STP 1498-2011

Journal of Testing and Evaluation
Selected Technical Papers
STP 1498
Condensation
in Exterior
Building Wall
Systems
JTE Guest Editors:
Bruce Kaskel
Robert J. Kudder
Journal of Testing and Evaluation
Selected Technical Papers STP1498
Condensation in Exterior Building
Wall Systems
JTE Guest Editors:

Bruce S. Kaskel
Robert J. Kudder
ASTM International
100 Barr Harbor Drive
PO Box C700
West Conshohocken, PA 19428-2959
Printed in the U.S.A.
ASTM Stock #: STP1498

Library of Congress Cataloging-in-Publication Data
Condensation in exterior building wall systems / JAI guest editors, Bruce S. Kaskel,
Robert J. Kudder.
p. cm. -- (Journal of testing and evaluation selected technical papers; STP1498)
Includes bibliographical reference and index.
ISBN: 978-0-8031-4471-2 (alk. paper)
1. Dampness in buildings. 2. Exterior walls--Protection. 3. Waterproong. I. Kaskel,
Bruce S. II. Kudder, Robert J., 1945-
TH9031.C663 2001
693.893--dc22 2011006935
Copyright 2011 ASTM INTERNATIONAL, West Conshohocken, PA. All rights

reserved. This material may not be reproduced or copied, in whole or in part, in any printed,
mechanical, electronic, lm, or other distribution and storage media, without the
written consent of the publisher.
Journal of ASTM International JAI Scope
The JAI is a multi-disciplinary forum to serve the international scientic and engineering
community through the timely publication of the results of original research and
critical review articles in the physical and life sciences and engineering technologies.
These peer-reviewed papers cover diverse topics relevant to the science and research that
establish the foundation for standards development within ASTM International.
Photocopy Rights
Authorization to photocopy items for internal, personal, or educational classroom use, or
the internal, personal, or educational classroom use of specic clients, is granted by
ASTM International provided that the appropriate fee is paid to ASTM International, 100
Barr Harbor Drive, P.O. Box C700, West Conshohocken, PA 19428-2959, Tel:
610-832-9634; online: http://www.astm.org/copyright. The Society is not responsible, as
a body, for the statements and opinions expressed in this publication. ASTM
International does not endorse any products represented in this publication.
Peer Review Policy
Each paper published in this volume was evaluated by two peer reviewers and at least
one editor. The authors addressed all of the reviewers comments to the satisfaction of both
the technical editor(s) and the ASTM International Committee on Publications. The
quality of the papers in this publication reects not only the obvious efforts of the authors
and the technical editor(s), but also the work of the peer reviewers. In keeping with
long-standing publication practices, ASTM International maintains the anonymity of the
peer reviewers. The ASTM International Committee on Publications acknowledges
with appreciation their dedication and contribution of time and effort on behalf of ASTM
International.
Citation of Papers
When citing papers from this publication, the appropriate citation includes the paper
authors, paper title, J. ASTM Intl., volume and number, Paper doi, ASTM International,
West Conshohocken, PA, Paper, year listed in the footnote of the paper. A citation is
provided as a footnote on page one of each paper.
Printed in Baltimore, MD
May, 2011
Foreword
THIS COMPILATION OF THE JOURNAL of TESTING and EVALUATION
(JTE), STP1498, on Condensation in Exterior Building Wall Systems
contains only the papers published in JTE that were presented at a
symposium in San Antonio, TX, October 1011, 2010 and sponsored by
ASTM Committee E06 on Performance of Buildings.
The Symposium Co-Chairmen and JTE Guest Editors are Bruce S.
Kaskel, Wiss, Janney, Elstner, Associates, Inc., Chicago, IL and Robert J.
Kudder, Raths, Raths & Johnson, Inc., Willowbrook, IL.
Contents
Overview ........................................................................ vii
Insulation Draws Water
W. B. Rose ........................................................... 1
Testing/Analysis
Laboratory Tests of Window-Wall Interface Details to Evaluate the Risk of Condensation
on Windows
W. Maref, N. Van De Bossche, M. Armstrong, M. A. Lacasse,
H. Elmahdy, and R. Glazer............................................... 31
Drying Characteristics of Spray-Applied Cellulose Fiber Insulation
M. Pazera and M. Salonvaara ............................................. 59
Moisture Damage in Vented Air Space of Exterior Walls of Wooden Houses
T. Umeno and S. Hokoi. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Moisture Measurements and Condensation Potential in Wood Frame Walls in a
Hot-Humid Climate
T. A. Weston and L. C. Minnich ........................................... 94
A Review of ASHRAE Standard 160Criteria for Moisture Control Design Analysis in
Buildings
A. TenWolde .......................................................... 119
Moisture Response of Sheathing Board in Conventional and Rain-Screen Wall Systems
with Shiplap Cladding
F. Tariku and H. Ge ..................................................... 131
Investigation of the Condensation Potential Between Wood Windows and Sill Pans in a
Warm, Humid Climate
G. P. Stamatiades, III ................................................... 148
Case Studies
Interior Metal Components and the Thermal Performance of Window Frames
................................................
S. K. Flock and G. D. Hall 169
Controlling Condensation Through the Use of Active and Passive Glazing Systems
................................
A. A. Dunlap, P. G. Johnson, and C. A. Songer 187
Case Study of Mechanical Control of Condensation in Exterior Walls
C. M. Morgan, L. M. McGowan, and L. D. Flick. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
Considerations for Controlling Condensation in High-Humidity Buildings: Lessons
Learned
S. M. OBrien and A. K. Patel ............................................. 247
Fenestration Condensation Resistance: Computer Simulation and In Situ
Performance
E. Ordner ............................................................ 269
Improving the Condensation Resistance of Fenestration by Considering Total Building
Enclosure and Mechanical System Interaction
P. E. Nelson and P. E. Totten ............................................. 286
Condensation Problems in Precast Concrete Cladding Systems in Cold Climates
..........................................................
T. A. Gorrell 299
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317

Overview
This STP represents the peer-reviewed papers first presented at the October
1011, 2010 symposium on Condensation in Exterior Wall Systems in San
Antonio, Texas, sponsored by ASTM E06 Building Performance, Subcom-
mittee E06.55 Exterior Wall Systems. The symposium and this STP repre-
sent the continued efforts of this subcommittee to exchange state-of-the-art
knowledge through symposia on topics related to the performance of exte-
rior wall systems. Past symposia of this subcommittee include water leak-
age, repair and retrofit, faade inspection and maintenance, and perfor-
mance of exterior wall systems. Condensation in walls is a timely topic for
ASTM E06 to address. Advancements in building sustainability, energy ef-
ficiency, and new wall systems have progressed significantly in recent years,
while the consequential changes in wall moisture behavior resulting from
these advancements are less well understood.
Although the topic of condensation, per se, is not addressed in this sub-
committees prior symposia, it has been a related topic in much of the work of
this subcommittee and the ASTM E06 committee at large. Numerous previ-
ous papers, available through ASTM, have addressed this topic. Seminal
manuals and prior symposia presented by ASTM E06, and ASTM committee
C16 on Thermal Insulation, chaired solely or in part by Heinz Treschel,
serve as background to our current work. A sampling of those volumes in-
cludes:
MNL 40 Moisture Analysis & Condensation Control in Building Enve-
lopes - Treschel, ed. 2001;
MNL 18 Moisture Control in Buildings - Treschel, ed. 1994; and
STP 1039 Water Vapor Transmission Through building Materials and
Systems Treschel and Bomberg, eds. 1987.
Manual MNL 40 described some of the now-established computer simu-
lations for condensation control such as WUFI (ORNL/IBP). Given the now
nine years time since that work was published, E06 believed that the state-
of-the-art had advanced and that practical experiences have been gained
from the use of analytical products that were presented in the 2001 manual.
This symposium provided the opportunity for leading scientists and practi-
tioners to again advance the body of knowledge on the topic of condensation
in exterior wall systems.
Beyond ASTM, organizations such as ASHRAE have offered longstand-
ing input on the issue of condensation control. Other organizations have
grown more recently, such as BETEC; and USGBC along with their LEED
certification system. These organizations are interested, directly or periph-
erally, in the issue of condensation. They too have offered recent workshops
on the topic of condensation. Code writing organizations such as IBC, in
their energy code IECC, as well as their under-development green code,
vii
IgCC, are actively codifying issues related to condensation control, which
were brought to light in prior ASTM publications and in the work of these
other organizations. E06 believed in presenting this symposium, that these
current papers on condensation could have a similar impact in future build-
ing codes.
This STP is organized, in the same presentation as the October 2010 sym-
posium, into two parts:
Testing/Analysis 7 papers that concentrate on testing/analysis of materi-
als and mock-ups to predict and prevent condensation in common exterior
wall systems and
Case Studies 7 papers that document condensation problems found in the
real-world and their solutions.
In addition, there is one keynote paper by William Rose, which presents
the history that has lead to the present state-of-the-art and some of the er-
roneous concepts that have advanced to today. This paper sets the tone that
common-place thinking does not well serve the industry, and when it comes
to the on-going discussion of condensation control, new ideas, and concepts,
the consistent application of the principles of physics and the use of appro-
priate analytical techniques need to be embraced.
Although not included in this STP, the symposium attendees also ben-
efited from a first-day tutorial session offered by Wagdy Anis and Robert
Kudder on condensation. This primer provided the science of condensation
formation and present technologies used to control its formation. For those
without this background, this tutorial served as necessary background for
the technical presentations.
An ASTM symposium and STP are a team-effort, which warrants the rec-
ognition of those who spend much time and energy in their success. First,
recognition goes to the many unnamed reviewers who, solely to better the
industry, spent many hours reviewing and re-reviewing the submitted pa-
pers. ASTM and JTE efforts were spearheaded by Dorothy Fitzpatrick and
Susan Reilly, respectively, with able assistance by Hannah Sparks and
Christine Urso. Upon Dorothys retirement, Mary Mikolajewski ably
stepped in. Finally, special recognition goes to WJE staffer, Amber Stokes,
who assisted the Editors keep to the ambitious review and symposium
schedule, and the numerous email correspondences necessary to pull this all
together.
Bruce S. Kaskel
Wiss, Janney, Elstner Associates, Inc.
10 S. LaSalle Street, Chicago, IL
Robert J. Kudder
Raths, Raths & Johnson, Inc.
835 Midway Drive, Willowbrook, IL
viii
Reprinted from JTE, Vol. 39, No. 1
doi:10.1520/JTE102972
Available online at www.astm.org/JTE
William B. Rose1
ABSTRACT: In the late 1930s, an architect and two researchers created a

version of hygrothermal building science for the United States that focused
on moisture conditions in exterior materials during cold weather. The version
they created was partial, and it was biased: It highlighted the importance of
vapor transport, while it obscured the importance of temperature impact.
They based their argument on the prevention of condensation, yet they
failed to provide a denition of condensation sufcient for use as a perfor-
mance measure or criterion. They produced prescriptive recommendations
that later became code requirements, and these prescriptions embodied the
incomplete and biased nature of their analysis. They supported their argu-
ment with a awed and misleading analogy. They and their followers left a
legacy of consumer fear of ill-dened moisture effects in buildings and of
designers assigning excessive importance to prescriptive measures. Their
version provides inadequate preparation for the anticipated re-insulation of
millions of U.S. buildings in the years to come. This paper will provide a short
description of the hygrothermal issues involved. It will trace the development
of the condensation version by Rogers, Teesdale, and Rowley and the efforts
that followed up to 1952. It will explain the legacy and impact of this ap-
proach related to existing building re-insulation and professional practice in
design and architecture. It will propose a framework for reviewing the link
between moisture control prescriptive requirements and performance out-
comes.
KEYWORDS: condensation, moisture control, insulation
Condensation
In 1901, in the course of the design of the Minnesota State Capitol Building, the
architect Cass Gilbert was in discussion with Mr. Guastavino, a highly regarded
supplier of ceiling tiles, and a Mr. Butler, the contractor. Gilberts notes indicate
Manuscript received January 21, 2010; accepted for publication June 14, 2010; published
online August 2010.
1
Research Architect, Univ. of Illinois at Urbana-Champaign, Champaign, IL 61820.
Cite as: Rose, W. B., Insulation Draws Water, J. Test. Eval., Vol. 39, No. 1. doi:10.1520/
JTE102972.
Copyright 2011 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West
Conshohocken, PA 19428-2959.
1
2 JTE STP 1498 ON EXTERIOR BUILDING WALL SYSTEMS
I urged that great difculty might be encountered in condensation, in

which Mr. Butler agreed, and that this condensation would drip and in-
jure the work below it or form as ice and cause the upper work to heave.
Mr. Guastavino said he had considered this. I had several times, during
the conversation, mentioned the danger from condensation. I then pro-
posed thatwe should abandon the idea of an opening in the center of
this inner bell with a canopy over it, and should make it a continuous
vault of Mr. Guastavinos material, and asked Mr. Butler if he were willing
to go forward on this basis, and if he had any doubts as to Mr. Guastavi-
nos material for the inner bell. He said he had no doubts about it and was
willing to go forward with it 1.
This architectural conversation occurred more than 3 decades prior to the
appearance of building condensation as an issue of concern in the technical
engineering literature. Mr. Gilbert grasped, in a general sense, the conditions
under which frosting, melting, and paint peeling might occur and how these
conditions are associated with air ow through openings and with chilled sur-
faces. Condensation, in the sense of this discussion, was a visceral concern for
the architect, and it represented a range of possible phenomena rather than a
specic phenomenon. Mr. Gilbert did not need a scientic understanding of
condensation; he had a practical problem, which was resolved by practical
assurance from experienced collaborators. The appearance of condensation
as a vaguely dened worry about buildings predates by several decades the
appearance of technical studies related to condensation.

When insulation was introduced into wood frame houses in the late 1920s and
early 1930s, the paint began to peel. House painters often refused to paint
insulated houses 2. The painters developed a pithy expression to describe
what happens: Insulation draws moisture. The residential paint-peeling prob-
lem was assigned in 1929 to F. L. Browne, a chemist with the U.S. Forest
Products Laboratory in Madison, Wisconsin. He recognized that the paint-
peeling problem was occasional and was associated with two types of abnor-
mal conditions.
1 Type ARainwater seeping through leaky joints left by poor carpentry
work or faulty design
2 Type BMoisture originating within the building and carried by air
circulating within the hollow outside walls. When moisture laden air
comes in contact with surfaces at sufciently lower temperature, water
condenses 3.
Brownes nding that bulk water effects were the cause of most problems
Type A bears repeating in every discussion of the subject. Regarding Type B
conditions, he notes that sufciently low temperature is a precondition for
condensation. He suggests that the source of the moisture was from the interior
and air circulation was the transport mechanism.
Does insulation draw water? The Forest Products Laboratory 4 had devel-
oped the wood sorption isotherm several years earlier see Fig. 1. It shows a
ROSE, doi:10.1520/JTE102972 3
!"#$%!& ()!$*+", -!" .!!/

-"!, !!" $%&"'!!(
%!
$!
84790:-, .4/0,/0 45 ;44<
=" >?699;60,-@?699<-A 67- B
<,C-,,9 D
# <,C-,,9 D
$
% <,C-,,9 D
#!
"!
"
!& "!& #!& $!& %!& !!& '!& (!& )!& *!&
+,-.,/0 23 45 67-
FIG. 1Wood sorption isotherm from Ref 4.
somewhat-linear relationship between the equilibrium moisture content in

wood and relative humidity RH of air that surrounds the wood for tempera-
tures above freezing. Figure 2 shows the same data as in Fig. 2 but plotted as
lines of constant wood moisture content on a psychrometric chart. This repre-
sentation allows temperature control and vapor control to be viewed indepen-
dently.
RH ratio vp/svp of vapor pressure, vp, to saturation vapor pressure, svp
can be raised in two ways of courseby increasing the vapor pressure nu-
merator or by lowering the temperature denominator. So at the same vapor
pressure, a cold piece of wood will be wetter than a warm piece of wood. We
FIG. 2Lines of constant wood moisture content plotted on a psychrometric chart.

Horizontal arrows show the impact of change in temperature; vertical arrows show the
impact of change in vapor.
may expect that upon adding insulation to a wall, the exterior materials during
Winter will get cold, and by virtue of being cold, they will get wet. How wet is
a matter for analysis of course. Also, freezing events in exterior materials will
be more common and more severe. Insulation draws cold, and cold draws
wetness. So one cannot quarrel with the painters claim that insulation draws
moisture, at least on physical grounds. Insulated buildings in cold climates
have wetter cladding and sheathing materials than similar uninsulated build-
ings.
Tyler Stewart Rogers

Rogers was an architect and one of the founders of the reference work Time-
saver Standards for Architects. He wrote a seminal article on garages, as Ameri-
cans were turning from carriage houses toward the use of the automobile. In a
1936 article 5, he held the title of Director of Technical Service, though the
organization is not identied. In 1938, he was Director of Technical Publica-
tions for Owens Corning Fiberglass.
In November 1936, the opening salvo of the condensation paradigm was
provided in American Architect and Architecture, Insulation: What we know
and ought to know about it, by Rogers. His main intent was to present heat
transmission factors for building materials, which constituted new informa-
tion, never before presented to the architectural profession. He cited the need
for better knowledge regarding properties, installation techniques, testing
methods, rating methods, and amounts needed. Then he began a discussion of
moisture.
The advent of air conditioning2 is opening up still another eld for con-
structive research. Technicians know that when indoor relative humidities
are articially controlled as they should be for comfort and health there
is a theoretical dew point temperature somewhere within the exposed
walls or roof at which point air-borne moisture is condensed into water.
It is further known that vapor pressures tend to move this internal mois-
ture toward the cold side of the wall.
These facts set up a number of speculations that cannot be answered
without further research. It seems important to know how much air-
borne moisture permeates building sections of different types. It would
appear also that plaster, building paper or any other impervious curtain
on the warm side might be a sufcient barrier to prevent any measurable
accumulation of dampness. Much research is being undertaken, more is
planned for 1937. Progress is being made 5.
Rogers was setting the stage for decades of understandingsand
distortionsto come, so his words should be reviewed carefully. The following
represent distortions contained in his writing.
Although insulation was introduced about the same time as humidica-
2
This includes Wintertime humidication.
ROSE, doi:10.1520/JTE102972 5
tion air conditioning, here Rogers associated wetness with humidi-

cation rather than with insulation. The moisture increase associated
with insulation itself was not discussed.
Dewpoints have locations within the wall.
The terms condensed and condensation can be applied in the absence
of denition. Rogers saw no need to distinguish condensation phenom-
ena on sorptive and non-sorptive materials.
Wetness on the outside requires a humidied interior as the moisture
source.
Prescriptions sufcient barrier to prevent measurable accumulation of
dampness may be suggested or offered prior to the completion, publi-
cation, and discussion of research.
Rogers then wrote Preventing Condensation in Insulated Structures for
the rst issue of Architectural Record, March 1938 6. It began
Architects, owners and research technicians have observed, in recent
years, a small but growing number of buildings in which dampness or
frost has developed in walls, roofs or attic spaces. Most of these were
insulated houses, a few were winter air-conditioned. The erroneous im-
pression has spread that insulation draws water into the walls and
roofs.
Obviously, insulation is not at faultat least not alone. Nor could winter
air-conditioning, creating comparatively high and sustained relative hu-
midities for health, be charged with sole responsibility, for not all struc-
tures reporting dampness were equipped with humidiers. The need for
research became apparent.
Rogers attributed the new problems to new conditions, including humidi-
cation, reduced inltration, weathertight construction, and efcient insula-
tion, and he stated that all four are highly desirable in terms of health, com-
fort, and economy. This framing of the problem meant that other means of
moisture control needed to be adopted.
In this article Rogers, provided no operational denition of condensation.
He stated, in fact, that the problem of condensation had been fully solved; this
was even before it was ever satisfactorily dened. His position was that these
new measures must and will be adopted. He proceeded to discuss measures
necessary to mitigate the wetness associated with adoption of the new ele-
ments. The measures he proposed were the vapor barrier and attic ventilation.
Rogers stated
Architects may avoid all technicalities in explaining this new vapor bar-
rier principle by using some parallel situation such as that shown in gure
13. Here are two basins into which water is running at the same rate. The
basin at the left has an outlet larger than the supply. In this no water
3
Figure 3 in this paper.
FIG. 3Flawed analogy comparing diffusion transport kinetic to bulk ow dynamic

from Ref 6.
accumulates. The one on the right has an outlet restricted in size to less
that that of the inow. Here water accumulates until it spills over the
sides.
So with the wall sections shown below these basins. The room between is
indicated as being warm and humid. In the wall at the left there is a vapor
barrier not completely perfect in its stoppage of vapor movement. How-
ever, it checks most of the vapor, and what little remains can pass out
through the colder side of the wall with little difculty. This wall shows no
accumulation of vapor.
This description is fundamentally awed. The funnel and faucet analogy
describes a dynamic system where the faucet water feeds all of the materials
along its path. In kinetic moisture diffusion, the entire surrounding air and
materials provide moisture to the materials, not just a high source at a dis-
tance. Rogers said that by checking interior moisture at the warm side, the
wall shows no accumulation of vapor. However, vapor will accumulate in
materials that move to low temperature, and the capacity for barriers to miti-
gate that accumulation is limited. Rogers analogy is captivatingand mislead-
ing. Arguing that under diffusion all of the water comes from the high vapor
pressure side is equivalent to claiming that all of the heat comes from the
high temperature side under heat conduction.
The remainder of the article dealt with practical matters of placing vapor
barriers and providing attic ventilation. These included two pages from Time-
saver Standards on Heat Transmission, listing coefcients of heat transmission
of common building materials, and two pages on Preventing Condensation in
Insulated Structures. These gures were widely reproduced in guidance litera-
ture that followed.
ROSE, doi:10.1520/JTE102972 7
FIG. 4Larry V. Teesdale, U.S. Forest Products Laboratory.
L. V. Teesdale
The rst of two researchers Rogers referenced was L. Larry V. Teesdale of the
Forest Products Laboratory see Fig. 4. His paper, Condensation in walls and
attics 1937 7, began
Condensation or moisture accumulation within walls and in attics or roof
spaces has become a subject of considerable concern to many home own-
ers and prospective builders, especially in the states north of the Ohio
River. There have been so many cases in recent years that any prospec-
tive builder may hear about ice in attics, stained ceilings and side walls,
plaster becoming loose, ruined decorations, decayed side wall, roof, studs,
and sheathing, oors that have bulged up, outside paint failures, and nu-
merous other manifestations of moisture resulting from condensation.
Obviously the question arises as to why we hear so much more about this
condition now than we used to just a few years ago. The answer is rela-
tively simple. During the last few years there has been a marked tendency
on the part of the architects, builders and home owners to improve homes
both new and old with the idea of increasing the comfort of the occupants
and decreasing operating expenses. Prominent among these improve-
ments are the increasing use of storm sash, insulation, weather strips,
calking around windows and doors, and other means of decreasing heat
loss and wind inltration. Because of the tighter construction the normal
humidity or vapor pressure within a house so constructed is higher than
in houses less tightly constructed. In addition, as a health and comfort
measure the normal humidity is usually augmented by evaporating water
or some other means of winter air conditioning. Improvements that add
to comfort and health are worth while and should not be discouraged, but
it so happens that they introduce the unanticipated moisture problem just
described.
In these opening paragraphs, Teesdale, like Rogers, emphasized how im-

provements have led to tighter envelopes and thereby to higher indoor vapor
pressures. Teesdale, again like Rogers, introduced the term condensation freely
and qualitatively without presenting a denition. The focus of his work was on
the moisture effect of added insulation, and he identied a moisture effect that
is independent of indoor vapor pressures, namely the wetting effect that comes
with colder material temperatures.
Comparing gures 1 and 24, it is at once evident that, within the stud
space, the temperature gradients are much steeper in gure 2 than in
gure 1, and that the respective sheathing temperatures are much lower
in gure 2 than in gure 1. This results from the addition of insulation in
gure 2. Because of the lower sheathing temperatures5 condensation will
occur on the sheathing with lower room humidities when insulation is
used than when it is not used. Conditions within the walls are actually
more complicated than the drawings and examples indicate, because they
are not static.
Now we come to one of the most bafing paragraphs in all of building
science.
There are a number of types and kinds of insulation on the market and
the potential buyer often hears that certain types draw water and be-
come wet. This is not true. Such insulation, because of its efciency in
reducing heat loss, lowers the temperatures within the wall and thus sets
up the condition that increases the amount of moisture that may accumu-
late. Once understanding the conditions that cause the moisture it is also
possible to provide means of prevention as discussed later.
A close look reveals that Teesdale contradicted himself herecompare the
phrases become wet and moisture that may accumulate. Insulation is ef-
cient in reducing heat loss. So temperatures within the wallactually the outer
materialsare lowered. Lowering those temperatures sets up the condition
that increases the amount of accumulated moisture in the cold materials. Tees-
dales objection to insulation drawing water cannot be an objection to the fact
of moisture accumulation; Teesdale agrees with that. Why then did he object to
the simplication that drawing water expresses? Is this an early example of
spin and counterspin regarding public opinion? Teesdale suggests an answer as
he moved toward a summary:
Moisture accumulation within a wall like those illustrated in gures 1 and
2 is affected by ve factors:
1 Outside temperature and humidity.
2 Efciency of the insulation.
3 Inside atmosphere temperature and humidity.
4 Resistance of the outer wall to vapor movement.
4
Figure 5 in this paper.
5
Emphasis added.
FIG. 5Figures from Ref 7, showing theoretical temperature proles in uninsulated and insulated assemblies.
ROSE, doi:10.1520/JTE102972 9
FIG. 6Figures from Ref 8, showing test data temperature proles in uninsulated and
insulated assemblies, compared to theoretical Fig. 7.
5 Resistance of the inside wall to vapor movement.

As the outside temperature and humidity cannot be controlled, and as
insulation adds to comfort, health and fuel economy, methods of preven-
tion are limited to the three other factors.
Teesdale expressed here an insulation imperative. According to him, we
must insulate and, given that, we must accept the consequences of colder and
wetter exteriors,; we must settle for whatever moisture improvements are pos-
sible by manipulating indoor humidity and access of that humidity to the ex-
terior materialsvapor barriers. Promoting the use of insulation is ne, of
course. What is not ne is that Teesdale appears to be willing to lend a voice
that insulation does not draw water, while demonstrating just the opposite,
very clearly and succinctly.
A few months later, an article without author attribution appeared in The
Architectural Forum, which contained language and drawings so similar to
Teesdales report that his authorship is unmistakable 8. This article contained
much of the material from his previous report, but this time it included mea-
sured values, shown in prole charts, where his previous charts were only the-
oretical see Fig. 6. The article explained how these charts are made and inter-
preted.
two words of caution with their interpretation are worth repeating here:
The rst of these is that the theoretical dewpoints shown are in all cases
room dewpoints. This assumes that the vapor pressure within the stud
space will be the same as that within the house, a condition which will
never actually exist. Actually the vapor pressure and dewpoint within the
stud space will be somewhat lower than that within the house. How much
lower will depend on how much the inner and outer wall surfaces resist
the passage of vapor; if the inner surface passes vapor readily and the
outer surface is relatively impervious, the condition will be substantially
as shown; if, however, the surface is relatively impervious, or the outer
ROSE, doi:10.1520/JTE102972 11
part of the wall passes vapor easily, actual dewpoints will be lower than
those shown.
The second word of caution is in regard to the location of the dewpoints,
which are in some cases shown within the insulation. This is purely a
matter of diagrammatic convenience and does not mean that condensa-
tion will actually take place at this point. Actually, the moisture in such
cases will collect on the nearest cold surfacethe inside of the wood
sheathing. This is because the condensation of moisture sets up a relative
vapor vacuum which draws4 vapor from the surrounding air. Whenever
condensation is actually taking place, the actual vapor pressure within the
stud space will be equal to that for saturated air at the temperature of the
inside of the sheathing, as indicated on the test data diagrams, and
condensation will be possible only at this point.
Bravo, Larry. Cold sheathing draws moisture; indeed it operates like a
vapor vacuum, sucking up vapor from the air that surrounds it. This is the one
instance of critical challenge to the developing paradigm, and it was buried in
an anonymous article, out of the mainstream. Note that the measured data
contains cavity vapor pressure values. The vapor pressure values do not corre-
spond to vapor permeances of the inner and outer skins: For one thing, those
values are not known. For another, the permeances are presumed to be the
same in the two cases shown in Fig. 6, yet the measured vapor pressure in the
cavity is quite different. The explanation under the second word of caution
correctly explains the measured data. And this is the explanation that claims
that cold materials draw moisture.
Teesdales gures Fig. 6 merit close study. They show clearly that what
determines the actual vapor pressure in a cavity is the temperature of the
sheathing, far more than the vapor permeance values of the assembly materi-
als. In his two examples, the material permeances are essentially the same,
while the cavity pressures differ greatly. This is consistent with how the Ameri-
can Society of Heating, Refrigerating and Air-Conditioning Engineers
ASHRAE prole method is conducted.6 That method requires that cavity
vapor pressure in excess of the saturation vapor pressure at the temperature of
the colder materials be set to a value equal to that saturation vapor pressure,
and net accumulation rates are then calculated. Many users who are not famil-
iar with the method may simply nd vapor pressure in excess of saturation
vapor pressure and suggest condensation occurs. This mistaken but widespread
approach has no support in ASHRAE Handbook Fundamentals 2009 9 and is
inconsistent with physical ndings including those of Teesdale.
Teesdale also pointed out, in the second word of caution, that language
referring to the location of dewpoint or where you reach dewpoint repre-
sents an interpretation of a graphic device and does not reect a representation
of actual conditions. There is no where regarding dewpoints.
This article has a curious introduction clearly not by Teesdale, most likely
6
ASHRAE 2009 Handbook Fundamentals. Chapter 27, Heat, Air and Moisture Control in
Building AssembliesExamples, Examples 9 and 10.
by Rogers apparently written to address the concerns of the architecture au-

dience.
Already validated by these reports7 are the following generalizations:
1 Condensation is a thoroughly predictable, understandable phenom-
enon which need hold no terrors4 for the architect or builder who mas-
ters its fundamentals.
2 It does not result from the use of any particular insulating material, nor
does it necessarily affect adversely one type of material more than an-
other.
3 Precautions against it are effective and relatively so inexpensive as to
seem desirable in any event.
Terrors? What should be apparent at this point is that there are two parallel
domains of the world of condensation: One, represented by Rogers and the
architects, which addresses visceral worries regarding a non-specic range of
moisture effects; the other represented by Teesdale and the emerging science/
engineering community, in which the puzzles offered by condensation phenom-
ena provide a fascinating glimpse into physical processes calling out for expla-
nation. As we will see, terrors come to play a role when vapor barriers are
marketed in 1950 using slogans such as The Menace of Moisture and War
Against Water.
There is a curious contrast between the statement here that
Condensationdoes not result from the use of any particular insulating ma-
terial and previous statements by both Rogers and Teesdale to the effect that
some types and kinds of insulation materials draw water. The type or kind
argument distracts from the more fundamental matter of insulation per se
leading to wetness conditions.
American Society of Heating and Ventilating Engineers Now ASHRAE

In 1937, Technical Advisory Committee IF-23Insulation was formed at
American Society of Heating and Ventilating Engineers ASHVE. The rst re-
port of this committee was presented at the 1937 annual meeting 10. That
report stated
The committee further gave consideration to the following questions con-
cerning the usage of insulation, but deferred until a later date in making
recommendations concerning the need for study of any of the particular
subjects mentioned or concerning their order of importance.
1 The effect of regain moisture on the conductivity of insulating mate-
rials. Quotes are from the original.
2 The effect of condensed moisture on the conductivity of insulating ma-
terials in a structure.
The report then listed 19 other subjects, none of which deal with moisture.
What is of interest here is the explicit distinction between regain moisture and
7
Rogers was referring to the work of Teesdale and Rowley.
ROSE, doi:10.1520/JTE102972 13
condensed moisture. The exact denitions were not given and are not known
because the distinction dropped from use. But it is tantalizing to speculate as to
the intended distinction. Thermal wetting versus source humidity? Sorbed
moisture versus hard-surface condensation? Assuming these are distinct wet-
ting processes and further assuming that vapor barriers restrict condensed
moisture, had this committee described a wetting process that does not lend
itself to vapor barrier control? We are unable to answer at this point.
Frank Rowley
As Professor of Mechanical Engineering at the University of Minnesota, Rowley

Fig. 7 had worked with National Mineral Wool Association funding to deter-
mine values for thermal resistance of materials. His work and that of his co-
researchers and students came from the laboratory and from theory.
In A Theory Covering the Transfer of Vapor Through Materials, ASHVE
July 1939 11, he posited that
For convenience, it has often been assumed that the laws for vapor trans-
mission are similar in form to those governing the ow of heat through
the walls of a building, and that coefcients of vapor transmittance may
be developed for materials or combinations of materials in the same man-
ner as coefcients of heat transmission. Before accepting a complete
analogy between the two problems an analysis should be made to deter-
mine those elements which are similar and those which may be
conicting.
Note the argument given for assuming at the outset that vapor moves by
diffusion: For convenience. To Rowleys credit, he did not consider that a suf-
ciently strong argument without further validation.
Rowley then sought to dene condensation. He distinguished non-
hygroscopic materials from hygroscopic. Dening condensation for non-
hygroscopic materials was, of course, quite simple. Then he distinguished non-
permeable from permeable hygroscopic materials. For permeable materials, he
stated, without demonstration, that vapor transport would be by diffusion only,
and frost or ice may form as condensation within the material if, at any point,
dewpoint exceeds sensible temperature. His discussion of non-permeable hy-
groscopic materials introduced capillary transport. His discussion was long,
involved, and at times speculative. At no point in this discussion did he state
what constituted condensation. In other words, this effort to provide a theoret-
ical underpinning for preventing condensation was unable to provide an op-
erational denition of condensation for capillary materials e.g., wood,
masonry.8 As a consequence, there was no metric available to determine per-
8
ASHRAE Handbook Applications 2013, Chapter 43, Building Envelopes, contains this
approved wording: Moisture condensation is the change in phase from vapor to liquid
water. Condensation occurs typically on materials such as glass or metal that are not
porous or hygroscopic and on capillary porous materials that are capillary saturated.
Use of the term condensation to refer to change in phase between vapor and bound
FIG. 7Frank Rowley, professor of Mechanical Engineering, University of Minnesota.
formance outcomes of prescriptive measures for normal building materials.

Rowley and his colleagues conducted experiments of wall and attic con-
struction in 1938 12. They made use of a refrigerated room capable of reach-
ing 20 F in which they placed test assemblies. They developed an ingenious
method of working around their inability to dene condensation for capillary
materials. The assemblies were designed for rapid dismantling so that an alu-
minum plate, located at the inside of the sheathing, could be removed and
weighed to determine frost accumulation or condensation at that location.
Thus, the appearance of frost on an aluminum plate became a surrogate for
condensation on wood or masonry.
The interior space was maintained at 70 F and 40 % RH. The assemblies
were all 2 4 studs, metal lath and plaster at the interior, Ponderosa pine ship-
lap sheathing, building paper, and redwood siding, with 3 5/8 mineral wool
between studs. To summarize their test facility ndings, see Table 1.
We may note the following from Rowleys data.
Accumulation is largely a product of sheathing temperature. Recall that
the inside of the sheathing temperature provided by Teesdale, with out-
door temperature of 20 F, was 1.6 F with insulation and 38.5 F
without insulation.
water in capillary or open porous materials is discouraged.

TABLE 1Temperature and frost accumulation condensation results from Rowley 12.
Outdoor Temperature, F
19.5 10 19.5 10
Inside Surface Condensation on Sheathing,
of Sheathing Temperature, F grams/ ft2 / 24 h
No paint, no vapor barrier 0.2 22.8 2.15 1.41
2 coats seal coat paint 0.8 23.6 0.20 0.00
2 coats white at paint 1.5 20.6 0.24 0.00
Glossy asphalt impregnated sheathing paper 2.0 20.5 0.07 0.00
303030 duplex paper 3.2 20.9 0.25 0.00
Asphalt felt paper 2.2 21.3 0.52 0.18
Duplex crepe paper 1.8 21.6 0.09 0.00
ROSE, doi:10.1520/JTE102972 15
Paint is an excellent vapor barrier, comparable to the protection pro-

vided by the papers available at the time.
Since the sheathing is below freezing in this test, the test methodfrost
on an aluminum plate at steady staterepresents actual conditions
since moisture sorption would be inactive. Under actual conditions, a
warming spell would melt the frost and allow sorption.
To gain a sense of the impact of sorption, we may calculate thus 1 ft2 of
3/4 in. thick Ponderosa pine weighs 1.0 kg. The safe moisture range
may be from 10 % to 25 % moisture content, or 150 g accumulation
within that range. So, with no vapor protection, the safe duration of a
cold spell would be 70 days at 20 F 150 g/2.15 g/sf/24 h. It should be
obvious that a painted wall can withstand a cold spell ten times longer.
The conditions at the inside of the sheathing are not a strong indicator
of the wetness at the layer of wood beneath the paint. There was a shift
in the site of concern from the paint substrate to the inside of the
sheathing, with no justication provided.
These results were the only published measured results regarding the im-
pact of vapor barriers and attic ventilation prior to the promulgation of vapor
barriers as a code requirement in 1942 see Housing and Home Finance Agency
HHFA below. These measurements, under the steady-state conditions Row-
ley imposed, lent support to the prescriptive measures that were on the table
vapor barriers and attic ventilation.
They could hardly do otherwise. The choice of 20 F for outdoor condi-
tions would naturally highlight the occurrence of condensation, in the case
where even small amounts of moisture are diffused or leaked onto the test
plate. Rowleys research was congured to nd performance thresholds where
complying structures fall on the desirable side and non-complying structures
show condensation. A research design that was not constrained to the prescrip-
tive measures suggested might well have been structured differently, and might
have reached other conclusions. The research appears to be post hoc; it appears
to rationalize a given set of design decisions rather than initiate design ap-
proaches that a fairer research design would suggest.
Furthermore, the vapor barrier research provided a clear demonstration of
the effectiveness of paint as a vapor barrier. If Rowleys results are correct, and
we may assume they are, then the use of paint appears equal in performance to
the barrier materials he studied. This should have been cause to suspend the
vapor barrier momentum. Indeed, were not the troubled houses painted at the
interior? It did not.
Rowleys co-researcher, Lund described their Minnesota research in 1952
13, saying, There have been many erroneous statements made that insulation
is one of the primary causes of structural condensation. That is as close as the
University of Minnesota team came to saying that insulation draws water.
Housing and Home Finance Agency
Throughout the 1940s, the Housing and Home Finance Agency had carried
through its mandate to facilitate low-cost home construction during and after
ROSE, doi:10.1520/JTE102972 17
World War II. It managed the FHA, and, in January 1942, it published Mini-
mum Property Requirements that contained the rst numerical values for
vapor barrier permeance, attic ventilation, and crawl space ventilation 14.
HHFA Technical bulletins from 1946 through the mid 1950s contain some of
the most valuable research results and guidance in the eld. The HHFA did a
remarkable amount of work in the last half of the 1940s, primarily under Ralph
Britton. He followed up on the prescriptions of the FHA January 1942 require-
ments which he may have authored, with several research campaigns.
HHFA proposed to write a booklet on Condensation Control in Dwelling
Construction, with drawings to illustrate vapor barrier placement and attic
ventilation, in easy-to-read three dimensional drawings. The 73-page booklet
appeared in August 1949 15. Prior to its issuance, a conference was held to
review the proposed content 16. Minutes from that conference were taken and
mimeographed.
Leonard Haeger from the National Association of Homebuilders chaired
the meeting. He began with
severe winters seem to bring out the most complaints and the milder
winters produce lesser troubles. The most common of these problems
involve exterior paint failures and damage to interior wall and ceiling
decorations. More serious damage may occur through decay or corrosion.
The problems involving thermal insulation and vapor barriers have been
studied in a number of public institutions such as the University of Min-
nesota, Pennsylvania State College, National Bureau of Standards and
Forest Products Laboratory. A considerable amount of work has been
done in private laboratories of producers of thermal insulation, and of
various types of papers and felt. These researches carried on over a period
of 10 to 15 years have led technicians generally to agree that it is desirable
to use a good vapor barrier on the warm side of exterior walls and
ceilings.
It is hoped that our new publication will explain thoroughly and in a
simple way that thermal insulation is a very valuable element in construc-
tion and that it does not, as many people believe, draw water. Moisture
problems develop in cold weather inside heated dwellings because greater
amounts of water can be held in the enclosed air space.
The conference began with the attendees showing slides of buildings with
moisture problems. The moisture problems were almost entirely wood frame
buildings with paint peeling, mostly located in the Madison, WI, area. A total of
45 buildings was shown. There was no uniformity in house description. But a
review of the descriptions shows that of the 45 houses, 28 of them were claimed
to have crawl spaces. The foundation types in the others were not mentioned
and may also have been crawl spaces. In many of the cases, the crawl space was
noted as the likely source of wetness. The fact that this foundation type was
called out in so many instances suggests that the participants knew of crawl
spaces as a major element in the wetness equation.
We may note that the justication for widespread, indeed nationwide,
adoption of the vapor barrier was based on case studies of houses, insulated or
not, largely centered around Madison, WI, with paint on clapboard or plywood,
largely on crawl spaces. The limited scope of case studies undermines the ne-
cessity of a vapor barrier requirement. These conferences, seeking to push for-
ward a regulation, based their conclusions on a most narrow range of case
studies. Nevertheless, the vapor barrier plan was applied broadly. Newkirk, for
example, a vapor permeance specialist from the American Reenforced Paper
Co. recommended the use of a vapor barrier in all climates.
Most of the conference discussion followed the intended thrust, to provide
support for a vapor barrier recommendation. Elmer Queer of Penn State was a
strong supporter of advancing the prescriptive measures. However, he did ex-
press the importance he attached to temperature as a determinant of outdoor
wetness.
Mr. Chairman, I would like to comment on the point about the effect of
outside humidities. Really, the humidity has very little effect on the vapor
transmission. The temperature is still actually the controlling factor, be-
cause in the majority of cases you have a condition approaching 100 per-
cent humidity on the outside, and even if you dont have 100 per cent at
these low temperatures, your vapor pressure differential will be virtually
the same; it varies very little.9
Teesdale also pointed to the importance of exterior temperature and to the
fact that sub-freezing behavior is quite different from warmer behavior.
In the sills of these highly insulated walls where the condensation was
made at the top, the fact that the moisture content was relatively low does
not mean that under variable state conditions or under actual conditions
of test that we wouldnt have had moisture contents way up above 35 or
40. They were low down there because the moisture never had a chance to
get down, because it was held there in frozen conditions.10
The conference did not take up the points that Queer and Teesdale ex-
pressed.
The conference ran overtime, so critical discussion was limited. One nal
matter needed to be settled howeverunits to be used for measurement of
vapor barrier permeance. It was agreed to proceed with grains per hour-square
foot-inch of mercury vapor pressure.11 They needed a name for this unit. It was
proposed to be a Rowley, a Britton, or a Queer, though the nal name selection
was delayed until later.
The HHFA brochure was printed in 1949 and distributed widely. It went
into eight printings. Figure 8 shows a drawing from that brochure.
9
HHFA 1948, p. 64.
10
P.115. Teesdale then offered an interesting observation: And, incidentally, some of the
walls, I believe, showed warmer inside surface temperature when we had that frozen
condition, when we had that frost made, than they did without the frost. The frost served
as an insulator which it wouldnt under variable state conditions.
11
Rowley had used grams rather than grains. But the psychrometric chart at the time
used grains of water in the expression of humidity ratio.
ROSE, doi:10.1520/JTE102972 19
FIG. 8Vapor barrier installation illustration from Condensation Control in Dwelling

Construction.
FIG. 9Brochure covers from the War Against Water campaign of the NPVA.
The War Against Water Campaign

We may recall that the incompatibility that precipitated the entire condensa-
tion effort was between insulation and exterior paint nishes. The insulation
industry supported Rogers and funded Rowley. And the paint industry? Their
work became apparent in 1950, with the rollout of the War Against Water
campaign by the National Paint and Varnish Association NPVA see Fig. 9.
This campaign consisted of brochures, press releases, reproductions of articles,
and slide shows. By 1952, 300 000 of the brochures had been distributed, and
the slide show had been seen by 50 000 people.12 The aim was to demonstrate
the need to comply with the new prescriptive regulations.
From Menace of Moisture brochure:
Oh, its a battle, all righta constant struggle between mankind and mois-
ture. The price of victory is eternal vigilance. Its up to you, of course, as it
is to every home owner.
First do everything possible to keep the moisture pressure within your
home below the danger point by providing adequate ventilation at all
times. As an added safeguard for the walls of your home, retard the inva-
sion of excess moisture from the room side. As allies you have vapor
barriers and water-repellent paint.
From the War Against Water brochure:
They seem innocent enough, these three pools of moisture: the milk from
the babys bottle, the steam from the shower, the vapor rising from the
whistling tea kettle. But are they? Oh, no! Like the mild-mannered mur-
deresses in Arsenic and Old Lace theyre up to no good!
12
BRAB 1952, p. 103.
ROSE, doi:10.1520/JTE102972 21
Where do they go from here? Believe it or not, they have an engagement.

At the dew-pointif you please.
Whats a dewpoint?
In a wall of a structure its the coolest surface. Here vapors and steam
condense, or turn into drops of water. Scientists must have been feeling
poetic when they named this point for the drops of dew that look like
diamonds on a cob-web at dawn. If you could cut a cross section through
the walls of your houseif you could get a clear view of this procedure,
you wouldnt feel poetic. Youd feel like screaming Dont! and christening
it the dont-point! Youd demand that moisture cease its villainy.
Up to now the moisture in the walls of your house can have stained plas-
ter, rotted wood and rusted metal. After the various vapors reach the
dew-point, theyre ready to proceed with concentrated devilment. They
can even force their way through outer siding, masonry and trim. They
can push the paint loose from its moorings and can cause it to blister,
then peel!
These brochures capture all of the hallmark misunderstandings and distor-
tions of the condensation paradigm, including 1 dewpoint has a where, 2
wetness outside comes from inside, 3 the condensation discussion can pro-
ceed in the absence of a technical operational denition, and 4 prescription
compliance is necessary and sufcient. But it is the tone of the brochures that
is most compelling, coming as they did at the height of the McCarthy era in the
United States. The concerns and worries about condensation expressed by Gil-
bert and others have at this point morphed into a fear of moisture. The NPVA
pushed the on-button for fear of moisture. There is no off-button for fear,
making it a marketers dream. Fears call for protection. Protection is provided
by prescriptive measures visibly employed. Science, on the other hand, can do
no more than estimate with probability the likelihood of moisture perfor-
mance. Science at the time lacked even a working denition of condensation.
By instituting a subculture of fear note the widespread use of the term trap
for moisture, a strong advantage is handed to those addressing the problem
prescriptively rather than with performance estimates.
Moreover, the NPVA marketed these prescriptions to homeowners, in addi-
tion to the design and construction industries.13 The impact of this is to under-
mine the normal relation between professionals with owners, wherein many
decisions about envelope components are judgment calls. Instead, there is a
perceived obligation for designers, contractors, and owners to meet a simple
compliance requirement providing a barrier against a bogeyman.
Wrapup, BRAB Condensation Conference 1952

The proceedings of the Building Research Advisory Board BRAB conference
Condensation Control in Buildings as related to Paints, Papers and Insulating
13
Compare this to the marketing of prescription pharmaceuticals to the public via tele-
vision and its impact on doctor-patient relations.
Materials, February 1952 2 may be read as a victory lap for the condensation
paradigm and its authors. By this time, the BOCA model building code had
been written, and it incorporated a requirement for a 1 perm vapor barrier in
all buildings. Insulation materials were sold with reminders of the importance
of vapor barriers. HHFA and NPVA had success stories to tell about their cam-
paigns. The prescriptive measures were beginning to nd their way into prac-
tice. The keynote presentation was by Rogers of Owens Corning. He gave some
background.
It is never in fashion to blame ourselves, of course. It is always some
other Joe who caused the trouble. So paint failures were at rst blamed on
insulation and condensation; and condensation was itself blamed on in-
sulation, until the insulation industry, in self defense, had to undertake
research to establish its innocence.
This merits a pause. The insulation industry was being blamed for conden-
sation, and it reacted by undertaking research, in self defense, to establish its
innocence. Decades later, we may remark on the frankness of such a disclosure,
but Rogers presentation and its reception all occurred in a more light-hearted,
and less litigious, era. Rogers continued,
While this research and similar work by the paint industry was going on,
there was a great deal of buck-passing. The insulation men blamed the
paints or the wet lumber and some painters retaliated by refusing to paint
an insulated house. Then the building paper manufacturers got caught in
the middle; their new sheathing papers were blamed for causing conden-
sation instead of shielding a building from dampness. The foils were soon
in the ring with the papers, while architects, builders, building owners
and the general public watched this battle royal and wondered if any of
the ghters was worth betting on.
At the top level of course there was no battle; the causes and prevention of
condensation were well understood after Mr. Teesdale and Prof. Rowley
both of whom will participate in this conference almost simultaneously
published their ndings in the winter of 193738. The paint industry
began to tell its members the facts, the better manufacturers of insulating
materials began to educate their people on the same facts. But even after
a decade of such effort, the layman is still confused.
The program turned to Haeger of the National Association of Home Build-
ers: I suppose in the beginning we should have a denition of condensation,
and to practical men condensation is what you nd on a high ball glass at 5:30
in the afternoon. He went on to describe paint failures, rotting lumber, corrod-
ing steel, and frost on plywood sheathing, thereby contributing to the non-
specicity of condensation denition that consistently characterized the discus-
sion. The non-specicity continued through the course of the conference, with
attendees all contributing war stories of varied types and causes.
The conference proceedings make for highly entertaining reading. Curling
shingles on a roof look like a hen with her back to the wind. Moore described
how a Mrs. Smith complained about condensation to the Secretary of Com-
ROSE, doi:10.1520/JTE102972 23
merce and this prompted the funding for the entire Penn State effort. Britton,
who managed the Penn State research, noted that We were appreciative of
her cooperation. The house had no sheathing, it had no oor in the cellar, it
had a tin roof. It didnt have anything it should have had. Moore from Rey-
nolds Aluminum used an analogy to describe marketing the vapor barrier.
I would suggest when you get home you ask your wife as to whether or
not she uses household aluminum foil. You will nd that two years ago
she had never heard of household aluminum foil. However, I think it is
quite analogous, to realize that in a period of less than twelve months,
through national advertising, through local advertising and through point
of purchase advertising, a product was created that it will take a great deal
of promotional work to kill.
The papers presented at the conference, for the most part, provided sum-
maries of work to date. There were a lot of loose ends noted, but the primary
work of implementing a radical change in building practice was underway and
unstoppable.
Rogers provided the summary of the conference.
My nal recommendation is, lets dare stick our collective necks and put
down our best opinions, based upon technical background, as the thing to
do. State it simply: this is what we believe you should do now. And then
have the courage to go out a year hence, or six months hence if we need
to, and say, I have learned a little better, so now do it this way. I would
rather see the lay-public, the building public, know what we now here
today, in the best way we can express it, than to wait until tomorrow or
the day after to nd a few more gaps lled in.
Also in 1952, Hutcheon of the National Research Council of Canada con-
ducted research on Quonset huts in northern regions and came, for the rst
time, to the conclusion that airow explains the occurrence of condensation far
better than diffusion 17,18. Once the condensation paradigm became insti-
tuted in U.S. construction, with its prescriptions for vapor barriers and attic
ventilation, challenges such as those from Canada were not at all handled with
exibility, as Rogers suggested. Because the prescriptive requirements arrived
in place with little science and no performance criteria, future challenges to the
technical requirements become quite difcult. Moore was correct; it would be-
come very difcult to kill the vapor barrier.
What Is Next?
The authors of the condensation paradigm created a framework, a way of ana-

lyzing moisture conditions in buildings that was partial, and therefore awed.
As it promoted vapor control, it masked an important physical principlehow
materials at cold temperatures are wetted, and how, once wetted, the possibili-
ties for vapor control mitigation are severely limited. So what falls to the cur-
rent generation to do?
First, it is essential to recover balance in presenting wetting phenomena,
giving temperature the importance it deserves. Recall Fig. 2. This is critically

important as the United States undertakes an effort to add insulation to its
existing structures. The exterior materials will be wetted by the addition of
insulation. In a well-insulated building, the exterior materials will have a wet-
ness that approaches the wetness of unheated buildings such as garages, out-
buildings and unheated portions of buildings such as porches, overhangs, and
parapets. This re-insulation effort is critical, and exterior wetting is inevitable.
In light of this, building professionals must be equipped to estimate the wet-
ting, predict the long-term impacts of the wetting, and present the tradeoffs
with technical expertise. As part of this effort, building professionals must rec-
ognize that building owners are likely to be unsettled by the prospect of added
wetness, and will respond with the fears, worries, and concerns that have been
planted in them over the previous decades. There is a sociology of moisture
performance that is just as important as the thermodynamics of moisture per-
formance. Professionals must be equipped to deal with the human side of mois-
ture and its effects. Professionals must somehow nd the off-button for mois-
ture fears.
Durability in exterior materials is a function of their wetness but may also
be a function of water effects at sub-freezing conditionsthe duration, cycling,
and severity of frost and ice formation. In this case, temperature control has
much more impact on durability than vapor control.
Second, architects and design professionals must not hide from poor build-
ing performance behind prescriptive requirements. Owners and the public ex-
pect and deserve building performance, and that is what professionals should
deliver. Prescriptive measures are intended as shortcuts to performance deliv-
ery. And they facilitate commerce. Someone must police the link between pre-
scriptive requirements and the performance outcomes. If prescriptive require-
ments fail to deliver performance, they must be abandoned. However,
prescriptive requirements have stakeholders, namely, the architects and design-
ers who cannot afford to wait around to see if their building performs as de-
signed. We might expect one of Rogers battle royals if the moisture control
prescriptive requirements were withdrawn or abandoned. There is a principle
of professional ethics at work here: Professionals must put the interest of the
client and public ahead of their personal or business interest. Client/public
benets from performance outcomes and designer/architect benets from pre-
scriptive shortcuts. If the link is broken and the benet that a prescription
offers accrues to the professional e.g., reduced exposure to litigation while the
performance outcome benet fails to accrue to the client/public, then the pro-
fessional may have crossed the line of unethical behavior.
Third, what is to be done about the prescriptive requirements currently on
the books? I propose a four-part test that a prescriptive requirement should
pass see Figs. 9 and 10.
Is it critical? That is, does it contribute materially to health safety and
welfareto life safety? At this point, I can only offer an opinion, that
moisture control in buildings does not meet the same level of criticality
that is met by structural, re, and sanitary elements in the building
codes. Health effects of dampness were studied in IOM 2004, and the
links between indoor humidity and health were, at most, associative
ROSE, doi:10.1520/JTE102972 25
0!)*$!"5#"6) 6%51! 8%!!")! $12) !)9-"!):)/#'
/1
!"#"$%&' 3"#42!%3
+)*
2)5)/2* /1
0)!.1!:%/$) !)9-"!):)/# ()$)**%!+' 3"#42!%3
+)*
2)5)/2* /1
0)!.1!:%/$) !)9-"!):)/# ,-.."$")/#' 3"#42!%3
+)*
/1
01&"$)2' 3"#42!%3
+)*
;<
FIG. 10Decision tree for prescriptive code requirements.
rather than causal. Like the moisture researchers, ever since the 1930s
we are inclined to assign criticality to the matters that are the focus of
our personal careers. But, to step back for a moment, is it really such a
big deal?
Is it necessary? That is, would we expect buildings that do not comply to
be essentially bad performers? Given that the vapor barrier was in-
vented in 1937, we have many examples of buildings without vapor
barriers, though of course it may be argued they are largely without
insulation. The need for a vapor barrier on the warm side of a walk-in
cooler is not in dispute. The need for a vapor barrier on the warm side of
a building is very much in dispute, to say nothing of the disputes sur-
rounding what is the warm side. Do buildings with foam insulation need
additional measures? Do all buildings not already have one or two or
many vapor barriers if we dene a vapor barrier as a material with a
permeance 1 perm ? Why is the vapor barrier such a fetish in the
United States and so insignicant elsewhere? The researchers we have
studied in this paper make the case that the vapor barrier is desirable.
There is quite a reach between what is desirable and what is necessary.
Is it sufcient? That is, are we assured that compliance with the pre-
scriptive requirement will result in satisfactory performance? The re-
searchers we studied here made no use of sufciency criteria, so they
had no means to answer this question. They never dened condensation.
Their argument seemed to be by default: Insulation makes exterior ma-
terials wet, vapor barriers keep wetted materials from becoming wetter
still, it is all we can do so we should do it, and the sufciency question is
moot. What does sufciency mean in the face of an imperative? Pres-
ently, we are equipped to make sufciency estimates using ASHRAE 160
19. It has limitations of course, but it represents an effort to ll what
has been a blind spot for the last 60 yearswhat performance criteria
are at work?
Is it policed? That is, is there an effective mechanism in place to revisit
the questions of criticality, necessity, and sufciency? This does not

meanare they enforced? It is presumed that code ofcials administer
code enforcement. Prescriptive shortcuts are predicated on assump-
tions about what delivers good performance. The assumptions never
strengthen over time; they weaken by necessity. And at some point, they
must no longer deliver performance and must be withdrawn or aban-
doned. Rogers said as much in his closing remarks. Prescriptive require-
ments have a shelf life. Unfortunately, the moisture control prescrip-
tions discussed here were put in place absent a well-dened set of
conditions from which their necessity and sufciency would be de-
duced. Who has the benet of the doubt? I would argue that it is those
who defend the requirements that must prove they deliver performance
rather than those who challenge them.
The current situation bears a strong resemblance to the situation in 1937.
An insulation imperative is upon us, and the moisture consequences must be
addressed. Back then, the researchers sought to obscure the growing realiza-
tion that painted wood siding simply could not withstand the chilling effect of
insulation without peeling. The industry responded on two levels. They created
a series of prescriptive measures that offered more consolation than perfor-
mance; the measures never returned exterior materials to their pre-insulation
dryness, but they at least had modest benecial effects at the outset. They ad-
dressed the need to do something. At a more fundamental level, the industry
quietly substituted new, more wetness-resistant exterior materials anyway:
First aluminum siding the tin men, then stained cedar in the 1960s, and
then vinyl siding in the 1980s.
We face a similar situation. Our worst option would be to seek to protect
ourselves behind prescriptive measures, as additional insulation leads to unan-
ticipated and unwelcome high levels of wetness for building owners. A better
option would be to paint a complete picture for owners: New insulation im-
poses the need cold weather, of course for more water resistant exterior ma-
terials or for a changed set of expectations regarding the wetness performance
of materials that have been in place.
Perhaps the guidance offered by Rogers remains the best, to go forward
with what we know now but to be prepared to adopt a new outlook as more
information becomes available. Building professionals in organizations such as
ASTM have the responsibility to police the links between prescriptive measures
and performance outcomes however comforting the prescriptive measures are
and however difcult the performance outcomes. These professionals must
have the courage to act on their understandings.
References
1 Gilbert, C., Notes dictated regarding meetings with Mr. Guastavino and Mr. Butler
regarding the construction of the Minnesota State Capitol Building, March 7,
1901, Provided by Brian Kiggins, SchooleyCaldwell Associates.
2 Rogers, T. S., Opening of the Conference, Proceedings: Condensation Control in
Buildings as Related to Paints, Papers, and Insulating Materials, Washington, D.C.,
ROSE, doi:10.1520/JTE102972 27
1952, Building Research Advisory Board BRAB, National Research Council/

National Academy of Sciences, Washington, D.C.
3 Browne, F. L., Some Causes of Blistering and Peeling of Paint on House Siding, U.S.
Forest Products Laboratory No. R6, Madison WI, 1933, 11 pp.
4 Forest Products Laboratory FPL, Wood Handbook: Wood as an Engineering Ma-
terial, United States Department of Agriculture, Forest Service, Agriculture Hand-
book 72, US Government Printing Ofce, Washington, D.C., 1999.
5 Rogers, T. S., Insulation: What We Know and Ought to Know About It, American
Architect and Architecture, Vol. 150, 1936, pp. 8384.
6 Rogers, T. S., Preventing Condensation in Insulated Structures, Architectural
Record, Vol. 152, 1938, pp. 109119.
7 Teesdale, L. V., Condensation in Walls and Attics, U.S. Department of Agriculture,
Forest Service, Madison, WI, 1937.
8 Teesdale, L. V., Condensation, The Architectural Forum, Vol. 91, New York, April
1938, pp. 2024.
9 ASHRAE, Handbook, American Society of Heating Refrigerating and Air Condi-
tioning Engineers, Atlanta GA updated annually. Includes four volumes: Funda-
mentals, Applications, Refrigeration, and HVAC Systems and Equipment.
10 ASHVE, Proceedings of the 43rd Annual Meeting, 1937, St. Louis, MO.
11 Rowley, F. B., A Theory Covering the Transfer of Vapor Through Materials,
ASHVE Transactions, Vol. 45, No. 1134, 1938, pp. 545560.
12 Rowley, F., Algren, A., and Lund, C., 1939, Condensation of moisture and its
relation to building construction and operation, ASHVE Transactions, Vol. 45,
No. 1115, pp. 231252.
13 Lund, C. E., Technological Aspects of Condensation Control in Building Struc-
tures, Proceedings: Condensation Control in Buildings as Related to Paints, Papers,
and Insulating Materials, Washington, D.C., 1952, Building Research Advisory
Board BRAB National Research Council/National Academy of Sciences, Wash-
ington, D.C.
14 Federal Housing Administration FHA, Property Standards and Minimum Con-
struction Requirements for Dwellings, Federal Housing Administration, Washing-
ton, D.C., 1942.
15 Housing and Home Finance Agency HHFA, Condensation Control in Modern
Buildings, Housing and Home Finance Agency, Washington, D.C., 1949.
16 Housing and Finance Agency HHFA, Conference on Condensation Control in
Dwelling Construction, Washington, D.C., 1948, Housing and Home Finance
Agency HHFA, mimeographed proceedings, provided by Heinz Trechsel.
17 Hutcheon, N. B., Vapor Problems in Thermal Insulation, Heating, Piping, and Air
Conditioning, August 1958, American Society of Heating and Air Conditioning En-
gineers, Atlanta, GA.
18 Handegord, G., Design Principles RevisitedThrough the Looking Glass, Pro-
ceedings Fourth Canadian Building Congress, Ottawa ON, 1985, National Research
Council of Canada, Ottawa NRCC9130.
19 ASHRAE 160, 2009, Criteria for Moisture Design Analysis in Buildings, Ameri-
can Society of Heating, Refrigeration, and Air-Conditioning Engineers, Atlanta,
GA.
TESTING/ANALYSIS
doi:10.1520/JTE103071
Wahid Maref,1 Nathan Van De Bossche,2 Marianne Armstrong,1

Michael A. Lacasse,1 Hakim Elmahdy,1 and Rock Glazer1
Laboratory Tests of Window-Wall Interface

Details to Evaluate the Risk of
Condensation on Windows
ABSTRACT: The development of alternative details to manage water intru-

sion at the window-wall interface has produced a number of novel ap-
proaches to detailing the interface between the window and the adjacent wall
assembly. Many of these approaches advocate the need to provide drainage
at the rough opening of the window subsill, given that the window compo-
nents themselves are susceptible to water entry over their expected life.
Depending on the types of windows used and the cladding into which the
windows are installed, there arise different methods to provide drainage that
may also affect air leakage through the assembly. This in turn may give rise
to the formation of condensation along the window, at the sill, or along the
window sash and glazing panels. Hence, there is a need to determine if,
under cold weather conditions, specic interface details that incorporate sill
pans provide a potential for condensation on the window components in
which air leakage paths may be prominent at the sill or elsewhere on the
window assembly. The paper reports on a laboratory evaluation of conditions
suitable for the formation of condensation at the window frame perimeter of
the interface assembly as a function of both temperature deferential and air
leakage rate across the test assembly. A summary of the laboratory test
protocol is provided, which includes a description of the test setup and ap-
paratus, fabrication details of the specimen, information on instrumentation
and calibration, and experimental results for one type of window ange win-
Manuscript received March 8, 2010; accepted for publication December 17, 2010; pub-
lished online xx xxxx.
1
National Research Council Canada, Institute for Research in Construction, Ottawa,
ON, Canada.
2
Dept. of Architecture and Urban Planning, Ghent Univ., Ghent, Belgium.
Cite as: Maref, W., Van De Bossche, N., Armstrong, M., Lacasse, M. A., Elmahdy, H. and
Glazer, R., Laboratory Tests of Window-Wall Interface Details to Evaluate the Risk of
Condensation on Windows, J. Test. Eval., Vol. 39, No. 4. doi:10.1520/JTE103071.
31
dow. In parallel, preliminary simulation results were presented and com-

pared to those obtained from the experiment using the commercially avail-
able thermal software BISCO.
KEYWORDS: air leakage, laboratory testing, window installation, wall-
window interface, window condensation
Introduction
There exist several standard methods for determining the potential for the for-
mation of condensation on windows, as provided in Table 1; however, the es-
sential aspects of the method were rst proposed by Sasaki 1, and the stan-
dardisation work carried out by the Architectural Aluminium Manufacturers
Association AAMA 2,3, ASTM 4, and Canadian Standards Association
CSA 5 follows on these initial efforts. These standards prescribe the overall
test protocol, temperatures of the room side and cold side, and the maximum
relative humidity RH under test conditions. A useful overview of these meth-
ods is given by Elmahdy 6.
There also exist simulation tools that could be used to assess the potential
for window condensation; such as, for example, FRAME 4.0 7, VISION 4.0 8,
and BISCO 10.0w 9.
The provided details of the window prole are available in the format ac-
cessible by the simulation software. Such tools permit standard window types
to be readily assessed from known boundary conditions, and rapid evaluations
on the energy efciency as well as condensation potential are possible. How-
ever, such software is not typically adaptable to measuring the performance of
window installation in which non-standard conditions and different ap-
proaches might be of interest, as, for example, where air leakage is considered
a testing and evaluation parameter 10.
Method for Determining Condensation Potential

