433 131 86MB
English Pages 489 [492] Year 2020
Lemieux | Keegan
ASTM INTERNATIONAL Selected Technical Papers Building Science and the Physics of Building Enclosure Performance
ISBN: 978-0-8031-7680-5 Stock #: STP1617 www.astm.org
STP 1617
ASTM INTERNATIONAL Helping our world work better
Building Science and the Physics of Building Enclosure Performance STP 1617 Editors: Daniel J. Lemieux Jennifer Keegan
SELECTED TECHNICAL PAPERS STP1617
Editors: Daniel J. Lemieux and Jennifer Keegan
Building Science and the Physics of Building Enclosure Performance ASTM STOCK #STP1617 DOI: 10.1520/STP1617-EB
ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959 Printed in the U.S.A.
Library of Congress Cataloging-in-Publication Data Names: Keegan, Jennifer, editor. | Lemieux, Daniel J., editor. Title: Building science and the physics of building enclosure performance / editors, Jennifer Keegan, Daniel J. Lemieux. Description: West Conshohocken : ASTM International, 2020. | Series: Selected technical papers STP1617 | Includes bibliographical references. | Summary: “This compilation of Selected Technical Papers, STP1617, Building Science and the Physics of Building Enclosure Performance, contains peer-reviewed papers that were presented at symposiums held October 21-22, 2018, and December 2, 2018, in Washington, DC, USA. The symposiums were sponsored by ASTM International Committee E06 on Performance of Buildings and Committee D08 on Roofing and Waterproofing”-- Provided by publisher. Identifiers: LCCN 2020007959 (print) | LCCN 2020007960 (ebook) | ISBN 9780803176805 (paperback) | ISBN 9780803176812 (pdf) Subjects: LCSH: Buildings--Thermal properties. | Dampness in buildings. Classification: LCC TH6025 .B827 2020 (print) | LCC TH6025 (ebook) | DDC 720/.472--dc23 LC record available at https://lccn.loc.gov/2020007959 LC ebook record available at https://lccn.loc.gov/2020007960 ISBN: 978-0-8031-7680-5 C 2020 ASTM INTERNATIONAL, West Conshohocken, PA. All rights reserved. This material Copyright V may not be reproduced or copied, in whole or in part, in any printed, mechanical, electronic, film, or other distribution and storage media, without the written consent of the publisher.
Photocopy Rights Authorization to photocopy items for internal, personal, or educational classroom use, or the internal, personal, or educational classroom use of specific clients, is granted by ASTM International provided that the appropriate fee is paid to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, Tel: (978) 646-2600; http://www.copyright.com/ ASTM International 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 reflects 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,” in STP title, book editor(s) (West Conshohocken, PA: ASTM International, year), page range, paper doi, listed in the footnote of the paper. A citation is provided on page one of each paper. Printed in Hanover, PA June, 2020
Foreword THIS COMPILATION OF Selected Technical Papers, STP1617, Building Science and the Physics of Building Enclosure Performance, contains peer-reviewed papers that were presented at symposiums held October 21–22, 2018, and December 2, 2018, in Washington, DC, USA. The symposiums were sponsored by ASTM International Committee E06 on Performance of Buildings and Committee D08 on Roofing and Waterproofing. Symposium Chairs and STP Editors: Daniel J. Lemieux Wiss, Janney, Elstner Associates, Inc. Falls Church, VA, USA Jennifer Keegan GAF Materials Corporation Parsippany, NJ, USA
iii
Contents
Overview
vii Energy Use and Conservation
Energy Codes and Standards: Do They Reflect Best Practice in Building Envelope Performance? Tat S. Fu, Jason S. Der Ananian, and Brent A. Gabby Changes in Exterior Envelope Systems’ Thermal Performance Requirements and Impacts upon Design, Materials, Products, and Systems William Jensen and Paul G. Johnson
1
19
Building Physics of HVAC Interaction with Enclosure Systems and People Paul E. Totten and Charlotte Metzler
49
A Window of Opportunity Kyle Normandin and Robyn Pender
67
Predicted Thermal and Hygrothermal Improvements for a Mid-Twentieth-Century Brutalist Landmark Niklas W. Vigener and Anthony J. Nicastro
88
Analysis and Modeling Integrating Computational Fluid Dynamics Simulations and Condensation Risk Assessments: Using Project Specific Analyses to Better Inform the Design of Buildings Ali Moussawi and Scott N. Bondi Role of Initial Moisture Content on Hygrothermal Models and Envelope Performance Carly May Wagner and Rex A. Cyphers
v
107
125
Approach to Incorporating Water Entry and Water Loads to Wall Assemblies When Completing Hygrothermal Modelling Travis V. Moore and Michael A. Lacasse
157
Solutions to Address Osmosis and the Blistering of Liquid Applied Waterproofing Membranes Elyse A. Henderson, Graham Finch, and Brian Hubbs
177
Solutions and a Path Forward Design Criteria and Solutions to Common Issues in Building Envelope Design Adam L. Rizor and Aaron Corn Unintended Consequences: A Review of Critical Details, Serviceability, and Durability of Modern High-Performance Facades Jennifer Keegan and Matthew Ridgway Two Agencies, BECx, and Design-Build: Challenges and Opportunities for BECx with Multiagency Involvement and Design-Build Project Delivery Andrea DelGiudice
195
211
263
Building Enclosure Performance and Commissioning BECx: A Case Study for Lessons Learned Patrick G. Giblin, Elizabeth O. Cassin, Larry R. Meyers, and Chelsea F. Ames
280
Ventilation and Moisture Control in Architectural Metal Panel Roofing Systems Eric K. Olson and Anthony J. Nicastro
317
Air Barrier Performance and Life Cycle from Inception to Installation Benjamin Meyer, Maria Spinu, and Elizabeth Cassin
338
Thermal Bridging and Linear Thermal Transmittance Calculations for Balconies Mehdi Ghobadi, Josip Cingel, and Michael A. Lacasse
386
Evaluation and Strategies for Existing Buildings Moisture Reduction Strategies for Building Envelopes Wade L. Vorley and Lauran Drown
396
Assessment of Water Damage in a Mass Masonry Wall Building E. Webb Wright
420
Evaluating the Impact of Moisture Content on Thermal Resistance of Mass Masonry Wall Assemblies Rex A. Cyphers and Jodi M. Knorowski Thermal Performance of Spandrel Assemblies in Glazing Systems John A. Jackson, Cheryl M. Saldanha, Gert Guldentops, Jin Rui Yap, Robert J. Abdallah, and Sarah B. Rentfro
vi
435
450
Overview The Joint Symposium on Building Science and the Physics of Building Enclosure Performance was held in Washington, DC, beginning with Part 1 on October 21 and 22, 2018, and concluding with Part 2 on December 2, 2018. Sponsored by ASTM Committee E06 on Performance of Buildings and Committee D08 on Roofing and Waterproofing, this first joint symposium brought together subject-matter experts from both committees in addition to students, teachers, research scientists, conservators, and practicing professionals from both the public and private sectors in North America, the United Kingdom, Europe, and Asia. The objective of the symposium was to provide a forum for the exchange of ideas on current research regarding building science and the physics of building enclosure and whole-building performance, including testing and assessment of building envelope heat, air, and moisture transfer, energy use, and our environment. Areas of inquiry included:
Building Science Fundamentals Physics of Heat/Air/Moisture Transfer Energy Modeling: Facts, Myths and Legends Human Comfort and Productivity in Our Built Environment Critical Review of Building Enclosure Standards and the Codes Material Selection for Climate-Specific Durability and Performance Adaptive Re-Use and Conservation in Truly Sustainable Design Building Enclosure Commissioning: Promises Kept Lessons Learned and the Road Ahead
Dr. John Straube (RDH Building Science, Inc.) served as keynote speaker for Part 1 of our joint Symposium. John is a professor of building science in the Civil Engineering Department and School of Architecture at the University of Waterloo, and his research focuses on energy-efficient, healthy, durable, and sustainable building design supported by advanced computer simulation, laboratory testing, and full-scale natural exposure performance monitoring. John built upon his message as keynote speaker for our previous ASTM/National Institute of Building Sciences (NIBS) Workshops on Building Science Education in North America (ASTM/NIBS Workshops on Building Science Education in North America, Toronto, Ontario, Canada in October, 2016, and Washington, DC in April, 2017) to challenge all of us to reconsider the role of building science
vii
education in delivering—rather than simply promising—quantifiable performance in our built environment. For John, standards development for standards-development sake is not the answer. Improved and more readily accessible opportunities to pursue graduate, post-graduate, and continuing education in building science fundamentals and a deeper appreciation for practical solutions to the challenges that we face in our built environment were the core of his message and helped fuel the discussion and debate that followed during our closing plenary session for Part 1 of the symposium. Kevin Kampschroer (Office of High Performance Buildings and US General Services Administration) was our second keynote speaker. As the largest property owner in the United States, Kevin brought the owners’ perspective to the discussion and attributed the apparent success of building enclosure performance to Building Enclosure Commissioning (BECx). He spoke of the cascading effect of design and encouraged the community to embrace high performance buildings and leverage the power of BECx, stating, “If you set a standard that is achievable, people will get excited about it.” Roy Wright (Insurance Institute for Business & Home Safety) encouraged listeners to focus on the people that rely on the practices and standards development, so that they can withstand these stronger weather-related events that have come and those we know will come. He challenged us to ask, how can we improve our standards to reduce the risk for future disaster? How can we improve the process to inspire pace, not haste, to get these standards approved more expeditiously? And how do we activate community leaders to create opportunities to build better? Resilience requires leadership. Rene Dupuis (Structural Research Inc.) encouraged us to continue to foster a more cross-disciplined approach at ASTM that reaches across silos. In his experience with failure analysis, this collaborative focus may reduce the amount of system-related failures in the future. Paul Johnson (SmithGroup) is an ASTM Fellow and active member of ASTM Committees E06 and D08. He brought the architects’ perspective to the discussion, encouraging the holistic view of the building in which the enclosure is a critical part. Buildings need to serve people, and it is incumbent upon our committees to collaborate in order to strengthen our standards from a building performance perspective. We would also like to take this opportunity to offer our sincere thanks to Steve Mawn and Joe Hugo, our respective ASTM Staff Managers for Committees E06 and D08, for their tireless patience and support. We share a common goal to encourage a more cooperative, cross-disciplined, and fully transparent approach to standards development at ASTM. Steve, Joe, and the ASTM Board of Directors are firm believers in that effort. This symposium and the symposia that will follow are in no small part a direct result of their insight, wisdom, and learned guidance.
viii
A special thanks to Dr. Robyn Pender (Historic England) for joining us from London to offer her insight and expertise on this topic. Robyn is a renowned author, building physicist, and conservator whose work includes landmark buildings and structures throughout greater London and across the United Kingdom. She served as a panelist alongside Kampschroer and Straube during Part 1 of our symposium and previously as a keynote speaker for the 2017 Symposium on Building Physics and Conservation at Southbank Centre, London. Many thanks also to Kelly Dennison, Alyssa Conaway, Sara Welliver, and all of the truly dedicated professional staff at ASTM for making this first joint symposium an overwhelming success for both committees. Teamwork and a considerable amount of patience—often beyond our view and behind the curtain—came together to make it all look so easy. Thank you. This symposium would not have been possible without our sponsors, who supported us in bringing these critical discussions to the industry. A very special thank you to our platinum sponsors: Building Envelope Technologies, RCI/IIBEC, Simpson Gumpertz & Heger, Wiss, Janney, Elstner Associates, Inc., and WSP. Thank you to our silver and bronze sponsors: CDC, Cetco, GAF, NRCA, Rimkus, Soprema, and Tremco. We are grateful for your support. And finally, a very special thanks to our keynote speakers, authors, presenters, technical reviewers, and invited and registered guests for donating your time, interest, passion and expertise. For three days in 2018 we were your students—humbled by your knowledge and grateful for your continued contributions to our industry and to the advancement of our profession. It has been a privilege to serve each of you and to serve ASTM as Co-Chair of this symposium. Symposium Chairs and STP Editors: Daniel J. Lemieux, AIA, NCARB, RIBA, MRICS Director and Principal Wiss, Janney, Elstner Associates, Inc. (USA) Wiss, Janney, Elstner Limited (UK) Jennifer Keegan, AAIA Director Building and Roofing Science GAF Materials Corporation
ix
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180086
Tat S. Fu,1 Jason S. Der Ananian,2 and Brent A. Gabby3
Energy Codes and Standards: Do They Reflect Best Practice in Building Envelope Performance? Citation T. S. Fu, J. S. Der Ananian, and B. A. Gabby, “Energy Codes and Standards: Do They Reflect Best Practice in Building Envelope Performance?,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 1–18. http://doi.org/10.1520/STP1617201800864
ABSTRACT
Energy codes and standards allow multiple paths for compliance. This paper addresses the prescriptive approach only, which stipulates minimum R-values of building enclosure insulation, fenestration thermal performance requirements, and maximum building envelope air leakage rates. Recent changes within energy codes and standards commonly reflect a “more is better” strategy regarding building envelope performance (e.g., wall and roof insulation). However, is this a sound strategy due to the diminishing energy savings as the building envelope performance improves (e.g., increasing minimum R-values)? This paper considers recent trends in energy codes and standards in the context of an energy modeling case study of an existing office building. The paper will discuss the results of parametric studies on a validated building energy model with different roof insulation minimum R-values, as well as building envelope air leakage rates and window glazing types. The case study building is an 86,400 ft2 single-story office building located in Climate Zone 5 (near Boston, MA) with a recently renovated building envelope that included new roofing, wall cladding, and windows. This building’s energy model was calibrated by incorporating data
Manuscript received October 15, 2018; accepted for publication April 26, 2019. 1 Structural Engineering/Structural Mechanics Division, Simpson Gumpertz & Heger Inc., 41 Seyon St., Building 1, Suite 500, Waltham, MA 02453, USA https://orcid.org/0000-0002-8743-2759 2 Building Technology Division, Simpson Gumpertz & Heger Inc., 41 Seyon St., Building 1, Suite 500, Waltham, https://orcid.org/0000-0002-6727-8539 MA 02453, USA 3 Brent Gabby Consulting, LLC, 360 North Emerson Rd., Lexington, MA 02420, USA 4 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
1
2
STP 1617 On Building Science and the Physics of Building Enclosure Performance
from a quantitative whole-building air leakage test, an on-site weather station, and metered energy consumption of electric and mechanical systems. The paper examines the following variables with respect to the predicted heating, ventilation, and air conditioning (HVAC)-related energy consumption: roof insulation minimum R-values, building envelope air leakage rates, window glazing types (e.g., single-, double- or triple-pane glazing, low-E coating types), and window-to-wall ratios. Finally, this paper compares the cost implications of some of these variables. While the lessons learned from this study can be applied to new construction, the results of this study are more applicable to renovations of existing buildings and their envelope systems. Keywords air leakage, energy, roof insulation, whole-building air leakage testing, thermal performance, building science, windows, glazing
Introduction As the demand for energy-efficient buildings and the cost of electricity and heating fuel have increased, energy codes such as ASHRAE 90.1, Energy Standard for Buildings Except Low-Rise Residential Buildings, and the International Energy Conservation Code (IECC) have increasingly become stricter as they relate to the building envelope. For example, since the 2007 ASHRAE 90.1 Standard1 and 2009 IECC,2 the minimum roof insulation prescriptive R-values have increased by up to 50% in some climate zones. Additionally, the maximum allowable window-to-wall ratios have been reduced and building enclosure airtightness has become a requirement in the energy code. This paper reports the results of a case study using whole-building energy simulation software to study the heating, ventilation, and air conditioning (HVAC)-related energy and cost impact of envelope code changes in an existing 86,400 ft2 office building located in Climate Zone 5. The case study office building is calibrated such that the simulation results closely match actual building energy consumption.3 Although the case study presented in this paper is similar to the Pacific Northwest National Laboratory’s (PNNL) 2015 cost-effectiveness analysis4 of the ASHRAE Standard 90.1-20135 changes (from 20106) in that they both include a costbenefit study of different building components in terms of their initial cost and their effects on the subsequent energy consumption,4 the following identifies several important differences between the two studies. First, the PNNL study covered a variety of buildings and climates, including six prototype buildings (small office, large office, retail, school, small hotel, and midrise apartment) in five climates (Climate Zones 2A—Houston, 3A—Memphis, 3B—El Paso, 4A—Baltimore, and 5A—Chicago). Due to its breadth, the PNNL study analyzed only the changes between the 2010 and 2013 versions of ASHRAE Standard 90.1, and much of its focus was not on envelope components. For example, in the large office prototype, only one of the 16 major addenda was related to envelope components, and
FU ET AL., DOI: 10.1520/STP161720180086
remaining addenda were related to HVAC and lighting. In this case study, we focused on envelope components (i.e., roof insulation, fenestration, and air leakage) and, because of the limited components and only one building, we examined not only the code changes but also a range of possible thermal performance values related to the components. For example, whereas the PNNL study only considered roof insulation changing from R-25 to R-30, this case study included ten different roof R-values between R-0 and R-50. A second difference is that our case study used installation costs provided by a large construction firm* in the United States to conduct a cost-benefit analysis on each individual building envelope component, while PNNL reported the cost results combining all changes to envelope, HVAC, and lighting systems. In other words, this paper examines the cost impact of each variable independently (i.e., roof insulation and fenestration glazing type), whereas the PNNL study examined overall effect of the code cycle changes for multiple variables. Our study excluded a costbenefit analysis for building enclosure airtightness due to unknown and variable costs associated with constructing tighter building envelopes. A third difference is that our study was based on applied-research on an existing office building by calibrating an energy model of the building and not utilizing a prototype PNNL building. Table 1 lists some characteristics of the case study building compared to the PNNL prototype that is most similar, namely the large office building. Previous studies7–9 showed significant discrepancies between predicted and actual energy usage. These discrepancies limit the applicability of purely simulation-based studies such as the PNNL study. On the other hand, simulationbased studies are almost the only reliable way to test the cost effectiveness of a building component. Another method, for example, is to build two identical buildings next to each other with different roof insulation minimum R-values and study the buildings’ energy consumptions over a long period of time to account for weather variability, a method applying mostly to experimental testbeds and not actual buildings. Our study threaded a compromise between validation- and simulation-based studies by calibrating a building energy model of an in-service building with a variety of data and conducting parametric studies of various envelope changes. This approach enabled simplified cost-benefit comparisons while allowing the flexibility of simulation-based studies.
Building and Energy Codes Building and energy codes provide both prescriptive and performance-based thermal performance requirements for commercial building envelopes in the United States. Building owners, such as the General Services Administration (GSA) or other institutional owners, may also establish more strict performance requirements. The *
Skanska Building USA, Boston, MA, 02210.
3
4
Case Study
Floor Area (ft2)
Number of Floors
Window-toWall Ratio
86,400
1
26%
Building
Exterior Wall
Occupancy (People/ 1,000 ft2)
Plug Load
Lighting Load (W/ft2)
Insulation
EIFS with
5.8
170 W/person þ
1.00
above deck
4 in. of EPS
Roof
0.005 W/ft2
insulation PNNL
498,640
12
40%
Insulation above deck
Mass
5
Note: EIFS = exterior insulation and finish system; EPS = expanded polystyrene; VAV = variable air volume.
0.73 W/ft
2
0.82
HVAC Systems
•
VAV with reheat (primary)
•
Gas (heating)
•
Rooftop unit electric (cooling)
•
VAV with reheat (primary)
•
Boiler (heating)
•
Chiller, cooling tower (cooling)
STP 1617 On Building Science and the Physics of Building Enclosure Performance
TABLE 1 Case study building versus PNNL’s large office prototype4
FU ET AL., DOI: 10.1520/STP161720180086
prescriptive energy code requirements for building enclosures generally have shown an increase in stringency over the past several code cycles. We summarize the major differences in prescriptive energy code requirements between the IECC and ASHRAE Standard 90.1 as they relate to roof insulation minimum R-values, fenestration thermal performance, and building envelope air leakage. Table 2 summarizes modifications of building envelope performance requirements for the 20092, 2012,10 and 201511 IECC versions, while Table 3 summarizes similar requirements for the 20071, 20106, and 20135 versions of ASHRAE Standard 90.1 for the case study building located in Climate Zone 5 and described in Table 2.
Case Study Building The building utilized for this case study is located in Waltham, MA, and is approximately nine miles west of Boston (2015 IECC Climate Zone 5A; 2013 ASHRAE 90.1 Climate Zone 5) (fig. 1 and fig. 2).* A representative building energy model was created and calibrated on the basis of metered electricity usage, data from an onsite weather station (fig. 3), and results from a building envelope air leakage test. The building includes punched and clerestory windows with thermally broken aluminum frames and a glass and metal curtain wall at the front entrance. A past building renovation project included installing exterior insulation and finish system (EIFS) cladding over a majority of the existing brick masonry, ethylene propylene diene terpolymer (EPDM) membrane roofing with a uniform R-20
TABLE 2 Prescriptive IECC limits on select building envelope components (Climate Zone 5)
Maximum Allowable Building
IECC 20092
IECC 201210
IECC 201511
No requirements
0.40 CFM/ft2 at 75 Pa
0.40 CFM/ft2 at 75 Pa
R-20
R-25
R-30
40% max
30% max (40% max
30% max (40% max
with daylighting
with daylighting
control)
control)
Air Leakage Rate Roofs—Insulation Entirely Above Roof Deck Vertical Fenestration —Window-to-wall ratio
—U-factor (fixed)
0.45
0.38
0.38
—U-factor (operable)
0.45
0.45
0.45
—SHGC
0.40a
0.40
0.40b (South, East,
(solar heat gain coefficient)
and West) 0.53b (North)
a
Glazing with a projection factor less than 0.25. Glazing with a projection factor less than 0.20.
b
*Simpson Gumpertz & Heger Inc., former headquarters, Waltham, MA 02453.
5
6
STP 1617 On Building Science and the Physics of Building Enclosure Performance
TABLE 3 Prescriptive ASHRAE 90.1 limits on select building envelope components (Climate Zone 5)
Maximum Allowable Building Air
ASHRAE 90.1-20071
ASHRAE 90.1-20106
No requirement
No requirement
Leakage Rate
ASHRAE 90.1-20135
0.40 CFM/ft2 at 75 Pa when field tested
Roofs—Insulation Entirely Above Roof
R-20
R-20
R-30
40% max
Deck Vertical Fenestration —Window-to-wall ratio
40% max
40% max
—U-factor (metal-framed fixed)
–
–
0.42
—U-factor (metal-framed operable)
–
–
0.50
—U-factor (metal framing—all other)
0.55
0.55
–
—SHGC
0.40
0.40
0.40
(solar heat gain coefficient)
FIG. 1 Exterior view of the case study building.
polyisocyanurate insulation over a sloped structural roof deck, and new punched windows, central clerestory windows, and front entrance glass and metal curtain wall with insulating glass units (IGUs). The renovation also included installing new forced-air HVAC systems with outdoor air ventilation. WHOLE BUILDING AIR LEAKAGE MEASUREMENT
In May 2015, we measured the building envelope air leakage rate (fig. 4). The primary air leakage path identified during the test was through the clerestory window perimeter. The measured building envelope air leakage rate of 0.26 CFM/ft2 at
FU ET AL., DOI: 10.1520/STP161720180086
FIG. 2 Interior view of the case study building with mostly open office areas.
FIG. 3 Weather station installed on the case study building roof.
75 Pa was below the IECC/ASHRAE 90.1 limit (0.40 CFM/ft2) and approximately equivalent to the U.S. Army Corp of Engineers’ requirement (0.25 CFM/ft2).12,13 We then coupled a computational fluid dynamics (CFD) analysis of the building envelope with the measured air leakage rate to model air infiltration and exfiltration through the building envelope under differing weather conditions.14 (Refer to Fu and Lyon3 for more information regarding the model calibration.) BUILDING ENERGY MODEL SIMULATION
The case study building includes 80 mechanical zones, as shown in figure 5. We used Blender, an open-source three-dimensional (3D) modeling software, to model the building geometry. ODS Studio, a plugin for Blender, was then used to
7
8
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 4 Air leakage test in the case study building.
FIG. 5 Blender model of the case study building.
convert the model into an EnergyPlus input file. EnergyPlus is a building energy simulation program developed by the U.S. Department of Energy that models both energy consumption and water use in buildings. Additionally, we used ODS Studio for a CFD analysis (fig. 6) and fed the results into an EnergyPlus model to
FU ET AL., DOI: 10.1520/STP161720180086
FIG. 6 Computational fluid dynamics model of the case study building.
create an airflow network incorporating weather data and the measured building envelope air leakage rate. We calibrated the EnergyPlus model to reduce the difference between the simulated and metered electricity usages using on-site weather data. The calibrated model exhibited a coefficient of variation of the root mean square error (CVRMSE) of 5% using the calculation method published in the ASHRAE Guide 14.15 For calibration, we used weather data from the on-site weather station. However, for the parametric analysis of building envelope air leakage rates, roof insulation minimum R-values, and fenestration configurations, we used the BostonLogan International Airport TMY3 weather file (the airport is located approximately 11 miles east of the case study building). TMY3 files are “typical meteorological year” weather files that cover the period of 1991 to 2005. Also, for the parametric study, we did not resize the mechanical systems for simplicity and consistency given that we calibrated the energy model to the as-built mechanical system. We recognize that utilizing more efficient mechanical systems for each of these configurations provides an opportunity to reduce building energy consumption. COST ANALYSIS
Utilizing cost data provided, we combined the average cost for each roof insulation and glazing configuration with their associated surface areas to calculate initial costs for the various options studied. To calculate the HVAC-related energy costs, we identified the average monthly costs for commercial electricity and gas for
9
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Massachusetts from the Energy Information Administration.16,17 To calculate the annual cost of energy, we multiplied the predicted monthly energy consumption by the corresponding average month electricity/gas cost and summed the monthly electricity and gas cost for the year. When we projected energy cost over a 20- or 25-year period, we assumed a 2.5% annual inflation rate. A limitation of this cost analysis is that it does not take into consideration potential costs savings related to downsizing of the HVAC system, duct work, or other systems associated by utilizing a more efficient building enclosure.
Results and Discussion Utilizing the calibrated energy model, we conducted several studies to examine the effect of prescriptive code requirements on HVAC-related energy usage with respect to building envelope air leakage rates, roof insulation minimum R-values, and fenestration configurations. We examined the cost benefit of various roof insulation minimum R-values and fenestration configurations using the cost data described previously. BUILDING ENVELOPE AIR LEAKAGE RATES Figure 7 shows HVAC-related energy usage for the same building model when sim-
ulated with building envelope air leakage rates ranging from 0.01 to 2.0 CFM/ft2 at 75 Pa. The baseline case was the model with the measured building envelope air leakage rate of 0.26 CFM/ft2 at 75 Pa. This figure also indicates that the case study FIG. 7 HVAC-related energy usage versus air leakage rate. 40 60 35.0%
50 30
27.2% 40
22.9% 20
17.7%
30
12.3% 20 10 5.9% 10 0.0% baseline
0
0
-4.2% -5.3% -5.8% -6.2%
-10
-10 0 0.08
0.26
0.4
0.6
0.8
1
1.2
1.4
Building enclosure air leakage rate (CFM/ft2 )
1.6
1.8
2
HVAC annual $x000 difference to baseline case
31.4%
HVAC annual usage as a % to baseline case
10
FU ET AL., DOI: 10.1520/STP161720180086
building with a higher building envelope air leakage rate increased the predicted HVAC-related energy consumption while the most airtight building used the least amount of predicted HVAC-related energy. A higher air leakage rate in a building allows more air exchanges across the building envelope and thus requires more heating/cooling capacity and energy to maintain the desired setpoint temperatures. Figure 7 also shows that the building envelope air leakage rate and predicted HVAC-related energy usage have an approximate linear relationship for the case study office building; the rate of change in the predicted HVAC-related energy usage was relatively consistent for both higher and lower air leakage rates. This trend aligns with the prescribed IECC and ASHRAE 90.1 requirements and other industry guidelines on limiting building envelope air leakage rates to certain values. For the case study building, reducing the building envelope air leakage rate from an existing 0.26 CFM/ft2 at 75 Pa to 0.08 CFM/ft2 at 75 Pa (PHIUSþ2015 requirement18) subsequently reduced the predicted HVAC-related energy usage by 4.6% annually. Increasing the building envelope air leakage rate up to 0.40 CFM/ft2 at 75 Pa (IECC/ASHRAE 90.1 limit) increased the predicted annual HVAC-related energy usage by 3.2%. We have tested existing buildings with envelope air leakage rates exceeding 2.0 CFM/ft2 at 75 Pa, which would increase the predicted annual HVAC-related energy usage of the case study building by 35% or greater. ROOF INSULATION MINIMUM R-VALUES Figure 8 shows that the simulated HVAC-related energy usage decreases as the roof Rvalue increases as indicated in figure 8. However, unlike figure 7, which illustrates the general linear relationship between building envelope air leakage rates and HVACrelated energy usage, there is a diminishing return with more roof insulation. For the case study building, most of the HVAC-related energy reduction occurred between R-0 and R-20, excluding the impact of thermal bridging sources through the insulation
FIG. 8 Predicted energy usage versus roof insulation HVAC annual usage as a % to baseline case
70 65.7
280
60
260
HVAC annual cost ($x000) 264
50
240
40 220 30 22.6
200
195
20 180
8.8
10
2.8 0
173 164
0.0
-1.9
-4.2 -5.0 -5.6 -6.1
-10
160
159 156
153 151 150 149
140 0
10
20
30
40
R-value of roof insulation
50
2
4
6
Insulation thickness (inch)
8
11
12
STP 1617 On Building Science and the Physics of Building Enclosure Performance
(i.e., roof fasteners and plates, dunnage, vent pipes, etc.). A recent study conducted by Olson, Saldanha, and Hsu showed a 17% reduction in the effective roof R-value when accounting for thermal bridging sources such as insulation fasteners and plates and other rooftop penetrations.19 However, our case study did not examine this reduction to due to inconsistencies in the roof construction. We conducted a simple cost-benefit analysis on HVAC-related energy and roof insulation minimum R-values. The installed cost of polyisocyanurate insulation was estimated to be $1.20 per in. per ft2, including material and labor. Assuming a 20-year service life for the EPDM roof assembly, we compared the 20-year cost of different roof insulation minimum R-values, factoring in the initial installation cost and the 20-year HVAC-related energy cost. Thinner roof insulation (i.e., less minimum roof R-value) would be less expensive to install but results in higher HVACrelated energy costs. In contrast, thicker roof insulation (i.e., higher minimum roof R-value) would have a higher initial cost but reduces HVAC-related energy usage to a limited degree. Understanding the cost analysis limitations described previously, the roof assemblies with the lowest overall cost included those with approximately 2.5 to 4.5 in. of polyisocyanurate insulation (approximately R-15 to R-25) for the case study building. Figure 9 also shows a cost saving in adding the first few inches of roof insulation due to the exponentially diminishing return in increasing roof insulation R-value (thickness) as shown in figure 8. The case study building showed that the optimal R-15 to R-25 range of roof insulation aligned closely with the R-20 and R-25 requirements in the 2009 and 2012 IECC, respectively. The 2015 IECC and ASHRAE-90.1-2013 requirement of R-30 may result in a less cost-effective design for the case study building based on the simple cost analysis described earlier.
FIG. 9 Expected 20-year cost versus roof insulation
HVAC$ 20 years + Insulation cost ($x1M)
7 6.74 6.5 6 5.5
5.08 5 4.61 4.5 0
10
4.45 4.42 4.44
20
30
Insulation R-value
4.66 4.73 4.53 4.59
40
50
FU ET AL., DOI: 10.1520/STP161720180086
FENESTRATION PERFORMANCE
For fenestrations, we modeled the framing with ten glazing configurations (Table 4) and assessed their impacts on HVAC-related energy usage. These configurations include some of the most common glazing types, such as single-, double-, and
TABLE 4 Whole fenestration assembly values
Glazing Options
# Panes
Gas Filler
U-Factor
Coating, Surface
ID
Btu h ft2 F
Allowed Per
W m2 K
SHGC
2015 IECC?
ASHRAE 90.1-2003?
1
n/a
none
1-p
1.025
20,953
0.818
No
No
2
air
none
2-p Air
0.474
9,689
0.704
No
No Yes
3
air
low-e, 2
2-p Air Low-E
0.291
5,949
0.392
Yes
argon
none
2-p Argon
0.448
9,158
0.705
No
No
argon
low-e, 2
2-p Argon Low-E
0.245
5,008
0.387
Yes
Yes
air
low-e, 2 & 4
2-p Air Low-E4
0.235
4,804
0.361
Yes
Yes
air
none
3-p Air
0.345
7,052
0.616
No
No
air
low-e, 2
3-p Air Low-E
0.266
5,438
0.331
Yes
Yes
argon
none
3-p Argon
0.314
6,419
0.617
No
No
argon
low-e, 2
3-p Argon Low-E
0.219
4,477
0.328
Yes
Yes
Note: Refer to figure 10 for glazing configuration details.
FIG. 10 Glazing by number of glass panes.
13
STP 1617 On Building Science and the Physics of Building Enclosure Performance
triple-pane IGUs as well as different filler gases and the presence of low-E coatings. Of the ten configurations, five of them do not satisfy the 2015 IECC requirements for buildings located in Climate Zone 5—mostly because the IGUs exhibit high solar heat gain coefficients (SHGCs). The case study building includes double-pane IGUs with a low-E coating on the Number 2 surface (or “2-p Air Low-E”) installed with a window-to-wall ratio (WWR) of 26%. We refer to this configuration as a baseline for comparing performance of various fenestration configurations. Figure 11 shows the predicted annual HVAC-related energy usage for the ten fenestration configurations listed in Table 4 with a 26% WWR. For the case study building, the single-pane fenestration exhibited the highest annual HVAC-related energy costs, which was primarily due to higher terminal reheat costs in the winter. Among the multipane configurations, triple-pane uncoated glazing resulted in lower HVAC-related energy costs than double-pane, uncoated counterparts primarily due to lower U-factors. On the other hand, despite having lower U-factors, triple-pane low-e coated glazing resulted in higher heating and overall energy costs than their double-pane counterparts (when both are low-e coated). This was likely because the SHGCs had greater impact on the heating costs compared to U-factors among the low-e coated glazing. The loss of solar heat gain outweighed the greater insulating value of the glazing and therefore increased the heating costs. Triple-pane fenestrations also had
FIG. 11 Predicted HVAC-related energy usage versus fenestration configuration.
5
Fan Cooling Heating
Total: 4.85 4.33
4.14
4.26
3.25
3.00
3.17
0.83
0.79
0.83
0.79
0.31
0.30
0.31
0.30
4.25
4.25
4.22
4.17
4.23
4.18
3.69
3.10
3.14
3.06
3.06
3.13
3.04
0.84
0.84
0.80
0.85
0.81
0.80
0.31
0.31
0.30
0.31
0.30
0.30
4
3
E n
Lo w
Ar go n go Ar p 3-
ow
E
p 3-
Ai r
Ai rL p
3-
3-
pa
ne
ow
w
Ai rL p
2-
go Ar p 2-
E4
E
n
Lo n
Ar p 2-
Lo
w
go
E
Ai r Ai r
ne 2p
pa 2-
0
pa
1
ne
2
1-
Annual HVAC energy usage (J x10 12 )
14
FU ET AL., DOI: 10.1520/STP161720180086
lower cooling costs compared to double-pane fenestrations, but the improvements were smaller than the increase in heating costs. Configurations “3-p Argon” and “3-p Air” exhibited the lowest and the thirdlowest overall HVAC-related energy usage, respectively, and yet they are not permitted for use in Climate Zone 5 per the 2015 IECC and ASHRAE 90.1-2013. Their lack of low-e coating in the glazing resulted in SHGCs that exceeded the limits of IECC and ASHRAE 90.1. Nonetheless, their high SHGCs also allowed more solar heat gain through the glazing and thus reduced the heating cost during the winter in a climate zone where heating loads dominate. We also conducted a cost-benefit analysis of the ten fenestration configurations assuming a 25-year lifecycle for the glazing systems. Figure 12 shows the initial cost of the configurations, while figure 13 shows the overall 25-year cost of the configurations combining both the initial and annual energy costs. Although Configuration “3-p Argon” was the most cost-effective option, both the 2015 IECC and ASHRAE 90.1-2013 would not permit this assembly because the SHGC exceeded the allowable prescriptive limit. For WWRs, we compared the predicted annual heating, cooling, and total HVACrelated energy use for each fenestration configuration to the baseline “2-p Air Low-E” configuration (fig. 14). In terms of overall HVAC usage, all fenestration configurations performed worse with increasing WWRs (fig. 14A). Single-pane fenestrations increased HVAC usage at the highest rate (i.e., slope in fig. 14A) with increasing WWRs, while double- and triple-pane fenestrations exhibited similar usage increase rates with
FIG. 12 Initial cost of fenestration configurations.
Fenestration cost ($x1M): material + labor 1.51 1.5
1.46
1.45
1.40 1.37
1.4 1.32
E Lo w
go n
3-
p
Lo w E
Ai r
Ai r
3-
p
pa ne
3-
E
Lo w E4
Ai r
Lo w p 2-
go n
go n
Ar Ar
Lo w
E 2p
2p
2-
p
Ai r
ne
Ai
r
an e
pa 2-
1p
Ar
1.21 1.2 1.17
go n
Low-E Single-pane Double-pane Triple-pane
1.26
Ar
1.27
3p
1.3
15
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 13 Expected 25-year cost of fenestration configurations.
HVAC + Fenestration cost ($x1M)
175 172.46
Low-E Single-pane Double-pane Triple-pane
170 165 160
155.34 155
151.73 151.75
152.82
151.46
150.67
150
149.80
149.17
148.45
E
n
2-
Lo w
3-
p
3-
Ar
3-
p
go n
Ar
rL ow
go
E
Ai r
Ai p
3pa ne
E Ai p 2-
p
rL ow
Lo w
go n Ar
2p
go n
Ar
E
r
rL ow
Ai
Ai p 2-
2pa ne
E4
145
1pa ne
FIG. 14 Annual HVAC usage related to WWR. (A) HVAC annual usage as a % to baseline case
(B) Heating annual usage as a % to baseline case (C) Cooling annual usage as a % to baseline case
35
25
40
30
no coating
35
}
20 30 1-pane 2-pane Air 2-p Air LowE 2-p Argon 2-p Argon LowE 2-p Air LowE4 3-pane Air 3-p Air LowE 3-p Argon 3-p Argon LowE baseline
25
20
15
10
25 15 20 15
10
}
10 5
Low-e coating
16
5 5 0 0 0
-5 20
-5
25
30
35
40
45
Window to Wall ratio (%)
50
55
-10 20
25
30
35
40
45
Window to Wall ratio (%)
50
55
-5 20
25
30
35
40
45
50
55
Window to Wall ratio (%)
increasing WWRs. Cooling loads increased mostly linearly with increasing WWRs, with low-e coated glazing increasing at a slower rate compared to uncoated glazing (fig. 14C). Heating loads increased at a much lower rate (i.e., flatter slopes) than the cooling loads (fig. 14B), except in single-pane fenestrations. In many configurations, the increases were effectively flat. Configuration “3-p Argon” was the best performing system in overall HVAC-related energy usage (fig. 14A). The heating load increases due to larger WWRs were relatively flat for most of the double- and triple-pane fenestrations. This is because there was a tradeoff
FU ET AL., DOI: 10.1520/STP161720180086
between the larger windows allowing more heat (solar) gain while losing more heat due to the lower insulation values in building envelopes with higher WWRs. On the other hand, in the cooling season, buildings with higher WWRs always performed worse because of heat gains from both direct sunlight and warmer outside temperatures. Overall, building envelopes with larger WWRs performed worse with respect to HVAC-related energy usage. This aligns with both the prescriptive IECC and ASHRAE 90.1 requirements that limit WWRs to 30% and 40%, respectively.
Conclusions Using a calibrated energy model of an actual building in Climate Zone 5, we studied the impacts of prescriptive energy code requirements on three building envelope components: air leakage rates, roof insulation minimum R-values, and fenestration. In our simulated results, we determined that the annual HVAC-related energy consumption consistently decreased when the building envelope’s air leakage rate decreased. This aligns with the IECC and ASHRAE 90.1 recent adoption of stricter building envelope airtightness requirements. For the roof insulation configurations, we identified an optimum insulation to be R-20 (excluding thermal bridging sources) and that higher R-values have a diminishing energy savings at a higher initial cost. The prescriptive R-30 roof insulation value required by the 2015 IECC and ASHRAE 90.1-2013 exceeds the optimum continuous roof insulation R-value identified by this study. The energy codes and standards also set restrictions on prescriptive fenestration performance, and our study showed general agreement with the codes on WWRs but not on SHGCs. The SHGC limits proved to be less energy efficient in the case study building in Climate Zone 5 because more heat gain (i.e., greater SHGC values) lowered the heating cost. This case study demonstrates the benefit of developing calibrated energy models to more accurately assess the cost-benefit of proposed envelope modifications to existing buildings. Using calibrated models can help architects, engineers, and building owners make more informed design decisions related to enclosure modifications and replacement HVAC systems using projected cost analysis. ACKNOWLEDGMENTS
We appreciate the support of Simpson Gumpertz & Heger Inc. for this work and the cost data provided by Skanska Building USA.
References 1.
Energy Standard for Buildings Except Low-Rise Residential Buildings, ASHRAE/IES Standard 90.1-2007 (Atlanta, GA: American Society of Heating, Refrigerating and AirConditioning Engineers, 2007).
17
18
STP 1617 On Building Science and the Physics of Building Enclosure Performance
2. 3.
4.
5. 6. 7. 8.
9.
10. 11. 12.
13. 14.
15. 16.
17.
18. 19.
International Code Council, International Energy Conservation Code (Country Club Hills, IL: International Code Council, Inc., 2009). T. Fu and E. Lyon, “Building Energy Model Calibration: A Case Study Using Computational Fluid Dynamics with Air Leakage Testing and On-Site Weather Data,” in Proceedings of the ASHRAE and IBPSA 2018 Building Performance Analysis Conference and SimBuild (Atlanta, GA: ASHRAE, 2018), 564–588. R. P. Hart, R. A. Athalye, M. A. Halverson, S. A. Loper, M. I. Rosenberg, Y. Xie, and E. E. Richman, National Cost-Effectiveness of ANSI/ASHRAE/IES Standard 90.1-2013 (Report No. PNNL-23824) (Richland, WA: Pacific Northwest National Laboratory, 2015). Energy Standard for Buildings Except Low-Rise Residential Buildings, ASHRAE/IES Standard 90.1-2013 (Atlanta, GA: ASHRAE, 2013). Energy Standard for Buildings Except Low-Rise Residential Buildings, ASHRAE/IES Standard 90.1-2010 (Atlanta, GA: ASHRAE 2010). J. H. Scofield and J. Doane, “Energy Performance of LEED-Certified Buildings from 2015 Chicago Benchmarking Data,” Energy and Buildings 174 (2018): 402–413. D. Oates and K. T. Sullivan, “Postoccupancy Energy Consumption Survey of Arizona’s LEED New Construction Population,” Journal of Construction Engineering and Management 138, no. 6 (2011): 742–750. C. Menassa, S. Mangasarian, M. El Asmar, and C. Kirar, “Energy Consumption Evaluation of US Navy LEED-Certified Buildings,” Journal of Performance of Constructed Facilities 26, no. 1 (2011): 46–53. International Code Council, International Energy Conservation Code (Country Club Hills, IL: International Code Council, Inc., 2012). International Code Council, International Energy Conservation Code (Country Club Hills, IL: International Code Council, Inc., 2015). D. Jones, B. Brown, T. Thompson, and G. Finch, “Building Enclosure Airtightness Testing in Washington State—Lessons Learned about Air Barrier Systems and Large Building Testing Procedures,” in Proceedings of the ASHRAE 2014 Annual Conference (Atlanta, ASHRAE, 2014), 43. L. Ricketts and G. Finch, Study of Part 3 Building Airtightness (Vancouver, BC: RDH Building Engineers Ltd., 2015). E. G. Lyon and C. M. Saldanha, “Integrating Whole Building Air Leakage Test Data into Energyplus Infiltration Models,” in Proceedings of SimBuild (Salt Lake City, UT: IBPSAUSA, 2016), 393–398. Measurement of Energy and Demand Savings, ASHRAE Guideline 14-2014 (Atlanta, GA: ASHRAE, 2014). U.S. Energy Information Administration (EIA), “Independent Statistics and Analysis: Average Retail Price of Electricity, Massachusetts,”2019, http://web.archive.org/web/ 20191220164830/https://www.eia.gov/electricity/state/massachusetts/ U.S. EIA, “Independent Statistics and Analysis Natural Gas. Massachusetts Price of Natural Gas Sold to Commercial Consumers,” 2019, http://web.archive.org/web/ 20191220165410/https://www.eia.gov/dnav/ng/hist/n3020ma3m.htm Passive House Institute U.S., PHIUSþ2015: Passive Building Standard–North America (Chicago, IL: Passive House Institute U.S., 2015). E. K. Olson, C. M. Saldanha, and J. W. Hsu, “Thermal Performance Evaluation of Roofing Details to Improve Thermal Efficiency and Condensation Resistance,” in Roofing Research and Standards Development: 8th Volume, ed. W. Rossiter and M. Sudhakar (West Conshohocken, PA: ASTM International, 2015), 44–67, http://doi.org/10.1520/ STP159020150021
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180114
William Jensen1 and Paul G. Johnson1
Changes in Exterior Envelope Systems’ Thermal Performance Requirements and Impacts upon Design, Materials, Products, and Systems Citation W. Jensen and P. G. Johnson, “Changes in Exterior Envelope Systems’ Thermal Performance Requirements and Impacts upon Design, Materials, Products, and Systems,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 19–48. http://doi.org/10.1520/ STP1617201801142
ABSTRACT
This paper addresses the evolution of ASHRAE Standard 90.1, Energy Standard for Buildings Except Low-Rise Residential Buildings, with regard to the continued long-term increase in thermal performance levels required of commercial building envelope and fenestration systems. The trend of increasing thermal performance levels for building envelopes has had a significant impact on design and the materials, systems, and products available for use in exterior building envelopes. This is particularly true when considering buildings of size or simplicity that do not warrant thermal modeling of the entire building or envelope system to validate compliance with mandated energy performance levels. As a widely recognized, and most often code-mandated, standard, ASHRAE 90.1 significantly impacts the selection of materials, products, and systems available for compliance with required energy standards. As increases continue through the most recent edition of ASHRAE 90.1 and as U-values become more stringent for materials, products, and systems in specific envelope applications, identifying options for the design of buildings becomes more
Manuscript received November 7, 2018; accepted for publication May 17, 2019. 1 SmithGroup, 500 Griswold, Detroit, MI 48556, USA 2 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
19
20
STP 1617 On Building Science and the Physics of Building Enclosure Performance
challenging. The demand for additional options also increases, which will hopefully spur greater interest in the development of new and improved envelope materials, systems, and products to meet the demands of aesthetics and the more stringent energy performance requirements. As requirements for energy performance increase, the number of buildings and envelope systems evaluated by thermal modeling to confirm compliance with new standards is also likely to increase. Compliance with the new ASHRAE 90.1 “prescriptive” requirements has been challenging and appears likely to become even more so. This paper includes consideration of a 20-year period (1999–2019) of evolving ASHRAE 90.1 standards and criteria for prescriptive building envelope thermal performance, and specific examples of the types of restraints currently being experienced by designers is included. Keywords exterior building envelope, ASHRAE 90.1, prescriptive value analysis
Introduction The American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) 90.1, Energy Standard for Buildings Except Low-Rise Residential Buildings,1 is and has been the reference standard for establishing maximum allowable energy usage in commercial buildings for many years. It is referenced by the International Energy Conservation Code (IECC)2 published by the International Code Council (ICC) and is the basis for most state-adopted energy codes in the United States. The standard was first published in 1975, and in 1999, the ASHRAE board of directors voted to place the standard on continuous maintenance based on rapid changes in energy technology, availability, and economic costs. ASHRAE 90.1 is currently reviewed and updated on a three-year cycle. As of this writing, the most recent publication date is 2019, which has yet to be adopted by most states and other governing authorities. ASHRAE 90.12016 is referenced by the ICC’s current International Building Code 2018 edition,3 which was finalized and published in 2018. ASHRAE 90.1 is wide-ranging and covers all aspects of commercial building energy performance, including energy usage during operations of the building. The minimal allowable levels of thermal performance criteria for building envelope systems are included under Chapter 5 of the ASHRAE 90.1, 2019 edition.1 Across the last 20 years, the minimal requirements for the performance of building envelope systems, as mandated by this standard, have been increased significantly. While the standard does contain three approaches for demonstrating compliance, this paper specifically addresses the “prescriptive” approach, which is the easiest approach and likely the most widely used by building envelope designers. There are several reasons why building designers are inclined to use the prescriptive method of compliance—the most significant being the level of expertise and time required to exercise the other two options of demonstrating compliance.
JENSEN AND JOHNSON, DOI: 10.1520/STP161720180114
When following the prescriptive approach to demonstrate compliance of the building exterior envelope with the goals of the code, there are significant challenges to building designs. This is in regard to attaining the minimum required thermal performance and still constructing buildings that will meet the basic tenants of architecture. Aside from thermal performance, these basic tenants are appearance, comfort, suitability for use, and the inclusion of daylighting. The prescriptive approach to compliance includes two basic criteria that must be met depending upon the specific element of the building being considered and the materials that will be used. The first is the solid/opaque element portion of the building being considered—for example, the roof, slabs, or the opaque walls. The second is the percentage of the envelope fenestration element allowed to be incorporated into the building envelope—including glazing systems such as windows, storefronts, curtain walls, and skylights. The specific requirements of the prescriptive approach for areas of the envelope systems, such as below-grade insulation, roof insulation, or soffit insulation, are easy to understand and implement and of little consequence to the aesthetic and functional design of the building. However, the requirements for other elements— such as glazing systems and skylights—can have a very large impact upon the architectural considerations of the building. This area is one that needs to be considered for many buildings and is the primary cause of building designers needing to consider the use of one of the other two options for demonstrating compliance for the building envelope minimum thermal performance requirements of ASHRAE 90.1. This paper and ASHRAE 90.1 specifically address the energy performance of only commercial and large residential buildings.
Background Under the current version of the U.S. Code, Title 42, Chapter 81, the U.S. Department of Energy (DOE) establishes energy policy and guidance for the United States, but this document does not mandate specific requirements for buildings. Specifically, the statute—Title 42 U.S.C. § 6834 The Energy Conservation and Production Act (ECPA), as amended—requires the DOE to review the most recently issued version of the commercial energy efficiency building codes and determine whether that version delivers improved energy efficiency when compared to the previously issued code. The codes that DOE must review are the IECC for residential and commercial buildings and ASHRAE Standard 90.1 for commercial and large residential buildings. The latter includes specific criteria for thermal performance of exterior building envelopes. The DOE supports and participates in the model building energy code development processes administered by ASHRAE and the ICC. DOE activities include developing and submitting code change proposals, conducting the analysis of building energy efficiency and cost savings, and formulating underlying evaluation methodologies.4
21
22
STP 1617 On Building Science and the Physics of Building Enclosure Performance
The DOE’s State Energy Program (SEP) provides funding and technical assistance to states, territories, and the District of Columbia to enhance energy security, advance state-led energy initiatives, and maximize the benefits of decreasing energy waste. SEP emphasizes the role of state governments as the decision makers and administrators for program activities that are tailored to the resources, delivery capacity, and energy goals of each state and by each state. The DOE Building Energy Codes program tracks the adoption of energy codes for residential and commercial buildings.5 State adoption is tracked based on the national model energy codes, primarily the International Building Code (IBC) and the IECC. The IECC establishes minimum standards for residential and commercial buildings, while ASHRAE 90.1 solely addresses commercial and large/high-rise residential buildings. The DOE also analyzes state energy codes to assess the savings associated with IECC updates. The resulting findings are intended to aid model code adoption by the states and as a means of supporting states that are working to update their energy code requirements.
ASHRAE 90.1 Requirements The energy requirements of ASHRAE 90.1 are established as law based on the following federal statutory requirements: • Statutory Authority: Energy Conservation and Production Act (ECPA) (Pub. L. No. 94-385), as amended6 • “Section 304(b) of ECPA [Energy Conservation and Production Act], as amended, provides that whenever the ANSI/ASHRAE/IESNA Standard 90.1-1989, or any successor to that code, is revised, the Secretary must make a determination, not later than 12 months after such revision, whether the revised code would improve energy efficiency in commercial buildings and must publish notice of such determination in the Federal Register. (42 U.S.C. 6833(b)(2)(A)) The Department, consistent with the International Code Council (ICC) and the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) considers high-rise (greater than three stories) multifamily residential buildings and hotel, motel, and other transient residential building types of any height as commercial buildings for energy code purposes. Low-rise residential buildings include one- and two-family detached and attached buildings, duplexes, townhouses, row houses, and low-rise multifamily buildings (not greater than three stories) such as condominiums and garden apartments.”6 • “If the Secretary determines that revision to the commercial code would improve energy efficiency then, not later than two years after the date of the publication of the affirmative determination, each State is required to certify that it has reviewed and updated its commercial building code regarding energy efficiency in accordance with the revised standard for which the Secretary has made a positive determination. (42 U.S.C. 6833(b)(2)(B)(i)) Such
JENSEN AND JOHNSON, DOI: 10.1520/STP161720180114
certification must include a demonstration that the provisions of such State’s commercial building code regarding energy efficiency meet or exceed such revised standard.”6 TITLE 42—THE PUBLIC HEALTH AND WELFARE OF THE US CODE
ASHRAE 90.1 was developed as the primary tool for establishing the required performance levels needed to improve energy efficiency in commercial and large residential buildings and to provide a process for continuous improvement of performance levels. First and foremost, it is important to understand that in the words of ASHRAE: Standard 90.1 is a fluid document. As technology evolves, the project committee is continually considering new changes and proposing addenda for public review. When addenda are approved, notices are published on the ASHRAE and IES websites. Users are encouraged to sign up for free ASHRAE and IES Internet listserv for this standard to receive notice of all public reviews and approved and published addenda and errata.1 And, in accordance with Section 1.1, the purpose of the standard is “to establish the minimum energy efficiency requirements of buildings other than low-rise residential buildings for design, construction, and a plan for operation and maintenance, and utilization of on-site, renewable energy resources.”1
Envelope Compliance Options ASHRAE 90.1 addresses building envelope requirements under Chapter 5, Building Envelope and Compliance Paths. There are three methods to achieve building envelope system compliance for minimum performance levels: 1. Prescriptive building envelope option (Section 5.5) 2. Building envelope trade-off (Section 5.6), sometimes referred to as COMcheck 3. Energy cost budget (Section 11) The basics of each option for ASHRAE 90.1 compliance are: Option 1 Under the prescriptive approach to compliance, specific values are identified by the standard as minimally acceptable levels of performance for various aspects of the building envelope, such as roof, fenestration, or exposed soffits, for example. This approach also includes maximum allowable areas of certain envelope components, such as skylights or windows. Based upon this approach, the designer’s task is simpler than it is with the other two options—building envelope trade-off or energy cost budget. While this approach is a very straightforward and relatively simple method of confirming compliance with the standard requirements, it also offers less flexibility in the design of the building envelope to meet other aesthetic, functional, and performance desires than the other two approaches. This is particularly true regarding the allowable area of glass walls and skylights within the building.
23
24
STP 1617 On Building Science and the Physics of Building Enclosure Performance
This option is the most straightforward and simplest approach to confirming building envelope compliance with ASHRAE 90.1 requirements. When a prescriptive analysis indicates a certain design and or systems will not meet ASHRAE 90.1 for energy usage on this simple basis, one of the other two options for analysis to demonstrate compliance may be checked, or the design must be revised. Option 2 The building envelope trade-off approach shifts from prescribed performance levels of pieces and parts of the building envelope to an acceptable level of performance for the entire building envelope. You can supplement the performance of one component, such as the roof, to make up for an area that may not be as high in performance as required for prescriptive compliance or that has more area than prescriptively allowed, such as windows or skylights. This, in effect, allows one to numerically balance performance levels and areas against thermally weaker performing materials. This approach also allows further latitude to the designer of the building in selecting materials, products, and systems to be used, as well as amounts and locations of specific system elements. One example is the addition of solar shading elements to reduce solar heat gain. This option is more complex than Option 1 and requires a set of balancing equations to demonstrate compliance. It does, however, have the benefit of allowing a wider range of choices in designing the building envelope and selecting the components. A common program or tool used for Option 2 is COMcheck, which is available for download from the DOE website.* Option 3 The energy cost budget approach to demonstrating compliance of the building exterior envelope with code requirements is the third option allowed by ASHRAE 90.1. Under this option, the energy usage of the building, including mechanical and electrical systems as well as the exterior envelope, is used to ensure compliance with the code based on building energy usage for the entire building and its systems. In this option, a less-than-compliant exterior enclosure may be found to be acceptable if there are enough savings in energy usage that can be demonstrated for other building systems, such as lighting, heating, or cooling. This approach allows the most latitude to the designer of building envelope systems in terms of materials, products, and systems, and the configurations in which they are utilized. It also places a greater level of importance on higher performance of the other building systems, such as mechanical and electrical. It is often used by offices in the design of larger and more complex buildings. Due to the degree of effort and expertise required to perform this type of compliance analysis to a high degree of accuracy, the use of this approach may be limited to projects of higher dollar value or size due to economic consideration regarding design fees.
*
https://www.energycodes.gov/comcheck
JENSEN AND JOHNSON, DOI: 10.1520/STP161720180114
There are several programs (or tools) available to facilitate these analyses. However, in the opinion of the authors, these tools are somewhat schematic in their level of analysis. While they may confirm compliance, they may not lead to a high degree of accuracy in the answers to the actual performance of the building. It can be noted that no matter how precise the energy modeling tools, the true analysis occurs postoccupancy when validated with actual energy consumption.
Code Adoption ASHRAE 90.1 has become the standard referenced by a large part of the design and construction profession for identifying the minimum acceptable values of energy use and consumption in buildings through adoption by state agencies. This includes consideration of energy usage of the entire building and its support systems such as mechanical and electrical systems. At this time, the tracking of state energy codes by the DOE indicates 43 states plus the District of Columbia have adopted some version of ASHRAE 90.1. This effectively establishes ASHRAE 90.1 requirements as the predominant code criteria for energy performance of building envelopes. The DOE also indicates seven states have not yet adopted an energy code as part of their building code or have generated their own energy code. However, each state is required by the DOE to develop its own version of an energy code or adopt by reference a published standard (fig. 1). The IBC under Chapter 13, Energy Efficiency, references the IECC. IECC references ASHRAE 90.1 as a minimum design requirement along with two options for IECC compliance under Chapter 4, Commercial Energy Efficiency, as follows: C401.1 Scope. The provisions in this chapter are applicable to commercial buildings and their building sites. C401.2 Application. Commercial buildings shall comply with one of the following: 1. The requirements of ANSI/ASHRAE/IESNA 90.1. 2. The requirements of Sections C402 through C405 and C408. In addition, commercial buildings shall comply with Section C406 and tenant spaces shall comply with Section 406.1.1. 3. The requirements of Sections C402.5, C403.2, C403.3 through C403.5.5. C403.7, C403.8.1 through C403.8.4, C403.10.1 through C403.10.3, C403.11. C403.12, C404, C405, C407, and C408. The building energy cost shall be equal to or less than 85% of the standard reference design building.2 Most architects and engineers will choose the ASHRAE 90.1 path of compliance due to familiarity with the standard. Note that the code is always the basis for minimum design standards, and it is up to owners, architects, and engineers to at least comply with these minimum standards and hopefully establish goals to go beyond them.
25
26
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 1 State energy code adoption of ASHRAE 90.1, United States map.7
The ICC family of codes, including IBC, are evaluated, revised, and published with revisions on a three-year cycle. The most current IBC (2018) references the 2016 edition of ASHRAE 90.1 through the 2018 Energy Conservation Code, Chapter 4, Commercial Energy Efficiency, C401.2. This 2018 version has greatly increased the required minimum envelope performance values compared to previous editions. Again, the states will establish their respective requirements and adopt building and energy codes at their own rate and modify the base code language and requirements as they determine appropriate to fit their own programs and goals. However, the states must adopt the revised versions of the standard and codes for them to become effective as law. In the opinion of the authors, in most cases, legislators do not often take expeditious action to adopt code changes. This may be a function of concern for financial burdens to owners associated with the process as well as implications due to construction costs associated with increased requirements. As an example, at the time of this writing (October 2018), the State of Michigan, Building Energy Codes Program8 had just adopted the 2013 version of ASHRAE 90.1 at the end of 2017 when they adopted the 2015 IBC, a four-year time lag for the energy code. Why such a slow adoption rate? Historically, building codes
JENSEN AND JOHNSON, DOI: 10.1520/STP161720180114
and standards change very slowly unless an overwhelming reason is found to make change occur quickly. In the case of energy conservation, in the past, the urgency came and went with the cost and availability of energy, which—at least recently— has not been the cause for concern in most areas of the United States. Another contributing factor is that governments tend to react slowly in most cases, again without an overwhelming need for change—such as increased cost in construction resulting from the increased thermal performance of envelope systems. Legislators are not anxious to burden projects with additional regulations and likely costs. Here is another example: For new projects in the state of Massachusetts, ASHRAE 90.1-2013 has been adopted with a minimum design requirement for new large-area and high-rise buildings set at 10% better than ASHRAE 90.1. This requirement is known as the “Stretch Energy Code”9 and is an example of some jurisdictions establishing higher energy code standards than those adopted by others. On a recent large government project, ASHRAE 90.1-2007 was determined to be the applicable energy code standard under Leadership in Energy and Environmental Design (LEED) 2009. As is common with many large government projects, funding lags and budgets suffer from escalation and changing design criteria. The project was reevaluated under LEED V4 and ASHRAE 90.1-2010 and it was found that a budget increase was required to meet new energy compliance standards. Ultimately with construction and fit-out, the first occupancy date will be behind energy standard developments over the ten-year design and construction schedule and user fit-out! This situation is not uncommon for large government projects and suggests a good reason to design above the energy code minimums applicable at the time of design. While many authorities having jurisdiction will accept the applicable energy code in force at the time of design, this is not always the case, which can lead to serious problems when it is time for permitting and construction. As a rule of thumb, using LEED standards for energy goals at 30 to 50% above base ASHRAE 90.1 criteria can provide a key benefit in terms of energy performance for long, drawn-out project schedules.
ASHRAE 90.1 Progression The first ASHRAE 90.1 energy standard was a reaction to the 1973 oil embargo and the United States no longer having ready access to cheap and plentiful oil. Since the first version published in 1975, the standard was followed by revised editions in 1980 and 1989. Since then, the standard has been revised many times, and solving energy challenges remains a national priority. ASHRAE 90.1 is considered a fluid document. As technology evolves, the project committee is continually considering new changes and proposing addenda for public review. In 1999, the ASHRAE 90.1 standard was placed on a continuous maintenance cycle that allowed the standard to be updated several times each year through the publication of approved addenda. With the 2001 edition, a schedule was established
27
28
STP 1617 On Building Science and the Physics of Building Enclosure Performance
to update and publish the standard on a three-year cycle with a fall season issuance. The three-year cycle follows the similar cycle of the ICC family of codes though not concurrently. For example, the 2016 edition of ASHRAE 90.1 is incorporated into the 2018 ICC code publications. Also, each new edition includes addenda and errata issued under the previous edition. ASHRAE divides the United States into seven distinct climate zones for the basis of energy design compliance. Not shown are the more recently added Climate Zones 0 and 8. Zone 0 is used in South America and Zone 8 in northern Canada (fig. 2). The 2016 edition of ASHRAE 90.1 is an aggressive update of the 2013 edition that includes numerous energy-saving measures resulting from continuous maintenance proposals submitted by the public and committee. More than 125 addenda to the 2013 edition were processed and approved by committee and ASHRAE and IES boards of directors. Appendix 8 lists and describes the addenda. The most significant 2016 changes for the building envelope are as follows: 1. Mandatory provisions include the addition of envelope verification in support of reduced air infiltration and increased requirements for air leakage of overhead coiling doors. 2. Prescriptive requirements include increased stringency requirements for metal building roofs and walls, fenestration, and opaque doors. Requirements were added for Climate Zone 0.1 Tables 1 and 2 show the progression of changes in required minimum thermal values from 1999 through the 2016 standard to demonstrate how the prescriptive building envelope requirements have changed over this 17-year period. The tables
FIG. 2 ASHRAE 90.1 climate zone map—United States.1
TABLE 1 ASHRAE 90.1, opaque elements
Year Opaque Elements
1975 1st Edition
1999
2001
2004
2007
2010
2013
2016
2019 Draft
% CHANGE
Assembly U Value Maximum
0.075
0.063
0.063
0063
0.048
0.048
0.032
0.032
0.032
49%
Insulation R Value Minimum
NR
15
15
15
20
20
30
30
30
100%
Assembly U Value Maximum
0.32
0.123
0.123
0.123
0.09
0.09
0.09
0.09
0.09
27%
Insulation R Value Minimum
NR
7.6
7.6
7.6
11.4
11.4
11.4
11.4
11.4
50%
Roof Insulation Entirely above the Deck
Walls, above Grade Mass
Walls, above Grade Steel Framed 0.32
0.084
0.084
0.084
0.064
0.064
0.055
0.055
0.055
35%
Insulation R Value Minimum
NR
13 þ 3.8 c.i.
13 þ 3.8 c.i.
13 þ 3.8 c.i.
13 þ 7.5 c.i.
13 þ 7.5 c.i.
13 þ 10 c.i.
13 þ 10 c.i.
13 þ 10 c.i.
37%
Assembly U Value Maximum
0.08
F0.73
F0.73
F0.73
F0.73
F0.73
F0.520
F0.520
F0.520
29%
Insulation R Value Minimum
NR
NR
NR
NR
NR
NR
R 15 for 24
R 15 for 24
R 15 for 24
NA
inches
inches
inches
Slab on Grade Floors Unheated
Note: c.i. = continuous insulation.
JENSEN AND JOHNSON, DOI: 10.1520/STP161720180114
Assembly U Value Maximum
29
30
Year Fenestration
1975 1st Edition
1999
2001
2004
2007
2010
2013
2016
2019 Draft
% Change
Assembly U Value Maximum
NR
NA
NA
NA
0.35
0.35
Assembly SHGC Maximum
NR
NA
NA
NA
0.4
0.4
0.32
0.31
NA
NA
0.4
0.38
NA
NA
Assembly U Value Maximum
NR
NA
NA
NA
0.45
Assembly SHGC Maximum
NR
NA
NA
NA
0.4
0.45
0.4
0.38
0.36
37%
0.4
0.4
0.38
0.36
8%
Vertical Fenestration 0% to 40.0% of Wall Nonmetal Framing, All
Metal Framing, Fixed
Fixed Glazing 30.1 to 40.0% of Wall Assembly U Value Maximum
NR
0.57
0.57
0.57
NA
NA
NA
NA
NA
NA
Assembly SHGC Maximum
NR
0.39
0.39
0.39
NA
NA
NA
NA
NA
NA
Note: For fixed glazing, values below 40% glazing remained the same and increased at 40.1% during this period. The 1975 and 1999 tables are based on HHD65 (6,167) CCD50 (3,046) for Detroit versus Climate Zone 5 used in subsequent publication years. F = F value/factor used by ASHRAE 90.1 for lineal foot of building perimeter. Assembly F factors are included in Appendix A of the standard. The 2001 version started the transition from HDD charts to Climate Zone charts in the addendum issue; ASHRAE 90.1-1975 did not have minimum glazing requirements; ASHRAE 90.1-1975, Walls, did not have a U-value difference based on construction type, mass, or other; ASHRAE 90.1-2019, Draft, proposes no differentiation among frame material types for U-value performance; 2019 roof, wall, and floor U-values are shown as no change over 2016; 2019 reduces fenestration U-values. Percent change values are a comparison between 1999 and proposed 2019 values. SHGC = solar heat gain coefficient.
STP 1617 On Building Science and the Physics of Building Enclosure Performance
TABLE 2 ASHRAE 90.1, fenestration
JENSEN AND JOHNSON, DOI: 10.1520/STP161720180114
also includes currently proposed U-values for the 2019 edition. For simplicity, Tables 1 and 2 indicate only changes for Climate Zone 5, covering the upper midwest region and western (noncoastal) states. As Tables 1 and 2 illustrate, there has been a stepped progression in increasing energy efficiency for building envelope components since 1999, and certain editions have taken more aggressive jumps for specific components. Roofs took a major jump in 2007 and 2013. Wall performance values have remained constant since 2007 and slab-on-grade values increased in 2013. There has been an aggressive reduction in fenestration (glazing) U-values and a decrease on the limits of allowable percentages over the last 20-year period that is predicted to continue; this increase in thermal performance will start to shape what envelope designs can be and will be. These fenestration targets make sense as infiltration/exfiltration and glazing contribute to the most heat loss and gain per square foot of the envelope component of the building. The table notes (not shown) expand on the historical table data presented. This trend of increasing thermal performance levels for the building envelope has had a significant impact on design and the materials, systems, and products available for use in exterior envelopes. This is particularly true regarding buildings of size or simplicity that would not warrant thermal modeling of the entire building or envelope system to validate compliance with mandated energy performance levels.
ASHRAE 90.1 Compliance Options As a widely recognized, and most often code-mandated standard, ASHRAE 90.1 in effect significantly impacts the selection of materials, products, and systems available for compliance with the required energy standard. As increases continue through the most recent edition of ASHRAE 90.1 and as U-values become more stringent for materials, products, and systems in specific envelope applications, the options for designers of buildings become more challenging. As requirements for energy performance levels increase, this is also likely to increase the number of buildings and envelope systems evaluated by thermal modeling to confirm compliance with the new standard because the prescriptive requirements have become more restrictive and challenging to meet. Envelope values present major challenges for designers under the ASHRAE 90.1-2016 prescriptive requirements as these performance values for systems continue to increase and options for areas of glazed systems become more stringent. Full curtain wall building design will be even more challenged to meet the opaque wall and fenestration values with currently available products, and costs are likely to increase dramatically for these systems. At present, one of our most-used common curtain wall framing systems no longer meets the prescriptive requirements under ASHRAE 90.1-2016 in certain cold climate zones. When a prescriptive analysis (Option 1 for demonstrating compliance) indicates a certain design or systems will not meet ASHRAE 90.1 for energy usage on
31
32
STP 1617 On Building Science and the Physics of Building Enclosure Performance
this simple basis (or both), one of the other two optional analysis methods to demonstrate compliance may be checked, or the design must be revised. The first alternative for consideration where the prescriptive requirements cannot be met is Option 2, building envelope trade-off. Building envelope trade-off permits the designer to offset envelope values by providing higher U-values in areas while using lower (below the prescriptive requirements) values in other areas of the building envelope. The common program modeling tool for this option is COMcheck. The COMcheck software compliance tool is available from the DOE’s Building Energy Codes Program. If a trade-off design option cannot be achieved, Option 3 will need to be considered. Programs like COMcheck will recognize some elemental trade-off features such as overhangs and shading devices that the minimum prescriptive approach does not. The energy cost budget, Option 3, requires an analysis of the whole building envelope and building’s heating, ventilation, and air-conditioning (HVAC) and electrical systems. This is an extensive and time-consuming design option that many, if not most, projects and owners cannot afford and many smaller design firms will lack the resources and experienced staff to properly perform. The inability for projects and design firms to undertake this level of effort may limit the number of designers pursuing and using this option, thereby losing an opportunity to greatly increase aesthetic and functional envelope designs, while still achieving better energy use. As with prescriptive values, envelope trade-off and energy cost budget design options are used and pursued to achieve energy code compliance, and all can be considered as limiting design options and opportunities. The post-1970s’ oil embargo period brought on a design aesthetic of massive buildings with limited glazed area percentages. Design moved away from this as technology caught up and offered more options and the “glass-box” reappeared. A common aesthetic in recent years has been the use of overhangs and shading devices on the south and west facades to limit solar heat gain to the building interior via large glazed areas. Use of low-iron, ultraclear glass became very popular and will now find limited use under current and projected ASHRAE 90.1 fenestration U-value requirements. If the building design is not found compliant under the prescriptive requirements or any of the options, the design will need to be revised until a design path meeting the code requirements is achieved. This effort of design and redesign until an energy codecompliant design is achieved can be very time consuming for the design professional, and standard or base design fees often do not support this effort.
Application From the beginning of a design, and as a baseline approach, the prescriptive building envelope values are the starting point for many architects. As the prescriptive values for thermal performance continue to increase, the designer is continually challenged to meet the minimum values. For years, we have added insulation thickness to roofs and walls as well as decreasing fenestration percentages while
JENSEN AND JOHNSON, DOI: 10.1520/STP161720180114
increasing thermal performance values. The performance increases under the 2013 and 2016 editions of ASHRAE 90.1 have sent rumblings through the design community and concern over where the future editions might go. An estimated 10% increase in performance values in each new edition of ASHRAE 90.1 is not unheard of. At the time of this writing, the ASHRAE 90.1-2019 edition is not targeting increases in roof and wall thermal performance but propose further limitations for fenestration systems. As we look to the increased performance values, we are at a point with the 2016 edition of ASHRAE 90.1 where wall and roof insulation increases will provide less of an impact in overall energy savings. On the other hand, the fenestration (glazing) values have reached a point where manufacturer standard products will no longer hit the minimum values, and we now need to look to what are considered thermally advanced products at additional cost. For a typical glazed window or curtain wall system, what was normally used recently will not meet the minimum 2016 required performance values. Similar to the shift from single- to double-glazed insulated glass units that occurred 40 to 50 years ago, we can now anticipate increased use of triple-glazed glass units to meet the minimum requirements. Triple glazing has become increasingly common in Climate Zones 6 and 7 to meet the minimum prescriptive U-values for fenestration. FENESTRATION AND IMPROVING THERMAL PERFORMANCE
Architects are still in love with glass. Transparency and low-iron glass, both skylights and windows or curtain walls are all key to bringing daylight indoors. Daylighting provides very real health benefits for the building occupants; however, it increases both heat loss and gain as compared to even minimal opaque wall or roof construction. There has already been a reduction to 40% in allowable fenestration in building walls under the ASHRAE 90.1 prescriptive requirements and it is hard to predict when ASHRAE will implement further reductions. ASHRAE 90.1 2016 fenestration U-values will be difficult to achieve with high-visibility light transmittance glass. Even with the best Low-E coatings available today, standard aluminum spacers used in double insulated glass (IG) units with air are no longer a design option to meet the prescriptive code requirements. We now need to look at the use of stainless steel or thermally broken spacers with argon to approach or exceed (or both) required U-values. Gases used for insulated glazing units also need to be considered where air values might not be enough to meet the prescriptive code requirements. Better Low-E coatings are now key to improving glass U-value performance, and double Low-E coatings are now becoming more common in cold climates. The more reliance we place on new technologies to improve the performance of glass, the less assurance we have of acceptable performance and useful service life, two of the key components of sustainability. Insulated glass units with argon are known to lose about 1% a year, with each loss lowering the thermal performance of the unit. The technology and performance of coatings, spacers, and third layers, either films or glass lites, to improve the thermal characteristics of glazing unit
33
34
STP 1617 On Building Science and the Physics of Building Enclosure Performance
performance for glass have continued to advance. However, framing systems appear to be lagging in development to achieve assembly thermal performance sufficient to keep pace with the requirements of ASHRAE 90.1. Published glass U-values are listed at the optimal level of the center of the glass. Required fenestration U-values are for the total system assembly. Very often, the designer incorrectly assumes that the glass performance value is sufficient. However, in almost all cases, the assembly performance will not meet the code criteria when only considering glass. The framing systems perform at a lower level than the glass and can dramatically reduce the overall system performance. Further, the published ratings of a system are for a specific size and window frame configuration. Adding or reducing mullions and changing the size of the window can dramatically impact the system performance in either direction, up or down. When looking at total building design and considering the number of mullion framing members and size of the glass panels used in that design, the resultant system design thermal performance U-value can be drastically reduced or improved. To demonstrate an example: Framing system manufacturers’ lab test system assemblies with a standard test unit size that includes framing members and glazing. The standard assembly size is established by the testing standard (ANSI/ NRFC-100),10 which is 78.75 in. by 78.75 in. (2,000 mm by 2,000 mm) and a single perimeter frame with a center mullion (fig. 3). Note that this is a very large unit
FIG. 3 ANSI/NFRC-100 standard test unit diagram.10
JENSEN AND JOHNSON, DOI: 10.1520/STP161720180114
assembly and often does not reflect actual building usage where smaller IG sizes and more framing members would be used, thus U-values would increase. Table 3 provides a baseline comparison of commonly available glazing values and resultant system assembly U-values using framing systems available in the current marketplace. The table lists three curtain wall framing options by a specific manufacturer. This and very similar curtain wall systems are available from multiple manufacturers. The table notes expand on the table data presented. Framing System Descriptions System 1: Curtain Wall System 1, “standard” pressure bar system with aluminum pressure bar System 2: Curtain Wall System 2, “high performance” pressure bar system with aluminum pressure bar System 3: Curtain Wall System 3, “high performance” pressure bar system with fiberglass pressure bar Glass Low-E Coating Types Used for Simulations Low-E Coating 1: Common Low-E coating available for the past 15-plus years Low-E Coating 2: Newer higher-performance Low-E coating available for the past 5 years Low-E Coating 3: Latest higher-performance Low-E coating technology Figure 4 provides a thermal modeling illustration of how varied system types perform on temperature flow from exterior to interior. The figure illustrates three types of glass and framing systems available from an aluminum curtain wall framing manufacturer. Thermal simulations showing temperature variations from the exterior/cold side to the interior/warm side. From left to right, this illustrates Systems 1, 2, and 3. Improving fenestration U-values needs to go beyond looking at the glazing values. What is demonstrated (Table 3) is that starting with glass selection based on good U-values can easily be set back with poor IG spacer selection and framing choices. The use of industry standard aluminum spacers does not offer thermal performance enough to meet the minimum performance values of the new requirements. Stainless-steel spacers are the next best choice, and thermally broken spacers offer even more thermal improvement. Thermally improved fenestration framing systems are not necessarily new to the marketplace and are readily available from major manufacturers—though at a premium in cost. Standard systems that have proven performance over a period of years provide a minimal reduction in thermal transfer and often result in significantly reducing the thermal performance of good glass. Meeting the most current fenestration assembly U-values in cold climate regions will require good glass, spacer, and framing selections just to hit the minimum prescriptive values, all at increased project cost. Some of these individual component changes will appear to offer minimal U-value improvement. However, when missing the minimal code required prescriptive values by 0.01, they will be beneficial to qualify the system. Table 4 and Table 5 illustrate, along with figure 5, how varying system components and design affect the thermal performance values of a glazing system.
35
36
No.
Unit Description
VLT
U-Value
Visible Light Transmittance
Winter Nighttime
SHGC
Assembly U-Values
Assembly U-Values
Assembly U-Values
System 1
System 2
System 3
NA
1
0.75-in. clear monolithic glass
89
1.02
0.82
NA
NA
2
1-in. clear insulated glass
79
0.47
0.70
0.605
0.54
NA
3
1-in. clear insulated low-iron glass
84
0.47
0.82
0.605
0.54
NA
4
1-in. clear insulated glass with
70
0.29
0.39
0.46
0.39
0.38
64
0.28
0.27
0.45
0.38
0.37
51
0.29
0.23
0.46
0.39
0.38
64
0.22
0.37
NA
0.33
0.32
57
0.21
0.25
NA
0.325
0.315
47
0.21
0.21
NA
0.325
0.315
Low-E Coating 1 5
1-in. clear insulated glass with high-performance Low-E Coating 2
6
1-in. clear insulated glass with ultrahigh performance, Low-E Coating 3
7
1.75-in. triple-glazed clear insulated glass with Low-E Coating 1
8
1.75-in. triple-glazed clear insulated glass with Low-E Coating 2
9
1.75-in. triple-glazed clear insulated glass with Low-E Coating 3
Note: ASHRAE 90.1-2016 minimum values for fixed metal frames; Climate Zone 5 assembly U-value 0.38, SHGC 0.38. Yellow highlight indicates glazing and system meeting ASHRAE 90.1-2016 minimum requirements for Climate Zone 5. ASHRAE 90.1-2019 minimum U-value for fixed metal frames; Climate Zone 5 assembly U-value 0.36. Insulated glazing values based on 0.5-in. air fill with aluminum spacer. U-values are based on 0.25-in. glass thickness. Glass U-values are to the center of the glass. Values are not assembly values including framing systems. Assembly U-values are based on aluminum curtain wall systems. System 1 values are based on a standard curtain wall system with an aluminum pressure plate. System 2 values are based on a high-performance curtain wall system with an aluminum pressure plate. System 3 values are based on a high-performance curtain wall system with a fiberglass pressure plate.
STP 1617 On Building Science and the Physics of Building Enclosure Performance
TABLE 3 Glazing system reference values11,12
JENSEN AND JOHNSON, DOI: 10.1520/STP161720180114
FIG. 4 Curtain wall thermal modeling.12
FIG. 5 ANSI/NFRC-100 design options: Base System, ANSI/NFRC-100 standard unit (A); Option 1, mullion added, (B); Option 2, mullion removed (C).10
Table 4 illustrates U-values for gas type selections of air, argon, and krypton and further shows how dual Low-E coatings and triple glazing can continue to add to the performance. Table 4 represents the IG unit, only absent of any framing system. Moving from air to argon to krypton gas fill continues to reduce the U-value, which is beneficial to meeting the new standard minimal performance requirements. The U-values show in Table 4 are based on using aluminum spacers and highlight a significant improvement in changing gas fill versus changing spacer types as illustrated in Table 5 and figure 5.
37
38
STP 1617 On Building Science and the Physics of Building Enclosure Performance
TABLE 4 Glass U-values based on air-gas fill using glass with Low-E Coating 2
Gas
Winter U-Value
SHGC
Notes
Air
0.28 BTU/hr/ft
2
0.27
Argon
0.24 BTU/hr/ft2
0.27
Krypton
0.22 BTU/hr/ft2
0.27
Argon
0.23 BTU/hr/ft2
0.27
Adding second Low-E on Surface #4
Argon
0.12 BTU/hr/ft2
0.25
Triple-glazed and dual Low-E coatings
Published U-values for assemblies generally are based on the ANSI/NFRC 100, Procedure for Determining Fenestration Product U-Factors.10 The system’s thermal performance will be less with more framing members being added and greater with the center mullion removed. Historically, architectural preference has been toward more use of mullions to break up large expanses of glass where current trends are leading more toward larger and even “oversized” glazing units. The insulated glazing units used in the tests are based on the use of an aluminum spacer and air as the gas fill. U-values for insulated glazing units can be improved as previously mentioned with the use of stainless steel and thermal spacers as well as argon and krypton gas use. Running optional frame simulations using the base system ANSI/NFRC-100 (fig. 5A), Options 1 and 2 (fig. 5B and 5C), the values shown illustrate how U-values are affected by variations in the amount of aluminum framing and substantiate that the reduction of aluminum framing members improves U-value performance and adding aluminum framing members increases the system U-value, reducing performance (fig. 5A, 5B, 5C and Table 5). Further, Table 5 shows U-value performance based on insulated glazing unit spacer material types in addition to adding and removing aluminum mullion framing members. As mentioned, the base system (ANSI/NFRC-100) is air-filled units with aluminum spacers and a set mullion configuration. Slight improvements are found with the use of stainless steel, where the best performance can be found in the use of warm-edge type spacers.
TABLE 5 Framing option performance U-values
Spacer Type System Assembly
Aluminum
Stainless Steel
Warm Edge
Base System (fig. 5A)
0.450
0.443
0.428
Option 1 (fig. 5B)
0.504
0.494
0.474
Option 2 (fig. 5C)
0.394
0.389
0.379
Note: U-values based on standard aluminum curtain wall system with 1-in. IG and Low-E Coating 2. The base system is the ANSI/NFRC-100 standard test model, figure 5A. Option 1 adds aluminum mullions to the ANSI/NFRC-100 standard test model (fig. 5B). Option 2 removes aluminum mullions from the ANSI/NFRC-100 standard test model (fig. 5C).
JENSEN AND JOHNSON, DOI: 10.1520/STP161720180114
Table 5 shows minimal U-value reductions in changing spacer options of 1% to
3% and an approximate 12% increase or decrease in U-values as the amount of aluminum varies. The best thermal improvement is with Option 2 and warm-edge spacers, providing a 16% reduction in U-value over the base aluminum system. Table 6 is similar to those provided by curtain wall manufacturers and are used to determine overall system performance values in tandem with glass performance values. There is a two-way approach to using the charts: select the desired system value and see the required glass value or select the glass value and find the resultant system value. Either way, the charts illustrate how the curtain wall system selection impacts the glass value selection. Looking at U-value performance, you see that the better the glass value, the worse the resultant system is. SHGC shows system improvement because mullions block solar heat gain better than glass. Visible light transmittance is also reduced as mullions block the amount of daylight coming in. Achieving a “goal” U-value, or even a minimum code requirement, is a balancing act of cost versus design and performance. For example, air is free, and argon
TABLE 6 Curtain wall system performance charts12
Thermal Transmittance
SHGC Matrix
Visible Transmittance
(BTU/hr ft2 F) Glass U-Factor
a
System U-Factor
Glass SHGC
Overall SHCG
Glass VT
0.47
0.54
0.75
0.68
0.75
0.67
0.46
0.53
0.70
0.64
0.70
0.63
0.44
0.52
0.65
0.59
0.65
0.58
0.42
0.50
0.60
0.55
0.60
0.54
0.40
0.48
0.55
0.50
0.55
0.49
0.38
0.47
0.50
0.46
0.50
0.45
0.36
0.45
0.45
0.41
0.45
0.40
0.34
0.43
0.40b
0.37b
0.40c
0.36c
0.32
0.42
0.35
0.32
0.35
0.31
0.30a
0.40a
0.30
0.28
0.30
0.27
0.28
0.38
0.25
0.24
0.25
0.22
0.26
0.37
0.20
0.19
0.20
0.18
0.24
0.35
0.15
0.15
0.15
0.13
0.22
0.33
0.10
0.10
0.10
0.09
0.20
0.32
0.05
0.06
0.05
0.04
0.18
0.30
0.16
0.28
0.14
0.26
0.12
0.25
0.10
0.23
Overall system thermal transmittance U-value reduction of 33%. SHGC system improvement of 7.5%. c Visible transmittance overall system reduction of 10%. b
Overall VT
39
40
STP 1617 On Building Science and the Physics of Building Enclosure Performance
costs a little more. However, krypton gas is expensive. When it comes to IG unit spacers, aluminum has long been the industry standard. However, it also is the lowest-performing thermally. Stainless steel spacers are readily available and not much of an overall cost increase, though they offer little in increased performance. Thermal or warm-edge spacers are the newest to the marketplace and carry a premium cost. For framing systems, standard aluminum curtain wall pressure bar systems are common and low cost. However, they perform poorly thermally, and they will no longer hit the U-value demands for cold climates. The newer “highly thermally improved” or “ultrathermal” curtain wall systems are now available for cold climates and meet minimum U-value requirements. Triple-glazed units perform the best—however, at a much higher cost for the system, including both the glass units and the framing. These systems also weigh significantly more than the more typical systems previously used and add dimension to the building exterior enclosure. Further, with triple-glazed IG units, the curtain wall framing system and supporting substructure must be designed to accept the additional glass unit depth and weight. It should be noted that most federal government projects require some level of blast-resistive design construction. Fenestration system selections can be limited when looking for highly thermally improved systems with blast design testing and approvals in place. As with thermal values, blast design requirements have also continued to advance with higher reaction load design values. Blast design fenestration system testing for a given project can be cost restrictive versus selecting available fenestration systems with testing and approvals in place.
New Technologies As the minimum level of thermal performance requirements for the building envelope continues to increase, new and improved technology for the systems composing the building enclosure must be advanced and made readily available. The building envelope consists of many components and ASHRAE 90.1 prescriptive requirements break them into two major categories: opaque elements and fenestration. The performance of the opaque elements is primarily driven by air barrier and insulation. Solid doors are classified under opaque elements; however, perimeter door/frame seals and vestibule provisions are also of primary importance. Fenestration is broken into vertical elements (windows and curtain walls) and horizontal elements (skylights). Glazed or metal framed doors fall under fenestration. Designers have reacted to increased performance demands on opaque elements by continually adding more insulation. Under ASHRAE 90.1-2016, in Climate Zone 7, roofs are required to have R-35 continuous insulation; this is approximately equivalent to 7 in. of polyisocyanurate foam insulation. Where this becomes a challenge is in linking the continuous insulation requirements from roof to wall. For the commercial building envelope, the major insulation types are as follows: polyisocyanurate, polyurethane, extruded polystyrene, and mineral wool/fiber. Fiberglass
JENSEN AND JOHNSON, DOI: 10.1520/STP161720180114
insulation is still common in the residential market and in the light wood- and light gauge steel-framed commercial market. As mentioned previously, we are at a point where increasing insulation thickness is adding little overall increase in value and performance. Advances in insulation technology have been slow to evolve, and manufacturers are challenged with meeting environmental production issues while maintaining product performance. Thinner, more R-value per unit thickness insulation is in demand today; however, these materials generally are not available for the building marketplace. In the schematic design phase on a recent project, the design team was challenged to evaluate total building energy performance in a good-better-best scenario where “good” was ASHRAE 90.1 minimum prescriptive requirements. Since the project was required to hit 30% increased performance over the ASHRAE 90.1 base requirements, the “better” value was increased by 15%, and the “best” value by 30%. All design disciplines were held to the same challenge. The project was further required to cost justify the life-cycle benefits of adding increased performance. When it came to the opaque building elements, the better and best options, where insulation value was increased, were not value/cost justified. For the fenestration system, selecting a good high-performing glass (U-value and SHGC) was a priority as well as a high-performing framing system. The design team found that due to the high amount of aluminum framing members in the design, the resultant system U-value performance was only hitting the minimum ASHRAE requirements. Triple-glazed IG units were not evaluated; they were determined to be cost prohibitive from the start, even though they would have offered the best thermal performance overall. Now, considering ASHRAE 2016 requirements, triple-glazing would certainly be an option for consideration for improving a Climate Zone 5 through 7 project and even more advanced technology glazing systems. While manufacturers have reacted to current ASHRAE 90.1 requirements for thermal performance increases of fenestration systems over the last 10 years, perhaps the question is whether they will continue to develop products and systems to meet the future performance requirements. Advances in both framing systems and glass technology have been introduced and are readily available. However, as previously highlighted, fenestration systems will continue to be looked at for increased performance; thus, even higher-performing systems will be required. The glass industry is focused on improving glass U-values and SHGC with advances in coating technologies, and they have also put a focus on the production of oversized glass units that increase thermal performance. This has been at least in part based upon recent design trends, which are at best transitory over time. Oversized insulated glass units will provide an improvement in total system U-value performance (less metal framing) versus improving the glass performance value. The framing system manufacturers need to develop even better thermal products to respond to what will be a continual reduction in the fenestration system U-values. Thermally broken metal-framed glazed doors have been available for a long time but are not yet commonplace in cold climate projects and should be used in Climate Zones 4 through 7
41
42
STP 1617 On Building Science and the Physics of Building Enclosure Performance
to meet metal-framed door prescriptive values. Fenestration systems are challenged by the number of manufacturers involved in one system. There are glass, spacers, insulated units, gaskets, and framing systems often all developed under separate industries. Partnerships in developing high-performing integrated fenestration systems is now a process that needs to be encouraged to meet the code-required performance demands and go beyond them. Improving thermal efficiency in fenestration systems is possible. Today, most walls are passive and not using available technology. However, walls can become active systems and adapt or change with thermal demands. The next wave of technology for fenestration systems, which can also include shading devices is the “smart” or “dynamic” wall that can adapt and change with the demand. These types of systems can offer both improved thermal resistance as well as optimum control over solar heat gain. The following list highlights some of the systems that are available and might be utilized to increase the thermal performance of the fenestration of the envelope. Not all of these are commonplace today due to cost and lack of proven durability/performance. However, these systems and approaches represent the kind of thinking and innovation that needs to be considered in the future. Note, not all of these will assist in prescriptive compliance; however, they can all be used with the compliance path option. • Photovoltaic glass to produce electricity • Electrochromatic or auto-tinting glass • Translucent insulating glass, utilizing nanogel or aerogel for high R-value • Increased use of double skin curtain wall systems • Smart shading devices • Vacuum-insulated glass units One of the more interesting of these systems, and just coming available in the U.S. production marketplace, is the vacuum-insulated glass (VIG) unit. A two-pane VIG unit with a single Low-E coating can approach sub 0.09 U-values for the glass, and unit depth is less than a standard 1-in. IG unit. The triple-pane VIG units with double LowE coatings are reported to approach 0.07 U-values.13 This is a significant U-value improvement over traditional IG units. Note, however, this is new production technology and current and near-term production capacity could be limited and expensive. The ASHRAE 90.1-2019 committee has addressed a key industry concern by setting requirements for thermal bridging similar to requirements for continuous insulation. The committee also proposed a reduction in the 40% limit in fenestration. However, it had not yet made it to a final proposal at the time of this writing. While reducing fenestration percentage will increase overall building thermal performance, it also reduces views and indoor daylighting quality. It is very likely that the design industry and the glazing industry will continue to push back on this type of reduction proposal. Note that reducing the minimal allowable fenestration percentage under the prescriptive requirements does not limit or restrict the designer from increasing the percentage. It will, however, require that performance be demonstrated by evaluation according to the envelope option paths.
