Skip to main content Skip to footer

A Discussion on Fenestrations Testing

May 15, 2018

A Discussion on Fenestrations Testing
José Estrada, RRO, PE
JRS Engineering
12721 30th Ave. NE, 2nd Floor, Seattle, WA 98125
Phone: 206-728-2358 • E-mail:
3 3 r d 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 r c h 2 2 – 2 7 , 2 0 1 8 E s t r a d a • 1 0 7
Standardized testing is a terrific way to validate the real-world performance of fenestration
systems. When combined with a thorough quality assurance plan, testing an enclosure
system with a standardized procedure allows design and construction teams to control for
quality. With that said, testing should not be specified or performed without consideration
of the needs of a project. Sometimes, an industry standard doesn’t quite match the design
intent of an assembly. In these cases, testing should be adjusted accordingly. The speaker
will cover the mechanics of fenestration water testing and how to critically analyze a water
penetration test against a fenestration’s design intent. The presentation will include a review
of common fenestration testing protocols from ASTM and AAMA, and exploration of their
origins and intent. Discussion will also focus on the building science behind common fenestration
systems. The dialogue will tie testing and science together to highlight areas where
the two don’t align. The presentation’s conclusion will feature a discussion of what could be
revised in test procedures to provide better value to design and construction teams.
José Estrada, RRO, PE — JRS Engineering, Ltd., Seattle, WA
JOSÉ ESTRADA is a professional engineer in the state of Washington
and a Registered Roof Observer. With a decade of experience as an
enclosure consultant under his belt, Estrada has worked on projects
of various types and sizes throughout the United States, Canada, and
China, focusing primarily on the Pacific Northwest and California. His
project experience includes all phases of design through construction,
including involvement with field and lab fenestration testing.
1 0 8 • E s t r a d a 3 3 r d 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 r c h 2 2 – 2 7 , 2 0 1 8
Standardized testing is a terrific way
to validate the real-world performance of
fenestration systems. When combined with
a thorough quality assurance plan, testing
any enclosure system to a standardized
procedure allows design and construction
teams to control quality on their project.
With that understanding, testing should
not be specified or performed without consideration
of the requirements of a project.
Sometimes, an industry standard doesn’t
quite match the design intent of an assembly.
In these cases, the testing should be
adjusted accordingly.
This paper aims to shine a light on the
mechanics of fenestration water testing and
guide the reader on how to critically analyze
a water penetration test when considering
the fenestration design intent. We do this
by first reviewing common fenestration testing
protocols from ASTM & AAMA. For the
selected protocols, we explore their origin to
gain a better understanding of their original
intent. We then jump into the building
science side of some common fenestration
systems. Finally, this paper ties the testing
and science together to discuss areas where
the two don’t align. We conclude with a
discussion on what might be revised in test
procedures to help the tests provide better
value to design and construction teams.
It is not uncommon for a group of
designers, owners, consultants, and contractors
to all be gathered at a construction
site staring anxiously at a drop of water on
a window frame during a field water penetration
resistance test, wondering where it
all went wrong. If the stage has been properly
set, the interested parties will be on the
same page with a consistent interpretation
of not only the test goals, but also the larger
project goals. Far too often, however, there
arises some conflict that usually stems from
a discrepancy in expectations among the
owners, designers, and fenestration manufacturers.
Owners often expect that the window
will test to its specified pressure without
any water intrusion to the interior. Window
test reports are often included in the record
documents that are submitted after the
building is complete and, understandably,
owners tend to be sensitive to blemishes on
this record that are the result of a string of
failed water leakage tests. Manufacturers,
on the other hand, supported by some
aspects of common industry test standards,
are perhaps held to a set of expectations
that are different from what the owner
expects. Industry-set performance expectations
generally allow some water intrusion
without failing the test. Designers are often
caught in the middle of this tug-of-war, having
to answer to owners who do not want to
see water on the interior of their window at
the design-specified conditions, while also
having to respond to manufacturers who
say that these expectations are not consistent
with standard industry terms.
Situations such as this compel the
entire team—each member with his or her
own unique perspective—to stand back
and consider the larger picture. What is the
goal of this testing, anyway? How did we
get here? What are we trying to achieve by
performing this test? This paper attempts
to gather the different perspectives and tie
them together.
The building and construction industry
widely subscribes to the notion that building
enclosure-related failures, when they
happen, tend to occur at interfaces and
penetrations. Indeed, the author’s anecdotal
experience supports this idea, not to mention
the bulk of industry records and case
studies that do the same. Perhaps the most
well known types of enclosure failures relate
to slow, concealed water intrusion at fenestrations.
1 The bulk of industry knowledge
regarding window failures and their causes
leads designers and owners of new construction
projects to, rightly so, scrutinize
newly installed window systems. This scrutiny
is guided by a series of standardized
test methods that are intended to hold all
window systems to the same basic criteria
such that designers and owners can compare
different systems appropriately. These
test standards have largely been developed
by the American Society for Testing and
Materials (ASTM) and further referenced by
the American Architectural Manufacturers
Association (AAMA) to provide the baseline
for how fenestration systems should be
tested throughout the progress of construction,
from design and manufacturing, to
installation, to post-construction.
