Rainwater Management Performance of Newly Constructed Residential Building Enclosures During August and September 2004

Transcription

1 Rainwater Management Performane of Newly Construted Residential Building Enlosures During August and September 2004 Prepared for the Home Builders Assoiation of Metro Orlando and the Florida Home Builders Assoiation By: Joseph William Lstiburek, Ph.D., P.Eng. Building Siene Corporation 70 Main Street Westford, MA January 11, 2005

2 Table of Contents Table of Contents 1 Exeutive Summary 2 Bakground and Approah 6 Performane of Stuo Claddings 7 Events of August and September Performane Issues With Water Resistant Barriers 13 Reverse Flashing Problems 21 Performane of Windows and Doors 27 Performane of Servie Penetrations 33 Performane of Soffit Vent Assemblies 39 Paint and Coating Performane 43 Review of Florida Building Code 61 Reommendations 64 Referenes 65 Page 1 of 66 Building Siene Corporation 1

3 Exeutive Summary Building Siene Corporation was engaged by the Home Builders Assoiation of Metro Orlando and the Florida Home Builders Assoiation to review the performane of residential assemblies in the entral Florida (Orlando) area during the three hurrianes in August and September The speifi fous of the review was on the water management details assoiated with stuo laddings. Two types of stuo laddings were reviewed: traditional three oat hard oat stuo and ementitious deorative finishes. Both are renderings applied to substrates and the fundamental physis apply to both equally. Stuo laddings leak (as do all laddings). When stuo laddings leak the penetrating water is traditionally managed in two fundamental ways: the water is direted to a water resistant barrier (WRB) and direted downwards and out of the building assembly; or the water is absorbed in a non water sensitive material, redistributed and released to the interior and exterior in a ontrolled way The first method is used with frame wall assemblies and the traditional WRB is building paper. The seond method is used with masonry blok onstrution. Both methods are ommon in the entral Florida (Orlando) area. The first method is typially limited to the seond floor and gable roof assemblies. The seond method is the standard first floor wall onstrution of the majority of homes onstruted in the region. The seond floor assemblies are drained assemblies and drain into the first floor assemblies. The first floor assemblies are mass assemblies where penetrating rainwater is stored in the mass of the onrete blok until it is released to both the interior and exterior during drying periods. The performane of a mass assembly is based on a rate-storage relationship. When the rate of wetting exeeds the rate of drying aumulation ours. Aumulation of water an our until the quantity of aumulated moisture exeeds the moisture storage apaity of the assembly. The moisture storage apaity is time, temperature and material speifi. Under normal onditions the amount of penetrating rainwater through stuo into a masonry blok wall is minor and easily absorbed, redistributed and released to both the interior and exterior. August and September 2004 was not a normal time. The mass assemblies were overwhelmed due to the extraordinary weather events. The mass assemblies were not able store the quantity of penetrating water and not able to dry rapidly enough between wetting events and in many situations water entered past the interior lining. The seond floor frame assemblies provided mixed performane. In many ases the seond floor assemblies were also overwhelmed prinipally for two reasons: drainage was poor due to the failure of plasti housewraps and other WRB systems to provide drainage and water holdout drained rainwater was not expelled to the exterior at the base of the seond floor frame assemblies The performane of the seond floor frame assemblies is also based on a rate-storage relationship. However, unlike the mass assemblies, very little moisture storage apaity is available. As suh for the seond floor frame assemblies the rate of drying must math or exeed the rate of wetting. The key drying method in the seond floor frame assemblies is drainage. This drainage depends on the provision of a drainage spae between the stuo rendering and WRB and the water repelleny of the WRB. Additionally, the drainage depends on the draining water being expelled to the exterior at the base of the frame assembly. Page 2 of 66 Building Siene Corporation 2

