Skip to main content Skip to footer

Concrete Deck Moisture Issues: Causes and Preventative Measures

May 15, 2016

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 S c hwe t z e t a l . • 2 1 7
Concrete Deck Moisture Issues:
Causes and Preventative Measures
Stephen Condren, PE
Simpson, Gumpertz & Heger Inc.
41 Seyon Street, Suite 500, Waltham, Massachusetts 02453
Phone: 781-907-9000 • Fax: 781-907-9009 • E-mail:
Joe Schwetz and Stan Graveline
Sika Corporation
100 Dan Road, Canton, Massachusetts 02021
Phone: 781-332-3224 • Fax: 781-828-5365 • E-mail:
The frequency of moisture-related claims in concrete roof decks has been increasing
rapidly over the past few years. The issues are typically encountered on roofs that have not
exhibited any leakage and generally appear to have been properly installed, yet when cut
open, are found to be damp to wet across the entire roof surface. Reasons for these issues
include changes in system buildups, year-round construction in all climates, and accelerated
construction schedules. Exacerbating the problem is the increased usage of lightweight
structural concrete. This presentation will review the key issues involved and provide practical
recommendations for preventing problems, as well as potential remedial approaches in
situations where problems have occurred.
Stephen Condren, PE — Simpson, Gumpertz & Heger Inc.
Stephen Condren, a senior project manager, joined Simpson Gumpertz & Heger in
1980. He specializes in building envelopes, including roofing of all types, masonry, flashings,
waterproofing, and thermal and moisture analyses. Condren has designed new and remedial
roofing, waterproofing, masonry, sealant systems, and curtain walls and is involved in
bid and construction-phase services for new and remedial construction. He has authored
articles on roofing and is involved with many construction industry organizations, including
the American Society of Civil Engineers and ASTM International.
Joe Schwetz — Sika Corporation
Joe Schwetz is the Director of Technical Services for Sika Roofing. He has a degree
in architectural engineering from SUNY and has worked in the roofing industry for over
30 years in various research and development, technical, and managerial capacities. He
is active in various technical standards and code development bodies, including SPRI,
SIGDERS, and ASTM, where he cochairs ASTM Subcommittee D08.18. Schwetz received the
ASTM Award of Merit in 2015.
Non-presenting coauthor
Stan Graveline — Sika Corporation
2 1 8 • S c hwe t z 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
The frequency of moisture-related
claims involving concrete roof decks has
been increasing rapidly over the past few
years. Problems have been reported under
all types of roof systems in most climates.
The issues are typically encountered on
roofs that have not exhibited any leakage
and that generally appear to have been
properly installed, yet when cut open, are
found to be damp to wet across the entire
roof surface. Legal experts specializing in
the construction industry expect this to be
the most frequent source of roofing-related
litigation in the coming years.
There are numerous reasons for
these issues, including changes in system
buildups, year-round construction in
all climates, and accelerated construction
schedules. Exacerbating the problem is
the increased use of lightweight structural
concrete (LWSC).
Much has been written on the topic,
and various trade groups have weighed
in on the matter. Most of the published
literature highlights potential causes and
recommends caution installing roofs on
concrete decks, but generally, they provide
little to no specific advice on how
to avoid problems. This paper will review
the key issues involved, provide practical
recommendations for preventing problems
from occurring, and pose potential remedial
approaches in situations where problems
have occurred.
The low-slope commercial roofing industry
is increasingly faced with moisture-related
claims over concrete decks. Roof assemblies
over concrete decks are typically no
more prone to leakage than similar systems
over other types of decks such as
steel or wood. There is, however, a general
consensus that problems resulting from
moisture migration from the deck into the
roof assembly are on the upswing. We have
been installing roofs over concrete decks for
many, many decades. Concrete is still basically
a mixture of Portland cement, aggregate,
and water. What, if anything, has changed?
Our quest for ever-more-aggressive construction
schedules may be a contributing
factor. Regardless of the time of the year
or a building’s location, builders are striving
to get structures “closed-in” as fast as
possible, often squeezing the time available
between when the concrete deck is poured
and when the roof is installed. The initiation
of interior work such as pouring floor slabs,
drywall installation, and application of interior
finishes, is constantly compressed, generating
large quantities of moisture that are
available to migrate into the newly completed
roofing assembly. Although these are no
doubt contributing factors, there are other,
likely more important mechanisms at play.
Perhaps of greater significance are the
changes that have occurred over the past
years in the way we construct roofs over
concrete decks. In the past, forms upon
which the wet concrete was placed to create
the deck were removed after the concrete
had achieved the desired strength. Once the
roof assembly was installed and the interior
space was conditioned and occupied, any
residual moisture in the deck could migrate
into the building, allowing the deck to dry
gradually over time. Traditionally, insulation
was bonded to concrete substrates in
a full mopping of hot asphalt. Regardless of
whether plies of felt were included to intentionally
create a vapor retarder, the continuous
asphalt “adhesive” provided the roofing
assembly with a degree of protection from
upward moisture migration from the deck.
The ASHRAE Book of Fundamentals
lists asphalt, applied at 22 lbs. per 100 sq.
ft., with a permeance of 0.1 perm, which
makes the film a Class I vapor retarder. In
contrast, low-rise urethane insulation adhesives,
which have significantly increased
in usage, are relatively porous and, more
importantly, are applied in ribbons that
are discontinuous across the roof surface,
thereby providing no barrier to moisture
Wintertime vapor drive typically moves
the moisture in the concrete upward toward
the roof membrane, where it can condense.
Here, changes in materials and technology
may be making our roofs more vulnerable
to damage. In the past, relatively massive,
absorbent materials such as rigid fiberglass
and wood fiber insulation were used.
Although excessive amounts of moisture
would ultimately doom the assembly to
failure, these products could safely store a
certain amount of moisture without significant
loss of performance, particularly if they
could seasonally dry under favorable vapor
drive conditions.
Products commonly used in today’s roof
systems, such as organic-faced polyisocyanurate
insulation and gypsum-based
cover boards, suffer from significant loss of
cohesive strength in the facer and/or have
dimensional stability issues when wetted to
as little as 2 or 3% moisture by weight. The
durability performance (i.e., freeze/thaw
and traffic resistance of gypsum products)
drops significantly when approaching these
moisture levels. There appear to be numerous
buildings protected by roof assemblies
built over concrete decks, with forms left in
place, without vapor retarders, incorporating
the components listed above that have
performed without issue over many years.