JENSEN AND JOHNSON, DOI: 10.1520/STP161720180114
In the design and construction industry, demand, code, and law requirements tend to drive new products and technology advancements. Building material manufacturers need to concentrate on the advancement and development of new thermally advanced products for the building envelope. While the United States still has relatively low energy costs, we see the other world markets developing new and more thermally advanced enclosure systems quicker than the United States. Often Europe, with much higher energy costs, is at the forefront of building envelope system and product development. With the ultimate goal of higher and perhaps even net-zero energy performance for buildings, it is the authors’ opinion that the time for advanced technology to be vigorously pursued in the United States is now not in the future.
Other Energy Standards and Programs for Buildings There are other energy standards and programs that can be adopted by building owners for guidance in designing buildings with increased energy performance and reduced energy usage for improved energy conservation. These include: LEADERSHIP IN ENERGY AND ENVIRONMENTAL DESIGN
LEED,14 developed by the U.S. Green Building Council (USGBC), is often established as the standard for a “basis of design.” LEED references ASHRAE 90.1 as well and uses it to establish minimum standards and point goals for exceeding the minimum requirements in energy conservation above the base standard. The current LEED system (LEED V4.1) references ASHRAE 90.1-2010. However, LEED 2009 references ASHRAE 90.1-2007 and is still active, though it is scheduled for anticipated sunset in 2021 and should no longer be considered for new projects. The most common LEED rating system for new construction is “Building Design + Construction.” NET-ZERO ENERGY BUILDINGS (NZEB) AND NET-ZERO-PLUS BUILDINGS (NEPB)
NZEB AND NEPB have been developed to promote energy efficiency well beyond LEED and ASHRAE 90.1 requirements in hopes of generating buildings that use little or no energy in daily operations or that are net providers of energy in an environmentally responsible manner. At the turn of the millennium, many groups promoted 2020 building challenges with the goal of accomplishing buildings with significantly less energy consumption and even net-zero. This initiative was set back by the substantial economic downturn of 2008 to 2011 and has been superseded by the 2030 challenges and more stringent goals. In the private or public sector, it takes a committed client to strive to hit net-zero goals. The federal government is leading the effort in the push for net zero by establishing policy and issuing executive orders for implementation. The government set policy in place under President Obama on October 9, 2009, with Executive Order 1351415 requiring all new federal buildings that are entering the planning process in 2020 and later be designed to achieve zero net energy by 2030. This aligns with the
43
44
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Energy Independence and Security Act of 2007,16 which calls for a 30% reduction in fossil-fuel consumption (relative to 2003 levels by 2015) for new federal buildings and major renovations. However, the federal government continues to change energy policy, as Order 13514 was replaced by 13693 and then again by 13834, no longer targeting NZEB. Most net-zero energy buildings require extensive on-site energy generation to offset energy consumption. EUI
Energy use intensity (EUI) is a common measure for NZEB that can also be used to obtain a beneficial analysis and rating factor for buildings by setting low energy use goals based on building use type. EUI measures a building’s annual energy use per unit area.17 Typically, it is measured in thousands of BTU per square foot per year (kBTU/ft2/yr). This approach provides good comparative values for building use types (i.e., a hospital versus a warehouse can have drastically different EUI values or energy consumption). Climate zones will also have an impact on EUI values; lower EUI values indicate better energy performance. The use of EUI as an indicator provides a method to equalize the way that energy use is compared among various types of building use and to evaluate potential means for reducing overall energy consumption for all buildings. EUI can also be calculated on existing buildings as a baseline for energy improvements in renovation projects. Figure 6 illustrates some typical EUI values based on research by the Environmental Protection Agency’s (EPA’s) Energy Star Program conducted on more than 100,000 buildings.18
FIG. 6 EUI values based on building type.17
JENSEN AND JOHNSON, DOI: 10.1520/STP161720180114
FEDERAL GUIDING PRINCIPLES
For federal government buildings, there are also the “Federal Guiding Principles for Sustainable Federal Buildings”19 issued by the Council of Environmental Quality (CEQ) in 2016. The new edition is updated from the 2008 version but has roots back to a 1962 document, “Guiding Principles for Federal Architecture.” The guiding principles follow a similar pattern of goals more recently established by the USGBC LEED program. Consistent in all programs is the goal to optimize energy performance in buildings and reduce the consumption of fossil fuels. This new edition references ASHRAE 90.1 for energy performance requirements; however, it no longer lists a year for the publication. The document now states, “For new construction, ensure energy efficiency is 30% better than the current American Society of Heating, Refrigeration and Air Conditioning Engineers (ASHRAE) 90.1 standard.”1 Achieving 30% above the 2016 edition of ASHRAE 90.1 is a lofty goal and will be a challenge for most designers and buildings to achieve by any means while still attaining the current thinking and desires for appropriate aesthetics of buildings. This is consistent with the current federal program policy on building energy conservation in lieu of previously targeting NZEB. With the number of energy standards and program options currently available, it is a challenge for building designers to keep up with changes and be proficient at them all. Often design firms will seek out energy consultants to assist in the design process and analyze the application of system options. It is important to establish the program and set the energy goals and budget for the project early in the design process, especially if NZEB is a project requirement, or even a target or goal. Renewable energy sources are fast becoming a requirement when high-performance energy-efficient buildings and NZEB goals are established. Energy consultants can provide the valuable and necessary expertise to assist the design team in selecting the best performance-for-value systems.
Summary Since the 1973/1974 energy crisis, conservation and reduction in energy usage have been among the driving forces in the ever-changing design and construction industry. ASHRAE took on the challenge to look forward on this issue and establish needed standards with the creation of the ASHRAE 90.1 standard and has continued to update and push for improving buildings’ overall thermal envelope and energy use performance. As increases in the required thermal and energy performance of buildings continue through the most recent edition of ASHRAE 90.1, and as U-values become more stringent for materials, products, and systems in specific envelope applications, the process of designing to compliance with these requirements becomes more challenging for designers of buildings. The challenge is not one that cannot be met on the technical side. It is rather one of attaining creative buildings that meet the aesthetic, functional, and enriching needs and desires of owners and society while improving the thermal and energy performance of the overall building.
45
46
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Some elements of the building envelope, such as opaque walls and roofs, are at or very near the point where simply “adding more insulation” will not be effective in increasing the energy performance of the building envelope. Other components of the envelope, however, such as windows and skylights, are becoming very limiting to designers in terms of allowing a broad-based and extensive use of views and daylighting for most areas of a building, when following only the prescriptive approach or Option 2, building envelope trade-off, for compliance. It is very probable that fenestration will continue to be an envelope component targeted for increased thermal performance and further limiting of the area when evaluated by the prescriptive or the area trade-off methods of compliance. At the time of this writing, the 2019 edition of ASHRAE 90.1 was out in draft form with proposed changes available for public review, to be followed by a final vote by the ASHRAE committee. There were changes proposed for the building envelope criteria that did not make it to draft and other items that did. Initially, for walls and roofs, the committee proposed minor increases to be incorporated, and due to time constraints, they did not become a part of the 2019 edition. For fenestration, the ASHRAE committee originally promoted a reduction in allowable glazed area; however, this was not approved. A slight improvement in fenestration U-values did pass and is now a part of ASHRAE 90.1-2019. A new item to address the issue of thermal bridging is also emerging for ASHRAE consideration. While the concept is simple, the successful application of the concept in design and construction will most likely be a complicated item for most design and construction professionals to address and for code authorities to review and enforce. With an issue date of publication in 2019, this new edition of ASHRAE 90.1 is scheduled to become the basis for the ICC’s 2021 IECC. As the 2021 IECC comes into effect, states and other agencies will then review, adopt as written, or modify the new version of ASHRAE 90.1 as they deem appropriate. Whatever the timeline, it is almost a certainty that the new code will be adopted by most authorities over time, and that future versions of ASHRAE 90.1 will mandate even more stringent requirements for energy performance of buildings. It is very likely that the requirements of ASHRAE 90.1 will continue to result in lower energy consumption by buildings and in even higher thermal performance requirements of building envelopes. We can reasonably predict that lower allowable percentage limits on fenestration prescriptive values are on the horizon; we may even see a return to different allowable values based on the facade demands, such as in north versus south elevations. Both types of limits could easily become a consideration for prescriptive building envelope requirements. Or a new position could be taken on fenestration whereby U-values are at one number during the occupancy of the building and at a lower U-value requirement during no occupancy. Note, most commercial buildings are at a state of being unoccupied for 8 to 12 hours a day, and with active wall systems, fenestration can be improved to a higher level of thermal efficiency at this period of nonoccupancy.
JENSEN AND JOHNSON, DOI: 10.1520/STP161720180114
While it does appear likely that the mandatory criteria for energy performance of buildings will continue to become more stringent over time, the question of how high performance values can go is one that should be considered. Is there a limit? If so, is it a limit that can be met with the design and analysis tools and materials available today? Is net zero the answer, or are energy-producing buildings the next requirement? Will the incentive/profit to manufacturers of energy-saving products, systems, and materials to produce the needed new materials be adequate to gain the needed attention and commitment to accomplish the task? Should we use life-cycle cost analysis as a basis for thermal performance decisions or energy conservation as a whole? These and other questions will require close attention on the part of building design and construction professionals, as well as those promulgating the new code criteria. There is an urgency for action on each of these questions. The requirements for higher performing and lower energy usage in buildings is real, it is current, and it is not likely to become less, but rather greater, in importance. Continuing to add layers and thickness, or even greatly improved thermal efficiency of opaque materials and insulation to improve the thermal performance of buildings, is a good first step; however, it is not the long-term solution for saving energy in buildings. Real and effective changes in our education, analysis, and design procedures, materials, products, systems, and regulatory procedures are needed to accomplish this goal.
References 1.
2.
3. 4.
5.
6.
7. 8.
ASHRAE 90.1, Energy Standard for Buildings Except Low-Rise Residential Buildings, ANSI/ ASHRAE/IES, 2019, https://web.archive.org/web/20200427115016/https://www.ashrae. org/technicalresources/bookstore/standard-90-1 International Code Council, International Energy Conservation Code, 2018, https:// web.archive.org/web/20200427115338/https://www.iccsafe.org/products-andservices/ i-codes/2018-i-codes/iecc/ International Code Council, International Building Code, 2018, https://www.iccsafe.org/ codes-tech-support/codes/2018-i-codes/ibc/ U.S. Department of Energy, Building Energy Codes Program, 2020, https://web.archive.org/web/20200427115547/https://www.iccsafe.org/products-andservices/i-codes/ 2018-i-codes/ibc/ U.S. Department of Energy, Building Energy Codes Program, State Code Adoption Tracking Analysis, 2020, https://web.archive.org/web/20200427120132/https://www.energycodes.gov/state-codeadoption-tracking-analysis Title 42–The Public Health and Welfare, Energy Conservation and Production Act, 6834 Federal Building Energy Efficiency Standards, 2008, https://web.archive.org/web/ 20200427121022/https://www.govinfo.gov/content/pkg/USCODE-2011-title42/pdf/ USCODE-2011-title42-chap81-subchapII U.S. Department of Energy, State Energy Program, 2020, https://web.archive.org/web/ 20200427122205/https://www.energy.gov/eere/wipo/stateenergy-program U.S. Department of Energy, Building Energy Codes Program, Michigan, 2020, https:// web.archive.org/web/20200427122458/https://www.energycodes.gov/adoption/states/ michigan
47
48
STP 1617 On Building Science and the Physics of Building Enclosure Performance
9.
10.
11. 12.
13. 14. 15.
16.
17.
18.
19.
State of Massachusetts, Stretch Energy Code, 2018, https://web.archive.org/web/ 20200427123505/https://www.mass.gov/regulations/780-CMRchapter-115-aa-stretchenergy-code National Fenestration Rating Council, NRFC-100, Procedure for Determining Fenestration Product U-Factors, 2017, https://web.archive.org/web/20200427123722/https://www. nfrccommunity.org/store/ViewProduct.aspx?id=1380591 Vitro Architectural Glass, 2020, https://web.archive.org/web/20200427125549/https:// www.vitroglazings.com/ Kawneer Company, Curtain Wall Systems, 2020, https://web.archive.org/web/ 20200427124745/https://www.kawneer.com/kawneer/north_america/en/info_page/ home.asp Guardian Industries, Vacuum Insulated Glass, 2020, https://web.archive.org/web/ 20180720064048/https://www.guardianglass.com/us/en/vig/index.html USGBC LEED Program, Building Design + Construction, 2020, https://web.archive.org/ web/20200427135245/https://www.usgbc.org/leed Executive Order 13514, Federal Leadership in Environmental, Energy and Economic Performance, https://web.archive.org/web/20200427135515/https://www.fedcenter. gov/programs/eo13514/ EPA, Laws and Regulations, Summary of the Energy Independence and Security Act, Public Law 110-140, 2007, https://web.archive.org/web/20200427135754/https:// www.epa.gov/lawsregulations/summary-energy-independence-and-security-act EPA, Energy Star U.S. EUI by Property Type, 2018, https://web.archive.org/web/ 20200427140615/https://portfoliomanager.energystar.gov/pdf/reference/US%20National %20Median%20Table.pdf Energystar.gov, https://web.archive.org/web/20200427140314/https://www.energystar. gov/buildings/facilityowners-and-managers/existing-buildings/use-portfolio-manager/ understand-metrics/whatenergy “Federal Guiding Principles for Sustainable Federal Buildings,” 2016, https://web.archive.org/web/20200427141808/https://www.energy.gov/eere/femp/sustainable-federalbuildings
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180067
Paul E. Totten1 and Charlotte Metzler1
Building Physics of HVAC Interaction with Enclosure Systems and People Citation P. E. Totten and C. Metzler, “Building Physics of HVAC Interaction with Enclosure Systems and People,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 49–66. http://doi.org/10.1520/STP1617201800672
ABSTRACT
Many building types are designed, especially in the commercial office field, as a cold dark shell where interior tenant fit-out will occur later. Many aspects of the fit-out, such as final heating, ventilation, and air-conditioning (HVAC) system layout, partition wall locations, and the location of larger assembly spaces like conference rooms or other amenities, may not be fully defined at the same time as the core and shell. If tenant spaces are designed separately from the core and shell, the disconnect between the two designs may unintentionally impact the comfort of the interior space where it does not perform at a high level. As people utilize and move through the space, they are constantly being impacted by the building physics of the interior environment. This includes moisture and heat transfer across their skin. Clothing and footwear can impact a person’s overall comfort within the space based on how heat transfers through these and, in some cases, on moisture load within the space. System furniture, layout of structure, and HVAC system type and distribution can also affect a person’s comfort by adversely impacting the intended building enclosure performance. As the art and science of building design becomes more complex, it is necessary for designers to maintain a suitable proficiency in both these spheres and understand how different building systems interact. The wealth of scientific and technical data relating to the performance of buildings is one of contemporary architecture’s Manuscript received October 7, 2018; accepted for publication June 9, 2019. 1 WSP, 1300 North 17th St., Suite 1000, Arlington, VA 22209, USA 2 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
49
50
STP 1617 On Building Science and the Physics of Building Enclosure Performance
primary assets. This paper will cover the aspects of heat, air, and moisture transfer that impact our built environment and the interaction of HVAC with interior spaces, enclosure systems, and people. The authors will use examples from several projects to identify the concerns and practical solutions that can be implemented to improve a person’s comfort within interior spaces. Keywords building science, HVAC design, building enclosures
Introduction Vitruvius noted in The Ten Books on Architecture, “The architect should be equipped with knowledge of many branches of study and varied kinds of learning, for it is by their judgment that all work done by the other arts is put to test. This knowledge is the child of practice and theory.” This concept is exhibited through the understanding of heating, ventilation, and air-conditioning (HVAC) interaction with space and people and its direct impact on occupant health and comfort and building performance. With ever-expanding building technologies, the need for expanded understanding and knowledge has never been more important. HVAC systems are varied in type, with many options for designers to choose from. This, coupled with climate change considerations, evolving enclosure system design, and building physics and its relation to macro- and microclimates, creates the need for increased knowledge about these impacts on building performance and occupant comfort in the profession. Every space we use in a building, whether through passive or active means of heating, ventilating, and cooling, is impacted by a constant dynamic of heat, air, and moisture movement within the walls, ceilings, roofs, floors, fenestration, and other assembly components. HVAC systems are designed for space comfort. This does not always equate to occupant comfort if the end-use group is not known at the time of design. In addition, the dynamic forces interacting with the space’s enclosure assemblies often creates a condition different from that which was designed. The comfort needs of children in day care or elementary school and adults in the same space, patients and doctors and nurses in hospitals, office employees, and those living in residential structures are all different. ASHRAE Applications Handbook1 provides great guidance on recommended conditions for the space, but as individuals each have different needs and perceptions, how this performs in the space can be more challenging. The examples that follow show a holistic approach and needed collaboration among the HVAC design team, the architecture and interior design teams, and the building science and enclosure specialists. It is this collaborative approach coupled with science and understanding of human psychology and perception of space that presents an opportunity to provide enhanced detail concepts and solutions to occupant discomfort.
TOTTEN AND METZLER, DOI: 10.1520/STP161720180067
Modes of Heat Transfer How occupants experience comfort is typically at a skin-to-air or skin-to-clothingto-air relationship of heat and moisture transfer. This includes diffusion of moisture through the skin, or in the case of sweat, from skin-to-air, and heat movement by conduction, convection, and radiation. Convection can be initiated by movement of a person within a space, air moving in a space driven by HVAC systems, and similar phenomena. Conduction can occur at different rates depending on the space’s stratification of temperature under stack effect. Knowing that heated air rises and cold air falls, inevitably at most times of the year, the coldest air in an interior space is almost always at our feet, with radiant heating in a floor or delivery of heat at the floor line in a raised floor system creating exceptions. Conduction is the transfer of heat from molecule to molecule. Through it, a temperature gradient in multiple dimensions occurs. We tend to view this in onedimensional (1-D) and sometimes 2-D relations, but how we experience it in reality as people is in 3-D. A temperature delta from one area of our arm, for example, to another of 1.667 C to 2.77 C (3 F to 5 F) is perceptible and, across a larger temperature range, is more than noticeable. As humans, we have a body temperature of 37 C (98.6 F) on average. Most interior spaces are maintained at 22.2 C (72 F). Therefore, we are warmer than most surrounding space and will radiate to any surface colder than us. Radiation helps with cooling the surface of our skin as well. If we are near a window that is quite a bit colder, we can radiate more heat through the line of site. This is why people can experience a temperature delta when positioned too close to a window even if the window is inoperable, such as in an office environment. In this situation, one side of the person feels colder than the other in winter months and warmer in summer months. When we move, we create movements in the air and air currents around us. This leads to convection between our warmer body temperature and the space around us. Airflow in the space caused by HVAC and other means (doors opening and closing and elevators acting like pressure-relief valves) also drives convection. These three constantly occurring modes of heat transfer not only define how occupants feel in a space but also how air and moisture move and impact directional movement of heat and air across our building enclosure systems. Deepening our understanding of this constant dynamic will drive forward better design for future new construction and retrofits.
Tall Building Stack Effect—Not Always What It Seems The general assumption for tall buildings with pressurization issues is that stack effect (buoyancy of warm air to rise and cold air that is heavier to fall) is a predominant cause. Stack effect must be examined alongside air tightness, including floorto-floor air tightness, HVAC systems type and layout, and building pressurization.
51
52
STP 1617 On Building Science and the Physics of Building Enclosure Performance
In the case of one 16-story building study we conducted, entry doors experienced operational issues when closing and opening, and widespread occupant comfort issues were reported throughout the building. Occupants complained about draftiness, and feelings of temperature discomfort throughout the year. Several previous studies had been completed that concluded that a combination of stack effect and air leakage through the building enclosure were the root cause of all the issues in the space. Yet adjustments to the HVAC system to compensate for stack effect in some seasons only made the issues worse. Descriptions of stack effect and diagrams describing the issues found in the previous studies did not align with the phenomena as described throughout the previous reports. Many of the diagrams showed that, in this 16-story building, all air was being drawn to a central shaft to the eighth floor throughout the year. The diagrams did not show how floor-to-floor compartmentalization, which was done in most cases, would reduce impact for most spaces due to the stack pressure being distributed floor to floor. The portions of the building that should have had a more positive pressure and at the middle section’s neutral pressure were not experiencing these conditions, which implied stack effect was not the predominate driver. Therefore, HVAC flow and system layout and operation required closer examination. This became very suspect because the eighth floor housed the HVAC mezzanine and most of the building’s air handlers, including heating units that did not have return fans and automated dampers. In addition, when walking the building and meeting with tenants, and further reviewing the building condition with facilities staff, we found that only variable air volume (VAV) boxes located at the building perimeter had a reheat coil, whereas all year interior VAVs only supplied ventilation air. Ventilation air was minimally heated and ranged from 12.78 C to 15.56 C (55 F to 60 F). This was much cooler than air delivered at the perimeter, which ranged from 29.4 C to 32.2 C (85 F to 90 F) at the diffuser in an attempt to maintain 22.2 C (72 F) in the space. This identified another potential cause of occupant discomfort because, in many spaces, perimeter and interior diffusers influenced the space. In testing for overall building air tightness, pressure mapping using differential pressure gauges (fig. 1) was used. Pressure mapping showed consistent zones of negative pressure on several floors, including those that typically would have been at or close to neutral pressure for the overall building near the eighth-floor HVAC mezzanine. At the top and bottom of the central shaft, which ran the full height of the building, pressure gauges with a data logger showed the same pressure amount, typically negative returning through the shaft that was intended as both a hard ducted and passive return air shaft. All signs pointed to mechanical flow to the eighth-floor HVAC mezzanine and not stack effect as the predominant driver. Localized conditions of stack effect were found at penetrations through a front elevation wall at Floors 4 to 7 where flagpoles had been mounted (fig. 2). The structural flagpole connection through the exterior wall was not inherently airtight at these penetrations. The connection was made with tube steel brackets that allowed exterior air to traverse from the open flagpole tubes inboard. No closure was
TOTTEN AND METZLER, DOI: 10.1520/STP161720180067
FIG. 1 Differential pressure gauge used for pressure zone mapping.
FIG. 2 Flagpole feature shown from exterior—horizontal supports in tube steel anchored back to slab inside were open to verticals of exterior flagpole.
provided at the brackets to the flagpole to prevent this airflow, and pressure readings aligned with stack effect and not wind flow. Readings showed a steady rise in pressure as the sun heated the flagpole tubes connected to the structure and a falling
53
54
STP 1617 On Building Science and the Physics of Building Enclosure Performance
pressure during nighttime that was indicative of stack effect during night cooling at the exterior. These were localized to a few areas of complaints and recommended to be air sealed at the structural tubes that traversed the enclosure systems to prevent inside to outside air draw and vice versa. They were minimal in number and did not align with the total building complaints. Another common building element that can experience stack effect is an elevator shaft. This is due to the height of the shaft and is also influenced by elevator cab movement creating the plunger effect of air movement, which is different than stack effect. While riding the top of an elevator car with an operator, pressure mapping and air leakage testing with small theatrical smoke-generating devices was completed. The pressure mapping showed a normalized stack pressure with a neutral pressure plane just above mid-height of the shaft and pressure readings above and below mid-height consistent with stack effect pressures at the time of year the investigation was undertaken. In addition, as the elevator cars on either side of the elevator we utilized were still active, the plunger effect of air movement was also experienced. This building’s elevator core was mostly disconnected from other building elements (it was located one corner of the building and only tied to a portion of the entry lobby) and therefore did not contribute to the pressure issues the building experienced overall. Many of the complaints came from zones well away from the elevator core. In reviewing the HVAC design, some systems functioning predominantly for cooling were located at the penthouse level. However, the main air handlers for heating were located on the eighth floor. A common way of modulating pressure through air handlers is to provide a return damper and return fan within the air handler on the return side of the system. Based on space layout limitations, however, the heating season air handlers within this building did not have return dampers and fans. The return side of the system was within 4 ft of the back of the exterior precast panel walls. In order to enhance return airflow to the system, the manual louvers for the air handlers were left wide open in winter. This resulted in larger draw back to the system than that being supplied via the VAV system. This imbalance in supply and return with a larger return air volume cycling through the air handler on a consistent basis resulted in a building operating under negative pressure. Air from lower floors was drawn up to the eighth-floor mechanical mezzanine and air from upper floors was drawn down. This explained why the building had a mostly negative pressure during the heating season but did not fully explain the draftiness. Air leakage testing at localized conditions such as the roof to rising precast wall interface and at a few other locations did not align with all of the thermal discomfort occupant complaints. At Floors 5 and 6, which had higher slab-toslab heights and a drop ceiling set about 8 ft from the slab, the open ceiling plenum saw localized stack effect in the winter where heat would be driven more readily into the ceiling plenum and would cause cooler conditions to be maintained in the occupied space. Changing the ceiling design to a hard ceiling that was air-tightened with airtight fixtures and a ducted return would help resolve some of the issues
TOTTEN AND METZLER, DOI: 10.1520/STP161720180067
occurring at these floors. However, at these and other floors, many of the complaints during the heating season were a combination of feeling hot and cold in the same space at varying locations. In reviewing the layout of systems for the VAV boxes and diffusers, it was determined that only the perimeter wall VAV systems were equipped with dualduct heating at the terminal unit. Single-duct VAV boxes inboard did not have the ability to reheat locally and thus many times were only blowing cooler ventilation air at the occupants. Note that this localized phenomenon can occur in many buildings near the diffuser, especially those that have higher air change per hour (ACH) requirements where ventilation air is supplied more often. The ventilation air was drawn in and heated to about 12.78 C (55 F) off the air handler’s preheat coil, which is very common. This resulted in differing conditioned zones within the space, many times within a single office or conference room. As with any HVAC system, sensors analyzing a space’s temperature and relative humidity and thermostats to set the temperature are important interactive requirements for proper HVAC operation coupled with proper design. An understanding of occupant needs in each space plus the ability of the control system and HVAC system to meet the demand is critical. Within this property, multiple-use groups also had differing needs for temperature, and the operation of most of the HVAC system did not readily allow easy adjustment of one space without negatively impacting another. The absence of a reheat coil at interior VAV boxes also led to many occupant complaints of draftiness. Although perimeter VAVs had the ability to reheat the ventilation air, the spaces that had perimeter zone and interior zone systems were constantly opposing each other for temperature balance. At some periods, this resulted in more heat being delivered. However, ventilation air delivery was also signaled and constantly resulted in the feeling by occupants of hot then cold in colder months, which they wrongfully attributed to draftiness through exterior walls. Although thought to be caused by stack effect and air leakage through the building enclosure, the building’s issues instead were HVAC issues that impacted how spaces were conditioned and experienced by occupants. Although the issues are now well understood, unfortunately, there is no easy way to fix the building, especially in a cost-effective way. Our recommended actions included localized air sealing at known enclosure issues including the flag pole connections and review of the option to make interior VAV boxes dual ducted with a localized reheat coil. These are components that require careful review during the design period, including allowing sufficient space for the air handler and other HVAC systems to have all components provided for more effective building operation.
HVAC Considerations in Spaces for Children For day care and classroom spaces for small children, we see issues related to spaces in the way they are designed to condition the space. HVAC designs, especially if the end use of the space is not yet known (more common with daycare spaces), are
55
56
STP 1617 On Building Science and the Physics of Building Enclosure Performance
predominantly focused around supply and returns that mostly benefit adults in the space. Children, who are smaller and shorter in stature than adults, do not always experience the same level of conditioning as adults within the same space, especially if a forced air system with high supply diffusers is utilized. This can relate to spaces like day care centers or elementary school classrooms having challenges with children becoming ill more often as air changes and temperatures are not cycled at the same level in the lower portion of the room as the higher portion of the room. The mechanically supplied air in these spaces does not adequately condition the space at the level of the occupants—small children. This is especially true if return vents and louvers are located in the mid to full height of the space and if diffusers are laid out to condition fenestration with the wash coming down from the top. The inadequacy is exaggerated at naptime or other times of the day when children may be sitting or lying at floor level, especially with floors at grade level with a slab on ground that may not be fully insulated below. These issues are occurring in spaces designed for this occupancy group. However, many day care or even elementary school spaces are in areas that were not initially planned to be used by this occupancy group, thereby furthering the issue. In addition, placement of furnishings and reconfigurations that can occur for more effective teaching will change how the occupants experience the space. Existing HVAC designs are rarely reconfigured for new occupancy types, let alone designed to accommodate children and adults within the same space. Radiant heating and cooling systems set at the slab level are one way to condition the low-height space occupied by children, if this is integrated into an improved control system for forced air needed for ventilation air and other conditioning. Providing insulation underneath the entire slab-on-ground also helps maintain consistent temperatures. Ceiling fans can also help with mixing air within the space. In order to provide adequate comfort levels for stratified conditions within a space, controls must be able to take readings from the stratifications for optimal conditioning for both children and adults. This concept is similar to climate control features within cars where each occupant is able to control the conditions of their own space. Ventilation air being delivered low across radiant heating and cooling systems coupled with a high return can help initiate convection within the space. This will provide more air changes at the occupant level than that achievable within typical system layouts. While the occupants in these spaces require a slightly customized approach to occupant comfort, the solution is attainable and can result in full-scale occupant comfort.
HVAC Flow and Interior Fit-Out When considering the interactions among window system placement, interior finishes and design and furnishings locations, HVAC system supply and return placement, and the collective impact on occupants, solutions applied to transform a building shell into tenant spaces do not always translate into occupant comfort the
TOTTEN AND METZLER, DOI: 10.1520/STP161720180067
FIG. 3 Convection that can occur in a space prior to tenant fit-out (cooling season shown).
way the designers intended. In a cold dark shell without furnishings, supply can be seen as shown in figure 3. The moment interior fit-out occurs without reconfiguration, or without examining the influence of system furniture, these flow paths can change drastically, resulting in occupant discomfort. On this project, numerous occupants with system furniture located along the exterior wall over a soffit condition had reported occupant discomfort predominantly in winter, but there were also summertime complaints. The soffit assembly, from exterior to interior, included exterior sheathing with a painted finish at the outside, insulation, sprinklers and plumbing pipes within a heated plenum space, and a concrete floor slab with carpet on the top surface. The complaints in winter ranged from cold sensations at the feet and lower extremities to exaggerated hot sensations at the upper body—all while within the same space. Summer complaints including hot sensations at the feet.
57
58
STP 1617 On Building Science and the Physics of Building Enclosure Performance
We were asked to investigate the space during the summer months due to other project work ongoing at the building. We measured air and surface temperatures using an infrared temperature gauge. We used an additional handheld gauge that could also read air temperature and relative humidity and, through a thermocouple, surface temperature. We installed data loggers connected to multiple thermocouples to log surface temperature, RH, and occupancy (using a lighting sensor) over a longer period of time. On the day of our walk-through, when we stepped out over the soffit zone on the south elevation and as we moved closer to the exterior wall, we experienced what occupants had been noting. We felt through our footwear an elevated temperature. High temperatures, up to 32.2 C (90 F), were recorded at the top surface of the carpet on the floor slab along the south elevation by both gauges, including utilizing the thermocouple adhered to the surface and normalized. Vertical precast cladding panels with a reddish color at the perimeter of the building were insulated at the inboard side of the wall above the floor slab (between metal studs) but not below the slab within the soffit space. A thermal bridge was therefore created at the interface between the vertical precast panels and the concrete slab and through the stud wall due to a lack of continuous insulation. In the summer, this allowed heat to transfer from the warm exterior conditions to the cooler interior conditions via conduction, and it was intensified on the south elevation due to outside-to-inside conduction and solar radiation. Conversely, in winter, heat transfer occurred from the interior to the exterior via conduction. Heat transfer was also occurring across the floor slab between the exterior soffit space and the interior. However, this interior-to-exterior heat transfer was minimized within the zone of influence of the heaters located within the soffit assembly. Due to heat transfer through conduction, occupants at least 6 to 8 ft from the heater locations experienced enhanced heat loss from their feet through the slab to the exterior. System furniture layout in the space also impacted occupant comfort. The walls of the system furniture at the desk acted like ductwork, which directed air to a gap between the base of the system furniture and the floor slab and acting like a diffuser (fig. 4). In summer, as air traverses this wall and the floor due to the thermal bridging and especially in daytime hours, heat would be picked up and transferred below the desk. In winter, two modes must be examined. When calling for heat, temperatures at the diffuser are 37.8 C (100 F) or higher many times; ventilation air is much cooler and is maintained on this building at 18.3 C (65 F) (fig. 5 and fig. 6). The space experiences the same conditions most spaces do when only ventilation air is supplied and, with mixing in the space, occupants experience about a 1 C to 2 C (2 F to 3 F) differential. The largest temperature differentials occur above and below the desk of the system furniture at times of heating. As air traverses the thermally bridged conditions, it cools, resulting in cold temperatures at the floor slab and below the desk to which the occupant conducts heat; warmer temperatures are found above the desk where the rate of human-to-space conduction (and radiation)
TOTTEN AND METZLER, DOI: 10.1520/STP161720180067
FIG. 4 Convection and conduction during summer months in the space after tenant fit-out but prior to insulating.
is slower. These differences, predominantly in heat conduction, were what was driving occupant comfort issues and complaints. Our solution involved installing insulation at the underside of the concrete floor slab within the soffit condition (fig. 7). Mineral wool insulation was installed in two layers to a total thickness of 6 in. The outboard side of the outboard layer of insulation had an integral air and vapor barrier to control vapor flow and airflow. This adjustment substantially reduced the thermal bridging at the slab, and occupant comfort complaints related to this issue were eliminated in subsequent years. As the sprinkler system and plumbing still required a semiheated condition within the winter months, the insulation above the exterior soffit sheathing remained within the assembly design so that the heaters could provide heat between the two locations of insulation.
59
60
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 5 Winter convection and conduction under heating mode that can occur in a space after tenant fit-out during calls for heat, prior to insulating.
While it appeared that the existing design provided continuous insulation at the building exterior, a discontinuity in the insulation occurred at the floor slab. Although the soffit insulation maintained a conditioned space suitable for the plumbing pipes, it did not reduce the heat transfer across the floor slab. Analyzing the building enclosure for possible discontinuities within the thermal boundary and examining the possible influences of interior furniture layout on intended HVAC performance are important ways to reduce the risk of occupant discomfort.
Odors in Spaces Unintentional odor flow can be caused by interior flow paths influenced by HVAC supply and return forces and external forces such as wind washing along the facade
TOTTEN AND METZLER, DOI: 10.1520/STP161720180067
FIG. 6 Winter convection and conduction that can occur under ventilation mode, which can occur after tenant fit-out and prior to insulating.
or wind direction carrying exhaust into fresh air intakes. At one particular property, occupants complained of a variety of unpleasant odors occurring in a number of rooms. The original HVAC design provided the main air supply into a large central atrium, with returns occurring in each corridor. Both ends of each corridor were connected to the atrium at each floor with doors that were originally kept open to access a balcony at each floor (fig. 8). Due to safety concerns, the fire doors that connected
61
62
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 7 Additional insulation provided as part of the repair design changed the dynamic and resolved the volume of complaints.
the corridors to the atrium were permanently closed at all floors. No alternate means to allow fresh air into the corridors from the atrium was provided. Therefore, the intended HVAC design was no longer achieved, and stagnation of air occurred within the corridors (fig. 9). This caused return flows to maintain a lower pressure than the atrium supply. Any odors that occurred in any room off of the corridor also stagnated, resulting in complaints from occupants in numerous rooms. Separate from the stagnation issues within corridors, the exterior window system had varying degrees of perimeter sealant failure. This not only created risk for water entry but also, based on testing with theatrical smoke, risk of exterior-to-interior airflow at the windows. Occupants within rooms located along the alley side of the building at multiple floors consistently reported complaints of unpleasant odors
TOTTEN AND METZLER, DOI: 10.1520/STP161720180067
FIG. 8 Mechanical system flow paths driven by original design that provided a better means for ventilation through the corridor space.
including vehicle and diesel exhaust and food waste. In examining the various sources of potential odors, we identified diesel generators, loading docks, and dumpsters for adjacent restaurants within the alley. In addition, the wind flow direction predominantly pointed toward the building’s wall within the alleyway where it would traverse vertically up the wall. Air sealing the windows with new perimeter seals was required along with wet sealing the glazing system to reduce the entry of these exterior smells. Air leakage paths transported exterior air as well as other environmental impacts including sounds and odors. The proposed solution for the corridors was to hard duct air supply to the corridors and install passive smoke and fire dampers at the end of each corridor over the fire doors. This would be challenging to install but would provide improved airflow through the corridors. The proposed solution within the rooms, along the alley especially, was to air seal the exterior windows from the inside and outside by wet sealing the fixed glazing units.
Glazing and HVAC Interaction Atrium spaces with their large expanses are challenging for HVAC design and enclosure considerations, especially if portions of the architecture are more
63
64
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 9 With the doors to the atrium closed, HVAC flow for ventilation was greatly minimized and more stagnation of air occurred.
complex. With large portions of many atriums having glazed exterior walls, it is important to prevent condensation from occurring at all times of the year. The heat loss occurring across a glazing unit is more rapid than the loss occurring across an opaque insulated wall. The glass assembly for the wall studied here is laminated (multiple layers of plate glass with two interplies of laminate), not an insulated glazing unit (IGU). Additionally, atrium spaces are multistory spaces within which consistent temperature control across these spaces is difficult to manage. Preventing condensation is even more challenging if the space adjacent to an atrium requires humidification to maintain conditions appropriate for a particular use. On this project, the glazing would be further impacted by wind washing across the exterior glass atrium wall much of the year. Wind washing further reduces the surface temperature of the glazing from ambient air temperature and, in winter, results in higher conduction rates across the glazing. This would result in colder temperatures at the interior face of the glass and subsequently would increase the risk of wintertime condensation at the interior. The property would experience a heightened occupant load during daytime hours within the atrium and a minimal occupancy load overnight. This increased occupancy load during the daytime hours would also increase the expected moisture load in the space.
TOTTEN AND METZLER, DOI: 10.1520/STP161720180067
In an effort to maintain consistent air conditions across the height of the atrium, an HVAC design method was proposed where return ducts were located high within the space and supply ducts were located at each interior floor slab angled into the space and pointed toward the glass. The glazed wall was vertical, although shaped across the horizontal plane. The architecture of the interior side of the space angled from top to bottom inward, meaning diffuser distance along this wall would be closer near the top but further away near the bottom. Diffusers near the top had some risk of short cycling, and thus additional layout and angle considerations for these rows of diffusers was provided. At the bottom, the distance away from the glass, considering flow rates and diffuser type, would have the risk of not enough heat supplied to the glazing. This would result in the risk of less heat delivery from the HVAC system at the lower atrium space, which adds to the risk of condensation occurrence already heightened by the cool exterior wall temperature and the heightened humidity load of the atrium and adjacent spaces. In order to mitigate the risk, the design team (architectural design team, glazing consultant, HVAC designer, and enclosure consultant) worked together to determine a solution. The first decision made was to provide doors to separate the adjacent highhumidity space from the atrium, with the potential to provide air curtains above the door. As air supplied into the atrium was not humidified, this separation was critical to help minimize extraneous humidity loads from adjacent spaces into the atrium space, thus lowering moisture loads. The next step was to examine better ways to bring heat to the glazed monumental atrium skin. We analyzed two modes of heat transfer—conduction and convection—for the space and attachment components for the glazed wall, which also had the potential of allowing the glass system and its support to act like a radiator. As discussed in earlier case studies, air movement that can initiate convection is not only created by HVAC flow but also by people moving through a space. Conduction is most effective through highly conductive materials. We examined ducting heated air into the steel framing tubes that supported the atrium glazing system in addition to diffusers to condition the space. The ducted air would heat the steel and, through conduction, would heat the aluminum framing supporting the glazing. This in turn would create a potential for the glass and frame support to also radiate heat. Considering supply air delivery would be well above 37.8 C (100 F), this temperature, even with losses across the steel, would result in more heat delivery directly to the glass. Coupled with diffusers to heat the space, air movement from the forced air system along with these new temperature differentials would drive convective loops in multiple orientations, resulting in more passive mixing of air and less risk for moisture stagnation at the glazing. From a conduction standpoint, the more wind washing the glass receives (the more conduction that will occur to the outside), the larger the heat draw from the space and drive for convection, using the wind washing as a passive natural enhancement to this combined system. In addition, the surface of the glass would
65
66
STP 1617 On Building Science and the Physics of Building Enclosure Performance
lose heat to the exterior at a faster rate within the zone of influence of the steel framing (higher localized temperature delta), thus creating differential rates of conduction across the glazed atrium wall. This typically would result in additional convective loops localized to this phenomenon. The air mixing would also drive localized diffusion off the glass, which for much of the winter will actually have a lower risk of condensation along the glass. In the end, the cost to use the steel tubes themselves as ductwork was too high. However, a compromise was determined to supply diffusers close to the steel and angled such that the supplied air would heat the steel to create a phenomenon similar to the originally proposed solution. The building is still under construction, but it will be monitored to ensure the performance meets the design intent. During the end of construction and start of occupancy, optimization can be evaluated and adjusted during additional HVAC system commissioning to reduce condensation risk, if needed.
Conclusions The lessons learned from the case studies presented here and numerous other projects with which the authors have been involved emphasize the critical importance of understanding the relationship of HVAC design, occupancy type, and the building enclosure. This involves a multidisciplinary approach and increased communication and discussions across the design and consulting teams. Discussion among the HVAC designer, architect, and enclosure and building science specialists is a key component of enhancing design for occupant comfort. The HVAC layout, furnishings layout, type of enclosure systems, and intended occupant zones within a particular type of space are all critical factors to review as an ensemble to provide the intended performance of a space. By performing these collaborative reviews early in the design process, and learning from examples of past concern, we can provide the intended occupant comfort and improve building performance.
Reference 1.
American Society of Heating, Refrigerating and Air-Conditioning Engineers, ASHRAE Handbook—HVAC Applications (Atlanta, GA: ASHRAE, 2015).
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180075
Kyle Normandin1 and Robyn Pender2
A Window of Opportunity Citation K. Normandin and R. Pender, “A Window of Opportunity,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 67–87. http://doi.org/10.1520/STP1617201800753
ABSTRACT
The demand for the adaptive reuse of prewar and postwar landmark buildings and structures is rising, and it will continue to rise: particularly with a changing climate, where energy conservation has become a significant influence on refurbishment. There is a widely held belief that upgrading, or reimagined occupancy, must be in practical or commercial conflict with preservation; but in fact, projects informed by a good understanding of the behavior of the building envelope and the significance (or otherwise) of its components can achieve remarkable results. This paper highlights the wider role that must be played by building science in developing a more nuanced understanding of energy use in the built environment: one that looks at the fundamental role of occupants and their control over their environments, as well as the fabric. This approach to minimizing energy use until very recently drove innovation in the built environment, and it retains its power. By looking at the example of architectural glazing through history, it is possible to see that energy conservation and conservation are twin goals. Together, they can deliver enormous benefits in terms of both sustainability and improved conditions for building use. As examples of this approach, the paper discusses two recent Californian examples involving the conservation of twentieth-century glazing: the Salk Institute for Biological Studies in La Jolla and the former Gibraltar Savings and Loan building in Beverly Hills. These case studies demonstrate how intervention and conservation can be balanced to maximize retention of original buildings’ fabric
Manuscript received October 12, 2018; accepted for publication April 15, 2019. 1 Wiss, Janney, Elstner Engineers & Architects, P.C., 1350 Broadway, Suite 910, New York, NY 10018, USA https://orcid.org/0000-0002-5310-5709 2 Technical Conservation, Historic England, 4th Floor Cannon Bridge House, 25 Dowgate Hill, London EC4R 2YA, UK https://orcid.org/0000-0001-5887-4438 3 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21-22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
67
68
STP 1617 On Building Science and the Physics of Building Enclosure Performance
and maintain their cultural and aesthetic significance, while ensuring long and productive futures for these buildings. Keywords building science, building enclosure, adaptive reuse, prewar landmarks, postwar landmarks
Introduction Glazed windows are recognized as a primary weak spot in building enclosures; but what does that "weakness" actually entail? It is certainly not merely the transfer of heat through the glass, especially not heat transfer as roughly approximated in R-values (or U-values, which are used more commonly outside the United States). Many other factors are at play. Treating these highly complex building units too simplistically compromises our ability to not only maximize performance but also to reduce energy and carbon use in the built environment. Indeed, it risks encouraging retrofit measures that, over the longer term, use considerably more energy than they save. This paper attempts to take a broader look at the question of windows. By looking at the evolution of architectural glazing over the centuries, it aims to better understand the impact that glazing has had on the building enclosure. Some examples of good practice from building preservation are used to examine how a conservation viewpoint can deliver gains in sustainability as well as benefits to the enclosure for all buildings not just those of historic importance.
A Brief History of Architectural Glazing Originally, the primary purpose of windows was ventilation not lighting (the very word window comes from the Old Norse for "windy eye"). Building enclosures were intended to keep out the rain and wind, and many were of a massive construction that was ideal for cutting heat loss to the exterior in cold climates and heat uptake from the exterior in hot climates. The walls were therefore pierced as little as possible. In cold climates such as those of northern Europe, the original window openings (fig. 1) were very small, positioned high in the walls, and had simple wooden shutters to keep out the weather and control air exchange.1 Work that required lighting had to be completed either outside in daylight hours under minimal cover or inside using the sparse illumination of rush lights, candles, and fireplaces. The obvious benefits of being able to work in daylight hours under all weather conditions meant that, prior to the adoption of glass, many types of translucent material were tried for sealing window openings: alabaster and mica, thin slices of horn, and oiled paper, for example. All proved variously unsatisfactory. These days we have become so accustomed to buildings filled with architectural glass that it is easy to forget that this technology is actually both extremely sophisticated and very new in terms of building history. Learning how to make glass, and then how to form it into flat sheets thin enough to allow transmission of daylight, took well over 7,000 years. The fundamental
NORMANDIN AND PENDER, DOI: 10.1520/STP161720180075
FIG. 1 Boxford St Andrews: During building work on this English church in 2010, an Anglo Saxon window was discovered compete with its original wooden shutter. It is around 1000 years old, and dates from a time when windows were almost entirely for ventilation rather than light.
problem is that the primary ingredient of glass, silica, does not become workable until it has been heated to more than 2,000 C (3,632 F), and furnaces capable of reaching these extremely high temperatures did not exist before the twentieth century. To soften silica—sand—at the temperatures achievable in the kilns and furnaces of the ancient world (850 to 1,300 C or 1,500 to 2,400 F) required a series of scientific discoveries to be made. First was the use of fluxes: materials such as soda ash, which could be added to the sand to reduce its melting point. The product that resulted was very unstable, however, so the second vital step was to discover that lime added to the mix acted to stabilize it and make it waterproof. This technology may well have grown out of experiments in waterproofing ceramics. Glass was first made in Mesopotamia almost 4,000 years ago, but for many centuries, it appears to have been confined to ceramic glazes and the production of small beads. When glass vessels finally appeared, they were produced by coiling molten glass over a mold. It was not until the first century AD that it was discovered that glass could be shaped by blowing the molten material into a bubble. The Romans quickly developed a sophisticated glass industry based on this new technology, which concentrated almost exclusively on manufacturing glass containers; they did not find a way of forming glass into flat sheets other than by pouring the molten material into molds. A small amount of window glass was made in this way, but it was extremely expensive and not particularly transparent. The development of glassblowing was limited by the nature of the blowpipes and other tools, which, until the fourth century AD, were made of ceramic. Effective glassblowing really demands tools that are both heat resistant and robust, which essentially
69
70
STP 1617 On Building Science and the Physics of Building Enclosure Performance
means metal and, more specifically, iron. (Bronze, for example, melts at 950 C [1,742 F] and is therefore unsuitable for working with molten glass.) Working iron, like glass, needs very high temperatures, and in Europe, it was not possible to melt and cast it until the blast furnace appeared in the fourteenth century. Until then, iron articles had to be made by softening and beating the ore in bloomeries. Making blowing pipes from this wrought iron presented a serious technical challenge. The primary technology for making thin sheet glass—the cylinder or broad glass, method—appeared sometime in the eighth or ninth century AD in monasteries in what is now Germany and Belgium. These were the research institutes and high-technology factories of their day. The method the monks developed, cylinder or broad glass, remained the primary way of producing window glass until the twentieth century, and it is still used today (fig. 2). A molten glass bubble is elongated by swinging it in a pit as it is being blown; the ends of the bubble are then
FIG. 2 Cylinder glass was the primary way of producing window glass for almost 1,000 years, and it is still used today. (A) Swinging the glass bubble to elongate it; (B) opening up the bubble to form a cylinder; (C) scoring the cylinder lengthwise; (D) opening out the scored cylinder to form a flat sheet by reheating it in a kiln.
NORMANDIN AND PENDER, DOI: 10.1520/STP161720180075
opened out to produce a cylinder of thin glass that is then scored down its length and returned to another furnace, where it opens out in the heat and flattens to form a thin sheet of even thickness. An alternative method of production, developed in France in the fourteenth century, produces an even thinner sheet—crown glass— where a large bubble of very hot glass is cut from the blowpipe and transferred to a rod, which is spun quickly until the bubble opens out into a flat disk. Being lighter, crown glass proved particularly popular in England, where window glass was taxed by weight. Although by the eleventh century glassmakers were making sheet glass in quantity, the energy needed to fire the furnaces, and the size of the furnace mouths, limited the size of sheet that could be produced. To make panels of glazing large enough use in windows, small pieces of glass had to be joined together with H-channels of lead, which in turn had to be supported by being tied to a supporting framework of iron (called the ferramenta). This was difficult and time-consuming work, but it produced a great art form; the monks also refined the coloring of glass with metal oxides and paints and began using their "stained glass" to infill the window openings in churches. The technology of glazing transformed building enclosures. Windows quickly became larger and larger, but as they did so some of the disadvantages began to make themselves felt. The materials of traditional walls (such as brick, stone, earth, and lime- or earth-based mortars) resist rainwater ingress by acting in the same manner as a greatcoat, where air trapped in the pores resists water uptake. Rain hitting the surface is held in the surface pores and does not run down the surface to any significant degree. Glass and metal, by contrast, act like raincoats; water hitting the surface will collect into flows and run downward and will be drawn into cracks or poorly sealed joints in the material by capillary action. Faced with the impact of large areas of glazing, architects had to find ways of minimizing the amount of rain hitting windows. They also had to deal with the impact of runoff on other parts of the fabric—particularly the areas of wall below the windows, which became prey to moisture problems. Working closely with the masons, and learning quickly from experience, they developed features such as cornices, hood moldings, label stops, and water-shedding frames, mullions, and projecting water-shedding window sills; and they learned to cut drips into the underside of these to stop water being carried back into the wall. All are as characteristic of Gothic architecture as the flying buttresses that so famously allowed the glass to occupy more and more of the wall area. Although they were also made very beautiful, these features of the enclosure were there to serve a practical purpose. Glazing was much too expensive to be used for windows in houses until the end of fifteenth century, when iron-framed glazed casements—effectively a glazed form of wooden shutters—began to appear as a mark of great wealth. It took at least two more centuries, and the production of glass in dedicated factories using furnaces burning coal rather than charcoal, for architectural glazing to become common. These glasshouses could make larger sheets that could be made into sashes using
71
72
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 3 The weakest point for glazed windows is where the glass rests in the frame, especially on the horizontal elements such as the sills. If rain and condensation running down the glass is able to penetrate, it will be trapped and can cause timber decay in wooden frames, or corrosion in metal.
simple timber frames rather than lead. Glass windows became larger, and their price began to drop. Timber framing had the additional advantage of being able to absorb the thermal expansion and contraction of glass (a problem that became more serious as the sheets of glass became larger and larger). For the most part, the wood used was oldgrowth heartwood and thus very resistant to rot and insects. The weak point was the joint between the glass and the frame, where water could be trapped. This therefore needed to be well sealed but without preventing thermal movement. The sealants used were soft and elastic putties made from oil and chalk that were finished with a coat of linseed-oil paint that prevented them from hardening and sealed the meeting point with the glass (fig. 3). Cheaper glass quickly led to the development of another new and sophisticated element for the building enclosure: the vertical sliding sash window, which as well as posing less risk of glass breakage than casements allowed for larger windows and very precise control of ventilation. As buildings began to be heated, and air exchange was reduced by more effective sealing of the enclosure, condensation on the interior of the glazing became more of a problem both for the glazing itself and for the building performance. Other new performance issues included: • Heating of the interior by the infrared component of sunlight. • Loss of heat energy from the interior.
NORMANDIN AND PENDER, DOI: 10.1520/STP161720180075
Drafts. Damage to interior fixtures and fittings from the ultraviolet component of sunlight. • Glare (although this was probably not a significant problem until the last quarter of the twentieth century, when computer screens began to be common). Windows were commonly protected by wooden shutters, which continued to be used both internally and externally even as glazing became standard. Many were split horizontally or had cutouts that allowed night flushing: the shutters could be closed for security, while the sliding sash windows were left ajar (fig. 4).2 To further counteract drafts and heat loss, blinds and curtains also became popular (the latter began as a way of counteracting drafts around doors). As soon as plain glass became cheap enough to make it possible, secondary glazing appeared, especially in colder climates such as in Russia and Germany. Glazed windows were given multiple layers of thermal and light protection, which occupants could freely adapt to suit their needs. Horace Walpole’s house at • •
FIG. 4 With vertically sliding sash windows, it is possible to upen the window but have the shutter closed. By piercing the shutter, the windows could be used for nightflushing of warm air without compromising security or privacy.
73
74
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 5 The thermal behavior of various traditional window assemblies was tested under laboratory conditions and shown to be much more effective than had been assumed.
Strawberry Hill in London provides a splendid example: the windows in the Tribune had both sliding shutters and sliding secondary glazing—the latter being fitted with decorative stained glass.3 Laboratory research (fig. 5) by Historic England and Historic Environment Scotland (among many others) has been able to confirm how effective shutters, blinds, and curtains are at reducing heat transfer.4,5 Thermal transfer is only a small part of the performance issues around glazing in buildings; it is critical to consider not only the window assembly and heat transfer equations but also the way building occupants experience thermal comfort and discomfort. Radiant heat loss and gain, drafts from air leakage, downdrafts caused by the cooling of air in front of large glass windows—all of these have a greater impact on comfort than air temperature, and hence, on the often energy-intensive actions such as space heating and air-conditioning taken by occupants to try to improve their interior environment. The importance of these other causes of discomfort is illustrated by one particular type of secondary glazing, used specifically to reduce condensation problems on stained glass windows. In environmental protective glazing (EPG), the “interspace” between the stained glass and a protective layer of glazing on the exterior is ventilated at top and bottom (usually to the building interior). Diurnal heating and cooling of the exterior glass drives an air flow that is surprisingly strong and effective, even when the interspace includes partial obstructions such as ferramenta. EPG was recently investigated by Historic England, who found that even in very damp buildings, condensation was almost entirely prevented on the historic stained glass.6 Modeling and measurement showed that the reduction in heat
NORMANDIN AND PENDER, DOI: 10.1520/STP161720180075
transfer, while not as great as for sealed secondary glazing, was still significant. However, church congregations installing EPG on their windows in the hope of improving thermal comfort have been bitterly disappointed; the air flow through the vents produces such strong drafts that they report themselves less, not more, comfortable. It is also important to recall that, although (in the UK at least) so much research into the impact of glazing on building performance has been devoted to heat loss, even in a cooler climate a more significant concern is heat gain. In recent research on comfort, occupants of an eighteenth-century terrace building in London reported overheating as their primary problem, even in winter.2 The impacts of solar gain were felt as soon as glazed windows increased in size in the eighteenth century, but a solution was quickly found in the form of awnings, which stop the glass warming and then radiating heat into the interior. Adjustable canvas awnings were commonplace throughout the eighteenth and nineteenth centuries, and engineers developed sophisticated models that also enhanced ventilation (fig. 6).3 As glass became ever cheaper, it began to occupy more and more of the building enclosure. Skylights and glass roofs were first introduced for shopping arcades (sunlight was critical in the days before clean lighting became possible with electricity) and became popular for the new railway stations. The first all-glass facades were for greenhouses, and this idea was adopted for the Crystal Palace, built in
FIG. 6 After their introduction in the eighteenth century, awnings quickly became very popular, although they were stripped away from most buildings in the second half of the twentieth century. This photograph of a Royal Garden Party in 1897 shows Buckingham Palace with ventilating awnings deployed on every window.
75
76
STP 1617 On Building Science and the Physics of Building Enclosure Performance
London for the 1851 Great Exhibition; but performance issues were immediately obvious.7 In June, the Times newspaper reported: “On Saturday, the oppressive heat proved too great even for the attraction of the Crystal Palace, and since it was opened we have hardly seen so small an attendance there… In vain did ladies appear in the thinnest muslin dresses, and gentlemen walk about with their hats in their hands. The wind would not blow in such a direction as to secure a thorough draft, and in the desperate effort to find relief from their suffering some clustered around the Fountains.”8 Despite these drawbacks, after World War I, modernist architects adopted glass as their principal facade material, and glazing continues to dominate building enclosures 100 years later.
Dealing with Performance Issues from Glazing As the history of architectural glazing reveals, adopting glass (a lightweight and temperature-sensitive material) as the boundary between two different hygrothermal conditions introduced new building performance issues. Problems such as condensation were exacerbated as the differences between the interior and exterior conditions were increased first by heating and then by air-conditioning (the latter being encouraged by overheating from solar gain). Early solutions were developed for glazed windows—including water-shedding features, shutters and awnings, and ventilation control—but these were lost over the course of the twentieth century for reasons of fashion. Clearly, such elements should not be seen as additions to windows; rather, a glazed window should be seen as an assembly of which these elements are an essential part. Good management of the enclosure should center on their retention or reconstruction, especially since they also have important contributions to make to the reduction in energy and carbon use in the built environment. To take the example of awnings, they are a particularly effective low-energy way to combat discomfort because they allow occupants to respond to local glare and overheating as the sun moves around the building. Unlike blinds, they do not cut daylight to the extent that artificial lighting must be used, nor do they block the views through the window. Awnings continue to be used successfully on buildings throughout continental Europe,9 although in the UK they all but entirely disappeared after World War II. Unfortunately, quantification of their benefits is still in its infancy, with most research into the thermal and lighting benefits being based on modeling rather than in situ measurement.10–12 This makes it challenging to optimize designs and materials. It is also important to find ways of quantifying energy and carbon benefits in terms of building occupation, recognizing that factors driving occupants to make use of energy-hungry internal systems such as heating, dehumidification, and airconditioning are never a simple response to temperature being exchanged through walls between the interior and the exterior.
NORMANDIN AND PENDER, DOI: 10.1520/STP161720180075
Through-life costing is another vital part of energy and carbon equations.13 Building conservators are particularly sensitive to this because they know the value of traditional materials, such as heartwood, which are not only very long-lived, but these days almost irreplaceable. This is as true of the glass itself as the materials of frames and other components; while handmade and older machine-made glass was highly recyclable, this is not the case for modern glass that has been laminated or given coatings or other treatments.14 Insulated glazing units (IGUs) are particularly problematic, being composed of high-energy materials and having short life spans—much shorter than the buildings.15 It is straightforward to demonstrate that, taking life span into account, triple-glazed units can never save the energy expended in their manufacture.16 In terms of through-life costing, it is important to take into account not only the embodied energy of the assemblies but also how easy—or difficult—they are to repair and maintain. In general terms, it should be considered poor practice to replace older materials or assemblies that could still be retained in functional form with shorter-lived and less flexible alternatives. If it is logical to see the approach to improving historic windows as being a return to traditional methods of management, what are the options for highly glazed enclosures that never incorporated such features but that may now have historical importance? John Straube writes eloquently of the thermal and daylighting problems induced by using glass over too great an area of the facade.17 Are there any effective remedial solutions that combine a respect for the original design with better building performance? Currently, the most common approach is to replace single glazing with IGUs, but unless the glass used is coated, this has little effect on solar gain and glare. It is also at the expense of life span; Mic Patterson has noted that recladding facades absorbs the equivalent of many years of operational energy expenditure, which may never be recouped.18 Potentially more flexible and interesting solutions are being introduced by conservation architects such as Wessel de Jonge, who uses secondary layers of glazing set far back into the building, which serves additional purposes such as providing corridor space and ventilation ducting.9 For less significant buildings, it is sensible to consider improvements that do visibly impact on the exterior, such as awnings or water-management systems. This is a lesson that can be learned from building history; vernacular building types were so successful because they incorporated lessons learned into the next generation of construction. One vital piece of the puzzle is that great gains that can be made by allowing occupants immediate control of their environment and by encouraging a better understanding (on the part of both occupants and engineers) of the real reasons for comfort and discomfort above and beyond air temperature (fig. 7).1,19,20 This “wholebuilding” approach has the potential to not only decrease energy use but at the same time to greatly increase occupant satisfaction.21 Patterson has recently predicted that, “[We] will see a steady progressive shift to human experience and comfort as the dominant performance criteria.”22 We certainly will need to do so if we are to cut energy use and increase building functionality.
77
78
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 7 Researchers at Berkeley have shown that controlling the temperature of the air is not as beneficial to comfort as is often assumed: by adding radiant panels for heating or fans for cooling, and crucially giving them control over the operation of these, the band of comfort could be extended from 21-24 C to at least 16-30 C.23
Case Studies of Sensitive Glazing Refurbishments The following two case studies make use of a standard conservation methodology and process that has long been applied to interventions on heritage buildings (including those involving modern architecture). This approach respects the embodied energy of original assemblies, prioritizing the repair and maintenance of any assemblies and materials that are still functional. It involves five steps: 1. Understanding the building and place, and determining what is of historic and other significance 2. Gathering together information about physical condition and about external requirements (such as codes, client requirements, and feasible user conditions) 3. Developing guidelines for intervention to retain or enhance the building’s significance 4. Applying these guidelines to develop a staged design response 5. Implementation of the staged design response CASE STUDY 1: TEAK WINDOW ASSEMBLES AT THE SALK INSTITUTE FOR BIOLOGICAL STUDIES
Situated on a Southern California bluff in La Jolla, overlooking the Pacific Ocean, the 1965 Salk Institute for Biological Studies is one of Louis Kahn’s finest works. Among the major architectural elements of the complex are 200-plus prefabricated teak window-wall assemblies, all set within openings in the concrete walls of the studies and offices that flank the Institute’s plaza. Kahn used prefabrication to both reduce cost and increase quality. The assemblies were largely crafted in a local cabinetmaker’s shop before being transported to
NORMANDIN AND PENDER, DOI: 10.1520/STP161720180075
FIG. 8 Salk Institute for Biological Studies, view of study towers.
the site and lifted into place by crane. They consisted of various combinations of horizontal sliding window sashes, louvers, and paneled shutters, with internal pockets to receive sliding-sash components. Kahn believed that introducing these elements would allow the occupants to access the natural elements of light and ocean air of the surrounding environment, responding to their own needs (fig. 8). Structural members were teak, with stud framing of softwood, and the units were sheathed on the interior with asbestos-cement “transite” board, similar in structural performance to sheathing (fig. 9). Exterior faces were clad in vertical tongue-and-groove teak siding, and interior faces (in oak paneling or gypsum board) often incorporated office shelving. Narrow glazed lights on either edge of the units allowed for additional light to wash the concrete walls of the interior and more practically served as shim space when the assemblies were set into the wall. After 50 years in an exposed marine environment, the assemblies had deteriorated and weathered and were in need of repair. Surface erosion and degradation varied across the different elevations. Shortly after the original construction, a black fungal biofilm had begun to develop on the surface of the teak. This was found to have spread from nearby eucalyptus trees; and it had resulted in a varied weathered appearance across the site depending on exposure. Although the Institute originally feared that total replacement of the window walls might be necessary, it was concerned that this would cause the loss of a significant amount of the building’s original fabric as well as of the occupants’ access to light and air. Replacement would not only have led to a loss of historic integrity, there were concerns about the viability and the ethics of any replacement materials because Southeast Asian teak is now a rare natural resource.
79
80
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 9 View of weathered teak window assemblies at south study towers.
In 2013, therefore, the Institute began a four-year conservation-based program to address the deterioration, with the aim of developing a long-term repair strategy that would allow the maximum retention of the original window assemblies. The resulting Teak Window Wall Conservation Project is an excellent demonstration of how a conservation-based approach can be applied to address the care of assemblies in a comprehensive, long-term manner.* Because the La Jolla site is of high value both architecturally and aesthetically, it was vital to maintain authenticity as well as material integrity—particularly in the most visible areas. In response to the varying material conditions, and concerns over weathering and environmental exposure, the program grouped interventions into three levels based on need: • Minor interventions—These included in situ cleaning and repair of assemblies that showed minor to moderate erosion of the teak cladding and that did not have signs of termite damage. • Moderate interventions—These included in-kind replacement of eroded tongue-and-groove teak siding of assemblies that also had minor deterioration of the internal framing. • Major interventions—These included in-kind replacement of any tongueand-groove teak siding (fig. 10) or other components that exhibited severe erosion as well as structural repair of major frame deterioration resulting from severe termite damage.
* The Salk Institute partnered with the Getty Conservation Institute and Wiss, Janney, Elstner Associates, Inc. to carry out investigation of the teak window assemblies and to develop and assess protocols for their conservation.
NORMANDIN AND PENDER, DOI: 10.1520/STP161720180075
FIG. 10 Weathered teak tongue-and-grove cladding with underlying termite decay.
For moderate to major interventions, improvements were introduced to reduce air and moisture infiltration. This enhanced the overall long-term performance, as well as the sustainability of the assemblies. Wherever possible, the transite panels were encapsulated in place. Where disturbance was unavoidable, they were removed and replaced with compatible materials. All three levels of intervention incorporated cleaning and removal of past surface coatings, application of new surface treatments to retard the growth of biofilms, and better integration of existing and replacement teak, as well as preventive treatment to control termites. In areas where the teak was considerably deteriorated, it was selectively replaced. Both reclaimed teak and new First European Quality (FEQ) teak from Southeast Asia were considered as replacement materials; FEQ was finally chosen on the basis of sourcing requirements and consistent quality. The teak was incorporated into a panel system designed to allow access to the window wall cavities to allow water leaks to be dealt with easily should they arise in the future. Finally, to maintain visual integrity and uniformity, differences in appearance among the three intervention levels were minimized by surface treatments. The overarching challenge of the project was balancing visual integrity with future care, especially the requirements for occasional material renewal or replacement. A unique treatment protocol was developed for each window location. This allowed the physical condition of each assembly to be precisely understood and addressed, including considerations about how materials could be maintained and how each treatment used might weather over time.
81
82
STP 1617 On Building Science and the Physics of Building Enclosure Performance
CASE STUDY 2: THE HISTORIC GLAZED FACADE OF THE GIBRALTAR SAVINGS AND LOAN ASSOCIATION BUILDING
The former offices of the Gibraltar Savings and Loan Association, located on Wilshire Boulevard in Beverly Hills, California, were designed in 1958 by Pereira & Luckman, Planning, Architecture and Engineering. This was one of their last commercial properties and is listed as a “local landmark” by the City of Beverly Hills. The building and its tower are both rectangular in plan and enclose some 140,000 ft2 of interior floor area. The original design specified the use of “tinted plate glass,” and the 0.25-in.-thick gray-tinted vision glass on floors two through ten is a significant and notable feature of the tower (fig. 11). The original curtain wall was constructed with extruded aluminum profiles glazed into the aluminum curtain-wall master frames. Horizontal and vertical frame profiles were mill finished, with dimensions measuring 2.5 in. in width and up to 4 in. in depth. The curtain-wall assembly of the upper stories of the tower intersperses rows of graytinted vision glass with rows of spandrel glass coated on the interior with an opaque finish.* The lower portion of the building is clad with a curtain wall that includes
FIG. 11 Gibraltar Savings and Loan office, circa 1960.
* In 1955, Pittsburgh Plate Glass introduced Spandrelite, the first ceramic-coated glass, and the Libbey, Owens, Ford Glass Company introduced Vitrolux spandrel glass, a color-fused, heat-strengthened polished plate glass.
NORMANDIN AND PENDER, DOI: 10.1520/STP161720180075
large sections of vision glass, together with stone panels. Vision glass along the ground floor storefront consists of clear plate and opaque-finish spandrel glass. The overall effect of the glass was to give a smooth, nonwavy appearance to the facade. Quite apart from this aesthetic function, however, the tinted float glass was intended to reduce solar gain, transmission of ultraviolet (UV) light, and glare. At the time of construction, glass manufacturers such as Pittsburgh Glass Products and Accessories and the Libbey, Owens, Ford Glass Company had been developing innovative tinted heat-absorbing glasses able to absorb infrared energy. This glass was available in gray and silver (products such as Solarex, and—for spandrel glass—Spandrelite and Vitrolux).24 Tinted float glass is now produced by adding metal oxides during production; this achieves the same performance by absorbing solar energy and is similar to the 1950s’ heat-absorbing glass. Over the years, there had been further additions of silver reflective solar control films to try to reduce solar gain, UV light, and glare. Vision glass had been cracked in some areas (fig. 12) and had been replaced “as needed” using a clear 0.25-in. tempered safety glass. However, their clear color, combined with roller-wave distortion, made these repairs very noticeable. Repairs to the spandrel glass had been carried
FIG. 12 View of crack in the vision glass.
83
84
STP 1617 On Building Science and the Physics of Building Enclosure Performance
out in a number of different shades; and investigation suggested that some replacement campaigns used a coating system that did not match the original. Cracking probably had been caused by the placement of aluminum shims in direct contact with the glass, creating stress points. The added UV films also may have been a contributing factor. Retrofit films of this type perform by both reflecting and absorbing energy from the sun’s rays; if they are added to some types of glass, the resulting absorption of heat energy can increase edge stresses, sometimes leading to glass breakage that would not have otherwise occurred. Glass cracking related to thermal stresses typically originates at the edges of the glass, with the cracks orientated perpendicular to the edge. At numerous locations throughout the building, other components of the building’s curtain wall, including gaskets and glazing tape, were in poor condition. Loose gaskets and failed sealants increased the risk of cracking of original glass. At the time of writing, it has not been possible to quantify this risk because it will depend on the extent to which the aluminum shims are found to be in contact with the glass. Despite these unknowns, it has been possible to develop a conservation repair program for the Gibraltar Savings and Loan building that not only retains the maximum materials of the original curtain wall system but enables it to be maintained with ease in the future. Similar to many types of listed heritage landmarks, a conservation-based approach can be applied here to address the care of assemblies in a comprehensive manner. As noted previously, this means using a standard conservation methodology that has long been applied to a variety of heritage sites, including maximum retention of original building materials from those of modern architecture and the recent past. In each of the case studies, a central part of the program is the addition of reinforcements for the original frame and of new gaskets and sealant detailing to improve performance. The new design detailing allows for replacement glass and even new frame components to be inserted easily should that prove necessary in the future. By taking a flexible approach to long-term maintenance and repair needs, not only has the life span of the existing materials been greatly prolonged, but the building’s use over a long period has been ensured—and all while retaining its historic integrity. In each case, the overall aesthetic and primary user functions of the building have been retained and preserved. Operability of the custom teak window walls and sashes at the Salk Institute for Biological Studies has been reinstated; the built-in louvers allow controlled access to air and light so critical to the unique coastal environment and setting. By comparison, given the new proposed use of the Gibraltar Savings and Loan building as a new hotel, the overall aesthetic and primary user functions will also be preserved. However, the curtain wall will not have operable components. The building will be rehabilitated for use as a new hotel, allowing for controlled interior climate where direct access to air and light is not required and is very limited.
NORMANDIN AND PENDER, DOI: 10.1520/STP161720180075
Conclusion Developing a good understanding of original materials and assemblies and of the causes of historic failures—and successes—is key to maximizing the life span of buildings while minimizing energy use, both for refurbishment and for day-to-day operation. Together with an understanding and respect for the historic significance, this approach has underpinned refurbishment programs for both the Salk Institute and the Gibraltar Savings and Loan Association building. It has not only reduced the costs of refurbishment—especially in terms of environmental impact—but also delivered buildings capable of sustaining useful lives into the long-term future, maximizing the benefit of the energy expended. Key has been the explicit incorporation of increased repairability and maintainability into the treatment of architectural glazing. This points the way toward a more sustainable future for buildings of all dates and types, and this point is as important for a new build as it is for refurbishment. For building performance specialists, it is clear that there must be many subtle lessons that can be learned by studying the past with greater attention. Now the climate emergency is making better and more efficient construction a high priority, solutions developed in the past—when energy was hard to tap, and cost a great deal1—should be actively analyzed by building scientists and tapped by building architects and engineers. In particular, the role of the occupants deserves greater focus. Much is currently being written about the opportunities provided by dynamic facades, which allow users to control their local environments. We would argue that this does not require high technology; opening windows, awnings, sliding sashes, and shutters are all intelligent and flexible solutions to problems that affect us today just as much as they did our forebears. They are available to be exploited once again and, indeed, improved upon. It is time to return the “people” element back to the heart of our understanding of building performance, where it sat so effectively for many millennia.
References 1.
R. Pender, B. Ridout, and T. Curteis, eds., Practical Building Conservation: Building Environment (Farnham, Surrey, UK: Ashgate/Routledge, 2014).
2.
S. Khan, Learning from History: Traditional Low-Energy Approaches to Comfort (London, UK: Historic England, forthcoming). R. Pender and S. Godfraind, eds., Practical Building Conservation: Glass & Glazing (Farnham, Surrey, UK: Ashgate/Routledge, 2011). C. Wood, B. Bordass, and P. Baker, “Research into the Thermal Performance of Traditional Windows: Timber Sash Windows,” Historic England Research Report 109, 2009, https://web.archive.org/web/20200512152222/https://research.historicengland.org.uk/ Report.aspx?i=16035 P. Baker, “Improving the Thermal Performance of Traditional Windows: Metal-Framed Windows,” Historic England Research Report 15, 2017, http://web.archive.org/web/ 20200428104751/https://research.historicengland.org.uk/Report.aspx?i=15568
3. 4.
5.
85
86
STP 1617 On Building Science and the Physics of Building Enclosure Performance
6.
7.
8.
T. Curteis and L. Seliger, “Conserving Stained Glass Using Environmental Protective Glazing,” Historic England Research Report, 2018, http://web.archive.org/web/ 20191210151538/https://research.historicengland.org.uk/Report.aspx?i=15635 H. Schoenefeldt, “The Role of Architectural and Environmental Experimentation in the Design of the Crystal Palace” (paper presentation, History Study Group, Institution of Structural Engineers, June 6, 2011), http://web.archive.org//web/20200428105305/ https://www.researchgate.net/publication/231892912_The_Crystal_Palace_environmentally_ considered Anonymous, “The Great Exhibition,” The Times, June 30, 1851.
9.
W. De Jonge, “Sustainable Renewal of the Everyday Modern,” Journal of Architectural Conservation 23, nos. 1–2 (2017): 62–105.
10.
E. Hroncˇekova ´, “Extinct Building Features—The Future of Historic Buildings? The Implications of Reintroduction of Historic External Blinds on Conservation and Visitor Experience at the Wallace Collection” (master’s thesis, Institute for Sustainable Heritage, University College London, 2016). C. Ganem and H. Coch, “Solar and Heat Protection Techniques: Evaluation and Design Recommendations for Different Types of Fabric Awnings,” in Proceedings of PALENC 2010 Passive and Low Energy Cooling for the Built Environment (Rhodes Island, Greece: PALENC, 2010), 1–9, http://web.archive.org/web/20191210155220/https://upcommons. upc.edu/handle/2117/11306 C. Kohler, Y. Shukla, and R. Rawal, Calculating the Effect of External Shading on the Solar Gain Coefficient of Windows, Energy Technologies Area LBNL-2001057 (San Francisco, CA: Lawrence Berkeley National Laboratory, 2017). P. Loussos, T. Konstantinou, A. van den Dobbelsteen, and R. Bokel, “Integrating Life Cycle Energy into the Design of Facade Refurbishment for a Post-War Residential Building in the Netherlands,” Buildings 5, no. 2 (2015): 622–649, http://web.archive.org/web/ 20191211075918/https://www.mdpi.com/2075-5309/5/2/622 M. Patterson and J. Vaglio, “Facade Retrofits: The Dilemma of the Highly Glazed High-Rise Fac¸ade,” in 2011 Building Enclosure Sustainability Symposium: Integrating Design & Building Practices (Pomona, CA: Cal Poly, 2011), http://web.archive.org/ web/20191210155533/http://www.enclos.com/assets/docs/Insight02-Chapter01-Facade_ Retrofits.pdf M. Patterson, J. Vaglio, and D. Noble, “Incremental Facade Retrofits: Curtainwall Technology as a Strategy to Step Existing Buildings toward Zero Net Energy” ISES Solar World Congress, Energy Procedia 57 (2014): 3150–3159. C. Jones and M. Fulford, “Choosing Low-Carbon Windows,” Building Magazine, September 2013, http://web.archive.org/web/20191211080316/https://www.building.co.uk/focus/ choosing-low-carbon-windows/5060079.article J. Straube, Insight BSD-006: Can Highly Glazed Building Fac¸ades Be Green? (Westford MA: Building Science Corporation, 2008), http://web.archive.org/web/20191210155908/ https://www.buildingscience.com/documents/insights/bsi-006-can-fully-glazed-curtainwalls-be-green M. Patterson, B. Silverman, J. Caspar, and K. Kensek, “Curtainwall Lifestyles: Evaluating Durability and Embodied Energy,” in CTBUH 2014 Shanghai Conference Proceedings (Chicago, IL: Council on Tall Buildings and Urban Habitat, 2014), 747–753. T. Hoyt, K.-H. Lee, H. Zhang, E. Arens, and T. Webster, “Energy Savings from Extended Air Temperature Setpoints and Reductions in Room Air Mixing” (paper presentation, International Conference on Environmental Ergonomics, Boston, MA, August 2–7, 2011).
11.
12.
13.
14.
15.
16.
17.
18.
19.
NORMANDIN AND PENDER, DOI: 10.1520/STP161720180075
20.
21.
22.
23. 24.
Y. Zhai, H. Zhang, Y. Zhang, W. Pasut, E. Arens, and Q. Meng, “Comfort under Personally Controlled Air Movement in Warm and Humid Environments, Building and Environment 65 (2013): 109–117. A. Martinez, M. Patterson, A. Carlson, and N. Noble, “Fundamentals in Facade Retrofit Practice,” International Conference on Sustainable Design, Engineering and Construction, Procedia Engineering 118 (2015): 934–941. M. Patterson, “Facade Futures 2018: Considering Long-Term Trends in Curtainwall Technology,” 2018, http://web.archive.org/web/20190614123017/https://facadetectonics.org/ news-articles/facade-futures-2018/ H. Zheng, E. Arens, and W. Pasut, “Air Temperature Thresholds for Indoor Comfort and Perceived Air Quality,” Building Research and Information 36, no. 2 (2011): 134–144. R. McKinley, “Spandrel Glass,” in U.S. National Park Service: Twentieth Century Building Materials: History and Conservation, ed. T. Jester (Washington, DC: U.S. Department of the Interior, 1995).
87
88
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180090
Niklas W. Vigener1 and Anthony J. Nicastro1
Predicted Thermal and Hygrothermal Improvements for a Mid-Twentieth-Century Brutalist Landmark Citation N. W. Vigener and A. J. Nicastro, “Predicted Thermal and Hygrothermal Improvements for a Mid-Twentieth-Century Brutalist Landmark,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 88–106. http://doi.org/10.1520/STP1617201800902
ABSTRACT
The Hirshhorn Museum, opened in 1974, is a landmark of Brutalist architecture and serves as the Smithsonian Institution’s museum of contemporary and modern art. In its time, the building, a squat concrete donut, nearly windowless on the exterior and elevated over a courtyard by four massive piers, was a controversial addition to a landscape of stately and monumental “traditional” museums. After more than forty years in service, the building is experiencing performance issues that are part and parcel of its 1970s’ technology: Its energy performance is poor by contemporary standards, and the mechanical humidification required to preserve its collection causes wintertime condensation on building enclosure components. While these issues can be ameliorated or resolved with contemporary technology, the Hirshhorn’s peculiar geometry and immutable architecture place limits on the addition of contemporary materials and equipment. Potential trade-offs include appearance changes, disruption of the collection, reduction in usable space and, of course, cost. The Smithsonian Institution is dedicated to preserving this architectural landmark and is assessing options for performance upgrades to the building but requires quantitative as well as qualitative information to support its decision-making. Using the building science–based options analysis for the
Manuscript received October 23, 2018; accepted for publication August 16, 2019. 1 Simpson Gumpertz & Heger, 1625 Eye St. NW, Suite 900, Washington, DC 20006, USA 2 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
VIGENER AND NICASTRO, DOI: 10.1520/STP161720180090
Hirshhorn’s enclosure as an example, the paper will work through a technical analysis and decision-making process, including: identification of required enclosure repairs to address performance issues; comparison of building science modeling tools to determine the cause of condensation and its extent, and correlation with field observations; analysis of building enclosure alterations to improve energy performance; establishment of minimum improvements to resolve wintertime condensation; and qualitative assessment of air and thermal barrier improvements and their potential to reduce energy consumption. Keywords mechanical humidification, historic museum, thermal analysis, hygrothermal analysis, air barrier, thermal barrier, energy efficiency, rehabilitation, Brutalism
Background The Hirshhorn Museum actually consists of two buildings. An approximately 100 ksf cast-in-place concrete below-grade box houses limited exhibit spaces, a gift shop, and an auditorium, along with art storage, offices, mechanical rooms, a loading dock, and other back-of-house spaces. This box supports the prominent abovegrade portion, an approximately 80 ksf, three-story hollow concrete cylinder that is elevated over a pedestrian plaza on four concrete piers. The cylinder’s exterior facade consists of precast panels outboard of a 2.75-ft-thick cast-in-place concrete backup wall that is cast integral with exposed concrete spandrels at the top and bottom of the building. Its only opening is a third-floor balcony, with floor-to-ceiling glazing that affords a view of the National Mall, on the north elevation. The cylinder’s interior facade consists of continuous floor-to-ceiling fenestration between projecting precast spandrels. The only first- or ground-floor space is a glassenclosed entrance lobby that connects above- and belowground spaces and fills the space between two of the support piers on the building’s south elevation. The lowslope roof of the above-grade portion is covered with an exposed built-up roofing membrane. When it opened in 1974, the Hirshhorn, which was designed by Skidmore Owings Merrill’s Gordon Bunshaft, was controversial largely because of its stark and unconventional exterior appearance in a landscape of more traditional museum buildings* (fig. 1). Critics agreed, however, that the building worked well as a museum—its three floors of circular galleries were deemed well suited to display and appreciate visual art in the galleries along the windowless exterior walls and sculpture on the naturally lit courtyard side. Over time, the museum became a firmly established fixture on the National Mall, making preservation of its appearance imperative.
* See, for example, “A Fortress of a Building that Works as a Museum” by Paul Goldberger, New York Times, October 2, 1974.
89
90
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Building Enclosure Components and Operating Conditions The above-grade portion of the Hirshhorn, which is the subject of this paper, includes the following major enclosure assemblies (components are listed from exterior to interior): • Roof: Four-ply built-up asphalt roofing membrane, 1-in. perlite cover board, 3.5-in. (average) sloped polyisocyanurate insulation, two-ply built-up asphalt vapor barrier, 5-in. reinforced concrete roof deck with deeper coffer ribs. • Exterior cylinder walls (fig. 2): Four-inch (average) precast concrete panels, 2.25-in. (average) air space, 2.75-ft reinforced concrete backup wall, uninsulated 3.5-in. metal framed cavity, 0.75-in. gypsum plaster with metal lath (art hanging wall); projecting reinforced concrete ring beams, integral with floor slabs and roof slab, at floor levels. • Cylinder walls at courtyard (fig. 3): One-inch-thick insulating glazing with continuous structural neoprene gaskets supported by extruded aluminum curtainwall framing; projecting reinforced concrete ring beams, integral with floor slabs and roof slab, at floor levels. • Elevated second-floor soffit (fig. 4): Five-inch reinforced concrete floor slab with deeper coffer ribs, 1-in. cellular glass rigid insulation, polyethylene sheet, 3-in. concrete topping slab with fluid-applied flooring system.
FIG. 1 The Hirshhorn Museum and Sculpture Garden under construction, circa 1972. (Image courtesy of the Smithsonian Institution.)
VIGENER AND NICASTRO, DOI: 10.1520/STP161720180090
FIG. 2 View of precast panel removed from exterior cylinder walls. The precast panels are mounted to the concrete backup wall with steel straps and relieving angles. The wall assembly lacks a dedicated air/water-resistive barrier or insulation.
FIG. 3 Interior cylinder walls clad with curtain wall set between precast panels covering projecting concrete slab edges.
•
Balcony: Two-inch concrete pavers on pedestals, fluid-applied waterproofing, flat cast-in-place concrete balcony slab and coffer beams integral with backup wall, separated from concrete floor slab and coffer beam by a 1/2 in. movement joint.
91
92
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 4 Elevated second-floor soffit and gallery topping slab over insulated structural slab.
Courtyard cornices: Flat, precast concrete pavers sloped to internal drains, sloped mortar setting bed, several layers of fluid-applied and sheet waterproofing, flat cast-in-place concrete perimeter beam, separated from concrete floor slabs and coffer beams by a 0.75-in. movement joint. The Hirshhorn maintains constant interior conditions that include mechanical humidification in order to comply with collections’ storage requirements. Actual conditions in the building run at 22 C (71 F) 6 1.5 C (2 F) and 48% RH 6 4% (approximate dew point 12 C [53 F]) year round.* Washington, DC, is in the American Society of Heating, Refrigerating, and Air Conditioning Engineering (ASHRAE) Climate Zone 4A and has a 99.6% heating design temperature of 8.2 C (17.3 F).{ •
*
Smithsonian Institution data logger data collected between June 2013 and May 2015. Value based on Ronald Reagan National Airport.1
{
VIGENER AND NICASTRO, DOI: 10.1520/STP161720180090
Building Enclosure Performance Issues The building enclosure design technology employed during the design of the Hirshhorn is consistent with mid-twentieth-century buildings. The exterior walls and the plaza assembly over the below-grade portion lack insulation. The building lacks a dedicated air and water-resistive barrier. The building is mechanically humidified, but preservation and nonpreservation spaces are not internally separated, which results in unnecessary energy demand for humidification. Partly as a result of its geometry, which maximizes exposed exterior surface area, partly because of its enclosure detailing and materials, and partly because of its mechanical system and operating parameters, the Hirshhorn is, on a per-square-foot basis, one of the largest energy consumers in the Smithsonian Institution’s (SI) portfolio.* In addition to high energy consumption, the Hirshhorn exhibits building enclosure performance problems that threaten artifacts or diminish the visitor experience. These include: • Water leakage that wets the floors and plaster art-hanging walls in limited locations on the exterior perimeter (fig. 5) • Widespread wintertime condensation that wets the floors and art-hanging walls on the exterior perimeter • Widespread wintertime condensation on glazing (fig. 6) • Moisture exiting the underside of the second-floor coffered ceiling
FIG. 5 Damage to plaster art wall caused by water leakage.
* International Association of Museum Facility Administrators (IAMFA) 2013 energy benchmarking data. At approximately 280 to 350 kBtu/ft2, the Hirshhorn has the fourth-largest energy use intensity among fortyeight national and international museums.
93
94
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Ineffective building pressurization control Spontaneous fracture of courtyard glazing due to nickel sulfide inclusions (not discussed in this paper) • Lack of ultraviolet (UV) and lighting control (not discussed in this paper) SI recently embarked on a project to study the causes of enclosure performance issues, develop rehabilitation options, and implement long-term enclosure repairs. • •
Building Enclosure Investigation and Analysis INVESTIGATION OF WATER LEAKAGE
We performed a water leakage investigation in general accordance with ASTM E2128-17, Standard Guide for Evaluating Water Leakage of Building Walls.2 Details of this investigation are beyond the scope of this paper; however, relevant findings include the following: • Water leakage into interior exhibit spaces is caused by failed membrane waterproofing at movement joints between the balcony slab and the backup wall and at movement joints between the courtyard cornice perimeter beams and the floor slabs. • A portion of this leakage enters below the second-floor concrete floor slab, spreads out on top of the second-floor elevated concrete soffit slab, exits through shrinkage cracks in the soffit slab, and wets interior finishes in an area directly below the balcony. • Water does not leak through the exterior walls or through the glazing.
FIG. 6 Interior surface condensation on lobby glazing, February 2015. Note uneven distribution of heat from fin tube radiators along the sill.
VIGENER AND NICASTRO, DOI: 10.1520/STP161720180090
SI recently performed limited waterproofing repairs at the balcony that have stopped the leakage. FIELD INVESTIGATION OF INTERIOR CONDENSATION
We visited the building on several cold days to observe interior condensation. The glazing in the exhibit spaces is typically covered with manually operated perforated blinds. All glazing includes fin tube radiant heating along the sill, but the heat distribution is uneven from window to window (fig. 6). In February 2015, SI observed widespread condensation on interior glazing surfaces, but not on interior aluminum framing surfaces, at reported exterior temperatures of approximately -6 C (21 F) (fig. 6). Also in February 2015, SI observed condensation on interior surfaces of the concrete backup wall after several days of sustained cold conditions when the daily maximum temperature did not exceed -4 C (25 F) (fig. 7). We measured interior glazing surface temperatures below the interior air dew point temperature (fig. 8). This demonstrates the vulnerability of the glazing assemblies to condensation at exterior temperatures that are above the ASHRAE Fundamentals1 winter design condition for heating (i.e., -8 C [17.3 F]). ANALYTICAL ASSESSMENT OF INTERIOR CONDENSATION
We performed hygrothermal and thermal modeling of the major exterior enclosure assemblies to complement the field investigation and understand the formation of condensation on exterior assemblies. Modeled assemblies included: • Second-floor soffit • Exterior walls
FIG. 7 Interior face of the concrete backup wall evenly covered with water droplets during the winter of 2015. Presence of water visually confirmed with waterindicating test paper.