These organizations typically build standards
in a consensus approach in which
they invite input from the industry players,
including manufacturers and designers. In
general, the intent of any test standard is
ultimately to set a common stage on which
different manufacturers can present their
product. In support of this goal, a test
standard typically isolates one component
of that product and tests it against some
measurable and consistent metric. The test
standard should respect the basic way such
products work and evolve as the technology
for that product does. Of course, test standards
should also be understood by those
performing and interpreting the test. Each
test standard has limitations, sensitivities,
and a fundamental goal; these should be
considered during both testing and interpretation
of test results.
The intent of this paper is to focus on
fenestration water penetration testing, for
which there are currently a handful of test
standards that are widely in use today.
The idea for this paper stems from field
discussions surrounding field water penetration
testing, but this paper will discuss
water penetration testing more conceptually.
Before we continue, the window test
methods, standards, and voluntary specifications
that we will discuss are briefly
defined below:
• ASTM E331 – Standard Test Method
for Water Penetration of Exterior
Windows, Skylights, Doors, and
Curtains Walls by Uniform Static Air
Pressure Difference
• ASTM E547 – Standard Test Method for
Water Penetration of Exterior Windows,
Skylights, Doors, and Curtain Walls by
Cyclic Static Air Pressure Difference
A Discussion on Fenestrations Testing
3 3 r d 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 r c h 2 2 – 2 7 , 2 0 1 8 E s t r a d a • 1 0 9
• ASTM E1105 – Standard Test Method
for Field Determination of Water
Penetration of Installed Exterior
Windows, Skylights, Doors, and
Curtain Walls, by Uniform or Cyclic
Static Air Pressure Difference
• AAMA 501.1 – Standard Test Method
for Water Penetration of Windows,
Curtain Walls, and Doors Using
Dynamic Pressure
• AAMA 502 – Voluntary Specification
for Field Testing of Newly Installed
Fenestration Products
• AAMA 503 – Voluntary Specification
for Field Testing of Newly Installed
Storefronts, Curtain Walls, and
Sloped Glazing Systems
The University of Miami claims the
earliest discoverable standardized window
test procedure in a 1952 published paper
that documents a “Study of Glass Jalousie
Windows Under Hurricane Conditions.”2 This
study has since been referenced by multiple
publications as being the origin of regular
standardized window testing in the industry.
The test procedure in the glass jalousie windows
study consisted of a window specimen
that was tested for water intrusion using
an aircraft propeller to generate wind with
water injected into the air stream to simulate
rain. The wind and water mix were intended
to simulate wind-driven rain conditions that
might be encountered during a hurricane
event.3 A similar test method was ultimately
included in the AAMA 501, Methods of
Test for Exterior Walls specification, and
was pulled out of this specification to form
a stand-alone test method, AAMA 501.1,
Standard Test Method for Water Penetration
of Windows, Curtain Walls and Doors Using
Dynamic Pressure,” in 20054
(Figure 1).
A few years after the
early glass jalousie windows
testing, a test method
was developed and used in
Norway that consisted of a
static air pressure difference
applied to one side of a window
with a spray of water
applied to the exterior. Due
to the complicated nature of
the apparatus used in the
Norwegian testing, the testing
method was not practical
for large specimens.5 In the
United States, early pressure
chamber testing included an
apparatus that consisted of
a perforated pipe that poured
water onto the test specimen
from above, while a chamber
applied a pressure gradient
across the specimen.
This test procedure was also
deemed impractical because the water
application method was driven by gravity
alone; thus, water could not easily wet the
window portion under any projections.6
At the end of the 1950s, the water spray
apparatus was modified to include spray
nozzles that allowed water to cover the full
specimen during pressure chamber testing.
This basic test procedure, including a pressure
chamber with a water spray in a grid
pattern, has gone on to form the foundation
for window testing across the industry.6
In 1967, ASTM first published this
basic procedure as a test standard, E331,
67T, Tentative Method of Test for Water
Resistance of Windows by Uniform Static
Air Pressure Differential. The standard was
revised in 1970 to include curtainwall
systems and doors.7 The test procedure in
ASTM E331 standardized several variables
in the test method, including the rate of
1 1 0 • E s t r a d a 3 3 r d 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 r c h 2 2 – 2 7 , 2 0 1 8
Figure 1 – Example of AAMA 501.1 dynamic test setup.
1950s 1960s 1970s 1980s 1990s 2000s Present
University of Chamber ASTM E331 ASTM E1105 AAMA 501.1
Miami glass testing and E547 first published; established
jalousie developed published. AAMA 501 as stand-alone
windows in U.S. published. procedure.
standard and Europe.
testing. First chamber
test method
drafted in U.S.