The challenge is that in combination, these
elements allow for a very small, if any, margin
of error with regard to the amount of
excess moisture remaining in the concrete
at the time they are installed. Although the
potential for problems exists on all projects
that have concrete decks, the risk is higher
when using LWSC, which is made up of
very porous aggregate that retains much
more water.
Reroofing over concrete decks presents
its own unique challenges. Often, buildings
are only reroofed after significant leakage
has occurred over long periods of time,
allowing significant amounts of moisture to
accumulate in the roofing system and the
concrete deck. By necessity, as the existing
roof system is removed, the new system is
installed as quickly as practical thereafter
in order to protect the interior space. The
wet concrete deck may be exposed for a few
hours before being covered with the new
system. This may allow for surface drying,
but is not sufficient to adequately dry a wet
concrete deck.
Concrete Deck Moisture Issues:
Causes and Preventative Measures
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 S c hwe t z e t a l . • 2 1 9
Industry-wide, the issue is compounded
by the lack of a reliable test method for
determining the amount of moisture in
concrete roof decks and by our inability to
answer the fundamental question of how
dry a concrete deck needs to be before
installing the roofing assemblies.
The roofing industry has attempted to
notify and educate roofing professionals of
the potential issues involved with moisture
Various roofing industry trade organizations,
such as ARMA, the National
Roofing Contractors Association (NRCA),
the association representing the Single-
Ply Roofing Industry (SPRI), RCI, and the
Polyisocyanurate Insulation Manufacturers
Association (PIMA), have released position
statements regarding the issue of moisture
in concrete decks. Basically, these organizations
suggest the designer of record be
aware of the moisture issues with new concrete
and select components and systems
that can accommodate the moisture.
The Midwest Roofing Contractor
Association (MRCA) T & R Advisory Bulletin
1/2011, “Structural Lightweight Concrete
Roof Decks,” suggests its contractors provide
the designer of record and general
contractor with a copy of the MRCA
bulletin that discusses issues on moisture
in concrete, along with having them
review the American Concrete Institute
(ACI) document ACI 214R-03 (“Guide for
Structural Lightweight Aggregate Concrete”)
and the Portland Cement Association (PCA)
Engineering Bulletin 119 (“Concrete Floors
and Moisture – 2008”).
The NRCA, in its August 2013 “Industry
Issue Update,” suggests that the project
structural engineer, general contractor,
concrete supplier, and concrete placement
contractor recommend when the new concrete
is ready to be covered. Their reasoning
is that this group will have the best knowledge
of the concrete’s cure and moisturerelease
rates. The NRCA also recommends
that LWSC not be used for roof decks or
toppings for roof decks. They suggest that
if LWSC is used, that designers specify the
concrete’s drying parameters, using ASTM
F2170 to determine the relative humidity
(RH). They suggest decks receiving roofing
systems should have a maximum RH of
75% until there is an industry consensus as
to what the value should be.
In reroofing situations where the existing
deck is LWSC or the existing deck is
known to be wet, the NRCA recommends
two alternative roof designs. Either provide
for above-deck venting with a venting
base sheet with a loose-laid ballasted roof
system and perimeter venting, which may
allow release of the moisture; or seal the
moisture into the deck by using an adhered
vapor retarder, followed by an adhered roof
FM Global initially did not recognize
LWSC as an approved substrate, as it did
not meet their definition of structural concrete.
In the current version of 4470 (dated
2012), FM defines structural concrete as
having a density of approximately 150 lbs./
ft.3; LSSC has a density of 90 to 130 lbs./
ft.3. FM Global will be addressing the use of
LWSC in its revised Loss Prevention Data
Sheet (LPDS) 1-29. In the revision, FM will
allow the use of the LWSC if the structure
cannot support normal-weight concrete.
They will suggest that removable forms be
used whenever possible. If the forms cannot
be removed, they will recommend the lowest
water/cement ratio possible.
If LWSC is used, FM Global will require
a test to be performed to ensure that the
moisture migration will be reduced to the
point where any damage to the above-deck
components is minimal. Unfortunately, FM
Global does not offer a test method to determine
the moisture in concrete.
They will also require above-deck components
to be resistant to moisture or the
installion of a vapor retarder that will limit
moisture migration to the components.
The Steel Deck Institute (SDI) issued a
Position Statement in May 2012 to address
inquiries regarding the use of vented steel
form decks. They note that vented steel form
decks traditionally have been used to drain
“excess” mix water from the cementitious
slurry of lightweight insulating concrete
(LWIC). The venting allows for the reduction
of vapor pressure once the roof cover is
installed over the LWIC. They caution that
LWIC is quite different from LWSC. When
designing for LWSC, SDI notes:
While it is known that the inclusion
of slots has little effect on
the strength of the steel deck, it is
unknown what the effect of draining
mix water through the bottom
of the deck has on the properties of
the cured concrete and the bond of
the concrete to the deck. Specifiers
should proceed with caution when
requiring slots in this application.
The steel deck acts as a vapor
retarder, preventing diffusion of
water vapor out of the bottom of the
slab. Some publications note that
the amount of diffusion is directly
proportional to the open area in
the vapor retarder (Ficks Law). For
example, providing a hypothetical
1.5% open area will increase the
diffusion of water vapor by 1.5%, an
inconsequential amount. This has
been experimentally verified through
testing sponsored by the Expanded
Shale Clay and Slate Institute, which
has shown that the rate of concrete
drying is not increased by venting
the steel deck.
To summarize, the roofing industry has
generally taken the position that the building’s
designers must be aware of the issues
and take responsibility for the selection
and timing of roofing system placement
when there is a concern with moisture
in concrete. Some industry stakeholders
have taken the position that if at all possible,
LWSC should not be used for the
roof deck. The designer should contact all
parties involved with the roofing portion
of the project to establish requirements on
the schedule for covering the concrete. The
method used may be to allow for venting of
the moisture or encapsulating the moisture
by using a well-secured vapor retarder, both
of which are discussed below. Testing of the
concrete is suggested by some organizations,
specifically using ASTM F2170, even
though this test method is not designed for
an exposed concrete slab.