95
96
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 8 Infrared image and visible light reference image of glazing assembly indicating large surface temperature differences between the center of the glazing and the sill. The cold area near the center of the glazing is an anomaly related to the failed insulating glazing seal. Note that obtaining accurate surface temperature readings from reflective surfaces is notoriously difficult, so the image serves best to illustrate the large temperature differences between glazing areas.
Roof Glazing We used the WUFI* computer program to analyze the hygrothermal performance of the opaque assemblies to determine the risk of moisture accumulation and condensation associated with moisture migration through each assembly. The risk of condensation is affected by the thermal performance of an assembly and the effects of two-dimensional thermal bridging elements, such as concrete floor slabs or light-gauge metal framing, which are not accounted for in the onedimensional hygrothermal model. To address the risk of condensation associated with assembly component surface temperatures, we also ran thermal analyses of the four envelope assemblies and several system details using the THERM{ computer program. We based our model simulations on the most current version of the • •
*
WUFI (an abbreviation for “Wa¨rme und Feuchte Instationa¨r”) Pro Version 5.3 by the Fraunhofer Institute for Building Physics, Germany. This program derives time-dependent, one-dimensional distributions of temperature, relative humidity, and water content across a component assembly using historical weather data in addition to material properties and environmental conditions. WUFI is an appropriate tool to calculate the transient one-dimensional heat and moisture transport through the field (i.e., not including details or transitions) of the typical exposed second-floor structure, exterior wall, and roof assemblies. WUFI includes the effects of wetting and drying in component materials of the modeled assembly, solar radiation, and winddriven rain exposure. { THERM Version 7.4 by the Lawrence Berkeley National Laboratory. This program derives a two-dimensional temperature distribution across a component assembly based on steady-state heat flow calculations and constant interior/exterior environmental conditions. Our analysis was intended to estimate surface temperatures and compare them to the interior dew point to determine the probability of condensation on interior surfaces.
VIGENER AND NICASTRO, DOI: 10.1520/STP161720180090
National Fenestration Rating Council (NFRC) Simulation Manual3 as well as ASHRAE 90.1-2013, Energy Standard for Buildings Except Low-Rise Residential Buildings. The thermal analysis tool does not model the effects of airflow or water vapor diffusion, both of which may affect condensation risk. BOUNDARY CONDITIONS
To develop the interior boundary conditions for our models, we reviewed temperature and relative humidity (RH) readings gathered by SI from data loggers from seventeen locations throughout the building. Based on the logger data gathered between 2013 and 2015, we established interior conditions at 22 C (72 F) and 51% RH, conditions that correspond to a dew point temperature of 12 C (53 F). For exterior boundary conditions, we used hourly annual weather data from the WUFI database and modeled conditions on the north (likely coldest and wettest) elevation using the WUFI cold-year weather file and the south (warmest and highest solar exposure) using the WUFI warm-year weather file. When analyzing for condensation risk, we used the 99.6% winter design condition heating temperature, which includes the effects of a 15 mph wind. Note that the heating design temperature represents a temperature that, on average, will be exceeded for all but 0.4% of a typical year. This equates to approximately 35 h per year and is especially relevant to the results described here and corroboration of field observations. RESULTS FOR ROOFS, GLAZING, AND SECOND-FLOOR STRUCTURE
Modeling using either WUFI or THERM produced results that informed our understanding of enclosure performance for the second-floor soffit, roof, and glazing. The models showed that, consistent with SI’s and our observations, and at reasonably expectable cold temperatures, condensation does not occur on the underside of the roof slab or the top of the second-floor slab, both of which incorporate some insulation. The results also showed that, at sustained colder temperatures, condensation will occur on the edge of curtain wall glazing when the exterior temperature is less than 6 C (21 F) and, to a greater degree, at the ASHRAE 99.6% heating design temperature ( 8 C [17.3 F]). This two-dimensional, steadystate analysis is appropriate and correlated well with our observations of actual condensation for the glazing assemblies, which have very limited thermal mass and therefore equilibrate quickly with the surrounding ambient conditions. UNDERSTANDING EXTERIOR WALL OBSERVATIONS USING A COMBINATION OF MODELING TOOLS
Accurate representation of the condensation phenomenon on the exterior walls proved difficult to capture with a single modeling tool. For the exterior walls, a onedimensional hygrothermal software program was insufficient to capture the effects of the thermally massive concrete structure across two dimensions, while THERM could not account for transient moisture migration through the exterior walls. These effects skewed results in individual software tools. Therefore, to accurately
97
98
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 9 One-dimensional exterior wall section modeled in WUFI.
represent the potential formation of condensation and to reflect field observations, we used a combination of hygrothermal and thermal modeling to capture the effect of both transient moisture migration and two-dimensional thermal effects of the massive concrete structure, respectively. We modeled a typical exterior wall section including precast panels over an air gap and thick concrete backup wall (fig. 9). Based on field observations and original construction drawings, we modeled the existing exterior wall assembly as shown in figure 9. Hygrothermal simulation of the exterior walls using WUFI showed that the north elevation of the wall assembly would experience a relative humidity above 80% (maximum of approximately 84%) inboard of the dampproofing at the concrete backup wall (fig. 10 and fig. 11). However, condensation would only be predicted by the model if one of the following conditions occurred: • The exterior temperature decreases below the lowest temperature in the TMY3* weather data file used in our WUFI model ( 17 C [1 F]) or deviates from the TMY3 and is colder for an extended period of time. • Exterior air infiltrates and circulates behind the precast concrete panel at a higher rate than the five air changes per hour (ACH) that we modeled. • The interior RH of the museum increases to approximately 58% or higher, which may happen periodically based on the data logger results. Results from the THERM analysis (fig. 12) show that the dew point isotherm, associated with the defined interior temperature conditions, extends inboard of the concrete backup wall at the existing exterior wall assembly. The area inboard
*
Typical meteorological year, Version 3.
VIGENER AND NICASTRO, DOI: 10.1520/STP161720180090
FIG. 10 WUFI output film (thirty-year simulation) showing temperature, relative humidity (RH), and moisture content in the exterior wall assembly. Note that the light green area, which shows the sweep of the RH curve, does not indicate condensation on the interior of the backup wall, which is inconsistent with our observations.
FIG. 11 Temperature and RH plots from the last five years of the thirty-year WUFI simulation at the interior face of the concrete drum show that RH (in green) peaks yearly at approximately 84% RH.
of the dew point isotherm on the left portion of figure 12 indicates locations of predicted condensation. Note that the steady-state THERM simulation cannot include transient effects of heat flow. In other words, this model cannot account
99
100
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 12 Results of THERM analysis of the exterior wall at a floor slab. Colored portions of illustration on the right indicate areas below the dew point under heating design temperatures.
for the thermal mass of the concrete backup wall, which causes considerable time lag of the response of the interior assembly components to exterior temperature changes. The results show that the surface temperatures of the concrete backup wall fall below the dew point temperature at sustained exterior temperatures below 4 C (24 F). Because the thick backup wall represents a significant thermal mass, and it is not directly exposed to either the interior or the exterior environment, it does not respond quickly to temperature changes. The exterior temperature would have to remain at or below the exterior design temperature ( 8 C [17.3 F]) for an extended period of time (likely several days) for the interior surface of the concrete backup wall to drop to the minimum concrete surface temperature (11 C [51 F]) derived by the THERM analysis. This is consistent with field observations of condensation on a 6 C (21 F) day following several days of sustained cold temperatures.
VIGENER AND NICASTRO, DOI: 10.1520/STP161720180090
FIELD INVESTIGATION OF AIR LEAKAGE
The Hirshhorn is an occupied museum and conducting a whole-building air leakage test is not advisable because the large volume of unconditioned and unfiltered exterior air that is required to run the test represents an unacceptable risk to its collection. In the absence of such test data, we qualitatively assessed air leakage using the following approaches: • Review of Infrared (IR) Thermography: IR images made available by SI indicate likely air exfiltration paths (indicated by elevated surface temperatures during a nighttime IR scan) around the perimeter of the balcony glazing and the perimeter of the courtyard glazing (fig. 13). • Qualitative Air Leakage Testing: We performed qualitative air leakage testing of the fenestration and precast cladding in general accordance with ASTM E1186-17, Standard Practices for Air Leakage Site Detection in Building Envelopes and Air Barrier Systems,4 except that we were not able to increase building pressurization during the test (building pressure is generally neutral or moderately positive). These tests yielded inconclusive results. • Visual Assessment of Existing Construction: Where practical, we performed visual assessments to identify obvious air leakage paths across the building envelope. The concrete backup wall on the exterior of the drum is continuous and the fixed windows facing the courtyard have continuous glazing gaskets, so there are no obvious air leakage paths. We had the opportunity to observe the fenestration tie-in at the balcony perimeter where prior IR imaging detected a likely air exfiltration source (fig. 13). Upon removal of precast concrete panels adjacent to the balcony fenestration, we identified
FIG. 13 IR image and corresponding digital image (for context) during February 2015 showing a thermal anomaly (crosshairs) at the east end of the balcony that appears warmer than the adjacent exterior cladding. The anomaly cools gradually with distance from the balcony corner. The anomaly is likely due to warm air exiting the building. Temperature scale is approximate.
101
102
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 14 View of the balcony curtain wall jamb. The jamb lacks an air barrier but was partially covered with batt insulation (which is not an effective air barrier material and was removed prior to taking the photo).
that the fenestration lacked an airtight transition with the adjacent concrete backup wall, allowing for unabated passage of air across the building enclosure (fig. 14). REQUIRED ENCLOSURE REPAIRS TO ADDRESS CONDENSATION
Project goals that are unrelated to the enclosure performance and energy consumption dictate the removal and reinstallation of the precast panels. Therefore, the enclosure rehabilitation can provide thermal insulation on the exterior of the concrete backup, where it will be continuous and most advantageous for the improvement of thermal performance. Adding continuous insulation on the exterior side of the concrete structure in the wall cavity will provide the most benefit to hygrothermal and thermal performance because it will draw the dew point outward and reduce thermal bridging. THERM analysis of proposed new, additional insulation indicates that the addition of a relatively small thickness of rigid insulation placed
VIGENER AND NICASTRO, DOI: 10.1520/STP161720180090
in the exterior wall cavity will be sufficient to elevate wintertime surface temperatures on the interior of the backup wall to prevent condensation. The available space between the panels and the backup wall allows installation of up 1.5 in. to 2 in. (average) of rigid insulation, which should be maximized to improve energy performance. We did not consider other workable approaches to prevent condensation, such as circulating warm supply air over the backup wall to raise its surface temperature or providing closed-cell spray polyurethane foam (ccSPF) insulation only on the interior. These approaches are less energy-efficient because they cannot resolve the thermal bridging at floor slabs. Preventing condensation on the glazing also requires elevating surface temperatures above the dew point. Since the glazing has little thermal mass and responds very quickly to environmental conditions, this requires both decreased thermal conductivity (reduced U-value) of the glazing and uniform heat delivery. Even highperformance glazing systems will experience condensation if beneficial heat delivery in the form of moving air from the interior is prevented (e.g., by curtains or other obstructions) and, conversely, thermally inefficient glazing systems will not condense if they are continually warmed by moving air. Consequently, energy-efficient prevention of condensation on the glazing requires provision of thermally improved glazing, thermally improved glazing perimeter details, and thermally broken frames. For the larger windows at the lobby and balcony glazing, the provision for moving warm air,* while not energy efficient, will be required to circulate conditioned air across the glazing interior to warm frame-glass junctures sufficiently to avoid condensation. We recommended providing triple glazing in thermally broken aluminum frames and improved mechanical heat delivery. ADDITIONAL BUILDING ENCLOSURE REPAIRS TO IMPROVE ENERGY PERFORMANCE
Enclosure-related opportunities for repairs that improve energy performance are limited by SI’s mandate to respect the immutable architecture of the Hirshhorn. Nonetheless, the principal contributors to building enclosure energy inefficiency, air leakage and lack of insulation, can be improved through remediation with relatively little impact on museum appearance. We considered and recommended the following: • Reduced air leakage: The IR images indicate significant air leakage at window perimeters. This is not surprising because, at the time of Hirshhorn’s design, the importance of air barriers as a critical tool to limit energy consumption and condensation risk was not well understood. The rehabilitation design, consistent with current building code requirements and installation *
THERM captures the effect of air movement across the glazing surface using a surface film coefficient, Hc with units of BTU/(h*ft2* F). The derivation of this surface coefficient requires specific analysis, such as computational fluid mechanics modeling, that accounts for heat sources and three-dimensional surface geometry. This discussion is beyond the scope of this paper.
103
104
STP 1617 On Building Science and the Physics of Building Enclosure Performance
•
practices, should include provisions for air barrier continuity between the replacement fenestration and the concrete backup wall, which is the de facto air barrier in the field of the exterior walls. Increased thermal insulation: The rehabilitation design can provide more efficient insulation in the roof assembly (thicker than existing, subject to water drainage and architectural considerations), in the elevated secondfloor soffit (more thermally efficient extruded polystyrene [XPS] insulation to replace the existing foam glass insulation but without additional thickness), and ccSPF insulation applied to the interior side of the backup.
ENCLOSURE PERFORMANCE ASSESSMENT—BASELINE AND ENERGY IMPROVEMENTS
Improving energy performance of the Hirshhorn is a priority for the Smithsonian Institution. Making improvements to the enclosure, particularly with respect to airtightening around fenestrations and adding insulation to opaque wall components, will also translate to operational cost savings. This section discusses analytical assessment of enclosure improvements. ANALYSIS OF U-FACTOR IMPROVEMENTS
In addition to using THERM for condensation analysis, we used it to calculate the average U-factor of the modeled building envelope assemblies and to allow a qualitative assessment of thermal efficiency compared to current codes and standards. Obviously, the museum’s building enclosure lacks the insulation to deliver energy performance in line with current standards. We studied several optional improvements to energy efficiency for each enclosure system where each improvement is benchmarked against current energy code (IECC5) standards or is maximized based on a feasible revitalization option. We then derived potential percentage U-factor improvements, sorted by major enclosure system, as summarized in Table 1.
Discussion and Summary The Hirshhorn Museum’s thermally inefficient (by today’s standards) enclosure design has caused wintertime condensation and is a contributor to poor energy performance. The condensation can be resolved through provision of exterior wall insulation, along with high-efficiency glazing and mechanical system improvements that deliver sufficient heat to the glazing to maintain surface temperatures above dew point conditions. These limited air barrier improvements and improved insulating value of exterior walls and fenestration also deliver a, likely modest, energy efficiency improvement. More significant enclosure-related energy efficiency improvements will likely be achieved through additional air-tightening of the building enclosure. The architecture of the building, which prioritizes exterior surface area and has thermal bridges that cannot be resolved with the addition of insulation, limits possible energy performance improvements.
VIGENER AND NICASTRO, DOI: 10.1520/STP161720180090
TABLE 1 Summary of average U-factors of major enclosure assemblies
Assembly
Average U-Factor (BTU/hr-ft2- F)
Approximate Performance Improvement
Comments
(E) Second-Floor Soffit
0.17
-
1-in. foam glass insulation
(N) Second-Floor Soffit
0.15
13%
1-in. XPS insulation
(E) Exterior Wall Assembly
0.16
-
(N) Exterior Wall Assembly
0.08
200%
(no interior insulation) (N) Exterior Wall Assembly
2-in. XPS insulation on exterior, no interior insulation
0.04
400%
(interior insulation)
2-in. XPS insulation on exterior, 3-in. interior ccSPF insulation
(E) Courtyard Curtain Wall
0.51
(vertical mullion) Vision Glazing (unit size: 4 ft,
-
Existing
1 3/16 in. by 11 ft 6 in.) Spandrel Glazing (unit size: 4
0.56
ft 1 3/16 in. by 1 ft 4 in.) (N) Courtyard Curtain Wall
0.38
and Spandrel Glazing
34%
Based on 2015 IECC maxi-
47%
mum prescriptive allowable U-factor for Climate Zone 4 for both glazing assemblies
(E) Roof
0.04
-
Existing
(N) Roof
0.03
33%
Using 2015 IECC requirement for R-30 insulation
Note: The average U-factor values for glazing assemblies are calculated using an area-weighted analysis per NFRC 1003 and are representative values based on a typical unit in which the glazing is captured on all four sides by a mullion with the same basic geometry. (E) = existing; (N) = proposed.
Future Work and Research Needs This study is based on limited data. The project team plans to perform, at a later date and if the museum will be closed and exhibits put in storage, whole building or selective air leakage testing to establish actual air leakage rates of the existing assemblies and compare them to future improvements. This data can then be used in a whole building energy simulation to predict likely energy use improvements. ACKNOWLEDGMENTS
The authors wish to acknowledge the enthusiastic support of the Smithsonian Institution and Simpson Gumpertz & Heger in the preparation of this study.
References 1.
American Society of Heating, Refrigerating and Air-Conditioning Engineers, ASHRAE Handbook—Fundamentals (Atlanta, GA: ASHRAE, 2017).
105
106
STP 1617 On Building Science and the Physics of Building Enclosure Performance
2.
3. 4.
5.
Standard Guide for Evaluating Water Leakage of Building Walls, ASTM E2128-17 (West Conshohocken, PA: ASTM International, approved November 1, 2017), http://doi.org/ 10.1520/E2128-17 National Fenestration Rating Council, NFRC Simulation Manual (Greenbelt, MD: NFRC, 2017). Standard Practices for Air Leakage Site Detection in Building Envelopes and Air Barrier Systems, ASTM E1186-17 (West Conshohocken, PA: ASTM International, approved July 15, 2017), http://doi.org/10.1520/E1186-17 International Code Council, International Energy Conservation Code (Washington, DC: ICC, 2015).
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180079
Ali Moussawi1 and Scott N. Bondi1
Integrating Computational Fluid Dynamics Simulations and Condensation Risk Assessments: Using Project Specific Analyses to Better Inform the Design of Buildings Citation A. Moussawi and S. N. Bondi, “Integrating Computational Fluid Dynamics Simulations and Condensation Risk Assessments: Using Project Specific Analyses to Better Inform the Design of Buildings,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 107–124. http:// doi.org/10.1520/STP1617201800792
ABSTRACT
The risk of condensation on the interior of building enclosure systems can be an important driver in the design and selection of both building enclosure and mechanical systems. Condensation risk is an especially important performance metric in sensitive spaces and humidified spaces, such as museums and healthcare facilities, where uncontrolled interior moisture has significant consequences. Traditional analyses of condensation risk use thermal modeling to predict the onset of condensation. Fenestration assemblies often have the lowest thermal performance, and thus the analyses of these systems are the most important for making informed decisions. Fenestration products typically are analyzed for their thermal performance through existing standards developed by the National Fenestration Rating Council (NFRC) and using tools such as finite element analysis (FEA) to estimate detailed performance of complex assemblies. These methods use fixed environmental conditions to enable a comparative analysis of Manuscript received October 20, 2018; accepted for publication May 29, 2019. 1 Simpson Gumpertz & Heger, 550 Seventh Ave., New York, NY 10018, USA A. M. https://orcid.org/00000002-5978-7990, S. N. B. https://orcid.org/0000-0001-9351-9693 2 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
107
108
STP 1617 On Building Science and the Physics of Building Enclosure Performance
different products and systems. However, these environmental conditions do not necessarily correlate to the air movements and localized temperatures near the fenestration for a specific project condition for which the mechanical systems were designed, and therefore may not accurately predict condensation risk. This leaves designers with an unclear understanding of how the actual space will perform. This paper will discuss the use of computational fluid dynamics (CFD) analyses to more closely estimate the conditions inside a space relative to the current standards. The paper will also describe how the results of CFD models can be used as inputs for the thermal analyses typically used to predict the onset of condensation on fenestration systems. A case study will be presented to describe the approach to modeling both the enclosure and mechanical systems. It will further discuss how these studies can be correlated to statistics for weather events to provide owners and designers with better information to make more informed design decisions. Keywords condensation, thermal analysis, fenestration, computational fluid dynamics, film coefficient
Introduction Condensation on the interior of building enclosure assemblies can be detrimental to the durability of the building and can create favorable conditions for microbial growth that may be harmful to occupants. In cold climate regions, and in humidified buildings, the risk of condensation in building enclosure assemblies is increased. The risks to building owners can range from inconvenience, such as moisture obscuring the vision area of a window, to significant impingement on the intended function of the building, such as increased potential for microbial growth in a hospital or the potential of condensate dripping on sensitive art exhibits in a museum. The frequency of condensation on building enclosure assemblies is typically an increased risk at thermal bridges through the building envelope. For fenestration systems, these bridges often occur in the metal framing elements supporting glazing. In most building enclosure systems, the interior surfaces of the fenestration framing are the coldest during heating climate conditions because they are the least-insulated components within the enclosure. Condensation occurs when the temperature of a surface is less than the dew point temperature of the air. For building enclosures, the risk of interior condensation occurs when the surface temperatures of the enclosure framing or glazing are below the dew point temperature of the space. For many buildings, persistent condensation is avoided by a combination of factors including introducing thermally isolated enclosure systems, keeping the interior relative humidity (RH) low, and directly heating the perimeter zone (the zone near the envelope). Thermally insulated systems, including thermally broken framing and insulating glazing, allow interior component surfaces to be isolated from the exterior weather conditions.
MOUSSAWI AND BONDI, DOI: 10.1520/STP161720180079
The reduction in RH is typically a passive phenomenon that occurs because the heating, ventilation, and air-conditioning (HVAC) systems in most buildings introduce significant quantities of outside air that is naturally dry when exterior temperatures are low. The introduction of heat at the building perimeter also aids in reducing the frequency and risk of condensation as it both locally adds heat directly to the interior surfaces of a facade system as well as forces more air movement near the facade of the building enclosure. In older enclosures that do not contain continuous air barriers, air infiltration may also play a role in reducing interior humidity during cold weather. The process for determining condensation risk can be evaluated using numerical analysis. Standards for fenestration systems such as those developed by the National Fenestration Rating Council (NFRC) have been established to allow consumers and designers to evaluate the performance of different systems relative to one another. These standards do not easily allow the designer to understand the condensation risk specific to a particular building. To enable a simple comparison between fenestration products, NFRC 500-20171 allows a designer to evaluate the Condensation Resistance Factor (CRF). This analysis process makes general assumptions on the conditions inside and outside of the product being evaluated that may not match the project’s specific conditions but makes direct comparisons among products possible. While this can be helpful for the designer, it ultimately may be difficult to relate the results of the analysis to what the owner can expect with their building. The NFRC 500 analysis process utilizes several analytical tools for comparative analysis. These include tools such as WINDOW and THERM developed by Lawrence Berkeley National Laboratory (LBNL). WINDOW is a calculation program that allows for the simulation of energy flow and surface temperatures in glass and insulating glass assemblies. THERM is a finite element analysis (FEA) program that has been optimized to simulate heat transfer through fenestration framing assemblies. It uses a two-dimensional heat transfer FEA solver that allows designers to evaluate the U-factor and local temperature profiles at detailed sections of the building enclosure. WINDOW and THERM have built-in libraries for interior and exterior boundary conditions for use in the analysis. These conditions are generally the same as those used in NFRC 500-2017 to evaluate the thermal performance of fenestration assemblies. The boundary conditions for these analyses include both temperatures, as well as air film coefficients that relate to localized air movement near the fenestration assembly. These criteria are rarely descriptive of project-specific conditions; however, the simulation portion of the analysis process is well suited for analyzing the condensation risk as it provides detailed information regarding temperatures through complex building fenestration assemblies. The authors propose that this analytical process could be improved upon to provide designers and owners with better information as to how their building enclosures will perform with respect to condensation risk. We based our approach on the fundamental tools that were developed for the NFRC standards but with the
109
110
STP 1617 On Building Science and the Physics of Building Enclosure Performance
substitution of project-specific information. Specifically, if project-specific design temperatures and associated air movement are known for the local climate and interior of the building, these boundary conditions can be substituted into the NFRC analysis process. For the exterior climate of a project, resources such as ASHRAE’s Fundamentals2 handbook and the National Oceanic and Atmospheric Administration (NOAA) provide detailed and historic climate data for thousands of locations. With this data, it is a straightforward process to associate a given set of exterior conditions to a statistical occurrence. This can in turn be used to provide the owner with an estimate of what a given design condition can relate to with regard to the number of hours and occurrences that condensation may occur in an average year. In an ideal process, if the analysis predicts condensation risk at a frequency that is untenable for an owner, a combination of better insulating fenestration components or different interior design criteria can be incorporated into the design. The interior conditions of a building, temperature, humidity, and air movement, are influenced by the building’s HVAC system as well as the building enclosure itself. Some spaces require atypical interior conditions, making the estimation of what these conditions might be more difficult. Tall spaces, and those that limit mechanical conditioning of the air near the building enclosure systems, may have atypical and nonuniform interior conditions throughout the space. Certain building types typically have a relatively small tolerance for condensation under design conditions. One example is a museum, which typically contains tall open areas where artifacts and paintings are better preserved when RH is maintained at elevated levels such as 45% RH.3 Fluctuations of temperature and RH in museum spaces typically are limited to preserve artifacts. The mechanical systems in many museums are atypical of other building types, and thus the assumptions for near-fenestration air temperatures and movements can be difficult to estimate. In these spaces, a conservative assumption of air temperature and film coefficient near the interior surface will most often lead to predictions of condensation, despite well-designed and insulated fenestration components. While design guides such as ASHRAE’s handbooks,2 and physical experiments by others,4 provide some guidance on what quantities of air movement may be expected in these spaces, the custom nature of the mechanical and fenestration systems limits the applicability of the values provided. Computational fluid dynamics (CFD) analysis is a powerful tool for the analysis of fluid flow and associated physics, including heat transfer. In the built environment, CFD analysis has the ability to predict airflow patterns as well as temperature and humidity gradients in spaces. These analyses can include building enclosure thermal performance, interior building thermal sources (e.g., plug load, lighting load, radiators), HVAC systems, and complex interior geometry that are difficult to analyze with other traditional analyses. The models can also track all conjugate heat transfer across the boundaries and can account for energy flow, velocities, and turbulence in the interior environment. From the results of the analysis, including
MOUSSAWI AND BONDI, DOI: 10.1520/STP161720180079
fields of temperature and fluid flow, it is possible to extract the film coefficient at the surface of interest. The air movement and temperatures near the fenestration system can be estimated via numerical simulations using CFD simulations to determine localized film coefficients and associated air temperatures throughout a space. The results of these analyses can be used directly with statistical weather data to provide a digital estimate of how the building enclosure and mechanical system will perform under a given set of conditions. Others5 have integrated the use of CFD simulations in the direct prediction of condensation for fenestration in a simulated space; the authors propose the integration of CFD with more traditional tools to enable the simulation of larger spaces with finite computational resources. The results of this analysis can be used directly with FEA analysis with programs such as THERM to determine detailed surface temperatures and a reasonable estimate of condensation potential. This paper presents the details of one such analysis the authors have performed. THE FILM COEFFICIENT
Heat transfer through building enclosures involves all modes of heat transfer, including convection due to natural airflow outside and air movement inside of a space, conduction through the enclosure assembly, and radiation heat transfer to the exposed surfaces on the interior and exterior, as shown in equation (1). qtotal ¼ qconduction þ qconvection þ qradiation
(1)
The conductive heat transfer is controlled by the components and arrangement of components in the building enclosure assembly itself. The convective and radiative portions of heat transfer have a significant impact on the total quantity of heat transfer through an assembly. A large quantity of the convective and radiative components occurs at the interfaces of the interior and exterior surfaces of the enclosure assembly. Convection and radiation also occur in air cavities of assemblies. An air film coefficient can be used to quantify heat transfer on each boundary. As defined by ASHRAE and in the context of fenestration systems, the film coefficient, hi , is a description of the convective and radiative heat flow between the interior glazing surface and the adjacent air. The film coefficient depends on the indoor air temperature and speed, the temperature of the surface of the glazing, the emissivity of the glass surface, and the radiant temperatures to which it is exposed (fig. 1). Chapter 15 of ASHRAE’s Fundamentals2 provides equation (2) for determining hi hi ¼ hic þ hir ¼ 0:3ðDT=LÞ0:25 þ erðTf4 Ts4 Þ=DT
where: DT is Ts Tf ,
(2)
111
112
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 1 Heat transfer mechanisms through an enclosure assembly, with convection shown in blue and radiation heat transfer shown in red.
L is the height of the glazing in feet, r is the Stefan-Boltzmann constant, and e is the surface emissivity. Effectively, the radiative and convective components of the energy transfer are normalized with respect to the surface temperature and the reference temperature of the fluid. In other words, the convective film coefficient, hi ; at the interior surface is described by the total heat flux at the surface, qs , the temperature at the interior glazing surface, Ts , and a reference temperature in the fluid, Tf . The relationship is defined as: qs ¼ hi ðTs Tf Þ
(3)
Currently, there is no established method to obtain values for the interior film coefficient for building enclosure assemblies. Chapter 17, Table 2, of ASHRAE’s Fundamentals2 provides reference interior film coefficients to use along with temperatures for both winter and summer conditions for different glazing heights and glazing system types. ISO 15099 provides several empirical equations for heat transfer by natural convection in conditions typically found in and near fenestration.6 In equation (3), the reference temperature to be used, Tf , is not precisely defined. Conventionally, a reference temperature is chosen at a measurable location from a surface or at the extents of the boundary layer of the flow along that surface. Note several different combinations of the film coefficient and reference temperature can be used with the same overall heat transfer and surface temperature.
MOUSSAWI AND BONDI, DOI: 10.1520/STP161720180079
For the exterior, environment empirical relationships are available to relate wind speed to film coefficients. ASHRAE and NFRC standards reference several different relationships; for the purposes of our case study, we utilize the default relationship used in NFRC standards and LBNL tools that is derived from ISO 15099. The wind speeds can be selected with a statistically coincident wind speed to provide a measure of recurrence. There are a variety of empirical relationships in literature for forced natural convection modes of heat transfer. Most of these relationships, some of which are presented in ASHRAE’s Fundamentals,2 relate to primitive geometry forms that only describe the most rudimentary geometric applications; relating these to building interiors can be difficult. It is difficult to accurately predict a convective film coefficient for interior spaces with a high degree of confidence using these same equations. The importance of the convective component of heat transfer combined with the difficult nature of quantifying it make performing accurate assessments of heat transfer, and thus enclosure surface temperatures, difficult. Computational fluid dynamics analysis provides a method for determining these values based on specific interior conditions. MECHANICAL SYSTEM DESIGN
Building mechanical systems traditionally are developed via the integration of manufacturers’ equipment that transports heating and cooling to spaces throughout a building. These systems typically are designed to offset thermal loads in the building—both from internal loads as well as heat gains and losses through the building enclosure. Various types of equipment can be used to deliver heating; among the most common are systems that directly apply heated air or elements that are used to heat the air. Systems can be designed to provide occupant comfort and to direct heating to building enclosure elements. Heating systems are often designed to offset heating losses at the building enclosure, particularly at fenestration as that is where the majority of the losses occur. The application of heat locally to an enclosure assembly on the interior space of a building can help reduce the risk of condensation through promoting more convection to the enclosure assembly and directly increasing the surface temperatures of the assembly. It is the authors’ experience, however, that using standard mechanical heating design, even when heat is provided at the perimeter of the building, does not necessarily minimize condensation risk as not enough heat may be delivered to or distributed throughout the building enclosure. For buildings with humidification, the risk of condensation is greater than an equivalent nonhumidified building because of the increased dew point temperature. Manufacturers of mechanical equipment provide generic performance data that may not accurately reflect how that equipment will perform in a space. This can especially be the case in large, tall spaces, limiting the designer’s ability to predict, without detailed analysis, the temperature distribution in the space. The use of manufacturers’ standard equipment to develop mechanical systems inherently limits what types of equipment can be placed into an architectural design.
113
114
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Case Study: Analysis of a Museum Environment Our example case study consists of a museum gallery with extensive fenestration, including a metal-framed curtain wall and skylight system. The approximate plan dimensions are 100 ft. by 100 ft. with a height of 62 ft. The architectural design mandated that the mechanical equipment be concealed at the perimeter of the space as much as possible. The design team was tasked with trying to optimize both the mechanical and fenestration systems for cost while maintaining a minimal frequency of condensation. The driver for avoiding condensation was primarily a need to prevent condensation from dripping on exhibits. Another concern was for the safety of visitors who might slip on the smooth flooring if it was wet. The building fenestration systems consist of double-pane insulating glazing units (IGUs) on the facade and triple-glazed IGUs on the skylight. The building structure includes a steel space frame immediately inboard of the skylight and curtain wall, which supports these systems. As part of the architectural and lighting design, there are also fixed shades just beneath the skylight within the space truss. The HVAC airside supply system in the gallery consists of air nozzles located at distances of approximately 22 ft. above the ground and a set of diffusers just below the skylights. In addition, there are fan coil radiators (forced air heaters) located at the base of the curtain wall as well as at 22 ft. and 40 ft. above the ground. Airside return openings are located high in the space on the walls. The geometry of the space and the HVAC system terminals are shown in figure 2 and figure 3. The baseline interior design temperature and RH for our modeling assumptions are 20 C (68 F) and 46 6 3% RH, while the most conservative case for
FIG. 2 Geometry of the interior space as modeled.
MOUSSAWI AND BONDI, DOI: 10.1520/STP161720180079
FIG. 3 Geometry of curtain wall and skylight areas.
condensation analysis was 20 C (68 F) and 56% RH. The exterior temperature and wind speed were selected as the 99.6% heating design condition for the project location, which resulted in 8.3 C (17 F). The mean coincident wind speed for this design temperature was 17 mph. This condition accounts for all but 35 h of an average year, which are more severe. An initial study was performed to test the risk of condensation with various estimates for air movement inboard of the curtain wall and skylight systems. These studies used thermal FEA models using LBNL THERM to simulate the surface temperatures of the fenestration systems with the design exterior conditions and an array of potential conditions. The studies’ results showed a level of uncertainty regarding whether condensation risk could be avoided during a design condition with natural to mild forced convection. If a large film coefficient was used in the analysis, then a relatively low near-enclosure air temperature (Tf) could be assumed without a significant risk for condensation under the design conditions. Using a film coefficient more representative of natural convection, a temperature very close to the space temperature would be required to achieve a low frequency of condensation. The design team determined that a more refined analysis was needed to provide the owner with an understanding of when condensation might occur and to attempt to optimize the mechanical and fenestration design. To accomplish this task, we employed a CFD analysis of the space to accurately determine the near-wall conditions. Our CFD modeling approach included the detailed geometry of the space (including some of the larger artifacts that might disrupt airflow in the space) and
115
116
STP 1617 On Building Science and the Physics of Building Enclosure Performance
the simulation of all of the mechanical elements mentioned previously that supply heat to the space. We simulated the design-day condition used during the sensitivity analysis. As the environmental conditions selected for the analysis typically occur in middle-of-the-night conditions, we assumed no occupancy and insignificant plug loads within the space at the time of the analysis. The mechanical equipment required the specification of heat and airflow volumes. We worked with the mechanical designer to tune each of the systems in the space to achieve an average interior set-point temperature of 20 C (68 F). The process of achieving the balance of heat supply from each system was tuned manually within the model via an iterative approach. The thermal performance of the building fenestration systems was simulated directly in the model. The analysis was performed under steady-state conditions, a reasonable assumption giving the desired constant interior conditions and continued HVAC system operation. The objective of the analysis was to determine the risk for condensation potential with the addition of heating elements along the curtain wall. We also set out to consider the effect of using warm-edge spacers to the IGUs. Warm-edge spacers typically are made of either stainless steel or of polymer materials, reducing the conduction around the IGU perimeter from the base case where standard aluminum spacers are used. Because the perimeter of the IGU typically is most susceptible to condensation, warm-edge spacers present a convenient passive method by which we can reduce potential for condensation. MODEL DEVELOPMENT
The location of mechanical equipment and geometry of the skylight of the space resulted in the need for a three-dimensional model. We built the geometry using Solidworks and exported the volume of the interior air space into our CFD simulation tool, Star-CCMþ (fig. 3). We included part of the gallery exhibit in the model, particularly where the exhibit may have interrupted airflow in the interior of the space. We used Star-CCMþ’s internal polyhedral meshing tool to discretize the geometry. We used a prism layer mesher to develop a refined grid near the boundaries of the model and near detailed geometry. We refined the mesh to obtain wall values of yþ < 20 for all surfaces of interest on the curtain wall and on the skylight. We limited the largest cells in the model to approximately 6 in. in size. The total size of the mesh is 57 million cells with a polyhedral mesh type (fig. 4). We employed the Realizable k-epsilon turbulence model and a pressure-based segregated solver. We also included the surface-to-surface radiation model to account for heat transfer between surfaces within the model. Given the size of the model, the simulation was run on a parallel processing compute cluster; several hundred cores were used to decrease to solve time. CFD INPUTS
The geometry of the model included both the framing and glass surface, some of the interior exhibits and structure, and the HVAC terminals (fig. 2 and fig. 3). We
MOUSSAWI AND BONDI, DOI: 10.1520/STP161720180079
FIG. 4 CFD model mesh.
first defined the velocities and temperatures of the air from the wall diffusers according to the mechanical design drawings. These values represented the operating conditions under exterior design conditions. Working with the mechanical engineers for the project, we varied the capacities of the fan coil units at the base and at the 22-ft. level of the curtain wall, and we slightly varied the performance of the HVAC system to ensure that we achieve the set-point temperature of 20 C (68 F) inside the space. We simulated the thermal conductivity through the fenestration directly for the insulated glazing and indirectly for the framing. We specified effective conductivity of the IGUs and the exterior temperature and heat transfer coefficient, and we allowed the CFD analysis to compute the interior surface temperature of the glass. We used WINDOW to obtain the effective conductivity of the IGU system. The interior fenestration framing, unlike the IGUs, is not uniform through its thickness and cannot be assigned an effective conductivity. Therefore, we defined the interior surface temperatures of the mullions manually using the two-dimensional finite element analysis performed in THERM. We updated these surface temperatures as the CFD analysis updated the nearsurface air temperatures on an iterative basis, using the THERM models to develop updated surface temperatures as the CFD model converged. The CFD model was set up to allow for variations in surface temperatures from one segment of framing to another. The variations of temperature in this model were found to be small. As the goal of the output of the CFD model was to provide film coefficients and local air temperatures near the framing and glazing, small variations in surface temperatures would not be significant in determining the overall air movement near the fenestration of this space. We did not integrate the detailed thermal modeling of the fenestration framing in the model directly as it would have significantly increased the size of the model and thus the computational resources required to solve the model.
117
118
STP 1617 On Building Science and the Physics of Building Enclosure Performance
We conservatively assumed a value of 0.9 for the emissivity of the glass and 0.72 for the emissivity of the framing elements as inputs for the radiation modeling of the space. We selected the exterior sky radiation temperature as 17.8 C (0 F), which was included in the exterior environmental boundary conditions. OVERALL CFD ANALYSIS RESULTS
The results of the CFD analysis showed a significant impact of fan coil radiators on the interior temperature of the IGUs as well as on the patterns of airflow in the space. Figure 5 shows interior surface temperatures colored with a range of 8.9 C (48 F) to 23.9 C (75 F) and air velocity vectors colored with a range of 0 to 194 ft./min. on a plane perpendicular to the fenestration. The IGUs directly adjacent to the radiators (shown in orange) at the 22 ft. height are washed by the upward forced convective current. A downdraft exists on IGUs one bay above the position of the radiators. There is also a significant increase in the temperatures of the IGUs that are washed by the radiator output. A similar pattern exists at the skylight, where IGUs closer to the HVAC diffusers are washed by the warm conditioned air and those farther away are colder. Figure 6 shows a plot of streamlines originating at the HVAC diffusers. The streamlines show that the air washes the skylight IGUs closer to the wall (left of image) with warm air, and the IGUs that are farther away (right of image) do not receive the same heating and have a lower extent of air movement. USING CFD ANALYSIS RESULTS TO PREDICT FILM COEFFICIENTS
The complexity of the fenestration systems made it impractical to simulate the detailed heat transfer through the framing systems of the curtain wall and skylight
FIG. 5 Analysis of air velocities and surface temperatures at fenestration.
MOUSSAWI AND BONDI, DOI: 10.1520/STP161720180079
FIG. 6 Streamlines at the skylight colored by temperature.
directly in the CFD model. Our goal was therefore to obtain the film coefficient and the air temperature near the surface from the CFD model, such that we could input these as boundary conditions into our two-dimensional finite element analysis models of the fenestration section details. We could then determine the interior surface temperatures of the framing and estimate the number of hours per year for which condensation is expected. Our first step was to extract the values in equation (3) from the CFD analysis to determine the film coefficient. Step 1: Determining the Film Coefficient We used the IGU as a reference for the film coefficients and the near-surface air temperatures required for the two-dimensional finite element analysis. We began our procedure by defining surfaces offset by 1 in. for each of the IGUs in the curtain wall and in the skylight of the model (fig. 7). We inspected the average temperatures of the air at these imaginary planes and determined the IGUs with the lowest near-surface air temperature, Tf , for further analysis. At these designated IGUs, we determined the average surface temperature over the interior IGU surface, Ts, and the average thermal and radiative energy flux per unit area, qs. From these values, we calculated an estimate for the film coefficient, hc, using equation (1). Step 2: Determining the Lowest Near-Surface Air Temperatures We then determined the lowest near-surface air temperatures to use in our two-dimensional thermal analysis for the same IGUs selected in Step 1. The
119
120
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 7 Interior surface temperatures (left) and near-wall air temperatures at a distance of 1 in. (right).
near-surface air temperature varies along the imaginary plane that is offset 1 in. from the IGU, so extracting a value for our two-dimensional thermal analysis boundary condition is a subjective process. We found that very small areas around the edge of the plane may have significantly lower temperature than the remainder of the plane. Therefore, we employed a histogram of temperatures of cells intersected by the plane and used a 5% cut-off value to obtain a lower bound on the near-surface air temperature. The process is shown in figure 8 and figure 9 for one of the skylight IGUs selected.
FIG. 8 Near-surface air temperature distribution for a skylight IGU.
MOUSSAWI AND BONDI, DOI: 10.1520/STP161720180079
FIG. 9 Histogram of CFD temperature distribution shown in figure 8.
We employed this two-step process to determine the film coefficients and nearsurface temperatures for several of the worst-case IGUs. We then selected the worst-case conditions, the lowest near-surface air temperatures, and the lowest film coefficients for use in our two-dimensional thermal analysis. Our analysis resulted Btu in a value of 0:9 hrft 2 F and a near-wall temperature of 13:3 C ð56 FÞ for the skyBtu light, a value of 1:2 hrft 2 F and a near-wall temperature of 13:3 C ð56 FÞ for several locations on the curtain wall away from the radiators, and a value of Btu 1:4 hrft 2 F and a near-wall temperature of 12:8 C ð55 FÞ at the base of the curtain wall away from in-floor radiators. The values for the film coefficients are consistent with the greater air movement near the curtain wall due to the radiators and the lower air movement at the center of the skylight area away from the diffusers. In addition, the larger film coefficient at the base of the curtain wall is consistent with air movement at the large inground fan coil radiators at the base of the curtain wall. The values for the curtain wall and skylight were significantly larger than the published standard values in NFRC 500. TWO-DIMENSIONAL THERMAL ANALYSIS RESULTS
After obtaining the film coefficients and the near-wall air temperatures from the CFD analysis, we simulated the two-dimensional thermal analysis of the framing details using these values as boundary conditions. The exterior boundary conditions for the two-dimensional thermal analysis are the same as those used in the CFD model, and they are based on statistical weather data. Figure 10 shows the different section details for which we performed two-dimensional thermal modeling on the building envelope.
121
122
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 10 Two-dimensional thermal analysis models analyzed for the curtain wall and skylight details.
Figure 11 shows the detail at the head of the curtain wall. The left image in figure 11 shows the colored temperature distribution within the outline of the section. The temperature distribution is cut off at the interior dew point temperature of 8.11 C (46.6 F) to allow easy identification of areas where there is potential for condensation. As shown in figure 11, there is a small area at the edge of the IGU where the interior surface temperature is 14.7 C (5.6 F) below the interior dew point temperature and at which we expect condensation to occur. We performed a similar analysis for the remainder of the details.
FIG. 11 Two-dimensional thermal analysis results for the head-of-curtain-wall condition with a standard spacer (left) and with a warm-edge spacer (right).
MOUSSAWI AND BONDI, DOI: 10.1520/STP161720180079
To help illustrate for the owner the likely risk of condensation under design conditions, we used the overall temperature results calculated directly in the CFD model for the IGUs to determine the locations that had a higher expected risk of condensation. These locations, combined with the statistical information that the exterior conditions were based on, provided a basis for the owner to make decisions regarding which additional mechanical equipment and glazing types would be selected for the final design of the project. We analyzed the section details in our two-dimensional thermal analysis using standard spacers and warm-edge spacers and found a significant reduction in the number of IGUs affected by condensation using the warm-edge spacer. The analysis results for the detail at the head of the curtain wall are shown in figure 11 (right image) for comparison to the standard spacer case. The dashed line indicates the interior surface that should always be above the dew point temperature in order to avoid condensation. We cut off the temperature plot at the dew point of 8.11 C (46.6 F) to make it possible to determine if condensation would occur by inspection. The black circles indicate locations where the dew point temperature falls on the interior surface. For this detail, the difference between the surface temperature and the dew point temperature decreased from 14.7 C (5.6 F) using a standard spacer to 16.4 C (2.4 F) using a warm-edge spacer. Our two-dimensional thermal analysis results for the complete set of details using warm-edge spacers indicated that 13% of the curtain wall IGUs in the current design and 4% of the skylight IGUs are affected by condensation. By contrast, the thermal analysis using standard spacers indicated that 64% of the curtain wall IGUs and 50% of the skylight IGUs would be affected by condensation. We note that, while the analysis we performed compared the surface temperature to the exact dew point temperature of the space, variations and tolerances during construction may change the results. For the purposes of this project, the results were being used to compare two options for the glazing spacer in conjunction with the required mechanical operating conditions to minimize the areas that are predicted to condense. If this procedure was not being used to compare fenestration design options, then the authors would advise including a buffer temperature between the predicted surface temperature and the dew point temperature to define whether a system will likely condense. CONCLUSIONS
In this paper, we demonstrate the use of CFD modeling of building spaces to inform the thermal analysis of condensation on building enclosure assemblies. The method presented can be used to provide project-specific evaluations that can be related to a statistical probability of occurrence. For high-humidity environments, odd geometry conditions, or in buildings where condensation potential is a major consideration in the design, a comprehensive analysis such as that described here provides a methodology to obtain the interior temperature and film coefficient near the building envelope for condensation analysis.
123
124
STP 1617 On Building Science and the Physics of Building Enclosure Performance
For the example project discussed, the results of the incorporation of the CFD model into the condensation resistance analysis provided project-specific air movement and near-wall temperature inputs into the condensation risk assessment for the fenestration systems. The results of the CFD model showed that different air film coefficients and near-wall air temperatures were predicted relative to standard film coefficients provided in NFRC standards. This analysis approach allowed for the coordination of the design HVAC operating parameters with a digital simulation of the performance of the design fenestration assemblies proposed for the building prior to construction. The results of the analysis also allowed the design team to estimate for the owner how the proposed fenestration systems would perform. We hope to gather physical data for this project in the future such that we can publish the accuracy of using CFD to predict interior environmental conditions as they relate to statistical condensation risk relative to the behavior of the actual installed mechanical and glazing systems.
References 1. 2. 3.
4.
5.
6.
NFRC 500, Procedure for Determining Fenestration Product Condensation Resistance Values (Greenbelt, MD: National Fenestration Rating Council, Inc., 2017). American Society of Heating, Refrigerating and Air-Conditioning Engineers, ASHRAE Handbook: Fundamentals (Atlanta, GA: ASHRAE, 2017). M. F. Mecklenburg, “Determining the Acceptable Ranges of Relative Humidity and Temperature in Museums and Galleries,” 2007, http://web.archive.org/web/20191204190222 /https://repository.si.edu/handle/10088/7055 A. J. N. Khalifa and R. H. Marshall, “Validation of Heat Transfer Coefficients on Interior Building Surfaces Using a Real-Sized Indoor Test Cell,” International Journal of Heat Mass Transfer 33, no. 10 (1990): 2219–2236. J. L. Boone and E. R. Pugh, “Practical Application of Coupled FEA and CFD Steady State Analysis for the Prediction of Nighttime Building Enclosure Surface Temperatures,” in Advances in Hygrothermal Performance of Building Envelopes: Materials, Systems and Simulations, ed. P. Mukhopadhyaya and D. Fisler (West Conshohocken, PA: ASTM International, 2017): 344–361, https://doi.org/10.1520/STP159920160103 Thermal Performance of Windows, Doors and Shading Device—Detailed Calculations, ISO 15099 (Geneva, Switzerland: International Organization for Standardization, 2015).
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180070
Carly May Wagner1 and Rex A. Cyphers1
Role of Initial Moisture Content on Hygrothermal Models and Envelope Performance Citation C. M. Wagner and R. A. Cyphers, “Role of Initial Moisture Content on Hygrothermal Models and Envelope Performance,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 125–156. http://doi.org/10.1520/STP1617201800702
ABSTRACT
Countless inputs are used in transient hygrothermal analysis, just as countless variables determine the long-term hygrothermal performance of a real-world building envelope assembly. One variable often overlooked, oversimplified, and inaccurately accounted for is the initial moisture content of materials. This paper discusses the impact initial moisture content has on both the real-world performance and the accuracy of the modeled hygrothermal performance. ASHRAE Standard 160, Criteria for Moisture-Control Design Analysis in Buildings, provides guidelines for assumptions of initial moisture content inputs in hygrothermal analysis. These guidelines are generic in nature and are not necessarily representative of in-service conditions. The authors conducted a study to explore the role that a range of inputs have on the accuracy of hygrothermal models. Temperature and humidity data across a wall assembly along with all necessary interior and exterior climate data were collected over the course of several seasons. The data were used to develop a series of WUFI models, attempting to validate these models and the default or recommended inputs. The study indicated that the models did not accurately reflect the data recorded in the field unless values for the initial moisture content matched what was recorded in the field rather than the default ASHRAE 160 recommended values. This study, and several other real-world cases where excessive initial Manuscript received October 9, 2018; accepted for publication August 15, 2019. 1 WDP & Associates Consulting Engineers, Inc., 335 Greenbrier Dr., Suite 205, Charlottesville, VA 22901, USA C. M. W. https://orcid.org/0000-0001-6913-3010, R. A. C. https://orcid.org/0000-0003-0034-7840 2 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
125
126
STP 1617 On Building Science and the Physics of Building Enclosure Performance
moisture content has led to envelope failures, highlights the importance of accurately accounting for initial moisture content. Particularly in the cases of renovations and restoration projects, methods are available to determine realistic values of the initial moisture content within the existing materials. Appropriate methods for determining the initial moisture content of materials comprising existing envelope assemblies are discussed, including the methods contained within the new standard ASTM E3069-19, Standard Guide for Evaluation and Rehabilitation of Mass Masonry Walls for Changes to Thermal and Moisture Properties of the Wall. Keywords moisture content, hygrothermal analysis, vapor permeance, insulation, hygroscopic materials, sorption-isotherm curve, moisture storage function
Introduction Industry best practice stipulates avoiding the use of “damp,” “wet,” or “moist” materials in many construction applications. Protect materials from moisture. Do not install over wet or damp substrate. Remove any wet material. All these recommendations are repeated commonly in material technical bulletins, project specifications, technical product data sheets, and industry association position statements. However, most design and construction professionals, manufacturers, and fabricators cannot define the terms damp, wet, moist, or other similar terms with any universally quantifiable parameters because these terms turn out to be extremely complicated and technical to define, cannot be universally defined, and vary from material to material. How wet is too wet, how damp is too damp, and how moist is too moist not only depends on the individual material but also on the application of that material as well as the other materials within the envelope assembly in which it is installed. As it relates to building construction, moisture comes in two forms: vapor and liquid, both of which have an impact on the moisture content of materials. Sometimes, it can be seen or otherwise sensed; but other times, it cannot be detected without instrumentation. Even with instrumentation, it is difficult to accurately determine a material’s moisture content without taking samples and comparing the in situ weight to the oven-dried weight. Moisture is always present to some degree, but it is only at certain levels that the moisture will become problematic. Prolonged periods of elevated moisture content within envelope assemblies is directly associated with corrosion, mold growth, structural decay, condensation, and a variety of other problematic conditions. This paper begins with an overview of the relationship between moisture in the vapor form and material moisture content and then discusses a range of moisture sources during construction that can have a direct impact on the initial moisture content of an envelope assembly. A brief literature review of the industry recommendations for specific materials known to be sensitive to moisture content is
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
presented. Next, the paper discusses envelope assemblies that are particularly sensitive to moisture content, demonstrating the role initial moisture content plays in these assemblies through hygrothermal computer simulation models. Finally, the paper presents a study that used field data to calibrate the commonly used hygrothermal modeling software and the default assumptions. This study indicated that the assumptions regarding initial moisture content were more critical variables with respect to the accuracy of the models then would have been anticipated. HYGROSCOPIC NATURE OF MATERIALS
It is first necessary to recognize that most building materials are hygroscopic in nature, meaning that any discussion of moisture must look beyond liquid water and consider water in the vapor form. Hygroscopic materials will, by definition, take on moisture as the ambient relative humidity (RH) increases through the process of adsorption. Figure 1 shows examples of how the equilibrium moisture content of a material generally increases with ambient relative humidity. These are graphical representations of material data published in Moisture Analysis and Condensation Control in Building Envelopes (ASTM Manual 40).1 The graphs in figure 1 are often colloquially referred to as the moisture storage function of a material, referring to how much moisture the material will take on at a given relative humidity. More technically, the graphs are called sorption-isotherm curves, which reflect the amount of water a material will take on at a fixed temperature. The graphs are made up of data from two distinct regions: 1. The data up to approximately 90 to 95% humidity represent the hygroscopic and super hygroscopic portion of the sorption-isotherm curve. The data for this region of the curve should be developed based on ASTM C1498, Standard Test Method for Hygroscopic Sorption Isotherms of Building Materials.2 2. The data from 95 to 100% are relative humidity and represent the capillary portion of the sorption isotherm curve. The data for this region of the curve should be developed based on ASTM C1699, Standard Test Method for Moisture Retention Curves of Porous Building Materials Using Pressure Plates.3 The term sorption refers to the process of both adsorption (the process of taking on moisture) and desorption (the process of releasing moisture). For some materials, the desorption process requires a greater reduction in ambient relative humidity to reach the same moisture content. This phenomena in the difference between adsorption and desorption is referred to as hysteresis. It should also be noted that the moisture content shown in the graphs is the equilibrium moisture content (EMC). There is a time aspect to sorption, and a material will not instantaneously change its moisture content with an instantaneous change in relative humidity. At a fixed relative humidity, the material will slowly take on or give up moisture, eventually reaching the equilibrium moisture content for that given humidity level.
127
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 1 (A) Moisture storage (sorption-isotherm) for brick. (B) Moisture storage (sorption-isotherm) for mortar. (C) Moisture storage (sorption-isotherm) for gypsum board. (D) Moisture storage (sorption-isotherm) for stucco. (E) Moisture storage (sorption-isotherm) for oriented strand board (OSB). (F) Moisture storage (sorption-isotherm) for plaster. (G) Moisture storage (sorption-isotherm) for pine. (All as published in ASTM Manual 40.1)
Brick
(A) Water Content (lb/ft3)
15.0 12.0 9.0 6.0 3.0 0.0 0
10
20
30
40
50
60
70
80
90
100
70
80
90
100
Relave Humidity
Mortar
(B) Water Content (lb/ft3)
21.0 18.0 15.0 12.0 9.0 6.0 3.0 0.0 0
10
20
30
40
50
60
Relave Humidity
Gypsum Board
(C) Water Content (lb/ft3)
128
36.0 33.0 30.0 27.0 24.0 21.0 18.0 15.0 12.0 9.0 6.0 3.0 0.0 0
10
20
30
40
50
60
Relave Humidity
70
80
90
100
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
FIG. 1 (continued)
Stucco
(D)
Water Content (lb/ft3)
21.0 18.0 15.0 12.0 9.0 6.0 3.0 0.0 0
10
20
30
40
50
60
70
80
90
100
70
80
90
100
70
80
90
100
70
80
90
100
Relave Humidity
OSB
Water Content (lb/ft3)
(E) 27.0 24.0 21.0 18.0 15.0 12.0 9.0 6.0 3.0 0.0 0
10
20
30
40
50
60
Relave Humidity
Plaster
(F) Water Content (lb/ft3)
12.0 9.0 6.0 3.0 0.0 0
10
20
30
40
50
60
Relave Humidity
Pine
(G) Water Content (lb/ft3)
21.0 18.0 15.0 12.0 9.0 6.0 3.0 0.0 0
10
20
30
40
50
60
Relave Humidity
129
130
STP 1617 On Building Science and the Physics of Building Enclosure Performance
From the graphs in figure 1, it can be seen that the increase in moisture content is roughly directly proportional to the increase in moisture content for humidity levels up to 90 to 95% relative humidity. At humidity levels above 90 to 95%, the moisture content tends to increase exponentially. This is due to the physical nature of the pore structure of the material and the process by which the pores adsorb moisture. Figure 2 provides a conceptual graph of the various phases of the sorption-isotherm curves. At relatively low relative humidity, up to about 60 to 80%, water molecules simply bind to the walls of the pores within materials. This is referred to as the hygroscopic range. At humidity levels above 60 to 80%, condensation within the pores begins to occur, and unbound liquid water is stored within the pore structure of the material. This is referred to as the super hygroscopic range or the capillary water range. The material continues to adsorb far more moisture from far smaller increases in humidity until 100% relative humidity. The equilibrium moisture content at 100% is known as free water saturation. However, at free water saturation, 100% of the pores are not yet filled with moisture. Any pores that are completely closed off can still be filled if pressure drives moisture into these pores. At a theoretical humidity above 100%, a material goes from free water saturation to maximum moisture content in what is referred to as the supersaturated region of the sorption-isotherm curve. A material’s maximum moisture content will be greater than the free saturation water content. The exact curve of a material’s sorption-isotherm curve is unique and is related to the pore structure of the material: open pores versus closed pores, ratio of large pores to small pores, and other factors.
FIG. 2 Conceptual graph illustrating the regions of a sorption-isotherm curve.
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
Impacts of Moisture Content within a Material Many material properties are directly related to the material’s moisture content. It follows that many material properties are then also related to the ambient relative humidity because moisture content is a function of relative humidity. The first material property that is directly impacted by moisture content is the material’s thermal resistance. Water is highly thermally conductive, so as moisture content of a material increases, the thermal resistance of the material decreases. The concept of moisture content decreasing a material’s thermal resistance seems to be fairly well appreciated by design professionals and relatively common knowledge. Unfortunately, there is limited published data that provide quantitative correlations between moisture content and thermal resistance for most materials. This is likely because there are many different test methods used to determine a material’s reported R-value (capacity to resist heat flow), some of which require a certain relative humidity for the testing procedures and some of which do not stipulate a relative humidity for the test procedure. The vapor permeability of a material is also reliant upon moisture content. As the moisture content of a hygroscopic material increases, the vapor permeance tends to increase as well. However, a material’s vapor permeance is typically stated as a single value, implying that vapor permeance is a fixed material property. More often than not, a material’s vapor permeance is reported and sometimes noted whether that value was found using ASTM E96, Standard Test Methods for Water Vapor Transmission of Materials,4 Method A, often called the “dry cup method,” or Method B, often called the “wet cup method.” Vapor permeances found using the wet cup method will be greater than vapor permeances found using the dry cup method even though both methods result in roughly the same vapor pressure difference across the specimen. For the dry cup, the vapor pressure inside the cup is assumed to be 0 inHg (0% relative humidity, 734 F, 23 C), and the vapor pressure outside the cup is roughly 0.42 inHg (50% relative humidity, 73.4 F, 23 C), resulting in a 0.42 inHg vapor pressure gradient across the specimen. For the wet cup, the vapor pressure inside the cup is assumed to be roughly 0.83 inHg (100% relative humidity, 73.4 F, 23 C), and the vapor pressure outside the cup is roughly 0.42 inHg (50% relative humidity, 73.4 F, 23 C), resulting in a 0.41 inHg vapor pressure gradient across the specimen. Thus the relative increase water vapor transmission rate found using the wet cup method as compared to the dry cup method, is not attributable to an increased vapor pressure difference. The greater permeance for the wet cup is primarily a result of the specimen having a greater moisture content. Although the relationship between vapor permeance and moisture content is often obscured by the manner in which most manufacturers report vapor permeance, various published data sources show how dramatically moisture content can influence a material’s vapor resistance. For example, ASHRAE Handbook—Fundamentals5 reports a permeability of 0.14 perm-in for 0.25-in. (6.35 mm) fiber cement board (0.56 perm permeance) at 10% relative humidity. The reported permeability
131
132
STP 1617 On Building Science and the Physics of Building Enclosure Performance
increases to 10.1 perm-in (40.4 perm permeance) at 90% relative humidity, a 72fold increase. Unfortunately, it is typically only a relatively small portion of design professionals that spend considerable time doing hygrothermal analysis that see this type of published material data and fully appreciate the dramatic impact that moisture content can have on vapor permeance. In addition to changing a material’s properties, elevated moisture content can play a detrimental role in the serviceability and durability of materials. Although other variables impact the overall risk, with all other variables being equal, higher moisture content will lead to a greater potential for mold growth as well as corrosion of embedded metallic items. It is also well known that the potential for structural decay of wood increases at moisture contents greater than 19%. Based on the density and moisture storage functions included in WUFI’s* material databases, this equilibrium moisture content is typically reached between 85 to 90% relative humidity. Finally, a material’s moisture content can also have an impact on a material’s dimensional stability. Some materials tend to swell and warp at greater moisture contents. APA–The Engineered Wood Association (APA) provides good testing data and model equations for determining linear expansion and thickness swell for oriented strand board (OSB) and plywood.6
Moisture Sources during Construction and Initial Moisture Content When evaluating the moisture content of a material, both liquid water sources and water vapor sources must be considered because most building materials are hygroscopic and will by nature take on moisture as a direct function of the ambient relative humidity. Design and construction professionals must be cognizant of all moisture sources and exercise professional judgement regarding materials that are sensitive to moisture content. When initial moisture content is discussed, it is referring to the moisture content of the materials at Time Step 0 of transient hygrothermal analysis. This is typically taken to represent the time in which a building has all materials installed and the heating, ventilation, and air-condition (HVAC) systems go into operation. In the case of new construction, this implies that construction sequencing and scheduling can have a large impact on the initial moisture content. In the case of retrofits and renovations, the initial moisture content would be the moisture content of the existing building materials when the new insulation or air and vapor retarding materials (or both) are installed. The initial moisture content can dramatically influence the long-term hygrothermal performance of building envelope assemblies. To understand what factors influence the initial moisture content, construction moisture sources need to be discussed.
* Wa¨rme Und Feuchte Instationa¨r (WUFI) is a commonly used, one-dimensional, transient, hygrothermal software package developed by Fraunhofer IBP.
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
The most obvious moisture sources during new construction are rain or other sources of liquid water. It is common practice to store and protect most materials not intended for permanent exposure from the rain. Additionally, it is best practice to “dry-in” a building—to install the weather-resistive barriers and roofing—as fast as possible to limit direct rain exposure. As an example of one of many professional and industry standards dealing with this, APA states that, “APA-trademarked engineered wood products are bonded with moisture-resistant adhesives and are suitable for limited moisture exposure during construction delays. ‘Limited’ is key—construction should proceed with minimal interruptions, and the wood products should be protected by roofing and/or weather-resistive barrier as soon as possible.”7 In practice, it is rare if all materials stored on a construction site are fully protected from the elements during shipping, on-site storage, and throughout the construction process. However, storing materials in excessive relative humidity can also result in excessively high initial moisture content as well. Best practice is to store moisturesensitive materials in fully conditioned space. As an example, information from the Gypsum Association states that, “Gypsum panel products shall not be stored in areas of excessive humidity.” It continues, “The plastic covering provided for product protection [during shipment] is not suitable for storage of gypsum panel products and shall be removed upon arrival at the destination.” It further states, “Failure to remove this plastic covering can result in damage to the gypsum panel products due to moisture, condensation, and/or mold. Exposure of gypsum panel products to rain and other high moisture levels may result in water stains, discoloration, mold, paper delamination, and sag. This sensitivity of most gypsum panel products to moisture requires that gypsum panel products NOT be stored outdoors.”8 Other industry standards and manufacturers’ literature for other hygroscopic building materials contain similar language that require the materials to be stored in dry conditions. (Additional discussion on what dry versus high moisture levels means is included in the next section.) A third moisture source during construction that must be considered is water within site-mixed materials, such as concrete and masonry mortar. This includes the water used to mix concrete roof slabs, concrete floor slabs, and concrete walls, as well as brick and concrete masonry unit (CMU) grout or mortar (or both). CMU and other plant-cast concrete can also be problematic if placed in service prior to sufficient drying after the manufacturing process. Different concrete, grout, and mortar mixes will dissipate this moisture at a different rate. Sometimes, it takes weeks or months to dissipate a majority of this initial moisture. In other cases, it may take months to reduce the moisture to a level that is acceptable for installing the other materials of the assembly, particularly in the case of lightweight aggregate concrete roof decks. MATERIAL-SPECIFIC RECOMMENDATIONS FOR ACCEPTABLE MOISTURE CONTENT
Insulation, gypsum wall board, and wood products are some of the more commonly used moisture-sensitive materials installed in the building envelope. These materials
133
134
STP 1617 On Building Science and the Physics of Building Enclosure Performance
can be damaged because of excessive moisture. A review of the industry standards related to these materials shows that there is an awareness of the issue with excessive moisture, but unfortunately, there are little to no qualitative references to what excessive moisture is. As mentioned previously, the Gypsum Association is clear that gypsum panels must be stored in conditions without excessive moisture content; however, numerical values for what is considered dry are lacking. Similar observations can be made from other industry guidelines. The following excerpts are a few examples of available guidelines of the many industry associations that provide qualitative recommendations regarding acceptable moisture contents for a material that lacks quantitative backing: • The Facts About Mold Growth, published by the North American Insulation Manufacturers Association, states, “NAIMA does not provide specific recommendations addressing the issue of replacing wet insulation. It is very subjective whether insulation needs to be replaced after being exposed to water.”9 The document continues on about conditions that would necessitate replacement, including if mold growth is present, if the intensity of the water exposure was strong enough to “change shape” of the insulation, and if the water was clean. • In its Guidelines for Prevention of Mold Growth on Gypsum Board,10 the Gypsum Association states: Job site conditions that can expose the gypsum board to water or moisture must be avoided. Provisions must be made to keep gypsum board dry throughout application. Gypsum board must not be applied over other building materials where conditions exist that are favorable to mold growth. Immediate and appropriate remediation measures must be taken as soon as water leaks or condensation sources are identified. • Assessing Water Damage to Gypsum Board11 notes that, “Gypsum board may experience limited intermittent exposure to moisture from a variety of sources, such as improper storage, construction or design defects, water leaks, and janitorial activities. Gypsum board exposed to water should be replaced unless all of the following conditions are met.” One of the listed conditions includes, “The gypsum board can be dried thoroughly before mold growth begins (typically 24 to 48 hours depending on environmental conditions).” However, the report never stipulates what moisture contents are considered wet and which moisture content would be considered dry. Instead, it contains a long discussion about how handheld moisture meters cannot be used with any degree of accuracy to identify wet gypsum. • Another APA publication, Service Life of Oriented Strand Board (OSB) Sheathing,12 includes the following warnings: If exposed to conditions that result in the OSB panels having a moisture content in excess of 20 to 25% for an extended period, the OSB is subject *
*
*
*
*
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
to fungal decay and a commensurate loss in strength and serviceability as are any other untreated wood products in similar circumstances. Moisture content changes in OSB sheathing, if severe, may lead to dimensional and structural changes. OSB can withstand exposure conditions associated with construction delays of limited duration because the moisture content fluctuation is within a limited range. • Wood Moisture Content and the Importance of Drying in Wood Building Systems,13 another APA publication, states: Good design and construction practices protect against water leaks, control moisture-laden air infiltration and condensation potential, influence the rate at which air moves around and through the building systems, and mitigate the effects of humidity and temperature differentials between the inside and the outside of the structure. Improper design, construction, or maintenance can result in moisture build-up in the structure and lead to problems with mold, mildew, decay, or other moisture related problems, such as dimensional stability issues. To mitigate moisture exposure, APA recommends covering wood structural panel sheathing as soon as possible after installation. After the building exterior envelope is complete, but before enclosing the wall cavity, roof cavity, or installing interior finishes, the roof and wall sheathing, and lumber framing should be allowed to dry (to less than 18%) to minimize moisture absorption during construction. This is especially important when substances that inhibit drying have been applied to the sheathing or assembly. Although subtle, the last excerpt touches on another topic related to acceptable initial moisture content of individual materials; the acceptable material moisture content is often dictated by the overall assembly. This concept is explored in greater detail and illustrated with actual analysis in the following section. *
*
*
Initial Moisture Content in Hygrothermal Models Hygrothermal computer simulations require a range of inputs and assumptions from material properties to interior climate data, exterior climate data, surface transfer coefficients, and more. ASTM E3054, Standard Guide for Characterization and Use of Hygrothermal Models for Moisture Control Design in Building Envelopes,14 requires that the initial moisture distribution used for modeling be stated in the report. However, the standard does not provide guidelines on what initial moisture contents should be assumed. The standard generally refers to ASHRAE Standard 160-2016, Criteria for Moisture-Control Design Analysis in Buildings,15 one of the most commonly used standards followed for conducting hygrothermal analysis, for detailed guidelines for many inputs. ASHRAE 160 outlines what variables must be accounted for when conducting hygrothermal analysis and provides guidance on how to accurately account for each variable. With respect to initial moisture content, ASHRAE 160 states:
135
136
STP 1617 On Building Science and the Physics of Building Enclosure Performance
The initial moisture content of construction materials in new construction to be used in calculations for this standard shall be two times [equilibrium moisture content at 90% relative humidity] EMC90 for concrete and two times EMC80 for all other materials, unless procedures to dry construction materials and/or procedures to protect construction materials and assemblies from wetting during construction are specified, in which case EMC90 for concrete and EMC80 for all other materials shall be used. In retrofit applications, EMC90 for concrete and EMC80 for all other materials shall be used, unless measured moisture content values are available.15 The commentary on this section notes that the EMC80 is based on International Energy Agency guidelines and practice as outlined in Annex XIV: Condensation and Energy.16 This Annex is a summary of the discussions and proceedings of a workshop held in 1985 at Leuven University in Belgium that brought together building scientists from Belgium, German, Italy, The Netherlands, and the UK. The workshop was focused on mold and surface condensation and was “focusing on the state of the art in different countries” but “revealed a real lack of overall knowledge and understanding on the levels of data, modeling, and measures.” The Annex was aimed at providing researchers and professionals with better knowledge about mold growth and condensation. The introduction of the Annex states: Contrary to what might be expected, the booklet does not contain worldwide applicable rules of the thumb or ready to use performance criteria, to be applied for the avoidance of mold and surface condensation. This is in fact, given the difference in outside climate within and between countries, impossible. What is given is the philosophy behind the way performance criteria could be formulated and a methodology to analyze and solve problems with existing buildings.16 The Annex is referenced by ASHRAE 160 for the rationale behind recommending EMC80 as the default initial moisture content assumption. However, based on the actual text of the Annex, it seems that this EMC80 was not intended by the workshop participants to be taken as an initial moisture content but rather as the threshold for mold germination or a failure criteria. After noting: (1) the fact that the mold growth is temperature dependent; (2) the mold species needing the least moisture requires EMC75 (the Annex uses the term “a-value” and refers to the moisture content as the “water activity”); (3) that some substrates provide more nutrition for mold; (4) that at lower humidity, longer time is needed before mold is visible; and (5) that hygroscopic inertia can allow for short periods of high humidity within the interior of materials without resulting in the same humidity levels at outside surfaces of materials, the group proposed: Take for new design as threshold RH for mold germination “a” = 0.8; use this value all over the year on monthly mean basis, introduce a substrate correction only for regularly cleaned surfaces with porosity zero (glass, metallic, glazed
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
tiles….). There, put a = 1, turning the mold germination condition to the instantaneous surface condensation equation: as soon as the vapor pressure against a surface equals the saturation pressure on it, surface condensation starts.16 Based on the authors’ understanding of the Annex, it does not seem that the group intended for EMC80 to be taken as the initial moisture content but rather to be taken as the threshold for which mold growth can occur, considering other variables such as nutrient value of the surface, temperature, and duration of elevated humidity level. Furthermore, the Annex does not include information to back up the rationale behind ASHRAE 160’s recommendation to use two times EMC80 as the initial moisture content. It should be noted that ASHRAE 160 does allow for lower initial moisture content values to be used for new construction if drying procedures are used and materials are protected from wetting, although no quantitative guidance as to what wet means and what initial moisture content would be appropriate is provided. Similarly, in the case of retrofits, ASHRAE 160 allows for the initial moisture content to be based upon measured moisture contents. ASTM E3069, Standard Guide for Evaluation and Rehabilitation of Mass Masonry Walls for Changes to Thermal and Moisture Properties of the Wall,17 includes provisions for measuring the existing moisture content of mass masonry walls to determine an appropriate assumed initial moisture content. This standard recommends using in situ probes to determine the humidity within various depths of the mass masonry wall. The recorded relative humidity would then need to be connected back to the sorption-isotherm curve for the masonry materials to determine the existing moisture content of the masonry. One downside to this is the sorption-isotherm curves for masonry, especially historic masonry, can vary dramatically. Without actual physical testing data of the specific masonry’s moisture storage function for the existing project, the recorded relative humidity can only be connected back to published data for different masonry; and the moisture contents may or may not be accurate. ASTM E3069 recommends developing sorption-isotherm curves for masonry from the existing building using ASTM C1498, but more often than not, project constraints do not allow for this. Cyphers et al.18 provide case studies where in situ humidity probe data were correlated to laboratory-determined sorption-isotherm curves for project-specific brick. The paper also compares generic published data to the laboratory data.
Envelope Assemblies Critically Sensitive to Initial Moisture Content The impact of moisture content on material properties and serviceability has been discussed earlier in this paper. However, it has not yet been established if initial moisture content can actually have an impact on the long-term moisture content of envelope assemblies. Could moisture within an envelope assembly at the time of
137
138
STP 1617 On Building Science and the Physics of Building Enclosure Performance
construction impact the performance of the assembly five, six, or seven years into the building’s service life? Based on theoretical modeling and computer simulation, it seems that the answer to this depends heavily on the overall assembly. Assemblies with inherently limited drying potential are more likely to be impacted by initial moisture content than assemblies with high drying potential. Drying potential of an envelope assembly can come in two primary modes: (1) drying by convection, meaning that the assembly incorporates some type of ventilation, and (2) drying by diffusion, meaning that the placement of any vapor retarders or vapor-retarding materials is such that the assembly will not prohibit drying via vapor diffusion. One common example of an envelope assembly that can have its long-term performance detrimentally impacted by the initial moisture content at the time of construction is a low-sloped roof with a concrete deck. This issue is one that the National Roofing Contractors Association (NRCA) has been battling:19 “Reported problems include roof systems moisture accumulation, adhesion loss, adhesive issues with water-based and low-volatile organic compound adhesives, metal and fastener corrosion, insulation R-value loss, and microbial growth.” The cause of these issues is installing the roof membrane or vapor retarder (or both) prior to allowing sufficient drying out of the concrete slab. Because there is such a range of concrete mix designs and jobsite conditions, there cannot be a uniform cure time specified that will result in sufficiently dry concrete. In order to avoid these issues, the NRCA recommends ASTM F2170, Standard Test Method for Determining Relative Humidity in Concrete Floor Slabs Using in situ Probes,20 humidity probe testing and states that 75% relative humidity is the default threshold for installing roofing but that may be lower for some types of roofs. To illustrate the role that initial moisture content can have on the long-term performance of a roof assembly, one dimensional, transient hygrothermal models using WUFI Pro 6.2 were developed for the following roof assembly: • Thermoplastic polyolefin (TPO) membrane (assumed white) • 0.5-in. (12.7 mm) gypsum cover board • 4-in. (101.6 mm) polyisocyanurate insulation • 0.1-perm vapor retarder • 4-in. (101.6 mm) concrete deck • Non-perforated corrugated steel deck The roof was modeled using Baltimore, MD, as the exterior climate. No moisture sources were included. The interior climate was assumed to vary as a sine curve between 20 and 22.2 C (68 and 72 F), and the interior relative humidity was assumed to vary as a sine curve between 30 to 60%. Multiple iterations of the analysis were conducted, each time changing only the initial moisture content. Figure 3 shows the temperature (red) and relative humidity (green) as a function of time at the midpoint of the insulation. Figure 3A shows the results when ASHRAE 160 initial moisture contents were assumed: two times EMC80 for all materials except concrete, which was assumed to have two times EMC90. From this graph, it can be seen that the relative humidity is anticipated to remain above 80% for prolonged
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
FIG. 3 (A) Temperature and relative humidity at midpoint of insulation assuming ASHRAE 160 initial moisture contents (two times EMC80 for all materials except concrete at two times EMC90). (B) Temperature and relative humidity at midpoint of insulation assuming ASHRAE 160 initial moisture contents but EMC75 assumed for concrete.
periods and reaches 96%. This can create issues with corrosion, adhesion loss, decreased insulation performance, or other failures. Figure 3B shows the results of the same analysis, except that the concrete initial moisture content was taken as EMC75 as recommended by the NRCA. This graph shows drastically reduced humidity levels and a drying trend over the years. By the third summer, the
139
140
STP 1617 On Building Science and the Physics of Building Enclosure Performance
humidity will remain within limits recommended to reduce the potential for corrosion and other failures. Similarly, wall assemblies with limited drying potential can have their longterm performance impacted by the initial moisture content of the assembly. Walls that feature vapor retarders—particularly in mixed humid climates—or materials that will function as a vapor retarder even if not intended as such, such as vinyl wall coverings, have limited drying potential. To illustrate this point, a transient, onedimensional hygrothermal analysis was conducted on the following wall assembly using WUFI Pro 6.2: • 3-5/8-in. (92.1 mm) brick veneer • 1-in. (25.4 mm) air space • 1.5-in. (38.1 mm) rock wool insulation • 0.1-perm air/water barrier • 5/8-in (15.9 mm) exterior gypsum sheathing • 6-in. (152.4 mm) fiberglass batt insulation • 0.05-perm interior facer of batt insulation (typical of foil-scrim-kraft [FSK]) • 0.5-in. (12.7 mm) interior painted drywall (paint assumed to have vapor resistance of 8 perms) The wall was modeled using Baltimore, MD, as the exterior climate. No moisture sources were included. The interior climate was assumed to vary as a sine curve between 20 and 22.2 C (68 and 72 F), and the interior relative humidity was assumed to vary as a sine curve between 30 and 60%. The analysis was run multiple times, each time changing only the initial moisture content. Figure 4A shows the temperature and relative humidity at the inside face of the batt insulation against the interior vapor retarder when the initial moisture content was assumed per the ASHRAE 160 recommendation of two times EMC80 for all materials. The result showed prolonged periods with a high potential for corrosion and condensation. When the analysis was run again with all the same inputs but half the initial moisture content (assuming EMC80 for all materials), the assembly was predicted to have no potential for mold growth, corrosion, or condensation, as can be seen in figure 4B.
Field Data Collection and Model Calibration Assumptions regarding initial moisture content can have a dramatic impact on the long-term performance of envelope assemblies, at least as shown in the computer simulation of these envelope assemblies presented in the preceding section. Without field data to validate the computer simulation, the models represent a theoretical risk assessment for unfavorable long-term performance. A field study, originally aimed at examining the potential for condensation within exterior wall assemblies containing a vapor retarder in mixed humid climates,21 found that the assumptions regarding the initial moisture content within the computer modeling were critical in validating the simulation results.
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
FIG. 4 (A) Temperature and relative humidity at midpoint of insulation assuming ASHRAE 160 initial moisture contents (two times EMC80). (B) Temperature and relative humidity at midpoint of insulation assuming half of ASHRAE 160 initial moisture contents (EMC80 for all materials).
Part of this study compared recorded field data to “blind” WUFI analysis conducted using generic materials and the default inputs and assumptions outlined in ASHRAE 160. However, even when the blind WUFI models were revised to include the recorded temperature and relative humidity as the interior climate and the recorded exterior conditions as the exterior climate, the results and conclusions of the WUFI models did not match the actual recorded temperature and relative
141
142
STP 1617 On Building Science and the Physics of Building Enclosure Performance
humidity recorded across the wall assembly. Only when the assumption about the initial moisture content was reduced did the WUFI results match the recorded data well. In other words, it was not until accurate initial moisture contents were included in the computer models that the modeling outputs matched the field recorded data. The following section recaps the portions of the findings of the prior field study that related to the role that initial moisture input played in the accuracy of the WUFI results. Additional details about the study can be found in Wagner, Cyphers, and Whitlock.21
Outline of Field Study
Eight months of field data were collected at the northeast wall of an occupied office building located in Manassas, VA. The wall assembly consisted of: • 3-5/8-in. (90.1 mm) brick veneer • 2-in. (50.8 mm) air space • Building felt as the weather-resistive barrier • Spun bonded polyolefin air barrier • 0.5-in. (12.7 mm) fiberglass-faced gypsum sheathing • R-19 foil-faced fiberglass batt insulation within the 6-in. (152.4 mm) stud cavity with the studs spaced 16 in. (406.4 mm) on center • Painted interior gypsum wall board The wall assembly and the general field setup are shown schematically in figure 5. A total of ten temperature and humidity data loggers were installed at the site on June 19. These loggers were set to record the temperature and the relative humidity every hour. The loggers were left in place until February 27 of the following year, allowing the extreme times of both the summer and winter to be captured. Two sets of loggers were installed across the depth of the wall assembly. Each set was installed at a midpoint between studs. There were four data loggers in each set, and they were installed as follows: (1) against the exterior surface of the building felt, (2) at the interior side of the sheathing, (3) between the foil facer and the interior side of the batt insulation, and (4) on the outer surface of the interior gypsum. The ninth logger, an exterior grade logger, was installed on the exterior surface of the brick veneer. The final logger was suspended from the ceiling at the center of the room to determine the interior ambient conditions within the office space. Additionally, a weather station was installed to record the hourly exterior weather just outside of the location where the loggers and the thermocouples had been installed. The hourly ambient temperature, relative humidity, and barometric pressure were recorded over the entire data collection period, beginning on June 29, when the weather station was fully installed and operational. The weather station recorded all necessary data for creating a WUFI climate file including a rain gauge that reported hourly precipitation, an anemometer that recorded average hourly wind speed and wind direction, and a radiation sensor that measured total
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
FIG. 5 Schematic of wall assembly and field data collection setup.
radiation. This sensor was mounted on the face of the brick so that the total radiation on the brick surface could be recorded. Measuring the direct total radiation does not provide the level of detail needed (a separation of direct radiation versus diffuse) to develop a radiation rose. As such, the climate file written, based on the collected weather data, could only be applied to the location of the field study and could not be extrapolated to other building elevations and heights, which could have significantly different radiation. Presentation of Field-Collected Data Figure 6A–F consists of a series of graphs depicting the field-recorded temperatures
and relative humidity at locations indicated in each caption. Based on the results
143
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 6 (A) Recorded temperature and relative humidity at the exterior surface of the brick veneer. (B) Recorded temperature and relative humidity at the exterior face of the building felt (logger part of Data Logger Set 1). (C) Recorded temperature and relative humidity at the exterior face of the building felt (logger part of Data Logger Set 2). (D) Recorded temperature and relative humidity at the interior face of the exterior sheathing (logger part of Data Logger Set 1). (E) Recorded temperature and relative humidity at the interior face of the exterior sheathing (logger part of Data Logger Set 2). (F) Recorded temperature and relative humidity at exterior face of the foil facer of the insulation (logger part of Data Logger Set 1). (G) Recorded temperature and relative humidity at exterior face of the foil facer of the insulation (logger part of Data Logger Set 2). (H) Recorded temperature and relative humidity at the interior face of the interior gypsum wall board (logger part of Data Logger Set 1). (I) Recorded temperature and relative humidity at the interior face of the interior gypsum wall board (logger part of Data Logger Set 2). (J) Recorded temperature and relative humidity at the interior ambient conditions. (A)
Temperature Humidity
Recorded Exterior Surface
120 110 100 90
F/%RH
80 70 60 50 40 30 20 10 0 9/17/12
Time
8/18/12
7/19/12
6/19/12
(B)
Temperature Humidity
Recorded Exterior Face of Building Felt (1)
110 100 90 80 70 F/%RH
144
60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
FIG. 6 (continued) Temperature
(C)
Recorded Exterior Face of Building Felt (2)
Humidity
110 100
Logger baery power lost
90 80
F/%RH
70 60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time Temperature
(D)
Recorded Interior Face of Sheathing (1)
Humidity
110 100 90 80
F/%RH
70 60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time
Temperature
(E)
Recorded Interior Face of Sheathing (2)
Humidity
110 100 90 80
F/%RH
70 60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time
145
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 6 (continued) Temperature
(F)
Recorded Exterior Face of Foil (1)
Humidity
110 100 90 80
F/%RH
70 60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time
Temperature
(G)
Recorded Exterior Face of Foil (2)
110
Humidity
100 90 80
F/%RH
70 60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time Temperature
Recorded Interior Surface of Gypsum (1)
(H)
Humidity
110 100
Temporary logger malfuncon
90 80 70 F/%RH
146
60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
FIG. 6 (continued) Temperature
(I)
Humidity
Recorded Interior Surface of Gypsum (2)
110 100 90 80
F/%RH
70 60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time
Temperature
(J)
Recorded Interior Ambient
110
Humidity
100 90 80
F/%RH
70 60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time
presented in the figure, it can be seen that no condensation was recorded at any locations. Furthermore, no location within the wall assembly was recorded to have a 30-day running average relative humidity above 80% during the collection period. The collected data conflicted with the results of blind WUFI analysis conducted using the default assumptions of ASHRAE 160 (refer again to Wagner, Cyphers, and Whitlock21 for more information on the default assumption analysis). There are obvious reasons that explain the differences, namely:
147
148
STP 1617 On Building Science and the Physics of Building Enclosure Performance
1. The blind WUFI analysis used a climate reference year for the exterior climate, which did not match the data recorded by the weather station. 2. The blind WUFI analysis used a sine curve for the interior climate, which did not match the data recorded by the interior data loggers. Thus, the following revisions were made to the blind WUFI models: 1. A weather file was created from the data collected from the weather station, and this weather file was used in place of the climate reference file for the exterior climate. 2. The hourly ambient temperature and humidity recorded within the office space with the tenth data logger was used to create an interior climate file, and that file was used in place of the default model interior climate. The results of the WUFI simulation using the field-collected weather files are presented in figure 7A–E. When figure 7 is compared to the analogous figure 6, it is clear that accounting for the actual weather resulted in the simulated performance matching very closely to the recorded performance. There were only a couple of quantitative differences between the simulated and recorded performance as outlined here: • Generally, the daily fluctuations in the recorded data were greater than the simulated fluctuations in the data. This is likely due to heat and moisture transport via air movement throughout the wall assembly, which were not taken into account in the model. However, these daily fluctuations did not change the overall trend and ranges of the data. The general trends, longterm performance, and conclusions still translated well between the recorded and simulated data except as noted within the next bullet. • From the middle of June through the middle of August, the simulated performance, when accounting for the actual recorded interior and exterior climates, showed prolonged periods of humidity in excess of 80% at the exterior face of the interior vapor retarder (fig. 7D), which does not comply with the ASHRAE 160 recommended limits to minimize the potential for corrosion. However, the recorded data at this depth in the wall did not show conditions that would constitute a failure under ASHRAE 160 criteria. The recorded 30-day running average humidity remained below 80% at this location as shown in figure 6F and 6G. Even when the actual weather is accounted for, the simulated conditions at the vapor retarder were not a close match to the recorded conditions, specifically at the vapor retarder. The initial moisture contents of the materials within the wall assembly had been assumed to be two times the expected moisture content at 80% relative humidity in accordance with ASHRAE 160 for the simulation with results shown in figure 7. The field study was conducted approximately five years after construction of the office building was complete. As such, these initial moisture contents were not an appropriate and accurate representation of the actual
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
FIG. 7 Predicted temperature and relative humidity at (A) the exterior surface of the brick veneer, (B) the exterior face of the building felt, (C) the interior face of the exterior sheathing, (D) the exterior face of the foil facer of the insulation, and (E) the interior face of the interior gypsum wall board, as per ASHRAE 160 when collected weather data were used for interior and exterior climate.
Temperature
Predicted Exterior Surface
(A) 120
Humidity
110 100 90
F/%RH
80 70 60 50 40 30 20 10 0 9/17/12
8/18/12
7/19/12
6/19/12
Time
Temperature
(B)
Predicted Exterior Face of Building Felt
Humidity
110 100 90 80
F/%RH
70 60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time
149
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 7 (continued) Temperature
(C)
Predicted Interior Face of Sheathing
Humidity
110 100 90 80
F/%RH
70 60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time
Temperature
Predicted Exterior Face of Foil
(D)
Humidity
110 100 90 80 70 F /% R H
150
60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
FIG. 7 (continued) Temperature
(E)
Predicted Interior Surface of Gypsum
Humidity
110 100 90 80
F/%RH
70 60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time
initial moisture contents at the time the data collection began. To account for more representative initial moisture contents, the average humidity over the first week of analysis was found at each data logger location. This average humidity was then used to set the initial moisture content of that material within the models. The results of the simulated performance accounting for a more accurate initial moisture content are presented in figure 8A–E. This change resulted in the simulated performance even more closely matching the recorded data. The conditions at the exterior face of the interior vapor barrier were reduced when the initial moisture contents were reduced. The simulated humidity at this location within the summer was still slightly higher than the recorded humidity at this location but would not have significant impacts on the overall conclusions regarding the hygrothermal performance of the wall assembly. This slight decrease in recorded humidity as compared to the simulated at this point in the wall is likely the result of an imperfection in the installed vapor retarder or air movements throughout the stud cavity that can dissipate this moisture (or both).
Conclusions and Recommendations The construction industry often uses terms such as moisture, wet, damp, and dry. However, the industry does not seem to have quantitative definitions for these terms. Aside from oven-dried conditions, there will always be moisture present in
151
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 8 Predicted temperature and relative humidity at (A) the exterior surface of the brick veneer, (B) the exterior face of the building felt, (C) the interior face of the exterior sheathing, (D) the exterior face of the foil facer of the insulation, and (E) the interior face of the interior gypsum wall board, as per ASHRAE 160 when collected weather data was used for interior and exterior climate and initial moisture contents were reduced. Temperature
(A)
Predicted Exterior Surface with Reduced MC i
120
Humidity
110 100 90 80 F/%RH
70 60 50 40 30 20 10 0 9/17/12
8/18/12
7/19/12
6/19/12
Time
Temperature
Predicted Exterior Face of Building Felt Reduced MCi
(B)
110
Humidity
100 90 80 70 F/%RH
152
60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
FIG. 8 (continued) Temperature
(C)
Predicted Interior Face of Sheathing Reduced MC i
110
Humidity
100 90 80
F/ % R H
70 60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time Temperature
(D)
Predicted 2" Into Ba Reduced MC i
110
Humidity
100 90 80
F/%RH
70 60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time
Predicted Interior Surface of Gypsum with Reduced MC i
(E)
Temperature Humidiity
110 100 90 80
F/%RH
70 60 50 40 30 20 10 0 2/14/13
1/15/13
12/16/12
11/16/12
10/17/12
9/17/12
8/18/12
7/19/12
6/19/12
Time
153
154
STP 1617 On Building Science and the Physics of Building Enclosure Performance
most building materials because of their hygroscopic nature. The impact of humidity on a material’s moisture content has been explained. Excessive humidity (greater than 85%) can have almost as much impact on a material’s moisture content as liquid water. Further, a material’s moisture content impacts many material properties and can reduce the material’s useable service life. The initial moisture content of materials can have a dramatic impact on the long-term performance of certain materials in certain assemblies. With respect to hygrothermal simulation, ASHRAE 160 provides recommendations for default initial moisture contents of materials. Based on the references stated in the standard, these default assumptions may be taken out of context. Based on an array of field collected data, it appears that the default assumptions for initial moisture content may be unreasonably high for some materials and assemblies. Additional field studies should be conducted to specifically examine the role that initial moisture content has on actual performance. The field studies should be coupled with computer-simulated hygrothermal analysis to validate the findings. Until the default recommended assumptions for initial moisture content assumptions are further validated, hygrothermal models should be undertaken for a range of initial moisture contents in the case of new construction. Conducting iterative hygrothermal analyses with a range of initial moisture content assumptions will provide the designer with an appreciation of how sensitive the assembly is to initial moisture content, and if needed, additional requirements regarding storing materials, acceptable material moisture content, and construction sequencing should be specified. Otherwise, alternative assemblies that are more forgiving to a range of initial moisture contents should be designed. For retrofit and renovation projects, whenever possible, actual field recorded data regarding the ambient relative humidity should be used for initial moisture content assumptions for hygrothermal analysis. The recorded relative humidity should be used in conjunction with sorption-isotherm curves to arrive at the most realistic initial moisture content assumptions.