The essential elements of the method, briey described, consist of testing a
window in a hotbox chamber, measuring the window and frame surface tem-
peratures at specied locations on the window, and calculating a weighted av-
erage of the interior surface temperature. The temperature index I is then
determined based on the following relationship and on CSA Standard
TABLE 1List of standards for determining condensation potential of windows.
Room Cold
Standard Side Temp. Side Temp. Test Period
Org. Designation C C h Pressure/%RH
AAMA 1502.3-1972/
AAMA AAMA 1503-98 21.170 F 17.80 F Nil/ 15 %
ASTM ASTM C1199-00 21.170 F 17.80 F Nil/ 15 %
29
CSA CSA A440.2-04 20 1 30 1 51 C 0 5 Pa/ 15 %
MAREF ET AL., doi:10.1520/JTE103071 33
I = Ts To/Ti To 100 1
where:
Ti , To = indoor and outdoor air temperatures in C and
Ts = average room side surface temperature measured in the test.
For Eq 1 to be used internationally, we used a different denition of the
temperature index according to EN ISO 10211 11
I = T s T o / T i T o 2
The temperature index is non-dimensional and represents the interior sur-
face temperature relative to the interior and exterior air temperatures. I = 0
implies that Ts = To, which is the same as having no window at all because the
interior surface temperature is the same as the outdoor temperature. When the
temperature index I = 1, this indicates that Ts = Ti, the same as the room side air
temperature, thus providing the best possible rating. Based on Eq 2, I may
range between 0 and 1, with a typical value for a clear double-glazed window
having a metal frame being ca. I = 0.4. Thus, using Eq 2 to predict the conden-
sation potential of a given window in a room, the following information is
required:
The I temperature index value for the window,
the indoor room air temperature Ti and outdoor temperature To at
the location of interest, and
the RH or dew point temperature of the room air.
Based on this information, the estimated room side surface temperature
Ts can be determined and thereafter compared to the dew point temperature.
The dew point temperature of the indoor air can be calculated according to Eqs
3 and 4 EN ISO 13788:2001 12
237.3 loge
psat
610.5

= for psat 610.5 Pa 3
psat
17.269 loge
610.5
265.5 loge
psat
610.5

= for psat 610.5 Pa 4
psat
21.875 loge
610.5
If the value of Ts is less than that of the dew point temperature inside the
room, then condensation on the window is expected. An overall condensation
risk assessment methodology is presented in EN ISO 13788:2001 12.
Overview of Approach
The development of alternative details to manage water intrusion at the
window-wall interface has produced a number of new approaches to detailing
the interface between the window and the adjacent wall assembly. Many of
these approaches advocate the need to provide drainage at the rough opening
of the window sill, given that the window components are susceptible to water
entry over their expected life. Depending on the types of windows used and the
cladding into which the windows are installed, there arise different methods to
provide drainage that may also affect air leakage through the assembly. This in
turn may give rise to the formation of condensation along the window, at the
sill, or along the window sash and glazing panels. Hence, there is a need to
determine if, under cold weather conditions, specic interface details that in-
corporate sill pans provide a potential for condensation on the window com-
ponents in which air leakage paths may be prominent at the sill or elsewhere on
the window assembly.
As has been shown, several methods have been devised to evaluate the
potential for condensation at the window proper; however, there have not yet
been any methods specially derived for evaluating the formation of condensa-
tion at windows given different window installation details. Such a method
would permit determining if a window installation provides adequate thermal
resistance and reduced risk to the formation of condensation. It would also
permit comparative evaluations amongst different installation methods using
the same window and cladding types in a given wall assembly. It may also offer
a means to benchmark the results of a eld-testing method, should one be
developed in the future.
This paper reports on a laboratory evaluation of conditions suitable for the
formation of condensation at the window frame perimeter of the interface as-
sembly as a function of both temperature deferential and air leakage rate
across the test assembly. A summary of the laboratory test protocol is provided,
which includes a description of the test setup and apparatus, fabrication details
of the specimen, and information on instrumentation, calibration, and experi-
mental results for one type of window ange window. In parallel, preliminary
simulation results were presented and compared to those obtained from the
experiment using the thermal software BISCO version 10.0w 9.
Outline of Evaluation Program

Testing was conducted in a hotbox test facility 13 and the CSA A440.2-04 test
method for determining window condensation 5 was adapted to subject speci-
mens to specied temperature differences. A suitable specimen incorporating a
window and related interface details was subjected to temperature differentials
from which surface temperature measurements on specic components of the
window were determined. This information permitted establishing whether
there existed conditions suitable for the formation of condensation given speci-
ed interior and exterior conditions. A description of the hotbox was provided
by Brown et al. 14 and details on the experimental procedure, the calibration
of the hotbox, specimen instrumentation, and data acquisition are provided in
subsequent sections.
Given that the primary interest was to determine whether different window
installation details affect the potential risk to the formation of condensation on
a window, the different issues arising from this relate to the following:
Selection of type of wood frame assembly and assembly components
FIG. 1Schematic of anged window installation in window opening showing the

planes of thermal resistance of window in relation to that of nominal wall assembly.
including size of frame, sheathing board, insulation, and window types,

location of window in window opening,
choice of installation details, including sill pan ashing and other ash-
ing, and
incorporation of deciencies in the specimen.
One might consider that in North America, a typical wood frame assembly
for cold climates would consist of a 6 in. wood stud wall, sheathed with an
oriented strand board OSB panel and incorporating a polyolen sheathing
membrane, berglass insulation, polyethylene vapour barrier, and interior gyp-
sum board nish. This was assumed to be a representative assembly in which
non-operable PVC windows of two different types were installed: Windows in-
corporating a xing installation ange i.e., anged window and those not
having a ange i.e., box window. PVC windows were selected as these were
considered to be the most commonly specied in large commercial housing
projects.
This paper reports on results derived from testing anged windows. For
anged windows, the location of the window is xed in relation to the plane of
thermal resistance of the wall Fig. 1; this necessarily affects the overall ther-
mal insulation R-value in SI units RSI at the opening. When a window in-
corporating a xed ange is installed, as shown in Fig. 1, the location of the
plane of thermal resistance is predetermined, based on the location of the
ange in relation to the windowpane and window frame. In this instance, the
plane of thermal resistance of this window may very well be forward of that of
the nominal plane of thermal resistance of the wall.
As the difference between the locations of these two planes increases, so
too does the likelihood for the formation of condensation on the window or
frame for windows installed in cold climates. This is due to the fact that there
is less thermal insulation in the cavity between the window and the window
opening as compared to the wall proper, thus giving rise to decreases in the
surface temperature of the window or window frame on the interior in relation
to the interior room temperature. The continuity of the thermal plane at the
window interface with the wall must, in principle, be reconciled; nonetheless,
this is limited by the amount of space in which insulation can be applied even
though the products typically used have themselves a high thermal resistance.
In respect to the choice of installation details, consideration was only given
to those details that had in a previous study 15,16 demonstrated an ability to
adequately manage rainwater entry. Such installation details typically include a
sloped sill with sill pan ashing incorporating a back dam.
Finally, given the interest in using installation details that include a sill pan,
thought was given to possible paths of air leakage through the assembly at the
sill and the type of deciencies that might arise at these locations due to im-
proper installation of components or premature failure of seal components.
Two possible paths were considered: A short path at the sill and another longer
path along the interior jamb that enters the room side at the window head.
Detailed information is provided in the subsequent section on the congu-
ration of the test specimens, the window installation and location of the win-
dow in the specimen, and the incorporation of deciencies in the specimens to
simulate air leakage problems.
Experimental Setup and Procedures
Conguration of Test Specimen and Mounting Specimen in Test Assembly

Figure 2 provides the nominal test specimen setup showing the wood frame
assembly and the size of the opening and its location within the assembly of a
non-operable PVC window. The nominal size of the test frame incorporating
the window-wall interface was 1.22 m 4 ft wide by 2.44 m 8 ft high. The test
assembly was intended to be representative of typical North American wood
frame construction practice. The wall assembly was framed with 51
152 mm 2 6 in. spray-in-place polyurethane foam SPF lumber in the
conguration shown in Fig. 2. The exterior cladding of the assembly was hard-
board wood composite siding, installed in accordance with current building
practice and directly to the sheathing membrane spun bonded polyolen. The
membrane overlays an OSB 11 mm sheathing panel afxed over the wood
frame. Fiberglass batt type insulation was placed in the stud cavities adjacent
to the window opening and the interior nish was gypsum board 12.7 mm.
A non-operable PVC window 610 1220 mm incorporating a xing in-
stallation ange i.e., anged window was installed in the specimen. These
were the same type of window as was used in a previous study to evaluate the
watertightness of different window installation details 15,16, and are typical
of those specied for North American construction practice. The location of a
anged window in the window opening, as was previously discussed, is neces-
sarily predetermined by the conguration of the ange in the window frame.
FIG. 2Nominal test specimen setup showing wood frame assembly of the test speci-
men and size and location within assembly of non-operable vinyl window.
Table 2 provides information on the respective values of RSI for each of the
relevant wall components from which the location of the plane of thermal re-
sistance was determined; the line in the table that separates the two values of
berglass batt indicates the plane of thermal resistance, i.e., the location within
TABLE 2RSI value of wall components and calculation location of plane of thermal
insulance of wall.
Thermal Thermal
Thickness Resistivity Resistance Cumulative
Component m in. m K / W RSI m2 K / W RSI
Wall-outside air lm, 24 km/h 0.03 0.030 0.030
Sidinghardboard 0.0111 7/16 10.75 0.120 0.150
SheathingOSB 0.0111 7/16 11 0.122 0.272
Fiberglass batt 0.0685 2.7 26 1.798 2.070
Fiberglass batt 0.0712 2.8 26 1.872 3.942
Drywall 0.0127 0.5 6.2 0.079 4.021
Inside air lm
non-reective, vertical 0.120 0.120 4.141
Total 0.175 6.9 4.141
FIG. 3Vertical section of specimen showing installation details for non-operable

anged windowshows plane of thermal resistance of window unit as compared to
wall.
the wall assembly on either side of which the cumulative values for RSI are
equal i.e., 2.7 in.. Figure 3 provides the installation details of a anged win-
dow incorporating the self-adhering exible pan ashing membrane, sloped
sill, up-stand, or related component details that help promote the drainage of
water from the windowsill if subjected to inadvertent water entry. Figure 3 also
TABLE 3Test series for anged window conguration.
Flanged Installation Test Set 4 Test Set 5 Test Set 6
Position Flange Flange Flange Flange Flange Flange

Insulation type None None Batt Batt SPF SPF
Deciency type/location No Yes No Yes No Yes
D1 + D3 D1 + D3 D1 + D3
shows that the location of the insulated glass unit IGU, assumed to be the
plane of thermal resistance of the window, differs from that of the wall and
indeed is located closer to the exterior of the wall.
A summary of the different interface details for anged windows are given
in Table 3, in which information is provided on the
type of window anged and the position of the window in the rough
opening; in Table 3, position refers to the position of the window in
the rough opening; ange implies that the plane is determined on the
basis of where the ange of the window is located in relation to the
respective planes of thermal resistance;
use, or not, of insulation and the type of insulation used when specied;
batt refers to berglass batt insulation, whereas the SPF designation
indicates that polyurethane spray-in-place-foam was used in the same
cavities located between the window opening and the window frame;
incorporation, or not, of deciencies in the assembly; D1 in Table 3 was
located at the exterior of the wall-window interface and at the juncture
of the cladding and window frame at the lower extreme comer of the
window Fig. 4; D3 in Table 3 was located at the interior of the assembly
and at the interface between the window frame and the interior nish
but located at the upper most corner of the window assembly.
The introduction of deciencies at the wall-window interface provided a
means to evaluate whether air leakage across different components of the win-
dow assembly could cause condensation to form on the warm side of the wall
assembly when leakage was induced in the test assembly. The intent was to
demonstrate the vulnerability of the assembly to the formation of condensation
on the interior in instances where, for example, a deciency was located as
shown in Fig. 4 as D1 on the exterior and D3 on the interior.
Instrumentation and Data Acquisition
Measurement of Temperatures and Relative HumidityThe chambers envi-

ronmental conditions were monitored continuously over the course of a test
sequence; likewise, surface temperatures on the interior and exterior surface of
the window and window frame were measured; the following provides infor-
mation on the accuracy of measurements and the means of capturing data.
Monitoring Chamber Conditions. Both the temperature and RH were con-

tinuously monitored over the course of a test sequence in the warm side cham-
FIG. 4Isometric view of specimen; shows location of deciency at lower corner of

window and alternate leakage path at top of window.
ber and only temperature in the cold side chamber. Measurements of tempera-
ture were made to an accuracy of 1.5 C and that of RH to 1 % RH. The data
was recorded on the data acquisition system, then subsequently used to ensure
that steady state conditions have been maintained over the course of a test
sequence.
FIG. 5Nominal location of thermocouples on exterior and interior face of test

specimen.
Monitoring and Recording Surface Temperature Condition. Surface tempera-

ture conditions on either side of the window i.e., exterior and interior and on
specied window components e.g., window lite; window frame at sill, frame at
jambs, etc. were continuously monitored with the use of a set of 40 thermo-
couples; 20 were used on the exterior and 20 on the interior of the specimen.
Measurements of temperature were made to an accuracy of 0.5 C. The loca-
tion of each of these thermocouples on the interior and exterior face of the
specimen glass lite and frame follows that which is specied in the standard
CSA A440.2-04, as given in Fig. 5. Thermocouples were also placed within the
cavity between the window frame and the window opening; the intent of plac-
ing thermocouples at these locations was to monitor the temperature in the
space, and in proximity to the window frame, along the path of expected air
leakage and through which cold air would migrate.
In additional to continuous monitoring of surface temperature conditions,
an Infra Red IR camera was used to scan the surface temperature of the
widow frame. The data was thereafter compared to that provided by the ther-
mocouples. Acquiring an IR scan of the interior of the window requires the use
of a bafe to minimize variations in surface conditions that may result from the
presence of a camera operator in the warm side room.
Measurement of Pressure Differentials and AirowTests were carried out

under a pressure differential that required the pressure difference across the
test assembly to be continuously monitored during the test sequence. A pres-
sure transducer having a 250 Pa range 1 in. water and accuracy of 1 Pa was
placed in the warm side chamber. Additionally, a pressure transducer was also
used to monitor the pressure in the interstitial space between the window
frame and the window rough opening in the wall. The requirement for a speci-
ed pressure differential across the test assembly necessitated the use of an air
pump that expelled air from the warm side chamber. The amount of airow
was not monitored; however, the ow was adjusted to accommodate the speci-
ed pressure differential.
Experimental Procedures
The basis for this test is the observation of surface temperatures sufcient to
cause the formation of condensation on window components located on the
warm side of the test assembly. Actual visual observation of condensation is not
required, nor desired, as the formation of condensation on thermocouples can
affect measurements taken of affected sensors. Hence, temperatures measure-
ments on the warm side were undertaken in conditions where the humidity
levels were sufciently low to preclude the formation of condensation.
Calibration of Test FacilityThe guarded hotbox test facility is calibrated

according to the protocol described in EN ISO 13788:2001 12. The lm heat
transfer coefcient on the room side and weather side surfaces was determined
from the calibration of the hotbox with the use of the Calibration Transfer
Standard CTS. For calibration, the CTS was mounted ush with the room side
surface of the surround panel.
Test Procedure Under No Pressure DifferentialThe tests in this instance

were carried out without pressure differences across the specimen. The tem-
perature differential for the initial test was set at 50 C 1.5 C and the tem-
perature sensor measurements are recorded once steady state conditions
have been achieved following a period of 15 min in these conditions
20 C 30 C = 50 C. This test represents standard test conditions as de-
scribed in CSA A440.2-04 5. The humidity on the warm side chamber was
maintained at 5 % RH to ensure that no condensation occurred on any of the
interior exposed surfaces of the widow frame. A time recorded scan using the
IR camera was taken for later analysis.
The test was repeated at a lower temperature differential of 45 C and
thereafter at 40 C to investigate the extent to which such conditions altered the
surface temperatures on the warm side of the window assembly. As was the
case for the test conducted in standard conditions, temperature sensor mea-
surements were recorded once steady state conditions had been achieved fol-
lowing a period of 15 min in these conditions and, as well, the humidity on the
warm side chamber was maintained at 5 % RH to ensure that no condensation
occurred. As before, a time recorded IR camera scan was taken for subsequent
analysis.
Test Procedure when Applying a Pressure DifferentialThe procedure for the

test conducted under pressure differential rst included incorporating decien-
cies of a specic type, size, and location in the specimen as previously de-
scribed. Once this had been completed, the prescribed temperature differential
under no pressure differential was rst attained, and thereafter, differential
pressure was slowly applied to the chamber by evacuating the warm side cham-
ber to reach an initial pressure differential of 20 Pa 0.5 Pa and thereafter, a
threshold of 40 Pa 1 Pa. The basis for this pressure threshold is the study
conducted by Cornick 17 to establish hygrothermal test conditions for build-
ings located in northern Canada. The wind velocity pressure derived from an
annual mean wind speed of 22 km/h is 27 Pa at 30 C 17. Given that the
calibration tests on the test setup described above revealed that 40 Pa was the
highest achievable pressure difference within reach of the test facility, this was
set as the possible upper threshold for the test pressure differential; measure-
ments were also completed at 20 Pa 18. The intent of taking measurements at
two different pressure levels was to gain insight into the effect of varying pres-
sure differences across the envelope on increasing the risk to the formation of
condensation at the window. As was the case for the test conducted in standard
conditions i.e., at no pressure difference, temperature sensor measurements
were recorded once steady state conditions had been achieved following a pe-
riod of 15 min in these conditions.
Numerical Simulation
Steady state thermal simulations were done with the numerical software pro-
gram BISCO 10.0w 9. This program allows a simple conversion from
computer-aided design-based drawings into a user-dened grid size based on a
triangulated grid model in conformity with EN ISO 10211 11. The system
nodes are located at the triangle vertices, at which the temperatures are calcu-
lated and from which heat uxes can be deducted. Material properties were
adopted from EN ISO 10077-2 19 and EN ISO 10456 20. Radiation is mod-
eled iteratively in a non-linear way based on view factors conned to 100 ray
traces per cavity, whereas convection in cavities is calculated according to EN
ISO 10077-2 19. The value for heat transfer coefcient at the weather side was
33.0 W / m2 K, as measured in the hotbox. The heat transfer on the room side
was calculated according to the formula hc = 1.776 T0.25 21. The surface
temperature and heat transfer coefcients were calculated in ve iterations a
uniform heat transfer coefcient was assumed. The black radiation heat trans-
fer coefcient was set to 4.38 W / m2 K, and the grid size was 0.1 mm.
FIG. 6IR picture of test setup.
Results
Test Set 4 No Insulation

Without a pressure difference over the specimen, there was a very homoge-
neous temperature distribution over the wall, window frame, and glass on the
outside cold side of the specimen 28.2 to 29.1 C. On the inside warm
side of the specimen, the temperature of the wall shows a little more variation,
as could be expected; on the outside, the air space between the WRB and the
hardboard siding blocks any thermal bridging by wood studs to appear on the
measurements. The temperature on the window frame and the IGU has a
strong thermal gradient, whereas the temperatures on the upper half of the
window range from 11.9 to 13 C, temperatures between 4.1 and 7.7 C are
found near the sill. The argon inside the IGU-cavity will, when heated, rise
towards the top of the unit, resulting in a typical thermal gradient; this is
evident in Fig. 6, in which is given an IR photograph from the interior side of
the test setup. Note that the convection in the interior cavity of the IGU would
be even more pronounced in the case where it is lled with a gas with lower
density, such as air. Figure 7 shows a picture from the window taken from the
inside.
As the cavity between the anged window and the wood framing is
12 mm 1/2 in. wide and 75 mm 3 in. deep, internal convection within the
cavity might induce thermal stratication in the vertical air space, which is
essentially open to the outside at the ange in front of the subsill. The cavity is
located at the plane of thermal resistance of the window where the highest
thermal gradient from inside to outside is evident, thus intensifying the result-
ing convection. Any cold air entering into the cavity will gather heat from the
inner side of the window frame and thereafter rise due to its descending mass
density. However, due to this phenomenon, one might expect a similar thermal
gradient at the outside of the windowthis was not observed. A possible expla-
nation can be found by analyzing the window frame conguration. The specic
design of the window prole lets the outside chamber partially act as a geo-
metrical cooling n, short-circuiting the location of the thermocouple. Further-
more, due to the high heat ux caused by the spacer of the IGU, the effect of
FIG. 7Picture inside view test setup.
other components on the temperature at the thermocouple is outclassed. A

steady state simulation of the window-wall interface shows the temperature
distribution in the interface in Fig. 8 30 to 20 C, and Fig. 9 shows the heat
ux density 0 200 W / m2. A comparison of the simulation results and mea-
surements are provided in Fig. 10; the results show good correlation except for
the temperature in the cavity between the window frame and wood stud wall
caused by convective effects in the window prole and in the cavity between
frame and wall. However, an evaluation of static and dynamic simulation tech-
niques does not lie within the scope of this paper and will be part of future
research reports. Figure 9 shows the high heat ux through the spacer of the
IGU, and next to that, the heat ux around the groove intensies the thermal
separation of that part of the frame from the cavity between frame and wall.
The data retrieved from the six thermocouples located in the cavity conrm
a temperature gradient over the height of the cavity Fig. 11. However, when
compared to the experiments with SPF in the cavity, in which both natural
thermal stratication and forced air ow was eliminated, it seems that a part of
the observed temperature gradient is caused by other effects. The mounting
brackets were not close enough to the thermocouples to affect the measure-
ment, and given that this was a non-operable window, hence without hinges or
stays, it is apparent that neither of these items can affect results. The vertical
thermal gradient in the cavity is monitored throughout all measurements and is
consistent in nature.
The thermocouples are mounted onto the vinyl frame, so perhaps the win-
dow frame itself can account for the measured effects. The cross section of the
window frame reveals that the section is comprised of two large cavities; the
FIG. 8Simulation section: Isothermal lines.
thermocouples are attached to the exterior surface of the largest cavity but
closer to the interior side of the assembly. Experimental analysis and numerical
modeling of convection in a tubular 1 in.2 vinyl frame has been previously
discussed by Gustavson et al. 22. In this work, it was evident that for a tem-
perature difference of 20 C over a vertical tube of length 800 mm forming part
of a 800 800 mm vinyl frame, there was a vertical thermal gradient of about
2 C, excluding conductive effects at the top and bottom of the tube. Whereas in
the test setup to assess the condensation risk of interest in the present study, the
temperature difference is 50 C, and it is most likely that a large part of that will
act over the vertical window frame cavity because it is located at the plane of
highest thermal resistance of the IGU and the wall. Although this cavity is
smaller than that evaluated from Gustavsons work about half the size and
results are possibly compensated by a larger temperature difference, this is the
most plausible explanation to account for the observed thermal gradient along
the vertical axis of the window frame.
Condensation Risk AssessmentWith regards to assessing the risk to the

formation of condensation, differentiation can be made between surface con-
densation at any visible location and interstitial condensation inside the cavi-
ties between the different components. As cold air passes through the interior
construction, it will not only cool the adjacent components, but also dry the air
vapour pressure remains constant, but the RH is diminished as the air gets
FIG. 9Simulation section: Heat ux intensity.
warmer. This effect will not be present at the interior visible surface, as the air
might enter the room at a different location. However, one should take into
account that pressure differences over building envelopes are induced by a set
of causes. Wind loads, stack effects, and HVAC systems might be able to change
the pressure any time and reverse the air ow through deciencies. Obviously,
this situation would increase the risk to condensation inside the assembly itself,
which in principle should be avoided any time. It may be assumed that long-
lasting pressure differences sufciently long to cool the adjacent components
are not sequenced by opposite long-lasting static pressure differences. Such a
situation is only likely to occur due to malfunctioning HVAC-systems. Hence,
only the risk to the formation of surface condensation is considered in this
paper.
In this respect, the risk to the formation of condensation depends only on
two parameters assuming constant pressure conditions: Temperature and hu-
midity. For any given surface temperature, there can be condensation if the
humidity is sufciently high. A practical approach to analyze the data is to
calculate the dimensionless temperature index for every test case. This index
value provides normalized results and thus an objective basis of comparison of
results derived from testing the different setups. Furthermore, given that heat
transfer is linearly correlated with temperature difference, this index is inde-
pendent of the specic temperatures used in the experiments. For any bound-
ary conditions, the indoor surface temperature can be calculated using the
FIG. 10Validation of steady state simulation.
temperature index. A straightforward psychometric calculation offers the maxi-

mum RH for the specic indoor temperature in order to avoid surface conden-
sation on the window frame.
Figure 11 shows an overview of the different temperatures on the IGU, the
window frame, and the cavity between the window frame and the wall. The
circles on the wall show the locations of the different thermocouples on the
surface, e.g., to check if no thermal stratication would occur in the test cham-
ber for reasons of clarity, these are not reported here. The locations of the
different thermocouples on the glass and frame are marked and corresponding
temperatures are shown in result boxes. In every result box, the top value refers
to the situation without a pressure difference over the wall, the middle number
was measured at 20 Pa, and the bottom value corresponds to 40 Pa pressure
difference. On the upper half of the window frame, the temperature is 11.9 C,
temperature index 0.85 without pressure difference Figs. 11 and 12; down at
the window sill, temperatures are lower and the corresponding temperature
and index are 4.1 C and 0.69, respectively. The temperature index is rounded
at two digits, which corresponds to a rounding of the temperature to 0.5 C.
For a temperature difference of 50 C 30 to 20 C, the surface temperature is
0.69 50 C 30 C = 4.5 C. The difference between the 4.5 and 4.1 C is
caused by a deviation of the indoor temperature was slightly lower than the
20 C set point. As such, the temperature index is a more reliable measure to
compare results and is independent of the boundary conditions of the experi-
mental test setup. For other boundary conditions, the surface temperature can
similarly be calculated, offering a comprehensive method to calculate conden-
sation risk for a specic indoor condition. The component of the window frame
FIG. 11Temperatures on the frame and inside the cavity for different pressure
differences.
that consistently permitted establishing the risk to the formation of condensa-

tion was the lower part of the window frame. Hence, for reasons of clarity, only
the results for the lower portion of the window frame and IGU at the sill will be
reported. Merely for illustration, according to this calculation, the temperature
index in the cavity between window frame and wood framing would be 0.55.
Effect of Pressure Differences No DecienciesIn order to analyze the ef-

fect of cold air inltrating into the construction, a pressure difference of 20 and
40 Pa was applied. Due to this unbalance, cold air was drawn into the warm
side of the assembly through small cracks and openings in the different com-
ponents of the specimen. As there were no wilful deciencies in the initial test
i.e., purposely incorporated deciencies not functional in this setup, air could
only enter through local imperfections of the wall, window-wall interface, or
window frame. The results from the tests indicate that the temperature distri-
bution is very similar to the test without pressure difference taking into ac-
count a slightly lower room temperature. However, a lower temperature inside
the cavity at the corners of the window was noted. Already divergent from the
general trend in the rst measurement, this becomes more pronounced as the
pressure difference is increased. Although this effect may partially be explained
by the three-dimensional heat transport in the corners of the window frame
FIG. 12Temperature indexes when deciencies are present.
introducing a point thermal bridge, that should only be proportional to the

temperature difference, which is clearly not the case here. A close examination
of several samples of the window used in this study revealed specic imperfec-
tions at the mitred joint of the heat welded vinyl frame. The mitred joint was
chamfered after welding but in some instances was apparently cut off, thereby
revealing a small opening slit at the exterior corners of the window frame. On
the other hand, at the top and bottom side of the window, there were minor
perforations caused by staples that held the wood protection strapping in place
during transport. As well, and contrary to good practice, there were no weep
holes at the bottom side of the window. It may be stated that the windows used
in this study were of low quality and additionally were obtained having several
deciencies present specically at the corners both on the inside and outside of
the frame. These rendered it possible for cold air to enter the window frame at
any opening on the outside of the frame between the WRB and the cladding
blocking the effect from the thermocouple measurements, and the air could
then exit the frame at exterior corners, in the cavity behind the sealed window
anges. Tests with smoke pencils and IR photographs of the window exclude
any unintended deciencies to be the cause. Perhaps there were some cracks or
slits at the perimeter of the wall specimen mounted in the hotbox. A laminar
ow of air between the air barrier and the drywall could remain undetected,
allowing the air to exit the cavity. As will be explained in the analysis of the test
setup with berglass insulation in the cavity, this will affect the temperatures in
the overall cavity as well.
When a 20 Pa or 40 Pa pressure was applied, the temperature on the
window frame decreased by 0.8 C 0.9 C at the top side and 1.4 C 1.8 C at
the bottom side of the window frame. Although the pressure difference is
doubled, one should take into account the power law in respect to the air
leakage rate, specically
Q = C . pn
where:
Q = air ow rate L/s,
C = flow coefcient L / s Pan, and
n = flow coefcient .
It can reasonably be assumed that the ow exponent of the specic de-
ciency lies between 0.55 and 0.65 for building applications. Within this range of
ow exponents, the air ow would rise between 46 % to 57 % when the pressure
difference is doubled. Hence, the temperature drop cannot be expected to be
directly proportional to the pressure difference. Based on the pressure differ-
ence, discharge coefcient of the deciencies, and internal friction, it is pos-
sible to estimate effective airow rates. Any assumed velocity prole inside the
cavity will affect the convective heat transfer coefcient at the boundary layer
with the window prole. However, this analysis requires a more elaborate fun-
damental study that is beyond the scope of this study but may be investigated
in future research.
With regards to the condensation potential, the results are similar to the
tests without pressure difference: A temperature index of 0.69 at 0 Pa, 0.66 at
20 Pa, and 0.66 at 40 Pa. The low temperatures in the cavity would cause severe
problems during a rapid change of air ow direction: A temperature of 8 C
corresponds to a temperature index of 0.44, so humid indoor air could easily
condense on that surface.
Effect of DecienciesDuring the second series of tests, two deciencies

were installed: One at the lower right corner on the outside D1 and one on the
upper left corner on the inside when looking from the inside of the window
D3. When no pressure difference is applied, the results are identical to the
previous test. The temperature in the cavity exactly at the deciency on the
outside drops 3 C; however, other than this result, no differences are recorded.
Due to the big temperature difference between the inside and the outside, there
is a difference in the mass density of the air. Any direct contact between both
spaces through an opening will result in a local stratication, pressure differ-
ence, and air exchange. The local temperature drop can be traced back to the
deciency connecting the cavity to the outside. As one looks at the other de-
ciency D3, a similar effect is not found. First of all, the temperature difference
between the cavity and the outside at the lower deciency D1 is about 30 C,
whereas the temperature difference at the top corner is only 14 C cfr. the
temperature stratication in the vertical section of the cavity. As well, the
deciency at the lower side D1 is close to the slit between the ange and the
WRB, offering limited resistance to air ow. Measurements conrm that the
pressure difference recorded at the bottom 0.5 Pa is lower than the pressure
difference at the top 1.3 Pa; note as well that there is no vertical temperature
gradient in both rooms, only in the cavity itself.
Effect of Deciencies and Pressure DifferencesWhen a pressure difference

of 20 Pa is applied whilst deciencies are present see Fig. 12, the outside
surface temperatures remain the same but the surface temperature of the win-
dow prole on the inside goes down about 1.5 C compared to the test without
deciencies about 2 C for 40 Pa. A temperature drop of 1 C corresponds to a
drop of 0.02 points in the temperature index for the given boundary conditions
of the test setup. Compared to the test with the deciencies active but no
pressure difference, the upper part of the window cools down slightly less than
1 C at 20 Pa and the lower part 2 C; at 40 Pa, the corresponding values are 1
and 3 C, respectively. In the cavity, the effect is most pronounced in the cor-
ners. At 20 Pa, the temperature drops 1.5 and 3 C at the bottom corners and
7 C at the top left corner. A temperature drop of 2.5, 6, and 13.5 C, respec-
tively, were measured inside the cavity at 40 Pa. The results permit suggesting
that a constant pressure difference of 40 Pa has a signicant effect on the
indoor surface temperatures, but is limited to about 3 C the temperature in-
dex drops, at most, 0.08. The surface temperatures at the inside of the window
will decrease somewhat, imposing a temperature index of 0.66 0.63 at the
bottom half of the window at 20 Pa 40 Pa. The initial temperature index for a
window with neither deciencies nor pressure differences was 0.69. As could be
expected, the effect was more pronounced in the cavity near the corners, result-
ing in temperatures as low as 7.3 C 11.4 C at 20 Pa 40 Pa, correspond-
ing to temperature indexes of 0.45 0.38.
Test Set 5 Fiberglass Insulation

For this test setup, the cavity between the window frame and the wood framing
was lled with berglass insulation and caulked on the interior side. The sur-
face temperatures on the outer side do not differ from those obtained in the test
setup without insulation Test Set 4. On the inside, the effect is most pro-
nounced for the surface temperature of the window frame, which is consis-
tently 0.5 to 1 C higher due to the insulation. However, this is not the case for
the IGU, whereby the temperature index stays at 0.69. In fact the surface tem-
perature of the glass perimeter is primarily determined by the centre-of-panel
R-value of the IGU, the IGU-spacer along the edges, and the removable glazing
stop of the window frame. Indeed, the effect on the IGU will be negligible given
the limited lateral thermal resistance of the frame and unless the temperature
around the frame is below a certain threshold. The cavity between the window
and the window rough opening, now lled with insulation, is about 0 to 2 C
warmer compared to the setup without insulation. It should be noted that there
is still a minor thermal gradient present, albeit less pronounced then earlier.
This conrms the ndings concerning the effect of convection inside the verti-
cal mullions of the window frame. A detailed analysis of the results indicates
that the calculated thermal gradient is smaller than the one observed without
insulation but slightly larger than the one observed with SPF insulation. Either
this is caused by thermal stratication despite the resistance of the berglass
insulation, or there is an upward air ow caused by a leak. In the latter case, it
would also account for a part of the thermal gradient in the setup without
insulation. In the horizontal cavity at the sill, the temperature is signicantly
higher than the test without insulation, consistent with the assumed convection
effects.
Effect of Pressure DifferencesBoth for 20 and 40 Pa, there is no important

effect on the outdoor surface temperatures, not for the wall, not for the window
frame, and not for the IGU. The window frame temperature on the inside di-
minishes by 1.2 C at 20 Pa and 2.1 C at 40 Pa. In the cavity between the
window frame and the window rough opening, the temperature drop is the
same as an assembly without insulation. Without insulation, the average tem-
perature drop in the vertical cavity due to a 40 Pa pressure difference was
0.8 C, and with insulation it was 0.6 C. The resistance of the berglass insu-
lation to air ow can account for this small change. However, an explanation
for temperature variations at the corners differs from that provided above;
whereas the upper left corner responds in a similar fashion to the case without
insulation, the two bottom corners get extremely cold during this test. At 20 Pa,
these cool to 7.3 and 6.5 C compared to 3.5 and 3.1 C without insula-
tion, and at 40 Pa even to 11.2 and 11.4 C compared to 5.4 and 4.0 C.
One would assume that the berglass insulation would reduce the ow of air
through the cavity, and the corresponding temperature drop should hence be
lower. During this test, the ve thermocouples on the lower side of the drywall
register an average temperature drop of 0.9 C, whereas the room side tempera-
ture is exactly the same. There must have been a local deciency in the caulking
along the sill, through which the cold air could ow at much higher ow rates,
in between the drywall and the air barrier towards any deciencies in the wall
installation perimeter.
Effect of DecienciesThe results of the test with the two deciencies D1

and D3 in the construction show a general temperature rise of 0.2 C, which
can be attributed to measurement errors and rounding errors. Without batting.
there was a temperature drop at one deciency of 3 C. Even near the decien-
cies no temperature drop is recorded in this test. The berglass insulation
blocked some movement of air, and as well, may have insulated the thermo-
couple from what was occurring in the middle of the cavity. However, this
should only have dampened the effect, not caused it to disappear. As mentioned
before, the window frames were found to be of low quality with several de-
ciencies near the corners; it may be possible that a deciency causing the tem-
perature drop in the previous case became neutralized due to the caulking that
was redone after the insulation was installed. Nonetheless, there is no effect on
the condensation risk.
Effect of Deciencies and Pressure DifferencesWhilst there is no effect on

the outside surface temperatures, a pressure difference has a distinct effect on
the window frame temperatures on the inside. Figure 13 shows the tempera-
FIG. 13Temperature indexes at 40 Pa with deciencies for the different test sets.
ture indexes for the case where deciencies are present and a pressure differ-
ence of 40 Pa is applied across the test specimen. The results, provided from
top to bottom, include, respectively, the test without insulation between win-
dow frame and wall, with berglass insulation, and with SPF. Along the path
running from the outside deciency in the lower right corner to the deciency
on the inside top left corner, the temperatures on the window prole decrease
from about 2 to 3.5 C at 20 Pa and 2.5 to 4.5 C at 40 Pa. During the test
without insulation, the effect of air leakage across the window frame at 40 Pa
was less pronounced and was limited to 1 3 C. Thus, contrary to expectations,
higher ow rates caused lower temperatures, despite the fact that berglass
insulation was present and should reduce air movement. This can only be
caused by uncertainties in the experimental setup such as small cracks or other
deciencies. Figure 13 shows the lowest temperature indexes inside the cavity
for the case with berglass insulation in the cavity. The effect on the tempera-
ture index at the interior surface is negligible. Although it may not be clear
what causes the higher air ow rates, it is established that air leakage around
the window perimeter can considerably lower the window surface temperature.
Peak values of temperature shift between test pressure conditions are 3.3 and
5.4 C for 20 Pa and 40 Pa, respectively. The temperature index during a pres-
sure difference of 40 Pa is 0.63. This is very similar to the earlier results when
there is no insulation in the cavity. However, note that the temperature index
for an installation without insulation, without deciencies, and without pres-

sure differences was just slightly higher: 0.69. Even though a static pressure
difference of 40 Pa is quite high and not likely to occur, the effect on the indoor
surface temperatures is about 10 % of the overall applied temperature differ-
ence.
Test Set 6 Spray Polyurethane Foam

The application of SPF inside the cavity should prevent any convection from
occurring. The surface temperatures are similar to the case with berglass in-
sulation and thus the same hygrothermal criteria apply. The thermal gradient
along the vertical axis in the cavity between window and window rough open-
ing is slightly lower than in the previous cases. The SPF blocks all convection in
that cavity, but there still remains thermal stratication inside the window pro-
les as was discussed previously.
Effect of Deciencies and Pressure DifferencesAlthough there may be no

convection in the cavity, a pressure difference of 20 Pa allowed a reduction in
the surface temperature of the window prole, on average, of 1.2 C, and for 40
Pa, this reached 2.3 C. Again, in the absence of thermal effects in the cavity
around the window, the cause of this temperature drop lies within the window
frame itself. Only deciencies in the window could permit an airow that af-
fects the surface temperature at that location. The temperature drop was very
similar to the case with berglass insulation. At 40 Pa, the temperature drop
was more pronounced in the case of the SPF insulated window. As SPF gener-
ally insulates better than berglass, the cold air can extract less heat from the
cavity around the window, causing a greater heat ux towards the inside
through the window frame. This conrms the general principle that a thermal
bridge can be accentuated by insulating the surrounding components. The tem-
perature index was 0.67 and 0.65 for 20 Pa and 40 Pa, respectively. Deciencies
caused no change in the temperature prole with or without pressure differ-
ence. The SPF blocked any possible air ow around the cavity; hence, pressure
differences have no inuence on surface temperatures.
Concluding Remarks
A test protocol has been developed to determine the condensation potential of

windows based on the existing CSA A440.2-04 test standard, but that also in-
cluded a means to determine the effects of air leakage on the risk to condensa-
tion on windows. The windows were installed in a 2 4 ft opening of a 2
6 in. wood frame assembly, the assembly being typical of cold climate North
American construction practice. The installation details were those that pro-
mote the management of rainwater entry and incorporate such features as a
sloped sill, sill pan ashing membrane, and back dam. Air leakage across the
wall-window interface may increase the likelihood that condensation may form
on the window. Hence, deciencies simulating either the improper installation
of components or the premature failure of critical seals have been included in
the evaluation to verify the degree to which such openings inuence the risk to
condensation. The risk to condensation was rst determined in conditions
where no deciencies were present at the wall-window interface, and thereafter,
a series of defects were included that permitted air, in varying degrees, to pen-
etrate the interface. In each instance, the surface temperatures of the window
were monitored to establish any changes in comparison with the instance
where no defects were present. This series of experiments was rst conducted
with no insulation in the cavity between the window unit and the window
rough opening and thereafter with berglass batt insulation and SPF. This per-
mitted comparing the relative importance of insulating the cavity on the per-
formance in regard to the risk to condensation of these approaches to window
installation practice. The information developed from these tests provides guid-
ance to window manufacturers, window installers, and knowledgeable practi-
tioners investigating window deciencies and the effects of such deciencies on
thermal performance at windows.
The experimental results for determining the condensation potential of
anged windows when installed in a wood frame wall assembly are reported in
this study. The following observations and analysis were made through the set
of experiments carried out in this study.
The exterior side of the conguration was not sensitive to thermal ef-
fects induced by air leakage to the inside. This means that, e.g., the use
of IR scans may not be useful for visualizing convective effects in the
window-wall interface from the outside.
The temperatures on the IGU showed a signicant vertical thermal gra-
dient, and the spacer around the perimeter acted as an additional ther-
mal bridge causing low surface temperatures in all congurations. How-
ever, the IGU was not sensitive to the response that occurred inside the
cavity between the window frame and the rough opening.
The anged window used in the measurements was of lesser quality, as
several cracks and deciencies in the window frame, in certain in-
stances, directly affected results. Even for the installation with SPF
without deciencies, it was observed that the surface temperature on the
window prole dropped 2.3 C, possibly caused by insufcient airtight-
ness of the window frame.
Due to air ows around the window, there was a temperature drop up to
3.3 C 20 Pa and 5.4 C 40 Pa on the interior window prole. This
corresponds to a change in temperature index of 0.07 and 0.11, respec-
tively. As mentioned, half of this can be attributed to the low perfor-
mance window. It appears that the effect of cold air ow and air leakage
through the window frame have the same order of magnitude.
As the window installation was tested under severe conditions 50 C
and 40 Pa difference between interior and exterior climate, the overall
effect of air ows is rather limited. However, these results are only valid
for the vinyl frame window used in this study.
Convective air transport around the window was not sufciently re-
tarded by the installation of berglass insulation. Only the use of SPF
insulation correctly provided a seal to the perimeter, thereby avoiding
cooling the window prole.
This paper reports experimental results for window-wall interfaces with

anged windows; results on box windows will be analyzed and reported
in a subsequent publication.
Future research will focus on the validation of dynamic simulation tools
and convective heat transfer coefcients to assess the condensation risk
for different interfaces. Based on the experimental data and dynamic
simulations, comprehensive design guidelines will be published to as-
sess the condensation risk based on outdoor temperature, indoor tem-
perature, and internal moisture loads.
Acknowledgments
The writers wish to thank the Canada Mortgage and Housing Corporation for
partial funding of the project and acknowledge the contributions of Mr. Silvio
Plescia, Senior Researcher at the Canada Mortgage and Housing Corporation,
to this paper.
References
1 Sasaki, J. R., Developing a Standard Test for Window Condensation Perfor-

mance, Mater. Res. Stand., Vol. 1110, 1971, pp. 1720 NRCC-11847 DBR-RP-
473.
2 AAMA 1502.3-1972, 1972, Voluntary Test Method for Condensation Resistance of
Windows, Doors and Glazed Window Sections, Architectural Aluminium Manu-
facturers Association, Schaumburg, IL.
3 AAMA 1503-98, 1998, Voluntary Test Method for Thermal Transmittance and
Condensation Resistance of Windows, Doors and Glazed Wall Sections, Architec-
tural Aluminium Manufacturers Association, Schaumburg, IL.
4 ASTM C1199-00, 2000, Standard Test Method for Measuring the Steady-State
Thermal Transmittance of Fenestration Systems Using Hot Box Methods, Annual
Book of ASTM Standards, Vol. 04.06, ASTM International, West Conshohocken, PA.
5 CSA A440.2-04, 2004, Energy Performance of Windows and Other Fenestration
Systems, Canadian Standards Association, Mississauga, ON, Canada, 108 p.
6 Elmahdy, A. H., A Universal Approach to Laboratory Assessment of the Conden-
sation Potential of Windows, Sustainable Energy Choices for the 90s: 16th Annual
Conference of the Solar Energy Society of Canada, Halifax, NS, Canada, 1990, Solar
Energy Society of Canada, Ottawa, ON, Canada, pp. 165173.
7 FRAME, A Computer Program to Evaluate the Thermal Performance of Window
Frame Systems, Version 4.0. 1996. Enermodal Engineering, Ltd, Kitchener, ON,
Canada.
8 VISION 4.0, Glazing System Thermal Analysis: Users and Reference Manual. 1996.
University of Waterloo Advanced Glazing System Laboratory, Waterloo, ON,
Canada.
9 BISCO 10.0w Computer Program to Calculate Two-Dimensional Steady State Heat
Transfer in Free-Form Objects. 2009. Physibel, Adegem, Belgium.
10 McGowan, A. G. and Wright, J. L., Computer Simulation of Window Condensa-
tion Potential, Proceedings of Thermal Performance of the Exterior Envelopes of
Buildings VII, Sheraton Sand Key Hotel, Clearwater Beach, FL, 1998, American
Society of Heating, Refrigerating and Air Conditioning Engineers ASHRAE, At-

lanta, GA.
11 EN ISO 10211, 2007, Thermal Bridges in Building ConstructionHeat Flows and
Surface TemperaturesDetailed Calculations, CEN, Brussels, Belgium.
12 EN ISO 13788:2001, 2001, Hygrothermal Performance of Building Components
and Building ElementsInternal Surface Temperature to Avoid Critical Surface
Humidity and Interstitial CondensationCalculation Methods, ISO, Geneva, Swit-
zerland.
13 Elmahdy, A. H. and Bowen, R. P., Laboratory Determination of the Thermal Re-
sistance of Glazing Units, ASHRAE Trans., Vol. 942, 1988, pp. 13011316.
14 Brown, W. P., Solvason, K. R., and Wilson, A. G., A Unique Hot-Box Cold-Room
Facility, ASHRAE Trans., Vol. 67, 1961, pp. 561577.
15 Lacasse, M. A., Manning, M. M., Rousseau, M. Z., Cornick, S. M., Plescia, S.,
Nicholls, M., and Nunes, S. C., Results on Assessing the Effectiveness of Wall-
Window Interface Details to Manage Rainwater, 11th Canadian Building Science
and Technology Conference, Banff, AB, Canada, March 22, 2007, pp. 114 NRCC-
49201.
16 Lacasse, M. A., Rousseau, M. Z., Cornick, S. M., and Plescia, S., Assessing the
Effectiveness of Wall-Window Interface Details to Manage Rainwater, 10th Cana-
dian Conference on Building Science and Technology 10BS&T, Ottawa, ON,
Canada, May 12, 2005, pp. 127138 NRCC-47685.
17 Cornick, S. M., Task 5: Proposed Test Protocol for Walls of Houses in Extreme
Cold Regions. Part 1: Dening Exterior Conditions, Report No. B-1239.5, National
Research Council of Canada, Institute for Research in Construction, Canada, 2008,
pp. 147.
18 ASCE/SEI 7-05, 2005, Minimum Design Loads for Buildings and Other Struc-
tures, Structural Engineering Institute of the American Society of Civil Engineers,
ASCE, Reston, VA.
19 EN ISO 10077-2, 2003, Thermal Performance of Windows, Doors and Shutters
Calculation of Thermal TransmittancePart 2: Numerical Method for Frames,
CEN, Brussels, Belgium.
20 EN ISO 10456, 2008, Building Materials and ProductsHygrothermal Properties
Tabulated Design Values and Procedures for Determining Declared and Design
Thermal Values, CEN, Brussels, Belgium.
21 Khalifa, A.-J. N., Natural Convective Heat Transfer CoefcientA Review. 1. Iso-
lated Vertical and Horizontal Surfaces, Energy Convers. Manage., Vol. 42, 2001,
pp. 491504.
22 Gustavson, A., Grifth, B. T., and Arasteh, D., Three-Dimensional Conjugate Com-
putational Fluid Dynamics Simulations of Internal Window Frame Cavities Vali-
dated Using Infrared Thermography, ASHRAE Trans., Vol. 107, 2001, pp. 112.
doi:10.1520/JTE102971
Marcin Pazera1 and Mikael Salonvaara2
Drying Characteristics of Spray-Applied

Cellulose Fiber Insulation
ABSTRACT: Cellulose ber insulation CFI can be installed as a loose-

blown or spray-applied product. Spray-applied CFI is installed in a wet or
damp form with water and sometimes adhesives used as bonding agents.
The spray application leads to a more uniform and homogenous product with
lower density in comparison to loose-blown insulation. It is self-supporting
and does not need permanent retainers and can lead to cost savings for the
installers. In new construction, spray-applied CFI is gaining acceptance and
popularity in northern regions of the United States. Installations are per-
formed year round even in cold regions of the country. Walls are often en-
closed with interior nishes shortly following the installation of spray-applied
CFI. It is commonly assumed that moisture contained in the CFI will dry out
within a short period of time. Limited information exists in the public domain
regarding the drying characteristics of walls with spray-applied CFI insulation
and how proper drying rates can be achieved. The hygrothermal response of
a typical residential wood-framed wall was investigated using combined nu-
merical and empirical approaches. Hygrothermal computer model was em-
ployed to examine drying rates and residual moisture contents in the CFI and
exterior sheathing. Laboratory tests were conducted to quantify and validate
hygrothermal storage and transport properties of CFI. The results show slow
drying rates for spray-applied CFI installed in cold weather. The drying period
for CFI to reach moisture content levels safe for the installation of interior
nishes can exceed the recommended 2448 hr. In fact, the walls will not
fully dry unless the relative humidity of the interior air is maintained below the
critical limit, which is dependent on the outdoor temperature. The current
practice of measuring the moisture content at the surface or at the center
depth of the cavity is not representative of moisture contents at other loca-
tions and is a misleading indicator whether the insulation in its full depth is
sufciently dry for the installation of gypsum wallboard.
Manuscript received January 11, 2010; accepted for publication August 14, 2010; pub-
lished online September 2010.
1
Ph.D., Simpson Gumpertz & Heger, Rockville, MD 20850.
2
Owens Corning.
Cite as: Pazera, M. and Salonvaara, M., Drying Characteristics of Spray-Applied
Cellulose Fiber Insulation, J. Test. Eval., Vol. 39, No. 2. doi:10.1520/JTE102971.
59
KEYWORDS: cellulose ber insulation, built-in moisture, condensa-

tion, moisture transport, vapor pressure
Introduction
Spray-applied cellulose ber insulation CFI has gained popularity during the
past several years. The installers perceive numerous benets with spray appli-
cation in comparison to dry blown installation. The density of a cured spray-
applied product is lower; it is self-supporting and does not require permanent
retainers, which translates into cost savings for the installers. In order to avoid
costly delays interior nishes may be installed shortly following the application
of spray-applied CFI. Limiting the drying out period can result in high quantity
of residual moisture to remain in CFI, which might lead to moisture related
problems such as condensation on the inboard side of the exterior sheathing. In
instances when high moisture contents persist for prolonged period of time, the
risk for mold and microbial growth on materials adjacent to the CFI increases.
Spray-applied CFI is installed damp within the stud cavity. During instal-
lation, water is sprayed in a controlled manner onto the dry insulation at the
exit point of the applicator hose. Moisture activates the dry adhesives within
the CFI matrix and imparts cohesion between bers. The excess CFI is evenly
trimmed with the inboard surface of the studs and reclaimed. Prior to the
installation of interior nishes i.e., gypsum wall board, CFI remains exposed
for a period of time to allow the residual moisture to dry out. In current prac-
tice, the CFI manufacturers typically recommend 2448 hr timeframe as a suf-
cient period to dry the residual moisture. The CFI manufacturers and Cellu-
lose Insulation Manufacturer Association CIMA 1 provide limited
information regarding the drying periods and safe moisture content levels in
CFI for installation of interior nishes, specically in cold climate/weather ap-
plication. Typically, manufacturers recommend a maximum of 3540 % initial
moisture content at the time of installation and recommend the installation of
gypsum wallboard when moisture content reaches 25 %. One manufacturer
prescribes initial moisture content to be 30 % with 10 % reduction prior to the
installation of gypsum wallboard. Another manufacturer species two moisture
content levels for permeable walls and walls containing low permeability layers
on both sides of the wood-framed wall. Controlling moisture content at the
time of installation is not simple, owing to the fact that measuring the moisture
content of the brous insulation in the eld is not accurate. The technical lit-
erature provided by some manufacturers includes detailed recommendations
for the use of heating devices in cold weather/climate applications and dehu-
midiers in high-humidity conditions.
Despite these recommendations, detailed guidelines that provide compre-
hensive drying recommendations for varying climatic conditions specically
cold weather/climate applications, types of wall systems, and initial moisture
contents are missing. CFI installed in warm and relatively dry environments
can exceed 1 or 2 weeks before moisture content decreases to acceptable mois-
ture content levels. The drying rates can increase substantially, prolonging the
PAZERA AND SALONVAARA, doi:10.1520/JTE102971 61
drying period from days to weeks. In addition, the measured moisture contents
in the eld can be in excess of the recommended levels. A higher moisture
content in the CFI at the time of installation can result in higher dry densities,
and thus the actual volumetric moisture contents can be two to three times
greater than recommended. In addition to higher densities and reduced cover-
age, prolonged drying of the installed product is expected. Inadequate drying of
CFI in cold climates can result in high moisture levels and interstitial conden-
sation near or at the inboard surface of the exterior sheathing. High moisture
content can lead to linear expansion and swelling of oriented strand board
OSB sheathing as well as increased risk for mold and microbial growth on
surface of materials adjacent to CFI. These factors not only impact the service
life and durability of building enclosure but also could affect indoor air quality.
Objectives
This paper is focused on drying characteristics of spray-applied CFI following

its installation under different exterior and interior climatic conditions. A resi-
dential wood-framed assembly with OSB for exterior sheathing and spray-
applied CFI was investigated with specic focus on the drying rates and mois-
ture content proles as a function of outdoor temperature.
Approach
A combined numerical and experimental approach was used in examining the

drying characteristics of spray-applied CFI. First, laboratory tests including
sorption isotherm, water vapor transmission, and water absorption coefcient
were conducted to quantify moisture storage and transport properties. Second,
drying experiments with a 44.5 121.9 cm2 wood-framed cavity lled with CFI
were carried out under controlled temperature, relative humidity RH, and air
ow. Third, the measured hygric properties were used as input data to the
subsequent drying simulations, and the calculated RH was compared with that
determined in laboratory experiments. This critical step in the analysis pro-
vides validation of material properties. If the comparison between the mea-
sured and predicted drying rates shows good agreement, simulations of walls
can be carried out to examine the risk for interstitial condensation on the in-
board surface of the exterior sheathing, drying characteristics, and impact of
moisture in the CFI on adjacent materials in the wall system.
Drying Experiment
A drying experiment was carried out to determine the drying rates for CFI and
monitor changes in temperature and RH at the upper surface of the CFI and at
the transition between the upper surface of the OSB sheathing and the lower
surface of the CFI. The experimental set-up consisted of a rectangular cavity
FIG. 1Drying experimental set-up consisting of wood-frame cavity, OSB sheathing,

and CFI with sensors mounted on the upper surface of the OSB sheathing between the
sheathing and the insulation and on the upper surface of the cellulose.
measuring 44.5 121.9 cm2 constructed out of 3.80 8.90 cm2 wood-frame
and OSB sheathing on one side Fig. 1. All wood-framed cavity surfaces were
painted with ve coats of vapor impermeable paint to impart unidirectional
drying and prevent moisture from being absorbed into the frame and sheath-
ing. Prior to the installation of spray-applied CFI, three wireless temperature
and RH sensors were mounted on the inboard surface of the sheathing two at
both ends and one at the center of the cavity to log conditions during the
drying test. Following the CFI installation, three more wireless sensors were
placed on the top surface, and the test cavity was immediately moved into the
chamber and the drying test was started. Since the surfaces of the test frame
were sealed and evaporation only occurred through the top CFI surface, mois-
ture near the bottom of the cavity at the inboard surface of the sheathing
remained at higher moisture content the longest.
The drying experiment was conducted in an environmental chamber with
temperature, RH, and air ow rate set to 23 0.2 C, 50 2.5 %, and
0.2 0.02 m / s. The upper surface of the test assembly was exposed to the air
ow, and drying occurred through the top surface of the CFI. Constant condi-
tions in the chamber were maintained using a dedicated environmental control
system mounted on the top of the chamber Fig. 2. The system consisted of a
centrifugal fan, ultrasonic humidier, heaters, and cooling coil. The chamber
functioned on the basis of displacement ventilation with one side of the cham-
ber serving as air inlet and the opposite side functioning as air outlet. In a
closed loop, the conditioned air was forced through the inlet using a centrifugal
fan. At the chamber outlet, the air was recirculated through the HVAC system.
Part of the air was bypassed in the system, and the remaining part was recon-
ditioned to maintain the required temperature RH and air ow. At the chamber
outlet, a portion of the air is bypassed, and a portion of the air is recirculated
through the environmental control system where it is cooled and dehumidied
as it passes through the cooling coil, mixed with the bypassed air, reheated, and
FIG. 2Schematic showing the environmental chamber and environmental control

system used in controlling temperature, RH, and air ow.
humidied. The instantaneous on/off response of the humidier is a critical

factor in a precise process control, an ultrasonic stainless steel humidier,
model ENU 1200 was utilized. The temperature and RH were measured in the
duct upstream of the chamber inlet using HMT100 probe with 0.2 C tem-
perature accuracy at 20 C and 1.7 % and 2.5 % accuracy for RH ranging
between 0 % and 90 % and 90 % and 100 %, respectively. The air ow was
measured at the center of the chamber and 5 cm above the surface of the
specimen using HHF2005HW hot wire anemometer with an accuracy of 0.02
m/s. The chamber was insulated to reduce heat transfer through its walls and
provide improved temperature uniformity. The environmental system was con-
trolled using a SERIES SD proportional-integral-derivative controller.
An air stream straightener was installed at the inlet to the chamber to
provide laminar air ow above the specimen surface and limit variations
brought about by turbulence.
Hygrothermal Simulations
The drying simulations were performed to validate CFI properties by matching

the measured RH. Simulations of wall assemblies were carried out to examine
short and long term hygrothermal responses of walls with spray-applied CFI.
The parameters in the drying simulations were selected from the drying experi-
ment.
100
Relative Humidity [%]

75
50
Monitorpos. 3
Monitorpos. 4
25 Measured Course 2
Measured Course 3
Measured Course 5
0
0 2.3 4.6 6.9 9.2 11.5 13.8
Time [days]
FIG. 3Drying of CFI lled wood-framed assembly. Comparison of measured and pre-
dicted RH from simulations near exterior sheathing surface facing the interior.
Results and Discussion
Drying Experiments in the Laboratory

Figure 3 shows measured RH at three locations in the CFI near the inboard
surface of the exterior sheathing and RH predicted with simulation. Two moni-
tors were set in the CFI near the sheathing surface 1.25 and 2.54 cm from the
sheathing. The resultant plots show a good agreement between measured and
simulated RH. Under controlled laboratory conditions, 23 C temperature and
50 % RH a period of 1 week were required to reduce RH across the entire
insulation depth from 100 to 50 %. These results indicate a relatively slow
drying period even with favorable drying conditions such as exposed CFI to a
continuously ow of air above its surface.
Effect of Material Properties on Drying Spray-Applied Cellulose Fiber

Insulation
The accuracy of predicted results is strongly affected by the precision of input
parameters. Moisture storage and transport in the hygroscopic range of mois-
ture contents are critical material properties. Salonvaara et al. 2 employed
stochastic modeling to evaluate how variations in material properties affect
moisture distribution in a stucco wall. The results indicated that a 20 % varia-
tion in material properties of the exterior stucco layer resulted in a 45 %
variation in the moisture content of the OSB layer. Varying the properties of
both exterior stucco layer and OSB by 20 % resulted in 50 % range in the
predicted moisture content 50 110 kg/ m3 of the OSB sheathing. Salonvaara
and Karagiozis 3 attributed the accuracy of sorption curve as one of the im-
portant factors affecting reliability of simulated results. Small variations in
sorption curves were also reported to have a large effect on simulated drying
100
RH3 no liquid transfer
RH4 no liquid transfer
RH3 with liquid transfer

75 RH4 with liquid transfer
50
25
0
0 2.3 4.6 6.9 9.2 11.5 13.8
Time [days]
FIG. 4Drying of cellulose lled CFI stud wood-framed stud cavity. Comparison of
predicted from simulations RH levels near the exterior sheathing surface facing the
interior.
4. These research ndings highlight that accurate material properties are im-
portant to the improvement of modeling results.
Liquid transport properties in low density and brous materials such as
insulations are rarely measured and are typically deemed unnecessary for sev-
eral reasons. First, these insulations are installed in the dry cavity inboard of
the wetness plane, and it is assumed that these will not come into contact with
bulk moisture. Second, the materials are non-homogenous and compressible
and pose difculties in measuring moisture transport. Third, materials are typi-
cally not capillary active and do not absorb liquid water into the material ma-
trix. Fiberglass and mineral wool insulations are examples of insulations that
do not absorb water in any considerable amounts. CFI insulation is strongly
absorptive, and liquid transport is a critical material characteristic that cannot
be neglected if reliable simulation results are expected.
A drying experiment was simulated again with two sets of material prop-
erty data, sets that included and excluded liquid transport property. The RH
plots in Fig. 4 show a good agreement between the measured and simulated
results with CFI that included liquid transport properties. However, a consid-
erable deviation from the measured data was obtained in simulations with
material property data that excluded the liquid transport. The data shows a
slower drying i.e., higher RH levels of the CFI with exclusion of liquid trans-
port, which is expected.
The results emphasize the importance of reliable and accurate material
characteristics in simulations and highlight the critical role in modeling that
verication methods have as quality control steps. The validated CFI properties
including liquid transport were used in subsequent simulations.
Effect of Exterior and Interior Climate on Drying Spray-Applied

Cellulose Fiber Insulation
Series of simulations were carried out to assess the effect of exterior and inte-
rior climatic conditions on the drying characteristics of spray-applied CFI and
FIG. 5Materials in the simulated wall assembly.
the adjacent materials in typical residential wall. The focus was on evaluating
the ability of spray-applied CFI to dry immediately after its installation. The
wall system consisted of the following components listed from the exterior to
the interior of the building enclosure: Spun bonded polyolen water resistive
barrier, 1.25 cm OSB sheathing, and 3.8 14.0 cm2 wood-frame wall with
spray-applied CFI. The cladding was omitted since insulation is often installed
prior to the installation of exterior nishes. The gypsum wallboard was ex-
cluded to allow the CFI to dry out to the interior. Figure 5 shows the compo-
nents of the simulated wall assembly.
Effect of Exterior Climate

