Thermal Performance of Foam Retrofitted Vented Residential Attic

MISSING IMAGE

Material Information

Title:
Thermal Performance of Foam Retrofitted Vented Residential Attic
Physical Description:
1 online resource (131 p.)
Language:
english
Creator:
Shreyans,Sushmit
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Civil Engineering, Civil and Coastal Engineering
Committee Chair:
Prevatt, David O.
Committee Members:
Gurley, Kurtis R
Masters, Forrest

Subjects

Subjects / Keywords:
attic -- ccspf -- energy -- hvac -- hygrothermal -- residential -- retrofit -- roof -- temperature -- thermal -- vented -- wufi
Civil and Coastal Engineering -- Dissertations, Academic -- UF
Genre:
Civil Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
This study documents the findings of an experimental investigation and field evaluation of the thermal performance of closed-cell spray-applied polyurethane foam (ccSPF). This research was sponsored by the Florida Division of Emergency Management, Florida and this field study was performed on an occupied single family residence located in Gainesville, Florida. Temperature, relative humidity, weather data, electricity and gas consumption were continuously monitored over before and after ccSPF installation in vented attic of the test house. The objective was to quantify the energy consumption in heating/cooling the house and evaluate the benefits of ccSPF application in a residential house. A thermal scan was performed to determine the interior ceiling temperatures before and after installation. In addition, air tightness testing was carried out on the building envelope and a duct leakage done prior to insulation installation. It was found that installing ccSPF to the wood deck resulted in approximately a 200F decrease in peak attic temperatures during summer months. Further, energy analysis showed that there was a small reduction (~5%) in the energy consumption of the test house in 2010 from that in the year 2009. Three different approaches for normalizing the energy consumption of HVAC were developed to determine the changes in cooling energy usage pattern in the house due to foam installation alone. The results showed about 49% reduction in daily cooling energy requirements of the house after ccSPF installation. A numerical model was developed using finite difference based software WUFI Pro 4.2, simulating the hygrothermal performance of the roof and attic of the subject house. The roof model was validated using the data collected from test house and a comparative study of long term moisture performance of twelve roof systems in case of an incidental leakage was also performed. The results of numerical study showed that ccSPF retrofitted vented and unvented roofs were found to be prone to moisture accumulation in roof sheathing for a long time duration potentially leading to mold growth on the wood.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Sushmit Shreyans.
Thesis:
Thesis (M.S.)--University of Florida, 2011.
Local:
Adviser: Prevatt, David O.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2011
System ID:
UFE0043491:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 T HERMAL PERFORMANCE OF FOAM RETROFITTED VENTED RESIDENTIAL ATTIC By SUSHMIT SHREYANS A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Sushmit Shreyans

PAGE 3

3 To my parents Anil Kumar and Sarita Verma and my sister Sneha Subhi

PAGE 4

4 ACKNOWLEDGMENTS I wish to express my sincere gratitude to my mentor and thesis committee chair, Dr. David O. Prevatt for giving me the opportunity to pursue this research under his guidance. I also thank my committee members Dr. Kurtis R. Gurley and Dr. Forrest J. Masters for their guidance throughout the course of this research. I am also grateful to the faculty and staff of the Civil and Coastal Engineering Department for their assistance and support during my stay at the University of Florida Gainesville I acknowledge the financial support provided by the Department of Community Affairs, FL through DCA grant ID #: 10 RC 26 13 00 22 210 without which this study could not have been possible. I would also like to thank Mr. Mirjam Bakker, Fraunhofer Institute for Building Physics, Germany, for providing the non commercial license of WUFI Pro. 4.2 used in this research. I extend my thanks to Dr. Peter Datin and Zach Ferrall for helping me with the installation of instrumentation for the experimental s tudy. I am indebted to my parents for their constant love and encouragement all my life making my dreams of pursuing research come true. A special thanks to my colleagues Kenton McBride, Cra ig Dixon, Luping Yang, Xinlai Pe ng and all my friends for their su pport, help and motivation.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 8 LIST OF FIGURE S ................................ ................................ ................................ ......................... 9 ABSTRACT ................................ ................................ ................................ ................................ ... 12 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .................. 14 Background and Motivation ................................ ................................ ................................ ... 14 Objectives ................................ ................................ ................................ ............................... 18 Scope of Work ................................ ................................ ................................ ........................ 18 Organization o f Report ................................ ................................ ................................ ........... 19 2 LITERATURE REVIEW ................................ ................................ ................................ ....... 20 Background ................................ ................................ ................................ ............................. 20 Construction of As phalt Shingle Roofing Systems ................................ ................................ 21 Attic Ventilation ................................ ................................ ................................ ..................... 23 Performance of Vented and Unvented Attics in Hot and Humid Climates ............................ 28 Attic Temperature and Heat Flux ................................ ................................ .................... 28 Sheathing and Shingle Temperatures ................................ ................................ .............. 29 Moisture Performance ................................ ................................ ................................ ..... 30 Effect of Roofing Systems on Energy Efficiency of Houses ................................ ................. 31 Field Studies ................................ ................................ ................................ .................... 32 Analytical Studies ................................ ................................ ................................ ............ 33 Wind Uplift (Structural) Resistance of Spray Foam Insulation ................................ .............. 35 3 REVIEW OF EXISTING FOAM INSTALLATION IN FLORIDA HOMES ...................... 37 4 TEST HOUSE DESCRIPTION ................................ ................................ ............................. 42 Gene ral Layout ................................ ................................ ................................ ....................... 42 Type of Construction ................................ ................................ ................................ .............. 43 Attic Space and Vent Details ................................ ................................ ................................ .. 46 Roofing and Attic Inspection ................................ ................................ ................................ .. 47 Thermal Inspection ................................ ................................ ................................ ................. 48 ccSPF Installation in the Test House ................................ ................................ ...................... 49

PAGE 6

6 5 INSTRUMENTATION OF TEST HOUSE AND DATA COLLECTION ........................... 51 Temperature and Relative Humidity Monitoring ................................ ................................ ... 51 Energy Monitoring ................................ ................................ ................................ .................. 55 Weather Station ................................ ................................ ................................ ...................... 57 6 RESULTS OF EXPERIMENTAL INVESTIGATION ................................ ......................... 60 Temperature Measurement ................................ ................................ ................................ ..... 60 Relative Humidity Measurement ................................ ................................ ............................ 65 Electricity Consumption ................................ ................................ ................................ ......... 68 Gas Consumption ................................ ................................ ................................ .................... 78 Discussion of Experimental Results ................................ ................................ ....................... 79 7 ANALYTICAL MODELING IN WUFI PRO 4.2 ................................ ................................ 82 Introduction to WUFI Pro 4.2 ................................ ................................ ................................ 82 Model Inputs ................................ ................................ ................................ .................... 83 Calculation Proc edure and Results ................................ ................................ .................. 84 Application of WUFI to Predict Roof Performance ................................ ............................... 85 Numerical Study Using WUFI ................................ ................................ ............................... 88 Mo del 1 ................................ ................................ ................................ ............................ 88 Model 2 ................................ ................................ ................................ ............................ 89 Roof Assembly ................................ ................................ ................................ ................ 90 Input and Simu lation Parameters ................................ ................................ ..................... 91 Results of WUFI model ................................ ................................ ................................ ... 93 Prediction of Long Term Hygrothermal Performance of Different Roof Configurations ..... 97 Roof Configurations ................................ ................................ ................................ ........ 97 Simulation Parameters ................................ ................................ ................................ ..... 99 Results ................................ ................................ ................................ ............................. 99 8 CONCLUSIONS AND RECOMMENDATIONS ................................ ............................... 105 Summary ................................ ................................ ................................ ............................... 105 Concluding Remarks ................................ ................................ ................................ ............ 106 Limitati ons of this Study ................................ ................................ ................................ ...... 108 Recommendations for Future Work ................................ ................................ ..................... 108 APPENDIX A BASELINE SURVEY REPORT OF THE TEST HOUSE BY THERMO SCAN INSPECTIONS PRIOR TO CCSPF INSTALLATION ................................ ....................... 110 B BASELINE SURVEY REPORT OF THE TEST HOUSE BY THERMO SCAN INSPECTIONS POST CCSPF INSTALLATION ................................ ............................... 117 C WEATHER STATION DATA RECORDED AT THE TEST HOUSE .............................. 123

PAGE 7

7 D MONTHLY ELECTRICITY AND NATURAL GAS CONSUMPTION DATA OBTAINED FROM GAINEVILLE REGIONAL UTILITIES (GRU) FOR 2009 AND 2010 ................................ ................................ ................................ ................................ ...... 127 LIST OF REFERENCES ................................ ................................ ................................ ............. 128 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 131

PAGE 8

8 LIST OF TABLES Table page 5 1 Date of installation of the instruments. ................................ ................................ .............. 51 5 2 EK H4 sensor identification ................................ ................................ .............................. 55 5 3 Update interval for data transmission from the ISS to the receiver ................................ .. 58 6 1 Mean of daily maximum temperatures before and after closed cell spray applied polyurethane foam (ccSPF) installation ................................ ................................ ............. 63 6 2 Monthly temperature difference between attic and ambient during 2009 2010 ................ 65 6 3 Average daily electricity consumption a week before and after ccSPF installation .......... 71 6 4 Normalized cooling energy ................................ ................................ ................................ 78 7 1 Comparison of predicted drying times of sheathing in roof models developed at UF to those predicted by Nelson and der Ananian. ................................ ................................ 89 7 2 Roof components of the numerical model ................................ ................................ ......... 90 7 3 Peak moisture content and drying times of oriented strand board (OSB) and plywood sheathing in the simulated vented and unvented roof models at UF ............................... 103