Figure 2 – Timeline of test standards and voluntary specifications development.
water spray to the exterior and the definition
of water penetration. The test standard
included a pressurized chamber, which
applied a static pressure differential
across the specimen
in an effort to simulate
wind conditions in a more
controlled and repeatable
manner than can be achieved
with a propeller in AAMA’s
dynamic test. In 1975, ASTM
followed up with an additional
test standard, E547, Test for
Water Penetration of Exterior
Windows, Curtain Walls, and
Doors by Cyclic Static Air
Pressure Differential, which
basically took the method in
E331 and applied pressure to
the chamber in cycles instead
of consistently to simulate
variable wind.16
ASTM E331 (static) and
ASTM E547 (cyclical) were
developed to test new window products in
a laboratory setting but were found to be
impractical by many users in a field setting.
To address this, ASTM developed its
E1105 test method in 1986 as a way of
standardizing window tests in the field:
ASTM E1105, Standard Test Method for
Field Determination of Water Penetration of
Installed Exterior Windows, Curtain Walls,
and Doors by Uniform or Cyclic Air Pressure
Difference. The E1105 test method carries
forward definitions and procedures from
both ASTM E331 and E547, and is now
what is referenced in several AAMA voluntary
specifications that pertain to window
testing for new and existing windows in the
field. See Figure 2.
There are many numbers and test codes,
which can get confusing, but the method of
test, in general, is similar across the board.
Pressure chamber test methods basically
consist of the following components (Figures
3, 4, and 5):
• Specimen: A w indow s pecimen
installed plumb into a rough opening
to a chamber.
• Water Spray: A pplied t o t he s pecimen
at a rate of five gallons per
minute through a series of spray
nozzles set up in a grid pattern. The
magnitude of flow—five gallons per
minute—was based on what the
test method authors thought was
reasonable to achieve full coverage
of the window specimen; ultimately,
the goal is full coverage to uncover
any water leaks.8
• Pressure Chamber: The test chamber
described in ASTM E331, a
lab-based procedure, was designed
to apply a positive pressure to the
exterior of the window as may be
expected in a wind event. Since
applying a chamber to the exterior of
a building in the field is impractical,
ASTM E1105, the field version of the
test, instructs that the test chamber
is to be applied to the interior of the
3 3 r d 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 r c h 2 2 – 2 7 , 2 0 1 8 E s t r a d a • 1 1 1
Figure 3 – Example of ASTM E331 chamber test on curtainwall,
with chamber on interior side of specimen.
Figure 5 – Example
of ASTM E105
field chamber test
with chamber on
interior of punch
window specimen.
Figure 4 – Example
of ASTM E331
chamber test on
window wall unit
with chamber on
exterior side of
specimen with an applied vacuum.
• Time: Water is then applied using
either static pressure for 15 minutes
or cyclic pressure with five-minute
intervals. The authors of ASTM
E331 proposed that 15 minutes was
a sufficient time period to exhibit
weaknesses in the fenestration system.
Although the ASTM E331 standard has
been updated four times since its creation,
the basic components of the test method
have not changed. Indeed, the procedure
in ASTM E331 is now so standard that it is
used widely in test methods and voluntary
specifications throughout the industry.
The ASTM E1105 standard has also
been updated several times over the years,
but much like the E331 standard, the
basics of E1105 have remained the same.
For the purpose of this paper, the author
intends to discuss ASTM E1105 specifically,
as it is the standard usually followed during
field fenestration tests and the document
that is often scrutinized when a field failure
This field test standard has some specific
wording, as well as some notable ambiguities.
For example, being strictly a test
procedure, ASTM E1105 does not specify a
minimum test pressure (though both ASTM
E331 and E547, the lab equivalents, do).
The test pressure, which one can argue is
a design requirement rather than a procedural
item, is left up to the industry in the
field and could be the subject of a paper on
its own. Instead, the standard specifies how
to achieve a pressure differential. Given the
origin of the test, it is safe to assume that
the intent of the test is, and has always
been, to provide a standardized method of
testing that can be performed at various
levels of intensity. Specifically, the intensity
of the pressure chamber test is determined
by the pressure differential applied to the
chamber. The pressure differential is the
key variable; all else being equal, different
windows will test to different pressure differentials.
This approach is consistent with
standard scientific method, in which a limited
number of variables are monitored and
controlled in a repeatable test.
On the other hand, the test procedure
is clear on the rate of water spray to be
applied to the test specimen—
a rate of five gallons
per hour per square foot.
Since this is very clearly
called out in the ASTM
test, it is sometimes scrutinized
during testing as
having a more significant
impact on the performance
of the specimen
than the ASTM E1105
authors intended. Some
observers have attempted
to convert the spray rate
in this standard to inches
per hours of rainfall.
ASTM, in their non-mandatory
information note
“X1., Spray Rack Rate,”
indicate that “the purpose
of the water-spray system
is to wet the specimen
uniformly for the purpose
of evaluating water resistance.