Simulations were carried out to evaluate the impact of exterior climate on
moisture response of walls with specic focus on determining the drying char-
acteristics of CFI and its impact on the adjacent exterior sheathing. The loca-
tion i.e., city and state and time of year installation were considered as the
two critical parameters that affect the hygrothermal response of walls. A series
of 12 simulations i.e., one for each month of the year represented installation
at different time during the year. Beginning with October, each simulation was
started at the beginning of a subsequent month and run for 3 months. Only the
starting and the ending points were changed to account for year round instal-
lation of CFI. The simulations were carried out with weather data cold year
for Detroit Michigan. The interior climate conditions were derived from exte-
rior weather data in accordance with ANSI/ASHRAE 160 5. The air condition-
ing system was set to heating mode only with temperature set-point of 40 F
and ventilation rate of 2 air changes per hour ACH with low moisture genera-
tion rate representing typical moisture generation rate of a two bedroom
house. The base case simulations assume a well ventilated house that has
almost the same absolute humidity indoors as in the outdoor air. These param-
FIG. 6Predicted moisture content proles in OSB sheathing during a 3 month simu-
lation period with CFI installed during each month of the year.
eters represent typical recommended interior climatic conditions during the

installation of CFI insulation. Typical built-in moisture i.e., 80 % RH was
selected as the initial condition in all assembly components except in CFI. The
initial moisture content in CFI was set to 40 % by weight and represents the
upper limit of as-installed moisture content recommended by CFI manufactur-
ers and CIMA i.e., with CFI density equivalent to 40 kg/ m3 this implies adding
16 kg/ m3 of water.
The 14 cm insulation layer was subdivided into a 2.54 cm layer layer ad-
jacent to inboard surface of the exterior sheathing and 11.4 cm layer. Figures
68, respectively, show moisture content proles in OSB, 2.54 cm and 11.4 cm
layers of CFI. The plots show change in moisture content in the OSB for the 3
month simulation period and change in moisture content in the insulation
layers for a 1 week duration following its installation. In all cases, the moisture
content in the OSB increases during the initial 1 week period as moisture gra-
dient develops in the wall. The moisture content in the OSB increases from the
initial 12.515.5 % by weight with CFI installed during late Spring and early
Summer months i.e., May or June and up to 18 % by weight with CFI in-
stalled in the Fall and early Winter i.e., October or November. For CFI in-
stalled during late Spring or early Summer, the moisture content in the OSB
peaks after 12 weeks and stabilizes between 9 % and 11 % approximately after
1 month. For CFI installed during Fall and early Winter, the moisture content in
the OSB sheathing peaks and remains above 18 % for the 2 months. For CFI
FIG. 7Predicted moisture content proles in the 2.54 cm CFI layer 1 week following
its installation.
installed between October and December, the moisture content in the OSB
remains above the as-installed moisture content levels at the end of the 3
month simulation.
Moisture content proles in two insulation layers show a signicant differ-
ence 1 week into the simulation period. The 2.54 cm layer of CFI adjacent to
the exterior sheathing surface facing the building interior is at higher moisture
content than the remaining insulation. Moisture content proles in the insula-
tion i.e., layer adjacent to the exterior sheathing installed during the months
of December through March indicate no change during the rst week of drying.
During the same period, moisture content proles in the 11.4 cm insulation
layer show a reduction from initial 40 % to below 20 % in all simulated cases.
The CFI closest to the surface of the exterior sheathing remains at highest the
moisture content for the longest period. The plots also show that with late Fall
and early Winter installations, the moisture content in CFI near the sheathing
surface remains unchanged for several weeks, indicating slow if any drying
during this time period. This leads to a higher risk for occurrence of moisture
condensation on the surface of the sheathing. When exterior temperatures fall
to below freezing levels, the condensed moisture can freeze.
Effect of Indoor Climate

The effect of indoor temperature and RH was investigated with CFI installed in
January. Three different indoor temperatures heating set-points at 40, 50, or
FIG. 8Predicted moisture content proles in the 11.4 cm CFI layer 1 week following
its installation.
60 F, and three different ventilation rates 2.0, 1.0, and 0.5 ACH were intro-
duced to remove indoor moisture sources. The resulting indoor conditions are
presented in Fig. 9 for the month of January. The lower ventilation rates result
in higher absolute humidity of the indoor air than outdoors, subsequently re-
ducing the potential for the insulation to dry.
Maintaining a well ventilated house with 2.0 ACH, the expected average
indoor RH is 55, 37, or 26 % at respective indoor air temperatures of 40, 50,
and 60 F. Figure 7 shows the lower quartile, median, and upper quartile of the
RH range for the nine different indoor climates.
The results show that moisture content in the OSB increases throughout
the rst month following the installation of CFI. The two parameters evaluated
specically the indoor temperature and the ACH rate has a different affect on
moisture content in the OSB. Indoor temperature has a substantially lower
inuence on moisture content in the OSB sheathing in comparison to the ACH
Fig. 10. Increasing the indoor temperature from 40 to 50 and 60 F changes
the moisture content in the OSB by fraction of a percent. Increasing the ACH
on the building interior from 0.5 to 1.0 and 2.0 ACH decreases the moisture
content in the OSB from 19 to less than 17 % at the end of 1 month drying
period Fig. 10.
The moisture content trends show that the 11.4 cm CFI layer facing the
building interior dries out faster Fig. 12 in comparison to the 2.54 cm CFI
FIG. 9Indoor temperature and RH matrix for the sensitivity analysis of indoor cli-
mate on drying of CFI. Ventilation rate decreases from left to right, and temperature
increases from top to down 40, 50, and 60F.
layer adjacent to the OSB sheathing Fig. 11. The greatest temperature effect
occurs in the insulation layer closest to the building interior, which allows the
insulation to dry faster and results in lower nal moisture content. The interior
climatic conditions i.e., increase in heating set-point had the least effect on
drying of the 2.54 cm insulation layer adjacent to the exterior sheathing. The
higher indoor ACH provides faster drying of the CFI layer facing the building
interior.
Drying Versus Moisture Redistribution
The above results highlight the complex and interactive nature of moisture
response in walls with spray-applied CFI. Following its installation, the mois-
ture movement in walls is dominated by two processes: Drying and moisture
redistribution. Several factors including exterior and interior climatic condi-
tions, initial moisture content of the CFI, and permeability of the adjacent
material layers affect the drying characteristics of the spray-applied CFI. The
simulation results show that drying of CFI to the building interior is not always
FIG. 10Moisture content of OSB as a function of time with CFI installed in a house
with different indoor temperatures and ACHs.
the predominant direction of moisture transport. Although, consistent with CFI

installed during warmer Summer months or in warmer southern regions of the
United States, this is not the case when installations are completed in cold
weather. Figure 1315 show the temperature, RH, and moisture content distri-
butions in the wall after 1 week of drying in June, October, and January.
In 1 week, the total moisture in the walls decreased from 3.13 kg/ m2 to
2.18, 1.73, and 2.33 kg/ m2 in January, June, and October, respectively, result-
ing in 0.95, 1.4, and 0.8 kg/ m2 of total drying. Part of the moisture dried from
the surface of the CFI toward the interior of the building inward drying. Part
of the moisture dried through the exterior sheathing to the exterior of the build-
ing outward drying. The total moisture content is summed from the hourly
uxes calculated by the software 6 for the entire duration of the simulation
and represents the combined inward and outward drying.
Table 1 shows the amount of outward and inward drying at the interior and
exterior wall surfaces during a 3 month period with CFI installed in three dif-
ferent months. The data shows that drying to the interior dominates.
This shows that approximately, twice the amount of moisture dried out in
one week with June installation in comparison with colder month installation
i.e., January even in low indoor moisture load conditions. The results also
show that 90 % of the drying happens inward toward the building interior.
The stress on the exterior sheathing due to moisture is highest in the Winter
FIG. 11Moisture content of the 2.54 cm CFI layer closest to OSB as a function of
time after the installation of CFI in a house with different indoor temperatures and
ACHs.
months due to redistribution of moisture into the sheathing from the insulation
and low potential for drying toward the building exterior. Low temperatures
lead to low vapor pressures in the wall closest to the exterior. Potential for
drying through the exterior sheathing increases exponentially with outdoor
temperature increase.
Figures 1315 respectively show simulated temperature and moisture con-
tent proles 1 week following the installation of CFI in June, October, and
January. In all three cases, moisture dries out albeit at different rates as indi-
cated by the varying moisture prole histories. During cold weather installa-
tion, moisture in CFI is redistributed toward the exterior side of the assembly
Fig. 13. A steep moisture gradient develops in the insulation layer adjacent to
the exterior sheathing. A slightly steeper moisture content prole also develops
TABLE 1Amount of outward and inward drying during a 3 month period with CFI
installed in January, June, and October.
Outward Drying Inward Drying

Time of Installation kg/ m2 kg/ m2
January 0.24 1.84
June 0.26 1.82
October 0.15 1.59
FIG. 12Moisture content of the 11.4 cm CFI layer closest to indoor air as a function
of time after the installation of CFI in a house with different indoor temperatures and
ACHs.
in the OSB sheathing adjacent to the CFI. Figure 13 indicates a larger tempera-
ture difference in comparison to temperature proles in Figs. 11 and 12. During
cold weather, as interior temperatures falls below the 40 F set-point, the heat-
ing system turns on to maintain the interior temperature above the set-point.
The steeper temperature gradient in the wall contributes to greater vapor pres-
sure difference between the interior which is now maintained above 40 F and
colder exterior and a stronger driving potential for moisture transport. In cold
climate zones during late Fall and Winter, vapor pressures are higher on the
building interior than the exterior. The predominant direction of vapor pres-
sure drive is toward the exterior.
Initially, the interior surface of the CFI can dry toward the interior consecu-
tively; moisture inside the insulation is driven toward the exterior side of the
wall. Even in the presence of permeable exterior sheathing, cold weather sig-
nicantly reduces the drying potential of the built-in moisture due to the low
partial pressure of water vapor at these temperatures. Simulation results show
that drying toward the building interior is the predominant mode drying spray-
applied CFI. To allow moisture to dry to the building interior the partial pres-
sure of water vapor inside the insulation must be higher than that in the indoor
air.
In cold conditions, this is not often possible. The exterior surface of the CFI
is cold, and the partial pressure of water vapor is lower than on the building
interior. The required indoor RH as a function of outdoor temperature to allow
FIG. 13Temperature, RH, and moisture content proles after 1 week of drying with
CFI installed in June. A 0.26 kg/ m2 of moisture dries toward the building exterior, and
1.82 kg/ m2 of moisture dries toward the building interior in 3 months.
for drying of the exterior layers of the insulation toward indoors is shown in
Fig. 16 for warm indoor conditions. Maintaining the indoor temperature sev-
eral degrees higher than outdoor temperature and providing dehumidication
are the preferred approach of moisture removal of CFI. If dehumidication is
not a feasible option, then providing continuous ventilation using drier outdoor
air is also possible.
Moisture Content Measurements

In accordance with standard practice, moisture measurements should be per-
formed to verify that CFI is dry enough prior to the installation of gypsum
wallboard. In practice, either surface type moisture measuring devices or pin
type devices are recommended for measuring moisture content. With pin type
moisture metres, the center depth of CFI insulation is the recommended loca-
tion. The standard practice calls for moisture content in the insulation to be
equivalent to or less than 25 %.
The simulation results indicate that in cold weather installation, steep
moisture content proles develop in CFI layer adjacent to the exterior sheath-
ing, and the drying time can be reduced. When looking at the average moisture
contents in the insulation layers, it was found that in January, it took 232 and
CFI installed in October. A 0.15 kg/ m2 of moisture dries toward the building exterior,
and 1.59 kg/ m2 of moisture dries toward the building interior in 3 months.
69 hr to dry the insulation layers below 25 % 25.4 and 101.6 cm layers, respec-
tively. Similarly, in June, the drying took 58 and 47 hr. In October, the drying
occurred in 113 and 101 hr. The drying of the layer closest to the exterior
sheathing is slowest in the Winter during cold months, whereas the insulation
layer closest to the indoors dries slowest during the Fall.
The center depth location in the insulation layer dried to below 25 % mois-
ture content level in 118 hr in January, in 74 hr in June, and in 128 hr in
October. With CFI installed in Winter months, moisture content measurements
at the center depth are lower than moisture content of the insulation layer near
the exterior sheathing. The insulation layer closest to the exterior sheathing
and its drying rate becomes a critical factor during the colder months. Depend-
ing on the indoor conditions, this layer may not dry prior to the onset of
warmer weather. During the cold months, vapor drive from indoors in walls
without vapor retarders may increase the moisture content to even higher lev-
els.
Conclusions
Modeling results show that drying characteristics of walls with spray-applied
CFI are signicantly affected by the interior/exterior boundary conditions.
CFI installed in January. A 0.24 kg/ m2 of moisture dries toward the building exterior,
and 1.84 kg/ m2 of moisture dries toward the building interior in 3 months.
Warm weather conditions rarely cause any difculties in drying, but cold
weather can lead to prolonged high moisture contents in the insulation and
exterior sheathing. Spray-applied cellulose should not be installed during cold
seasons unless drying can be achieved in the full depth of the insulation layer
prior to installing interior nishes or unless it can be demonstrated that redis-
tribution of the initial moisture will not lead to high moisture contents in any
material layer. The spray-applied CFI installed in typical wood-frame walls will
dry more inward than outward. The walls will not fully dry unless vapor pres-
sure of the interior air is maintained below the critical limit, which is depen-
dent on the outdoor exterior sheathing temperature. Water vapor perme-
ability of the exterior sheathing will also affect the drying rate of the CFI lled
cavity.
The current practice of measuring moisture content at center depth of the
cavity is not representative of moisture contents at other locations and is a
misleading indicator whether the insulation in its full depth is dry enough.
Guidelines describing how drying of walls with CFI are dependent on the in-
door humidity and the outdoor temperature and how moisture content of the
insulation should be developed. Ignoring the moisture content of CFI and in-
T indoor=68F
Maximum allowable indoor Relative Humidity
100
Maximum allowable RH in doo r , % 90
80
70
60
50
40
30 Indoor RH that allows panel to dry
20
10
0
23 32 41 50 59 68
T outdoo r Tp an e l, F
FIG. 16Maximum indoor RH that allows any drying from the panel surface toward
the indoor climate which is at 68F. If indoor humidity is higher the panel will keep
getting wetter and drying can happen toward the outside only. Drying at cold tempera-
tures will be slow.
stalling gypsum wallboard soon after its installation without allotting enough
drying time under favorable drying conditions can lead to a multitude of mois-
ture related performance problems including dimensional changes in the exte-
rior sheathing and framing, moisture condensation and frost build up on the
surface of the sheathing as well as increased risk for mold growth.
Most material data sets exclude critical material properties such as liquid
transport for CFI. Typically, sophisticated and costly equipment such as
gamma-ray or nuclear magnetic resonance is required to quantify liquid trans-
port as a function of moisture content. These measurements have successfully
been performed with rigid materials such as sandstone, limestone, and brick.
However, large inaccuracies can be expected when measuring with compress-
ible and non-homogeneous materials such as CFI. Signicant experience and
knowledge base are required to perform these measurements correctly and to
analyze the results to derive the properties. Simpler water uptake tests quanti-
fying total moisture content as a function of time are typically used to estimate
the liquid transport properties. There is a need for another test, e.g., drying test,
to be used in conjunction with the water uptake test to validate the derived
property. Standardization of the test methods currently not available is needed.
References
1 Cellulose Insulation Manufacturers Association CIMA, Standard Practice for the

Installation of Sprayed Cellulosic Wall Cavity Insulation, CIMA Technical Bulletin
No. 3, Dayton, OH, 1998.
2 Salonvaara, M., Karagiozis, A., and Holm, A., Stochastic Building Envelope
ModelingThe Inuence of Material Properties, Performance of Exterior Enve-
lopes of Whole Buildings VIII CD-ROM, Clear Water, FL, Dec. 27, 2001,
ASHRAE, Atlanta, GA, p. 8.

3 Salonvaara, M. and Karagiozis, A., Inuence of Material Properties on the Hygro-
thermal Performance of a High-Rise Residential Wall, Proc. 1995 ASHRAE An-
nual Meeting, Chicago, January 1995, ASHRAE, Atlanta, GA, pp. 647655.
4 Wang, J. and Kumaran, M. K., How Well Should One Know the Hygrotermal
Properties of Building Materials?, Proceedings of CIB W40 Meeting, Prague, Czech
Republic, 1999, CIB W40, pp. 4752.
5 ANSI/ASHRAE 160, 2009, Criteria for Moisture Design Analysis in Buildings,
ASHRAE, Atlanta, GA.
6 Fraunhofer Institut Bauphysik, Moisture Design Tool for Architects and Engi-
neers, WUFI Pro 4.2., ORNL/IBP 2009.
79
ERRATUM for JTE102971, Journal of Testing and Evaluation, Volume 39, Issue 2,
March, 2011 and STP1498, Condensation in Exterior Building Wall Systems, published
May, 2011, ASTM International
DRYING CHARACTERISTICS OF SPRAYED-APPLIED CELLULOSE FIBER

INSULATION, M. Pazera and M. Salonvaara
Vol 39, Iss 2 Journal of Testing and Evaluation Correction:

Page 219, line 23 increase should be decrease
Page 220, FIG. 2, Colling Coil should be Cooling Coil
STP1498 Correction:
Page 60, last line increase should be decrease
Page 63, FIG. 2, Colling Coil should be Cooling Coil
doi:10.1520/JTE102995
Tetsuya Umeno1 and Shuichi Hokoi2
Moisture Damage in Vented Air Space of

Exterior Walls of Wooden Houses
ABSTRACT: Most of the exterior walls of wooden houses in Japan have a

vented air layer between the exterior cladding and the insulation. This vented
air layer is designed to dehumidify the exterior walls by discharging humidity
to the outside and allowing outdoor air to enter, thereby decreasing the risk
of condensation on the exterior wall during winter. It is assumed that the
source of this moisture is the indoor air, and that the outdoor air is drier.
However, the outdoor air is often highly humid during the rainy season and
may become a source of moisture. The vented air layer also allows rain
water to drain away. Any rain water penetrating through the exterior cladding
is drained away through the vented air layer. However, rain can also enter
the vented air space through the air inlets. Since it takes a certain amount of
time for all the rain water to drain away, water may accumulate in the vented
air layer and produce high humidity in the exterior wall. In order to evaluate
the effectiveness of an exterior wall with a vented air layer, its hygrothermal
characteristics should be investigated, taking the effect of rain water into
consideration. This paper describes a case of moisture damage where stain
appeared on the outer surface of the plywood wall in a wooden residential
building. Experiments were carried out in climate chambers to clarify the
conditions that are causing stain. Hygrothermal conditions in the vented air
layer were simulated using heat, air, and moisture model, and the causes of
staining were investigated.
KEYWORDS: vented air layer, wooden residential building, hygrother-
mal characteristics, moisture damage, climate chamber, numerical
analysis, stain
lished online October 2010.
1
Graduate School of Engineering, Kyoto Univ., Kyoto 619-0224, Japan.
2
Professor, Graduate School of Engineering, Kyoto Univ., Kyoto 619-0224, Japan.
Cite as: Umeno, T. and Hokoi, S., Moisture Damage in Vented Air Space of Exterior
Walls of Wooden Houses, J. Test. Eval., Vol. 39, No. 2. doi:10.1520/JTE102995.
80
UMENO AND HOKOI, doi:10.1520/JTE102995 81
FIG. 1Exterior wall with a vented air layer.
Introduction
Most of the exterior walls of wooden houses in Japan have a vented air layer
between the exterior cladding and the insulation Fig. 1. This vented air layer is
designed to dehumidify the exterior walls by discharging humidity to the out-
side and allowing outdoor air to enter, decreasing the risk of condensation in
the exterior wall during winter. A windbreak layer is installed between the
vented layer and the insulation layer to prevent any degradation of insulation
performance caused by the inow of outdoor air.
The vented air layer also aids waterproong by draining away any rain
water that penetrates the exterior walls. A pressure equalizing design mini-
mizes the amount of rain water penetrating through the cladding due to out-
door wind pressure, and any rain water that does enter is drained away through
the vented air space. The windbreak layer also acts as a waterproong layer.
Recently, the number of residential buildings that use structural plywood
on their exterior walls has been increasing. The plywood works as a windbreak
layer, and materials such as house wrapping sheets are then attached over the
plywood for waterproong purposes. The plywood also has the ability to absorb
and desorb moisture.
Several studies have been carried out on the vented air layer and the most
effective thickness for discharging water vapor to the outside. However, most of
these studies have only investigated the performance of vented air layers under
experimental conditions, not in a eld environment. This paper describes a case
of the moisture damage involving stain on the outer surface of the plywood of
the exterior wall in a wooden residential building. Two climate chambers were
used to clarify the conditions under which the stain appeared. The hygrother-
mal characteristics of an exterior wall with a vented air layer were examined
using numerical analysis, and the possible causes of condensation in the vented
air layer were investigated.
Moisture Damage in a Vented Air Layer of an Exterior Wall
Stain from Plywood

Figure 2 shows a case of moisture damage caused by dampness in the vented
air layer of an exterior wall in a 4-year-old wooden residential building. Red-
(a) Stain on the flashing (b) Bottom edge of house wrapping sheets
(c) Outside of house wrapping sheets (d) Inside of house wrapping sheets
(e) Distribution of stain under windows
(c),
Outdoor Indoor
Stain (e)
(f) Section (g) Distribution of stain on east elevation
FIG. 2Stain generation in a vented air layer.
dish brown water was found owing between the plywood and the house wrap-
ping sheets. The stain was shaped like marble pattern and distributed uni-
formly. The stain consists of organic materials, mainly phenol. Though it is not
clear whether the stain is derived from the veneer or the binder, it can be said
that the stain is from the plywood. The house wrapping sheets were moist and
wrinkled. The stain was mainly seen at the contact points between the plywood
and the house wrapping sheets. Furthermore, screws and staples in the ply-
wood had rusted.
Figure 2g shows the distribution of the stain on the east wall. Much stain
was found at the lower part of the full two-story walls. In contrast, the amount
of stain found under the windows was insignicant. The stain seemed to have
run down over the surface of the plywood. There was no evidence of conden-
Solar
radiation
Rain
Rain intrusion Elevated humidity in the Condensation on Stain entering the

through joints vented air layer plywood condensed water
House wrapping sheets

Vapor

Liquid Plywood
External
cladding Water Absorbed
Vented
Highly humid air from outside Air water
layer Stain
Wrinkle
FIG. 3Schematic diagram of stain generation.
sation on the inner surface of the plywood, insulation, or the posts and beams.
The walls near the eaves showed no sign of either condensation or leakage of
rain.
The stain usually appeared in houses with a thin vented air layer and was
mainly found on the east-facing walls. The stain was usually seen during the
rainy season or in Summer.
The stain does not necessarily damage the wall materials. However, the
appearance of the stain means the plywood is wet, which is strongly related to
fungi and decay of plywood. Also, many residents regard the stain as some
trouble, and claim to have been caused by the builders. The phenomenon that
residents require improvement should be treated as damages, and the causes of
stain are presumed as below.
Cause of Stain
Since the stain was distributed uniformly all over the surface of the plywood, it
was unlikely that rain water had directly leached out the chemical components
of the plywood. It seems more plausible that condensation occurred on the
outer surface of the plywood. As the inner surface of the plywood and the
insulation had not been damaged, the possibility of condensation due to mois-
ture from the indoor air was low. The initial water content of the construction
materials could not have been the source of the condensation problem because
the stain appeared a few years after the construction. For these reasons, we
concluded that condensation had occurred during the rainy season on the out-
side surface of the plywood. Figure 3 shows the proposed condensation pro-
cess. This is the phenomenon studied as solar vapor drive 1. The moisture
source of solar vapor drive is usually the moisture contained in the exterior
walls, which is originally from rain or condensation due to night cooling by
long-wave radiation. In the present case, the main source of water seems the
rain water penetrating through the joints of the cladding or the highly humid
air from the outside because the outside surface of the exterior walls is usually
well waterproofed in Japan.
1 Humidity of the vented air layer rises due to either rain intrusion or
highly humid outdoor air.
2 When the temperature of the exterior walls rises due to solar radiation,
Temperature of vented air layer

House Temperature of plywood
wrapping Plywood Relative humidity of vented air layer
sheets
50 100
Climate Climate
Temperature []
40 80
Relative humidity [%]

chamber 1 chamber 2
30 60
Vented air layer External surface of 20 40

insulation layer
10 20
0 0
6:00 9:00 12:00 15:00 Time
(b) Measured hygrothermal conditions
(a) Test wall in climate chamber
FIG. 4Experimental setup.
vapor desorbed from the exterior walls raises the humidity of the
vented air layer.
3 Condensation occurs on the outer surface of the plywood whose tem-
perature has not yet risen.
4 The condensed water is held for several hours on the outer surface of
the plywood due to the waterproong action of the house wrapping
sheets, and the stain then separates from the plywood.
5 The condensed water containing the stain is drained away through
gaps between the plywood and the house wrapping sheets.
The moisture source for the condensation may either be rain water or out-
door air or both. A temperature humidity difference between the vented air
layer and the plywood surface seems to be caused by solar radiation in the
morning, which explains why the stain is usually found on the east-facing
walls. Since humidity increases readily in spaces with a small volume and low
ventilation rate, the stain tends to appear in walls with a thin vented air layer.
Experiment Simulating Stain Generation

Experiments designed to simulate stain generation were performed in climate
chambers in order to check whether the proposed scheme above is valid.
Experimental Procedures
The experimental procedures were as follows. An external test wall, consisting
of a wooden post-and-beam framework covered in plywood and house wrap-
ping sheets, was constructed between two climate chambers Fig. 4a. The
temperature and humidity of climate chamber 1 were controlled in order to
simulate the vented air layer, and chamber 2 was used to simulate the external
surface of the insulation layer of the external wall. Figure 4b shows tempera-
ture and humidity measured in a vented air layer of a residence without any
stain appearance during the rainy season in Japan. The temperature of vented
air layer and surface of plywood in the experiment were based on these mea-
TABLE 1Experimental conditions.
Case 1 Case 2 Case 3

Vented air layer 30 C, 70 % RH 35 C, 70 % RH 40 C, 70 % RH
Insulation layer 20 C, 55 % RH 24 C, 55 % RH 26 C, 55 % RH
External surface
of plywood 26 C 30 C 34 C
surements. Table 1 lists the experimental conditions used. It is estimated that

the humidity of the vented air layer where the stain appeared was higher than
that shown in Fig. 4b and thus the relative humidity of vented air layer in the
experiment was assumed to be 70 % higher than that in Fig. 4b. Each case of
the experiments was performed for 1 week after 2 days exposure to 30 % rela-
tive humidity. The temperature, humidity, and moisture content of the plywood
were measured every 30 min, and the visual appearance of the test wall was
observed.
Experimental Results
In cases 1 and 2, neither condensation nor stain was observed on the plywood
surface. In case 3, under the severest test conditions, a small amount of con-
densation was observed on the plywood surface immediately after the start of
the experiment. The amount of condensation increased over time. After 24 h,
reddish brown water was observed on the ashing. Figure 5 shows the change
in the appearance of the test wall during the experiment in case 3. Condensa-
tion was seen on the plywood surface and stain appeared on the inside surface
of the house wrapping sheets, accumulating along the lines of wrinkles Fig.
5b.
The stain accumulated on the waterproof tapes at the bottom of the house
wrapping sheets Fig. 5c, and some drained through the gap between the
waterproof tapes Fig. 5d. After 1 week, mold was observed on the plywood
surface, and rust was found on the staples of the house wrapping sheets Fig.
5e and 5f.
Figure 6a and 6b shows the wall temperature and the dew point tem-
perature of the vented air layer Room 1 in cases 2 and 3. In case 2, where stain
did not appear, the temperature of the plywood was higher than the dew point
temperature throughout the experiment. In case 3, where stain was generated,
the surface temperature of the plywood was lower than the dew point tempera-
ture.
The moisture content of the plywood is shown in Fig. 6c. It was around 15
wt % before the experiment in case 3. However, the moisture content increased
rapidly during the course of the experiment and eventually reached 30 %.
The stain formation observed in the experiment was similar to the moisture
damage seen in the exterior walls of actual residences. This indicates that stain
could appear in the vented air layer of exterior walls when condensation oc-
curred on the plywood surface, and the condensed water remained there for
longer than 1 day.
(a) Stain on the flashing (b) Inside surface of a house wrapping sheet
gap
tape
(c) Stain backed up on waterproof tape (d) Draining through the gap in the tape
(e) Mold on plywood (f) Rust on a staple
FIG. 5Change in the appearance of the test wall in case 3.
Analysis of Condensation in the Vented Air Layer

Hygrothermal characteristics in the vented air layer were evaluated by numeri-
cal analysis, and the possibility of condensation caused by outdoor airow was
investigated.
Hygrothermal Calculations for an External Wall with a Vented Air Layer

Figure 7 shows the schematics of the simulated wall. One-dimensional heat and
moisture ow was considered. The hygrothermal model for simultaneous heat
and moisture transfer formulated by Matsumoto 2,3 was used. The moisture
balance equation 1 and the heat balance equation 2 were solved using a nite
difference method
X T
0 a + = X 1
t t
X T
r + c + r = T 2
t t
where:
35 40
Vented air
layer Vented air
layer
Temperature []
Plywood
Temperature []
30 35
Dew point
Dew point
Plywood
Insulation
25 30 Insulation
20 25
5/2
5/4
5/6
4/30
4/21
4/23
4/25
4/27
Date Date
(a) Temperature and dew point temperature (b) Temperature and dew point temperature
of vented air layer (case 2) of vented air layer (case 3)
40
Case 1 Case 2 Case 3
35
30
Moisture content [wt%]
25
20
15
10
Stain occured
5
0
4/14 4/21 4/22 4/28 5/1 5/2 5/7
Date
(c) Moisture content of plywood
FIG. 6Experimental results.

= w , = w
X T
0 = porosity m3 / m3,
a = density of dry air kg/ m3,
= moisture capacity to variation of humidity ratio kg/ m3 kg/ kg,
= moisture capacity to variation of temperature kg/ m3 K,
FIG. 7Schematics of the simulated wall.

TABLE 2Hygrothermal properties of wall materials.
Thermal Vapor Specic Heat

Conductivity Conductivity Density Capacity Porosity
W/mK kg/ms kg/ kg kg/ m3 J/kg K m3 / m3
Fiber cement 0.19 1.6 106 1000 840 0.53
Plywood 0.16 3.1 107 550 1880 0.60
Glass ber 0.038 2.2 105 16 840 0.99
Gypsum board 0.22 9.13 1010 700 870 0.70
t = time s,
X = humidity ratio kg/ kg,
T = temperature K,
= vapor conductivity kg/ ms kg/ kg,
r = heat of phase change J/kg,
c = heat capacity J / m3 K c = 0macpma + scps,
ma = density of moist air kg/ m3,
s = density of a solid kg/ m3,
cpma = specific heat of moist air,
cps = specific heat of a solid J/kg K,
= thermal conductivity W/mK, and
= moisture content kg/ m3.
The hygrothermal properties of the wall materials used are listed in Table
24,5, and Fig. 8a shows the sorption isotherms of ber cement and plywood.
With respect to the house wrapping sheets, the vapor transfer resistance was
omitted and the heat transfer resistance was included into the heat transfer
coefcient for the space between the plywood and the vented air layer. The
temperature and humidity for the vented air layer were calculated based on the
air exchange rate obtained from the ventilation calculation, taking into account
the chimney effect 6. The air velocity in the vented air layer was calculated by
Eq 3 considering buoyant force due to the temperature difference between the
vented air layer and the outdoor air
v=
B2
2B1
+ PT B22
B1
+
4B12
3
where:
B1 =
a
2
1 + Zin + Zout, B2 = 12
lh
l2w
, PT = T g lh
T a T o
Ta To
, Zin = 1
0.6in

1 ,
2
Zout = 1
0.6out
1 2
Zin = pressure loss coefcient of the inlet of the vented air layer -,
Zout = pressure loss coefcient of the outlet of the vented air layer -,
u = viscosity coefcient of the air Pa s,
lh = height of the vented air layer m,
lw = thickness of the vented air layer m,
50
Moisture content [vol %]

40 Fiber cement
Plywood
30
20
10
0
0 20 40 60 80 100
Relative humidity [%]
(a) Sorption isotherms
Temperature Relative humidity
100
80
Temperature []
60
40
20
0
3/11
3/21
3/31
4/10
4/20
4/30
5/10
5/20
5/30
6/19
6/29
7/19
7/29
8/18
8/28
3/1
6/9
7/9
8/8
(b) Outdoor temperature and humidity
East West South North
800
Solar Radiation [W/]
600
400
200
0
6/10 6/11 6/12
(c) Solar radiation measurements on walls
FIG. 8Material properties and outdoor conditions.
PT = pressure difference due to temperature difference N / m2,

g = acceleration of gravity m / s2,
Ta = air temperature of the vented air layer K,
To = outdoor temperature K,
in = opening ratio of the inlet of the vented air layer -, and
out = opening ratio of the outlet of the vented air layer -.
The height of the vented air layer was set at 6 m and the thickness was set
at 18 mm, as commonly used in wooden houses in Japan. A vented air layer
thickness of 4 mm was used for the moisture-damaged house. The discharge
coefcient was 0.7 at both ends of the vented air layer. The calculated tempera-
ture and humidity of the vented air layer were used as the boundary conditions
for the simulation of simultaneous heat and moisture transfer in the wall ma-
terials.
The calculation was carried out for the period from March to August the
period in which it was suspected that stain generation caused by condensation
occurred. Reference weather data for Tokyo 7 were used for outdoor tem-
perature and humidity values, and the monthly average temperature and hu-
midity ratio were used for the indoor air values. The solar radiation on the
external wall was estimated using global solar radiation values obtained from
FIG. 9Temperature of the plywood and dew point temperature of the vented air layer.
TABLE 3Air velocity in the vented air layer.
Thickness
18 mm 4 mm
Max. m/s 0.239 0.037
Min. m/s 0.000 0.000
Average m/s 0.039 0.004
Air exchange rate 1/h 23.65 2.47
reference weather data for Tokyo. Figure 8b shows the outdoor temperature
and humidity, while Fig. 8c shows the solar radiation on the test walls used in
the simulation.
Results
The calculated temperature of the external surface of the plywood and the dew
point temperature of the vented air layer for an east-facing wall are shown in
Fig. 9a and 9b. The temperature of the plywood was always higher than the
dew point temperature, and in both cases studied, condensation did not occur
on the plywood surface during the 6-month period. Figure 9c and 9d shows
the humidity ratio of the vented air layer. The humidity ratio varied due to the
absorption and desorption of moisture by the plywood and the ber cement.
The humidity ratio for a vented air layer 4 mm thick was lower than that for a
vented air layer 18 mm thick because the moisture inow from the outdoor air
was low.
The air velocity in the vented air layer is shown in Table 3. In the case of a
vented air layer 4 mm thick, the average velocity was 4 mm/s, corresponding to
an air exchange rate of 2.47 1/h. In the case of an air layer 18 mm thick, the
corresponding values were only 10 % as large.
Figure 10 shows the time variation of the temperature and humidity distri-
butions inside the wall on July 23rd for a vented air layer 18 mm in thickness.
The plywood temperature varied slowly and fell below that of the vented air
layer after 7 a.m. The humidity ratio of the vented air layer rose higher than
that of the outdoor air after 6 a.m. due to incoming moisture from the ber
cement.
Figure 11 shows the calculated difference between the plywood tempera-
ture and the dew point temperature of the vented air layer. The temperature of
the plywood surface was higher than the dew point temperature in every case,
regardless of the orientation of the exterior walls and the thickness of the
vented air layer.
The calculated results show that condensation on the outer surface of the
plywood did not occur due to the moisture inux from the outside and that the
possibility of stain generation was low. However, since the calculated results
showing that the possibility of condensation was lower with an air gap 4 mm
thick than with an air gap 18 mm thick do not agree with the observed results,
rain penetration through the exterior cladding rather than condensation
could be the reason for the stain generation.
FIG. 10Time variation of the temperature and humidity ratio 18 mm.
Conclusions
This study claried stain generation on the plywood facing the vented air layer
of the external wall of a wooden residential building and investigated the
mechanism of stain generation by means of experimentation and numerical
analysis.
Experiments carried out in climate chambers simulated actual eld situa-
tions and claried the conditions responsible for stain generation. Stain forma-
tion was observed on the plywood when condensation on the plywood re-
mained in place for longer than one day.
The hygrothermal characteristics of the vented air layer were evaluated by
means of numerical analysis, and the possible causes of condensation in the
vented air layer were investigated. It was concluded that outdoor airow into
the vented air layer could not be the cause of condensation leading to the stain
generation on the plywood. Instead, rain penetration through the exterior clad-
ding was a more probable cause of stain generation.
FIG. 11Temperature differences between the surface temperature of the plywood and
the dew point temperature of the vented air layer.
References
1 Wilson, A. G., Condensation in Insulated Masonry Walls in Summer, Proceedings

of RILEM/CIB Symposium Moisture Problems in Buildings, Helsinki, 1965, NRCC,
Canada.
2 Matsumoto, M., Heat and Moisture Transfer in Porous Materials with Airow,
Summaries of Technical Papers of Annual Meeting, 1971, Kinki Chapter of Archi-
tectural Institute of Japan, Japan, Ossaka, pp. 9396.
3 Matsumoto, M., Hokoi, S., and Ka, E., An Analysis of Coupled Heat and Moisture
Transfer in Buildings Considering the Inuence of Radiant Heat Transfer,
ASHRAE Trans., Vol. 1, Part 1, 1997, 573583.
4 Kumaran, M. K., Heat, Air and Moisture Transfer Through New and Retrotted
Insulated Envelope Parts, Final Report, Vol. 3, IEA ANNES 24, 1996.
5 Architectural Institute of Japan, Architectural Design Data Corpus, No.1 Environ-
ment, 1978.
6 Akasaka, H. and Takeda, K., Calculation Method for the Evaluation of Sunshad-
ing and Insulating Effect of the Walls and Roofs with Vented Air-Layer, J. Envi-
ron. Eng., Vol. 595, 2005, pp. 33-40.
7 Architectural Institute of Japan, Expanded AMeDAS Weather Data, Maruzen Ltd.,
Tokyo, 2000.
doi:10.1520/JTE103053
Theresa A. Weston1 and Liza C. Minnich2
Moisture Measurements and Condensation

Potential in Wood Frame Walls in a
Hot-Humid Climate
ABSTRACT: It has long been noted that interior vapor barriers in wood frame
walls in hot-humid climates can lead to interstitial condensation within walls.
The bases for this recognition are predictive simulations, anecdotal observa-
tions, and a limited number of experimental studies. This paper describes an
experimental study conducted in a hot-humid climate that investigated the
inuence of an interior vapor retarder and compares observed performance
with simulation predictions. The wall performance data reviewed here was
gathered as part of a larger test program evaluating the performance of a
range of typical wood frame, residential wall constructions in a hot-humid
climate. The approach chosen was to use real-time eld exposure using a
test hut located in Tampa, Florida. The test hut had two long sides, which
provided the ability to test 16 wall specimens each. Wall specimens were
instrumented with a variety of temperature, humidity, and moisture sensors.
In addition to natural weather exposure, the wall specimens could be manu-
ally wetted by a water injection system to simulate rain leakage. More spe-
cically, this paper focuses on using the data collected before and after the
installation of an interior vapor barrier vinyl wallpaper to show the change in
moisture loading and the potential condensation within the walls resulting
from the installation. The eld data is compared with predictions of the wall
behavior using a commonly available hygrothermal model. There is increas-
ing reliance on the use of predictive models to assess the moisture perfor-
mance of building assembly designs. These predictive models need to be
validated against real data to test their variance from real systems.
KEYWORDS: condensation, water management, wall assemblies
Manuscript received February 26, 2010; accepted for publication October 8, 2010; pub-
lished online xx xxxx.
1
PhD., Research Fellow, DuPont Building Innovations, Richmond, VA 23261.
2
Senior Research Technologist, DuPont Building Innovations, Richmond, VA 23261.
Cite as: Weston, T. A. and Minnich, L. C., Moisture Measurements and Condensation
Potential in Wood Frame Walls in a Hot-Humid Climate, J. Test. Eval., Vol. 39, No. 3.
doi:10.1520/JTE103053.
94
WESTON AND MINNICH, doi:10.1520/JTE103053 95
Introduction
It has been long noted that the use of an interior vapor retarder or vapor im-
permeable vinyl wallpaper on internal wall surfaces can lead to moisture ac-
cumulation and mold growth in the gypsum wall-board within the wall systems
13. Water may enter the wall through leaks or when humid air leaks into the
wall. This water is then trapped in the wall by impermeable building materials.
An extensive simulation study of the use of vapor barriers was conducted to
support changes in the requirements for vapor barriers and retarders in the
International Energy code. The results showed that although interior vapor
retarders were benecial in cold climates, there was signicant potential for
moisture damage from condensation when vapor retarders were used in
warmer, mixed-climates 4. Hot-humid climates were not included in this
study.
Cautions of the use of interior vapor barrier are more important in hot-
humid climates 5,6. Hot-humid climates, as dened in the International Resi-
dential Code IRC, include a signicant area of the United States including
Florida, the gulf coast regions of Texas, Lousiana and Missippi, and the Atlantic
Coast of South Carolina and Southern North Carolina 7. The IRC exempts
hot-humid climates from interior vapor barrier requirements.
The use of hygrothermal modeling has increased in recent years. It is being
promoted to examine the robustness of building system design. Accompanying
the increase in the use of models is the increase in the publication activity of
standards and manuals, which govern the use of simulation models. This is
illustrated in the publication of ASTM Manual 40 in 2001 and ASHRAE Stan-
dard 160 in 2009 8,9. Although many commercially available simulation mod-
els have been benchmarked, their use requires exact material properties and
weather conditions for good agreement between simulations and measured
data 5.
This study examines the predictions of one commonly used model and
compares the measured eld data to examine the accuracy of its predictions of
condensation and moisture accumulation. More specically, it examines the
prediction of the effect of the application of vapor impermeable wall covering
to typical residential walls in hot, humid climates.
Experimental Method
Relocatable Building Enclosure Test Station

The Relocatable Building Enclosure Test Station RBETS is an ongoing test
program that supports the real-time eld exposure of residential wall systems.
Experimental data for this study was obtained from an RBETS located in
Tampa, Florida. Tampa, Florida, represents a hot-humid climate.
The test station was created from a sea-land container and was constructed
with regional building practices in mind. The long sides of the container were
modied and tted with 2 4 studs. These sides were long enough to accom-
modate 16 wall specimens on each side, with each wall specimen 2.4 m 96 in.
in height and 0.6 m 24 in. wide see Fig. 1. In order to take advantage of the
FIG. 1RBETS.
local weather patterns, the unit was orientated with the wall systems facing east
and west. The 16 different wall specimens were duplicated on each side of the
unit to assess the exposure differences. The test station conguration provided
for no overhang or other protection from roof run-off, which would typically be
in place in residential construction. To limit roof run-off a drip cap was in-
stalled at the top of the wall specimens, as shown in Fig. 2. The most problem-
atic area was the brick clad specimens, which protruded further than the other
FIG. 2Drip cap at top of wall.

FIG. 3Detail at top of brick wall specimens.
wall specimens see Fig. 3. The brick wall specimens on the west wall were the
only specimens that were observed to have signicant roof run-off.
The wall specimens are shown generically in Fig. 4. The wall specimens
contained R-13, unfaced berglass batt and painted gypsum on the interior.
Four different claddings were studied on this unit: Vinyl siding, ber cement
siding, stucco, and brick. Several different water-resistive barriers housewraps
were used behind the claddings and over oriented strand board OSB sheath-
ing. All materials were widely available and commercially purchased.
Each wall specimen was equipped with a number of sensors measuring
temperature, relative humidity RH, and moisture content. Each sensor was
placed at a predetermined location and within each layer of the wall system
i.e., gypsum, interstitial, cladding face, etc.. The interior of the test station had
a HVAC system, which allowed the interior of the unit to have a controlled
temperature and RH. Additionally, the units weather station tracked the wind
speed and direction, rainfall, temperature, and RH. Each side of the test station
had a driving rain gage and a horizontal solar sensor. All of this information
was monitored remotely through the test units data acquisition system. Lastly,
in order to include the effects of incidental water intrusion from possible con-
struction defects, each wall assembly contained a manual water injection sys-
tem located 1.2 m 48 in. from the bottom of the wall specimen. This system
simulated a point defect in the construction of the wall and water intrusion,
which may likely occur. A known amount of water was introduced either be-
tween the wall assemblys sheathing and water-resistive barrier or between the
wall assemblys sheathing and cavity insulation.
The walls in the RBETS were exposed to natural weather conditions for 18
FIG. 4Generic description of wall specimen construction gure not to scale.
months. The data for the full 18 month period is reported elsewhere 10. This
study examines only an 2 month period Aug. 14Oct. 11, 2007, in the
middle of which vapor impermeable vinyl wallpaper was installed in the inte-
rior of the wall assemblies. At the beginning of this period August 14, 700 mL
of water was injected between the OSB sheathing and the water-resistive bar-
rier. A duplicate water injection was conducted on September 11 after the in-
stallation of the vinyl wallpaper on the inside of the walls. Moisture conditions
at the junction of the insulation cavity and the interior of the gypsum wall-
board were measured by electrical resistance moisture pins in a wooden wafer
see Fig. 5. The use of wooden wafer moisture sensors as surrogates in appli-
cations where the materials do not allow for direct moisture measurement is a
common practice. The performance of this type of sensor has been evaluated
for use in eld moisture monitoring 11. The actual measurements were found
to depend on the material properties of the wood wafer, resulting in a slower
FIG. 5Wood wafer moisture sensors.

FIG. 6Measured exterior and interior temperature.
wetting response than drying response. This allows for better indication of
longer-term wetting rather than short-term or diurnal wetting events.
Experimental Results
Exterior and Interior Climate Conditions

The temperature and humidity measure by the on-site weather station and
inside the RBETS unit are shown in Figs. 6 and 7.
Moisture Conditions at the Gypsum Wall-board

Figures 813 show the moisture content of wood wafer, which monitored the
moisture conditions at the junction of the insulation and the gypsum wall-
board for the time period of August 14October 11. The differences in wall
performance due to the water-resistive barrier properties were found not to be
statistically signicant, and therefore moisture content was averaged across all
of the wall specimens of a given cladding for the purposes of this study. Diurnal
cycling of moisture conditions was observed at the back of the back surface of
the wall-board. The magnitude of this cycling may have been underestimated
by the previously noted wetting and drying time response of the wood wafer
sensors.
FIG. 7Measured exterior and interior humidity.
Effect of Impermeable Wall CoveringThe installation of impermeable wall

covering caused a dramatic increase in the moisture content of all for the wall
specimens. Whereas prior to the installation of the impermeable wall covering
the water injected was able to dry out of the wall, after its installation injected
water was trapped within the wall and collected in the gypsum wall-board,
causing its moisture content to continue rising. A time lag was observed be-
tween the water injection and the start of the rise in moisture content of the
wood wafer sensor. None of the walls reached condensation conditions by the
end of the test period. However, as program constraints required the testing to
be discontinued in October, it is not known if the moisture would have met
condensation conditions or whether eventually drying would have taken place.
Effect of Exposure Direction

The effect of exposure direction is shown in Figs. 811 for each of the tested
claddings. For each of the vinyl siding, ber cement siding, and stucco clad-
dings, the west facing exposure exhibited lower moisture content than the east
exposure. A higher level of moisture is consistent with weather data for Tampa,
Florida, which shows the prevailing driving rain load is from the east. Site
weather station measurements of total rain and directional driving rain are
shown in Fig. 14. These measurements conrm higher driving rain on the east
FIG. 8Moisture conditions at gypsum wall-board, vinyl siding wall assemblies.
exposure. The opposite effect of exposure was seen with the brick cladding
walls, which exhibited signicantly higher moisture content on the west expo-
sure. The cause of this difference in behavior was discovered at the end of the
test period when it was discovered that the west exposure brick walls had a
water leak at their roof interface. This leak caused a much higher level of water
entry into the wall than that caused by climatic exposure and the intentional
water injection combined. Because of this water leak, the west exposure brick
clad specimens were removed from the further analysis
Effect of Cladding
Figures 12 and 13 show the effect of cladding on the moisture content of the
gypsum wall-board. The highest moisture content is in the stucco walls
brick walls fiber cement siding walls vinyl walls. Reservoir cladding
showed higher moisture than non-reservoir cladding or sealed reservoir clad-
ding. The moisture content is also inuenced by the amount of air movement
or ventilation of the cladding, the stucco having the least ventilation and the
vinyl being a known leaky cladding. The effect of cladding seen in this experi-
ment was in agreement with past research 4,1214.
Visual Observations
The wall assemblies were disassembled November 1215, 1 month after the
cessation of data collection mid-October. Wall assemblies were visually exam-
FIG. 9Moisture conditions at gypsum wall-board, ber cement siding wall

assemblies.
ined for signs of moisture accumulation or mold growth. The only walls in
which visible mold was observed on the gypsum wall-board was the brick clad-
ding walls with western exposure. These walls exhibited much higher moisture
contents than the other walls and were later determined to have a leak at the
roof wall interface.
Computer Simulations of Wall Performance in Tampa, Florida

A series of modeling simulations was run to evaluate the potential for conden-
sation and/or moisture accumulation with and without vinyl wallpaper in walls
in the Tampa climate. The simulation model and inputs are described in
Table 1 and in the following sections.
Simulation Model
WUFI 4.2 Pro3 was chosen as the simulation model because it is a well-
validated and benchmarked model for hygrothermal applications. It is impor-
tant to note that due to the inherent limitations of the model, results of the
3
WUFI Wrme und Feuchte instationir is a software family developed by the Fra-
nhofer Institute of Building Physics and adapted for use in North America by the Oak
Ridge National Laboratory under contract with the U.S. Department of Energy.
FIG. 10Moisture conditions at gypsum wall-board, stucco wall assemblies.
FIG. 11Moisture conditions at gypsum wall-board, brick wall assemblies.

FIG. 12Moisture conditions at gypsum wall-board, east exposure.
FIG. 13Moisture conditions at gypsum wall-board, west exposure.

FIG. 14Site rain measurements.
simulations were only expected to be predictive of relative performance and not

specic material moisture content. The model only considers vapor and liquid
diffusion. It does not consider moisture transport by liquid ow or by mass
transport of water vapor air currents. This model considers the surface wet-
ting of materials and can include an initially wet component.
Simulations were run for the portion of the year in which we collected data
and desired a comparison with: August 14October 11. The simulations were
run in two parts.
1 August 14September 10: Wall assembly without impermeable wall
covering. Water injection simulated by initial moisture content of the
OSB sheathing.
2 September 11October 11: Wall assembly with impermeable wall cov-
ering. Water injection simulated by initial moisture content of the OSB
sheathing.
Environment
The model uses historical weather data 30 years continuous data for a par-
ticular region to calculate the hygrothermal response of the wall system. The
simulation engine uses hourly weather data for that region to calculate the
expected dynamic moisture response of the wall system over time based on the
varying weather conditions. Weather les for a 10 % cold year and a 10 % warm
year for Tampa, Florida, which are embedded in the WUFI software, were cho-
sen for the simulations. The criterion is the yearly average temperature, and
TABLE 1Simulation inputs.
Simulation Inputs
Simulation model WUFI 4.2 Pro
Material data
Material Source
Interior surface S-D value set as 100 m
Vinyl wallpaper vapor barrier default
0.0125 m gypsum board WUFI Generic N. A. Database
0.089 m ber glass WUFI Generic N. A. Database
0.0157 m OSB WUFI Generic N. A. Database
SBPO WRB WUFI Generic N. A. Database
WUFI Generic N. A. Database for ber
cement board modied with a higher vapor
permeability to account for the air
Fiber cement siding movement through the siding
Brick WUFI Generic N. A. Database
Stucco WUFI Generic N. A. Database
Shield to rain water and no
Vinyl siding resistance to vapor diffusion
Orientation East, west
WUFI dened typical built-in
moisture with the exterior 0.003 m
raised to 148 kg/ m3 to simulate
Initial conditions water injection
Two periods: from 8/14 to 9/11
Calculation periods and from 9/12 to 10/11
WUFI Database: Tampa, cold year
Exterior climate and Tampa, warm year
WUFI SPC 160 option using
AC with dehumidication with the
following settings: Floating indoor
temperature shift 2.8 oC,
set-point for heating 21 oC,
set-point for cooling 22 oC, RH
Interior climate control set-point 50 %
thus the selected les did not necessarily represent the years with tenth percen-
tile coldest and warmest temperatures over the mid-August to mid-October
portions of the year. A summary of the differences between the two weather
les is shown in Table 2.
Interior conditions were specied by selecting the ASHRAE Standard 160
option within the WUFI software with air-conditioning with dehumidica-
tion. ASHRAE Standard 160 states that if the design or operating specications
for the building specify indoor operating temperatures, or the indoor design tem-
peratures are specied by applicable code, regulation, or law, these temperatures
shall be used. The WUFI software requests both heating and cooling set-points.
TABLE 2WUFI weather les for Tampa, Florida.
Weather File Tampa, Cold Year Tampa, Warm Year

Mean T C 20.9 22.8
Min T C 0 5.6
Max T C 34.4 36.7
Mean RH % 74 72
Min RH % 18 15
Max RH % 100 100
Based on the experimental data, a heating set-point of 21 C and a cooling

set-point of 22 C were selected. ASHRAE Standard 160 has three methods of
specifying interior RH. The method used embedded in the WUFI software is
the intermediate method with several different options for air-conditioning
and humidication control. Because the interior of the test box had controls on
temperature and humidity air-conditioning with dehumidication was cho-
sen with the dehumidication set-point was 50 %. This produced constant in-
terior conditions for the period simulated.
Exterior and interior conditions used in the simulations are shown in Figs.
1518. Although the temperature and RH were generally representative of mea-
sured values, there were some minor variations. Measured temperatures were
slightly higher than those in the simulation. The 10 % warm year was closer to
the measured values than the 10 % cold year. The RH used in the simulations
was constant at the set-point and did not show the diurnal cycling of the RH
visible in the measured values. Rain events for the two climate les are shown
in Fig. 19.
Simulations were conducted on wall assemblies with both east and west
exposure.
FIG. 15Exterior and interior temperaturesTampa, 10 % cold year simulations.

FIG. 16Exterior and interior humiditiesTampa, 10 % cold year simulations.
Wall Construction
The wall construction used in the simulation was representative of the experi-
mental wall assemblies. Material properties were assigned from the material
property database included as part of the modeling software. The wall system
from exterior to interior was as follows.
1 Claddings: Material properties for stucco and brick were obtained from
the generic North American material database provided within WUFI
program. The material properties for ber cement siding were esti-
mated using properties for ber cement board from the WUFI generic
North American database modied with a higher vapor permeability to
account for the air movement through the siding. Because of its high
degree of air leakage and non-absorptive properties, vinyl siding was
FIG. 17Exterior and interior temperaturesTampa, 10 % warm year simulations.

FIG. 18Exterior and interior humiditiesTampa, 10 % warm year simulations.
represented only as a shield to rain water and no resistance to vapor

diffusion was included.
2 Water-resistive barrier: The material data for spun-bonded polyolen
SBPO in the generic North American WUFI database was used for all
wall assemblies except the stucco walls. For the stucco wall assemblies,
paper-backed lath was used in addition to the housewrap, and there-
fore those assemblies were simulated using a layer of building paper
and a layer of housewrap using material properties from the WUFI
database for building paper and SBPO, respectively.
3 Sheathing: Material properties for OSB obtained from the generic
North American material database provided within WUFI program.
There are three OSB materials with densities ranging from 575 to
725 kg/ m3 listed in this database. The OSB material with the middle
density of 650 kg/ m3 was selected. The original source of this data is
listed as ASHRAE Research Project 1018.
4 R-13 berglass insulation: The material property data was taken from
the embedded WUFI generic North American material database.
5 Interior gypsum wall-board: The material property data was taken
from the embedded WUFI generic North American material database.
6 In the second part of the simulation, the vapor barrier wall nish was
simulated by the use of a generic vapor barrier.
Simulation Results
The simulation results are analyzed by examining two responses:
1 The surface RH at the insulation cavity side of the gypsum wall-board
and
2 The moisture content of the gypsum wall-board.
FIG. 19Rain in simulation weather les.
Gypsum Wall-Board Surface Relative Humidity

Figures 2024 show the RH on the insulation cavity face of the gypsum wall-
board. In the rst simulation period, the RH is below 60 % for the entire period.
In the second simulation period, the RH rose quickly and was greater than 80
% for most of this period. This clearly showed the effect of the vapor barrier
reducing drying and trapping moisture in the wall.
ASHRAE Standard 160 states the following:
In order to minimize problems associated with mold growth on the sur-
faces of components of building envelope assemblies, all of the following
conditions shall be met:
a 30-day running average surface RH 80 % when the 30-day running
average surface temperature is between 5 C 41 F and 40 C
104 F, or
b 7-day running average surface RH 98 % when the 7-day running
FIG. 20Simulated humidity at gypsum wall-board/insulation cavity surface, vinyl

siding walls.

104 F, or
c 24-h running average surface RH 100 % when the 24-h running av-
erage surface temperature is between 5 C 41 F and 40 C 104 F
8.
Although the RH does not reach 100 % and predict actual condensation,
when compared against the ASHRAE Standard 160 mold growth minimization
criteria, the conditions for mold growth are exceeded when an interior vapor
barrier is present although not when the interior wall is vapor permeable.
Effect of Climate File and Exposure

The choice of climate le and the exposure produced only small differences in
the simulation results. See Figs. 2023.
Effect of Cladding
The results appear to be independent of cladding. See Fig. 24.
Gypsum Wall-Board Moisture Content

The simulated moisture content of the gypsum wall-board is shown in Figs.
2529. The moisture content is low and relatively constant in the rst simula-
FIG. 21Simulated humidity at gypsum wall-board/insulation cavity surface, ber ce-

ment siding walls.
tion period when no interior vapor barrier is present. The moisture content
rises dramatically in the second simulation period and, in many cases, contin-
ues to rise at the end of the period.
Effect of Climate File and Exposure

The choice of climate le and the exposure produced only small difference in
the simulation results, although the differences were greater than those seen in
the RH response. See Figs. 2528. The largest difference was seen between the
two climate les, although both were directionally the same. Virtually no dif-
ference was seen between the two exposures that were simulated.
Effect of Cladding
There was very little effect of the cladding on the simulated moisture contents.
The stucco walls had slightly higher moisture content than walls with other
claddings. The non-stucco cladding wall moisture contents were all approxi-
mately the same. See Fig. 29.
Comparison of Simulated Results to Experimental Results

Comparisons of the simulated results with the eld test results show agreement
in overall trends. Both clearly show the expected moisture trapping effect of
adding an interior vapor barrier. The simulated results differed from the eld
results in a few ways.
FIG. 22Simulated humidity at gypsum wall-board/insulation cavity surface, stucco

walls.
Simulations did not show the effects of exposure observed in the eld
test. This indicates that local site conditions such as shielding and shad-
ing of adjacent structures may not be adequately assessed by simula-
tions.
Simulations did not predict the differences between claddings that were
observed during the experiment. There are several possible explanations
for this. The main differences in performance due to cladding are ex-
pected to result from differences in either the absorptivity or the venti-
lation of the respective claddings. Differences due to the absorptivity of
the claddings were likely minimized in the simulations as the weather
les had very few rain events during the second simulation period see
Fig. 19. Cladding effects may have been observed if simulations are run
over longer periods of time. Additionally, it is difcult to specify clad-
ding ventilation in the simulations, as air movement is not input di-
rectly. It is possible that a better representation of the air movement
through and ventilation behind claddings would have resulted in more
differences between claddings in the simulations. Finally, differences
may have been due to the selection of generic material properties rather
than actual material properties.
Bulk water leaks, such as the roof leak over the west exposure brick
walls, overwhelm other water entry mechanisms and are not included in
standard wall performance simulations.
FIG. 23Simulated humidity at gypsum wall-board/insulation cavity surface, brick

walls.
FIG. 24Simulated humidity at gypsum wall-board/insulation cavity surface, Tampa

warm year, east exposure.
FIG. 25Simulated gypsum wall-board moisture content, vinyl siding walls.
FIG. 26Simulated gypsum wall-board moisture content, ber cement siding walls.
FIG. 27Simulated gypsum wall-board moisture content, stucco walls.
FIG. 28Simulated gypsum wall-board moisture content, brick walls.

FIG. 29Simulated gypsum wall-board moisture content, Tampa warm year, east
exposure.
Conclusions
The results of both the experimental data and the simulations conrmed that
the use of interior vapor retarders in hot-humid climates reduced wall drying
and was detrimental to the moisture performance of wall systems. RH pre-
dicted by the simulations was in excess on the mold growth minimization cri-
teria according to the ASHRAE Standard 160 criteria for all of the wall assem-
blies once an interior vapor barrier was installed. Although mold was only
observed in the experimental walls that had an additional roof leak, it is pos-
sible that mold would have been present if the test walls had been exposed for
a longer period of time after the vapor barrier installation. Additionally, the
experimental data indicates that cladding system has a distinct effect on mois-
ture conditions at the back surface of the interior wall-board when an interior
vapor retarder is present. This difference was not predicted in the simulations
run in this study. Possible causes included the use of generic material proper-
ties and short durations on weather les. Additionally the characterization of
ventilation of claddings is difcult to specify using the simulation software. A
parametric simulation study would be required to further explore these issues.
References
1 Glass, S. V. and TenWolde, A., Review of In-Service Moisture and Temperature

Conditions in Wood-Frame Buildings, General Technical Report FPL-GTR-174,
USDA, Forest Service, Forest Products Laboratory, Madison, WI, 2007.