PAGE 9

9 LIST OF FIGURES Figure page 1 1 Percentage of existing residential structures by year built ................................ ................. 17 2 1 Typical cross section of wood residential roof ................................ ................................ .. 22 2 2 Common attic insulations ................................ ................................ ................................ .. 23 2 3 Typical venting techniques in residential houses ................................ ............................... 24 2 4 Mechanical ventilation techniques ................................ ................................ ..................... 24 2 5 Thermal interaction and air movement in a typical Florida home ................................ ..... 27 3 1 Sketch showing the three protection levels of ccSPF structural retrofit of wood roof sheathing panels ................................ ................................ ................................ ................. 38 3 2 Front view of the house in Pinellas Park, FL ................................ ................................ ..... 39 3 3 Level II ccSPF sprayed on the underside of roof deck ................................ ...................... 40 3 4 House 1 with ocSPF installed in the attic ................................ ................................ .......... 41 3 5 Office roof with ocSPF installed ................................ ................................ ........................ 41 4 1 View of subject house looking from North west corner ................................ .................... 42 4 2 Layout and orientation of the test house ................................ ................................ ............ 43 4 3 Truss and soffit details ................................ ................................ ................................ ....... 44 4 4 Main gable end truss and interior truss over the house ................................ ...................... 45 4 5 Trusses in the garage ................................ ................................ ................................ .......... 45 4 6 Typical corner section in the house ................................ ................................ ................... 46 4 7 Blown insulation in the attic ................................ ................................ .............................. 47 5 1 LogTag temperature and relative humidity recording kit ................................ .................. 52 5 2 Installed LogTag sensors. ................................ ................................ ................................ .. 52 5 3 Position of LogTag temperature/relative humidity sensors in t he test house ................... 53 5 4 EK H4 temperature/relative humidity sensors for measuring surface temperatures ......... 54

PAGE 10

10 5 5 Position of EK H4 temperature/relative humidity sensors in roof ................................ ... 54 5 6 Configuration of LD 1203AESG recorder components ................................ .................... 56 5 7 Electricity and gas monitoring ................................ ................................ ........................... 57 5 8 Weather station installed on metal pole near NE corner of building ................................ 58 5 9 Data acquisition laptop and Davis Vantage Pro 6162 weather station console ................. 59 6 1 Ti meline of events during the course of research ................................ .............................. 60 6 2 Daily mean temperature time history ................................ ................................ ................. 61 6 3 Ambient temperature time history ................................ ................................ ..................... 61 6 4 Attic temperature time history ................................ ................................ ........................... 62 6 5 Daily maximum temperatures ................................ ................................ ............................ 62 6 6 House interior temperature time history of living and study room ................................ .... 64 6 7 Ambient RH (%) time history ................................ ................................ ............................ 66 6 8 Attic RH (%) time history ................................ ................................ ................................ .. 67 6 9 House interior RH (%) time history ................................ ................................ ................... 68 6 10 GRU electricity consumption data of test house during 2009 2010 ................................ .. 68 6 11 HVAC energy consumption history (15 minute interval) ................................ .................. 70 6 12 HVAC electricity vs. ambient temperature during summer ................................ .............. 70 6 13 Average daily cumulative HVAC energy consumption ................................ .................... 72 6 14 Average distribution of 4 ton HVAC unit energy consumption at 15 minute interval during May19 May 24, 2010 and ambient interior temperature difference ...................... 73 6 15 Average distribution of 4 ton HVAC energy consumption at 15 min interval during May 25 June 8, 2010, ambient interior and attic interior temperature difference ........... 73 6 16 Average distribution of 3.5 ton HVAC unit energy consumption at 15 min interval during Aug 9 Sep 30, 2010, ambient interior and attic interior temperature difference ... 74 6 17 Average HVAC energy consumption at 15 minut e interval for positive ambient interior temperature difference for 4 ton unit during May19 May 24, 2010 ..................... 75

PAGE 11

11 6 18 Normalized HVAC e nergy against ambient house interior positive temperature difference ................................ ................................ ................................ ........................... 76 6 19 Average daily cumulative cooling energy consumption before and after the installation of ccSPF ................................ ................................ ................................ .......... 77 6 20 Calculation of time integrated temperature difference by integrating the ambien t interior temperature difference over time ................................ ................................ .......... 77 6 21 GRU gas consumption data of test house during 2009 2010 ................................ ............ 79 7 1 Unvent ed asphalt shingle roof model ( after Nelson and Der Ananian, 2009) ................... 86 7 2 Vented asphalt shingle roof model (after Nelson and Der Ananian, 2009) ....................... 86 7 3 Test roof model ................................ ................................ ................................ .................. 90 7 4 Screenshot of exterior climate file for Gainesville (2010) ................................ ................. 92 7 5 Screenshot of interior climate file (from recorded attic temperature/relative humidity at the test house) ................................ ................................ ................................ ................. 92 7 6 Predicted temperature and relative humidity of plywood at the plywood foam interface in test house roof ................................ ................................ ................................ 93 7 7 Predicted surface temperature and relative humidity of ccSPF layer in test house roof ... 94 7 8 Measured vs. predicted surface temperature of plywood at the plywood foam interface in the attic of test house ................................ ................................ ...................... 95 7 9 Measured temperature (EK H4 Sensor 2) vs. predicted surface temperature of ccSPF exposed in the attic of test house ................................ ................................ ....................... 95 7 10 Measured temperature (EK H4 Sensor 3) vs. predicted surface temperature of ccSPF exposed in the attic of test house ................................ ................................ ....................... 96 7 11 Predicted water content of OSB sheathing in vented, non retrofitted roof ...................... 100 7 12 Water content of oriented strand board (OSB) sheathing in simulated asphalt shingle roof models ................................ ................................ ................................ ...................... 101 7 13 Water content of plywood sheathing in simulated asphalt shingle roof models .............. 102

PAGE 12

12 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science THERMAL PERFORMANCE OF FOA M RETROFITTED VENTED RESIDENTIAL ATTIC By Sushmit Shreyans August 2011 Chair: David O. Prevatt Major: Civil Engineering This study documents the findings of an experimental investigation and field evaluation of the thermal performance of closed cell spray applied polyuretha ne foam (ccSPF) This research was sponsored by the Florida Division of Emergency Management Florida and this field study was perf ormed on a n occupied single family residence located in Gainesville, Florida. Temperature, relative humidity, weather data, electricity and gas consumption were continuously monitored over before and after ccSPF installation in vented attic of the test house. The objective was to quantify the energy consumption in heating/cooling the house and evaluate the benefits of ccSPF app lication in a residential house A thermal scan was performed to determine the interior ceiling temperatures before and after installation. In addition, air tightness testing was carried out on the building envelope and a duct leakage done prior to insulat ion installation. It was found that installing ccSPF to the wood deck resulted in approximately a 20 0 F decrease in peak attic temperatures during sum mer months Further, e nergy analysis showed that there was a small reduction (~5%) in the energy consumpti on of the test house in 2010 from that in the year 2009. Three different approach es for normalizing the energy consumption of HVAC w ere developed to determine the changes in cooling energy usage pattern in the house due to

PAGE 13

13 foam installation alone. The results showed a bout 49 % reduction in daily cooling energy requirements of the house after ccSPF installation A numerical model was developed using finite dif ference based software WUFI Pro 4.2, simul at ing the hygro thermal performance of the roof and atti c of the subject house The roof model was validated using the data collected from t e st house and a comparative study of long term moisture performance of twelve roof systems in case of an incidental leakage was also performed The results of numerical study showed that ccSPF retrofitted vented and unvented roofs were found to be prone to moisture accumulation in roof sheathing for a long time duration potentially leading to mold growth on the wood.

PAGE 14

14 CHAPTER 1 INTRODUCTION Background and Motivation Residential buildings consume more than one fifth of the energy used in the United States today (U.S. Energy Information Administration 2010) and 50% of this energy is used for heating and cooling rooms. A significant reduction in energy use can be made only by substantial improvements to the energy use characteristics in existing 20 million homes with wood framed roof structures located in hurricane prone areas (U.S. Census 2007) T he majority of these wood framed houses are single story structures with vented attics that enable air circulation of ambient air to moderate roof deck temperatures and to remove/prevent moisture build up in the wood In tradit ional roof construction, insulation is placed immediately above the ceiling so as to isolate the conditioned air in the interior of the home from the hot, humid attic air. In many homes, the ducting that delivers the conditioned air is housed within the at tic itself. Attic temperatures can be as much as 4 0 to 5 0 0 F above the ambient temperatures (Parker and Sherwin 1998) and has been suggested that such high temperatures may reduce the efficiency of the cooling syst em. A critical issue with existing residential structures in hurricane prone regions is their structural resistance to wind uplift forces. Observations after hurricanes h ave shown that many of these structures are inadequately built to resist hurricane for ce winds (Gurley et al. 2006) with many lacking continuous load paths with adequate strengths in all links to resist the forces. While newer homes may have more robust connections (i.e. ring shank nails and closer fastener spacing in roo f decks, metal roof to wall strap connections), many older, existing structures do not have these structural details and would therefore be more susceptible to wind damage in the next hurricane. Thus in order to reduce the elevated vulnerabilities of a large proportion of older