Concrete Composition
Concrete is a mixture of Portland
cement, aggregates, air voids, water, and
other additives. Concrete hardens by the
occurrence of a chemical process; a portion
of the mix water reacts with the cement and
any pozzolanic additives present to form the
hydrated binder of the hardened concrete.
Concrete does not harden by drying.
Concrete contains additional water
to provide the fluidity for placement and
finishing of the freshly mixed concrete.
Although chemical admixtures may reduce
2 2 0 • S c hwe t z 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
the amount of additional water required, all
freshly placed concrete will contain more
water than what is consumed in hydration
The mix water added to the concrete
mixture is expressed as the water-tocementitious
material ratio (w/cm). Most
concrete mixes are batched at 0.40 to 0.55
w/cm. The cement hydration process (most
of which occurs within the first month)
consumes about 0.25 w/cm. This leaves an
additional 0.15 to 0.30 w/cm, which will
remain in the concrete as liquid water in the
capillaries and pore spaces; this is sometimes
referred to as “free water.”
In addition, the aggregates absorb water:
0.1 to 2% for normal-weight aggregates, and
between 10 and 30% for lightweight aggregates.
This water will also be contained as
free water within the concrete after it has
been placed.
Differences Among Concretes
Concretes used in roof decks fall into
three categories: structural concrete (SC),
structural lightweight concrete (SLWC),
and cellular lightweight concrete (CLWC).
Structural and lightweight structural concrete
contain graded aggregates. Cellular
concrete is made from cement, water, and
a foaming agent and does not typically contain
any aggregates. Lightweight insulating
concrete (LWIC), containing a large quantity
of water, is also used in roof decks, but we
did not include the material in this study.
The aggregates in SC are graded, dense
stone with a density of 110 pounds per
cubic foot (pcf), resulting in concrete that
weighs 150 pcf. The aggregates in SLWC are
porous stone, either natural or man-made,
with a density of less than 70 pounds per
cubic foot (pcf), resulting in concrete that
weighs 90 to 115 pcf. Precast concrete can
be made from SC or SLWC.
Cellular concrete used in roofing cures
to produce a medium of closed-cell cement
foam with a dry density ranging from 26 to
32 pcf and compressive strength of 180 to
280 psi. Cellular concrete is not structural
and is typically used to provide slope to
structural roof decks to promote drainage.
Moisture Sources in Concrete Decks
New Construction
New concrete roof deck construction will
contain varying amounts of water, based on
the type of concrete and the aggregates used
in the mix. Most concrete roof decks constructed
today are placed over a steel deck
that serves as a form that remains in place.
The metal deck essentially prevents moisture
from evaporating from the underside
of the deck. Although some metal decks are
slotted along the ribs (flutes) on the corrugated
forms, the area of these perforations
is only 0.25% to 1.5% of the metal deck
surface, allowing only a minimal amount
of moisture evaporation through the metal
deck. Precast concrete will not have forms.
Table 1 presents a summary of conditions
for typical concretes that are used for
roof decks. The numbers are based on a
6-in. thickness for the roof deck. The last
line includes how much free water remains
in the concrete after the concrete has cured
for 30 days by hydration of the cementitious
A new 6-in.-thick concrete deck will
contain 0.9 to 2.6 quarts of the original mix
water per square foot at one month of age.
This assumes that no additional water is
provided during placement and finishing
and the curing process. This is the typical
time when the construction schedule will
expect the application of the roofing system.
In reality, additional water is often added
to the fresh concrete to assist with pumping
and placement. Finished concrete needs to
be wet-cured by sealing the surface to prevent
evaporation or by applying water onto
the surface until the concrete has developed
sufficient strength to prevent surface cracks
due to surface drying. It is also unreasonable
to assume that the exposed deck will
not be exposed to precipitation between its
construction and full cure. Placement, curing,
and precipitation will add an undetermined
amount of water to the deck.
When the roofing system is being
installed, the new roof deck will contain a
minimum of 0.9 to 2.6 quarts of free water
(for a 6-in.-thick deck) that will be available
to migrate into a roofing system.
The conditions described above are what
can be expected for new construction. The
assumptions (i.e., no water added during
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 S c hwe t z e t a l . • 2 2 1
Normal-Weight Lightweight Lightweight
Concrete Structural Structural
With 0.5% Concrete Concrete Cellular
Aggregate With 10% With 20% Concrete
Absorption Aggregate Aggregate
Absorption Absorption
Total cementitious material (lb./cubic yard [cy]) 650 710 710 779
w/cm ratio 0.39 0.43 0.43 0.50
Added batch water (lb./cy) 242 308 308 384
1% fine aggregate absorption 11 13 13 n/a
(lbs. water/cy)
Coarse aggregate absorption (lb. water/cy) 10 73 145 n/a
Total water in batch (lb./cy) 263 394 466 384
Water consumed in hydration reactions (lb./cy) 163 178 178 195
Water remaining after hydration (lb./cy) 100 216 288 189
Water remaining in 6-in. concrete deck (qt./sq. ft.) 0.9 1.9 2.6 1.7
Table 1 – Concrete components and water remaining in example concrete mixes.
placement, curing, or from precipitation)
may not be realistic, but it is prudent to
recognize that a significant amount of water
remains entrapped in a concrete deck that
is about to receive a new roofing system.
However, reroofing an existing building
with a concrete deck presents a dilemma.
The amount of water contained within the
deck is not known, nor is there a reasonably
effective test to measure how much
water is present in the deck. If the building
had a history of leakage, it is reasonable to
assume the roof deck contains entrapped
water. However, the absence of leakage is
not an assurance that the deck is dry.
Time Required to Dry Concrete
The interior voids of the freshly placed
concrete (i.e., the pores, capillaries, and
aggregate porosity) are initially saturated
with water. When drying begins, the concrete
loses water by evaporation from its
surface. The surface of the concrete may
appear to be dry, but it is only an illusion.
As the surface loses water, diffusion begins,
drawing water and alkaline salts from the
wet interior to the surface. When the roofing
materials are installed, the water evaporating
from the top surface of the concrete
becomes sealed within the roofing assembly.
