INTRODUCTION AND BACKGROUND Roofing systems—particularly in the western part of the United States—are often installed directly to a wood deck, with the building’s insulation below the deck. This construction assembly has a long track record of performance and generally has been used without consequence with traditional built-up roof systems. In recent years, there has been an increase in “conversation” in the roofing industry about anecdotal instances of wood deck deterioration being discovered a relatively short period of time after a roofing system has been installed. Investigations into these occurrences have generally concluded that interior moisture is condensing on the wood deck and causing rot. Hygrothermal modeling of specific roofing systems has confirmed these conclusions. These failures are not being caused by roofing system leaks but by the movement of moisture through the insulation and into the deck and roofing assembly. Unfortunately, these types of failures are becoming more common as the roofing industry (driven by a variety of influences, including regulatory mandate) moves from traditional built-up roofing systems in these markets to reflective membranes—either white or highly reflective “cool”-colored singleply membranes or white reflective cap sheets. The application of reflective coating can also result in this type of failure. In early 2012, Interstate Roofing in Portland, Oregon, approached GAF seeking assistance to help address increasing occurrences of condensation issues in roofing systems, particularly with reflective membranes. One of the commonalities identified that appeared to have a higher incidence of moisture problems was this type of construction assembly (wood deck with belowdeck insulation) in low-rise multifamily residential housing. These buildings often have a high occupancy in a limited space, they lack adequate ventilation below the roof deck, and are occupied by residents who normally generate large amounts of interior moisture. In meeting with Interstate Roofing’s management team, it became apparent that while everyone understood the mechanism(s) leading to condensation issues within roofing systems installed over wood decks, little was available to the roofing contractor to provide guidance on how to avoid potential problems or even to indicate when a roofing contractor or consultant should be aware of the potential for condensation issues. Review of the current literature on this phenomenon offered little practical guidance for roofing professionals. While hygrothermal analysis has provided answers postmortem and could be used to address the needs raised by Interstate Roofing, it had not been used in a large study to further the body of work that would help roofing professionals, given a set of different variables. Based on the need for some practical guidelines, a research project was initiated with Oak Ridge National Laboratory’s (ORNL) Building Envelope Group to model the potential for wood deck deterioration under a variety of conditions and to provide guidelines for the amount of insulation that should be installed over a wood deck to reduce this risk. This modeling took into consideration: • Six different geographic locations representing the climates of the most populous areas on the West Coast • Four different levels of interior moisture loads, from a low relative humidity to a high relative humidity, in compliance with a two-bedroom residence (ASHRAE 160, 2011) • Two different rafter depths (8 and 10 in.) that are in common use • Two different common wood deck types: plywood and oriented strand board (OSB) • Three different levels of below-deck fiberglass insulation thickness (R-11 to R-30) to represent reroofing situations where lower insulation values may be present to situations marked by higher insulation values more in compliance with current energy codes • Two different roofing membranes (one white and one black) to simulate going from a nonreflective to a reflective California Energy Commission Title 24-compliant membrane Using the hygrothermal calculation tool WUFI (Künzel, 1995), a total of 288 different simulations based on the above parameters were performed. Ma r c h 2 0 1 4 I n t e r f a c e • 9 HYGROTHERMAL SIMULATIONS AND THE DEFINITION OF FAILURE WUFI performs onedimensional transient hygrothermal calculations to evaluate the long-time energy and moisture performance and durability of building envelopes, including roofs. WUFI has been successfully validated repeatedly over the past two decades (Kehrer and Schmidt, 2008) and has a large database of material properties and exterior climates from all U.S. climate zones. For the purposes of this study, the moisture content of the wood deck material was the basis of defining “failure.” The 2013 ASHRAE Handbook of Fundamentals (ASHRAE, 2013) states that decay in wood decks typically requires over 30% moisture saturation; thus, for the simulations, “failure” was any set of parameters that resulted in over 30% moisture in the deck in years two to three of the simulation. In addition, those sets of parameters that resulted in 20-30% moisture in the deck were defined as “risky.” These criteria are summarized in Table 1. When performing hygrothermal modeling, three years is the typical simulation period. The logic behind this time selection is that the first year can be impacted by the assumed initial moisture contents of the roofing components. Comparing years two and three will indicate whether the moisture contents of the roofing system components have exceeded threshold values or are increasing at a rate that will eventually lead to failure. While