There is no evidence
that the developers
of Test Methods E1105,
E331, and E547, intended
to reproduce or simulate any given
rain event.”8 Indeed, the author’s anecdotal
experience with field fenestration testing
suggests that the rate of water application
to a vertical window surface alone does not
usually govern, since water is shed quickly
from the surface and does not typically
accumulate sufficient volume to generate
enough hydrostatic pressure on the exterior
of the fenestration to make a significant
impact. The number, 5 gal/hour/sq. ft., is
merely a target.
Time is also defined in the test standard:
15 minutes continuous in Procedure
A (static), and at least 15 minutes of pressure
difference total in Procedure B (cyclical).
The 15-minute duration of spray was
included in the initial ASTM E331-70 definition
of “water leakage” and has since carried
forward to some degree in the ASTM
E1105 standard. The duration of spray—
much like the rate of water application—is
intended to provide enough time to allow a
discontinuity or leak to present itself and is
not necessarily intended to be representative
of real-world rain conditions.7
Along with defining the test procedure,
including specifics on water spray and
method of achieving pressure differential,
the test method defines a series of terms
and definitions. One notable definition that
is included is for the term “water penetration,”
which the standard defines as:
3.2.3 water penetration, n—penetration
of water beyond a plane parallel
to the glazing (the vertical plane)
intersecting the innermost projection
of the test specimen, not including
interior trim and hardware,
under the specified conditions of air
pressure difference across the specimen.
For products with non-planar
surfaces (domes, vaults, pyramids,
etc.) the plane defining water penetration
is the plane defined by the
inner most edges of the unit frame.8
Based on this definition, at any given
specified test pressure, water can enter the
test chamber and remain on the window
specimen. If that water does not overflow
the “innermost vertical plane” of the specimen,
the water intrusion is not considered
“water penetration” (Figure 6). Though the
test is very specific on the definition of water
penetration, it is ambiguous on allowable
1 1 2 • E s t r a d a 3 3 r d 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 r c h 2 2 – 2 7 , 2 0 1 8
Figure 6 – Example of ASTM E1105 “water penetration”
plane overlain on window shop drain. Water to the
exterior of this plane does not constitute water
penetration per the standard.
Additionally, the standard defines a failure
criterion as:
…water penetration in accordance
with 3.2.3 [definition of water penetration].
Failure also occurs whenever
water penetrates through the
perimeter frame of the test specimen.
Water contained within
drained flashing, gutters, and sills is
not considered failure.8
The challenge with these definitions
is that, depending on their interpretation,
they may not always align with the expectations
of designers, specifiers, and the end
users of the windows, and may conflict with
building science-based design objectives,
discussed further in this paper (Figure 7).
Following research on the origins of the
test standards and their wording, the origin
of the definition for water penetration was
not clear to the author. One may be able
to surmise that it might have something to
do with the very early standardized testing
of glass jalousie windows at the University
of Miami, which intrinsically don’t have a
water or air control layer (more on these
later). If a definition for water penetration
must be global enough to work for both a
jalousie window on the one hand, and a
unitized curtainwall on the other, then this
might be the right definition.
Or, perhaps a different reasoning for
this definition might be how the authors
thought the test should be applied. For
example, the authors may have intended
this to be a “limit states” test—a test to
bring the specimen to near ultimate loads
under which some leakage, if contained,
would be considered reasonable performance.
Perhaps, when the test standard
was developed following early testing in
Florida that simulated hurricane conditions,
the mindset of the authors may have
been that the window, under such conditions,
should be allowed to leak and capture
water to some degree—thinking that
the window would seldom be exposed to
such intense weather events and would not
leak under normal conditions. Surely, it is
reasonable to say that the window should
have some design pressure under which
no water intrusion should occur, and some
limited state during which a finite amount
of controlled water leakage may be reasonable
to anticipate.
Whatever the origin of the wording,
ASTM E1105 (as well as E331 and E547)
ultimately allows the designer and specifier
some room for modification, if needed,
to align with their project goals, indicating
that the failure criteria is as defined “unless
otherwise specified.”18,13,16
The AAMA501.1 test method (the propeller
test) has a similar, though slightly more
specific, definition of water penetration
(called “water leakage” in the test method),
defining it as:
…any uncontrolled water that
appears on any normally exposed
interior surfaces, that is not contained
or drained back to the exterior,
or that can cause damage to
adjacent materials or finishes. Water
contained within drained flashings,
gutters, and sills is not considered
water leakage. The collection of up
to 15 ml (1/2 oz) of water in a 15-minute
test period on top of an interior
stop or stool integral with the system
shall not be considered water leakage.