PAGE 15

15 existing homes within the current inventory of residential housing underscore the need to develop reliable retrofit strategies are required to improve the wind resistance of these roofs. During hurricanes hou ses are also susceptible to intense rainfall, which when within the high winds becomes wind driven rain that penetrates the building envelope through cracks and holes, and eventually developing leak paths for water to enter the house. Water within a home w ill damage the interior structure of the building, initiate or accelerate growth of mold or fungi on building surfaces, and if water enters in sufficiently large volume, it may also damage the interior contents, finishes etc. In fact, studies have shown th at losses from water ingress can be as large a contributor to economic loss from a hurricane as does the structural damage caused by the win d (FEMA 2005) Inadequate structural design of roofs and framing in residential structures h a s no doubt contributed to the high costs and extent of damage occurring in Florida and other states after hurricanes. Unless something is done, many existing structures will retain a higher than anticipated vulnerability to wind loads as a result of these inadequacies. Thus, in order to mitigate hurricane damage the conditions of existing homes must be considered. If those structures are not structurally adequate, they ought to be retrofitted, and at the same time, retrofits should be installed which minim izes potential for water intrusion to occur. Such retrofits would have maximum benefits if the existing structures are simultaneously retrofitted to upgrade their energy efficiency, which has been the subject of several federally sponsored programs (FEMA 2010; IBHS 2011) One retrofit method which has shown promise in addressing all three issues is closed cell spray applied polyurethane foam (ccSPF). The benefits of closed cell spray applied polyurethane foam (ccSPF) for thermal insulation of wood roof attics are well known. ccSPF is an excellent thermal insulator (~R6 per

PAGE 16

16 in.) which provides beneficial energy savings for heating and cooling of residen tial structures. Recent r esearch at the University of Florida has shown that a ccSPF fillet sprayed along the joint between wood framing ( truss top chord or rafter ) to roof sheathing increases the wind uplift capacity of a roof deck by a factor of 2.0 to 5 .0, depending on the fastener size and schedules used Furthermore, ccSPF is impermeable to water and so it acts as a secondary water barrier that reduces water leakage in the event the roof covering is compromised during hurricanes. Currently studies are underway at the University of Florida to evaluate the effect of ccSPF on the durability of wood in direct contact with ccSPF, should roof leaks occur in the structure. It is claimed (by ccSPF installers and homeowners) that ccSPF installed in a typical ven ted attic of a residential home will reduce the ambient attic temperature by 20 0 F to 40 0 F on hot summer days. In hot humid climates, such as the state of Florida, reduc ing attic temperatures from 140 0 F to 95 0 F creates a more efficient (semi conditioned) sp ace for the heating and cooling system to operate. Independent verification of these claims is needed and will go a long way to encourage homeowners to retrofit their older homes, which in the long term is the most efficient way of reducing the extensive d amage that is annually sustained by Florida housing inventory. Studies have established the energy efficient performance of un vented attics which are thermally isolated from the ambient air ( by using make up airflow provided by the mechanical air system). Lstiburek ( 2006 ) provides comprehensive reviews of the attic ventilation issues like moisture accumulation and heat gain pertaining to vented versus non vented attic configurations. However, t he cost to convert the existing residential roof attic to an unvented attic may be beyond the level of homeowners as there are many steps involved Housing the HVAC units and air ducts within a semi conditioned attic can reduce the thermal load and moisture load which increases its energy efficiency and longevity. Parker (2005) showed that

PAGE 17

17 energy consumption for heating and cooling of buildings can be reduced by 20% to 40% using non ve nted attics. While the current ccSPF insulation market focuses primarily on new construction of non vented attics (and to lesser extent exterior wall insulation), a vast potential market exists for ccSPF in retrofit applications of existing residential str uctures. As shown in Figure 1 1 approximately 80% of the existing housing stock was built before 1994, when more stringent wind resistant provisions were first introduced to building codes in Florida (U.S. Census 2003) Figure 1 1 Percentage of existing residential structures by year built In summary, as a retrofit application, ccSPF in w ood roof structures may prov ide three very important benefits; 1) significant improvement of the wind uplift (structural) resistance of

PAGE 18

18 the roof, 2) significant reduction in energy consumption for heating and cooling of the home and 3) a durable secondary water barrier to limit water penetration through the roof structure during hurricanes. However, test data are needed to compare the energy p erformance of the vented and unvented roof configurations. Specifically, this study will test the hypothesis that vented attic installations of ccSPF can lower energy bills by 40% in residential homes. Objectives The primary objective of this study was to test the hypothesis that a 1 in. thick layer of ccSPF sprayed onto the underside of roof deck in a vented residential attic reduces the peak attic temperatures by at least 20 0 F and also leads to energy savings of up to 40% in cooling/heating the house A secondary objective is to validate a 1 D hygrothermal roof model in WUFI using the experimentally collected tempera ture and relative humidity data and analyze the long term thermal and moisture performance of several foam retrofitted and un retrofitted roo f systems in a hot humid climate. Scope of Work To test the hypothesis instrumentation was set up in a test house located in Gainesville, FL, to quantify the energy performance of the house, monitor the temperature and relative humidity and ambient weather conditions and to monitor the energy usage for house for over 25 months before and after the installation of the ccSP F The data were used to evaluate the effect of foam installation on the performance of vented attic as well as energy costs for the test house from January 2009 through December 2010. In parallel to the experimental studies, a numerical simulation of the roof attic was developed using the WUFI Pro 4.2 software. This program, jointly developed by the Fraunhofer Institute of Building Physics Germany and Oak Ridge National Laboratory, USA is used to

PAGE 19

19 develop a one dimensional hygrothermal model of the roof cross section and to evaluate potential for water build up through the cross section. A climate file was developed usin g the weather data collected at the site. The model was used to study and compare the performance of twelve asphalt shingle roof s with plywood/ oriented strand board ( OSB ) sheathing in six different configurations a) vented unretrofitted roof, b) unvented unretrofitted roof, c) vented roof retrofitted with 1 in. ccSPF, d) unvented roof retrofitted with 1 in. ccSPF, e) vented roof retrofitted with 3 in. ccSPF and f) unvented roof retrofitted with 3 in. ccSPF Organization of Report Chapter 2 presents the bac kground and literature review relevant to this research Chapter 3 presents observations from field visits cataloging the use of spray foam insulations in Florida homes, and the construction specifications for installing these sy s tems The general layout a nd description of the construction systems used in the test house are provided in Chapter 4 as well as observations for roofing inspections and thermal scanning It also describes the installation of foam in the test house. The instrumentation and data collection of the test house for collecting temperature, relative humidity, and weather and energy data is described in Chapter 5. Chapter 6 presents the results and analysis of data from the long term monitoring of the test house. Validation of a 1 D hygrothermal roof model of the test house using a finite difference met hod based software WUFI Pro 4.2 is discussed in Chapter 7 of this thesis. In addition, the hygro thermal performance of several ccSPF retrofitted and non retrofitted roof models are also discussed in this chapter Finally, the conclusions and limitations of this study along with recommendat ions for future work are discuss ed in Chapter 8

PAGE 20

20 CHAPTER 2 LITERATURE REVIEW Background During the 2004 hurricanes, McCarty ( 2005 ) found that almost half of a survey population of nearly 12,000 Florida residents had to evacuate at least once prior to a hurricane, despite a lmost all respondents living beyond surge prone coastal regions. The wind induced damage occurred to nearly one third of the study areas homes, which sustained major damage and median losses of $11,000 (2004 dollars). These losses are occurring to the roof structures due to inadequate structural systems, improper design and poor construction. The structural retrofit is the one approach that can significantly reduce future losses to Florida homes from hurricanes and reduce energy consumption. Many post storm surveys and reports, (FEMA 2006; Graettinger 2006; Gurley et al. 2006) h ave documented the failures of wood framed roofing in recent hurricanes. Once the roof covering fails, there is much more damage to roof structure du e to water ingress and subsequent failure of other building components. The construction of wood roofs using plywood sheathing was particularly vulnerable because the sheathing was attached either with smooth shank 6d common (2 in. long) nails or metal sta ples that had very low wind uplift resistance. The loss of roof sheathing, particularly from roof corners in residential construction is prevalent particularly in older homes due in large part to the inadequacies in connection detailing prevalent in pre 19 94 building codes. Tests at UF showed the structural inadequacy of using 2 in. long smooth shank nails in hurricane prone regions. Although minimum code requirements have since been strengthened since 1994, the vast majority of homes today still contain th e inadequate connections. In addition, field studies show that there is high possibility that fasteners installed (nailed blind) through the roof deck will many times miss the wood member, resulting in poorly installed and vulnerable roof panels.

PAGE 21

21 The use o f c losed cell spray applied polyurethane foam (ccSPF) as a building thermal insulation system is growing thermal insulating properties (R6 per inch thickness) and it is a durable (water tolerab le) building product. Typical ccSPF formulations 2 inches thick have a permeability of 1 perm or less (i.e. practically impermeable to water), and they also act as a vapor barrier. ccSPF is the only FEMA approved insulator for use in their Flood Insurance program in below grade locations. The thermal properties of spray foam insulation have tremendous potential to reduce heat gain, heat loss and energy use in homes yet questions remain as to appropriate fire retardation approaches suited to residential home s. Recent testing at the University of Florida (Prevatt 2007a; Prevatt 2007b) have quantified the relatively high adhesive strength of one ccSPF formulation installed to retrofit wood panels. The retrofitted pane ls have approximately 2.0 to 2.5 times the uplift strength of original wood panels connected using 2 in. long nails. Thus, ccSPF is also suitable as a structural retrofit system to mitigate hurricane damage to existing structures. The structural benefits o f ccSPF retrofit are realized even if all original fasteners are missing or ineffective. Furthermore, t his chapter discusses about the construction of asphalt shingle roof systems in residential buildings in the US, common insulation methods, attic ventilation techniques and code requirements. The findings of several studies on thermal and energy performance of vented roof systems have also been discussed. Finally, previous research at the University of Florida on wind uplift resistance of ccSPF retr ofitted roof panels has been summarized. Construction of Asphalt Shingle Roofing Systems Asphalt shingled roof systems have been traditionally used in the construction of residential buildings. A 2001 study found that 49.7 % of new steep slope roof construction and 55.5% of re roofing projects consist of asphalt shingle roofing systems (Cash 2003) Several