Simpson Gumpertz & Heger Inc. (SGH)
used WUFI, a one-dimensional heat and
moisture transport software program, to estimate
the drying of concrete decks cast over
vented metal decks for one month prior to
the application of the roofing system (Table
1). The drying time for cellular concrete is
two weeks. SGH selected three climate zones
across the United States as benchmarks for
its calculations, and we have assumed that
there will be no added water from a curing
process or from rain during this initial drying
period. To estimate the best-case conditions
for drying, SGH based the drying simulation
on local climate data beginning July 1.
SGH found that the amount of water
evaporating from the exposed concrete roof
decks during the month before applying
the roofing system, with the exception of
cellular concrete, is only a small portion of
the available excess water remaining in the
concrete roof decks. The water that remains
in the deck after this initial period of evaporation
is shown in Table 2.
The results show that there is no location
in the United States that will dry all of
the excess water contained in the concrete
within 30 days after placing.
In Figure 1, we show how much drying
can occur if the evaporation time is
extended to one year. The graphs are based
on the assumption that the deck can be left
exposed for one year without precipitation,
an unlikely condition without providing
temporary protection. Both the normalweight
and lightweight concrete decks will
still contain free water if allowed to dry for
one year in all regions.
In the real world, water from curing and
precipitation must be taken into account,
which means the free-water content within
the concrete can be expected to remain significantly
higher for prolonged periods.
To evaluate the effect that the free water
in the concrete deck has on the roofing
materials, we need to consider the vapor
drive of the moisture into the roofing materials
from the concrete in a roofing assembly
that does not contain a vapor retarder. We
use RH to evaluate the vapor drive. When
the concrete begins drying, the RH within
2 2 2 • S c hwe t z 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
Normal-Weight Lightweight Structural Cellular Concrete
Concrete (qt./sq. ft.) Concrete (20% Aggregate @ 14 days (qt./sq. ft.)
(% of evaporable water) Absorption) (qt./sq. ft.) (% of evaporable water)
(% of evaporable water)
Cold-moist climate 0.7 2.2 0.5
International Falls, MN (80%) (87%) (33%)
Warm-humid climate 0.7 2.2 0.6
Miami, FL (80%) (187%) (35%)
Warm-dry climate 0.6 2.1 0.3
Phoenix, AZ (70%) (83%) (15%)
Table 2 – Water remaining in concrete roof decks exposed for one month.
Figure 1 – Water loss by evaporation from vented concrete decks exposed a year.
Cellular concrete must be roofed-over between 3 and 14 days for sites 1, 2, and 3
(value in table reported at 14 days). Table assumes a July 1 summer casting date
and no rain during the entire year.
the concrete is 100%. As drying continues,
the RH becomes less. After the roofing materials
are installed, the moisture in the concrete
raises the RH of the air trapped within
the roofing system above the concrete deck
and begins to cause problems. Typical
organic roofing components are susceptible
to mold and other decay organisms when
the RH is above 80% for prolonged periods
of time. RH is also important in determining
when condensation may occur within
the roofing system, and when condensed
water may begin deteriorating the moisturesensitive
roofing components.
To demonstrate the moisture drive from
the concrete to the roofing materials, we
calculated the RH within the concrete at
one month of age, at the time the roofing
materials are installed. We found that the
RH within the concrete is above 95% in
all locations (Table 3), indicating that the
moisture-sensitive roofing materials will be
exposed to deleterious amounts of moisture.
Since most roof decks receive their roofing
system within a month of placement,
when the concrete is still wet, an alternative
approach is needed to address the free
water (water that remains even if the deck
surface appears to be dry).
How Wet Is the Concrete ?
Current Testing Procedures
With the changes in construction methods
and schedules described previously,
the question becomes how to determine if
and when the deck is dry enough to apply
the roofing system. There are various test
methods, both qualitative and quantitative,
as well as standards and practices that have
been used within the construction industry
to evaluate the moisture content in concrete.
Following are a few of the most common
methods, most coming from the flooring
industry. We will discuss some concerns
when using the methods for the application
of the roofing system.
Qualitative Methods
Qualitative methods include:
ASTM D4263, Standard Test Method
for Indicating Moisture in Concrete
by the Plastic Sheet Method
This method uses a transparent
polyethylene sheet approximately 4
mils thick, secured to the concrete
substrate with a 2-in.-wide adhesive
tape. The secure plastic sheet
remains in place for a minimum of
16 hours. After the allotted time has
elapsed, the film is removed and the
surfaces are visually inspected for
the presence of moisture—typically
a darkened concrete surface or condensation
on the underside side of
the plastic sheet.
There is a note in Section 4.0 of
the standard that cautions that the
test should be conducted when the
surface temperature and ambient
conditions are within the parameters
for application of the coating system,
and to avoid direct sunlight, direct
heat, or damage to the plastic sheet.
These conditions are not feasible on
a roof deck exposed to the elements.
The Application of Hot Bitumen
Directly to the Concrete
While this is not a standard test
method, this procedure has been
used by the roofing industry for
many decades to determine if components
may be hot-mopped to the
concrete. The concept is if there is
foaming of the hot asphalt or the
asphalt is easily removed after it
has cooled, the concrete is too wet
to install the roof system. Not being
a standard test method, there is no
direction as to any precautions that
should be taken for this procedure.
In addition, conditions may change
as the work moves into a different
area of the deck.
Comments Regarding the
Qualitative Methods
Both of these methods have limitations,
including locations of the test areas, such
as in direct sun or shaded areas, as well as
primarily evaluating only the moisture at
the upper thickness of the roof deck.
The plastic film test is no longer considered
a valid test, especially with LWSC. The
NRCA stated in an Industry Issue Update
titled “Moisture in Lightweight Structural
Concrete Roof Decks” that this method is
unreliable. The NRCA notes the difficulty in
achieving an airtight seal between the film
and the concrete deck. It also states that
if the temperatures on both the top and
bottom of the concrete slab are not nearly
identical, the pressure difference can result
in a false “dry” result. Additionally, Mark
Graham of the NRCA (in a presentation,
“Problems and Risks Posed by Concrete
Roof Decks,” at the September 2013 NRLRC
Conference) noted that the “historical guidelines”—
including the plastic film test and
the application of hot bitumen—are not
appropriate for current generations of concrete
Quantitative Methods
Quantitative methods include:
ASTM F1869, Standard Test Measuring
Moisture Vapor Emission
Rate of Concrete Subfloor Using
Anhydrous Calcium Chloride
This test method measures the rate
of moisture vapor emitted from concrete,
in pounds of moisture over a
1,000-sq.-ft. area during a 24-hour
period. The requirement for this
test is for it to be conducted at
the same temperature and humidity
expected during normal use. If
this is not possible, then the test
conditions will be 75º ± 10ºF and
50 ± 10% relative humidity, for 48
hours prior to and during testing.