2 Harriman, L. G. III and Lstiburek, J. W., The ASHRAE Guide for Buildings in Hot
and Humid Climates, American Society of Heating, Refrigerating and Air-
Conditioning Engineers, Inc., Atlanta, GA, 2009.
3 Kunzel, H. M., Adapted Vapor Control for Durable Building Enclosures, Interna-
tional Conference on Durability of Building Materials and Components, Lyon,
France, 2005, 10 DBMC, Lyon, France.
4 Karagiozis, A. N. et al., Scientic Analysis of Vapor Retarder Recommendations
for Wall Systems Constructed in North America, Proceedings of the Performance of
Exterior Envelopes of Whole Buildings X International Conference, Clearwater
Beach, FL, December 2008, ORNL, Oak Ridge, TN.
5 Derome, D., Karagiozis, A., and Carmeliet, J., The Nature, Signicance and Con-
trol of Solar-Driven Water Vapor Diffusion in Wall SystemsSynthesis of Re-
search Project RP-1235, ASHRAE Trans., Vol. 116, Part 1, 2010.
6 Lstiburek, J., Builders Guide, Hot-Humid Climates, Energy and Environmental As-
sociation, Eden Prairie, MN, 2000.
7 International Code Council ICC, International Residential Code for One- and
Two-Family Dwellings, International Code Council, Inc., Country Club Hills, IL,
2009.
8 ASHRAE Standard 160, 2009, Criteria for Moisture-Control Design Analysis in
Buildings, American Society of Heating Refrigeration and Air Conditioning Engi-
neers, Atlanta, GA.
9 Moisture Analysis and Condensation Control in Building Envelopes ASTM Manual
40, H. Trechsel, Ed., ASTM International, West Conshohocken, PA, 2001.
10 Weston, T. A., Minnich, L. C., Smegal, J., VanMullekom, J., Schumacher, C., and
Conlon, J. S., Evaluation of Cladding and Water Resistive Barrier Performance in
Hot-Humid Climates Using a Real-Weather, Real-Time Test Facility, Proceedings
of the Performance of Exterior Envelopes of Whole Buildings XI International Con-
ference, Clearwater Beach, FL, December 2010, ORNL, Oak Ridge, TN.
11 Ueno, K. and Straube, J., Laboratory Calibration and Field Results of Wood Re-
sistance Humidity Sensors, The First Biennial International Conference on Build-
ing Enclosure Science and Technology, June 1012, 2008, The Building Enclosure
Council BEC-Minnesota, Minneapolis, MN.
12 Carll, C., TenWolde, A., and Pilon, C., Moisture Performance of a Contemporary
Wood-Frame House Operated at Design Indoor Humidity Levels, Proceedings of
the Performance of Exterior Envelopes of Whole Buildings X International Confer-
ence, Clearwater Beach, FL, December 2008, ORNL, Oak Ridge, TN.
13 Salonvaara, M., Karagiozis, A., Pazera, M., and Miller, W., Air Cavities Behind
CladdingsWhat Have We Learned?, Proceedings of the Performance of Exterior
Envelopes of Whole Buildings X International Conference, Clearwater Beach, FL,
December 2008, ORNL, Oak Ridge, TN.
14 Finch, G. and Straube, J., Ventilated Wall Claddings: Review, Field Performance,
and Hygrothermal Modeling, Proceedings of the Performance of Exterior Envelopes
of Whole Buildings X International Conference, Clearwater Beach, FL, December
2008, ORNL, Oak Ridge, TN.
doi:10.1520/JTE102896
Anton TenWolde1
A Review of ASHRAE Standard 160Criteria

for Moisture Control Design Analysis in
Buildings
ABSTRACT: Many of our current recommendations for moisture and conden-

sation control are not based on quantitative analysis under a consistent set
of design assumptions, even though a growing number of computer tools are
beginning to make such analysis practical. To address this issue, the Ameri-
can Society of Heating, Refrigerating, and Air-Conditioning Engineers
ASHRAE published ASHRAE Standard 160, Criteria for Moisture-Control
Design Analysis in Buildings, in January 2009. The standard provides
performance-based procedures and criteria for moisture design analysis for
buildings. It sets criteria for moisture design loads, moisture analysis meth-
ods, and satisfactory building performance. The standard can be used for the
design analysis of the building envelope or to help guide specications for
HVAC equipment and controls. Eventually, it should form the basis for pre-
scriptive moisture design rules based on a uniform set of design assump-
tions and loads. It is intended to help reduce building failures in service,
provide consistency in design approach and recommendations, offer more
exibility in design for moisture control and better ability to incorporate new
materials, and provide greater transparency by requiring reporting of design
assumptions. This paper describes the rationale behind this standard, what
is in it, its potential uses, and areas of uncertainties. A standing ASHRAE
Standard Project Committee has now been formed to update the standard.
The paper discusses some of the main changes that are likely to occur in the
next update.
KEYWORDS: building, moisture, mold, design
Manuscript received November 29, 2009; accepted for publication June 14, 2010; pub-
lished online August 2010.
1
Madison, WI 53705.
Cite as: TenWolde, A., A Review of ASHRAE Standard 160Criteria for Moisture
Control Design Analysis in Buildings, J. Test. Eval., Vol. 39, No. 1. doi:10.1520/
JTE102896.
119
Introduction
In January 2009, the American Society of Heating, Refrigerating, and Air-

Conditioning Engineers ASHRAE published a new standard, ASHRAE Stan-
dard 160, entitled Criteria for Moisture-Control Design Analysis in Buildings
1. The work on the standard began in 1996 and arose from the increased use
of computer-based moisture analysis tools that predict movement and accumu-
lation of moisture in building components and materials. Although analytical
tools were becoming more sophisticated and accurate, relatively little attention
had been paid to the choice of appropriate inputs and boundary conditions.
This may not be an issue when using these tools in forensics to analyze a failure
of an existing building because only the data for conditions during the period
before the failure are needed, although obtaining accurate and sufcient data is
difcult enough. However, the choice of appropriate input values is even more
uncertain when the analysis is used for design purposes before the building has
been built and before actual loads can be measured.
Another reason for creating this standard is that many of our current rec-
ommendations and rules for moisture control are not based on a set of consis-
tent underlying assumptions. This has at times led to largely futile discussions
about the need for various moisture control strategies or design features be-
cause the need often depends on what indoor or outdoor conditions are as-
sumed. For instance, a computer analysis by Tsongas et al. 2 of moisture
accumulation in a wood frame wall in Madison, Wisconsin, showed that the
need for including a vapor retarder in the design completely depended on the
selection of indoor humidity. The level of indoor humidity is equally important
when considering the risk of condensation on windows or wall surfaces or the
need for attic ventilation. Thus, the choice of input values for a design analysis
is critical. Whether a design analysis will show acceptable or unacceptable per-
formance of a particular design or moisture control strategy largely depends on
the design loads operating on the building. It is widely accepted that structural
building design should be based on reasonable assumptions for structural de-
sign loads, and ASHRAE Standard 160 is introducing an analogous approach
for moisture design to the extent feasible.
TenWolde 3 showed how the use of a moisture design standard such as
ASHRAE Standard 160 might have alerted manufactured-home builders to the
potential of widespread decay of plywood sheathing that occurred in the mid-
1980s in a group of manufactured homes in the Midwest. The article also
shows how the use of the standard could have led to the solution and preven-
tion of problems that occurred and might have circumvented a lot of the dis-
agreements and litigation that took place following the discovery of the build-
ing failures.
In summary, the standard is intended to bring moisture design out of the
realm of purely prescriptive measures and turn building moisture design into a
performance-based procedure, with the potential for greater exibility and a
better ability to incorporate new designs and building materials. In addition to
uniformity of design assumptions, the standard also seeks to make the mois-
ture design analysis procedure more transparent by requiring documentation
TENWOLDE, doi:10.1520/JTE102896 121
of the assumptions, material properties used and other choices made for the
analysis.
Summary of ASHRAE Standard 160
ASHRAE Standard 160 applies to new buildings, additions, or retrot and reno-
vation of existing buildings and includes all types of buildings, building com-
ponents, and materials. The purpose of ASHRAE Standard 160 reads as fol-
lows: Given the role that moisture plays in the degradation of building
envelope materials, components, systems and furnishings, the purpose of this
standard is to specify performance-based design methods for predicting, pre-
venting, mitigating, or reducing moisture damage depending on climate, con-
struction type and system operation. These methods include a criteria for
selecting analytic procedures, b design input values, and c criteria for evalu-
ation and use of outputs.
The standard species the minimum attributes for analytical procedures,
depending on the building design and other parameters. For instance, if the
construction is a brick wall, the analytical procedure should be able to handle
water absorption and redistribution in the brick. In contrast, in the unlikely
event that the construction does not contain hygroscopic materials and is air-
tight, a simple vapor diffusion analysis may be sufcient. The standard also
denes design input values, or design moisture loads, primarily by prescrib-
ing default values in case the designer does not have actual design specica-
tions. This includes interior as well as exterior such as rain and humidity
loads. The introduction of the concept of moisture load parallels the use of
mechanical loads in structural analysis. A load here is used in the sense of a
burden or demand on the building. The response of the building to this load
can be analyzed, and the performance can be judged acceptable or unaccept-
able failure. In the standard, the performance of the building design is evalu-
ated using the design loads specied in the standard, and the results are com-
pared with the performance criteria described in the standard. If the
performance is unacceptable, for instance if the analysis indicates high poten-
tial for mold growth, the design should be changed and reevaluated. The stan-
dard may also be used to design or evaluate the heating, ventilation, and air-
conditioning HVAC equipment and controls. The indoor design loads may be
manipulated by varying the design of HVAC equipment and controls, and the
design analysis is used to evaluate the building performance in response to this
manipulation. An example of a HVAC design change that affects moisture de-
sign loads is the addition of a dehumidier to lower indoor humidity.
As with structural design loads, moisture design loads should be more se-
vere than average loads. If average loads were used, actual loads would likely
exceed design loads 50 % of the time. An international consensus has emerged
that design analysis should be based on loads that will not be exceeded 90 % of
the time. This standard has adopted this approach whenever feasible. Follow-
ing is a description of the main features of the standard. Of course, it is not
possible to present all the details in this paper.
Design Initial Moisture Conditions

Some building materials, such as concrete, wet-spray cellulose insulation, and
wood, may contain large amounts of water at the time of building enclosure.
This moisture is often called construction moisture. ASHRAE Standard 160
accounts for this initial moisture load by prescribing high initial moisture con-
tents for those materials, unless specic plans have been included in the con-
struction cycle to dissipate this moisture or to prevent this moisture from ac-
cumulating in the materials through proper storage and protection from rain
and ooding during construction. If such measures are included in the design
and construction plans, the initial conditions to be used are the equilibrium
moisture content EMC of each material at 80 % relative humidity RH
EMC80. This level was chosen because the standard denes this as the highest
possible moisture level that does not lead to mold growth see Performance
Criteria section, thereby allowing for a reasonably high but not destructively
high moisture content. The prescribed design initial moisture content of con-
crete is EMC90 EMC at 90 % RH if specic care is taken to limit initial
moisture conditions. Such measures include requirement for dry storage of
building materials and a plan for drying out the building before enclosure in
case the building gets wet during construction. If no such measures are
planned, the design moisture contents must be doubled i.e., 2 EMC90 for
concrete and 2 EMC80 for all other materials. The factor of 2 was chosen
somewhat arbitrarily, but little quantitative information is available on con-
struction moisture, and actual conditions likely vary substantially.
In case the design analysis concerns building retrots, the lower values
EMC80 and EMC90 are to be used, or measured values, if available.
Internal Loads
The choice of indoor environmental conditions is extremely important, espe-
cially for design analysis of buildings in cold climates. The standard encourages
designers to use their own design parameter values if they are known and part
of the design, or if they are prescribed by code, regulation or law, to use those
values. If they are unknown or not included in the design, the standard pro-
vides a simplied procedure or default values. In residential buildings, indoor
humidity is rarely explicitly controlled, and default design assumptions are
usually needed for these buildings. Internal moisture design conditions include
temperature, humidity, and air pressures. In case design values are available
from HVAC or other design specications, they should be used. Whenever pos-
sible, the standard prescribes higher than average loads because the standard
applies to design analysis.
Design Indoor TemperatureThe default indoor temperatures are shown in

Table 1. The design indoor temperature depends on the outdoor temperature
24-h average and on the HVAC equipment present heating only or heating
and air-conditioning.
Design Indoor HumidityThe standard provides three alternative default

pathways for the determination of indoor design humidity: The Simplied
TABLE 1Default design indoor temperatures.
Indoor Design Temperature,

C F
24-h Running Average of Outdoor Tempera-
ture Heating Only Heating and Air-Conditioning
To,24 h 18.3 C To,24 h 65 F 21.1 C 70 F 21.1 C 70 F
18.3 C To,24 h 21.1 C 65 F To,24 h
70 F To,24 h + 2.8 C To,24 h + 5 F To,24 h + 2.8 C To,24 h + 5 F
To,24 h 21.1 C 70 F To,24 h 70 F To,24 h + 2.8 C To,24 h + 5 F 23.9 C 75 F
Note: To,24 h = 24-h average outdoor temperature.
FIG. 1Design indoor RH, Simplied Method.
Method, the Intermediate Method, and the Full Parametric Calculation.

The Simplied Method for indoor design humidity is based on measured
data from residential buildings without air-conditioning, primarily in Northern
Europe. It provides a simple correlation between outdoor temperature and in-
door humidity see Fig. 1. This approach is used in the European standard,
CEN EN 15026 4, applicable to residential and apartment buildings. Although
the standard allows use of this method for air-conditioned buildings, it should
be understood that the values obtained with the Simplied Method are very
high for buildings with AC. The Simplied Method is also likely to produce
high values for dry climates. The Intermediate Method provides more exibility
and more realistic indoor conditions for air-conditioned buildings. While the
current standard allows use of the Simplied Method for commercial build-
ings, it would be far more preferable to use intended indoor design humidity if
the HVAC controls are included in the design.
The Intermediate Method distinguishes between indoor design humidity
with dehumidication or air-conditioning and indoor design humidity without.
The basis for the methodology is largely described by TenWolde and Walker 5.
If moisture removal is only by ventilation with outdoor air, the Intermediate
Method uses a simple mass balance between moisture sources and ventilation
Eq 1. Residential source rates are provided in the standard see Table 2, but
the designer is encouraged to use appropriate source rates for non-residential
construction. Ventilation rates are those specied in the design or by code, or
default values are given in case they are not. The calculation is done with 24-h
running average outdoor vapor pressure to account for the effect of moisture
storage buffering in the building
TABLE 2Residential design moisture generation rates.
Number of Number of
Bedrooms Occupants Moisture Generation Rate
1 bedroom 2 8 L/day 0.9 104 kg/ s 0.7 lb/h
2 bedrooms 3 12 L/day 1.4 104 kg/ s 1.1 lb/h
Additional
bedrooms +1 / bedroom +1 L / day +0.1 104 kg/ s +0.1 lb/ h
cm
pi = po,24 h + 1
Q
where
pi = indoor vapor pressure, Pa in. Hg,
po,24 h = 24-h running average outdoor vapor pressure, Pa in. Hg,
c = 1.36 105 Pa m3 / kg 10.7 in. Hg ft3 / lb,
m = design moisture generation rate, kg/s lb/h, and
Q = design ventilation rate, m3 / s cfm
The simple mass balance approach does not work when a dehumidier or
air-conditioner are operating because the rate of moisture removal is difcult to
quantify. The standard therefore resorted to a correlation between indoor de-
sign humidity and outdoor design humidity for cooling 5. This approach does
produce reasonable levels of indoor humidity, which are related to outdoor
humidity conditions, and are more realistic than the values prescribed under
the Simplied Method.
The situation is much simpler when air-conditioning or dehumidication
equipment is controlled with a de-humidistat, in which case the anticipated
humidity setting can be used the Standard sets a default of 50 % RH in case
the setting is not known.
The Full Parametric Calculation allows the designer to go beyond the pre-
vious two methodologies and use more sophisticated tools. This may include
building ventilation modeling, using design weather data see the next sections
and equipment models that can estimate moisture removal rates. It also may
involve models that include the effect of adsorption and desorption of water
vapor in various building materials and furnishings. However, the principle of
using design loads i.e., higher than average loads must be adhered to.
External Loads
External design loads include loads from wind, rain, temperature, humidity,
and solar radiation. To ensure that the analysis is done with appropriately se-
vere weather conditions, the standard requires using 10 consecutive years of
weather data or the use of Moisture Design Reference Years MDRY. In the
current standard MDRYs are dened as the 10th-percentile warmest and 10th-
percentile coldest years from a 30-year weather analysis. Unfortunately, such
weather data are currently not available for most locations. Current ASHRAE
research will most likely lead to a change in the denition of MDRY in the
standard.
RainThe standard includes simple formulas for design rain loads on

walls for those users who are not inclined, or capable to perform a full wind-
driven rain analysis. The formulas are based on work by Lacy 6. The standard
assumes that a small amount of this rain water will penetrate behind the clad-
ding even when adequate ashing is included in the design. The reason is that
claddings are usually not completely water tight, especially around windows,
doors, and other penetrations. Very few data are available on the amount of
rain that penetrates through various types of claddings. In the absence of spe-
cic full-scale test methods and data, the default penetration rate is 1 % of the
rain incident on the cladding. The default deposition site for this water is the
water-resistive barrier WRB. If no WRB is present, the designer needs to
specify where the water is deposited.
Air Pressures and Air Flow

How to account for the effect of air pressures and air ows in the standard
presented a special difculty. On the one hand, air ows are usually much more
important to moisture distribution than vapor diffusion and often completely
dominate. On the other hand, analytical procedures to calculate these effects
are not readily available, and appropriate design air pressure boundary condi-
tions are not easily established. For that reason, the standard made the inclu-
sion of air ows in the analysis optional. If the user chooses not to include air
ow analysis, the standard requires that this be explicitly stated in the analysis
report. If analysis is included, the standard contains default values for compo-
nent air tightness in case better data are not available and prescribes minimum
pressures and direction to be used.
Performance Criteria
Performance criteria are needed to evaluate the results from the design analy-
sis. The standard focuses on surface mold growth because under most circum-
stances, they are likely to be the most stringent of all performance criteria. If
mold is of no concern rarely, if the material is not conducive to mold growth
or if temperatures are too cold or warm for mold growth, other criteria such as
surface condensation or structural degradation may become critical.
Conditions for mold growth are complex and depend on mold species, and
the rate of growth varies with temperature and other parameters. However, as
with the rest of the standard, the standard needed to strike a balance between
the complex reality and needed simplicity of use of the standard. The criteria
for mold growth in the standard have been adapted from the International
Energy Agency IEA, Annex 14 7. The IEA criteria were based on observa-
tions of the most common mold species in the indoor environment, data from
the literature, and the collective judgment of the Annex 14 participants. The
criteria in ASHRAE Standard 160 apply to the humidity and temperature con-
ditions at the surface of building materials. To avoid mold growth, the current
ASHRAE Standard 160 requires that the following conditions must be met:
1 30-day running average surface RH 80 % when the 30-day running
104 F,
2 7-day running average surface RH 98 % when the 7-day running av-
erage surface temperature is between 5 C 41 F and 40 C 104 F,
and
3 24-h running average surface RH 100 % when the 24-h running av-
erage surface temperature is between 5 C 41 F and 40 C 104 F.
The standard allows less stringent criteria for materials that are naturally
resistant to mold growth e.g., concrete, masonry, glass, and metals or have
been chemically treated to resist mold growth.
Reporting Requirements
One of the more important aspects of ASHRAE Standard 160 is its requirement
for extensive documentation. The designer needs to provide a description of the
building envelope assembly and building materials. In addition, data need to be
provided on relevant properties of the building materials, such as thermal prop-
erties, moisture properties, density, and airow permeability. Some additional
information about the building, its operation, and the construction process is
also needed. And nally, choices made in the moisture analysis need to be
described.
Changes in the Updated Standard

Many items in ASHRAE Standard 160 are based on incomplete data or when
no data were available based on the professional judgment of the Project Com-
mittee members. The development of the standard has pointed to the need for
specic research, and some of the research projects are beginning to provide
information that can be included in the next version of the standard. The fol-
lowing is a partial list of areas where the standard is in need of improvement
and may be substantially changed or augmented.
Residential Moisture Generation

Indoor humidity is one of the most important parameters related to winter
moisture problems, and the expected indoor humidity is greatly inuenced by
the expected indoor moisture generation rate. The Intermediate Method for
determining design indoor humidity contains a table of residential design
moisture generation rates see Table 2. Those values were arrived at by calcu-
lating the 32 % exceedance level from available historical data i.e., moisture
production is expected to be higher in 32 % of homes 4. More recently, cal-
culations of the moisture balance in homes have been made based on long-term
measurements. These measurements indicate that the production rates in Table
2 are too high 8. Recent experience with applying ASHRAE Standard 160 to
buildings in coastal climates also seem to afrm that the rates in Table 2 may
be too high 8. A preliminary analysis of the limited amount of the more recent
measured data suggests that the 32 % exceedance level for a household of one
to two persons is 67 kg/day 1316 lb/day, not 8 kg/day 18 lb/day as listed in
Table 2 8. Moreover, the results from measured data also suggest that the
addition of 4 kg/day 9 lb/day for the third occupant Table 2 is too high and is
more likely to be on the order of 12 kg/day 24 lb/day. Using these lower
values for moisture production drastically lowers design indoor humidity lev-
els, especially for larger homes multiple bedrooms in cold humid climates.
However, additional measurements are still ongoing, and the results of those,
as well as the results of analysis of other existing data will guide the revision of
the design moisture production rates. However, at this time there seems to be
little doubt that they will be revised downward.
Performance Criteria
The current standard contains an error in the criteria for mold growth. The
recommendations by IEA Annex 14 6 were not transferred correctly: The sec-
ond requirement that the 7-day running average surface RH should be less than
98 % is not correct and should be changed to 89 %. Thus, in the revised version,
the second criterion will read: 7-day running average surface RH 89 % when
the 7-day running average surface temperature is between 5 C 41 F and
40 C 104 F. However, more likely the committee may decide instead to sim-
plify this criterion to one: 30-day running average surface RH 80 % when the
30-day running average surface temperature is between 5 C 41 F and 40 C
104 F and eliminate criteria 2 and 3.
The standard would benet from additional performance criteria for mate-
rials, such as corrosion and structural degradation. However, data on the mois-
ture tolerance of building materials are very hard to come by, and it will be a
challenge to include more criteria in a succinct manner.
Design Weather Data

One problem we encountered during the development of the standard is that
obtaining appropriately severe weather data for use with the standard is very
cumbersome. ASHRAE Weather Years for Energy Calculations do not include
rainfall and are intended to represent average years for the purpose of energy
calculations. As mentioned earlier, ASHRAE Standard 160 requires the use of
ten consecutive years of actual weather data or using a MDRY. Running 10
years of data is cumbersome. Weather les with MDRYs are not available, and
thus obtaining a MDRY from 30 years of weather data is even more time-
consuming. In response to this problem ASHRAE initiated research project RP
1325, Environmental Weather Loads for Hygrothermal Analysis and Design of
Buildings. The project, lead by Mikael Salonvaara, is close to completion and
will produce design weather data for moisture design analysis for 100 locations
in North America. It will also provide a methodology to generate additional
weather data for locations not included in the initial 100. This methodology
will be included and referenced in the next standard.
Effects of Air Flow

It has been obvious for decades that air ow is an important parameter for
moisture ow in buildings and building envelopes. Even small air ows that
may be of no importance to the energy performance of the envelope, still have
the capability of transferring more moisture than diffusion. The rudimentary
handling of the effects of air ow is therefore a serious weakness in the stan-
dard. This weakness stems from a lack of capability in available analytical
moisture analysis tools, and, more importantly, a lack of data on appropriate
air pressures and air ow patterns in building envelopes. Despite the fact that
we have known about the importance of air ows on moisture ows for many
years, only limited computational uid dynamics CFD modeling of air ow
has been done, and air ows and pressures in and through building envelope
components in real buildings in service have been measured only sporadically.
It does not look that this situation will substantially change over the next few
years, and it is therefore likely that the updated standard will be as weak in this
area as the current one is.
Summary
Many of our current recommendations for moisture and condensation control
are not based on quantitative analysis under a consistent set of design assump-
tions, even though a growing number of computer tools are beginning to make
such analysis practical. To address this issue, ASHRAE published ASHRAE
Standard 160, Criteria for Moisture-Control Design Analysis in Buildings, in
January 2009. The standard provides performance-based procedures and crite-
ria for moisture design analysis for buildings. It hopes to help accomplish a
reduction in building failures in service, provide consistency in design ap-
proach and recommendations, offer more exibility in design for moisture con-
trol and better ability to incorporate new materials, and provide greater trans-
parency by requiring reporting of design assumptions. It sets criteria for
moisture design loads, moisture analysis methods, and satisfactory building
performance. The standard can be used for design analysis of the building
envelope or to help guide specications for HVAC equipment and controls.
Eventually, it should form the basis for prescriptive moisture design rules based
on a uniform set of design assumptions and loads. This paper describes the
rationale behind this standard, what is in it, and its potential uses and areas of
uncertainties.
A standing ASHRAE Standard Project Committee has now been formed to
update the standard. Substantial changes are expected in many sections of the
standard, especially in the sections dealing with design indoor humidity calcu-
lations and design weather data. The handling of air ow and performance
criteria is less likely to change substantially, with the exception of the criteria
for mold growth.
References
1 ASHRAE Standard 160-2009, 2009, Criteria for Moisture-Control Design Analysis

in Buildings, American Society of Heating, Refrigerating, and Air-Conditioning
Engineers ASHRAE, Atlanta, GA.
2 Tsongas, G., Burch, D., Roos, C., and Cunningham, M., Parametric Study of Wall
Moisture Contents Using a Revised Variable Indoor Relative Humidity Version of
the MOIST Transient Heat and Moisture Transfer Model, Thermal Performance of
the Exterior Envelopes of Buildings VI, American Society of Heating, Refrigerating,
and Air-Conditioning Engineers, Atlanta, GA, 1995.
3 TenWolde, A., Durability Case Studies RevisitedMold and Decay in Tri-State
Homes, Proceedings of the Second Annual Conference on Durability and Disaster
Mitigation in Wood-Frame Housing, Nov. 68, 2000, Madison, WI, Forest Products
Society, Madison, WI, 2001.
4 CEN EN 15026, 2007, Hygrothermal Performance of Building Components and
ElementsAssessment of Moisture Transfer by Numerical Simulation, Interna-
tional Standards Organization, Geneva, Switzerland.
5 TenWolde, A. and Walker, I., Interior Moisture Design Loads for Residences,
Performance of the Exterior Envelopes of Whole Buildings VIII, American Society of
Heating, Refrigerating, and Air-Conditioning Engineers, Atlanta, GA, 2001.
6 Lacy, R. E., Driving-Rain Maps and the Onslaught of Rain on Buildings, Building
Research Station Current Paper 54, 1965, Building Research Station, Garston, UK.
7 International Energy Agency IEA, 1991, Guidelines and Practice, Report Annex
XIV, International Energy Agency, Leuven, Belgium, Vol. 2.
8 Glass, S. V. and TenWolde, A., Review of Moisture Balance Models for Residential
Indoor Humidity, Proceedings of the 12th Canadian Conference on Building Sci-
ence and Technology, Montreal, Quebec, 2009, Quebec Building Envelope Council,
Montreal, Quebec, Canada, pp. 231245.
doi:10.1520/JTE102973
F. Tariku1 and H. Ge2
Moisture Response of Sheathing Board in

Conventional and Rain-Screen Wall
Systems with Shiplap Cladding
ABSTRACT: Building enclosures are subjected to a random climatic loading

on the exterior surface and a relatively stable indoor condition on the interior.
These loadings result in a transport of heat, air, and moisture across the
building enclosure. In this paper, the drying and wetting of sheathing board in
two exterior walls, more specically 26 in.2 wood-frame conventional no
strapping between sheathing membrane and cladding and a rain-screen
wall system with vertical strapping, are investigated through an experimen-
tal eld study. The experiment is carried out at British Columbia Institute of
Technology eld exposure test facility, where the test walls are exposed to
the coastal climate Vancouver weather on the exterior and controlled indoor
temperature and relative humidity conditions in the interior. The eld experi-
mental results indicate signicant moisture accumulation on the exterior
sheathing boards plywood during the Winter period. During the 9-month
monitoring period from March 13 to Dec. 6, 2009, the plywood underwent a
process of drying and wetting. In both the conventional and rain-screen wall
systems, the plywood dried to a comparable moisture level during the Sum-
mer before the wetting process started. For the wall systems considered in
this study, the plywood in the rain-screen wall has a tendency of faster drying
and wetting in the Spring and Fall seasons, respectively, in comparison to
the plywood in the conventional wall, which is attributed to the presence of
an air gap in the rain-screen wall between the sheathing membrane and the
cladding. A similar trend is observed during the monitoring period from De-
cember 7 to June 15, 2010.
1
Building Science Centre of Excellence, British Columbia Institute of Technology, 3700
Willingdon Av., Burnaby, BC V5G 3H2, Canada Corresponding author, e-mail:
Fitsum_Tariku@bcit.ca
2
Dept. of Architectural Science, Faculty of Engineering, Architecture and Science, Ry-
erson Univ., ON M5B 2K3, Canada.
Cite as: Tariku, F. and Ge, H., Moisture Response of Sheathing Board in Conventional
and Rain-Screen Wall Systems with Shiplap Cladding, J. Test. Eval., Vol. 39, No. 3.
doi:10.1520/JTE102973.
131
KEYWORDS: rain-screen wall, cavity ventilation, eld-experiment, hy-

grothermal performance, wetting and drying potentials
Introduction
Excessive moisture in the building envelope reduces a buildings durability and

compromises the quality of the indoor environment. Rain load is one of the
most important outdoor climatic parameters and rain penetration is the major
source of moisture in building envelopes, which has led to extensive building
failures in the region of Lower Mainland British Columbia and other countries
with similar rain conditions 1. The massive building envelope damage in Brit-
ish Columbia in the 1990s, which is well known as Vancouvers leaky-condo
crisis, was estimated to be $1 billion 2. In response to this massive building
failure crisis, the National Building Code 3 and the Provincial Building Code
4 enforce rain-screen wall design, which creates a capillary break and drain-
age plane between the rst line of defense cladding material and second line
of defense sheathing membrane, in the wet climate regions. The airow
through the capillary break gap might facilitate wetting or drying of the clad-
ding and the sheathing membrane depending on the climatic conditions and
the orientation of the wall system.
Extensive research has been done on evaluating the moisture removal by
cavity ventilation through laboratory testing, eld measurements, and simula-
tions 512. The general conclusions are that ventilation drying is benecial for
wet panel cladding and for solar-driven inward vapor diffusion in Summer. The
drying provided for Winter is minimal. The climate in southern British Colum-
bia is characterized by a long rainy Winter. Whether the provision of a ventila-
tion cavity in the rain-screen wall can assist the drying for this particular cli-
mate is the focus of this study. This study assesses the drying and wetting
potentials of sheathing boards in two wall systems: One with no strapping,
referred in this paper as conventional wall system, and a rain-screen wall
with 19 mm strapping that provided air gap between the shiplap cladding and
the sheathing membrane. The study is carried out at British Columbia Institute
of Technologys BCIT eld exposure test facility Building Envelope Test Fa-
cility BETF, where the test wall panels are installed and their hygrothermal
responses to Vancouver weather conditions and controlled indoor climatic con-
ditions are collected along with the respective hygrothermal loadings
When the wall systems were installed on the eld exposure test facility in
July 2008, the electrical resistance measurement method, with a 100 M re-
sistor in parallel with moisture pins, was adapted for moisture content MC
measurement of the plywood. Inspection of the measured data suggested that
the MC reading with the adapted method was high and unreliable. On March
13, 2009, the MC measurement wiring was switched to a series circuit with a
5 M resistor, a method that is commonly used by other researchers 1315.
In this paper, the measurements obtained after March 13, 2009, are used for
discussion.
TARIKU AND GE, doi:10.1520/JTE102973 133
FIG. 1BCITs BETF northwest view.
Field Experiment
Test Facility
The experimental study is being carried out at BCIT BETF photo shown in Fig.
1. The research facility is designed to evaluate the hygrothermal performance
of full-scale building envelope assemblies under simulated indoor and real cli-
matic outdoor conditions. The 44 28 ft2 two-story structure can accommo-
date in total 62 4 8 ft2 1.2 2.4 m2 panels. The panels are removable, al-
lowing for ease of implementation of any type or location of testing required.
Two mechanical systems are tted within the facility allowing the separation of
interior spaces into two conditioned horizontal zones, thus allowing control of
the indoor boundary conditions, namely, temperature and relative humidity, at
the desired values. Each system can maintain indoor temperature within the
range of 18 26 C, with a precision of 2 C, and relative humidity within the
range of 4080 %, with a precision of 5 %. The facility is equipped with a data
acquisition system with over 600 channels, allowing for the monitoring of hy-
grothermal conditions within wall assemblies including temperature, relative
humidity, MC, heat ux, air velocity, wind-induced pressure, and incidence of
condensation and rain penetration. The facility is also equipped with a weather
station mounted on the rooftop of the facility to measure the outdoor boundary
conditions including wind speed, wind direction, solar radiation on both hori-
zontal and vertical surfaces, and horizontal rainfall. Driving rain on wall sur-
faces is also collected. More information about this facility can be found in Ref
16.
Test Panels and Instrumentation

Two 4 8 ft2 test panels were fabricated as conventional and rain-screen wall
systems and installed in the northwest section of the BETF. The conguration
FIG. 2Vertical cross-section of the instrumented conventional wall system.
of the 2 6 in.2 38 140 mm2 wood-frame test panels from exterior to inte-
rior, in sequence, is as follows: Horizontal shiplap ber cement siding, two
layers of 30 min rated asphalt-impregnated building papers as a weather bar-
rier, 12.5 mm plywood as a sheathing board, 138 mm glass ber insulation,
6-mil polyethylene sheet as a vapor and air barrier, and interior nish gypsum
board, 12.5 mm. The rain-screen wall has a 19 mm air gap between the sheath-
ing membrane and cladding. The schematic diagrams of the vertical cross-
sections of the conventional and rain-screen walls, along with the correspond-
ing sensors that are installed to measure the MC and temperature of the
plywood, are shown in Figs. 2 and 3, respectively.
The cores of the test walls including framing, insulation, polyethylene
sheet, and plywood sheathing were fabricated and instrumented in a controlled
environment in mid-June 2008, and therefore, a good workmanship has been
achieved. The test panels were stored inside the shop for about 1 month before
being installed on the test facility, therefore, the initial MC of plywood sheath-
ing boards and wood-frame members can be deemed uniform. The building
FIG. 3Vertical cross-section of the instrumented rain-screen wall system.
papers, the cladding ber cement siding, and the interior layer gypsum
board were installed after the walls were in place on the test facility. To provide
the thermal and moisture separation from the surrounding existing walls, the
polyethylene sheet was wrapped around the edge of the stud to overlap with the
building papers. The 2 in. gap between each test wall was tted with rigid
insulation and sealed to the side with sealant and backing rod.
To measure the MCs of the plywood during the monitoring periods at dif-
ferent heights, three pairs of moisture pins were installed on each test panel
from the inside along its center line. The three moisture measurement points
were at the lower, middle, and upper position, more specically, at the one-
quarter, half, and three-quarter wall height. Figures 2 and 3 show the locations
of the moisture pins along with the thermocouple, which were installed at the
middle height of the wall. In addition to providing information about the ther-
mal responses of the two wall systems, the thermocouple readings are used for
conversion of the three electrical resistance measurements of the correspond-
ing wall systems. The MC measurement system was developed and calibrated
in the building science laboratory of BCIT with an accuracy of 2 % in the range
of 625 % MC. This system was used in a previous study 14, in which both
gravimetric and moisture pin measurements were taken and the discrepancy is
within 2 % 12.
30-gauge premier grade type T copper and constantan thermocouple
wires were used to measure temperature. They were calibrated by using an
isothermal bath accuracy of 0.1 C with Agilent Switch Unit model 34970A
at three different temperatures: 10, 20, and 30 C. The system measurement
accuracy is 0.5 C. The MCs of plywood and stud were measured by using
electric moisture pins. The moisture pins are stainless steel screws and, using
gravimetric samples, the measuring system was calibrated to a range of 625 %
for plywood and 730 % for wood stud with an accuracy of 2 %. The MC and
temperature measurements are scanned every 5 min and recorded by the data
acquisition system.
Climatic Conditions
The test panels are exposed to Vancouver weather conditions on the exterior
and controlled indoor temperature and relative humidity conditions on their
interior surfaces. The local outdoor climatic conditions including temperature,
relative humidity, wind speed and direction, global solar radiation, and hori-
zontal rainfall are measured with a weather station that is mounted on the
rooftop of the BETF. The wind-driven rain that impinges the test panels is also
measured with a rain gauge that is vertically mounted adjacent to the test
panels. The measured climatic conditions are presented and discussed in the
Results and Discussion section.
The indoor temperature and relative humidity conditions are controlled by
thermostat and humidistat, respectively. The temperature set point is 21 C and
has been kept constant throughout the monitoring period. The test facility is
equipped with humidication systems and it is possible to control the indoor
relative humidity during the Winter period. But, as can be seen in Fig. 4, the
indoor relative humidity during the Summer period is considerably higher than
the set point of 40 %. This is due to the fact that the ventilation fan was con-
tinuously running and there is no dehumidication unit to remove the excess
moisture. Moreover, the moisture removal by the air conditioning unit might
have been limited due to the mild outdoor temperature and part-load operation
of the equipment, which generally happens in mild, wet climates like Vancou-
ver. The maximum relative humidity inside the BETF during the Summer pe-
riod was 71 %, which was recorded on July 28, 2009.
Results and Discussion

In the wall systems considered in this study, the sheathing board is the critical
layer that is susceptible to moisture damage due to high moisture accumula-
FIG. 4Indoor temperature and relative humidity of the BETF.
tion. Thus, the MC and temperature of the plywood in the two different wall
systems are discussed.
Figure 5 shows the MC of the sheathing board in the conventional wall
system. The MC of the plywood at the upper position is consistently higher
than at the middle and lower positions. In fact, until the end of April, the MC at
the upper position is higher than the 19 % MC level that is recommended to
avoid moisture related durability problems. In the rst two months, the ply-
wood was in a slow drying process. During this period, the indoor temperature
and relative humidity were relatively stable at 21 C and 36 %, respectively Fig.
4, and the exterior surfaces of the walls were exposed to ten wind-driven rain
events all under 0.1 mm/h, Fig. 6, higher outdoor relative humidity Fig. 7
and low solar radiation Fig. 8, which might have contributed to slow drying of
the sheathing board. The drying process accelerated in the month of May as the
ambient temperature Fig. 7 and solar radiation increased. The MC changes in
the plywood during the Summer months of June and July were minimal. This
is expected since the material was relatively dry and further drying was a very
slow process. During this period, the ambient temperature and solar radiation
FIG. 5MC of the sheathing board plywood in the conventional wall system at the
lower, middle, and upper positions.
FIG. 6The hourly wind-drive rain load on the test walls collected by a wind-driven
rain gauge next to the test panels.
as well as the indoor relative humidity were rather high, and there were no
wind-driven rain events see Fig. 4 and Figs. 68. The effect of indoor vapor
pressure on the drying process is expected to be low since there is a 6-mil
polyethylene sheet behind the interior layer gypsum board and good work-
manship was achieved in constructing the test wall. The MC of the plywood
starts to increase at the beginning of August and continues through the Winter
period. The plywoods MC increase during the late Summer and Fall seasons is
due to the reduction in the ambient temperature and solar radiation, which
reduce the plywood temperature, coupled with the presence of relatively humid
air.
As can been seen in Fig. 5, the MC difference between the upper and lower
positions on March 13 is about 4 %. The difference slowly decreases, reaching
the lowest value of 2 % at the end of July, and then increases during the Fall
season to 3 %. The MC of the plywood at the middle section is slightly higher
FIG. 7The daily average outdoor air temperature and relative humidity.
FIG. 8The hourly average and daily total horizontal global solar radiation.
than at the lower section. The difference in MC between upper and middle
positions is slightly higher than that between middle and lower positions. A
number of factors may have contributed to the vertical prole of MC in ply-
wood such as convection loop within insulation cavity, higher moisture expo-
sure on cladding at the upper level due to wind-driven rain, and under-cooling
induced surface condensation; however, further investigation is required before
the actual cause can be identied. Tables 1 and 2 show the maximum and
minimum MC readings, as well as the maximum MC changes, observed during
the drying March 13 to Aug. 1, 2009 and wetting August 1 to Dec. 6, 2009
periods, respectively. Although the lowest MC reading during the drying period
Table 1 is at the lower position 5.8 %, the drying rate is higher at the upper
section of the plywood 14.4 %. During the wetting period, the upper position
also has a higher wetting rate 8.1 % compared to the lower and middle posi-
tions 6.1 % and 6.6 %, respectively.
Figure 9 shows the drying rates of the upper and middle positions of the
plywood. The drying rates are relatively high during the month of May and
June. The maximum drying rate is 0.19 % per day corresponding to May 26,
2009, and decreases as the plywood gets drier below 70 % relative humidity.
In general, the upper section of the plywood shows a relatively higher drying
rate in July and a higher wetting rate in October as compared to the middle
section.
Figure 10 shows the hourly and daily average temperature measurements
of the plywoods interior surface at the middle section. The daily average am-
bient air temperature is also superimposed on the gure. During the March 13
TABLE 1Extreme MCs of plywood sheathing in conventional test wall during the drying
period March 13 to August 1.
Moisture Content Lower Position Middle Position Upper Position

Maximum % 16.8 18.4 21.5
Minimum % 5.8 6.0 7.1
29
Difference % 11.0 12.4 14.4
TABLE 2Extreme MCs of plywood sheathing in conventional test wall during the wetting
period August 1 to December 6.
Moisture Content Lower Position Middle Position Upper Position

Maximum % 11.9 12.6 15.2
Minimum % 5.8 6.0 7.1
29
Difference % 6.1 6.6 8.1
to Dec. 6, 2009, monitoring period, the surface temperature of the plywood

varies from the lowest value of 5.8 C on December 2 to 62.8 C on July 21.
The daily average temperature difference between the ambient air and the ply-
wood is higher during the Summer months when the solar radiation is high.
Figure 11 shows the MC of plywood in the rain-screen wall system at the
lower, middle, and upper positions. Prior to April 8, the MC measurements of
the plywood at the upper location were over 25 %. Since the measurement
method adopted in the study does not yield reliable readings for MC over 25 %,
the respective data is excluded from the analysis. In general, the plywoods MC
in Fig. 11 can be categorized into three sections: Drying period March 13 to
June 4, stable period June 4 to August 4, and wetting period August 4 to
December 6. Faster drying is observed at the upper section of the plywood
followed by the middle and lower sections, respectively. During the stable pe-
riod of two Summer months, the MC of the plywood decreases by only 2 %.
During the wetting period, the MC of the plywood at the upper section in-
creases signicantly more than the middle or lower sections. This might be
attributed to an air leakage at the upper section of the test panel where the
measurement wires are routed out of the test panel to be connected to the data
acquisition system.
Relative Comparison of the Drying and Wetting Potentials of Plywood in the

Conventional and Rain-Screen Wall Systems
The MC and temperature readings at the middle positions of the conventional
and rain-screen walls are used to assess the relative drying and wetting poten-
FIG. 9Drying rate curves of the upper and middle positions.

FIG. 10Plywoods interior surface and outdoor air temperatures in the conventional
wall system.
tials of the plywood in the two wall systems. Figures 12 and 13 show the daily
average MC and the drying rate curves of the plywood in the conventional and
rain-screen wall systems, respectively. Although the plywood in the rain-screen
wall starts with slightly higher MC, by June 4 it reaches lower MC compared to
the conventional wall due to its higher drying rate as shown in Fig. 13. The
higher drying rate might be the result of extra moisture removal from the wall
by the airow through the gap between the sheathing membrane and the clad-
ding. During this period March 13 to June 4, the daily average MC of the
ambient air, shown in Fig. 14, is below 8.0 g/kg, and can potentially remove
moisture from the moist sheathing membrane and thereby result in lower MC
in the plywood.
Figure 15 shows the monthly total wind-driven rain on the prevailing wind-
driven rain direction southeast and test wall direction northwest for March
to November 2009. Since the test walls discussed in this paper were installed in
the northwest orientation, which is opposite to the prevailing wind-driven di-
rection, the amount of the wind-driven rain that impinges the exterior surfaces
FIG. 11Transient MC of the sheathing board plywood in the rain-screen wall system
at the lower, middle, and upper positions.
FIG. 12MC of the plywood in the conventional and rain-screen wall systems data
presented is the daily average of the middle location.
of the test walls is signicantly lower in most cases under 0.1 mm/h and, in
some months, none at all. For example, between April 26 and September 8 see
Fig. 6, the test walls were not exposed to wind-driven rain load, although there
were rain events during the same period of time as shown in Fig. 16. Between
March 13 and December 6, 2009, the maximum wind-driven rain load is 0.95
mm/h, which is about one-seventh of the horizontal rain that is recorded in the
same rain event. The insignicant wind-driven rain exposure of the test walls
suggests that the prominent effect of the air gap in the experimental study
reported here might be more on providing cavity ventilation than providing
capillary break or drainage. The effect of the air gap in walls in other orienta-
tions or climatic conditions can be different, and possibly result in higher varia-
tions of moisture accumulation between the conventional and rain-screen wall
systems than observed in this study.
During the following two months June 4 to August 4, the drying rate of
the plywood in the rain-screen wall is lower than in the conventional wall see
Fig. 13. This effect is also probably related to the airow through the cavity in
FIG. 13Drying rate curves of the plywood in the conventional and rain-screen wall
systems middle location.
FIG. 14The daily averaged absolute humidity ratio of the outdoor air.
such a way that the air that ow through the cavity during the Summer period
has high vapor pressure as reected in Fig. 14 as high humidity ratio with a
maximum value of 13 g/kg on July 26 and can also have had the effect of
cooling the plywood Fig. 18 and thereby reducing the drying capacity of the
plywood by vapor diffusion. During the wetting period after August 4, the MC
of the plywood in the rain-screen wall increases more than the conventional
wall. Also, the heat gain due to solar radiation and ambient temperature, as
well as the plywood temperature, are continuously decreasing, which may fa-
cilitate moisture accumulation as the moist air ows through the cavity. The
drying rate curve see Fig. 13 shows further increase in the wetting rate of the
plywood in the rain-screen wall in the month of November
To assess the long-term moisture responses of the sheathing boards in the
two wall systems, the experimental study was extended for another 6 months
December 2009 to June 2010. The MC of the middle section of the sheathing
boards in the conventional and rain-screen wall systems during the 15 months
monitoring period are presented in Fig. 17. Similar to the discussion already
FIG. 15Wind-driven rain on the prevailing wind-driven rain direction southeast and
test wall direction northwest.
FIG. 16The hourly rainfall horizontal rain measured at the roof top of the BETF.
presented in the Relative Comparison of the Drying and Wetting Potentials of

Plywood in the Conventional and Rain-Screen Wall Systems section, the ply-
wood in the rain-screen wall system reaches to its peak MC sooner than the
plywood in the conventional wall, and started drying two weeks in advance.
The two sheathing boards peak MCs differ by only about 1 %, and in general,
the MC differences in the two sheathing boards during the 15 months monitor-
ing period are insignicant. This can be attributed to the test panels orienta-
tion that lead to minimal wind-driven rain load exposure and the type of clad-
ding used in the study. In the conventional wall system, the shiplap cladding
provides compartmentalized airspace which would not exist in other type of
cladding such as stucco cladding that can have an effect in the drying and
wetting processes of the sheathing board.
The temperature of the plywood in the rain-screen wall is similar to that of
the plywood in the conventional wall see Fig. 10. To investigate further, the
temperature difference between the interior surfaces of the plywood in the two
wall systems is plotted in Fig. 18. In the same gure, the daily total solar radia-
FIG. 17MC of plywood sheathing boards in the conventional and rain-screen wall
systems from March 13, 2009, to June 15, 2010.
FIG. 18Temperature difference between the plywood in the rain-screen and conven-
tional wall systems and the daily total solar radiation.
tion is also plotted. As can be observed from the gure, the temperature differ-
ence in the two walls seems to depend on the magnitude of solar radiation.
When the solar radiation is low, so is the temperature difference. Moreover, at
high solar radiation the plywood in the rain-screen wall has a relatively lower
temperature than the plywood in the conventional wall system. Since there was
no airow measuring probe installed in the current experiment, it was not
possible to verify whether higher solar radiation increases cavity ventilation
and, consequently, results in cooling of the plywood. In theory, cavity ventila-
tion may result in cooling or heating of the sheathing layer when the exterior
surface of the cladding receives high solar radiation or loses signicant heat by
long-wave radiation heat exchange with the sky and surroundings, respectively.
This is because the ambient temperature will be lower compared to the sheath-
ing layers during high solar gain, and the converse is true for the case of sig-
nicant heat loss by long-wave radiation. At low or an absence of solar radia-
tion, the latter may dominate and may result in a relatively higher temperature
reading in the rain-screen wall sheathing layer compared to the conventional
wall. In the result presented in Fig. 18, the hourly average temperature of the
plywood in the rain-screen wall is higher than in the conventional wall for 75 %
of the monitoring period. The temperature deviations of the two wall systems
during this part of the monitoring period are relatively small less than 3.5 C.
During the rest of the monitoring period, the temperature differences are rela-
tively high and occur between 4 p.m. and 7 p.m. in the afternoon, which may be
associated to the high solar radiation that reaches the test walls as they are
oriented in the northwest direction. At this time, the ambient air will be at a
lower temperature compared to the sheathing layer temperature and may pro-
vide cooling to the sheathing layer as it passes through the rain-screen wall air
gap.
Conclusion
An experimental study of the drying and wetting processes of the sheathing
boards in the conventional and rain-screen wall systems was carried out at
BCITs eld exposure test facility. Similar drying and wetting processes are
observed in both wall systems. For the wall types, orientations, climatic condi-
tions, and monitoring period considered in the study, the MCs of plywood at
the middle and lower positions have relatively low MC readings compared to
the upper position, but the upper position has the highest drying rates and
reaches the same MC level as the other two positions at the beginning of the
Summer season. An analysis of the drying and wetting of the sheathing boards
suggests that the plywood in the rain-screen wall has a tendency of faster dry-
ing and wetting in the Spring and Fall seasons, respectively, in comparison to
the plywood in the conventional wall. The airow through the air gap in the
rain-screen wall system might have facilitated the drying and wetting pro-
cesses. But, in the Summer, the plywood in both wall systems dried to about the
same level of MC.
In the experimental study reported here, the prominent effect of the air gap
in the rain-screen wall system might be more on providing cavity ventilation
than providing capillary break or drainage since the wall systems were exposed
to wind-driven rain loads of low magnitude. The effect of the air gap in walls in
other orientations, cladding type or climatic conditions can be different, and
possibly result in higher variations of moisture accumulation between the con-
ventional and rain-screen wall systems than observed in this study.
Acknowledgments
The writers would like to acknowledge the nancial support received from the
School of Construction and the Environment, BCIT, and Pacic Building Sys-
tems and the assistance of Stephen Roy and Wendy Ye.
References
1 Canadian Mortgage and Housing Corporation, Survey of Building Envelope Fail-

ures in the Coastal Climate of British Columbia, CMHC, Vancouver, BC, 1996, 27
pp.
2 Barrett, D., The Renewal of Trust in Residential Construction, Report Submitted
to the Lieutenant-Governor in Council Government of British Columbia, Crown
Publication, Victoria, BC, June 1998, http://www.qp.gov.bc.ca/condo/ Last ac-
cessed September 2010.
3 National Research Council of Canada, 2005, National Building Code of Canada,
Ottawa, ON, Canada, http://www.nationalcodes.ca/eng/nbc/index_e.shtml Last ac-
cessed September 2010.
4 Province of British Columbia, 2006, British Columbia Building Code, Queens
Printer, Victoria, BC, http://www.bccodes.ca/default.htm Last accessed September
2010.
5 Onysko, D. M., Drying of Walls with Ventilated Stucco Cladding: A Parametric Analy-
sis. Housing Technology Series, Canada Mortgage and Housing Corporation, Ot-
tawa, 2000.
6 Hansen, M., Nicolajsen, A., and Stang, B., On the Inuence of Ventilation on
Moisture Content in Timber Framed Walls, Proceedings of the Sixth Symposium

on Building Physics in the Nordic Countries, Norweigian University of Science and
Technology, 2002, Trondheim, Norway.
7 Hazleden, D. and Morris, P., The Inuence of Design on Drying of Wood-Frame
Walls Under Controlled Conditions, Proceedings of Thermal Performance of Build-
ing Envelopes VIII, 2002, American Society of Heating, Refrigeration, and Air-
Conditioning Engineers, Clearwater, FL.
8 Bassett, M. R. and McNeil, S., Drained and Vented Cavity WallsMeasured Venti-
lation Rates, IRHACE Conference, Nelson, New Zealand, 2005, Institution of Re-
frigeration, Heating and Air Conditioning Engineers, Nelson, New Zealand.
9 Bassett, M. R. and McNeil, S., The Theory of Ventilation Drying Applied to New
Zealand Cavity Walls, IRHACE Conference, Nelson, New Zealand, 2005, Institu-
tion of Refrigeration, Heating and Air Conditioning Engineers, Nelson, New
Zealand.
10 Shi, X. and Burnett, E., Ventilation Drying in Walls with Vapor Impermeable
Claddings, Proceedings of the Third International Conference in Building Physics
and Building Engineering, 2006, Concordia University, Montreal, QC, Canada, p.
395402.
11 Ye, Y., Ge, H., and Fazio, P., Inuence of Cavity Ventilation on the Drying and
Wetting of Plywood Sheathing in the Rainscreen Walls in the Coastal Climate of
British Columbia, Proceedings of the Fourth International Conference in Building
Physics, 2009, The Physical Environment Control Unit of Istanbul Technical Uni-
versity, Faculty of Architecture, Istanbul, Turkey.
12 Ge, H., Ye, Y., and Fazio, P., Field Investigation of the Impact of Cavity Ventilation
on the Wetting and Drying of Brick Veneer Wall Systems in the Coastal Climate of
British Columbia, Proceedings of the 12th National Building Science and Technol-
ogy Conference, May 2009, National Building Envelope Council, Montreal, QC,
Canada.
13 Hazleden, D., Envelope Drying Rates Experiment, Final Report Contract No. 99
2221, Canada Mortgage and Housing Corp., Ottawa, 2001.
14 Simpson, Y., 2010, Field Evaluation of Ventilation Wetting and Drying of Rain-
screen Walls in the Coastal Climate of British Columbia, M.S. Sc thesis, Concor-
dia University, Montreal, QC, Canada
15 Maref, W., Lacasse, M., and Booth, D., Experimental Assessment of Hygrothermal
Properties of Wood-Frame Wall AssembliesMoisture Content Calibration Curve
for OSB Using Moisture Pins, J. ASTM Int., Vol. 71, 2010, pp. 112.
16 Ge, H., Krpan, R., Fazio, P., and Ye, Y., A Two-Storey Field Station for Investiga-
tion of Building Envelope Performance in the Coastal Climate of British Colum-
bia, Proceedings of the Symposium in Building Physics, Leuven, Belgium, October
2008, Leuven, Belgium.
doi:10.1520/JTE102970
George Pericles Stamatiades III1
Investigation of the Condensation Potential

Between Wood Windows and Sill Pans
in a Warm, Humid Climate
ABSTRACT: As building envelope designers have become more attuned to
the need to prevent moisture intrusion, window ashing has become more of
an issue. Sill pans are an important component of many window ashing
systems. These pans are constructed from a variety of materials, including
metal, and are located beneath the window sill. Sill pans are also typically
formed to have back dams and end dams so that they can redirect water that
leaks through the window to the exterior. Draining the water to the outside
requires a vented area between the sill pan and the window sill, and there is
speculation that environmental humidity can condense within this space in
warm, humid climates. This condition may be most severe when the sill pans
are made from materials with high thermal conductivities, like metal. If water
were to condense in this cavity, condensate may periodically contact the
exposed underside of a window sill, which, if wooden, may cause decay. In
order to determine if condensation-related moisture could negatively impact
the performance of a window, a wood window was installed with a metal sill
pan at the Oak Ridge National Laboratorys Natural Exposure Testing Facility
near Charleston, SC, to measure parameters including temperature, humid-
ity, and wood moisture content. Based on the testing results, recommenda-
tions are made to maximize moisture performance of wood windows installed
with sill pans in warm, humid climates.
KEYWORDS: condensation, sill pan, window, decay
Background
Moisture control is paramount when designing and constructing a building.
According to the Builders Guide to Hot-Humid Climates 1, controlling mois-
ture effects on exterior building walls is the most important factor in making a
structure durable. Creating an effective drainage plane requires the seamless
1
Academic Magnet High School, Charleston, SC 29407.
Cite as: Stamatiades, G. P., Investigation of the Condensation Potential Between Wood
Windows and Sill Pans in a Warm, Humid Climate, J. Test. Eval., Vol. 39, No. 2.
doi:10.1520/JTE102970.
148
STAMATIADES, doi:10.1520/JTE102970 149
FIG. 1Rough opening decay due to leakage at a window jamb/sill interface.
integration of ashing at all rough openings in walls 2. Rough openings are

very susceptible to accumulating moisture, and this could lead to decay. Thus,
it is critical to understand the sources of moisture and to determine ways to
handle this moisture so that no damage is done.
The Problem: Water Leakage in Windows

According to a study conducted by the Canada Mortgage and Housing Corpo-
ration, over 90 % of tested windows leaked 3. This statistic indicates that
water leakage may be occurring in many window assemblies. If installed prop-
erly, but without sill ashing, the window may not leak initially. However, some
note that there seem to be two types of window installations in North America:
those that now leak and those that will leak later 2. Over time, the integrity
of non-leaking window installations is likely to be lost, and water will likely nd
its way in. Thus, it is of utmost importance to address leakage associated with
windows.
Figure 1 reveals decay resulting from leakage at a windows jamb and sill
interface. The Canada Mortgage and Housing Corporation study 3 classied
window leakage types into six different categories Fig. 2. The decay shown in
Fig. 1 is a result of leakage along paths L4 and L5 shown in Fig. 2. If all of these
types of leakage are addressed, then the damage resulting from such leakage
would be eliminated.
The Solution: Sill Pans

If windows are prone to water leakage, then controlling that leakage is very
important. The decay in Fig. 1 may have been avoided if the water leakage
along paths L4 and L5 was collected and redirected to the exterior. This could
be accomplished with a sill pan. For additional information on sill pans, refer
to ASTM E2112-01 4.
FIG. 2Possible window leakage paths.
The Speculation: Sill Pan Condensation

Sill pans may keep water from leaking into the interior of a wall, but they may
be a factor in causing moisture accumulation as well. One wood window manu-
facturer claims that metal sill pans may induce condensation in the cavity
above the pan. Condensate resulting from the sill pan may be absorbed by
exposed wood components of the window and, over time, cause decay. The
reasoning behind this idea is logical: many sill pans are made of metal, and
metal is a highly conductive material. Thus, in a hot and humid climate where
homes are cooled with air-conditioned air, indoor air may cool a metal sill pan
enough to create conditions where water vapor in the hot and humid air above
it will condense. Condensate on the exposed underside of a window sill can
eventually cause decay, especially if this portion of the window is not treated.
Hypothesis
This thesis explores condensation on metal sill pans in hot/humid climates and
the potential for decay of wood windows. The following sections will discuss
the science behind the thesis, the experimental set-up, the ndings, and the
subsequent implications Figs. 321.
Scientic Basis
Since this thesis relies upon observations of condensation and its inuence
upon materials, it is critical to have an understanding of the science behind it.
FIG. 3Psychrometric chart. Notice how the dot at 75F has a RH of 50 %. As the
temperature decreases represented by the arrow, the RH increases.
Condensation is directly related to the properties and behavior of air. The eld
of science associated with this topic is known as psychrometrics.
Psychrometrics
The American Society of Heating, Refrigerating, and Air-Conditioning Engi-
neers ASHRAE 5 denes psychrometrics as the science that deals with the
thermodynamic properties of moist air and uses these properties to analyze
conditions and processes involving moist air. ASHRAE 5 notes that there are
three types of air compositions: atmospheric air, dry air, and moist air. Atmo-
spheric air is the most common type: it includes the typical gasses that com-
pose air along with water vapor and all of the contaminants found in air. Dry
air contains only the gasses typically found in air, while moist air is dry air with
water vapor added. Thus, air is typically mixed with water vapor. The amount
of water vapor that exists in air, known as humidity, depends on the tempera-
ture and pressure of the air. When moist air is saturated with humidity at a
relative humidity RH of 100 %, condensation results. Excessive condensation
occurring on moisture sensitive materials may cause decay.
FIG. 4NET facility.
Mold Growth and Condensation

According to Hens 6, mold growth requires ve different factors. First, fungal
spores must settle on a surface. Then, the surroundings must provide adequate
oxygen. Next, the mold must have the optimal temperatures, which are typi-
cally warm. Then, the mold must have nutrients. Finally, the mold must have
moisture 6. Mold growth, as Evans 7 agreed, is caused by fungi living a
moist environment. In humid environments, mold growth is more likely. In
fact, Hens 6 claimed that mold can survive in areas where the average RH is
about 80 %. Consequently, this thesis must determine the average RH on sur-
faces within the sub-sill cavity.
An important trend to note with RH is that when atmospheric air tempera-
ture decreases, the RH increases. This can be proven with a psychrometric
chart Fig. 3: a critical tool for architects and engineers.
Weather in Charleston
According to the exterior insulation and nishing system EIFS performance
study 8, Charleston has very warm summers, with daytime dry bulb tempera-
tures typically ranging in the 90s. Coupled with this, the RH in the atmosphere
is typically above 50 %. As these weather conditions indicate, Charleston has a
hot and humid climate, specically a mixed, coastal, Zone 3 climate 8.
These climate zones were dened by the U.S. Department of Energy to aid
engineers with energy efcient designs of buildings.
Methodology
The Natural Exposure Testing Facility

The Natural Exposure Testing NET Facility is a research building located on
the Baptist Hill High School campus in Hollywood, SC. The building was de-
FIG. 5The rough opening is adjusted by attaching additional wood.
signed for testing building materials in walls, such as insulation and vapor
retarders, as it is equipped for collecting numerical data, such as temperature
and moisture content. The EIFS performance study 8 expands on this by
writing that up to now, what has been lacking is a full understanding of the
hygrothermal temperature and moisture performance of all types of wall sys-
tems for typical climactic effect, such as wind driven rain, rainwater penetra-
tion, condensation, solar and night sky radiation, wind speed, and site/wall
orientation. This is attributed to the lack of data and research on this particu-
lar area, so the NET Facility was created. According to the EIFS performance
study 8, the building has computers that log data hourly. The recorded data is
then sent to Oak Ridge National Laboratory ORNL and processed 8.
ORNL was gracious enough to allow this experiment to be conducted at the
NET Facility, which is an ideal building since it is located in a hot and humid
environment. The facility has 30 interchangeable wall panels that can be exten-
sively instrumented to measure thermal and moisture performance character-
FIG. 6Placing the WRB.
istics. One panel on the northwest elevation had been previously used to assess
the performance of a prototype window, but the testing was completed several
years ago. This panel was modied to work with a new wood window; the
exterior cladding was removed, and the rough opening was adjusted to accom-
modate the test window.
Window Installation
Once the panel construction was modied, a spun bonded polyolen weather
resistive barrier WRB membrane was placed over the opening, and it was
subsequently cut and folded to comply with the manufacturers installation
instructions. The membrane was weather lapped with existing WRB layers
above and below the window opening.
FIG. 7Constructing and placing the sill pan.

FIG. 8Placing shims, securing the window, and insulating it.
A metal sill pan was custom made for this window opening. Care was taken
to bend the metal to provide leak-proof corners where the end dams and back
dams met. Consequently, when water was poured on the sill pan during a test,
the water was unable to leak past the back and end dam intersections. It owed
toward the exterior as desired. The sill pan was then bedded into several beads
of plastic sealant. Self-adhering exible ashing was wrapped around the rough
jambs and lapped into the pan at the end dams. This conguration ensures that
water leaking past the window is directed to the pan for drainage to the exte-
rior.
The window was then placed into the rough opening and shimmed to en-
sure that it was plumb, level, and square. Next, it was secured through the head
and jamb members. Rigid ashing was placed over the window head to shield
the window and to direct water on the WRB to the exterior. Insulation was
installed within the cavities along the head and jambs. Additional ashing tape
and sealant were applied between the rough opening and the window to ensure
that the installation was leak proof and airtight on the exterior and interior,
respectively.
FIG. 9The nal window and sensors.

