT MODELING WITH
HYGROTHERMAL SIMULATION
MODEL
The features of a complete moisture analysis
model include transient heat, air, and moisture
transport formulations. Dynamic indoor and
outdoor conditions as established by ANSI/
ASHRAE Standard 160,5 should be used for
simulation. Extensive data on material properties
are available in published literature.6
In this study, simulations were conducted on
one of the types of roof decks recommended by
the ASA/ANSI standard,2 a composite roof deck
assembly for high-humidity environments, to
assess the performance of the selected type of
roof deck in cold climates and the role of vapor
retarders in that performance. The simulations
were modeled with WUFI (WUFI is an acronym
for Wärme und Feuchte Instationär [heat and
moisture transiency]).7
This hygrothermal simulation model has
been developed over the past three decades by
Fraunhofer Institute for Building Physics and Oak
Ridge National Laboratory to perform transient
hygrothermal calculations to evaluate the
long-term thermal and moisture performances of
building enclosures, including roofs. The model
has been validated repeatedly over the past
two decades and has an extensive database of
material properties and exterior climates from all
US climate zones.8,9
Figure 2 presents the schematic of the
roof assemblies studied for transient heat
and moisture flow. The basic assembly (from
inside to outside) consisted of a wood-fibercement
composite, extruded polystyrene
(XPS) insulation, oriented strand board (OSB),
and an ethylene propylene diene monomer
(EPDM) roofing membrane. A vapor retarder
(0.032 perm) was installed at two different
locations, one on the warm side (Fig. 2b), and the
other after the OSB (Fig. 2c).
A retarder with a very low perm-rating
(<0.1 perm) was chosen to overstate the role of
the vapor retarder. (The general understanding
is that the lower the perm rating of the retarder,
the less chance there is of condensation in the
assembly.) The wood-fiber composite inside
the pool space provided acoustical absorption,
which is one of the requirements for good indoor
environmental quality (IEQ) in the space. Table 1
lists the primary transport properties of each
of the components. The rest of the properties—
moisture- and temperature-dependent transport
properties and the equilibrium moisture content
for each element—were obtained from the
database provided in WUFI.
Fargo, North Dakota, in Climate Zone 7,
was chosen for the location to simulate the roof
performance. Table 2 presents the boundary
conditions and the main surface transfer conditions.
The hygrothermal simulation software
includes the actual hourly averages of outside
temperature, humidity, rain, wind, and the solar
(short- and long-wave) radiation used in the
simulation. To simulate a high moisture load in
the natatorium, the inside conditions ranged
from 40% to 65% relative humidity and 68°F
to 80°F (20°C to 31.1°C).5 For all components,
the initial moisture content was assumed to be
equilibrated at 50% relative humidity.
The roof deck assembly had a total
R-value of 37 ft2 · h ft2 °F (6.52 m2K/W). To
emulate realistic conditions, an air leak at the
rate of 0.99 ft3/h ft2 (0.084 L/s·m2 @ 50 Pa)
was included in the model.5 The air leak
was introduced at the interface of OSB and
insulation, and it was set to start from the wood
composite (inside) up to the EPDM layer for
all the three cases. The stack height was set
Figure 1. A typical metal roof deck assembly.
March 2024 IIBEC Interface • 19
very high.11 The simulation was carried out
from 2000 to 2020.
SIMULATION RESULTS AND
DISCUSSION
Moisture Accumulation in OSB
The primary concern regarding the roof’s long-term
performance would be moisture condensation and
accumulation in the OSB over its service life. Decay
and rot in wood occur when its moisture content
exceeds 30%.3 However, to provide an adequate
Table 1. Thermodynamic and transport properties of roof components
Properties EPDM roofing membrane
(0.06 in.)
OSB
(0.5 in.)
XPS foam
(6 in.)
Wood-fiber board
(1 in.)
Vapor retarder
(0.25 in.)