References 1. 2.
3.
H.R. Trechsel, ed., Moisture Analysis and Condensation Control in Building Envelopes (West Conshohocken, PA: ASTM International, 2001). Standard Test Method for Hygroscopic Sorption Isotherms of Building Materials, ASTM C1498-04a (2016) (West Conshohocken, PA: ASTM International, approved August 15, 2016). http://doi.org/10.1520/C1498-04AR16 Standard Test Method for Moisture Retention Curves of Porous Building Materials Using Pressure Plates, ASTM C1699-09 (2015) (West Conshohocken, PA: ASTM International, approved May 1, 2015). http://doi.org/10.1520/C1699-09R15
WAGNER AND CYPHERS, DOI: 10.1520/STP161720180070
4.
5. 6.
7.
8. 9. 10. 11. 12.
13.
14.
15.
16. 17.
18.
19.
20.
Standard Test Methods for Water Vapor Transmission of Materials, ASTM E96/E96M-16 (West Conshohocken, PA: ASTM International, approved March 1, 2016). http://doi.org/ 10.1520/E0096_E0096M-16 American Society of Heating, Refrigerating and Air-Conditioning Engineers, ASHRAE Handbook—Fundamentals (Atlanta, GA: ASHRAE, 2013): 26.17–26.18. APA–The Engineered Wood Association, Moisture-Related Dimensional Stability, Technical Topics TT-028C (Tacoma, WA: APA–The Engineered Wood Association, May 2016). APA–The Engineered Wood Association, Wood Moisture Content and the Importance of Drying in Wood Building Systems, Technical Topics TT-111B (Tacoma, WA: APA–The Engineered Wood Association, November 2016). Handling and Storage of Gypsum Panel Products: A Guide for Distributors, Retailers, and Contractors, GA-801-2017 (Hyattsville, MD: Gypsum Association, 2016). North American Insulation Manufacturers Association, The Facts About Mold Growth, Insulation Facts #72 (Alexandria, VA: NAIMA, November 2004). Gypsum Association, Guidelines for Prevention of Mold Growth on Gypsum Board, GA238-2016 (Hyattsville, MD: Gypsum Association, 2016). Gypsum Association, Assessing Water Damage to Gypsum Board, GA-231-2015 (Hyattsville, MD: Gypsum Association, 2015). APA–The Engineered Wood Association, Service Life of Oriented Strand Board (OSB) Sheathing, Technical Topics TT-052C (Tacoma, WA: APA–The Engineered Wood Association, November 2013). APA–The Engineered Wood Association, Wood Moisture Content and the Importance of Drying in Wood Building Systems, Technical Topics TT-11B (Tacoma, WA: APA–The Engineered Wood Association, November 2016). Standard Guide for Characterization and Use of Hygrothermal Models for Moisture Control Design in Building Envelopes, ASTM E3054/E3054M-16 (West Conshohocken, PA: ASTM International, approved March 15, 2016). http://doi.org/10.1520/ E3054_E3054M-16 Criteria for Moisture-Control Design Analysis in Buildings, ANSI/ASHRAE 160-2016 (Atlanta, GA: American Society of Heating, Refrigerating and Air-Conditioning Engineers, 2016). International Energy Agency, Annex XIV: Condensation and Energy, Volume 2: Guidelines & Practice (Paris: IEA, August 1990). Standard Guide for Evaluation and Rehabilitation of Mass Masonry Walls for Changes to Thermal and Moisture Properties of the Wall, ASTM E3069-19 (West Conshohocken, PA: ASTM International, approved February 1, 2019). http://doi.org/ 10.1520/E3069–19 R. A. Cyphers, A. M. Cyphers, J. M., Knorowski, and A. M. Skertic, “Evaluation of Existing Moisture Content in Brick for Hygrothermal Models in Rehabilitation of Mass Masonry Walls,” in Masonry 2018, ed. N. Krogstad and W. McGinley (West Conshohocken, PA: ASTM International, 2018): 1–12. http://doi.org/10.1520/STP161220170167 M. S. Graham, Moisture in Concrete Roof Decks: Normal-Weight and Lightweight Structural Concrete Cause Some Concern (Washington, DC: National Roofing Contractors Association, September 2017). Standard Test Method for Determining Relative Humidity in Concrete Floor Slabs Using in situ Probes, ASTM F2170-19 (West Conshohocken, PA: ASTM International, approved March 1, 2019). http://doi.org/10.1520/F2170-19
155
156
STP 1617 On Building Science and the Physics of Building Enclosure Performance
21.
C. M. Wagner, R. A. Cyphers, and A. R. Whitlock, “Evaluation of the Potential for Corrosion, Mold Growth, and Moisture Accumulation within Typical Brick Veneer Wall Assemblies Designed per 2006 International Energy Code in a Mixed Humid Climate,” in Building Walls Subject to Water Intrusion and Accumulation: Lessons from the Past and Recommendations for the Future, ed. J. Erdly and P. Johnson (West Conshohocken, PA: ASTM International, 2014): 1–67. http://doi.org/10.1520/STP154920130051
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180088
Travis V. Moore1 and Michael A. Lacasse1
Approach to Incorporating Water Entry and Water Loads to Wall Assemblies When Completing Hygrothermal Modelling Citation T. V. Moore and M. A. Lacasse, “Approach to Incorporating Water Entry and Water Loads to Wall Assemblies When Completing Hygrothermal Modelling,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 157–176. http://doi.org/10.1520/STP1617201800882
ABSTRACT
The long-term performance in respect to moisture management of wood-frame wall assemblies depends on the hygrothermal response of the wall to local climate loads. Critical factors in estimating the longevity of wood-frame structures include limiting the temperature range, wood moisture content, and time of exposure to conditions suitable for the onset, growth, and propagation of mold and rot to occur. Increasingly, hygrothermal simulation tools are being used to estimate the risk of formation of condensation and the presence of moisture from inadvertent water entry to a wall that, in turn, may bring about the formation of mold or wood rot to wood-frame assemblies. More recently, standards have evolved to permit systematically undertaking performance evaluations of wood-frame walls, such as ASHRAE S160, Criteria for MoistureControl Design Analysis in Buildings, and ASTM E3054, Standard Guide for the Characterization and Use of Hygrothermal Models for Moisture Control Design in Building Envelopes. Defined within these methods are approaches to include water entry loads to the wall assembly. In this paper, a review of different methods to incorporate water entry and water loads to wood-frame wall
Manuscript received October 17, 2018; accepted for publication July 28, 2019. 1 NRC-Construction, Building Envelope & Materials Division, 1200 Montreal Rd., Building M24, Ottawa, ON, https://orcid.org/0000-0002-4920-9193, M. A. L. https://orcid.org/0000-0001K1A 0R6 T. V. M. 7640-3701 2 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by the National Research Council of Canada. ASTM International, 100 Barr Harbor Drive, PO Box C700, Copyright V West Conshohocken, PA 19428-2959.
157
158
STP 1617 On Building Science and the Physics of Building Enclosure Performance
assemblies is given and an approach is provided that is consistent with assessing the long-term performance of wall assemblies following ISO 13823, General Principles on the Design of Structures for Durability. Examples are used to demonstrate the validity of the proposed approach and compare the outcomes when using existing standards. Keywords durability, hygrothermal modeling, moisture load, wall performance, water entry, water loads, wind-driven rain, standards
Introduction Assessment of long-term performance of building assemblies can be determined in several ways. For a specific location, experience can be used based on past performance of building envelope assemblies and components such as the prescriptive solutions in the National Building Code of Canada.1 Long-term testing can also be used in which a representative building envelope assembly is subjected to the real local climate for a long duration (one to several years) and performance can be monitored.2,3 These strategies can be used effectively assuming that knowledge of long-term performance is only required for the specific climate area in question, and that one assumes the climate on which the opinion is based is representative of future climate scenarios. Generally, performance for building envelope components and assemblies is sought for a variety of climate locations and building types. To assess performance of building envelope components and assemblies for a variety of climate locations’ hygrothermal simulation can be utilized. Hygrothermal simulations, or heat air and mass transfer simulations, are computer models that represent the physical nature of moisture and heat transport in materials. The hygrothermal equations used in the computer model balance the heat and moisture transport through building materials when those materials are exposed to interior (indoor) and exterior (climate) boundary conditions for a specific climate location. Conducting a hygrothermal simulation for a building component requires knowledge of several factors of the climate location of interest and materials in the building envelope being investigated. These factors include: material properties of building envelope components, climate conditions, and indoor conditions to which the building envelope is exposed. Standards such as ASHRAE 160,4 ISO 15927,5 and ASTM E3054, Standard Guide for the Characterization and Use of Hygrothermal Models for Moisture Control Design in Building Envelopes,6 provide information on the material property, climate, and indoor conditions used to conduct hygrothermal performance assessments. For hygrothermal assessments, the parameters required to be defined for the material properties consist of the sorption/desorption isotherm (moisture retention function), liquid conductivity, water vapor permeability, thermal conductivity, heat capacity, density, and—in some cases—the air permeability. The interior boundary
MOORE AND LACASSE, DOI: 10.1520/STP161720180088
conditions require definition of the relative humidity and temperature of the indoor air, whereas the exterior climate parameters consist of hourly values of solar radiation, temperature, relative humidity, rainfall, wind speed, and wind direction. An important parameter that is also required for hygrothermal simulations is the rain load that impinges on the exterior cladding surface, that is, the wind-driven rain (WDR) load. This can be determined through several methods,5,7,8 using the hourly rainfall rate, wind speed, and wind direction combined with the wall orientation under consideration. The value of wind-driven rain is of particular importance because it is used to determine the water entry rate through the cladding system to within the wall assembly toward more moisture-vulnerable portions of the wall, and it is considered one of the most significant climate loads when considering long-term hygrothermal performance of wall assemblies.4,9 Water entry can then occur through the cladding itself, if it is hygroscopic, or through deficiencies that occur at through-wall penetrations such as vent pipes, electrical outlets, and windows (or both).10,11 Although conceptually simple, applying water entry due to wind-driven rain in a hygrothermal simulation is not straightforward. The reason for this is that common hygrothermal computer codes (e.g., WUFI,* DELPHIN,{ and COMSOL{) do not model liquid droplets acting on a wall; liquid moisture is represented as a “moisture load” within the layers of a wall assembly. To further complicate the implementation of moisture to the simulation, it has also been demonstrated12,13 that small air gaps between the cladding and the sheathing membrane in a wall assembly can exhibit significant drainage of any water entry that potentially occurs beyond the exterior cladding surface. Although the drainage efficiency of a wall assembly is of some importance, more significance should be given toward the water retained in the wall assembly after drainage has occurred such that one can determine whether the presence of this quantity of moisture load to the wall is detrimental to the long-term durability of the wall assembly. This implies that to properly represent the moisture load in a hygrothermal simulation, water entry due to wind-driven rain in combination with information on the drainage characteristics of the wall assembly should be used to determine the water retention rate. The water retention rate can then be used to calculate the moisture source, as this value is the moisture load that ought to be applied to the wall system due to wind-driven rain events. In this paper, a method is described from which the quantity of moisture within a cladding system arising from the effects of wind-driven rain may be used as a “moisture load” for hygrothermal simulations from experimental results. This method is dependent on two independent experiment test methods. The first test method is used to determine the percentage of water entry that can be expected
*
Fraunhofer IBP, WUFI(R) Pro, 2018. Bauklimatik-dresden, DELPHIN, Dresden, Germany: Bauklimatik-dresden, 2018. COMSOL AB, COMSOL Multiphysics (R), Stockholm, Sweden: COMSOL AB, 2018.
{ {
159
160
STP 1617 On Building Science and the Physics of Building Enclosure Performance
beyond the cladding in a wall assembly due to various conditions of wind-driven rain. The second test method considers the water retention rate of the cladding system for different quantities of water entry to the wall assembly after drainage has occurred. The combination of these methods is used to determine the water entry retention rate of the cladding, which is dependent on the wind-driven rain load (driving rain deposition rate and driving rain pressure) as determined from climate data for a location of interest when undertaking hygrothermal simulations. Using these methods, the moisture load for a National Building Code (NBC) of Canadacompliant vinyl-clad wall assembly is determined for three climate locations in Canada having varying degrees of local wind-driven rain loads and drying potential and is then compared to that obtained using criteria for determining the moisture load as described in ASHRAE 160.4
Description of Test Methods A description is provided of test methods used to determine the following: (i) Water entry characteristics of the cladding system due to wind-driven rain (ii) Water retention characteristics of the cladding system considering water entry and drainage of the cladding system In this paper, the water entry test method is briefly described and previously published examples are referenced. The drainage test, and methods by which the water entry and drainage test results can be used to define a moisture source for hygrothermal simulations, is described in more detail. WATER ENTRY CHARACTERISTICS OF THE CLADDING SYSTEM
To determine the rate of water entry beyond the cladding in a wall assembly, several methods can be utilized including, for example, in situ water entry testing,14 laboratory water entry testing,10,11,15 and long-term monitoring of site-installed claddings.16,17 For the purposes of determining the water entry rate as a function of wind-driven rain loads (rain deposition rate on the cladding surface and driving rain wind pressure), the method used for the results in this paper followed that described by Lacasse et al.11 The water penetration testing described by Lacasse et al.11 considers water entry characteristics for a wall assembly measuring 8 ft by 8 ft (2.44 m by 2.44 m), containing a 2-ft-by-2-ft (600 mm by 600 mm) window, a 4-in. (100 mm) diameter round pipe, and a standard exterior electrical outlet box. The wall assembly is subjected to water deposition rates of 1.0, 2.0, and 3.4 L/min-m2. For the wall assembly presented in this paper, the wall (an assembly configured for a cold climate) consists of, from the exterior to the interior layers: vinyl cladding, air space (less than 1/8 to 7/16-in., about 3 mm to 11 mm), sheathing membrane (i.e., water-resistive barrier; e.g., asphalt-impregnated kraft paper), 3/8-in. (9.5 mm) clear polycarbonate sheathing (replicating the exterior sheathing panel), 2-in. by 6-in. (50 mm by 300 mm) wood (spruce-pine-fir [SPF]) studs spaced at 24 in. (610 mm) on center, and 3/8-in. (9.5 mm) clear polycarbonate sheathing
MOORE AND LACASSE, DOI: 10.1520/STP161720180088
(replicating the interior sheathing panel and acting as the air/vapor barrier) having an air leakage rate of 2.36 ft3/h-ft2 (0.2 L/s-m2 at 75 Pa). The use of clear polycarbonate sheathing to replicate the exterior sheathing, interior sheathing, and the air/ vapor barrier in the wall assembly enable the test operator to visually observe water entry beyond the sheathing membrane or into the stud cavity. Previous to water entry testing, the wall assembly was conditioned to simulate mechanical loads during service life and the air leakage was measured and adjusted to be 2.36 ft3/h-ft2 (0.2 L/s-m2 at 75 Pa) by drilling holes in the clear polycarbonate interior sheathing board. The testing without deficiencies consisted of testing the NBC-compliant vinyl test specimen in an “as built” condition and with the caulking in place around all cladding penetrations (i.e., window, pipe duct, and electrical outlet). The second test scenario, testing with deficiencies, consisted of implementing defects in the caulking at the cladding penetrations and, thereafter, completing the water entry tests (fig. 1). The deficiencies implemented in the caulking products at the cladding penetrations consisted of: i. Boring a 3-mm (1/8-in.) diameter hole at the top center of the caulking around the pipe ii. Boring a 3-mm (1/8-in.) diameter hole at the bottom left corner of the caulking around the electrical outlet iii. Boring a 3-mm (1/8-in.) diameter hole at the bottom corner of the window frame to simulate failure of a window frame
FIG. 1 Water collection troughs.
161
162
STP 1617 On Building Science and the Physics of Building Enclosure Performance
The water entry test method was used to determine the water entry rate for a wall system as a function of the wind-driven rain (water deposition rate and driving rain wind pressure). This relationship is used to establish the expected hourly water entry rate into a wall system for input to a hygrothermal simulation given the hourly wind-driven rain climate data for the geographic location of interest. WATER RETENTION CHARACTERISTICS OF A WALL ASSEMBLY
Similar to determining water entry, several methods exist for determining the drainage and retention characteristics of a clad wall assembly, including both laboratory testing12,13,18 and in situ testing.19 In all tests, a controlled amount of water is injected into the drainage space between the cladding and the sheathing membrane. Through laboratory testing, it has been demonstrated that, for even small drainage spaces (less than 1 mm), drainage can occur with high efficiency in wall assemblies.12,13 However, in these tests, and ASTM E2273, Standard Test Method for Determining the Drainage Efficiency of Exterior Insulation and Finish Systems (EIFS) Clad Wall Assemblies,18 the quantity of water injected into the drainage cavity typically is greater than that expected in the field. For example, during ASTM E2273 testing, between 7,950 g and 8,725 g of water is deposited into the drainage space over 75 min. For the study by Van Linden, Lacasse, and Van Den Bossche,13 769 g of water was deposited into the drainage space during the test. Straube and Smegal12 followed ASTM E2273 and correspondingly injected 8 L of water into the drainage space over 75 min. Straube and Schumacher7 conducted an analysis on driving rain and wind following an extensive review of Canadian weather data. They found that the average driving-rain deposition rate for all monitored Canadian cities was 0.012 L/min-m2 and that the 1% extreme driving-rain events ranged from 0.05 L/min-m2 to 0.17 L/ min-m2. For the test specimens selected in ASTM E2273 (4 ft by 8 ft, 1219 mm by 2438 mm) this corresponds to a total wind-driven rain load on average of 2 L/h and extreme 1% values ranging from 8 to 30 L/h. If it is expected that, conservatively, 5% of the wind-driven rain load bypasses the cladding (this is five times that which is estimated from ASHRAE 1604 and 2.5 times that which was noted by Olsson16), then for a 4-ft–by-8-ft (1219 mm by 2438 mm) specimen, the 1% extreme water entry should not exceed 1.5 L/h (approximately 1,500 g in 1 h). On average, this water entry rate would be 0.1 L/h (approximately 100 g in 1 h). Straube and Smegal12 also determined that the same maximum amount of water retention was observed if the amount of water injected into the drainage space was halved to 4 L. Van Linden, Lacasse, and Van Den Bossche13 noted that the drainage rate of the systems tested in their work reached steady state after approximately 100 s, corresponding to approximately 128 g of water being retained in the system. Given that the system likely has a maximum retention quantity (i.e., saturation of the cavity), and from these two studies for which different size specimens were tested, the maximums were demonstrated to have been attained. These values would represent the upper limits of water entry retention in the drainage
MOORE AND LACASSE, DOI: 10.1520/STP161720180088
cavities of each respective wall system. This suggests that the drainage efficiency of a wall assembly can be altered depending on the quantity of water injected into the drainage cavity. This implies that, for the purposes of undertaking long-term durability assessments using hygrothermal simulations, the quantity of water retained in the system for different water entry rates is of more importance than the drainage efficiency given that the retained water is the water “load” the wall system must manage to avoid the risk to the initiation of degradation mechanisms, as may occur from the presence of mold on wall components, decay of wood-based elements due to wood rot, and corrosion of metal components. The purpose of the water retention test described in this paper is to attempt to determine a method by which the water retention of a wall system can be defined for a variety of water entry events, including water entry events that are in quantities lower than the maximum retention of the drainage space, which perhaps more closely represents water entry events in the field. Water Retention Apparatus
The test specimen was evaluated by depositing water into the cavity between the cladding and sheathing membrane of the wall assembly using a water deposition tube and through which water was evenly distributed across the width of the specimen at the uppermost opening of the test specimen drainage cavity. The water deposition tube consisted of a 12.7-mm-diameter copper tube in which evenly spaced (at 50 mm) 0.5-mm-diameter holes had been bored along its length. The test specimen was dosed with water that was ejected under pressure from the tube onto an angled metal tray that was used to redirect the water into the space between the cladding and the sheathing membrane (drainage cavity). The use of a metal tray allowed the water to enter the drainage cavity with minimal kinetic energy, thereby permitting the drainage process to be regulated purely by gravitational forces. The deposition tube was supplied water by a centrifugal pump (Stenner, Model 85M4, 0–10 L/h, 6 2%) that was calibrated to provide a dosage rate of 6 L/h. The water supply to the pump was from a plastic container; the weight of the container was measured gravimetrically during the test sequence. Water that was dosed to the test specimen and that subsequently drained from the system was collected in a separate container that was also monitored gravimetrically during the test. The gravimetric measurements were provided by a balance scale (Mettler, PM30000-K, 0–30 KG, 6 0.5 g). A photo of the test apparatus is provided in figure 2. Water Retention Test Specimen Configuration
The test specimen for water retention, likewise provided in figure 2, measures 4 ft wide by 6 ft high (1219 mm wide by 1828 mm high); a vertical sectional view of this specimen is provided in figure 3. These specimens were constructed to resemble the field of cladding assembly and therefore included no penetrations. The specimens were constructed to include all details of the wall assembly from the sheathing board outward; therefore, there is no stud cavity or interior sheathing board.
163
164
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 2 NBC vinyl cladding drainage specimen and drainage test apparatus: (1) water dosage container, (2) centrifugal dosage pump, (3) water deposition tube and redirection tray, (4) test specimen, and (5) water collection container.
Water Retention Test Method
The test method consisted of determining the relationship between the quantities of water entering (water entry) the space between the cladding and the sheathing membrane (drainage cavity) and the corresponding quantity of water retained in the drainage cavity. This was achieved by water dosage to the drainage cavity at specific intervals of time with the use of the water deposition tube and redirection tray. The time intervals that were used in the drainage test are given in Table 1. The selection of specific time intervals for dosage, as given in Table 1, were intended to permit capturing a high degree of resolution of the quantity of water retained during the early part of the test and a reduced resolution toward the end of the test. Prior testing had shown that the greatest degree of change in the quantity of water retained in drainage assemblies occurred at the lower dosage rates, with the greatest quantities retained at the lowest dosage rate. These findings corresponded to those of Van Linden, Lacasse, and Van Den Bossche.13 This effect is increasingly lessened as the water dosage rate to the drainage cavity is increased.
MOORE AND LACASSE, DOI: 10.1520/STP161720180088
FIG. 3 Test specimen configuration—vertical cross section.
The time over which the test was conducted was determined to ensure that 8 L of water was dosed to the drainage cavity by the end of the test (6 L/h for 82 min greater than or equal to 8 L, accounting for 6 2% pump repeatability). This was completed such that end results of retention could be compared to those from ASTM E2273 if required.19 The test sequence consisted of: (1) weighing the water dosage and collection containers, (2) initiating the pump for the duration of the time interval, and (3) stopping the pump and measuring the corresponding change in weight of the dosage and collection containers. The retention quantity could then be calculated as the difference between the quantity (weight) of water dosed to the drainage cavity and the water collected in the container. The rate of evaporation of water for each water container was also taken into consideration. This was accomplished by filling a container of the same size as the
165
166
STP 1617 On Building Science and the Physics of Building Enclosure Performance
TABLE 1 Drainage test time increments Time Interval (min.)
0
Cumulative Test Time (min.)
0
0.5
0.5
1
1.5
1.5
3
2
5
2.5
7.5
3
10.5
3.5 4
14 18
4.5
22.5
5
27.5
5.5
33
6
39
6.5
45.5
7
52.5
7.5
60
22
82
dosage and collection containers with water and thereafter measuring the loss in weight due to evaporation as might occur in laboratory conditions over the duration of a typical test sequence. The evaporative losses were then averaged over the test time to account for evaporation at each time increment.
Results and Discussion The results of the water entry test protocol and the water retention test protocol for a vinyl-clad wall assembly are provided in this section. WATER ENTRY TEST RESULTS
The water entry testing on the NBC-compliant vinyl-clad wall assembly consisted of two test conditions: water entry testing without deficiencies and water entry testing with deficiencies. The results for each deposition rate are given in figure 4, figure 5, and figure 6. The results for normalized water entry as a percentage of WDR (L/min-m2) are presented in figure 7. In the test condition where no deficiencies were incorporated in the caulking at the cladding penetrations, the water entry test of the NBCC-compliant vinyl-cladding assembly resulted in 0 ml of water entry under any tested deposition rate or applied pressure. For the deficiency condition, testing resulted in water entry at the windowsill location only. The other deficiencies did not cause any water entry to occur, which was likely due to deposited water not encountering any of the deficiencies.
MOORE AND LACASSE, DOI: 10.1520/STP161720180088
FIG. 4 Water entry results for NBC vinyl-clad wall assembly with deficiencies for deposition rate of 1.0 L/min/m2.
FIG. 5 Water entry results for NBC vinyl-clad wall assembly with deficiencies for deposition rate of 2.0 L/min/m2.
167
168
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 6 Water entry results for NBC vinyl-clad wall assembly with deficiencies for deposition rate of 3.4 L/min/m2.
FIG. 7 Water entry (percentage of WDR) as a function of applied pressure for a typical window deficiency.
MOORE AND LACASSE, DOI: 10.1520/STP161720180088
TABLE 2 Water entry percentage for vinyl-clad wall with deficiencies Water Entry as Percentage of Water Deposition Rate 1.0 L/min-m
2.0 L/min-m2
3.4 L/min-m2
0
0.17%
0.01%
0.31%
0.16%
50
1.45%
0.24%
0.88%
0.85% 0.97%
Applied Pressure (Pa)
2
Average
75
1.22%
0.58%
1.09%
150
2.16%
0.45%
0.64%
1.08%
300
1.18%
0.57%
1.23%
0.99%
500
3.31%
0.64%
1.05%
1.67%
700
2.26%
1.13%
1.72%
1.70%
Additionally, the results presented in Figures 4 to 7 indicate that, for higher water deposition rates (2.0 and 3.4 L/min-m2), there is a link between the applied pressure and the water entry rate. However, for the low water deposition rate (1.0 L/ min-m2), there does not seem to be a link between the applied pressure and the water entry rate. This is likely caused by the low water deposition rate not supplying a consistent rate of water to the deficiency location, causing the water entry rate to be more dependent on whether or not water is present at the deficiency. For the higher water deposition rates, increased water is present at the deficiency location; therefore, the water entry is more dependent on the applied pressure. Table 2 shows the water entry as a percentage of wind-driven rain for each applied pressure tested. Comparing the results from the water entry testing at a window deficiency presented in figure 7 and Table 2 with the method of assuming water entry is equal to 1% of wind-driven rain from ASHRAE 1604 shows reasonable agreement, with 1% being an overestimate for wind pressures below 150 Pa and an underestimate for water entry for wind pressures above 150 Pa. WATER RETENTION TEST RESULTS
The drainage test results are presented as percentage of water retained in the cladding system due to a given water entry quantity normalized in area and in time. The water retention test results for the NBC vinyl-clad wall assembly are shown in figure 8, and observations during the drainage test show the presence of water on the exterior surface due to weep holes in the vinyl cladding. A picture of one weep hole in the NBCC vinyl cladding is shown in figure 9. The water retention results show that there is a relationship between the water entry quantity and the retention rate of the vinyl-clad wall assembly drainage space up to a certain threshold. The results indicate that, for small water entry events (less than 0.024 L/min-m2, the initial value measured in the experiment), retention rates are high (greater than or equal to 50%), whereas, for large water entry events (greater than 0.5 L/min-m2), the retention rate quickly reduces to less than 10%.
169
170
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 8 Water retention test results for NBC vinyl-clad wall assembly.
FIG. 9 NBC-compliant vinyl cladding. Weep holes drain water from the cladding.
MOORE AND LACASSE, DOI: 10.1520/STP161720180088
WATER RETENTION AS A FUNCTION OF WIND-DRIVEN RAIN
Using the experimental data to determine the water entry percentage as a function of wind-driven rain, and the water retention as a function of the water entry is known, the amount of water retained by the cladding as a function of wind-driven rain for each cladding type can be determined. These functions will form inputs for the hygrothermal simulation evaluations of each cladding type. Table 3 presents the vinyl-clad wall water entry percentage of WDR as a function of pressure (P). It also presents the same function for the deficiency types selected in this evaluation. Table 4 presents the function for the total water entry for the vinyl-clad wall assembly, which is the combination of the water entry through the cladding and the water entry through the deficiencies. Table 5 presents the retention percentage functions (RET%) for the vinyl-clad wall assembly as a function of the total water entry. The function for retention of the vinyl-clad wall assembly is a piecewise function, with a linear interpolation between the first point in the experiment and 100% and, following that, a power law relationship for the remaining values. Finally, combining the water entry functions and the water retention function gives the water retention function as a function of wind-driven rain and wind induced pressure for the vinyl-clad wall assembly (Table 6).
TABLE 3 Water entry functions (WE%) for cladding type and incorporating deficiencies WE% = a*Pb
a
NBC vinyl-clad wall Vinyl-clad wall with deficiencies
R2
b
0
0
-
1.79E-03
3.51E-01
0.96
(average values used for correlation)
TABLE 4 Total water entry (WE) for cladding WEtotal = (WEcladding% þ WEdeficifences%)*WDR
WEVinyltotal = (0 þ a2*(Pb2))*WDR
NBC vinyl cladding
TABLE 5 Percentage of water retention in a vinyl-clad wall assembly as a function of the percentage of water entry to the wall NBC Vinyl Cladding
R2
RET%
For WE 0.024 L/h-m
2
For WE > 0.024 L/h-m2
RET% = 19.858*WEVinyltotal þ 1
0.96
RET% = 0.0411*WEVinyltotal0.621
0.96
TABLE 6 Retention percentage of vinyl siding as a function of WDR For WE 0.024 L/h-m2
RET% = 19.858*(a2*(Pb2))*WDR þ1
2
RET% = 0.0411*(a2*(Pb2))*WDR 0.621
For WE > 0.024 L/h-m
171
172
STP 1617 On Building Science and the Physics of Building Enclosure Performance
The retention function provides information for input to hygrothermal simulations that denotes the quantity of water dosing the wall system for a given rain event. As presented in this paper, this function represents the quantity of water remaining in the cladding system after drainage has occurred from water entry through the cladding and deficiencies in the wall. Additionally, as the RET% accounts for both WE and drainage as a function of WDR, the values obtained from this correlation can be directly compared to the ASHRAE 160 1% WDR value. COMPARISON OF RESULTS WITH ASHRAE 160
The resulting equation for the percentage of water retention is used to determine the moisture load for durability assessments using hygrothermal simulations. The use of the equation for determining the percentage of water entry of a vinyl-clad wall assembly as presented in this paper is compared with the industry standard of 1% moisture load (i.e., 1% of WDR load to the wall) recommended by ASHRAE 160 for three coastal locations in Canada having high wind-driven rain loads. The climate data in each location have been used to calculate the wind-driven rain loads in accordance with ISO 15927.5 For this comparative case, the wind-driven rain load is calculated as the airfield hourly index and does not account for wall factors. The climate locations selected are Halifax, NS (fig. 10), St John’s, NL (fig. 11), and Vancouver, BC (fig. 12). Plots of the respective moisture loads (1% WDR, retention method) for each of these locations are depicted in figures 10 through 12, and are converted into units of
FIG. 10 Moisture load comparison for a vinyl-clad wall assembly in Halifax, NS.
MOORE AND LACASSE, DOI: 10.1520/STP161720180088
FIG. 11 Moisture load comparison for a vinyl-clad wall assembly in St. John’s, NL.
FIG. 12 Moisture load comparison for a vinyl-clad wall assembly in Vancouver, BC.
kg/(m3-s), which corresponds to the moisture flux load used in hygrothermal simulations. As can be noted from these plots, for all three cities, the 1% WDR method in ASHRAE 160 provides substantially larger loads than that obtained using the
173
174
STP 1617 On Building Science and the Physics of Building Enclosure Performance
TABLE 7 Average, maximum, and minimum values for moisture load by city and calculation method Location (City)
Calculation Method
Halifax
St. John’s
Vancouver
Average
Maximum
Minimum
Retention
1.86E-02
6.05E-02
7.20E-05
1% WDR
4.79E-02
8.12E-01
2.59E-04
Retention
8.75E-03
3.00E-02
1.65E-04
1% WDR
4.95E-02
8.55E-01
3.93E-04
Retention
6.23E-03
4.25E-02
4.45E-05
1% WDR
1.45E-02
4.99E-01
2.62E-04
experimental methods for the retention percentage as a function of WDR described earlier. To provide a broad overview of the results, the average, minimum, and maximum moisture load values for each city and moisture load calculation method are presented in Table 7. From Table 7, it can be seen that there is a significant difference between the experimental retention method of determining moisture load and the 1% method described in ASHRAE 160 for large, wind-driven rain events. For Halifax, the 1% moisture load is, on average, 2.6 times that of the retention method. For St. John’s, this number increases to 5.7 times, and for Vancouver, it is 2.3 times. The difference among the minimum values is much less significant, with Halifax and Vancouver displaying similar results regardless of methods, and the 1% method in St. John’s having 2.9 times the moisture load as the retention method. The maximum water entry scenarios show that the moisture load using the 1% method is an order of magnitude greater than that using the retention method. Overall, these results seem to correspond to the results noted in the retention method. Specifically, the retention drainage method indicates that, for the smaller wind-driven rain events, and therefore smaller water entry events (corresponding to the minimum values from Table 7), the retention percentage is high (fig. 8), so the majority of the water entry is retained in the system. For the higher wind-driven rain and water entry events (corresponding to the maximum values in Table 7), the retention drainage method indicates that a larger percentage of the water drains (fig. 8); therefore, less of the water from the entry event is retained in the system.
Conclusions This paper presented a method using experimental data for water entry testing and retention drainage tests on a representative vinyl-clad wall assembly to determine moisture loads for hygrothermal simulations that accounts for both water entry and drainage characteristics of the wall assembly. The moisture loads calculated using the experimental results were compared to the standard value of 1% WDR as specified in ASHRAE 160.4
MOORE AND LACASSE, DOI: 10.1520/STP161720180088
Overall, as has been noted in other publications,12,13 small gaps between the cladding and sheathing membrane can show significant drainage capabilities. However, the relationship was expanded upon in this paper, demonstrating that the drainage efficiency of the cladding is directly related to the rate of water entry. Using the ASHRAE 160 method to determine water entry as 1% of the winddriven rain load corresponded reasonably well with the failed window deficiency (1/8-in. or 3.1 mm, hole drilled into the corner of the window frame) tested in the water entry experiments of the vinyl-clad wall assembly. It should be noted, however, that this type of deficiency is a failure of window condition and is a conservative assumption for field installations. However, comparing the 1% of the WDR estimate as a moisture load with experiment results that account for drainage aspects of a vinyl-clad wall assembly demonstrated that 1% is a significant overestimate as a moisture load for hygrothermal simulations. The results of ASHRAE 160 and the experiment methods were compared for three high-moisture-load coastal cities in Canada—Halifax, NS; St. John’s, NL; and Vancouver, BC. This overestimate is especially significant (an order of magnitude higher) in situations with high wind-driven rain events (and correspondingly high water-entry events). The results of this paper indicate that drainage and retention characteristics of cladding systems should be accounted for when determining moisture loads for hygrothermal simulations, especially for vinyl-clad wall assemblies.
References 1. 2.
3.
4. 5.
6.
7.
National Research Council Canada, National Building Code of Canada (Ottawa, ON: National Research Council Canada, 2015). M. Bartko, M. Lacasse, G. Ganapathy, and M. Nicholls, Evaluation of Thermal and Moisture Response of Highly Insulated Wood-Frame Wall Assemblies: Part I, Experimental Trials in the Field Exposure of Walls Test Facility (Ottawa, ON: National Reserach Council Canada, 2016). W. Maref, M. Armstrong, H. Saber, M. Swinton, G. Ganapathey, M. Nicholls, K. Abdulghani, and M. Rosseau, Field Exposure of Walls Facility (Ottawa, ON: National Research Council Canada, 2011). Criteria for Moisture-Control Design Analysis in Buildings, ASHRAE 160-2016 (Atlanta, GA: American Society of Heating, Refrigerating and Air-Conditioning Engineers, 2016). Hygrothermal Performance of Buildings—Calculation and Presentation of Climatic Data—Part 3: Calculation of a Driving Rain Index for Vertical Surfaces from Hourly Wind and Rain Data, ISO 15927-3:2009 (Geneva, Switzerland: International Standards Organization, 2009). Standard Guide for Characterization and Use of Hygrothermal Models for Moisture Control Design in Building Envelopes, ASTM E3054/E3054M-16 (West Conshohocken, PA: ASTM International, approved March 15, 2016), http://doi.org/10.1520/E3054_E3054M-16 J. Straube and C. Schumacher, Driving Rain Loads for Canadian Building Design (Ottawa, ON: CMHC, 2006).
175
176
STP 1617 On Building Science and the Physics of Building Enclosure Performance
8.
9.
10. 11.
12.
13.
14.
15.
16.
17.
18.
19.
B. Blocken and J. Carmeliet, “A Review of Wind-Driven Rain Research in Building Science,” Journal of Wind Engineering and Industrial Aerodynamics 92, no. 13 (2004): 1079–1130. K. Carbonez, N. Van Den Bossche, H. L. G. Ge, and A. Janssens, “The Spell Definition in ISO-15927 and Its Impact on the Rain Deposition on the Building Facade,” Energy Procedia 78 (2015): 2548–2553. L. Olsson, “Rain Intrusion Rates at Facade Details—A Summary of Results from Four Laboratory Studies,” Energy Procedia 132 (October 2017): 387–392. M. Lacasse, T. O’Connor, S. Nunes, and P. Beaulieu, Report from Task 6 of MEWS Project: Experimental Assessment of Water Penetration and Entry into Wood-Frame Wall Specimens—Final Report (Ottawa, ON: National Research Council Canada, 2003). J. Straube and J. Smegal, “The Role of Small Gaps Behind Wall Claddings on Drainage and Drying,” in Proceedings of the 11th Canadian Conference on Building Science and Technology (Banff, Alberta: British Columbia Building Envelope Council, 2007). S. Van Linden, M. Lacasse, and N. Van Den Bossche, “Drainage and Retention of Water in Small Drainage Cavities: Experimental Assessment,” in Proceedings of the 7th International Building Physics Conference (Syracuse, NY, International Association of Building Physics, 2018): 91–96. Standard Test Method for Field Determination of Water Penetration of Masonry Wall Surfaces, ASTM C1601-14a (West Conshohocken, PA: ASTM International, approved July 1, 2014), http://doi.org/10.1520/C1601-14A M. A. Recatala, “Proposal for Assessing a New Test Methodology for Assessing the Performance of Rear-Ventilated Facades against Wind-Driven Rain (WDR) and Driving Rain Wind Pressures (DRWP)” (PhD diss., Universteit Ghent, 2017). L. Olsson, “Long-Term Field Measurements of Moisture in Wooden Walls with Different Types of Facades: Focus on Driving Rain Tightness,” Energy Procedia 78 (November 2015): 2518–2523. E. Ngudjiharto, F. Tariku, and P. Fazio, “Preliminary Results from Field Experimental Study of Rain Load and Penetration into Wood-Frame Wall Systems at Window Sill Defects,” in Proceedings of the 14th Canadian Conference on Building Science and Technology (Toronto, Ontario: Ontario Building Envelope Council, 2014): 257–266. Standard Test Method for Determining the Drainage Efficiency of Exterior Insulation and Finish Systems (EIFS) Clad Wall Assemblies, ASTM E2273-18 (West Conshohocken, PA: ASTM International, approved September 1, 2018), http://doi.org/10.1520/E2273-18 Standard Test Method for Evaluation of Water Leakage Performance of Masonry Wall Drainage Systems, ASTM C1715/C1715M-15 (West Conshohocken, PA: ASTM International, approved July 1, 2015), http://doi.org/10.1520/C1715_C1715M-15
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180085
Elyse A. Henderson,1 Graham Finch,2 and Brian Hubbs1
Solutions to Address Osmosis and the Blistering of Liquid Applied Waterproofing Membranes Citation E. A. Henderson, G. Finch, and B. Hubbs, “Solutions to Address Osmosis and the Blistering of Liquid Applied Waterproofing Membranes,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 177–194. http://doi.org/10.1520/STP1617201800853
ABSTRACT
The transport of water through waterproofing membranes over concrete substrates resulting in water-filled blisters and leaks has been demonstrated by the authors to be caused by osmosis. Although this issue has now been studied for more than a decade, there is currently no industry standard to test for the risk of osmosis in waterproofing membranes. The authors have developed a protocol to measure the osmotic flow and evaluate the risk of osmotic blistering in waterproofing membranes including a standardized osmotic flow rate test, ASTM E96, Standard Test Methods for Water Vapor Transmission of Materials, inverted wet cup vapor permeance testing, and modified ASTM long-term absorption testing. This testing protocol has measured osmotic flow rates and ASTM E96 inverted wet cup vapor permeance for a range of different waterproofing membrane types. The authors propose that this set of testing protocols or another proxy test be adopted by ASTM to determine the risk for osmosis, including thresholds above which a membrane may be deemed “high risk.” To reduce the potential for osmotic blistering over concrete, it is recommended that waterproofing membranes used in inverted roofing
Manuscript received October 15, 2018; accepted for publication April 11, 2019. 1 RDH Building Science Inc., 4333 Still Creek Dr. #400, Burnaby, British Columbia, Canada, V5C 6S6 https://orcid.org/0000-0002-2079-4974 E. A. H. 2 RDH Building Science Inc., 740 Hillside Ave. #602, Victoria, British Columbia, Canada, V8T 1Z4 3 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
177
178
STP 1617 On Building Science and the Physics of Building Enclosure Performance
assemblies should have an osmotic flow rate near 0.0 g/m2/day when tested using the proposed protocol, an ASTM E96 inverted wet cup vapor permeance less than that of the substrate (i.e., less than 0.1 US perms on a concrete slab), and minimal long-term water absorption uptake. In addition to these thresholds, the long-term aging effects of the membrane should be determined. Keywords waterproofing, membrane, osmosis, absorption, vapor permeance testing procedure
Introduction Waterproofing membrane failures due to osmotic blistering have occurred in some types of protected membrane/inverted concrete substrate waterproofing and roofing assemblies and have been observed in many parts of the world over the past two decades. Water blisters have been observed between the waterproof membrane and the concrete substrate and are often under considerable pressure. These selfcontained, pressurized water blisters have no identifiable leakage path through or around the membrane. Blisters have ranged from millimeters to an entire roof area and contain significant quantities of water under pressure. Due to the hidden nature of a membrane within an inverted roofing assembly, the issue can go unnoticed for some time until other, more visible problems result from the large quantities of water held by the blisters. For example, large blisters have displaced concrete pavers, creating hazardous walking conditions, and water leaks have occurred as the blisters expand over cracks and joints in the concrete substrate. The issue is not exclusive to horizontal surfaces; blisters have also been observed on vertical applications in planter boxes and water features. Issues have not been observed with membranes over steel or wood substrates or in conventional roofing assemblies where the membrane typically is exposed or directly applied to insulation or a cover board substrate. Several years ago, the authors set out to understand the cause of this phenomenon. Hygrothermal analysis shows that vapor diffusion can transport water through membranes with relatively high vapor permeance such as asphalt-modified polyurethane. Even though the membranes that were found to have the most water blisters also typically had high vapor permeance, the quantity of water found in the blisters is orders of magnitude higher than what can be expected from vapor permeance alone. Additionally, the water vapor pressure differences on either side of the membrane are not great enough to explain the high hydrostatic pressure that exists inside the blisters. The hypothesis that was later confirmed with further research was that the physical process of osmosis was drawing water through the semipermeable waterproofing membranes such as asphalt-modified polyurethane.1,2 The solute concentration under the membrane (i.e., from the concrete/substrate or from the membrane itself) was measured and confirmed to be high enough to generate extremely high osmotic pressure within the blisters—up to 1,500 kPa (15 bar).3
HENDERSON ET AL., DOI: 10.1520/STP161720180085
Not all waterproofing membranes are at risk for developing osmotic blisters. For example, blistering has not been observed for hot rubber or styrene butadiene styrene (SBS) membranes, which are very common in this application. Research by the authors into the systematic failure of asphalt-modified polyurethane waterproofing membranes in the Pacific Northwest has led to the development of a testing procedure to estimate the risk for osmotic blistering of various other waterproofing membranes. The internally developed testing methodology has been used to measure osmotic flow through—and risk for osmotic blistering of—dozens of different waterproofing membranes including SBS, hot rubberized asphalt, poly(methyl methacrylate) (PMMA), ethylene propylene diene monomer (EPDM), thermoplastic polyolefin (TPO), high density polyethylene (HDPE), polyurea, asphalt emulsion, asphalt-modified polyurethane, and various other twocomponent cold-applied membranes. There is currently no industry-wide, standardized test to assess the risk of osmotic blistering of waterproofing membranes. The authors recommend that the testing methodology described in this paper be adopted by the ASTM Committees E06 and D08, including recommended target maximums for tested osmotic flow rate, ASTM E96, Standard Test Methods for Water Vapor Transmission of Materials,4 inverted wet cup vapor permeance, and a modified ASTM procedure for long-term water absorption. WHAT IS OSMOSIS?
The process of osmosis can be described as the flow of a solvent, usually water, across a semipermeable membrane from a solution of low solute concentration to a solution of high solute concentration. This is possible when the membrane separating the two solutions is slightly permeable to water yet impermeable to the solutes. Thus, the water flows across the membrane to balance out the solute concentrations on either side of the membrane, as shown in figure 1. If the vessel is open such as in the diagram in figure 1, the water level on one side rises until the hydrostatic pressure equals that of the osmotic pressure as defined by equation (1): p¼uCRT
(1)
where: p = osmotic pressure (bar, SI unit of pressure), u = osmotic coefficient (unitless, value that characterizes the dissolution of the individual salts in solution), C = concentration of all dissociated solutes (mol/L where mol is the standard unit of measurement for an amount of substance), R = universal gas constant (0.083145 Lbar/molK), and T = temperature (Kelvin, absolute measure of SI temperature equal to Cþ273). Osmotic pressure is a colligative property, meaning that the property depends on the concentration of the solutes and not on their identity. In other words, the osmotic flow across a system with 1.0 M sodium chloride (NaCl) as the solute is the
179
180
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 1 Diagram showing the process of osmosis and reverse osmosis. Water flows across a semipermeable membrane to dilute a high concentration of solutes on the other side (osmosis) until the hydrostatic head equals the osmotic pressure (equilibrium). Water can be forced back across the membrane by application of pressure (reverse osmosis).
same as an identical system with 1.0 M potassium iodide (KI) or a 1.0 M mixture of dissolved solids that come off a concrete slab when water is trapped within a blister below a waterproof membrane. Typical solutes that have been measured in osmotic blisters from inverted roof assemblies include calcium, carbonate, magnesium, potassium, sulfur, and silicon from the concrete substrate and, in many cases, organic compounds from the membranes themselves.3
Methodology Currently, there is no industry standard to evaluate the osmotic risk of waterproofing membranes. The authors have shown in previous research that vapor permeance and water absorption are often related to the osmotic flow potential of waterproofing membranes.1 As such, the proposed protocol includes both ASTM E96 wet cup and inverted wet cup vapor permeance testing and long-term water absorption testing in addition to the internally developed osmotic flow rate methodology. The proposed osmosis test protocol includes three parts: 1. Osmotic flow rate testing (method developed by the authors) 2. Water absorption testing (method adapted from ASTM D570, Standard Test Method for Water Absorption of Plastics,5 for a longer time frame) 3. Vapor permeance testing (by wet cup and inverted wet cup per ASTM E96) OSMOTIC FLOW RATE TEST
The test methodology to measure osmotic flow rate has been refined by the authors over the past decade. A solution of 1.0 M NaCl (equivalent to total dissolved solids [TDS] of 58,500 ppm) is placed in a glass container; the test membrane is cut to fit the top of the container and sealed to separate the salt water inside the container
HENDERSON ET AL., DOI: 10.1520/STP161720180085
from the distilled water in a water bath outside the container. The apparatus is designed so the osmotic flow of water from the freshwater side to the saltwater side is easily measured by the mass increase within the container. The test containers are removed for gravimetric measurements taken at regular intervals, and the osmotic flow of water into the container (g/m2/day) is calculated after some baseline adjustments. The following procedure was used to measure the osmotic flow rate through waterproofing membranes: 1. The samples of membrane are cut into circular discs to fit within a powdercoated, corrosion-resistant, screw-top lid fitting of glass containers. Each membrane sample is initially weighed, and the thickness is measured at a minimum of five points to determine an average thickness. 2. A known volume (approximately 80 mL in a 250-mL glass jar) of salt water is poured into the glass containers. The salt water is typically 1.0 M NaCl, but other salts (to represent specific solute conditions at a site) and varying concentrations (to represent different osmotic pressures) may be tested. • Triplicate blank samples (with distilled water, 0.0 M NaCl) and triplicate test samples (1.0 M NaCl) should be produced for each membrane type being tested. Triplicate samples are used as a balance between having enough samples to see if there are outliers due to jar leaks and few enough samples to manage in a small lab space and minimize measurement time. 3. Two component fast-set epoxy is applied to the perimeter of both sides of the cut membrane disk to create a sealed gasket between the lid and membrane and the glass container. Once the lid with membrane has been sealed to the container, the system is let to set for 24 h. 4. After the epoxy has cured, the container is leak tested by placing it upside down on an absorbent material such as shop towel for 24 h, and checking the absorbent material for signs of water, which may indicate a slow leak through the epoxy seal. Samples that show signs of leakage should be rejected from the test. 5. The initial mass of the container, membrane, and salt water together is measured using a scale with at least 0.01 g precision. 6. A freshwater bath is prepared using distilled water, and the samples are placed inverted on a rack in the bath to allow for fresh water under the containers where the membrane is located. • The water bath should be filled to a level so that the waterline is approximately level with the internal waterline of the inverted sample containers to eliminate the effect of hydrostatic water head on the samples. 7. At regular intervals (typically weekly), the containers and blank samples are removed from the freshwater bath, dried thoroughly using a durable and absorbent material such as a shop towel, and weighed using a scale with
181
182
STP 1617 On Building Science and the Physics of Building Enclosure Performance
0.01 g precision. This process is repeated approximately once a week for several months. 8. Throughout the experiment, the TDS levels in the water bath are monitored regularly using a TDS meter, and water is changed when the TDS is elevated above 10 ppm, or approximately every two weeks, to maintain fresh water outside the sample containers. 9. The flow of water through the membrane is estimated by subtracting the mass of the initial container from each of the measurements throughout the experiment. There are also baseline corrections that may be carried out (i.e., sample containers with distilled water). Pros and cons of each baseline correction option are discussed later in this paper. The duration of the osmosis flow rate experiment is typically four to eight months and may be shorter for membranes that have high osmotic flow rates or longer for membranes that are very impermeable to water (fig. 2). The sample containers in the osmotic flow test undergo an initial wetting process during which they absorb water, although this does not necessarily contribute to water permeating through the material into the osmotic cell. The internally developed osmosis test procedure decouples the test container wetting from actual osmotic flow into the container by carrying out simultaneous absorption testing and control samples to subtract out this effect. In addition to blank samples with no salts (i.e., containers with distilled water), control samples with impermeable metal lids sealed with the same epoxy may be added to the experiment and their small weight gain throughout the experiments can be subtracted from the sample osmotic measurements.
FIG. 2 Diagram of an individual sample container in water bath for osmotic flow experiment. If the membrane is semipermeable, fresh water can flow into the jar of saline water over the course of the four- to eight-month experiment due to the process of osmosis.
HENDERSON ET AL., DOI: 10.1520/STP161720180085
LONG-TERM ABSORPTION TEST
Water absorption measurements are part of the authors’ proposed standard osmosis testing procedure for two main reasons: a) To understand the long-term effects of contact with liquid water The absorption of water by waterproofing membranes can change their properties over time. Water absorption can dissolve some components of the material over time as well as loosen the adhesion of layers including fiber reinforcement. These changes to the chemical and physical properties of membranes can lead to decreased performance and failures in the field. Very high moisture absorption rates have been shown to fail waterproofing membranes on their own without osmosis occurring due to swelling, reemulsification, softening, or material degradation.6,7 b) To calibrate the osmosis results Most membranes in the osmosis experiments go through a wetting process during which they absorb water, although this does not necessarily contribute to water permeating through the material. The osmosis test procedure developed decouples these two processes. The absorption testing procedure generally follows industry standard water absorption tests (immersion of sample in room temperature water), yet for a longer time than most procedures specify. The 24-h moisture absorption specified in various ASTM standards (including ASTM D570 for plastics) is insufficient to accurately assess the long-term moisture uptake of a waterproofing membrane in an inverted roofing application that can be installed for 30 to 40 years. Longer-term testing is important as long-term water absorption into a waterproofing membrane may affect its durability and material properties (e.g., vapor permeance, susceptibility to osmosis, material strength). 1. The samples of membrane are cut into circular discs of similar size as for the concurrent osmotic flow rate test. Each membrane sample is initially weighed, and the thickness is measured at a minimum of five points to determine an average thickness. 2. Membrane samples are placed on a grate within a clean waterproof container. The grate serves to prevent membranes from sticking to the base of the container and to maintain water flow underneath the membranes. 3. The container is filled with fresh water until membrane samples are just covered with water (approximately 5 mm above samples). If the membranes float, a light grate is placed on top to maintain membrane submergence in the water bath. 4. At regular intervals (typically weekly), the membrane samples are removed from the freshwater bath, dried gently using a durable and absorbent material such as a shop towel, and weighed using a scale with at least 0.01 g precision. This process is repeated approximately once a week for several months, on the same days as gravimetric measurements for the concurrent osmotic flow test.
183
184
STP 1617 On Building Science and the Physics of Building Enclosure Performance
As part of the proposed standard osmosis testing protocol, the water uptake and moisture content of a membrane is measured for the duration of the concurrent osmotic flow rate test, typically four to eight months. VAPOR PERMEANCE TEST
The third and final piece of the proposed standard osmosis test procedure is vapor permeance testing in general conformance with ASTM E96 for both wet cup and inverted wet cup vapor permeance measurements. Distilled water is placed within a glass container, and the material being tested is sealed on top such that it separates the interior of the cup to the controlled relative humidity (RH) conditions of a climate chamber. The vapor pressure gradient created between the water in the cup (100% RH) and climate chamber conditions (50% RH) results in the moisture leaving the cup through the test material. Wet cup measurements typically are recommended over dry cup for describing the in-service properties of water resisting barrier sheathing membranes as they are exposed to high RH levels for most of the year. The inverted wet cup test is different from the regular wet cup method in that it inverts the standard wet cup apparatus in the climate chamber and exposes the top side of the membrane to liquid water (see diagram in fig. 3). The average RH the material sees in this case is the same as the wet cup; however, liquid water and capillary flow are present. Inverted wet cup measurements are recommended for the in-service properties of waterproofing membranes that are in contact with liquid water for significant periods of time, especially those used in protected membrane/inverted roofing assemblies, although this is currently not a common industry practice (typically dry cup measurements are used for product specifications as they generate lower permeance results). The change in mass of the apparatus is measured using a laboratory scale, and the resulting vapor permeance is calculated from the loss of mass over time per unit area of material (see ASTM E96). The preparation of the samples for the wet cup and inverted wet cup tests follows the same steps as sample preparation for the
FIG. 3 Wet cup (left) and inverted wet cup (right) vapor permeance test procedures per ASTM E96.
HENDERSON ET AL., DOI: 10.1520/STP161720180085
osmosis measurements, although the water inside the containers is distilled water instead of salt water as in the case of the osmosis test samples.
Results and Discussion OSMOSIS PROTOCOL TESTING RESULTS
The osmotic flow testing protocol developed and refined by the authors has been used to estimate the risk for osmotic blistering of waterproofing membranes in the field. Although at first the concern was specifically for two-component cold-applied membranes, a wide range of membrane types has been tested including other coldapplied polymer membranes and hot rubber and torched SBS.1–3,8 A summary of some of these results is shown in figure 4. Waterproofing membranes may be grouped into different levels of risk for osmosis to occur in the field using osmotic flow rates determined through the proposed testing protocol. In figure 4, high-risk membranes are shown with solid lines and typically have osmotic flow rates between 5 and 20 g/m2/day or higher. Membranes that fall into this high-risk category are typically two-component, coldapplied systems such as asphalt-modified polyurethanes (especially aged samples) and asphalt emulsions or other semipermeable materials. Membranes that fall into the medium-risk category (shown with dashed lines in the figure) typically have osmotic flow rates between 2 and 4 g/m2/day and include some new asphaltmodified polyurethanes and polyureas. These medium-risk membranes may have unknown long-term performance in the field due to aging and degradation
FIG. 4 Graph of osmotic flow test results from a subset of experiments over the past decade using the authors’ proposed testing protocol. The membranes determined to be at high risk for osmotic blistering in the field are shown with solid lines. The medium-risk membranes are with dashed lines, and the low-risk membranes are with dotted lines.
185
186
STP 1617 On Building Science and the Physics of Building Enclosure Performance
(discussed later). Membranes that typically test near 0.0 6 1.0 g/m2/day are deemed to be low risk and have not been known to exhibit osmotic blistering in the field (shown with dotted lines in the figure). These low-risk membranes include hot rubber, SBS, PMMA,* EPDM, TPO, and HDPE sheet (60 mil). To minimize the risk for osmotic blistering in the field, the authors propose that a testing protocol such as the one presented in this paper be adopted by ASTM for waterproofing membranes. It is recommended that an osmotic flow rate—tested using this methodology—not exceed 0.0 6 1.0 g/m2/day for a membrane to be considered low risk and acceptable for application in areas with high moisture and water exposure such as inverted or protected roof assemblies. VAPOR PERMEANCE AND MEMBRANE AGING
The past decade of osmosis testing by the authors has revealed some trends regarding which membranes perform better or worse in the osmotic flow rate test. Two key membrane characteristics have been linked to high osmotic flow rates: vapor permeance and aging. At first, a trend with membrane thickness was discovered, though with further experiments it became clear that the correlation of membrane thickness with osmotic flow rate was in fact linked to the difference in vapor permeance. Thicker membranes generally have lower vapor permeance. Since the process of osmosis hinges on the permeability of the membrane for water transport, this result further corroborates the findings that water-filled blisters are due to osmosis. Graphs showing the correlation of osmotic flow rate with both membrane thickness and membrane permeance are provided in figure 5 and figure 6. A more consistent correlation with permeance (rather than with thickness) is observed. The trend of higher osmotic flow rate for membranes with higher vapor permeance makes sense because the process of osmosis depends on the membrane being semipermeable to water yet impermeable to dissolved solids (e.g., salts, organic compounds). The higher the permeance of the membrane, the more easily water may be transported across it. Another key finding from the osmosis testing by the authors is the uncertainty in performance of aged membranes. Figure 7 shows the osmotic flow rates and vapor permeance of three sets of asphalt-modified polyurethane membranes. As seen in figure 7, both aged membranes (shown in triangles) perform worse than the lab-cured new membrane (shown in open diamonds), which indicates membrane degradation and increase in vapor permeance of the membranes over time. These results illustrate that a new membrane may perform adequately in testing, although a membrane that has been installed in the field for five to ten years (i.e., from Sites 1 and 2 in the figure) may have significantly different properties. Even between two sites, from which asphalt-modified polyurethane membrane is recovered and tested, results can vary
*Although PMMA membranes are listed as low risk in these osmotic flow rate results, some PMMA membranes have been tested to have high long-term water absorption and may have unknown long-term aging impacts.
HENDERSON ET AL., DOI: 10.1520/STP161720180085
FIG. 5 Osmotic flow rate determined by the testing protocol developed by the authors versus membrane thickness. High-, medium-, and low-risk membranes are shown with triangles, circles, and squares, respectively. The correlation between thinner membranes and higher osmotic flow rates is further explained by the difference in membrane permeance as shown in figure 6.
FIG. 6 Osmotic flow rate determined by the testing protocol developed by the authors versus vapor permeance determined through inverted wet cup testing per ASTM E96. High-, medium-, and low-risk membranes are shown with triangles, circles, and squares, respectively.
187
188
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 7 Osmotic flow rate determined by the testing protocol developed by the authors versus vapor permeance determined through inverted wet cup testing per ASTM E96 for three sets of asphalt-modified polyurethane membranes, two aged membranes recovered from sites, and one unaged lab-cured membrane.
significantly. This illustrates that the locations and conditions in which membranes age may influence how much their properties change over time. The unknown impact of aging in different locations and conditions (i.e., Site 1 versus Site 2 in fig. 8) highlights the importance of long-term testing for degradation of membranes over time. When membranes are exposed to liquid water, ultraviolet (UV) light, salts, metals, or organic substances over long periods of time, the properties of the membranes may be altered. For example, a membrane that has been in contact with water throughout an eight-month osmosis experiment, as described in this research methodology, has been shown to release organic compounds into the adjacent water. To quantify the amount of organic material that can be released from membranes in contact with liquid water over time, the water within a sample container from an osmosis experiment was collected and sent to a third-party analytical laboratory. The measurable degradation of membranes that are submerged in water for even eight months cannot be ignored. Most membranes are tested for compliance with ASTM protocols as new membranes or after simulated aging for a limited amount of time. The results here show that some membranes may be losing some of their mass into the surrounding water over longer periods of time. This effect is currently being studied through ongoing experiments by the authors.* It is important to consider
* RDH Building Science, Inc., has nearly one dozen membranes currently in a membrane soaking experiment. The membranes have been submerged in separate containers of water for more than one year. The water in each container will be tested for compounds emitted from the membranes in the second year of the experiment.
HENDERSON ET AL., DOI: 10.1520/STP161720180085
FIG. 8 The total organic carbon (TOC) measured by a third-party analytical laboratory for various water samples. The 1.0 M NaCl water solution from before an osmosis experiment as well as from several osmosis test jars (hot rubber, PMMA, and aged asphalt-modified polyurethane—eight months later) and a sample from a real blister (approximately ten years after membrane installation) were measured.
how membranes may degrade over time when they are in contact with liquid water, especially with different alkalinity from concrete or other substrates. Further study on this effect may elucidate and help predict how membrane characteristics can change in the field. Another method to detect membrane degradation is long-term water absorption testing. Water absorption measurements are taken gravimetrically throughout the length of the osmotic flow rate testing. The weight of membrane samples is recorded, which generally shows an initial uptake of water, then equilibration as the membrane saturates. Figure 9 shows that in some cases, such as for asphalt-modified polyurethane, there is very high water absorption (more than 15%) followed by a gradual decrease in mass over the course of the measurements. This decrease in mass indicates that some of the membrane material is being lost to the surrounding water bath. The other membrane that shows a decrease in mass over time is the HDPE sheet with an adhesive that dissolved in the water over time. In the field, the adhesive side of the membrane should not be in contact with liquid water. OPTIONS FOR BASELINE ADJUSTMENTS
The unknown effects of membrane degradation can impact the methodology for osmotic flow measurements, specifically for baseline subtraction. The following section provides an overview of different methods for baseline corrections that can be carried out for consistent osmotic flow estimates.
189
190
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 9 Example data from absorption measurements concurrent to the osmotic flow measurements. Categories of high-, medium-, and low-risk membranes as determined by the level of water absorption are shown in solid lines, dashed lines, and dotted lines, respectively.
a) Option 1: Blank sample subtraction. In this methodology for baseline correction, triplicate sample containers are assembled for the osmotic flow test with 0.0 M NaCl solution (i.e., distilled water) instead of 1.0 M NaCl solution. In this methodology, the weight of water taken up by the sample by osmosis may be estimated by subtracting the blank sample mass increase from the osmosis sample mass increase using equation (2) for each measurement date. Mo ¼ DMc DMb
(2)
where: Mo ¼ mass of water uptake into sample container due to osmotic flow, DMc ¼ change in mass of sample container, and DMb ¼ change in mass of blank container (with distilled water). Recommendation: Include blank samples in each experiment as best practice for quality assurance, though using the results from these blanks may not be ideal for baseline correction for the reason stated as follows. Issues: Total dissolved solids concentration inside the container may be nonzero by the end of the experiment due to membrane degradation, as discussed in the previous section. This option also increases the number of containers in the osmotic flow test. b) Option 2: Absorption subtraction. This methodology for baseline correction utilizes the absorption measurements that are taken concurrently with
HENDERSON ET AL., DOI: 10.1520/STP161720180085
the osmotic flow rate testing. If absorption measurements are not taken on the same days as the gravimetric analysis for osmosis, the absorption measurements may be used to create a best-fit curve and estimate absorption at other dates to match the osmosis analysis. The amount of water absorption may be subtracted from the gravimetric measurements of the osmosis sample jars using equation (3) for each measurement date. Mo ¼ DMc ð%H2 O Mm f Þ
(3)
where: Mo ¼ mass of water uptake into sample container due to osmotic flow, DMc ¼ change in mass of sample container, %H2O ¼ percent of change in mass determined by water absorption measurements, Mm ¼ mass of membrane sample, and f ¼ a factor, typically near 0.25, to account for less water absorption by the membrane once it is sealed in the sample container compared to what it experiences in the water bath for absorption measurements (because water absorption from only one side and not the edges will impact the weight of the sample container). Recommendation: This baseline correction is adequate for membranes that exhibit high osmotic flow (i.e., 5 to 20 g/m2/day or higher). In all cases, absorption measurements should be used to gauge when water absorption has equilibrated, at which point the osmotic flow can be determined from the slope of the gravimetric analysis from the osmosis test. Issues: Membranes with very low or no osmotic flow have lower signalto-noise ratios, which reveals a need for further baseline correction to remove the initial wetting impact of the sample container and epoxy (see Option 3). c) Option 3: Epoxy blank subtraction. This baseline correction methodology uses triplicate control samples with impermeable lids in place of the membranes being tested. Gravimetric measurements of these controls reveal a slight weight increase over time of the sample containers even when no membrane is present, indicating that the epoxy had a minor water absorption that needs to be accounted for in cases of low signal-to-noise ratios. The amount of weight increase from water absorption of the epoxy blank may be subtracted from the gravimetric measurements of the osmosis samples using equation (4) for each measurement date. Mo ¼ DMc DMe
(4)
And, optionally, with additional membrane absorption subtraction, using equation (5): Mo ¼ DMc DMe ð%H2 O Mm f Þ
(5)
191
192
STP 1617 On Building Science and the Physics of Building Enclosure Performance
where: Mo ¼ mass of water uptake into sample container due to osmotic flow, DMc ¼ change in mass of sample container, DMe ¼ change in mass of epoxy blank container, %H2O ¼ percent change in mass determined by water absorption measurements, Mm ¼ mass of membrane sample, and f ¼ a factor, typically near 0.25, to account for less water absorption by the membrane once it is sealed in the sample container compared to what it experiences in the water bath for absorption measurements (because water absorption from only one side and not the edges will impact the weight of the sample container). Recommendation: This baseline correction is adequate for most membrane types, though the additional samples and analysis may not be necessary for membranes with high osmotic flow rates (i.e., 5 to 20 g/m2/day or higher). Issues: This option increases the number of sample containers that are included in the methodology. Care must be taken to not overcorrect with compounding the epoxy blank and the membrane absorption subtraction—a low f factor may be used in the absorption subtraction to minimize this. Each of these baseline adjustment options has its pros and cons. Having an estimate for how high the expected osmotic flow rate will be can help in choosing an appropriate baseline correction option. Because the database of osmotic flow risk for different membrane types is currently being developed, it may not always be possible to estimate a risk for osmosis prior to conducting the osmotic flow test. As such, it may be useful to develop a simpler proxy to estimate osmotic risk prior to a long experiment, such as using vapor permeance as an indicator of approximate osmotic risk. USING A PROXY FOR OSMOTIC RISK
One recommendation for estimating the risk level for osmosis is to consider the inverted wet cup vapor permeance of a membrane. As discussed in this paper, there is a correlation between vapor permeance and osmotic flow rate due to the dependence of osmosis on membrane permeability to water and not solutes. For vapor permeance to be used as a proxy metric for osmotic risk, the authors recommend that inverted wet cup testing be required for waterproofing materials. It is currently common industry practice to report vapor permeance testing using the dry cup protocol and—at times—the wet cup protocol; yet, the inverted wet cup protocol in ASTM E96 is very rarely used. This inverted wet cup methodology reports measurably higher vapor permeance than dry cup testing because it includes conditions for capillary flow and is more representative of the environment in which waterproofing membranes are installed in situations such as protected/ inverted roof assemblies.
HENDERSON ET AL., DOI: 10.1520/STP161720180085
Conclusion Currently, there is no industry standard to test for the risk of osmosis in waterproofing membranes. The authors have developed a protocol to measure the osmotic flow and evaluate the risk of osmotic blistering in waterproofing membranes including an osmotic flow rate test, inverted wet cup vapor permeance testing, and long-term absorption testing. This paper summarizes the results from a decade of osmosis testing for various waterproofing membranes. It is recommended that the osmotic testing protocol developed by the authors be adopted by ASTM as a standardized approach to estimate osmotic risk. Ongoing osmotic testing is continuing to help develop a database of test results for different membrane types. The authors are also further investigating the degradation of membranes after long-term exposure to liquid water. Nearly one dozen membranes are currently in a membrane soaking experiment. These membranes have been submerged in separate containers of water for more than year. The water in each container will be tested for compounds emitted from the membranes in the second year of the experiment to better understand how the membrane properties may be changing over time. It is also recommended that accelerated aging testing be considered to better understand how membranes perform after years on sites. Additional analysis of membranes after soaking in water with different alkalinity (or in contact with concrete) is important to understand membrane characteristics in the field, including risk for osmosis. Further testing should be carried out on aged versions of membranes as a comparison to the new, lab-cured samples to identify potential changes in properties such as vapor permeance or osmotic flow rate. DEVELOPING A STANDARDIZED APPROACH
The testing protocol described herein has been used to measure osmotic flow rates ranging from near zero to more than 20 g/m2/day for more than a dozen different waterproofing membrane types. In addition to recommending that this protocol be adopted by ASTM, the authors propose to establish thresholds above which a membrane may be deemed high risk. To reduce the potential for osmotic blistering over concrete, the authors recommend that a low-risk waterproofing membrane for use in inverted roofing assemblies should have an osmotic flow rate near 0.0 g/m2/day when tested using the proposed protocol (6 1.0 to account for experimental error). Additionally, the inverted wet cup vapor permeance should be less than that of the substrate (i.e., less than 0.1 US perms on a concrete slab), and the long-term water absorption should plateau around 5% or less. In addition to the standardization of this osmotic testing protocol, the authors recommend that inverted wet cup vapor permeance testing be required for waterproofing membranes that are used in inverted roof assemblies. This test is more representative of the conditions that these membranes experience in the field.
193
194
STP 1617 On Building Science and the Physics of Building Enclosure Performance
The results from this test may also be used as an interim proxy to estimate the risk of osmosis because it has been determined that the inverted wet cup vapor permeance and osmotic flow rate are correlated.
References 1.
2.
3.
4.
5.
6.
7. 8.
B. Hubbs, G. Finch, and R. Bombino, “Osmosis and the Blistering of Polyurethane Waterproofing Membranes” (paper presentation, Twelfth Conference on Building Science and Technology, Montreal, Quebec, Canada, May 6–8, 2009). G. Finch, B. Hubbs, and R. Bombino, “Moisture Transport by Osmotic Flow through Waterproofing Membranes—Toward the Development of Osmosis-Resistant Membranes,” in Proceedings of the ASHRAE Conference on Buildings XI (Atlanta, GA: ASHRAE, 2010). E. Henderson, G. Finch, and B. Hubbs, “Solutions to Address Osmosis and the Blistering of Liquid Applied Waterproofing Membranes” (paper presentation, Fifteenth Canadian Conference on Building Science and Technology, Vancouver, British Columbia, Canada, November 6–8, 2017). Standard Test Methods for Water Vapor Transmission of Materials, ASTM E96/E96M-16 (West Conshohocken, PA: ASTM International, approved March 1, 2016), http://doi.org/ 10.1520/E0096_E0096M-16 Standard Test Method for Water Absorption of Plastics, ASTM D570-98(2018) (West Conshohocken, PA: ASTM International, approved August 1, 2018), http://doi.org/ 10.1520/D0570-98R18 X. F. Yang, C. Vang, D. E. Tallman, G. P. Bierwagen, S. G. Croll, and S. Rohlik, “Weathering Degradation of Polyurethane Coating,” Polymer Degradation and Stability 74, no. 2 (2001): 341–351. G. T. Howard, “Biodegradation of Polyurethane: A Review,” International Biodeterioration and Biodegradation 49, no. 4 (2002): 245–252. G. Finch, “Osmosis: The Bane of Liquid Applied Waterproofing Membranes” (slide presentation, Westford Symposium Summer Camp 2014, Westford, MA, August 6, 2014).
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180064
Adam L. Rizor1 and Aaron Corn1
Design Criteria and Solutions to Common Issues in Building Envelope Design Citation A. L. Rizor and A. Corn, “Design Criteria and Solutions to Common Issues in Building Envelope Design,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 195–210. http://doi.org/10.1520/STP1617201800642
ABSTRACT
As third-party consultants, we retain a unique perspective through participation in the design, construction, assessment, and forensic analysis of building envelopes. Our familiarity with the building envelope has shown us that the performance and life span of a building is remarkably dependent on the quality of its design—not only in the selection of materials and systems but also in the quality of their detailing. Considering modern architectural building design, building envelope systems are now relied upon heavily in ways beyond their basic function as environmental control layers. High-risk areas such as elaborately designed green roofs raise the stakes yet further and threaten the integrity of the building envelope when function follows second to aesthetics. A few small decisions in how building envelope systems are implemented can mean the difference between a successful or failing design, and the difference may not be immediately obvious to the untrained eye. Cost is another factor, and often a limitation, guiding the design of facade, roofing, and waterproofing systems. Special attention also needs to be given to cost- or value-oriented decision-making in building envelope design. Having the correct knowledge and tools to design building envelope systems can offer protection against costly mistakes and provide balance between function and architectural appeal.
Manuscript received September 30, 2018; accepted for publication March 16, 2019. 1 Curtainwall Design and Consulting, Inc., 1602 Village Market Blvd. SE, Suite 260, Leesburg, VA 20175, USA A. L. R. https://orcid.org/0000-0003-3921-8915, A. C. https://orcid.org/0000-0002-3003-4215 2 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
195
196
STP 1617 On Building Science and the Physics of Building Enclosure Performance
By providing general design criteria and solutions to commonly observed issues, our objective is to afford others a better opportunity to identify potential issues and, ultimately, avoid problems that develop during the service life of a building. Keywords curtain wall, roofing, waterproofing, expansion assemblies, building envelope, consultant, design, testing, air barrier, electronic leak detection
Introduction The perspective of a third-party building envelope consultant allows for the unique opportunity to provide information widely acknowledged as unbiased and free of hidden motivations. The documentation of the following subjects is an effort to provide applicable information in such a manner so that others may benefit from its usage. Although this information is intended mainly for those directly involved in the design of building envelope systems, it can also be of value to parties concerned with their procurement, construction, and maintenance. The primary objective of presenting this information is to encourage cognizant and knowledgeable decisions in the design and construction of building envelope systems. Historically, these systems have served as basic weatherproofing and environmental protection for structures of all types. Modern architectural trends have pushed building envelopes far beyond their basic functions and frequently overshadow the key concepts integral to their implementation and performance. Without recognizing the potential risks associated with building envelopes, issues will become increasingly common, and costly, as architecture continually progresses toward aesthetically motivated design.
Value of a Building Envelope Consultant The building envelope represents a significant percentage of construction costs, but just how significant can these costs be? Up to 20% of the whole building cost for multistory buildings.1 This figure may rise further in situations where the building envelope becomes a more substantial and complex part of a building’s design. Working with a building envelope consultant can help to reduce overall cost and can protect against expensive mistakes. When considering a consulting firm, it is important to research their available services. Firms that offer a wide variety of different services are more valuable as they can be called upon to assist in various situations. Unique project conditions may also require that the consultant provide highly specialized services. Reference Table 1 for common functions of a qualified building envelope consultant.2 DESIGN ASSISTANCE
Design professionals engage consultants to utilize those consultants’ expert proficiency in understanding complex building envelope systems. Early in the design of
RIZOR AND CORN, DOI: 10.1520/STP161720180064
TABLE 1 Functions of a building envelope consultant
Design Assistance
Bid Review
Construction Administration
Building Information Modeling (BIM)
Peer Review
Laboratory Mock-Up Evaluation
Performance Testing
Building Envelope Commissioning
Existing Condition Surveys
Nondestructive Testing
Engineering
Litigation Support
Drafting
Thermal Modeling
a construction project, important decisions are made that may eventually affect many aspects of the project including cost, schedule, and the overall success of the project. Taking advantage of a consultant’s expertise is especially valuable at this stage of construction. A qualified building envelope consultant can help identify problematic situations as well as influence design decisions such as structurally sloping a roof deck. A single decision such as this can greatly impact the total cost of a roofing assembly, resulting in significant material cost savings by excluding unnecessary and costly tapered insulation. A consultant’s level of involvement has a direct correlation with the opportunity for incremental cost savings. Also, the fees associated with hiring a building envelope consultant may easily be recouped multiple times over during the design of a project through savings in total envelope costs. Specifying building materials or building envelope systems is another opportunity for consultants to provide beneficial input. A 2015 survey by Kassem and Mitchell stated that up to 69% of industry participants felt that they were not given enough initial technical information prior to selecting or specifying curtain wall systems.3 A consultant’s experience with the individual materials and systems being considered for a project is likely to be very extensive. For example, consultants often make visits to the production facilities where building products are manufactured and can give testimony as to the quality of materials based on their manufacturing process. Due to the misleading nature of some published product information, hands-on experience such as this is an invaluable asset in understanding real-life performance and constructability. On paper, related materials may appear to be equals, but their quality and performance in the field may be entirely unalike. Specifying detailed installation requirements can help guarantee a quality application of the selected products. Building material manufacturers regularly publish information on the installation requirements for their products, but these generally represent minimum requirements and do not focus on the adjacent materials that interface with the product. Specifying a higher level of detail and quality can result in a superior installation, and in some cases, is necessary to meet desired performance characteristics.
197
198
STP 1617 On Building Science and the Physics of Building Enclosure Performance
TABLE 2 Benefits of building envelope commissioning
Documentation of the Owner’s Project Requirements (OPR) Focused pursuit of the owner’s goals for the project Fewer change orders Reduced punch list
Prevention of costly mistakes in the execution of the project Lower overall project costs through design to occupancy Reduced rework Better trained and equipped operations and maintenance staff
BUILDING ENVELOPE COMMISSIONING
There has been more focus on commissioning of the building envelope in recent years, specifically with the adoption of the certification program Leadership in Energy and Environmental Design (LEED), Version 4, which includes LEED points for commissioning of the envelope.4 Building Enclosure Commissioning (BECx) is a structured quality assurance process designed to deliver a building based on an owner’s specific requirements. The commissioning process for the building envelope is an extension of mechanical, electrical, and plumbing (MEP) commissioning that focuses on reduced energy consumption. It has been developed to function in concert with standards such as ASHRAE Guideline 0, ICC 1000, NIBS Guideline 3, and ASTM E2813-18, Standard Practice for Building Enclosure Commissioning.5 The process provides building owners with a strategy for achieving the desired level of performance of the envelope throughout the design, construction, and service life of a building. Building envelope commissioning translates and applies the Owner’s Project Requirements (OPR) through a process of review, verification, sampling, testing, and documentation. Commissioning of a building envelope differs from other quality control processes as it is not focused on individual systems but rather the complete functionality of all envelope systems as a whole. The process is adaptable among projects of varying size, expense, risk tolerance, and complexity. Reference 6,7 Table 2 for benefits of building envelope commissioning. FULL-TIME OBSERVATION
The presence of a qualified observer during the installation of building envelope systems is beneficial to all parties involved, including the installer, and is the best way to guard against installation deficiencies. Daily observations should include visual monitoring as well as various forms of testing, such as adhesion tests or thickness measurements for hot fluid applied (HFA) rubberized asphalt roofing (fig. 1). The documentation of these observations is invaluable and can prevent unexpected financial loss to parties who are not responsible for failures, but who would otherwise lack proof. In the specific instance of a green roofing system, much—if not all—of the roofing system is covered by overburden. This prevents periodic maintenance, repair, or even access to the roofing system. Full-time roofing
RIZOR AND CORN, DOI: 10.1520/STP161720180064
FIG. 1 Material temperature reading taken during HFA roofing installation.
observation significantly limits the risk of incurring damaging losses related to the repair of green roofing systems after the placement of overburden.
System Integration Design There is an increasing number of building envelope systems and products available in today’s market. While individual materials or systems may be suitable for a specific application, the selection of these materials is not always coordinated as part of a complete building envelope design. This can result in the selection of individual systems that are not compatible with each other. Common issues include drainage paths in different planes, voids in the envelope air barrier, and direct transitions between incompatible materials. As an example, consider a mid-rise multifamily residential building. This type of construction could include window systems of varying size and complexity, opaque walls that combine multiple exterior materials, occupied below-grade space, and accessible roof areas such as green roof terraces. Each of these systems has different primary and secondary considerations and may be appropriately selected per their individual location. However, the building envelope must function as a whole, and consideration must be given to how each system will function in relation to the other systems. To further explore this example, each individual envelope system is available in various forms. Air barrier membranes are manufactured as self-adhering sheets, fluid-applied membranes, expanding foam, and so on.8,9 The air barrier system will inevitably encounter the other building envelope systems. When the air barrier system terminates at a window opening, is the selected product suitable to integrate
199
200
STP 1617 On Building Science and the Physics of Building Enclosure Performance
with the drainage plane of the window system? This type of question must be considered. If transitions between systems are not considered as part of the initial building design, they must be coordinated during product selection and procurement or, in some cases, even after procurement (fig. 2). In either situation, problems typically will include discontinuity of the air and water control planes and incompatibility of materials. To coordinate the use of incompatible materials, additional materials may need to be introduced and supplemental details developed. In the absence of prior considerations, rework of installed materials can become necessary. Such conditions add cost, delay, and unnecessary complexity to the building envelope system. Vetting each individual system as part of the greater building envelope will result in simplified detailing, limited rework in the field, and reduced costs. MATERIAL COMPATIBILITY
A common oversight in designing a suitable building envelope is the compatibility of materials at transitions. When an air barrier system meets the waterproofing system, how do the systems interact and complete the envelope? When the roof system interfaces the curtain wall, how are the roofing materials applied to the aluminum framing of the curtain wall? These are examples of interfaces that need to be considered when designing the envelope. Design conditions commonly result in perimeter window weather sealants interfacing air barrier membranes. Depending on the air barrier system materials, window sealant joints can experience compatibility issues (fig. 3). Silicone is a reliable option for compatibility between adjacent materials and the sealants used in
FIG. 2 Transition between materials at a single-ply roof not considered during the initial building design. Additional materials applied in an attempt to detail the transition, resulting in further problems.
RIZOR AND CORN, DOI: 10.1520/STP161720180064
FIG. 3 Sealant and expansion joint materials deteriorating due to incompatibility.
the window fabrication. Even so, certain silicone sealants may not provide adequate adhesion to some materials, such as polyethylene-faced membranes. Also, lightcolored sealants can change color when exposed to asphaltic self-adhering sheet goods. Concerns such as this require that individual system materials be reviewed for use as part of the complete building envelope. Many manufacturers publish historical data on their products’ compatibility with other materials. The published data generally focus on materials their products encounter most frequently. For example, materials used for the perimeter sealant joints of windows will likely have published test data indicating adhesion to air barrier membranes. Although information like this can be a valuable resource, individual testing should be included in the submittal requirements for verification of product compatibility unique to each specific project. DRAINAGE
How each building envelope system drains must also be considered when designing a complete building envelope. Coordinating the way different systems will handle water individually is important to properly manage water control layers and drainage pathways. If drainage layers are not properly transitioned, water intrusion may occur. A masonry cavity wall is designed to collect water from the backside of the masonry and drain down the cavity to weeps at the base of the wall. If this drainage plane is interrupted by a curtain wall or window system, the drainage layer of the masonry wall must be coordinated where it interfaces the glazing system. Leaks frequently occur where drainage within a masonry cavity encounters a window frame
201
202
STP 1617 On Building Science and the Physics of Building Enclosure Performance
behind the masonry wall’s drainage plane; water that cannot be drained out will become trapped within the wall assembly and eventually find its way to the interior of the building. Drainage planes must also be coordinated and connected at flashings, closures, diverters, or any combination of these conditions.10 MOVEMENT
Buildings experience structural movement that ultimately transfers to the building envelope. These inherent movements must be incorporated into the envelope design to prevent damage as well as to maintain the air and water barriers of the envelope. There are different types of movements that need to be accounted for, such as dead loads, live loads due to occupant activities, creep from concrete shrinkage, wind loads, seismic movement, moisture absorption expansion of clay masonry, and thermal expansion and contraction. Individual movement dimensions related to creep or thermal movement may be smaller, and movement related to live load or seismic events may be more substantial. Imposed individually, these movements are more easily accounted for, but they must also be considered cumulatively. A building deflection joint may not perform properly if only one of these movement dimensions is accounted for; it could be stretched beyond its movement capability, resulting in a failure of the air and water barrier. Another factor for consideration is the way each individual system accommodates movement. Typically, an opaque wall will be dead loaded on each floor, but a unitized curtain wall system will be hung from the floor above. These conditions create differential movement between the individual systems. A standard silicone sealant joint may not accommodate the anticipated movement between systems, and the resulting shear force could jeopardize the sealant joint. In this type of condition, attention must be given to how much each system will move and in which directions movement will occur. Exterior Expansion Assemblies
Exterior expansion assemblies are responsible for some of the most complex exterior design difficulties because they are relied upon to span substrates of all sorts and to retain a number of important performance characteristics. Reference Table 3 for performance characteristics of exterior expansion assemblies.11,12 Designing exterior expansion assemblies involves much more than experience and knowledge relating to the systems themselves. To understand how these
TABLE 3 Performance characteristics of exterior expansion assemblies
Waterproof Fireproof
Trafficable Wind and air resistant
Soundproof
Seismic resistant
Insulating
Visually appealing
RIZOR AND CORN, DOI: 10.1520/STP161720180064
systems are utilized, the designer must be familiar with the substrates and conditions surrounding the expansion assembly; examples include seismic loading, structural movement, hydrostatic resistance, and thermal expansion and contraction.11 Furthermore, the nature of exterior expansion assemblies requires an understanding of the systems they interact with. A single exterior expansion joint will commonly interconnect masonry, panelized facade systems, curtain walls, roofing, below-grade waterproofing, and even structural components (fig. 4).12 In the case of precompressed expansion joints, stresses can also be introduced by the joint assembly itself. Accounting for these loads is important for potentially fragile substrates. Even after considering each of these aspects, it is still common to come across situations where “off-the-shelf” products cannot be implemented. Each manufacturer has a different ability to produce and supply custom products for these types of situations, and a familiarity with their individual abilities is vital. Production and shipping times also become significant in situations where timely solutions are needed to meet construction schedules. Some manufacturers also have the ability to assist in the design of custom exterior expansion assemblies by providing engineering, drafting, and field services. Having field representation from a manufacturer is particularly important. Representatives will often provide assistance in measuring and assessing the profile of existing joints and substrates. Information on existing substrates may not be readily available, leading to situations where field evaluation is required.
FIG. 4 Complex expansion joint transitions and changes in direction. The shown expansion joint is interfacing with single-ply roofing, HFA roofing, green roofing components, curtain wall, metal copings, and glass railings.
203
204
STP 1617 On Building Science and the Physics of Building Enclosure Performance
The design of exterior expansion assemblies can be simplified considerably by identifying their locations early in the design of a construction project.11 Doing so will create an opportunity to visually observe existing substrates on adjoining structures before they are inaccessible. Related trades and materials can also be properly scheduled, sequenced, and coordinated to promote successful implementation of the expansion joints. Each joint can be evaluated ahead of time to review sizing along with movement, insulation, and fire resistance requirements. Colors and joint types can be more strategically planned while considering aesthetic designs.12 Multiple manufacturing sources can be eliminated to ensure compatibility between intersecting expansion joints. Also, identifying in advance any conditions requiring custom solutions will help prevent delays and penalties related to construction schedules.
Testing Building Envelopes Testing is frequently employed to evaluate building envelope systems before, during, and after their installation. Performance testing, leak detection, and qualitative imaging are just a few examples of tests that have been performed on countless buildings around the world. The success of testing relies heavily on the knowledge and experience of the party performing the tests. In some cases, collected information is open entirely to interpretation. Infrared thermograms, for example, are easily misinterpreted by inexperienced eyes. Certain types of testing are commonly performed incorrectly or utilized in conditions that prevent accurate readings. Understanding the basic concepts and technologies related to testing building envelopes is the best way to prevent complications. ELECTRONIC LEAK DETECTION VERSUS FLOOD TESTING
Many mistakenly consider flood testing to be the tried-and-true form of leak detection but fail to understand that flood testing is not leak detection at all. Flood testing simply demonstrates the ability of a roofing or waterproofing membrane to resist hydrostatic pressure, whereas electronic leak detection seeks out and accurately locates breaches in the membrane. The accuracy of electronic leak detection can locate numerous breaches in an area as small as a few square feet. This technology is generally not well understood and is underutilized as a result, especially considering the sequencing and scheduling challenges that regularly leave roofing membranes defenseless to heavy construction traffic. Electronic leak detection (fig. 5) is the most reliable and effective technology currently available to test for damage or installation deficiencies in roofing and waterproofing systems. Flood tests conducted per ASTM D5957-98(2013), Standard Guide for Flood Testing Horizontal Waterproofing Installations,13 last for a minimum period of 24 h. This short testing period does not account for many of the most common causes of leaks. The development of a leak is largely affected by environmental conditions such as freeze–thaw cycles. Leaks may take weeks, months, or even years to fully develop and become noticeable during the life of a roofing or waterproofing system.
RIZOR AND CORN, DOI: 10.1520/STP161720180064
FIG. 5 Repaired breaches located using electronic leak detection. HFA roofing over a concrete deck.