three years is a short period of time relative to the expected performance life of a roofing system, and some conditions could result in saturation of the wood deck in a longer time period, certainly the results presented that indicate failure or risk to the deck provide roofing professionals with those conditions under which caution is necessary and steps to protect the roof deck are likely warranted. SIMULATION PARAMETERS The sites used for this study were San Diego, Los Angeles, San Francisco, and Sacramento in California; Portland, Oregon; and Seattle, Washington. These locations represent a broad array of different climates along the western part of the country, where wood deck construction is prevalent. Additionally, the weather data needed to perform simulations are available for these cities. As found in the results, the varying climates do have a significant effect on the moisture accumulation in wood decks. As anticipated, the amount of belowdeck insulation also was an important factor, because higher levels of insulation result in “colder” wood decks under winter/ cooler temperature conditions, which are conducive to moisture condensation on the wood deck. Because the intent of this study was to provide guidelines for additional insulation thickness above the deck in order to decrease the influence of the exterior climate, the amount of insulation below the deck is critical to the resultant guidelines. Three combinations of rafter dimension and cavity insulation were used for each simulated outdoor climate. A 2×8 rafter with R-11 cavity insulation, a 2×10 rafter with R-11 insulation, and a 2×10 rafter with R-19 insulation were used in the simulated cities of Climate Zone 3. San Francisco, Sacramento, Portland, and Seattle were instead simulated with a 2×8 rafter with R-19 cavity insulation; a 2×10 rafter with R-19 insulation; and, finally, a 2×10 rafter with R-30 insulation. The air cavities created in these constructions are assumed to be unventilated. Since the modeling tool is one-dimensional, all of the simulations are performed at the center of the cavity. Previous experience has shown that this location is most sensitive to hygrothermal damage since the wood studs add additional safe moisture storage. Two different types of deck were use: plywood and OSB with a ½-in. thickness. The roof assemblies were designed with an exterior single-ply membrane with varying solar radiation absorptivity in accordance with either a “cool” white color or a traditional dark surface. For the purpose of this hygrothermal analysis, the properties of the membrane are representative of all major membrane types used in the market today. The solar absorptivities for the white membrane surface and the dark membrane surface were set to 0.35 and 0.90 respec- 1 0 • I n t e r f a c e Ma r c h 2 0 1 4 Maximum Water Content Evaluation Result in Second and Third Years of Simulation Value ≤ 20% “Pass” 20% < Value ≤ 30% “Risk” Value > 30% “Failure” Table 1 – Evaluation criteria for the simulation results. Figure 1 – Simulated annual variations of the moisture content in both the OSB and the plywood decks in the climate of San Diego. Figure 2 – Simulated annual variations of the moisture content in both types of wood decks in the climate of Portland. tively (solar reflectances of 0.65 and 0.10 respectively) and are indicated in the results as “white” or “black” for simplicity. The indoor air humidity was simulated under four different scenarios: with a low, medium, and high moisture load (WTA, 2004), and as an excess of moisture in compliance with a two-bedroom residence (ASHRAE 160, 2011). In all assemblies, an appropriate air space was applied to fulfill the thermal insulation requirements. On the interior side of the roof, a ½-in. gypsum board with a 10-perm latex paint was applied. The roof assemblies were assumed to be airtight and constructed with satisfactory workmanship; i.e., no water leakage due to membrane flaws was simulated. RESULTS Comprehensive information about the study and the simulations conducted has been published in “ORNL TM Report 2013- 551.” The results from the simulations Table 2 – Evaluation of the results from simulating the hygrothermal performance of the wood deck roofing system in the climate of San Diego. Table 3 – Evaluation of the results from simulating the hygrothermal performance of the wood deck roofing system in the climate of Los Angeles. Table 4 – Evaluation of the results from simulating the hygrothermal performance of the wood deck roofing system in the climate of San Francisco. Blindside/Underslab Membrane PRECON is a composite sheet membrane comprised of a nonwoven fabric and elastomeric membrane bonded to an exclusive plasmatic matrix. This product provides a permanent seal between the membrane and poured concrete wall or floor. It helps prevent moisture migrat ion into the st ructure and improves resistance to termites, methane and radon gas. Make PRECON part of your next foundation waterproofing system. © W. R. MEADOWS, INC. 2014 For a solution based on your needs, visit wrmeadows.com or call 1-800-342-5976. PRECON® follow us: WR_Meadows_ Precon_RCI_Interface_2014b_Layout 1 Ma r c h 2 0 1 4 I n t e r f a c e • 1 1 demonstrate that the more northern climates are more sensitive to the radiative properties of the membrane when exposed to high interior moisture loads. Comparison of the simulated annual variations between San Diego and Portland illustrates this result in Figures 1 and 2. Using the evaluation criteria of “pass,” “risk,” or “fail” as defined above, Tables 2-7 summarize the results by parameter for each