Bulk density, lb/ft3
(kg/m3)
53.1
(851)
34.5
(553)
1.79
(28.7)
18.7
(300)
54.6
(875)
Porosity 0.001 0.64 0.99 0.8 0.001
Specific heat capacity, BTU/lb°F
(kJ/kg·K)
0.45
(1.88)
0.33
(1.38)
0.35
(1.46)
0.334
(1.4)
0.38
(1.6)
Thermal conductivity, BTU/h ft°F
(W/mK)
0.12
(0.66)
0.07
(0.12)
0.014
(0.024)
0.29
(0.5)
0.87
(1.5)
Permeability, perm-in. 0.0014 0.96 1.3 10.3 0.008
Note: EPDM = ethylene propylene diene monomer; OSB = oriented strand board; XPS = extruded polystyrene. 1 in. = 25.4 mm.
Figure 2. Roof deck assemblies: (a) without vapor retarder; (b) with vapor retarder after XPS foam installation; (c) with vapor retarder after the OSB
installation. The exterior surface is on the left side in all cases. Note: EPDM = ethylene propylene diene monomer; OSB = oriented strand board;
XPS = extruded polystyrene. 1″ = 1 in = 25.4 mm.
to 23 ft (7 m), and there was no mechanical
ventilation overpressure. Air leaks (typically
due to static pressure difference) can result in
significant moisture transport from the pool
space to the outside.
A no-coating option for permeance was
selected on the left side, and a permeance of
10 perm was selected for the right side. The
short-wave radiation absorptivity coefficient
is an important variable when determining
the outside surface temperature.10 The
roof deck had a dark roofing membrane
(short-wave radiation absorptivity = 0.8)
to absorb incident solar radiation, which
would significantly elevate the daytime (and
summer) surface temperatures since the
pool was located in Climate Zone 7. Typically,
white roofs are used in Climate Zones 1
through 3 to save energy (by lowering surface
temperatures). If a white roofing membrane
is used in a colder climate, the likelihood of
exceeding a moisture content of 20% will be
20 • IIBEC Interface March 2024
margin of safety, it is preferable to not exceed more
than 20% moisture saturation.12 Figure 3 presents
the simulation results for moisture content in the
OSB for the three cases.
As seen in Fig. 3a, in the roof assembly
without a vapor retarder, the average moisture
content in the OSB reached a dynamic steady
state in 5 years, and the maximum accumulation
of water in the OSB did not exceed 11.6%.
Therefore, there would be no risk of rot or decay
since the percentage of moisture was well
under 20%.
With the use of a vapor retarder on the
warm side of the roof assembly (after the
installation of foam), the maximum moisture
content in the OSB was reduced to 11.3%
(Fig. 3b). With the vapor retarder installed
after the OSB, the maximum moisture
content was around 10.8% (Fig. 3c). The vapor
retarder primarily minimized the magnitude
of the swing in moisture accumulation
between winter and summer months without
significantly affecting the average moisture
accumulation over the 20-year period. These
Table 2. Primary boundary and surface conditions
Conditions
Exterior climate (left side) Fargo, ND; ASHRAE Year 1
Roof orientation/inclination North/2 deg.
Heat resistance, h ft2 °F/BTU (m2K/W) 0.3 (0.053)
Short-wave radiation absorptivity 0.8
Long-wave radiation absorptivity 0.9
Interior (right side) EN 15026, high moisture load
Heat resistance, h ft2 °F/BTU (m2K/W) 0.7 (0.123)
results show that using a vapor retarder in
the roofing assembly offered no significant
advantage, thus contradicting popular
assumptions.
The simulations indicate that, even in
Climate Zone 7, a roof assembly without the
retarder could perform just as well as the one
with the vapor retarder so long as there were
a sufficient number of summer days with
warm temperatures (where the temperature
gradients cause the vapor drive from outside
to the inside). In these conditions, the moisture
content would reach a dynamic steady state,
with the accumulation of moisture during the
winter months being equal to the “drying”
of the OSB during the summer months. This
concept of a “self-drying roof” has been
recognized by previous investigators.12,13
Notably, the simulation shows that the vapor
retarder in the roof assemblies had minimal
effect on moisture accumulation. Although
the simulation illustrated the performance of
a wooden composite deck, the conclusions on
the role of the vapor retarder also hold true for a
metal deck.