These conditions will be very difficult,
if not impossible, to achieve
when testing an exposed concrete
roof deck. The ASTM standard also
notes, “The results obtained reflect
the conditions of the concrete at the
surface at the time of the testing
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 S c hwe t z e t a l . • 2 2 3
Normal-weight Lightweight Concrete
Concrete (20% Aggregate Absorption)
(% RH) (% RH)
Phoenix, AZ 94 98
(hot and dry)
International Falls, MN 97 99
(cold and moist)
Miami, FL 97 99
(hot and humid)
Table 3 – RH of one-month-old concrete.
and may not indicate future conditions.”
The National Ready-Mixed
Concrete Association (NRMCA), in
its Concrete in Practice technical
bulletin CIP-28, comments on ASTM
F1869, “… [F1869] has some major
shortcomings: It determines only a
portion of the free moisture at a
shallow depth of concrete near the
surface of the slab.” While this test
method is relatively inexpensive, it is
not very accurate, is sensitive to air
temperature and humidity, and only
measures the effect of moisture from
minimum or shallow depths.
ASTM F2170, Standard Test
Method for Determining Relative
Humidity in Concrete Floor Slabs
Using In-Situ Probes
This test method requires cores for
placing the liner/sleeve into the central
region of the concrete deck for
determining the RH of the material
(Figures 2 and 3). It requires that the
holes reach thermal and moisture
equilibrium before starting the measurement
for RH. This time may be
anywhere from several hours to several
days, depending on temperature
differences between the probes and
the concrete in a stable interior environment.
When testing an exposed
roof deck, however, the changing
conditions on the topside of the
concrete roof deck make obtaining
accurate, reproducible readings difficult
to achieve.
ASTM F2659, Standard Guide
for Preliminary Evaluation of
Comparative Moisture Condition
of Concrete, Gypsum Cement, and
Other Floor Slabs and Screeds
Using Nondestructive Electronic
Moisture Meter
This guide notes in its scope,
“Results from this guide do not provide
vital information when evaluating
thick slabs…[of] lightweight
aggregate concrete.” The guide notes
that the depth of the signal penetration
will vary, depending on the
material and moisture content, and
they generally will read between 0.5
and 1.0 in. (Figure 4). This method
may provide a moisture level based
on the reading of the top 1-in. level,
but it will not measure moisture
conditions within the deeper regions
of the deck.
Core Sampling and Gravimetric
Moisture Content
This procedure is used within the
concrete industry to determine
moisture content in cured concrete.
However, it is not a consensus test
method, nor could we find an ASTM
test method for determining the
moisture content in concrete after
removing core samples. Ideally, the
core sample should be the full depth
2 2 4 • S c hwe t z 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 2 – Placement of RH
probes into concrete slab.
Figure 3 – Preparation
for use of RH probes.
Figure 4 – Example of the depth of the concrete slab that is evaluated.
(Figure 5), with the core diameter
being at least three times the size of
the largest aggregate (Figure 6). The
core sample should be dry-cut to
avoid including additional water to
the core. The core must be completely
wrapped with an impermeable
material—typically a metal foil—to
keep the moisture content stable
during transport and storage.
To begin the test, the initial weight
of the core is taken and then heated
at 220ºF until the loss in mass is
not more than 0.1% in a 24-hour
drying period. The calculated weight
loss is expressed as a percentage of
the dry weight. This test procedure
produces an accurate measurement
of the weight of the free water present.
While this is an accurate test
method, there are reasons why this
test is not feasible for evaluating
concrete roof decks:
1. Most concrete samples are wetcored;
due to the added
water, these cores cannot be
used in this test method.
2. Dry-coring of concrete generates
heat that will drive moisture
out of the sample, giving
false low readings.
3. While the lab results determining
free moisture of the samples
are accurate, because
of the condition discussed
above, the results will vary
from actual field conditions.
Test methods ASTM F1869 and F2170
were developed for the flooring industry,
where a controlled interior environment can
be maintained to minimize the changes in
temperature and humidity. Trying to use
these test methods in an exposed environment
at the roof deck will be problematic, as
both the temperature and humidity will be
constantly changing. There is little chance
of equilibrium being achieved, as required
by the procedures, when testing a deck
exposed to the weather. These tests may
provide inconsistent or inaccurate information.
Given that there are no reliable testing
options for determining the moisture in
exposed concrete, if a designer decides to
use one of these tests, he or she should be
aware of the possible incorrect data.
Test methods ASTM F1869 and F2659
are, at best,
shallow deck
test methods,
providing information
down to
a depth of one
inch. These two
methods will
not give any
indication of
any moisture
available in the
deeper regions of the concrete and may provide
erroneous information, allowing one to
assume the deck is “dry.”
Based on the review of the typical concrete
moisture test methods noted above,
there is not a reliable, accurate method
available for determining the moisture
conditions in concrete decks exposed to
the weather, and there is no agreed-upon
acceptable moisture value for roofing systems
as there is for the flooring industry.
Concrete Admixtures
Of the many admixtures used in concrete,
there are chemical additives available
that allow the reduction of water
content while still achieving or improving
on the required workability needed to place
and finish fresh concrete. Water-reducing
admixtures and mid-range and high-range
water-reducing admixtures (superplasticizers)
can typically be used to reduce the
design water content by about 20%. Many
applications take advantage of the fluidifying
characteristics of these products
to make the concrete mix more workable
for placement. Most concrete placed today
has some form of water-reducing admixture
incorporated in the mix, and therefore has
less water present and less porosity than
an untreated concrete. Concrete admixtures
can reduce but will not eliminate the water
needed to work the concrete.
Curing and Sealing Compounds
There is confusion at times to understand
the difference between curing compounds
and sealing compounds.
Curing compounds are used to slow or
reduce the evaporation of moisture from the
concrete to prevent cracks due to drying
shrinkage. A liquid-type, membrane-forming
curing compound typically consists of
waxes, chlorinated rubber, resins, or similar
materials applied to retard the evaporation
rate. These curing compounds are applied
with spray equipment immediately after the
finishing of the concrete, while the surface
is still damp. Some curing compounds
may affect the adhesion of products to the
concrete surface. Since curing compounds
reduce the rate of evaporation, they keep
the excess water entrapped for a longer
period of time.