4 In the mass assemblies water penetrated the stuo via miro raks (as the water also did in the frame assemblies). Typial paint finishes are unable to span miro raks. Under normal onditions this is not an issue for the reasons previously mentioned (the huge moisture storage apaity of masonry blok assemblies). As stuo buildings age and are suessively repainted the water entry is redued after eah layer of paint is added. In general this is why many older buildings onstruted with mass walls performed somewhat better. In the seond floor frame assemblies water also penetrated the stuo via miro raks. Again, as previously mentioned, typial paint finishes are unable to span miro raks. In frame wall assemblies it is expeted that this penetrating water will be drained bak to the exterior. However, in many ases the penetrating rainwater was not drained to the exterior due to adhesion between WRB s and the stuo renderings preventing drainage between the stuo renderings and WRB s, a loss of water repelleny of the WRB s and the lak of effetive flashing at the base of the drained assemblies. There appear to be signifiant performane issues with WRB s relating to manufature, testing and approval. All plasti housewraps and some building papers tend to bond to stuo renderings thereby negating drainage. Additionally, many plasti housewraps loose their water repelleny when in ontat with sheathing and stuo renderings. The Florida Building Code in Setion sets out the requirements for WRB s and stuo renderings applied over wood sheathing. Setion reads as follows: Moisture Barrier. Wood shall be overed with 15-1b roofing felt, or other approved equally moisture-resisting layer, and metal reinforement as set forth herein. Unfortunately, equally moisture-resisting is not expliitly defined in the ode by referene to any ASTM standards. In fat, some of the problems identified in our review are not urrently addressed in any existing ASTM standard, partiularly the water hold out of WRB s under the influene of surfatants, or the loss of water repelleny due to adhesion, or the drainage harateristis of stuo renderings applied over WRB s. There are signifiant ode enforement issues regarding the onstrution of flashing at the base of the seond floor frame assemblies. Code offiials are enforing a provision of ASTM C1063 (Standard Speifiation for Installation of Lathing and Furring to Reeive Interior and Exterior Portland Cement-Based Plaster) that is not intended to apply where seond floor frame assemblies interset first floor mass assemblies. For reasons that are not expliable or defensible reverse flashing is mandated rather than drainage to the exterior at this loation. The water management of penetrations and openings in stuo laddings were also reviewed partiularly window and door openings, and servie penetrations. With respet to window and door openings it is instrutive to realize that windows and doors installed in residential buildings in Florida are rated for water holdout at 15 perent of the design wind load and no lower than 140 Pa or approximately 35 mph. These servie limits were learly exeeded many times during August and September Windows and doors are expeted to leak during hurrianes; they are not expeted to blow out. However, our testing indiates that many windows leaked at onditions well below their rated value i.e. under no wind ondition. In a similar vein wind driven rain also entered dryer vent openings, eletrial servie panels, and bathroom fan vent openings. But, again, our testing indiates that many servie penetrations leaked in the absene of wind pressures. The method of testing windows is speified in ANSI/AAMA 101 speifially ASTM E 547. The ASTM standard learly views leakage at orners to onstitute failure, however, it is our ontention that the mounting frames used by many window testing groups in the testing of windows for ompliane to the ASTM standard obsure the view of orners and subsequently many windows are listed as passing this test when in fat they fail. Additionally, many windows are sold as mulled or double windows or omposite windows. However, they are tested as single units. Every mulled window unit tested leaked with no applied wind load. Page 3 of 66 Building Siene Corporation 3

5 Finally, with respet to windows and doors, it is our ontention that the installation instrutions regarding window and door installation are inadequate with respet to water management. The windows and doors themselves under the Florida Building Code are subjet to an ASTM standard. The interfae between the window and door and the wall assembly is urrently not. Servie penetrations suh as dryer vents, eletrial panel boxes, eletrial boxes, vent fan hoods, and roof vents are urrently not rated or designed for wind driven rain. Anedotal evidene indiates that a signifiant amount of rainwater entered soffit vent assemblies. This is onsistent with the physis of the applied wind loads and the geometry of the soffit assemblies. Soffit geometries are urrently not designed to address extreme wind driven rain exposures. Additionally, unvented roof designs whih an address this mode of rainwater entry are urrently not expliitly allowed in the Florida Building Code although they are allowed in the International Residential Code. The use of paint as a water management tehnique for stuo renderings applied to mass assemblies was also examined. As the mil thikness of paint inreases, the ability of some paints to span miro raks also inreases. However, this applies primarily to mostly smooth surfaes. Highly textured surfaes are almost impossible to oat in a manner to seal miro raks. As the mil thikness of paint oatings inreases, the water vapor permeability of these oatings dereases leading to problems with blistering and re-emulsifiation of some stuo renderings. The appropriate mil thikness and water vapor permeability relationship is urrently unknown. Finally, as part of our review of the water management provisions of the Florida Building Code we identified a provision relating to roof membranes, Setion , that appears to make no sense: If the underlayment is a selfadhering membrane, the membrane shall be applied over a mehanially attahed anhor sheet, attahed in ompliane with This provision enourages roof membrane blow off by requiring a bond break between the selfadhering membrane and the strutural roof dek. This negates the entire reason to use a self-adhering membrane. The requirement for a mehanially attahed anhor sheet under a self-adhering membrane has no basis in physis or logi and prevents the onstrution of highly water resistant roof assemblies that an provide a high degree of seondary protetion during extreme wind driven rain events. The following reommendations are offered: The moisture storage apaity of mass walls be inreased by providing a seat at the base of these assemblies. A bond break be provided between primary drainage planes and stuo renderings in drained assemblies. In simple terms this will require two layers of building paper or a layer of building paper over a plasti housewrap. The speifiation, rating and testing of WRB s be onsistent with their installed exposure i.e. tested and rated as part of a stuo assembly. Appropriate performane speifiations need to be developed for WRB s used with stuo renderings and the Florida Building Code altered to require them. Code offiials be instruted regarding the orret interpretation of ASTM C1063 and the Florida Building Code be expliitly altered to require drainage where drained assemblies interset mass assemblies. Windows and doors be orretly rated and tested aording to ANSI/AAMA 101. Mulled window units, double windows or omposite windows be tested and held to the same requirements as single units. Water managed window and door installation requirements be developed and the Florida Building Code altered to require them. Page 4 of 66 Building Siene Corporation 4

6 Pressure relieved/baffled soffit assemblies be developed for vented roof assemblies and the Florida Building Code altered to require them. The Florida Building Code be altered to ome into ompliane with the International Residential Code to expliitly allow for the onstrution of unvented roof assemblies. Water managed details for dryer vents, eletrial panel boxes, eletrial boxes, vent fan hoods be developed and the Florida Building Code Altered to require them. It is unlikely that a pratial paint speifiation an be developed in the short term to address miro-raking stuo issues as the relationships among water vapor permeability, mil thikness and elastiity are not known. It is reommended that these relationships be explored and that until these relationships are understood the Florida Building Code not be altered to require elastomeri paints on stuo renderings. Repeal of Setion of the Florida Building Code requiring a mehanially attahed anhor sheet between a self-adhering membrane and roof sheathing. Page 5 of 66 Building Siene Corporation 5