This definition is mostly adopted by the
AAMA 503 voluntary specification (which
is intended for curtainwalls),18 while the
AAMA 502 voluntary specification maintains
a definition similar to that in ASTM
E1105 and E331.17
The different interpretations and definitions
of water penetration, water leakage,
and fail criteria in the standards compel
the industry to take another look at these
standards and consider whether a global
change may be in order. The definition of
water intrusion is simply perceived by some
in the design community as being “too lax,”
and as a result, designers and specifiers are
frequently applying more stringent definitions
in their project specifications,9 which
in turn often leads to contention when
manufacturers don’t fully understand the
idiosyncrasies of a project specification and
bid their product based on standard AAMA
or ASTM definitions.
In 1967, when ASTM E331 was first
being developed and published, the field of
building science was still very much in its
infancy. The National Institute of Building
Sciences (NIBS) was founded in 1974, with
a mission to connect the U.S. government
with the private sector in an effort to
“improve the built world.”10 In Canada, the
Canada Housing and Mortgage Corporation
(CHMC), founded in the 1940s to house
war veterans returning home from World
3 3 r d 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 r c h 2 2 – 2 7 , 2 0 1 8 E s t r a d a • 1 1 3
Figure 7 – Example of water collecting on mullion during an ASTM E1105 field
test. White dashed line shows innermost vertical plane of specimen. Per ASTM
E1105, this did not constitute water penetration or a failure.
War II, didn’t start to focus on the performance
of building enclosures until the
late 1970s. Indeed, the CMHC credits the
1990s as being the “new era of building
science,” in which the building enclosure
became a big part of the conversation as
it pertains to building performance.11 The
Roof Consultant’s Institute (RCI) joined the
picture in 1983, focusing first on roofing,
and growing into the international group
we know today, whose mission now incorporates
the full building enclosure.12
The standards, not having changed
significantly, have remained largely idle
through substantial changes and leaps of
knowledge in the building industry. This is
not to say that the standard must change for
the sake of change itself. What the author
suggests is that the building enclosure discipline
has experienced a paradigm shift
since the 1990s, and the route of normal
science is to revisit practices as paradigms
change. The gains in knowledge of fenestration
design and manufacturing that have
largely developed alongside the gains in the
building science discipline compel the standards
to adjust such that they fit into the
larger picture and remain relevant.
Fundamentally, the intent of a building
enclosure system is to separate environments.
The exterior environment is uncontrolled,
and has conditions that can be
widely variable. Interior conditions, on the
other hand, are often intended to be more
stable, controlled, and less variable. In
order to separate these environments and
enclose a space, we need to control the variables.
These variables, which are addressed
by the enclosure, include air and its component
parts: water, heat, harm (people
and impact), fire, and sound. Sometimes,
the enclosure even acts as structure while
simultaneously controlling these variables.
For the purpose of this discussion, we will
focus on the separation of environments
necessary to control weather—specifically
air, water, thermal, and vapor variables.
In this regard, when designing a building
enclosure, the building science professional
focuses on what are known as
“control layers.” Basically, these are defined
boundaries in the building enclosure that
work together to separate the exterior environment
from the interior environment.
These control layers are continuous around
the full enclosure and can be traced from
the walls, to the roofs, to the floors. They
are also closely related to each other, as
you will read below. When considering
one of the control layers, we must always
remember to step back and consider how it
relates to the others, and how it relates to
the larger building system.
The primary control layers that a building
science professional reviews are air,
water, thermal, and vapor.
Air is a mix of gasses,
consisting mostly of gaseous
nitrogen and oxygen
and other trace gasses,
such as carbon dioxide
and gaseous water (water
vapor). The energy that
we feel in the environment
as temperature is
contained within the bulk
mixture of these gasses;
so, in order to maintain
a controlled interior
environment, it is first
important to control the
exchange of this gas mixture.
Because of the fluid
nature of air, air barriers
must be continuously
sealed, much like a balloon
needs to be continuously
sealed to prevent its
air from escaping.
Control of bulk water is important for
longevity of the building structure. Because
of the nature of most building materials,
water is responsible for much of the damage
associated with a building enclosure
failure. Uncontrolled water intrusion is also
often responsible for deterioration of interior
air quality. Because of the fluid nature of
water, the water barrier must be continuous,
though it does not need to be continuously
sealed. Redundancy and containment
are important for water control, as well as
lapping and shedding. Liquid water is often
carried within moving air, in the form of
wind-driven rain, so even though a water
barrier does not necessarily need to be air
tight to be waterproof, one must consider
the location and detailing of both the water
and air barriers together when designing for
water control.
The energy we feel as temperature is
held largely in the bulk gasses of air. So, a
well-designed and constructed air barrier
is usually the first step for thermal control
across the building enclosure. The building
enclosure is intended to keep occupants
at a comfortable temperature despite the
cold or heat outside. To do this, a thermal
insulation system, used to control heat
exchange across the enclosure, is necessary
in addition to an air barrier. Heat is
transferred primarily in three ways: convection
(movement of fluids—addressed by
the air barrier and thermal insulation),
conduction (through solids—addressed by
thermal breaks and through thermal insulation),
and radiation (through space itself—
addressed by reflective coatings). In order
for the thermal enclosure to be effective, it
needs to be continuous.