Sealing compounds used for exterior
concrete is usually an optional procedure,
typically used to protect the concrete from
freeze/thaw, corrosion of reinforcing steel,
or acid attack by reducing the absorption
of liquids such as water. Sealers used on
interior concrete allow for easier cleaning,
reduce dust, and protect from the absorption
of spills. Sealers are applied to the
finished, hardened concrete after the cure
time—typically 28 days.
Sealers are typically a surface treatment,
with minimal penetration into the
concrete. Sealers may be acrylics, polyurethanes,
or epoxies. Penetrating-type sealers
include silanes and siloxanes, based on
silicone chemistry. While these sealers are
penetrating, they will allow moisture vapor
from the concrete to vent out.
In summary, curing compounds retard
the movement of water from the concrete,
while sealers protect the absorption of
liquids into the concrete. In both cases,
the films used may adversely affect adhesion
of the roofing components, which will
affect the uplift performance of the system.
Neither of them will prevent moisture migration
from the roof deck into the roof assembly
after it has been installed.
Vapor Retarders
Philosophies regarding the use of vapor
retarders have been somewhat fluid over
time, ranging from “when in doubt, use
one,” to “when in doubt, leave it out,” to
“when in doubt, think it out.” Typically,
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 S c hwe t z e t a l . • 2 2 5
Figure 5 – Coring into the
Figure 6 – Removed concrete
decisions as to whether to use a vapor
retarder have been based primarily on
temperatures (interior and exterior design
values) and the moisture occupancy (RH) of
the interior. With the exception of situations
where significant interior moisture may be
generated after the roof has been installed,
this approach may have been successful for
steel and wood decks.
Whether to include a vapor retarder—
particularly on a concrete deck—is one of
the most significant roofing system assembly
decisions that a design professional
must face today. A troubling habit in the
industry is to design the roofing system to
warranty coverage. Many elements of roof
design revolve around the designer’s and/or
the owner’s expectations regarding warranty
coverage (duration, wind speed, hail, etc.)
and the roof system manufacturer’s requirements
for issuing warranties meeting these
expectations. Most, if not all, roof system
suppliers disclaim any and all responsibility
for deciding whether to include a vapor
retarder in any project, and most go so far
as to explicitly exclude damage resulting
from condensation within the roof assembly
from warranty coverage.
The hygrothermal modeling (WUFI) that
we used to determine the drying rates for
the various types of concrete can also be
used to design the components of the roofing
system for various climatic regions.
We demonstrate herein that virtually all
concrete decks contain more moisture
than many modern roofing systems that
don’t include a vapor retarder can tolerate.
Providing a vapor retarder on top of the
concrete is the only practical solution available
to control the water entrapped in the
concrete deck and prevent it from entering
and condensing in the roofing assembly.
The type of vapor retarder must be selected
based on the relative permeability between
the membrane and the vapor retarder.
The WUFI model software is a powerful
tool, using the thermal and hygrothermal
properties of building components, combined
with regional weather records, which
can provide realistic predictions of the seasonal
performance of buildings. The model
can be extended to determine cumulative
effects as the building ages.
However, as with any computer tool, the
results are only as reliable as the data that
is provided. Proper application of the model
requires experience in the field of hygrothermics.
As a starting point, WUFI has a
database of hygrothermal material data for
many of the components used in building
construction and weather data derived from
ASHRAE records. The data can be further
refined if the specific material properties
are known.
WUFI computes the evolution of the
temperature and moisture conditions in
the building components, which allows the
user to observe predicted trends in overall
behavior of the system. Components can be
adjusted, and materials added, deleted, or
replaced, to observe their effect on the overall
performance. Provided the material properties
and weather data input are accurate,
WUFI is a tool that can be used to design for
the long-term moisture performance of roofing
and other building envelope systems.
Running the hygrothermal model will confirm
the performance of the selected vapor
retarder in a roofing assembly or guide the
designer toward a proper solution.
It is beyond the scope of this paper to
conduct an in-depth analysis of the various
types of vapor retarders available and
their relative merits. All standard technologies
(self-adhered, hot-applied, cold-applied,
torched-on, etc.), properly used in a given
assembly typically can be effective. The
ultimate solution will depend on a number
of variables, including ambient conditions
at the time of installation, the build-up
of and attachment methods to be used,
the components installed above the vapor
retarder, wind uplift requirements, insurance
or other approvals to be fulfilled, and
system manufacturers’ available options, to
name but a few.
The use of a vapor retarder is not to
be used as a shortcut around sound roofing
practices. A vapor retarder will only be
effective in preventing future moisture or
condensation-related damage if properly
designed and installed at the outset. The
surface of the concrete deck must be sufficiently
clean, dry, and properly primed
to achieve the required level of adhesion. If
adhered, the adhesive should be stable in a
moist environment so the attachment does
not degrade. It is strongly recommended
to conduct on-site adhesion tests prior to
the actual installation. The vapor retarder
must be detailed and installed such that
it is sealed at all parapets, curbs, and
other penetrations to prevent moisture from
“short-circuiting” it and migrating into the
roof system at these locations. Where the
vapor retarder is to also perform as an air
barrier, it must be tied into the wall air barrier
system to achieve continuity.
If the concrete is dry on the surface, in
all likelihood it still contains a significant
amount of moisture. If the forms are left
in place, the moisture is effectively trapped
between two highly impermeable surfaces
after the application of the vapor retarder,
2 2 6 • S c hwe t z 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 7 – Loose membrane/ insulation facer billowing under wind load.
which will not create any issues, provided
the interface between the vapor retarder
and the concrete remains at a temperature
well above the dew point. For example, a
problem could occur if the vapor retarder is
installed and left exposed in winter conditions
and the space below the deck is conditioned.
This could result in the moisture
migrating to the outer surface and condensing
at the deck level, resulting in a loss of
bond. Conversely, during hot summer conditions,
dark-colored vapor retarders can
heat up sufficiently to cause blisters to form
at any void between the vapor retarder and
the deck, leading to blister formation, which
if not addressed prior to installing the balance
of the system, could eventually impact
the finished system’s resistance to wind
uplift and other forces over time.