FIG. 10Sensor panel and moisture content sensor circled.
Lastly, the exterior of the test panel was nished with vinyl siding. This
acted as an exterior cladding that would screen the panel from undesirable
environmental conditions 9. In conjunction with the installation, RH, tem-
perature, and moisture content sensors were placed on various components in
order to obtain data. A more detailed description of the sensor placement is
given on the following page.
Data Collection
With these sensors, computers collected the corresponding data at hourly in-
tervals. This data was then sent electronically to the researcher via e-mail for
conversion, organization, and analysis. Data collection began immediately after
the window installation was completed. From February to September, data was
recorded every hour of every day with one exception. A power failure caused an
FIG. 11Sensor placement picture and diagram. The lower shaded region represents
the sill pan. The upper shaded region represents the underside of the window sill.
FIG. 12Moisture content: FebruarySeptember. Trend lines are second order

polynomials.
interruption in data collection from June 19 to June 30. Retrieved data was
ultimately organized in Microsofts Excel, and corresponding graphs were
made within the same software.
Data Observation: Analysis

The data collected in this experiment included many different critical proper-
ties of the window assembly and the surrounding environment. Sensors applied
to the assembly and sensors within the NET facility collected the interior RH,
the interior temperature, the exterior RH, the exterior temperature, the surface
temperature on the interior gypsum wallboard, the surface temperature of the
exterior wall sheathing oriented strand board, the temperature of the cavity
below the window sill, the RH of the cavity below the window sill, the tempera-
ture of the sill pan, and the moisture content in the wood components of the
window near the sill pan. A diagram of the sensor placement in the window
assembly can be seen in Fig. 11.
Sensors RH1, RH2, T3, and T4 were placed on the underside of the window
sill. Sensors MC1, MC2, T1, and T2 were placed on the jambs beneath the sill
that rested in the sill pan.
The data, which was digitally collected and analyzed, was used to calculate
other critical properties, including the interior dewpoint, the dewpoint of the
FIG. 13Comparison of all calculated dewpoints FebruarySeptember.
cavity under the sill, the exterior dewpoint of the exterior, and the surface RH at
the point T1 in Fig. 11.
Analysis Part 1: Moisture Content

The analysis commenced with graphing the moisture content in the sub-sill
cavity for the entire duration of the collection period. The objective of this was
to search for spikes in the moisture content. If there were any particularly
prominent spikes, then an explanation for this surge would be needed.
Figure 12 shows that the moisture content did spike quite a few times.
Moisture content averaged at about 8.5 % over the entire experiment, and it
spiked above 10 % in some specic areas. The moisture content surged to a
high of 12.9 % on April 4. The 10 % mark was used as a threshold for this graph
because moisture content rarely declined below 7 %, which is 1.5 % less than
the average. The 10 % mark is 1.5 % greater than the average. Consequently, all
values that spiked over this value needed to be closely analyzed for an explana-
tion. The following dates crossed this threshold: February 19, March 29, April
3, April 6, April 14, May 6 May 7, May 11, May 17, May 24, May 25, May 28,
May 30, June 6, June 7, June 9, June 15, June 17, July 14, July 16, July 17, July
18, August 1, August 3, August 4, September 1, and September 3.
According to the National Climactic Data Center, it rained in Charleston on
the following dates: February 19, April 14, May 17, May 24, June 6, June 7, June
9, June 15, July 16, July 17, August 1, and August 4 10. For the majority of the
FIG. 14Dewpoint: FebruaryJune. Trend lines are second order polynomials.
remaining dates, rain was reported the day prior. This may be the reason for
the surge in moisture content on these particular days. However, on May 7 and
September 3, no rain was even reported the day before. Hence, an explanation
is needed for the spike in moisture content.
Two additional items of interest can be gathered from this graph. First, a
trend line shown in the graph as the black line peaks in the summer and
decreases as the winter approaches, indicating that the moisture content fol-
lows a seasonal trend rather than an accumulative trend. Second, there are
some noticeable divergent spikes in the graph between the two moisture con-
tent pins, MC1 and MC2. This shows that moisture was greater at one end of
the sill pan, indicating that water leakage, rather than atmospheric properties
in the cavity, may have caused this increase.
Analysis Part 2: Dewpoint Comparison

After conrming that there were spikes in moisture content, a determination on
whether or not this could be caused by sill pan condensation was necessary.
Thus, the next step was to determine the atmospheric properties of the air
within the cavity below the sill. Since this cavity acts as a median between the
interior and the exterior, it was essential to know if the air within the cavity was
mostly inuenced by interior or exterior conditions. If the exterior, uncondi-
tioned air was most prominent in the cavity, then there would be a signicant
chance that condensation and mold growth could occur in the cavity. Thus, a
calculation of the dewpoint of the three zones was completed since this would
FIG. 15Dewpoint: JulyAugust. Trend lines are second order polynomials.

be an ideal value for comparison dewpoint incorporates both temperature and
RH. Following this calculation, graphs of the three sets of dewpoint values
were created and compared to search for trends. Similarities between the three
sets of dewpoint values would be used to determine what inuenced the cavity
air most. The results indicated that the cavitys dewpoint followed the dewpoint
of the exterior more often than the interiors dewpoint. However, there were
some interesting trends in the data.
Figure 13 shows that the dewpoint of the cavity and the exterior were
nearly identical for the rst few months of the experiment. However, as the
Summer approached, the two seemed to slowly diverge. In the midst of the
summer, the cavitys dewpoint consistently started to show a pattern in its os-
cillations: its highest dewpoint values matched the exteriors dewpoint, but its
lowest dewpoint values matched the interiors dewpoint values. Thus, the dew-
point oscillated between matching the two zones. Further evidence of this can
be seen in Figs. 14 and 15.
An analysis of a few particular days shows that the dewpoint of the cavity
tended to match the interior dewpoint early in the morning. The dewpoint
would then rise to match the exterior dewpoint in the middle of the day and fall
to match the interior dewpoint as the morning approached. This indicated that
there may have been some air leakage from the interior that caused this trend
Figs. 14 and 15.
Analysis Part 3: Relative Humidity Comparison

The next step in the analysis was to examine the RH in the cavity. One possible
trend in this examination was of utmost importance: if the RH in the cavity
FIG. 16RH: FebruarySeptember. Trend lines are second order polynomials.
especially the surface RH at the wood averaged at 80 % for about 30 days,

then it could safely be concluded that the environment is capable of inducing
mold growth and subsequently causing decay 6. Figures 1621 provide that
information.
While the RH of the cavity did spike over 80 % quite a few times, it aver-
aged at about 60 % over the whole data collection period. This was not close
enough to the threshold needed to induce mold growth. The surface RH exhib-
ited the same behavior. The RH of the exterior averaged at about 80 % quite
often, while the interior, like the cavity, averaged at about 60 %.
In conclusion, the exterior air, with an average RH of 80 %, mixed with the
cavity air in greater amounts than the interior air did, especially towards the
beginning of the data collection period, but it did not inuence the cavity air
enough to bring its atmospheric or surface RH up to the 80 % threshold level
for over 30 days. This is likely attributed to interior air leakage, consequently
keeping the moisture content level of the wood relatively low.
Conclusion and Recommendations

From the preceding ndings, it can reasonably be concluded that sill pan in-
duced condensation does not occur often enough to induce related mold
growth. Hens 6 noted that an RH of 80 % over 30 days would promote mold
growth on wood. However, the experiment indicated that the cavity RH and the
surface RH averaged at about 60 %, just below the required threshold. Al-
FIG. 17RH at sensor RH1 from February to September. Trend line is a second order
polynomial.
though the RH values of these areas surged over 80 % in some instances, they
were quite brief, and a longer period of time under this condition would have
been required to promote mold growth. This low RH value may be attributed to
signicant interior air leakage which the researcher attempted to prevent. The
data indicated that leakage into the cavity occurred since the air properties of
both zones matched quite often, resulting in the oscillations seen in the graphs.
Hence, further research is needed, for the possibility of sill pan condensation in
conditions without interior air leakage still exists. It is especially interesting to
note that the exterior air and cavity air were very similar until the interior air
leakage began. Thus, if the interior air leakage had never occurred, the trend in
the beginning would have continued, and the outcome of the experiment may
have been entirely different.
A few improvements could have been made regarding the experiment. An
air leakage test of the window before and after the data collection period would
have provided insight on the air tightness of the assembly and would have
helped to determine if air leakage was a substantial factor in the experiment.
Also, recording manometers could have been used to collect data regarding the
air pressure differences between the interior, the cavity, and the exterior. These
two steps could have proven whether or not the interior air leakage mentioned
above was a factor. Additional steps could have been taken to reduce the air
leakage during the experiment.
One topic to reect upon is the inherent limitations for this experiment.
Unfortunately, only one panel was available for usage at the NET Facility. A
FIG. 18RH at sensor RH2 from February to September. Trend line is a second order
polynomial.
second concurrent window test would have been ideal, for this could further
validate the data collected. This would have also increased the chances of col-
lecting data from an installation that was not affected by interior air leakage.
Furthermore, an experiment that could have tested the effectiveness of differ-
ent designs/congurations and that was least susceptible to high moisture con-
tent levels and high RH levels would have been excellent. This may be an idea
for future research. Lastly, time was a limitation on the experiment: ideally, the
experiment should have collected data for another year to see if moisture con-
tent increased year over year. This would show that while the moisture content
trend is seasonal, it may also be accumulative to a certain degree.
Although sill pan induced condensation did not occur as often as expected,
the possibility still exists that it may occur often enough in an airtight window
assembly to cause mold growth. Had interior air leakage been eliminated in
this experiment, the results could have been much different. Additional re-
search on this topic could contribute useful data to the engineering eld.
Acknowledgments
The writer would like to thank ORNL for allowing him to use their facility and
equipment. The writer would also like to thank his advisor, Mrs. Floyd, for her
help and guidance throughout this process. Finally, the writer would like to give
FIG. 19Surface RH at sensor T1 from February to September. Trend line is a second

order polynomial.
FIG. 20Surface RH at sensor T2 from February to September. Trend line is a second

order polynomial.
FIG. 21Interior, exterior, and cavity RH comparison from February to September.

Trend lines are second order polynomials.
special thanks to his mentor, Mr. Larry Elkin, for his expertise, enthusiasm, and
patience from start to nish.
References
1 ASTM International, Builders Guide to Hot Humid Climates, ASTM International,

West Conshohocken, PA, 1998.
2 Harriman, L., Rodney, L., Patenaude, R., Chea, S. K., Elkin, L., Kosar, D., Murphy,
J., Rose, W., Werman, L., Bailey, R., Lawson, C., OBrien, S. Ow, C. S., Sekhar, C.,
Turner, S., The ASHRAE Guide for Building in Hot and Humid Climates, ASHRAE,
Atlanta, GA, 2009.
3 Canada Mortgage and Housing Corporation, Water Penetration Resistance of
Windows: Study of Manufacturing, Building Design, Installation and Maintenance
Factors, Research Highlight, CMHC, Ottawa, Ontario, 2003.
4 ASTM E2112-01, 2001, Standard Practice for Installation of Exterior Windows,
Door and Skylights, Annual Book of ASTM Standards, Vol. 1, ASTM International,
West Conshohocken, PA, pp. 4044.
5 American Society of Heating, Refrigeration, and Air-Conditioning Engineers, Inc.,
ASHRAE Handbook: Fundamentals, ASHRAE, Atlanta, GA, 2001.
6 Hens, H., Minimizing Fungal Defacement, ASHRAE J., Vol. 42, No. 10, 2000, p.
30.
7 Evans, S., Water Inltration, Mold Colonization, and Construction Defects, 2005,
http://www.mde.com/publications/H2OInltrate.pdf Last accessed March 20,

2009.
8 EIFS Industry Members Association, Executive Summary: Exterior Wall Cladding
Performance Study, 2007, http://www.eima.com/pdfs/
EIMA%20ORNL%20ExecSum%20Final.pdf Last accessed March 2, 2009.
9 American Architectural Manufacturers Association, Training Manual, ASTM Inter-
national, West Conshohocken, PA, 2000.
10 National Climatic Data Center, Quality Controlled Local Climatological Data for
Charleston, U.S. Department of Commerce, 2010, http://cdo.ncdc.noaa.gov/qclcd/
QCLCD Last accessed June 12, 2010.
CASE STUDIES
doi:10.1520/JTE102968
Sarah K. Flock1 and Garth D. Hall2
Interior Metal Components and the Thermal

Performance of Window Frames
ABSTRACT: In an effort to reduce the likelihood of condensation at metal

window frames and glazing, methods to improve the thermal performance
are now readily available, such as thermal breaks and warm edge technol-
ogy. Despite technological improvements, metal window frames may still ex-
hibit interior surface condensation if the installation details or interior environ-
mental conditions create circumstances that compromise or breach thermal
breaks. The U-factor, condensation resistance CR factor, and CR are often
utilized to compare the thermal performance of fenestration products. How-
ever, these performance parameters do not necessarily result in adequate
CR of windows in service. As a result, remedial measures are commonly
required to alter or modify the window details after installation. Computer
software programs can be used to evaluate the thermal performance of win-
dow components and various installation details under steady-state condi-
tions are available, as well as approaches to correcting condensation/frost
formation issues. Computer simulated repair approaches indicated that the
application of additional metal components on the interior may effectively
raise the interior surface temperatures of the window frame. Our study will
evaluate the feasibility of the metal n in practice, the impact of installation
techniques, as well as relate the ndings of simulations to mockups. In sum-
mary, this paper explores the affect of in situ repair efforts to raise the interior
surface temperatures of metal window frames and indirectly on the likelihood
of surface condensation on the frame.
KEYWORDS: condensation, thermal performance, modeling, THERM
program, interior exposure
Manuscript received January 10, 2010; accepted for publication October 8, 2010; pub-
lished online November 2010.
1
Senior Architect, Raths, Raths, and Johnson RRJ, 835 Midway Dr., Willowbrook, IL
60527.
2
Principal, Raths, Raths, and Johnson, Inc. RRJ, 835 Midway Dr., Willowbrook, IL
60527.
Cite as: Flock, S. K. and Hall, G. D., Interior Metal Components and the Thermal
Performance of Window Frames, J. Test. Eval., Vol. 39, No. 3. doi:10.1520/JTE102968.
169
Introduction
The term condensation is often misused and misunderstood with relation to
building performance. Condensation is the term used to represent the phase
change from vapor to liquid and forms when the surface temperature drops
below the dew point. Most people identify condensation as the fog that forms
on the bathroom mirror after a shower or the water droplets collecting on the
exterior surface of a cold glass on a hot, humid day.
Hygroscopic building materials such as precast concrete or brick masonry
rarely experience the formation of condensation and liquid water accumulation
when surface temperatures fall below the dew point. In these cases, the sorp-
tion of water vapor into the material becomes bound water and results in an
increase in the moisture content. Consequently, the moisture content of these
hygroscopic building materials is in constant ux due to sorption and desorp-
tion as the indoor and outdoor environments change 1. In many cases, the
sorption of moisture vapor from the surrounding air can be accommodated by
building assemblies without detriment. However, these same materials can ex-
perience frost formation when the exposed surface temperatures are simulta-
neously below freezing and the dew point. The accumulation and subsequent
thawing of the frost can contribute to damage to building materials.
Nonporous surfaces, such as glazing and frame components, can also pro-
duce condensation or frost formation that may contribute to localized deterio-
ration of building components, as well as reduce the overall system perfor-
mance. Although this problem can occur on the interior of fenestration
products regardless of frame material, the problem is often most prevalent with
metal frames. As a result, remedial measures are commonly required to alter or
modify the window details after installation. In the authors experience, repairs
that involve window removal to correct and mitigate condensation formation
can become costly and disruptive to occupants. Therefore, an effort to control
the condensation formation while the windows are in place is often a desirable
option. This paper will explore the affect of in situ repair efforts to raise the
interior surface temperatures of metal window frames to prevent surface con-
densation formation. Understanding the prevention of condensation formation
also relies on a basic knowledge of heat transfer mechanisms, dew point, meth-
ods of reducing heat transfer, and thermal performance ratings.
Heat Transfer
Heat ow occurs across the building envelope as the thermal energy tries to
reach equilibrium by moving from higher to lower temperatures. Heat is trans-
ferred through a wall or window system by a variety of means including con-
duction, convection, and radiation. The aforementioned heat transfer mecha-
nisms are dened as follows.
Conduction is the transfer of heat due to direct molecular contact within
a homogenous material or multiple materials. The rate at which this
transfer occurs depends on the conductivity of the materials or assem-
blies through which it ows and the temperature differential across the
materials 2.
FLOCK AND HALL, doi:10.1520/JTE102968 171
Convection is the transfer of heat energy due to the movement of a gas

or a liquid past a surface of different temperature. The amount of heat
transfer is dependent on the heat capacity of the uid transporting the
heat. When discussing buildings, the uid is normally air 2.
Thermal radiation is the process of heat transfer from one body to an-
other due to electromagnetic radiation 2.
In addition to the three transport mechanisms dened above, the bulk con-
vective heat transfer due to unconditioned or partially conditioned airow
through the exterior wall can be a signicant contributor to heat loss or gain in
buildings. Air migration can result in a localized increase or decrease in surface
temperatures of the building components due to convection and contribute to
the potential for condensation formation. For example, if outdoor air can mi-
grate through the exterior walls and reach the concealed portions of window
and door frames, micro climates can be created which vary signicantly from
the building design parameters.
Dew Point
Water will condense out of moist air and form frost or water droplets on a
nonporous surface when the surface temperature drops below dew point. The
dew point temperature for moist air is determined by the temperature and
moisture content of the air as measured by relative humidity RH. RH is the
ratio of water vapor in the air at a specic temperature to the maximum
amount that air can hold at that temperature and is expressed as a percentage.
Controlling Heat Transfer

As discussed previously, heat moves from higher to lower temperatures in an
attempt to reach equilibrium. For example, in heating climates, the direction of
heat transfer is from inside out during the Winter months. A common location
of heat loss and lower surface temperatures at the building envelope is at
window openings. Controlling the heat transfer at fenestration products often
requires study and consideration of all the inuential factors including the
materials used for frame construction, installation details, and operating con-
ditions at the interior, as well as proximity to a heat source.
Frame Materials
The conductivity of the material selected for the window construction has a
direct inuence on the surface temperature of the frame. For commercial and
high-rise residential design, aluminum emerged as the window and curtain
wall framing material of choice due to its favorable framing characteristics,
structural properties, and density. However, one drawback of metal window
frames is the signicant increase in the thermal conductivity in comparison to
wood, vinyl, and berglass. Based on the generic material properties published
in the American Society of Heating, Refrigeration and Air-Conditioning Engi-
neers ASHRAE Handbook of Fundamentals, the conductivity value for alumi-
num is 1526 Btu in. / h F ft2, in comparison to 0.063 Btu in. / h F ft2 for
wood and 2.8 Btu in. / h F ft2 for vinyl 3. Although, the conductivity rates
can vary as a result of temperature changes; higher conductivity values ulti-
mately result in greater heat transfer and corresponding heat loss. In addition,
the conductivity rate of solid materials is only one factor among many contrib-
uting to condensation formation.
Because window frames with high conductivity rates conduct heat more
readily from the interior to the exterior, interruptions or breaks are now pro-
vided in the cross-sections of the aluminum framing members to reduce the
heat ow and improve the interior frame temperatures. First introduced in the
1950s, the use of a poured-and-debridged thermal break, became one method
of overcoming the higher conductivity values associated with aluminum. Due
to a number of performance issues, such as thermal break shrinkage, manufac-
turers have further rened and improved the traditional technologies in recent
years, incorporating the use of thermal struts and polyamide in many newer
window products.
Glazing Materials
Prior to the 1960s, the windows installed in most buildings were glazed with a
single pane of glass. During Winter in northern climates, storm windows were
often installed in conjunction with single panes to address thermal issues and
reduce air inltration. As a result, signicant improvements for transparent
glazing components stemmed from assembling multiple glass panes into one
unit. These congurations are identied as insulating glass units or IGUs. The
addition of each air pocket and glazing material can result in a reduction of
thermal transfer. Although, double panes are the most common, triple glazed
products are sometimes used when the desire for increased energy efciency
exists. Continued developments with surface coatings i.e., low-e have also
improved the performance of IGUs. Increased thermal benets may also be
achieved by incorporating gas lls in the space between panes, such as argon or
krypton. However, concerns exist with gas lling practices and their mainte-
nance over time.
Spacers
The element that separates the panes of glass in an insulated glass unit is
known as the edge spacer. At one time, non-thermally broken aluminum box
spacers were used; however, the high thermal conductivity of aluminum cre-
ated greater heat loss at the edges of the IGU. This led to improvements in
spacers, commonly referred to as warm edge technology. Warm edge technol-
ogy can be achieved by a variety of materials and congurations including, but
not limited to, desiccated matrix, thermally broken aluminum, and thermally
improved stainless steel. A sealant is used in combination with spacer technolo-
gies at the perimeter of the glazing unit. Butyl, silicone, and urethane products
have all been utilized in an attempt to create durable seals; however, the overall
performance and durability of the seal system can be dependent on implement-
ing good workmanship during fabrication.
TABLE 1Interior RH and dew points.
Interior Relative
Humiditya % Dew Point F C
10 13.26 10.4
20 27.7 2.4
30 37.2 2.9
40 44.6 7.0
50 50.5 10.3
a
Using 70 F 21 C as the interior temperature.
Installation Details
The design of opaque walls and their integration with fenestrations also di-
rectly affect the thermal performance of window units. For example, the tem-
perature of the cavity air within a brick veneer wall may be close to that of the
exterior air. During Winter months, if the window frame is directly exposed to
the cold, cavity air, the surface temperatures at the frame can be lowered. Often
times, attempts at controlling low surface temperatures at the frame, as well as
subsequent condensation formation, involves the application of insulating
components and/or air seals adjacent to the fenestration product. The intent of
insulation is to prevent the heat migration away from the window components,
as well as provide an air barrier from cold cavity air. However, inappropriately
placed insulation can also have a detrimental effect, serving to preclude the
interior heat from reaching the building components susceptible to condensa-
tion formation.
Operating Conditions
Building operations can also have a substantial effect on fenestration perfor-
mance. Even though measures are incorporated to maintain relatively consis-
tent temperatures and interior RH within many facilities, this condition may
not be practical in all cases. In buildings with higher indoor humidity require-
ments or spaces where control is not provided, an elevated interior humidity
can contribute to moisture problems. As the RH increases, the dew point raises,
resulting in an greater potential for condensation or frost at exterior walls and
fenestration products. As shown in Table 1, when the interior temperature is
70 F 21.1 C and the interior RH is 30 %, the dew point is 37.2 F 2.9 C.
However, when the interior temperature is 70 F 21.1 C and the interior RH
is reduced to 20 %, the dew point is 27.7 F 2.4 C. A RH of 30 % may be
typical of a hospital operating room with mechanical humidication, whereas
20 % may be closer to that of an ofce space.
Heat Source
Heated air will fall as it cools due to natural convection. This phenomenon can
produce stratication along cold walls and windows, as well as contribute to
colder air temperatures at the sill. In the past, the installation of radiators
below window openings helped provide an upward ow of heated air that

washed across the window surfaces and raised the temperatures. However, the
benets of increased surface temperatures may be lost if the heating methods
are not positioned appropriately or are blocked by window treatments. Addi-
tionally, if the hot washing airow has a high RH, it can provide considerable
water vapor, which may augment the potential for condensation. Forced air
systems can supply heat directly over windows to provide some of the same
benets but must be designed accordingly. In some cases, heating elements,
such as heat trace, can be added directly to the interior framing members or
the glazing components to raise the surface temperatures and reduce the con-
densation potential. However, these options can impact energy costs and may
do little to affect surfaces not in direct contact with the heat source. Alterna-
tively, passive approaches, such as exposing more of the frame to interior air,
may be considered to capitalize on available heat rather than to create new
sources. This option will also be discussed in greater detail later in this paper as
a repair approach for existing windows.
Thermal Performance Ratings
Reducing heat loss in residential structures rst gained attention in the 1970s
due to the high costs associated with energy use. This desire for enhanced
energy efciency led to advancements in wall assembly materials such as insu-
lation with higher R-values, windows with lower U-factors, as well as higher
condensation resistance factors CRFs. R-value is the measure of thermal re-
sistance. However, fenestration products are not accurately rated by R-values
due to the effects of solar radiation and air transfer. The U-factor
Btu/ h F ft2, also known as thermal transmittance, accounts for more
variables than the R-value and is typically associated with fenestration prod-
ucts. The U-factor is the total heat transfer based on conduction, convection,
and thermal radiation. In simplest terms, the lower the U-factor, the lower the
rate of heat transfer, and U-factors for windows are typically found between
0.20 and 1.20. However, windows are not homogeneous products; therefore, the
U-factor of the specimen is not constant across the sill, jamb, sash, or glazing.
Various test methods are available to obtain specimen U-factor including
AAMA 1503.1-97, Voluntary Test Method for Thermal Transmittance and Conden-
sation Resistance of Windows, Doors and Glazed Wall Sections 4, ASTM C1199,
Standard Test Method for Measuring the Steady-State Thermal Transmittance of
Fenestration Systems Using Hot Box Methods 5, and NFRC 100, Procedure for
Determining Fenestration Product U-Factors 6. These test methods standardize
specimen size, test conditions, instrumentation products, and calibration tech-
niques required to produce the U-factor of a fenestration product. The thermo-
couples are used to measure the interior surface temperatures of the frame and
glass of standard-sized window specimens. The U-factor is then calculated
based on the average difference between the warm and cold sides relative to the
time rate of heat ow. These test standards are intended to evaluate the perfor-
mance of the window product alone.
The CRF was established by American Architectural Manufacturers Asso-
ciation AAMA in 1972 in order to allow designers and speciers to compare

the laboratory generated condensation resistance of window products 7,8.
The CRF allows for a more direct comparison than the U-factors. The CRF
calculation methods are also addressed in the AAMA 1503. The number is in-
tended to provide a relative comparison of condensation resistance. However,
the rating is based on an average of values around the window, therefore, any
localized areas of poor thermal resistance commonly the sill are less apparent
in the overall rating. Though AAMA 1503-98 9 includes provisions for testing
units to simulate the eld conditions, it is often difcult to nd data for these
conditions and to determine what accessories are included in published values.
These limitations have contributed to the development of other rating systems
within the industry.
In 2001, the National Fenestration Rating Council NFRC developed an
alternate rating for the condensation resistance, known simply as the CR con-
densation resistance and dened in the standard NFRC 500 10. This rating is
based on similar interior and exterior temperatures as the AAMA CRF; how-
ever, this value also accounts for a range of interior RHs. In addition, this rating
is based on the lowest of the frame temperatures, edge-of-glass, and center-of-
glass as opposed to an average of frame and glazing temperatures utilized in
the CRF calculations. The CR also considers the differences in vapor pressure
at the condensing surface. Because vapor moves from high to low, the vapor
pressure at the condensing surface is reduced, therefore increasing the water
vapor drive toward the fenestration product. The NFRC considers this effect by
taking into account the difference between the dew point and surface tempera-
ture. The NFRC 500 also includes the provisions for both computer simulation
and test methods. Unfortunately, no known relationship exists between the
NFRC CR and the AAMA CRF to date. The designers and manufacturers alike
have begun to specify according to both ratings but without a complete under-
standing of how to compare the values.
Despite technological advancements in window systems to reduce heat ow
and the development of industry ratings to assist designers in product selec-
tion, the condensation formation and accumulation on windows in cold cli-
mates remains a signicant problem. In addition to the design and manufac-
ture of the fenestration product, many site variables exist that affect the
thermal performance. Variables such as installation details, interior operating
conditions, or proximity to a heat source are not accounted for by the afore-
mentioned test methods or industry ratings; therefore, a window may not per-
form in as-built conditions as predicted by the test methods. As a result, de-
signers and consultants are often enlisted to alter or modify the window details
to achieve enhanced thermal performance after installation. In the authors
experience, solving condensation problems often involves investigation to de-
termine as-built conditions, evaluation of repair concepts with simulation pro-
grams, as well as mockups of proposed solutions prior to large-scale implemen-
tation.
Simulation Program and Modeling

Throughout the authors professional careers investigating the building prob-
lems, numerous condensation-induced moisture problems in structures in cold
climates have been encountered. Condensation is commonly a problem in con-

cealed wall and roof assemblies but is also readily observed on the occupied
interior at fenestration products. The authors employ a variety of methods for
analysis when evaluating condensation problems.
Simulation modeling is one tool available to designers to evaluate in situ
performance or options for remediation under steady-state conditions.
Throughout the course of the authors experience, many remedial options have
been modeled to improve surface temperatures at metal window frames in an
effort to reduce the potential for condensation. Such repair options have in-
cluded adding insulation to the frame components, altering the adjacent build-
ing materials, and applying metal components at the interior. Comparatively,
the installation of metal components at the inside of the frame yielded signi-
cantly higher interior surface temperatures than others in most instances.
The computer simulation program THERM was employed in order to
study the effects of the metal n application at the interior of the window
frame. THERM is a software package developed by Lawrence Berkley National
Laboratory, which applies the nite-element method to steady-state, two-
dimensional details to evaluate the effects of heat transfer 11,12. The surface
temperatures determined by the simulations can be used to estimate the behav-
ior of windows and potential of condensation formation by applying the psy-
chrometric principles of moist air based on the interior environmental condi-
tions.
In addition to numerical results, the program can produce various graphic
outputs for interpretation. When performing modeling in accordance with
NFRC, the simulations are based on an assumed exterior temperature of 0.4 F
18 C and interior temperature of 69.8 F 21 C. The interior RH is con-
trolled at 15 % maximum in the simulations. These are industry standard pro-
cedures for comparative purposes and are not intended to be representative of
any particular installation or exposure condition. In practice, such controlled
conditions are not always realistic.
It is also important to understand the limitations of the computer software.
As stated previously, THERM evaluates the window details under steady-state
conditions in only two dimensions, which does not account for three-
dimensional heat transfer phenomena. In addition, THERM does not account
for the effects of airow or building orientation. Consequently, the authors built
mockups to monitor and understand if the results indicated by modeling with
the metal n in place were achievable in practice. At the conclusion of the
mockup monitoring period, the temperature improvement ndings from the
simulations were reviewed with the mockup ndings for relative comparison
only.
Simulations of Fin Installation
The simulation study undertaken evaluated two different window products,

each of slightly different size and conguration. Both windows were commer-
cially available, thermally broken, aluminum hung units with anodized bronze
frames. The window congurations were widely used throughout the United
TABLE 2THERM models with exterior temperature at test site.
Surface Temperature,
Conditiona F C
Window 1, without n 37.7 3.2
Window 1, with n 3.5 in. vertical leg 43.0 6.1
Window 1, with n 3.5 in. vertical leg
and thermally conductive paste 42.7 5.9
Window 2, with n 3.5 in. vertical leg
and thermally conductive paste 44.4 6.9
a
Using 10 F 12 C and 70 F 21 C as boundary conditions.
States in residential and light commercial structures and, based on many build-
ing investigations, have experienced interior condensation problems at the
frame. Window 1 measures 25 3/4 in. 64.4 cm wide by 38 in. 96.5 cm high.
The frame extrusion is 41/ 2 in. 11.4 cm deep and is fastened to the wood
buck surround at the jambs only. The window does not incorporate a nailing
ange or other accessory components. Window 2 measures 23 in. 58.4 cm
wide by 36 in. 91.4 cm high, and the frame extrusion is 3 1/2 in. 8.9 cm in
depth. Again, the window is fastened to the adjacent construction at the jambs
only and did not incorporate a nailing ange. No solar sheeting or low-e coat-
ings were utilized at either window specimen.
In this study, the authors used an outdoor temperature of 10 F 12 C as
the exterior temperature for the simulations. This temperature value was se-
lected to coincide with the lower limit of the measured outdoor temperatures
during the initial mockup. This temperature falls between the 99 % heating dry
bulb exterior design temperature for suburban Chicago of 0 F 17.8 C as
published by ASHRAE 2 and the average minimum temperature of 20 F
6.7 C during the month of January 13.
Several iterations of installation details with and without interior ns at the
windowsills were modeled. The metal n detailed in the simulation was in
direct contact with the frame and with the application of thermal conductive
paste. The results of the simulations at the sill section of windows are listed in
Table 2. For the purpose of comparing the results of the various computer
simulations, the results are characterized by the surface temperature at the
lower edge of the sill frame. As indicated by the results, the application of a
metal n at the interior produced increased surface temperatures at the win-
dow frames. Additionally, increasing the size of the metal n resulted in an
additional rise in surface temperatures due to increased exposure to the inte-
rior heat. Therefore, if the results shown in Table 2 are compared against the
dew point conditions listed in Table 1, window 1 without the metal n may
result in condensation formation at an approximate interior RH of 30 % and
above. With the n in place, the surface temperatures are raised so that the
TABLE 3THERM models at exterior temperature of F.
Conditiona F C
a
Using 0 F 17.8 C and 70 F 21 C as boundary conditions.
potential for condensation formation at window 1 is reduced, requiring the

interior RH to approach 40 % for the dew point to occur. At window 2 without
the n in place, the condensation formation may result at an approximate in-
terior RH of 35 %, whereas with the addition of the n, the interior RH can be
raised closer to 40 % without creating the potential for condensation.
In order to achieve the direct contact evaluated in the simulation models,
mechanical attachment would likely be incorporated in practice. However, be-
cause mechanical attachment may also result in areas where the n is not in
direct contact with the frame, the use of thermally conductive paste was also
analyzed. The paste utilized had a conductivity of 52.04 Btu in. / h F ft2
7.5 W / m K.
Another consideration is the impact of the exterior temperature on the
interior frame surface temperature. As the outside temperature falls, a lower
interior humidity level is needed to prevent the condensation formation at the
interior. To illustrate this, the models were also calculated for window 1 with
and without the n by using an exterior temperature of 0 F 17.8 C, as
indicated in Table 3. Under these conditions, window 1 may result in a conden-
sation formation at an approximate interior RH of 25 % and above without the
metal n applied. With the n in place, window 1 reduces the potential for
condensation until the interior RH approaches 30 %.
Mockups of Fin Installation

The authors installed the window specimens in a mockup chamber that was
specically constructed to house the two fenestrations and expose them to the
Winter environment in the Chicago suburbs. The interior chamber conditions
were generally maintained at 70 F + / 5 F and were monitored under various
test congurations during this exposure. Initially, the baseline temperatures of
both specimens were established without the metal n in place. A metal n was
then installed at one jamb of each window specimen with fasteners, and data
was collected from both the modied and unmodied portions of the frame.
The metal n was then removed and reinstalled with thermally conductive
paste in conjunction with fasteners and monitored in the same fashion. Next,
the metal n was removed from the jambs and installed at the sill frame of both
windows by using fasteners only. Data was collected at the modied portion of
the sill frame, as well as immediately adjacent to the modied portion. Lastly,
the addition of the thermally conductive paste was installed at the backside of
FIG. 1S. elevation of mockup chamber.
the metal n and reattached at the sill frame only. Data was collected in the
same manner as with the prior conguration.
The wall construction of the mockup chamber, from exterior to interior,
consisted of a weather-resistive barrier over wood sheathing, wood stud fram-
ing, and 2 in. expanded polystyrene insulation installed between framing mem-
bers. The windows were anchored to wood-frame surrounds and sealed to the
wood-frame members at the exterior and interior perimeters. It is the authors
experience that buildings often integrate windows into the wall system that are
ush with the cladding at the exterior, resulting in an interior recess at the
perimeter. The placement of the window and the presence of a return on the
interior can be signicant as it has a direct impact on the exposure of the
window frame to the interior heat source. Therefore, the mockup was cong-
ured with interior returns to evaluate the conditions often investigated for con-
densation problems. No nishes were installed at the interior of the mockup
chamber.
The exterior of the chamber was exposed to the elements during the Winter
climate in suburban Chicago Figs. 1 and 2. The exterior temperatures were
monitored with the use of thermocouples and recorded at regular intervals. The
mockup was oriented facing north so that the exterior of the windows was not
exposed to direct sunlight in an attempt to limit the effects of solar radiation
and heat gain. The interior of the mockup was heated by a radiant source
placed equidistant between both window units. No additional heating or venti-
lation was provided. The RH of the interior space was monitored; however, no
attempt at control was initiated as a part of this evaluation. The average inte-
rior RH in the hut for January and February was 18 %.
Thermocouples were installed at the interior of the window frames, and
surface temperatures were recorded at regular intervals. In order to compare
the surface temperatures of the window frames from one mockup congura-
tion to another, data points were selected at comparable interior and exterior
air temperatures. The baseline surface temperature data recorded at the win-
dowsill without the metal n in place were found to be generally representative
of the THERM models for window 1 and window 2 as shown in Table 4, when
FIG. 2N. elevation of mockup chamber.
the interior and exterior temperatures for these windows were similar to the
simulated boundary conditions.
The rst modication to the mockup windows consisted of the addition of
a metal n to the interior of the window frame. The n consisted of a 1-5 / 16
3-1 / 2 1 / 8 in.3 aluminum angle mechanically fastened to the frame using
self-tapping screws through the jambs at the top, middle, and bottom. Thermo-
couples were placed at the n, the adjacent jamb frame, and the unmodied
jamb without a n applied. The window temperatures shown in Figs. 3 and 4
were recorded. The temperatures enclosed within the parentheses are baseline
temperatures without the n installed, while the others are surface tempera-
tures with the n in place. The temperatures were veried with an infrared
thermometer and found to be in agreement with those obtained by the thermo-
couples.
The installation of the ns on both windows in the mockup identied the
relative temperature increases along the jamb of 3.2 3.9 F on window 1 with
a 4 1/2 in. frame section with the n in place. At window 2 with a 3 1/2 in
frame section, the temperature differentials with the n in place along the
jamb were 3 4.9 F. No impact on the sash and glazing temperatures were
noted upon the application of the n at the jamb in comparison to baseline
conditions; therefore, no further monitoring was performed on these compo-
nents. The installation of the ns on both windows in the mockup identied the
TABLE 4Simulation and in situ surface temperatures.
Conditiona F C
Window 1, THERM 37.7 3.2
Window 1, chamber 37.9 3.3
Window 2, THERM 40.5 4.7
Window 2, chamber 40.5 4.7
a
Using 10 F 12 C and 70 F 21 C as boundary conditions.
FIG. 3Window 1 with jamb n.
temperature differentials between the n and adjacent frame of 1.2 3 F at

window 1. At window 2, the temperature differentials along the jamb were
0.7 2.1 F between the n and adjacent frame at the jamb. The localized af-
fects of the mechanical fasteners and their impact on the transfer of heat be-
tween the n and the window frame were also evaluated. The temperature data
did not show a notable difference in the frame temperatures immediately ad-
jacent to fasteners.
Although the temperatures obtained with the n in place at the mockup
windows showed relative improvement, the THERM models showed a lesser
temperature differential between the n and the frame. In THERM, the tem-
perature difference between the n and frame, 1 / 2 in. away from the n, was
0.7 F for window 1 and 0.5 F for window 2 Figs. 5 and 6. Therefore, the
mockup was modied to include the use of the thermally conductive paste
between the n and the window frame, in addition to mechanical fasteners, in
an attempt to improve the interface continuity in the as-built conguration.
The ns were installed by using one line of a thermally conductive paste.
By using the thermally conductive paste, the temperature difference at win-
dow 1 was reduced to a range of 0.1 1 F from the n to the adjacent jamb
frame. For window 2, the temperature differences were reduced to 0.4 1.4 F
FIG. 4Window 2 with jamb n
FIG. 5Window 1.
FIG. 6Window 2.
at the jamb. The data reveals an improvement in heat transfer between the n
and the window frame with the thermally conductive paste applied between the
n and the frame.
The next conguration at the mockup windows consisted of the removal of
the jamb n and the addition of a metal n to the sill. The thermocouples were
again placed at the n and the adjacent frame. No monitoring was performed at
the head or jambs during this phase of the study. The following window tem-
peratures were recorded and again, the temperatures obtained with the n in
place at the sill of the mockup windows showed a relative improvement that
resulted in higher sill extrusion temperatures. The installation of the sill ns on
both windows in the mockup identied the temperature differences between
the n and frame to be 1.2 3.5 F at the sill on window 1 and 0.9 2.6 F at
window 2. However, THERM predicted less of a temperature difference be-
tween the n and the frame. The mockup was again modied to include the use
of a thermally conductive paste between the n and the window frame. The
addition of the thermally conductive paste at the interface between the n and
the frames reduced the temperature difference between the two components
and resulted in relative improvements closer to those suggested by THERM. By
using the thermally conductive paste, the temperature difference at the sill of
window 1 was reduced to 0.2 0.9 F, and at window 2, the temperature differ-
ence was reduced to 0.3 1.2 F at the sill. The temperatures enclosed within
the parentheses are baseline temperatures without the conductive paste in-
stalled, whereas the others are surface temperatures with paste Figs. 7 and 8.
Conclusions
The ndings described in this paper are intended to provide information for the
consideration of repair designers and consultants when modifying existing win-
dows that display condensation problems at the frame. However, further study
is needed to rene and understand other variables of this concept including the
FIG. 7Window 1 with sill n.
impact of having ns applied to all sides, the conguration of the n, alternate

attachment methods, the depth of the window frame, and the implications of
heated, forced air directed at the windows. Future study is planned by the
authors to address these issues.
The data from the mockups indicates a relative increase in surface tem-
peratures when the interior metal ns are applied. However, the magnitude of
improvement is impacted by a variety of factors including, but not limited to,
conguration of the frame, size of interior n, and attachment method. As with
all approaches, this method also has limitations that must be understood by
designers, owners, and users prior to implementation in order to benet from
the improvements.
In some instances, the constructability issues presented by existing condi-
tions can result in circumstances where the optimal performance calculated by
THERM may not be attainable. While variations between simulation ndings
and chamber results can be anticipated due to certain factors not accounted for
in the THERM models, a lesser improvement is also likely due to difculties in
FIG. 8Window 2 with sill n.
achieving full contact between the remedial n and installed windows The ad-
dition of a thermally conductive paste between the remedial n and frame
resulted in a greater thermal improvement, closer to that of the computer simu-
lations, but also impacts the costs associated with remediation.
Although window conguration and installation can impact performance,
the exposed surface area of the n may also impact the relative temperature
increase at the frame. Increasing the exposed length of the n in THERM mod-
els to 5 1 / 2 in. 14 cm suggested a potential increase in frame temperatures
in comparison to window units without the n of 7 F 12.2 C to 9 F
12.7 C depending on the window conguration, in contrast to 4.3 F
15.4 C to 5.3 F 14.8 C with a 3 1/2 in. n. Covering the n with stools
or other interior nishes may be desirable for aesthetic concerns, but may
reduce the relative effectiveness of the n in improving surface temperatures on
the frame. For this reason, the use of stools, interior nishes at the window
return, or other materials that reduce the exposed window area should be ap-
proached with great care when attempting to increase the frame temperatures.
In conclusion, although the application of a metal n is not a substitute for
specifying appropriate products with good detailing in new construction, this
approach does provide a viable remediation approach for consideration when

attempting to reduce the potential for condensation formation.
References
1 Rose, W., Water in Buildings: An Architects Guide to Moisture and Mold, John Wiley
& Sons, Inc., Hoboken, NJ, 2005.
2 Straube, J. and Burnett, E., Building Science for Enclosures, Building Science
Press, Westford, MA, 2005.
3 American Society of Heating, Refrigerating, Air Conditioning Engineers, Inc.,
ASHRAE Handbook: 2001 Fundamentals, ASHRAE, Atlanta, GA, 2005.
4 AAMA 1503.1-97, 1998, Voluntary Test Method for Thermal Transmittance and
Condensation Resistance of Windows, Doors and Glazed Wall Sections, AAMA,
Schaumburg, IL.
5 ASTM C1199, 2000, Standard Test Method for Measuring the Steady-State Ther-
mal Transmittance of Fenestration Systems Using Hot Box Methods, Annual Book
of ASTM Standards, Vol. 04.06, ASTM International, West Conshohocken, PA, pp.
621639.
6 NFRC 100, 2001, Test Method for Determining U-Factors of Fenestration Sys-
tems, National Fenestration Rating Council, Silver Spring, MD.
7 AAMA 1502.3-1972, 1972, Voluntary Test Method for Condensation Resistance of
Windows, Doors and Glazed Window Sections, AAMA, Schaumburg, IL.
8 AAMA 1503.1-80, 1980, Voluntary Test Method for Thermal Transmittance and
Condensation Resistance of Windows, Doors and Glazed Wall Sections, AAMA,
Schaumburg, IL.
Condensation Resistance of Windows, Doors, and Glazed Wall Sections, AAMA,
Schaumburg, IL.
10 NFRC 500, 2002, Procedure for Determining Fenestration Product Condensation
Resistance Values, National Fenestration Rating Council, Silver Spring, MD.
11 THERM 2.0: A PC Program for Analyzing the Two-Dimensional Heat Transfer
Through Building Products, LBL-40682. 1998. Lawrence Berkeley Laboratory,
Windows and Daylighting Group, LBNL, Berkeley, CA.
12 WINDOW 5.0: User Manual, LBL-44780. 2001. Lawrence Berkeley Laboratory,
Windows and Daylighting Group, LBNL, Berkeley, CA.
13 NOAA, Maximum/Minimum Temperature and Precipitation for Chicago, January
2008.
doi:10.1520/JTE102967
Andrew A. Dunlap,1 Paul G. Johnson,2 and Curt A. Songer3
Controlling Condensation Through the Use of

Active and Passive Glazing Systems
ABSTRACT: Special building occupancies, those with a high interior relative

humidity, have a specic set of performance requirements to control conden-
sation for exterior window and curtain wall systems. High levels of humidity,
if not properly accommodated by glazed exterior wall systems, can result in
condensation on and within glazing systems and adjacent construction. The
performance of glazing has typically been viewed and analyzed based on
overall system performance that generally results in a set of values that do
not apply to project specic applications. With this approach, important as-
pects of a system, as applied to specic buildings, are often overlooked. In
this circumstance, the result too often is unacceptable levels of condensa-
tion. Unfortunately, many within the design, manufacture, and construction
communities do not fully understand the importance of treating each building
for the unique set of conditions that it is, and the resultant consequence of
not analyzing and treating each humidied building as a unique set of mate-
rials, systems, and environmental conditions. Similarly, the analysis methods
and tools necessary to predict and prevent condensation are even less fa-
miliar within the design and construction industry. As a result, decient sys-
tems and materials may often be installed in these applications, resulting in
failures ranging from minor inconvenience to complete loss of service. There
are, however, methods that can be used to reduce or eliminate such prob-
lems. This paper describes methods and procedures that can be utilized to
understand and identify the performance levels required for high humidity
spaces and analysis methods including computer modeling to predict per-
lished online January 2011.
1
AIA, Associate, SmithGroup, Inc., Detroit, MI 48226.
2
FAIA, Principal, SmithGroup, Inc., Detroit, MI 48226.
3
P.E., Vice-President, SmithGroup, Inc., Detroit, MI 48226.
Cite as: Dunlap, A. A., Johnson, P. G. and Songer, C. A., Controlling Condensation
Through the Use of Active and Passive Glazing Systems, J. Test. Eval., Vol. 39, No. 4.
doi:10.1520/JTE102967.
187
formance of systems and materials, and this paper also describes both ac-
tive and passive technologies that have been successful in meeting these
needs. Passive design, through the use of high performance glazing, and
active technologies, such as heat tracing and heated glass, are considered.
Benets, risks, and appropriate uses of each are identied. Examples are
included to illustrate these approaches.
KEYWORDS: condensation, active glazing control, passive glazing
control, wall systems, high humidity, glazed exterior wall systems, con-
densation control, hygrothermal analysis, thermal analysis
Background
The following incorrect combinations of certain conditions can and often
does lead to signicant problems of condensation on and within exterior wall
glazing materials and assemblies Fig. 1. These combinations can include the
following:
Interior humidity
Interior temperature
Exterior temperature
Exterior humidity
Wall assemblies and materials
As a concept, it is a simple mix of common sense and physics; however, it is
a concept that is too often overlooked or misunderstood or simply ignored by
building designers, glazing system manufacturers, constructors, and building
owners.
This fundamental concept is the starting point for this discourse. Unfortu-
nately, it is also a concept that is not well understood nor adequately considered
by all who design, manufacture, and construct buildings and exterior glazed
wall systems and assemblies.
While these introductory statements may seem to be overly simplistic to
those familiar with these concepts as they apply to both buildings and exterior
FIG. 1Example of signicant condensation problems.

DUNLAP ET AL., doi:10.1520/JTE102967 189
wall design, our industry, design, and construction of buildings has known
about condensation for at least 60 years, and yet there is still confusion about
the physical realities of this issue and how to mitigate or control condensation.
In 1991 19 years ago!, Carlson wrote, Numerous books and technical
reports have been written on methods of preventing condensation-related
maintenance in buildings, yet the problem still persists. The simplest and most
illustrative book on the design of insulated buildings to prevent condensation
related maintenance problems was written by T. S. Rogers, architect, in 1951.
1 Since 1951 and certainly since 1991, there has been an abundance of re-
search and excellent development and distribution of analytical tools to ad-
equately evaluate and prevent condensation. There has also been abundant
discussion and education regarding these problems by committees within or-
ganizations such as the American Society for Testing of Materials committees,
specically Committee E06 on Performance of Buildings formed in 1946. This
does not even consider the work of the American Society of Heating, Refriger-
ating and Air-Conditioning Engineers ASHRAE and The National Institute of
Building Sciences groups, not to mention those outside of the United States,
even more concerned with condensation. However, given all of this information
and the wonderful tools, what is missing is a simple commitment from design-
ers, owners, manufacturers, and constructors to understand the issue and take
appropriate actions during the design, construction, operation, and mainte-
nance of buildings.
This paper addresses methods to accommodate and control the conse-
quences that will occur when signicant differences between interior humidity
and temperature, and exterior humidity and temperature exist for a building.
We will specically describe glazing systems and their adjacent construction.
These are described in a simple and direct way that will be useful to those
seeking information.
This paper does not address the following:
When buildings should be humidied
What types of buildings should be humidied
What level of humidication is required for specic building types
Whether or not some condensation is acceptable
How much condensation represents an acceptable amount
At what outside temperature condensation will start to occur and
How much of a typical year will be at or below that temperature
It should also be noted that the bulk of the experiential examples presented
in this discussion are based on commercial and institutional buildings and the
environments, materials, and systems normally associated with such buildings.
The problems are the same for other building and systems types, as are the
approaches for analysis and design. What may be different are the application
of a solution and perhaps the extent of the condensation problem.
Introduction
Properly designed buildings recognize and accommodate interior environmen-

tal conditions appropriate to their intended use and purpose. The building
owner and designer should evaluate building uses and function to determine
the temperature and humidity conditions that will be provided.
These two measurable interior environmental properties, in combination
with exterior climatic conditions, must be considered in the design of the exte-
rior envelope and therefore must be properly understood and evaluated. Proper
evaluation of these conditions will provide the building and exterior wall or
envelope designer with the information required to properly design and specify
the wall systems.
The results of these differences may be condensation at an unacceptable
level within or upon the surfaces of wall systems or components. In order to
avoid this condensation, or limit its occurrence, it is necessary to identify a
number of important physical conditions of the building spaces and wall sys-
tems and components. These include the following:
Interior air temperatures
Interior relative humidity RH
Exterior air temperature
Exterior RH
Moisture permeability of each glazed wall component or system
Heat ow properties of the various materials within and in contact with
the glazed wall system
When each of these conditions are adequately understood, there are meth-
ods that can be employed to determine where condensation will occur, under
which conditions it will occur, in what amounts it will occur, and therefore the
risk. By utilizing the evaluation tools available today to determine these factors,
a wall designer can start to develop physical properties of the wall system that
can control the risks attendant to uncontrolled condensation. This process may,
and most often does, require multiple evaluations and wall system modica-
tions in order to develop an assembly with the requisite properties to provide
the appropriate performance.
The important issue is that modeling of the wall systems can and should be
performed before the wall design is complete, and in effect the model provides
the necessary tool to guide the wall design. This is a very important point to
recognize. It is very difcult to change the wall design after this point without
considerable difculty. After the wall design is completed, the path from there
to construction is pretty much a straight line with little room to adjust systems
and components without impacting schedule, cost and coordination.
We will not describe or analyze opaque walls in this paper. There are many
systems and materials that can be utilized to avoid the occurrence and subse-
quent problems of condensation in these types of walls. The approach is pretty
much the same as utilized for glazed wall systems; however, it is generally a
much easier process when properly applied. For glazing systems, such as win-
dows or curtain walls, there are fewer alternatives available. By denition, glaz-
ing systems are intended to have large amounts of transparent vision surface
combined with small amounts of opaque framing materials. Remember that
glass is limited in its ability to provide high thermal resistance levels. The ma-
terials essentially establish the basic properties of the window or curtain wall
systems: Metals, glass, and plastics of various types and sealants in small quan-
tities. Most of these are very resistant to moisture vapor ow through the
material. If condensation does occur, the questions will be: How much and
where?
However, there are options, including thermal performance or the degree
of thermal resistance provided by the insulating glass units IGUs and the
framing system. We have identied two distinctly different approaches to con-
trolling condensation in exterior window and curtain wall systems: Active and
passive. For purposes of this discussion, we have developed the following de-
nitions for these approaches that are specic to moisture migration control:
Passive approachA method of controlling moisture migration and re-
sultant condensation in and on glazed wall systems which relies upon
the specic physical properties of the various materials of the system.
Active approachA method of controlling moisture migration and re-
sultant condensation in and on glazed wall systems, which relies upon
input of heat into or onto the glazed wall system or components, in
addition to the specic physical properties of the various materials
within the systems. Such input may include heat applied close to or in
direct contact with the glazed wall system.
As in all building systems, the success of either approach lies in a compre-
hensive analysis of the physical properties for each material, including their
heat ow characteristics and ability to resist moisture migration. Some of the
approaches to resisting moisture condensation within or upon window and
curtain wall systems seem to y in the face of the accepted denitions of
green or sustainable as used in todays construction industry. This is due at
least in part to the simple fact that the most effective and reliable methods of
resisting condensation often rely upon taking heat from the inside of the build-
ing, in some form and transferring it into and through the window or curtain
wall system to avoid a dew point temperature 2.
Certain building uses in certain climates will require what at rst may seem
like an unreasonable approach and an extraordinary amount of analysis and
detail to properly design the exterior wall construction. However, upon
thoughtful consideration, it will be seen that the solutions are attainable
through the use of readily available analysis tools to identify what is required,
followed by the application of reasonable design and construction techniques
and appropriate materials.
Analysis Methods
A glazing systems ability to control condensation must not be confused with

the condensation resistance of the system. Condensation resistance is the ability
of a glazing system to resist the occurrence of condensation under limited cir-
cumstances. This does not mean that it will prevent condensation from occur-
ring at all times. Currently, our industry has more than one rating system for
identifying condensation resistance. The rating systems rely upon a standard-
ized testing method that provides an end result in the form of a number that is
assigned to a particular glazing system. This number is only a representation of
an average of temperatures observed or modeled on the interior surface of the
system being tested. The rating systems may provide a useful tool for compar-
ing relative performance of various systems. However, the rating methods used
do not address the system as installed in a specic building application. The
test methods are standardized to allow for comparisons of various systems.
An inherent problem with these methods is that they are standardized.
Buildings are not designed to standard sized glazing systems, they are not in-
sulated test boxes, and the mullion congurations and interior conditions all
combine to result in a very specic set of uniquely constructed individual as-
semblies. These varying conditions contribute to the project specic needs of a
particular building, and also contribute to the divergence away from the results
determined by the standard tests for condensation resistance.
Proper condensation control analysis includes consideration of the entire
glazing system. It evaluates the project specic components, materials, and
in-service conditions that are part of the exterior glazed wall. This form of
analysis requires the design professional to evaluate the specic components
and combination of components in order to determine if there is a possibility
for condensation to occur. Condensation control analysis does not look at av-
erage temperatures. It is quite the opposite; it is a very detailed look at a glazed
wall system to nd even small possibilities of condensation. Evaluating the
glazing system for locations in which the surface temperature is below the dew
point temperature accomplishes this.
Condensation Control Approaches
Several passive and active approaches are available to prevent and avoid con-
densation on and within exterior glazed wall systems. Analyses presented in
this paper will be limited to consideration of colder climates, similar to those in
Climate Zones 3 and colder, as dened by ASHRAE Standard 90.1 3, and those
interior spaces where the RH is actively applied or generally above 20 %. Gen-
erally speaking, if condensation were to occur on the exterior of a system, it
would likely be located in a warm southern climate, and it would typically not
be detrimental to the exterior envelope as it would drain to the ground. How-
ever, in colder northern climates, if condensation were to occur, it would likely
be on the interior side of the glazing system. This will have detrimental effects
on the buildings durability and functionality and perhaps indoor air quality. In
order to avoid condensation on the interior surfaces of the glazing system,
these surfaces of the system must remain above the interior dew point tempera-
ture due to the relatively low permeability of most glazing systems.
A common misunderstanding when performing analyses of glazed exterior
walls is that the heat from the sun will provide additional warmth needed to
keep the system above the dew point temperature. This is a true statement in
some cases; however, the additional heat is only available for a few hours dur-
ing each day, and it will not be applied evenly or to all exterior elevations of the
building. Analysis of systems and the components should be performed to rep-
resent a worst-case condition, typically nighttime. The coldest temperatures
typically occur at night. This is when radiation is lost to a dark night sky, a
property of heat ow commonly known as black body radiation. The black
body, in this case, is essentially a completely dark night sky. Since it is com-
pletely dark, it reects no light or radiant energy back to the glazed wall. This
insures a worst-case or maximum radiant energy heat loss from the glazed
wall. The rate of radiant energy transfer is essentially proportional to the color
temperature of the receiving body. The night sky acts as an innite sink absorb-
ing any radiant energy that is transmitted in its direction. For the purposes of
this paper, the analyses of the various systems and components assume a night-
time condition.
Our descriptions of passive and active methods of controlling condensation
utilize various analysis tools to show comparisons between the methods avail-
able. The following parametric conditions were established for each analysis:
Static interior and exterior temperatures and RH
Exterior temperature of 18 C 0 F
Interior temperature of 21 C 70 F
Passive Approach
A dependable approach to avoid and prevent condensation is to utilize passive
methods. Passive methods do not rely on supplemental systems to be effective.
Once in place, and when properly maintained, passive methods will provide a
specic level of condensation control, no more and no less. Systems that can be
employed to control condensation are numerous and can range from varying
the constituent materials used in the assembly to being very selective of the
location of the glazing assembly within a wall system in/out to details of the
structural anchorage and transition of the glazing system to the adjacent wall
system.
A common and perhaps obvious passive approach is simple: Provide the
proper materials and components, design the assembly, and install it properly.
The performance attribute with the most effect on condensation control is ther-
mal conductivity. IGUs and the framing system have multiple options for com-
ponents that can be utilized to provide a range of condensation control. The
following are some of the variables that effect performance of IGU systems:
Glass type: Clear or tinted
Glass assembly type: Single pane, double pane, triple pane, vacuum in-
sulating glass, and combinations
Glass coatings: Low E, reective, surface location, and multiple coatings
IGU spacer type: Aluminum, stainless steel, rubber, and warm-edge
technologies
IGU gas interspace types: Air, argon, and krypton
Framing material type: Steel, aluminum, wood, plastics, and combina-
tions
Framing type: Non-thermally improved TI, TI, and thermally broken
TB
Glazing method: Wet or dry
Of these variables, the type of IGU utilized has the potential for the largest
impact on the ability of the system to accommodate elevated levels of humidity
mainly because the glass constitutes the majority of the system area. There are
numerous types of IGUs available, each with a specic performance level.
Three basic glass types are considered Fig. 2 to demonstrate the differences in
FIG. 2Three basic glass types.
performance. All three units are composed of clear uncoated glass. Given an
exterior temperature of 18 C 0 F and an interior temperature of 21 C
70 F, the three glass types were analyzed with WINDOW 5.2 4 software to
determine the maximum interior RH that can be tolerated without formation
of condensation. As expected, when the insulating value of the glass unit in-
creases, so does its ability to tolerate higher levels of RH. This is identied by
the glass units indicated in Fig. 2. In the three cases shown, the single pane
glass unit, Fig. 2a, can accommodate 11 % interior RH, while double pane
IGU, Fig. 2b, and triple pane IGU, Fig. 2c, can accommodate 40 % and 55 %
RH, respectively, for the specic conditions indicated. This analysis only pro-
vides the center-of-glass COG temperature. Additional analysis is required to
determine how the frame and edge spacer affect the performance as a whole.
The next step in improving the condensation control of the system would
be to add a coatings to one or more surfaces of the unit. In typical designs, an
IGU often will employee one low E coating to improve the insulation value of
the glazing unit. To demonstrate the improvement that can be achieved in this
way a high quality low E coating is applied to the No. 2 surface of the unit
shown in Fig. 4 refer to Fig. 3 for description of glazing unit surface designa-
tions. Again, as suspected, the unit with the low E coating can tolerate a higher
level of interior RH. Of interest, the double pane IGU with a low E coating
performs almost as well as the uncoated triple pane unit in Fig. 4.
FIG. 3Glazing unit surface designation.