Water vapor (gaseous water) is a special
trace gas within the mix of gasses that
make up air. Water vapor is unique in that
it ties the water, air, and thermal control
layers of the enclosure together. Consider
that water gas molecules are much smaller
than the rest of the molecules in bulk air;
as a result, water vapor can diffuse through
some materials that would otherwise be
impermeable to bulk air. The ability of air
to hold water vapor is a function of the air’s
temperature; more heat means more energy
and thus more water vapor. We experi-
1 1 4 • E s t r a d a 3 3 r d 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 r c h 2 2 – 2 7 , 2 0 1 8
Figure 8 – Building science control layers overlain onto
a typical curtainwall assembly.
ence the water vapor in the air as humidity. While water vapor can diffuse
through materials that the rest of bulk air cannot, it usually does so quite
slowly. Bulk air movement, on the other hand, can carry water vapor across
an enclosure much more effectively than diffusion alone. So, a continuous air
barrier remains important in control of water vapor—but a well-placed vapor
retarder is also important.
In general, these four primary control layers need to translate at all
interfaces, including at fenestrations, to make an effective enclosure system.
Figure 8 applies these control layers to a typical fenestration system.
Different types of fenestration systems treat the various control layers differently,
but in general, there are three ways to construct a fenestration system.
14 For the purpose of this paper, we have assumed that all the fenestration
systems discussed are designed with an air barrier in mind, as required for
conformance with most commercial energy and building codes.
• Face Sealed – Face-sealed systems have no plan for water within the
outermost surface of the system. All seals and gaskets in a face-sealed
system are assumed to be perfectly tight. In a face-sealed system, no
water should enter the system, as there is usually no means to drain
the water back to the exterior.
• Concealed Barrier – In a concealed barrier system, the primary water
and air control layers are concealed by some cladding component. The
cladding component is not necessarily intended to drain and manage
water; it simply covers the underlying layers and conceals them.
• Rainscreen – A rainscreen system employs a two-stage approach to
managing water. There is an exterior water-shedding surface that is
mostly continuous with intermittent weeps; a concealed, protected,
and vented drainage cavity; and a backup water-control boundary
beyond which no water is intended to pass to the interior.
In each case, there
is a defined plane for all
the discussed control
layers. We will highlight
the defined water-control
boundary, beyond which
the design intent is for
no water to be permitted.
Refer to graphics in
Figures 9, 10, and 11 for
examples of how the water
control boundary looks in
each type of fenestration.
When considering the
building science side of the
equation, one finds a clear
disconnect between the
industry-adopted definitions
of water penetration
in the standards and the
design intent as described
in building science fundamentals,
which include
a water control boundary
that is often not planar.
The scientific method
3 3 r d 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 r c h 2 2 – 2 7 , 2 0 1 8 E s t r a d a • 1 1 5
Figure 9 – Example of face-sealed fenestration
system. Notice the water control boundary. The
design intent is for 100% of water to stop at exterior
Figure 10 – Example of concealed barrier
fenestration system. Notice that water control
boundary is partially protected, but there is not
a defined water-shedding plane.
Figure 11 – Example of rainscreen fenestration
design. Notice exterior water-shedding surface,
drained cavity, and interior water control
boundary beyond which water is not permitted.
requires repeatable tests with controlled
and limited variables that test a specific
item to prove or disprove a hypothesis.
Each fenestration design and assembly
can ultimately be seen as a hypothesis of
performance. It’s important to understand
what the intent of the test is, or consider
what it should be: What, specifically, is the
test standard trying to prove or disprove?
Ultimately, the test standards referenced in
this paper are a validation of the water control
boundary for the fenestration system.
Consider again the definition of water
penetration in the referenced test standards.
When looking at
the various boundaries of
the water control for different
fenestration systems,
as is done in Figures
9-10, it is evident that the
water control boundary as
defined by the building science
doesn’t align with the
boundary used to trigger
the term “water penetration”
as defined by the fenestration
system test standard
(Figures 12 and 13).
The author notes that
control layers are not necessarily
absolute. The air
control layer, for example,
does not stop all air from
passing in an uncontrolled
manner; rather, it limits
the rate of air penetration
to a manageable level. This
is similarly the case for the
building enclosure thermal
and the vapor control layers;
these do not necessarily
completely stop the
passage of energy or vapor,
respectively. Rather, they
reduce the uncontrolled
passage to a manageable
level. One may argue that
the water control layer is
no different and that some
water intrusion onto the
window, provided that it
is controlled and manageable,
is sufficient for
performance. While this
is a fair conclusion, the
author suggests that liquid
water control in the building
enclosure is special.