To achieve the best performance, the
vapor retarder should be insulated by installing
the insulation and membrane soon after
installing the vapor retarder. In instances
where the vapor retarder is to be used as a
temporary roof, pending completion of other
construction activities, careful consideration
should be given to all relevant factors and
impacts and addressed accordingly.
New Construction
A new office building was constructed
in New England. The building had a modest
footprint with a total roof surface of 10,000
sq. ft. on two levels.
The roof assembly consisted of a 14-in.-
thick concrete deck and tapered polyisocyanurate
insulation installed in two to four
layers. The layers of insulation were bonded
to the concrete deck and to each other with
low-rise urethane foam adhesive applied in
ribbons. A felt-backed thermoplastic membrane
was adhered to the insulation using
a solvent-based adhesive.
For approximately seven years after
the roof system was installed, it performed
problem-free, and no leaks had been reported.
In the fall of the seventh year, after a
storm with wind speeds up to 65 mph, the
membrane was observed to have become
unattached across practically the entire
surface of both levels of the roof (Figure 7).
The assembly is rated for 465 lbs./sq. ft.
of uplift pressure. A wind speed of 65 mph
should only have exerted an uplift force of
approximately 17 lbs./sq. ft. in the field of
the roof and approximately 36 lbs./sq. ft.
in the corners.
Test cuts revealed that the felt-backed
membrane was very well-bonded to the
underlying insulation, although the felt
backing was damp. Failure occurred within
the top facer of the top layer of insulation.
The facers were organic and were damp,
and had clearly been in that condition for
quite some time (Figures 8 and 9). They were
severely wrinkled in all locations observed,
and in some areas, exhibited mold growth.
The lack of any of record or evidence of
leakage, and the presence of moisture across
essentially the entire roof surface, clearly
pointed to condensation as the source of
water that allowed the failure. Based on
the occupancy of the building and the local
climate, it was concluded
that residual
moisture migrating
from the concrete
deck into the roofing
system was in
all likelihood the source of the problem.
The involved parties ultimately decided
to secure the membrane with fixation bars
and cover strips.
An office tower located in the Midwest
consists of three distinct, isolated roof levels.
The original roof construction consisted
of a structural concrete deck, topped with
a tapered LWSC screed, a single layer of
wood fiberboard insulation, and a built-up
membrane. After approximately 13 years,
the roof was removed and replaced with an
adhered single-ply system.
The original fiberboard roof insulation
was reportedly wet across large sections of
each roof level when it was removed. The
fiberboard and built-up roof were removed
down to the deck. Although much of the
original mopping asphalt came off with
the insulation, sporadic patches remained
stuck to the concrete at random locations.
The replacement system consisted of a 2-in.
layer of organic-faced polysisocyanurate
insulation and a ¼-in. layer of a glass-faced
gypsum protection board, both of which
were secured with low-rise polyurethane
foam adhesive. A thermoplastic membrane
was adhered to the cover board using a
solvent-based adhesive.
A storm during reroofing resulted in
localized wetting of the newly installed roof
assembly and the occupied space below it.
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 S c hwe t z e t a l . • 2 2 7
Figure 8 – The organic insulation facer,
which was damp, failed cohesively.
Figure 9 – Cohesive separation of
the damp organic facer. The deep
“notch” is a deeper-than-usual “knit”
line from the manufacturing process.
The affected area was reportedly replaced
with new, dry materials prior to the completion
of the new roof. After the roof was
completed, one small, isolated leak was
reported and repaired. For approximately
two years, the roof performed without any
reported issues. A tenant wanted to install
a rooftop terrace above their offices. During
the installation of structural elements to
support the terrace and associated planters,
etc., the surface of the cover board was
found to be damp. Subsequent investigations
revealed similar conditions across all
three roof levels.
The cover board was generally found
to be dry in a narrow band around all roof
perimeters and penetrations.
Although the moisture level in the gypsum
cover board was low, it was sufficient to
cause separation between the glass facer—
which was well bonded to the underside of
the membrane—and the gypsum core of the
cover board, leaving the thermoplastic sheet
essentially loose over most of the roof surfaces.
The roofs were temporarily ballasted
while various investigations were carried
out (Figure 10).
The owner hired a consultant who, after
analyzing the moisture content of the air
within the plenums on the underside of
the deck (Figure 11), concluded that condensation
due to air migration from within
the building was highly unlikely to be the
source of the moisture observed in the cover
boards. There was some debate as to whether
sufficient material had been replaced
after the storm damage during construction.
Even if this had in fact been the case, it
did not explain the widespread moisture in
the roofs on the other two levels. The most
likely source was residual moisture in the
concrete deck that did not
dry in the minimal time each
section was exposed to ambient
conditions between tearoff
and replacement. Deckto-
parapet joints and penetrations
through the deck
had not been sealed, and it
is assumed that these voids
allowed sufficient venting to
keep the cover board facers
dry in these locations.
Twenty months after the
condition was discovered, the
parties agreed that the small
area in the location of the
proposed terrace (approximately
15% of the total roof area) would
be replaced with new materials, and the
balance of the surfaces would be ballasted.
The remediation cost approximately $6.00
per square foot. The combined attorney and
consultant fees and other expenses of all
parties no doubt exceeded the repair costs.
It is estimated that the installation of a
vapor retarder in the new roof system would
have added approximately $1.00 to $1.25
per sq. ft. to the cost, had it been installed
during the reroofing project.
In extreme situations, the only appropriate
remedy to such issues may be the
complete removal of the entire roof system
and replacement with a new roofing system,
including a vapor retarder to prevent
a reoccurrence. As noted previously, damage
resulting from condensation, regardless
of the source of the moisture, is explicitly
excluded from the coverage of most manufacturers’
warranties. The
involved parties will need to
agree upon the funding of the
roof replacement, or litigation
inevitably ensues.
Complete removal and
replacement may not be necessary,
and in some cases
may even be undesirable,
due, for example, to access
issues once new construction
is complete, or the potential
impact on the occupants
and/or operations below the
affected areas.
With adhered roof assemblies
on concrete decks, the
most common issue in such
cases is a loss of cohesive strength, typically
at an insulation or cover board facer,
leaving the roof vulnerable to catastrophic
failure due to wind blow-off. Providing longterm
resistance to wind loads will generally
drive the choice of remediation strategy
implemented in such cases.