Flood testing excludes any vertical surfaces above the height of the water and may not be a feasible option due to the imposed structural load of flooding a roof deck. Testing after the installation of interior finishes can also result in costly damage if water intrusion occurs. The financial and time-related commitments associated with flood testing are significantly greater in comparison to electronic leak detection. Electronic leak detection performed per ASTM D7877-14, Standard Guide for Electronic Methods for Detecting and Locating Leaks in Waterproof Membranes,14 can be performed as quickly as 5,000 ft2/h with one to two people. Located breaches can be repaired the same day and retested within minutes of being repaired. Flood testing involves building dams, plugging drains, flooding the test area, waiting for the duration of the test, and draining the test area. Once the test area is drained, any areas of concern must be dried out and repaired before performing the test again. The process of building dams can also result in damage to the roofing or waterproofing system. Costs associated with flood testing include materials to build dams and plug the drains, labor to install and remove the materials, the cost of flooding large areas with multiple inches of water, and documentation of the testing. Specialists performing electronic leak detection generally will provide digital reports that include plans and photo documentation of any located breaches. INFRARED THERMOGRAPHY
In the hands of an experienced thermographer, infrared cameras are a powerful tool for testing the building envelope and can be used to identify the presence of moisture in roofing assemblies. Via infrared radiation, it is possible to nondestructively detect minute temperature variances within a roofing assembly. Temperature
205
206
STP 1617 On Building Science and the Physics of Building Enclosure Performance
differentials visible in the infrared spectrum are interpreted to locate the presence of moisture, damaged materials, and leak sources. Along with extensive knowledge in roof design, thermographers must be well-versed in nondestructive testing, interpreting thermograms, understanding thermal anomalies, and in evaluating the influence of ambient weather conditions. Infrared thermography is compatible with a wide range of roofing systems, including both new and existing roofs, and can be effectively performed by a skilled thermographer throughout most of the year (fig. 6). During a typical infrared survey, verification of accuracy can be achieved with nondestructive electrical impedance scanning, along with intermittent destructive testing. Located thermal anomalies are marked on the surface of the roof with high visibility paint and documented on a roof plan provided digitally with thermograms and visible light images. Infrared thermography for roofs is one of the numerous varieties of nondestructive testing technologies applied to roofing and waterproofing systems, and it has been utilized throughout the roofing industry for decades due to its reputation for ensuring the performance of roofs. But this type of nondestructive testing is equally as valuable when applied to other parts of the exterior building envelope, such as vertical facades. In this application, infrared thermography can be used to locate air leakage, trapped moisture, leaks, thermal bridges, condensation, and missing insulation in buildings (fig. 7). Information from infrared thermograms may be integrated with powerful analysis software to further diagnose problems. Infrared thermography for building envelopes is also useful when there are no specific concerns or issues. Qualitative information can be gathered to simply
FIG. 6 Moisture in a roofing assembly located using infrared thermography.
RIZOR AND CORN, DOI: 10.1520/STP161720180064
FIG. 7 Thermal bridging in a masonry cavity wall located using infrared thermography.
confirm the quality of a new installation. Similar testing is often utilized as an important part of periodic maintenance when employed at regular intervals to verify the integrity of existing building envelope systems. Annual, weather-related, preconstruction, and postconstruction inspections can all greatly benefit from the use of this technology.
Green Roofs Green roofs are an investment due to their benefits and advantages, but a poorly executed design can easily dissolve projected returns. Reference Table 4 for benefits of green roofs.15 The easiest and most effective way to enhance value in a design is to involve a professional who understands, equally, the benefits and risks related to green roofing; this individual may be an architect, consultant, manufacturer’s representative, or installer. Although the following concepts but scratch the surface of responsible design, any individual associated with specifying, installing, or maintaining green roofs should have a basic understanding of them. Perhaps the most common misapprehension associated with green roofing is simply that a functional, economical roofing design cannot also serve as a highly modernized architectural feature. Green roofing design, in many ways, has moved beyond straightforward performance applications; but the integrity of the roof system does not need to be compromised to allow for elaborate amenity spaces or landscaping. Not only can each expectation be achieved simultaneously, but they can even benefit from one another.
207
208
STP 1617 On Building Science and the Physics of Building Enclosure Performance
TABLE 4 Benefits of green roofs
Improved
Aesthetics and
Building energy
usable space
efficiency
Air quality
Water quality and controlled stormwater runoff
Increased
Wildlife habitat
Income from
Sound absorption and
Employment in the
occupancy comfort
maintenance and
rentable square
urban agriculture
footage
industries Reduced
Urban heat island
Expense related to
Expense related to
effect
stormwater
energy consumption
management Extended
Service life of protected roofing system
No matter how attractive or how impressive modern green roofs have become, their basic and most central function remains the same. More than any other type of roofing, a high-performance green roof design demands longevity, durability, and efficiency. Compromising just one of these characteristics can result in a failing design. Not all roofs are created equal, and the weaknesses of individual forms of lowslope roofing become more apparent once they are subjected to the concepts of green roofing. In Washington, DC, as with many urban districts, concrete roof decks are commonly constructed. Concrete affords significant design advantages and provides an exceptional foundation for the most highly performing inverted roof membrane assemblies. These systems, implemented with materials such as hot fluid applied rubberized asphalt, create dependable green roofs. The concrete deck provides efficient slope, the rubberized asphalt membrane provides durability, and the protective overburden provides longevity. For hot fluid applied rubberized asphalt roofing, how should design difficulties, such as metal roof decks, be handled? The most important aspect of such a circumstance is how substantially the risk factor has increased due to something seemingly unrelated to the roof system. Due to the nature of concrete, sloped concrete roof decks have a reduced potential for developing leaks, even if the roofing has been compromised. By introducing a metal roof deck into the construction of a roof, not only are leaks more likely to develop, water can now migrate throughout the roof by way of the metal flutes. Sourcing a leak through a metal roof deck is difficult because water may travel a lengthy distance from its origin before entering an interior space. Often, the only corrective possibility is costly destructive investigations. The cost of removing and replacing green roofing systems to source and repair leaks is too frequently left unconsidered by the design authority. At an initial cost range of $6/ft2 to $20/ft2, this cost can increase incrementally for the removal and replacement of green roofing.15 Understanding increased risk factors means making
RIZOR AND CORN, DOI: 10.1520/STP161720180064
conservative design decisions. Specifying a two- or even three-ply modified bitumen roofing system over a metal roof deck is a grounded and responsible decision. Specifying a green roof over an ethylene propylene diene monomer (EPDM) single-ply membrane with a metal roof deck is, at the very least, an unnecessary risk.
Conclusion The function and performance of building envelope systems should be elevated to a level of great importance in their design and construction. The previously mentioned subjects represent only a selection of significant matters relating to this issue. Established good design practices will continually become more important as architectural trends increase the potential for risks and failures. Looking forward, it is the responsibility of design professionals to regularly advance these practices to embrace future challenges. Those who are knowledgeable about these concepts should also seek out and create opportunities to educate others on their importance. The combined efforts of past, present, and future design professionals will benefit all who interact with commercial and residential building projects.
References 1.
2. 3.
4. 5.
6.
7.
8. 9.
C. Arnold, “Building Envelope Design Guide—Introduction,” 2016, http://web.archive. org/web/20180923163737/https://www.wbdg.org/guides-specifications/building-envelopedesign-guide/building-envelope-design-guide-introduction Curtainwall Design Consulting, “Value of a Facade Consultant,” 2018, https://web.archive. org/web/20180923171026/http://cdc-usa.com/value-facade-consultant-0 M. Kassem and D. Mitchell, “Bridging the Gap between Selection Decisions of Facade Systems at the Early Design Phase: Issues, Challenges and Solutions,” Journal of Facade Design and Engineering 3, no. 2 (2015): 165–183, http://doi.org/10.3233/FDE-150037 U.S. Green Building Council, “LEED v4,” 2018, https://web.archive.org/web/20180923172432 /https://new.usgbc.org/leed-v4 Standard Practice for Building Enclosure Commissioning, ASTM E2813-18 (West Conshohocken, PA: ASTM International, approved October 1, 2018), http://doi.org/10.1520/ E2813-18 Whole Building Design Guide (WBDG) Project Management Committee and Commissioning Industry Leaders Council, “Building Commissioning,” WBDG, 2016, https://web. archive.org/web/20180923203051/https://www.wbdg.org/building-commissioning B. Casheri, “What Is Building Enclosure Commissioning? Benefits and Tips,” Building Commissioning Association, 2017, https://web.archive.org/web/20180923203628 /https://www.bcxa.org/blog/2017/09/07/building-enclosure-commissioning-benefits-tips/ W. Anis, “Air Barrier Systems in Buildings,” 2016, https://web.archive.org/web /20180923171503/https://www.wbdg.org/resources/air-barrier-systems-buildings J. Lstiburek, “BSD-104: Understanding Air Barriers,” 2006, https://web.archive.org/web/ 20180924144637/https://buildingscience.com/documents/digests/bsd-104-understandingair-barriers
209
210
STP 1617 On Building Science and the Physics of Building Enclosure Performance
10.
11.
12. 13.
14.
15.
L. Dalgleish, “Achieving a High Performance Air Barrier System: Proper Design, Installation, and Field Quality Control,” Air Barrier Association of America, 2018, https://web. archive.org/web/20180924150211/http://www.airbarrier.org/technical-information/abaaarticles-standards/ M. Maing, “In Between: Designing Joints within Facades,” Symposium on Building Envelope Technology, Roof Consultants Institute (RCI), 2010, https://web.archive.org/web/ 20180923194405/http://rci-online.org/wp-content/uploads/2010-BES-maing.pdf Emseal, “Expansion Joints are UGLY!” 2016, https://web.archive.org/web/20180923195719 /https://www.emseal.com/article/expansion-joints-ugly/ Standard Guide for Flood Testing Horizontal Waterproofing Installations, ASTM D595798(2013) (West Conshohocken, PA: ASTM International, approved May 1, 2013), http:// doi.org/10.1520/D5957-98R13 Standard Guide for Electronic Methods for Detecting and Locating Leaks in Waterproof Membranes, ASTM D7877-14 (West Conshohocken, PA: ASTM International, approved August 1, 2014), http://doi.org/10.1520/D7877-14 GSA, “The Benefits and Challenges of Green Roofs on Public and Commercial Buildings,” GSA Green Roof Benefits and Challenges, 2011, http://web.archive.org/web /20180923170147/https://www.gsa.gov/about-us/organization/office-of-governmentwidepolicy/office-of-federal-highperformance-buildings/resource-library/integrative-strategies/ green-roofs
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720190154
Jennifer Keegan1 and Matthew Ridgway2
Unintended Consequences: A Review of Critical Details, Serviceability, and Durability of Modern High-Performance Facades Citation J. Keegan and M. Ridgway, “Unintended Consequences: A Review of Critical Details, Serviceability, and Durability of Modern High-Performance Facades,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 211–262. http://doi.org/10.1520/ STP1617201901543
ABSTRACT
The design of modern high-performance facades for the construction industry’s largest and most significant projects often involves the use of prefabrication, modern glazing, or curtain wall technologies (or combinations thereof). As a means of quality assurance and control, we depend on familiar tools: applicable codes and industry standards, specified performance requirements, Building Enclosure Commissioning, best practices, warranties, performance mock-ups, performance verification testing, and field inspection. Yet even systems designed with the highest level of detail and the most rigorous quality assurance/quality control programs may still be undermined by piecemeal value management, shortcomings in coordination, and cost and schedule issues. The forensics of building enclosure science is replete with cautionary tales of building enclosures that were designed with the highest of aspirations but that failed to perform as intended. Using an institutional building constructed in 2001 as a case study, we will take a deep dive into the critical details (transitions within the assembly as it evolves from roof to wall, joints, and water management systems) and other Manuscript received December 2, 2019; accepted for publication January 14, 2020. 1 GAF, 1 Campus Dr., Parsippany, NJ 07054, USA https://orcid.org/0000-0003-2735-4856 2 Intertek, Building Science Solutions, 130 Derry Ct., York, PA 17406, USA 3 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21-22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
211
212
STP 1617 On Building Science and the Physics of Building Enclosure Performance
sometimes vague requirements such as serviceability and durability. The advantage of hindsight during an evaluation is that we are able to evaluate individual failures as well as the relationship and cascading effects that failures can have on other time-dependent durability requirements. We will present a taxonomy for understanding and analyzing each individual failure, the relationships such failures have with one another, and the decisions made during a project. These organizational strategies may be useful tools when making decisions during early project phases. While the ultimate service life of many “modern” curtain wall systems is expected to exceed the careers of practitioners tasked with design, manufacture, and installation of the systems, a critical look at these issues will become increasingly important as the building stock of highperformance facades and architectural glass continues to age. Keywords high-performance facades, four orders of failure, anticipating failures
Introduction The design of modern high-performance facades for the construction industry’s largest and most significant projects often involves the use of prefabrication, modern glazing, or curtain wall technologies (or combinations thereof). As a means of quality assurance and quality control for these projects, we depend on some familiar tools: applicable codes and industry standards, specified performance requirements, Building Enclosure Commissioning (BECx), best practices, product and workmanship warranties, performance mock-ups, performance verification testing, and field inspection. Yet even systems designed with the highest level of detail and the most rigorous quality assurance/quality control (QA/QC) programs may still be undermined by piecemeal value management, shortcomings in coordination, and cost and schedule issues. The forensics of building enclosure science is replete with cautionary tales of building enclosures that were designed with the highest of aspirations but that failed to perform as intended. Using an institutional building constructed in 2001 as a forensic case study, we will take a deep dive into the critical details (transitions within the assembly as it evolves from roof to wall, joints, and water management systems) and other sometimes vague requirements such as serviceability and durability. The advantage of hindsight during a forensic evaluation is that we are able to evaluate individual failures as well as the relationship and cascading effects that failures can have on other time-dependent durability requirements. We will present a taxonomy for understanding and analyzing each individual failure, the relationships such failures have with one another, and the decisions made during a project. These organizational strategies may be useful tools when making decisions during early project phases. While the ultimate service life of many “modern” curtain wall systems is expected to exceed the careers of practitioners tasked with design, manufacture, and
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
installation of the systems, a critical look at these issues will become increasingly important as the building stock of high-performance facades and architectural glass continues to age.
Case Study—Analysis of Modern Building Enclosure Failures Approximately twenty years ago, the Philadelphia-based Regional Performing Arts Center conceived of a significant civic project to house new performance venues for their resident companies, including the Philadelphia Orchestra and Pennsylvania Ballet among other groups. Via traditional project delivery methods, an institutional complex was constructed composed of two standalone theater buildings within a larger complex and topped with a skylight spanning approximately 3.5 acres (fig. 1). The lower four levels of the building enclosure consist of traditionally built masonry cavity walls that were originally topped with ethylene propylene diene terpolymer (EPDM) roofs over which the skylight is framed. The skylight forms a folded-plate barrel vault that spans north-south approximately 168 ft and extends approximately 360 ft in the east-west direction. The superstructure employs 5-in. steel tubes arranged as folded Vierendeel truss frames with adjacent curved top and bottom chords of the truss forming peaks and valleys, approximately 9 ft, 11 in. from peak to peak (figs. 2 and 3).
FIG. 1 Building plan showing interior buildings within vaulted skylight roof.
213
214
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 2 Folded plate skylight viewed from the west looking east.
FIG. 3 Folded plate skylight viewed from the south looking north.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
Preassembled unitized glazing panels (also known as a carrier system or cassette system) are set on an extruded aluminum subframe installed over a steel strongback system. A typical unitized panel, or cassette, consists of four structurally glazed units, and each unit is constructed with 7/16-in.-thick lites of laminated glass (3/16-in. panes with a 0.090 polyvinyl butyral [PVB] interlayer) that are stacked and captured by extruded aluminum ribs. Individual lites were originally factory glazed with structural silicone sealant to intermediate aluminum frames and, with other weather sealants, also factory applied. Field-applied perimeter sealant joints were installed at the framing joints between adjacent unitized panels and the extruded aluminum perimeter. At end wall conditions of high roofs, the skylight unitized panels turn 85 from vertical to horizontal to a structurally supported horizontal extruded aluminum frame and similarly glazed triangle skylight component that terminates at an EPDM roof curb. At lower roofs, the EPDM was originally designed to terminate vertically behind the unitized panels to the structural steel frame (figs. 4 through 6). POSTOCCUPANCY OPERATIONAL CHALLENGES
Following occupancy of the case study building, several leaks were noted beneath the skylight; and since those initial reports of water intrusion, various continuing shortcomings of exterior enclosure performance have been noted with varying degrees of maintenance, investigation/remediation efforts, and capital renewal. In summary, the initial operational challenges included:
FIG. 4 Exterior view of skylight terminations at low roof.
215
216
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 5 Exterior view of skylight termination at high roof.
FIG. 6 Construction of typical unitized assembly.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
2001 (Summer) The vaulted skylight roof was reportedly completed in accordance with original construction documents. 2001 (Fall) The general contractor reported that five or six leaks were apparent in the barrel vault skylight immediately following substantial completion. 2001–2004 The subcontractor performed isolated repairs and selective replacement of sealants (“spot repairs”) on a recurring basis in the vicinity of reported leak activity, with varying degrees of success. 2005 The original factory-glazed exterior glazing joints within the unitized panels were found to be prematurely failing, and it was determined that replacement was required. 2006 (Fall) Exposed factory-glazed sealants at the intermediate aluminum frames of the unitized panels were replaced with a high-performance silicone sealant. 2007 (Fall) Operations staff reported an increase in observed leak activity magnitude and locations. INITIAL INVESTIGATION AND REPAIR EFFORTS
An initial postoccupancy evaluation attributed water leakage primarily to improperly installed sealant joints at locations within the skylight, which had more than 100,000 linear feet of sealant. Findings of the limited initial investigation and repair efforts included: 2008 (Spring) A preliminary investigation was performed. A report was delivered identifying causes of the previously characterized leak activity and making general recommendations for a path forward. The leaks were characterized as the following general types of observed water infiltration: 1. Ends of copings at high roofs: Leaks emanating near the end condition of the barrel vault terminations at the high roof (fig. 7). 2. Perimeter leaks at the barrel vault terminations along low roofs: Leakage emanating from ponding water at the barrel vault/roof termination. 3. Field leaks: Leakage dripping from anywhere within the field of the skylight onto interior lobbies or theater roofs (see typical test results in fig. 8). 4. Valley leaks: Leaks that track from the field of the skylight and collect at the framing of the barrel vault structural connection to perimeter concrete curbs (fig. 9). 2009 As part of a pilot study, preformed silicone seals were installed on the south elevation of three ribs. The silicone seals spanned over the sealant joint of extruded aluminum profile to the unitized panel joint. Testing was not performed to confirm the remediation of leaks in the pilot area, but the owner reported general improvement of leak activity in the affected area.
217
218
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 7 Transition from aluminum coping to EPDM parapet terminations at high roof is one of the most weather-prone areas on the building.
FIG. 8 Multiple leak locations (shown with pads) revealed during test of isolated sealant failure above.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
FIG. 9 Leaks traveled down the interior of the skylight and structure and collected at perimeter framing members.
2009
2010
2012
The low roof was replaced with a new styrene butadiene styrene (SBS) modified bitumen inverted roof membrane (IRMA) system with new polymethyl methacrylate (PMMA) curb flashing at the base of the barrel vault. Repairs were reportedly successful; however, no testing was performed to verify the efficacy of the repairs. Horizontal glazing at the base of the south elevation barrel vault required replacement following impact damage. Snow and ice cascaded down the valley of the barrel vault causing breakage (fig. 10). An event space on the roof of one of the interior buildings was enclosed and provided with supplemental air conditioning. This was due to: 1. Elevated daytime temperatures rendering the space unusable and making it impossible to keep plants alive. It was noted that operable vents through the barrel vault skylight were not being used due to operational failures and water intrusion (fig. 11). 2. Noise pollution between the elevated event space and lobby below.
219
220
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 10 Horizontal skylight sections required selective replacement and generally deflected at midspan due to cascading snow and ice.
FIG. 11 Stock photo of planted roof.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
SUBSEQUENT INVESTIGATION AND REMEDIATION EFFORTS
Subsequent analysis years after the completion of the case study building revealed more systemic water infiltration issues that were all interrelated. Findings and remediation efforts of this subsequent investigation included: 2015 A detailed investigation (figs. 12 and 13) was performed on previously reported leak activity. A report was delivered further characterizing leaks previously identified and recommending prioritized repair considerations. The following leaks were characterized: 1. “Field leaks” were further clarified to consist of: a. Metal-to-metal butt joints of adjacent cassettes of the unitized panels (fig. 14). b. Metal-to-metal butt joints of adjacent aluminum subframe ribs (fig. 15). c. Field-glazing joints between lites of the vision glass (fig. 16). d. Metal-to-metal joints of the cassette panels at the aluminum subframe (fig. 17). e. Factory-glazed joints within the unitized panels (fig. 18). 2. “Ends of copings at high roofs” was further clarified and attributed to aluminum subframe terminations through the structurally supported aluminum frame (fig. 19). 3. Probes at the high roof interface with horizontal transitions of the skylight coping revealed insufficient drainage and negative slope within the structural member (fig. 20). 4. A pinhole leak was identified in the newly replaced lower roof PMMA flashings (fig. 21). 2015 Selective reglazing of failed sealant joints and installation of preformed silicone flashings was completed throughout the field of the skylight to address failing butt joints of cassette and subframe constructions. These repairs resolved an estimated 95% of field leaks. 2016 A roofing replacement program was implemented at the high roofs that included selective remediation of the horizontal transitions of the skylight system (which remained in place rather than being replaced—or removed, salvaged, and replaced). 2017 A roofing installation program was initiated at a metal panel roof at an elevator tower to address: • Failures of horizontal metal panel roof sealant joints and expansion control at cable-supported curtain walls (fig. 22). Additionally, interior water management systems were installed at various interior leak locations due to: • Ventilation louvers above elevator towers that were installed incorrectly, causing water to corrode elevator components (fig. 23). • Interior water intrusion that overwhelmed horizontal skylight transitions, condensation pans, and weeps (fig. 24).
221
222
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 12 Simultaneous up-close inspection and testing of representative rib locations.
FIG. 13 Representative forensic testing near identified sealant deficiencies.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
FIG. 14 Metal-to-metal butt joints of adjacent cassettes of the unitized panels.
FIG. 15 Metal-to-metal butt joints of adjacent aluminum subframe ribs.
223
224
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 16 Field-glazing joints between lites of the vision glass.
FIG. 17 Metal-to-metal joints of the cassette panels at the aluminum subframe.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
FIG. 18 Factory-glazed joints within the unitized panels.
FIG. 19 Geometry issues at the primary sealant joints between the aluminum coping, extruded ridge, and horizontal section of skylight.
225
226
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 20 Insufficient drainage at front edge of coping.
FIG. 21 Overview of discontinuity in vertical flashing allowing water to infiltrate to the interior.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
FIG. 22 Horizontal metal panel roof.
FIG. 23 Elevator tower with vertically aligned louver blades that permitted wind-driven rain to enter, overrun, and leak to the interior.
227
228
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 24 Multiple leak locations along sloped glazing transition.
Economics and Importance When an investment in the construction of a building is agreed upon, it is generally done so with an understanding of costs, expectations of service life, durability, and maintenance. These forward-looking practices are no longer abstractions but are now viewed as industry best practice and part of the initial discussion to define the Basis of Design (BOD). The Owner Project Requirements (OPR) are expected to anticipate costs of ownership (including capital and maintenance expenditures) in a fashion similar to budget requirements for initial construction. A review of the project archive and interviews with staff did not turn up any pertinent information on the formulation of project-specific OPR and BOD documentation specific to observed failures at the case study building, but there were avoidable compromises made during the project that would have reduced the economic impact of the ongoing remediation. In broad terms, the construction process can be rife with errors and omissions, issues with sequencing, cost-savings measures that reduce building performance, and any other number of compromises that result in a decrease in the quality of the finished product. The industry has generally recognized that the earlier an exterior enclosure consultant is involved in the design process (either via BECx or other design assistance), the more value the practitioner can provide. One of the earliest and most significant findings during the investigation the case study building was that water penetrating through exterior joints of the skylight was often unmanaged
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
or it quickly overwhelmed the intended moisture management systems. This led to time- and volume-dependent leaks to the interior that were disruptive to occupants (figs. 25 and 26). Failures of building enclosure performance can become increasingly difficult and more expensive to resolve as the time increases from that decision (fig. 27). This is especially true for early design decisions related to system selection and performance criteria. Layers of complexity are added as a project is designed and eventually built. For this building’s skylight, adding in a robust interior moisture management system would have been relatively inexpensive compared to ongoing service life maintenance costs or a capital program that might seek to add in such costs. There are economic incentives in the use of integrated project delivery and prefabrication, especially for high-performance facades* for the construction industry’s largest and most significant projects where modern wall and glazing technologies are used. The industry is trending toward more early and integrated design and shop drawing phases for advanced wall assemblies in order to permit timely manufacturing. As such, final design decisions often occur earlier in the project FIG. 25 Exterior sealant joints corresponded with all steel truss framing members and failures leading to water intrusion followed the path of least resistance.
* High-performance commercial building facades, according to Lawrence Berkeley National Laboratory’s Building Technologies Program Senior Advisor Stephen E. Selkowitz, are comprehensive systems that incorporate daylighting, solar heat-gain control, ventilation, and space conditioning. One could argue that these strategies, as well as others, would vary for each unique project. The intelligent combination of relevant high-performance strategies delivers the greatest performance when engineered as an essential part of a fully integrated building design.
229
230
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 26 There were numerous potential pathways of water intrusion, many of which were found to change depending on time and volume. This figure shows water traveling from a single sealant failure.
FIG. 27 Defects in the design and construction phase as well as renewal in the service phase of a building’s life cycle will vary in magnitude. Complexity and expense for remediation generally increases the earlier a mistake or defect is initiated in the design and construction phase.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
than was traditionally the case with conventionally built buildings. Piecemeal value management can be identified through testing under controlled conditions, intended to simulate some or part of the natural environment, or, if not fully understood and left unevaluated, can lead to enclosure failures during a buildings service phase (actual use and operations). The case study building program fell victim to value management decisions that resulted in a decrease in performance. Through several snowstorms, horizontal skylight elements at end conditions of the skylight field were broken due to snow and ice cascading down the valleys of the barrel vaults. Subsequent water intrusion investigations also noted that the interior water velocity along skylight components was very difficult to control and often dripped to the interior at the 85 turn from vertical to horizontal. Following the investigation, the initial recommendation for a long-term and durable repair was to replace horizontal members with sloped members in order to further improve durability and drainage. Such a repair would have involved supplemental steel framing installation and new skylight components and was ultimately over budget for the institution (figs. 28 and 29). After making this repair recommendation, it was discovered that the proposed repair would have been very similar to the initial design but for a reported value management decision made during construction (figs. 30 and 31) that removed more steeply sloped valley glazing units from the scope of work. There is an order of magnitude difference between the initial cost, likely savings, and long-term solution to the problem created during the reported value management. As practitioners, we endeavor to avoid unintended consequences; managing risk can limit exposure, saving time, costs, and reputations. As high-performance building enclosures demand more forethought than building codes or zoning can anticipate, project teams rely on project-specific performance strategies defined in the BOD and OPR. These documents are the foundation of all BECx projects and are useful tools in advocating for the high performance of the facade throughout the duration of the project, from design to value management and through construction, whether or not BECx is formally utilized on the project. Development of a tool to help predict potential shortcomings or failures of high-performance buildings would be beneficial in considering how design decisions, material selection, sequencing, and so on could lead to failure. A tool that could facilitate educated decisions throughout the design and construction process, especially with regard to decisions surrounding value management, could improve upon the success of high-performance buildings. This was the impetus for the four orders of failure.
The Four Orders of Failure The four orders of failure is a taxonomy or classification system of building enclosure failures based on four dimensions of space and time, as related to component interfaces, system capacity, and temporal degradation, and is described at length in
231
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 28 Proposed repair strategy to create steeply sloped glazing.
232
FIG. 29 Proposed repair strategy to create steeply sloped glazing.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
233
234
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 30 Sloped glazing reportedly removed from the project as part of the value management process.
FIG. 31 Horizontal glazing used throughout the project on upper level roofs.
the authors’ publication, “The Four Orders of Failure in Building Skin Design and Construction,”1 which was presented at the Roof Consultants Institute’s symposium on Building Envelope Technology in November 2017. For the purposes of this discussion, we will review the orders of failure in abbreviated form. The four orders of building enclosure failures allow us to spatially correlate multidimensional interfaces of building enclosure components with failures. They help pinpoint relationships between failures by identifying common characteristics and remedies.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
This classification system can be used as a communication tool during the design, value management, and construction of building enclosures by organizing relevant information in a clear, succinct, and hierarchical manner, limiting shortsighted decisions made at the expense of long-term performance. First-order failures consist of linear openings, discontinuity, or nonuniformity within a component or contiguous components of the building enclosure assembly. Examples include holes in a sealant joint, blocked weep holes, unsealed penetrations in a barrier membrane, or damaged structural members. Generally, these failures are local defects, typically found at a discrete point in the component’s surface, running from the exterior to the interior. They can be repetitive and may be the result of deficiencies on-site or in the fabrication plant. First-order failures observed at the case study building included poor sealant joint geometry at setting blocks (fig. 32) and adhesion issues related to workmanship (fig. 33). These failures might have been captured with a robust QA/QC program. Second-order failures generally consist of defects or deficiencies in assembly components within the enclosure, such as planar gaps or joint discontinuity. Examples include lack of sealant adhesion to incompatible substrate, improper lapping of flashing materials allowing water to enter in an uncontrolled manner, or lack of provision for differential movement at structural connections. Second-order failures at the case study building resulted in a significant volume of water infiltration that was disruptive to operations. The butt joints and panel end
FIG. 32 Shadow of setting block, resulting in inadequate sealant joint profile.
235
236
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 33 Removed sealant joint showing adhesion failures at edge.
cap butt joints of the extruded aluminum capture system were sealed to one another such that the primary weather seal had less than 1/8-in. bite when installed (fig. 34), which was found to create widespread linear sealant joint failures over time.
FIG. 34 Failed joint at cassette butt joint due to minimal sealant bite.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
Following characterization of the leak activity, it was determined that the second-order failures were causing the most systemic problems related to the volume of water infiltration. Addressing the issues of transition joints resolved approximately 95% of the most significant leakage. Third-order failures are those of size, volume, or design capacity. Improper material selection, omission of preinstallation laboratory testing, or improper engineering or design may lead to a failure to comply with performance requirements of the building enclosure system. A third-order failure can be a failure of a system that was installed in accordance with design documents but whose capacity or performance capability has been exceeded by actual environmental conditions, user expectations, or regulatory requirements. These failures may be due to forces not anticipated by building codes that overload design capacity, such as extreme wind loads or weathering, or they may be the result of conflicting or unanticipated performance requirements. Third-order failures can also be a result of limitations in capacity or performance caused by aesthetic requirements, such as low or no curb height for proper termination of roof membranes, oversized panels, and complex geometries. Examples include improper flashing assembly height for required water resistance, inadequate insulation thickness, and excessive deflection of oversized glass panels. The skylight at the case study building was designed with approximately 100,000 linear feet of weather seals, approximately 80% of which were field glazed. The skylight was designed to manage a small volume of condensation through back dams of the extruded aluminum subframe, but designers did not anticipate the potential for future water leaks through the nearly twenty miles of sealant joints. This unanticipated drainage requirement means that upon any aging or failure of sealants, the skylight quickly becomes overwhelmed by water intrusion (refer to fig. 35 for a representative water intrusion plan). This results in a third-order failure, exceeding the design capacity of the system, where water immediately drips to interior floors of the building. This third-order failure could have been anticipated and compensated for in the design phase by facilitating collection and drainage of ancillary water, if workmanship and long-term performance had been part of the conversation. Fourth-order failures cannot be avoided; they are related to progressive degradation of assemblies and components over time caused by repeated use, exposure to elements, seasonal changes, freeze/thaw cycles, weather, and environmental conditions. Examples include premature degradation of factory-glazed weather seals and corrosion of metal components. Design redundancy, postinstallation performance verification testing, proper maintenance programs, and timely repairs are important for reducing fourth-order failures. Consideration of future building codes and industry standards in the design of adaptable high-performance building enclosures will help ensure long-term performance. Unanticipated water infiltration within the skylight system at the factoryinstalled sealant joints has led to the premature degradation of the PVB interlayer at intermittent laminated glass edges (fig. 36), resulting in a fourth-order failure.
237
238
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 35 Significant leaking before and during remediation. Blue areas denote previous work areas, green areas denote resolved leaks, and red areas denote ongoing leak activity during work period.
FIG. 36 Up-close view of interlayer delamination at glass edge that may indicate future edge stability issues of glass as the skylight ages.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
The factory-installed weather seals failed prematurely and required replacement five years postoccupancy. This failure could have been prevented in the design phase through careful review of the specifications and in the construction administration phase through a thorough review of the submittals and fabrication quality assurance reviews. Fourth-order failures can be addressed through considerations for serviceability of elements that will degrade more readily than others, such as sealants compared to glass. Incorporation of a custom house stage to readily access all of the panels within the skylight for routine maintenance could also prolong the onset of fourth-order failures. Fourth-order failures can result in the occurrence of lower-order failures. Longterm exposure of the roof coping to ultraviolet (UV) light can result in first-order cohesive failure and second-order adhesive failure of the sealant joints. Freeze/thaw cycles can cause second-order cracking in the mortar joint below the coping. Weathering and movement can cause third-order cracking and displacement of the coping. High temperature variations, combined with radiant energy reflected from neighboring structures, can cause deterioration of the through wall flashing, which can result in first-order water intrusion around the reinforcing bars supporting the coping. These de-escalations of the orders of failures can be characterized as a cascade of the orders of failures. Conversely, lower-order failures can result in the appearance of fourth-order failures. First- and second-order water infiltration in laminated glass can cause third-order delamination. This can result in fourth-order preliminary yellowing or bubbling of the PVB interlayer at the edges of the skylight glass, as seen at the case study building.
Characterization of Failures The four orders of failure helped to categorize and understand the failures observed at the case study building. Failures were not as they initially seemed. For example, what was initially purported to be a series of intermittent leaks resulting from individual sealant failures turned out to be sealant failures at the aluminum frame butt joints, resulting in a capacity issue. The amount of water from sealant joint failures overwhelmed the capacity of the skylight to collect and drain the water because the design was limited to condensation-related moisture. Understanding the orders of failure and how they can be compounded to a cascading effect of orders of failure was helpful in our investigation, analysis, and remediation design. The failures of the barrel vault skylight roof system were multifaceted and involved several different modes of failure. Our investigation revealed the design and construction deficiencies discussed in the following sections. Ends of Copings at High Roofs
Leaks emanating near the end condition of the barrel vault terminations at the high roof copings was a second-order failure. As previously discussed, water was found to penetrate through the exterior seals directly to the interior of the building where the EPDM transitions beneath the structural coping supported horizontal skylight
239
240
STP 1617 On Building Science and the Physics of Building Enclosure Performance
sections. The complication of this detail was that the substructure of the skylight (extruded aluminum ribs) penetrated the transition detail and was continuous from interior to exterior space. This interruption in the assembly created a transition of three materials (glass, aluminum, and roofing) in three different planes, leading to a complication in the termination sealant geometry and to linear second-order failures in the plane of watertightness. To remedy this second-order failure, the copings were removed, and fins of the skylight substructure were trimmed back to simplify termination sealant transitions and geometry. The roofing material at this interface was changed to PMMA liquidapplied flashings so that transitions between assemblies could be simplified. The structural coping was fully flashed to create a better defined linear waterproofed curb (figs. 37 and 38). This failure could have been avoided if the unitized skylight assembly design had provided accommodations for end terminations with special units or custom shapes of the subframe ribs. The system could have been formally terminated at a structural member and curb on the roof line rather than continuing through the plane of watertightness. Perimeter Leaks at the Barrel Vault Terminations along Low Roofs
Leaks emanating from ponding water at the barrel vault-to-roof termination was a third-order failure. EPDM roofing was installed over an insulated deck with only a few inches of vertical separation or turn-up at the glazed skylight terminations. FIG. 37 Existing coping conditions prior to remediation.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
FIG. 38 Coping remediation in progress.
Over time, ponding water at the roof perimeter adjacent to the skylight created service conditions that allowed the ponding water to bypass base flashings, becoming a regular source of water intrusion. The EPDM was removed, and the system was changed to an IRMA using an SBS-modified roof with PMMA flashings. This permitted an increased vertical roof termination height at the edge of the skylight by installing the roof membrane below the insulation. The vertical roof termination occurred behind unitized skylight panels and subframe in order to permit drainage of moisture that bypassed the skylight at higher elevations and was cascading down the interior skylight framing components to the exterior roof (figs. 39 and 40). Because it is an industry standard to provide a minimum of 8 in. of vertical clearance at roof terminations, the unitized skylight should have been terminated above the roof at a reasonable distance to ensure roof drainage issues would not affect skylight performance. Similar to the high roofs, the unitized skylight assembly should have accommodated the end terminations with special units over the structure to provide an uninterrupted waterproof transition. Field Leaks
Water leaks dripping from within the field of the skylight onto interior lobbies or theater roofs was a pervasive issue. At first, they were thought to be the result of intermittent sealant failures, but upon investigation (figs. 41 and 42), they were found to be the compounded result of failures of all orders (first through fourth) acting together to create water intrusion (figs. 43 and 44).
241
242
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 39 Revised roof termination. Note primary waterproofing is located at deck to increase clearances.
FIG. 40 Typical section through skylight showing unitized panel installation over steel frame.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
FIG. 41 Representative forensic spray testing using BOT with multiple spray nozzles directed at the skylight on wheels to ride ribs.
FIG. 42 AAMA 501.2 testing in representative areas from crane basket.
243
244
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 43 Example of sloped members corresponding to exterior weather joints.
FIG. 44 Example of a potential pathway of water intrusion.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
The skylight was designed with approximately 100,000 linear feet of weather seals, approximately 80% of which were field glazed. Field leaks occurred for a variety of reasons including: 1. Metal-to-metal butt joints of adjacent cassettes of the unitized panels failed linearly throughout the field of the skylight due to the thin bite of the butt joint to the cassettes (1/8 in., typically) creating a second-order failure as the skylight moved and aged. This issue was discovered during up-close observation along the field of the skylights and representative testing of failed field-glazed butts of cassette joints, revealing that the failures proved to be one of the main sources of water intrusion through the skylight. This item was addressed via selective installation of preformed silicone that was bedded over the cassette joint to create a more durable barrier than the sealant joint spanning the reduced bite of the cassette ends (figs. 45 and 46). 2. Metal-to-metal butt joints of adjacent aluminum subframe ribs failed linearly throughout the field of the skylight due to the thin bite of the butt joints of the framing elements (1/8 in., typically) creating a second-order failure as the skylight moved and aged. The failure mode and contribution to overall leakage to the building interior was all very similar to the identification of the metal-to-metal butt joints of the glazing cassettes. This item was
FIG. 45 Use of preformed silicone to span cassette butt ends. Detail view.
245
246
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 46 Use of preformed silicone to span cassette butt ends. Field view.
addressed via selective installation of preformed silicone that was bedded over a prepared subframe assembly by trimming ribs at the face of the frame to permit a clear span of the preformed silicone (figs. 47 and 48). This created a more durable barrier than the sealant joint spanning the reduced bite of the subframe ends. 3. Field-glazing joints between lites of the vision glass were found to have intermittent failures, which appeared to be installation failures created at various points where defective installation of sealant tooling or backer rod installation was noted (fig. 49). Such point failures are considered first-order failures. During up-close inspection and representative testing, these failures were noted to be very minor contributors to the overall water intrusion around the building. Where observed during the remediation program, these failures were selectively cut out and replaced with new sealant. 4. Metal-to-metal joints of the cassette panels at the aluminum subframe were initially intended to be sealed with backer rod and sealant; however, due to field tolerances of steel and the skylight system, the joints were typically sealed with a fillet joint between a 1/4-in. flange of the subframe and the cassette (fig. 50). Representative testing revealed that what appeared to be a linear second-order failure of the joint actually contributed very little to water intrusion within the building. We suspect this was due to how little individual cassettes move in relation to local subframes and that sealant was found to extend into even small joints that should have had backer rod.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
FIG. 47 Use of preformed silicone to span butt joints of aluminum subframe at ribs. Detail view.
FIG. 48 Use of preformed silicone to span butt joints of aluminum subframe at ribs. Field view.
247
248
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 49 Intermittent variations of sealant geometry due to twisted backer rod caused first-order failures observed within the field of glazing.
FIG. 50 Reduced tolerances created linear second-order failures intermittently as the sealant butt joint varied to fillet joints at the 0.25-in. cassette flange.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
5. Factory-glazed weather seals at the perimeters of the unitized panels deteriorated prematurely. This fourth-order failure caused significant water intrusion throughout the field of the skylight and led to a capital program to replace the sealants five years after completion of the project. As noted previously, perhaps the most significant finding within the field of the skylight was not that sealant failure contributed to water bypassing the barrier system of the skylight but instead that the skylight was not designed to manage anticipated water intrusion to accommodate this water leakage through the sealant joints. Only small amounts of condensation were capable of being managed through back dams of the extruded aluminum subframe. This unanticipated drainage requirement meant that upon any aging or failure of sealants, the skylight quickly became overwhelmed by water intrusion, exceeding the design capacity of the system and resulting in a third-order failure. The water immediately dripped to interior floors of the building. Unfortunately, this drainage capacity issue will likely be exceedingly complicated, invasive, and cost prohibitive to remediate. Long-term water intrusion through the sealant joints may also lead to glass edge stability issues if the PVB interlayer of the laminated glass is exposed to longterm moisture ingress. This fourth-order failure is time dependent but can be exacerbated by failures of the sealant joints. During the design and selection of the system, sealant failures should have been anticipated and a more robust moisture management system implemented within the skylight. Mock-up testing could have been used to validate performance of moisture management systems at different service angles, with the selective removal of primary sealant joints from the mock-up simulating anticipated future aging and failures of sealants. Valley Leaks at Upper Roof Horizontal Skylight Transitions
Aside from leaking directly into the lobby, time- and volume-dependent leaks causing water intrusion to the interior from the field of the skylight also had the tendency to track down the steel strong back (Vierendeel truss) structural frames supporting the skylight. On the exterior of the building, water, snow, and ice formation followed a similar drainage path, melting along the valleys of the skylight and tracking down to the base of the barrel vault. The velocity with which the system shed water, snow, and ice accumulation was problematic for in-service conditions, especially because of the vertical-to-horizontal transition that was made at the skylight support and horizontal lite on the upper tier roofs (figs. 51 and 52) (Note: As previously noted, this vertical-to-horizontal skylight transition was constructed in lieu of a more steeply sloped glazing element, which was a decision made based on value management for the project.) For both interior and exterior drainage, this value management decision created several third-order service failures (issues of design capacity).
249
250
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 51 Vertical-to-horizontal transition on exterior of building. Note low clearance of coping to roofing.
FIG. 52 Detail section of vertical-to-horizontal transition at valley of skylight.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
One of the most acute service issues noted by ownership with regard to the horizontal glazing is that in rare cases, where snow and ice form along the barrel vault of the skylight (this happens rarely due to the nonthermally broken frame of the skylight), accumulations that eventually slip can cascade down the valley of the barrel vault and cause breakage of the horizontal glazing members. This has happened on several occasions, leading to glass replacement programs. These events also have the tendency to deflect the extruded aluminum frame that supports the skylight, leading to standing water that can sit against weather sealants on the exterior of the coping (figs. 53 and 54). Probes at the high roof interface with horizontal transitions of the skylight coping also revealed insufficient drainage and negative slope within the structural pan, creating an additional third-order failure and further reducing the ability of the skylight system to drain. Interior water intrusion was found to overwhelm horizontal skylight vertical-to-horizontal (85 transition) condensation pans and weeps (a third-order failure) due to the unanticipated volume of water entering the interior and the velocity with which it traveled down the steel frame. To further exacerbate the water intrusion, however, there were also several fastener connections made through the primary horizontal drainage plane, creating various first-order failures (fig. 55) that also contributed to increased leak activity and precipitated the early degradation of intumescent fireproofing at columns (fourth-order failure)
FIG. 53 Standing water at horizontal copings at base of skylight.
251
252
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 54 Extruded aluminum frame supporting horizontal glazing element.
nearest the leak activity. Another factor further complicating the detail and overall watertightness of the system was that the extruded aluminum subframe rib was continuous through the entire drainage assembly (fig. 56), creating an awkward geometry for the existing weather sealants as well as for the eventual repair. The main challenge of the remediation was that the existing structure and skylight system were required to remain largely the same (primarily due to the expense of a replacement program) while improving watertightness of the exterior and drainage capacity of the interior. On the exterior of the horizontal skylight assembly, the prime directive was to simplify the assembly and to align with industry best practices where possible. The decision was made to trim the rib from the extruded subframe in order to reduce the risk of a direct path of water to the interior (fig. 57). The simplified geometry also permitted more robust horizontal waterproofing terminations (fig. 58) to encapsulate fasteners through the horizontal drainage plane. Additionally, an inverted roof membrane system was chosen in order to improve termination heights of the roofing. In order to improve the drainage capacity of the skylight system, weeps were added to the end condition of the extruded aluminum subframe in line with previous weep holes in the extruded aluminum structural coping (fig. 59). This addition permitted additional drainage above the new waterproofing and below the new coping (fig. 60).
FIG. 55 Water leakage pathways due to fasteners installed through horizontal skylight transition.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
253
254
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 56 Geometry issues at the primary sealant joints between the aluminum coping, extruded subframe, and horizontal skylight.
FIG. 57 Trimmed ribs and preparation for reroofing.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
FIG. 58 Liquid-applied flashings installed over the extruded structural aluminum coping.
FIG. 59 Weeps and diverters installed at base of extruded aluminum subframe.
255
256
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 60 New coping installed over structural coping permitting continuous sealant joint across base of ribs.
On the interior of the building, internal gutters were added beneath end conditions in order to capture remaining water that bypassed any of the existing interior moisture management systems (fig. 61). Other Miscellaneous Failures
Various other elements of the building enclosure suffered from failures that have necessitated capital programs and an increased maintenance budget, including: 1. Elevated daytime temperatures rendered an interior roof garden unusable, making it impossible to keep plants alive. It was noted that operable vents through the barrel vault skylight that may have helped with this issue were not being used due in equal parts to operational failures, water intrusion, and perceived energy costs. The overwhelming heat exceeded the anticipated service condition of the space (a third-order failure) and necessitated the removal of the garden and the addition of a second building enclosure at the top of the theater where the garden once existed. The subsequent enclosure has separate environmental conditioning and electrochromic glass in order to better control glare and solar heat gain. 2. Failures of horizontal metal panel roof sealant joints and expansion control at cable-supported curtain walls were noted to cause water intrusion through a connector bridge between an elevator tower and the roof garden space. The primary failure responsible for the water intrusion was related to the failure of horizontal sealant joints at a metal panel roof (a second-order
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
FIG. 61 An interior gutter constructed of stainless steel was added to help manage water intrusion.
failure). The assembly did have an internal gutter system to manage water through sealant joint failures; however, this gutter was overwhelmed and was not large enough to capture water from all joints. The metal panel roof was reroofed with a more durable PMMA roofing system in order to reduce the time between maintenance cycles (figs. 62 and 63) 3. Significant and long-term water intrusion was noted at the elevator tower, causing corrosion of elevator controls that necessitated significant maintenance and replacement of electrical components. Upon investigation, it was discovered that ventilation louvers at the top of the elevator tower were installed with louver blades oriented vertically rather than horizontally, thereby causing an accumulation of water in the sill pan of the louvers that was uncontrolled (a third-order failure) and that leaked into the adjacent curtain wall, onto interior gypsum, and into the elevator cabs. The mobilization of the louvers in the elevator tower proved cost prohibitive, and an interior water-diverting system was installed to protect elevator components.
257
258
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 62 Metal panel roofing at connector between building and elevator tower.
FIG. 63 PMMA flashing and preformed silicone counterflashing at curtain wall installation, in progress.
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
Strategies to Predict and Prevent Building Enclosure Failures SERVICEABILITY AND DURABILITY
Building enclosure components will deteriorate over time due to seasonal changes, severe weathering, repeated usage, and exposure to UV light. These temporal degradations generally become evident in multiple dimensional interfaces of components, from linear gaps where air can infiltrate and planar adhesive failures of sealants to three-dimensional overload of a system’s ability to resist bulk water intrusion. Without reconstructing entire portions of the building enclosure, which is often not feasible from an operational or cost perspective, many repairs of these fourth-order failures involve limited or interim solutions that diminish the intended design performance of the system. To fully understand potential failures and the effect on a building enclosure, a designer must take a holistic view of building systems, components, and assemblies and the impact they can have on overall performance. If we consider the approximately twenty miles of sealant joints on the case study skylight, the cost and access limitations on regular maintenance, as well as the intensity of UV exposure, consideration of failing sealant joints could have led designers to increase the capacity of the skylight frame to capture water from failed joints and drain to the exterior. Unanticipated performance and serviceability requirements resulted in cascading orders of failures defined by a compounding effect that lower-order failures might have in causing higher-order failures, such as decreasing the service life of enclosure components and assemblies (as can be demonstrated in the section detailing valley leaks at upper roof horizontal skylight transitions). When failures cascade, characterization of deficient conditions may be complicated because the relationship between cause and effect of related assemblies and failures can be convoluted. Of primary importance to the higher-order failures are issues that are often sorted out early in the design process through identification of the OPR and design intent. While cost and schedule may dominate the construction process, it is recommended that the project requirements and design intent be referenced throughout the project, not just during the design phase. Efforts to drive cost and schedule may adversely affect previous decisions related to service life, durability, and redundancy. Designers and fabricators have a responsibility to understand how a building enclosure will fail or at least anticipate how decisions might affect performance requirements and other failure typologies. As architects and building enclosure consultants, we must continue to be advocates for performance throughout the entire project and educate project teams so the impact of design changes are clearly understood and informed decisions are made.
259
260
STP 1617 On Building Science and the Physics of Building Enclosure Performance
VALUE MANAGEMENT
High-performance building enclosure design is often modified through value management. Sessions are often conducted in high-stress conditions when the project is about to go to bid or has started construction, meaning they are often given lower priority and insufficient time. Consequently, there is often not sufficient time and priority to fully understand the impact on the performance requirements of the project or the cascading effect on the design and performance of adjacent systems. Evaluations are often performed piecemeal, without a thorough understanding of system impacts, and focus on cost and schedule reduction, increasing the risk of reduced performance and inefficiency. This can compromise quality assurance and long-term performance. The unintended consequences of these decisions can lead to additional time and cost to compensate for proceeding out of sequence. Consider the horizontal glazing at the base of the barrel vault. This detail originally provided substantial slope in anticipation of managing the water, snow, and ice that would cascade down the valleys of the barrel vault. However, it is apparent that during a value management exercise, the skylight end units were modified from an offset sloped panel to a horizontal panel (figs. 30 and 31). This caused the weep system to be unable to manage water infiltration cascading down from the interior side of the large field of the skylight and resulted in several broken glazing panels as snow and ice cascaded down the valley. Twenty years later, the hard costs and operational disruption to provide slope at the horizontal termination of the skylight is prohibitive. Working within the capital repair budget and limited ability to disrupt operations, the repair options are restricted in their capacity to address the design limitations. The damage sustained on the interior of the building steel, drywall, and other finishes remains an ongoing operations and maintenance problem due to the inability of the weep systems to facilitate serviceability requirements of the aging skylight sealants. Building enclosures are designed with good intentions but often succumb to performance failures as a result of uninformed value management decisions. These performance issues are often repaired with solutions that are unable to address life cycle considerations and system durability due to financial limitations and a low tolerance for operational disruptions. Better remedies and maintenance options, informed processes, and better management of design and construction can allow new building enclosures to perform at a higher level. If the project team had considered the long-term impact of the value management decision regarding the horizontal termination of the case study skylight, it is possible ownership might have considered another approach to cost savings. QUALITY ASSURANCE AND QUALITY CONTROL
Quality assurance can be defined as understanding the requirements necessary to achieve quality and developing the road map to get there. Control, as defined by ANSI/ISO/ASQ A3534-2, is “an evaluation to indicate needed corrective
KEEGAN AND RIDGWAY, DOI: 10.1520/STP161720190154
responses,” indicating that quality control, as defined in the realm of construction, is comprised of the operational techniques and activities used to fulfill requirements for quality.2 Dr. Joe Lstiburek elaborates on the difference between quality assurance and quality control. If you do the wrong thing right, it’s still wrong. Right? This is the basic difference between quality control and quality assurance. Quality assurance is figuring out what the right thing to do is. Quality control is executing it.3 With this in mind, architects and building enclosure consultants are charged with quality assurance. In the design phase, we use tools such as building codes and standards and best practices to inform our designs and detail-specified performance requirements. Enhanced performance warranties are specified to hold manufacturers and contractors accountable for the performance of the installed systems. Manufacturers and contractors are responsible for the quality control aspect of construction. This can include a quality control manager walking the production line or the jobsite with a checklist to verify the crew is complying with the necessary operational activities and producing an end product that meets the quality requirements for the project. Quality assurance efforts continue in the construction phase through performance mock-ups, performance verification testing, and periodic field observations. The more robust these efforts, the greater the assurance the quality control efforts are fulfilling the requirements for quality. As architects and building enclosure consultants, we can provide more vigorous and well-defined requirements within our specifications to drive quality control efforts. The factory-glazed sealant joints within the unitized panels at the case study building were replaced five years post occupancy; this short-term fourth-order failure resulted in the need to reglaze the approximately 20,000 linear feet of shopglazed primary weather sealant joints in the field. This should have been properly specified in the design process, rejected in the construction administration phase, and escalated during manufacturing plant inspections. A rigorous QA/QC program could have addressed this systemic failure before the unitized panels were installed.
Conclusion The four orders of building enclosure failure is a classification system that spatially correlates the multidimensional interfaces of building enclosure components with the building enclosure failures. This classification system can assist building enclosure professionals to present relevant information about project requirements and potential failures in a clear, ordered form to our owners, contractors, and project teams and thereby contribute to holistic decision-making to ensure long-term performance. This methodology can be used as an analytical tool to promote innovative adaptable high-performance building enclosures that can address changing environmental and regulatory requirements.
261
262
STP 1617 On Building Science and the Physics of Building Enclosure Performance
This case study demonstrated that there is a relationship between various types of failures and the effects on one another. In the case of the studied building, third-order failures (failures of capacity, such as was illustrated in general moisture management and vertical-to-horizontal transitions of the skylight system) and fourth-order failures (such as deterioration of spray interior fireproofing and delamination at PVB interlayer) were not apparent until unanticipated first- and second-order failures at the primary weather seals of the skylight occurred. The fact that so many of these lower-order failures occurred relatively soon in the service period of the building should be a cautionary tale. This is an example for what can happen when practitioners do not anticipate actual service conditions of the building. This case study also illustrates the importance that anticipating failures within the organization of the BOD can have on the design, construction, value management, and QA/QC programs to minimize unanticipated modes of degradation that may later drive unbudgeted costs.
References 1.
2. 3.
J. Keegan, M. Ridgway, and J. Ng, “The Four Orders of Failure in Building Skin Design and Construction” (paper presentation, Roof Consultants Institute’s Symposium on Building Envelope Technology, Raleigh, NC, November 13–14, 2017). American Society for Quality, “Six Sigma Terminology,” 2018, http://web.archive.org/web/ 20181118193443/http://asq.org/sixsigma/quality-information/termsq-sixsigma.html J. Lstiburek, “BSI-039: Five Things,” Building Science Corporation, April 2010, http:// web.archive.org/web/20180502152014/https://buildingscience.com/documents/insights /bsi-039-five-things
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180065
Andrea DelGiudice1,2
Two Agencies, BECx, and Design-Build: Challenges and Opportunities for BECx with Multiagency Involvement and Design-Build Project Delivery Citation A. DelGiudice, “Two Agencies, BECx, and Design-Build: Challenges and Opportunities for BECx with Multiagency Involvement and Design-Build Project Delivery,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 263–279. http://doi.org/10.1520/ STP1617201800653
ABSTRACT
Building Enclosure Commissioning (BECx) is not a new concept, and many documents, standards, and codes reference or attempt to define it. References to and discussion of BECx can be found in myriad documents, from codes such as the International Green Building Code (IgCC) and California Title 24 to voluntary standards such as Leadership in Energy and Environmental Design (LEED) v4. The process is discussed in detail in the National Institute of Building Sciences (NIBS), Guideline 3, and referenced in ASHRAE 202 (ANSI/ASHRAE/IES Standard 202-2013, Commissioning Process for Buildings and Systems). It is used on a wide variety of projects both private and public, and various entities also have their own requirements or approaches to BECx. Further, the process is intended to be possible regardless of project or owner type or project delivery method. However, the applicability, efficiency, and even viability of BECx for a design-build project is frequently questioned and challenges to effective BECx in design-build are cited. Further challenges can be raised when multiple BECx Manuscript received October 3, 2018; accepted for publication April 2, 2019. 1 Wiss, Janney, Elstner, and Associates, Inc., 2941 Fairview Park Dr., Falls Church, VA 22042, USA 2 National Institute of Building Science BECx Education Program, 1090 Vermont Ave. NW # 700, Washington, DC 20005, USA 3 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
263
264
STP 1617 On Building Science and the Physics of Building Enclosure Performance
requirements come into play or where there are multiple entities, such as multiple government agencies, involved in the project. This paper will discuss the challenges as well as the benefits of implementing BECx on a design-build project for the federal government with multiagency involvement, specifically through the lens of two projects developed by the U.S. Army Corps of Engineers (USACE) for a federal agency using the ASTM E2813, Standard Practice for Building Enclosure Commissioning, enhanced process. Keywords building enclosure commissioning, design-build, ASTM E2813
Introduction The Building Enclosure Commissioning (BECx) process is not a simple one-sizefits-all checklist. The process is simple: document project requirements and evaluate their inclusion in the project throughout the entire design and construction process and provide further review of the completed and occupied building to validate that the final product meets the documented requirements. The variety of requirements, users, design and construction processes, and project-specific conditions is expansive. Therefore, all commissioning (Cx) requires that the Owner’s Project Requirements (OPR) are established and documented (the OPR document) and a projectspecific plan is established and documented (the Commissioning Plan). ASTM E2813, Standard Practice for Building Enclosure Commissioning,1 provides core competencies required of the BECx Group (BECxG) in Section 4.2.1 because this group must be able to appropriately apply the listed competencies to the project at hand. Building type and expected use, life span, and project location influence the OPR as do the wants and needs of the stakeholders. Project delivery method influences the project-specific commissioning plan as do the type and arrangement of stakeholders. Many atypical contracting or project-delivery combinations raise questions about how the BECx process can be effectively applied to a particular project. This paper will explore the BECx process specifically through the lens of commissioning a design-build project, primarily comprised of office and support space for a federal agency, where the project is functionally managed through construction by the U.S. Army Corps of Engineers (USACE), the end user is a second government agency, and the BECx process is in general accordance with enhanced commissioning per ASTM E2813.
Exemplar Projects The author has been a member of the BECxG for two projects that fit the aforementioned description—one completed circa 2012 and one ongoing, where USACE retained a Commissioning Provider (CxP) for the project, who in turn hired a BECxG that included the author (see fig. 1 for roles and relationships). The end user for both projects is the same federal agency (hereinafter referred to as the owner
DELGIUDICE, DOI: 10.1520/STP161720180065
FIG. 1 USACE total building commissioning process organization roles and responsibilities for design-build acquisitions.3
agency). The projects are herein after referred to as the completed project and ongoing project.* Both of these projects fit the contractual relationship defined previously. What follows is not a case study of these projects; however, the author’s experience informs the concerns discussed relating to design-build BECx and multiagency involvement. Wiss, Janney, Elstner, and Associates (WJE) was involved early enough on both projects to provide OPR input and provided extensive input into the design-build request for proposal (RFP) on the ongoing project. Further, the author’s participation in ASTM as the technical contact for the ASTM E06.55.09 Task Group, which houses both ASTM E2813 and its companion guide, ASTM E2947, Standard Guide for Building Enclosure Commissioning,2 influences her understanding of challenges and perceived obstructions to providing BECx services. In particular, comments regarding the applicability of ASTM E2813 to “alternate delivery methods” (where the standard method is defined as Design-Bid-Build) have been submitted to and evaluated by the Task Group multiple times during the ASTM balloting process. * Actual project names, agency names, and locations are not releasable information and have limited relevance to the information discussed in this paper.
265
266
STP 1617 On Building Science and the Physics of Building Enclosure Performance
The owner of the exemplar projects is the federal government; however, the government is represented by two agencies: USACE shepherds construction, and the owner agency occupies and maintains the building. The design-build and commissioning teams are both contracted through USACE, as shown in figure 1. Note that USACE Design-Bid-Build relationships are illustrated in figure 2. In both diagrams, the project team is entirely contracted beneath USACE, and the owner agency role is not listed in either diagram. USACE acts in a construction manager (CM)-like role but with total contractual authority. The owner agency is more like an owner in that it will occupy as well as maintain the building; however, the government, not any particular agency, ultimately retains building ownership. This is more complicated than projects with a single entity owner who hires a separate entity as the CM and a separate design-build contractor and BECx G. Private sector owners can choose to contract directly with the design-build contractor and are not obligated to run that contract through a CM, nor are they obligated to hand over total contractual control to a CM.
Selecting a Standard While there are many documents that govern Department of Defense (DOD) projects, they are far from prescriptive about if or how BECx should be implemented.
FIG. 2 USACE total building commissioning process organization roles and responsibilities for design-bid-build acquisitions.3
DELGIUDICE, DOI: 10.1520/STP161720180065
Election to implement BECx is not required on all USACE projects but is provided as an option that may be selected. The federal guidance and policy documents governing commissioning are intended to cover a broad variety of projects; the United Facilities documents are intended to cover construction projects for all branches of the military—extending across land, sea, air, and space,*4 and they include a wide variety of project types and requirements (or lack thereof) relative to the building enclosure. Once an owner decides to implement BECx on their project, the BECx scope must be determined and a BECx Provider brought into the project. For USACE, BECx scope may be influenced by a number of documents including: executive orders, owner agency policy, and Department of the Army Standards. Most notably, the BECx scope will be influenced by Engineering Regulation 1110-345-723, Engineering and Design: Total Building Commissioning Procedures3 and the United Facilities Guide Specification, Section 01 91 00.15, Total Building Commissioning.5 While both of these documents significantly influence the entire building Cx, including the BECx scope, neither are completely rigid prescriptive documents, and their treatment of enclosurespecific requirements is very limited. In addition to the federal requirements, there are myriad documents and guides about BECx that the owner may choose to apply to the BECx scope of their project; however, as Hopps and Burhoe6 discuss, ASTM E2813 provides an industry accepted standard developed through the ASTM consensus process. Unlike many BECx documents in the industry, ASTM E2813 is not a guide but, as the name of the standard indicates, a standard practice written with mandatory language. The ASTM E2813 standard provides two enforceable levels of BECx—fundamental and enhanced—and specifically outlines tasks and performance tests for each level. Additionally, ASTM E2813 may be used with the aim of meeting Leadership in Energy and Environmental Design (LEED v4—often applicable to federal projects) BECx goals.{7 While adaptations must be implemented when applying ASTM E2813 scope on a federal project where there are BECx representatives on the design, construction, and federal government teams, enhanced BECx (per ASTM E2813 scope) was selected by the BECxG as the foundation for the BECx process for both exemplar projects precisely because of the mandatory language and defined scope.
* The United Facilities Guide Specification (UFGS) is approved by MIL-STD-3007F for use by the Departments of the Army, Navy, and Air Force (which includes the National Aeronautics and Space Administration [NASA]). { It should be noted that LEED v4 fundamental and enhanced levels do not correlate to fundamental and enhanced ASTM E2813 levels. ASTM E2813 can be implemented at the fundamental level to meet the requirements of LEED v4 enhanced enclosure commissioning.
267
268
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Multiagency Stakeholders—Challenges and Opportunities STAKEHOLDERS
All projects have a wide variety of stakeholders. Federal projects with the twoagency arrangement explored here layer the complications of merging multiple standards onto a stakeholder arrangement where there are two government entities but only one has contractual decision-making power. While multiagency stakeholder arrangements can be analogized to relationships such as developer or construction manager and owner (e.g., USACE and the projects discussed here), or landlord and tenant (e.g., the General Services Administration as “the government’s landlord”), the relationship is not quite the same as that description may imply in the private sector. Both parties are “the government” and may also have divergent priorities. USACE may primarily judge the success of a project on schedule and budget management, while the end user, the owner agency, is likely to base their estimation of building success upon factors such as operation, comfort, aesthetic appeal, and maintenance requirements, which may not be fully evaluated until after the building is occupied. USACE has pointed out that many of the USACE personnel who work on these projects will never spend any time inside the completed building. In the completed project, members of the USACE team judged the project as largely unsuccessful because of budget and schedule overruns, while the owner agency viewed the project as highly successful because they generally like their building. As the agency and USACE began development of the OPR and designbuild RFP for the ongoing project, it became very important for the BECx and the greater Cx team to understand this varying opinion and the effect these views had in driving proposed OPR requirements. Further, the BECxG must also bear in mind that, while the owner agency occupies the building in a way that is like a typical owner, the overall government (the true owner) may choose different priorities than the owner agency. The government will continue to own and may rehabilitate and use the building for fifty or more years, potentially for a very different user, long after the owner agency moves on to another building.
Managing Multiple Stakeholders and OPR Development—Practical Concerns In order for the BECxG to effectively fulfill their role on the USACE design-build team, the BECxG must help USACE to fulfill their responsibilities to the owner agency. To achieve this, it is critical for the owner agency to be party to BECx discussions, and it is critical that their priorities are tracked along with USACE concerns throughout the entire process. In many respects, this is similar to a lesscomplex BECx project where all stakeholders are brought to the table during early phases of OPR development. So that the project can continue to move forward, it is
DELGIUDICE, DOI: 10.1520/STP161720180065
necessary to have a select group of decision makers; however, during the OPR development phase, it is typical for the BECxG to assist the owner in identifying a wide group of stakeholders as well as assist in soliciting input from them in development of the OPR. Rank-and-file building maintenance personnel are rarely in this small circle of decision makers, but they provide invaluable input about potential service life maintenance. Similarly, in the case of the two agencies’ structure, the BECxG must find strategies to facilitate discussion with the owner agency without going outside of their contractual bounds. The BECx plan is an excellent tool to exploit for this purpose. In addition to tailoring the BECx plan to overall budget and project-specific goals, a strong understanding of the contractual relationships at play can also inform the plan. The plan can then include strategies such as: workshop meetings with requirements for various stakeholder representatives; requirements for review and comment opportunities for the owner agency; and a process for recording, tracking, vetting, and resolving owner agency comments. While the owner agency may not direct the BECxG, their input can be gathered and evaluated through these methods and discussed with decision makers. In the ongoing project, the BECx plan included planned document reviews as well as workshops and meetings that included both the owner agency and USACE in order to collect and track both parties’ comments; all comments were then tracked in proprietary software and resolved by a select group of decision makers, which included select representatives from the owner agency. This cooperative process was facilitated by USACE, who contributed significantly to successful outcomes. The workshops proved invaluable for highlighting and resolving conflicting priorities. In particular, a USACE value engineering comment arose suggesting eliminating all window performance testing on the building, which was contrary to previous discussions with representatives from the owner agency. After serious discussion including the owner agency, USCE, and the BECxG, the group was able to balance both the owner agency’s desire for performance assurance as well as USACE’s concerns about cost and schedule implications of larger testing quantities by determining a lower quantity of window performance testing for inclusion in the OPR. The BECxG should consider the challenges to commissioning projects with relationships such as the USACE/owner agency relationship, where the decision makers will not occupy the building. The BECxG must also consider that the owner agency has comparatively little or no construction experience and is therefore illequipped to execute construction contracts on behalf of the federal government, particularly for large or complex projects (or both). Further, as the government is ultimately the owner, USACE may be in a more neutral position to prioritize requirements in a way that is aligned with the government’s priorities, which may not always align with the owner agency’s preferences. Therefore, a balanced approach with communication among all parties for issue resolution is essential. It should be noted that resolution does not mean “acceptance.” In any project, some project requirements will be prioritized over others. Planning for the owner agency to have a voice (through reviews, meetings, and representation in decision-making),
269
270
STP 1617 On Building Science and the Physics of Building Enclosure Performance
how it will exercise that voice (recorded and tracked comments), as well as how to resolve conflicts between that voice and USACE (selected decision makers with decisions tracked in comment responses) was critical for successful OPR and RFP development of the ongoing project.
Challenges and Opportunities with Design-Build This paper discusses projects where the enclosure design is complex (or anticipated to be complex) and design-build RFP documents are extensive and substantially developed to include building enclosure requirements such as: fenestration performance criteria, BECx testing requirements for the enclosure, laboratory testing requirements, enclosure mockup requirements, excluded materials, and requirements for allowable materials should they be selected (e.g., dual lines of sealant required at precast architectural panels, required minimum slopes for low slope roof assemblies, etc.). The RFP is put out to bid through the public bidding process. In order to win the design-build contract, teams first submit qualifications and are down-selected to a small group. This final group then responds to the design-build RFP by providing a design concept that is developed into early phases of design development as well as an overall project price. The winning team is selected and the project cost is established with the design-build contract award. Regardless of the design submitted by the selected team, ultimately, the requirements for the contract are the RFP; anything not included in the RFP is not owed on the project. The government must have clear criteria for selecting teams and must be able to maintain a high level of accountability regarding impartiality and fairness in the contracting process.8 In the exemplar projects, the project cost is set upon the awarding of the design-build RFP.
Design Build BECx—Potential Challenges During the ongoing project, stakeholders questioned early in RFP development how the Cx process would work for a design-build project and whether or not it was even a viable approach. One member of the owner agency team referenced the ASHRAE article by Turner, June, and Hwang9 to support the proposal that Cx as a whole, including BECx, be removed from the project due to its likely inefficiency. The article cites concerns and challenges with the overall Cx process in a designbuild project, which may be generally summarized where Turner, June, and Hwang state: “Since design/build delivery blurs the lines between design phase and construction phase, the Guideline 0-2005 Cx process can be more costly and less effective, directing Cx resources from the substantive technical work that is so essential to an effective Cx approach.”9 In addition to the concerns raised during this project, ASTM Task Group E06.55.09 has discussed multiple ballot comments that raise concerns that the standard does not fully address multiple project delivery types, including design-build. These concerns are addressed here.
DELGIUDICE, DOI: 10.1520/STP161720180065
Project Timeline, Influence Potential, and Cost Turner, June, and Hwang discuss cost and influence curves for design-build projects and conclude that while the design for a design-build project is not fixed until much later in the project schedule (as compared to design-bid-build), project budget generally is set very early in the schedule. Turner, June, and Hwang provide discussion and cost and influence curves that demonstrate the ability to change design versus the cost to make changes over the timeline of a project. Since the design is not fixed until arguably much later in the process for a design-build project than it would be for a design-bid-build project, it would seem that there should be an opportunity to have a greater influence on design later in the design-build process. However, as Turner, June, and Hwang point out, the cost basis for the design and its construction are generally fixed at the time of the project award. Turner, June, and Hwang conclude: “As a result, on many projects, the cost to change has already jumped, and the ability to influence has already dropped substantially, at the time of the design/build contract award.”9 If we accept Turner, June, and Hwang’s conclusions about design influence and cost relationships, then changes to requirements in order to meet OPR after the design-build contract award are likely to result in a change in contract and cost. If the project cost is set at the time of the design-build contract award, then the money must be in the job at the time of award. Therefore, the OPR must be developed and fully reflected in the design-build RFP prior to award of the design-build team contract. In a design-bid-build project, the design team participates substantially in OPR development with the owner and stakeholders, and OPR refinement may continue through design. Fundamental BECx in ASTM E2813 allows for the BECx to be brought on as late as commencement of design development, meaning that some amount of OPR decision-making has already occurred during schematic design, and some portion of OPR documentation is a retroactive recording of OPR development performed prior to onboarding the BECx. In contrast, if the OPR is to be effectively enforced through design and construction of a design-build project without threat of significant change order, it must be developed prior to completion of the contract with the design-build team. If the BECxP is to assure this portion of the process without changing the design-build contract and cost, it may be necessary to implement a more stringent timeline for involving the BECxP. For enhanced BECx, ASTM E2813 requires that BECx engagement occur: “during the Pre-Design Phase of the BECx process, but no later than commencement of the Schematic Design sub-phase.” If this is applied strictly by the letter, then the BECxP may be engaged after the design-build RFP is developed and placed out for bid or even after it is awarded and design begun. Although depending upon the project, there still may be great value in providing BECx services even at this later BECxP engagement, in order to enforce inclusion of the OPR in the project design and construction, the OPR must already be reflected in the RFP. Therefore, for maximum ability to effect the project, the BECx P should
271
272
STP 1617 On Building Science and the Physics of Building Enclosure Performance
be involved at the early stages of RFP development and the OPR should be developed and finalized prior to finalizing and issuing the RFP for bid. It should be noted that early involvement of the BECxP is a theme that is supported by the timeline required in ASTM E2813, for both fundamental and enhanced commissioning, as well as the timeline either required or recommended by many other BECx documents including LEED v47 and is supported in numerous articles and presentations on BECx and Cx in general, including those of Turner, June, and Hwang9 as well as Hopps and Burhoe.6 In cases where the owner elects to have a retroactive OPR developed after the RFP award, there may still be value in the BECx process. The BECxP can establish a retroactive OPR and, if needed, provide assistance to the team in developing changes to the design-build contract in order to reflect the OPR in the design-build contract requirements. In this case, the BECxP should advise the owner on the expected efficacy of that approach for the particular project as well as prepare the owner for the potential project cost change depending upon the extent to which the RFP already reflects (or does not reflect) the owner’s requirements. In the private sector, an owner may be able to work repetitively with the same team and, without creating stringent RFPs, be able to obtain pricing that reflects at least most of the desired OPR (e.g., based upon the team’s previous experience with the owner). Similarly, a sophisticated owner may be able to develop their own OPR or hire a bridging architect, or similar experienced professional, to create an OPR during RFP development. Provided the OPR or the design-build contract pricing already contains all of the owner’s requirements relative to the building enclosure, the BECxP joining the team after the award of the RFP may be a viable approach with limited cost or enclosure performance implications. ASTM E2813 fundamental BECx requires only that the BECxP review and document the preliminary OPR (ASTM E2813, Section 4.1.1.2). However, USACE, as with all public sector contracting, is subject to a high standard of contracting fairness and cannot rely on being able to consistently work with a familiar team, nor can they enforce an unclear intent in the RFP. A more disciplined approach to the RFP and OPR process is required on federal projects if the project is to deliver the OPR and maintain federal contracting standards. If something is not reflected in the contract, it is not owed; further, there are numerous complications, including the appearance of bribery, associated with doing “free work” for the federal government.
BECx for Design-Build—High Costs? Looking at Turner, June, and Hwang’s work on cost and influence curves, note how late involvement of BECx and late development of enclosure-related project requirements may be drivers for the perception that BECx drives up construction costs. Ideally, the OPR process is completed prior to pricing the project and must be prioritized within and constrained by the overall project budget. In the cases where the OPR has not been developed prior to awarding the design-build RFP and
DELGIUDICE, DOI: 10.1520/STP161720180065
enclosure performance requirements were not defined or were unclear, achieving the performance that the owner actually needs or wants may require additional cost. Similarly, a BECx budget for the actual BECx scope should be established at the beginning of the project; this budget constrains decisions such as the level of BECx selected, number of BECx site observation visits, number of performance tests (if any) beyond the minimum required in ASTM E2813, and should be developed with the owner as a function of the level of service desired and appropriate for the building and its use as well as the available budget. While there is no set guideline for BECx budgets, the General Services Administration’s website (www.gsa.gov) cites total building Cx costs between 1.25% and 2.25% of the total construction costs.10 Therefore, BECx would be some portion of that overall percentage—a very small portion of overall project budget. So, while bringing on a BECxP after the award of the design-build RFP (or even after preliminary cost estimates on documents that do not adequately define the required enclosure OPR) may begin a process that results in change orders or cost increase as a function of poorly defined scope, BECx is not directly raising the cost of the project. The cost increase is a function of the cost of providing systems and assemblies that meet the owner’s requirements but that were not previously budgeted. The BECx process itself is not significantly adding expenses to the overall project costs and is designed to provide assurance aimed at preventing costs associated with enclosure assembly failures, either during construction or building service life. The author is also familiar with the concern that the contractor will add some premium to cost estimates for a project that is being commissioned. Presumably, this concern is based upon the premise that the contractor believes that they will need extra funds to execute commissioning. There are several reasons the contractor may believe this. If the BECxG costs are a small percentage of the entire project, we may surmise that there is also some cost to the contractor in order to coordinate with the BECxG. The contractor may include some additional cost in order to provide personnel to walk with the BECxG on their site walks, accommodate/coordinate performance testing of the enclosure, and complete mockups specified as part of the BECx plan. It should be noted that while there may be some cost associated with these items, in many cases, the BECx process should also be able to identify issues early in construction before building-wide implementation thus limiting rework related to finding the issues later in the construction process or after the building is occupied. Large cost differentials to deliver a commissioned project should be interrogated. Providing BECx is not intended to create additional responsibility for the contractor to deliver a project that meets the project documents—that obligation is already fundamental to the contract with the owner. The BECx process should be aligned with the project goals, which should be reflected in the contract documents. Installation per the contract documents and assemblies capable of meeting specified performance requirements are required by the contract, whether or not a BECxG member completes additional site walks or the BECx plan includes performance
273
274
STP 1617 On Building Science and the Physics of Building Enclosure Performance
testing of the assemblies. For example, if window testing is part of the BECx plan, the plan should establish performance testing that is aligned with the specified performance of the windows and uses the industry standard test methodologies included in ASTM E2813 (e.g., ASTM E1105, Standard Test Method for Field Determination of Water Penetration of Installed Exterior Windows, Skylights, Doors, and Curtain Walls, by Uniform or Cyclic Static Air Pressure Difference11). Therefore, suggesting that projects with BECx cost more, for example, because fenestration tends to fail (costing time and money to repair it adequately to pass testing), suggests that were the project not commissioned and the windows never tested there would be installed windows that could not meet the specified performance criteria.
Design-Build—Challenges and Advantages Like all delivery methods, design-build delivery has inherent benefits and drawbacks. While the focus of this paper is to address and present potential solutions to challenges with implementing BECx on design-build projects, it is also worth exploring some of the advantages that may entice an owner to select this project delivery method as well as advantages that can be exploited in the BECx process when a design-build delivery is selected. Design-build projects are able to allow for design to occur for each element as it is needed in the project rather than require that design is finished prior to starting the project. This delivery method promises that the overlap of the design and construction phases for the project can result in shorter overall project schedules. Additionally, design-build team structure provides opportunities for communication between designer and builder throughout design, which is something that typically is not feasible in the same manner during design-bid-build and could be exploited for more cost-efficient construction solutions by facilitating combined construction and design expertise in problem-solving. While there are numerous subtle variations in exact contracting of a designbid-build project, and often a contractor is brought in for pricing and preconstruction work during the design process, the design-bid-build process fundamentally necessitates that the design is complete prior to competitively bidding the project. For federal work, it is likely that any contractor providing preconstruction assistance would be considered conflicted and unable to bid on the construction contract. In this case, the team responsible for executing the design is unable to provide input during the phases of design because they have not yet been selected. However, as Turner, June, and Hwang explain, the design phase is where the opportunity to make a change and the cost associated with that change are optimized.9 Designbuild fundamentally provides opportunities for contractor input during design; the contractor is not only available to work with the designer but, with the structure shown in figure 1, the design team is contracted beneath the contractor. Certainly, this does not fully immunize against poor communication, but the integration of contractor and designer can create the opportunity.
DELGIUDICE, DOI: 10.1520/STP161720180065
Similarly, this early onboarding of the contractor affords the BECxG the opportunity to work directly with both the designer and builder from relatively early design phases through the end of the project. Design review comments can be discussed with the entire team, ideally allowing for more synergy between the designer and builder expertise and the BECx process. This provides an opportunity to utilize contractor expertise for BECx items such as: logistics for performance testing and problem-solving schedule challenges to accommodate project mockups. This also provides an opportunity to develop builder buy-in at very early stages of the project, which tends to facilitate a smoother BECx process during construction. However, there are disadvantages to the design-build delivery method as well. A commonly cited concern is the removal of what is often viewed as a checks-and-balances system between the design team and contractor in the traditional design-bid-build model. Looking at the relationships in figures 1 and 2, design-bid-build projects, in contrast with design-build, have no contractual relationship between the design team and contractor; both parties work directly for the owner toward the same end of completing the project. This relationship provides contractual independence to both the design and construction teams to advocate separately and directly for the goals that they advance. While both parties work toward the end of a completed project, each team typically places greater and lesser emphasis on individual goals within the larger goal. While more nuanced, this generalization may apply: Construction teams are often believed to be driven primarily by schedule and cost, while the architect is viewed as charged with maintaining design intent, which may include aesthetics as well as performance characteristics of the design. For all projects, cooperation among the project team members is an asset to the entire project and the most likely path to meeting all of the project goals. However, the cooperation inherit in the contracting of design-build can also be viewed as eliminating the designer’s power to advocate for, among other things, required enclosure performance characteristics. For projects where this is true, an independent BECx process can fortify advocacy for enclosure performance.
BECx and Procurement and Project Delivery Core Competency ASTM E2813, like many documents that discuss BECx, organizes the BECx practice (summarized in Section 4 and outlined in Section 5 of the standard) roughly by project phase in what would be chronological order for a design-bid-build project. The standard takes care to indicate its intended use in all project delivery methods and includes a requirement for the BECxP to assemble the appropriate team members to provide sufficient competency in implementing the standard on any project delivery method, including design-build. This is illustrated in Section 4.2.1, BECxG Core Competencies, which outlines four main areas of competence: building and materials science
275
276
STP 1617 On Building Science and the Physics of Building Enclosure Performance
(Section 4.2.1.1); procurement and project delivery (Section 4.2.1.2); contract documents and construction administration (Section 4.2.1.3); and performance test standards and methodology (Section 4.2.1.4). Per ASTM E2813, Section 4.2: “This practice establishes that the BECxP shall assemble a team (BECxG) that, at a minimum, demonstrates a level of proficiency in the core competencies listed below that meets or exceeds the requirements of building codes, standards, guidelines, and regulations applicable to or otherwise voluntarily adopted by the Owner to govern enclosure-related design, construction, integration, and performance.” While there are challenges in executing BECx that are unique to the design-build process, the knowledge requirements listed under Section 4.2.1.2, Procurement and Project Delivery, clearly demonstrate the intention for the BECxG to be able to adapt the standard for use for a number of project delivery methods, including design-build: 4.2.1.2 Procurement and Project Delivery, including, at a minimum, demonstrated knowledge of the: (1) Influence of the project delivery method* selected by the Owner on the scope, adaptation, implementation, and cost of the BECx process as defined in this practice; (2) Influence of the number and type of contracts{ established between the Owner and the design and construction teams on the role and responsibilities of the BECxP and individual members of the BECx Group; (3) Influence of design and construction scheduling, phasing, and sequencing of the work on the scope, adaptation, implementation, and cost of the BECx process as defined in this practice; (4) Influence of the experience, qualifications, technical depth, and commitment of the design and construction teams to the BECx process on the role and responsibilities of the BECxP, the range and technical depth required of the BECx Group, and the anticipated scope and cost of the BECx process. This language clearly requires extensive knowledge about project delivery and influence of contracts as well as scheduling and sequencing. Since there are many variations, not just in contracting and project delivery type but in any number of project-specific requirements for any given building, the BECxGmust use the OPR and BECx plan to create a project-specific and holistic approach to executing BECx on each project. Both of these documents play a role in managing the challenges
*Including but not limited to: Design-Build; Design-Bid-Build; Design-Negotiate-Build; Construction Management; and Owner-Build as defined by CSI Project Resource Manual and Manual of Practice. { Including, but not limited to Single-Prime Contract and Multiple-Prime Contracts, with basis-of-payment provisions that may include; Stipulate/Lump Sum; Cost-Plus Fee; Fixed Fee, and Guaranteed Maximum Price, with penalties, bonuses, and incentives for early completion of the work and liquidated damages for any delays in substantial or final completion of the project.
DELGIUDICE, DOI: 10.1520/STP161720180065
associated with executing BECx for design-build projects. As project-specific standards of practice cannot be written for every variation of project, it remains the responsibility of the BECxP to assemble a sufficiently competent team to develop an appropriate plan based upon the OPR and project budget and to execute an effective BECx process, regardless of the project-specific challenges.
Practical Considerations—Logistics Logistics and organization are not explicitly discussed in ASTM E2813 BECxG core competencies, but these skills must be merged with the listed competencies in order to effectively execute BECx on any project. Design-build may present unique challenges for the BECxG to navigate, and efficient BECx processes that minimize potential for error will be an asset to the BECxG and project and will effect overall project success. First, the team must be able to successfully execute piecemeal design reviews. Design reviews must be logistically coordinated to fulfill BECx requirements, which include providing reviews when required to facilitate a fragmented design schedule as well as ultimately providing the equivalent of three full design reviews (if following ASTM E2813 enhanced BECx requirements). The team must also be skilled in considering potential project-wide implications of decisions made during these isolated reviews. Unlike in a design-bid-build project, no one on the team will review a 100% construction document set to aid in considering a design decision’s implications across the broader scope of the project; many decisions must be made prior to finishing the enclosure design. While it is not the BECxG’s job to supplant the designer in their role neither is forecasting the future a core competency listed in ASTM E2813. The team’s expertise in reviewing details as well as placing these details within the context of the overall project will be important in facilitating and verifying the OPR’s inclusion in the project documents. Similarly, various components of the enclosure may be at various stages of design and construction at one time. For example, the BECxG may still be providing design review services on the roof while curtain wall is in shop drawing phases and below-grade waterproofing installation has begun. The team must be able to track design items, submittal items, and construction issues as well as track preconstruction and performance testing all at the same time. Ideally, the BECxG will work with the design-build team and the owner to establish procedures and documentation methods such as tables, site visit reports, web-based tracking software, and construction trailer bulletin boards that meet the project needs and can be utilized by all parties participating in BECx. Clear communication and organized tracking are always required for successful BECx; however, design-build brings with it a large number and variety of tasks as different project phases run in parallel at the same time. Again, the BECx plan must address this organization and communication in a manner that meets the project-specific needs and goals; plan revision may be required as the project
277
278
STP 1617 On Building Science and the Physics of Building Enclosure Performance
progresses in order to effectively address all of the required tracking and resolution processes.
Conclusions Early involvement, competence, and project specific are watchwords for all BECx projects. These concepts also summarize what is required to meet the challenges of providing BECx for a project with design-build or complex stakeholder or contracting arrangement (or any combination thereof). The BECxG must be sufficiently competent to make and execute a BECx plan for the specific project. Further, while there are other project team members who may begin the OPR development process, the BECxG is most effective when given the opportunity to influence the project early enough to allow for inclusion of a meaningful OPR and BECx plan in the RFP (for design-build). Involvement later in the project may add cost to the project or limit the BECx process. Early involvement for maximum influence is a BECx best practice, which is supported by the conclusions of Turner, June, and Hwang7 among many other articles, papers, and the language of BECx guidelines and standards. For the exemplar projects referenced, a strict RFP process was necessary for USACE contracting, necessitating early involvement of the BECxP, in order to confirm inclusion of enclosure goals in the OPR and RFP. When utilized by a competent and experienced BECxG, ASTM E2813 is readily adaptable to design-build and complex government stakeholder relationships through astute use of the OPR and BECx plan. In fact, a competent BECxG should be able to exploit some of the advantages of design-build delivery to enhance the BECx process. Similarly, BECx may be able to enhance the design-build process. If there are concerns that the designer will be unable to successfully advocate for enclosure performance goals because of the contracting arrangement, the BECx process is well-suited to help fill that gap. Even when the BECxP is not brought on early enough in the process to influence the design-build RFP, the BECxG can advise the owner on the extent of their ability to influence the process moving forward and evaluate, on a project-specific basis, the viability of applying the BECx process without contract modifications. Depending upon the project, there may be great benefit in bringing on a BECxG, even later in the design-build contracting process. Ultimately, standards cannot address granular details for each specific project. This is true whether the project challenges are related to climate zone, unusual building use, innovative design, project delivery method, or complex stakeholder relationships. Awareness of relationships, contracting, and delivery method must inform the BECxG’s approach, whether they are involved from the very first discussions of the project or after design has begun. For successful BECx, the BECxP and their team must be sufficiently competent to execute the art and science of building enclosure commissioning, to mitigate the challenges, and to exploit the advantages afforded by all the particulars for the project at hand.
DELGIUDICE, DOI: 10.1520/STP161720180065
References 1.
2.
3.
4. 5.
6.
7.
8.
9. 10.
11.
Standard Practice for Building Enclosure Commissioning, ASTM E2813-18 (West Conshohocken, PA: ASTM International, approved October 1, 2018), http://doi.org/ 10.1520/E2813-18 Standard Guide for Building Enclosure Commissioning, ASTM E2947-16a (West Conshohocken, PA: ASTM International, approved June 1, 2016), http://doi.org/10.1520/ E2947-16A U.S. Army Corps of Engineers, “Engineering and Design: Total Building Commissioning Procedures,” U.S. Army Corps of Engineers Official Publications, March 31, 2017, http:// web.archive.org/20191205003437/https://www.publications.usace.army.mil/Portals/76 /Publications/EngineerRegulations/ER_1110-345-723.pdf Department of Defense, “Whole Building Design Guide,” 2018, https://web.archive.org /web/20181003011914/http://www.wbdg.org/FFC/FEDMIL/std3007f.pdf U.S. Army Corps of Engineers, “Whole Building Design Guide,” National Institute of Building Science, 2018, https://web.archive.org/20191205003752/https://www.wbdg.org /FFC/DOD/UFGS/UFGS 01 91 00.15.pdf E. R. Hopps and A. M. Burhoe, “Balancing Your Building Enclosure: The Building Enclosure Commissioning Scope that Matters” (paper presentation, BEST 5 Conference: Building Enclosure Science & Technology, Philadelphia, PA, April 15–28, 2018). U.S. Green Building Council, “LEED BDþC: New Construction v4—LEED v4: Enhanced Commissioning,” U.S. Green Building Council, 2018, http://web.archive.org/ 20191205004517/https://www.usgbc.org/credits/new-construction-core-and-shell-schoolsnew-construction-retail-new-construction-healthca-17 Department Ethics Office, “Government-Wide Ethics Laws,” U.S. Department of the Interior, http://web.archive.org/20191205004734/https://www.doi.gov/ethics/governmentwide-ethics-laws S. C. Turner, M. H. June, and S. H. Hwang, “Commissioning Design/Build Projects,” ASHRAE Journal 54, no. 10 (2012): 54–60. U.S. General Service Administration, “Commissioning Agent Costs,” August 13, 2017, http:// web.archive.org/20191205005127/https://www.gsa.gov/real-estate/design-construction /commissioning/commissioning-program/building-commissioning-process/planningstage/establish-initial-budget/commissioning-agent-costs Standard Test Method for Field Determination of Water Penetration of Installed Exterior Windows, Skylights, Doors, and Curtain Walls, by Uniform or Cyclic Static Air Pressure Difference, ASTM E1105-15 (West Conshohocken, PA: ASTM International, approved August 1, 2015), http://doi.org/10.1520/E1105-15
279
280
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180068
Patrick G. Giblin,1 Elizabeth O. Cassin,1 Larry R. Meyers,1 and Chelsea F. Ames2
BECx: A Case Study for Lessons Learned Citation P. G. Giblin, E. O. Cassin, L. R. Meyers, and C. F. Ames, “BECx: A Case Study for Lessons Learned,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 280–316. http:// doi.org/10.1520/STP1617201800683
ABSTRACT
Building Enclosure Commissioning (BECx) is a quality-oriented process intended to verify that a building enclosure’s design and construction meet the Owner’s Project Requirements (OPR). The BECx process also helps to enhance the performance of building enclosure components and assemblies through technical assistance provided by the BECx Provider (BECxP) to the project team. However, without proper execution or integration into the project, BECx can be an underutilized tool. For instance, the BECxP may be introduced to the project after the design and project budget have been established, thereby narrowing the opportunities to provide technical recommendations to the project team on design alternatives that more efficiently meet the OPR. Additionally, the full project team must be committed to and active in the process to realize its full benefit. In some cases, the BECx process is viewed as a box to check and the BECxP is not wholly integrated into the project team. Finally, recommendations made by the BECxP may be construed as undermining decisions made by other project stakeholders. This paper will discuss the challenges to full integration of BECx into a project and, through a recently completed case study example, will illustrate successes, failures, and lessons learned regarding the BECx process. Key milestones in the BECx Manuscript received October 8, 2018; accepted for publication June 11, 2019. 1 Wiss, Janney, Elstner Associates, Inc., 10 South LaSalle St., Suite 2600, Chicago, IL 60603, USA, P. G. G. https://orcid.org/0000-0003-3300-3016 E. O. C. https://orcid.org/0000-0003-2563-9807 L. R. M. https://orcid.org/0000-0002-1491-9662 2 Wiss, Janney, Elstner Associates, Inc., 605 North Highway 169, Suite 1000, Minneapolis, MN 55441, USA, https://orcid.org/0000-0001-8441-7222 3 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
GIBLIN ET AL., DOI: 10.1520/STP161720180068
process will be discussed to illustrate where the project fell short of expectations and where potential issues were successfully addressed through recommendations by the BECxP and responsive actions by the project team. We will identify how the BECx process affected design and construction details such as window-to-wall interfaces, air and water control layer detailing, and roofing, some of which were successful and some of which fell short of expectations. The BECx team will share lessons learned from this case study that are now being utilized to improve the process for its projects moving forward. Keywords Building Enclosure Commissioning, BECx, roofing, precast concrete, building envelope, case study, storefront, curtain wall, testing
Introduction Complex architectural designs, an endless variety of ever-changing proprietary enclosure systems, and increasing demands for energy performance are among many factors that have contributed to more complicated and technically challenging building enclosures. Increased air tightness and insulation results in less drying across the building enclosure (which, if improperly managed, may lead to wetter materials and added potential for degradation); lighter walls hold less water and show leaks more readily; and in general, many current materials are less moisture tolerant than those used historically. As a result, the margin for error in building enclosure detailing and construction has decreased and the risk of problem development or failure has increased. The Building Enclosure Commissioning (BECx) process is one way to help reduce the risk of failure by validating throughout design and construction that the building enclosure meets the owner’s project requirements, whether they be air tightness, water tightness, thermal efficiency, durability, low maintenance, or similar criteria. The intent of the process is to establish, promote, and verify the owner’s desired performance objectives, as recorded in the Owner’s Project Requirements (OPR) document, by observing, documenting, and making recommendations related to the design, construction, testing, and operation of the building enclosure from predesign through postoccupancy. This paper presents a general background of the BECx process and then utilizes a case study example from a recent project to illustrate benefits and limitations of BECx in the context of the project specific challenges, with particular focus on challenges stemming from lack of BECx integration into the overall project.