Moisture Profiles in the Assembly
The development of moisture profiles across
the roof assembly provides insight into the
12.0%
11.5%
11.0%
10.2%
10.0%
9.5%
9.0%
8.5%
0 5 10 15 20 25
Time (years)
Moisture Content (%) in OSB Over 20 Years
(a) Without retarder (b) retarder after foam (c) retarder after OSB
Figure 3. Moisture content in the oriented strand board, 2000–2020.
March 2024 IIBEC Interface • 21
Figure 4. Moisture distribution profiles in the roof assemblies, 2000–2020: (a) without vapor retarder; (b) with vapor retarder after the installation of XPS
foam; and (c) with vapor retarder after installation of the OSB. The green regions represent the range of the relative humidity profiles in each component of
the roof deck. These profiles are calculated based on partial vapor pressure corresponding to the equilibrium moisture content in each component at a given
instant. The green lines represent the instantaneous relative humidity profiles. The blue lines represent the moisture content in each component (primary
axis). Note: EPDM = ethylene propylene diene monomer; OSB = oriented strand board; XPS = extruded polystyrene. 1 kg/m3 = 0.062 lb/ft3.
22 • IIBEC Interface March 2024
dynamics of moisture migration and buildup
inside the roof during the 20-year study period.
Figure 4 shows the distribution of the moisture
profiles through each layer during the 20 years
of simulated life. It is important to note that
when the roof deck assembly did not have a
vapor retarder (Fig. 4a), the maximum relative
humidity reached in the assembly did not exceed
about 80%.
However, when the roof deck assembly had
a vapor retarder (Fig. 4b and 4c), the maximum
relative humidity at the foam-retarder
interphase reached about 100%, indicating
that moisture condensation was quite likely
to occur at that location. While the vapor
retarder lowered the moisture flux from inside
to the outside during winter, its presence
inside the assembly retarded the downward
(from outside to inside) moisture flux through
the foam during the drying cycle (summer).
With the occurrence of condensation at the
foam-retarder interface, there would be a risk of
mold or mildew buildup in the assembly.
In the case of a metal roof deck, with the
retarder on the warm side, condensation
would be associated with an increased risk
of corrosion of the metal as the condensed
water could seep through pathways created
by screws and other anchoring hardware.
Therefore, in general, using a vapor retarder
would pose a higher risk and would not
provide the intended safety of preventing
condensation in the roof assembly in cold
climates. These conclusions, which contradict
common understanding and practice, can be
generalized for natatorium roofs in Climate
Zones 1 through 7.
In general, in addition to the roof deck
attributes mentioned in this article, a good
basis for natatorium roof design should include
the following: (a) a continuous air barrier,
(b) appropriate HVAC design and operation,3
and (c) minimization of stack pressure by
fan control.
CONCLUSION
The steady-state dew point analysis model
used by many practitioners of roof deck design
does not accurately characterize moisture
transport in roof decks. As a result, its use
can lead to suboptimal design. Hygrothermal
modeling can accurately predict roof deck
performance. Simulation results of a roof deck
assembly in a natatorium in Climate Zone
7 derived from a widely used hygrothermal
simulation model show that—contrary to the
assumptions behind common practice—a
vapor retarder does little to affect moisture
transport and condensation in a roof deck
assembly in the long run.
In Climate Zone 7, despite a
high-moisture load in the natatorium,
the roof deck assembly with or without
a vapor retarder can be self-drying,
with the moisture content in the OSB
maintained below the “at-risk” level of
20%. Furthermore, this study shows
that a vapor retarder can increase the
risk of condensation by hindering the
downward drying (outside to inside)
and thus increase the risk of corrosion
in a metal roof deck. Thus, contrary to
common belief, a vapor retarder provides
a false sense of safety (with regard to
moisture transport and condensation) as
it does not prevent or reduce the risk of
moisture condensation and buildup in
roof decks in cold climate zones.
REFERENCES
1. Lochner, G., and L. Wasner. 2017. “Ventilation
Requirements for Indoor Pools.” ASHRAE
Journal. 59 (7): 16–24.