location. Based on these results, it is clear that the level of indoor moisture plays a significant role in the level of moisture that can accumulate in the deck. This is graphically seen when the results are summarized for the highest moisture level defined in ASHRAE 160 in Table 8. THE USE OF INSULATION TO REDUCE DETERIORATION RISK The intent of this research was to determine the potential for wood deck deterioration and then to provide guidelines for the amount of insulation R-value that should be installed over a wood deck to eliminate this risk for western coastal climates where this construction practice is employed. Therefore, additional simulations were performed to determine the amount of additional insulation necessary to reduce the moisture content of the wood deck after Ma r c h 2 0 1 4 I n t e r f a c e • 1 3 Enhanced Performance Roof Hatches for Green Building Applications For more information call 800-366-6530 LEED is a registered trademark of the or log on to www.bilco.com U.S. Green Building Council Available in a number of standard sizes • Energy effi ciency exceeds standard hatches by more than 48% • Meets LEED® standards for recycled content • High solar refl ective index (SRI) • Ozone-friendly polyisocyanurate insulation • EPDM gasketing with enhanced weather-resistance Our Most Energy-Effi cient Model Yet! BLC625 IF Ad.indd 1 6/25/10 12:11:24 PM Table 5 – Evaluation of the results from simulating the hygrothermal performance of the wood deck roofing system in the climate of Sacramento. Table 6 – Evaluation of the results from simulating the hygrothermal performance of the wood deck roofing system in the climate of Portland. three years to a level below 20% for the excessive interior humidity level in ASHRAE 160. This additional insulation is placed above the wood deck in order to reduce the effects of the below-deck insulation on the deck’s temperature in the hygrothermal modeling. Table 9 summarizes these simulations. Because the simulations performed to determine how much additional R-value is needed to reduce the potential for deck deterioration were based on the highest level of interior moisture of the four levels included in the study, Table 9 is also the most conservative of the results this type of simulation will generate for the input variables chosen in this study. Adding this level of insulation to the other failed scenarios caused by lower interior moisture content levels would mitigate their problems as well. DISCUSSION AND CONCLUSIONS The results of this study indicate that the moisture content of wood decks can be expected to vary greatly under different conditions, with the most influential of conditions being the location of the building (climate) and the interior moisture content. While there have been significant numbers of anecdotal claims that the radiative properties of roofing membranes (e.g., highly reflective) are the single-source causes of wood deck deterioration in the western part of the United States, this study clearly shows that this is not a factually based claim. However, the more reflective a roofing membrane, the higher the risk for moisture accumulation and deck failure when there is higher interior moisture. It is interesting to note that deck type was not a major contributor; however, plywood seems slightly more prone to failure than OSB. When not caused by a roof leak, the propensity for water to accumulate in a roof system and/or the deck is the result of moisture movement via vapor drive from the building interior. In its simplest form, this is explained by the movement of water vapor from the interior up through the vaporpermeable insulation and into the deck when the exterior temperature falls, (e.g., at nighttime and during the colder months of the year). This movement is then reversed when the exterior temperature rises, driving the moisture back into the interior of the building. Understanding that interior water vapor “seeks the cold” explains this movement and also explains why those variables/parameters that raise the temperature at the deck (such as roof color) can reduce the potential for moisture accumulation, as well as why a reduction in the level of interior relative humidity reduces the amount of moisture available to cause problems with the deck and roofing system. While this study showed a low potential for interior moisture-related deck deterioration in mild climates such as San Diego and Los Angeles, the potential for deck deterioration does exist in these climates with high interior humidity and different insulation/ deck constructions. Likewise, in northern climates, even moderate levels of interior humidity can significantly affect the potential for deck deterioration. Each location and circumstance should be evaluated for its own merit. For roofs with below-deck insulation, where the addition of ventilation or a vapor retarder can be difficult, this study provides directional indications that roofing professionals can use to prevent problems in roofing system assemblies where interior moisture levels are suspect. This study shows that the performance of roofing systems can be improved by adding insulation above the deck to raise the temperature of the deck and make it less affected by the outward 1 4 • I n t e r f a c e Ma r c h 2 0 1 4 Table 7 – Evaluation of the results from simulating the hygrothermal performance of the wood deck roofing system in the climate of Seattle. Table 8 – Summary of the evaluation of all the roof assemblies simulated with an excess of indoor moisture according to standard (ASHRAE 160, 2011). movement of moisture. Additionally, when reroofing, if