The BECx Process BECx STANDARDS AND GUIDES
There are several reference documents that provide guidance on the BECx process. This paper references the following two companion documents: ASTM E2813-18, Standard Practice for Building Enclosure Commissioning,1 and ASTM E2947-16a,
281
282
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Standard Guide for Building Enclosure Commissioning.2 ASTM E2947-16a provides recommendations for the process from predesign through the occupancy and operation phases. ASTM E2813-18 is intended to serve as an authoritative practice for BECx that is based on an OPR document, clearly defined and enforceable levels of BECx (fundamental and enhanced), and minimum core competencies required of the BECx Provider. Definitions for key players in the process are established in ASTM E2947-16a. The BECx team includes the BECx Provider (BECxP), who is the authorized person or firm retained by the owner to develop and manage the BECx process; the BECx Specialist (BECxS), who has the applicable technical knowledge of building enclosure performance; and the BECx Technologist (BECxT), who performs the testing. For simplicity purposes, BECxP will be used to represent each of these roles throughout the paper. BECx ROLES AND RESPONSIBILITIES
Many players, each with their own unique responsibilities, serve different roles to successfully construct a new building or structure. The project team structure varies depending on project delivery method, including traditional Design-Bid-Build, Construction Manager at Risk, Integrated Project Delivery, or Design-Build, such as the case study discussed in this paper. While the project delivery method affects the contractual responsibilities of the owner, architect, and contractor, in all project delivery methods, the BECxP is retained by the owner and maintains third-party independence from the design and construction teams throughout the project to best serve the owner and to avoid conflicts of interest. ASTM E2947-16a notes, “The BECxP… should have no contractual relationship to any firm providing design or construction related services to the project and have no known or potential conflicts of interest.”2 Additionally, both ASTM E2947-16a and ASTM E2813-18 make clear that the BECx process is not intended to supersede or replace the contractual obligations or alter the roles and responsibilities of the design and construction team. As such, the BECxP does not have authority over any of the project team members. The definitive authority on the project is the owner or the owner’s authorized representative. The owner relies on the BECxP to provide the project team with technical guidance and verification that the building enclosure design and construction meets the OPR. Similarly, the BECx process is not intended to replace the design team and the contractor’s quality assurance and quality control (QA/QC) processes. Each member of the project team is responsible for their own QA/QC and has an obligation to deliver their respective services at the level of quality agreed upon. While each team member has their own role and responsibilities, it is imperative that the BECx process be collaborative to be successful. The most likely path to delivering a building that meets the OPR is full collaboration of the entire project team and full integration of the BECx process within the project as a whole. However, collaboration is sometimes stifled for various reasons, including timing of the BECxP involvement, resistance to alternate design options or materials, perceived
GIBLIN ET AL., DOI: 10.1520/STP161720180068
infringement on the roles and responsibilities of another team member, lack of trust in the BECxP by the designer or the contractor, impact on the project schedule, and potential impacts on project costs. If quality is a primary goal of all team members involved, and if team members recognize the goal of BECx is to improve quality, then the value of BECx to the project can be substantial. The BECx process is most effective when the project team uses the BECxP as a resource for quality and is committed to the process, instead of seeing the BECx process as a hindrance to the schedule or as a mandatory fulfillment of a requirement.
Case Study PROJECT BACKGROUND
The subject project is a dormitory building that is eight stories tall and totals approximately 14,307 m2 (154,000 GSF). The project is located in Climate Zone 5, which is a cold climate that generally is defined as having between approximately 5,400 and 12,600 heating degree days.3 The building structure consists of a cast-inplace concrete frame with precast hollow core concrete plank floor slabs. The commissioned components of the exterior cladding include load-bearing insulated precast concrete sandwich panels (3 in. [7.62 cm] insulation core with 4 in. [10.16 cm] of concrete on each side) with punched single-hung windows, aluminum and glass curtain walls, and storefront windows. The roofing system consists of polyisocyanurate insulation (over top of the concrete plank), cover board, and fully adhered, fleece-backed thermoplastic polyolefin (TPO) membrane. Other enclosure systems that were commissioned include below-grade waterproofing, sheet metal flashing and coping, and joint sealants. The project was delivered under the Design-Build bridging format, whereby the owner retained a bridging architect to provide pricing documents and a designbuilder to complete the design and construct the building based on the pricing documents. The BECxP was retained directly by the owner. LATE INVOLVEMENT
ASTM E2813-18 and ASTM E2947-16a both state that the owner should engage the BECxP early in the process to be most effective. In general, when retained early in the process, the BECxP, in collaboration with the other team members, is better positioned to advise on appropriate performance requirements and to offer pertinent suggestions relating to the OPR and the resulting design that affect material selection and detail concepts. Waiting to involve the BECxP after OPR development and initial design can result in missing or incomplete OPRs, predesign work that may not meet the goals of the owner, or missed opportunities to enhance the enclosure systems (or any combination thereof). Recommendations made later in the process that are not part of the predesign discussions can be met with resistance from project team members due to adverse effects on the budget and construction schedule. For the case study project, the owner retained the BECxP late in the design phase. The bridging architect had already established the basis of design (BOD)
283
284
STP 1617 On Building Science and the Physics of Building Enclosure Performance
through bridging documents and the design-builder had already provided a guaranteed maximum price (GMP) based on the bridging documents. The later involvement limited the BECxP’s ability to influence design decisions that had already been made, which, in several cases, could have helped to avoid several challenges faced during construction (some of which will be discussed under the “Select Identified Concerns” section of this paper). ISSUE RESOLUTIONS
Since the bridging architect had limited involvement in the project delivery following initial design, dispute resolutions were left to the design-builder, the BECxP, and the owner, without input from a third-party designer to weigh in and enforce their design as there would be in a traditional design-bid-build approach. In general, the owner is ultimately responsible for accepting unresolved items, which may require revision of the OPR or enforcing resolution of the items. Due to the potentially contentious nature of dispute resolution, it is important for the BECx Plan or Specification (or both) to outline the process for resolving open issues and that this process be discussed and agreed upon at both the design and construction phase BECx kickoff meetings. For the case study example, when the BECxP and design-builder disputed a resolution or when the design-builder did not respond to an action item, the owner frequently did not use their authority to direct the design-builder to respond or address nonconforming items. Since the BECxP’s role was not to direct the design-builder, consistent with ASTM E2947-16a, issues oftentimes remained unaddressed or inadequately addressed. Several examples of how this misunderstanding of roles and responsibilities resulted in a more challenging process and likely in a lower performing enclosure (compared to the potential enclosure performance if action items were promptly addressed) are discussed in the “Select Identified Concerns” section of this paper. BECx SCOPE OF WORK
Even when referencing the available industry standards, the BECx scope will vary among projects depending on a number of factors including project delivery method, sensitivity of the building to enclosure issues or failure, risk aversion of the owner, level of complexity and uniqueness of the enclosure, proficiency of the design and construction teams, level of comfort with enclosure materials or systems, project schedule, and budget. These factors inform the final OPR for the project as well as the projectspecific BECx plan. The following is a discussion of the BECx process that was implemented for the case study based on the documents referenced earlier and includes the experience of the authors from the perspective of the BECxP. The project generally employed fundamental BECx with some modifications to the process as outlined in ASTM E2813-18, which stemmed from late involvement of the BECxP. Predesign Phase
The OPR is the seminal document that defines the owner’s performance expectations for the enclosure among other requirements. It is used to evaluate the design and construction of the building enclosure throughout the BECx process, and as such, the BECxP should develop or review this document in the predesign phase.
GIBLIN ET AL., DOI: 10.1520/STP161720180068
The OPR was developed without the input of the BECxP as is often the case for institutional clients with campus standards that establish the OPR by precedent. The BECxP was not yet retained during the predesign phase, so the review of the OPR was performed concurrent with the design phase. Design Phase
The BECxP reviewed the OPR and the Request for Proposal (RFP) issuances for the Design-Build solicitation to become familiar with the requirements for the enclosure. Since the OPR was an established standard for the client and the designbuilder was already engaged based on its RFP response, updates to the OPR were not part of the BECxP’s scope. The OPR included general preferences for types of systems but did not provide performance metrics or service life expectance. The owner indicated to the BECxP that they generally desired an airtight, watertight, energy-efficient, and durable enclosure. During the design phase, the BECxP performed a review of the 100% review set provided by the design-builder. The BECxP provided review comments as notations on the drawings and specifications and tracked comments with an issues and resolutions log. Review comments typically addressed items that were nonconforming with the OPR as well as provided recommendations to enhance performance and constructability in alignment with the stated goals in the OPR. The BECxP comments and recommendations were provided to the design team and owner for their review and consideration. The BECxP also provided a BECx specification and conducted a design phase kickoff meeting, during which the BECxP established that the BECx process does not replace the QA/QC process of the Design-Build team, that the BECxP’s role is to identify nonconforming items or items of concern (or both) for resolution by the Design-Build team and the owner, that the design review comments were for the design-builder’s consideration, and that the owner would be responsible for accepting or resolving unresolved issues. Despite this, many issues identified by the BECxP were left unresolved, which will be further discussed in the “Select Identified Concerns” section of this paper. One efficient way to help in resolving issues is for representatives from the design team and owner to attend regular meetings with the BECxP for discussion of issues identified during the design reviews and other questions or issues that arise from changes to the design. The BECxP may identify advantages and disadvantages of certain design decisions so that the project team can make informed decisions. These project meetings were not included in the scope of the case study example. Preconstruction Phase
The BECxP and project team conducted a preconstruction phase kickoff meeting to discuss upcoming BECx related activities, the BECx process, and roles and responsibilities, including issue resolution. The BECxP also participated in preinstallation meetings for the enclosure systems to discuss material specific requirements, testing, detailing, open issues, installation schedule, and similar items.
285
286
STP 1617 On Building Science and the Physics of Building Enclosure Performance
The BECxP attended laboratory testing of one window randomly selected by the owner to verify window air and watertightness performance prior to building-wide installation. The BECxP made the recommendation to test a window prior to shipment to the site based on lessons learned from a similar project. The intent was to discover any issues with fabrication of the window for resolution by the manufacturer prior to delivery of the remaining 392 windows. During the testing, fabrication issues related to weather-stripping discontinuities and inadequate compression of the weather-stripping by the locking mechanism were discovered and were able to be addressed prior to installation of the windows on the project. The BECxP also reviewed enclosure submittals to compare against the OPR, to assist in coordination of trades, to review constructability, and to help resolve material or system issues. Due to the fast-track nature of this project, several of the submittals were reviewed prior to the BECxP’s back-check review of the construction documents. The BECxP provided comments as notations on the submittals. The design-builder did not respond to all comments, and many issues were left unresolved, which will be further discussed in the “Select Identified Concerns” section of this paper. Construction Phase
In accordance with ASTM E2947, the BECxP performed periodic construction observation shop and site visits at certain milestones and at regular intervals. Each visit was documented with a site visit report indicating work in progress and significant observations such as nonconforming work or concerns about workmanship. All issues were also tracked in an issues and resolutions log. Timely identification, documentation, and resolution of issues are critical to the success of the process. Representatives from the design team, owner, and contractor should attend site visits with the BECxP or participate in team meetings immediately following each site visit (or both) to discuss and resolve open and new issues; however, neither happened with any regularity in the case study example. Instead, following each visit, prior to leaving the site, the BECxP communicated issues and items of concern to the Design-Build representative on site. This more limited communication had a negative impact on the BECx process, as will be further outlined in the “Select Identified Concerns” section of this paper. Performance testing of the enclosure components or systems began during the construction phase as soon as key components were installed to evaluate in situ performance for comparison with the OPR as reflected in the project specifications. The BECxP coordinated, observed, and documented specified air leakage testing (ASTM E783-02[2018], Standard Test Method for Field Measurement of Air Leakage Through Installed Exterior Windows and Doors4) and water leakage testing (ASTM E1105-15, Standard Test Method for Field Determination of Water Penetration of Installed Exterior Windows, Skylights, Doors, and Curtain Walls, by Uniform or Cyclic Static Air Pressure Difference5) of windows, storefronts, and curtain walls.
GIBLIN ET AL., DOI: 10.1520/STP161720180068
The BECxP also helped to identify and resolve issues that were discovered during testing. While it is beneficial for representatives of the owner, design team, and contractor to be present to witness testing with the BECxP, this did not occur for the case study project. Testing included 12 out of 392 single-hung windows (approximately 3%), eight locations for three different types of storefront, and five locations for curtain walls at various phases of installation completion. The single-hung windows had a 100% success rate for testing, which may have been partially attributable to the lessons learned during the laboratory testing. However, five out of eight storefronts failed their initial static water test (ASTM E1105), and one of those storefronts failed five times before finally passing. All five curtain wall areas failed their initial static water test. Failures were a result of poor workmanship, including missing or incomplete sealant, as well as incomplete tie-ins between the fenestration and the opaque wall, as will be further described in the “Select Identified Concerns” section of this paper. The BECxP also performed specified electrical impedance testing (ASTM D7954/D7954M-15a Standard Practice for Moisture Surveying of Roofing and Waterproofing Systems Using Non-Destructive Electrical Impedance Scanners6) and electronic integrity testing (ASTM D7877-14 Standard Guide for Electronic Methods for Detecting and Locating Leaks in Waterproof Membranes7) of the roofing system upon completion. The BECxP identified moisture within the roofing system, which ultimately required 100% replacement, as will be further discussed in the “Select Identified Concerns” section of this paper. Occupancy and Operations Phase
After construction and testing, the BECxP provided the owner with a BECx report that included all documentation of the process as well as maintenance and warranty information for the enclosure. Prior to expiration of warranties, the BECxP performed a site visit and met with the building staff that were familiar with the performance of the enclosure to discuss and identify outstanding or newly discovered issues with the enclosure. Following this review, the BECxP provided a report containing observations and recommendations for maintenance, repair, or replacement to the owner. The only significant issue observed was evidence of water infiltration at the storefront and curtain wall assemblies. It was not known if water stains were from leaks or from original construction staining, but the BECxP recommended further investigation. SELECT IDENTIFIED CONCERNS
The following examples include action items that were identified by the BECxP during the design or construction phases (or both). The examples are intended to illustrate how the BECx process was successful in addressing several potential issues, but how it was also difficult to implement or unsuccessful in fully addressing concerns.
287
288
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Precast Cladding Joint Sealant
During the initial design review, the BECxP recommended dual-stage sealant for joints between precast panels (fig. 1). Dual-stage seals provide redundancy to the design rather than relying on a single line of defense at the sealant joints. This meets the goals of the owner as stated in the OPR for an air- and watertight enclosure that will require less frequent maintenance. The exterior seal of the dual-stage seal typically is located close to the exterior face of the panel and is the weather seal that prevents most direct rain entry. The interior seal of the dual-stage seal typically is located a few inches back from the exterior seal or, for an insulated sandwich panel, near the outside face of the inner wythe. The interior primary seal is the air and vapor seal and is protected from ultraviolet light, thermal stress (when placed inboard of the insulation in an insulated precast sandwich panel), and direct wetting, which leads to a longer service life. The interior seal must be continuous and integrated with the surrounding air control layer (fenestration, roofing, waterproofing, air barrier, etc.). The air gap between the interior and exterior seals is wept to the exterior, typically at the intersection of horizontal and vertical joints, at the interface with the roofing system, and at the base of the wall. Weeps at every floor line help to compartmentalize the air gap, which minimizes the pressure differential across the outer seal and minimizes the risk for subsequent water penetration. The dual-stage sealant joints should be detailed to ensure that the air and water seals are continuous and integrated with the air and water control layers of the surrounding construction.8,9 Details for precast sealant joints were not provided in the design drawings and were not specified. During the design-phase drawing review, the BECxP recommended
FIG. 1 Double seal at precast joints.
GIBLIN ET AL., DOI: 10.1520/STP161720180068
that details include dashed lines in the plan and a section to show where the inner and outer sealant joints of the dual-stage seals would be placed and how they would integrate with window/door perimeter seals, the roofing membrane, and the waterproofing at the base of the wall, including weeps to drain the air space between the two joints at each floor line and at the roofing and waterproofing interfaces. The BECxP also recommended that the sealant be installed from the exterior to avoid access issues at floor slabs, steel embedments, and columns because these planes could prevent installation of a continuous seal and therefore compromise the OPR requirement for “airtight” and “watertight.” The BECxP also recommended that the joints between panels be approximately 1 in. (2.54 cm) wide to allow for construction tolerances and sealant installation from the exterior, which would help to avoid the potential for “tight” joints during construction where the dual seal would not be able to be reliably installed. The BECxP reviewed the second (and final) round of drawings concurrently with submittals and as the enclosure was being installed due to the fast-track nature of the design-build project. The second set of reviewed drawings included a few details of precast joints that indicated an inner polyurethane sealant joint at the inner face of the precast installed from the interior and an outer silicone sealant joint at the outer face of the precast installed from the exterior (fig. 2). The insulation in the sandwich panel was shown to extend the full width of the panel, and no weeps were shown. Additionally, the drawings clarified that the precast panels were load bearing with grout to fill horizontal joints the full length of panels at the lower three floors and with shims and partial grout fill at the upper floors. The BECxP reiterated the initial design review recommendations to provide additional sealant details and install the seals from the exterior, as well as the following additional recommendations: • Both the interior and exterior seals of the dual-stage seals should be silicone because silicone has a longer anticipated service life than polyurethane, which addresses the OPR requirement for durability. Since polyurethane does not bond to silicone, each weep location where these seals would meet would pose a compatibility or sequencing issue (or both). • Provide more information about weep detailing, including where weeps were to be located and if the sealant would be compatible with the precast insulation as it “sweeps” across the air gap. • If the interior sealant joint was to be installed from the interior, provide more information to indicate how the sealant would be continuous at floor lines and columns, where access was problematic. • Provide more information about dual-stage seals at fully grouted joints and partially grouted and shimmed joints. During preinstallation meetings and initial installations, the BECxP continued to request more information about the dual-stage seals, including how they would be detailed and the anticipated trade sequencing to allow for continuity of the joints. The design-builder decided against the BECxP’s recommendation to use silicone for both interior and exterior sealant as well as to install sealant from the exterior. The design-builder chose to install the inner sealant joint from the interior due
289
290
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 2 Dual-stage sealant joints between precast panels shown in a few plan details.
GIBLIN ET AL., DOI: 10.1520/STP161720180068
to schedule constraints because exterior access was not available at the time they were getting the building dried in, concerns about the placement of grout or shims (or both) at horizontal joints preventing exterior access, and concerns about the 11-in. (27.94 cm) depth of the panels and the narrow joints between panels not allowing reliable installation from the exterior. Instead of a continuous interior joint at floor lines, steel embedments, and columns, which was not possible due to access, the design-builder proposed installation of closed-cell spray polyurethane foam (SPF) insulation from the exterior at floor lines and columns in place of the sealant (fig. 3 and fig. 4). The BECxP requested verification of adhesive and chemical compatibility between the SPF and the silicone and polyurethane sealants and noted concerns about the ability to make the interior joint continuous. The design-builder also indicated that weeps would only be installed at grade, in lieu of at every floor line as recommended by the BECxP, in addition to locations where the vertical sealant joint would be interrupted (e.g., at fenestration and above roofing). Fewer weeps result in a greater pressure differential across the exterior sealant joint, which could lead to increased water penetration across deficiencies in the exterior joint and could also result in more water being managed by single weeps. Despite recommendations for weeps at each floor line, the design-builder only installed weeps at the base of the wall and detailed them with SPF to fill the vertical joint across the air gap. To verify the continuity of the inner sealant joint and the base of wall weeps, the BECxP performed visual observations and water testing at select locations, which included introducing a small amount of water behind the exterior weather seal (similar to the ASTM C1715/C1715M-15 Standard Test Method for Evaluation of Water Leakage Performance of Masonry Wall Drainage Systems10) and verifying
FIG. 3 SPF installed in the vertical joint at a floor line.
291
292
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 4 Interior SPF observed at the exterior face of a wall.
that it did not bypass the inner seal. The BECxP did not observe water infiltration but identified the following concerns during installation of the sealant and SPF: • Steel embedments were grouted solid and the interior sealant joint was discontinuous across the grout (fig. 5). The design-builder proposed routing back the grout to install bond breaker and sealant. Precast joints were as small as 1/16 in. (1.5875 mm) at some locations, preventing the installation
FIG. 5 Discontinuous interior precast panel sealant at grout patches.
GIBLIN ET AL., DOI: 10.1520/STP161720180068
•
of SPF from the exterior at floor lines. The BECxP provided recommendations for the design-builder’s consideration, and ultimately, they decided to grind the joints to widen them and allow the SPF to be installed. Interior sealant joints were not installed at several locations prior to installation of topping slab (fig. 6), metal stud framing (fig. 7), or sheathing. The BECxP noted concern with continuity of interior seals and risk for water infiltration or air leakage (or both) at the discontinuities. Ultimately,
FIG. 6 Topping slab flowed into an unsealed interior precast joint.
FIG. 7 Missing sealant at horizontal precast joint behind metal framing.
293
294
STP 1617 On Building Science and the Physics of Building Enclosure Performance
•
•
•
the design-builder reportedly removed interior sheathing and framing to address missed seals, but the BECx team did not observe these repairs and the design-builder never clarified how concealed locations were reviewed. Precast parapets were initially installed without dual-stage sealant joints to weep the joints above the level of the roofing membrane. Sealant was removed and replaced to include dual stage seals. SPF at the base of wall weeps was tooled from the exterior before curing (similar to sealant) in an effort to provide a positive slope to drain. The BECxP was concerned that tooling of spray foam might alter the closed cell structure of the foam and that the convex profile of the SPF (due to continued expansion following tooling) would direct water to the bond line between the precast and SPF, where the BECxP observed locations of adhesive failure between the SPF and precast (fig. 8) and where the BECxP was concerned about long-term adhesive bond. Because the base of the wall was the only weep location for the entire eight stories of the building, maintaining drainage at these weeps was critical for performance. The design-builder added silicone sealant over top of the SPF, but the sealant was incomplete with holes and gaps at several locations (fig. 9). Sealant repairs were reportedly made by the design-builder. Weeps were not installed at several locations at the interface with the roofing membrane and at window/door heads (fig. 10). Sealant repairs were made by the design-builder. Figure 10. Dual-stage weep missing above storefront assembly
FIG. 8 Dual-stage weep with SPF that has a convex profile and separations at the bondline with the precast.
GIBLIN ET AL., DOI: 10.1520/STP161720180068
FIG. 9 Sealant over spray foam not continuously tooled and not married to the precast sealant below.
FIG. 10 Dual-stage weep missing above storefront assembly.
Fenestration Perimeter Detailing
The continuity of the air/water line of fenestration and the opaque wall air control layer interface is a critical detail that is oftentimes detailed incorrectly or overlooked. Two common issues with this interface include (1) misaligning the primary air/water line of the fenestration with the air barrier of the opaque wall and (2) not
295
296
STP 1617 On Building Science and the Physics of Building Enclosure Performance
understanding where the primary air/water line of the fenestration is and mislocating the perimeter sealant joint. The continuity of the air barrier is critical at this interface to achieve air- and watertightness. In this case, that continuity was not achieved at several locations, and despite comments made during initial and followup design reviews, during submittal reviews, and during preinstallation meetings, these issues were not addressed until construction of the enclosure commenced. Due to the fast-track nature of the project and continued unresolved BECx comments regarding discontinuities throughout the process, field retrofits had to be installed by the design-builder, which were more challenging, more costly, and likely lower performing than options that could have been implemented had the issues been addressed during design. During the design phase, the BECxP identified that there was minimal overlap between the primary air/water line of the windows and the inner wythe of the precast (fig. 11 and fig. 12) or that the primary air/water line of the windows and curtain
FIG. 11 Minimal overlap between primary perimeter sealant and precast inner wythe.
GIBLIN ET AL., DOI: 10.1520/STP161720180068
FIG. 12 Minimal overlap between primary perimeter sealant and precast inner wythe.
wall was not aligned with the inner wythe of the precast (figs. 13–15). As such, the air/ water control layer was discontinuous at fenestration-to-opaque-wall interfaces. As a field retrofit, self-adhered membrane or sheet metal bedded in sealant was installed against the precast to bridge between the air/water line of the fenestration and the air/ water control layer of the precast panel (figs. 13–15). Had this been addressed during design, the fenestration could have been moved inward to seal directly to the precast to
297
298
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 13 Revised shop drawing detail showing retrofit transition membrane bridging between window and precast air/water seal.
avoid the need for a supplemental material at this condition and the need to rely on a blind bed seal at the metal for water and air continuity. This would have been a more straightforward and likely a less costly and more durable option. Issues related to this field retrofit that were identified by the BECxP included the following: • Self-adhered membrane or sheet metal at rough openings was missing at several locations where the fenestration had been installed (fig. 16). Fenestration was removed and reinstalled as needed to facilitate installation of membrane/sheet metal. Sheet metal at the rough opening was bedded in sealant along the interior edge only and not along the outer edge (fig. 17), which resulted in water infiltration during testing at the sills of the curtain wall where the sheet metal terminated. As a retrofit, the outer edge of the
GIBLIN ET AL., DOI: 10.1520/STP161720180068
FIG. 14 Revised shop drawing detail showing retrofit transition membrane bridging between window and precast air/water seal.
sheet metal was also bedded in sealant. Lap joints in the sheet metal or selfadhered membrane were reverse shingled or had open gaps that allowed water infiltration during testing (or both). During the design phase and throughout the construction phase, the BECxP also noted concern with discontinuity in the air/water control layer at the storefront and curtain wall sills and with the exposed metal sill flashing being a single line of defense against water infiltration. Additionally, details were not provided to indicate how the metal sill flashing would be terminated at its ends where it met the precast, which was important to the design if the sill flashing was to form the air/water control layer. As detailed, water infiltration and durability of the plywood beneath the sill flashing was a concern. As such, the BECxP recommended self-adhered membrane flashing below the finished metal flashing sills to provide the continuity of the air/water control layers (fig. 18 and fig. 19). The design-builder addressed the
299
300
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 15 Revised shop drawing detail showing retrofit transition membrane bridging between curtain wall and precast air/water seal (circle). Curtain wall primary air and water seal to be located at inside side of glazing pocket (arrow).
FIG. 16 Minimal overlap between window and inner wythe of precast and missing transition membrane at rough opening.
GIBLIN ET AL., DOI: 10.1520/STP161720180068
FIG. 17 Sheet metal not bedded in sealant at exterior edge.
FIG. 18 Discontinuity between precast and storefront. Detail relied on exposed metal flashing as single line of defense. BECxP recommended membrane flashing (heavy dashed line).
301
302
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 19 Discontinuity between precast and storefront. Detail relied on exposed metal flashing as single line of defense. BECxP recommended membrane flashing (heavy dashed line).
concern with the sill details toward the end of the construction phase by installing sealant between the metal and the precast at the ends of the exposed sill flashing (essentially relying on the sill flashing to be the air/water control layer) (figs. 20–22). During ASTM E1105 water leakage testing, water infiltration occurred at breaches in the sealant at exposed flashing joints (fig. 23). Subsequently, the BECxP recommended preformed silicone seals over the sealant joints at splices and ends of the exposed flashing for added protection for the sealant, which would reduce risk for water infiltration and degradation of the plywood below the flashing. These seals were never installed. During the design phase and throughout construction, the BECxP also identified discontinuous insulation and discontinuous air/water control layers at several precastto-storefront interfaces due to the precast configuration (figs. 24–26). These issues were not addressed until after interior stud framing installation started. Due to water leakage testing failures, the design-builder ultimately removed framing to retrofit each storefront jamb to correct the issue. Retrofits generally included a combination of SPF and self-adhered membrane flashing (fig. 27) to create the continuity of the air/water and thermal control layers across the storefront to precast interface.
GIBLIN ET AL., DOI: 10.1520/STP161720180068
FIG. 20 Sill flashing not sealed to precast, incomplete seal at precast joint.
FIG. 21 Seals installed at sill flashing and precast joints.
Roofing System
During the design phase reviews, the BECxP recommended a continuous vapor retarder on top of the precast concrete plank structural roof deck, both as a temporary roof and as a means to prevent exfiltration of interior air into the roofing
303
304
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 22 Sealant installed at edge of sill flashing to precast.
FIG. 23 Interior water infiltration at sill during ASTM E1105 water leakage testing.
system. Air exfiltration during the wintertime is a concern for roofing systems in Climate Zone 5 because there is a potential for condensation on the underside of the roofing membrane if interior air can flow across the bottom of the roofing
GIBLIN ET AL., DOI: 10.1520/STP161720180068
FIG. 24 Discontinuous air/water seal between precast and storefront.
FIG. 25 Discontinuous air/water seal and insulation between precast inner wythe and storefront.
305
306
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 26 Discontinuous air/water seal and insulation between precast inner wythe and storefront and lack of suitable substrate for inner seal of storefront.
membrane, whose surface temperature is below the dew point temperature of the interior air. A fully adhered vapor retarder detailed to provide air control (sealed at all penetrations, etc.) can be used both to prevent airflow into the roof assembly as well as to provide a temporary roof to help make the building watertight prior to installing the rest of the roofing assembly (fig. 28). The project schedule is also less of a concern because other trades can work on the roof and a damaged vapor retarder can more easily be repaired (as compared to the roofing membrane and materials below it) prior to installation of the remainder of the roofing system. The design team responded to the BECxP’s recommendation by stating that the joints and penetrations in the deck would be sealed or grouted in lieu of including the recommended vapor retarder, primarily due to cost. During the construction phase,
GIBLIN ET AL., DOI: 10.1520/STP161720180068
FIG. 27 Self-adhered membrane retrofit to provide continuous air/water seal between storefront and precast.
FIG. 28 Diagram showing condensation potential without a vapor retarder on the structural deck (left) and how a vapor retarder reduces condensation potential (right).
the BECxP observed several unsealed/ungrouted precast plank joints and isolated gaps at penetrations through the precast planks and between precast planks and parapet walls. During construction, the BECxP also documented several issues that would allow bulk water to be introduced into the roofing system. In addition to the obvious potential for interior leaks, water in a roofing assembly can have many adverse effects, including corrosion of metal components, deterioration of moisturesensitive roofing components (i.e., gypsum, paper facers, and adhesives), reduced
307
308
STP 1617 On Building Science and the Physics of Building Enclosure Performance
thermal efficiency, biological growth, or freeze-thaw damage of concrete or masonry substrates (or any combination thereof). Issues identified throughout the duration of construction that could allow moisture to be introduced and trapped within the roofing system included the following: • Moisture-sensitive insulation and cover board roofing materials stored onsite were not adequately protected from weather with tarps or some other breathable means of covering (figs. 29–31). Wet materials were not always discarded; the BECxP observed moisture-stained materials within completed roofing installations. Insulation boards, cover board, and roofing membrane were installed over standing moisture on top of the roof deck. FIG. 29 Moisture-sensitive roofing materials stored in snow and uncovered.
FIG. 30 Moisture-sensitive roofing materials stored in snow and uncovered.
GIBLIN ET AL., DOI: 10.1520/STP161720180068
FIG. 31 Moisture-sensitive roofing materials stored on roof deck in water and uncovered.
FIG. 32 Water at edge of completed work area.
This moisture will become trapped within the system. Night seals and incomplete flashing conditions were often not watertight, and snow, ice, and water were present on top of the roof deck at the edge of the completed work area (fig. 32) and behind flashings (fig. 33). This water can traverse laterally into the roofing system. Some incomplete flashing conditions were left unsealed for up to two months, which can allow bulk water to enter the roofing system during precipitation events. A large amount of debris and fasteners was present on the finished roofing membrane, which could result in punctures to the roof membrane with foot traffic as work at parapets and penthouses continued.
309
310
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 33 Snow behind incomplete membrane flashing.
The BECxP performed testing of the roofing system at the end of the construction phase to assess if moisture was present within the roofing system. An impedance scan of 100% of the roofing system was performed (fig. 34) using a
FIG. 34 Initial moisture scan (hatched area indicates potential moisture).
GIBLIN ET AL., DOI: 10.1520/STP161720180068
nondestructive electrical impedance scanner (ASTM D7954/D7954M-15a6) and readings were correlated with inspection openings. At areas with high readings during the impedance scan, inspection openings revealed wet cover board and insulation. The cover board core and insulation facers were deteriorated and liquid water was present on the insulation facers (fig. 35). Impedance scanning and inspection openings revealed that 60 to 70% of the overall roof area demonstrated a potential for excessive moisture. The BECxP performed a follow-up scan with additional inspection openings several months later to determine if the affected area of moisture was increasing. The area had grown to approximately 90% of the total roof area (fig. 36). The BECxP also performed a scan of the membrane with high-voltage electronic leak detection to locate any breaches in the roofing membrane (ASTM D7877-147). The scan revealed more than 180 breaches in the membrane (e.g., punctures and incomplete welds) (fig. 37). Due to the widespread moisture issues identified in the roofing system, the system did not meet the owner’s expectations of a watertight, low-maintenance enclosure or the project requirements to receive a manufacturer’s roofing system warranty. As such, the BECxP recommended that 100% of the system be removed and replaced and also recommended inclusion of a vapor retarder in the replacement system (fig. 38). While the BECxP did not conduct testing or investigation to
FIG. 35 Water present on insulation facers.
311
312
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 36 Follow-up moisture scan (hatched area indicates potential moisture).
FIG. 37 Breaches in membrane.
identify the exact sources of moisture, any or all of the following site observations noted by the BECxP were potential causes: • Water/snow entering the system through poor or missing seals (e.g., unsealed/ unclamped flashings, incomplete edge details, inadequate night seals) • Water/snow entering the system at breaches in the membrane
GIBLIN ET AL., DOI: 10.1520/STP161720180068
FIG. 38 Inclusion of vapor retarder with roofing replacement.
•
•
Moisture built into the system by installing potentially wet materials over a wet deck Conditioned interior air entering the roofing system through unsealed penetrations or joints in the deck and condensing on the underside of the roofing membrane
Summary The BECx process is intended to improve quality and reduce risk by verifying that the design and installation of the building enclosure meets the owner’s requirements for performance. Proper implementation of the BECx process can help reduce the potential for water infiltration and air leakage, improve energy performance and durability, and reduce maintenance requirements. When the process is performed with commitment and active participation from all parties involved, it usually is successful at improving quality. However, as evidenced by the case study example, the process does not come without challenges, especially when quality is second tier to cost and schedule and when communication and collaboration are not prioritized. For the case study example, even with design reviews, shop drawing reviews, mock-ups, preinstallation meetings, construction observations, and testing, there were missed opportunities in the BECx process that likely had a negative impact on project schedule and initial cost for the design-builder and, in some cases, resulted in decreased durability or performance. For example, the primary sealant at precast joints likely was not as durable (because SPF was used instead of sealant) or as
313
314
STP 1617 On Building Science and the Physics of Building Enclosure Performance
complete as what could have been achieved if the sealant was installed from the exterior. The discontinuous air/water control layers at the storefront perimeters, which remained open/unresolved issues that required removal of interior finishes, were challenges that could have been avoided if the interface details were addressed during the design phase. The curtain wall and storefront sill that relied on the exposed sill flashing as a single line of defense did not meet the low-maintenance enclosure performance requirement of the owner (because sealant will have to be maintained more frequently than self-adhered membrane flashings—had they been included). The finished roofing system contained trapped water, which did not satisfy the OPR and, due to the extensive amount of water beneath the membrane and breaches in the membrane, required 100% removal and replacement of the roofing system to remediate the issue. While BECx strives to prevent issues such as those noted here, the proactive efforts of the BECx process can be hindered if the owner does not enforce the process and the architects and contractors do not commit to the process or fulfill their contractual obligations. The BECx process requires participation from all parties for the process to be of most value. In addition to the importance of participation and commitment to the process by all project team members, other lessons learned in the case study example that could be used to improve the process on future projects include the following: • The BECxP should be engaged early in the design phase or predesign phase before decisions are made that might negatively affect enclosure performance. If the BECxP was involved earlier in the case study example, there might have been opportunities to positively affect design changes, in collaboration with the bridging architect and design-build team, which could have been zero or low-cost improvements. In several cases, resolving identified issues earlier in the design phase would have resulted in less schedule setbacks and fewer costs during construction. • For a Design-Build project where there is no other party (e.g., an architect) to help resolve issues or enforce the design intent, the owner’s role may increase, and they should be made aware as such. The owner is ultimately responsible for resolving or accepting nonconforming issues when there is a dispute in the resolution. The owner is also responsible for enforcing the process. If project team members are not responsive and the owner does not enforce the team members’ responsibility to respond to action items, issues will go unresolved and untimely changes may be too costly to implement. As such, the owner may choose to settle for a lower-performing building. • Providing clear communication about the roles and responsibilities of all parties is important throughout the process so that all parties understand what is expected. The owner then needs to hold each party accountable if they are not fulfilling their contractual responsibilities.
GIBLIN ET AL., DOI: 10.1520/STP161720180068
QC/QA are critical primary responsibilities that each designer and contractor must maintain. The BECx process cannot be expected to effectively serve as a substitute for those contractual responsibilities. • Increased communication, including more frequent BECx meetings and site walks with the project team, including the owner, provide more opportunities for the project team to resolve issues promptly. Additionally, the BECxP should keep an action item log up to date with all correspondence related to issue resolution and should discuss this action item log with the project team on a regular basis. This is especially important for a Design-Build project that moves quickly. By recommending more than a single line of defense at precast joints, promoting continuous control layers at the fenestration to opaque wall interfaces, identifying moisture within the roofing system, and providing other recommendations and solutions in addition to those discussed in the “Select Identified Concerns” section of this paper, the BECxP was able to help the project team deliver a betterperforming building enclosure than perhaps would have been the case if the BECx process was not implemented. •
References 1.
2. 3. 4.
5.
6.
7.
8.
Standard Practice for Building Enclosure Commissioning, ASTM E2813-18 (West Conshohocken, PA: ASTM International, approved October 1, 2018), http://doi.org/10.1520/ E2813-18 Standard Guide for Building Enclosure Commissioning, ASTM E2947-16a (West Conshohocken, PA: ASTM International, approved June 1, 2016), http://doi.org/10.1520/E2947-16A International Code Council, International Energy Conservation Code (Washington, DC: ICC, 2017). Standard Test Method for Field Measurement of Air Leakage through Installed Exterior Windows and Doors, ASTM E783-02(2018) (West Conshohocken, PA: ASTM International, approved October 1, 2018), http://doi.org/10.1520/E0783-02R18 Standard Test Method for Field Determination of Water Penetration of Installed Exterior Windows, Skylights, Doors, and Curtain Walls, by Uniform or Cyclic Static Air Pressure Difference, ASTM E1105-15 (West Conshohocken, PA: ASTM International, approved August 1, 2015), http://doi.org/10.1520/E1105-15 Standard Practice for Moisture Surveying of Roofing and Waterproofing Systems Using Non-Destructive Electrical Impedance Scanners, ASTM D7954/D7954M-15a (West Conshohocken, PA: ASTM International, approved June 1, 2015), http://doi.org/10.1520/ D7954_D7954M-15A Standard Guide for Electronic Methods for Detecting and Locating Leaks in Waterproof Membranes, ASTM D7877-14 (West Conshohocken, PA: ASTM International, approved August 1, 2014), http://doi.org/10.1520/D7877-14 Canada Mortgage and Housing Corporation, Architectural Precast Concrete Walls: Best Practice Guide (Ottawa, Ontario: Canada Mortgage and Housing Corporation, 2002).
315
316
STP 1617 On Building Science and the Physics of Building Enclosure Performance
9.
10.
Canada Precast/Prestressed Concrete Institute, Architectural Precast Concrete Sealant & Joint Guide (Ottawa, Ontario: Canada Precast/Prestressed Concrete Institute), http:// web.archive.org/web/20200428012431/http://www.tri-krete.com/wp-content/uploads /2016/08/Sealant_and_Joint_Brochure.pdf Standard Test Method for Evaluation of Water Leakage Performance of Masonry Wall Drainage Systems, ASTM C1715/C1715M-15 (West Conshohocken, PA: ASTM International, approved July 1, 2015), http://doi.org/10.1520/C1715_C1715M-15
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180069
Eric K. Olson1 and Anthony J. Nicastro2
Ventilation and Moisture Control in Architectural Metal Panel Roofing Systems Citation E. K. Olson and A. J. Nicastro, “Ventilation and Moisture Control in Architectural Metal Panel Roofing Systems,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 317–337. http://doi.org/10.1520/STP1617201800693
ABSTRACT
Architectural metal roofing systems (AMRS) are considered to be a very durable type of steep-slope roofing system. Architectural metal roofing systems are commonly used in compact or unvented roofing assemblies over moisturesusceptible, nailable wood-based substrates placed directly over or laminated to rigid insulation or over nailable roof decks with spray-applied air- and vaporimpermeable polyurethane foam (SPUF) insulation applied directly to the underside of the deck. Use of AMRS without ventilation to relieve trapped moisture within the compact roofing assembly is commonplace, but under certain conditions, this practice can be risky. As is the case with most roofing systems, water trapped within roofing systems can lead to problems such as corrosion of fasteners and degradation of the substrate and supporting structural elements. Potential sources of water or moisture intrusion are many and can include intrusion of warm, humidified indoor air into the roofing system and condensation on cold surfaces, water leakage through the roofing system, or water entry during construction. Without adequate ventilation to relieve moisture, water trapped in the system can quickly cause damage to the underlying materials in the roof system. Ventilation of roofs over uninsulated decks and attic space is straightforward and commonly practiced, particularly when asphalt shingles are used. However, compact roof assemblies and Manuscript received October 9, 2018; accepted for publication March 20, 2019. 1 Simpson Gumpertz & Heger Inc., 480 Totten Pond Rd., Waltham, MA 02451, USA 2 Simpson Gumpertz & Heger Inc., Washington, DC 20036, USA 3 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21-22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
317
318
STP 1617 On Building Science and the Physics of Building Enclosure Performance
assemblies with insulation above the deck require more sophisticated solutions that consider airflow continuity, placement of air and vapor retarders, and an understanding of the hygroscopic performance of the roofing system. This paper examines what makes architectural metal panel roof systems different from other systems in terms of construction techniques and moisture management, and it presents case studies and computer-based modeling of moisture flow to illustrate the need for ventilation. The authors discuss factors that increase the risk of moisture problems in AMRS and present effective design methods for ventilating and mitigating moisture in these roofing systems. Adequate protection of these systems from the effects of internal moisture using proper ventilation can help provide a robust, durable, and resilient AMRS. Keywords metal roofing, roof ventilation, vented metal roof, compact metal roof, nailbase insulation, roofing moisture, resilient roofing
Introduction Architectural metal panel roofing systems (AMRS) are considered among the most durable steep-slope roofing systems available and are expected to provide a service life of about 40 years.1 AMRS are commonly attached to a nailable sheathing, such as plywood, with a water-resistant underlayment and concealed flashings beneath the AMRS to control water that may pass through open seams, laps, and penetrations through the roofing panels. The nailable sheathing may be secured directly to insulated enclosed rafters or may be integral with a nail base insulation system where the sheathing is laminated to rigid foam insulation and installed directly to a structural deck (a “compact roof assembly”). The AMRS relies on the integrity of the nailable sheathing to transfer wind and gravity loads to the building structure. Therefore, the nailable sheathing and underlying materials must have a service life matching or exceeding the anticipated lifespan of the AMRS. Nailable sheathing and other components, however, may be degraded by prolonged exposure to moisture. If sufficiently severe, decay of these elements can compromise the integrity of the AMRS or its structural support and resistance to loads. Degradation of these elements is generally concealed—one cannot usually detect it by visual inspection. Building codes, including the 2018 International Building Code (IBC),2 specify ventilation of attics and rafter spaces, with some exceptions and will be discussed here. Design and construction methods to provide ventilation beneath the nailable sheathing layer are well established, and ventilation of asphalt-shingle-covered roofing systems is commonly practiced. Guidance regarding AMRS ventilation is more limited; the Metal Construction Association recommends ventilation to control condensation and, in cold climates, to reduce formation of ice dams.3 However, use of ventilation in AMRS directly beneath the sheathing is, in the authors’ experience, infrequent. The possibility of condensation due to trapped moisture within AMRS
OLSON AND NICASTRO, DOI: 10.1520/STP161720180069
is not sufficiently considered when the roof meets exceptions to ventilation requirements allowed by code or when the roof is considered a compact roofing assembly. In some applications, or when used in conjunction with vapor-retarding underlayment (or both), the decision not to ventilate beneath an AMRS can be risky, even when code-allowed ventilation exclusions apply, due to the possibility of trapping or allowing buildup of trapped moisture from various unplanned sources that can occur during construction or during the life of the building. There are many potential sources of water intrusion and moisture accumulation. These sources may include exfiltration of or intrusion of moisture-laden interior air leading to condensation within the assembly, water leakage through or resulting from damage to the AMRS, water entry during construction through an uncompleted/inadequately protected roof, or drying of green (wet) materials following roofing installation. Without adequate ventilation to relieve moisture, and when the roof otherwise lacks the ability to dry, water can become trapped within the roofing system. Consequences of trapped moisture can be dire. Moisture-related problems include corrosion of fasteners; deterioration of the sheathing, insulation, and supporting structural elements; and, if left untreated, disengagement of the roof. The following factors increase the risk of potential damage due to trapped moisture and should be considered in the decision to provide ventilation: • Cold climate where the predominant moisture drive is to the exterior when a vapor-impermeable underlayment is used below the AMRS • Interior vapor retarder, frequently used in colder climates, that retards drying of roofing materials to the interior • Vapor-impermeable metal deck supporting roofing materials, which also retards drying of roofing materials to the interior • Ceiling construction consisting of air-permeable material or with numerous ceiling penetrations that can allow interior air to infiltrate the roofing system • Pressurized or mechanically humidified buildings, particularly in colder climates In addition, there are factors outside the control of the designer that should be considered: • Concealed imperfections and defects in the roofing underlayment system relied on to protect sheathing and underlying materials from incidental water penetration through the AMRS, leading to water infiltration into the roofing system • Concealed imperfections and defects in an internal air barrier system or vapor retarder designed to protect the roofing assembly from air intrusion or air exfiltration or vapor diffusion to the exterior, exposing the system to moisture and condensation • Water entry into the uncompleted roofing system during roofing construction
319
320
STP 1617 On Building Science and the Physics of Building Enclosure Performance
The objective of this paper is to illustrate, with case studies and computerbased heat and moisture transfer models, factors that increase the risk of roofing system degradation, sources and consequences of moisture, and effectiveness of ventilation on mitigating moisture and corresponding risks.
Industry and Building Code Requirements for Ventilation CODE PROVISIONS
Use of ventilation in attic spaces has long been recognized as an effective means to control moisture within attic spaces.4 Where roof deck or attic ventilation is required by the IBC, and where no ceiling vapor retarder is used, the IBC requires inlet (eave) and outlet (ridge) vent openings at a rate of 1 ft2 (0.093 m2) per 150 ft2 (13.94 m2) projected roof area, with vent area split equally between inlet and outlet vents. With a ceiling vapor retarder, the vent area is halved (1 ft2 [0.093 m2]) per 300 ft2 [27.87 m2]). Where ventilation is provided by means of a plenum between rafters and above insulation, or by means of a ventilated deck above rafters, the continuous plenum space must be 1 in. (2.54 cm) deep, continuous between inlet and outlet vents, and provide cross-ventilation around obstructions (such as dormers and chimneys). Building codes contain provisions clearly applicable to ventilation of attic spaces and enclosed rafter spaces, but they also allow exceptions to those ventilation requirements when certain conditions are met. In the United States, requirements governing the ventilation of attics and roofing systems are found in Sections 1203.2 and 1203.3 of Chapter 12, “Interior Environment,” of the IBC.2 Enclosed rafter spaces formed where ceilings (and presumably insulation) are applied directly to the underside of the roof framing spaces are required to be vented by an airspace not less than 1 in. (2.54 cm) in depth between the roof sheathing and insulation. However, ventilation of the attic or enclosed rafter assembly is not required when certain conditions are met (i.e., the attic space is entirely within the building thermal envelope, there is no Class I vapor retarder on the attic floor, and there is a Class II vapor retarder beneath the insulation). If air-permeable insulation is used in rafter spaces, an air-impermeable insulation may also be required above the structural sheathing for condensation control, with the required R-value based on climate zone. These exclusions do not apply to steep-sloped AMRS in buildings humidified above 35% relative indoor humidity in Climate Zones 5 to 8 during wintertime or in special use structures such as swimming pools, art galleries, or data processing centers. Since, by definition, compact steep-slope roofing assemblies do not have insulated rafter spaces and insulation lies entirely above the structural deck (i.e., the unvented attic is entirely within the thermal envelope), the authors do not believe that the code intends for ventilation of the compact AMRS assembly unless the building is humidified or falls under special use categories as described
OLSON AND NICASTRO, DOI: 10.1520/STP161720180069
previously. Also note that IBC Chapter 15, “Roof Assemblies and Rooftop Structures,” references Section 1203.2 for ventilation of attics and rafters and does not explicitly specify ventilation of sheathing below compact AMRS. INDUSTRY GUIDANCE
The effectiveness of roofing ventilation in removing moisture from the assembly is largely dependent on the rate of airflow through the vented assembly.4 Mathematical methods for calculation of ventilation requirements for purposes of removing excessive heat from below compact, steep-sloped roof decks to mitigate ice dams are established by Tobiasson, Tantillo, and Buska.5 These methods can also be used to predict drying airflow beneath a vented, steep-sloped roof deck. Tobiasson, Tantillo, and Buska’s research focuses on the sizing of vent space and inlet areas, given the roof slope and slope length, to achieve ventilation rates sufficient to maintain a roofing surface cold enough (i.e., below 0 C [32 F]) to reduce ice dam formation. One can use formulas presented by Tobiasson, Tantillo, and Buska to predict ventilation airflow by selecting the vent area and calculating airflow. A survey of AMRS manufacturers’ published details finds that manufacturers typically include ridge and eave vent details for application of metal roofing over vented attic space, with no insulation on the deck, similar to details used with asphalt-shingle-covered roofs. However, we find no details for venting sheathing beneath AMRS in a compact AMRS or recommendations or requirements to do so. The Polyisocyanurate Insulation Manufacturers Association (PIMA) promotes the use of vented nail base insulation (VNB) to provide ventilation below the roof sheathing and AMRS.6 VNB consists of a nailable oriented strand board (OSB) or plywood sheathing adhered to wood spacers over the rigid foam insulation. The wood spacers separate the sheathing from the insulation to provide a vented space below the sheathing.
Case Studies CONSTRUCTION MOISTURE TRAPPED IN ROOFING ASSEMBLY
The first case study examined a pitched AMRS in the western United States, in a region defined as a dry, Climate Zone 5 in Chapter 3 of the 2015 International Energy Conservation Code. The roofing system consisted of, from interior to exterior, a corrugated steel deck, an impermeable self-adhered air and vapor retarder, two layers of paper-faced isocyanurate insulation totaling 6 in. (15.2 cm) thick, 0.5 in. (1.3 cm) thick plywood sheathing fastened through insulation to the steel deck, a loose-laid roofing underlayment sheet, and 16 in. (40.6 cm) wide standingseam AMRS panels attached with metal clips (fig. 1). No ventilation space was provided within the assembly. Due to aesthetic issues regarding the AMRS panels unrelated to moisture, an investigation of the roofing system was performed, including sample openings through the roof and insulation. The sample openings revealed widespread corrosion of fasteners securing panels to the plywood
321
322
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 1 Unvented roofing assembly consisting of metal deck, vapor retarder, insulation, nailable substrate, and architectural metal roofing system (AMRS).
sheathing, extensive water staining on sheathing, and mold growth on damp paper insulation facers throughout the roofing system (fig. 2A and 2B). A review of project construction records and weather data during construction revealed several severe rainstorms accompanied by strong gusts during periods of the month-long insulation and underlayment installation period, prior to metal panel installation. Construction records also indicated deficiencies, including reversedshingled underlayment, in roof valleys. However, the existence of widespread damp materials within the roofing system was not detected during construction. According to the roofing underlayment manufacturer, the loose-laid underlayment used to cover the plywood sheathing exceeded “requirements of ASTM D226 [Standard Specification for Asphalt-Saturated Organic Felt Used in Roofing and Waterproofing7]” but has a reported vapor permeance of 0.059 perms (3.36 ng/PaS-m2) and is thus a Class I vapor retarder. We note that #15 asphalt-saturated
FIG. 2 (A) and (B) Mold growth due to moisture trapped in roofing system insulation system as shown in figure 1. (Photos courtesy of The Vertex Companies)
OLSON AND NICASTRO, DOI: 10.1520/STP161720180069
organic felt underlayments are classified by ASTM D226 as semipermeable, with a permeance of about 5.1 perms.8 The roofing assembly was not ventilated. When coupled with the self-adhered air barrier applied directly to the steel deck, the resulting assembly contains insulation and plywood sheathing sandwiched between two Class I vapor retarders. The investigation concluded that water entry into the roofing system occurred primarily during construction, prior to installation of the metal roofing panels. Lack of venting and an underlayment with low vapor permeance (i.e., a vapor retarder) trapped this water between the underlayment and the roof deck vapor retarder/air barrier, preventing drying to the exterior. Use of the self-adhered air and vapor retarder on the steel deck was appropriate for the roofing assembly in the project climate zone, but its presence and relative watertightness minimized water intrusion into the uncompleted building, helping to conceal the water entry into the roofing system. Ultimately, the entire roofing system and insulation required full replacement to remove wet and damaged materials. While poor coordination and scheduling of roofing installation were among the main contributing factors to water entry, rapid drying of the roof afforded by proper ventilation would have allowed drying, minimized damage, and probably would have eliminated the need to replace the entire roofing assembly. HUMIDIFIED AIR EXFILTRATION/DIFFUSION INTO ROOFING SYSTEM
This case examines an occupied wood-framed residence with AMRS in the northeastern United States, Climate Zone 5. The homeowner reported water dripping from recessed light fixtures and joints between ceiling planks in wintertime, which appeared to the authors to likely be condensation. The subject roof assembly consisted of an AMRS applied over self-adhered vapor-impermeable membrane underlayment and a rafter-supported plywood roof (fig. 3). Rafter spaces were insulated with 10 in. (25.4 cm) thick, open-cell, low-density, spray-foam insulation. A wood plank ceiling was secured to the underside of the rafters. Recessed lights penetrated the ceiling and projected into the insulated rafter space. The pitch of the roof was approximately 1:12 (8%), with an overall eave-to-ridge elevation difference of approximately 5 ft (1.52 m). An investigation into the causes of the condensation revealed indoor relative humidity (RH) levels of about 35% and lack of a vapor retarder directly above the ceiling. The open-cell spray foam, intended to provide an air barrier, contained voids around recessed light fixtures, against the plywood roof deck, and at rafter bearings on exterior walls, and lacked uniform thickness, all of which compromised its effectiveness as an air barrier and created air paths into and throughout the ceiling and insulated rafter space. Since low-density, open-cell foam insulation is vapor permeable, diffusion of moisture through the foam is possible. Removal of metal roofing panels during wintertime investigations revealed that the entire unvented plywood roof deck was covered with a continuous self-adhered, asphaltic ice dam protection membrane, which is also a Class I vapor retarder. The plywood was damp beneath the self-adhered membrane and had areas of decay.
323
324
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 3 Unvented roofing assembly over plywood deck and vapor-permeable insulation in rafter space. Roofing underlayment is vapor impermeable.
Decay was severe at eaves beyond exterior walls due to exfiltration of warm, humid interior air into the cold, unconditioned eaves (fig. 4). Replacement of the roofing system was required to correct structural damage and defects in the air barrier and spray-foam insulation and to replace wet materials. Due to occupancy of the building, disruption, and cost involved, the owner decided against removing the plank ceiling to provide a vapor retarder. Computer-based thermal and moisture transfer modeling supported the conclusion that it would be necessary to reconstruct the roofing system to correct and fill voids in the insulation, provide air barrier continuity at eaves, and ventilate the roof assembly. To provide adequate ventilation, ensure air barrier adequacy, and reliable waterproofing beneath the low-sloped AMRS, the ventilated assembly (fig. 5 and fig. 6) included the following: • New structural plywood deck on rafters following repair of spray-foam insulation voids. • Permeable, self-adhered, sheet-applied air barrier membrane. • 1.5-in. (3.8 cm) ventilated air space. • Plywood roof deck. • Continuous, self-adhered, asphaltic ice dam protection membrane covering the entire deck, as is appropriate with the 1:12 roof pitch. Although the ice dam protection membrane is a vapor retarder, we demonstrated by computer modeling (as discussed and shown as follows) that the introduction of the new, ventilated air space within the assembly allowed drying and prevented condensation.
OLSON AND NICASTRO, DOI: 10.1520/STP161720180069
FIG. 4 Moisture damage in roofing system described in figure 3. Due to interior air intrusion into rafter space, there is diffusion of interior moisture through permeable insulation. Impermeable underlayment prevents drying.
FIG. 5 Ventilation space added to roofing assembly shown in figure 3. Ventilation space allows use of impermeable underlayment such as ice dam protection membrane with low risk of condensation.
The new assembly provided sufficient air exfiltration reduction and ventilation to compensate for diffusion through the open-cell spray-foam insulation while eliminating the need for installation of an interior vapor retarder. The new roof assembly, which has been in service for five years, has no reported moisture accumulation or condensation problems.
325
326
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 6 Photograph of partially installed system as shown in figure 5 with air barrier on structural deck, ventilation space furring, and vented eave blocking. Tarpaulin is temporary.
Supporting Computer Modeling of Heat and Moisture Transfer Mechanisms The ability for ventilation to relieve moisture accumulation in AMRS is demonstrable through the use of computer-aided simulations. As an example, assemblies approximating those described in the previous case studies can be modeled with and without a ventilation layer to show its potential benefit. To this end, the authors used WUFI Pro 6.1 developed by the Fraunhofer Institute for Building Physics (Germany) to calculate the transient one-dimensional heat and moisture transport to determine the potential for condensation and moisture accumulation within varying AMRS roof constructions and boundary conditions. The following simulations compare ventilated and unventilated AMRS. The results of computer simulations using WUFI corroborate the authors’ field experience and show the advantages of circulating air to improve the drying potential of roofing materials. Note that the models described here are a qualitative comparison tool among various assemblies used to assess relative performance of unvented and vented assemblies and risk of moisture buildup in the modeled assemblies under assumed conditions. Quantitative performance varies based on project-specific conditions. ROOF ASSEMBLY AND BOUNDARY CONDITIONS FOR MODELS
Modeled systems use basic geometry shared by many AMRS. WUFI models described here include the following components, from interior to exterior:
OLSON AND NICASTRO, DOI: 10.1520/STP161720180069
Corrugated metal deck Glass-mat-faced gypsum board • Sheet-applied vapor retarder • Polyisocyanurate insulation, 5 in. thick • Plywood (ventilated model only) • Varied air gap for ventilation • 0.5-in. plywood nail base • Underlayment (varied between permeable and impermeable) • Standing-seam metal roofing system To approximate interior conditions, the models use a sinusoidal curve with a one-year period. The models used the WUFI preset for medium moisture load interior conditions with a mean temperature set to 21 C (69.8 F) with an amplitude of 1.0 C (1.8 F) and RH set with a mean value of 50% and an amplitude of 10%. The interior moisture load will vary with occupancy, size of space, and use; however, the medium moisture load preset is an appropriate approximation for a generalized simulation. For exterior conditions, we selected a predefined WUFI weather file corresponding to a cold year in Boston, MA. • •
MATERIALS AND ASSUMPTIONS •
•
•
•
•
•
The models generally use predefined materials included in the WUFI Pro 6.1 program material databases; however, the models assume a permeance of 0.1 perms for the self-adhered vapor retarder. All models assume a 22.62 inclination (5:12 slope). The simulation time period is five years to analyze long-term performance trends (e.g., potential moisture accumulation) in the moisture-sensitive layers (i.e., plywood sheathing) of the assembly. Each case is run for a simulated time period of two years to stabilize the initial moisture content and is then reviewed based on the results of the following three-year period. The models assume constant initial moisture conditions across all assembly components at an RH of 80% to account for built-in construction moisture. Except where manually introduced, the models do not account for air leakage, two-dimensional effects of heat or moisture flow, bulk water leakage through the roofing system, or airflow into the roofing system. As noted earlier, in the second case study, these effects can cause or greatly exacerbate moisture accumulation within roof assemblies. Metal roofing is simulated with WUFI’s predefined “unperforated metal deck” material. The WUFI definition for this material assigns a vapor permeance of 0.64 perms, which is a Class II vapor retarder. Some vapor permeability is appropriate in this layer because joinery in the metal roofing will not be airtight and some vapor exchange approximates hydrothermal function of the metal. Note that the models examine variation in air changes within the ventilation layer because experimental research has not yet quantified real airflows in
327
328
STP 1617 On Building Science and the Physics of Building Enclosure Performance
•
•
vented metal roofs. However, Tobiasson, Tantillo, and Buska5 give steep-sloped roofing temperatures and predict corresponding required ventilation rates to achieve them. Using this technique, a 5:12 sloped roof with a 30-ft (9.1 m) rise, 70-ft (21.3m) run, and 1.5-in. (3.8 cm) deep vent space will approach 60 air changes per hour (ACH), the chosen upper limit adopted herein. We also run vented models using an arbitrary 5 ACH to help determine drying effectiveness at ventilation rates well below those used for cold roof design. Unvented models include a water leak source corresponding to 0.1% of the adhering fraction of rain, while the ventilated models include a water leak source of 1.0% of the adhering fraction of rain. The authors selected a water leak source an order of magnitude greater for the vented models to demonstrate their resiliency compared with the unvented assemblies.
CASES MODELED
The computer simulation includes four cases that explore the hygrothermal performance of AMRS with vapor-impermeable or vapor-permeable underlayments in both unvented and vented assemblies. Secondly, the cases were developed to demonstrate that ventilation rates, specifically ACH, as well as manual introduction of a moisture source (i.e., leakage) can have a drastic effect on the potential for moisture accumulation. For each case, models were run for roofs with and without leaks. Modeled cases include the following: Unvented Assemblies • Case 1A: Unvented assembly with permeable underlayment (fig. 7). • Case 1B: Unvented assembly with permeable underlayment and simulated water leak in plywood layer corresponding to 0.1% adhering fraction of rain (fig. 7).
FIG. 7 Unvented assembly with permeable underlayment used in Case 1A and Case 1B, table 1.
OLSON AND NICASTRO, DOI: 10.1520/STP161720180069
FIG. 8 Unvented assembly with impermeable underlayment used in Case 2A and Case 2B, table 1.
Case 2A: Unvented assembly with impermeable underlayment (corresponding to observed field failures) (fig. 8). • Case 2B: Unvented assembly with impermeable underlayment and leak in plywood layer corresponding to 0.1% adhering fraction of rain (fig. 8). Vented Assemblies • Case 3A: Vented assembly with permeable underlayment. Ventilation rate of 5 ACH to assess relationship between ventilation rate and hygrothermal performance (fig. 9). • Case 3B: Vented assembly with permeable underlayment. Ventilation rate of 60 ACH to assess relationship between ventilation rate and hygrothermal performance (fig. 9). •
FIG. 9 Vented assembly with permeable underlayment used in Cases 3A–3C, table 1.
329
330
STP 1617 On Building Science and the Physics of Building Enclosure Performance
•
•
•
•
Case 3C: Vented assembly with permeable underlayment. Ventilation rate of 60 ACH. Leak introduced in ventilation layer corresponding to 1.0% adhering fraction of rain (fig. 9). Case 4A: Vented assembly with impermeable underlayment. Ventilation rate of 5 ACH to assess relationship between ventilation rate and hygrothermal performance (fig. 10). Case 4B: Vented assembly with impermeable underlayment. Ventilation rate of 60 ACH to assess relationship between ventilation rate and hygrothermal performance (fig. 10). Case 4C: Vented assembly with impermeable underlayment. Ventilation rate of 60 ACH. Leak introduced in ventilation layer corresponding to 1.0% adhering fraction of rain (fig. 10).
RESULTS
ASHRAE Standard 160-2016, Criteria for Moisture-Control Design Analysis in Buildings, provides guidelines and criteria for performing hygrothermal analyses. The standard defines a criteria for determining the level of biological growth on material surfaces with a “mold index” number on a scale from zero (no growth) to six (heavy and tight growth, coverage about 100%). The mold index is calculated from simulation results at any surface of an assembly and incorporates time, surface temperature, and surface relative humidity. The mold index calculation also varies with the sensitivity class of the building material surface being evaluated. The four material sensitivity classes are very sensitive, sensitive, medium resistant, and resistant. The standard states that the mold index shall not exceed three (the threshold for visible biological growth) regardless of the sensitivity class. For each case described earlier, the mold index was reviewed for the exterior surface of the plywood roof sheathing, which is the most moisture-sensitive plane
FIG. 10 Vented assembly with impermeable underlayment used in Cases 4A–C, table 1.
OLSON AND NICASTRO, DOI: 10.1520/STP161720180069
in the roof asembly and is where condensation potential is highest based on our initial review of WUFI output for RH at these surfaces. The plywood sheathing is also a critical structural element for attachment of the AMRS, so occurrence of decay in the sheathing can be detrimental to roofing integrity and longevity. For Cases 2 through 4, the glass mat facer of roofing insulation was also examined in the plane of the self-adhered vapor retarder. We also reviewed the direct WUFI output RH profiles for indications of potential condensation. Given the uncertainties of material properties, we generally consider RH in excess of 95% as having a potential for condensation occurrences. Results from the analysis are summarized in table 1. Cases 1 and 2 show hygrothermal performance of unvented assemblies. The models corroborate field observations where unvented assemblies result in moisture accumulation (e.g., Case 2A). The models also show that unvented assemblies have no tolerance for incidental moisture as Cases 1B and 2B introduce a leak in the plywood layer corresponding with 0.1% adhering fraction of rain. Introduction of moisture causes these assemblies to fail even if permeable (i.e., felt paper) underlayment is used. With the vented assemblies, choice of permeable or impermeable underlayment makes little difference in hygrothermal performance because both types of underlayments are associated with low risk of condensation or moisture accumulation. Varying air changes between 5 ACH and 60 ACH resulted in little consequential difference between models, although higher ACH rates appear to correlate with faster drying of assembly components in the wintertime but result in a higher RH in the vent space during winter months. Note that higher ventilation rates marginally increase the average relative humidity within the vent space as the moisture laden exterior air circulates into the vented cavity. Vented models with a leak source of 1.0% adhering fraction of rain (an order of magnitude greater than the assumed leak volume for unvented assemblies) maintain an acceptable RH in all layers inboard of the vented space. Introduction of a leak in the plywood layer may result in local degradation of the plywood in the vicinity of the leak but will not affect the overall assembly performance. Comparison of the unvented and vented assemblies with introduced leakage (Cases 2B and 4B) shows the ability of a vented assembly to handle moderate water leakage without detrimental effects except localized plywood degradation, while unvented assemblies fail even with far less manually inserted moisture.
Discussion AMRS are intended to provide a service life that often exceeds that of typical singleply or asphalt-shingle roofing systems. To achieve the intended service life, moisture-susceptible components of the roofing system, including insulation, structural wood sheathing, and metal fasteners, should be ensured an environment free of trapped or excessive moisture. Ventilation of the AMRS helps to achieve this
331
332
Ventilation
Underlayment Type
Risk of Condensation or Moisture Accumulation
Mold Growth Potential (ASHRAE 160)
Leak Source
Comments
Case 1A
Unvented
Permeable
Low
Low
No added moisture.
No leak assumed. Critical assembly components lower
Case 1B
Unvented
Permeable
High-Certain
High-Certain
Leak at top of plywood
Assembly experiences condensation in plywood and
layer corresponding to 0.1%
95% RH in vapor retarder with year-over-year moisture
than 80% RH.
Case 2A
Unvented
Impermeable
(See fig. 11)
High- ASHRAE criteria
High-Certain
adhering fraction of rain.
accumulation.
No added moisture.
Condensation potential high during summer months on top of self-adhered vapor retarder/base of insulation.
exceeded 65% of the
Thirty-day running average RH exceeds 80% for
year
approximately 28% of the year in the plane of the vapor retarder/base of insulation. Similarly, at base of insulation, the mold index criteria exceeds three (high risk) 49% of the year on average across the five-year simulation. Model corroborates field observations of failures where either trapped construction moisture or leakage can rapidly deteriorate moisture-sensitive roofing components. Case 2B
Unvented
Impermeable
High-Certain
High- Certain
Leak at top of plywood
100% RH and condensation in multiple layers within
layer corresponding to 0.1%
roofing assembly.
adhering fraction of rain. Case 3A
5 ACH
Permeable
Low
Low
No added moisture.
Assembly does not approach 80% RH or higher. Lower RH maintained in critical layers compared to unvented counterpart in Case 1A.
(continued)
STP 1617 On Building Science and the Physics of Building Enclosure Performance
TABLE 1 WUFI analysis results summary
TABLE 1
(continued)
Case 3B
Ventilation
Underlayment Type
Risk of Condensation or Moisture Accumulation
Mold Growth Potential (ASHRAE 160)
Leak Source
Comments
60 ACH
Permeable
Low
Low
No added moisture.
Vented cavity maintains slightly higher RH in winter months compared with Case 3A. Vented cavity peaks at approximately 84% RH for brief periods during the wintertime.
Case 3C
60 ACH
Permeable
Low but localized
High near leak point,
Leak at top of plywood
Vented cavity peaks at value similar to Case 3B during
decay in plywood
low elsewhere
layer corresponding to
the wintertime. Leak in plywood will cause local
1.0% adhering fraction
degradation.
of rain. 5 ACH
Impermeable
Low
Low
No added moisture.
Assembly does not approach 80% RH or higher. Lower RH maintained in critical layers compared to unvented counterpart in Case 1.
Case 4B
60 ACH
Impermeable
Case 4C
60 ACH
Impermeable
Low
Low
No added moisture.
Vented cavity maintains slightly higher RH in winter
Leak at top of plywood
Vented cavity peaks at value similar to Case 3B during
layer corresponding to
the wintertime. Leak in plywood will cause local
1.0% adhering fraction
degradation.
months compared with Case 4A. (See fig. 12)
Low, localized failure in High near leak point, plywood
low elsewhere
of rain.
OLSON AND NICASTRO, DOI: 10.1520/STP161720180069
Case 4A
333
334
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 11 Illustrative example of hygrothermal model output, Case 2A.
FIG. 12 Illustrative example of hygrothermal model output, Case 4C.
OLSON AND NICASTRO, DOI: 10.1520/STP161720180069
goal, provides extra protection, and adds resiliency against moisture that may not be accounted for in the roofing system design. Hygrothermal modeling discussed here corroborates field observations of successful and unsuccessful assemblies. Ventilation is helpful in reducing risk of condensation within all AMRS assemblies, but it is critical to reducing condensation in conditions where extraneous airflow or incidental water leakage may enter the roofing system. Once in the assembly, leakage becomes trapped with no means of relief because the modeled assemblies have vapor-retarding materials on the winter-warm (interior) side. The models presented here are based on AMRS in colder climates, where a vapor retarder on the winter-warm side of the roofing assembly typically is specified. Warmer climates that omit this vapor retarder are outside the scope of our analysis, but the risk of internal moisture buildup should be lower under these conditions because the predominant vapor drive is to the interior, provided they are free to dry to the building interior (i.e., no vapor retarder on the underside of the thermal insulation). However, in all cases, additional consideration must be given to buildings with controlled or high interior humidity, substantial pressurization, or other special instances. The authors note the use of vapor-impermeable underlayment with AMRS. In some instances, self-adhered ice dam protection underlayment meeting ASTM D1970, Standard Specification for Self-Adhering Polymer Modified Bituminous Sheet Materials Used as Steep Roofing Underlayment for Ice Dam Protection,9 is used, particularly on AMRS with pitch lower than 3:12 where added protection against water ingress through seams or flashings is desired. Use of ice dam protection underlayment is generally appropriate for these circumstances; however, the design must take into account the impermeable nature of this type of underlayment. Ventilation below this material can provide adequate protection from moisture that may enter materials below the underlayment. Use of loose-laid, low-permeance, sheet-applied polymeric underlayment intended to be used in lieu of ASTM D226 asphaltic felt underlayment is also common. Underlayment with low vapor permeance lacks a key property of traditional asphalt felt underlayment—breathability. Users must carefully consider the permeance of the underlayment. If an impermeable underlayment is used in a cold climate in conjunction with a roof deck vapor retarder or where construction leaves the possibility of air intrusion into the roofing assembly, ventilation should be provided. An airflow rate through the vented deck space of 5 ACH provides ventilation sufficient to reduce moisture levels. However, actual airflow rates through the vent space can vary widely depending on roof pitch and temperature differences between exterior and interior. Our models assume that the predetermined airflow rates used in the models will be achieved by the described roofing and venting geometries, but this assumption is not based on field measurement of airflow. Further research into airflow through vented deck assemblies is needed to develop guidance helpful for roofing system design.
335
336
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Conclusions Based on the field observations and hygrothermal analyses discussed here, the following conclusions may be drawn regarding assembly performance. • Impermeable underlayments in compact, unvented AMRS assemblies can increase risk of trapped moisture and system degradation, even with no additional water introduced in service, if installed materials contain some moisture. Design and installation of impermeable underlayments, particularly in colder climates, must consider this risk and should use roof deck ventilation to reduce this risk. • Where an interior vapor retarder is used, avoid use of impermeable underlayment except in vented assemblies. • A ventilation layer improves drying potential, even at relatively low ACH rates. The models and examples presented here represent limited situations and conditions and do not anticipate the numerous variations and conditions that can affect the design and performance of an AMRS. Hygrothermal performance and risks associated with varying materials, configurations, environmental conditions, and other factors will fluctuate on a case-by-case basis. However, use of ventilation, assemblies, and materials that promote drying greatly reduce the risk of trapping moisture within the system and improve chances of an extended service life.
References 1.
2. 3. 4.
5.
6.
7.
Fannie Mae, Instructions for Performing a Multifamily Property Condition Assessment– Appendix F (Estimated Useful Life Tables), Form 4099F (Washington, DC: Federal National Mortgage Association/Fannie Mae, October 2014). International Code Council, “Interior Environment,” in International Building Code (Falls Church, VA: International Code Council, 2018), 321–322. F. B. Rowley, A. B. Algren, and C. E. Lund, “Condensation of Moisture and Its Relation to Building Construction and Operation,”ASHVE Transactions 45 (1939): 231–252. T. W. Forrest and I. S. Walker, “Attic Ventilation Model,” in Proceedings of the ASHRAE/ DOE/BTECC Fifth Conference on Thermal Performance of Exterior Envelopes of Buildings (Atlanta, GA: ASHRAE, 1992), 399–408. W. Tobiasson, T. Tantillo, and J. Buska, “Ventilating Cathedral Ceilings to Prevent Problematic Icings at Their Eaves,” in Proceedings of the North American Conference on Roofing Technology (Rosemont, IL: National Roofing Contractors Association, 1999), 84–97. “Ventilated Nail Base for Commercial and Residential Sloped Roofs,” Technical Bulletin #114, Polyisocyanurate Insulation Manufacturers Association, July 2011, https://web. archive.org/web/20191204131250/https://cdn.ymaws.com/www.polyiso.org/resource/resmgr /Tech_Bulletins/tb114_Feb2017.pdf Standard Specification for Asphalt-Saturated Organic Felt Used in Roofing and Waterproofing, ASTM D226/D226M-17 (West Conshohocken, PA: ASTM International, approved June 15, 2017), http://doi.org/10.1520/D0226_D0226M-17
OLSON AND NICASTRO, DOI: 10.1520/STP161720180069
8.
9.
American Society of Heating, Refrigeration and Air Conditioning Engineers, “Heat, Air, and Moisture Control in Building Assemblies–Material Properties,” in ASHRAE Fundamentals Handbook (Atlanta, GA: ASHRAE, 2017), 26.1–26.23. Standard Specification for Self-Adhering Polymer Modified Bituminous Sheet Materials Used as Steep Roofing Underlayment for Ice Dam Protection, ASTM D1970/D1970M-19 (West Conshohocken, PA: ASTM International, approved November 1, 2019), http:// doi.org/10.1520/D1970_D1970M-19
337
338
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180105
Benjamin Meyer,1 Maria Spinu,2 and Elizabeth Cassin3
Air Barrier Performance and Life Cycle from Inception to Installation Citation B. Meyer, M. Spinu, and E. Cassin, “Air Barrier Performance and Life Cycle from Inception to Installation,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 338–385. http://doi.org/10.1520/STP1617201801054
ABSTRACT
An air barrier system is the combination of interconnected materials, assemblies, sealed joints, and components of the building enclosure that control airflow between conditioned and unconditioned spaces or between spaces that are conditioned differently. For it to perform its intended function to control airflow across the building enclosure, the air barrier must be air impermeable, must be continuous, must maintain structural integrity, and must be durable in its installed and long-term application. When these performance requirements are met, a continuous air barrier will provide energy savings, comfort for the building occupants, durability, and a reduced environmental footprint. The first portion of this paper will describe the air barrier’s environmental benefits by comparing the “environmental cost” of an air barrier system throughout its life cycle assessment (LCA) with the “environmental benefits” due to building envelope airtightness during the building use phase (estimated through whole building energy simulations). The analysis of two air barrier systems shows that the environmental payback period is between a few months to a year and that the energy saved through airtightness over the operational phase of the building amounts to
Manuscript received October 31, 2018; accepted for publication July 11, 2019. 1 ECS Mid-Atlantic, LLC, 2119-D N. Hamilton St., Richmond, VA 23230, USA http://orcid.org/0000-00016629-8397 2 DuPont Protection Solutions, 974 Centre Rd., Wilmington, DE 19805, USA 3 Wiss, Janney, Elstner Associates, Inc., 10 South LaSalle St., Chicago, IL 60516, USA https://orcid.org/ 0000-0003-2563-9807 4 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
MEYER ET AL., DOI: 10.1520/STP161720180105
significant energy and CO2 emission credits. The second portion of this paper will discuss design- and construction-related items that are critical to air barrier and, ultimately, to whole building performance. Since the interface between the air barrier materials and adjacent systems (e.g., roofing, fenestration, waterproofing, penetrations, etc.) is oftentimes where whole building performance is most influenced, the designer must clearly illustrate the continuity of the air barrier across all interfaces, transitions, and penetrations. Examples of such details will be discussed. Keywords air barrier, life cycle, LCA, performance, BECx, quality assurance, verification, commissioning, enclosure
Background AIR BARRIER BACKGROUND
The air barrier system is included at all six sides of the building enclosure. It consists of the air barrier material itself at the opaque building walls (e.g., fluid-applied membrane, self-adhered sheet membrane, mechanically fastened wrap, rigid insulation board, medium-density closed-cell spray polyurethane foam, or factorybonded membranes to sheathing) and associated air barrier material accessories (e.g., fasteners, adhesives, tapes, flashing, transition membranes, and sealants) as well as adjacent systems, such as fenestration, roofing, cladding that may function as the air barrier (i.e., precast, insulated metal panel, etc.), and below-grade waterproofing and the interfaces between different systems. For it to perform its intended function to control airflow across the building enclosure, the air barrier must be air impermeable, must be continuous, must maintain structural integrity, and must be durable in its installed and long-term application. Air impermeability is the ability for the material or the assembly to resist air leakage. Continuity of the air barrier is critical at interfaces, transitions, and penetrations, and as such, design details must indicate how the continuity is intended to be achieved. Additionally, installation of the air barrier requires multiple components and multiple trades’ coordination. An air barrier’s structural integrity allows the air barrier to withstand pressure loads by being supported and to transfer these pressures to structural elements of the building enclosure without leakage. Finally, air barriers must be durable in order to perform over the service life of the enclosure, withstanding environmental exposure, including ultraviolet (UV) and thermal exposure, thermal cycling, repeated exposure to water, abrasion, and mechanical stresses. AIR BARRIER TESTING AND COMPLIANCE OPTIONS
Air barriers can be tested as an air barrier material, air barrier assemblies, or whole building airtightness. The simplest compliance option for air barriers is through materials. The standard method for air barrier materials testing is ASTM E2178-13, Standard Test Method for Air Permeance of Building Materials.1 While materials’
339
340
STP 1617 On Building Science and the Physics of Building Enclosure Performance
testing is necessary, it does not guarantee the performance of installed air barriers. The ultimate goal of a continuous air barrier is to achieve an airtight building, which is dependent not only on the air barrier material but also on the continuity at penetrations and transitions to adjacent systems, as well as the airtightness of other systems that may also function as the air barrier. As such, manufacturer evaluation testing for airtightness of air barrier assemblies (ASTM E2357-18, Standard Test Method for Determining Air Leakage Rate of Air Barrier Assemblies,2 and ASTM E1677-19, Standard Specification for Air Barrier (AB) Material or Assemblies for Low-Rise Framed Building Walls3) and for whole buildings (ASTM E779-19, Standard Test Method for Determining Air Leakage Rate by Fan Pressurization,4 ASTM E1827-11[2017], Standard Test Methods for Determining Airtightness of Buildings Using an Orifice Blower Door,5 ASTM E3158-18, Standard Test Method for Measuring the Air Leakage Rate of a Large or Multizone Building,6 and Air Barrier Association of America [ABAA]-2016, Standard Method for Building Enclosure Airtightness Compliance Testing) are also important tests that are used to determine the performance of installed air barrier systems. Testing an air barrier assembly is an important intermediate step toward achieving an airtight building envelope because it measures the performance of installed air barriers, and testing can take place in the lab. Air barriers can have very different performance characteristics depending on the installation details, which in turn affect the ultimate performance of the building as a whole. As such, when selecting an air barrier material or system, it is important for the designer to specify the desired performance level. There are two accepted performance levels for installed air barrier assemblies in commercial buildings, determined by the design criteria for wind loads for the building envelope: • ASTM E1677-19 for building envelope designs requiring up to 65 mph equivalent structural loads and up to 15 mph equivalent wind-driven rain water infiltration resistance • ASTM E2357-18 for higher wind loads (above 65 mph) The desired performance level must be specified by the design team. As a general guideline, ASTM E1677-19 is acceptable for low-rise, light, commercial construction while ASTM E2357-18 is often required for high-performance, commercial construction. Whole building airtightness testing requires pressurizing and depressurizing the building envelope using blower door fans or the building’s mechanical system. The test protocol is based on ASTM E779-19, which was originally developed for measuring the airtightness of building envelopes of single-zone buildings. In addition to measuring airtightness, whole building testing uses diagnostics tools (thermal infrared and smoke) to identify locations of air leakage and to facilitate continuous improvement in air barrier practices. Once airtightness performance requirements are specified, either based on code requirements or other performance needs, the designer must consider several other
MEYER ET AL., DOI: 10.1520/STP161720180105
air barrier characteristics. Two essential characteristics include water resistance and water vapor permeability, both critical for building enclosure durability and moisture management. Other physical properties that are relevant for installed air barrier performance may include puncture resistance, tensile strength, peel or stripping strength of adhesive bonds, lap adhesion, low temperature flexibility, crack bridging, flame spread, dimensional stability, pliability, and others. These material properties are generally provided in the manufacturer’s technical data sheet. Other considerations include UV resistance, application temperature, inservice temperature, surface preparation requirements, combustibility, compatibility with adjacent materials, required detailing (penetrations, deflection joints, etc.), and the designer’s and local contractors’ familiarity and comfort with the various products and assemblies. CODE REQUIREMENTS AND STANDARDS
Due to the increased recognition of the importance of air barriers to improving building performance, air barriers are required by code in most building types and climate zones. Even though air leakage can impact many aspects of building performance, air leakage control is regulated through energy codes. A focusing on the energy aspects of air leakage in the code, rather than moisture, is simpler because the impact on energy use can be estimated through whole building energy simulations, while moisture intrusion and its impact on durability are harder to quantify. It’s helpful to compare the ASHRAE Standard 90.1, Energy Standard for Buildings Except Low-Rise Residential Buildings,7 baseline energy efficiency performance requirements over time (fig. 1).
FIG. 1 Energy standard savings, efficiency code minimum.
341
342
STP 1617 On Building Science and the Physics of Building Enclosure Performance
ASHRAE Standard 90.1 has continuously increased its requirements for improved energy efficiency with each published version, and continuous air barriers have contributed to the progress (fig. 2). The energy efficiency improvements adopted in ASHRAE Standard 90.1 require rigorous technical justification and sophisticated cost justification that takes into account net present value (NPV) and service life of the components. The most current (as of this writing) published ASHRAE Standard 90.1-2016 has the most improved baseline energy efficiency requirement (highest percentage of energy savings/lowest energy use) when compared to all previous versions, and it is cost effective. The technical requirements and methods of compliance for air barriers in the ASHRAE 90.1 Standard regularly change, and it is updated based on improved technical knowledge and installation costs. Specifically, the ASHRAE 90.1-2007 publication included limited guidelines for air leakage control, without quantifiable requirements for building envelope airtightness, by generally following the prescriptive guidance in the "Building Envelope Sealing Design" section. In the next version (ASHRAE 90.1-2010), performance criteria for materials (ASTM E2178) and assemblies (ASTM E2357, ASTM E1677, ASTM E1680, Standard Test Method for Rate of Air Leakage through Exterior Metal Roof Panel
FIG. 2 Continuous air barriers in ASHRAE Standard 90.1-2016.
MEYER ET AL., DOI: 10.1520/STP161720180105
Systems,8 and ASTM E283, Standard Test Method for Determining Rate of Air Leakage Through Exterior Windows, Skylights, Curtain Walls, and Doors Under Specified Pressure Differences Across the Specimen9) were added. This 2010 compliance addition aided designers and manufacturers in specifying products to meet the energy codes. To meet the materials performance criteria, an air barrier material must have an air infiltration rate not exceeding 0.004 cfm/ft2 (cubic feet per minute per square foot) at 0.3 in. wc (inch water column) pressure differential (0.02 L/[s m2] at 75 Pa—liters per second per square meter at 75 pascal). For assemblies, most current codes require that the average air leakage rate not exceed 0.04 cfm/ft2 under a pressure differential of 0.3 in. wc (0.2 L/[s m2] at 75 Pa) when tested in accordance with ASTM E2357-18 or ASTM E1677-19. The ASHRAE 90.1-2013 version included two significant additions regarding air barrier compliance requirements. The first 2013 change provided extensive air barrier verification procedures and documentation requirements to help ensure the installed performance of the entire building enclosure. The second change in 2013 added a whole building airtightness test (ASTM E779) as a compliance option and added a modeling path in Appendix G to allow project teams to take credit for improved air barrier design and installation. The ASHRAE 90.1-2016 version reconfigures the air barrier prescriptive compliance methods, placing the whole building testing method as the primary continuous air barrier path. Previously, the whole building airtightness method was referenced as an "exception" in the 2013 version. The move of the whole building air testing path in the standard helps to clarify the intention of air barrier compliance and aligns more closely with the modeling and performance compliance path available in Appendix G of ASHRAE 90.1. ASHRAE Standard 90.1-2013 and 2016 reference a maximum whole building air leakage rate of 0.40 cfm/ft2 under a pressure differential of 0.3 in. wc (2.0 L/ [s m2] at 75 Pa) when tested by an independent third party in accordance with ASTM E779 or ASTM E1827. The U.S. Army Corps of Engineers (USACE) Protocol was the first whole building airtightness testing requirement developed for projects under the jurisdiction of the Unified Facilities Criteria and “provides planning, design, construction, sustainment, restoration, and modernization criteria, and applies to the Military Departments, the Defense Agencies, and the DoD Field Activities.”10 The USACE Protocol came into effect based on the publication of the USACE Engineering and Construction Bulletin No. 2012-16, issued in May 2012.10 The most recent USACE Protocol requires less than or equal to 0.25 cfm/ft2 at 75 Pa for whole building airtightness. The General Services Administration (GSA) PBS-P100, Facilities Standards for the Public Buildings Service, is GSA’s mandatory facilities standard. It applies to design and construction of new federal facilities, major repairs and alterations of existing buildings, and lease construction facilities that GSA intends to own or has the option to own (Design Excellence in Leasing). P100 users span the entire
343
344
STP 1617 On Building Science and the Physics of Building Enclosure Performance
spectrum of building professional disciplines, and the P100 informs and regulates decisions made throughout a project’s life.11 The GSA/PBS-P100 2017 standard requires less than or equal to 0.40 cfm/ft2 at 75 Pa for “baseline” whole building airtightness, 0.25 cfm/ft2 at 75 Pa for “Tier 1 High Performance” buildings, 0.15 cfm/ft2 at 75 Pa for “Tier 2 High Performance” buildings, and 0.10 cfm/ft2 at 75 Pa for “Tier 3 High Performance” buildings. A more stringent requirement of less than 0.25 cfm/ft2 at 75 Pa for whole building airtightness is included in Naval Facilities Engineering Command (NAVFAC)2011 and International Green Construction Code (IgCC)-2012. A continuous air barrier will provide significant performance and economic and environmental benefits to a building. The next section will discuss the environmental benefits of a continuous air barrier.