PAGE 22

22 factors that affect the performance of asphalt shingle ro of systems are geographic location, roof slope, roof ventilation, material and proper detailing and construction (Nelson a nd der Ananian 2009) The framework of the roof consists of wooden trusses or rafters (commonly nominal 2 in. by 4 in. for single story residential houses) typically spaced 24 in. o n c enter Figure 2 1 shows the cross section of a typical shingle roof s ystem. Plywood or oriented strand b oard (OSB) sheathing (typically 4 ft by 8 ft) is installed on these framing members using fasteners (nails or screws). The roofing or roof covering consists of asphalt shingles installed over a weather resistant barrier ( WRB) in a vented or unvented attic configuration. WRB is a membrane placed on the exterior surface of roof sheathing and protects the building envelope from severe weather damage (due to wind, rain or snow). Felt underlayment and self adhering rubberized a sphalt membrane (SRAM) are the most commonly used weather barriers in residential constructions. Figure 2 1 Typical cross section of wood residential roof The roof systems are required to have a minimum level of insulation by most of the building codes in North America to control the thermal interaction between conditioned (interior) and unconditioned (exterior) space. Attic insulation reduces the loss of he at from the building envelope ensuring the thermal comfort of occupants and minimizes the conditioned space heating/cooling demand thus saving energy. Insulation level of a material is often expressed in terms of R value which is a measure of its thermal r esistance. Higher the R value, greater is the

PAGE 23

23 resistance to heat flow across the material and vice versa. The insulation requirements vary by climatic regions and are specified in the regional building codes across the United States. Florida Building Code (2007) requires a minimum of R 19 ceiling insulation in residential buildings (N1104.AB.1, FBC 2007 ). Blown fiber glass and batt insulations (approximately R 3 to 3.5 per in.) have conventionally been used for attic insulations. The use of closed cell (R 6 per in.) and open cell (R 3.5 per in.) spray applied polyurethane foams in roof and wall insulation has gained momentum in the past few years owing to their additional benefits ( Figure 2 2 ). This study focuses on the performance of closed cell foam retrof itted residential vented attics. A B C Figure 2 2 Common attic insulations A) Blown fiber glass insulation on ceiling, B) Batt insulation (Courtesy: http://www.showroom411.com/ViewEntryPicture.aspx?id=12317 ) and C) ccSPF insulation sprayed to underside of wood deck Attic Ventilation Ventilated roof systems have been conventionally used in the construction of low rise residential buildings in North America. Attic ventilation can be achieved through proper ventilation techniques installing soffit, ridge and gable vents in the roof struc ture (Figure 2 3 ) Mechanical ventilation techniques such as the use of exhaust fans on gable end walls and turbine attic ventilators can also be used to increase the rate of attic ventilation ( Figure 2 4 ).

PAGE 24

24 A B C Figure 2 3 Typical venting techniques in residential houses A) Soffit vent (Courtesy: http://www.hvacinstaller.com/how to install 16 inch soffit vents B) Ridge vent (Courtesy: http://www.americananimalcontrol mnwi.com/sub%20pages/Attic%20Ventilation.html ) and C) Gable vent (Courtesy: http://www.trappingdept.com/IMAGES/43.jpg ) A B Figure 2 4 Mechanical ventilation techniques A) Turbine attic ventilator and B) Gable mounted venti lator ( Courtesy: http://activerain.com/blogsview/2058820/a vent ) The primary purpose of ventilation is to remove the attic solar heat gain which occurs through conduction across the roof components in order to reduce the cooling load of house. Some of the main reasons for using a vented attic in both hot humid and cold climates are as follow s : Reducing the attic temperature in hot climate by providing a continuous flow of air from soffits to the ridge removing the warm air from attic. Removing excess moisture from the attic and roof members (sheathing, framing members, insulation, etc.) which may otherwise could cause condensation in the attic. Improving the life span of roof materials (sheathing and shingles) by reducing their surface temperatures. Controlling the form ation of ice dams in cold climates.

PAGE 25

25 The attic ventilation requirements vary from 1 sq. ft. net free ventilation area per 300 sq. ft. of attic floor area to 1 sq. ft. net free ventilation area per 150 sq. ft. of attic floor area. 2009 International Resident ial Code lays down the following minimum ventilation requirements and for the construction of unvented roof assemblies : R806.1 Ventilation required. Enclosed attics and enclosed rafter spaces formed where ceilings are applied directly to the underside of r oof rafters shall have cross ventilation for each separate space by ventilating openings protected against the entrance of rain or snow. R806.2 Minimum area. The total net free ventilating area shall not be less than 1/150 of the area of the space ventilat ed except that reduction of the total area to 1/300 is permitted p rovided that at least 50% and no t more than 80% of the required ventilating area is provided by ventilators located in the upper portion of the space to be ventilated at least 3 feet (914 mm ) above the eave or cornice vents with the balance of the required ventilation provided by eave or cornice vents. As an alternative, the net free cross ventilation area may be reduced to 1/300 when a Class I or II vapor barrier is installed on the warm in winter side of the ceiling. R806.3 Vent and insulation clearance. Where eave or cornice vents are installed, insulation shall not block the free flow of air. A minimum of a 1 inch (25 mm) space shall be provided between the insulation and the roof sheathin g at the location of the vent. R806. 4 Unv ent ed a ttic assemblies. Unvented attic assemblies (spaces between the ceiling joists of the top story and the roof rafters) shall be permitted if all the following conditions are met : 1. The unvented attic space is completely contained within the building thermal envelope. 2. No interior vapor retarders are installed on the ceiling side (attic floor) of the unvented attic assembly. 3. Where wood shingles or shakes are used, a minimum inch (6mm) vented air space separates the shingles or shakes and the roofing underlayment above structural sheathing.

PAGE 26

26 4. In Climate Zones 5, 6, 7, and 8, any air impermeable insulation shall be a vapor retarder, or shall have a vapor retarder coating or covering in direct contact with the unders ide of the insulation. 5. Either Items 6 7 or 8 shall be met, depending on the air permeability of the insulation directly under the structural roof sheathing 6. Air impermeable insulation only. Insulation shall be applied in direct contact with the underside o f the structural roof sheathing. 7. Air permeable insulation only. In addition to the air permeable ins ulation installed directly below the structural sheathing, rigid board or sheet insulation shall be installed directly above the structural roof sheathing a s specified in Table R 806.4 for condensation control. 8. Air impermeable and air permeable insulation. The air permeable insulation shall be applied in direct contact with the underside of the structural roof sheathing as specified in Table R806.4 for condens ation control. The air permeable insulation shall be installed directly under the air impermeable insulation. Figure 2 5 shows the thermal interaction and air movement in a typical vented attic in hot and humid climate. Solar radiation is absorbed by the s hingles or other roof covering and is conducted through sheathing to the unconditioned attic space. This raises the attic temperature up to 40 50 0 F higher than the ambient temperature in a hot and humid climate during the day time (Parker and Sherwin 1998) The heat built up in the attic space is radiated to the attic floor and then to the conditioned living space underneath it. Cooler air entering through soffit vents gains heat in attic and rises upwards where it exits through ridge vent. The heat is also transferred across any le ak that might be present in the attic floor/house ceiling. As such, the amount of insulation on attic floor plays an important role in reducing the heat transfer between the conditioned (living) and unconditioned (attic) space.

PAGE 27

27 Figure 2 5 Thermal interaction and air movement in a typical Florida home Mechanical systems and air conditioning ductwork are often located in the attic space. Inadequate insulation of ducts in the attic can lead to absorption of attic heat during summer and incre ase the energy load on HVAC unit vice versa is true in winter. Poorly insulated and leaky ducts in the attic can cause the condensation of incoming hot and humid water vapor when it comes in contact with the cold surfaces (Rudd and Lstiburek 1996) Long term presence of condensed water on roof components can cause the growth of mold and mildew and also potentially lead to degradation of wooden components. There is a growing trend of using unvente d attics (no soffit or ridge vents) in homes with insulation applied on the underside of the sheathing rather than within the top floor of the ceiling assembly. The advantage of using this configuration is that the space between ceiling floor and sloped ro of plane can be conditioned and used for living or storage purposes. The performance of unvented roof construction has been studied by several researchers and has been compared with

PAGE 28

28 that of vented roofs. Both types of roof systems have their own advantages and limitations. The following section discusses the previous research done to understand the performance of vented and unvented attics in hot and humid climates. Performance of Vented and Unvented Attics in Hot and Humid Climates Several field and analyt ical studies have been done in the past to understand the performance of vented and unvented attics in hot and humid climates. The effects of high attic temperatures and moisture built up on durability and energy performance of several types of roof system s has been extensively investigated. The thermal and energy performance of roof systems with different shingle types, ventilation strategies and insulation methods have been documented. However, the performance of ccSPF insulation in a vented, asphalt shin gle roof system has not been investigated until now, which is the focus of this study. The effect of attic ventilation on attic temperatures, heat flux, shingle and sheathing temperature and energy performance of houses is discussed below. Attic Temperatur e and Heat Flux Attic ventilation plays an important role in reducing the attic temperature as well as that of roof sheathing and asphalt shingles. Parker et al. ( 1998 ) performed an extensive experimental investigati on to study the impact of roof system on the attic temperatures. Mid attic temperatures in a conventional black shingle roof system with 1:300 attic ventilation can peak up to 135 0 F on a hot day with peak ambient temperature of about 92 0 F. A similar roof s ystem with 1:150 ventilation had 12 0 F lower attic temperatures on an average. Roof surface color (shingles or tile) has a significant impact on the attic temperatures. Roofs with light colored shingles and tiles were found to have significantly lower attic temperatures than dark colored roofs as more heat is absorbed by black or other dark surfaces than light surfaces.