an existing wood deck shows signs of rot or deterioration not associated with a specific roof leak cause, caution should also be taken to consider the potential for moisture migration, and a hygrothermal analysis should be performed to avoid potential problems with the new roofing system. This is particularly important when components of the roofing system are changed, such as the type or color of roofing membrane being installed. Comprehensive information about the study and the simulations conducted has been published in “ORNL/TM Report 2013-551.” It is the authors’ intent to open a dialogue within the industry—not on what has happened when there is deck deterioration with this type of constructed assembly, but how to avoid problems at the outset. Clearly some education of property owners is in order, Ma r c h 2 0 1 4 I n t e r f a c e • 1 5 Table 9 – Required R-values of the insulation boards situated between the wood deck and surface membrane. American Hydrotech, Inc. 303 East Ohio | Chicago, IL 60611 | 800.877.6125 | www.hydrotechusa.com © 2013 Monolithic Membrane 6125 is a registered trademark of American Hydrotech, Inc. Monolithic Membrane 6125, the original rubberized asphalt membrane, has been entrusted with keeping high profile structures across the country and around the globe watertight for 50 years. More than 2 billion square feet of membrane is still performing today as it did the day it was installed. 50 YEARS OF PROVEN PERFORMANCE – MONOLITHIC MEMBRANE 6125® RCI-50-Half-Horiztonal.indd 1 12/11/2012 5:15:41 PM as well as awareness on the part of roofing professionals that several variables can come into play to cause condensation and deck deterioration besides whether the roof has a white membrane or is leaking. REFERENCES ASHRAE 160 (2011). ANSI/ASHRAE Addendum A to Standard 160-2009, Criteria for Moisture-Control Design Analysis in Buildings. Atlanta, GA, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. ASHRAE (2013). Handbook Fundamentals. Chapter 25, “Heat, Air, and Moisture Control in Building Assemblies.” Atlanta, GA, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. Phil Dregger (2012). “‘Cool’ Roofs Cause Condensation – Fact or Fiction?” 2012 International Roof Coatings Conference, Baltimore, MD, IRC. M. Kehrer and T. Schmidt (2008). “Radiation Effects on Exterior Surfaces.” Proceedings of Nordic Symposium on Building Physics, 2008, Copenhagen. H.M. Künzel (1995). “Simultaneous Heat and Moisture Transport in Building Components – One- and Two-Dimensional Calculation Using Simple Parameters.” Dissertation, IRB Verlag. University of Stuttgart. S. Pallin, M. Kehrer, and A. Desjarlais (2013). ORNL/TM-2013/551, “Hygrothermal Performance of West Coast Wood Deck Roofing Systems.” WTA Publications (2004). Simulation of Heat and Moisture Transfer. Guideline 6-2-01/E. Munich, Germany. 1 6 • I n t e r f a c e Ma r c h 2 0 1 4 André Desjarlais is the group leader for the Building Envelopes Research Program at the Oak Ridge National Laboratory (ORNL). He has been involved in building envelope and materials research for over 40 years, first as a consultant and, for the last 23 years, at ORNL. Areas of expertise include building envelope and material energy efficiency, moisture control, and durability. Desjarlais is a member of ASTM and was awarded the title of ASTM Fellow in 2011. He is a member of ASHRAE and served as chairman of Technical Committee 4.4 on Thermal Insulation and Building Systems. William Woodring has been with GAF for 44 years, holding various positions in research and development, marketing, and field technical support in connection with built-up, modified-bitumen, and single-ply roofing, as well as shingles. In his current position as Director of Technical Ser-vices, he is responsible for providing technical information, training, and assistance to contractors in resolving application/installation issues. He is active in RCI, the Asphalt Roofing Manufacturers Association (ARMA), and the Single Ply Roofing Industry (SPRI) association. Helene Hardy Pierce is vice president of technical services, codes, and industry relations for GAF Materials Corporation. She has spent over 33 years in the roofing industry and has been very active in many of the industry’s organizations. Pierce received the ASTM Award of Merit and the title of Fellow from ASTM Committee D08, the James Q. McCawley Award from the Midwest Roofing Contractors Association, and the title of Fellow of RCI. She is secretary of the RCI Foundation U.S. Helene Hardy Pierce Simon Pallin is a postdoctoral associate at Building Technologies Research and Integration Center at the Oak Ridge National Laboratory. He obtained his undergraduate degree in civil engineering and a master’s in structural engineering from Chalmers University of Technology in Sweden (while also studying at California State University of Long Beach). He received his PhD from Chalmers in the spring of 2013 through his work in risk assessment of energy and moisture performance in residential buildings. Simon Pallin André Desjarlais William Woodring ASTM International Committee D08 on Roofing and Waterproofing has issued a call for papers for its Eighth Symposium on Roofing Research and Standards Development, which will be held December 6, 2015, in Tampa, Florida. The symposium will be held in conjunction with the committee’s standards development meetings. Potential authors should submit a 250- to 300-word abstract by April 8, 2014. Authors will be notified of accepted papers for presentation by September 8. For more information, visit www.astm.org/D08RoofingCFP12-2015. Call for Papers by ASTM D08