7 Bakground and Approah Building Siene Corporation was engaged by the Home Builders Assoiation of Metro Orlando and the Florida Home Builders Assoiation to review the performane of residential assemblies in the entral Florida (Orlando) area during the three hurrianes in August and September The speifi fous of the review was on the water management details assoiated with stuo laddings. The performane of stuo laddings was alled into question after the events of August and September 2004, speifially the effet of rain from hurrianes Charley, Franes and Jeanne. The review was onduted through field investigation, field testing and experimentation, mok-up testing and experimentation, benh top testing, review of literature, review of odes and standards, and interviews with builders, ontrators, materials suppliers, manufaturers, and ode offiials. Page 6 of 66 Building Siene Corporation 6

8 Performane of Stuo Claddings The speifi fous of the review is on the water management details assoiated with stuo laddings. Two types of stuo laddings were reviewed: traditional three oat hard oat stuo and ementitious deorative finishes. Both are renderings applied to substrates and the fundamental physis apply to both equally. Stuo laddings leak (as do all laddings). ASTM C a Standard Speifiation for Appliation of Portland Cement-Based Plaster reognizes this: Resistane to rain penetration is improved where plaster has been adequately densified during appliation and properly ured. Plaster shall not, however, be onsidered to be waterproof. When stuo laddings leak the penetrating water is traditionally managed in two fundamental ways: the water is direted to a water resistant barrier (WRB) and direted downwards and out of the building assembly; or the water is absorbed in a non water sensitive material, redistributed and released to the interior and exterior in a ontrolled way The first method is used with frame wall assemblies and the traditional WRB is building paper (Figure 1). The seond method is used with masonry blok onstrution (Figure 2). Both methods are ommon in the entral Florida (Orlando) area. The first method is typially limited to the seond floor and gable roof assemblies (Photograph 1). The seond method is the standard first floor wall onstrution of the majority of homes onstruted in the region. The seond floor assemblies are drained assemblies and drain into the first floor assemblies (Figure 3). The first floor assemblies are mass assemblies where penetrating rainwater is stored in the mass of the onrete blok until it is released to both the interior and exterior during drying periods. The performane of a mass assembly is based on a rate-storage relationship. When the rate of wetting exeeds the rate of drying aumulation ours. Aumulation of water an our until the quantity of aumulated moisture exeeds the moisture storage apaity of the assembly. The moisture storage apaity is time, temperature and material speifi. Under normal onditions the amount of penetrating rainwater through stuo into a masonry blok wall is minor and easily absorbed, redistributed and released to both the interior and exterior. Page 7 of 66 Building Siene Corporation 7

9 Figure 1: Drained Assembly Figure 2: Mass Assembly Page 8 of 66 Building Siene Corporation 8

10 Figure 3: Seond Floor Drained Assembly Drains Into First Floor Mass Assembly buildingsiene.om Photograph 1: Typial Constrution buildingsiene.om e.omm Page 9 of 66 Building Siene Corporation 9

11 Events of August and September 2004 August and September 2004 was not a normal time. Four hurrianes hit the State of Florida, and three of the hurrianes hit the Orlando region (Figure 4, Figure 5, Figure 6 and Figure 7). Unpreedented quantities of rain fell during high wind events. The mass assemblies were overwhelmed due to the extraordinary weather events. The mass assemblies were not able store the quantity of penetrating water and not able to dry rapidly enough between wetting events and in many situations water entered past the interior lining (Figure 8). The seond floor frame assemblies provided mixed performane. In many ases the seond floor assemblies were also overwhelmed (Photograph 2) prinipally for two reasons: drainage was poor due to the failure of plasti housewraps and other WRB systems to provide drainage and water holdout drained rainwater was not expelled to the exterior at the base of the seond floor frame assemblies The performane of the seond floor frame assemblies is also based on a rate-storage relationship. However, unlike the mass assemblies, very little moisture storage apaity is available. As suh for the seond floor frame assemblies the rate of drying must math or exeed the rate of wetting. The key drying method in the seond floor frame assemblies is drainage. This drainage depends on the provision of a drainage spae between the stuo rendering and WRB and the water repelleny of the WRB. Additionally, the drainage depends on the draining water being expelled to the exterior at the base of the frame assembly. In the mass assemblies water penetrated the stuo via miro raks (as the water also did in the frame assemblies). Typial paint finishes are unable to span miro raks. Under normal onditions this is not an issue for the reasons previously mentioned (the huge moisture storage apaity of masonry blok assemblies). As stuo buildings age and are suessively repainted the water entry is redued after eah layer of paint is added. In general this is why many older buildings onstruted with mass walls performed somewhat better. In the seond floor frame assemblies water also penetrated the stuo via miro raks. Again, as previously mentioned, typial paint finishes are unable to span miro raks. In frame wall assemblies it is expeted that this penetrating water will be drained bak to the exterior. However, in many ases the penetrating rainwater was not drained to the exterior due to adhesion between WRB s and the stuo renderings preventing drainage between the stuo renderings and WRB s, a loss of water repelleny of the WRB s and the lak of effetive flashing at the base of the drained assemblies. Page 10 of 66 Building Siene Corporation 10