The two case studies noted above highlighted
the two most common options:
fastening the existing assembly to the structural
deck or ballasting the now-loose roof
system. Ballasting will generally be the
less expensive solution if the structure can
accommodate the additional load. Even
when there is sufficient structural capacity,
this may not be an appropriate solution
in some jurisdictions. For example,
in hurricane-prone regions, IBC does not
allow aggregate ballast, and paver ballast
can only be used if it complies with RP-4.
Similarly, some buildings, such as high-rise
structures without parapets, may not be
suitable candidates for ballasting by code,
or the owner’s insurance carrier may not
allow it.
Mechanically attaching the assembly
through the existing roof membrane may,
in some instances, be a more viable option.
Beyond doing pullout tests, an assessment
must be carried out to determine the risk
of hitting rebar or tensioning cables in the
deck. Tapered insulation can increase the
cost of this option significantly.
Although both options will avert the
potential for catastrophic failure under wind
load, neither addresses the moisture that is
entrapped within the system. In all cases,
serious consideration must first be given
to the amount of moisture in the system
and what impact this moisture will have
2 2 8 • S c hwe t z 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 10 – Temporary ballast installed to secure
loose membrane.
Figure 11 – Assessing the moisture content of the
gypsum cover board.
on the different components in the system
over time.
Some success has been reported with
the use of one-way vents (i.e., an air-pressure
equalized system) in such situations.
Wind moving across the roof surface draws
air containing moisture from the assembly
and exhausts it to the atmosphere.
Furthermore, the air being drawn from the
assembly equalizes the pressure on the
top and bottom surfaces of the membrane,
thereby holding the membrane in place
and eliminating the need for additional
fixation or ballast. The effectiveness of these
systems is highly dependent on having
an air-impermeable substrate below the
roofing system to prevent billowing of the
membrane. The degree of drying will depend
on the potential for lateral movement of the
moisture within the roofing system so it can
reach the vents, as well as the number of
vents and their locations. The insulation or
another medium needs to be air-permeable
to allow air movement within the sealed
system. Such an approach should only be
considered after consultation with experts
in the use of the approach.
Although the costs of the approaches
listed will typically be significantly less than
the cost to remove and replace the roof,
ultimately someone still needs to pay for
them. Experience has shown, however, that
the lower the cost to resolve these problems,
the more amenable various stakeholders are
to working out some form of cost-sharing
arrangement and avoiding litigation.
The roofing industry must address the
issues of moisture in concrete, beginning
with educating the designers, construction
managers, general contractors, roofing
contractors, and material suppliers of the
problems that may occur due to moisture
migrating from the deck into the roofing
system. The structural engineer selects
the type and strength of the concrete, and
the supplier typically designs the concrete
mix to achieve the required compressive
strength and mix properties required to
place and finish the concrete to meet the
construction schedule. We commonly look
at this to be the 28-day cure period. While
the concrete will meet the specified compressive
strength in the designated time
frame, as noted above, the new concrete
still has almost all of its initial water. The
construction industry must accept this fact
and not use cure time as the notice to begin
roofing. The same can be stated for an
existing concrete roof deck that has been
subjected to leaks for a period of time and
is likely wet. Removing the existing roof system
and letting the deck dry for a few hours
will not allow the free water within the deck
to evaporate.
Developing a moisture test for new concrete
roof decks is likely to be an exercise
with little to no return. As demonstrated
above, the free water in the deck escapes so
slowly, and construction schedules are so
demanding, that any test that is developed
will take too long and will be impractical.
Therefore, we must accept that the roofing
system will be installed over a wet deck.
The authors believe that using the WUFI
software or similar modeling software, with
proper and adequate data, will provide the
best solution for designing a roofing system
that contains the vapor barrier needed to
provide long-term performance over a wet
concrete deck.
ASHRAE, 2013 ASHRAE Handbook of
Fundamentals, Chapter 26.
ASHR AE Standard 160, Criteria for
Moisture Control Design Analysis in
ASTM Standards D4263, F1869, F2170,
F2659, ASTM International.
Stephen J. Condren, “Vapor Retarders
in Roofing Systems: When Are They
Necessary?” Moisture Migration in
Buildings, ASTM STP 779, M. Lieff
and H. R. Treschel, Eds., American
Society of Testing and Materials,
1982, pp. 5-27.
René M. Dupius and Mark S. Graham,
“The Shortcomings of Using Prescriptive
Specifications With Emerging
Roofing Technologies,” Sept. 7,
Mark S. Graham, “Concrete Deck
Dryness,” Professional Roofing,
January 2012.
Mark S. Graham, “Moisture in Concrete
Roof Decks,” Professional Roofing,
February 2010.
Mark Graham, National Roofing Legal
Resource Center (NRLRC), “Problems
and Risks Posed by Concrete Roof
Decks,” September 2013.
H. Koester and I. Odler, “Investigations
on the Structure of Fully Hydrated
Portland Cement and Tricalcium
Silicate Pastes in Bound Water,
Chemical Shrinkage and Density
of Hydrates,” Cement and Concrete
Research, v 16, n 2, p 207-214,
March 1986.
H. Kuenzel, A. Karagiozis, and A. Holm,
2011. Wärme Feuchtetransport
Instationär (WUFI) Pro 5.1, Oak
Ridge National Laboratory and
Fraunhofer Institute for Building
MRCA T&R Advisory Bulletin 1/2011,
“Structural Lightweight Concrete
Roof Decks,” Sept 1, 2011.
National Ready Mix Concrete Association
(NRMCA), “Concrete in Practice,”
Technical Bulletin CIP-28 – 2004.
National Roofing Contractors Association
Industry Update, “Moisture in
Lightweight Structural Concrete
Roof Decks,” August 2013.
Steel Deck Institute, “Venting of
Composite Steel Floor Deck,”
Position Paper, May 2012.
Steel Deck Institute, “Venting of
Composite Steel Floor Deck,”
Position Statement, November 2008.
Sverre Smeplass and Asa Selmer, “Drying
of Light Weight Aggregate Concrete,”
Second International Symposium
on Structural Lightweight Aggregate
Concrete, Helland, Steinar, Ed., Vol.
2nd, pg. 833-843, 2000.
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 S c hwe t z e t a l . • 2 2 9