FIG. 4Basic glass compared to glass with low E coating.
If additional performance is required, the interspace of the IGU can be

lled with an inert gas. Argon is most commonly used. The benet that can be
obtained by adding argon gas to the interspace indicated in Fig. 4 is shown by
Fig. 5a5d.
Although the addition of multiple panes of glass, coatings, and inert gases
allows the tolerable interior RH levels to increase, it becomes apparent that as
we continue to add performance enhancing options, we also begin to reach a
point of diminishing returns. Figure 6 provides a summary of the options pre-
viously discussed.
The next component of glazed wall systems that can signicantly reduce
the potential for condensation is the framing system. The framing is often one
of the weak links of glazed exterior walls that compromise performance. In
most cases, when using IGUs, the framing is colder than the IGU. In fact, the
framing can have a negative effect on the IGU, causing the surface of the glass
that is close to the frame to be colder than the COG. This area of the IGU is
identied as the edge-of-glass EOG. In many cases, the temperature difference
between the EOG and COG can be 5.5 C 10 F or more. This effect will
greatly reduce the RH that can be tolerated by a given system. Computer mod-
FIG. 5Comparison of air lled interspace and argon lled interspace.
FIG. 6Performance summary of glazing options.

eling, in combination with specic analysis and design criteria, is required to

determine the maximum RH that can be tolerated for specic framing and
glass combinations.
The type of material used to construct the framing, the interior and exterior
frame depths, and the frame type affect the performance level of the entire
glazed wall. Three methods of modifying the framing system are considered.
The rst method is to change the type of material used for framing. Several
materials can be used, such as aluminum, steel, plastics, and wood, as well as
combinations. Based purely on the thermal conductivity of the materials,
frames constructed from wood will perform the best in limiting the amount of
condensation that will occur. However, wood frames often are not considered
due to the porosity of the material, strength limitations, and service life, in
addition to other considerations. Thermally, the next best materials to be con-
sidered for framing are berglass or plastic often polyvinyl chloride. Both
materials have excellent thermal properties and are not porous; however, they
are also limited by the strength of the material, and other properties.
For purposes of this paper, aluminum and steel are the framing materials
considered. These comprise the bulk of experience in our practice. In large
commercial and institutional construction, aluminum and steel are by far the
most commonly used materials. Both of these materials have much higher
strength and potential service life than wood or plastics. However, when un-
modied, both are less thermally efcient than other materials. When com-
pared, steel will perform better than aluminum in controlling condensation. In
addition to improvement of thermal conductivity lower heat conductivity,
steel frames also have another advantage over aluminum, higher strength. This
can result in smaller frame proles and larger panes of glass for the system. As
indicated previously, the frame is the weak link for most glazed wall systems. In
general, less metal framing results in higher performance relative to condensa-
tion control.
Metal window frames can be produced in three different forms:
Thermally unimproved members
Thermally Improved TI members
Thermally Broken TB members
Frames constructed from thermally unimproved members do not provide
any thermal enhancement. To combat the poor thermal resistance of the metal
frames, TI and TB frames must be used. As indicated below, the National Fen-
estration Rating Council NFRC denes TI, TB, and thermal break in NFRC 100
5.
TI members: System members with a separation of 1.60 mm 0.062 in.
provided by a material where thermal conductivity is 0.5 W / mK
3.6 Btu in. / h ft2 F or open air space between the interior and exte-
rior surfaces. Such systems include members with exposed interior or exte-
rior trim attached with clips and all skip/debridged systems.TB members:
System members with a minimum of 5.30 mm 0.210 in. separation pro-
vided by a low conductance material where thermal conductivity is
0.5 W / mK 3.6 Btu in. / h ft2 F or open air space between the inte-
rior and exterior surfaces. Examples of such systems include, but are not
FIG. 7Comparison of glazing frame types.
limited to, pour and debridged urethane systems, crimped-in-place plastic

isolator systems and pressure glazed systems with intermittent fastener-
s.Thermal break: A material of low thermal conductivity that is inserted
between members of high conductivity in order to reduce the heat transfer.
Thermal barrier material conductivity shall not be more than 0.52 W/mK
3.60 Btu in. / h ft2 F
As indicated by these denitions, the differences between the two types of
thermally enhanced frames are the thickness and the thermal conductivity of
the material used to improve the thermal resistance of the metal framing. Glaz-
ing systems constructed from TB members will provide better control of con-
densation than those that are constructed from TI members. To illustrate this,
three THERM 5.2 4 computer models are presented: one with a typical
thermally unimproved member, one with a TI member, and the last with a TB
member Fig. 7. All other aspects of the models are the same, including the
glass type, frame size, and interior/exterior temperature conditions. Similar to
the glass analysis, an exterior temperature of 18 C 0 F and an interior
temperature of 21 C 70 F were used for these computer models. For the
purposes of this paper, the models represent generic types of framing systems
that are not specic to any manufacturer and basic two pane IGUs.
The coldest point on the interior surface of framing for all three glazed wall
systems is near the interface of the IGU with the framing. As expected, the wall
system utilizing the thermally unimproved members performed the worst. The
framing has a cold point of 1 C 31 F, which would be able to tolerate a
maximum interior RH of 23 %. The glazed wall system with the TI members
has a cold point of 2 C 36 F, which would be able to tolerate a maximum
RH level of 28 %. Providing the best control against condensation and hav-
ing the warmest surfaces, the system with the TB members has a cold point of
6 C 43 F, which would be able to tolerate a maximum RH of 37 %. This
analysis shows that if higher levels of RH are a requirement for a specic space,
then TB frames will likely be required to control condensation.
FIG. 8Comparison of glazing frame depth.
The framing components considered here are all modeled at the midpoint
of the glazing framing. In many situations, this condition is the best-case sce-
nario. Often, the surfaces temperatures at intersections of vertical and horizon-
tal framing members can be at least 3 C lower, which will also reduce the level
of RH that can be tolerated. Currently, none of the industry standard computer
modeling software programs has the capabilities to simulate three-dimensional
conditions such as these intersections.
Another modication that can be implemented to the framing to inuence
the condensation control of the system is the depth of the framing members.
When using metal framing members, the amount of metal that is exposed to
the interior and exterior of the building will have an effect on the surface tem-
perature of the glazing system. The metal components of the framing essen-
tially function like radiators that conduct or transfer heat from the warmer side
to the colder side. Given this function of the framing, it follows that having a
large amount of metal on the interior side of the wall system relative to the
exterior side will accomplish an enhanced level of condensation control.
Figure 8 illustrates results of computer models of two glazed exterior wall
systems using the same IGU and TB frames. The only difference between the
two models is that the depth of the metal on the interior side of the system has
been signicantly increased. By limiting the amount of metal exposed to the
exterior and maximizing the amount exposed to the interior, the metal frame
can provide a benecial inuence on the entire system. The additional framing
transfers the heat from the interior to the IGU, which can provide warmer
surfaces that can then tolerate higher interior RH. The shorter framing has a
cold point of 3 C 38 F, which would be able to tolerate a maximum RH
level of 30 %. The deeper system has a cold point of 8 C 47 F, which
would be able to tolerate a maximum interior RH of 45 %.
Another passive modication to the glazed wall components is the IGU
edge spacer. This edge spacer component is used to keep the individual panes of
the IGU separated, thereby creating an air cavity. The air cavity provides the
insulating property of the standard IGU and must be sealed air and vapor tight
FIG. 9Comparison of glazing edge spacers.
to keep the IGU performing properly. The IGU is warmest at the COG and
coldest at the EOG. As previously discussed, this condition is partially caused
by the framing, but it is also partially caused by the type of spacer used. Essen-
tially, the spacer creates a thermal bridge between the two panes of glass, pro-
viding a direct path for heat loss to the cold exterior and thereby reducing
surface temperatures of interior components.
There are multiple types of spacers available. A common type of spacer
used in the commercial industry is fabricated from aluminum. These spacers
function as a thermal bridge at the EOG, which contributes to potential con-
densation. However, there are many new technologies that have been developed
to reduce the thermal bridging. As a group, these are generally referred to as
warm-edge spacers. This category of spacers is comprised of various types of
materials such as stainless steel, foam rubber, and even spacers that incorpo-
rate poured and debridged thermal break methods. Other types of warm-edge
spacers are also available. The basic goal they all have is to reduce thermal
transfer at the EOG.
Often use of warm-edge spacers will provide an increase in surface tem-
perature at the EOG of at least a few degrees. When designing and analyzing
glazing systems that need to tolerate high levels of humidity, the use of this
technology is sometimes required to reduce the potential for condensation.
Figure 9 indicates the computer model output for two glazed wall systems with
the same IGU and the same TB frames. The only difference is the spacers. One
is modeled with a typical aluminum spacer, and the other with a warm-edge
spacer. A noticeable increase in surface temperature is indicated for the warm-
edge spacer. This permits the glazed wall system to tolerate higher RH. The
system with the standard spacer has a cold point of 1 C 34 F, which would
be able to tolerate a maximum interior humidity level of 26 %. The one with
the warm-edge spacer has a cold point of 4 C 40 F, which would be able to
tolerate a maximum RH of 33 %, a signicant improvement.
FIG. 10Comparison of basic versus enhanced systems.
Looking at the differences in performance of these glazed wall system com-

ponents individually can sometimes be misleading. To fully understand the
benets provided or limitations imposed, they need to be modeled and ana-
lyzed as whole systems. Each component will inuence and have an effect on
the others. They must be viewed as specic combinations to see the full poten-
tial of a system and to nd the weak point of the entire system. The surface
temperature at the weak point will help determine the maximum level of inte-
rior RH the system can tolerate without the formation of condensation. In Fig.
10, two more computer models indicate the considerable range of performance
that can be attained. The rst example indicates the use of thermally unim-
proved materials that do not have substantial performance, and the second
includes materials and components with higher thermal performance proper-
ties. The basic system has a cold point of 3 C 27 F, which would be able
to tolerate a maximum interior RH of 20 %. The enhanced system has a cold
point of 8 C 46 F, which would be able to tolerate a maximum interior RH
of 42 %. Again a signicant difference.
Additional Passive Approaches to Control Condensation

So far, we have discussed various ways to improve the condensation level of
glazed wall systems through the selection of system components. However,
there are also some additional passive approaches that can be implemented,
which are associated with the relationship between the glazed wall system and
the opaque wall systems. Often, the majority of a particular glazing system can
handle a high RH; however, the perimeter conditions may have been designed
in such a way that they are detrimental and can cause condensation to occur at
the sill, jamb, and head conditions. The next three passive approaches describe
the relationship between the glazing system and adjacent exterior wall con-
struction.
FIG. 11Comparison of system placement.
A common oversight in designing the exterior envelope is the position of

the glazing system within the wall assembly. The poorest performing location
to place a glazing system is on the outboard side of the adjacent wall assembly
insulation. In this position, the warm side of the glazing systems is often ex-
posed to cold temperatures, and its thermal resistance properties will likely be
negated. An ideal location is to position the system by aligning the IGU with the
insulation layer of the adjacent wall assembly.
Figure 11 illustrates the difference in surface temperatures that can be
expected when placing the same glazed wall system in two different locations
relative to the adjacent wall assembly. As indicated by the computer models, the
glazing system that is installed on the outside of the wall assembly insulation
has a cold point surface temperature of 4 C 39 F, which would only be able
to tolerate a maximum interior RH of 32 %. The other has a cold point
temperature of 7 C 44 F, which would be able to tolerate a maximum inte-
rior RH of 39 %.
Another common condition that can cause condensation problems with
glazing systems is the lintel at the head of the glazing system often used in
typical masonry cavity wall assemblies. The lintel is commonly constructed
from steel and typically is used to support the masonry above the glazing sys-
tem. This condition is not necessarily a direct glazing system issue; however, it
can become one depending on the particular application. In some cases, the
steel lintel is also used to secure the glazing system. In most typical details the
FIG. 12Comparison of lintel types.
steel lintel is a continuous piece from the interior to the exterior that essentially
creates a thermal bridge, again negating the purpose and capabilities of using
TI or TB glazing systems.
To eliminate or reduce the negative effect of steel lintels on glazed wall
systems, special attention is required for the lintel design. The goal is to ther-
mally break the lintel, and there are various methods that can be utilized to
provide a thermal break in the lintel assembly. One approach spaces the steel
lintel from the face of the backup wall. In this approach, the lintel is only
attached periodically with clips, allowing the full thickness of cavity wall insu-
lation to be installed between the lintel and the backup wall. Although there are
still thermal bridges at the clip locations, the overall effect of this method of
attachment will greatly reduce thermal transfer and thus the potential for con-
densation.
Another method to thermally break the lintel is similar to the rst one.
Again, the rst step is to disconnect the lintel from the backup structure. Plac-
ing a plastic or berglass reinforced plastic shim between the steel lintel and
the backup structure at connection points will provide a thermal break to help
reduce the risk of condensation. The plastic material could be continuous or a
shim with insulation installed between the shim locations.
Figure 12 illustrates the difference in surface temperature that can be ex-
pected when the same glazed wall system is used with two different lintel con-
ditions. As indicated by the computer models, the glazing system with the typi-
cal lintel has a cold point surface temperature of 3 C 37 F, which would only
be able to tolerate a maximum interior RH of 30 %. The other, installed with
a TB lintel, has a cold point temperature of 6 C 43 F, which would be able to
FIG. 13Comparison of glazed wall with steel frame added.
tolerate a maximum interior RH of 38 %. Depending on the wall assembly

construction, this type of thermal bridging can also occur at the sill and jamb
conditions. Similar solutions can be applied to remedy other problematic con-
ditions.
The last passive method that will be discussed is not a method that is
commonly utilized. For some high RH interior environments, using the best
performing window system, optimal placement of the system within the wall,
and reducing perimeter thermal bridging still may not provide a construction
that can sufciently reduce the potential for condensation. For many systems
the potential for condensation exists at the perimeter condition, not at the eld
of the glazing system, as previously discussed. The addition of deeper glazing
framing can reduce the potential for condensation for a given glazing system. A
similar concept can be applied to the glazing system without adding depth to
the framing. The addition of a continuous steel frame attached directly to the
perimeter of the glazing system can help reduce the risk of condensation. The
addition of the steel frame maximizes the amount of mass on the interior side
of the thermal break thereby functioning similar to a radiator absorbing and
transferring heat. Figure 13 provides a comparison between two similar glazing
systems. One, however, has additional steel applied to the backside of the glaz-
ing system. As shown, the system with the additional steel has higher surface
temperatures and will be able to tolerate higher levels of RH.
Summary of Passive Approaches

Several methods to improve condensation control in glazed wall systems are
available. These passive approaches are dependable because they are built into
the glazing system by design. They rely solely upon their physical properties
and not supplemental sources of heat to reduce the risk of condensation. These
methods do not use independently applied energy to keep surfaces warm. It
becomes apparent that as we continue to add performance to the glass and
framing materials, position the systems in the optimal locations, and modify
the adjacent surfaces, we will reach a point of diminishing return for perfor-
mance. The level of RH that is to be tolerated will be a key determinant in
deciding which and how many passive approaches are necessary. However, a
wide variety of interior humidity levels can be tolerated by proper selection and
application of these passive approaches.
Active Approach
The previous section described several methods to limit condensation potential
using the physical properties of the various materials within and adjacent to
the glazed wall system. In some situations, the passive approach may not pre-
vent condensation from occurring. In those cases, additional measures must be
taken to tolerate a higher RH. These situations require addition of an active
system that relies upon heat input from an independent source into the glazing
system. An active system may include application of heat close to or in direct
contact with the glazed wall system.
The following situations often require applications of active systems to con-
trol condensation.
Glazed wall systems that are offset from the primary plane of the exte-
rior wall. In these cases, the warm air from the room may not be able to
reach the surface of the system to adequately elevate the glazing and
frame temperatures sufciently to prevent condensation.
Existing glazing systems that were never intended to tolerate a high RH,
such as historic buildings with new uses. This may be the situation
when retrotting existing buildings to accommodate a new occupancy
or function. However, this can also arise due to poor initial system se-
lection or design, or improper operations during construction.
Buildings that lack a perimeter heat source to provide more proximate
heating of glazed wall systems to help prevent condensation.
Inaccurate or non-existent mechanical system controls, which result in
uctuations in the interior RH beyond the ability of a specic glazing
system to avoid condensation.
Presence of window treatments, such as roller shades, which partially
insulate the window from the heat of the room, thus lowering the sur-
face temperature of the glazed wall system.
These are just some of the common scenarios that are found to be prob-
lematic for typical glazed wall systems. When passive systems cannot ad-
equately prevent condensation in conditions such as these, there are active
control approaches that can be used to help correct a problem. Active ap-
proaches are, however, inherently more complex and by denition include the
application of energy through the use of electrical or mechanical devices and
varying levels of control. Even though an active approach may provide excellent
results, the use of these devices and systems can substantially elevate the degree
of difculty in analyzing and predicting the surface temperatures of the glazing
system.
The following discusses various active methods and their application to a
variety of conditions and provides insight to the analysis techniques that can be
used to accomplish a functional design.
Convective and Radiant Supplemental Heating

The key for many of the active approaches is to provide a supplemental source
of heat applied directly to, or on, the glazed wall system, thereby keeping sur-
face temperatures above the dew point temperature and preventing condensa-
tion. This can take many forms, and depending on the specic approach, the
analysis tools available, the experience and the ability of the evaluator, the
results can be difcult to accurately predict. Given this, many of the active
system approaches are typically designed to be conservative.
One of the most common and basic methods to actively heat a glazed wall
system is to provide a heat source above or below the glazing system. In some
cases, a combination of both may be used. This is not an innovative idea. Good
practice indicates and a good look at many existing buildings reveals radiators
below nearly every window. The main reason that radiators were located in that
position in the past was the poor thermal properties of windows. This is the
same approach that we are suggesting. Radiators helped eliminate cold down
drafts and put heat at the point where it was most needed. As the technology of
modern glazing systems improved, the use of heat sources at windows began to
disappear. This may be acceptable design for standard, low RH spaces, but it
may not be sufcient for spaces with higher RH. Radiators that provide heat
through convection and panels supplying heat through direct radiation are
types of heat sources that can be benecial to use at glazed exterior walls. These
heating sources can use electricity, hot water, or steam as the primary heating
energy with each source having its own characteristics and engineering and
design requirements. The goal is the same for each, to apply heat to the glazed
wall system Fig. 14.
There are several difculties in determining the amount of heat needed.
When below window radiators or convectors are used, we rely primarily on the
buoyancy of the heated air to rise up past the glazed wall system and to elevate
the wall surface temperatures adequately. When these heating devices are sized
to fully offset the calculated heat loss of the glazing system and its surrounds, it
would seem to be a reasonable assumption that typical heat transfer calcula-
tions will fairly approximate the resulting surface temperatures. Unfortunately,
there are many variables that can cause these calculations and assumptions to
be invalid. Issues such as the height of the window, depth of window setbacks,
thick wall surrounds or adjacent column enclosures, the use of shades and
blinds, and methods of window frame attachment can all have major impacts
on the resulting glass and frame temperatures.
Application of Warm Air Supply

A traditional and generally effective approach to heating the inside surfaces of
glazed wall systems is to blow heated air directly on the window surfaces either
FIG. 14Heat applied indirectly to glass and framing.
from above or below. This approach would seem to be far more reliable than
stationary radiators or xed radiant panels, knowing that the heated air will be
more reliable in impacting the window surfaces. Unfortunately, this too pre-
sents its own set of difculties.
In many cases the building architecture prevents us from locating air sup-
plies in the most advantageous and reliable locations. Deep window surrounds
and unique ceiling designs with complex soft constructions may present the
most challenges. In such cases, it is very difcult to accurately predict the
outcome and resultant surface temperatures without advanced analysis. Com-
putational uid dynamics CFD analysis, in combination with other thermal
analysis tools, may be the best approach to most accurately predict the out-
come and result in a successful design. However the difculty of this analysis is
most often prohibitive.
Warm air supplies must also provide dry air, not just warm air. If this air
supply is highly humidied, it may elevate the RH in the vicinity of the window
surfaces, elevating the local dew point, which in turn, can result in condensa-
tion on glass and frame surfaces. This is why certain old rules of thumb, such
as providing a minimum of 15 m 50 ft per min of air velocity across window
surfaces will prevent condensation, are not valid. In reality, the humidity of the
air, the physical geometry of the window and architecture, and numerous other
conditions make it impossible to provide such simple guides to avoiding con-
densation.
Window Frame Heat Tracing

Another method that can be successful, without requiring the amount of space
and disruption that radiators, duct work, and radiant panels require, is heat
tracing. Heat tracing can be provided in two different forms, electrical and hot
water. Both systems are typically installed on or within the wall system fram-
ing. Both systems also function primarily through simple heat conduction. The
need for more complex heat transfer functions involved with radiation and
FIG. 15Heat applied directly to framing.
conduction are minimal. The idea is to heat the water or wire, have it placed
directly against the metal frame, which will conduct the heat into the framing
system and raise the surface temperature above the dew point temperature.
The elevated frame temperature, in turn, heats the EOG and, to some extent,
can even elevate temperatures at other regions of the glass. These active sys-
tems can be very useful for counteracting the negative effect of window shades
or non-TB window frames. In fact, in some conditions, the window shade can
help trap the heat generated by the heat tracing system between the glass and
the shade itself.
The analysis of the resultant window frame surface temperatures is not a
straight forward application within industry standard heat transfer analysis
programs such as THERM. As described in the included case study, tech-
niques can be used that allow heat trace applications to be modeled using these
programs. However, this type of unique analysis must be approached with care
and conservatism.
When evaluating whether to use either electric or hot water heat tracing,
there are risks and benets to both systems that must be considered.
ElectricalFig. 15 For electrical systems in contact with metal frames

exposed to the public, the risk of electrical shock must be considered. While
this type of installation can be protected through the use of listed electrical
components and providing ground-fault current interrupters for each circuit,
some jurisdictions are hesitant to allow electric heat tracing of glazing systems.
Others will allow this approach only if the electrical heat trace cables are Un-
derwriters Laboratories UL listed specically for window applications. Unfor-
tunately, no such cable exists as of the publication of this paper. Electric heat
trace cables are available with a variety of approvals by various organizations
such as UL and Factory Mutual but not specically for installation in, or on, a
window frame. This application can, however, provide very efcient solutions
to complex condensation issues, and manufacturers need to provide additional
FIG. 16Natatorium in Michigan.
testing and labeling for use on glazing systems.

The life expectancy of electrical heat trace is similar to that of IGUs, prob-
ably in the range of 20 years if installed carefully and properly operated. But
just as glazed wall systems must be designed to allow glass replacement, heat
trace applications for windows must also be designed to allow cost effective
replacements. An electrical heat tracing system can also be easily controlled.
The heating cables can be turned on at a specic exterior temperature below
which condensation could likely be expected to start forming. In addition to the
switching capability, this method can also provide a self-regulating system.
That is to say that the heat output of the heat trace cable will automatically vary
depending on the need. As it gets colder outside and the metal adjacent to the
heating cables becomes cooler, the physical make-up of the electrical cable
automatically provides higher output, helping to ensure that sufcient tem-
peratures are achieved without overheating or excessive energy use.
Hot WaterIn most applications of hot water heated glazed wall systems,
the frame extrusions are too complex to easily accommodate hot water piping
even when exible tubing is considered. More often, such applications provide
heat directly to the larger supporting members or larger primary frame ele-
ments. Figure 16 shows just such an application where the skylight supporting
frame system incorporates extruded metal panels with channels to accommo-
date hot water circulation. Water temperatures in the metal panels can vary
between 50 and 70 C 120 160 F. The glass and frames of the skylight are
then heated through the combined actions of conduction, convection, and ra-
diant heating. This skylight, above an 80 % RH natatorium, located in the cold
climate of Michigan, does not experience condensation and the frames and
glass remain dry throughout the Winter.
The hot water system approach can be easily controlled to turn on and off
at a specic exterior air temperature and also can be used as a primary heating
element for the space or room. Heat output can be closely matched to the need.
The analysis of this approach can be very difcult, even more so than for elec-
tric heating elements. This is because in most cases, a more substantial amount
of the heating energy is transferred by radiant heat exchange. The analysis
must incorporate specic applications of the heat exchange program as well as
the use of radiant transfer surfaces. This type of application must be ap-
proached with a high degree of conservatism.

The principal physical risk of a water heat trace system is that there is
potential for the system to develop a leak. If a leak does occur, the repairs can
be costly. More importantly, if used in critical facilities such as museums, water
leakage could cause substantial damage to nearby collection items. Healthcare
is another facility type that may not be an appropriate use for this application
due to the risk of failure and resultant disruption to occupancy. There is also a
concern of freezing pipes in a water system. If a loss of power occurs or a
mechanical device fails, there is the potential for the water piping or channels
to freeze and burst. These systems must be designed with redundancy and
protection against such failures. Emergency power generators serving the as-
sociated mechanical equipment, redundant circulation pumps, and stand-by
boilers or heat exchangers providing heating hot water are methods that can
provide the necessary redundant systems. The addition of piping system tem-
perature sensors for early warning of temperature drops may also be prudent.
A Case Study for Electrical Heat TracingThe following process was fol-
lowed in developing an electrical heat tracing system as a solution to a building
glazing system that experienced a condensation problem. The process included
dening the problem, determining the environmental conditions, discovering
additional considerations, developing options, and evaluating the options. The
nal results are also presented.
The Problem An existing library in a mid-south state, used to house a collec-

tion of historical archives, was experiencing an excessive amount of condensa-
tion on the interior surfaces of the curtain wall system during the colder Winter
months. The facility was designed and built in the mid-1990s and almost im-
mediately began experiencing excessive and uncontrolled condensation on the
curtain wall system. Due to the type of collections housed by the building, the
interior environment was required to be maintained at elevated levels of RH
throughout the year.
Conditions The Winter time interior RH was typically controlled to 50 %,

and the interior temperature was held at 21 C 70 F. At these interior con-
ditions, the dew point temperature is 10 C 50 F.
Considerations In addition to the elevated levels of RH, a few other condi-

tions contributed to the condensation that was experienced in this facility.
ASHRAE 3 and NOAA 6 both indicate the exterior Winter design
temperature to be 10 C 14 F for this geographic location. As the
exterior air temperature decreases, the potential for condensation on
the interior surfaces begins to increase.
Due to the nature of the collection stored in the facility, interior blinds
were installed and were often closed to limit exposure of the collection
to natural light. With the blinds closed, the heat of the room was
blocked from warming the curtain wall system above the dew point
temperature.
The curtain wall was constructed from a TI framing system. In interior
FIG. 17Summary of condensation control options.
environments with elevated interior RH and low exterior temperatures,

TI framing systems generally will not be able to resist condensation
without the use of some type of supplemental heat source. At a mini-
mum, TB frames are normally required to assist in preventing the con-
densation from occurring. Note that even though higher performing cur-
tain wall systems were available when this facility was designed and
built, such a system may still not have been able to resist the condensa-
tion without the addition of an active method. Often, when a high per-
formance system is used in environments with elevated RH, the addition
of window treatments such as blinds renders the system less effective.
Development of Options Options considered to stop or control the condensa-

tion include the following:
Option 1: Complete removal and replacement of the existing curtain
wall systems with a new higher performing curtain wall system.
Option 2: Retrot of the existing curtain wall system to improve its ther-
mal performance through the addition of a thermal break and new
higher performing IGUs.
Option 3: Addition of supplemental HVAC systems in close proximity to
the curtain wall system.
Option 4: Application of a heat tracing system installed directly onto the
existing curtain wall system.
A preliminary review of needs was performed to identify the criteria to be
used in determining the most viable options for remediation. This criterion was
then matched against the probable outcome of implementing each of the pos-
sible options. A summary of this evaluation is provided in Fig. 17. Upon nal
evaluation, it was determined to pursue the heat trace application. This re-
quired additional analysis and development to validate the feasibility of this
preferred approach.
Analysis and Development of Electrical Heat Trace SystemIn order to verify

that heat tracing could be successfully implemented, computer modeling was
performed to evaluate various conditions. THERM 5.2 was used extensively
throughout the evaluation to produce the various computer models. Due to
FIG. 18CRF/THERM model temperature comparison.
inherent limitations of the modeling software, this evaluation was approached

in a conservative manner.
The rst step of the evaluation process was to model the typical compo-
nents: Head, jamb, sill, and intermediate vertical mullions. In order to conrm
the modeling process, the components were then combined in a model that
replicates an existing CRF test AAMA 1503 7 that was performed on the
system. The same glass type and interior and exterior temperatures were used
in the computer models as those that were used in the actual CRF test. As
indicated in Fig. 18, the temperatures of the modeled system all were within
0.1 3.1 C 0.1 5.7 F of the temperatures recorded in the CRF test. The
correlation between the two sets of temperatures allowed for a factor of safety
to be determined for this specic application.
After determining the degree of accuracy for the modeling process, the
project specic horizontal and vertical sections were modeled. The interior and
exterior design temperatures were assigned to the models to conrm that the
existing condition would be expected to experience condensation even with the
blinds open. Figure 19 shows a THERM model of the typical jamb detail
without the blinds simulated, and as indicated, the surface temperatures of the
glass and frame are below the dew point temperature. Blinds were then added
to the computer model in order to determine the resulting temperatures. Figure
20 shows the same jamb detail but with a blind added. As anticipated, some of
the surface temperatures dropped, further increasing the potential for
condensation.
The nal step in the modeling process was to determine if heat tracing
would be an effective method to reduce condensation. This required additional
computer modeling, small desk-top mockups, and full-scale on-site mockups.
The rst step of this process was to apply a heat source to the existing condition
FIG. 19Existing curtain wall system without blinds.
computer models that would accurately replicate a heat trace cable. Unfortu-
nately, the computer modeling software used does not include an option de-
signed specically to accomplish this task. Replication of the heat trace cable
was modeled through careful and considered use of the options available
within the software. THERM 5.2 includes constant heat ux boundaries that
can be applied to the surfaces of the modeled components. The units used for
this type of boundary condition are expressed in BTU/ h ft2. When applied to a
material surface, this type of boundary will result in a hotspot on the model.
In order to replicate the effect of the heat trace cable, the published heat output
of the cable was converted to units that are compatible with the constant heat
ux boundary condition and applied to the models. The new computer models,
with the heat trace cables replicated, indicated that the surface temperatures
would be above the dew point temperature and should greatly reduce the risk
of condensation Fig. 21.
Desk-top mockups were then created to verify that the heat trace cable
would actually perform as predicted by the modeling. The desk-top mockup
was composed of a 12 12 in2 corner intersection of the curtain wall system
FIG. 20Existing curtain wall system with blinds.

FIG. 21Existing curtain wall system with blinds and heat trace.
with an IGU very similar to those used on the project. The heat trace cable and
attachment/concealment system that was planned was also installed on the
mockup. During the desk-top mockups, minor adjustments of the mechanical
attachment method used to secure the heat trace cable were required in order
to ensure sufcient contact of the cable to the curtain wall framing surface Fig.
22. After rening the attachment method, it appeared that the heat trace cable
would be an effective method of raising the curtain wall system surface tem-
peratures. One of the keys to a successful installation was that the cable must
have substantial contact with the framing system in order to adequately con-
duct the heat generated by the cable. If insufcient contact area is achieved,
then the effectiveness of the cable is dramatically reduced.
After reviewing the successful computer modeling and preliminary desk-
top mockups, it was determined that a full-scale on-site mockup would be in-
FIG. 22Heat trace system plan detail.

FIG. 23Installed heat trace.
stalled on the curtain wall system within the library to conrm that the heat
tracing would effectively eliminate the condensation Figs. 23 and 24. Note
that the heat trace application provides for relatively simple replacement if it
should become necessary. The on-site mockup was installed in the winter
months which allowed for results to be observed and reviewed immediately.
Interior temperatures, exterior temperatures, and surface temperatures of the
curtain wall system were recorded with the heat trace system turned on. The
information was then compared to computer modeling for validation of results.
The physical temperatures recorded and the temperatures that were obtained
from the models varied within 0.5 2.2 C 1 4 F Fig. 25. This installation
demonstrated the validity of the proposed corrections and provided a good
correlation to the THERM analysis.
The on-site mockup was left in place for 1 year to verify that it would
function as predicted and eliminate the risk of condensation. During the trial
period, the mockup area did not experience condensation. Based upon this, the
heat trace system was installed on the remainder of the building curtain wall.
The heat trace system was connected to the buildings direct digital control
system, which allowed for the system to be automatically turned on and off at
predetermined outside air temperatures. Additionally, the heat trace cable uti-
FIG. 24Heat trace cover installed.

FIG. 25Comparison of actual and modeled temperatures.

lized is of the self-regulating type. The amount of current the cable draws in-
creases as the surface temperatures of the materials it is in contact with de-
creases. This type of cable conserves energy, avoids damage to adjacent
materials and the cable itself from overheating, and provides efcient opera-
tion. As the outside air temperature drops, the temperature of the curtain wall
system drops and as a result the heat trace cable will draw more power to
compensate for the demand.
In this case, electrical heat trace was chosen as the method for controlling
condensation due to
Cost effectiveness
Minimal disruption
Ease of installation
Minimum amount of waste original windows retained
The understanding of a particular facilitys interior and exterior environ-
ment is critical to the success of any fenestration system. The installation of the
electric heat trace system proved to be successful in providing a means of con-
trolling the condensation that was being experienced on the facilities curtain
wall system. The system has been in operation for over 5 years without any
reports of condensation.
This type of installation has since been utilized in other projects, both new
and existing. Although, further development and study have been completed as
we utilize the system, there are still additional areas that can be pursued to
better understand the opportunities and advantages the system presents:
Redundancy through the use of multiple cables
The number of cables required to warm various types of materials, and
maintain a proper temperature range
System controls for use with multiple cable
Monitoring of the electric usage required to run the system
Identication of cost impact for new and retrot applications
Heated Glass Applications

Another method of applying heat to the glazed wall system that has proven
successful is the use of heated glass. Heated glass is a relatively new technology
FIG. 26Heated glass IGU.
but is emerging as a potentially excellent approach to solving certain complex

condensation issues. While still a very new application for commercial con-
struction, there are now a sufcient number of installations to demonstrate a
history of reliable service. The development and use of the technology that we
are most familiar with evolved from the freezer industry, where the heated glass
was installed at the small windows in freezer doors to keep them from condens-
ing and frosting. Currently, there are only a few manufacturers producing this
type of glass for architectural purposes, and at this time, the ones that provide
UL listed products with architectural applications are even more limited.
Heated glass utilizes an electrical current applied to a coating located on
the #2 or #3 surface of an IGU. To our knowledge, the coatings are limited to
the hard-coat process at this time. The ow of the imposed electrical current
through the coating generates heat as a result of resistance and elevates the
temperature of the glass. The heated IGU can be provided with the heated glass
pane either inboard to heat the room or outboard to heat an exterior element
such as an existing window surface Figs. 26 and 27.
Heated glass can directly prevent condensation from occurring on its own
surface by applying sufcient heat for the glass to remain above the dew point.
It can also be heated further to supplement the heating requirements for the
indoor space if desired.
At issue, however, is its ability to also heat the glazed wall framing materi-
als. Due to the low conductivity of the glass and glazing gaskets, the heated
glass does not substantially conduct heat to the frame members. Although,
some amount of heat is transferred to the frame by simple convection. In our
preliminary trials, this did not sufciently elevate the frame temperatures,
which can present a signicant drawback if the heated IGU is intended to be
the only protection for the framing system. A TB framing system can be utilized
to provide good condensation resistance; however, TB framing alone may not
be fully condensation resistant in a high RH environment. When the heated
glass is used with interior shades, the shading devices help trap the heat and
provide improved heating of the framing.
FIG. 27Heat direct applied to glass.
At this time we are not aware of any calculation methods or analysis pro-
grams that can predict the expected frame temperatures.
Heated glass can help solve issues with lower performing windows, such as
those found in historic buildings. When historic windows need to be exposed to
humidied environments, heated glass can be installed as an inboard storm
window. The heated pane is used to directly heat the historic window main-
taining its surface temperatures above the dew point. The heated glass interior
storm may not need to be tightly sealed, and the much warmer temperatures of
the historic window can withstand elevated RH. In this application, the heated
glass also directly heats the outboard historic window and frame as well as its
own frame, providing condensation resistance Fig. 28.
While this type of application may change the appearance of the existing
FIG. 28Historic window with heated glass storm window.

historic windows to some extent due to the addition of the interior window
frame, the exterior appearance can be reasonably maintained.
This approach can be used in new construction as well to accommodate
very high RH environments and conditions that require interior shading de-
vices.
Seasonal Control and Facility Operations Applications

This approach is different in that it relies upon the control of the interior envi-
ronment rather than the behavior of the glazing system. Depending on the
specic RH and climate control needs of a facility, it may be possible to reduce
the inside RH during Winter overall or during very cold outside conditions.
This would lower the dew point of air adjacent to cold glazed wall surfaces,
reducing the probability of condensation during the coldest times of the year.
To accomplish this, the building HVAC design and controls must be capable of
utilizing increased amounts of outside air which during very cold outside con-
ditions is also very dry to reduce the building RH. Alternatively, this can also
be accomplished by increasing the cooling provided by the mechanical equip-
ment to provide increased dehumidication. This approach can be energy in-
tensive, and most mechanical systems are not designed to provide this degree
of dehumidication. As a third option, desiccant heat wheel dehumidication
systems can also be used to provide RH control. A comprehensive discussion of
these systems is beyond the scope of this paper.
This approach, providing a seasonal reduction in the building RH seasonal
offset control, or drift can result in relatively quick uctuations in building RH
corresponding to uctuations in outside temperature and humidity uctua-
tions. During a very cold day, low humidity may be achieved; however, if a
Winter time warm-up occurs, such as a 4 C 40 F day during December, the
inside RH could quickly elevate to near 50 %.
This type of short-term uctuation is not likely to result in glazed wall
system condensation because during the warmer weather and elevated inside
humidity, the glass and frame also experience proportionately warmer surface
temperatures. Unfortunately, certain facility types, such as museums, may not
be able to tolerate short-term uctuations in space RH.
Another control option that can sometimes relieve a condensation issue is
to simply open shades and blinds at night. For window and glazing systems
that have sufcient condensation resistance if not for the added impact from
interior shading devices, these devices can sometimes be opened during the
coldest conditions, alleviating some, or even most, of the condensation issue.
Since the lowest outside temperatures occur during the nighttime, and the
shades and blinds are primarily intended for solar control, this approach can
often be utilized. Unfortunately, for large facilities with many windows, this can
be operationally difcult as well as adding operational costs. If the windows do
not have adequate daytime condensation resistance when shades are up, this
approach will not be helpful, except to possibly reduce the amount and time of
condensation.
Regardless of the space RH to be provided or the types and positions of
heating devices, adequate attention must be paid to the accurate control of
HVAC systems. It is relatively easy for HVAC systems having active humidica-
tion to over humidify the spaces served. Likewise, perimeter heating systems
may not be controlled accurately or be adjusted properly to provide the best
opportunity for the window systems to remain condensation free.
Analysis Methodologies and Tools

While there are various assumptions that can be made to help anticipate the
effects of these conditions, a denitive predictive analysis cannot be provided
through routine engineering calculations and possibly not even with the use of
sophisticated analysis programs such as THERM. The THERM program is
an excellent two-dimensional heat transfer analysis program and one of the
best tools available for analyzing the resultant temperatures of complex win-
dow and framing systems. Other programs of similar capabilities are also avail-
able. Unfortunately, this type of analysis cannot incorporate the effects of inte-
rior heat sources and complex air ow geometries. When radiators or warm air
supplies are utilized, we cannot accurately predict the inside air temperature
conditions adjacent to glazing systems in order to input to a THERM analysis.
A more denitive analysis can be provided by the use of CFD analysis in
conjunction with programs such as THERM. This approach utilizes complex
simultaneous heat transfer calculations to simulate the full variety of interre-
lated temperature effects and related physical conditions of the installation.
The temperatures of the heating elements, heating of the air, changes to its
buoyancy, exterior wall and window heat transfer, air ow patterns in the space
and around the window surrounds, and other factors as necessary to accurately
model the space and system response can be modeled in CFD software to more
accurately predict the temperatures in the vicinity of the windows.
In the past, this type of analysis has been unavailable or has been far too
expensive to apply to this type of engineering application. However, while still
somewhat costly, CFD programs are becoming more available to building ar-
chitects and engineers and will hopefully gain more prominence in the analysis
of glazing system temperatures Fig. 29.
The combined use of multiple computerized analysis programs such as
THERM and CFD cannot denitively calculate the inside temperature condi-
tions of architecturally complex installations with various heating sources and
complex framing systems. As new methods of analysis become available to
designers, it should be understood that each will have its own benecial appli-
cations and its own limitations and drawbacks.
CFD analysis could include a detailed model of a glazing system with the
intricacies of the framing systems and wall constructions, as necessary to pro-
vide a highly predictive outcome. To do this would require considerable devel-
opment time and normally an unacceptable cost. CFD computation time could
also easily become excessive. In that regard, a more appropriate application
can use CFD analysis with a simplied and conservative glazed wall system
model, provided by preliminary THERM modeling, and then use the resultant
air temperatures calculated by the CFD model to enhance the accuracy of the
THERM model.
This type of analysis, utilizing multiple analysis tools to help improve the
FIG. 29Computational Fluid Dynamics.
accuracy and predictability of calculations is becoming more common as ar-

chitects and engineers learn more about the complexities of these types of cal-
culations and the corresponding short comings of various analysis tools. It is
also important that the level of detail utilized in each analysis tool be kept
consistent with the programs capabilities, and it is important to not draw false
expectations from these analysis tools simply because more detail was in-
cluded. Often added detail and complexity in these models can actually detract
from the outcome rather than add accuracy.
Through careful and considered use of the options available within thermal
modeling software, there are methods to partially simulate the addition of a
heat source in the model, and possibly avoid more difcult and rigorous analy-
sis techniques. The THERM program can also incorporate the application of
radiant surfaces, or radiation enclosures, located away from the glazing sys-
tem. This type of modeling, while not intended by the program creators, can
provide insight to the resultant temperature effects provided by radiant panels
installed within the sight lines of the glazing system.
These types of unique modeling are experimental and essentially trick the
software programs into doing an analysis it was not intended to perform. While
these techniques can prove very useful and add to our abilities to predict win-
dow and curtain wall surface temperature, this type of unique use of these
programs must be carefully analyzed and used only with caution. These unique
applications should be coupled with a level of experience and validated in situ
data from similar models to improve the reliability of the analysis.
The Future of Analysis

Eventually, with enough applications of sophisticated of modeling techniques
and resultant physical test data, we will be able to improve the reliability of
certain modeling tricks and techniques.
The use of these analysis tools for complex physical conditions cannot be
thought of as an exacting science, at least not at this time. For critical applica-
tions and possibly also for applications of substantial numbers of repetitive
assemblies, approaches that provide physical models of the intended applica-
tion may be warranted. Unfortunately, building physical mockups and perform-
ing real world testing are costly and time consuming and are not considered a
practical approach in our industry at this time.
Summary
The two basic approaches to controlling or eliminating condensation on and
within glazed exterior building walls are dramatically different in approach,
limitations, initial costs, energy demands, appearance considerations, and their
long term ability to perform and be maintained. The following provides a sum-
mary for each consideration.
Passive
Approach Protection from condensation provided by the physical prop-
erties and arrangement of the materials used for the glazed wall system.
Limitations Can accomplish medium performance for vision walls.
Condensation resistance for walls with vision glass Fig. 30 will be lim-
ited by the ability of glass and framing systems to resist condensation.
Appearance considerations Traditional appearance for glazed wall sys-
tems will be attained.
Initial costs Relatively low cost impact for reasonable levels of conden-
sation resistance.
Energy Supplemental use of applied energy is not required outside of
normal space conditioning needs.
Maintenance needs No special maintenance needs to retain initial con-
densation resistance performance.
Active
Approach Protection from condensation provided by a combination of
physical properties and arrangement of the materials used in addition to
input of heat energy applied to the glazed wall system.
Limitations Can accomplish high performance. Condensation resis-
tance for walls with vision glass Fig. 30 can be greatly increased by
introduction of heat energy applied to the wall system.
Appearance considerations Can be coordinated into the architectural
aesthetic with good planning.
Initial costs High cost impact for increased levels of condensation
resistance. Cost impact varies depending upon degree of performance
required and active systems approach selected.
Energy Active use of supplemental energy applied to the system to
attain enhanced performance. Degree of energy usage proportional to
performance levels and environmental requirements Fig. 31.
Maintenance needs Maintenance required for systems applying energy
For purposes of this paper glazed wall systems refers to walls with vision glass which allows
light to pass from the exterior of the building to the interior. This precludes the use of insulating
products or insulation with other materials which would be used as spandrel or non-vision
panels. If spandrel glazing is used without insulation or other means to seal the interior side of
the spandrel glazing from the heat and moisture effects of the building interior environment, they
can be considered as vision panels even if an opacifier coating is applied to the interior surface.
FIG. 30Glass types.
to exterior walls to assure continued performance. Degree of mainte-

nance highly dependent upon active system or systems utilized. Main-
tenance in some form and to some degree will always be required to
ensure continued and adequate performance.
While the tools and methods for analysis and solution necessary to prevent
condensation on exterior glazed wall systems are certainly available today,
there is as previously indicated a gap in the awareness and commitment re-
quired to eliminate these problems from the design and construction industry.
When the designers, constructors, and building owners are committed to deal-
ing with this condition, there are two basic approaches to consider: Passive and
Active. Each approach has its appropriate uses and limitations; however, there
are solutions available.
It is likely that the tools for analysis will become more sophisticated and
readily available and that the materials and new systems for application will
also become more readily available, more reliable, and more familiar. However,
designers, manufacturers, owners, and constructors must become prepared
and committed to understanding and solution of the conditions.
The degree of energy required to counteract the natural tendency of condensation to occur on or
within the glazed wall system is directly proportional to the RH of the interior air, the difference
between interior and exterior temperatures, and the difference between the interior and exterior
RH. With currently available materials, in all cases energy will be required to resist condensation
in very high humidity environments if the temperature difference causes the dew point to occur
in a location where the surface(s) are exposed to the water vapor in the air.
FIG. 31Graphic representationapplied energy to resist condensation.
It currently appears that model codes will mandate additional applications

for elevated interior RH in buildings in the future. It is likely that this particular
building functional requirement is here to stay.
Acknowledgments
Graphics are through the courtesy of Ryan Asava, Jerry Carter, and Zach Rusu
of SmithGroup, Inc.
References
1 Carlson, A. R., Computer Simulation of Wall Condensation Problems, Water in

Exterior Building Walls: Problems and Solutions ASTM STP 1107, T. A. Schwartz,
Ed., ASTM International, West Conshohocken, PA, 1992.
2 American Society of Heating, Refrigerating, and Air-Conditioning Engineers,
ASHRAE Fundamentals, American Society of Heating, Refrigerating, and Air-
Conditioning Engineers ASHRAE, Atlanta, GA, 2009. In summary, Chapter 1 de-
scribes the dew point temperature as the temperature at which the water vapor in
the air becomes saturated and condensation begins.
3 ASHRAE Standard 90.1, 2007, Energy Standard for Buildings Except Low-Rise
Residential Buildings, ASHRAE, Atlanta, GA.
4 THERM 5.2 and WINDOW 5.2. 2006. Thermal modeling software programs de-
veloped by Lawrence Berkeley National Laboratory LBNL, Berkeley, CA; see
http://www.lbl.gov/, http://windows.lbl.gov/software/therm/therm.html, and http://
windows.lbl.gov/software/window/window.html Last accessed November 2009.
5 NFRC 100, 2004, Procedure for Determining Fenestration Product U-Factors,
National Fenestration Rating Council NFRC, Greenbelt, MD.
6 Engineering Weather Data. Version 1.0. Dec. 23, 1999. National Oceanic and At-
mospheric Administration NOAA, Camp Springs, MD.
7 AAMA, 1503.1, 1998, Voluntary Test Method for Thermal Transmittance and Con-
densation Resistance of Windows, Doors and Glazed Window Sections, American
Architectural Manufacturers Association, Schaumberg, IL.
doi:10.1520/JTE102959
Christopher M. Morgan,1 Linda M. McGowan,2 and

Loren D. Flick3
Case Study of Mechanical Control of

Condensation in Exterior Walls
ABSTRACT: An electronics manufacturing facility in Colorado was con-

structed with typical commercial steel framing containing berglass batt in-
sulation with foil-facing as the vapor retarder. A large portion of the facility
required constant positive space pressurization and an indoor relative hu-
midity RH of 48 % RH at 21C 70F. Due to a lack of containment of the
moist interior air, condensation formed in the exterior walls, in the roong, on
the window and door systems, and on the skylights of the facility, particularly
along the north and east elevations of the building, which resulted in signi-
cant damage to the wall components and the inability of the HVAC system to
remain properly balanced. The as-constructed exterior wall systems were
evaluated for performance relative to the existing building mechanical sys-
tem. It was determined that the as-designed and as-constructed exterior wall
assemblies and components could not perform satisfactorily under the nec-
essary indoor environmental requirements. Options were explored to im-
prove the performance of the exterior wall assemblies and related details.
Due to a number of constraints, it was decided that constructing new walls
inside the as-constructed exterior walls and windows, thereby creating an
interstitial space to separate the exterior walls from the interior building
space for their full height and mechanically supplying warm, dry air to this
1
P.E., Principal, Building Consultants & Engineers, Inc., 1520 West Canal Court, Suite
240, Littleton, CO 80120.
2
P.E., A.I.A., President and Principal, Building Consultants & Engineers, Inc., 1520 West
Canal Court, Suite 240, Littleton, CO 80120.
3
P.E., Vice-President and Senior Principal, Building Consultants & Engineers, Inc., 1520
West Canal Court, Suite 240, Littleton, CO 80120.
Cite as: Morgan, C. M., McGowan, L. M. and Flick, L. D., Case Study of Mechanical
Control of Condensation in Exterior Walls, J. Test. Eval., Vol. 39, No. 2. doi:10.1520/
JTE102959.
226
MORGAN ET AL., doi:10.1520/JTE102959 227
interstitial space, was the desired method to correct the condensation prob-
lems that were occurring. These repairs have proven to be effective. This
paper discusses the original design and construction, the method of evalua-
tion of the condensation problems that occurred, and the design and results
of the implemented repairs.
KEYWORDS: condensation, exterior walls, windows, vapor barrier
General Description of Building
The subject building was a two-story electronics manufacturing facility near

Denver, CO, which housed both laboratory and ofce spaces. The steel and
concrete framed structure was constructed as a design/build project in 2000
2001 with a total of 41 800 m2 450 000 ft2. The south elevation featured a
large aluminum-framed curtain wall at the two-story entry and reception area.
The long dimension of the building north-to-south was bisected by a two-
story central corridor illuminated by four large skylights mounted onto a raised
portion of the roof along the corridor. Figure 1 shows the general layout of the
oor plan of the building.
The exterior walls were constructed with a combination of precast concrete
panels with light-gauge steel framing on the interior side, conventional light-
gauge steel framing clad with metal panels, and aluminum-framed curtain
walls. Aluminum-framed windows were positioned into punched openings
through the precast concrete panels and the conventional steel, and formed a
number of bay windows, which extended beyond the face of the exterior walls
along the north elevation. At the exterior walls, foil-faced berglass batt insu-
lation was installed within the cavities of the steel framing and covered with
gypsum board, which extended from the oor to the bottom of the perimeter
steel beams. The rst oor consisted of concrete slabs-on-grade, and the second
oor was constructed of concrete on steel deck supported by steel framing. The
roong system consisted of a build-up roong membrane over rigid and ta-
pered insulation over a steel roof deck supported by structural steel framing.
No vapor retarder was installed in the roong assembly. A roof expansion joint
extended east-to-west across the entire width of the building.
The east half of the building featured a large laboratory containing multiple
clean rooms and assembly line spaces, which required an interior air relative
humidity RH level of 48 % RH at 21 C 70 F under a slightly positive
space pressurization of about 12 Pa 0.05 in. H2O. The west half of the build-
ing, which primarily consisted of ofce space, the central corridor, and the
entry and reception area on the south end of the building were not humidied
and were not pressurized.
The mechanical system for the east side of the building provided mechani-
cally humidied air to both the rst and second oors of the laboratory space
through ductwork and supply vents mounted above the suspended ceiling. Re-
turn air was handled in the air plenum above the suspended ceiling. In most
areas, the suspended ceiling and air plenum extended up to and terminated
against the exterior walls. The suspended ceilings were constructed of with a
steel grid and drop-in ceiling tiles. Above the second oor, the suspended ceil-
FIG. 1As-built typical oor plan.
ing and air plenum terminated about 3 m 10 ft south of the north exterior
wall, exposing the steel roof framing and steel roof deck in this area. The ves-
tibules and stairwells were typically open to the surrounding space and had
exposed framing at the second oor and roof levels.
Evaluation of Condensation-Related Problems

The evaluation of the condensation-related problems began with an interview
of knowledgeable personnel associated with the facility and a brief walk-
through around the building interior and exterior. A detailed examination of
FIG. 2Condensation on the sill of the window frame and glazing.
the original design and construction documents followed in order understand

the details of the construction and to gain insight into possible causes for the
problems. A systematic examination commenced including detailed visual ex-
amination and partial disassembly of some components of the existing con-
struction. Most observations were visual and tactile, although hand-held infra-
red thermometers and digital pin-style moisture meters were also used.
Observations of Moisture-Related Damage

Shortly after construction and during cooler weather, condensation was ob-
served on the interior surfaces of the aluminum window frames and on por-
tions of the glazing on the east and north elevations at the laboratory space.
Additionally, water was observed seeping out from below the precast wall pan-
els directly above the concrete foundation stem wall and water damage around
the skylights in the central corridor was noted. Prior to the involvement of the
authors, initial attempts were made to remedy these problems, which included
wet-sealing of some of the windows, installation of berglass batt insulation
and gypsum board at some perimeter conditions, the addition of mechanical
ductwork and supply vents at select locations, and the repair of roong termi-
nations. Despite these attempts, the problems did not abate, and condensation-
related problems were being reported in areas of the building outside the labo-
ratory space.
During the investigation, signs of moisture-related damage were observed
throughout the building. The apparent damage was particularly signicant
along the north and east elevations of the building. The following is a summary
of the moisture-related damage that was observed.
During the Fall, Winter, and Spring months, condensation was observed
on the aluminum window frames and portions of the glazing along the
majority of the north and east sides of the building adjacent to the
laboratory spaces in the building. Figures 2 and 3 show examples of the
condensation formation on the window glazing and frames. Water from
the condensation pooled on the sills and ran down the interior surface
of the walls and onto the oors below the windows.
FIG. 3Condensation on the head of the window frame.
Along the north and east exterior walls, the interior surface of the pre-
cast concrete wall panels and the berglass batt insulation were visibly
wet. The open-cell backer rod behind the sealants at joints in the precast
concrete wall panels and at joints around the window frames was satu-
rated with water.
Water-damaged and stained gypsum board and interior nishes indi-
cated mold contamination within the exterior wall assembly. Mold was
found within exterior walls along the majority of the north and east
elevations of the building. Figure 4 shows an example of moisture-
related damage to the gypsum board at the head condition of a bay
window.
Where the oor and roof structure met the north and east exterior walls,
the embedded steel plates and the exposed steel beams above the sus-
pended ceiling were covered with surface rust. Water was observed drip-
ping down the interior face of the steel at the second oor at the open
stairwells at the northeast corner of the building.
Along a portion of the exterior side of the east exterior wall, water was
observed draining out of the joint between the bottom of the precast
FIG. 4Damaged gypsum board and interior nishes at head of bay window.
FIG. 5Water leakage between precast concrete wall panels and foundation wall.
concrete wall panels and the top of the concrete foundation stem wall.
Figure 5 shows an example of this condition.
Along the exterior side of the north elevation, evidence of moisture was
observed discharging out of joints between the window frames and the
precast concrete wall panels and between joints in the metal panel clad-
ding system near the bay windows. Figure 6 shows an example of this
condition.
Water was observed discharging from the roof expansion joint located
above the east half of the building and the central corridor. On the east
half of the building, water was also observed discharging from below the
roof counterashing and above the base ashings of the roong mem-
brane along the parapet walls. During times of cooler weather, this
water formed icicles.
Signs of moisture were observed around the interior of the skylight glaz-
ing assemblies located along the central corridor, and water was ob-
served discharging from the sill framing member of the skylights, both
on the interior and exterior of the building. Figure 7 shows an example
of this condition.
FIG. 6Moisture at window-to-precast interface along north elevation.

FIG. 7Water discharging from the sill framing member of the skylight.
In many locations on all building elevations, the sealant joint between

the precast concrete wall panels had failed. Debonding and splitting of
the sealant were observed.
Although the mechanical system was designed with excess capacity, the
mechanical system had been operating at maximum capacity in order to main-
tain the interior temperature and humidity conditions required for the labora-
tory space on the east half of the building.
Identication of Causes of Moisture-Related Damage
Discontinuous Vapor Barrier at Exterior Precast Concrete Wall PanelsFoil-

faced berglass batt insulation was intended to serve as both the vapor barrier
and the insulation for the exterior walls. The insulation, about 9 cm 3 1/2 in.
thick with a listed insulative value of R-11, was installed between steel framing
on the interior side of the precast concrete wall panels. The design drawings
indicated the vapor barrier foil-facing with a dashed line but did not indicate
how terminations were to be detailed. The specications required the termina-
tions of the anges of the foil-faced berglass batt insulation be sealed to
produce a continuous, overlapping of facing material for an airtight
installation.
The foil-facing was installed continuously over surface of the steel stud
anges and was taped with adhesive-backed foil tape. However, the foil-facing
was terminated short of the edges of the cavity at window and door openings in
the precast concrete wall panels creating discontinuities in the vapor barrier. At
these openings, the foil-facing was not wrapped into the rough opening and
was not otherwise sealed to the precast concrete wall panels or aluminum
framing. In addition, penetrations through the foil-facing, such as at electrical
junction boxes and steel beam penetrations, were not sealed in any manner.
Further, the foil-facing was not sealed to the oor, to the bottom of the perim-
eter steel beams, or to the underside of the steel decking.
On a day when the exterior temperatures were 3 C 27 F, the interior
surface of the precast concrete wall panels and the adjacent surface of the
berglass batt insulation was visibly wet, as was the open-cell backer rod be-
FIG. 8Gaps between metal decking and top of corridor wall.
hind sealant in joints in the precast concrete wall panels. The temperature of
the interior surface of the precast concrete wall panels was measured using a
hand-held infrared thermometer to be about 7 C 44 F.
Lack of Vapor Barrier at Central Corridor WallsUnfaced berglass batt in-

sulation had been installed within the cavities of the steel framing of the corri-
dor walls. Neither the drawings nor the specications indicated the use of a
vapor barrier in the walls between the laboratory space on the east half of the
building and the central corridor walls. These walls separate the humidied
space of the laboratory from the non-humidied space of the remainder of the
building.
At the underside of the second oor and roof level framing, a continuous
opening was discovered along the top of the each of the center corridor walls at
the intersection with the steel roof deck. The corridor walls were oriented at an
angle to the direction of the steel deck utes; therefore, openings for air ow
were created as the utes extend past the top of the wall. No seal was provided
to reduce the air transfer through these gaps. Figure 8 shows an example of this
condition. Additionally, all of the electrical and mechanical penetrations
through the gypsum board were not sealed.
Discontinuous Vapor Barrier and Insulation at Exposed Framing and

StairwellsAbove the second oor, the suspended ceiling and air plenum ter-
minated about 3 m 10 ft south of the north exterior wall, exposing the steel
framing and steel roof deck. The steel framing and steel roof deck were also
exposed above the stairwell located at the northeast corner of the building.
Steel beams span parallel to the exterior precast concrete wall panels along the
north and east sides of the building. Unsealed gaps were observed where the
utes in the steel deck bear on the top anges of the perimeter beams, permit-
ting warm, moist air from the laboratory space to ow into the roong assem-
bly along the parapet walls on the north and east sides of the building.
The steel framing inside of the exterior precast concrete wall panels termi-
nated at the underside of the perimeter steel beams. The foil-facing serving as
the vapor barrier for the exterior walls was not continuous behind the steel
beams nor was it terminated to the underside of the steel beams. One layer of
unfaced berglass batt insulation, about 9 cm 3 1/2 in. thick with a listed
insulative value of R-11, had been installed between the perimeter steel beams
and the precast concrete wall panels; however, in some locations the insulation
was poorly installed and did not extend the full height of the beams. Where
these conditions were checked, the interior surface of the precast concrete wall
panel was wet to the touch and the berglass batt insulation appeared to be
moist. In a few locations, the insulation between the beams and the exterior
precast wall panels had not been installed.
At the east wall, where the perimeter steel beam transitioned from the
enclosed area above the suspended ceiling to the area of exposed framing, the
space between the beam and the precast concrete wall panel was not sealed nor
had gypsum board been installed the close the gap around the beam. Near this
gap, the berglass batt insulation and the exposed surface of the precast con-
crete wall panel were wet to the touch.
Curtain Wall Framing System at StairwellsThe specications indicated

that the aluminum curtain wall framing system was to be thermally broken,
and the assembly was specied to have a condensation resistance factor CRF
of 51 or greater. The curtain walls were specied to be exterior-glazed with 2.5
cm 1 in. thick insulated glass units, but no U-factor was specied for the
glazing. The aluminum frames for the doors were also thermally broken; how-
ever, the doors and thresholds were not thermally broken.
Along the jambs, the curtain wall framing was positioned completely
within the thickness of the exterior precast concrete wall panel and the alumi-
num framing on the interior and exterior sides of the thermal break were ad-
jacent to the edges of the precast concrete wall panels. Therefore, the alumi-
num framing on the interior side of the thermal break was not insulated. The
design drawings did not indicate where the thermal break in the curtain wall
framing system was to be positioned within the depth of the wall assembly or
how the curtain wall framing was to be insulated.
The curtain wall framing extended the full height of the exterior wall. A
spandrel panel was positioned at the upper portion of the curtain wall from
about the elevation of the steel roof deck to the top of the parapet. No attempts
were made to seal the interior side of the curtain wall framing system at the
perimeter steel beam. Moisture was observed behind the spandrel panel at the
parapet.
Several repair attempts had been made to the curtain wall system at the
stairwells at the northeast corner and east side of the building following the
original installation to try to reduce the formation of condensation on the alu-
minum frames and glazing. These included wet-sealing of the curtain wall
framing and glazing and the installation of electric baseboard heaters along the
bottom of the curtain walls. Despite these attempts, water was observed drip-
ping down the interior face of the glazing from the top of the curtain wall on
the north elevation, and moisture stains on the sills of curtain wall were ob-
served at multiple locations. The temperature of the interior surface of the
aluminum curtain wall framing was measured using a hand-held infrared ther-
mometer to be about 8 C 46 F on a day when the exterior temperatures were
3 C 27 F.
Punched WindowsThe punched windows were specied to consist of an

aluminum thermally broken curtain wall framing system with a minimum CRF
for the assembly of 57. The glazing specication called for 2.5 cm 1 in. thick
insulated glass units, but no U-Factor was specied. The aluminum window
frames were positioned within the thickness of the exterior precast wall panels.
The frames were bearing on the precast concrete panels and the sills were
shimmed up 1 cm 1/2 in. above the top of the rough opening in the precast
concrete. Although the design drawings graphically indicated that the alumi-
num frames were to be closed and thermally separated from the precast con-
crete wall panels, in the as-constructed condition, the outer edges of the alumi-
num frames were open and no insulation had been installed between the
aluminum frames and the rough openings.
The steel framing on the interior side of the precast concrete wall panels
surrounded the punched windows on four sides, which was covered with gyp-
sum board. The gypsum board returned toward the inside face of the alumi-
num window frames on four sides where it was sealed with sealant. Foil-faced
berglass batt insulation had been installed within the steel framing; however,
the foil-facing did not wrap into the rough opening at the sill, jamb, or the head
conditions and was not sealed to the aluminum frames.
Condensation was observed on the aluminum frames and glazing around
the windows, but appeared to be worse at the sill and head conditions. The
interior surfaces of the precast concrete wall panels adjacent to the windows
were wet to the touch, as was the berglass batt insulation in contact with the
panels. The open-cell backer rod behind the sealants around the window
frames was also saturated with water.
Bay Windows along North ElevationThe bay windows were similar to the
punched windows, except that the aluminum framing system was positioned
15 cm 6 in. beyond the face of the precast concrete wall panels. At each bay
window, small side lights returned to enclose projected portions of the windows
to the exterior surfaces of the precast concrete panels. Steel framing and metal
wall cladding was constructed to inll portions of the bay windows at the oor
lines and above the bay windows.
The design drawings showed a large steel angle supporting the bottom of
the bay window assembly and indicated a thermal separation between the alu-
minum framing and the steel support structure. In the as-constructed condi-
tion, the bottom of the bay window frames were supported by steel shim plates
on a continuous structural steel assembly welded to steel plates embedded in
the face of the precast concrete wall panels.
Although the design drawings indicated that the foil-facing on the ber-
glass batt insulation was to extend into the rough opening at the sill and be
sealed to surfaces at other locations, this was not done in the as-constructed
details. The structural steel under the bay windows was wet and signicantly
rusted. The interior surface of the precast concrete wall panels above the bay
FIG. 9Ductwork and diffusers installed along north wall in attempt to mitigate con-
densation on bay windows.
windows was wet, and berglass batt insulation was saturated with water. The
inside surface of the metal cladding was also wet as the result of condensation
from the building interior. Mold also appeared to be growing on the top surface
of the interior gypsum board soft.
New Mechanical Ductwork and Supply Vents Install as a RepairAs part of

the initial attempts to remedy condensation problems on the north-facing bay
windows at the second oor level, new mechanical ductwork and supply vents
were installed along the upper portions of the windows. The diffusers were
oriented such that warm air would be directed toward the windows. Figure 9
shows the ductwork and diffusers along the north wall, which had been in-
stalled in an attempt to mitigate condensation on the bay windows. However,
the supply air for this new ductwork was provided by an existing system that
was mechanically humidied. Therefore, this increased the condensation prob-
lems on the bay windows by supplying additional moisture to the air near the
bay windows.
RoongSix test cuts into the built-up roong system were made to exam-
ine the details and components of the roong assembly and to observe whether
moisture inltration into the roong assembly had occurred. Three of the test
cuts were made adjacent to the parapet walls; one test cut was made at the
northwest corner above a non-humidied portion of the building, and the other
two test cuts were made above the east half of the building. Two test cuts were
made in the eld of the roof above the east half of the building 6 m 20 ft
from the exterior walls, and one test cut was made near the roof expansion
joint.
The perimeter test cuts revealed that the steel roof deck terminated 2 cm
3/4 in. short of the precast concrete wall panels and that no air barrier or
vapor retarder had been installed along this gap. Rigid polyisocyanurate insu-
lation board with asphalt-adhered tapered perlite insulation was installed over
the steel roof deck and terminated ush with the precast concrete wall panel.
Fiberglass felt plies were adhered to and terminated at the top of a perlite cant
strip and were covered by a granular-surfaced modied bitumen base ashing
membrane. The base ashing membranes were adhered onto the roof-side sur-
face of the precast concrete wall panels and terminated 18 cm 7 in. above
the roof surface. Sheet metal counterashing and a rubberized plastic sheet
membrane lapped over the top of the base ashing and extended up the vertical
face of the parapet wall perimeter wall and under the sheet metal parapet cap.
The test cut that was made near the northwest corner of the building over
non-humidied ofce space revealed voids in the base ashing, which were
lled with water. Using a pin-style digital moisture meter, moisture was found
in the tapered perlite insulation, but the polyisocyanurate insulation was dry.
Air ow from the interior of the building was evident where base ashing was
removed.
At the second and third perimeter test cuts, water owed out from the
berglass felt plies and the perlite cant strip when the base ashings were cut
and removed. Also, the surface of the precast concrete wall panel behind the
base ashing was visibly wet. Using a pin-style digital moisture metre, the ta-
pered perlite insulation was found to be saturated, but the polyisocyanurate
insulation was dry.
A test cut at the expansion joint above the east half of the building revealed
conditions similar to the perimeter test cuts. Moisture was found behind the
base ashing and in the perlite cant strip and berglass felt plies. One differ-
ence was that wood blocking forming the expansion joint was also saturated
and moisture was found on the underside of the elastomeric ashing that cov-
ered the expansion joint. Moisture meter readings revealed that the tapered
perlite insulation was saturated, but the polyisocyanurate insulation was dry.
At the two test cuts made in the eld of the roong above the east half of
the building, no visible moisture was observed and the components were dry to
the touch. Using a pin-style digital moisture meter, both the tapered perlite
insulation and the polyisocyanurate insulation were found to be dry.
SkylightsThe four large skylights located in the central corridor consisted

of translucent, insulated berglass panels mounted onto an integral aluminum
framing grid supported on a raised roof surface. Condensation was observed to
have formed on the underside of the skylights, and water from the condensa-
tion dripped from the skylights onto the corridor oor during the cold winter
months. In some areas, the berglass panels had delaminated from the alumi-
num framing grid. Deections in excessive of 2.5 cm 1 in. in 3.7 m 12 ft were
measured across the span of the skylights.
Discussion of Causes of Moisture-Related Problems

The high indoor relative humidity coupled with the slightly positive space pres-
surization the laboratory spaces within the building placed signicant demands
upon building enclosure in terms of resisting the formation of condensation
and controlling air exltration. Widespread condensation problems developed
as the result of the warm, moist air from within the laboratory spaces coming
into contact with cold surfaces of the building enclosure. While the condensa-
tion problems were concentrated at the exterior walls, windows, and roof areas
located adjacent to the laboratory, condensation problems developed at interior
spaces adjacent to the laboratory, such as at the central corridor. Additionally, it
appeared that some moisture-related problems were beginning to occur at non-
humidied spaces throughout the building.
Condensation is the change of water from its gaseous form water vapor
into liquid water. The formation of condensation on portions of the building
envelope is a well-documented phenomenon in the building industry, particu-
larly with humidied buildings in cold climates. Negating possible mitigating
factors such as air movement, in simple terms condensation requires two con-
ditions to form 1 a cold surface on which water vapor can condense and 2 a
source of interior moist air.
The relationship between the temperature of interior surfaces and the
amount of moisture in the air inside the building typically measured as relative
humidity that results in condensation is well known. For a given interior rela-
tive humidity, there is a surface temperature, called the dew point temperature,
at and below which condensation will occur. The higher the indoor relative
humidity, the warmer the surface temperature at which condensation will
occur. For the laboratory portion of this building that was intended to maintain
an indoor relative humidity of 48 % RH at 21 C 70 F, the surfaces of the
building enclosure should remain above 10 C 50 F to prevent the forma-
tion of condensation. Measurements of the surface temperatures of the interior
surface of the precast concrete wall panels, the aluminum window framing,
and structural steel framing supporting the bay windows were all below this
dew point temperature.
The laboratory space within this building required a high indoor relative
humidity and a narrow range of indoor temperatures and with a slightly posi-
tive space pressurization. The exterior walls, windows, doors, and roof assem-
blies should have been designed and constructed with a high level of regard to
maintain these indoor environmental conditions without negative effects on
the building enclosure. While the remaining portions of the building did not
require mechanical humidication, the division of the laboratory space from
the other portions of the building interior should have been designed and con-
structed to maintain adequate division between the areas. It is clear that this
was not achieved.
Several options are available to reduce the risk for the formation of con-
densation on and within the building enclosure: 1 Raise the surface tempera-
tures of the condensing surfaces of the building enclosure, 2 lower the indoor
relative humidity, or 3 construct a continuous barrier that prevents the inte-
rior moisture-laden air from reaching cold exterior elements of the building
enclosure.
Vapor BarrierThe installation of a perfect, continuous, and air-tight va-

por barrier on the interior side of the exterior walls and roof assembly around
the laboratory space would have reduced the potential for moist indoor air
from coming into contact with the cold surfaces of the building enclosure.
However, with the conventional construction utilized for this building, it is
exceptionally difcult to construct the necessary perfect air-tight vapor barrier,
which would prohibit the passage of moist indoor air via diffusion and air ow
from reaching the cold portions of the building enclosure. As a result, the as-
constructed vapor barrier proved to be ineffective. Additionally, the lack of any
type of continuous air and vapor barrier between the laboratory space and the
remainder of the building interior exacerbated moisture problems in other por-
tions of the building. Once condensation had occurred, the formation of mold
within the wall cavity created health-related risks beyond the performance is-
sues described previously.
Punched Window, Bay Windows, and Curtain Wall AssembliesThe speci-

cations for the window and curtain wall framing systems indicated a minimum
CRF of 5157 for the entire assemblies. According to information published by
the window and curtain wall manufacturers, the CRF value for the thermally
broken aluminum frames was about 72. However, due to the relatively lower
thermal performance of the 2.5 cm 1 in. thick insulated glass units, the CRF
for the assemblies ranged from 51 to 57. The wintertime outdoor design tem-
peratures for the Denver metropolitan area range from 19 C 2 F 1 to
16 C 3 F 2. Using an outdoor temperature of 16 C 3 F, the minimum
CRF for the entire assemblies should have been in the range of 70 3 to signi-
cantly reduce the risk for the formation of condensation on the window and
curtain wall assemblies in the laboratory during the majority of the coldest
wintertime temperatures.
The curtain wall areas did not have any type of window covering, and the
aluminum mini-blinds at the punched windows and at the bay windows were
reportedly completely open most of the time including during the evaluation.
Nevertheless, signicant condensation was observed on the interior aluminum
framing and portions of the glazing for both the window assemblies and the
curtain wall assemblies on a day when the exterior temperature was 3 C
27 F. At this outdoor temperature, and assuming an assembly CRF value of
5157, condensation should not have formed on the assembly at the given in-
door relative humidity and temperature of the laboratory. Therefore, it ap-
peared that the placement of the framing systems within the precast concrete
wall panels had lowered the anticipated performance of the aluminum framing.
Other factors such as air leakage around the perimeters and through the alu-
minum framing systems, contact of the aluminum framing with conductive
materials, lack of insulation at or near the aluminum framing, and the com-
plete lack of integration of the foil-facing vapor barrier with the framing system
also likely reduced the anticipated performance of the aluminum framing sys-
tems for all of the window types.
Solutions to Address Condensation Problems