PAGE 29

29 Beal et al. ( 1995 ) compared the performance of soffit and ridge venting on ceiling heat flux in a residential black shingle roof system. They found that a highly ventilated soffit (1 sq. ft. of soffit per 50 sq. ft. of attic floor) reduced the ceiling heat flux by 66% whereas a perforated soffit (1 sq. ft. of soffit per 230 sq. ft. of attic floor) resulted in 25% less ceiling heat flux than a sealed attic. Lower ceiling heat flux would imply lower cooling energy requirements, however, the relation between ce iling heat flux and energy consumption was not quantified by the authors. From the above discussion, it is clear that attic ventilation has a significant effect on the attic temperatures and heat flux. Higher attic temperatures and heat built up in the att ic, would mean that greater heat transfer occurs by conduction across the ceiling (between conditioned and unconditioned space) if there is insufficient insulation leading to greater cooling energy requirements. Sheathing and Shingle Temperatures Higher s urface temperature of shingles accelerates their aging thus reducing their service life. Many shingle manufacturers do not warrant their products for unvented attics as they are believed to cause higher shingle temperatures. Rudd et al. ( 1996 ) found peak shingle temperatures up to 180 0 F over conventional shingle roof with a ventilation of 1:300. It was also found that the temperatur es measured on the underside of plywood in sealed attics were up to 17 0 F higher than the vented attic. In another study, Rose et al. ( 1992 ) had found that shingle temperatures over vented attics were lower by about 10 0 F than those over unvented attics. In an analytical study, Rudd et al. ( 1998 ) compared the shingle and plywood temperatures for several parametric simulations in hot and dry climate of Las Vegas, NV and hot and humid climate of Orlando, FL. The roof plywood tem perature in sealed attic was 13 0 F higher than the reference 1:300 vented attic for Orlando. Ho wever, the use of white tile instead of black shingles

PAGE 30

30 resulted in 43 0 F decrease in the plywood temperature with respect to the reference attic. Similar observations were made for the parametric simulations for Las Vegas. TenWolde et al. ( 1999 ) reviewed several issues associated with the venting of attics and sealed attic constructions in a variety of climates including cold inland climates, wet, cold coastal climates and warm, humid climates. They investigated issues related to moisture control, formation of ice dams, durability of shingles and heating and cooling loads. They suggested ventilation only as a design option for hot and humid and wet coastal climates. They also concluded that attic ventilation may s lightly increase winter heating loads as ventilation allows direct air pressure variations across the ceiling plane. The effects of good insulation, use of light colored surfaces (shingles), efficient cooling equipment and reduced solar and appliance heat gains are more significant in cooling energy savings. Moisture Performance In addition to reducing the attic temperature, attic ventilation plays another important role of removing excess moisture from the attic space and roof members. A recent study at th e helps to remove the moisture from the attic in hot and humid climates (Porter 2003) It was also found that moisture content levels in roof sheathing were within acceptable limits in both vented and unvented attic spaces. The use of ccSPF as an insulation in unvented attics has been rising in the recent few years as it can be used as a secondary water barrier and is also impermeable to air. Spray applied foams are easy to apply and seal all the leaks and gaps as they expand to 30 50 times of their volume after application. Prevatt (2007) had raised some concerns about the long term durability of roof systems retrofitted with closed cell foam related to water being trapped between foam layer and sheathing. Nelson and der Anan ian ( 2009 ) performed an analytical study using a one

PAGE 31

31 dimensional hyg rothermal analysis tool WUFI and found that attic ventilation reduces the drying time of roof components in case of incidental water leakage. The vented roof assemblies are more tolerant and durable in case of incidental water leakage than the unvented roof assemblies in both hot and cold climates. The results of one of the vented roof models with asphalt shingles, felt underlayment, OSB/Plywood sheathing and ccSPF insulation had a drying time of less than 1.5 months whereas a similar roof w ith unvented attic had a drying time of over 7 months in case of an incidental leakage in a hot climate of Miami, FL. The drying times for similar roofs were less than 2 months and 14 months respectively for a vented and unvented attic in a cold climate of Boston, MA. Straube et al. ( 2010 ) conducted a hygrothermal modeling study to address the issue of moisture build up on the underside of sheathing in unven ted attics during winters. WUFI simulation results of unve nted roofs showed that several roofs with open cell and closed cell sprayed foam insulations performed well if air leakage and internal humidity levels are controlled. The author also raised issues with workmanship and occurrence of failures due to lack of building air tightness. Effect of Roofing Systems on Energy Efficiency of Houses The cooling energy requirements of a house typically depend on the transfer of ceiling heat flux from the attic to house interior, heat gain to the duct system located in at tic and air leakage from attic to the conditioned space Parker et al. ( 2005 ) As the heat transfer from unconditioned to the conditioned space inc reases the energy demand for cooling the conditioned space also increases. Properly insulated attic with no air leakage to conditioned space is the most energy efficient way of constructing a roof. However, proper workmanship will always be a consideratio n in construction and greater energy efficiency can be achieved with better workmanship.

PAGE 32

32 S ome of the field and analytical studies to understand the energy performance of roof systems in hot and humid climate are discussed below Field Studies Parker et al. ( 2002 ) evaluated the impact of roofing systems on residential cooling energy demand in Florida by studying seven roof systems designed to reduce attic heat gain. Sealed attic construction with R 19 insulation (5 in. of open cell foam with R 3.8/in.) on the underside of roof deck produced an annual savings of about 6 11% in cooling energy use over conventional vented roof with dark colored shingles. Results also showed 18 26% cooling energy consumption reduction for ro ofs with white reflective surfaces (white S tile, white flat tile and white metal roof) over the conventional roof. The results of an experimental study by Rudd et al. ( 1996 ) indicated an average energy savings of 19% in cooling for two sealed attic homes (with high performance windows which may have been responsible for as much as 3% energy savings) over the reference 1:150 vented attic house for a hot and dry climate of Las Vegas As previously discussed, duct insulation and tightness also affects the efficiency of HVAC systems in cooling/heating the house. Improper insulation causes loss of heat from the ducts to attic space and thus leading to greater energy consumption. Jump et al. ( 1996 ) conducted a field study to determine the potential energy savings due to sealing and insulating the duct systems in Sacramento, CA. 24 houses were tested for air leakage through building envelope and ductwork, flow through individual registers, duct air temperatures, ambient temperatures, surface area of ducts, and HVAC energy consumption pre and post retrofit of the duct systems. The results from a 2 week monito ring period showed an average 18% reduction in the energy consumption due to sealing and insulating the duct systems. The energy consumption reduction in the test

PAGE 33

33 houses varied largely from 5% to 57% suggesting that effectiveness of duct retrofits is highl y dependent on individual systems. Analytical Studies This section discusses some analytical studies on thermal and energy performance of vented and unvented attic roof systems. Using the field data to validate analytical models is the most common approach taken by researchers to perform comparative studies on roof systems. Rudd et al. ( 1996 ) conducted a computer modeling study to i nvestigate the cooling energy use in vented and sealed attics in Las Vegas, Nevada. The results showed a 4% energy savings in sealed attics over a vented attic with no duct leakage assumed in both cases. However, when a typical duct leakage (10% return and 5% supply leak) was accounted in the model, the sealed attic showed 10% energy savings on space conditioning. In another analytical study, Rudd et al. ( 1998 ) developed a computer model to simulate the annual and peak cooling loads for a 1500 sq. ft. wood frame house for Orlando, Florida and Las Vegas, Nevada. They used the one dimensional, finite element computer model FSEC 3.0 for this inves tigation. The base case house models for Orlando and Las Vegas differed in amount of attic ventilation, ceiling and wall insulation and glazing. All other parameters in the building model were kept same. Several attic models were simulated with varying amo unt of insulation, ventilation and duct leakage in a parametric study for the two cities. Annual simulation results for Orlando showed that white tile, sealed attic with R 28 and R 19 insulations saved 12.4% and 7.9% annual heating and cooling energy respe ctively as compared to the reference vented attic. Attic models with duct leakage resulted in loss of energy in cooling and heating the house. A duct system with 10% return and 5% supply leakage resulted in 15.7% higher energy consumption while a system wi th 15% return and 10% supply leakage resulted in 25.9% higher energy consumption than the reference attic without any duct leakage. Similar models for Las

PAGE 34

34 Vegas showed 9.5% and 16.4% higher energy consumption respectively than the reference vented attic mo del without duct leakage. Sealing the attic showed a reduction of 8.9% in the annual energy consumption of the house. National Renewable Energy Laboratory (NREL) conducted field testing and numeri cal modeling experiments using DOE 2.2 building simulation software to evaluate the energy related effects of vented and unvented attics (Henderon et al. ( 2004 ) ) Field tests were carried out on houses in hot dry climates of Las Vegas, NV and Tucson, AZ while analytical simulations were developed for houses in Phoenix, AZ and mixed dry climate of Sacramento, CA to determine the effect of duct leakage on performanc e of unvented attics. The estimated energy savings for unvented were 8% and 20% at 100 cfm and 200 cfm supply duct leakage respectively for hot dry climate of Las Vegas. Computer simulation results show that unvented attics lead to energy savings in hot dr y Phoenix whereas, for the same house in Sacramento (mixed dry), vented attics have cooling energy savings over unvented attics. Cost analysis of cooling energy for unvented attics in a variety of hot dry, hot humid and mixed climates showed clear benefits in very hot and sunny climates such as Las Vegas, Phoenix and Tucson while there could be energy penalties in other climates. The authors concluded that energy savings that can be achieved with an unvented attic are dependent on the weather conditions, du ct leakage, roof R value, attic air exchange rate and roof solar heat absorption. The studies discussed above talk about the thermal performance of vented and unvented attics, effect of shingles, insulation, duct leakage issues, etc., but none ha s quantifi ed the thermal performance of ccSPF retrofitted vented attic. This thesis discusses the performance of a vented roof retrofitted with a 1 in. layer of ccSPF.