It is fair to say that some
water passage inboard of
the outer face of the fenestration
system is usually
inevitable. Indeed, building
enclosure design allows for
water to be controlled and
managed through the rainscreen principle,
which is increasingly the method of design
for enclosure components. Fundamentally,
the water control layer actually consists of
multiple layers: a water-shedding layer, a
drainage space, and a boundary of watertightness
beyond which water is not permitted
to pass at the specified design pressure
Consider Figure 13, which shows a typical
operable casement window. This figure
illustrates one example of a fenestration
system allowing water to bypass the outermost
plane in such a way that it is contained
within drained gutters that are intended to
take and discharge water. Building science
fundamentals acknowledge that it is nearly
impossible for water to not bypass the outermost
plane of any system. This is why
internal drains and gutters are fairly standard
practice in many fenestration systems.
Allowing water to bypass the secondary
water control boundary—even if that water
is contained on the frame of the window
inboard of the water penetration boundary
defined by ASTM—would defeat the design
intent of the system. Water inboard of that
secondary water control boundary is simply
not planned for, and thus, not intended to
drain or be discharged in any way. Allowing
water to bypass that line without constituting
a failure simply because the water is
inboard of the innermost vertical plane of
the specimen, does not acknowledge the
design intent of such a system.
Looking at the adjacent building enclosure
components, such a definition that
allows water to bypass the boundary of
watertightness at the specified pressure
difference would simply not suffice. The fenestration,
being an extension of the building
enclosure, is often subject to the same rigor
and performance expectation. It is with
this perspective that the building enclosure
designer comes to the table.
Consider the face-sealed fenestration
system in Figure 9 and the concealed barrier
fenestration system in Figure 10, which
I’ve shown again in Figure 14 with the
“water penetration” boundary as defined
by the test standard. While these two fenestration
design approaches do not employ
a rainscreen principle, they, by design, also
do not allow for the passage of liquid water
inboard of their water-control boundary. In
these designs, water intrusion into the window
beyond that boundary would constitute
a failure of the design intent, even though
1 1 6 • E s t r a d a 3 3 r d 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 r c h 2 2 – 2 7 , 2 0 1 8
Figure 12 – Water control boundary as defined by
building science compared with “water penetration”
boundary as defined by ASTM E1105.
Figure 13 – Water control boundary as defined by
building science compared with “water penetration”
boundary as defined by ASTM E1105.
it may not constitute a failure by letter of
the standard. This approaches the heart of
the misalignment in expectations and the
origin of much contention during interpretation
of field test results. The standard has
a set criterion of failure that is based on a
definition of water penetration that does not
match the design-side definition of water
penetration. The boundaries defined by the
building science design community simply
do not match the boundaries established by
ASTM in the late 1960s. Until these boundaries
agree, there will remain a potential
for misunderstanding, unmet expectations,
and contention in our industry.
ASTM does not purport to be a steadystate
organization, where, once a standard
is published it is held indefinitely. Indeed,
ASTM invites the industry to comment on
its standards and provide feedback and
criticism as may be appropriate with the
changing times. To this point, ASTM has
already been considering draft edits to
the definition of water penetration as is
described in ASTM WK27894, which was
initiated in 2010, is still in draft form, and
states that “The current definition of water
penetration in E1105—as well as E331 and
E547—does not align with end user expectations
when associated with service wind
loads. This work item will provide a revision
to more accurately align the definition of
water penetration with end user expectations.”
Being a consensus organization,
changes often take time while the committee
considers the perspectives of various
players and considers potential unforeseen
impacts from any changes.
In the interim, designers, specifiers, and
owners may consider a few steps that can
be taken to mitigate the potential for misunderstanding
and contention as they relate
to window water penetration performance
testing in the lab and field. Some possible
steps are included below:
1. Request and read the test reports
from the fenestration manufacturer,
and inquire what water penetration
test was performed in the lab,
what the specific results were, and
how the test was performed. Even
though the ASTM lab test procedure
defines water penetration the
way it does, many manufacturers
test their windows such that no
water intrusion occurs inboard of
the water control boundary.
Understanding how
the specific fenestration
was tested in the lab will
provide a more realistic
performance expectation
that the manufacturer
will be more likely to
honor in the field.
2. Define the expected
performance clearly
in your project specification,
including a reasonable
and appropriate
pressure difference that
the fenestration system
must (and can) meet in
the lab and in the field,
a specific definition of
water penetration that
respects the fenestration
design intent and the
boundary of watertightness,
and criteria for a
failure that also respects
the design intent for the
specific fenestration.
3. Consult with your fenestration
window installer,
and owner early in
design and construction
to relay the project performance
The project performance
expectations should be
set amongst the owner,
design team, and window
manufacturer prior to
any testing. It may be the
case that under test conditions,
a limited amount
of water in the test definitions
is acceptable to
an owner if there are
trade-offs with cost and
the expectation is set.