Several constraints were placed on the design of the repairs for the building.
These included the following.
The ofce and laboratory activities within the building would stay in full
operation, 24 h per day, 7 days per week, with minimal interruptions.
The repairs would not include the loss of a signicant amount of oor
space within the laboratory.
Natural light into the laboratory spaces from the windows and curtain
walls would need to be preserved.
Large, expensive, sensitive, and precision equipment in the laboratory
would remain in-place and would need to be protected.
Repair efforts would need to be well-planned and coordinated to allow
for the systematic, temporary relocation of work activities and equip-
ment during the mold mitigation and installation of repairs/
modications in a given area.
An industrial hygienist was retained to better dene the nature and extent
of the mold contamination and to develop a remediation plan to mitigate the
mold contamination during the repairs. Sampling was performed on every ex-
terior wall section around the perimeter of the humidied spaces. The majority
of the exterior wall spaces was found to be in need of complete mold mitiga-
tion, which would require the removal and replacement of the majority of the
interior nishes, interior gypsum board, and insulation along the exterior walls.
A mechanical engineer was retained to evaluate the existing mechanical
systems for the building and to assist in developing the repair plan. Modica-
tions to the existing mechanical system were found to be necessary, and new
mechanical and electrical systems were required in the repair.
The design team developed a plan for the remediation of the mold contami-
nation and for the repairs. In general, this included creating dry corridor
spaces around the mechanically humidied laboratory space and modifying the
existing mechanical system to supply separate dry air into these corridor
spaces and moist air into the laboratory space. Figure 10 shows the general
layout of the building indicating the separation of dry and moist air spaces.
The following describes the repairs for each of the causes of moisture-related
damage identied above.
Exterior Precast Concrete Wall Panels

The typical exterior wall was modied by constructing a new wall 36 cm 14 in.
inside of the existing exterior wall, thus creating a new dry corridor between
the laboratory space and the existing exterior wall. The new wall was con-
structed just inside of the existing perimeter steel columns, which were about
30 cm 12 in. deep, to provide a continuous interstitial space while resulting in
minimal loss of oor space. The new wall extended from the oor to the under-
side of the steel oor or roof deck and consisted primarily of wall panels manu-
factured specically for clean rooms with glass panels to allow natural light
from the existing windows and curtain walls to enter. The clean room wall
panels were gasketed, and the perimeter edges were sealed to minimize air
leakage through the walls. Above the suspended ceiling, light-gauge steel fram-
FIG. 10Modied typical oor plan.
ing, gypsum board, and a vapor impermeable membrane were utilized to allow
the upper portions of the walls to conform to the irregularities of the steel deck,
and the new wall was sealed to the underside of the steel deck. Figure 11 shows
the approximate cross-section through the modied wall section.
A new, separate mechanical system was installed to supply warm, dry non-
humidied air into the dry corridor space. Air supplied from this new me-
chanical system was supplied in new ductwork and delivered the warm, dry air
to the dry corridor near the oor level and return air vents were positioned in
the upper portion of the dry corridor space. The new mechanical system was
sized to provide a sufcient number of air exchanges within the dry corridor
and to account for air loss from the laboratory space into the interstitial space.
The exhaust-side ductwork was equipped with sensors to monitor and auto-
FIG. 11Typical exterior wall sectionmodied.
matically adjust the number of air exchanges should excessive moisture accu-
mulate within the dry corridor. The laboratory spaces were maintained at the
required 48 % RH at 21 C 70 F and at a slightly positive space pressuriza-
tion relative to the dry corridor. The dry corridor was also maintained at a
slightly positive pressure relative to the building exterior. This ensured that the
quality of the air in the clean-rooms was maintained by not allowing non-
ltered air from within the dry cavity or from the building exterior to ow
into the laboratory.
To further monitor the performance of the new dry corridor space, tem-
perature and humidity sensors were installed within the interstitial space at
multiple locations. These monitoring sensors could be observed through the
glass panels in the clean room walls and were monitored regularly following
the repair work.
For the existing exterior walls, all of the interior gypsum board and insula-
tion was removed, allowing the interior surfaces of the precast concrete wall
panels and the light-gauge steel framing to be cleaned. All of the joints in the
precast concrete wall panels and the metal cladding were re-sealed with closed-
cell backer rod and polyurethane sealant. Following mold remediation activi-
ties on the interior of the precast concrete wall panels, new unfaced berglass
batt insulation and gypsum board were installed. New light-gauge steel framing
was used to box around the perimeter steel beams parallel to the precast con-
crete wall panels, and unfaced berglass batt insulation and gypsum board
were installed. The gypsum board was sealed at all perimeter conditions to
create an air seal and was painted with a vapor-permeable paint.
Unavoidable penetrations in the new clean room walls and in the repaired
exterior walls, such as electrical outlets and junction boxes, were designed with
compressible gaskets at the outlet covers and rubber gaskets around the junc-
tion boxes box to reduce the amount of air loss through the walls.
Central Corridor Walls

At the interior walls separating the laboratory from the non-humidied por-
tions of the building, no provisions to control air leakage or vapor diffusion
were provided in the original design and construction. To rectify this condition,
several modications were made.
The continuous opening along the top of the wall formed by utes in the
steel deck was lled with a compressible material and coated with a
vapor impermeable membrane coating. This was done to prevent the
uncontrolled ow of air from the laboratory into the central corridor.
To reduce the air ow through the interior walls, penetrations through
the walls were sealed with closed-cell backer rod and polyurethane seal-
ant and outlet cover gaskets were installed.
All of the doors between the laboratory and the central corridor were
tted with new door sweeps and weatherstripping to reduce the amount
of air leakage.
To reduce vapor diffusion through the interior walls assembly, the labo-
ratory sides of the walls were painted with a low vapor permeance paint.
The corridor side of the walls had previously been painted with a vapor-
permeable paint to allow moisture within the corridor walls to dry via
vapor diffusion into the non-humidied corridor space.
Area of Exposed Framing and Stairwells

Because the oor area below the exposed framing at the second oor level was
relatively minimal and since the activities in this space, including the northeast
stairwell, did not require humidica-tion, this portion of the building was con-
verted to non-humidied space. This space was separated from the laboratory
space with similar clean room walls as was used for the dry corridors, and
warm, dry air was supplied into this space from the dedicated mechanical
system for the dry corridors.
At the interior stairwell located on the east side of the building, a new dry
corridor was constructed around three sides of the stairwell similar as was used
for the typical exterior walls and was made continuous with the dry corridor
along the exterior walls. Where necessary, the dry corridor width was increased
to accommodate egress requirements. Additional doors were installed to sepa-
rate the new dry corridors from the laboratory spaces, and the doors were
equipped with door sweeps, auto-closers, and weatherstripping. These new dry
corridors and the adjoining stairwell were supplied with dry air from the dedi-
cated mechanical system.
Curtain Walls, Punched Windows, and Bay Windows

The majority of the issues with the window systems were addressed by the
addition of the new dry corridor. The existing aluminum framing assemblies
and glazing were capable of controlling condensation at the relative humidity
levels maintained for the dry corridors. Therefore, in general, no specic re-
pairs were required at these windows other than those described above for the
exterior precast concrete wall panels. Routine maintenance such as replace-
ment of the perimeter joints with closed-cell backer rod and polyurethane seal-
ant and replacement of some gaskets was performed.
At one laboratory space near the northeast corner of the building, it was
not possible to construct the dry corridor between the laboratory space and the
exterior wall or to convert the space to a non-humidied area. Therefore, fol-
lowing the typical remediation activities, the exterior wall was reconstructed
similar to the existing wall; however, a vapor-impermeable membrane was in-
stalled behind the new gypsum board. Measures were taken to ensure continu-
ity of the vapor-impermeable membrane with the other laboratory walls and
with the window framing system. Additionally, the existing window within this
laboratory space was modied to signicantly improve its thermal performance
and increase its resistance to the formation of condensation. This was accom-
plished by installing a new aluminum frame to support a piece of new single-
pane glazing on the interior side of the existing window. Butyl tape was in-
stalled around the perimeter of the new frame to create an air-tight seal and a
thermal break between the glazing and the frame.
New Mechanical Ductwork

Throughout the building, mechanical ductwork was modied to separate the
mechanically humidied air supplied to the laboratory from the remaining
non-humidied portions of the building.
Roong
To reduce the risk of moist air from entering into the roong system, all of the
perimeter base ashings, expansion joint, drain penetrations, and mechanical
unit penetrations located above the east half of the building were repaired and
modied. The openings in the steel roof deck along and around these areas
were sealed to minimize air exltration. Damaged materials were removed and
replaced.
Skylights
The damage to the skylights was twofold. The structural assemblies were com-
promised by the delamination of the berglass panels from the aluminum
framing grid, and leakage was discovered along the edges of the skylight units.
The manufacturer of the skylights provided a structural repair, which reduced
the spans of the skylights by the installation of new structural beams below the
skylights. The leakage problems were repaired by re-sealing the perimeter
joints and re-working the transition from the skylight to the roong system.
The skylights were water spray tested to conrm that the leakage problems had
been corrected. Potential condensation problems with the skylights were re-
duced due to the improvements to the corridor walls.
Conclusions
Monitoring of the installed systems revealed that conditions immediately im-
proved within the building during the rst winter following the repairs. Con-
densation formation was eliminated on the exterior windows, walls, and sky-
lights throughout the building. The operation of the existing mechanical system
for the humidied laboratory spaces was reduced to a level of about 50 % of
maximum capacity. A number of air samples were taken by the industrial hy-
gienist in the exterior wall cavities 1 year following the repairs. No moisture
accumulation, contamination, or air-quality problems were identied.
Although the cost of the repairs to the building was high, in excess of $2.5
million dollars, the overall long-term maintenance costs of the repairs are ex-
pected to be relatively minor over the age of the structure.
It is clear that the original design and construction failed to account for the
signicant affect mechanically humidication and positive space pressuriza-
tion have on the building enclosure in cold climates. These factors need to take
primary importance in the design and construction of all aspects of the enclo-
sure elements and transition details in particular. By creating an interstitial
space supplied with warm, dry air between the mechanically humidied inte-
rior spaces and the exterior building walls, the need to create a perfect vapor
barrier at the exterior walls was eliminated. This also eliminated the need for
high-performance window assemblies.
References
1 99% Winter Design Dry Bulb Temperature, Handbook of Fundamentals, Ameri-

can Society of Heating, Refrigerating and Air Conditioning Engineers, Inc.
ASHRAE, Denver, CO, 1972, Chap. 33, Table 1.

2 97.5% Winter Design Dry Bulb Temperature, Handbook of Fundamentals, Ameri-
can Society of Heating, Refrigerating and Air Conditioning Engineers, Inc.
ASHRAE, Denver, CO, 1972, Chap. 33, Table 1.
Condensation Resistance of Windows, Doors, and Glazed Wall Section, American
Architectural Manufacturers Association AAMA, Schaumburg, IL, Fig. 14.
doi:10.1520/JTE102999
Sean M. OBrien1 and Amrish K. Patel2
Considerations for Controlling Condensation

in High-Humidity Buildings: Lessons
Learned
ABSTRACT: Condensation problems in general use i.e., non-humidied

buildings such as ofces, schools, and condominiums typically manifest
themselves as visible staining on window and door perimeters or minor drip-
ping from overhead components. The relatively low interior moisture levels
that are typical of these types of buildings are generally insufcient to cause
severe water damage in the short term. At the other end of the spectrum,
high-humidity buildings such as museums and natatoriums can suffer ex-
treme damage due to condensation, sometimes within weeks, not years, of
completion. High interior moisture levels and, in many cases, differential air
pressures, contribute to condensation in both visible locations, such as win-
dows and curtain walls, and concealed locations within walls and roofs.
Signs of concealed condensation such as dripping water or rust stains may
only become visible after moderate to heavy damage has already occurred
within the enclosure. These problems are typically more severe in cold cli-
mates, but high-humidity buildings may experience condensation problems
in mild climates, where such issues are often not considered by designers
due to the lack of prolonged cold weather during the winter. This paper will
review the severe and immediate condensation problems that are unique to
high-humidity buildings, including surface condensation and concealed con-
densation due to air leakage through the enclosure. It discusses the theoret-
ical mechanisms by which condensation forms in building enclosures, and
illustrates these concepts through various case studies in both cold and mild
lished online January 2011.
1
P.E., LEED AP, Building Technology, Simpson Gumpertz and Heger, Inc.,19 West 34th
St., Suite 1000, New York, NY 10001.
2
P.E., Building Technology, Simpson Gumpertz and Heger, Inc.,19 West 34th St., Suite
1000, New York, NY 10001.
Cite as: OBrien, S. M. and Patel, A. K., Considerations for Controlling Condensation in
High-Humidity Buildings: Lessons Learned, J. Test. Eval., Vol. 39, No. 4. doi:10.1520/
JTE102999.
247
climates. This paper will focus on design strategies for avoiding problems,
but also discusses remedial work to existing buildings, drawing on the au-
thors experience with the investigation and repair of high-humidity buildings.
KEYWORDS: condensation, high-humidity, museum, natatorium,
hospital, airow
Introduction
When discussing condensation and high-humidity buildings, museums and na-

tatoriums often come to mind. These buildings are typically maintained at rela-
tively high temperature and/or relative humidity RH levels, which can lead to
signicant condensation if sufcient building enclosure and mechanical sys-
tems are not designed and installed. Condensation problems are commonly
more severe and immediate i.e., visible within weeks of construction than
similar problems in general use, or non-humidied, buildings. The most com-
mon forms of condensation in high-humidity buildings include surface conden-
sation and concealed condensation in exterior wall and roof assemblies. Spe-
cic design strategies must be considered in high-humidity buildings for
avoiding condensation problems, often requiring close coordination with the
projects mechanical system design. Careful design is critical, since performing
remedial work to address condensation problems in buildings after construc-
tion is generally more expensive and less effective than addressing these con-
cerns on paper during design. Further, the budgetary and physical constraints
in existing buildings often prohibit the types of changes that can be easily
incorporated into a design to address condensation, such as upgrades to win-
dow or air barrier systems, and may result in the need for active systems such
as specic mechanical system controls or electric heat trace that can become
long term maintenance items. In some buildings, it may be possible to reduce
risk by lowering interior humidity levels. However, this is often not the case
with swimming pools or museums, where high humidity levels are a basic
requirement for the use of the building 1,2. Although not considered special-
use buildings, many air-conditioned buildings in the southeastern United States
develop problems due to the combination of extremely high exterior moisture
levels and low interior temperatures. Air leakage and moisture migration can
lead to signicant damage to these buildings if the enclosure design and me-
chanical system operation do not adequately control the exterior moisture 3.
The problems in these buildings are often similar to those encountered in
humid buildings within cold climates, with the notable exception that the pri-
mary source of damaging moisture is on the exterior, not the interior, of the
building.
Although not discussed in this paper, low-temperature buildings such as ice
rinks 4 and cold storage facilities 5 can also experience moderate-to-severe
condensation problems resulting from inadequate building enclosure design or
construction. In some ways, these problems are more challenging to address
during the design process than heated, high-humidity buildings, as they require
designers to think backwards in terms of insulation and air/vapor barrier
OBRIEN AND PATEL, doi:10.1520/JTE102999 249
systemsboth of which are necessary to prevent high-humidity conditions on

the exterior of the building from causing problems within the enclosure 6.
Condensation Overview
The following summary is not intended to fully describe the theory and mecha-
nisms of condensation, and is presented as a general background for a discus-
sion of specic condensation problems in high-humidity buildings only.
Condensation describes the phase change process by which water vapor
becomes liquid water. A given volume of moist air a mixture of dry air and
water vapor will begin to undergo this process at a temperature known as the
dew point. The dew point is a function of the absolute moisture content of the
air and is independent of air temperature. This is different from RH, which is a
function of both moisture content and temperature, making the dew point a
better parameter for comparison of moisture contents at different tempera-
tures. Thus, a higher dew point equates to higher moisture levels regardless of
interior temperature. In buildings, condensation occurs on surfaces that are
colder than the dew point temperature. To some degree, the nature of a surface
will affect condensation, as porous surfaces may absorb moisture as it forms
and before it becomes visible, while impermeable surfaces show condensation
almost as soon as it begins to form.
Modes of Condensation
Condensation in buildings can occur on both visible and concealed surfaces.
Condensation on visible surfaces such as the interior of windows and doors
occurs when those surfaces are colder than the dew point of the interior air.
Condensation on concealed surfaces can occur by one of two mechanisms.
The rst is water vapor diffusion. Water vapor in the interior air can diffuse
through vapor permeable material, such as gypsum wallboard, most brous
insulation materials, and many types of concrete and masonry. The process of
water vapor diffusion is relatively slow, and is dependent on the water vapor
permeability of the building enclosure and the interior and exterior tempera-
ture and moisture conditions. If water vapor reaches a cold enough surface in
the enclosure, it will condense. Condensation typically occurs at material sur-
faces i.e., sheathing or insulation facers, although it can occur within a ma-
terial under the right conditions. For reference, at constant interior conditions
of 70 F 21.1 C/30 % RH and constant exterior conditions of 20 F
6.6 C/80 % RH typical of a Winter day in the northeast United States, the
theoretical maximum accumulation due to water vapor diffusion through a
vapor permeable insulated exterior wall is 0.2 oz 6 mL/day per square foot
0.09 m2 of wall area. This example assumes a 5/8 in. 0.015 m thick gypsum
wallboard on either side of 6 in. 0.1524 m metal studs and R-19 RSI-3.34
glass ber batt insulation, with condensation occurring at the inner face of the
exterior wallboard. Actual accumulations will vary, as vapor diffusion rates are
dependent on interior and exterior conditions, the latter of which are rarely
constant. The potential magnitude of water vapor ow is determined by the
difference in water vapor pressure a property of moist air that, similar to the
dew point, is a measure of absolute moisture content across a component.
Greater differences in water vapor pressure generally equate to a stronger driv-
ing force behind vapor diffusion.
The second mechanism of condensation occurs due to air leakage through
the building enclosure. Moving air can transport both heat and water vapor. In
buildings without continuous air barriers or buildings with defective air bar-
rier systems, airow through cracks and gaps in the building enclosure can
transport signicant amounts of water vapor from the interior to the exterior,
often through circuitous paths in the enclosure. Unlike vapor diffusion, which
is a relatively slow process, moving air can quickly transport a large quantity of
moisture from the interior environment to components in the enclosure that
are colder than the interior dew point. Using the same example and interior/
exterior conditions described above, the theoretical maximum accumulation of
moisture due to a 2 cfm 0.94 L/s air leak is 13 oz 385 mL/day per square
foot 0.09 m2 of wall area. This represents a 60-fold increase in moisture
accumulation as compared to vapor diffusion alone, and shows that air leakage
is typically the dominant mechanism for concealed condensation in building
envelopes. For high-humidity buildings with elevated temperature and mois-
ture levels, the theoretical moisture accumulation due to air leakage may be
double or even triple the accumulations seen in typical non-humidied build-
ings.
High-Humidity Buildings
Moisture Levels
Most general use buildings without specic humidication will experience rela-
tively low interior RH levels during Winter operation. Since the level of mois-
ture in the interior air is dependent on occupancy and moisture in the exterior
air, interior RH levels during dry Winter months in colder climates may be as
low as 1020 %, at a temperature of 70 F 21.1 C, due to both ventilation
and incidental air leakage from the exterior to the interior. This is benecial to
reducing condensation potential, as the interior RH levels are generally lowest
when outdoor temperatures are low and the condensation risk is highest. Con-
versely, extremely tight buildings with little to no ventilation may experience
abnormally high RH levels during the Winter due to the buildup of occupant-
generated moisture. Even without humidication, occupant-generated mois-
ture can lead to high RH levels if not dissipated by air exchange with the
exterior.
Museums and natatoriums are the two most common types of high-
humidity buildings that a typical architect will design. The ambient airborne
moisture levels within these spaces are 2 and 3 times, respectively, greater
than those in a typical ofce building during Winter operation. Operating con-
ditions in museums and similarly, archival storage facilities do not typically
vary by season, and are generally in the range of 68 2072 F 22.2 C and
4050 % RH. Although the unofcial museum environment is often taken as
FIG. 1Comparison of interior moisture levels in typical high-humidity buildings.
70 F 21.1 C/50 % RH the origins and applicability of these conditions are

the subject of much debate, but that debate is beyond the scope of this paper
7,8, specic storage or display areas may have different requirements de-
pending on the materials being stored or displayed. The interior temperature in
natatoriums may vary depending on the temperature of the pool water which
in turn varies based on the use of the pooltherapeutic, recreation, competi-
tion, etc.. For most natatoriums, interior conditions of 80 F 26.6 C/60 %
RH are maintained year-round 1. Although some museums or natatoriums
may utilize seasonal reductions in interior RH levels, for the purposes of this
paper we do not consider those reductions as they are more the exception than
the rule.
A third building type, hospitals, is rarely thought of as requiring high-
humidity design, although in practice, condensation problems in hospitals are
relatively common. The American Society of Heating, Refrigeration, and Air
Conditioning Engineers ASHRAE provides guidelines for health care and out-
patient facilities in their 2007 Handbook of Applications 9. Conditions for
most spaces are generally in the range of 68 F 20 C to 75 F 23.9 C and
3060 % RH annually i.e., with seasonal variation. However, some special-use
spaces have more extreme temperature and RH requirements, such as radiol-
ogy 78 F 25.6 C to 80 F 26.7 C and 40 % Winter RH and post-operative
suites 75 F 23.9 C and 4555 % winter RH. Other areas, such as special-
ized equipment rooms for medical imaging, may similarly require humidica-
tion to control electrostatic discharge that could cause damage to the sensitive
equipment.
Figure 1 shows a comparison of the moisture levels expressed as mass of
water per set air volume within the building types discussed in this paper.
Table 1 below presents typical dew point temperatures for these building types.
High-humidity conditions are a signicant concern in cold climates, but
TABLE 1Comparison of interior dew point temperatures.
Building Type Dew Point F C

Ofce 27.1 2.72
Hospital: typical 37.2 2.88
Museum 50.6 10.3
Hospital: radiology 51.8 11
Hospital: post-op 52.3 11.27
Natatorium 64.9 18.27
also create a moderate condensation risk in mixed climates and even some
warm climates during parts of the year. This can be extremely problematic, as
interior condensation is rarely a concern in typical buildings in warmer cli-
mates and may not be taken into account in special-use buildings as a result.
Pressure Differentials
While interior temperature and RH levels have a direct impact on surface con-
densation risk, air pressure differentials can signicantly affect the risk of con-
densation due to airow. Museums and archives are often maintained at a
positive interior air pressure relative to the exterior as a contaminant control
measure. Positive interior pressure tends to force air out through small open-
ings in the building enclosure unsealed laps in barriers, window perimeters,
etc. as well as through large openings such as windows and doors. This pre-
vents the uncontrolled entry of exterior contaminants into storage and display
spaces 2. Similar pressure differentials in hospitals may be maintained for
similar reasons, such as protection of immune-compromised patients from
contaminants or potential infections 9. Negative pressure may be used for
containment in some areas. A unique feature of hospitals is that varying pres-
sure differentials are often maintained in adjacent areas, making airow con-
trol necessary both between internal zones and between the interior and the
exterior. Natatoriums are often maintained at a negative pressure with respect
to the exterior to prevent the migration of humid air from the interior to the
exterior. However, negative pressure systems are often incapable of overcoming
the stack effect in tall spaces caused by the buoyancy of warm air. This is
discussed in greater detail in the following sections.
Design Guidelines for High-Humidity Buildings
This section presents common problems and design guidelines for addressing
condensation in high-humidity buildings during the design phase, based on
both theory and the authors practical experience investigating problems in
high-humidity buildings.
FIG. 2Light condensation at glazing perimeter.
Surface Condensation Problems

Surface condensation is most common on glazed components such as win-
dows, skylights, and curtain walls and can vary from slight surface condensa-
tion and perimeter condensation on glazing, or picture-framing Fig. 2, to
frost and/or ice accumulation on interior surfaces of the glazed components
FIG. 3Heavy condensation and ice buildup on aluminum-framed curtain wall.

FIG. 4Skylights mounted high within a space. Arrow indicates location of closest
heat source, a linear diffuser.
Fig. 3. Such condensation may be highly visible, especially if glazed compo-

nents are a prominent feature of the design as is often the case with museums
or institutional natatoriums. Condensation on fenestration can also be a sig-
nicant problem in hospitals because hospitals typically have large numbers of
windows for corridors, patient rooms, etc. Skylights can be especially prob-
lematic, as they are often located far from dedicated heat sources such as ra-
diators or diffusers Fig. 4 and have lower-than-expected interior surface tem-
peratures and an increased condensation risk as a result. For natatoriums, the
lack of localized heating is often due to the skylight location directly over the
pool which creates both practical and aesthetic problems with running duct-
work, piping, or other services to the skylight location. The lack of localized
heat is exacerbated by the increased heat loss via radiation to the sky from
near horizontal components, which can lower the effective exterior surface
temperature by several degrees F below the ambient air temperature 10.
The net effect of these two issues is the skylight being subjected to lower inte-
rior and exterior temperatures than it was intended for and a reduction in
interior surface temperatures as compared to tested values. Although heat loss
to the sky is less of an issue at clerestory windows often used as a means of
providing natural lighting, distance from heat sources may still contribute to
lower localized temperatures in these locations. Solar heat gain, especially on
skylights, can be signicant enough to overcome low temperatures and conden-
sation risk, but the typically-constant interior humidity levels in high-humidity
buildings require a more constant and reliable means of keeping skylight and
clerestory window systems warm.
For most museums, the balance between the desire for natural lighting and
FIG. 5Schematic illustrating the mechanism of condensation on shaded window.
prevention of artifact degradation via ultraviolet radiation is difcult to main-

tain. In most cases, blackout shades are used to completely block all incident
radiation, as well as visible light, from entering the space when drawn. Al-
though fairly simple from a design and operation standpoint, blackout shades
can signicantly affect the condensation potential on windows and skylights.
The issue with most blackout shades is that they create a relatively captive
airspace microclimate between the glazing and the interior. This airspace acts
as an insulator, reducing heat ow from the interior and creating lower surface
temperatures on the glazing and framing. Since they are rarely completely air-
tight and may even be vapor permeable, the shades do not prevent interior
moisture from reaching the component Fig. 5. Since the surface temperatures
on the component will be greatly reduced as much as 10 F 5.6 C to 20 F
11.1 C in some cases, even small amounts of interior moisture that migrate
into the captive space can result in signicant condensation, or even ice forma-
tion. A small amount of interior air will also be trapped against the component
when the shade is initially closed, creating an additional source of condensa-
tion moisture. As an alternate means of lighting control, some museums will
utilize translucent laylights inboard of the skylights to diffuse natural light, or
to visually conceal shading or other mechanisms. Laylights and even vertical/
horizontal window blinds can create a similar effect as blackout shades, creat-
ing an insulating pocket of air that reduces skylight surface temperatures by
blocking heat from the interior.
Condensation in the eld of opaque walls or roofs may occur in framed
systems, which utilize the insulation between wood or metal studs, due to the
increased heat loss and reduced interior surface temperatures at the stud loca-
tions. For systems that utilize continuous insulation, which greatly reduces or
eliminates thermal bridging, surface condensation is much less of a concern
than on framed systems or fenestration components. Condensation on opaque
assemblies, regardless of insulation type, is more likely to occur in discrete
FIG. 62D thermal analysis results temperatures showing difference in window per-
formance with perimeter accessories installed. Interior and exterior conditions are
based on NFRC 100procedure for determining fenestration product U-factors 16.
areas such as corners, intersections, or window/door perimeters, where local

detailing may interrupt the continuity of the insulation and lead to reduced
interior surface temperatures.
Surface Condensation Solutions

Addressing condensation on fenestration components requires both the selec-
tion of appropriate systems and adequate detailing at the interface of those
systems with the surrounding construction components. Even for well-
designed, thermally broken framing systems, heat loss through perimeter com-
ponents can signicantly lower interior surface temperatures. While this reduc-
tion, which can vary signicantly depending on the adjacent construction, may
not be sufcient to cause problems in typical buildings, high-humidity condi-
tions create an environment when temperature drops of just a few degrees can
make the difference between satisfactory performance and damaging conden-
sation.
Although most fenestration manufacturers provide test data for their prod-
ucts, such data is typically inadequate for high-humidity applications because
it does not account for reductions in surface temperature due to adjacent con-
struction 11. Further, test reports may not include the effects of accessory
components such as window receptors, which can have a signicant impact on
product performance regardless of perimeter wall construction Fig. 6 12.
The window shown in Fig. 6 may be adequate for a 30 % RH space in the
tested condition, but the addition of perimeter window retainers produces a
signicant drop in surface temperatures enough so that ice may form on the
sill in some conditions. Although some but not all test reports will include
interior surface temperature data, this data only describes performance at a
single set of conditions, 0 F 17.8 C exterior and 70 F 21.1 C interior
for typical reports. An exterior temperature of 0 F 17.8 C is too conserva-
tive for much of the United States, and an interior temperature of 70 F
21.1 C is not representative of conditions within most natatoriums and even
some museums or archival storage facilities.
Designing fenestration for high-humidity buildings begins with the selec-
FIG. 72D thermal analysis results temperatures showing difference in performance

due to alignment with insulation. Interior and exterior conditions are based on NFRC
100procedure for determining fenestration product U-factors.
tion of high-performance systems based on available test data; although the test
data itself may not be sufcient, it provides a good starting point for a design.
Once a system is selected, both the system and the perimeter conditions which
are not addressed in the performance data should be analyzed using two-
dimensional 2D thermal analysis tools to estimate interior surface tempera-
tures based on the actual design conditions for the project. This type of analysis
provides more realistic estimates of system performance than product-only test
data.
The thermal performance of fenestration systems is almost always maxi-
mized by aligning the insulating component of the fenestration i.e., the insu-
lating glass units with the insulation in the building enclosure Fig. 7 12.
However, that detailing is still unlikely to bring the performance of the compo-
nent up to the same level as its tested performance. In the event that the se-
lected fenestration systems cannot by itself resist interior condensation, a
secondary means of condensation control will be necessary. This is typical in
natatoriums due to the extreme interior moisture conditions, but may also
occur in some museums in cold climates. Providing supplemental heat to fen-
estration systems in the form of warm air directed at the system can signi-
cantly reduce or eliminate interior condensation by providing both increased
local temperatures, increased heat delivery to the system via moving air, and
increased evaporation rates for condensation that may form on the surface
again, due to moving air. This approach requires careful coordination with
the project mechanical engineer, as duct layouts and system capacities need to
be designed accordingly to provide sufcient heating of the affected fenestra-
tion components Fig. 8.
Other options for supplemental heating include electric resistance heat
trace and radiant heating systems. Heat trace, although effective at delivering
heat directly to frame components where it is needed the most, is relatively
inefcient and may require periodic maintenance, making it more suited to
FIG. 8Linear slot diffuser intended to prevent condensation on clerestory windows.

Solid line indicates location of condensation, dashed line indicates area where airow
is effective at preventing condensation. Arrows highlight the transition between wet and
dry areas.
retrot applications where replacement fenestration or revised ductwork is not

practical 13. Further, it may be difcult to design heat trace systems capable
of meeting electrical or building codes, as few products are specically de-
signed or certied to function in this application.
Radiant heat systems may require signicant input of heat to effectively
warm all surfaces on the fenestration being heated. The primary benet of
radiant heat is that it does not require direct contact with the window like
heating cables or ductwork like air curtains, and can often be fed with a small
electric cable or hot water pipe, making satisfactory aesthetics easier to
achieve. Since radiative heat transfer requires the surfaces being heated to be
visible to the heat source, inverted or partially concealed surfaces may not
receive the same level of heat as surfaces with direct line-of-sight to the heater.
Although similar effects may occur with warm air systems, air will tend to ow
around components and can reach some but not all surfaces that are invis-
ible to radiant heat. The space requirements for radiant panels and the need
for line-of-sight make radiant heating more suited to new design than to retrot
applications.
The simplest way to address condensation due to shading devices or inte-
rior laylights is to provide heat directly to the space between the fenestration
component and the interior shade. As discussed above, this requires coordina-
tion with the mechanical engineer to provide the correct ductwork and system
capacities at or near the fenestration. Adding heat becomes difcult in areas
where the shades are extremely close to the fenestration, as there may not be
sufcient clearance to install supply air grilles or n-tube radiators in the space,
but is typically the only way to reduce the incidence of condensation other than
eliminating shading devices altogether or reducing the interior RH. If heat can-
not be provided, the use of shades should be minimized especially at night,
FIG. 9Basic mechanisms of condensationair leakage and vapor diffusion.
when condensation risk is the highest due to low exterior temperatures and
moisture-resistant nishes should be used in the vicinity of the shaded areas.
A last option for reducing condensation on glazed components is to mini-
mize their use in high-humidity areas. Although simple and cost effective, this
option is often at odds with the aesthetic intent of an architectural design and
may not be seen as a viable option by the design team.
For opaque walls and roofs, the use of continuous thermal insulation and
an effective air barrier is the best defense against surface condensation. Pre-
venting condensation at interface details or potential thermal bridges often
requires thermal analysis to estimate both interior surface temperatures and
the effectiveness of design solutions.
Concealed CondensationProblems
As previously discussed, concealed condensation in building enclosures may
occur through one of two mechanisms: Vapor diffusion or air leakage. Figure 9
illustrates these basic mechanisms of condensation within wall systems. Al-
though vapor diffusion is a relatively slow process in typical buildings, it can be
greatly accelerate in high-humidity buildings to increase the vapor drive from
the interior to the exterior. Problems may also occur in milder climates if an
adequate vapor control is not provided or a vapor retarder is placed on the
building exterior both common practices in these types of climates.
Condensation due to vapor diffusion is typically much less severe than that
caused by airow through the enclosure. At typical natatorium conditions dur-
ing the Winter, a relatively small air leak 5 cfm can theoretically transport up
to 1 gal 3.78 L of water through the building enclosure over the course of a
day. For buildings without continuous air barriers, leakage rates of several hun-
dred cfm are not uncommon resulting in massive amounts of water being
transported through the enclosure 14. Since the dew point of the airow will
be essentially the same as that of the interior air, comparing the interior con-
ditions to the annual exterior temperatures can provide a rough assessment of
condensation risk i.e., whenever the exterior temperature is below the interior
dew point, condensation due to airow is a potential risk. Figure 10 shows a
FIG. 10Comparison of exterior temperatures in New York and Miami and interior
dew points in a typical museum and natatorium.
comparison of dew points and exterior temperatures for New York, NY and
Tampa, FL, and demonstrates not only the severe risk associated with colder
climates but the moderate risks in warmer climates as well due to extreme
interior environments.
The risk of condensation due to multiple small airows can actually be
greater than that from large concentrated airows. This is due to the nature of
moving air, which carries heat as well as moisture. A large volume of warm air
moving through a wall may actually raise the surface temperatures of sur-
rounding components, reducing the incidence of condensation. However, large
airows still present a signicant problem in terms of heat losses or gains,
depending on the season.
In addition to air leakage though the exterior walls and roofs of a building,
internal airows can also be problematic in high-humidity buildings. This is
most common in natatoriums that are part of a larger athletic complex, and
include both humidied and non-humidied space, but also occurs in hospitals
where temperature, humidity, and air pressure differentials may exist between
adjacent rooms Fig. 11. Partition walls are often problematic, since designers
do not regularly design air and vapor barrier systems for interior walls. How-
ever, the lack of airow control through partition walls can lead to problems
ranging from elevated humidity levels in adjacent spaces which can lead to
condensation in the enclosure systems of those spaces or condensation on/
within the partitions.
A second form of internal airow that can lead to problems is the migration
of interior air to colder locations in the enclosure that may still be inboard of
the air barrier. The most common example of this is at parapets, which may be
FIG. 11Mechanism of condensation due to internal airows between spaces.
completely airtight but, despite insulation on the interior or exterior, may still
remain cold. In taller parapets, the upper portions of the construction may be
close to the exterior temperature due to the distance from interior heat sources
Fig. 12 12. Although airtight, normal convective currents within the parapet,
even relatively short parapets, may bring humid interior air into contact with
these surfaces, resulting in condensation if a suitable air barrier is not provided
Fig. 13 12.
The consequences of concealed condensation within walls can vary signi-
cantly, from simple staining to severe deterioration that goes unnoticed until a
component e.g., interior gypsum wallboard fails. On the exterior, eforescence
of masonry Fig. 14 due to increased wetting and, in extreme cases, ice forma-
tion Fig. 15 on the exterior of the building, can occur. While some of these
problems may occur in non-humidied buildings, they rarely occur with the
same speed and severity as in buildings with elevated interior moisture levels.
Concealed Condensation Solutions

Vapor diffusion can be controlled through the use of vapor retarders in the
exterior enclosure systems. It may be also possible to provide suitable vapor
ow resistance through the use of multiple materials, or thick layers of material
such as closed-cell spray-applied foam insulation as opposed to thin layers
such as plastic lm. When designing vapor control systems, both the interior
and exterior climates need to be taken into account, as moisture migration will
vary seasonally.
Controlling airow through the enclosure is simple in conceptprovide a
continuous air barrier in the building enclosurebut can be extremely difcult
in practice. The difculty stems not only from the complexity of modern build-
ings and the level of detailing required to provide an effective air barrier, but
also the incredibly slim margin of error in high-humidity buildings. An air
barrier that is 98 % effective may be more than adequate for a condominium
building but still allow for signicant airow-driven condensation in a natato-
rium. Air barriers are required in both exterior walls/roofs and interior parti-
tions that separate humidied from non-humidied space. In addition, design-
ers must carefully evaluate the building enclosure and identify areas where low
temperatures could occur inboard of the air barrier plane.
It is common practice in natatoriums to maintain the interior space at a
FIG. 122D thermal analysis results temperatures for a tall parapet. Top half of para-
pet is at or near exterior temperature. Interior temperature=70F 21.1C, exterior
temperature=0F 17.8C. Surface lms based on ASHRAE 2005 Handbook of Fun-
damentals, Chapter 25 Table 1.
negative pressure with respect to the exterior 1. This measure prevents moist
air from escaping through the building enclosure, instead causing exterior air
to be drawn in through any discontinuities in the enclosure. Although air inl-
tration will increase heating and cooling loads, that increase is less of a prob-
lem than widespread condensation within the building enclosure. Maintaining
negative pressure within a natatorium is only practical if the enclosure is fairly
FIG. 13Convective loops within wall cavity lead to condensation within cold parapet.
Diagonal line indicates dew point isotherm for interior conditions. Interior temperature
=70F 21.1C, exterior temperature=0F 17.8C. Surface lms based on ASHRAE
2005 Handbook of Fundamentals, Chapter 25 Table 1.
airtight. The leakier the enclosure, the greater the amount of mechanical
exhaust necessary to maintain negative pressure and the greater the risk of
localized condensation due to the cooling effect of incoming exterior air. After
a certain point, maintaining negative pressure becomes impractical due to ei-
ther equipment capacity limitations or excessive added heating/cooling loads.
The mechanical system must also be designed to overcome stack pressure
in tall spaces typical competition natatoriums have ceiling heights of 30 ft
9.14 m or greater. If the mechanical system is only designed to maintain
neutral or negative pressure at the pool deck level, the actual pressure at the
FIG. 14Eforescence on exterior masonry due to airow-driven condensation in

brick cavity.
underside of the roof may be positive to the buoyancy of warm interior air 15.
The relationship between a tight building enclosure and the ability of the me-
chanical system to maintain negative pressure underscores the close coordina-
tion required to achieve a working design.
As previously discussed, maintaining negative pressure is typically not an
option in museums due to contaminant control issues. Similarly, contaminant
control or patient protection in hospitals may control pressure balances, mak-
ing negative pressure not an option in all areas. In these cases, effective air
barriers and moisture-tolerant materials must be used to control airow-driven
condensation.
FIG. 15Severe icing at roof eaves due to airow-driven condensation at roof-to-wall

interface.
FIG. 16Condensation on original exterior windows due to air leakage past interior
storm window.
Retrotting Existing Buildings

A common type of building retrot is the upgrading of an existing high-
humidity building to address condensation concerns. In this type of retrot, the
options available to the designer are limited by several factors, including bud-
get e.g., full replacement of windows in a 2 year old building will rarely be
considered a viable option and existing construction/geometry constraints. So-
lutions to condensation problems often include the addition of remedial heat
trace systems on fenestration, addition of interior insulation or barriers, modi-
cation of mechanical systems to maintain negative pressure, and in extreme
cases, reduction in interior humidity levels if no other options can be practi-
cally implemented. The latter choice, maintaining lower RH levels, almost al-
ways provides the best combination of lower rst-cost, lower operating cost,
and reduced condensation risk. However, lower RH levels may not be practical
or acceptable based on the use of the particular building.
The addition of exterior insulation or barriers is complicated by the need to
remove roong or cladding systemsinterior gypsum wallboard is much easier
to remove and replace than brick veneer or metal panels. Interior storm win-
dows are often considered as a means of reducing condensation on existing
windows. However, as with shading devices on windows, storm windows may
end up causing more problems than they solve if they are not effective at pre-
venting moisture from reaching the existing window which is now much
colder due to the insulating effect of the storm window Fig. 16.
Reducing humidity levels is often seen as a last resort, especially in muse-
ums or some hospital areas where humidity is required for artifact preservation
or occupant health. Natatoriums may also have difculty lowering interior RH
levels due to issues of occupant comfort; lower RH levels tend to increase sur-
face evaporation from swimmers skin, consequently increasing evaporative
cooling. Further, the signicant moisture load created by the pool itself makes
dehumidication to low levels extremely energy-intensive. Due to the limita-

tions of existing buildings, retrot systems rarely address condensation prob-
lems completely, and are likely to require some type of ongoing maintenance or
monitoring to remain effective e.g., sealant-based repairs, electric heat trace.
A second type of retrot is the addition of humidication to an existing
building. This can have disastrous consequences if the building enclosure is not
carefully evaluated during the process to determine either maximum safe RH
levels or appropriate building enclosure upgrades to accommodate the increase
in interior moisture levels. Unfortunately, adding humidication is a relatively
simple task as compared to upgrading the building enclosure to accommodate
increased RH levels. As discussed above, retrot options may be limited by
existing building geometries as well as project budgets. When considering the
humidication of an existing building, designers need to carefully consider the
level of upgrade necessary to maintain the integrity of the building itself. Such
an analysis may reveal that the available funding is insufcient to bring the
existing systems up to an acceptable level of performance. With sufcient fund-
ing, heat trace systems, insulation, new air/vapor barriers, and other upgrades
can be effective at preventing condensation as long as they take into account
the existing building enclosure construction and new interior RH levels.
The special case of converting an existing, solid masonry building to an art
museum requires particular attention. In many cases, the exterior walls of the
building cannot be insulated due to the risk of accelerating freeze-thaw damage
to the masonry or trapping moisture within the interior nishes. In these cases,
condensation may occur in the eld of the wall i.e., widespread throughout the
building due to the relative lack of thermal resistance in the masonry. Win-
dows and other components become difcult to detail since there is no insula-
tion with which to align them. The lack of insulation contributes to increased
heat loss at fenestration perimeters.
For opaque walls, a ventilated return air wall strategy can be used to ow
interior air over the walls on its way to a return air plenum for the mechanical
system. The air within the cavity is maintained at a negative pressure with
respect to both the interior and exterior. This eliminates the ow of moist inte-
rior air to cold locations within the enclosure, while air that may be drawn into
the building is immediately captured by the return airstream and delivered to
the mechanical system for ltering. A well-detailed, continuous air barrier is
necessary in the walls to minimize the amount of inltration across the wall
and allow for closer control over interior air pressures, temperatures, and RH
levels. At the same time, positive pressure can be maintained in the occupied
space, so that exterior air does not ow into the space when doors, etc. are
opened. The moving air delivers more heat to the walls, preventing condensa-
tion even if it may occur under stagnant conditions Fig. 17. In the example
shown in Fig. 17, insulation was not added to the wall due to the risk of accel-
erating freeze-thaw damage throughout the thickness of the wall although
this is a fairly energy-intensive solution, adding insulation to the wall would
likely have short term benets only and lead to advanced degradation of the
building enclosure. Careful design is necessary to ensure sufcient airow as
well as consistent negative pressure within the cavity. Detailing at windows is
difcult, as it requires routing of the return air around the openings. For this
FIG. 17Schematic diagram of return wall system for solid masonry retrot.
reason, mockup testing is often required to ne tune such a system and validate
the performance of the built assembly.
Summary
The design of high-humidity buildings, especially in colder climates, inevitably
requires the selection of building enclosure systems that offer higher perfor-
mance and, by nature, cost more than the types of systems included in general
use buildings. This is especially problematic in warmer climates, where the
additional expense of high-performance systems may not be seen as justiable
by designers or owners who are not familiar with the potential problems asso-
ciated with high-humidity buildings. Despite these perceptions, high-
performance systems and more detailed design and analysis are absolutely es-
sential to preventing condensation and the associated problems of degradation,
loss-of-use, and damage to a facilitys reputation. This is especially true in mu-
seums, where even a small amount of condensation in the wrong place could
cause irreparable damage to a priceless artifact.
The following is a summary of the design steps necessary for addressing
condensation in high-humidity buildings:
As a general rule, avoid the use of moisture-sensitive nishes in areas
that may experience condensation. Incidental condensation may be ac-
ceptable in some locations if it will not cause damage to the surrounding
construction.
Design continuous insulation systems for opaque walls and roofs, and
align insulation with fenestration components to maximize their effec-
tiveness.
Select high-performance fenestration systems and analyze in-place per-
formance using thermal analysis tools.
Provide supplemental heating systems where necessary to boost fen-
estration performance, including systems to provide heat between fen-
estration and shading devices or laylights.
Design appropriate air and vapor barrier systems to address moisture

migration both by diffusion and via airow.
In retrot applications where an existing building is humidied, care-
fully evaluate existing conditions and design appropriate measures to
address both surface and concealed condensation within the enclosure.
References
1 ASHRAE, Natatoriums, ASHRAE Handbook of Applications, Chap. 4, ASHRAE,

Altanta, GA, 2007 sec. 4.6..
2 ASHRAE, Museums, Galleries, Archives, and Libraries, ASHRAE Handbook of
Applications, Chap. 21, ASHRAE, Altanta, GA, 2007.
3 ASHRAE, The ASHRAE Guide for Buildings in Hot and Humid Climates, Chaps. 15
ASHRAE, Atlanta, GA, 2008.
4 ASHRAE, Ice Rinks,ASHRAE Handbook of Applications, Chap. 4, ASHRAE, At-
lanta, GA, 2007 sec. 4.5..
5 ASHRAE, Refrigerated Facility Design/Loads, ASHRAE Handbook of Refrigera-
tion, Chaps. 2324, ASHRAE, Atlanta, GA, 2010.
6 Nelson, P. E., Condensation Problems in a Refrigerated Storage BuildingA Case
History, Water Vapor Transmission Through Building Materials and Systems,
Mechanisms and Measurement, ASTM STP 1039, H. R. Treschel and M. Bomberg,
Eds., ASTM, West Conshohocken, PA, 1989, pp. 5160.
7 Mecklenburg, M. F., Tumosa, C. S., and Pride, A., Preserving Legacy Buildings,
ASHRAE J., Vol. 46, 2004, pp. S18S23.
8 Wood, R. K., 2006, Determining Correct Temperature and Relative Humidity
Ranges for Historic House Museums, Thesis for Master of Arts in Historic Pres-
ervation, Goucher College, Baltimore, MD.
9 ASHRAE, Health Care Facilities, ASHRAE Handbook of Applications, Chap. 7,
ASHRAE, Atlanta, GA, 2007.
10 Hagentoft, C., Introduction to Building Physics, Studentlitteratur AB, Lund, Swit-
zerland, 2001, pp. 5356.
11 OBrien, S. M., Finding a Better Measure of Fenestration Performance: An Analy-
sis of the AAMA Condensation Resistance Factor, RCI Interface, May, 2005, pp.
410.
12 Lawrence Berkeley National Laboratory, Thermal Analysis Performed Using the
Therm 5.2 and Window 5.2 Computer Programs, http://windows.lbl.gov/software/
default.htm Last accessed August 2010.
13 Lyon, E. G., IBP Real World Glazing and Window Performance Designing to
Mitigate Condensation, Research in Building Physics and Building Engineering,
2006, Taylor & Francis, London.
14 ASHRAE, Commercial and Institutional Air Leakage, ASHRAE Handbook of
Fundamentals, Chap. 16, ASHRAE, Atlanta, GA, 2009 sec. 16.25.
15 ASHRAE, Stack Pressure, ASHRAE Handbook of Fundamentals, Chap. 16,
ASHRAE, Atlanta, GA, 2009 sec. 16.6.
16 NFRC 100, 2001, Procedures for Determining Fenestration Product U-Factors,
National Fenestration Rating Council NFRC, Greenbelt, MD.
doi:10.1520/JTE102983
Elizabeth Ordner1
Fenestration Condensation Resistance:

Computer Simulation and In Situ
Performance
ABSTRACT: The purpose of this paper is to examine condensation rating

procedures, specically the procedures as dened by the National Fenestra-
tion Rating Council NFRC, and to compare the standardized procedure
with in situ performance data. The American Architectural Manufacturers As-
sociation AAMA was the rst to develop standardized methods to compare
fenestration system performance with respect to the formation of condensa-
tion. Following this, the NFRC developed its own procedure for developing
condensation resistance values. The AAMA test method involves laboratory
testing, while the NFRC method involves computer simulation. Both methods
provide a rating system to enable relative ease in performance comparison
between similar fenestration products. The ratings are based on standard-
ized conditions i.e., fenestration size and shape, fenestration surround, and
environmental conditions that do not account for the in situ conditions and
therefore may not accurately predict actual performance. Recent literature
has discussed the limitations of condensation resistance ratings relative to
as-built conditions that can impact the formation of condensation on a fen-
estration system. For example, 1 installing spandrel panel insulation, al-
though necessary to improve the thermal performance of a curtain wall, can
lower surface temperatures of adjacent framing members; 2 existing inte-
rior blinds or drapery act as an insulator, which can cause the framing mem-
bers to be colder; and 3 if the fenestrations thermal break is not aligned
with the walls insulation, the framing members can be subject to colder
temperatures. This paper will focus on the NFRC rating procedure and ex-
amine a case study to demonstrate the differences between NFRC ratings,
simulated surface temperatures using THERM and WINDOW computer soft-
1
Wiss, Janney, Elstner Associates, 10 South LaSalle, Suite 2600, Chicago, IL 60603.
Cite as: Ordner, E., Fenestration Condensation Resistance: Computer Simulation and
In Situ Performance, J. Test. Eval., Vol. 39, No. 2. doi:10.1520/JTE102983.
269
ware, and in situ measurements of a high-rise ofce building curtain wall.

The case study will provide a representative illustration of how actual condi-
tions can impact the fenestration products in situ condensation resistance
and provide insight on the condensation resistance of a representative cur-
tain wall system.
KEYWORDS: condensation, CR rating, fenestration, THERM
Condensation Resistance Background
The American Architectural Manufacturers Association AAMA and the Na-

tional Fenestration Rating Council NFRC have developed standardized meth-
ods to compare fenestration system performance with respect to the formation
of condensation. The AAMA test method involves laboratory testing, while the
NFRC method involves computer simulation. Both methods provide a rating
system to enable relative ease in performance comparison between similar fen-
estration products. Neither method takes into account installation, congura-
tion shape and glass/frame arrangement, size, or environmental conditions of
in-service products.
AAMA Publication 1502.3-1972 1 was developed in 1972 to evaluate and
compare the condensation resistance CR of thermally improved aluminum
fenestration products, and AAMA Publication 1503.1-1980 2 was instituted in
1980 to measure thermal transmittance. The current voluntary test method is
AAMA Publication 1503-09, Voluntary Test Method for Thermal Transmittance
and Condensation Resistance of Windows, Doors and Glazed Wall Sections
3, which involves placing a standard conguration of a specimen into a test
chamber designed to maintain a constant temperature differential. A CR factor
CRF is calculated using the lower of either the measured average glazing
temperatures or the weighted frame temperatures. The CRF is a numerical
index generally in the range of 3580, with the larger the CRF number, the
greater the resistance to condensation. The AAMA provides guide-line CRF val-
ues for various climatic regions and interior relative humidity RH levels. For
example, in Chicago, for a typical ofce building with a relatively low RH of 20
%, AAMA guidelines suggest a minimum CRF of 43.
The NFRC 500 procedure was developed in 2002 and uses computer simu-
lation to calculate CR values. The NFRC 500-2010, Procedure for Determining
Fenestration Product Condensation Resistance Values 4, is based on NFRC-
approved software tools to calculate a CR rating concurrently with U-factor,
solar heat gain coefcient, and visible transmittance. Like the AAMA approach,
the CR rating is based on standardized test conditions. The CR rating is a value
that considers the relative area of condensation at three RH levels 30 %, 50 %,
and 70 % and the degree to which the frame and glazing surface temperatures
are below the dew point. The CR rating is based on the lower of the frame,
edge-of-glazing, or center-of-glazing values and is reported on a scale of 1100,
with a higher number indicating a better resistance to condensation. Currently,
NFRC certied products have ratings that vary from about 10 to 61 5. The
NFRC has not published guidelines on the use of CR ratings, and therefore it is
ORDNER, doi:10.1520/JTE102983 271
difcult for a specier to know what an appropriate CR rating for a specic

climate is.
CRF and CR ratings serve as a means to select fenestration products based
on their ability to resist the formation of condensation. CRF and CR ratings are
not interchangeable, and no method currently exists to convert a CRF value to
a CR value or vice versa. Because of the standardized parameters to develop CR
ratings that do not reect variable and sometimes unpredictable in situ condi-
tions, CRF and CR ratings merely provide a consistent means of fenestration
product comparison and should not be used to predict if interior surface con-
densation will occur.
NFRC Simulation Parameters2
The NFRC procedure enables comparisons between fenestration products by

using standardized simulation parameters. There are several limitations to the
standardizations and rating procedure.
1 The use of steady-state and standardized environmental conditions not
reecting reality
2 The use of standardized fenestration sizes and congurations not re-
ecting actual sizes and congurations
3 The neglect of heat transfer between the fenestration products and the
surround
The following is a discussion of the standardizations for NFRC CR ratings
and how the standardizations may be different from in situ conditions, thereby
potentially affecting thermal performance.
Environmental Conditions
Since both temperature and surface air lm conditions affect results, the NFRC
500 simulation procedure species the standardized boundary conditions
shown in Table 1 above and as referenced in NFRC 100-2010, Procedure for
Determining Fenestration Product U-Factors 6.
The standardized air temperatures and lm coefcients, though necessary
for a comparison standard, rarely reect in-service conditions, as temperatures
and wind speeds are not steady-state and only perchance meet
18.0 C 0.4 F outside and 21.0 C 69.8 F inside. Localized interior air
temperatures adjacent to a fenestration product are affected by heat register
placement, window treatments, and furniture placement. Heating systems are
generally designed to warm the interior surfaces of the fenestration product. If
the heat source is inappropriately placed or blocked, the fenestration surface
temperatures can signicantly decrease. Furniture and window treatments can
2
Note that the standardized NFRC simulation parameters are similar to the AAMA test
method procedures as described in the paragraph above, but for the purposes of this
paper, only the NFRC parameters are discussed.
TABLE 1Standardized boundary conditions for NFRC simulations.
Convective Film Coefcient, W / m2 K

Boundary Condition Temperature, C F Btu/ h ft2 F
NFRC 100-2001 exterior 18.0 0.4 26 4.578 5.5 m / s 12.3 mph wind speed
Interior aluminum frame convection only 21.0 69.8 3.29 0.579
Interior thermally broken frame convection
only 21.0 69.8 3.0 0.528
Interior thermally improved frame convection
only 21.0 69.8 3.12 0.549
Interior wood/vinyl frame convection only 21.0 69.8 2.44 0.429
ORDNER, doi:10.1520/JTE102983 273
TABLE 2Standardized fenestration sizes and congurations for NFRC simulations.
Opening (X) and Size (width x height)

Product Type Configuration
Non-Operating (O) mm (in.)
1200 x 1500
Casement- Double XX
(47 x 59)
1200 x 1500
Fixed O
(47 x 59)
2000 x 2000
Glazed Wall OO
(79 x 79)
1500 x 1200
Horiz. Slider XO or XX
(59 x 47)
Projecting (Awning, 1500 x 1200

XX
Dual) (59 x 47)
1200 x 1500
Vertical Slider XO or XX
(47 x 59)
act as insulators or prevent heat from reaching the fenestration, also decreasing
the surface temperatures. Wherever surface temperatures are decreased, the
potential for condensation increases.
Fenestration Size and Conguration

The NFRC 500 procedure species that the simulation be performed with a
specic conguration and size, as referenced in NFRC 100. Table 2 below illus-
trates a few common fenestration products.
Fenestration products come in all shapes, sizes, and congurations, which
affect CR ratings since the ratings are established by the following equation 4:
CR ratingframe, edge of glass, or center of glass = 1

Abelow dew point temperature at frame, edge of glass, or center of glass
Aframe, edge of glass,or center of glass/Atotal

1/3
100
The nal CR rating is the lowest value of the CR rating of the frame, CR rating
of the edge of glass, and CR rating of the center of glass. Reducing the area of
the glass typically decreases the CR rating because the frame is a larger per-
centage of the total area, and the frame typically has the lowest CR of the
system. Therefore, the simulation typically is not an accurate representation of
an in situ fenestration product.
Heat Transfer Between Fenestration Products and Surround

The NFRC simulation ignores heat transfer that occurs between the fenestra-
tion product and the surrounding wall system. This is illustrated in Figs 1 and
2.
FIG. 1NFRC simulation with no heat transfer between frame and surround solid
black line=plane of zero heat transfer.
In situ fenestration products are installed in a wide variety of openings, and

the heat transfer that occurs between the fenestration product and the sur-
round can signicantly affect the thermal performance of the fenestration
product. Consider the effects of a window head being directly connected to a
thermal bridging element, such as a steel lintel, which will cause the surface
temperatures of the window frame to decrease. Another example of a common
installation issue is when wall insulation is installed further inboard than the
thermal break of the window frame, which will cause the interior surface tem-
peratures of the window to be signicantly colder 7,8. These are installation
issues that can decrease surface temperatures, but are not considered in NFRC
rating procedures.
A necessary component of a curtain wall systems thermal performance, yet
also not considered in NFRC rating procedures, is spandrel insulation. In the
FIG. 2In situ simulation with heat transfer occurring between frame and surround.
Note that due to misalignment of thermal planes, the interior side of the frame is colder
than when surround is not modeled solid black line=plane of zero heat transfer.
ORDNER, doi:10.1520/JTE102983 275
case of a curtain wall, for the purposes of NFRC simulation procedures, the
vision lite is the fenestration product and the insulated spandrel panel is the
surround. Spandrel insulation moves the thermal plane of the curtain wall
inward, thereby lowering interior surface temperatures of framing members
adjacent to the insulation. The heat transfer between the surround and the
fenestration product is not considered in the NFRC rating procedure; hence,
the rating procedure does not provide an accurate representation of in situ CR.
Case Study
A case study illustrates the differences between

1 NFRC CR ratings,
2 computer simulations of the in situ construction, and
3 measurement of in situ performance.
These three studies are identied as Method 1, Method 2, and Method 3.
THERM 5.2 and WINDOW 5.2 computer software programs were used to com-
pare NFRC CR ratings Method 1 with a simulation of an in situ high-rise
ofce building curtain wall Method 2. The computer simulated surface tem-
peratures Method 2 were then compared with measured in situ temperatures
for the curtain wall Method 3. Interior conditions were varied in Method 3 to
demonstrate how actual conditions affect the thermal performance of the cur-
tain wall.
Description of Software
NFRC-approved computer software includes WINDOW and THERM 9, which
were developed by the Lawrence Berkeley National Laboratory. THERM incor-
porates the nite element method to analyze two-dimensional, steady-state
heat transfer through a cross section. THERM is used to calculate frame and
edge of glass U-factors and can be used to demonstrate temperature and heat
ow patterns. WINDOW is used in conjunction with THERM to calculate cen-
ter of glass U-factors and total product U-factors, solar heat gain coefcients,
visible transmittance, and CR ratings.
Description of Curtain Wall

The subject curtain wall encloses the fth through 37th oors of a downtown
Chicago ofce building. The curtain wall consists of thermally improved
painted aluminum framing with insulated glazing IG units at the vision lites,
a single pane of glass at the spandrels + / 1 m 3 ft above and below the oor
slab, and painted aluminum panels glazed into the framing at the columns and
oor slabs. The aluminum framing consists of a hollow tube with screw spline,
thermal break, pressure plate, and snap cover. The IG units consist of a 6 mm
1/4 in. blue-tinted outer lite with stainless steel reective coating on surface 2,
12 mm 1/2 in. air space, and 6 mm 1/4 in. clear inner lite. The spandrel glass
consists of a 6 mm 1/4 in. blue-tinted lite with stainless steel reective coating
FIG. 3Overall exterior view of curtain wall.
on surface 2. Both the glass spandrels and the aluminum panels are insulated
with 100 mm 4 in. of mineral wool insulation with a galvanized steel back
pan.
An overall exterior view of the curtain wall is shown in Fig. 3. Curtain wall
shop drawings are included in Fig. 4.
Method 1: National Fenestration Rating Council Simulation

NFRC 100 states that curtain walls are simulated as two lites with one vertical
mullion. The jambs and the mullion are modeled as intermediate vertical
frames, and the head and sill are modeled as intermediate horizontal frames.
The standardized curtain wall size is 2000 2000 mm2 79 79 in.2. The
spandrel panels, slab covers, and column covers are not modeled.
The NFRC simulation of the high-rise ofce building curtain wall provides
a CR rating of 45 using WINDOW and THERM softwares. The interior surface
temperatures associated with a CR rating of 45 for the simulated curtain wall
are shown on the left cross sections in Fig. 7.
Method 2: Computer Simulation of In Situ Construction and Comparison to

Method 1
The curtain wall was modeled using THERM with the in situ surrounding con-
ditions of the spandrel glass and column covers and with the effects of closed
blinds inboard the curtain wall. The spandrel was modeled per the curtain wall
shop drawings by removing the IG unit in one vision lite of the Method 1 NFRC
model, revising the frame to reect in situ conditions at the spandrel, and
adding the spandrel glass, insulation, spandrel back pan, and interior drywall
sheathing, as shown in Fig. 5. The closed blinds were modeled generically by
adding a 1 / 16 in. sheet of vinyl at the plane of the interior surface of the
FIG. 4Curtain wall shop drawings. Sill detail is on left and jamb detail is on right.
ORDNER, doi:10.1520/JTE102983 277
FIG. 5THERM model of in situ spandrel and blinds conditions.
window frame in THERM, as shown in Fig. 5. Such modeling of the blinds is a

conservative approach to the simulation since the in situ blinds are not a con-
tinuous sheet rather, horizontal-slatted vinyl mini-blinds with the ability to
allow a small amount of interior air to bypass the blinds. The simulation is for
comparative purposes between Method 1 and Method 2, and the effects of the
blinds will be further examined when measuring in situ temperatures for the
curtain wall Method 3.
The resulting interior surface temperatures from the Method 2 model were
then compared to those of the Method 1 NFRC simulation, as shown in Fig. 7.
At 18.0 C 0.4 F exterior and 21.0 C 69.8 F interior temperatures, the
following results were found.
Interior curtain wall frame surface temperatures simulated in Method 2
were, on average, 5 F lower than that of Method 1 due to the effects
of the spandrel and column cover insulation.
When including the effects of closed blinds, the interior surface tem-
peratures decrease another 7 8 F at the frame and 16 18 F at the
center of glass.
By including the effects of the surrounding enclosure and lowering the
blinds alone, the interior surface temperatures decrease by 10 20 F.
To predict surface temperatures at various exterior and interior tempera-
tures, THERM simulations were also performed at 0, 10, 20, and 30 F exterior
and 65, 70 and 75 F interior temperatures. At each temperature, interior sur-
face temperatures were recorded at specic locations corresponding to the ther-
mocouple TC locations for the in situ measurements described later in this
paper for Method 3. A linear relationship between interior surface tempera-
tures and interior and exterior ambient temperatures was established, and
equations were generated to calculate the interior surface temperature at a
given location for given interior and exterior temperatures. An example for TC
location 1 is shown in Fig. 6.
ORDNER, doi:10.1520/JTE102983 279
FIG. 6Example of equation generated for TC 1, showing relationship of surface tem-

perature at TC 1 to interior and exterior temperatures.
Method 3: Measurement of In Situ Performance with Variations of In Situ

Conditions and Comparison to Method 2
To compare simulated performance Method 2 to in situ performance Method
3 and to demonstrate the effects of various conditions south or north facing
and presence of diffuser and/or window blinds, TCs were placed on the interior
surfaces of two lites of the high-rise ofce building to measure actual surface
temperatures over time. Eight TCs were placed on one vision lite on the north
facade, and eight TCs were placed on one vision lite on the south facade, as
shown in Fig. 8. The curtain wall specimens are located on the 26th oor of the
high-rise ofce building. The vision lites are 1470 1470 mm2 58
58 in.2. The north curtain wall lite is located at the corner, and the south
curtain wall lite is located approximately in the center of the building.
Horizontal-slatted vinyl mini-blinds existed at both windows.
At both test specimens, a supply air diffuser existed adjacent to the curtain
wall head: At the north, the diffuser was centered on the adjacent lite of glass;
and at the south, the diffuser was centered on the lite being measured. The
supply air temperature is 32 38 C 90 100 F. A smoke pencil was used to
determine the supply air wash on the specimen, which was found to spread
outward to the left/right 457 mm 18 in., outward away from the specimen
610 mm 24 in., and downward 914 mm 36in.. The supply air fully dif-
fused with room ambient air before it reached the sill.
Interior surface temperature data from two specimens was collected to
allow comparison between the effects of a diffuser adjacent to the head at the
south curtain wall and with the blinds open and closed. TC temperature read-
ings were recorded one to two times per day between Dec. 7, 2009, and Jan. 8,
2010. Temperature data recorded when the sun was shining was ignored be-
cause the interior surface temperatures on the south curtain wall were in-
FIG. 7Comparison of NFRC simulation Method 1, in situ simulation with blinds

open Method 2, and in situ simulation with blinds closed Method 2.
creased by 4 C 25 F or more. Exterior temperatures recorded from Na-

tional Weather Service data from a nearby location ranged between 16 1 C
2 34 F and averaged 5.5 C 22 F. Interior air temperatures were mea-
sured with a hand-held temperature meter and in both the north and south
ofces ranged from 19 24 C 66 75 F and averaged 21.6 C 71 F at the
north ofce and 21.1 C 70 F at the south ofce.
Table 3 is a summary of the Method 2 and 3 data, comparing the north and
south curtain walls with the blinds open and closed. For each condition, the
table includes the Method 3 average measured surface temperatures Actual
Tave, average simulated surface temperatures which was calculated as previ-
ously noted in Method 2 THERM Tave, and the difference between the aver-
age measured and simulated surface temperatures at each TC location.
ORDNER, doi:10.1520/JTE102983 281
FIG. 8TC locations on in situ curtain wall.
When comparing the north and south curtain wall with blinds open and
closed, the following results were found.
The south curtain wall with the adjacent diffuser was 4F warmer at
the head and 3F colder at the sill than the north curtain wall with the
blinds open. As expected, the head at the south curtain wall was warmer
than the north curtain wall due to the closer proximity of the diffuser at
the south curtain wall washing more warm air on the upper surfaces.
Closing the blinds on the north curtain wall lowered the surface tem-
peratures at the sill by 2 3F at the frame and edge of glass and 4F at
the center of glass. Surface temperatures were slightly higher or un-
changed at the head.
Closing the blinds on the south curtain wall increased the frame surface
temperatures by 8 14F at the frame and edge of glass and 68F at
the center of glass. The increase in temperatures is due to incidental
solar radiation warming the air gap between the blinds and the glass. To
support this conclusion, readings were taken at night with the blinds
closed, and similar to the north curtain wall, the sill surface tempera-
tures were 3 4F lower than with the blinds open.
This data demonstrates the potential effect that diffuser placement and the
presence of window treatments have on the curtain wall. It is difcult to pre-
cisely determine the potential change in surface temperatures in the case study
due to the dynamic nature of the supply air supply air from the diffuser at the
adjacent lite had a warming effect on the north curtain wall head, and a curtain
wall specimen without supply air wash was not measured, the supply air by-
passing the vinyl blinds, and the effect of solar radiation when the blinds were
closed surface temperature readings were recorded during daylight hours
only; however, we can conclude that surface temperatures are indeed affected
by changes in surrounding environmental conditions.
When comparing the measured temperatures Method 3 to the predicted
temperatures Method 2, the following results were found.
TABLE 3Comparison of average measured surface temperatures Method 3, average simulated surface temperatures Method 2, and the
difference between the average measured Method 3 and simulated Method 2 temperatures for the north and south curtain wall with the
blinds open and closed.
North Curtain Wall
Blinds Open Blinds Closed
Actual Tave. THERM Tave. Actual Tave. THERM Tave.