PAGE 35

35 Wind Uplift (Structural) Resistance of Spray Foam Insulation In 2007, wind uplift tests were condu cted to determine the structural capacity of 30 ccSPF retrofitted wood roof panels at the University of Florida (Prevatt 2007a) The wood roof panels consisted of in. thick by 4 ft. by 8 ft. OSB sheath ing fastened to 2 in. by 4 in. southern yellow p ine (SYP) framing members spaced 2 ft. apart using screws which were removed prior to uplift testing. The objective of this study was to determine the additional uplift strength provided by ccSPF retrofit alone. Following three configurations of r oof panels with cc SPF retrofits were tested per ASTM E330 procedure 1. Configuration A: 3 in. ccSPF f illet along sheathing to wood joints in panel 2. Configuration B: 3 in. ccS PF f illet plus in. thick foam layer. 3. Configuration C: 3in. ccSPF foam layer. The mean failure pres sures of the ccSPF retrofit ted panels were 209 psf, 178 psf and 199 psf, respectively for Configurations A, B and C which are up to 2.5 times the strength of an unretrofitted panel nailed with 6d smooth and 8d ring shank nails A similar experimental invest igation was also conducted in which the foam retrofitted roof panels had sheathing fastened to the framing members by 6d smooth and 8d ring shank nails (Prevatt 2007b) However, none of the 34 panels tested in this study had a Configuration B retrofit. Roof panels with 6d smooth shank fasteners and Configuration B foam retrofit showed the highest uplift capacity of about 250 psf (up to 3 times of unretrofitted panels wi th 6d common nails) among all other configurations tested. The authors concluded that nail type did not appear to affect the uplift capacity of retrofitted panels. Several issues regarding the long term performance and durability of ccSPF retrofitted ro ofs were raised by the authors that need further investigation.

PAGE 36

36 An ongoing research study at the University of Florida to evaluate the structural performance of roofs retrofitted with ccSPF but subjected to intense water leakage has shown similarly high upli ft resistance of roof panels. However, the moisture built up in the foam is an issue which is likely to decay the roof sheathing and framing members in the long run and is being investigated in another study.

PAGE 37

37 CHAPTER 3 REVIEW OF EXISTING FOAM INST ALLATION IN FLORIDA HOMES Within the past five years there has been an increase in the use of ccSPF to insulate roof attics of residential structures in Florida, perhaps due to the research work by Parker and others. There are several large companies like NCFI Polyurethanes, BASF, CertainTeed, Icynene Inc, etc., which manufacture ccSPF Usually these manufacturers license the process for installing ccSPF to local contractors who are trained in the application of the products. ccSPF is a combination of polyisocyanate and polyhydroxyl (blowing agent) compounds which is applied at high pressure and temperature using a compressed air spray gun The liquids (polyisocyanate and polyhydroxyl) are sprayed through two separated hoses which chemically combine on the substrate surface and the mixture inc r eases in volume by about 30 times by forming tiny closed cell bubbles and hardens The level of structur al protection and insulation provided by closed cell foam varies by the thickness and location of foam in stallation Typically there are three protection levels approved in the Florida Building Product Approvals for application in residential roofs diagrammatically represented in Figure 3 1 Level I consists of fillet only at the sheathing seams and along wood framing joints. Level II consists of the f illet at wood joints and a minimum in. thick layer covering protrusions. Level III protection consists of fillet at wood joints plus a 3 in. thick fill.

PAGE 38

38 Level I Level II Level III Figure 3 1 Sketch showing the t hree p rotection levels of ccSPF structural retrofit of wood r oof sheathing p anel s Three residential houses and one office building was investigated for the Spray applied Polyurethane Foam (SPF) installations in Florida. One house was located in Pinellas Park, FL and the other two houses and office building were located in Tampa, FL.

PAGE 39

39 Th e owners of the house visited in Pinellas Park, FL ( Figure 3 2 ) decided to install closed cell foam in their attic primarily to increase the uplift capacity of their roofs as the house had suffered water damage and loss during the recent storms This house had a 4 in 12 roof slope with intersecting gable structure terminating in 12 in. wide eaves. Continuous perforated soffits were installed at the eaves and there were small gable end and ridge vents. The roof structure consisted of OSB sheathing on wood an d metal plate trusses spaced 24 in. apart. There was approximately 6 in. thick layer of blown insulation on the ceiling. The ccSPF was installed in a level II configuration ( Figure 3 3 ). The installers covered the soffit vents prior to spraying to prevent excess spray from escaping and damaging the walls or exterior vegetation. The installers used an exhaust fan to remove attic fumes from the spray process during the work. On a typical sized house (say 1800 sf) the foam installation job usually takes about 3 4 hours to complete. Figure 3 2 Front view of the house in Pinellas Park, FL

PAGE 40

40 Figure 3 3 Level II ccSPF sprayed on the underside of r oof deck Another insulating product that is also in common use in Florida is the open cell spray applied polyurethane foam The two houses and one office building visited in Tampa, FL had open cell foam insulation in the attics. The chemical composition is similar to that of ccSPF except that the cells formed after the foam hardens are partially open and fill ed with air unlike in closed cell f oams where the cells are closed The density of open cell foam is about 0.5 pcf and thus is much lighter than close d cell foams which have a density of about 2 pcf. Open cell spray foam is used in walls and other areas an d has the benefit of forming continuous insulation layers without major thermal breaks. A 3 in. thick foam layer can be applied as is the ccSPF, however the open cell product does not have as high insulating value (it is around 3 .5 per in.) as compared to the ccSPF The open cell foam has a spongy feel and it is believed it does not have significant structural resistance to wind load. During the inspection water stains on the ceiling tiles below the installed open cell foam suggested that it is also not im pervious to water Figure 3 5

PAGE 41

41 Figure 3 4 House 1 with ocSPF installed in the attic Figure 3 5 Office roof with ocSPF installed

PAGE 42

42 CHAPTER 4 TEST HOUSE DESCRIPTI ON General Layout The test house is a single family detached structured with a total floor area of about 2400 square feet, located in a suburban neighborhood of Gainesville, FL. The house was constructed in 1972 and most of the houses in the neighborhood are of similar construction and vintage. The front of the house faces to the w est and the main roof ridge is oriented in a north south direction. The house is amply shaded by large live oak trees with extensive canopies located on the east and west sides (Figure 4 1) Figure 4 2 illustrates the general plan and layout of the test ho use on the site. Figure 4 1 View of subject house looking from North west corner

PAGE 43

43 Figure 4 2 Layout and orientation of the test house The axial length of the house is parallel to the street. As viewed from the street, the garage is located on the right (southern end) and connects perpendicularly to the main house. On the left side of the garage is a covered walkway leading to the front entrance. The test house is 38 ft. wide by 63 ft. long. The exterior walls are 8 ft. 6 in. tall and there is a 24 in. wide ea ve extending beyond the walls. The roof ridge is 14 ft. 9 in. above grade for the main roof and about 12 ft. 4 in. above grade over the garage. The garage is located at the south west corner of the house and measures 21 ft. 10 in. wide and 18 ft. long in plan. Type of Construction The house has cinderblock exterior walls with a stucco finish. All gable end walls are clad with a painted wood siding material ( T 111) The roof consists of asphalt shingles place d over felt underlayment on in. p lywood sheathing with a slope of 4:12. Metal plate roof trusses span between the exterior wall, and are fastened to them with galvanized metal straps.

PAGE 44

44 Figure 4 3 Truss and soffit d etails The trusses at different sections of the roof over the house and garage are shown in Figure 4 4 and Figure 4 5 The gable trusses in the house ( Figure 4 4 ) are built using two 2x4 in. SYP members stacked vertically on top of each other for the top member. The top 2x4 in. is connected to the main truss by 8 in. long, 2x4 in. wood blocking. There are five blocks spaced 4 ft. on center of each side of the truss. Both the gable end and interior truss es span 42 ft. with the exterior walls supporting the trusses being 38 ft. apart Figure 4 5 show s the truss design at different sections (Section CC, DD and EE, Figure 4 3 ) in the garage. The garage walls are 21 ft. 10 in. wide, while the trusses extend 2 ft. further on each side to the eaves.

PAGE 45

45 Figure 4 4 Main g able end truss and interior truss over the house Figure 4 5 Truss es in the garage

PAGE 46

46 1 in. wide galvanized metal straps are anchored into the walls where the trusses attach to the exterior walls as shown in Figure 4 6 Figure 4 6 Typical corner section in the house Attic Space and Vent Details The attic is an unconditioned space ventilated by three gable end vents, soffit vents and three off ridge vents. The air conditioning ventilation ducts and electrical wiring run all through the attic. The soffit vents along the eaves are 3x12 in. in size, and have in sect screens and rounded ends. They are sp aced approximately 4 ft. apart. On the drywall ceiling there is approximately, 3 to 5 in. thick blown fiberglass insulation (Figure 4 7). This insulati on is of variable thicknesses with numerous gaps where the ceiling surface is completely exposed. The insulation is blown only in the attic area over the conditioned space and there is no insulation in the attic space over garage

PAGE 47

47 Figure 4 7 Blown i nsulation in the attic Roofing and Attic Inspection In January 2010 the test house was inspected by a licensed professional roofing contractor to evaluate the current condition of the roof prior to the foam installa tion. The roof sheathing was found to b e in good condition, with no active l eaks or water stains observed. As part of the inspector the roofer suggested the following: Provide Moisture strip paper to detect moisture intrusion. To alleviate concern for leakage after foam installation there paper strips can be installed at critical areas (penetrations, transitions, valleys, or areas that had past leaks). The paper would turn a different color and alert the owner of a potential leak problem. Consider stiff paper barriers around all roof penetration s (plumbing, chimneys, heat pipes). The barrier would limit the contact of foam at these penetrations and allow water to flow away from the wood. Critical areas could be exposed for a visual inspection at any time. Install rulers or tags to measure ccSPF foam thickness. Owner should be assured that fillet sizes and thickness of layers are adequately maintained throughout the roof.