Keep in mind that while
a current owner may
accept the performance
of a fenestration for a cost
savings, a future owner
(if the building gets sold)
may not accept such performance
On the larger industry scale, a revision
to the ASTM standard is due, in the author’s
opinion. The definition of water penetration
in the referenced standard and subsequent
3 3 r d 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 r c h 2 2 – 2 7 , 2 0 1 8 E s t r a d a • 1 1 7
Figure 14 – Water control boundary as defined
by building science compared with “water
penetration” boundary as defined by ASTM
E1105 on face-sealed (top) and concealed barrier
(bottom) fenestration.
standards is too prescriptive and does not
respect the design intent of many modern
fenestration systems or the current practices
of building science. In order to adjust
for this mismatch, many designers provide
project-specific performance criteria, sometimes
including a more stringent definition
of water leakage, but this often leads
to contention in the field, especially when
there is a water leak and the manufacturer
contends that their product is held to a different
standard. Fundamentally, the point
of a standard is to standardize wording so
that such contention is avoided. An update
to the standard can be as simple as revising
the definition of water penetration to “any
water that passes the defined boundary of
watertightness for the fenestration system
at the specified pressure.” This definition
would allow manufacturers to define their
own boundary of watertightness clearly,
such that owners and designers would better
understand what to expect and could
rely on their windows not exhibiting water
leakage of any kind inboard of the boundary
of watertightness under specified conditions.
Ultimately, when there are misunderstandings
or contention in the field, it is
usually the result of poorly set expectations.
In such cases, it is best for all players to
step back, consider the perspective of those
around them, and find a common middle
ground that is practical, appropriate, and
respects the project goals. In this regard,
a common understanding of fenestration
systems, building science principles, along
with the design team and end-user expectations,
may help bridge the misunderstanding
and move the industry as a whole one
step further.
1. E.C.C. Choi. “Criteria for Water
Penetration Testing.” Water Leakage
Through Building Facades. ASTM
STP 1314. R.J. Kudder and J.L.
Erdly, Eds., American Society for
Testing and Materials. 1998.
2. A.A. Sakhnovsky. “Testing for Water
Penetration,” Window and Wall
Testing. ASTM STP 552. American
Society for Testing and Materials,
1974. pp. 31-35.
3. H.H. Sheldon. “A Study of Glass
Jalousie Windows Under Hurricane
Conditions.” University of Miami,
Coral Gables, FL, 1952.
4. AAMA 501-15, Methods of Test for
Exterior Walls.
5. S.D. Svendsen and R. Wigen. Testing
Window Assemblies, ASTM STP 251.
American Society for Testing and
Materials, 1959. pp. 36-38.
6. F.J. Spagna and B. Buchberg. “Water
Testing Misconceptions: Fenestration
Product Certifications and Forensic
Investigations of Building Leakage.”
27th RCI International Convention
and Trade Show Proceeding. 2012.
pp. 95-96.
7. ANSI/ASTM E331, Standard Test
Method for Water Penetration of
Exterior Windows, Curtain Walls,
and Doors by Uniform Static Air
Pressure Difference. 1970 (reapproved
8. ASTM E1105-15, Field Determination
of Water Penetration of Installed
Exterior Windows, Skylights, Doors,
and Curtain Walls, by Uniform or
Cyclic Static Air Pressure Difference.
9. B.S. Kaskel, M.J. Scheffler, and I.R.
Chin. “Critical Review of Curtain
Wall Mockup Testing for Water
Penetration,” Water Leakage Through
Building Facades, ASTM STP 1314,
R. J. Kudder and J. L. Erdly, Eds.,
American Society for Testing and
Materials, 1998. pp. 11.
10. 2017 National Institute of Building
Sciences. About the institute.
11. 2017 Canada Mortgage and Housing
Corporation (CMHC). History of
CMHC. https://www.cmhc-schl.
12. RCI, Inc. About RCI. http://rcionline.
org/about-rci/. 2017.
13. ASTM E331-00 (Reapproved 2016).
Standard Test Method for Water
Penetration of Exterior Windows,
Skylights, Doors, and Curtain Walls
by Uniform Static Air Pressure
14. Canada Mortgage and Housing
Corporation. Water Penetration
Resistance of Windows—Study of
Manufacturing, Building Design,
Installation and Maintenance
Factors, November 2003. p 2.
15. AAMA 501.1-05, Standard Test
Method For Water Penetration Of
Windows, Curtain Walls And Doors
Using Dynamic Pressure.
16. ASTM E547-00 (Reapproved 2009),
Standard Test Method for Water
Penetration of Exterior Windows,
Skylights, Doors, and Curtain
Walls by Cyclic Static Air Pressure
17. AAMA 502-08, Voluntary Specification
for Field Testing of Newly Installed
Fenestration Products.
18. AAMA 503-08, Voluntary
Specification for Field Testing of
Newly Installed Storefronts, Curtain
Walls and Sloped Glazing Systems.
1 1 8 • E s t r a d a 3 3 r d 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 r c h 2 2 – 2 7 , 2 0 1 8