12 Figure 4: Hurriane Charley (Aug 9-14, 2004) Figure 5: Hurriane Franes (Aug 25-Sept 25, 2004) buildings siene.om Figure 6: Hurriane Ivan (Sept 2-24, 2004) Figure 7: Hurriane Jeanne (Sept 13-28, 2004) buildings siene.om Page 11 of 66 Building Siene Corporation 11

13 Figure 8: Rainwater Entry into Mass Assembly buildingsiene..om Photograph 2: Rainwater Entry into Seond Floor Frame Assembly buildingsiene.om Page 12 of 66 Building Siene Corporation 12

14 Performane Issues With Water Resistant Barriers There appear to be signifiant performane issues with WRB s relating to manufature, testing and approval. All plasti housewraps and some building papers tend to bond to stuo renderings thereby negating drainage (Photograph 3). Additionally, many plasti housewraps loose their water repelleny (Figure 9) when in ontat with sheathing and stuo renderings (Photograph 4). The Florida Building Code in Setion sets out the requirements for WRB s and stuo renderings applied over wood sheathing. Setion reads as follows: Moisture Barrier. Wood shall be overed with 15-1b roofing felt, or other approved equally moisture-resisting layer, and metal reinforement as set forth herein. Unfortunately, equally moisture-resisting is not expliitly defined in the ode by referene to any ASTM standards. In fat, some of the problems identified in our review are not urrently addressed in any existing ASTM standard, partiularly the water hold out of WRB s under the influene of surfatants, or the loss of water repelleny due to adhesion, or the drainage harateristis of stuo renderings applied over WRB s. All perforated plasti housewraps fail a simple test for water repelleny and water holdout (Photograph 5 and Photograph 6). Many non perforated housewraps fail a similar test when exposed to surfatants (Photograph 7, Photograph 8 and Photograph 9). The problem of adhesion of stuo renderings to plasti housewraps and to building papers an be address by installing a bond break (Figure 10). Tests showed that a thin layer of asphalt saturated felt ats as an effetive bond break for both building papers (Photograph 10) and plasti housewraps (Photograph 11). The felt used is similar to the baking on paper baked lath (Photograph 12). The tehnique of using two layers of building paper is typial in other regions in the United States and signifiant trade literature exists delineating the pratie (Figure 11). Drainage between the bond break and primary drainage plane is signifiantly enhaned when a drainage spae is provided suh as through the use of a textured or profiled drainage plane (Photograph 13). A test for drainage should be developed and performane requirements established. Page 13 of 66 Building Siene Corporation 13

15 Figure 9: Water Repelleny buildingsiene.om Photograph 3: Stuo Rendering Bonding to Building Paper buildingsiene.om Page 14 of 66 Building Siene Corporation 14

16 Photograph 4: Stuo Rendering Bonding to Plasti Housewrap buildingsiene.o gsiene.omm Photograph 5: Water Drop Test For Water Repelleny buildingsiene..om Page 15 of 66 Building Siene Corporation 15

17 Photograph 6: Plasti Housewrap Failing Water Drop Test buildingsiene. ingsiene.om Photograph 7: Absorptive Surfae For Water Drop Test buildingsiene e.om Page 16 of 66 Building Siene Corporation 16

18 Photograph 8: Plasti Housewrap Undergoing Water Drop Test buildingsiene.om Photograph 9: Extensive Failure of Plasti Housewrap During Water Drop Test buildingsiene iene.om Page 17 of 66 Building Siene Corporation 17

19 Figure 10: Bond Break Best Pratie For Stuo Renderings Applied Over Frame Wall Assemblies m o. e n e i ings d l i u b Photograph 10: Asphalt Saturated Felt As Bond Break Over Building Paper m o. iene s g n i d l i u b Building Siene Corporation 18 Page 18 of 66

20 Photograph 11: Asphalt Saturated Felt As Bond Break Over Plasti Housewrap buildingsiene.om m Photograph 12: Paper Baked Lath As A Bond Break buildingsiene. e.om Page 19 of 66 Building Siene Corporation 19

21 Figure 11: Trade Literature Delineating Two Layers of Building Paper m o. e n e i ings d l i u b Photograph 13: Enhaned Drainage With A Profiled Plasti Housewrap m o. iene s g n i d l i u b Building Siene Corporation 20 Page 20 of 66

22 Reverse Flashing Problems There are signifiant ode enforement issues regarding the onstrution of flashing at the base of the seond floor frame assemblies. Code offiials are enforing a provision of ASTM C1063 (Standard Speifiation for Installation of Lathing and Furring to Reeive Interior and Exterior Portland Cement-Based Plaster) that is not intended to apply where seond floor frame assemblies interset first floor mass assemblies. For reasons that are not expliable or defensible reverse flashing is mandated rather than drainage to the exterior at this loation. The typial method of installation is to extend the building paper downwards over the joint between the frame assembly and the mass assembly (Photograph 14). The building paper ats as a bridge flashing between the two assemblies. Control joint aessories are not typially designed as water shedding devies i.e. flashings (Photograph 15). They are designed in most ases to be installed over the top of building paper rather than being tuked behind building paper in a shingle or water shedding fashion. Standard appliation has these ontrol joint aessories installed over the top to the building paper bridge flashing (Photograph 16). This reates a dam and water is direted inwards. This pratie results in leaks during wind driven rain events (Photograph 17) as the moisture storage apaity of the frame wall assembly is easily overwhelmed in the absene of drainage. The seond floor frame assembly, whih is a drained assembly is effetively unable to drain anywhere exept to the interior. The appropriate pratie is to use a flashing between floors that effetively drains the water to the exterior (Figure 12). This detail is desribed in detail in Photograph 18, Photograph 19, Photograph 20, Photograph 21 and Photograph 22. Page 21 of 66 Building Siene Corporation 21