Life Cycle Assessment of Air Barrier Systems DEFINITIONS: LIFE CYCLE ASSESSMENT, ENVIRONMENTAL PRODUCT DECLARATION, PRODUCT CATEGORY RULES
A life cycle assessment (LCA) evaluates potential environmental impacts of a product throughout its life cycle, from raw materials extraction and processing, transportation, manufacture, installation, and use through disposal, recycling, or reuse of a product. An environmental product declaration (EPD) is a summary of the LCA report. The primary intent of an EPD is to describe environmental attributes in a consistent way so that a direct comparison between Product A and Product B is possible. This consistency is established by the product category rules (PCR). LCAs and EPDs can include the entire life cycle, from “cradle to grave,” or can address only the upstream portion of the life cycle, “cradle to gate.” The product category rules (PCR) are a standardized set of rules describing which characteristics should be disclosed for a particular product type/function. The PCR provides the parameters for conducting the LCA and EPD to increase consistency within a product category. The PCR for air barriers is based on the EN 15804 standard for plastic and elastomer roofing and sealing sheet systems that describes the format and requirements for creating an EPD for self-adhesive and non-self-adhesive plastic and elastomer roofing and sealing sheeting. The PCR also defines the “functional unit,” the quantity of product needed to serve an intended purpose, to allow for direct comparison between products within a given category. LCAs of building products often address only the upstream stages of the life cycle—cradle to gate—omitting the building use phase and end-of-life data if the PCR allows it. The use stage for some building products (such as air barrier systems) can actually dominate the rest of the life cycle: They can stay in place and in use for many years. They can affect building durability and hence avoid replacement, which can generate solid waste. They can affect energy consumption during
MEYER ET AL., DOI: 10.1520/STP161720180105
the building use. In fact, 75% to 85% of the energy associated with the building life cycle is estimated to be consumed during its operational phase.12,13 As a result, the cradle-to-gate LCA approach for an air barrier system does not give a complete picture of its overall environmental impact because its potential contribution to energy savings during the building use phase is not captured. This section will assess the overall environmental benefits of a continuous air barrier with effective air leakage control by comparing the energy cost of the air barrier system throughout its life cycle with the energy saved through building envelope airtightness during the building use phase. LIFE CYCLE ASSESSMENT
A holistic approach to the environmental impact of different energy efficient strategies should include a comparison between the environmental benefits of energy efficient measures and the environmental cost of implementing such measures. This has been a challenge for the construction industry due to the complexity of the task, lack of information, and need for multidisciplinary project teams to perform such complex assessments. To further complicate the matter, there are many materials available for any given building function, but only few materials have completed the life cycle impact assessment (LCIA) that would provide information on the environmental cost of producing, transporting, installing, and disposing of materials at the end of their service life. This section will show an example calculation of the overall environmental impact of a continuous air barrier by comparing the environmental cost of an air barrier system from LCA with the environmental benefits from energy savings due to building envelope airtightness estimated through whole building energy simulations. The analysis results are shown as a payback period for an average U.S. case—the time needed to payback the energy and greenhouse gas impacts associated with manufacturing, installing, and disposing of air barrier systems. Because an LCA is product specific, the current analysis was performed for two air barrier systems for which LCAs were available: a fluid applied (FA) and mechanically fastened building wrap system with a low-rise system and one with a high-performance system. The environmental impact assessment includes the steps outlined in figure 3. This paper does not describe all six steps in detail as they are described elsewhere,14–16 but it summarizes the data used for calculating the payback period. EXAMPLE CALCULATIONS: SMALL HOTEL BUILDING PROTOTYPE
The environmental cost of an air barrier system from an LCA is compared with the environmental benefits from energy savings due to building envelope airtightness estimated through whole building energy simulations. The air barrier environmental cost in this analysis is based on the small hotel building prototype, with characteristics shown in figure 4.14–16
345
346
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 3 Steps for calculating energy and carbon footprint for a commercial prototype building.
FIG. 4 Small hotel building characteristics.
Energy costs associated with the air barrier systems for this building prototype were calculated from an LCA performed by thinkstep, Inc., and independently verified in accordance with ISO 14025, Environmental Labels and Declarations—Type III Environmental Declarations—Principles and Procedures, ISO 14044, Environmental Management—Life Cycle Assessment—Requirements and Guidelines, and the reference PCR by Underwriters Laboratories. The LCAs were performed for a fluid applied and two building wrap systems (high-performance and low-rise
MEYER ET AL., DOI: 10.1520/STP161720180105
systems—based on design criteria and performance requirements). The LCA was based on the PCR for plastic and elastomer roofing and sealing sheet systems published by the German Institute of Construction and Environment. LCA calculations were performed for the functional unit defined for this PCR category, which is 1 m2 of air barrier system. Based on the representative building characteristics in figure 4, 1,640 m2 of weatherization system are required to seal the building envelope of the prototype building. The energy cost associated with 1 m2 of air barrier (functional unit) was obtained from the LCA and then calculated for the representative building. The average energy savings were then calculated for two airtight building standards (using whole building energy simulations) and compared with the energy cost from the LCA to estimate an average payback period. A similar analysis was conducted to calculate the CO2e (CO2 emissions) payback period for air barrier systems. The CO2 emission reductions due to energy savings are compared with the carbon footprint for the representative building from the LCA to calculate a payback period. Energy and CO2 payback periods are summarized in table 1. The payback period calculated by comparing energy saved with energy cost in the referenced study14–16 indicates that, from a primary energy standpoint, the energy cost to produce, transport, and dispose of the air barrier systems at the end of their service life can be recovered within the first year of building use (1.8 to 11.2 months). From a carbon standpoint, carbon emissions generated to produce, transport, and dispose of the three air barrier systems at the end of their service life can also be recovered within the first year (1.5 to 10.3 months). This basically means that, after the initial two to twelve months, the energy savings and carbon footprint reductions due to building envelope airtightness over the building’s life represent energy and carbon credits and that the air barrier systems have a significant positive impact on the building’s environmental footprint. The environmental payback period is ten to twenty times less than the economic payback period (12–20 years) required by ASHRAE 90.1 scalar of 12 to 20 years, proving that a continuous air
TABLE 1 Payback periods for air barrier systems for the representative building (small hotel)
Payback Periods (Months) Low-Rise Mechanically Fastened System Air tightness standard @ 75 Pa
Energy
CO2
High-Performance Mechanically Fastened System
Energy
CO2
Fluid Applied System
Energy
CO2
1.0 cfm/sf
--
--
--
--
--
--
0.4 cfm/sf
1.8
1.5
3.4
10.3
11.2
10.3
0.25 cfm/sf
--
--
1.8
5.5
6
5.5
347
348
STP 1617 On Building Science and the Physics of Building Enclosure Performance
barrier system is a cost-effective and environmentally beneficial solution for improved building envelope performance. EXAMPLE CALCULATIONS: DETAILS FOR PERFORMANCE
The overall environmental impact of a continuous air barrier was estimated by comparing the environmental benefits with the environmental cost of the materials and applications involved. To achieve whole building airtightness, the details matter, and this includes the performance attribute beyond just the main air barrier membrane material. In order to achieve the whole building airtightness savings, the necessary flashing and accessories need to be included in the LCA calculation to represent the anticipated performance of the system. Details that should be captured for the whole building use phase LCA model for airtightness, shown in figure 5, include the roof-to-wall interface, window detailing, sealing penetrations and cladding fasteners, door flashing, and terminations at the base of the wall system. For the same base building (U.S. Department of Energy [DOE] small hotel), when the flashing details and accessory air barrier materials are included, the FA and mechanically fastened building wrap systems with both a low-rise system and high-performance system option have very different overall system composition (table 2). When the building is modeled with the construction details included, the mechanically fastened low-rise system has roughly twice as much air barrier membrane as flashing by overall mass (61% water-resistive barrier [WRB] versus 31% flashings); the mechanically fastened high-performance system has an almost even split in mass between the WRB and the flashings (45% WRB versus 45% flashings); and the fluid-applied system is primarily composed of the WRB membrane relative to the flashing material (84% WRB versus 14% flashings). The increase in flashing mass percentage as the performance level of the air barrier increases can be
FIG. 5 Details modeled for representative building (small hotel DOE prototype).
MEYER ET AL., DOI: 10.1520/STP161720180105
TABLE 2 Air and water barrier system component products and materials
Low-Rise Mechanically Fastened System Mass %
High-Performance Mechanically Fastened System Mass %
High-Performance Fluid Applied System Mass %
WRB Material
61%
45%
84%
Flashings
31%
45%
14%
Accessories
8%
10%
2%
100%
100%
100%
Total
Source: http://www.dupont.com/products-and-services/construction-materials/building-envelopesystems/articles/sustainable-buildings.html
explained by the more extensive detailing at each interface condition. The increased WRB percentage of the fluid applied system relative to the mechanically fastened high-performance system comes from the increased mass per comparable area of WRB as well as the relatively higher density of the silyl terminated polyether (STPE) fluid-applied material evaluated, compared to the high-density polyethylene (HDPE) of the wrap materials. To better illustrate the range of air barrier system composition proportions for the same DOE small hotel base building example, it is helpful to consider the most common window opening shown in the model. A 3.5-ft (1.1 m)-wide and 5.0 (1.5 m)-ft-tall flanged window is the most common window type in the model building, with 77 units included in the calculation (fig. 6). For this standard window opening, mechanically fastened low-rise systems utilize 0.5 kg of flashing products to install the window unit, while a high-performance
FIG. 6 Flanged window installation material comparison.
349
350
STP 1617 On Building Science and the Physics of Building Enclosure Performance
installation can require 1.0 kg of flashing materials combined with the mechanically fastened system and 1.6 kg with a fluid-applied flashing installation of the window unit respectively. When considering air barrier, the designer should choose the correct installed performance expectations to meet the building’s needs. An LCA or EPD modeled air barrier system (or both) is able to capture the energy and CO2 reduction benefits when the use phase is included in the assessment and results. When including the use phase of the air barrier system, it is critical to model more than just the primary air barrier membrane material. The flashings and accessories can result in more than half of the overall contribution in mass to the installed system. Including the proper details for interfaces such as windows, doors, penetrations, and top/bottom of wall terminations is critical to actually achieving the use-phase modeled performance.
Air Barrier Design, Installation, and Verification As shown in the LCA, an air barrier is a cost-effective and environmentally sensitive solution to promote energy efficiency in buildings. However, if the air barrier is not designed and installed appropriately, its impact on energy efficiency can be significantly reduced because that air barrier may no longer be able to resist airflows across the building enclosure. AIR BARRIER DESIGN
In order to accomplish a successful air barrier system, the designer must note on drawings and in specifications which materials or combination of materials are intended to function as the air barrier. It is typically beneficial for an air barrier to be located on the exterior side of the backup wall assembly so that it can be more easily made continuous. Additionally, the designer must consider where the enclosure boundaries are located. For example, unconditioned or semiconditioned spaces or spaces that are difficult to control their condition (i.e., a vestibule or loading dock) typically require air separation from adjacent conditioned space. Not addressing these kinds of separations oftentimes results in large leaks that significantly affect performance. Likewise, spaces that are conditioned differently (i.e., a natatorium within a larger facility) may require separation from adjacent spaces. Interfaces between the air barrier materials and adjacent systems (e.g., roofing, fenestration, waterproofing, penetrations, etc.) also influence the success of the air barrier system. The designer must clearly illustrate the continuity of the air barrier from the center of the roof, through the exterior wall systems, and down to the foundation in all plan and section design details. Enlarged details should be provided at locations where there is a penetration in the air barrier, where the air barrier transitions to a different system, or where the substrate changes. To verify continuity, consider the pen test method, in which a pen should be able to be run continuously along the air barrier profile without interruption (i.e., without picking
MEYER ET AL., DOI: 10.1520/STP161720180105
up off the paper).17 This is a good quality control check during construction detail development when considering the moisture and air control assembly continuity. The ability of the designer to envision three-dimensional interfaces and convey this approach in their design also helps to verify that the air barrier system design is continuous. In addition to verifying continuity with the pen test, enlarged details can help to coordinate across trades and resolve potential issues such as compatibility, sequencing, and other issues. Multiple building interface conditions that require enlarged details to clearly illustrate continuity of the air barrier in building enclosure design are presented in the following sections. These conditions include locations where continuity of the air barrier is most frequently not achieved by designers. Note that the discussion of these details focuses on air barrier materials applied to backup walls. Roof to Wall Details
Air barrier continuity must be maintained across the roof-to-wall interface. This paper limits detailing considerations to low-sloped roofing systems where the air barrier is usually the fluid-applied or self-adhered vapor retarder on top of the structural deck, the structural deck itself, the roofing membrane (if adhered), or a combination thereof. In cold climates with a risk for condensation on the underside of the roofing membrane during the wintertime (due to exfiltration of interior air), roofing air barriers (typically identified as vapor retarders in the roofing industry) are recommended on top of air-permeable decks and sealed penetrations, and joints are recommended (as a minimum) at air-impermeable decks. The wall air barrier can connect directly to the roof air barrier or deck or the roofing membrane can transition between the roof air barrier/deck and the wall air barrier. Other items besides continuity that need to be considered in roof-to-wall design include compatibility between different membranes, sequence of construction, and potential for differential movement between the roof deck and the wall. For example, figure 7 includes balloon metal framing extending past the roof deck. The primary air barrier system includes a fluid-applied air barrier with selfadhered transition membranes (on the exterior wall sheathing), a self-adhered air barrier/vapor retarder (on substrate board on metal deck), and adhered roofing membrane. The continuity is achieved by connecting the roof air barrier/vapor retarder to the roofing membrane and the roofing membrane to the wall air barrier. This detail requires consideration of deflection between the roof and wall due to the balloon framing, with a deflection joint that must be detailed. Incompatibility between the roofing membrane and the wall air barrier is accounted for with a “sheet metal separator.” Also note that the metal studs within the parapet include closed-cell spray polyurethane foam (SPF) to limit the migration of interior air into the parapet, which can lead to condensation within the parapet during the wintertime in cold climates, typically when the parapet is very tall and the thermal plane short circuits the parapet.
351
352
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 7 Parapet detail with balloon framing.
In another example, in figure 8, which includes platform metal framing on top of the roof deck for the parapet, the air barrier materials include a selfadhered roof air barrier/vapor retarder (on substrate board on metal deck) and an air barrier on exterior wall sheathing. The continuity of the air barrier system is maintained at the roof-to-wall interface with the roof air barrier/vapor retarder extending across the deck to the exterior wall prior to installation of the parapet. This requires the parapet to be built out of sequence and trades to return to the area to complete the work. To provide watertightness to the roofing system, the air barrier extends up the parapet and connects to the roofing membrane that wraps the parapet. When aluminum and glass curtain wall, precast, or insulated metal panel parapets act as the wall’s air barrier and form the parapet, the designer must know where the air control line is in those systems and must marry the roofing air barrier to that line. Usually, in those instances, compatibility among sealants, gaskets, and membranes as well as differential movement between the wall and the roof are items to be considered, among others. Air Barrier Membrane to Fenestration Details
Sealants or various types of flexible transition membranes typically create the continuity between the air barrier at the opaque wall and the air/water line of the fenestration system. The continuity of the air barrier is critical at this interface to achieve air- and watertightness. Generally, for fluid-applied or self-adhered air barriers at
MEYER ET AL., DOI: 10.1520/STP161720180105
FIG. 8 Parapet detail with platform framing.
opaque walls, the air barrier is wrapped into the rough opening, the fenestration is put into place, and the joint between the two is sealed (with sealant or membrane). At these transitions, designers frequently do not align the primary air/water line of the fenestration with the air barrier at the opaque wall, which can be an issue if there is not a means to provide a bridge at the change in plane (fig. 9). Other times, designers might not understand where the primary air/water line of the fenestration is and incorrectly locate or omit the primary perimeter sealant joint (fig. 10 and fig. 11). Similarly, when precast concrete cladding or insulated metal panel cladding that act as the wall’s air barrier interface with the fenestration, the designer must know where the air/water control line is in those systems and marry the adjacent air barrier membrane to that line (fig. 12). Compatibility between sealants and gaskets that are part of the air/water control line and the air barrier membrane needs to be considered.
353
354
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 9 Curtain wall primary air/water line outboard of opaque wall air barrier membrane. Primary perimeter sealant joint for curtain wall is bonded to insulation (not air barrier).
Air Barrier Details at Shelf Angles
Sequencing of construction, constructability, and continuity of the air and water barriers are challenges to be addressed at shelf angle details. Air barrier membranes need to be integrated with through-wall flashing to create a continuous air barrier system that also manages water, and penetrations through the flashing or air barrier need to be considered. Shelf angles can be held off the backup with stand-off clips of various kinds to allow for continuous insulation behind the shelf angle, or they can be anchored directly to the backup.
MEYER ET AL., DOI: 10.1520/STP161720180105
FIG. 10 Primary perimeter sealant joint for curtain wall incorrectly located. It should be located in alignment with the primary air/water line of the curtain wall (dashed line).
In figure 13, the shelf angle is hung from the floor slab to align with the window heads and is projected outboard of the backup wall with T-sections that are bolted to steel knife plates to allow continuous insulation behind the shelf angle. Fluidapplied air barrier on the exterior sheathing integrates with a two-piece stainless steel flashing above the hung structural steel, which allows the flashing to be continuous across the masonry cavity. The air barrier also extends into the window rough opening to allow the window perimeter sealant to be married to the air barrier. The overlap between the steel knife plate and the T-section needs to be considered. If the overlap occurs too close to the backup wall, the air barrier will lap onto both the knife plate and the T-section and will need to be detailed around each joint between the knife plate and the T-section, and possibly at each anchor bolt, to prevent a breach in the air barrier. As such, it is recommended that the T-section overlap
355
356
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 11 Air barrier does not wrap into rough opening and is not integrated with storefront perimeter sealant.
FIG. 12 Incomplete air barrier system between insulated precast panel cladding and storefront jamb. Thin red dashed lines represent dual-stage seal at cladding. Heavy blue dashed line represents transition between systems to provide continuity.
MEYER ET AL., DOI: 10.1520/STP161720180105
FIG. 13 Hung shelf angle detail.
with the knife plate be as far outboard as possible so that the air barrier only laps onto the knife plate. Additionally, construction sequence should be considered. In this instance, the fluid-applied air barrier needs to be installed prior to the T-section due to limited access following T-section installation, but the masonry through-wall flashing needs to be positioned with consideration of the future location of the T-section. As such, coordination between trades was critical to the success of this detail. In figure 14, the shelf angle is anchored to the concrete slab that contains an embedded dovetail slot. In this case, the designer must consider how the air barrier will be extended into the steel dovetail and how the deflection joint between the bottom of the slab and the top of the stud wall (which is nearly in line with the bottom of the shelf angle) is detailed. At other shelf angle anchorage locations that may not use a dovetail slot, consideration must also be given to how the bolt penetrations through the air barrier will be sealed. Similarly, in figure 15, where the shelf angle stands off the backup wall to allow the insulation to be continuous behind the shelf angle, the air barrier must be detailed at each hollow structural steel (HSS) standoff. Since the exterior sheathing extends over the exterior face of the slab edge, the stand-off assembly must be
357
358
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 14 Shelf angle detail with embedded dovetail slot connection.
FIG. 15 Shelf angle with HSS stand-off.
installed before the air barrier assembly. As such, detailing at each HSS is a challenge because access behind the assembly is limited. Additionally, the HSS must be continuously welded, or otherwise sealed, to either the outer steel plate or the inner
MEYER ET AL., DOI: 10.1520/STP161720180105
steel plate to prevent each HSS from being a conduit for air/water leakage. Finally, the exterior sheathing and air barrier must have a deflection joint detail at the stud slip-track below the slab edge to accommodate the 0.75-in. (19 mm) anticipated vertical deflection. Movement Joint Details
Deflection joints are oftentimes necessary at edges of slabs or similar locations to accommodate differential movement between the structural system and the exterior wall. The air barrier (and backup wall) must be detailed to accommodate this differential movement. Oftentimes, air barrier deflection joints include transition membranes with a bellows or flexible “boot” flashing that must be integrated with the air barrier material. In the example shown in figure 16, the air barrier includes a rigid
FIG. 16 Deflection joint detail at rigid board air barrier.
359
360
STP 1617 On Building Science and the Physics of Building Enclosure Performance
insulation board. Board joints are sealed with silicone liquid flashing, and deflection joints are detailed with butyl tape. Because butyl will not adhere to silicone, sequencing is a concern, and the deflection joint detail must be installed before board joints are sealed. This could become problematic if trades work in sequence from bottom to top. The deflection joint may align with window heads, in which case the deflection joint must extend into the window head rough opening and be integrated with shelf angles, window perimeter sealant, or cladding trim that might be present. An example of one such condition is shown in figure 17. This detail becomes very complicated to execute at the upper corners of the window heads where the deflection joint, which in this case was detailed with a bellowed transition membrane over the precast slab to exterior sheathing interface, meets the window receptor and window perimeter sealant joint. Air Barrier Details at Soffits or Overhangs
Other details that are oftentimes overlooked in terms of air barrier continuity are exterior soffits or overhangs. At these conditions, it is important to understand if the soffit is considered exterior or interior conditioned space. For the former, the insulation and air barrier should be included at the underside of the slab and integrated with the wall air barrier, and the soffit/overhang should be ventilated. For the latter, the insulation and air barrier should follow the plane of the suspended soffit/overhang cladding. In both cases, how the air barrier is integrated with the wall air barrier system above and below is a critical detail to ensure air barrier continuity. Figure 18 is an example where SPF was provided at the underside of the slab above and the soffit was ventilated. The SPF did not extend across the underside of the slab to the curtain wall on the exterior side of the steel spandrel beam. As such, vertical curtain wall tubes that extend into the soffit present conduits for air leakage.
FIG. 17 Deflection joint that aligns with window head.
MEYER ET AL., DOI: 10.1520/STP161720180105
FIG. 18 Soffit detail with incomplete air barrier between underside of floor slab and curtain wall.
Additionally, the floor slab, the edge of slab fire safing, and the curtain wall spandrel backpan are relied upon to create the air barrier continuity on the exterior side of the spandrel beam, which likely was not the intent of the design. In figure 19, it is not clear if the intent was for the overhang to be vented or conditioned, but because the insulation and air barrier were not included at the
361
362
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 19 Detail at overhang with incomplete air barrier between storefront head and underside of roof deck.
suspended soffit cladding, one could assume the soffit would be ventilated/unconditioned space. As such, it is critical that the air barrier extend from the window head rough opening to the metal deck above. Detailing of the air barrier at the fluted metal deck, particularly when the flutes run perpendicular to the exterior wall, is important to ensure continuity. One example of a “conditioned” overhang is shown in figure 20. In this case, the air barrier and the insulation wrap around the overhang. However, continuity between the exterior sheathing on the underside of the overhang and the storefront head below is not achieved. Additionally, one might anticipate that the thermal plane will follow the path of least resistance, which in this case, is a vertical plane that extends from the storefront head to the insulated precast concrete above, thereby short-circuiting the overhang. Since the overhang will not be actively conditioned with forced air and since air permeable insulation is indicated in the plane of the exterior wall, condensation could be a concern in a cold climate, if interior air can reach cold surfaces. As such, it is important to understand where to best locate the air barrier to minimize risk for performance-related problems. Louver and Plenum Details
Louver and plenum detailing requires coordination with the mechanical design and an understanding that louvers are not watertight. Water management needs to be
MEYER ET AL., DOI: 10.1520/STP161720180105
FIG. 20 “Conditioned” overhang with incomplete air barrier between storefront head and soffit (circle). Dashed line indicates possible better location for air barrier and insulation plane in overhang.
incorporated inboard of louvers. Louvers contain sill pan flashing (for water management) and insulated blank off panels that must be part of the air barrier system and integrated with the insulated mechanical ductwork inboard of the louvers. Louver anchorage at the sill pan needs to be detailed as air- and watertight. In the example detail shown in figure 21, a sheet metal sleeve inserted into the rough opening is integrated with the air barrier membrane surrounding the rough opening and with the mechanical ductwork. In this example, blank off panels are included but are not part of the air barrier system. The duct sleeve is intended to manage water penetration inboard of the louver and be airtight as part of the air barrier system. Plenums, which contain drains and roofing membrane or, similar, manage water inboard of the louvers, must contain a continuous air barrier system surrounding the plenum (i.e., side walls, interior wall, floor, and ceiling of the plenum space). Structural columns are typically within the plenum space, and these elements also need to be detailed to show how the air barrier is integrated with them (fig. 22 and fig. 23). In summary, providing enlarged details to clarify design intent in regard to air barrier continuity, compatibility, and constructability at all air barrier transitions, terminations, and penetrations is critical for air barrier performance and durability. Designers must pay particular attention to air barrier membrane transitions to adjacent systems (precast cladding, insulated metal panel cladding, fenestration, roofing, and waterproofing). The examples provided here are among the most common locations where air barrier discontinuity is encountered.
363
364
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 21 Louver detail with duct sleeve tie-in between air barrier and duct.
Specifications
Just as important as enlarged details that show air barrier continuity are the specifications that outline air barrier requirements. Specifications must clearly identify which material or combination of materials comprise the air barrier system. An air barrier assembly (i.e., air barrier membrane, transition membrane, sealants,
MEYER ET AL., DOI: 10.1520/STP161720180105
FIG. 22 Section detail of plenum with heavy red line showing location of air barrier assembly.
through-wall flashings, etc.) should be a tested/proved system with a total system warranty. Specifications must also provide allowable leakage rates for the air barrier membrane and assembly as well as the adjacent system, such as fenestration and doors (included in Construction Specification Institute MasterFormat Division 08 Openings), and they must provide other performance requirements such as vapor permeance and water penetration resistance. Other important requirements that
365
366
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 23 Plan detail of plenum with heavy red line showing location of air barrier assembly. Note that additional detail should be provided at column within plenum to show continuity of air barrier assembly.
must be specified include submittals (including project-specific shop drawing details and compatibility letters), storage and environmental limitations, warranties, surface preparation, installation, inspections and testing, and protection. AIR BARRIER INSTALLATION
Because of the effect that the air barrier has on building performance, the air barrier must be installed continuously as specified in the contract documents and must be installed with a high level of workmanship. To help ensure this, project-specific shop drawing details should be provided to help resolve coordination issues and verify constructability and continuity prior to installation. Additionally, preinstallation meetings, where all enclosure trades, the architect, and the contractor participate, are an important means to discuss the significance of the air barrier with all trades that interface with it or come into contact with it. Items that are discussed during the meetings typically include interface detail coordination; sequencing and scheduling; manufacturer’s and specification requirements for preparation, installation, and protection; and testing and verification. Tested mockups may also be required as part of the project to resolve unanticipated field conditions prior to being installed on the building; identify sequencing, constructability, or compatibility issues; and provide a basis for acceptability of remaining construction.
MEYER ET AL., DOI: 10.1520/STP161720180105
During installation, it is important for all enclosure and mechanical, electrical, and plumbing (MEP) trades to recognize the importance of the air barrier, coordinate with each other, and make sure any deficiencies are addressed prior to covering them up. For the installing air barrier subcontractor, specifications oftentimes require that the installer be accredited by manufacturers or third-party programs such as the ABAA Quality Assurance Program or have a minimum number of projects of similar scope and complexity with evidence of training of field personnel (or both). Even with complete and accurate contract documents that clearly identify the air barrier and even with the most qualified contractors, for various reasons, continuity of the air barrier is oftentimes not achieved in the field. Several examples are shown in figures 24 through 27. Other issues that are commonly identified during air barrier installation include the following, depending on the type of air barrier being installed: • Improper surface preparation, including irregular mortar joints in masonry backup (fig. 28), unfilled bugholes in concrete backup, unfilled sheathing board joints and board fastener penetrations (fig. 29), discontinuous substrate, and wet or dirty substrate (fig. 30) • Improper environmental conditions for storage and application, including surface and ambient temperatures above or below that required and forecasted precipitation within the specified time frame
FIG. 24 Incomplete air barrier at structural penetration.
367
368
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 25 Incompatibility between air barrier and silicone sealant splice joint.
FIG. 26 Incomplete air barrier seal at flashing boot.
MEYER ET AL., DOI: 10.1520/STP161720180105
FIG. 27 Discontinuous air barrier at the end of a shelf angle.
FIG. 28 Improperly tooled mortar joints for air barrier application.
369
370
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 29 Sealed sheathing board joints for application of fluid-applied air barrier—but note unsealed countersunk fasteners.
FIG. 30 Dirt on sheathing board prior to application of air barrier.
MEYER ET AL., DOI: 10.1520/STP161720180105
Improper application, including inadequate thickness (fig. 31), reverseshingle laps (fig. 32), incorrect transition and penetration detailing (fig. 32), fishmouths (fig. 33), blisters, inadequate adhesion (fig. 34 and fig. 35), incorrect board joint detailing leading to separations in the air barrier, and incompatible materials (fig. 36 and fig. 37) • Lack of coordination with trades, including utility penetrations (fig. 38) and cladding anchorage (fig. 39) • Lack of protection of in-place air barrier, including lack of protection at end of day’s work, damage from trades (fig. 40 and fig. 41), trapped moisture behind the completed air barrier (fig. 42 and fig. 43), and overexposure to UV radiation (fig. 44) If these items go unchecked, the performance and long-term durability of the air barrier could potentially be compromised. •
AIR BARRIER QUALITY ASSURANCE AND VERIFICATION
Due to the importance of the air barrier’s performance, oftentimes, third-party consultants or building enclosure commissioning providers (BECxP) are retained by owners, architects, or contractors to assist in evaluating and verifying that the design and installation of the air barrier will meet airtightness requirements for the project. Tasks that may be performed by the consultant or BECxP, as related to the air barrier system, include peer reviews of the design documents, reviews of the submittals, participation in preinstallation and preconstruction meetings, observation
FIG. 31 Inadequate thickness of fluid-applied air barrier.
371
372
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 32 Reverse-shingled lap joint with missing sealant at T-joint.
FIG. 33 Fishmouth at air barrier lap joint.
MEYER ET AL., DOI: 10.1520/STP161720180105
FIG. 34 Inadequate adhesion between rigid board air barrier and board joint seals.
FIG. 35 Inadequate adhesion between self-adhered membrane and required membrane lap seals.
373
374
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 36 Butyl tape installed over silicone sealant will not achieve required adhesion.
FIG. 37 Self-adhered flashing installed over silicone transition boot will not adhere.
MEYER ET AL., DOI: 10.1520/STP161720180105
FIG. 38 Plumbing penetration installed after air barrier, not detailed.
FIG. 39 Cladding anchors missed stud backup, creating holes in finished air barrier.
375
376
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 40 Damage to air barrier.
FIG. 41 Damage to air barrier due to welding.
MEYER ET AL., DOI: 10.1520/STP161720180105
FIG. 42 Moisture penetration within CMU backup can become trapped behind air barrier as it migrates downward.
FIG. 43 Trapped moisture behind air barrier forming icicle at air barrier membrane lap and penetration.
377
378
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 44 Cracked air barrier due to overexposure to UV radiation.
of mockup construction and testing, periodic observation of installation throughout construction, and field performance testing throughout construction or at project completion (or any combination thereof). Such quality assurance (QA) tasks are used to verify conformance with code requirements or the contract documents (or both) and help to recognize and resolve issues with the air barrier systems prior to covering them up with finish materials, after which repair to the air barrier would potentially be costly to achieve and disruptive to occupants. QA processes also help to comply with the installation verification requirements for continuous air barriers in ASRHAE 90.1-2013. ASHRAE 90.1-2013 requires that the air barrier design and installation be verified with (1) a design review and periodic field inspection or (2) a whole-building air leakage test. Likewise, both ASHRAE 189.1, Standard for the Design of High-Performance Green Buildings Except Low-Rise Residential Buildings, and the IgCC have provisions for air barrier installation inspections and whole-building air barrier performance testing for those projects seeking “green” certifications. Green Globes has an extensive set of credits for field conformance or performance testing (or both) of many building enclosure assemblies, including below-grade waterproofing, roofing, flashing, fenestration, and air barrier systems. It is also notable that Leadership in Energy and Environmental Design (LEED) v4 has building enclosure commissioning credits available, in which verification of the design and
MEYER ET AL., DOI: 10.1520/STP161720180105
installation of the air barrier system would be required for a building with airtightness performance requirements. The verification of the air barrier installation is important for overall airtightness performance for the building. The inspector must review the air barrier for continuity, workmanship, and installation according to the contract documents and manufacturer’s installation instructions. Areas of particular focus are noted in the previous design and installation sections of this chapter. Additionally, the third party who performs the inspections/testing can help the project team to resolve unanticipated field conditions that may otherwise result in an incomplete air barrier system. In addition to field observations, there are several types of tests that can be performed in the field to verify that air barrier assemblies meet specified or coderequired performance requirements. Quantitative airtightness testing helps to verify that the measured air leakage rate through the air barrier assemblies are at or below maximum airtightness requirements at a certain pressure differential across the enclosure. Quantitative tests include the following: • ASTM E783-02(2018), Standard Test Method for Field Measurement of Air Leakage Through Installed Exterior Windows and Doors (fig. 45 and fig. 46)18 • ASTM E779-19 • ASTM E1827-11(2017) (fig. 47) • ASTM E3158-18 • ABAA-2016
FIG. 45 ASTM E783 at opaque wall self-adhered air barrier.
379
380
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 46 ASTM E783 testing at fenestration.
FIG. 47 ASTM E1827 blower door testing.
MEYER ET AL., DOI: 10.1520/STP161720180105
Qualitative airtightness tests to identify air leakage sites in a building envelope in the field include ASTM E1186-17, Standard Practices for Air Leakage Site Detection in Building Envelopes and Air Barrier Systems (fig. 48 and fig. 49).19 Other tests that may be performed to verify adhesion include ASTM D4541-17, Standard Test Method for Pull-Off Strength of Coatings Using Portable Adhesion Testers (fig. 50),20 and ASTM C1193-16, Standard Guide for Use of Joint Sealants, Method A, FieldApplied Sealant Joint Hand Pull Tab (fig. 51).21 In summary, a quality assurance program for an air barrier assembly is an important part of the project to help ensure that airtightness is achieved. The thirdparty agent that is typically part of the QA program helps to identify and resolve issues such as those previously identified in the design and installation sections of this chapter. Identifying issues during the design or installation of the air barrier can help to minimize costs and disruption because repairs/resolutions can be difficult to implement once the building is complete and occupied.
FIG. 48 ASTM E1186 testing with smoke.
381
382
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 49 ASTM E1186 “bubble gun” testing.
FIG. 50 ASTM D4541 adhesion testing.
MEYER ET AL., DOI: 10.1520/STP161720180105
FIG. 51 ASTM C1193 adhesion testing.
Conclusions The overall environmental impact of a continuous air barrier was estimated by comparing the environmental benefits from energy savings and CO2 emissions reduction during the building use phase with the environmental cost of the materials involved (product energy cost and carbon footprint from an LCA). For the air barrier systems in this analysis, the payback period for energy and carbon emissions ranges from one month to a year, which means that, after the first year, the air barrier systems will have a net positive environmental impact: energy and CO2 emissions credits. While some assumptions and approximations had to be made in order to estimate energy savings due to building envelope airtightness through energy simulations, there is no doubt that air barriers are very effective measures for improved building envelope performance. The current analysis was performed for fluidapplied and building wrap (mechanically fastened) air barrier systems for which an LCA has been completed. The embodied energy of an air barrier material will determine the payback period and will enable the choice of air barrier materials with the smallest environmental footprint. When including the use phase of the air barrier system in an LCA, it is critical to model more than just the primary air barrier membrane material. The flashings and accessories can result in more than half of the overall contribution in mass to the installed system. Including the proper details for interfaces such as windows, doors, penetrations, and top/bottom of wall terminations are critical to actually achieving the use phase modeled performance.
383
384
STP 1617 On Building Science and the Physics of Building Enclosure Performance
In addition to energy savings and CO2 emissions reductions that were quantified in this analysis through whole building energy simulations, air and water barriers provide substantial additional benefits by preventing risks associated with water intrusion and air infiltration. Potential risks associated with water intrusion can range from health risks (due to mold and dampness that could affect the indoor environmental quality), to building envelope durability/material degradation (such as rotting of wood or corrosion of metals), and material performance (such as loss of insulation R-value). Potential risks associated with air infiltration include impact on the occupant’s comfort (e.g., cold drafts) and transport of outdoor contaminants into the conditioned space, affecting the indoor air quality. A continuous air barrier system is a cost-effective measure for avoiding risks associated with water intrusion and air infiltration. While these additional air barrier benefits are more difficult to quantify because incidental moisture intrusion and air infiltration events are unpredictable, the impact on building envelope durability/longer building life (which will obviously impact material resources and the building’s environmental footprint), as well as comfort and health of the building occupants, could be significant. Complying with the energy code continuous air barrier requirements has many facets. Successful project teams should determine the air barrier layer’s ability to meet the critical material attributes of performance when installed, as well as identifying any additional hygrothermal functions to be performed by the air barrier layer. In addition, project teams should establish a project’s building enclosure performance expectations, determine which version of the energy code the project must meet, and understand if there are air barrier performance project goals that exceed or meet the minimum described in the applicable standards. The full benefits of air barriers are only achieved if the air barrier is properly installed and is durable over the life of the building.
References 1.
2.
3.
4.
5.
Standard Test Method for Air Permeance of Building Materials, ASTM E2178-13 (West Conshohocken, PA: ASTM International, approved February 1, 2013), http://doi.org/ 10.1520/E2178-13 Standard Test Method for Determining Air Leakage Rate of Air Barrier Assemblies, ASTM E2357-18 (West Conshohocken, PA: ASTM International, approved October 1, 2018), http://doi.org/10.1520/E2357-18 Standard Specification for Air Barrier (AB) Material or Assemblies for Low-Rise Framed Building Walls, ASTM E1677-19 (West Conshohocken, PA: ASTM International, approved July 1, 2019), http://doi.org/10.1520/E1677-19 Standard Test Method for Determining Air Leakage Rate by Fan Pressurization, ASTM E779-19 (West Conshohocken, PA: ASTM International, approved January 1, 2019), http://doi.org/10.1520/E0779-19 Standard Test Methods for Determining Airtightness of Buildings Using an Orifice Blower Door, ASTM E1827-11(2017) (West Conshohocken, PA: ASTM International, approved September 1, 2017), http://doi.org/10.1520/E1827-11R17
MEYER ET AL., DOI: 10.1520/STP161720180105
6.
7.
8.
9.
10.
11. 12. 13. 14.
15.
16.
17.
18.
19.
20.
21.
Standard Test Method for Measuring the Air Leakage Rate of a Large or Multizone Building, ASTM E3158-18 (West Conshohocken, PA: ASTM International, approved December 1, 2018), http://doi.org/10.1520/E3158-18 U.S. Department of Energy, Building Energy Codes Program, “ANSI/ASHRAE/IES Standard 90.1-2016,” http://web.archive.org/web/20180227061322/https://www.energycodes. gov/development/determinations Standard Test Method for Rate of Air Leakage through Exterior Metal Roof Panel Systems, ASTM E1680-16 (West Conshohocken, PA: ASTM International, approved March 1, 2016), http://doi.org/10.1520/E1680-16 Standard Test Method for Determining Rate of Air Leakage through Exterior Windows, Skylights, Curtain Walls, and Doors under Specified Pressure Differences across the Specimen, ASTM E283/E283M-19 (West Conshohocken, PA: ASTM International, approved August 1, 2019), http://doi.org/10.1520/E0283_E0283M-19 USACE/NAVFAC/AFCEC/NASA, “Unified Facilities Guide Specifications,” UFGS-07 08 27.00 10, July 2018, http://web.archive.org/web/20181031195129/http://www.wbdg.org /FFC/DOD/UFGS/UFGS%2007%2005%2023.pdf PBS-P100, “Facilities Standards for the Public Buildings Service,” April 2017, http:// web.archive.org/web/20170711114645/http://wbdg.org/FFC/GSA/p100_2017.pdf K. Adalberth, “Energy Use during the Life Cycle of buildings: A Method,” Building and Environment 32, no. 4 (July 1997): 317–320. K. Adalberth, “Energy Use during the Life Cycle of Single-Unit Dwellings: Examples,” Building and Environment 32, no. 4 (July 1997): 321–329. B. Meyer, “Environmental Benefits of Continuous Air Barriers: Building Use Phase” (paper presentation, Symposium on Whole Building Air Leakage: Testing and Building Performance Impacts, Sponsored by ASTM Committee E06 on Performance of Buildings, San Diego, CA, April 8–9, 2018). UL Environment, “TyvekV Mechanically Fastened Air and Water Barrier Systems,” June 21, 2017, http://web.archive.org/web/20180411173913/http://www.dupont.com/content/ dam/dupont/products-and-services/construction-materials/building-envelope-systems /documents/102-1_DuPont_EPD_Tyvek_Mechanically-Fastened.pdf UL Environment, “TyvekV Fluid Applied Weather Barrier Systems,” June 21, 2017, http:// web.archive.org/web/20180411173909/http://www.dupont.com/content/dam/dupont/ products-and-services/construction-materials/building-envelope-systems/documents/ 101-1_DuPont_EPD_Tyvek_Fluid-Applied.pdf U.S. Environmental Protection Agency, “Moisture Control Guidance for Building Design, Construction and Maintenance,” EPA 402-F-13053, December 2013, http://web.archive. org/web/20181031195818/https://www.epa.gov/sites/production/files/2014-08/documents /moisture-control.pdf Standard Test Method for Field Measurement of Air Leakage Through Installed Exterior Windows and Doors, ASTM E783-02(2018) (West Conshohocken, PA: ASTM International, approved October 1, 2018), http://doi.org/10.1520/E0783–02R18 Standard Practices for Air Leakage Site Detection in Building Envelopes and Air Barrier Systems, ASTM E1186-17 (West Conshohocken, PA: ASTM International, approved July 15, 2017), http://doi.org/10.1520/E1186-17 Standard Test Method for Pull-Off Strength of Coatings Using Portable Adhesion Testers, ASTM D4541-17 (West Conshohocken, PA: ASTM International, approved August 1, 2017), http://doi.org/10.1520/D4541-17 Standard Guide for Use of Joint Sealants, ASTM C1193-16 (West Conshohocken, PA: ASTM International, approved January 1, 2016), http://doi.org/10.1520/C1193-16 R
R
385
386
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180063
Mehdi Ghobadi,1 Josip Cingel,1 and Michael A. Lacasse1
Thermal Bridging and Linear Thermal Transmittance Calculations for Balconies Citation M. Ghobadi, J. Cingel, and M. A. Lacasse, “Thermal Bridging and Linear Thermal Transmittance Calculations for Balconies,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 386–395. http://doi.org/10.1520/STP1617201800632
ABSTRACT
The thermal performance of building envelopes can be significantly affected by thermal bridging. Thermal bridges are localized areas of high heat flow through walls, roofs, and other insulated building envelope components. Thermal bridging is caused by highly conductive elements that penetrate the thermal insulation or misaligned planes of thermal insulation (or both) within the building envelope. These paths allow heat flow to bypass the insulating layer and thereby reduce the effectiveness of the insulation. The architectural look of exposed slab edges and protruding balconies or eyebrow elements in contemporary buildings is becoming increasingly common; however, the impact of floor slab edges and balconies on the thermal performance of the building is not well regulated. In this study, the impact of adding a balcony to a wood frame wall assembly was studied experimentally. Two wood frame wall assemblies were tested in the National Research Council’s guarded hot box test facility in Ottawa, ON, in accordance with ASTM C1363-19, Standard Test Method for Thermal Performance of Building Materials and Envelope Assemblies by Means of a Hot Box Apparatus. The first test included a wall assembly without a balcony. The second test consisted of a wall assembly with a balcony. The linear transmittance value was calculated for the balcony. COMSOL Multiphysics
Manuscript received September 28, 2018; accepted for publication April 12, 2019. 1 Fac¸ade Systems and Products, National Research Council Construction, 1200 Montreal Rd., Ottawa, ON, K1K 2E1, Canada 2 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22, 2018 and December 2, 2018 in Washington, DC, USA. This work is not subject to copyright law. ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959.
GHOBADI ET AL., DOI: 10.1520/STP161720180063
software was also employed to model the wall assemblies, and the results from the three-dimensional simulations are included. Keywords balcony, building envelope, guarded hot box, thermal bridging
Introduction The Pan-Canadian framework1 on clean growth and climate change was adopted in December 2016 by the Canadian government. The Pan-Canadian framework is a collective plan to grow the economy while reducing emissions and deploying measures to modify the National Building Code of Canada that will help ensure building resilience when adapting to a changing climate. To support the ambitious 2030 target of reducing greenhouse gas (GHG) emissions from federal operations by at least 40% below 2005 levels, the 2017 government budget proposed resources for Natural Resources Canada to provide expertise to other federal departments on best approaches to implement energy efficiency and clean energy technologies. Of Canada’s total GHG emissions, 17% comes from homes and buildings whereas 12% are from direct emissions (e.g., combustion of natural gas for heating) and an additional 5% are emissions associated with electricity generation that is consumed in the built environment.1 In this context, moving toward high-performance buildings that integrate and optimize all major high-performance building attributes, including energy efficiency, durability, life cycle performance, and occupant productivity, is essential to meet the national goal. Building envelopes are the main components of buildings that can be enhanced to increase the thermal performance of buildings. It is generally recognized that the thermal performance of building envelopes can be significantly affected by thermal bridging. Thermal bridges are localized areas of high heat flow through walls, roofs, and other insulated building envelope components. Thermal bridging is caused by highly conductive elements that penetrate the thermal insulation and may also be caused by misaligned planes of thermal insulation within the building envelope. These paths allow the movement of heat to bypass the insulating layer, thereby reducing the effectiveness of the insulation in providing resistance to heat loss to the building exterior. While the industry’s understanding of thermal bridging has improved in recent years, current North American building codes and energy standards—including the National Building Code of Canada, National Energy Code for Buildings (NECB) in Canada, and the International Energy Conservation Code (IECC) in the United States are referencing and following a procedure similar to ASHRAE Standard 90.1, Energy Standard for Buildings Except Low-Rise Residential Buildings, and have no specific prescriptive requirements for thermally broken slabs. Modeling slab edges is recommended, but no exact procedure is described. ASHRAE 90.1 assesses the thermal performance of a building for compliance in three pathways: a prescriptive path, a building enclosure trade-off path, and an energy modeling path.2 The latest Canadian Building Code (NBC)3 also requires
387
388
STP 1617 On Building Science and the Physics of Building Enclosure Performance
higher energy efficiency in all building types. Although builders often choose the trade-off or energy efficiency paths and try to achieve the energy requirements by using high-efficiency mechanical equipment, all of these compliance paths still rely on the effective R-values of the building enclosure assemblies. These codes and standards do not explicitly address how thermal bridges at interfaces between assemblies, such as exposed slab edges and balconies, should be addressed in thermal transmittance calculations (U-values) that are necessary when assessing code compliance. This might be due to different reasons such as the belief that slab edges occupy only a small area of the total enclosure and do not have a significant impact on the overall thermal performance, the lack of comprehensive thermal transmittance data for standard details such as balcony slab edges, and the fact that it would take too much effort to quantify slab edges, which require threedimensional thermal simulations. Structural penetrations, such as balconies, with a cross-sectional area less than 2% of the wall area do not need to be taken into account in the calculation of the effective thermal resistance of the wall assemblies. The European standard, EN ISO 10211, Thermal Bridges in Building Construction—Heat Flows and Surface Temperatures—Detailed Calculations,4 describes a numerical simulation methodology to calculate the heat flow and surface temperatures through a thermal bridge, which requires using a numerical simulation package in most cases. Several countries, however, are adopting catalogues and thermal bridging tables such as ISO 14683, Thermal Bridges in Building Construction—Linear Thermal Transmittance Simplified Methods and Default Values,5 to avoid the time-consuming and complex numerical simulations. The problem with these catalogues is they have fixed values or fixed parameters. To improve the flexibility of the simplified methods, Larbi6 developed regression correlations for three typical two-dimensional thermal bridge configurations based on the numerical simulations. The linear transmittance was calculated as a function of design parameters such as dimensions, thermal conductivity of the materials, and so on. Roels et al.7 used a parametric approach to avoid calculating the number or length of thermal bridges required by the simplified method. Clear field assembly thermal transmittance, linear hat transmittance coefficients, and point transmittance coefficient concepts are being used in this study. Slab edge balconies can be categorized as linear thermal bridges categories and a linear heat transmittance coefficient can be employed for this study. Linear thermal bridges typically occur at interface details and, since they are not distributed uniformly, they should be taken into account individually. ASHRAE Standard 90.1 indicates that, “Any other envelope assembly that covers less than 5% of the total area of that assembly type (e.g., exterior walls) need not be separately described.”2 In this project, the impact of adding a balcony slab, occupying only 2% of the total area of the wall, to a wood frame wall assembly was studied. The edge of the slab was replicated using a set of gypsum board panels. The reason gypsum boards were used instead of concrete was that they had the closest steady-state thermal properties to concrete and were easier to use in the lab.
GHOBADI ET AL., DOI: 10.1520/STP161720180063
A wood frame wall was manufactured and an opening within it was left for insertion of the slab’s proxy. The opening was first filled with expanded polystyrene (EPS) panels and, thereafter, a set of gypsum board panels were inserted representative of a balcony slab edge. Laboratory thermal tests were conducted on these wall assemblies in NRC’s guarded hot box (GHB) test facility following the test procedures given in ASTM C1363-19, Standard Test Method for Thermal Performance of Building Materials and Envelope Assemblies by Means of a Hot Box Apparatus.8 The impact of adding the balcony slab on the total thermal performance of the wall assembly was investigated by using the linear thermal transmittance concept. The heat transfer in the walls was also simulated in COMSOL Multiphysics software. Three-dimensional simulations were chosen to capture the lateral heat transfer in the wall assemblies that occur due to the thermal bridging elements.
Experiment Setup The NRC GHB test facility, for which a schematic is given in figure 1, is a test apparatus specifically designed to determine the thermal resistance of building envelope assemblies and components by subjecting a test specimen to a temperature difference and measuring the amount of energy required to maintain interior set point conditions (i.e., the amount of heat the test specimen consumes to maintain the imposed temperature difference is measured and the thermal resistance is determined on the basis of the rate of heat transfer across the specimen and the unit area of the test specimen). To determine the overall thermal resistance of the test specimen, measurements were taken of the interior (room-side) and exterior (weather-side) air temperature at the surface of the test specimen, as well as the heat input to the calorimeter, thereby allowing the air temperature within the guard to be maintained to that of
FIG. 1 Schematic of GHB setup and primary facility elements showing: room-side (interior) and weather-side (exterior) chambers, wall test specimen (sample), adiabatic mask, calorimeter, and guard (left); the balcony setup in the NRC GHB (right).
389
390
STP 1617 On Building Science and the Physics of Building Enclosure Performance
the room-side (interior) temperature. The thermal resistance can then be calculated using equation (1): RSI ¼ A DT Q
(1)
where: Q = the heat input to the calorimeter (W), A = the specimen area normal to the direction of heat transfer (m2), and DT = the absolute temperature difference between the interior and exterior air (K). The interior and exterior surface temperatures of the wall assemblies were also measured using thermocouples. One wall assembly was constructed for testing in the GHB test facility. The wall assembly measured 2.44 m (96.0 in.) in height and 2.44 m (96.0 in.) in width. The wall assembly layers and dimensions are given in figure 2. The area for the balcony during the construction was left empty through the wall up to the gypsum board. For Test One, the balcony area at the wall was filled with EPS and then tested in the GHB. For Test Two, the EPS was removed and a balcony was constructed in the following manner and another series of GHB tests were conducted on the test assembly. The tests were conducted at two different exterior temperatures (i.e., 20 C and 35 C). Table 1 summarizes the applied boundary conditions (interior and exterior air temperatures, convective heat transfer coefficient used) and the average measure temperatures on the cold and warm side of the wall.
Results and Discussion Employing the correlation for the GHB, the thermal resistances for the walls without and with balcony were calculated for two exterior temperatures of 20 C and 35 C.
FIG. 2 Exploded drawing of the wall assembly (left) and the dimensions of components (right).
GHOBADI ET AL., DOI: 10.1520/STP161720180063
TABLE 1 Exterior and interior boundary conditions Tcold air (8C)
hcold
w m2 K
w m2 K
Thot air (8C)
Tcold surface (8C)
Thot surface (8C)
5
21
4.6818
19.824
32.7
9.2
20
21
19.513
19.19
34.0
8.9
35
21
34.355
18.589
34.0
8.9
hhot
The results are shown in table 2. The thermal resistance values increase by decreasing the exterior temperature, which is due to the better thermal performance of the insulation in colder temperatures. An average decrease of 5.5% can also be seen when the balcony is added to the wall. To better evaluate the incremental increase in heat flow associated with the linear thermal bridge (the balcony slab), the linear transmittance can be used. With this approach, the heat flow through the interface detail assembly is compared with and without the thermal bridge, and the difference in heat flow is related to the detail as heat flow per a linear length. The difference in overall heat flow between the two assemblies is the additional amount of heat flow caused by the balcony/floor slab bypassing the thermal insulation.3 Dividing by the assembly width (linear length of the slab edge) creates the linear transmittance of the slab, which is a heat flow per linear length. Knowing the linear transmittance of a linear thermal bridging source and the clear field thermal transmittance of the wall assembly allows calculation of the total effective assembly thermal transmittance as shown in equation (2): P ðW LÞ UT ¼ þ Uo (2) ATotal where: Btu W UT is total effective assembly thermal transmittance (ft2 :hr: F or m2 K ), Btu W Uo is the clear field thermal transmittance (ft2 :hr: F or m2 K ), ATotal is the total opaque wall area (ft2 or m2), Btu W W is heat flow from the linear thermal bridge (ft:hr: F or mK ), and L is the length of the linear thermal bridge (ft. or m). The linear thermal transmittance for the two exterior temperatures is calculated and shown in table 2. The length of thermal bridging is 1.22 m (4 ft), and the area of the wall is 5.82 m2 (62.7 ft2). It can be seen that the linear transmittance is around 6.3% greater when the exterior temperature was set to 35 C, which means that the rate of heat transfer by the balcony slab is increased by reducing the mean temperature. We also employed COMSOL Multiphysics software to examine the heat transfer of the two tested walls. To normalize the temperature results for different outside temperatures, temperatures are nondimensionalized using the temperature index approach. The temperature index is the ratio of a surface temperature to the overall
391
392
Clear Field Wall R SI
Wall with Balcony Slab
R
U SI
U
RSI
R
U Si
U
W (w/mK)
W (Btu/ft hr.8F)
R AtA (258C)
5.30
30.11
0.189
0.033
5.12
29.1
0.195
0.0344
0.049104
0.028008
R AtA (2208C)
5.36
30.43
0.187
0.033
5.08
28.8
0.197
0.035
0.049104
0.028008
R AtA (2358C)
5.47
31.45
0.183
0.032
5.16
29.3
0.194
0.034
0.052446
0.036553
STP 1617 On Building Science and the Physics of Building Enclosure Performance
TABLE 2 Experimental results for the clear field wall, the wall with balcony slab, and linear transmittance
GHOBADI ET AL., DOI: 10.1520/STP161720180063
temperature difference between indoor and outdoor as shown in equation (3). A value of 0 is the outdoor air temperature and 1 is the indoor air temperature. Ti ¼
Tsurface Toutdoor Tindoor Toutdoor
(3)
Figure 3 and figure 4 show the interior and exterior temperature index contours on the exterior and interior of the EPS-filled wood wall assembly and balcony wall assembly, respectively. As can be seen from figure 3, the main source of thermal bridging that passes the majority of the heat flow is the wood stud and the difference between the maximum and minimum temperature index on the interior wall is around 2%. Whereas, for the wall assembly with balcony, the balcony should also be considered as a source of thermal bridging that increases the temperature index difference to 4% between the minimum and maximum values. Using the linear thermal transmittance and characterizing it for different types of sources of linear thermal bridging can make the calculation for the thermal resistance of wall assemblies simpler. Once the experimental results are used to benchmark a numerical method, the numerical method can then be employed to characterize different types of linear thermal bridging sources. Different ways are recommended to regulate the thermal bridging. We can prepare flexible catalogues for practitioners that the linear transmittance can be calculated as a function of different important parameters such as material thermal conductivities, dimensions, and so on. Another way is to define simplified correlations based on the numerical results. This method would require conducting numerous parametric studies to find the correlations between these parameters and the linear transmittance.
FIG. 3 Interior surface (left) and exterior surface (right) temperature index for wood stud wall assembly with the hole filled with EPS.
393
394
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 4 Interior surface (left) and exterior surface (right) temperature index for wood stud wall assembly with installed balcony.
ACKNOWLEDGMENTS
The authors acknowledge the financial support of National Resources Canada for the net zero thermal bridging project.
References 1. 2.
3.
4.
Pan-Canadian Framework on Clean Growth and Climate Change (Ottawa, ON, Canada: Environment and Climate Change Canada, 2016). Energy Standard for Buildings Except Low-Rise Residential Buildings, ASHRAE Standard 90.1 (Atlanta, GA: American Society of Heating, Refrigerating and Air-Conditioning Engineers, 2016). Natural Resource Canada, Energy Use Data Handbook (Ottawa, ON, Canada: Natural Resource Canada, 2009), http://oee.nrcan.gc.ca/corporate/statistics/neud/dpa/menus /trends/handbook/tables.cfm?wbdisable=true Thermal Bridges in Building Construction—Heat Flows and Surface Temperatures— Detailed Calculations, EN ISO 10211:2017 (Geneva, Switzerland: ISO, 2017).
5.
Thermal Bridges in Building Construction—Linear Thermal Transmittance—Simplified Methods and Default Values, EN ISO 14683: 2007 (Geneva, Switzerland: ISO, 2007).
6.
A. B. Larbi, Statistical Modeling of Heat Transfer for Thermal Bridges of Buildings, Energy and Buildings 37 (2005): 945–951.
7.
S. Roels, M. Deurinck, M. Delghust, A. Janssens, and D. V. Orshoven, “A Pragmatic Approach to Incorporate the Effect of Thermal Bridging within the EPBD-Regulation” in Proceedings of the Ninth Nordic Symposium on Building Physics, Vol. 3 (Tampere, Finland: Tampere University of Technology, Dept. of Civil Engineering, 2011), 1009–1016.
GHOBADI ET AL., DOI: 10.1520/STP161720180063
8.
Standard Test Method for Thermal Performance of Building Materials and Envelope Assemblies by Means of a Hot Box Apparatus, ASTM C1363-19 (West Conshohocken, PA: ASTM International, approved September 1, 2019), http://doi.org/10.1520/C1363-19
395
396
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180110
Wade L. Vorley1 and Lauran Drown1
Moisture Reduction Strategies for Building Envelopes Citation W. L. Vorley and L. Drown, “Moisture Reduction Strategies for Building Envelopes,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 396–419. http://doi.org/10.1520/ STP1617201801102
ABSTRACT
Building envelope standards and practices have evolved over the past few decades in response to increased energy efficiency goals and a better understanding of building science. New materials, systems, and detailing methods help to reduce thermal bridging and to reduce air leakage through building envelopes. A building envelope assembly that is properly designed, detailed, and installed to today’s standards typically will perform as expected, reduce energy usage, and provide a durable and long-lasting assembly. However, these technologies are not a panacea for deficiencies in the construction process, deferred maintenance, or unforeseen occupant uses and alterations. The sheer multitude of components involved in today’s building envelopes and the airtightness of the assemblies themselves can lead to vulnerabilities exacerbated by any of these factors. Having investigated many building envelope failures, we have found that air barrier, roofing, and waterproofing perfection is challenging to achieve. The goal of this paper is to outline strategies to reduce or evacuate moisture from building envelopes without extensive replacement of components or systems. This paper presents case studies of existing buildings equipped with moisture-monitoring data loggers to evaluate initial conditions and verify moisture reduction over time. The data loggers collected readings at 5-min. intervals for temperature, relative humidity, and moisture content and have been in place for more than six years in some buildings. The primary moisture reduction strategies employed in these studies include added thermal
Manuscript received November 16, 2018; accepted for publication April 14, 2019. 1 Wiss, Janney, Elstner Associates, Inc., 960 S. Harney St., Seattle, WA 98108, USA 2 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21-22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
VORLEY AND DROWN, DOI: 10.1520/STP161720180110
protection, modification of heating systems, the introduction of active air movement, and ventilation of the roof assembly. The success of these strategies is verified with empirical data. Keywords condensation, moisture, ventilation, air movement, air velocity, relative humidity, roof assembly, empirical study, building envelope
Introduction Building envelope standards and practices have evolved over the past few decades in response to increased energy efficiency goals and a better understanding of building science. New materials, systems, and detailing methods help to reduce thermal bridging and to reduce air leakage through building envelopes. A building envelope assembly that is properly designed, detailed, and installed to today’s standards typically will perform as expected, reduce energy usage, and provide a durable and long-lasting assembly. However, these technologies are not a panacea for deficiencies in the construction process, deferred maintenance, or unforeseen occupant uses and alterations. The sheer multitude of components involved in today’s building envelopes and the airtightness of the assemblies themselves can lead to vulnerabilities exacerbated by any of these factors. Elevated moisture levels in buildings or assemblies can then develop and, if not remediated, can lead to decay, corrosion, or organic growth. Having investigated many building envelope failures, we have found that air barrier, roofing, and waterproofing perfection is challenging to achieve. The goal of this paper is to outline strategies to reduce or evacuate moisture from building envelopes without extensive replacement of components or systems. It is a follow-up study to a previously published paper by this same author1 and contains additional data and research for two of the case studies presented in that paper. This paper presents empirical data gathered through case studies of existing buildings with known moisture problems. Moisture monitoring data sensors were installed in these buildings to measure temperature, relative humidity (RH), and moisture content of building components over time. The data loggers collected readings at 5-min. intervals and have been in place for more than six years in some buildings. Data were evaluated to compare initial conditions to moisture reduction over time and to verify the success of strategies employed. We studied moisture reduction strategies that included: Added Thermal Protection: Continuous roof insulation was added outboard of the roof deck to moderate the temperature on interior surfaces to reduce condensation potential. Modification of Indoor Relative Humidity: Lower RH was achieved with changes to heating and ventilation systems modified to reduce condensation potential as a lower RH would have a higher dew point.
397
398
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Movement of Indoor Air: Fans were installed to circulate stagnant indoor air to reduce condensation and the deposit of new moisture on material surfaces. This would allow materials to dry without actively removing moisture from the building interior. Added Ventilation: Active ventilation of the building envelope was achieved using induction fans and positive internal pressure combined with passive exhaust vents to force moisture-laden air out of the assembly. The first two strategies (continuous insulation and RH reduction) were covered in a previous paper1 that was presented at the Roofing Consultants Institute (RCI) National Convention in March 2018. The findings for those first two case studies are summarized in the “Background” section of this paper. The last two strategies (air movement and ventilation) were introduced in that previous paper; however, the studies were incomplete at the time of publication in March 2018. This paper further develops these two strategies and presents an evaluation of the completed data sets and findings of those studies. BACKGROUND
Over the past seven to eight years, we have evaluated a number of buildings with interior moisture and condensation issues. Early on, many of the buildings we studied were unconditioned warehouses in California, Nevada, and the Pacific Northwest that were constructed of tilt-up concrete panels, steel framing, and panelized wood roof systems. These buildings were insulated with fiberglass batts below the roof deck and had vapor barriers (VBs) that were discontinuous at the steel trusses (fig. 1). Many of
FIG. 1 Drawing detail of typical roof construction of tilt-up concrete warehouses.
VORLEY AND DROWN, DOI: 10.1520/STP161720180110
these buildings were unconditioned and unheated in the wintertime. Some of these warehouses had footprints of more than 300,000 ft2 and were divided into two or three tenant spaces with party walls. In a few of these buildings, one tenant space did not exhibit moisture problems while extensive wood decay at the ceiling was present in an adjacent tenant space with a similar use. In examining the history of buildings with this issue, we found that the affected section of the building was vacant or had often experienced a significant period of vacancy prior to our visit. This finding raised the issue of air stagnation and the lack of air movement within the spaces as a potential cause of these anomalies, and this issue was noted for future study. As mentioned previously, wood decay in the roof sheathing was observed in portions of the buildings we initially observed. In addition, some warehouse spaces did not serve their basic function of keeping stored goods dry because condensation was observed dripping off the steel trusses (fig. 2) during certain weather conditions. These problems needed to be investigated. We began our studies by monitoring wood moisture content, temperature, and relative humidity in the ceilings of a selection of these buildings (see the “Data Monitoring” section for details) and made recommendations to owners to address the conditions found. Continued monitoring of these buildings after modifications were complete provided the opportunity to verify with empirical data that the chosen strategy for each building was successful for reducing moisture and condensation.
FIG. 2 Condensation dripping off steel trusses at 9 a.m., June 18, 2011, PNW warehouse.
399
400
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Added Thermal Protection Strategy
This strategy was studied on two warehouses in the Pacific Northwest (PNW). The roofing on these buildings needed to be replaced (for unrelated reasons), which provided the opportunity to install continuous insulation and a vapor barrier above the wood roof deck. The new insulation affected the temperature on the underside of the deck, as demonstrated in the one week of temperature data collected during installation (fig. 3). Prior to the roof replacement, temperature on the underside of the deck was below the dew point for much of the day while immediately after the continuous insulation was installed, the temperature was raised about 3 F, bringing it above the dew point for the entire day and throughout the winter months. Shortly after the new roof was installed, one of the main warehouse tenants moved, leaving the space vacant for three years. The data loggers were disconnected for a portion of this time. When the sensor receiving devices were reconnected in 2016, we were able to verify that not only had the condensation potential (see the section “Condensation Potential” for an explanation of that subject) been reduced significantly and to a manageable level (fig. 4), but that the wood deck had dried to the interior (fig. 5) and was no longer gaining moisture from condensation in the winter months. The elevated levels of moisture content and condensation potential during the vacant period raised questions of the causation during that time. The long-term results of this study prove that continuous insulation and a vapor barrier placed outboard of the roof deck provide effective condensation control and can help remediate excessive moisture in roof assembly components.
FIG. 3 Temperature and dew point at underside of roof sheathing during week of roofing installation, PNW warehouse.
VORLEY AND DROWN, DOI: 10.1520/STP161720180110
FIG. 4 Six-year history of condensation potential, PNW warehouse. Note improvements after new roofing and continuous insulation added in 2012 and new tenant activity in 2015.
FIG. 5 Six-year history of moisture content in roof deck, PNW warehouse. Note improvements after new roofing and continuous insulation added in 2012 and new tenant activity in 2015.
RH Reduction Strategy
A second set of warehouses studied were located in Reno, NV, and were of similar construction to those studied in the PNW. However, unlike the PNW buildings, the Nevada warehouses were heated in the wintertime and the heat was provided from
401
402
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 6 Four years of relative humidity data, Nevada warehouse. Note the RH reduction after installation of indirect fired heaters.
direct-fired gas heaters. One of the products of combustion for direct-fired gas heaters is water vapor leading us to suspect that the elevated humidity levels at the ceiling experienced in winter were due in part to these units. Upon evaluating the available options to reduce condensation, the owners decided to replace these units with indirect-fired heating units. After the units were changed, significant reductions in relative humidity (fig. 6), condensation potential (fig. 7), and moisture content in the wood roof decks (fig. 8) were recorded in the winter months. This study proved that reducing interior RH was a reasonable intervention to resolve the problem of excess moisture for these buildings. It should be noted that the wood roof decking in some areas of the Reno warehouses was degraded by elevated moisture levels prior to our involvement and that some of these areas had been vacant for extended periods of time. The empirical evidence provided in the two studies summarized here demonstrates that properly located insulation or reduction of internal moisture sources (or both) are reasonable strategies to reduce condensation and excess moisture in buildings of this type of construction. While these findings may seem obvious or intuitive to some, installing new roofing with continuous insulation or replacing a heating system represents a major economic investment for a building owner. We began to explore additional cost-effective methods for reducing moisture and condensation and conducted further investigation of the effects of stagnant air in vacant warehouse spaces. Empirical evidence of these strategies can help owners and designers when comparing the benefits and costs of alternatives for reducing elevated moisture levels in similar situations.
VORLEY AND DROWN, DOI: 10.1520/STP161720180110
FIG. 7 Four years of daily condensation potential, Nevada warehouse. Note the reduction following the installation of indirect fired heaters.
FIG. 8 Four years of moisture content in wood deck, Nevada warehouse. Note the reduction following the installation of indirect fired heaters.
Hypothesis
Slow-moving air that passes over the surfaces of installed construction materials with elevated moisture levels will cause the moisture to evaporate and the moisture content in materials to be reduced to acceptable levels.
403
404
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Few studies in the building science field have provided empirical evidence from existing buildings to validate this hypothesis. In addition, many industry professionals have been skeptical of ventilation solutions since the 1980s when roof vents were shown to have a limited ability to dry out wet assemblies and were even shown to permit moisture into the assemblies.2 However, the typical flat roof assembly in use in 1981 was fully adhered with hot asphalt, and air movement below the membrane was limited due to the lack of air spaces. Mechanically attached roofing assemblies and layers of rigid insulation used today are far more likely to allow air movement within the assembly, and roof vents are much more effective. This study tests the hypothesis and addresses additional questions, such as: If air is stagnant for extended periods, is condensation more likely and why? If air movement causes evaporation, is condensation reduced or eliminated? What are the limiting factors for air speed and the optimal temperature and humidity thresholds of moving air? DATA MONITORING
The data monitoring equipment employed in the study recorded temperature, relative humidity, and moisture content over time. The equipment was acquired from Omnisense, Inc., and included the S-10 and S-16 data loggers. These were combined with data-receiving gateways with built-in cellular modems. The data were transmitted wirelessly from the sensors to the gateway on-site and uploaded through a cellular modem to the Omnisense website for analysis and download. The data collected include readings at 5-min. intervals for air temperature and relative humidity next to interior surfaces and similar data for an electrical resistancetype moisture measurement of wood materials. Data calculated through the Omnisense website include derived dew points and absolute humidity. Where appropriate, averages were taken on an hourly, daily, or weekly basis to aide graphical analysis of long-term data. In some cases, averages were calculated from readings of multiple sensors in a building to illustrate the aggregate effectiveness of repair strategies. CONDENSATION POTENTIAL
Condensation potential is defined and calculated in many ways by industry associations, computer-modeling software, and in written articles on the subject.3 In many cases, RH is the primary measure used to determine condensation potential and requires making some assumptions for computer modeling. For the purposes of this paper and our previous paper,1 we devised a relatively simple measure for condensation potential that, based on our studies, represents the condition effectively. The condensation potential is calculated from the difference between a temperature and its corresponding dew point temperature derived from the relative humidity at a given sensor location. We have defined condensation potential (CP) in equation (1) as: CP ¼ 1=ðT DPÞ
(1)
VORLEY AND DROWN, DOI: 10.1520/STP161720180110
where: T = temperature of the air directly beside the condensing surface, DP = the dew point temperature of the air directly beside the condensing surface, and CP is expressed as a percentage and is used throughout this paper. Although this measure could be criticized because the equipment used does not capture surface temperatures or the thermal mass and temperature retention of steel and other dense materials, it is effective in this paper as a comparative measure. RESEARCH
Published research on moisture within buildings or building envelopes that is based on empirical data from existing buildings is limited. Generally, prior studies on building envelope condensation are based on computer models that capture vapor diffusion through materials,4,5 although a few studies do include air movement in their work.6–8 Computer modeling programs in the fields of architectural design and building science are helpful in creating new designs that reduce the potential of condensation in buildings. Computer software allows users to change a single variable at a time and evaluate the results, which can be helpful in choosing one product, system, or performance level over another. However, not all variables are typically captured in standard computer models, including air movement within asbuilt assemblies, air leakage into and out of assemblies due to imperfections, and variable indoor humidity and temperature in unconditioned buildings. These variables can have a great effect on the moisture levels in existing buildings, so predictive models are not always reflective of actual conditions. It is common knowledge in the fields of atmospheric and oceanic study that, on a basic level, moisture in clouds is water vapor generated from the evaporation of water over lakes, oceans, and other bodies of water. Evaporation is the primary mechanism of creating water vapor and is facilitated by wind and air movement over liquid water. Theoretically, air movement across a wet surface, like a wet building material, will promote evaporation in the same way. The agriculture industry also has an interest in evaporation due to wind and air movement, especially during periods of drought and limited irrigation. In a 2014 study of the effect of wind on evaporation of moisture in soil, Davarzani9 found that evaporation rates were highest within the first 12 h, rates increased with increased wind speeds, and total evaporation was greater at any point with higher wind velocities. However, the wind speed effect was less pronounced after about two days and maximized at velocities of about 2 m/s (360 fpm). Wind speeds greater than that did not appear to provide significantly greater increases in evaporation. In a study presented in The Fabricator, Zeigler10 found that air movement at the ceiling level in warehouses, similar to those in our study, reduced surface temperatures by 2F (1.1 C) at very low air velocities of 0.3 m/s (50 fpm) and by up to 2.2 C (4 F) at 0.6 m/s (100 fpm). The temperature change at ceiling surfaces is mostly due to the mixing of air and a subsequent reduction of temperature
405
406
STP 1617 On Building Science and the Physics of Building Enclosure Performance
stratification. It is also apparent from the graph provided in the Zeigler study that the greatest rate of change in temperature occurs at air velocities of less than about 1 m/s (180 fpm). These two studies demonstrate that moving air at relatively low velocities of 0.6 to 2.0 m/s (100 to 360 fpm) promote evaporation and temperature moderation of surfaces, which, in terms of buildings, is applicable for condensation control. Air or wind speeds of 0.6 to 2.0 m/s (100 to 360 fpm) are beneficial for evaporation and temperature control of air in moist environments, but how do they compare to standards for occupant comfort related to indoor air movement? According to ISO 7730,11 indoor air movement in conditioned spaces should maintain a range of 0.10 m/s to 0.25 m/s (18 to 45 fpm), depending on outdoor temperature, climate, and time of year. ASHRAE Standard 55 has evolved over the years and currently recommends similar air velocities for occupant comfort. We examined a few studies of air movement standards12–15 and found that according to Boduch and Fincher,12 “Generally, airflow slower than 100 ft/min feels either pleasant or goes unnoticed. Higher than that, and the flow of air within an enclosed space can provoke distraction (up to 200 fpm) and annoyance (above 200 fpm).” This sentence is referenced as being drawn from Olgyay.16 Because all these wind speeds and velocities are based on different standards and presented in different units, a comparison of all of these standards and studies is provided in the conversion table (table 1) for easy reference. From these articles and standards, it appears that to achieve the maximal effect of evaporation of moisture from surfaces inside a building, velocities of just below 1.1 m/s (200 fpm) are optimal, which would surpass the ISO 7730 standards of 0.1 to 0.25 m/s (19 to 45 fpm). For occupied spaces such as offices, public spaces, and residential buildings, it will be challenging to provide air movement at a velocity great enough to remove moisture at a maximized rate while retaining occupant
TABLE 1 Air velocity standards and conversions Reference
m/s
ft/sec
ft/min
mph
0.1 to 2.5
0.3 to 0.75
18 to 45
0.22 to 0.54
2 F temperature effect10
0.3
0.8
50
0.6
4 F temperature effect10
0.6
1.7
100
1.2
1
3
180
2.17
1.1
3.3
200
2.41
2
6
360
4.34
ISO 7730 recommended range dependent on outdoor temperature and climate
Velocity that may go unnoticed12 Rate of temperature change maximized10 Velocity may be an annoyance12 Rate of evaporation maximized9
VORLEY AND DROWN, DOI: 10.1520/STP161720180110
comfort. However, if air movement was to take place in the evenings, weekends, or off-hours, higher air velocities could be used to dry building materials in these environments. Another strategy could be to provide air movement within building envelope cavities where it would have little effect on human comfort. Air movement at less than optimal rates could also be provided and still be effective, but it may take longer for moisture to evaporate. Air velocities in warehouse spaces are not subject to the same constraints as office spaces, so utilizing fans that move air at 1.1 m/s (200 fpm) may be acceptable in this type of building. To study the potential of air movement to reduce condensation and remove moisture from building components, we had opportunities to conduct testing on an existing warehouse and on a building that was under construction. The following empirical studies demonstrate the potential for air movement to reduce condensation and to remove moisture from building envelope components. CASE STUDY NO. 1—AIR MOVEMENT STRATEGY
Based on previous experience, warehouse spaces with greater activity experienced less condensation while spaces with lower activity had a higher frequency of moisture-related problems. Presumably, air movement was a big factor. In theory, air movement reduces the duration and thus the amount of moisture vapor allowed to condense on cold surfaces. Based on the research reviewed, air movement also promotes evaporation, thereby reducing the moisture content within the building materials. The first case study is of a tilt-up concrete warehouse building of a similar construction type to those in the background studies. This building is located in the South Seattle area and has a footprint of about 50,000 ft2. Condensation at the roof assembly was so prominent that it dripped off steel trusses onto shelves and aisle ways in the warehouse. Sensors were installed into the plywood at eight locations within the building. One area of elevated moisture in the wood roof deck occurred along the south wall in the shadow of the parapet where it was subject to less solar heating and thus a higher condensation potential at the interior. This south-southeast (SSE) sensor became the focal point of the study because it represented the worst-case scenario for condensation potential and moisture content in this building. No physical alterations were made to the roof or ceiling assembly, and the building was not ventilated to remove moisture-laden air. To generate air movement, the owner first installed caged propeller fans mounted to the trusses. This provided a reduction in condensation potential but only a slight decrease in the amount of condensate drips from the ceiling. Two weeks later, the owner installed large, industrial ceiling fans that directed airflow downward and generated air circulation throughout the building. This measure provided a further reduction in condensation potential, as shown in figure 9. The tenant reported no condensate drips after installation of the large ceiling fans.
407
408
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 9 Condensation potential for the first four weeks at the ceiling in the SSE corner, South Seattle warehouse.
A year-and-a half later, no drips had been reported since the fan installation, and the daily average condensation potential at the SSE sensor showed a marked improvement in the early months of the study (fig. 10). However, the period from late April to early May 2018 showed a rise in condensation potential, so we analyzed this data on an average hourly basis (fig. 11 and fig. 12). This revealed that the average hourly condensation potential reached above 50% for brief intervals, meaning that air next to the roof surface was within 2 F (1.1 C) of the dew point. Although the condensation potential reached above our desired threshold, it was during springtime when condensation potential is generally expected to be higher for Seattle, and the intervals were so short-lived that it was manageable. The moisture content in the wood deck increased during this period (fig. 13) indicating that some condensation was occurring throughout the early spring of 2018 and was elevated above 15% for about six months. However, the moisture content did not exceed 20%, a typical threshold for advanced wood decay. It is not known how high or for how long moisture content was elevated in the wood at this location prior to the start of this study. However, from studies of similar buildings (figs. 4, 5, 7, and 8) and the slope of the graphs at the beginning of this study (fig. 10 and fig. 13), it can reasonably be inferred that the moisture content prior to June 2017 likely exceeded 20% for extended periods in winter months. Relative humidity of the air near the SSE sensor was also elevated during April and May 2018 (fig. 14). Exterior temperatures during that time were not atypical for a Seattle spring, ruling out reduced temperature as a cause of high humidity, which is a common cause in wintertime. The absolute humidity of the interior air matched the increase in absolute humidity of the exterior air during this period, indicating
VORLEY AND DROWN, DOI: 10.1520/STP161720180110
FIG. 10 Daily average condensation potential for full 1.5-year study period at SSE sensor, South Seattle warehouse.
FIG. 11 Hourly average condensation potential for April/May 2018 at SSE sensor, South Seattle warehouse.
that the humid spring climate in Seattle was contributing to the high RH and moisture gain rather than temperature. The data collected in this study of the South Seattle warehouse demonstrate that air movement by itself can reduce condensation potential and that the problem of
409
410
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 12 Hourly average condensation potential for April 18 to 24, 2018, at SSE sensor, South Seattle warehouse.
FIG. 13 Daily average moisture content in roof sheathing for full 1.5-year study period at SSE sensor, South Seattle warehouse. Moisture content above 15% is identified in orange.
dripping condensate was eliminated. Although condensation potential and moisture content in the wood deck were not lowered to the extent realized in the background studies, both were lowered and maintained at manageable levels. The ceiling fans produced relatively low air velocities that were barely noticeable to the warehouse tenants.
VORLEY AND DROWN, DOI: 10.1520/STP161720180110
FIG. 14 Daily average relative humidity for full 1.5-year study period at SSE sensor, South Seattle warehouse.
It is possible that increased air speeds would have had a greater effect. We are continuing to monitor the conditions in this warehouse and plan to conduct a similar study in another building to further verify the success of this strategy. CASE STUDY NO. 2—VENTILATION STRATEGY
In this second case study, we were tasked with designing a method for removing moisture from a unique roof assembly during construction of a new building in West Seattle. The assembly included a white ethylene propylene diene terpolymer (EPDM) membrane fully adhered to a gypsum-based cover board over multi-ply rigid insulation of up to 12 in. in depth. The membrane, cover board, and top plies of insulation were attached with adhesives while the bottom layer of insulation was mechanically fastened to the roof deck. Beneath the insulation was a bituminous vapor barrier loose laid over 0.5-in. fire-retardant treated (FRT) plywood over 0.5-in. tongue-and-groove wood deck (fig. 15). The plywood had become wet during wintertime construction and, after the roof was complete, the moisture in the plywood was unable to escape. Six months after installation, moisture levels in the plywood were still elevated. This was worrisome because trapped moisture can contribute to organic growth and, if not reduced, can lead to wood decay. Opportunities to extract the excess moisture from the plywood were limited due to the full adhesion of the materials above and the concealment of the plywood from below. However, as the bottom ply of insulation was mechanically attached to the plywood, there was an opportunity to pass air underneath the vapor barrier and directly over the top surface of the plywood. The joint spaces between the tongueand-groove decking provided another opportunity for air movement and the
411
412
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 15 Roofing assembly for West Seattle case study.
plywood was installed with 0.5-in. gaps that would enable air to pass more effectively under the vapor barrier if pressure was great enough. Twelve data sensors were installed below the insulation directly into the top surface of the plywood. The contractor had already begun efforts to dry the plywood using dehumidifiers and box fans to induce air movement and extract moisture vapor from the building interior. Data gathered in the first few weeks of our involvement were analyzed to predict that the plywood would not reach acceptable moisture limits for many months using only the current strategies (fig. 16). In general, passive air vents have a limited ability to exhaust moist air from large roof areas because the zone served by each vent is limited by the attachment of the assembly and thermal conditions. Exhausting moist air with roof airextraction fans also has limitations because the negative suction causes spaces between materials to constrict, thereby limiting airflow. To avoid the constriction, it was necessary to induce positive airflow into the assembly rather than extract it. We recommended adding induction roof fans coupled with one-way exhaust vents to draw air in from the exterior and to circulate it through the roof assembly to help dry the plywood faster. Implemented in late July 2017, this strategy is referred to as Phase 2. The overall strategy is shown graphically in figure 17.
VORLEY AND DROWN, DOI: 10.1520/STP161720180110
FIG. 16 Predicted plywood drying times at several sensor locations, West Seattle case study.
FIG. 17 Diagram of moisture reduction strategy, West Seattle case study.
With few commercially available vent products to force air rather than extract it, we implemented attic fans modified for this purpose. Seven fans were installed along with twelve one-way exhaust vents (fig. 18 and fig. 19). The highest-rated attic fan that met the project requirements had an air flow of 1,800 CFM, which we
413
414
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 18 Modified attic fan prior to installation, West Seattle case study.
FIG. 19 One-way exhaust vent installed, West Seattle case study.
chose because air needed to be pushed under the vapor barrier and carried more than 50 ft in some instances. The vapor barrier was breached at the fan and exhaust vent penetrations to allow air to flow between the vapor barrier and plywood. The VB and roof openings were later repaired.
VORLEY AND DROWN, DOI: 10.1520/STP161720180110
Phase 3 was initiated when the HVAC system was active in this new building and included positive pressurization of the interior to force conditioned interior air up through joints in the wood deck, allowing it to exhaust through the one-way roof vents. The drying rate was evaluated after one month. At some locations, the predicted time to reach adequate moisture content was not acceptable to the contractor or owner so additional measures were taken. We implemented Phase 4 with the installation of more induction fans in late August, followed by Phase 5 with an increase in the positive pressurization of the building interior. Utilizing these strategies, the moisture content in the plywood was significantly reduced over a period of five to eight weeks in most locations and brought to within acceptable limits at all locations within four months. The most dramatic example of moisture reduction in this study is demonstrated in figure 20. At Location 2, a marked improvement in the drying rate occurred during Phase 2 and further accelerated during Phase 3 At Location 8 (fig. 21) and Location 1 (fig. 22), the Phase 4 addition of induction fans did not increase the drying rate, while the increase in positive pressure in Phase 5 did so significantly. The moisture content at Location 1 was reduced from more than 25% to around 15% within five to six weeks of the increased positive pressure of Phase 5. The results from Case Study No. 2 prove that induced air movement can be utilized to remove unwanted moisture from within roof assemblies and that the addition of positive pressure on the building interior significantly increases the rate of drying. Case Study No. 2 differs from the other studies presented in this paper because the moisture present in the assembly was acquired during construction and was not due to condensation generated during occupancy. However, we believe the results
FIG. 20 Moisture content in plywood at Location 2, West Seattle case study.
415
416
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 21 Moisture content in plywood at Location 8, West Seattle case study.
FIG. 22 Moisture content in plywood at Location 1, West Seattle case study.
achieved in this study using induced air and ventilation can be successfully employed for reducing condensation-related moisture from buildings. There are limitations to replicating this study, however. Case Study No. 2 was conducted during a hot, dry, Seattle summer without measurable rain and with outdoor relative humidity typically below 70%. The study certainly benefited from these climatic conditions. Modifications need to be made for this strategy to be successful in humid climates, during inclement weather, or to suit the circumstances of a particular assembly or occupancy situation. Modifications may include controls to
VORLEY AND DROWN, DOI: 10.1520/STP161720180110
allow air induction only when relative humidity is favorable, dehumidification of outdoor air brought into the assembly, solar-powered systems activated by the sun, and the incorporation of ventilation mats into roof assemblies to facilitate postoccupancy moisture extraction. Using heated, pressurized, interior air to circulate through a roof assembly that is exhausted to the exterior is not an energy-efficient strategy if required for extended periods and, therefore, should be used sparingly. Incorporating a cavity space or ventilation mat into the roof assembly would likely reduce drying times, though this would need to be anticipated by designers and installed during construction.
Summary Through empirical data, this paper described the success of different strategies for the reduction of condensation and construction-related moisture within buildings. Two previous studies summarized in the background included the addition of continuous insulation at the roof exterior and modification of heating systems. Two additional case studies demonstrated the use of air movement within a building interior and ventilation of the roof assembly to extract moisture. Monitoring of temperature, relative humidity, and moisture content over time in each of these studies provided a comparison of before-and-after conditions to verify the success of each strategy. The findings from the case studies are significant and demonstrate that moving air through an interior space and through a building envelope are viable strategies to remove unwanted water vapor and to maintain acceptable levels of moisture and condensation potential. Limitations of these strategies and how to tailor them were discussed to promote informed decision-making on which strategy to employ for a particular situation. Before implementing any strategy, it is important to first identify the root cause of elevated moisture levels and thoroughly evaluate potential solutions. The strategy for induced air and ventilation in the building envelope, as presented in Case Study No. 2, was a short-term measure to dry materials wetted during construction. While moisture gain in roof assemblies from interior condensation or rain during construction are not common in all climates, moisture gain from curing concrete in roofing assemblies with concrete decks or lightweight concrete fill is problematic in all climates. Many manufacturers offer vented base sheets for use over concrete decks to help diffuse excess moisture. However, venting of this moisture from under the vented base sheet to the exterior is not always accounted for in the roof design, and the base sheet typically is located on the warm side of the insulation so heat loss may result where exterior venting is provided. If air movement and ventilation strategies studied in this paper can be provided in roof assemblies over concrete decks, to vent the space directly under the roof membrane and outboard of continuous insulation, energy-efficient moisture reduction could be achieved in these particular roof assemblies and in other building envelope components.
417
418
STP 1617 On Building Science and the Physics of Building Enclosure Performance
As building science and the construction industry improve our abilities to tighten buildings and envelope assemblies against air leakage, removing unwanted moisture from interior spaces and within assembles will be an ever-important challenge. Continuous air and vapor barriers are essential to retain heated or cooled indoor air and keep out unwanted water vapor. Yet, as this paper demonstrates, we can reliably control air movement through the roof assembly and use it to our advantage to reduce moisture gain and allow excess moisture to escape. Providing ventilation under cladding and outboard of continuous insulation at other parts of the building envelope may also contribute to moisture reduction and control. Many manufacturers and designers currently are recommending natural ventilation and air movement within rainscreen assemblies, and this trend will likely continue. Innovative uses of air movement and ventilation to reduce condensation and promote evaporation have great potential to solve many moisture-related problems in buildings and envelope assemblies. Verifying the success and limitations of these strategies at other parts of the building envelope offers opportunities for further exploration.
References 1.
Vorley, W. L., “It’s Not the Heat, It’s the Humidity … Or is it?” in Proceedings of the 33rd RCI International Convention and Trade Show (Houston, TX, 2018), 223–235.
2.
W. Tobiasson, “Venting of Built-Up Roofing Systems,” in Proceedings of the Sixth Conference on Roofing Technology (Oak Park, IL: National Roofing Contractors Association, 1981), 16–21. L. Arena and P. Mantha, Moisture Research—Optimizing Wall Assemblies (Washington, DC: U.S. Department of Energy, 2013). H. H. Saber, M. C. Swinton, P. Kalinger, and R. M. Paroli, “Hygrothermal Simulations of Cool Reflective and Conventional Roofs,” in Proceedings of the 2011 International Roofing Symposium (Washington, DC, 2011), 1–28. M. Kehrer and S. Pallin, “Condensation Risk of Mechanically Attached Roof Systems in Cold Climate Zones” (paper presentation, 28th RCI International Convention and Trade Show, Orlando, FL, March 14, 2013). D. Auer, A. N. Karagiozis, and A. Desjarlais, “A Comprehensive Hygrothermal Investigation of an Unvented Energy-Efficient Roof Assembly in the Pacific Northwest” (paper presentation, Thermal Performance of Exterior Envelopes of Whole Buildings X Conference, Clearwater, FL, December 2, 2007). S. Molleti, B. Baskaran, P. Kalinger, and P. Beaulieu, “Air Intrusion and Its Effect on Moisture Transport in Mechanically Attached Roof Systems” (paper presentation, 2011 International Roofing Symposium, Washington, DC, September 7, 2011). A. Desjarlais, H. H. Pierce, and W. J. Woodring, “Practical Application of Hygrothermal Modeling of West Coast Wood Deck Systems,” RCI Interface 22, no. 3 (March 2014): 9–16. H. Davarzani, K. Smits, R. M. Tolene, and T. Illangasekare, “Study of the Effect of Wind Speed on Evaporation from Soil through Integrated Modeling of the Atmospheric Boundary Layer and Shallow Subsurface,” Water Resources Research 50, no. 1 (2013): 661–680, https://doi.org/10.1002/2013WR013952
3. 4.
5.
6.
7.
8. 9.
VORLEY AND DROWN, DOI: 10.1520/STP161720180110
10.
11. 12. 13. 14. 15. 16.