PAGE 48

48 The roofer also inspected the asphalt shingle roofing system, which was installed about 16 years ago. The tiles generally showe d no significant curling cuts or tears and there were no immediate areas of concern. The roofer recommended that a roof with ccSPF should be aggressively maintained to ensure any signs of leakage or failed flashing or shingles are promptly repaired. Therm al Inspection A thermal inspection of the house was conducted by Thermo Scan Inspections (TSI) to establish the baseline performance of the test house. It included visual inspection, blower door test, duct blaster test and infra red thermography. Blower door test helps to determine the airtightness of a home which is important for energy savings and maintaining indoor air quality. A large blower fan is installed on an exterior door frame and exhausts out the air from inside thus reducing the air pressure on the inside. Outside air at higher pressure gushes inside the house through leaks and cracks. An air manometer is then used to measure the airflow after a constant pressure difference (usually 50 Pascal) between outside and inside is achieved. T he total air leakage is reported as the volume rate of air flow at the set pressure difference. The total air leakage measured was 3290 cubic feet per minute (CFM) (at 50 Pa) which suggested that the house was drafty and leaky Several major and minor areas of air l eakage were identified and sealing options were recommended. Infrared thermography also helped to identify the areas of air leakage by capturing the images of temperature difference around those gaps. Appendix A shows the thermal images taken during the in spection. Duct blaster test is used to determine air leakage in the duct system on both supply and return side. Duct leakage in the unconditioned space can contribute to energy losses in heating and cooling of the house. For this test, a small fan is used to pressurize the ducts and with the

PAGE 49

49 help of pressure gauge and airflow manometer, the duct leakage is determined. Duct leakage for the supply and return side are determined individually and added together to get the total air leakage. The total duct leaka ge measured was 206 CFM (at 25 Pa) with 123 CFM on the return side and 83 CFM on the supply side. The amount of leakage to the outside of the house was 160 CFM (at 25 Pa). The infrared thermography confirmed the air leakage around metal duct in the attic g oing inside the wall cavity. After the installation of foam in the attic, p ressure diagnostics were performed on the test house to determine the air pressure differences between the house and outside, and the house and attic. Pressure readings between the house and the attic indicate that the house is under a slight ly positive pressure when the air conditioner is switched on possibly because of leaky returns in the attic or garage. Infrared scanning was performed to check the ceiling temperature and insula tion anomalies after foam installation. However, a small temperature difference between the ambient and inside made it difficult to determine any such anomaly. The infrared scanning images taken at various locations in the attic, ceiling and walls are show n in Appendix B. ccSPF Installation in the Test House A l evel II ccSPF structural retrofit was installed in the test house in May 2010, by technicians from Florida Fo am Adhesive of Cape Coral, FL. Attic insulation was used to block the soffit vents and limit escape of the foam. The gable vents were sealed with the plastic sheeting to prevent overspray. The technicians started to spray the foam from the far end (north side) of the attic and moved towards south end. Attic space over the garage was sp rayed at the end. The entire spraying process was completed in about 5 hours. An exhaust fan was used to eliminate the fumes creat ed during the foam application. The resulting work had some areas of non uniform

PAGE 50

50 foam thicknesses and fillets along the wood j oints that were le ss than minimum specifications. The foam installers had to return later with additional foam to correct th e spots with insufficient foam application.

PAGE 51

51 CHAPTER 5 INSTRUMENTATION OF T EST HOUSE AND DATA CO LLECTION Sensors were installed to measure the temperature, humidity and moisture content within the attic space. Energy measurement devices were installed to monitor the electricity and gas consumed in heating/cooling of the house. A weather station was i nstalled to provide a continuous record of the ambient weather conditions that affect attic temperatures. Finally, additional small footprint temperature/relative humidity sensors were installed at the interface between the ccSPF layer and the underside of the plywood sheathing to monitor surface conditions. Table 5 1 lists the number and installation dates of various sensors in the test house followed by a detailed description of the instrumentation of the test house. Table 5 1 Date of installation of the instruments. Parameter Instrument Number Date of Installation Temperature/RH LogTag Temp/RH sensors 5 11/08/2009 EK H4 Temp/RH sensors (for surface temp/RH in attic) 7 06/08/2010 Energy Digital Gas Meter (Elster AC 250) 1 04/06/2010 Enetics AESG 1203 Energy Recorder 1 04/23/2010 Weather Davis Vantage Pro2 Plus 6162 Weather Station 1 04/07 /2010 Temperature and Relative Humidity Monitoring Five data loggers, called LogTag humidity & temperature recorder (HAXO 8) and seven wired EK H4 temperature and humidity sensors (manufactured by Sensirion) were placed at different location s inside and outside the attic The latter sensors (EK H4) took readings at 15 minute intervals that are dir ectly downloaded to a laptop computer. The LogTag sensors are stand along data recorders set to record data every 15 minutes. With a memory capacity of over 8,700 point, the LogTag recorders were left in place for 3 month intervals during which it continuo usly recorded data every 15 minutes.

PAGE 52

52 Figure 5 1 shows the LogTag humidity and temperature recorder and the LogTag interface to transfer the recorded data to a laptop. Installed sensors at two differe nt locations have been shown in Figure 5 2 A B Figure 5 1 LogTag temperature and relative humidity record ing kit A) sensor and B) LogTag i nterface A B Figure 5 2 Installed LogTag sensor s. A) attic and B) outside the house (south wall) Figure 5 3 shows a schematic of the location of these temperature and humidity sensors throughout the attic. The sensors in the attic are located around 1 ft. below the roof decking.

PAGE 53

53 Figure 5 3 Position of LogTag t emperature/ relative humidity sensors in the test hou se EK H4 evaluation kit for temperature and relative humidity measurement comprises of: One EK H4 multiplexer box which has 4 channels for sensor cables, one port for power supply cable and one port for optional USB interface with a computer. Four Pin type S HT71 sensors. Four SMD (Surface Mount Device ) type SHT21 sensors. Four sensor cables with RJ45 plug for connecting sensors with multiplexer box. Figure 5 4 shows the configuration of the multiplexer and sensors and their connection to a laptop for data ac quisition and their location in the attic is shown in Figure 5 5.

PAGE 54

54 A B Figure 5 4 EK H4 temperature/ relative humidity sensors for measuring surface temperatures A) c onfiguration of EK H4 multiplexer & sensors and B) s ensor being installed on foam surface Figure 5 5 Position of EK H4 temperature/relative humidity sensors in roof Table Table 5 2 lists the identification of all the EK H4 sensors located in the attic. The two multiplexers which can support four sensors each are identified as communication channels 4 and 6 in t he EK H4 Viewer software. Four sensors are connected to the four ports in communication channel 4 and 3 sensors are connected to the communication channel 6.

PAGE 55

55 Table 5 2 EK H4 sensor identification Sensor Number Sensor Location Communication Channel Port # on Multiplexer 1 Sheathing 6 1 2 ccSPF 6 2 3 ccSPF 6 3 4 Ridge Vent 6 4 5 Ridge Vent 4 1 6 ccSPF 4 2 7 Sheathing 4 3 Energy Monitoring The Enetics LD Point End use Energy Disaggregation ) recorder ( Figure ) was used to record the electricity and gas consumption of the test house. The LD 1203AESG unit is wired directly to the circuit breaker panel using current transformers that generates the impulses in the unit. The recorder is compatible with standard 200A, 120/240V, 3 wire, 60 Hz electrical services. The Enetics energy recorder has an input port that interfaces with the pulse output from a gas meter to measure gas consumption. As such a n Elster American Gas Meter AC 250 (with standard 2 feet drive meter) was installed along with an IMAC domestic meter pulser (2 pulses per revolution resulting in 1 pulse/ft 3 of gas consumed). Once installed, th e recorder: Measures voltage and current on each service leg (incoming electricity in the house) and from the Alternate Energy Source (AES) Records and stores aggregate whole premises (grid + AES) kW; and, independently, generated AES kW at user selectable intervals Records and stores aggregate whole premises (grid + AES) kWh consumption; and, independently, generated AES kWh for time of use Records on/off events; (if the combined load (Watt + VAR) change is greater than a configurable th reshold, the recorder records the magnitude of the change, time stamp and the service leg on which the event occurred)

PAGE 56

56 Calculates and stores V rms I rms Watts and VARs (both aggregate grid/AES and alternate energy source) Communicates with the SPEED Fi eld Station or SPEED Master Station software to upload data, view latest readings and obtain latest configuration settings. Figure 5 6 Configuration of LD 1203AESG recorder components The software supplied with of the monitoring unit called Master Station processes the data and analyses the electrical properties of differ ent appliances in the circuit. These signature properties are matched against those in a standard appliance electrical load library to be identified. The load library can be modifie d to include the loads/appliances which have not been included. Once the loads are identified, SPEED Analysis Station software is used for data analysis and to obtain electricity consumption separately for each appliance and for the building premises as a an be extracted using MS Excel. Electricity and gas consumption was recorded at every five