2. Acoustical Society of America (ASA). 2019.
Acoustical Performance Criteria, Design
Requirements, and Guidelines for Schools,
Part 4: Acoustic Standards for Physical
Education Teaching Environments. ANSI/ASA
S12.60-2019/Part4. Melville, NY: ASA.
3. ASHRAE. 2021. Chapter 25: “Heat, Air, and
Moisture Control in Building Assemblies.” In
ASHRAE Handbook—Fundamentals. Peach Tree
Corners, GA: ASHRAE.
4. TenWolde, A. 1994. Chapter 11: “Design Tools.”
In Moisture Control in Buildings: The Key Factor
in Mold Prevention. West Conshohocken, PA:
ASTM International.
5. ASHRAE. 2021 Criteria for Moisture-Control
Design Analysis in Buildings. ANSI/ASHRAE
Standard 160-2021. Peach Tree Corners, GA:
ASHRAE.
6. Kumaran, M. K. 2006. “A Thermal and Moisture
Transport Database for Common Building
and Insulating Materials (RP-1018).” ASHRAE
Transactions. 112 (2): 485–497.
7. Fraunhofer Institute of Building Physics
(FIBP). WUFI Pro Version 6.5. 2020. Munich,
Germany: FIBP.
8. Desjarlais, A., H. H. Pierce, and S. Pallin.
2017. “Using Hygrothermal Modeling to
Resolve Practical Low-Slope Roofing Issues.”
In Advances in Hygrothermal Performance of
Building Envelopes: Materials, Systems and
Simulations, ASTM STP1599, edited by P.
Mukhopadhyaya and D. Fisler, 291–302. West
Conshohocken, PA: ASTM International.
GOURISH
SIRDESHPANDE, PHD
Please address reader comments to
chamaker@iibec.org, including
“Letter to Editor” in the subject line, or
IIBEC, IIBEC Interface Journal,
434 Fayetteville St., Suite 2400, Raleigh,
NC 27601.
9. Mundt-Peterson, S. O., and L. Harderup. 2013. “Validation
of a One-Dimensional Transient Heat and Moisture
Calculation Tool under Real Conditions.” In Proceedings
of the Thermal Performance of the Exterior Envelopes of
Whole Buildings XII International Conference, Oak Ridge
National Laboratory, Oak Ridge, TN. Peach Tree Corners,
GA: ASHRAE.
10. Bludau, C., D. Zirkelbach, and M. H. Kunzel. 2009.
“Condensation Problems in Cool Roofs. Interface.
27 (7): 11–16.
11. Kehrer, M., and S. Pallin. 2013. “Condensation
Risk of Mechanically Attached Roof Systems in
Cold Climate Zones.” Presented at the 28th RCI
International Convention and Trade Show, Orlando, FL,
March 14–19, 2013.
12. Desjarlais, A. 1995. “Self-Drying Roofs: What?! No
Dripping?!” In Proceedings of the Thermal Performance of
the Exterior Envelopes of Buildings, VI Conference, 1995,
Clearwater, Florida, 763–773.
13. Bludau, C., H. M. Kunzel, and D. Zirkelbach. 2010.
“Hygrothermal Performance of Flat Roofs with
Construction Moisture.” In Proceedings, Buildings XI,
Oak Ridge National Laboratory, Oak Ridge, TN. Peach Tree
Corners, GA: ASHRAE.
ABOUT THE AUTHOR
Gourish Sirdeshpande,
PhD, is a senior principal
scientist at STR
Resources in Lancaster,
Pennsylvania. Previously,
he was a senior principal
scientist in research
and development
at Armstrong World
Industries Inc., with over
35 years of experience.
His interests are in heat
and mass transfer modeling, composite materials,
building energy, and indoor environment quality.
He is the vice-chair of ASTM D22.05 on Indoor
Air Quality and a member of the US delegation
for ISO TC146-SC6 on Indoor Air Quality. He is a
member of ASHRAE, ISIAQ, ASTM, and AIChE. He
is also a member of ASTM D08.
March 2024 IIBEC Interface • 23