23 Photograph 14: Building Paper As Bridge Flashing buildingsiene dingsiene.om Photograph 15: Control Joint Aessories Installed Over Top of Drainage Plane buildingsiene e.om Page 22 of 66 Building Siene Corporation 22

24 Photograph 16: Reverse Flashing Between Floors buildingsiene dingsiene.om Photograph 17: Water Entry Between Floors buildingsiene.om Page 23 of 66 Building Siene Corporation 23

25 Figure 12: Best Pratie Drained and Flashed Control Joint buildingsiene..om Photograph 18: Flashing/Control Joint - Weep Sreed Flashing buildingsiene.o om Page 24 of 66 Building Siene Corporation 24

26 Photograph 19: Self Adhering Membrane Bridge Flashing buildingsiene dingsiene.om Photograph 20: Primary Drainage Plane Lapped Over Control Joint Flashing buildingsiene e.om Page 25 of 66 Building Siene Corporation 25

27 Photograph 21: Paper Baked Lath Bond Break Over Primary Drainage Plane buildingsiene dingsiene.om Photograph 22: Stuo Rendering Applied At Draining Control Joint buildingsiene e.om Page 26 of 66 Building Siene Corporation 26

28 Performane of Windows and Doors The water management of penetrations and openings in stuo laddings were also reviewed partiularly window and door openings. With respet to window and door openings it is instrutive to realize that windows and doors installed in residential buildings in Florida are rated for water holdout at 15 perent of the design wind load and no lower than 140 Pa or approximately 35 mph. These servie limits were learly exeeded many times during August and September Windows and doors are expeted to leak during hurrianes; they are not expeted to blow out. However, our testing indiates that many windows leaked at onditions well below their rated value i.e. under no wind ondition (Photograph 24 and Photograph 25). The method of testing windows is speified in ANSI/AAMA 101 speifially ASTM E 547. The ASTM standard learly views leakage at orners to onstitute failure, however, it is our ontention that the mounting frames used by many window testing groups in the testing of windows for ompliane to the ASTM standard obsure the view of orners and subsequently many windows are listed as passing this test when in fat they fail. Additionally, many windows are sold as mulled or double windows or omposite windows. However, they are tested as single units. Every mulled window unit tested leaked with no applied wind load (Photograph 26). Visual observations of hundreds of window units in the field and the disassembly of a dozen indiate a problem with the window sill to window jamb onnetion. It is our belief that quality ontrol problems exist with the manufaturing of window units prior to delivery to the site. The joint geometry is problemati in general the joint design an be signifiantly improved. The typial design makes it a very diffiult joint to seal, diffiult to inspet and results in an unneessarily high failure rate. Finally, with respet to windows and doors, it is our ontention that the installation instrutions regarding window and door installation are inadequate with respet to water management. The windows and doors themselves under the Florida Building Code are subjet to an ASTM standard. The interfae between the window and door and the wall assembly is urrently not. Many mass wall assembly window openings are onstruted with preast sill units that have a raised rib attahment that direts window water leakage inwards rather than outwards (Photograph 27, Photograph 28, and Figure 13). Figure 14 illustrates a preast sill profile that direts window water leakage outward. A similar approah is reommended for window openings in frame wall assemblies (Photograph 29). Door leakage an also be address in a similar fashion a seat is ast into the slab edge that ats as a pan flashing (Photograph 30). Page 27 of 66 Building Siene Corporation 27

29 Photograph 23: Window Openings buildingsiene. ngsiene.omm Photograph 24: Window Corner Leakage buildingsiene. e.om Page 28 of 66 Building Siene Corporation 28

30 Photograph 25: Window Unit Leaking Inboard of Preast Attahment Rib m o. e n e i ings d l i u b Photograph 26: Mulled Window Unit Leakage Paths m o. iene s g n i d l i u b Building Siene Corporation 29 Page 29 of 66

31 Photograph 27: Preast Sill With Illogial Attahment Rib buildingsiene.o om Photograph 28: Preast Sill That Channels Window Leakage Inward buildingsiene e.om Page 30 of 66 Building Siene Corporation 30

32 Figure 13: Current Pratie That Direts Window Leakage Inward buildingsiene. ingsiene.om Figure 14: Reommended Pratie That Direts Window Leakage Outward buildingsien e.om Page 31 of 66 Building Siene Corporation 31

33 Photograph 29: Frame Assembly Opening Reommended Pratie buildingsiene.om m Photograph 30: Sliding Door Unit In Seat in Slab Direting Door Leakage Outward buildingsiene.om Page 32 of 66 Building Siene Corporation 32