E. H. Ziegler and R. Oleson, “A (Re)moving Story: Eliminating Condensation,” April 2012, The Fabricator, https://www.web.archive.org/web/20200423212928/https://www. thefabricator.com/thefabricator/article/shopmanagement/a-removing-story-eliminatingcondensation Ergonomics of the Thermal Environment, ISO 7730 (Geneva, Switzerland: ISO Copyright Office, 2005), www.iso.org M. Boduch and W. Fincher, Standards of Human Comfort (Austin, TX: Center for Sustainable Development, University of Texas at Austin, School of Architecture, 2009). M. Fountain, “Air Movement and Thermal Comfort,”ASHRAE Journal 38, no. 8 (1993): 26-30. B. W. Oleson, “International Standards for the Indoor Environment. Where Are We and Do They Apply Worldwide,” Indoor Air 17, no. 7 (2004): 18–26. J. A. Orosa, Passive Methods as a Solution for Improving Indoor Environments (London: Springer-Verlag London Ltd., 2012). V. Olgyay, Design with Climate: Bioclimatic Approach to Architectural Regionalism (Princeton, NJ: Princeton University Press, 1963).
419
420
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180131
E. Webb Wright1
Assessment of Water Damage in a Mass Masonry Wall Building Citation E. W. Wright, “Assessment of Water Damage in a Mass Masonry Wall Building,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 420–434. http://doi.org/10.1520/ STP1617201801312
ABSTRACT
Following a modernization and expansion project of a historic school building constructed in 1927, facility personnel observed that moisture damage to interior finishes at the west wall of an auditorium had become more prominent. Damage was most evident in the cement plaster interior finish that had been applied to the interior surface of the exterior brick masonry mass wall. A comprehensive assessment of the mass wall assembly was performed to evaluate potential sources of moisture infiltration. This included a visual assessment of the exterior surface of the wall, which identified various distress conditions such as mortar joint cracks and cracked brick units as probable contributors to the interior moisture damage. In order to thoroughly evaluate other possible factors, an investigation was performed to assess the probability of vapor-driven moisture accumulation in the mass wall contributing to the moisture-related damage. The investigation included long-term monitoring of the interior environment with data loggers and hygrothermal modeling using the measured data to analyze the one-dimensional heat and moisture transfer behavior of the wall assembly. The hygrothermal analysis indicated it was unlikely that vapor-driven moisture accumulation would occur within the wall assembly to an extent necessary to contribute to the observed distress. Water penetration tests of the existing wall assembly and a mock-up of the repointed wall assembly in accordance with ASTM C1601, Standard Test Method for Field Determination of Water Penetration
Manuscript received November 30, 2018; accepted for publication June 12, 2019. 1 Walter P. Moore and Associates, Inc., 300 South Orange Ave., Suite 1200, Orlando, FL 32801, USA https://orcid.org/0000-0003-0662-9582 2 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
WRIGHT, DOI: 10.1520/STP161720180131
of Masonry Wall Surfaces, confirmed that repointing of mortar joints would significantly improve the water penetration resistance of the wall. Repair documents were prepared to address the conditions found to be contributing to the interior damage. Keywords data loggers, WUFI, repointing, mortar analysis, water penetration testing
Background Stuart-Hobson Middle School in Washington, DC, was constructed in 1927 and has undergone several improvements through the years. These include widespread renovations to the facility performed as part of a building modernization project and building additions completed on the west side of the school auditorium (fig. 1 and fig. 2). These improvements to the facility were completed between 2012 and 2014. The school auditorium has brick masonry mass walls with a direct-applied cement plaster interior finish (fig. 3). The auditorium had experienced moisture issues in the past, which reoccurred following the referenced building improvements, with moisture damage to interior finishes at the west auditorium wall (fig. 4) becoming more extensive over time. Moisture-related damage to plaster wall finishes
FIG. 1 General view of school auditorium.
421
422
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 2 West wall of auditorium.
FIG. 3 Wall section at auditorium exterior wall.
WRIGHT, DOI: 10.1520/STP161720180131
FIG. 4 Interior of west auditorium wall.
inside the auditorium and efflorescence on the interior face of sections of the west exterior wall inside the attic were observed during a preliminary walk-through in December 2015.
Investigation Following the preliminary walk-through, a detailed plan for investigation of the cause(s) of the moisture-related problems was developed. The plan consisted of a visual assessment phase immediately followed by a long-term monitoring phase. VISUAL ASSESSMENT
The visual assessment phase included observations of both the interior and exterior of the east and west walls of the auditorium and limited field testing to document current conditions. Exterior observations identified miscellaneous distress conditions at the building facade that could contribute to the moisture-related interior finishes’ distress. This included poor-quality mortar joints, cracked mortar joints, cracked brick, and spalled brick (fig. 5 and fig. 6). Moisture-related distress conditions at the building interior such as blistering/bubbling/peeling paint and deteriorated plaster were widespread on the west wall (fig. 7 and fig. 8) but notably absent from the east wall. Relative moisture readings were taken on the interior finished surfaces of the east and west walls of the auditorium with a capacitance-type
423
424
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 5 Exterior view of section of brick wall with typical distress conditions such as mortar joint cracks and cracked brick highlighted.
FIG. 6 Typical mortar joint crack at exterior.
WRIGHT, DOI: 10.1520/STP161720180131
FIG. 7 Moisture-related distress to plaster finishes at west wall.
FIG. 8 Moisture-related distress to paint and plaster finishes at west wall.
425
426
STP 1617 On Building Science and the Physics of Building Enclosure Performance
moisture meter. Readings corresponded with observations of finishes’ distress, with high readings being indicative of relatively higher moisture content taken at the west wall. Readings taken at the east wall were relatively low and were similar to readings taken at other interior plaster finishes not located at exterior walls. Infrared thermography imaging of the east and west walls (fig. 9) indicated relatively little variability in temperature across the wall surfaces except at windows, a condition likely attributable to air leakage or thermal bridging at window perimeters. No thermal signatures were detected that would have indicated specific hidden problems at particular wall locations or details contributing to elevated levels of moisture. The roofing membrane and associated flashings had been replaced as part of the renovations of the facility performed between 2012 and 2014. The roofing membrane, roof edge flashings, and gutter along the west wall of the auditorium were observed during the visual assessment with no apparent breaches in the roofing system or flashing or gutter deficiencies observed. Based on initial visual observations, it appeared that deterioration of the exterior mortar joints was a primary source for bulk water infiltration into the wall assembly. The relative differences in interior plaster damage between the east and west walls also suggested that vapor drive and differing drying rates of the walls may have been contributing factors. The blistered plaster on the west elevation was an indication that evaporative drying to the interior during the cooling season and
FIG. 9 Partial infrared (IR) image, west elevation. Arrows highlight thermal signatures corresponding to steel roof truss bearing locations.
WRIGHT, DOI: 10.1520/STP161720180131
resultant moisture accumulation within the wall cross section may have contributed to the interior distress. Long-Term Monitoring
A long-term monitoring program was developed in order to further evaluate whether vapor-driven moisture accumulation within the wall assembly was a contributing factor to the moisture-related plaster distress. Data loggers were installed in the auditorium during the initial site visit and programmed to take temperature and relative humidity readings every hour during the long-term monitoring period (fig. 10). Data loggers were positioned at both the east and west auditorium walls to capture any differences in the conditions of the two wall systems that might explain the relative lack of signs of distress on the east auditorium wall as compared to the west. The data loggers were retrieved after five months of monitoring from August through December 2016. ANALYSIS
WUFI* was used to model the one-dimensional heat and moisture transport, to evaluate the potential for moisture accumulation within the assembly due to vapor diffusion, in the typical wall assembly at the west auditorium wall, which had been observed to exhibit the most severe distress. WUFI allows for modeling of the transient hygrothermal behavior of multilayer building components exposed to natural climate conditions based on inputs of weather data and interior conditions for the specific building and location. Hygrothermal properties are those due to moisture absorption and temperature change. When air cools as it passes through the material layers of an assembly, moisture vapor carried in the air will condense on the first contacted cooler surface. Use of hygrothermal analysis can evaluate the conditions of the temperature and moisture that might prevail across and within a portion of any building enclosure over time. The one-dimensional WUFI software models vapor diffusion along with drying potential from exterior/interior conditions. The temperature and humidity data from the data loggers, as well as National Weather Service hourly outdoor weather condition data, were input into WUFI, so that the actual conditions at the site during the monitoring period were used in the analysis. The aforementioned model results indicated that, under most typical climatic conditions as ascertained by the data loggers, moisture accumulation within the typical finished wall assembly due to vapor diffusion did not appear to be a significant contributing factor to the observed distress conditions (fig. 11). These results were useful in determining an appropriate course of repair. The information indicated that the installation of a vapor barrier, or other vapor diffusion mitigation strategies, would not be required, indicating that the repairs should be focused on addressing bulk water infiltration. *Wa¨rme Und Feuchte Instationa¨r, which is translated as “heat and moisture transiency.”
427
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 10 Data logger.
FIG. 11 WUFI results indicating lack of moisture accumulation in wall due to vapor diffusion.
25 Brick (old) Cement Plaster (stucco)
20 Water Content [lb/ft3]
428
15
10
5
0 0
25.7
51.4
77.1 Time [days]
102.8
128.5
154.2
WRIGHT, DOI: 10.1520/STP161720180131
Based on the modeling results, it was determined that the repair design should be focused on restoring the exterior surface of the wall and introducing a means for water that enters the wall assembly to be routed back to the exterior. A testing program was undertaken to determine whether repointing would be an effective means of enhancing the resistance of the wall to water absorption and penetration. The program included mortar sample composition tests in accordance with ASTM C1324, Standard Test Method for Examination and Analysis of Hardened Masonry Mortar,1 and water penetration tests of the existing wall assembly and of a mock-up of the repointed wall assembly in accordance with ASTM C1601, Standard Test Method for Field Determination of Water Penetration of Masonry Wall Surfaces.2 TESTING
Laboratory examination of samples of the existing historic mortar was performed in accordance with ASTM C1324 to gain an understanding of the composition of the existing mortar. Based on detailed laboratory testing and calculations of mix proportions, it was determined that the existing mortar was a cement-lime mortar, equivalent to an ASTM C270, Standard Specification for Mortar for Unit Masonry,3 Type S cement-lime 1-part cement to less than 0.5-part lime to 3-part sand by volume. Based on the laboratory results, an ASTM C270 Type N cement-lime mortar made using Portland cement in conformance to ASTM C150, Standard Specification for Portland Cement,4 hydraulic lime in conformance to ASTM C207, Standard Specification for Hydrated Lime for Masonry Purposes,5 and sand in conformance to ASTM C144, Standard Specification for Aggregate for Masonry Mortar,6 was specified as the repointing mortar. Type N mortar was specified for repointing because it would be softer than the existing brick and existing mortar so that stresses that developed in the restored walls would be relieved in the mortar joints rather than potentially leading to distress in the brick units.7 ASTM C1601 chamber tests were performed to test the efficacy of repointing in enhancing the resistance of the auditorium walls to water absorption and penetration. Tests were done in both an area of the west wall that had had all mortar joints repointed and in an area of the wall where the existing mortar joints were unaltered. The results of the tests showed that the wall section with unaltered mortar joints allowed water absorption and penetration at significantly higher rates than the repointed wall section. The water absorption and penetration at the unaltered wall section was as much as four to five times higher than at the repointed wall section in liters of water per minute lost through the wall.
Repairs Repair drawings and specifications were developed based on the findings of the visual assessment and long-term monitoring, hygrothermal wall analysis, mortar analysis, and water penetration testing. The repair methodology centered on restoring the exterior face of the auditorium walls and introducing a means for water that does enter the wall cross section to be routed back to the exterior in order to
429
430
STP 1617 On Building Science and the Physics of Building Enclosure Performance
increase the resistance of the existing mass walls to water absorption and reduce the potential for damage to interior finishes. The specified and implemented repairs (fig. 12) consisted of removal and replacement of brick masonry units that were damaged, cracked, or spalled; repointing of all mortar joints (fig. 13); installation of wall flashing at the floor line of the attic space, at heads of windows, and below windows (fig. 14 and fig. 15); and application of a clear water repellant to the wall surfaces. Wall flashing was a partial through-wall flashing and was not extended through the full wall section (fig. 16).
Summary The Stuart-Hobson Middle School auditorium had a history of moisture issues, which reoccurred with damage to interior finishes becoming more prominent following building improvements completed between 2012 and 2014. Existing conditions were investigated by visual assessment and long-term monitoring of interior relative humidity and temperature over a five-month period. A comprehensive assessment was completed considering both bulk water infiltration and vapordriven moisture accumulation as potential contributing factors. The assessment resulted in the determination that bulk water infiltration into the mass masonry walls was the primary cause of the interior finishes distress and that hygrothermal
FIG. 12 Repairs in progress, east elevation.
WRIGHT, DOI: 10.1520/STP161720180131
FIG. 13 Repointed mortar joints.
FIG. 14 Installation of wall flashing.
431
FIG. 15 Elevation of west wall from repair documents showing locations of added partial through-wall flashing.
432 STP 1617 On Building Science and the Physics of Building Enclosure Performance
WRIGHT, DOI: 10.1520/STP161720180131
FIG. 16 Flashing installation detail at attic.
moisture transfer was not a significant source of moisture. This assisted with the development of an appropriate repair strategy. Field mock-up testing of both unaltered and repointed wall sections per ASTM C1061 confirmed that repointing of mortar joints could greatly increase the resistance of the wall to water absorption. Repairs were designed and implemented to restore the exterior face of the existing mass masonry walls and provide a means for removing water that does enter the wall cross section. Three paths for bulk water infiltration were addressed by the repairs: movement through mortar joint cracks through repointing and absorption of moisture into brick units and mortar joints by application of a water-repellent coating. Facade repairs were completed in August 2018, have been in place for approximately nine months as of the date of writing of this paper, and have thus far proven to be effective.
433
434
STP 1617 On Building Science and the Physics of Building Enclosure Performance
References 1.
2.
3.
4. 5.
6.
7.
Standard Test Method for Examination and Analysis of Hardened Masonry Mortar, ASTM C1324-15 (West Conshohocken, PA: ASTM International, approved December 1, 2015), http://doi.org/10.1520/C1324-15 Standard Test Method for Field Determination of Water Penetration of Masonry Wall Surfaces, ASTM C1601-14a (West Conshohocken, PA: ASTM International, approved July 1, 2014), http://doi.org/10.1520/C1601-14A Standard Specification for Mortar for Unit Masonry, ASTM C270-19ae1 (West Conshohocken, PA: ASTM International, approved May 1, 2019), http://doi.org/10.1520/C027019AE01 Standard Specification for Portland Cement, ASTM C150M-20 (West Conshohocken, PA: ASTM International, approved April 12, 2019), http://doi.org/10.1520/C0150_C0150M-20 Standard Specification for Hydrated Lime for Masonry Purposes, ASTM C207-18 (West Conshohocken, PA: ASTM International, approved October 1, 2018), http://doi.org/ 10.1520/C0207-18 Standard Specification for Aggregate for Masonry Mortar, ASTM C144-18 (West Conshohocken, PA: ASTM International, approved December 1, 2018), http://doi.org/10.1520/ C0144-18 R. C. Mack and J. P. Speweik, Repointing Mortar Joints in Historic Masonry Buildings (Washington, DC: U.S. Department of the Interior, October 1998).
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180071
Rex A. Cyphers1 and Jodi M. Knorowski1
Evaluating the Impact of Moisture Content on Thermal Resistance of Mass Masonry Wall Assemblies Citation R. A. Cyphers and J. M. Knorowski, “Evaluating the Impact of Moisture Content on Thermal Resistance of Mass Masonry Wall Assemblies,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 435–449. http://doi.org/10.1520/STP1617201800712
ABSTRACT
The thermal resistance of a wall assembly is one of the key factors that is considered when evaluating the contribution of the building envelope to the overall energy performance of new and existing buildings. The conductivity and thickness of the building materials comprising the wall assembly are used to determine the theoretical thermal resistance of a given wall assembly. While the thickness of the materials is generally unchanged, the conductivity of the material will vary depending on its exposure to moisture. The moisture content of a building material can be impacted by direct exposure to bulk water, such as during a rain event, or also through the transmission of water vapor. Because water will conduct heat at a greater rate than most building materials, an increase in moisture content within a wall assembly will result in a decrease in the overall thermal resistance. Published data for the thermal resistance of typical building materials do not account for changes in moisture content. Over time, the thermal resistance of a wall assembly will fluctuate and may not be captured accurately based on current design and modeling methodologies. Field data were collected for existing mass masonry wall assemblies using an assortment of data logging instrumentation to measure the in situ thermal resistance and moisture profile across the wall assembly. These data were analyzed to determine the impact that periods with elevated moisture levels within the wall assembly had on the measured R-value. Manuscript received October 10, 2018; accepted for publication April 14, 2019. 1 WDP & Associates Consulting Engineers, Inc., 335 Greenbrier Dr., Suite 205, Charlottesville, VA 22901, USA 2 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21-22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
435
436
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Keywords moisture content, thermal resistance, in situ measurements, mass masonry walls
Introduction Within the industry, there are a number of methods available to measure the moisture content and thermal resistance of a material, yet there is limited information on how these properties impact each other when such measurements are being taken. Variations in temperature, relative humidity (RH), or moisture levels could impact the measured thermal resistance at a given point in time. As such, the anticipated thermal resistance of a wall assembly based on laboratory testing may not be consistently realized in installed wall assemblies. MOISTURE CONTENT
Moisture content is the amount of moisture per unit volume of porous material and typically is measured in a laboratory setting by comparing the oven dry weight of a material sample to the weight when subjected to an amount of moisture. The moisture content of a material is most significantly impacted by exposure to bulk water, but hygroscopic materials can also be impacted by different levels of relative humidity. As the relative humidity increases, materials will take on and store water vapor from the air through adsorption. Different materials will adsorb moisture at different rates depending on the properties of the material. The relationship between relative humidity and moisture content can be determined by developing a sorption isotherm curve. These curves are developed in a laboratory setting by subjecting materials to increasing levels of relative humidity while maintaining a constant temperature and determining the moisture content of the material at each corresponding level once the material has reached equilibrium in accordance with ASTM C1498, Standard Test Method for Hygroscopic Sorption Isotherms of Building Materials.1 For building materials such as brick masonry, the increase in relative humidity generally has little impact on the moisture content of the material until a certain relative humidity is achieved at which point a significant increase in moisture content is observed. This is due to the pore structure of the brick masonry and the process in which it takes on moisture. As the relative humidity increases, the water vapor from the air becomes bound to the pore surfaces of the brick through adsorption. The amount of moisture being adsorbed by the brick is minimal until higher relative humidity values are reached, typically around 80%. As the relative humidity continues to increase, the smaller pores within the brick begin to fill with water through capillary condensation, which begins to increase the moisture content of the brick more rapidly. This generally occurs at moisture contents at a relative humidity above 95%, where water is not only bound to the surfaces of the pores but also is unbounded and fills the pores. At a relative humidity of 100%, the moisture within the brick has reached a point of free water saturation. While
CYPHERS AND KNOROWSKI, DOI: 10.1520/STP161720180071
FIG. 1 Sorption isotherm curve for porous building materials.2
many of the pores of the brick are filled with water, voids filled with air still remain within the structure of the brick. In order to go beyond this point, to a stage of supersaturation and to reach the maximum moisture content, water must be forced into the remaining pores of the brick. In a laboratory setting, this can only be achieved when a brick is placed in a vacuum or boiling water. A typical sorption isotherm curve for porous building materials is shown in figure 1, which illustrates these various stages of moisture content as they relate to the relative humidity.2 It should be noted that a sorption isotherm curve will vary for different materials, even different types of brick. For brick, the internal pore structure, to include the ratio of small and large pores, will be one of the key factors in the profile of the sorption isotherm curve. The moisture content of an existing building component can be difficult to determine, especially because it is constantly changing. Several methods are available to both qualitatively and quantitatively identify the amount of moisture within a given wall assembly. Material Testing
The moisture content of existing materials can be determined by extracting samples of the materials, sealing them in an airtight manner, and transporting them to a laboratory for testing. While this method is the most accurate for determining the
437
438
STP 1617 On Building Science and the Physics of Building Enclosure Performance
moisture content of that material, it only captures the moisture content at a specific point in time. Since the moisture content fluctuates, ideally samples should be taken at different times so moisture movement through the assembly can be better understood. This process can be fairly invasive depending on the number of samples that are taken, and it may be costly to perform a number of laboratory tests. Relative Moisture Content
Several moisture meters are available that can be used to determine the “dryness” or “wetness” of the outer surfaces of a material. However, without knowing certain properties of the material, this can only give you a relative indication of whether moisture content has increased or decreased but does not provide a quantitative value of the actual moisture content. Devices such as electrical impedance meters and electrical resistance meters use electrical circuit theories to determine the changes in moisture. These can be used in a nondestructive manner and generally can only identify moisture at an exposed surface, not deeper within the wall. Relative Humidity Probes
The use of relative humidity probes to determine moisture is common practice in flooring applications using methods outlined in ASTM F2170, Standard Test Method for Determining Relative Humidity in Concrete Floor Slabs Using in situ Probes,3 and has been adapted to similar building components. The use of data logging relative humidity probes to monitor conditions over time will provide a more accurate profile of the behavior of an existing wall in terms of moisture content in the hygroscopic range. Specific to mass masonry walls, methods have been developed where temperature and relative humidity probes are installed at various depths within a wall assembly to collect data over an extended period of time in accordance with ASTM E3069, Standard Guide for Evaluation and Rehabilitation of Mass Masonry Walls for Changes to Thermal and Moisture Properties of the Wall.4 This data can then be compared to sorption isotherm curves for a specific material to determine the approximate moisture content at a given time. Furthermore, with an understanding of both temperature and relative humidity, the actual vapor pressure can be calculated, and a vapor profile across the wall can be developed to understand how moisture is moving through the assembly. THERMAL RESISTANCE
The thermal resistance, better known as the “R-value,” of a material is the industry standard for how well the material is expected to perform from a thermal standpoint when installed in a building. There are a number of ASTM standards that provide methods for determining the thermal resistance of materials. Generally, the material is placed in an apparatus that creates a specific temperature differential across the material while the heat flux and surface temperatures are measured. This testing is performed in a controlled environment and accounts only for steady-state conditions.
CYPHERS AND KNOROWSKI, DOI: 10.1520/STP161720180071
While the thermal resistance is derived from variations relating to temperature, the impact of moisture on a material’s thermal performance should also be considered. As a material takes on more moisture, the thermal resistance typically will decrease because the thermal conductivity of water is much higher than most porous building materials and will allow heat to transfer through the material at a faster rate. Many of the ASTM standards do not provide specific conditions for relative humidity when testing for thermal performance, and guidance varies depending on the procedure being followed. ASTM C1363, Standard Test Method for Thermal Performance of Building Materials and Envelope Assemblies by Means of a Hot Box Apparatus, uses a hot box testing apparatus to determine the thermal resistance of materials or assemblies.5 As it relates to moisture exposure during testing, the standard states the “warm side relative humidity shall be kept below 15% or the laboratory shall verify that the dew point temperature of the metering side air is 2 C less than the minimum metering side surface temperature of the specimen.” ASTM C518, Standard Test Method for Steady-State Thermal Transmission Properties by Means of the Heat Flow Meter Apparatus, uses a heat flow meter apparatus to determine thermal resistance and references the manufacturer’s requirements for conditioning a specific material prior to testing.6 This standard states that, “Typically, the material specifications call for specimen conditioning at 22 C and 50% relative humidity for a period of time until less than a 1% mass change is observed over a 24-hour period.” The amount of moisture adsorbed by a material at 15% relative humidity and 50% relative humidity could impact the thermal resistance of the material depending on the material type and properties. The thermal resistance of an existing wall can also be measured using in situ data collection following procedures outlined in ASTM C1155, Standard Practice for Determining Thermal Resistance of Building Envelope Components from InSitu Data.7 The data collected will be representative of in-service conditions, which may vary from a controlled laboratory setting. The placement of the sensors on the building and exposure to different environmental conditions should be taken into consideration when determining the approximate R-value using in situ data. Data may need to be collected for extended periods of time to validate the results. MOISTURE CONTENT VERSUS THERMAL RESISTANCE
Research relating to the impact of moisture content on thermal resistance has been ongoing for at least 100 years. Early studies found a significant impact on the thermal conductivity of masonry materials depending on the amount of moisture exposure. Depending on the study, considerations were given to the moisture content by volume or to the development of a relative scale for moisture exposure. One such study found that the thermal conductivity of external walls in extremely wet climates was 65% above that of absolutely dry state brick.8 Another study also found an increase in thermal conductivity when exposing aerated concrete to various moisture contents; this study also noted that the changes in heat flow through the
439
440
STP 1617 On Building Science and the Physics of Building Enclosure Performance
material were a result of the moisture redistribution and saw fluctuations in the heat flow over time until the sample reached a state of quasi equilibrium.9 Currently, there is limited product specific data available on the impact that moisture content has on thermal resistance as most product manufacturer’s publish test data only at steady-state conditions as required by the ASTM testing standards. When performing hygrothermal analysis, understanding the thermal resistance of materials at various environmental conditions is important to calculate how heat is transferred through the assembly at each time iteration of the modeling simulation. As such, software programs used to perform these types of analyses generally include moisture-dependent thermal conductivity data for the materials in the material database. WUFI Pro* is one such modeling program, and data from the material database for several of the more common brick types are shown in figure 2. The sorption isotherm curves for the same brick can then be used to determine relative humidity corresponding to a given moisture content, in which case the approximate thermal conductivity can be calculated. For the “solid brick masonry” material from the WUFI database, which has a density of 118 lb/ft3, values for thermal conductivity are shown for different relative humidity values, as well as the
FIG. 2 Water content of various types of brick versus thermal conductivity.
*
WUFI Pro. Version 6.2, computer software. Stuttgart, Germany: Fraunhofer Institut Fu ¨r Bauphysik, 2017.
CYPHERS AND KNOROWSKI, DOI: 10.1520/STP161720180071
TABLE 1 Thermal conductivity of solid brick masonry at various relative humidity values
Relative Humidity
50% 3
Water Content (lb/ft ) Thermal Conductivity
60%
70%
80%
90%
95%
0.30
0.45
4.32
4.40
0.23 0
100%
0.68
1.12
2.26
3.94
11.86
4.52
4.75
5.35
6.23
10.40
0.23
0.22
0.21
0.19
0.16
0.10
1.8
4.6
10.0
23.9
44.4
140.8
(BTU-in/h-ft2- F) Thermal Resistance (h-ft2- F/BTU-in) % Increase from 50% RH
percent increase in thermal conductivity from a 50% baseline, in table 1. From this data, it should be noted that the percent increase in thermal conductivity of brick materials does not appreciably increase until the relative humidity is above 80%. As such, if the relative humidity remains relatively low, there should be limited impact on the thermal performance of the brick masonry due to moisture impacts. It should also be noted that the moisture content of the brick at 100% relative humidity is significantly greater than other relative humidity exposures. This is because the brick has reached the point of free water saturation and is beyond the hygroscopic range. Many of the voids within the internal structure of the brick are now filled with water instead of air. This impacts the thermal conductivity of the brick because water has a much higher thermal conductivity than air, thus allowing heat to flow more rapidly through the brick and lowering the thermal resistance. The impacts of the moisture content on the thermal performance of a multiwythe mass masonry wall can be demonstrated using the parallel method to calculate the U-factor of the overall assembly. A calculation for a mass masonry wall with three wythes of brick is shown in table 2 for an assembly where the relative humidity is consistently 50% throughout the assembly and for an assembly where the relative humidity and associated moisture content varies throughout the assembly. This calculation utilizes the thermal conductivity values found in table 1 to
TABLE 2 Theoretical impact of moisture content on thermal resistance Relative Humidity at 50% Thickness, in.
Thermal Conductivity, k
R-Value
Relative Humidity Varies Thermal Conductivity, k
R-Value
Exterior Air
…
Outer Wythe
4
4.32
0.93
0.17 6.23 (95% RH)
0.65
0.17
Middle Wythe
4
4.32
0.93
4.75 (80% RH)
0.84
Inner Wythe
4
4.32
0.93
4.32 (50% RH)
Interior Air
…
0.68
0.93 0.68
Sum of R-Values
3.64
Sum of R-Values
3.27
U-Factor
0.275
U-factor
0.306
441
442
STP 1617 On Building Science and the Physics of Building Enclosure Performance
calculate the R-value for each layer of the assembly. When accounting for the impacts of moisture, the U-factor for the overall assembly increases by about 11% in this example. While the focus of this paper is on the impact of moisture content on thermal resistance of mass masonry wall assemblies, it is important to note that modern construction of buildings incorporates many different materials beyond just masonry. Most notably, there are a number of insulation products with various material properties that are used for both interior and exterior cavity insulation. Although a number of studies have found that there is an impact from moisture on the thermal performance of insulating materials,2,8 the published data that are commonly referenced to make design decisions do not reflect these conditions. For example, most manufacturers publish the thermal resistance for a material only based on the steady-state conditions as outlined in the ASTM testing standards. While the material database in WUFI accounts for the moisture-dependent thermal resistance of a limited number of insulating materials, much of the data provided are only for dry conditions.
Case Study Walls To evaluate the impacts of moisture on thermal resistance, three mass masonry wall assemblies were evaluated. The walls were evaluated using the collection and analysis of field data utilizing data acquisition instrumentation. The three walls were studied in various locations and at different times of year. The following is a description of each of the wall assemblies. Wall Assembly 1 was part of a historic nineteenth-century mass masonry structure in Charlottesville, VA. The wall was constructed of two wythes of handmolded clay brick with lime mortar, an air space, and interior plaster with latex paint. For the purposes of this study, the instrumentation was installed on the interior and exterior surfaces of the brick, and the air space and interior plaster were excluded. The interior space was a single room approximately 13 ft by 13 ft in dimension. The space was unoccupied for the duration of the study. The wall studied was on the north elevation of the structure and did not have any protection from rain events. Instrumentation was installed during the summer months of June and July. A window-mounted air-conditioning unit maintained the interior temperatures between 21 and 23 C (69 F and 74 F) and 50% and 78% relative humidity. Wall Assembly 2 was part of a dormitory building constructed in the 1950s in Williamsburg, VA. The wall was constructed of a single wythe of exterior brick, 8in. concrete masonry unit block and interior plaster with paint. The interior space was approximately 12 ft by 18 ft in dimension; the space was unoccupied during the study while the remainder of the building was occupied. The wall studied was on the northeast elevation of the building and did not have any protection from rain events. Instrumentation was installed during the winter months from November to January. The interior conditions ranged from 20 C and 22 C (68 F and 72 F) and 35% to 60% relative humidity over the testing duration.
CYPHERS AND KNOROWSKI, DOI: 10.1520/STP161720180071
Wall Assembly 3 was part of an academic building constructed in the 1950s in Morgantown, WV. The wall was constructed of two wythes of exterior brick, an air space approximately 8 in. across, and 4-in. hollow clay tile finished with interior paint. The interior space was approximately 20 ft by 50 ft in dimension; the space was unoccupied during the study, although the rest of the building remained occupied. The wall studied was on the north elevation of the building and did not have any protection from rain events. Instrumentation was installed during both the winter and summer months from November to June. The building featured central air conditioning and radiators below the windows that were integral to the wall assembly. The interior conditions ranged from 21 C and 28 C (70 F and 83 F) and 9% to 46% relative humidity over the testing duration. IN SITU DATA COLLECTION
For each of the three wall assemblies, data acquisition instrumentation was installed to gather data to evaluate the impact of moisture on the thermal performance of the exterior wall assemblies. Interstitial temperature and relative humidity probes were installed at varying depths within the wall assemblies to record the movement of heat and moisture through the wall assembly. The probes were installed inside a temporary, impermeable liner that was sealed to the adjacent interior finishes to ensure the moisture measured was only at the end of the liner at a specific depth within the wall. Data collected were evaluated to understand the approximate moisture content of the brick by comparing data to published sorption isotherm curves. Additionally, the measured temperature and relative humidity values were used to calculate the actual vapor pressure and understand the vapor movement across the wall assembly. Ambient temperature and relative humidity data loggers were also installed at varying locations on the interior and exterior of the buildings. Heat flux and thermocouple sensors were installed to determine the approximate in situ thermal resistance of the assemblies in accordance with ASTM C1155.7 Heat flux sensors were installed at interior surfaces to document the heat flow through the assembly. Thermocouple sensors were installed at corresponding interior and exterior surfaces in close proximity to the heat flux sensor to measure the temperature differential across the wall assembly. When installed over an extended period of time, the data were analyzed to determine the approximate in situ R-value of the overall wall assembly.
Data Analysis For each wall assembly, the R-value at each measured time interval (approximately 5 min.) was calculated using the measured heat flux and temperature differential across the wall assembly at that same time interval. To determine the in situ Rvalue of the assembly, additional calculations are required to test for convergence between consecutive time steps and to test for the variance between these calculated R-values. Generally, the environmental conditions must be consistent over the
443
444
STP 1617 On Building Science and the Physics of Building Enclosure Performance
TABLE 3 Theoretical and calculated R-values for case study wall assemblies Wall Assembly
1
2
3
(Theoretical)
R-Value (Calculated)
R-2.80
R-2.25
U-0.36
U-0.44
R-3.91
R-4.23
U-0.26
U-0.24
R-3.19
R-7.25
U-0.31
U-0.14
course of several days in order for the data set to converge. The calculated R-value for each wall assembly is presented in table 3. These values are compared to the theoretical R-values for each wall assembly determined using the parallel path method. It was found that the calculated R-values were found to be agreeable with the theoretical values, with the exception of Wall Assembly 3. This discrepancy was attributed to the impacts of the air space and relative placement of the heat source within the wall assembly. For each wall assembly, data are presented for the calculated R-value, the temperature differential across the wall assembly, and the relative humidity measured in the exterior wythe of brick for each time interval for a portion of the monitoring duration. The relative humidity measurements were considered an indication of the moisture content within the brick; therefore, it was anticipated that there would be a correlation between higher relative humidity values and lower calculated R-values. WALL ASSEMBLY 1
The general trend for the calculated R-value for Wall Assembly 1 was found to be relatively consistent throughout the monitoring period. There was a period of time from July 3 through July 5 when the calculated R-value dropped significantly, as shown in figure 3. During this time, there was a heavy rain event. It was noted that there was limited drying of the exterior wythe of brick during this time, as was evidenced by the increasing relative humidity measurements that did not follow the typical daily wetting and drying pattern found throughout the monitoring period. Although the brick wythe did appear to be taking on more moisture during this time, the relative humidity only increased to about 80%. Based on the theoretical study, the moisture content of brick masonry does not have a significant impact on the thermal resistance until the moisture content is much greater. It was also noted that the temperature differential across the wall assembly was also significantly lower during this time, which could also cause the calculated R-value to decrease. Without a temperature difference across the wall assembly, there is limited heat flow through the wall, which are the two factors that are used to calculate the thermal resistance of the assembly. Based on the data that were collected, the periods where the R-value was calculated to be at a minimum during the testing period did not appear to be a direct result of increases in the relative humidity measured in the exterior brick wythe.
CYPHERS AND KNOROWSKI, DOI: 10.1520/STP161720180071
FIG. 3 Wall Assembly 1, temperature differential and relative humidity data compared to calculated in situ R-value.
WALL ASSEMBLY 2
When reviewing the data for Wall Assembly 2, it was noted that the relative humidity within the exterior wythe of brick generally remained above 80% as shown in figure 4. There were several noticeable instances where the calculated R-value was at a minimum (November 30, December 18, and January 12). These points in time did correspond with conditions where the relative humidity was at a maximum. There were rain events on November 30 and December 18; however, no precipitation was measured on January 12. Similar to the scenario of Wall Assembly 1, the temperature differential between the interior and exterior surfaces of the wall were at a minimum during the same time intervals where the calculated R-value was at a minimum, which may have a bigger impact on the calculated thermal resistance. It should also be noted that the monitoring period did not capture periods where the relative humidity within the exterior wythe of brick was consistently below 80% in order to compare the thermal performance to periods where the brick masonry was relatively dry. While the periods where the relative humidity was at a maximum were more consistent with periods where the calculated R-value was at a minimum than for Wall Assembly 1, the impact of the temperature differential at these same points could not be differentiated from the moisture content as a factor contributing to the lower thermal resistance of the wall assembly.
445
446
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 4 Wall Assembly 2, temperature differential and relative humidity data compared to calculated in situ R-value.
WALL ASSEMBLY 3
Data for Wall Assembly 3 was analyzed during the winter months (fig. 5) and the summer months (fig. 6). During the winter months, the relative humidity measured in the exterior brick wythe was below 60%, which correlates to a very low moisture content that should have little impact on the thermal performance of the wall assembly. The exterior conditions during the monitoring period were very inconsistent from day to day, resulting in very few cases where the calculated R-value would converge. This is also notable in the graph of the data, which does not indicate any patterns or correlation between the relative humidity in the exterior brick wythe or the temperature differential across the wall assembly with the calculated R-value. The relative humidity measured during the summer months was also relatively low and did not correspond with the periods in time when the calculated R-value was at a minimum. The calculated R-value during the summer months was noted to trend very closely with the temperature differential across the wall assembly. This wall assembly was unique from Wall Assembly 1 and Wall Assembly 2 in that the exterior wythes of brick masonry were separated from the interior clay tile furring by an air cavity. The low relative humidity values measured in the exterior brick wythe could be a result of the drying potential within the wall assembly that is created by the convective air movement within the wall cavity. It was also noted
CYPHERS AND KNOROWSKI, DOI: 10.1520/STP161720180071
FIG. 5 Wall Assembly 3, temperature differential and relative humidity data compared to calculated in situ R-value during winter months.
that the radiant heater for the space was located beneath the window assemblies in the room and was in plane with the wall assembly. As such, additional heat sources beyond just the solar heat gain from the exterior could have impacted the measurements, specifically during the winter months, which is why a consistent pattern is not observed between the temperature differential across the wall assembly and the calculated R-value, which was noted in the other wall assemblies that were studied.
Conclusion The impact of the environmental conditions on the thermal resistance of the mass masonry wall assemblies appeared to be driven more by the exterior temperatures and resultant temperature differential across the wall assembly than by the impacts of moisture for the cases that were studied. Each of the walls in the case study were well-constructed, mass masonry walls that did not have any signs of interior water leakage. The relative humidity within the exterior brick wythe measured using interstitial relative humidity probes generally was found to be below 80%, with the exception of Wall Assembly 2, which correlates to lower moisture content and less impact on the thermal conductivity of the masonry material. Wall Assembly 2, which did have relative humidity readings that were significantly higher than the other two wall assemblies, did have maximum relative humidity readings that
447
448
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 6 Wall Assembly 3, Temperature differential and relative humidity data compared to calculated in situ R-value during summer months.
corresponded to periods when the measured R-value was at a minimum; however, these instances also corresponded to periods when the temperature differential across the wall assembly was nearly zero; therefore, it is difficult to conclude that the impact on the thermal resistance was solely a result of the increase in moisture content of the brick masonry. FUTURE STUDIES
The data collected during the case studies provided insight into the behavior of the wall assembly when exposed to actual environmental conditions. Future studies are being developed to evaluate different types of wall assemblies in the field, as well as the impacts of moisture in a controlled, laboratory setting to evaluate a single variable at a time. The measured impacts to the thermal resistance of the wall assembly were most notable when the exterior ambient temperature was similar to the interior ambient temperature, reducing the thermal gradient across the assembly and thus lowering the measured R-value of the overall assembly. Therefore, the future studies will aim to mitigate the impacts of the temperature differential across the assembly such that measured changes within the assembly or material are related only to the moisture content of the material. The future studies will also aim to incorporate different materials, such as insulation, and wall assemblies beyond just brick masonry.
CYPHERS AND KNOROWSKI, DOI: 10.1520/STP161720180071
References 1.
2.
3.
4.
5.
6.
7.
8. 9.
Standard Test Method for Hygroscopic Sorption Isotherms of Building Materials, ASTM C1498-04a(2016) (West Conshohocken, PA: ASTM International, approved August 15, 2016), http://doi.org/10.1520/C1498-04AR16 American Society of Heating, Refrigerating and Air-Conditioning Engineers, “Heat, Air, and Moisture Control in Building Assemblies—Fundamentals,” in ASHRAE Handbook— Fundamentals (Atlanta, GA: ASHRAE, 2013), 25.1–25.19. Standard Test Method for Determining Relative Humidity in Concrete Floor Slabs Using In Situ Probes, ASTM F2170-18 (West Conshohocken, PA, ASTM International, approved March 1, 2019), http://doi.org/10.1520/F2170-18 Standard Guide for Evaluation and Rehabilitation of Mass Masonry Walls for Changes to Thermal and Moisture Properties of the Wall, ASTM E3069-17 (West Conshohocken, PA: ASTM International, approved February 1, 2019), http://doi.org/10.1520/E3069-17 Standard Test Method for Thermal Performance of Building Materials and Envelope Assemblies by Means of a Hot Box Apparatus, ASTM C1363-11 (West Conshohocken, PA: ASTM International, approved May 15, 2011), http://doi.org/10.1520/C1363-11 Standard Test Method for Steady-State Thermal Transmission Properties by Means of the Heat Flow Meter Apparatus, ASTM C518-17 (West Conshohocken, PA: ASTM International, approved May 1, 2017), http://doi.org/10.1520/C0518-17 Standard Practice for Determining Thermal Resistance of Building Envelope Components from In-Situ Data, ASTM C1155-95(2013) (West Conshohocken, PA: ASTM International, approved November 1, 2013), http://doi.org/10.1520/C1155-95R13 J. F. Van Straaten, Thermal Performance of Buildings (London, UK: Elsevier, 1967), 60–63. M. Bomberg and C. J. Shirtliffe. “Influence of Moisture and Moisture Gradients on Heat Transfer through Porous Building Materials,” in Thermal Transmission Measurements of Insulation, ed. R. Tye (West Conshohocken, PA: ASTM International, 1978), 211–233, https://doi.org/10.1520/STP35746S
449
450
BUILDING SCIENCE AND THE PHYSICS OF BUILDING ENCLOSURE PERFORMANCE
STP 1617, 2020 / available online at www.astm.org / doi: 10.1520/STP161720180107
John A. Jackson,1 Cheryl M. Saldanha,2 Gert Guldentops,3 Jin Rui Yap,1 Robert J. Abdallah,2 and Sarah B. Rentfro1
Thermal Performance of Spandrel Assemblies in Glazing Systems Citation J. A. Jackson, C. M. Saldanha, G. Guldentops, J. R. Yap, R. J. Abdallah, and S. B. Rentfro, “Thermal Performance of Spandrel Assemblies in Glazing Systems,” in Building Science and the Physics of Building Enclosure Performance, ed. D. J. Lemieux and J. Keegan (West Conshohocken, PA: ASTM International, 2020), 450–480. http://doi.org/10.1520/ STP1617201801074
ABSTRACT
Determining thermal performance of spandrel assemblies in glazing systems is a challenge for design teams. Most designers rely on current building energy codes and industry standards for defining and determining performance, but the industry as a whole recognizes that these codes and standards have several shortcomings regarding spandrel thermal performance. It is important for designers to understand the purpose of the codes and standards, as well as their limitations, when determining thermal performance. In this paper, we address the challenges in estimating the thermal performance of spandrel assemblies in glazing systems by reviewing current building energy code and industry standards, performing a parametric study using two-dimensional (2-D) and three-dimensional (3-D) analysis and calculation methods, and discussing considerations for future development of calculation methods, codes, and standards to address current shortcomings. Keywords fenestration, glazing systems, spandrel assemblies, thermal performance, U-factor, parametric analysis, THERM, ANSYS
Manuscript received October 15, 2018; accepted for publication May 3, 2019. 1 Simpson Gumpertz & Heger Inc., 1625 Eye St. NW, Suite 900, Washington, DC 20006, USA 2 Simpson Gumpertz & Heger Inc., 550 Seventh Ave., New York, NY 10018, USA 3 Simpson Gumpertz & Heger Inc., 480 Totten Pond Rd., Waltham, MA 02451, USA 4 ASTM Symposium on Building Science and the Physics of Building Enclosure Performance on October 21–22, 2018 and December 2, 2018 in Washington, DC, USA. C 2020 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. Copyright V
JACKSON ET AL., DOI: 10.1520/STP161720180107
Introduction Current U.S. building energy codes and standards (i.e., 2018 International Energy Conservation Code (IECC);1 ASHRAE 90.1-2016, Energy Standard for Buildings Except Low-Rise Residential Buildings2) define minimum prescriptive requirements for thermal performance of various building envelope systems, such as roofs, walls above-grade, and fenestration. Spandrel assemblies within fenestration and glazing systems,* however, are not well addressed within these codes. As opaque exterior wall elements, spandrels are categorized as “walls above-grade,” with minimum prescriptive code requirements for thermal performance that are typically unfeasible or impractical to achieve with conventional construction materials and systems. Therefore, designers often struggle to demonstrate the compliance of spandrel assemblies with prescriptive energy code requirements and compensate for the lower spandrel thermal performance with other building envelope or operating systems, such as mechanical and electrical systems. Furthermore, because U.S. industry standards do not provide clear guidance for accurate calculation of spandrel thermal performance (i.e., U-factor), spandrel U-factors used in trade-off analyses or building energy models are often inaccurate. These inaccuracies can lead to undersized mechanical equipment, increased operating costs, and thermal comfort issues for building occupants.
Definition of Spandrel Assemblies The American Architectural Manufacturers Association (AAMA) defines spandrels as “the opaque areas of a building envelope, which typically occur at floor slabs, columns, and immediately below roof areas.”3 Similar to a glass panel, spandrel assemblies are “glazed into” the framing (e.g., aluminum mullions) of fenestration. Spandrels typically incorporate some combinations of the following components, listed from exterior to interior (fig. 1): • Exterior Panel: Transparent (e.g., monolithic glass, insulating glass unit), semitransparent (e.g., ceramic frit-coated glass), or opaque panels (e.g., opacified spandrel glass, metal panel, terracotta, stone). • Air Cavity: Either fully sealed or vented/pressure-equalized to the exterior, typically 1 in. deep minimum.{ • Intermediate Panel: Typically, metal panel; included only when paired with transparent or semitransparent exterior panels (e.g., glass) to visually conceal insulation behind.{
*
We use the terms “fenestration” and “glazing systems” interchangeably, which include curtain walls, storefronts, windows, window walls, and other similar glazing systems. { The benefits of venting or sealing spandrel air cavities is dependent upon a variety of factors not discussed here. { This type of spandrel assembly of transparent or semitransparent exterior panels with an intermediate panel is called a “shadow box” assembly.
451
452
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 1 Diagram of spandrel assembly components.
• •
Insulation: Typically, semi-rigid mineral wool. Interior Membrane/Back Pan: Either a foil-facer membrane laminated to the insulation (functioning as an interior vapor retarder), taped to the perimeter framing, or a metal back pan (functioning as an interior air-vapor barrier or possibly, if detailed properly, as an interior air-water-vapor barrier), anchored and sealed to the perimeter framing.*
Building Energy Code and Referenced Energy Standards Designers often use the prescriptive requirements of the energy code as baseline or target thermal performance values for building envelope systems. The 2018 IECC lists prescriptive thermal performance requirements for opaque building envelope systems in Tables C402.1.3, C402.1.4,{ and, for fenestration systems, C402.4. Similarly, ASHRAE 90.1-2016 (a standard referenced by the 2018 IECC) lists prescriptive thermal performance requirements for building envelope systems in Tables 5.5-0
* The interior membrane/back pan is often selected based on factors such as whether the spandrel assembly is vented or sealed, the air-water-vapor management design of the spandrel assembly, and the desired durability. { Table C402.1.3 provides minimum R-values of the insulation components only, whereas Table C402.1.4 provides the maximum U-value of the entire assembly. Compliance with either table demonstrates compliance with the code.
JACKSON ET AL., DOI: 10.1520/STP161720180107
through 5.5-8. These energy codes and standards require one of the following paths for demonstrating energy code compliance: • Prescriptive Path: Building envelope systems must meet prescriptive thermal performance value requirements (e.g., R-value, U-factor) as defined in the energy code, among other prescriptive and mandatory requirements (e.g., maximum fenestration area, minimum solar and daylighting performance values, maximum air leakage). Performance values vary based on climate zones. This compliance path offers designers the least flexibility. • Building Envelope Trade-Off Path: This compliance path offers designers greater flexibility compared to the prescriptive path by allowing betterperforming building envelope systems to compensate for worse-performing building envelope systems.* • Performance Path: This requires an energy model to compare the proposed design to a baseline version of your building, using annual energy cost as the comparison metric. This compliance path allows trade-offs among other building envelope systems as well as other building systems, such as mechanical and electrical systems, allowing the greatest flexibility for designers. The 2018 IECC does not provide a clear definition of spandrel assemblies. The IECC defines a wall above-grade as “a wall associated with the building thermal envelope that is more than 15 percent above grade and is on the exterior of the building or any wall that is associated with the building thermal envelope that is not on the exterior of the building.”1 IECC, Section C402.1.1, states that, “The opaque portions of the building thermal envelope shall comply with the specific insulation requirements of Section C402.2 and the thermal requirements of either the R-value based method of Section C402.1.3; the U-, C-, and F-factor based method of Section C402.1.4; or the component performance alternative of Section C402.1.5.”1 Vertical fenestration is defined as “windows that are fixed or operable, opaque doors, glazed doors, glazed block, and combination opaque and glazed doors composed of glass or other transparent or translucent glazing materials and installed at a slope of not less than 60 degrees from horizontal.”1 ASHRAE 90.1-2016 defines a wall as, “That portion of the building envelope, including opaque area and fenestration, that is vertical or tilted at an angle of 60 degrees from horizontal or greater. This includes above- and below-grade walls, between floor spandrels, peripheral edges of floors, and foundation walls.”2 It goes on to define an above-grade wall as, “a wall that is not a below-grade wall,” a metal building wall as “a wall whose structure consists of metal spanning members supported by steel structural members (i.e., does not include spandrel glass or metal panels in curtain wall systems),” and a steel-framed wall as, “a wall with a cavity (insulated or otherwise) whose exterior surfaces are separated by steel framing members (i.e., typical steel stud walls and curtain wall systems).”2 Appendix A * U.S. Department of Energy COMcheck software is available to demonstrate an envelope trade-off compliance path.
453
454
STP 1617 On Building Science and the Physics of Building Enclosure Performance
notes, “Opaque mullions in spandrel glass shall be covered with insulation complying with the steel-framed wall requirements.”2 Fenestration is defined as, “all areas (including the frames) in the building envelope that let in light, including windows, plastic panels, clerestories, roof monitors, skylights, doors that are more than one-half glass, and glass block walls,” vertical fenestration as, “all other fenestration other than skylights,” and fixed metal framing in vertical fenestration as, “all types of vertical fenestration other than entrance door and operable, including, but not limited to, curtain walls, window walls, fixed windows, picture windows, glass block walls, non-openable clerestory windows, and non-openable sidelights and transoms.”2 Spandrel assemblies are opaque and, as such, they do not qualify as “fenestration” per the energy code definitions. However, they are part of metal framing in glazing systems and, as noted in ASHRAE 90.1-2016, are required to be categorized as “above-grade, steel-framed walls” with insulation that meets the listed prescriptive R-value and covers the interior surface of mullions.2 Assemblies that do not comply with these prescriptive R-value requirements must comply with the prescriptive assembly U-factor requirements for the entire spandrel assembly. However, the defined prescriptive thermal performances for “above-grade, steelframed walls” typically are unfeasible or impractical to achieve with conventional construction materials of spandrel assemblies. For example, the 2018 IECC requires the maximum prescriptive U-factor for metal-framed walls above-grade in commercial construction in Climate Zone 4 (except Marine) to be 0.064 btu/h*ft2* F.1 Although the thermal performance of spandrel assemblies is highly dependent on panel size, configuration, insulation thickness, and detailing, the calculated U-factor of spandrel assemblies can be as much as three to eight times higher than the maximum-allowable prescriptive Ufactor. Therefore, if a project attempts to use the prescriptive compliance path, the spandrels are unlikely to achieve this value, and designers must then turn to an alternate path to code compliance where the lower spandrel performance must be compensated by other building envelope and energy-consuming systems. The California Energy Code4 known as Title 24 defines spandrels as, “opaque glazing material most often used to conceal building elements between floors of a building so they cannot be seen from the exterior, also known as ‘opaque in-fill systems.’” The mandatory maximum allowable U-factor for spandrels in Title 24 is 0.28 btu/h*ft2*f.4 Title 24, Joint Appendix 4 includes a table for determining Ufactors for spandrel panel assemblies based on insulation R-value, frame type, and spandrel panel type. Four framing conditions are considered: aluminum without a thermal break, aluminum with a thermal break, structural glazing, and no framing (insulation is continuous). The following three spandrel panel types are considered: • Spandrel panel with little or no insulating value (e.g., single pane glass, stone veneer, metal panels, or precast less than 2 in. thick) • Spandrel panel with double-glazed insulating glass, and no low-e coating • Spandrel panel with triple-glazed insulating glass or double-glazed insulating glass with a low-e coating
JACKSON ET AL., DOI: 10.1520/STP161720180107
No interpolation is permitted when using the table. The U-factors for the spandrels were derived from a regression analysis of the values from the 2009 ASHRAE Handbook of Fundamentals, Chapter 15, Table 4.5 The U-factors assume the spandrel assembly includes gypsum board finishes and does not include a metal backpan. The values in this table demonstrate that achieving performance that is below the maximum allowable U-factor is possible with a number of configurations of panel insulation, framing conditions, and panel type, however, gives no guidance for calculating U-factors for panels that include a metal back pan.
Industry Standards Although spandrel assemblies are opaque, they are a component of a fenestration system, and as such, the industry generally recognizes the appropriateness of using fenestration calculation methods to estimate the thermal performance of spandrel assemblies. Various industry standards define methods to determine fenestration U-factors using calculation procedures and empirical test methods. (An example of an industry standard for empirical physical testing of thermal performance is ASTM C1363, Standard Test Method for Thermal Performance of Building Materials and Envelope Assemblies by Means of a Hot Box Apparatus.6) The primary objective for defining calculation and empirical test procedures is to allow for relative comparison of fenestration products. It is generally accepted that the procedures currently used may not always be accurate. For example, the calculation methods can underestimate the U-factor because they may not account for the impact of project-specific fenestration sizes or make appropriate approximations of three dimensional (3-D) heat flow. The industry continues to make progress in defining, simplifying, and standardizing fenestration calculation methods but does not yet provide clear guidance for accurate estimation of U-factors for spandrel panel assemblies. Consequently, the accuracy of designers’ representations of thermal performance of spandrels is affected by extrapolation of industry-standard procedures intended for fenestration and generalization of non-project-specific dimensions. ANSI/NFRC 100
The 2018 IECC and ASHRAE 90.1-2016 each require determination and certification of fenestration U-factors in accordance with American National Standard Insitute/National Fenestration Rating Council standard, ANSI/NFRC 100-2017, Procedure for Determining Fenestration Product U-Factors.7 ANSI/NFRC 100-2017 outlines a procedure for determining the overall U-factor of fenestration. The standard defines required module sizes (i.e., width versus height) of various fenestration product types (e.g., fixed window, curtain walls, skylights) as well as required boundary conditions (e.g., interior/exterior temperatures, film coefficients) for U-factor calculations; these requirements ensure consistency among different fenestration products (e.g., from manufacturer to manufacturer and product to product)
455
456
STP 1617 On Building Science and the Physics of Building Enclosure Performance
as well as consistency regardless of the climatic region (e.g., U-factors calculated for fenestration installed in Anchorage, AK, are the same as those in Miami, FL). The standard also provides a method for calculating U-factors based on project-specific module sizes based on an area-weighted average calculation.* Generally, this calculation procedure involves dividing the glazing system into the following three zones: • Frame Zone: The portion of the glazing system encapsulated by the width of the frame (often metal), typically ranging from 2 in. to 4 in. or more. This component usually has the highest U-factor (i.e., worst thermal performance) due to the thermal conductivity of the metal framing.{ • Edge Zone: The portion of the glazing system that transitions between the frame and the center-of-glass (COG) zone (i.e., beyond the sightline), stipulated to be 2.5 in. wide per ANSI/NFRC 100.{ • COG Zone: The remaining portion of the glazing system at the center of the glass panel.§ Once the U-factors for each of these zones are determined, the overall system U-factor (i.e., U-effective) is then calculated based on an area-weighted average of each component to comply with the standard (fig. 2): The NFRC calculation method does not require modeling the entire cross section of the fenestration product to determine the individual zone U-factors. NFRC requires modeling 150 mm (6 in.) of the glass panel beyond the sightline but only using the 63.5-mm (2.5 in.) edge zone U-factor for the calculation. The COG U-factor is obtained from a separate one-dimensional (1-D) model often performed by the glazing manufacturer.** The calculated overall U-factor must then be validated with physical empirical testing8{{ of the fenestration product in order to receive NFRC certification.{{ *The Building Envelope Trade-Off and Performance (i.e., energy model) compliance paths require fenestration U-factors to be calculated based on project-specific module sizes. Therefore, it is not appropriate to use ANSI/NFRC 100-defined module sizes for demonstrating code compliance with these paths, unless projectspecific module sizes happen to coincide with ANSI/NFRC 100 standard module sizes. { Thermal discontinuities within the metal framing (e.g., thermal breaks) improve the frame’s thermal performance. { Edge zones are utilized as a means for capturing the 2-D thermal influence of the frame on the glass before it transitions to the center-of-glass zone where the heat flow through the system more closely exhibits 1-D heat flow behavior. § ANSI/NFRC 100 requires modeling 150 mm (6 in.) of the glass panel beyond the sightline (i.e., interface with the frame). ANSI/NFRC 100 states that isotherms become parallel within the 6 in. extent-of-glass and that 2-D effects are adequately captured in the 2.5-in. edge zone. Therefore, a separate 1-D model can be used for calculating the center-of-glass U-factor. ** ANSI/NFRC 100 states that isotherms become parallel within the 6 in. extent-of-glass and that 2-D effects are adequately captured in the 2.5-in. edge zone; therefore, a separate 1-D model can be used for calculating the center-of-glass U-factor. {{ An available test method for determination of fenestration U-factor includes Architectural Aluminum Manufacturer’s Association (AAMA) 1503, Voluntary Test Method for Thermal Transmittance and Condensation Resistance of Windows, Doors, and Glazed Wall Sections. {{ ANSI/NFRC 100 requires the calculated U-factor and the tested U-factor to be within a certain margin of error, generally 10% or less.
JACKSON ET AL., DOI: 10.1520/STP161720180107
FIG. 2 Area-weighted average U-factor.
ANSI/NFRC 100 makes the following reference to spandrel panels in Section 5.9.6.3.1: “Spandrel panels shall be rated for U-factor at the size specified in Table 4-3 if the spandrel infill can be represented solely as a glazing assembly.” Table 4-3 goes on to define a spandrel panel module size of 2,000 mm by 1,200 mm (79 in. by 47 in.) with a single vertical mullion in the center of the panel. No additional guidance or alternative calculation methods are provided for calculating spandrel panel U-factors, other than using the same area-weighted average procedure described earlier for fenestration, which is often what designers use. The ANSI/NFRC 100-defined 63.5 mm (2.5 in.) wide edge zones for fenestration can be slightly inaccurate for glass panels and can significantly overestimate thermal performance for spandrel panels. These inaccuracies are addressed further in our parametric thermal analysis. AAMA 507
AAMA 507, Standard Practice for Determining the Thermal Performance Characteristics of Fenestration Systems Installed in Commercial Buildings,9 provides a standard practice for determining thermal performance of generic and product-specific fenestration systems in commercial buildings. The standard provides a methodology to calculate fenestration system U-factors for systems that include vision and spandrel areas based on the following parameters: • Frame type (e.g., nonthermal, thermally improved, thermal barrier, structurally glazed) for either aluminum or insulating spacers within insulating glass units
457
458
STP 1617 On Building Science and the Physics of Building Enclosure Performance
Center-of-panel (e.g., center-of-glass, center-of-spandrel) R-value Ratio of panel area (e.g., center-of-glass, center-of-spandrel) to area of total fenestration system AMMA 507 provides charts* for designers to estimate system U-factors for generic fenestration systems, which are useful for designers when product-specific systems are unknown. In addition, manufacturers can produce system U-factor charts for their specific fenestration product.{ The charts are useful but have some limitations. The charts are limited to a ratio of COP area/total area between 70% minimum and 95% maximum. In addition, the “spandrel area” is defined as “the area of the spandrel infill between the primary sash or frame members,” and the standard is unclear on how the edge-effects (i.e., the transition between the frame and center-of-spandrel) are considered.9 The standard also notes, “It is assumed that the spandrel insulation does not cover the framing members, is flush with the 6-mm (0.25 in.) glass and has a foil facing exposed to the interior” and provides assumptions for the R-value of insulation corresponding with various insulation thicknesses.9 It is unclear what type of insulation these R-values are based upon; however, they appear to be based on fiberglass batt insulation rather than mineral wool. In addition, the standard does not appear to consider the thermal benefit of the exterior panel (e.g., insulating glass), the thermal benefit of the air cavity between the exterior panel and the insulation, or the thermal effect of a metal back pan. • •
Research The limitations associated with ANSI/NFRC 100 and AMMA 507, the general lack of guidance available in the industry, and the code-mandated need for reasonable estimates of spandrel U-factors have prompted additional research and study. We summarize some of this recent work here. SPANDREL THERMAL PERFORMANCE SIMULATION
Stefan Elsholtz, a Master of Architecture student at the Massachusetts Institute of Technology (MIT), wrote an article in OAA Perspectives—The Journal of the Ontario Association of Architects titled, “Spandrel Thermal Performance” in the fall of 2014.10 In this article, the authors performed 2-D simulation{ of spandrel panels that included metal back pans. The analysis found the presence of a metal back pan (a common alternative to a foil-facer on insulation), significantly reduces the overall U-factor of the spandrel assembly, “such that the resulting effective U-value is almost as high as for a spandrel with no insulation at all.”10 The reduced
*
The charts are based on the frame type, center-of-panel (glass or spandrel), and ratio of panel (glass or spandrel) area to total area. { AAMA 507 requires that fenestration products be tested for U-factor in accordance with NFRC 100 or AAMA 1503. { Simulations utilized THERM analysis software developed by Lawrence Berkeley National Laboratories (LBNL).
JACKSON ET AL., DOI: 10.1520/STP161720180107
performance is due to the heat drawn from the building interior through the back pan and transmitted to the exterior through the mullion. The degree to which thermal performance is reduced due to the metal back pan depends on various parameters, such as the thermal properties of the mullion itself (e.g., no thermal break, thermally improved, thermally broken), insulation type and thickness, back pan material type (e.g., galvanized steel) and thickness, and anchorage detail of the back pan to the mullion (e.g., thermally separated or not). As a means of reducing this heat loss through the back pan, the authors suggest insulating on the inside surface of the back pan (i.e., helping to keep the back pan cold in a cold climate), as well as providing an alternative material for the back pan in lieu of metal, such as fiberglass. These alternatives help with improving thermal performance and are valid, but in practice, the location of insulation relative to the location of the back pan and the back pan material (i.e., rigidity, vaporpermeability, combustibility) need to be carefully evaluated to prevent condensation* and meet fire code restrictions for combustibility. BEST 4 CONFERENCE PAPER
The National Institute of Building Science BEST 4 conference in April 2015 included a research paper authored by members from the engineering firm Morrison Hershfield and the manufacturers Northern Facades and Dow Corning.11 In the paper, the researchers compared the thermal performance of spandrels in curtain walls utilizing 2-D and 3-D analysis methods, as well as physical empirical hotbox testing. The research utilized a 5-ft-by-5-ft-square spandrel panel with thermally broken aluminum framing. The surrounding construction (e.g., adjacent vision glass panels) was not included. Various spandrel assemblies were considered, including panels consisting of double- and triple-glazing and vacuum-insulated panels. All assemblies utilized 5-in.thick mineral wool insulation in front of a galvanized metal back pan behind the panels, attached directly to the aluminum framing. Comparing modeling methods to testing, the research paper noted the following: • 2-D modelling of the hotbox spandrel assembly using NFRC-100 can approach within 10–15% of the hot box test values as long as there are some adjustments to the approach […]. Further alterations, such as changing the edge distance, will likely provide even better agreement. The use of 2-D modelling is best used for quick comparisons between systems or where increased accuracy for thermal performance is less important. • The 3-D modelling approach in this paper, as per ASHRAE 1365-RP, provides the greatest agreement to the hotbox values (within 3%). 3-D modelling requires competent modelers and a desire for accuracy. In order to more accurately analyze designs as they would be installed in the field, 3-D modelling currently provides the most practical means of doing so. * Providing insulation inboard of the back pan reduces the temperature of the back pan, increasing risks of condensation.
459
460
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FENESTRATION ASSOCIATION OF BRITISH COLUMBIA REFERENCE PROCEDURE
The Fenestration Association of British Columbia (FENBC), in partnership with the building science and engineering firm RDH Building Science Inc. (RDH), published a document in September 2017 titled “Reference Procedure for Simulating Spandrel U-Factors” in an effort to better define calculation procedures for spandrel U-factors.12 Like ANSI/NFRC 100, the FENBC reference procedure involves dividing up the spandrel area into component parts (e.g., frame, edge, and center of spandrel), determining the U-factor for each component part,* and averaging the individual U-factors on an area-weighted basis to determine the overall assembly U-factor of the spandrel assembly.{ The reference procedure differs from ANSI/NFRC 100 primarily in that it increases the “edge zone” of the spandrel from 2.5 in. wide (as defined by ANSI/NFRC 100) to 6 in. wide, “A dimension found to be more accurate for spandrels.”12 The procedure notes, “Where isotherms converge to become parallel, the center of glass panel performance is achieved. For spandrel, similar to the simulation shown, this condition often occurs at a greater edge distance than is specified for glass infill.”12 The second way the procedure differs from ANSI/NFRC 100 is by including the slab edge construction inboard of the spandrel assembly. The procedure identifies the following three calculation methods depending on the relative position of the spandrel at slab edges: • Configuration 1—Uninterrupted Spandrel: A spandrel located beyond the face of the floor assembly such that it bypasses the floor slab (e.g., spandrel glazed into curtain wall) • Configuration 2—Partially Interrupted Spandrel: A spandrel located near the face of the floor assembly such that the spandrel is partially interrupted by the floor slab • Configuration 3—Entirely Interrupted Spandrel: A spandrel that is fully interrupted by the floor slab (e.g., slab edge cover integral to a window wall system) The procedure notes that Configurations 2 and 3 have significantly reduced thermal performance compared to Configuration 1 due to decreased insulation thickness at the slab edge and a decrease in the COP-to-frame area ratio. The authors state that the reason slab edges are included in the procedure is because their effect on spandrel Ufactor can be significant and results in “potentially large errors inherent in applying a simulated ANSI/NFRC 100 spandrel U-value to an interrupted spandrel configuration.”12 The procedure relies on 2-D models and area-weighted averages,{ similar to ANSI/NFRC 100’s procedure, and does not include physical empirical testing.
*
Component U-factors were simulated in THERM 7 and WINDOW 7, analysis software developed by LBNL. The FENBC procedure provides a spreadsheet tool to perform the area-weighted average calculation. { The authors of this paper provide an available Excel worksheet to users to perform the area-weighted average U-factor calculation. {
JACKSON ET AL., DOI: 10.1520/STP161720180107
Parametric Thermal Analysis Current industry standards have several well-known limitations for calculating accurate spandrel U-factors, as discussed in the previous sections. We summarize the key limitations here: • The NFRC-defined edge zone width of 2.5 in. is not appropriate for spandrel panel thermal calculations, specifically when combined with a separate 1-D COP thermal model. • The current standard’s 2-D calculation methods may not accurately approximate 3-D thermal effects of spandrels. • Typical spandrel material components and configurations (e.g., exterior panel types, air cavities, insulation type, and metal back pan) and their effect on thermal performance are not addressed in the current standards. To review the effect of these limitations and determine a potential range of more realistic spandrel U-factors, we performed a parametric thermal analysis utilizing commercially available finite element analysis (FEA) software tools (both 2-D and 3-D) and several calculation methodologies. We developed a set of models with incremental changes to evaluate the interdependence of multiple parameters, including spandrel size, insulation thickness, and thermal break at back pan, and their effect on the spandrel’s overall thermal performance. Our analysis excludes the following: • Evaluation of solar, daylighting, and condensation control • Thermal effects of adjacent assemblies (e.g., slabs, roofs, grade) • Physical empirical testing (e.g., for calibration and validation) • Thermal effects from fabrication and installation tolerances MODEL GEOMETRY AND ASSUMPTIONS
We used a typical curtain wall system (aluminum-framed, structurally glazed) as is common in contemporary building construction as the basis for our analysis, including spandrel assembly materials, components, and configurations as shown in figure 3: Because we modeled a curtain wall system, which most typically is installed in front of the slab edge, we did not include the thermal effects of adjacent construction (e.g., curtain wall anchors, edge-of-slab firesafing, slab edges, interior trim/ finishes) in our analysis.* We modeled spandrel panels of various module sizes within the curtain wall system. Mullion horizontal spacing was fixed at 5 ft wide, and we varied the height of the spandrel panel among 2 ft, 5 ft, and 10 ft. We leveraged symmetry and modeled a typical module that extends to the centerlines of the spandrel and adjacent vision panel, as shown in figures 4, 5, and 6.
* This is consistent with ANSI/NFRC 100, which does not account for the thermal effects of adjacent construction in calculation of U-factors.
461
462
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 3 Typical curtain wall spandrel.
FIG. 4 Elevation of 2-ft module.
JACKSON ET AL., DOI: 10.1520/STP161720180107
FIG. 5 Elevation of 5-ft module.
We varied the mineral wool insulation thickness among 2 in., 3 in., and 4 in. (without modifying the overall depth of the system), and varied the thermal-break thickness at the metal back pan among 0 in., 0.25 in., and 0.5 in. Figure 7 shows the various combinations of insulation and thermal break thicknesses we modeled. Note that we included the effect of fasteners to secure the metal back pan to the mullion and assumed continuous silicone sealant at the thermal breaks between the back pan and mullion. These variations in parameters resulted in 27 total different combinations (cases) shown in Table 1. METHODOLOGY
We analyzed the 27 cases using the following FEA computer software tools: • For 2-D analyses: WINDOW and THERM Version 7.6, developed by LBNL. • For 3-D analyses: ANSYS Version 19.1, developed by ANSYS, Inc.
463
464
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 6 Elevation of 10-ft module.
To calculate the overall spandrel assembly U-factor, we compared the following calculation methods: • Method 0—Center-Body: 1-D THERM model calculating the COP Ufactor only (fig. 8). When used alone, this does not comply with ANSI/ NFRC 100 or any other known accepted calculation methods.* We include this method to demonstrate the significant effects of frames and edges and also to combine with Methods 1 and 2 described here. • Method 1—2.5-in. Edge with 1-D COP: 1-D THERM model for the COP (Method 0 described here) and two 2-D THERM models (i.e., a vertical section through the head/sill frames and a horizontal section through the left/right frames) to obtain frame and edge U-factors. The frame, edge,
* The authors have seen this calculation method used erroneously on projects for calculating spandrel Ufactors.
JACKSON ET AL., DOI: 10.1520/STP161720180107
FIG. 7 Combination of insulation thicknesses and thermal break thicknesses.
and COP zones are slightly modified from NFRC, eliminating mitered corners (fig. 9). While this change likely has a minor impact on the overall U-factor, it may be more accurate for curtain walls where the vertical mullions extend beyond horizontals. Our methodology is also modified from NFRC because we include the spandrel metal back pan and insulation in the models. Notice the discrepancy between the isotherms in figure 10 (i.e., lines of constant temperature; plotted at 5 F increments) at the 2.5-in. edge interface to the 1-D COP model; the 2-D heat flow behavior changes abruptly at this interface, as traced with the dashed lines between the 1and 2-D models.
465
466
STP 1617 On Building Science and the Physics of Building Enclosure Performance
TABLE 1 Combination of parameters
Spandrel Height, ft
2
Insulation Thickness, in.
2
3
4
5
2
3
4
10
2
3
4
Thermal Break at Back Pan
Case
None
1
0.25 in. at 12 in. o.c.
2
0.5 in. at 12 in. o.c.
3
None
4
0.25 in. at 12 in. o.c.
5
0.5 in. at 12 in. o.c.
6
None
7
0.25 in. at 12 in. o.c.
8
0.5 in. at 12 in. o.c.
9
None
10
0.25 in. at 12 in. o.c.
11
0.5 in. at 12 in. o.c.
12
None
13
0.25 in. at 12 in. o.c.
14
0.5 in. at 12 in. o.c.
15
None
16
0.25 in. at 12 in. o.c.
17
0.5 in. at 12 in. o.c.
18
None
19
0.25 in. at 12 in. o.c.
20
0.5 in. at 12 in. o.c.
21
None
22
0.25 in. at 12 in. o.c.
23
0.5 in. at 12 in. o.c.
24
None
25
0.25 in. at 12 in. o.c.
26
0.5 in. at 12 in. o.c.
27
Note: All other components were held constant between cases; o.c. = on center. •
•
Method 2—6-in. Edge with 1-D COP: Similar to Method 1 but uses the 6-in. edge zone dimension noted in the FENBC procedure (fig. 11). The discrepancy at the interface between the 1- and 2-D models is less compared to Method 1 (fig. 12). Method 3—Thermal Symmetry: This method eliminates the need for an edge zone and 1-D COP model because the two 2-D THERM models are extended to their “point of thermal symmetry” (i.e., dimension at which the isotherms are mirrored). The frame and COP U-factors are obtained directly from each of their respective THERM models. However, the COP U-factors from the vertical and horizontal models are different; therefore,
JACKSON ET AL., DOI: 10.1520/STP161720180107
FIG. 8 Method 0—Center-body, excerpt from THERM showing isotherms (temperatures in F).
FIG. 9 Method 1—Spandrel elevation.
•
we employ a weighted average by drawing a 45 line extending from the spandrel corners toward the center (fig. 13). This approach equates the influence of the calculated COP U-factor proportionally to its side length. The isotherms are continuous within the model (without any staggers or abrupt shifts; fig. 14), indicating that the simulation is more accurate than the previous methods. Method 4—Three Dimensional: 3-D ANSYS that includes 3-D effects of system components. ANSYS captures both discrete and continuous elements and provides an overall U-factor without the need to apply the areaweighting average calculations of the previous 2-D methods (fig. 15).
467
468
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 10 Method 1—Excerpt from THERM showing isotherms (temperatures in degrees
F).
FIG. 11 Method 2—Spandrel elevation.
AUTOMATED OVERALL SPANDREL U-FACTOR CALCULATION
We automated the calculation of the 27 cases of overall spandrel U-factors for Methods 1, 2, and 3 using the visual programming software tool Grasshopper, a plug-in for Rhinoceros* (fig. 16). With these tools, we imported THERM results into *
Rhinoceros is a computer-aided design software program, and Grasshopper is a visual programming tool. Both were developed by Robert McNeel & Associates.
JACKSON ET AL., DOI: 10.1520/STP161720180107
FIG. 12 Method 2—Excerpt from THERM showing isotherms (temperatures in F).
FIG. 13 Method 3—Spandrel elevation.
Grasshopper to automate the calculation of overall spandrel U-factors for Methods 1, 2, and 3. This approach not only provides greater efficiency but also reduces data-entry errors in postprocessing calculations.
469
470
STP 1617 On Building Science and the Physics of Building Enclosure Performance
FIG. 14 Method 3—Excerpt from THERM showing isotherms (temperatures in F).
FIG. 15 Method 4—Three-dimensional excerpt from ANSYS.
FIG. 16 Automated parametric calculation.
JACKSON ET AL., DOI: 10.1520/STP161720180107
471
472
STP 1617 On Building Science and the Physics of Building Enclosure Performance
RESULTS Table 2 summarizes the results from our analysis following the methods described
here, and figure 17 plots the results for each of the three spandrel module sizes. We include the code-mandated prescriptive U-factor of 0.064 btu/h*ft2* F (2018 IECC, Climate Zone 4, except Marine, for “walls above-grade, metal framed” for commercial construction) in the plots for reference.
Observations and Discussion The results of our parametric analysis provide insight into the effects of the various parameters and calculation methods on the spandrel’s overall thermal performance as summarized here. BUILDING CODE REQUIRED SPANDREL U-FACTOR
Utilizing calculation methods other than Method 0 yields an overall U-factor that is higher than the prescriptive code-allowed maximum of 0.064 btu/h*ft2* F (Climate Zone 4). The U-factor of even the best-performing spandrel that we modeled is approximately 55% higher than the prescriptive code maximum. This finding supports the general understanding among designers that prescriptive code requirements for spandrel thermal performance typically are impractical to achieve or unachievable with contemporary construction materials and systems. We also observe that the size of the spandrel has an impact on U-factor. So, while using NFRC standard sizes to determine U-factors is prudent when comparing spandrel products to one another, it may not be appropriate for comparing a projectspecific system to the prescribed energy code U-factor or for use in an energy model. METHOD 0—1-D U-FACTOR
This calculation approach yields results that are lower than the code-prescribed maximum U-factor but does not account for the effects of framing; therefore, it is not an accurate representation of actual spandrel thermal performance. While we have seen some designers use this approach on projects, we recommend against its use because it significantly overestimates actual thermal performance. COMPARING 2-D METHODS
When comparing the three 2-D calculation methods (Methods 1, 2, and 3) to each other, we find that Method 1 yields the lowest U-factors, and Methods 2 and 3 yield similar results. When reviewing the plots in figure 17, it is important to consider the increments of the vertical axis. The scale has been stretched for the purposes of clarity, but practically, U-factors of 0.20 versus 0.23 btu/h*ft2* F are very close when trying to achieve a code requirement of 0.064 btu/h*ft2* F. The comparison of Methods 1 and 2 suggests that the NFRC calculation methodology for spandrel panels should include a larger “edge zone” dimension of at least 6 in. as recommended by FENBC/RDH. Method 3 shows that 2-D models could also be drawn to the point of thermal symmetry as an alternative means of capturing edge effects in the overall U-factor calculation.
TABLE 2 Table of overall U-factor results
Calculated Spandrel U-Factor Spandrel Height (ft)
2
Insulation Thickness (in.)
2
3
5
2
3
4
Case
Method 0 1-D COP
None
1
0.064
0.25 in.
2
0.5 in.
3
None
4
0.050
Method 1 2-D 2.5-in. Edge with 1-D COP
Method 2 2-D 6-in. Edge with 1-D COP
Method 3 2-D Thermal Symmetry
0.199
0.227
0.226
0.135
0.202
0.229
0.221
0.126
Method 4 3-D
0.196
0.216
0.212
0.123
0.192
0.222
0.222
0.132
0.25 in.
5
0.195
0.224
0.218
0.122
0.5 in.
6
0.192
0.216
0.210
0.120
None
7
0.25 in.
8
0.5 in.
9
None
10
0.042
0.064
0.187
0.218
0.220
0.130
0.188
0.217
0.212
0.120
0.181
0.207
0.202
0.117
0.157
0.178
0.182
0.123
0.25 in.
11
0.163
0.186
0.185
0.116
0.5 in.
12
0.161
0.181
0.179
0.114
0.148
0.170
0.174
0.116
0.154
0.177
0.176
0.109 0.107
None
13
0.25 in.
14
0.5 in.
15
None
16
0.25 in. 0.5 in.
0.050
0.151
0.172
0.170
0.141
0.163
0.167
0.112
17
0.147
0.171
0.169
0.105
18
0.144
0.163
0.162
0.102
0.042
JACKSON ET AL., DOI: 10.1520/STP161720180107
4
Thermal Break
(continued) 473
474
Calculated Spandrel U-Factor Spandrel Height (ft)
10
Insulation Thickness (in.)
2
3
4
Thermal Break
Case
Method 0 1-D COP
Method 1 2-D 2.5-in. Edge with 1-D COP
Method 2 2-D 6-in. Edge with 1-D COP
None
19
0.064
0.129
0.145
0.153
0.125
0.25 in.
20
0.138
0.157
0.159
0.126
0.5 in.
21
None
22
0.050
Method 3 2-D Thermal Symmetry
Method 4 3-D
0.145
0.161
0.163
0.117
0.119
0.136
0.144
0.117
0.25 in.
23
0.137
0.155
0.158
0.119
0.5 in.
24
0.133
0.150
0.151
0.108
None
25
0.112
0.130
0.137
0.111
0.25 in.
26
0.042
0.121
0.141
0.142
0.113
0.5 in.
27
0.126
0.143
0.144
0.102
STP 1617 On Building Science and the Physics of Building Enclosure Performance
TABLE 2 Table of overall U-factor results (continued)
FIG. 17 Graph of overall U-factor results.
JACKSON ET AL., DOI: 10.1520/STP161720180107
475
476
STP 1617 On Building Science and the Physics of Building Enclosure Performance
We also note that there are some minor variations in the anticipated trends. Specifically, we see that the U-factor increases when adding a 0.25-in. thermal break, where we would expect it to decrease. When increasing the thermal break thickness to 0.5 in., the U-factor decreases compared to the 0.25-in. case, but in some cases is equivalent to having no thermal break. The purpose of adding a thermal break is to separate the metal back pan from the mullions to limit heat flow through mullions. However, note that fasteners and the continuous sealant joint are required and somewhat counteract the benefit of the shims. The differences in Ufactor among these cases are very slight and, given the magnitude of the differences, could potentially be attributed to software error. To study this further, it would be necessary to calibrate these calculation results with physical empirical test data. COMPARING 2-D METHODS VERSUS 3-D METHOD
The 2-D analyses in THERM employ various calculation strategies to approximate 3-D heat transfer effects. When comparing the three 2-D methods (grouping Methods 1, 2, and 3) with the 3-D method, the overall trends (i.e., slope of the lines) are generally the same. As previously discussed, there are some discrepancies when comparing the thermal break cases with the 2-D calculation methods, which are not present with the 3-D method. The thermal break trends for the 3-D method show diminishing returns as expected, where the greatest U-factor decrease comes when adding a 0.25-in.-thick thermal break, and the incremental decrease is less when going to a 0.5-in. thermal break. As previously discussed, further calibration of calculation results with physical empirical testing would be required. Another finding is that when going from 2 ft to 5 ft to 10 ft, the lines cluster closer together. The clustering occurs because the frame effects are more diluted for a 10-ft-tall module compared to a 2-ft-tall module. We also observe that there generally is a point of diminishing returns for each parameter, where spandrel performance could potentially be tied to a spandrel size, aspect ratio, and insulation thickness. That is, if a spandrel is at least a certain size or aspect ratio, the thermal performance could be estimated. This finding could be the basis for a future code requirement where prescriptive code U-factors are established based on spandrel size, aspect ratio, and insulation thickness. As noted in the FENBC/RDH calculation procedure, the position of the slab edge could also have some bearing on prescriptive code requirements for spandrels interrupted by slabs. It is important to note that, for typical projects, the differences in modeling in 2-D versus 3-D may not have a substantial impact on an energy model’s results, particularly if envelope loads are not governing, and therefore may not justify the increased cost and complexity of 3-D modeling compared to the incremental improvement in accuracy (if any). The potential benefit of 3-D analysis could lie in helping to determine parameters by which the industry can define prescriptive Ufactors for spandrel panels that may be adopted by the energy codes and referenced standards in the future. Running multiple 3-D simulations is quicker and less expensive than constructing physical mockups and performing laboratory testing
JACKSON ET AL., DOI: 10.1520/STP161720180107
for the same purpose. Some physical empirical testing is required to verify whether 3-D simulations or perhaps some version of the 2-D methods is most accurate before this can be studied in more detail. THERM VERSUS ANSYS
To better understand the potential reasons for differences between the 2-D and 3-D methods, we need to also explore the differences between the THERM and ANSYS models.13* { In our attempt to produce consistent models in the two software tools, we made some simplifications to save on onerous mesh generation and simulation time,{ as noted here. • Noncontinuous Components: We model noncontinuous thermal bridging elements (e.g., fasteners and shims) according to NFRC Simulation Manual, Section 8.8, for the THERM models, whereas we can explicitly model those components in ANSYS. • IGU Spacer Bar: We model the spacer explicitly in THERM, but in ANSYS, it is modeled as a single solid block based on the effective conductivity calculated from a separate THERM model. • Shell Elements: Metal components (i.e., back pan and mullion) are modeled as thin shell elements in ANSYS, not solids, whereas THERM includes these as solid elements. • Boundary Conditions: The boundary conditions of the models are per ANSI/NFRC 100 but account for another difference between THERM and ANSYS. The interior and exterior temperatures and the exterior film coefficients are the same, using a black-body radiation model. The interior film coefficients again are the same per ANSI/NFRC 100; however, THERM uses an auto-enclosure radiation model, whereas ANSYS uses black-body radiation at the interior.§
* In general, thermal conductivities for all materials in both the THERM and ANSYS models are the same, following ANSI/NFRC 101, Appendix A. This includes air spaces—ANSI/NFRC 100 requires using the ISO 15099 method to calculate air space effective conductivity in THERM, and we assigned the same values to the corresponding ANSYS solid air space blocks. The glazing is also modeled the same in both tools; we used WINDOW to model the insulating glass unit (IGU) and imported it directly into THERM. For ANSYS, we assigned the conductivities from WINDOW to the glass and gas space solid blocks. { It is not within the scope of this paper to quantify each software tool’s range of numerical error. THERM and ANSYS use finite element theory to perform calculations. THERM includes a built-in algorithm to check for and reduce numerical error. ANSYS does not perform a similar check, and so to understand the numerical error due to discretization requires a full mesh refinement study. However, this likely has a negligible impact. We confirmed this by modeling a simple 1-D section (Method 0) in both THERM and ANSYS, with black-body radiation at both the interior and exterior of the model, and we found that the results matched. { Very small or thin elements within a large model increase the number of mesh elements, complicate mesh generation, and greatly increase the simulation time. § The auto-enclosure radiation model uses a view factor calculation based on an algorithm that is built into THERM. The authors were unable to confirm the exact calculation methodology to replicate that algorithm in the ANSYS models.
477
478
STP 1617 On Building Science and the Physics of Building Enclosure Performance
While, individually, each of these simplifications should have a minor impact on overall results, when combined, they may compound and increase the discrepancy among overall results.
Conclusions Based on the findings from our parametric analysis of different calculation methods, we conclude the following: • The 1-D calculation (Method 0) does not account for frame effects and should not be used to approximate spandrel thermal performance. • The 2-D calculation methods (Methods 1, 2, and 3) illustrate good agreement between Methods 2 and 3, recognizing that the 2.5-in. edge zone of Method 1 does not completely capture the impact of frame effects. • When comparing the 2-D calculation methods to the 3-D calculation method (Method 4), the results are closer for the larger-sized modules, in part because frame effects are diluted (area-weighted averages). In addition, discrepancies between results indicate that modeling simplifications may be compounding. • Further model refinement may be needed to more accurately model radiation and study the impact of mesh discretization between 2-D and 3-D modeling techniques. • Physical empirical testing is required to calibrate model results to better determine the accuracy of the various calculation methods. • There is a point of diminishing returns for improvement in overall U-factor for each parameter studied (e.g., module size, insulation thickness, thermal break). FUTURE DEVELOPMENT AND RESEARCH
Based on our review of the energy code and industry standards discussed here and the results of our parametric analysis, the industry is arriving closer to a consensus on how to more accurately calculate overall thermal performance for spandrel assemblies. While codes and standards currently treat spandrel assemblies as opaque elements with an unrealistic prescriptive U-factor, many designers recognize that actual thermal performance is more closely related to a fenestration system than an opaque wall system; and therefore, the U-factor calculation methodology should be similar. We provide the following recommendations for further development and research. • Future versions of the energy codes and referenced standards should include spandrel assemblies as a separate building envelope component with more realistic prescriptive requirements. • Optimization studies are needed to better define the point of diminishing returns for parameters such as spandrel panel size, aspect ratio, and insulation thickness. These could be used to help define “adjustment factors” as a
JACKSON ET AL., DOI: 10.1520/STP161720180107
•
•
way of estimating performance based on those parameters for comparison to prescriptive spandrel panel U-factors in a future energy code or standard. ANSI/NFRC 100 standard sizes should be used for comparing thermal performance among spandrels of different manufacturers and systems only. Project-specific spandrel sizes should be used for demonstrating energy code compliance and for energy modeling purposes. This should be clarified in ANSI/NFRC 100 as well as in building codes and referenced standards. Future versions of ANSI/NFRC 100 should be updated to include explicit U-factor calculation methodology for spandrels, including metal back pans and insulation components, and should also include a method similar to that proposed by FENBC/RDH, with a minimum 6-in. edge zone for spandrels, or thermal modeling to the point of thermal symmetry. Alternatively, a new NFRC standard could be developed for spandrel assemblies, complementary to ANSI/NFRC 100.
ACKNOWLEDGMENTS
We would like to thank the following contributors for their collaborative efforts in the development of this paper: Scott N. Bondi, Vince Cammalleri, Edward G. Lyon, Juhun Lee, Sean M. O’Brien, and Christian Sjoberg at Simpson, Gumpertz & Heger, Inc.; Chrystal Chern at the University of California, Berkeley; and Eric R. Iavarone at The Pennsylvania State University.
References 1. 2.
3. 4. 5. 6.
7. 8.
International Code Council, 2018 International Energy Conservation Code (Washington, DC: ICC, 2018). Energy Standard for Buildings Except Low-Rise Residential Buildings, ASHRAE 90.12016 (Atlanta, GA: American Society of Heating Refrigerating and Air Conditioning Engineers, 2016). North American Fenestration Standard, AAMA/WDMA/CSA 101/I.S.2/A440-11 (Schaumburg, IL: Architectural Aluminum Manufacturer’s Association, 2011). California Energy Commission, 2016 California Energy Code, Title 24 (Sacramento, CA: CEC, 2016). Handbook of Fundamentals (Atlanta, GA: American Society of Heating Refrigerating and Air Conditioning Engineers, 2009) Standard Test Method for Thermal Performance of Building Materials and Envelope Assemblies by Means of a Hot Box Apparatus, ASTM C1363-19 (West Conshohocken, PA: ASTM International, approved September 1, 2019), http://doi.org/10.1520/C1363-19 Procedure for Determining Fenestration Product U-Factors, ANSI/NFRC 100-2017 (Greenbelt, MD: National Fenestration Rating Council, 2017). Voluntary Test Method for Thermal Transmittance and Condensation Resistance of Windows, Doors, and Glazed Wall Sections, AAMA 1503-09 (Schaumburg, IL: Architectural Aluminum Manufacturer’s Association, 2009).
479
480
STP 1617 On Building Science and the Physics of Building Enclosure Performance
9.
10. 11.
12. 13.
Standard Practice for Determining the Thermal Performance Characteristics of Fenestration Systems Installed in Commercial Buildings, AAMA 507-12 (Schaumburg, IL: Architectural Aluminum Manufacturer’s Association, 2012). S. Elsholtz, D. Harvey, and P. Dowsett, “Spandrel Thermal Performance,” OAA Perspectives—The Journal of the Ontario Association of Architects 22, no. 3 (Fall 2014): 38–39. N. Norris, L. Carbary, S. Yee, P. Roppel, and P. Ciantar, “The Reality of Quantifying Curtain Wall Spandrel Thermal Performance: 2D, 3D and Hotbox Testing” (paper presentation, National Institute of Building Science’s Best 4 Conference: Performance Driven Architectural Design, Kansas City, MO, April 13–15, 2015). Fenestration Association of British Columbia, Reference Procedure for Simulating Spandrel U-Factors (and associated User Guide) (Surrey, BC: FENBC, September 22, 2017). Thermal Performance of Windows, Doors and Shading Devices—Detailed Calculations, ISO 15099:2003 (Geneva, Switzerland: International Standards Organization, 2003).
Lemieux | Keegan
ASTM INTERNATIONAL Selected Technical Papers Building Science and the Physics of Building Enclosure Performance
ISBN: 978-0-8031-7680-5 Stock #: STP1617 www.astm.org
STP 1617
ASTM INTERNATIONAL Helping our world work better
Building Science and the Physics of Building Enclosure Performance STP 1617 Editors: Daniel J. Lemieux Jennifer Keegan