TC Location C F C F C F C F C F C F
1 Corner of glass upper 9.2 48.5 6.9 44.5 2.3 4.0 9.5 49.1 4.4 39.9 5.1 9.1
2 Head 11.4 52.8 8.6 47.5 2.8 5.3 11.8 53.3 6.6 43.9 5.2 9.4
3 Jamb upper 8.1 46.6 7.1 44.7 1.0 1.9 9.2 48.6 5.4 41.7 3.8 6.9
4 Center of glass upper 13.3 55.9 13.3 55.9 0.0 0.0 11.3 52.4 6.7 44.0 4.6 8.3
5 Jamb lower 7.1 44.8 7.4 45.3 0.3 0.5 7.4 45.3 5.4 41.7 2.0 3.7
6 Center of glass lower 12.4 54.4 13.5 56.3 1.1 1.9 10.3 50.6 6.7 44.0 3.6 6.5
7 Corner of glass lower 5.8 42.5 7.3 45.1 1.5 2.6 4.7 40.5 4.5 40.1 0.2 0.4
8 Sill 8.7 47.7 9.1 48.3 0.4 0.6 7.2 45.0 6.7 44.1 0.5 0.9
South Curtain Wall

1 Corner of glass upper 11.1 52.0 8.7 47.7 2.4 4.4 17.1 62.8 4.6 40.2 12.5 22.6
2 Head 13.7 56.6 10.2 50.3 3.5 6.3 17.9 64.3 6.8 44.2 11.1 20.0
3 Jamb upper 8.1 46.6 8.8 47.9 0.7 1.3 15.9 60.6 5.6 42.0 10.3 18.6
4 Center of glass upper 14.6 58.3 14.3 57.7 0.3 0.6 19.1 66.3 6.9 44.4 12.2 21.9
5 Jamb lower 8.8 47.8 8.3 47.0 0.5 0.8 14.4 58.0 5.6 42.0 8.8 16.0
6 Center of glass lower 9.8 49.7 13.9 57.1 4.1 7.4 13.0 55.4 6.9 44.4 6.1 11.0
7 Corner of glass lower 6.9 44.4 8.3 46.9 1.4 2.6 12.8 55.1 4.7 40.4 8.1 14.6
TABLE 3 Continued.
North Curtain Wall

8 Sill 6.9 44.5 9.9 49.8 3.0 5.3 12.8 55.1 6.9 44.4 5.9 10.7
ORDNER, doi:10.1520/JTE102983 283
Generally, the interior surface temperatures on the test specimen were

within 5 F of the temperatures from the simulation when the blinds
were open. This data demonstrates that THERM simulated surface tem-
peratures provide a fairly accurate representation of in situ surface tem-
peratures for the representative case study. It should be noted that inte-
rior surface temperatures on the test specimen were recorded during
daylight hours, and although overcast, incidental solar radiation will
likely skew the surface temperatures slightly in the warm direction.
The test specimen was generally warmer than the simulation towards
the head of the curtain wall and lower than the simulation towards the
sill of the curtain wall when the blinds were open. This is a result of the
natural convection currents causing air to wash downward over the
product. Additionally, the head was generally warmer due to the supply
air diffuser placement adjacent to the head, and the sill was generally
colder possibly due to furniture placement adjacent to the sill.
THERM predicted that the surface temperatures would decrease more
signicantly 7 8 F at the frame and 16 18 F at the center of glass
compared to 2 4 F actual when the blinds were closed. This can be
attributed to the fact that THERM does not account for any interior air
circulation behind the blinds, to the diffuser placement adjacent to the
curtain wall, and/or to solar radiation effects.
Conclusion
The NFRC 500 simulation method used to determine CR ratings is based on an
individual fenestration product of a standard conguration and size, simulated
in an opening that neglects heat transfer between the surround and the fenes-
tration product, with steady-state interior and exterior temperature conditions.
These CR ratings allow the ability to select a fenestration product based on its
ability to resist condensation as compared to another product. However, be-
cause various project-specic installation details and environmental conditions
are not simulated, CR ratings should not be used to predict if condensation will
actually occur after installation.
The case study presented demonstrates that when in situ conditions i.e.,
fenestration surround and environmental conditions are taken into account,
interior surface temperatures can be signicantly lower 20 F for this case
study than the surface temperatures used to calculate CR ratings. Therefore,
understanding how installation details affect a fenestration products thermal
performance, in combination with the rating methods discussed, is necessary
to help avoid the formation of condensation. In some cases where CR is critical
or where environmental conditions are more severe, a computer simulation
modeling actual installation conditions or a test mockup may be necessary to
verify the performance of fenestration products. In all cases, good judgment in
designing and specifying fenestration products/assemblies can help prevent
condensation problems.
References
1 AAMA Publication 1502.3-1972, 1972, Voluntary Test Method for Condensation

ORDNER, doi:10.1520/JTE102983 285
Resistance of Windows, Doors, and Glazed Wall Sections, American Architectural

Manufactures Association, Palatine, IL.
2 AAMA Publication 1503.1-1980, 1980, Voluntary Test Method for Thermal Trans-
mittance of Windows, Doors, and Glazed Wall Sections, American Architectural
Manufactures Association, Palatine, IL.
3 AAMA Publication 1503-09, 2009, Voluntary Test Method for Thermal Transmit-
tance and Condensation Resistance of Windows, Doors, and Glazed Wall Sec-
tions, American Architectural Manufactures Association, Palatine, IL.
4 NFRC 500-2010, 2010, Procedure for Determining Fenestration Product Conden-
sation Resistance Values, National Fenestration Rating Council, Greenbelt, MD.
5 Lewis, J., Limit Window Condensation, Commercial Building Products, Con-
Source LLC, Barrington, IL, 2007.
6 NFRC 100-2010, 2010, Procedure for Determining Fenestration Product
U-Factors, National Fenestration Rating Council, Greenbelt, MD.
sis of AAMA Condensation Resistance Factor, Interface: Journal of the Roof Con-
sultants Institute, RCI, Inc., Raleigh, NC, 2005, pp. 410.
8 Kudder, R. J., Babich, S. K., and Johnson, D. K., Effect of Installation Details on
the Condensation Performance of Window Frames, J. ASTM Int., Vol. 2, No. 10,
2005, 17 p.
9 LBNL, THERM 5.2/WINDOW 5.2 NFRC Simulation Manual. 2006. Lawrence Ber-
keley National Laboratory, Berkeley, CA.
doi:10.1520/JTE102969
Peter E. Nelson1 and Paul E. Totten2
Improving the Condensation Resistance of

Fenestration by Considering Total
Building Enclosure and Mechanical System
Interaction
ABSTRACT: Many condensation problems at fenestration are related to the

adjacent building components the system interacts with. This includes adja-
cent cladding and wall systems, structural systems used for the connection
of the fenestration, and interaction with the buildings mechanical system.
We will discuss how the condensation resistance of windows and curtain
walls can be greatly improved by aligning the thermal elements in the wall
with the thermal aspects of the windows such as thermal breaks, how air
leakage at window systems can increase the volume of condensation, and
how thermal bridges adjacent to the fenestration can have a signicant im-
pact on the overall performance. We will discuss how Winter time cold air
ow from adjacent wall cavities can cause elements at window or curtain wall
surrounds sill, jambs, and head to condense when they would not other-
wise. Thermal improvements through the application of carefully applied and
designed insulation systems will also be discussed, as over insulating in
some cases can exasperate the condensation problems by further isolating
the systems from heat sources. We will discuss the relationship between the
heating source position and distance from the window or curtain wall system.
The discussion in this paper will be mostly related to Winter time condensa-
tion. We will use examples to show 1 how thermal breaks, like rubber and
plastic shims below the windows, can disassociate the window from heat
loss to colder large thermal wall masses and improve performance; 2 how
elements like spandrel panels and shadow boxes can have condensation
due to air leakage into the shadow box and proposed solutions for these
elements to reduce their condensation risk; 3 how interior curtains, furni-
lished online December 2010.
1
P.E., Simpson Gumpertz & Heger Inc., Waltham, MA 02453.
2
P.E., Simpson Gumpertz & Heger Inc., Rockville, MD 20850.
Cite as: Nelson, P. E. and Totten, P. E., Improving the Condensation Resistance of
Fenestration by Considering Total Building Enclosure and Mechanical System
Interaction, J. Test. Eval., Vol. 39, No. 3. doi:10.1520/JTE102969.
286
NELSON AND TOTTEN, doi:10.1520/JTE102969 287
ture, and interior furnishings can block heat ow and cause condensation; 4
how the two-dimensional thermal analysis can be used to evaluate the con-
densation potential and evaluate possible solutions; 5 the results of infrared
thermography used to evaluate both condensation problems and evaluate
the solutions; 6 the importance of long-term data logging and monitoring of
the results; and 7 the review of options for reducing condensation occur-
rences through passive heat n and insulation and active heat trace
means.
KEYWORDS: condensation resistance, air leakage, diffusive vapor
transport, thermal bridging, HVAC interaction, fenestration
Introduction
Condensation can occur at fenestration for numerous reasons. This includes
thermal performance of the fenestration product, the tie-in to adjacent struc-
tural and building enclosure components including insulation systems, and air
ow to the window from mechanical systems. Other considerations including
the effects of wind washing of the fenestration and air leakage that may occur
at the window surround can also affect the fenestration product and lead to
lowering of temperatures at the surface such that condensation occurs. We are
also nding many more projects that have elevated interior humidity levels to
improve human comfort or meet the control demands for industrial or com-
mercial applications with condensation problems. Sometimes, the elevated in-
terior humidity levels are linked to newer air tight buildings, which reduce the
outside air ventilation exchanges through a leakier building enclosure.
Window designers typically rely on the condensation resistance factor
CRF in choosing a window or curtain wall system. However, the CRF number
alone will not determine whether the fenestration product window, curtain
wall, storefront, etc. will condense. As noted by OBrien 1, the window posi-
tioning within the assembly can greatly affect its overall performance with
respect to condensation resistance. As such, the CRF number is only a starting
point in the selection criteria.
To improve the overall resistance to condensation at fenestration, several
factors must be reviewed including the following:
The climate outside the building
The interior environmental conditions temperature and relative humid-
ity RH
The method of anchoring/attaching the fenestration
The attachment of the air barrier and locations of all air seals
The method of gasketing/sealing the fenestration product
The tie-in to the building enclosures thermal barrier
The interaction with structural components that form part of the build-
ings structural system
The location of mechanical system supplied heat sources, the type of
heat source, and the method by wish the heat is delivered
The type of glazing and the construction of the glazing system i.e.,
insulated glazing unit IGU, heat mirror, etc.
Type and location of window framing and thermal breaks
Any and all potential thermal bridges, including those that may be
caused by sun shade, light shelve anchorage systems, and other archi-
tectural features surrounding the windows
In addition, thermal analysis such as the calculation performed by using
THERM 2 two-dimensional 2D analysis and HEAT 3 3 three-dimensional
3D analysis are useful to assess the potential for condensation at design tem-
peratures, as well as at typical temperatures, for example, Winter temperatures,
for a particular climate. They can be used to examine the potential options to
provide remedial solutions to correct the condensation problems. With addi-
tional calculations, an estimate of the duration and the potential volume of
accumulating condensation can be projected. We discuss each of these factors
to show ways to improve the fenestrations resistance to condensation below.
FenestrationInteraction with Surrounding Enclosure Systems
The fenestration products as tested for CRF do not include the surrounding
construction, such as ashing systems, structural support and anchorage, the
adjacent wall elements, and insulation systems. The test is only used to exam-
ine the fenestration product. The CRF test, therefore, does not accurately pro-
vide the design team with data on the resistance of the overall construction. As
such, a second test using the entire intended assembly should be evaluated.
However, the use of thermal analysis tools, such as THERM 2, can provide the
design team useful information to understand what elements may need to be
rened to address the potential condensation problems without performing a
test. The case studies at the end of the paper provide several examples of using
the analysis tools to further examine thermal inefciencies.
The surrounding structure to the fenestration has a great inuence on its
overall performance 1,4,5. As fenestrations are typically the least energy ef-
cient component of a building enclosure with respect to the overall thermal
resistance, optimizing the fenestration placement to maximize the overall per-
formance is recommended. In doing so, it is then necessary to examine the
interaction of the fenestration with the surrounding elements to limit the peri-
ods of condensation. The ASHRAE 189.1 standard 6 for high performance
green buildings introduces an increased potential for condensation for fenes-
tration due to the requirements in some climate zones to super insulate walls
i.e., some to as high as R-40 overall as it makes fenestration placement more
difcult in certain wall types. A review of other concerns on overall energy
efciency will not be discussed in this paper.
To examine the fenestration placement within an opening, the following
should be studied.
What type of heat source will supply heat to the fenestration in Winter,
and how far is the heat ow source from the window?
Is the fenestration isolated from heat ow sources due to a temporary or
permanent obstruction, such as blinds or furniture?
Is the thermal break in the frame aligned with the IGU, and if any por-
tion of an accessory is also thermally broken, is it in line with the pri-
mary thermal break?
Have potential air ows around the perimeter of the fenestration been
blocked off such that cold Winter time air and, in Summer, potentially
humid warm outdoor air are not allowed to come in contact with por-
tions of the window that are at or below the dew point temperature? Are
the air seals used to close off air ow paths durable, or will they require
regular maintenance?
Are there any framing elements or accessories/hardware that under-
mines the thermal break by bridging across the break and thus short
circuiting the system?
Will the structure, such as framing steel, concrete, stone, masonry, or
similar more thermally conductive elements short circuit the insulation
system at the fenestration interface?
Are any components such as attachments for interior light shelves or
exterior sun shades attached in such a way that the thermal path has
been altered to bypass the thermal break, and if so, do these create a
thermal bridge that increases the condensation potential?
Is the fenestration thermally attached to the exterior cladding such as
architectural precast concrete, brick masonry, stone, metal panel, or
similar materials, which might undermine the intent of the thermal
break of the fenestration?
Is the thermal break of the system aligned approximately with the mid-
point of the insulation system? Is the frame positioned such that the
inboard portion of the frame and the center of the IGU are inboard of at
least the midpoint of the insulation system?
We will discuss each of these items in the following sections.
Mechanical System Interaction
In addition to placement of the window in the window opening, the effective-

ness of heat ow on the surface of the window is an important consideration to
limit Winter time condensation. An important aspect to examine is how the
CRF test is run; a cold side i.e., exterior Winter side and hot side interior heat
source are held quite close to the window or curtain wall or similar component
being tested. The test, therefore, is valid when the heating sources are quite
close to the window and similar in application to the heat supplied during the
test. As the heating source is moved further away, the method by which heat is
intended to be supplied i.e., from below versus overhead or forced air versus a
conductive element versus a radiant panel or the path of heat ow is ob-
structed by elements such as blinds or furniture, then the results of the stan-
dard test are no longer directly applicable.
In systems where a radiator is positioned below a window n tube or
other and blinds or furnishings are positioned inboard of the window and the
radiator, a direct heat ow is supplied to the window. This method of delivery
most closely replicates the method by which the systems are evaluated during
the CRF test. Many buildings constructed in the 1950s through 1970s used this
methodology to supply heat to the fenestration mainly because of the issues
with non-thermally broken single glass glazing and its propensity to condense
in Winter. In many cases, a conductive element such as an interior metal sill

extender was tied into the fenestration product, or a conductive stone sill was
extended inboard of the interior nishes forming a heat n or component to
readily deliver heat to the sill of the window. The conductive non-thermally
broken frame and the methods used for glazing in many of these systems
readily delivered heat and reduced much of the condensation risk at the fenes-
tration. However, these congurations are not very energy efcient. On a num-
ber of projects we have examined, where the radiator system was abandoned
and a new mechanical system implemented with new energy efcient windows,
the new systems are condensing, and in many cases, for longer duration and in
greater volume than the original system. On several of these projects, we have
utilized the same heat n concept but have provided localized thermal improve-
ments such as carefully placed new insulation to reduce the risk of heat ow to
the exterior and thus energy loss. We will describe these methods further in the
case studies.
Radiant heat panels are being used as a method on some projects to direct
heat to the fenestration. The radiant heat panels are typically placed at the head
of the window and aimed down at the window surfaces. They rely on hot water
typically at temperatures upward of 120 F that run through copper tubing to
heat the panel. On projects we have investigated, we have found that the sills
were too far from the radiant panel at the head to supply adequate warmth to
the sills. In another case, the position of the furniture and window shades
blocks the line of site of the radiant heat panels causing the window sills to
become cold and condense.
Forced air ventilation systems still ultimately rely on the ow of air to
transport conditioned air to the fenestration. As with radiant heating, blinds
and furniture can block air ow. A solution is to allow for air ow behind the
blinds sometimes accomplished by holding the blind short at head and soft or
locking tilt in on tilted blinds to allow for air ow through the blades or place-
ment of the diffuser such that the blinds are inboard of the heat source.The
distance of the nearest diffuser and its orientation of the supply air ow are
important; diffusers should be set no more than a few feet from the fenestration
and run parallel to the exterior wall. This closeness of the heat source to the
window is especially important if using a passive heat n to solve the conden-
sation problem. The throw of the diffuser can be evaluated through either com-
putational uid dynamics CFDs or eld evaluation. In addition to placement,
the temperature of the air washing the fenestration is important. Mechanical
systems need to be set up to provide a constant wash of warm air; this means
controls and the duration of applied heat to obtain an interior space set point
requires careful review. Systems with temperature setbacks will have long cool-
ing periods, which can cause condensation wetting cycles.
In addition to heat, the humidity levels also are important. In humidied
buildings, such as museums, natatoriums, hospitals, and similar places, the
risk for humidity build up caused by mechanical systems can increase the du-
ration and volume of condensation, especially if air stagnates within a space.
The air changes per hour i.e., ventilation rate in the space as well as near the
fenestration are important to keep the surface of the fenestration above dew
point temperature by helping reduce the stagnation of air. This is especially
important in conned or enclosed spaces. The air tightness is important to

reduce other inuences, such as air leakage at window surrounds or into
shadow boxes that can result in condensation, but requires careful review of
the intended ventilation rates.
Window to Wall Interface

Installing a fenestration in a wall assembly requires joining two different types
of materials and systems. On a very basic level, the windows or curtain wall
relies on the wall opening for structural support to resist window loads, vibra-
tions, and pressure differences. But there are other basic functions of the wall
that need to carry over to the window and vice versa, including water resis-
tance, thermal comfort, noise control, air resistance, condensation control, and
constructability.
Standards like ASTM E2112, Standard Practice for Installation of Exterior
Windows, Doors, and Skylights 7, primarily dene how the window and the
wall systems should be joined so that they are watertight, as typically rainwater
leakage is the owners biggest complaint. As buildings strive to become more
energy efcient, designers, detailers, and installers as well as all the construc-
tion parties should endeavor to improve the overall performance of buildings.
In particular, the gap between the window and the wall needs more detail fo-
cused on its design and coordination. ASTM E2112 strives to make the win-
dows performance in the eld equal the performance in the laboratory for
water tightness. The overall goals mandated by energy improvements now
should include addressing the owners other complaints of thermal comfort,
condensation, durability, and maintenance.
Air Leakage at Window Surrounds

In addition to water tightness, the gap between the window and the wall needs
to be air tight and thermally efcient. Like the wall, the gap should be con-
structed with multiple air and weather seals to mimic the performance of the
wall. An external weather and air seal is needed to shed water. An internal seal
is needed to provide aesthetic continuity, as well as an air seal, by joining to the
interior nishes. Within the wall, the designer and construction parties all need
to identify where the other air barriers and weather resistant barriers are and
assure that they are also sealed to the windows. The overall air tightness of the
assembly should match the general air tightness of the window and of the wall.
Extraneous air leakage from a wall cavity to the side of the window assembly
should be avoided as this will lead to extraneous air leakage through the win-
dow, which could lead to condensation problems in the Summer and/or Winter,
and possibly rain water leakage. If the sides of the window face the wall cavity,
they should be air tight to provide continuity between all the components of the
assembly.
Thermal Bridges
Ideally, the location of thermal insulation of the wall should be aligned closely
with the IGU and the thermal break of the window. The misalignment of the
thermal layers will cause the internal portions of the window to be cold and
condense. The amount of condensation can range from small localized zones of
beading water to heavy frost accumulation and wetting and water damage to
adjacent nishes. Additionally, the misalignment can lead to thermal discom-
fort from cold air ow and an increase in utility costs. Summer warm weather
conditions can also cause condensation on cold surfaces. Extraneous air ow
typically accompanies thermal alignment problems as the air barrier of the wall
cannot be adequately sealed to the windows.
The thermal pathways can be modeled in 2D or 3D heat transfer software,
such as THERM 2 or HEAT 3 3. These programs calculate the heat ow and
show how thermal bridges provide a path of lesser resistance through insula-
tion layers, thus lowering or raising temperatures. The colored heat ow dia-
grams help visualize the effect of thermal shortcuts and the location of poten-
tial condensation.
Common examples of how thermal misalignment can occur include mis-
alignment of the thermal insulation of the wall with the insulation break of the
window, thermal discrepancies in the window opening caused by stopping the
wall insulation too far from the edge of the window, constructing the window
opening with masonry or metal that forms a thermal short circuit, or designing
sun shades or architectural ns that act as the thermal short circuit for the
windows or the walls. The attachment of aluminum window frames to ther-
mally conductive components of the wall that extends to the outside may inu-
ence the window by allowing the cold to draw up into the window assembly
due to the increased heat loss through the more conductive components. Ther-
mally breaking the joint between the aluminum window and the thermal sur-
rounds may help improve the performance of the window. However, condensa-
tion may still occur in hidden areas. In such cases, additional evaluation is
needed to study the effects of this concealed condensation to determine if this
wetting will cause decay and durability problems and if adding heat to these
areas is needed to reduce the condensation effect.
Insulation Considerations at the Fenestration Surround
Insulation tie-ins at the fenestration surround are typically quite difcult to

achieve as other membranes for air ow control and water control are some-
times in the way of providing continuity at the surround. In most cases, insu-
lation within the frame will be somewhat disconnected from insulation in the
remainder of the building enclosure. However, insulation introduced into the
frame, especially in certain applications where the window sill is condensing,
may help in reducing the instances of condensation. When utilizing a heat n
to address condensation problems, this insulation in the frame may be needed
to reduce heat loss to the exterior.
Retrotting insulation into an existing window frame without removing the
window introduces its own complexities. We have developed a method using
closed cell low-expansion spray polyurethane foam canisters with the plastic
straw applicator and plastic tubing to inject new insulation into the frame
through a hole in the nearby construction of less than 1/4 in. by sliding the
FIG. 1Condensation on the window frame and glazing gasket at the sill.
tubing to the far side of the frame and slowly retracting it back to the entry-
point to ll the frame. The insulation used to retrot existing frames, if in the
wet zone, needs to be moisture tolerant.
Examples
We give below three examples where condensation occurred and the proposed
or implemented solution for each based on numerous past project experiences.
Example 1Condensation at the Window Sill of New Construction

We have been asked to evaluate window condensation at the sill on many
projects; many of these projects involve humidied buildings. In the example
given below, the interior temperature was intended to be 70 F + / 2 F and RH
35 % + / 5 %. The building was positively pressurized. The outside wall is
precast concrete mass stone wall, cast in place concrete, concrete masonry
unit block walls, brick masonry mass walls, and certain brick masonry cavity
walls where the window is pushed outboard of the insulation system will per-
form similarly. The window is over the precast and outboard of the buildings
insulation system. No shims were installed below the aluminum thermally bro-
ken window with IGU, and the frame was not insulated. Heating and air con-
ditioning was supplied via forced air systems; the nearest diffusers for many of
the condensing windows was 15 ft or greater from the windows. The windows
have blinds inboard and inadequate interior side air seals. See Fig. 1 below for
one of the windows.
We completed an onsite survey as well as a thermal analysis. The analysis
tools did not assess the effects of air leakage. Therefore, repair caulking of the
air seals was required where the seals were decient. On several projects we are
working on, this wall cavity air leakage path around the window surround
greatly increases the condensation events duration and the volume of conden-
sation. Our thermal analysis examined a passive heat n solution i.e., tying a
section of conductive metal into the window system and its surrounding con-
struction with an active heat trace. See Fig. 2 for the heat n conguration and
Fig. 3 for an example of the analysis results for the heat n. Window shimming
could not be added to these windows, nor could retrotting insulation within
the frame. On other projects where insulation within the frame and shimming
could be applied, we have found a temperature improvement at the sill as much
as 4 5 F. Here, we were able to add localized insulation to the window sur-
round. However, the heat trace was ultimately used due to the long distance to
many of the condensing windows from the nearest forced air diffuser. Had heat
from a diffuser been positioned closer to the window the passive solution
would have worked. The thermally conductive n takes heat from one location
and transfers it to another location. When insulated outboard of the n, this
transfer is quite efcient with minimal loss. The mock-ups on this project were
monitored by using temperature and RH sensors, as well as thermocouples
with data loggers.
Example 2Condensation at the Window Sill of Retrot Window

In Example 2, the project team was tasked with retrotting a new energy ef-
cient window into an existing glazed and metal panel curtain wall system. The
building was constructed in the 1960s. New structural attachments and up-
grades for blast and seismic considerations were tied into the new thermally
broken aluminum window. See Figs. 4 and 5 below for thermal analysis of the
sill and jamb conditions. Additional insulation and a localized passive heat n
were added to reduce condensation risk and minimize heat loss and gain. The
original window system used a heat n tied into the curtain wall with radiators
below the n, which had the effect of reducing condensation. The radiators
were scheduled to be removed as part of a renovation project in process when
we were contacted to assist with condensation problems at the new windows.
The mock-ups were constructed onsite to further evaluate the additional insu-
lation and heat n, and based on the mock-ups a full scale repair program was
implemented. Thermography was completed by others on the design consulting
team to evaluate the solution and found good correlation with the thermal
analysis models and the predicted reduced risk for condensation. Through sev-
eral cold Winters, the window system is performing as expected with signicant
reductions in condensation volume and duration almost to the point of full
elimination.
Example 3Shadow Boxes

On more than one project, we have investigated complaints of condensation at
spandrel panels or shadow boxes. The introduction of insulation inboard of the
glazing thermally isolates the glazing from any interior heat sources and places
the glazing system close to exterior temperatures. Air leakage from the interior
into the shadow box space between the insulation element and the glass can
result in condensation. The volume and duration will be affected by the humid-
ity and temperature of the air inside the shadow box, the rate of leakage, the
rate of possible ventilation to the exterior sometimes allowed by short gaskets
FIG. 2Aluminum heat n conguration.

FIG. 3Temperature distribution for remedial thermal improvement with inclusion of

spray foam insulation, aluminum heat n, and thermal conductive gel.
or extra weeps at the head as well as the sill to vent the box, and the interior
pressurization of the building. The higher the interior humidity, the greater the
volume of condensation within the shadow box.
In these cases, the air sealing of the frame joinery and of the back pan/
insulation system as well as the venting of the box with dry outside air has been
effective in improving the condensation problems if there are tight internal
seals.
Conclusions
Based on our experience, we provide the following conclusions.
CRF numbers should only be used as a starting point in choosing fen-
estration for condensation resistance; the positioning of the product, the
interaction with the mechanical system, and the adjacent enclosure el-
ements can greatly affect its performance.
Analysis tools such as THERM 2 and HEAT 3 3 as well as CFD can be
used to evaluate both congurations for new work as well as retrot
work to evaluate condensation potential.
Air leakage through window surrounds can increase the risk for conden-
sation. The location and position of the air seals are critical to keeping
cold exterior air from entering into and short circuiting the thermal
insulation zone around the windows. An air leak from the outside to
inside can cause a consistent feed of colder air and result in increased
levels of condensation.
Details only provided in 2D may not adequately show the necessary
FIG. 4Thermal analysis of the window system after thermal improvements at sill.
FIG. 5Thermal evaluation of structural retrot of new energy efcient windows into
an existing curtain wall.
tie-in at fenestration surrounds for the air seals at the sill-to-jamb and
jamb-to-head interface.
Interior air leakage into spandrel panel and shadow boxes in fenestra-
tion systems, especially humidied air, can result in condensation in the
shadow box. The amount of condensation is dependent on the pressure
difference between the interior and the exterior, the size of the holes in
the air barriers, and the ow path.
A careful review of furniture and blind placement is necessary to avoid
obstructing heat ow that is needed to warm cold windows.
The interaction of the mechanical system with the building enclosure is
many times overlooked. Windows that rely on constant heat ow from
the HVAC system are at higher risk for condensation when the HVAC
systems are set back i.e., reduced temperature.
References

sis of the AAMA Condensation Resistance Factor, RCI Interface, May, 2005, pp.
410.
2 THERM Finite Element Simulator. 2003. Regents of the University of California,
Lawrence Berkely National Laboratories, Berkeley, CA.
3 Blomsberg, T. and Claessen, J., HEAT 3. 2003. Building Technology GroupMIT
and Department of Building PhysicsChalmers Institute of Technology, Cam-
bridge, MA.
4 Totten, P. E., OBrien, S. M., and Pazera, M., The Effects of Thermal Bridging at
Interface Conditions, Building Enclosure Science & Technology 1, Minneapolis,
MN, June 2008, NIBS, Washington, DC.
5 Totten, P. E. and Pazera, M., Thermal Inefciencies in Building Enclosures
Causes of Moisture Related Performance Problems, ASCE Fifth Forensic Engineer-
ing Congress, Washington, D.C., November 2009, ASCE, Reston, VA.
6 ASHRAE 189.1, 2009, Standard for the Design of High-Performance Green Build-
ings Except Low-Rise Residential Buildings, ASHRAE, Atlanta, GA.
7 ASTM E2112, 2007, Standard Practice for Installation of Exterior Windows,
Doors, and Skylights, Annual Book of ASTM Standards, Vol. 04.12, ASTM Inter-
national, West Conshohocken, PA.
doi:10.1520/JTE103017
Todd A. Gorrell1
Condensation Problems in Precast Concrete

Cladding Systems in Cold Climates
ABSTRACT: The use of precast concrete panels as a cladding material can

provide an economical and attractive method to enclose contemporary build-
ings. In cold climates, improper detailing or inappropriate construction of
precast clad walls can result in widespread water problems due to conden-
sation forming within the wall construction. Condensation may form at the
interior of the wall assembly due to inadequate protection against air ow
through the wall, such as exltration of warm, moist indoor air, or by inltra-
tion of cold outdoor air reaching the interior wall components. The conden-
sation within the wall assembly can also be caused by water vapor diffusion
into the wall from the building interior. In addition, inadequate separation
and/or insulation between precast panels and adjacent cladding compo-
nents, such as windows, curtain wall, or metal framing components, can
result in surface condensation on these wall elements. The moisture from
condensation can cause extensive damage to nish materials, including in-
sulation and gypsum wallboard; can result in corrosion and deterioration of
metal components, such as structural connections and metal wall framing;
and may result in mold and mildew. This paper will 1 present examples of
precast clad building wall construction where signicant condensation prob-
lems have occurred; 2 discuss typical causes of condensation in precast
clad buildings in cold climates; 3 review design considerations to reduce
the potential for condensation in new buildings; and 4 present repair meth-
ods to address condensation in existing buildings.
KEYWORDS: precast concrete, cladding, condensation, air leakage,
vapor retarder
Manuscript received February 2, 2010; accepted for publication October 19, 2010; pub-
lished online December 2010.
1
Senior Associate, Klein and Hoffman, 150 South Wacker Dr., Suite 1900, Chicago, IL
60606.
Cite as: Gorrell, T. A., Condensation Problems in Precast Concrete Cladding Systems in
Cold Climates, J. Test. Eval., Vol. 39, No. 4. doi:10.1520/JTE103017.
299
Introduction
Panelized precast concrete is commonly used as a cladding material on build-
ings throughout North America, and for very good reason, it is an economical
and attractive cladding system for contemporary buildings. The low cost of
concrete materials and relatively fast installation can reduce the cost of precast
panel cladding versus other more labor intensive systems, such as unit ma-
sonry. The wide variety of colors and textures available, as well as optional
applied nishes thin applied stone and brick, for example, allows for a wide
range of architectural possibilities. The precast concrete cladding is used on
many building types, both low and high rise, and is often found on hospital
buildings, institutional facilities, and multi-family dwellings, such as condo-
minium buildings 1.
Some of the building types where precast concrete panels are specied can
have high interior relative humidity RH conditions whether intentional or
not, such as hospitals, museums, or residential buildings. In cold climates,
condensation problems can occur within the exterior walls of precast concrete
clad buildings, and if uncontrolled, the condensation water accumulation can
cause visible and extensive damage to interior nishes. The water within the
wall system can damage concealed structural components, such as panel an-
chor systems, structural steel framing, and metal wall framing. Moisture accu-
mulation within the wall can also lead to mold growth, which in severe occur-
rences may cause health-related injuries.
The condensation forming within the precast concrete walls is usually
caused by improper design and detailing of the wall system, and by inappro-
priate construction. This condition is largely preventable if the principles of
vapor and water management and thermal transfer are understood and ad-
dressed by both the designer and contractor. Condensation can also be caused
by the building owner maintaining improper environmental conditions after
the building is occupied. This too is avoidable with proper education of the
building owner and maintenance staff.
Condensation
The air around us contains signicant water in vapor form. The amount of
water vapor that air holds referred to as RH is affected by the temperature of
the air; warm air can retain more moisture than cold air. Water vapor is ab-
sorbed and released from an air mass by the processes of evaporation and
condensation, respectively. In cold climates during Wintertime, moisture is
often added to a buildings interior air with humidiers to increase the comfort
level for building occupants and to protect the nish materials, furniture, and
art work. Condensation occurs when warmer, moist air contacts a surface that
is colder than the dew point temperature of the air mass. This results in the
water vapor in the air transitioning to liquid water on the cold surface 1,2.
Wintertime condensation is typically a greater concern in cold regions, such as
U.S. Regions 5 through 8 as shown in Fig. 18 of Ref 1 in the Precast/Prestressed
Concrete Institutes Designers Notebook for Energy Conservation and Con-
densation Control 1.
GORRELL, doi:10.1520/JTE103017 301
FIG. 1Water damaged ceiling tiles caused by condensation on precast panels above.
This building is in Climate Zone 5A, and the precast panels are insulated with foil-faced
rigid insulation applied directly to the precast panels.
During Wintertime, when warm, moist interior air is able to leak through
the exterior envelope of a building through holes and unsealed joints exltra-
tion, it can form condensation on the surfaces of cold building components
within the wall. Water vapor diffusion through a wall assembly without an
effective vapor retarder can also result in condensation within the wall, al-
though this is usually less of a problem than condensation caused by bulk air
leakage, which quickly transports far greater amounts of moisture than diffu-
sion alone 2.
In precast clad wall systems in cold climates, condensation is most often
observed within the wall assembly forming on the inside face of the precast
panel, which is nearly the same temperature as the exterior face except for
insulated sandwich panels. Condensation can also form on other exposed sur-
faces to the exterior of the wall insulation, such as steel anchors, steel framing
members, and other metal components.
The condensation that forms within the wall assembly will ow downward
until it meets an obstruction, such as the smoke seal at a oor slab, or a win-
dow head, which then results in the water being directed inward where it soaks
into and damages interior nishes, as shown in Figs. 1 and 2. Condensation
water can also soak into some insulating materials, thereby reducing their in-
sulating value and exacerbating the overall problem.
During extended periods of subfreezing temperatures, portions of the inter-
nal wall components will be below freezing and condensation will form within
the wall as ice, as shown in Figs. 3 and 4. As long as temperatures remain below
FIG. 2Damaged gypsum wallboard due to condensation on back of precast panels.

This building is in Climate Zone 6A, and the precast panels are insulated with batt
insulation between metal studs with a polyethylene vapor retarder behind the gypsum
wallboard.
FIG. 3A small portion of ice accumulation on back face of precast concrete panel
after most of the ice had already melted. This precast wall is in Climate Zone 6A and
is insulated with batt insulation between metal studs with a polyethylene vapor retarder
behind the gypsum wallboard. The batt insulation was saturated with water.
GORRELL, doi:10.1520/JTE103017 303
FIG. 4The ice accumulation on metal curtain wall framing in direct contact with
precast panel insulation removed for inspection.
freezing, the condensation ice will accumulate until temperatures rise above
freezing and the ice melts, releasing large amounts of water into the building
interior at one time 2. This sudden melt condition is sometimes mistaken as
exterior water leakage through the wall instead of condensation.
Insulation, Vapor Retarder, and Air Barriers Systems

Insulation systems, vapor retarders, and air barriers are used to control tem-
perature differentials across a wall assembly and to control air movement and
vapor diffusion through the wall 1. There are several different techniques and
materials used to insulate and provide a vapor retarder for precast panel walls.
The most common systems include the following.
Batt insulation in a stud wall usually separated from the precast panels
by an air cavity. The vapor retarder is typically either an integral facing
sheet on the insulation e.g., foil scrim or a separately applied sheet
e.g., polyethylene 1, as shown in Fig. 5.
Rigid or semi-rigid insulation applied to the back of the precast with or
without an air space. The vapor retarder is often an integral foil scrim
facing sheet on the insulation 1, as shown in Fig. 6. The joints of the
insulation boards must be fully taped and sealed.
Spray-applied polyurethane foam insulation applied directly to the pre-
cast panels. Spray foams used in this application are typically closed-cell
foam and are installed to a thickness to achieve a perm rating less than
1.0 in order to also perform as a low-level vapor retarder. Depending on
the product used, this thickness is usually in the range of 35 in. 76127
mm. If a signicantly lower perm rating is needed due to the planned
building use requiring high RH levels, this may be achieved with a
supplemental vapor retarder material, such as a vapor impermeable
coating applied to the inside face of the foam insulation. The thickness
of the spray foam insulation alone that would be necessary to achieve a
low perm rating e.g., 0.05 perms would be of the order of 40 in., which
is impractical for a typical building wall.
FIG. 5The precast cladding is insulated with batt insulation in a stud wall and cov-
ered with polyethylene vapor retarder. This building is in Climate Zone 6A.
Insulated precast sandwich panels with rigid insulation cast integrally

within the precast panel. The inner layer of concrete along with an inner
line of joint sealant acts as the vapor retarder for this system. There are
generally two types of sandwich panels: those with edge-to-edge insula-
tion with the two concrete layers connected with ties and those with
discontinuous insulation with the outer layers of concrete connected
with concrete ribs at panel edges and internal ribs. The sandwich panels
FIG. 6The precast cladding is insulated with direct-applied mineral wool insulation
with integral foil scrim vapor retarder with taped joints. This building is in Climate
Zone 5A.
GORRELL, doi:10.1520/JTE103017 305
with edge-to-edge insulation have improved thermal performance if the

ties are of non-conductive material, such as plastic or berglass. The
author has not witnessed condensation problems with insulated sand-
wich panels; however, ribbed panels can have similar thermal/
condensation problems as conventional solid panels due to thermal
bridging at the concrete ribs.
A properly designed and installed vapor retarder system also acts as an
effective air barrier, and generally is the best solution to control interior air
exltration and moisture transfer through a wall system. A continuous gypsum
wallboard assembly, with taped seams and sealed perimeter joints, can be an
effective air barrier 2.
Examples of Design, Installation, and Occupancy Problems That Cause

Condensation
No Vapor Retarder/Air Barrier

The most basic design error for precast clad buildings in cold climates is the
omission of specifying a vapor retarder system where one is needed. This oc-
curs most commonly when the designer is not familiar with the potential for
condensation in precast clad wall systems in cold and transitional climates.
Incomplete Installation of Insulation and Vapor Retarder

A more common occurrence is the lack of continuity of the insulation or the
vapor retarder, or both, at perimeter conditions of a wall. Often, the design
drawings do not include the level of detail to delineate the vapor retarder mem-
brane and specify that it must be continuous to and sealed against adjacent
systems, such as oor and roof structures, and door/window framing. When the
building enclosure is constructed with these gaps in the insulation and vapor
retarder systems, the lack of continuity allows for interior air to reach the
precast panel, as shown in Figs. 7 and 8, and cause condensation.
Penetrations and Thermal Bridges

Similar to the above, the items that penetrate through the vapor retarder must
be sealed against air leakage, as shown in Figs. 9 and 10. In addition, protection
should be taken against the penetrating element creating a thermal bridge that
could result in condensation. This is a common problem at precast anchors,
which are typically large and penetrate the interior wall layers 2, as shown in
Fig. 10. During Wintertime, the steel anchor exposed to the inside of the wall
insulation may be colder than the dew point temperature of the interior air
resulting in condensation on the cold steel.
Unanticipated Conditions
The design drawings typically include the details that are found within normal
line of sight, such as window jambs and heads, at oor lines, at column/wall
FIG. 8A view of an opening in gypsum wallboard at a window jamb detail adjacent to

a precast panel. The foil scrim vapor retarder is not continuous to the window frame,
and the window perimeter joint is uninsulated. The back face of the precast panel is
visible and exposed to interior air ow.
FIG. 9The unsealed vapor retarder at electrical penetration allows bulk air leakage
through the wall.
GORRELL, doi:10.1520/JTE103017 307
FIG. 7A view looking up through open ceiling tile at interface between window head
frame and a precast panel above. The foil scrim vapor retarder is not continuous to the
window frame, and the joint above the window is uninsulated. The bottom edge of the
precast panel is visible and directly exposed to interior air ow.
intersections, etc. Occasionally, a condition occurs at a location not considered

during the design phase that causes a discontinuity of the insulation and/or
vapor retarder and results in air leakage and condensation. For example, Fig.
11 is an example of a building specied with batt insulation between metal
studs and polyethylene vapor retarder membrane at the exterior walls. Unfor-
tunately, a ventilated soft condition typical throughout the building was not
detailed at the locations where the precast cladding meets the adjacent curtain
FIG. 10The unsealed vapor retarder at steel anchor penetration allows bulk air exl-
tration to the precast panel. In addition, the steel framing is uninsulated and will act as
a thermal bridge across the wall insulation.
FIG. 11A view looking up through an opening into a ceiling soft revealed the inter-
face between the curtain wall framing and the precast panels. The precast panel was
uninsulated and without a vapor retarder. Interior air was able to reach the back face of
the precast panels via the ceiling soft.
wall system, and during construction, this resulted in missing insulation and
vapor retarder at these interface locations. This omission created a direct path-
way for interior air to the precast panels and signicant condensation damage
occurred. Not all such conditions can be reasonably foreseen during the design
phase, but most can be addressed during construction if a building envelope
quality control program or a building envelope commissioning program is in
effect to help identify such deciencies 3.
Window Placement
Another form of condensation observed in precast clad buildings is at the inte-
rior surface of metal window frames, including thermally improved windows.
The accumulation of condensation water on the interior surfaces of the win-
dow frame is sometimes mistakenly attributed to water leakage through the
window, or a failure of the windows thermal break system, when in fact the
cause is related to either the design or installation of the window to the precast
interface detail.
Generally, this type of condensation problem is caused by poor placement
of the window assembly in relation to the precast panel, or by lack of thermal
separation between the precast panels and the thermally improved portion of
the window system. All too often, the thermally improved metal window sys-
tems are specied to be inset completely within the thickness of precast panel
cladding so that the interior portion of the window frame is directly adjacent to
GORRELL, doi:10.1520/JTE103017 309
FIG. 12Thermally improved window jamb inset into precast panel with no insulation
in the perimeter joint between window and precast panel. Under given temperature
differential, the interior surface of the window frame is below freezing and would likely
exhibit condensation.
the precast panel. In cold temperatures, the proximity of the cold concrete can
reduce the temperature of the window frame below the dew point temperature
of interior air.
To illustrate this scenario, a commonly observed thermally improved win-
dow jamb detail was modeled by using the THERM 5.2 heat transfer modeling
program 4, as shown in Fig. 12. The window system was placed directly ad-
jacent to a 6 in. 152 mm thick precast panel and no insulation was included in
the window perimeter joint. The model was analyzed at an exterior tempera-
ture of 10 F 23 C and an interior temperature of 70 F 21 C. If the
interior air is at 30 % RH, the dew point temperature of the air would be 37 F
2.8 C. The THERM model calculated the interior surface temperature of the
metal window frame to be 29.9 F 1.2 C, which is less than the dew point.
Therefore, this scenario would result in condensation, likely in the form of
frost, forming on the window frame.
Adding insulation to the window perimeter joint would improve the situa-
tion by increasing the temperature of the window frame by several degrees, as
shown in Fig. 13. Under some conditions, this added insulation alone may
prevent condensation on the window frame. However, repositioning the win-
dow inward to locate the thermal break of the window generally in line with the
wall insulation vastly improves the performance of the window system and
eliminates the likelihood of condensation on the window frame under normal
conditions, as shown in Fig. 14.
Interior Moisture Levels Too High for Specied Systems

This condition can be caused during the design phase by not specifying the
proper vapor retarder system for the anticipated level of interior moisture and
mechanical system requirements. However, all too often, the real problem is the
uncontrolled humidication of interior spaces by an inadequately designed and
controlled HVAC system and especially by the occupant after the building is
completed.
FIG. 13Same conguration as Fig. 12 with the addition of insulation in the perimeter
joint, resulting in the interior surface of the window frame rising by several degrees.
Design and Construction Considerations

To reduce or eliminate the potential for condensation in a precast clad building,
the following design considerations should be considered.
Design the exterior wall assembly to provide an interior air barrier to
prevent the inside air from reaching outside of the wall insulation and to
the back face of the precast panels. This may be achieved with a vapor
retarder membrane, or in some cases simply with continuous gypsum
wallboard with sealed perimeters and joints 2.
Determine if a vapor retarder system is necessary based on the climate
zone of the building, the interior humidication requirements, and the
local building code.
Select the insulation and vapor retarder system that is most appropriate
for the building. The selection should take into account the cost impact
and the constructability of the different systems as related to the skill
level of the local construction industry. The use of spray-applied, closed-
cell polyurethane foam to provide an airtight insulation envelope with
integral vapor retarder characteristics is fast becoming the material of
FIG. 14Window shifted inward to more closely align the window thermal break with
the wall insulation, thereby dramatically increasing the interior surface temperature of
the window frame.
GORRELL, doi:10.1520/JTE103017 311
choice for many projects, provided that the installed thickness of spray-
applied insulation can develop the vapor resistance required for the ap-
plication. The designer should keep in mind that re-rating require-
ments in some jurisdictions requires that additional steps should be
taken to protect or encapsulate this material in wall and roof assem-
blies.
Provide an appropriate level of drawing details to indicate the following:
Continuity of vapor retarder and insulation systems, including termi-
nation at wall openings. These details should clearly show that vapor
retarder membranes are sealed to window and door frames and do
not stop short in order to prevent internal air and vapor exltration
3.
Penetrations are properly sealed, and in the case of steel members,
insulated to prevent thermal bridging 3.
Locate window systems to prevent the thermally improved portion of
the windows from close proximity to cold wall components without
adequate thermal separation.
Consider the wall details that occur at locations other than the usual
line of sight locations. Cut the wall sections, both vertical and hori-
zontal, above ceilings and through soft areas to identify the problem
areas that could lead to air leakage and condensation.
Provide mechanical systems that properly monitor and regulate the in-
terior humidity conditions based on exterior temperatures.
Secondary drainage systems may be used to collect and drain incidental
water at the back face of precast panels, but these systems may not be
completely effective against condensation. These secondary drainage
systems typically consist of a gutter or reglet at the back of the precast
panels to collect water and drain it to the exterior through weep tubes.
However, there is usually a portion of the panel below these gutters
where condensation can form and ow down to the oor level where it
travels across the re/smoke seal and into the building. In addition,
when temperatures are cold enough that the condensation forms as ice,
the weep tubes will also be frozen and will not drain.
During the construction phase, the project team should incorporate
steps to identify and correct the potential air leakage and condensation
conditions as follows:
Specify and perform a quality control program or building envelope
commissioning program to inspect the installation of the exterior en-
velope systems and their components 3,5. Particular examination
should be performed at the interface between different systems and
different construction trades.
Specify and perform air leakage testing 3. Initial testing should be
performed during the building enclosure mock-ups to identify inad-
equacies in the design or construction prior to building-wide instal-
lation. Thereafter, perform regular intermittent testing during con-
struction at representative areas as determined based on the size and
complexity of the building enclosure design.
Upon completion of the exterior envelope for the building, consider
thermal infrared testing of the building enclosure to identify potential

locations of air leakage or thermal gaps that may contribute to con-
densation before the interior nishes are applied.
Once the building is occupied, the owner has the responsibility to main-
tain the building systems to prevent condensation as follows:
Monitor the interior environment and adjust accordingly to prevent
humidity levels from exceeding the design parameters of the building
enclosure systems.
Maintain the insulation and vapor retarder systems and seals. When
alterations are made to the building components, the continuity of
the insulation and vapor retarder systems should be carefully consid-
ered and remain intact.
Repairs of Existing Buildings
For the existing precast clad buildings that are experiencing condensation
problems, the resulting interior water damage is often mistakenly attributed to
water leakage from the exterior. An investigation of the problem area should be
performed to determine if the condensation is the actual cause, and should
include the following.
Review of the design drawings to identify possible causes for either
water inltration or the potential for condensation.
Determine the history and pattern of the observed water inltration.
Water leaks or damage that occurs only during Wintertime, particularly
after a thaw event, are likely due to condensation within the wall.
Perform an inspection of the wall exterior for possible avenues of water
leakage. This may require water testing to conrm or deny that leakage
from the exterior is occurring.
Investigate the as-built wall construction by using inspection openings
to identify pathways for air leakage. Also, inspect vapor retarder termi-
nations and seals, continuity of insulation materials, and internal signs
of condensation or water leakage. Pressurizing the building or specic
building areas and using a smoke pencil is useful to pinpoint air leaks.
Measure interior temperature and RH levels to determine if they are
possibly contributing to the condensation. These measurements should
be performed at the time of the condensation event. If available, the
historic HVAC data for the building temperatures and RH measure-
ments is also useful to determine the causes for observed condensa-
tion.
Once a condensation problem is identied, the repair options can be stud-
ied. Ideally, this would include removal of interior nishes to the extent neces-
sary to fully repair or replace the insulation and vapor retarder systems, and
sometimes that is the only effective solution. However, more often than not,
this extensive approach is either not possible, or is unnecessary. The simplest
solution may be modifying the mechanical systems, or the occupants habits, to
reduce the moisture levels in the building.
In most cases, the repairs will need to include some removal of interior or
GORRELL, doi:10.1520/JTE103017 313
FIG. 15A view looking up through a ceiling opening at spray foam insulation repair
installed on the back face of a precast concrete panel. The spray foam was applied to an
overall thickness to serve as both insulation and vapor retarder, as well as an air barrier
to seal openings in the wall system.
exterior nishes to expose and repair improper termination details of the insu-
lation and vapor retarder, such as at windows, doors, and at ceilings or oors.
These repairs typically consist of adding supplemental insulation materials to
complete the insulation envelope, and taping of gaps in the vapor retarder
membrane. As previously described, the use of spray-applied, closed-cell poly-
urethane foam is becoming a commonly used material in building walls, and
this is especially true for remediation of condensation problems, as shown in
Fig. 15. In addition to its physical characteristics, the ability of the expanding
foam product to be installed into tight joints and irregular openings, as well as
its ability to bond with many building materials, makes this product well-suited
to this application. However, careful evaluation and design of the application
and proper installation must be performed to provide an effective solution.
Where possible, it would also be prudent to perform and evaluate a trial
repair prior to embarking on a widespread repair project. A trial repair area
should be selected to include representative areas of the wall system that have
previously experienced condensation and are large enough to incorporate typi-
cal wall conditions joint conditions, interfaces with adjacent cladding systems,
etc.. The repair design should be fully executed in the trial repair areas by
using the same materials and techniques that would be used for a comprehen-
sive repair program. Depending on the conguration of the building and trial
repair areas, a special perimeter detailing may be required to isolate the trial
repair area from the adjacent wall to prevent inadvertent lateral air and water
vapor ow that may adversely affect the test results. The duration of a trial
repair should typically include all seasons for which condensation has previ-
ously occurred, which for cold climate regions usually means a full Winter
season. The trial repair area should be regularly monitored throughout the
duration of the test and compared with other wall areas to compare the relative
improvement of the repaired wall areas.
Conclusions
Condensation problems in precast concrete clad buildings are a potentially sig-

nicant and persistent problem that can result in extensive damage to building
interior nishes and systems, especially in buildings with high interior RH lev-
els. However, the basic causes for condensation can be understood and ad-
dressed. Proper design and specication of insulation and vapor retarder sys-
tems can reduce the potential for condensation in new buildings. This same
care and attention to detail must be carried through the construction phase and
even into occupancy practices. Consideration should be given to quality con-
trol, exterior commissioning, and testing programs during the construction
phase to verify the proper installation and performance of the specied sys-
tems. After completion of the building construction, the monitoring of interior
RH levels and maintenance of the building envelope systems should be per-
formed by the owner.
Existing precast clad buildings displaying condensation problems can be
repaired, but usually not without some level of invasive work that may be
costly. An investigation should be performed to determine if condensation is, in
fact, the problem and if the cause is related to the detailing and construction,
an improperly controlled level of interior RH, or possibly both. Consideration
should be given to the performance of trial repairs to measure the efcacy of a
repair design prior to proceeding with large-scale and comprehensive repair
programs.
References
1 VanGeem, M. G., Designers NotebookEnergy Conservation and Condensation

Control, Precast/Prestressed Concrete Institute, Chicago, IL, 2006.
2 Rousseau, M. Z. and Quirouette, R. L., Precast Panel Wall Assemblies, Building
Science Forum 82, a series of seminars presented in major cities across Canada in
1982, 1982, National Research Council Canada, Ottawa, ON.
3 Lemieux, D. J., Building Envelope Design GuideWall Systems, National Institute
of Building Sciences, Washington, D.C., 2010, http://www.wbdg.org/design/
env_wall.php.
4 THERM 5.2/WINDOW 5.2 NFRC Simulation Manual. July 2006. Lawrence Ber-
keley National Laboratory, Berkely, CA, Program software and users manual avail-
able at http://windows.lbl.gov/software/therm/therm.html.
5 Odom, J. D., Designers NotebookAvoidance of Mold, Precast/Prestressed Concrete
Institute, Chicago, IL, 2008.
STP1498-EB/May 2011
315
Author Index
A Morgan, C. M., 226-246
Armstrong, M., 31-58 N
D Nelson, P. E., 286-298
Dunlap, A. A., 187-225 O
OBrien, S. M., 247-268

E
Ordner, E., 269-285
Elmahdy, H., 31-58 P
F Patel, A. K., 247-268

Pazera, M., 59-78
Flick, L. D., 226-246
Flock, S. K., 169-186 R
G Rose, W. B., 1-27
Ge, H., 131-147 S

Glazer, R., 31-58
Gorrell, T. A., 299-314 Salonvaara, M., 59-78
Songer, C. A., 187-225
H Stamatiades, G. P., 148-166
Hall, G. D., 169-186 T

Hokoi, S., 80-93
Tariku, F., 131-147
J TenWolde, A., 119-130
Totten, P. E., 286-298
Johnson, P. G., 187-225 U
L Umeno, T., 80-93
Lacasse, M. A., 31-58 V
M Van De Bossche, N., 31-58
Maref, W., 31-58 W

McGowan, L. M., 226-246
Minnich, L. C., 94-118 Weston, T. A., 94-118
Copyright 2011 by ASTM International www.astm.org

STP1498-EB/May 2011
317
Subject Index
A G
active glazing control, 187-225 glazed exterior wall systems, 187-

air leakage, 299-314, 31-58, 286-298 225
airow, 247-268
H
B
high humidity, 187-225, 247-268
building, 119-130 hospital, 247-268
built-in moisture, 59-78 HVAC interaction, 286-298
hygrothermal analysis, 187-225
C hygrothermal characteristics, 80-93
hygrothermal performance,
131-147
cavity ventilation, 131-147
cellulose ber insulation, 59-78
cladding, 299-314 I
climate chamber, 80-93
insulation, 1-27
condensation, 59-78, 226-246,
269-285, 187-225, 299-314, 148-166, interior exposure, 169-186
247-268, 1-27, 169-186, 94-118
condensation control, 187-225 L
condensation resistance, 286-298
CR rating, 269-285 laboratory testing, 31-58
D M
decay, 148-166 modeling, 169-186

design, 119-130 moisture, 119-130
diffusive vapor transport, 286-298 moisture control, 1-27
moisture damage, 80-93
E moisture transport, 59-78
mold, 119-130
exterior walls, 226-246 museum, 247-268
F N
fenestration, 269-285, 286-298 natatorium, 247-268

eld-experiment, 131-147 numerical analysis, 80-93
Copyright 2011 by ASTM International www.astm.org

318
P V
passive glazing control, 187-225 vapor barrier, 226-246

precast concrete, 299-314 vapor pressure, 59-78
vapor retarder, 299-314
R vented air layer, 80-93
rain-screen wall, 131-147

W
S
wall assemblies, 94-118
sill pan, 148-166 wall systems, 187-225
stain, 80-93 wall-window interface, 31-58
water management, 94-118
T wetting and drying potentials, 131-
147
THERM, 269-285 window, 148-166
THERM program, 169-186 window condensation, 31-58
thermal analysis, 187-225 window installation, 31-58
thermal bridging, 286-298 windows, 226-246
thermal performance, 169-186 wooden residential building, 80-93
www.astm.org
Cover Photo courtesy of ISBN: 978-0-8031-4471-2
Keith Nelson, Wiss, Janney, Elstner Associates, Inc. Stock #: STP1498

STP 1498-2011

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

STP 1498-2011

Uploaded by

Copyright:

Available Formats

Journal of Testing and Evaluation

Selected Technical Papers

JTE Guest Editors:

Printed in the U.S.A.

ASTM Stock #: STP1498

Copyright 2011 ASTM INTERNATIONAL, West Conshohocken, PA. All rights

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317

Insulation Draws Water

ABSTRACT: In the late 1930s, an architect and two researchers created a

I urged that great difculty might be encountered in condensation, in

Insulation Draws Water

!"#$%!& ()!$*+", -!" .!!/

FIG. 1Wood sorption isotherm from Ref 4.

somewhat-linear relationship between the equilibrium moisture content in

FIG. 2Lines of constant wood moisture content plotted on a psychrometric chart.

Tyler Stewart Rogers

tion air conditioning, here Rogers associated wetness with humidi-

FIG. 3Flawed analogy comparing diffusion transport kinetic to bulk ow dynamic

FIG. 4Larry V. Teesdale, U.S. Forest Products Laboratory.

In these opening paragraphs, Teesdale, like Rogers, emphasized how im-

5 Resistance of the inside wall to vapor movement.

by Rogers apparently written to address the concerns of the architecture au-

American Society of Heating and Ventilating Engineers Now ASHRAE

As Professor of Mechanical Engineering at the University of Minnesota, Rowley

FIG. 7Frank Rowley, professor of Mechanical Engineering, University of Minnesota.

formance outcomes of prescriptive measures for normal building materials.

water in capillary or open porous materials is discouraged.

Paint is an excellent vapor barrier, comparable to the protection pro-

Housing and Home Finance Agency

FIG. 8Vapor barrier installation illustration from Condensation Control in Dwelling

The War Against Water Campaign

Where do they go from here? Believe it or not, they have an engagement.

Wrapup, BRAB Condensation Conference 1952

The authors of the condensation paradigm created a framework, a way of ana-

giving temperature the importance it deserves. Recall Fig. 2. This is critically

0!)*$!"5#"6) 6%51! 8%!!")! $12) !)9-"!):)/#'

FIG. 10Decision tree for prescriptive code requirements.

the questions of criticality, necessity, and sufciency? This does not

1952, Building Research Advisory Board BRAB, National Research Council/

Wahid Maref,1 Nathan Van De Bossche,2 Marianne Armstrong,1

Laboratory Tests of Window-Wall Interface

ABSTRACT: The development of alternative details to manage water intru-

dow. In parallel, preliminary simulation results were presented and com-

Method for Determining Condensation Potential

TABLE 1List of standards for determining condensation potential of windows.

Outline of Evaluation Program

FIG. 1Schematic of anged window installation in window opening showing the

including size of frame, sheathing board, insulation, and window types,

Experimental Setup and Procedures

Conguration of Test Specimen and Mounting Specimen in Test Assembly

FIG. 3Vertical section of specimen showing installation details for non-operable

TABLE 3Test series for anged window conguration.

Flanged Installation Test Set 4 Test Set 5 Test Set 6

Position Flange Flange Flange Flange Flange Flange

Instrumentation and Data Acquisition

Measurement of Temperatures and Relative HumidityThe chambers envi-

Monitoring Chamber Conditions. Both the temperature and RH were con-

FIG. 4Isometric view of specimen; shows location of deciency at lower corner of

FIG. 5Nominal location of thermocouples on exterior and interior face of test

Monitoring and Recording Surface Temperature Condition. Surface tempera-

Measurement of Pressure Differentials and AirowTests were carried out

Calibration of Test FacilityThe guarded hotbox test facility is calibrated

Test Procedure Under No Pressure DifferentialThe tests in this instance

Test Procedure when Applying a Pressure DifferentialThe procedure for the