34 Performane of Servie Penetrations In a similar vein wind driven rain also entered dryer vent openings, eletrial servie panels, and bathroom fan vent openings. But, again, our testing indiates that many servie penetrations leaked in the absene of wind pressures. Servie penetrations suh as dryer vents, eletrial panel boxes, eletrial boxes, vent fan hoods, and roof vents are urrently not rated or designed for wind driven rain. Eletrial panel boxes are typially installed without water protetion behind them (Photograph 31). This is typially not a major issue as these panel boxes are loated on garage walls. However, it is reommended that the stuo rendering be applied prior to the installation of the eletrial panel box. Most openings rely on the appliation of sealant rather than a flashing approah to ontrol rain entry (Photograph 32, Photograph 33 and Photograph 34). Sealants require maintenane. Using a flashing approah to address servie penetrations will provide improved performane (Photograph 35). Photograph 36 and Photograph 37 illustrate the use of flashing for eletrial boxes a premanufatured unit provides water holdout during wind driven rain events. Photograph 38, Photograph 39 and Photograph 40 illustrate effetive kik out flashing onstrution. Page 33 of 66 Building Siene Corporation 33

35 Photograph 31: Eletrial Servie Panel No Water Control Behind Panel buildingsiene.om Photograph 32: Typial Maintenane Intensive Servie Opening buildingsiene e.om Page 34 of 66 Building Siene Corporation 34

36 Photograph 33: Maintenane Intensive Servie Opening buildingsiene.om Photograph 34: Dryer Vent Opening buildingsiene iene.om Page 35 of 66 Building Siene Corporation 35

37 Photograph 35: Reommended Pratie For Dryer Vent Opening m o. e n e i ings d l i u b Photograph 36: Flashed Eletrial Box m o. iene s g n i d l i u b Building Siene Corporation 36 Page 36 of 66

38 Photograph 37: Eletrial Box Flashing Unit buildingsiene.omm Photograph 38: Reommended KikOut Flashing Detail buildingsiene. ne.om Page 37 of 66 Building Siene Corporation 37

39 Photograph 39: Kik Out Flashing buildingsiene.omm Photograph 40: Masti Adhered Kik Out Flashing On Mass Assembly buildingsiene. ingsiene.om Page 38 of 66 Building Siene Corporation 38

40 Performane of Soffit Vent Assemblies Anedotal evidene indiates that a signifiant amount of rainwater entered soffit vent assemblies. This is onsistent with the physis of the applied wind loads and the geometry of the soffit assemblies. Soffit geometries are urrently not designed to address extreme wind driven rain exposures. Additionally, unvented roof designs whih an address this mode of rainwater entry are urrently not expliitly allowed in the Florida Building Code although they are allowed in the International Residential Code. Figure 15 illustrates rainwater entry during wind driven rain events. Figure 16 illustrates the use of baffles to redue air pressure driving fores and failitate the use of air pressure hanges to deposit rain in soffit assemblies rather than in atti spaes. Photograph 41, Figure 17, Figure 18 and Figure 19 illustrate unvented onditioned atti onstrution. This tehnology has signifiant advantages in the Florida limate with respet to rainwater ontrol, energy onservation, moisture, and humidity ontrol, wind uplift and fire performane over standard vented atti roof tehnology. Page 39 of 66 Building Siene Corporation 39

41 Figure 15: Wind Driven Rain Entry In Vented Soffit Assembly buildingsiene.om Figure 16: Pressure Relieved Baffled Soffit Assembly buildingsiene.om Page 40 of 66 Building Siene Corporation 40

42 Photograph 41: Spray Foam Insulation In Unvented Conditioned Atti Assembly buildings iene.om Figure 17: Typial Vented Atti With Dutwork Outside of Conditioned Spae buildingsiene.om Page 41 of 66 Building Siene Corporation 41

43 Figure 18: Negative Pressure Indued Via Dut Leakage m o. e n e i ings d l i u b Figure 19: Conditioned Unvented Atti Assembly m o. iene s g n i d l i u b Building Siene Corporation 42 Page 42 of 66

44 Paint and Coating Performane The use of paint as a water management tehnique for stuo renderings applied to mass assemblies was also examined. As the mil thikness of paint inreases, the ability of some paints to span miro raks also inreases. However, this applies primarily to mostly smooth surfaes. Highly textured surfaes are almost impossible to oat in a manner to seal miro raks. As the mil thikness of paint oatings inreases, the water vapor permeability of these oatings dereases leading to problems with blistering and re-emulsifiation of some stuo renderings. The appropriate mil thikness and water vapor permeability relationship is urrently unknown. Water leakage testing of stuo renderings was onduted using numerous methods. These methods are memorialized in Photograph 42, Photograph 43, Photograph 44, Photograph 45, Photograph 46, Photograph 47, Photograph 48 and Photograph 49. The most useful method proved to be the use of a amera probe and a portable spray rak without the appliation of a pressure differene either air or hydrostati. Test assemblies were onstruted and baseline values were established. Stuo renderings, both painted and unpainted were tested to determine water hold out (Photograph 50, Photograph 51, Photograph 52, Photograph 53, Photograph 54, Photograph 55, Photograph 56 and Photograph 57). Three-hour water spray tests on unpainted walls systems rendered with stuo did not leak in the absene of raks. To state the obvious, stuo without raks does not leak paint or no paint (Photograph 58 and Photograph 59). However, the presene of raks (Photograph 60 and Photograph 61) aused all mass wall assemblies to pass water. The type of rak shown in Photograph 60 is indiative of typial and expeted stuo rendering rakage. Workmanship, quality ontrol and ure impat the number and extent of shrinkage raking. Soil onditions, the nature of the materials, geometry and aspet ratio of mass wall assemblies impat the number and extent of settlement raking. However, the nature of stuo appliation, materials and substrates make it impossible to onstrut rak free monolithi stuo renderings in other words shrinkage raks and settlement raks are to be expeted. These shrinkage raks and settlement raks traditionally are handled through ongoing maintenane. At present, tehnology offers no other pratial approah. Based on the field observations it is our belief that it is not possible to onstrut stuo assemblies without raks. It is also our belief that paint oating systems even the high build oatings - are unable to span the typial stuo raks enountered, both from initial drying and from subsequent settlement. Inreasing the moisture storage apaity of mass assemblies an partially address this issue. One approah is to onstrut a seat in the perimeter slab foundation to provide a reservoir for penetrating rainwater and to diret this rainwater to the exterior (Figure 20, Figure 21 and Photograph 62). The experiene with high build paint systems suh as elastomeri paints an lead to blistering (Photograph 63) and re-emulsifiation of additives. The use of suh oatings should remain a speialty tehnique speified and supervised by professionals. The inward migration of moisture an also be moderated by the use of non water sensitive insulation systems (Figure 22, Photograph 64 and Photograph 65). Page 43 of 66 Building Siene Corporation 43

45 Low build and high build paint system on both smooth and textured surfaes were arefully examined (Photograph 66, Photograph 67 and Photograph 68) and it is our onlusion that almost no oating or painting system is able to span settlement raks (Photograph 69). Only small hairline raks on relatively smooth, and relatively untextured surfaes are there likely to be signifiant performane benefits of high build systems. A more promising approah appears to lie with fiber reinforement to ontrol raking and premium polymer modified and polymer based ementitious renderings (Photograph 70). Most shrinkage raking ours over the first few weeks and most settlement raking happens over the first year. In other words the majority of the movement resulting in stuo raks happens within the first year. After the building and building systems have equilibrated standard paint finishes are often able to seal small raks as subsequent movement is typially minor. However, the larger settlement raks require flexible sealants typially brushed into the rak. After buildings are repainted and settlement raks are addressed in the normal ourse of home maintenane stuo leakage is typially redued signifiantly. Page 44 of 66 Building Siene Corporation 44

46 Photograph 42: Magnifying Glass Examination buildingsiene. e..om Photograph 43: Rilem Tube Hydrostati Head Test buildingsiene.om Page 45 of 66 Building Siene Corporation 45

47 Photograph 44: Infrared Camera Leak Detetion buildingsiene.om Photograph 45: Camera Probe Leak Detetion buildingsiene.om Page 46 of 66 Building Siene Corporation 46

48 Photograph 46: View Sreen On Camera Probe buildingsiene.om Photograph 47: Aess Openings In Mass Wall Camera Probe Leak Detetion buildingsiene.om Page 47 of 66 Building Siene Corporation 47

49 Photograph 48: Water Spray Rak On Masonry Mass Assembly m o. e n e i ings d l i u b Photograph 49: Power Washer Simulating Wind Driven Rain m o. iene s g n i d l i u b Building Siene Corporation 48 Page 48 of 66

50 Photograph 50: Test Wall Establishing Baselines buildingsiene.o om Photograph 51: Garages Used As Test Walls Varying Configurations buildingsien e.om Page 49 of 66 Building Siene Corporation 49

51 Photograph 52: Unpainted Stuo Rendering Test Wall buildingsiene.om Photograph 53: Painted Stuo Rendering Test Wall buildingsiene iene.om Page 50 of 66 Building Siene Corporation 50

52 Photograph 54: Spray Rak Without Indued Pressure Differene buildingsiene.o om Photograph 55: Visual Indiation of Water Entry At Base of Assembly buildingsiene e.om Page 51 of 66 Building Siene Corporation 51

53 Photograph 56: Water Entry Pattern of Saturated Mass Assembly buildingsiene.om Photograph 57: Spray Rak Unpainted Stuo Rendering buildingsiene iene.om Page 52 of 66 Building Siene Corporation 52

54 Photograph 58: Three Hour Water Test Unpainted Stuo Rendering buildings iene.om Photograph 59: Water Testing Control Joints buildingsiene.om Page 53 of 66 Building Siene Corporation 53

55 Photograph 60: Typial Crak In Stuo Rendering m o. e n e i ings d l i u b Photograph 61: Water Entry Assoiated With Typial Crak In Stuo Rendering m o. iene s g n i d l i u b Building Siene Corporation 54 Page 54 of 66

56 Figure 20: Mass Assembly With Inreased Moisture Tollerane buildingsiene.om Figure 21: Seat In Foundation Inreasing Moisture Storage Capaity buildingsiene.om Page 55 of 66 Building Siene Corporation 55

57 Figure 22: Vapor Semi Permeable Non Water Sensitive Rigid Insulation buildingsiene.om Photograph 62: Seat In Conrete Slab Foundation Inreasing Moisture Storage Capaity buildingsiene dingsiene.ooom Page 56 of 66 Building Siene Corporation 56

58 Photograph 63: Blistering of Elastomeri Paint Coating buildingsiene.o om Photograph 64: Rigid Foam Interior Lining Proteting Interior Surfaes buildingsiene. ne.om Page 57 of 66 Building Siene Corporation 57

59 Photograph 65: Rigid Insulation Extends Past Interior Framing buildings iene.om Photograph 66: Paint Coating Over Rough Surfae buildingsiene.om Page 58 of 66 Building Siene Corporation 58

60 Photograph 67: Paint Coating 3 Mil Thikness buildingsiene.om Photograph 68: Paint Coating 10 Mil Thikness buildingsiene.om Page 59 of 66 Building Siene Corporation 59

61 Photograph 69: Settlement Craks buildingsiene.o om Photograph 70: Fiber Reinfored Stuo Rendering to Control Craking buildingsiene.om Page 60 of 66 Building Siene Corporation 60

62 Review of Florida Building Code As part of our review of the water management provisions of the Florida Building Code we identified a provision relating to roof membranes, Setion , that appears to make no sense: If the underlayment is a self-adhering membrane, the membrane shall be applied over a mehanially attahed anhor sheet, attahed in ompliane with This provision enourages roof membrane blow off by requiring a bond break between the selfadhering membrane and the strutural roof dek. This negates the entire reason to use a self-adhering membrane. The requirement for a mehanially attahed anhor sheet under a self-adhering membrane has no basis in physis or logi and prevents the onstrution of highly water resistant roof assemblies that an provide a high degree of seondary protetion during extreme wind driven rain events. Self-adhering membranes that are diretly adhered to strutural deks suh as plywood and OSB sheathing are a proven tehnology in all limate zones in the United States (Photograph 71, Photograph 72, Photograph 73 and Photograph 74). Page 61 of 66 Building Siene Corporation 61

63 Photograph 71: Self Adhering Membrane buildingsiene.om Photograph 72: Diret Applied Self Adhering Membrane buildingsiene iene.om Page 62 of 66 Building Siene Corporation 62

64 Photograph 73: Large Institutional Roof Diret Applied Self Adhering Membrane m o. e n e i ings d l i u b Photograph 74: Complex Roof Geometry Diret Applied Self Adhering Membrane m o. iene s g n i d l i u b Building Siene Corporation 63 Page 63 of 66

65 Reommendations Based on our investigation, testing, review and analysis, the following reommendations are offered: The moisture storage apaity of mass walls be inreased by providing a seat at the base of these assemblies. A bond break be provided between primary drainage planes and stuo renderings in drained assemblies. In simple terms this will require two layers of building paper or a layer of building paper over a plasti housewrap. The speifiation, rating and testing of WRB s be onsistent with their installed exposure i.e. tested and rated as part of a stuo assembly. Appropriate performane speifiations need to be developed for WRB s used with stuo renderings and the Florida Building Code altered to require them. Code offiials be instruted regarding the orret interpretation of ASTM C1063 and the Florida Building Code be expliitly altered to require drainage where drained assemblies interset mass assemblies. Windows and doors be orretly rated and tested aording to ANSI/AAMA 101. Mulled window units, double windows or omposite windows be tested and held to the same requirements as single units. Water managed window and door installation requirements be developed and the Florida Building Code altered to require them. Pressure relieved/baffled soffit assemblies be developed for vented roof assemblies and the Florida Building Code altered to require them. The Florida Building Code be altered to ome into ompliane with the International Residential Code to expliitly allow for the onstrution of unvented roof assemblies. Water managed details for dryer vents, eletrial panel boxes, eletrial boxes, vent fan hoods be developed and the Florida Building Code Altered to require them. It is unlikely that a pratial paint speifiation an be developed in the short term to address miro-raking stuo issues as the relationships among water vapor permeability, mil thikness and elastiity are not known. It is reommended that these relationships be explored and that until these relationships are understood the Florida Building Code not be altered to require elastomeri paints on stuo renderings. Repeal of Setion of the Florida Building Code requiring a mehanially attahed anhor sheet between a self-adhering membrane and roof sheathing. Page 64 of 66 Building Siene Corporation 64

66 Referenes AAMA 101/I.S.2/NAFS-02, Voluntary Performane Speifiations for Windows, Skylights and Glass Doors, Amerian Arhitetural Manufaturers Assoiation, 1827 Walden Offie Square, Suite 550, Shaumburg, IL, ASTM E Standard Test Method for Water Penetration of Exterior Windows, Skylights, Doors, and Curtain Walls by Uniform Stati Air Pressure Differene, Amerian Soiety for Testing and Materials, 100 Barr Harbor Drive, PO Box C700, West Conshohoken, PA, ASTM E Standard Test Method for Water Penetration of Exterior Windows, Skylights, Doors, and Curtain Walls by Cyli Stati Air Pressure Differene, Amerian Soiety for Testing and Materials, 100 Barr Harbor Drive, PO Box C700, West Conshohoken, PA, ASTM C Standard Terminology Relating to Gypsum and Related Building Materials and Systems, Amerian Soiety for Testing and Materials, 100 Barr Harbor Drive, PO Box C700, West Conshohoken, PA, ASTM C a Standard Speifiation for Appliation of Portland Cement-Based Plaster, Amerian Soiety for Testing and Materials, 100 Barr Harbor Drive, PO Box C700, West Conshohoken, PA, ASTM C Standard Speifiation for Installation of Lathing and Furring to Reeive Interior and Exterior Portland Cement-Based Plaster, Amerian Soiety for Testing and Materials, 100 Barr Harbor Drive, PO Box C700, West Conshohoken, PA, AC38 Aeptane Criteria for Water-Resistive Barriers, ICC Evaluation Servies, In., Approved June 2004, 5360 Workman Mill Road, Whittier, CA, AC 132 Aeptane Criteria for Atti Vents, ICC Evaluation Servies, In., Approved January 2001, 5360 Workman Mill Road, Whittier, CA, Florida Building Code, State of Florida, May, 2001 Page 65 of 66 Building Siene Corporation 65