PAGE 57

57 minute interval throughout the duration of project. Figure shows the installed recorder in the test house. A B Figure 5 7 Electricity and gas monitoring. A) Enetics LD 1203AESG recorder installed in the test house and B) Gas Meter with digital pulsar (right) placed in parallel with existing meter (left) Weather Station Figure 5 8 ) was installed on a 15 ft. high metal pole near the north east corner of the building to continuously record the weather data. This station has an Integrated Sensor Suite (ISS) and a wireless receiver that wirelessly transmits all data to the receiver located within the house. The ISS consists of temperature and humidity sensors, rain collector, anemometer, solar radiation sensor, UV sensor and solar panel. The update interval for weather parameters vary with the sensors and all are listed in Table 5 3

PAGE 58

58 Table 5 3 Updat e interval for data transmission from the ISS to the receiver Weather Parameter Update Interval Temperature 1 min Humidity 1 min Barometric Pressure 1 min Dew Point 10 12 seconds Rainfall 20 24 seconds Wind Speed 2.5 3 seconds Wind Direction 2.5 3 seconds Solar Radiation 50 60 seconds UV Radiation 50 60 seconds Figure 5 8 W eather station installed on metal pole near NE corner of building The data from the receiver is logged using the in a computer WeatherLink software at every 15 min interval. The weather data is constantly logged on a laptop in the test house and also continuously uploaded on http://www.wunderground.com/ website. The weather station name on this we bsite is NW Gainesville with a station ID of KFLGAINE15.

PAGE 59

59 Figure 5 9 Data acquisition laptop and Davis Vantage Pro 6162 weather station console Figure 5 9 shows the installed laptop and the wireless weather station console. The laptop logs the data from weather station, EK H4 temperature/ relative humidity sensors and the energy meter. All the data is periodically downloaded from the laptop for analysis.

PAGE 60

60 CHAPTER 6 RESULTS OF EXPERIMENTAL INVE STIGATION The data coll ected and results from the test house during this study are presented in this chapter. Temperature, relative humidity and energy consumption time histories and their ana lyses has been discussed in detail. A timeline of events during the course of this research has been shown in Figure 6 1 and t h e effect of these changes on the temperature, relative humidity and energy consumption are discussed. Figure 6 1 Timeline of events during the course of research Temperature Measurement The daily mean ambient, attic and interior temperatures are shown in Figure 6 2 whereas the temperature time histories are shown in Figure 6 3 Figure 6 6 The average ambient summer (May September) temperature is 80 0 F during 2010 while the averag e attic temperature is about 85 0 F The average house interior temper ature is maintained at about 78 0 F during the same period. The attic temperatures before the installatio n of ccSPF peak up to about 130 0 F but after foam installation on May 25, 2010, the peak attic temperatures drop to about 110 0 F, a 20 0 F reduction in peak attic temperatures. It can also be noted that the re was a reduction in the variation of attic temperatures at the two locations after the installation of foam ( Figure 6 4 ) The mean

PAGE 61

61 temp erature difference at the two locations, before and after foam installation was 4.7 0 F and 1.3 0 F respectively. Figure 6 2 Daily mean t emperature t ime h istory Figure 6 3 Ambient t emperature t ime h istory

PAGE 62

62 Figure 6 4 Attic temperature time h istory The variation in peak attic and ambient temperatures for the months of May through August 2010, are shown in Figure 6 5 Figure 6 5 Daily maximum temperatures

PAGE 63

63 It was found that the peak attic temperatures dropped by about 13 20 0 F after the installation of ccSPF even though the peak ambient temperatures did not vary much during that period ( Table 6 1 ) Table 6 1 Mean of daily maximum temperatures before and after closed cell spray applied polyurethane foam ( ccSPF ) installa tion Temperature ( 0 F) 05/01 05/24 (Before foam installation) 05/25 06/08 (After foam installation) Attic South 115.5 10 2 .3 Attic North 124.5 10 4 .8 Ambient 93.5 94.5 The house interior temperature was controlled by the air conditioning system and hence was not affected by the temperature changes outside the conditioned space. The living room tempe rature fluctuates between 70 75 0 F during winter and 75 8 5 0 F during summer period. The study room temperature varies between 60 75 0 F during winters and 75 85 0 F during summers. A temperature difference of about 8 10 0 F between living room and study room was recorded during December 2009 February 2010 This temperature difference is most likely due to insufficient attic insulation over the study room and air le akage underneath the window sill on the west wall of the room The inflow of cold air from outside reduces the temperature inside study room. Closed cell f oam installation had no noticeable effect on house interior temperature measurements as the conditioned space (occupied living area) is separated from the unconditioned spa ce (attic) by the drywall and blown fiber glass insulation on attic floor. The attic had about R 10 fiber glass blown insulation which did not meet the code requirements of a m inimum R 19 (FBC 2007) insulation in the attic s Additional fiber glass insulation was blown in the attic on December 19, 2010 to meet the code standards. As a result of this, the temperature

PAGE 64

64 difference between living room and study room reduced by about 4 6 0 F from 8 10 0 F during winter in December 2009 February 2010 as shown in Figure 6 6 Figure 6 6 House i nterior temperature t ime h istory of living and study room T he attic gable vents were covered with plastic sheeting on 8 April 2010 to observe the effect o f gable vent ventilation on attic temperatures. T he sheeting was removed from the gable end vents on 25 May 2010 after the ccSPF was installed in the attic However, we could not observe any significant differences in the attic temperatures due to the gable attic vents. Note that ambient sensor was relocated to interior of the house from 16 November through 30 November 2009 The temperature/RH sensor in south attic is located near the ridge vent while the one in north attic is located approximatel y 6 ft. upfront from the gable vent on northern wall. The attic had a much more uniform temperature after the installation of ccSPF as the variation in the ambient and two attic temperatures is reduced after May 2010 ( as listed in Table 6 2 )

PAGE 65

65 Table 6 2 Mo nthly t emperature difference between attic and ambient during 2009 2010 Month Attic South Ambient Attic North Ambient Maximum Difference ( 0 F) Mean Difference ( 0 F) Maximum Difference ( 0 F) Mean Difference ( 0 F) November, 2009 30.6 8.4 22.0 5.6 December, 2009 24.3 8.7 17.0 4.8 January, 2010 48.4 13.0 20.5 6.2 February, 2010 26.2 8.4 27.2 7.8 March, 2010 25.2 7.0 33.1 7.7 April, 2010 32.6 7.7 39.7 9.5 May, 2010 32.4 7.2 43.0 9.7 June, 2010 19.4 4.2 22.2 5.1 July, 2010 32.7 5.6 35.0 6 .0 August, 2010 28.2 4.9 31.0 5.2 September, 2010 25.9 5.2 26.0 5.1 October, 2010 16.1 4.9 14.2 6.0 November, 2010 16.0 5.3 18.0 6.5 December, 2010 35.0 11.3 35.8 11.4 January, 2011 33.6 11.0 34.4 11.5 February, 2011 20.7 6.8 21.5 7.4 The maximum attic and ambient temperature difference of 48.4 0 F (attic temperature higher) occurs in January 2010 for the sensor in south attic while the maximum difference for the temperature sen sor in north attic is only 20.5 0 F for the same month. This difference is because of the proximity of the north attic sensor to the gable end vent through which the cold ambient air enters the attic and reduces the temperature However, the summer maximum temperature difference is for the north attic sensor (43 0 F) because the oak trees surrounding the house provide a shade on the roof over the south attic sensor, thus reducing the solar heat transfer around that region. The temperature relative humidity, rainfall and solar radiation data recorded by the weather station installed at the test house are shown in Appendix C Relative Humidity Measurement The relative humidity time histories recorded by the LogTag sensors in the ambient, attic and house interior are shown in Figure 6 7 through Figure 6 9 A mbient relative humidity varies

PAGE 66

66 from 0 to 100% during winter months A comparison of the house exterior relative humidity with the Gainesville Regional Airport data (obtained from h ttp://cdo.ncdc.noaa.gov/qclcd/QCLCD ) for the same period ( November 2009 February 2011 ) shows the differences in recorded data at the two locations. The monthly average relative humidity recorded at the airport during November January, 2010 is 2% higher tha n that recorded at the house. However, during the summer months, the ambient relative humidity fluctuates between 30 100% due to the humidity in air. The average ambient relative humidity during summer (June September) is 82% while that in the attic is abo ut 60%. During the winter of 2009 (November 2009 February 2010) the ambient and attic relative humidity were 75% and 66% respectively, whereas, in 2010 (November 2010 February 2011) ambient and attic relative humidity were 75% and 61% respectively. The wi nters are cold and dry during the day time but the humidity increases (due to condensation of water vapor) as the temperature lowers during night. Figure 6 7 Ambient RH (%) t ime h istory

PAGE 67

67 Figure 6 8 Attic RH (%) t ime h istory The re was a difference of 3.7% in relative humidity at the two attic locations before the installation of foam, whereas the differential reduced to 2.2% after the installation of foam ( Figure 6 8 ) The relative humidity variation is consistent with the temperature variation in the living and study rooms. At times, there is a dif ference in relative humidity of the two rooms up to about 10 12% both during summer and winter. During summer, the study room has lower relative humidity than the living room whereas, during winter, an opposite trend is observed. This is possibly due to the air leakage underneath the window sill in study room causing the entry of moisture laden cold air and thus increasing the relative humidity of the study room

PAGE 68

68 Figure 6 9 House interior RH (%) time history Electricity Consumption The monthly energy consumption data of the test house for the past two years obtained from the Gainesville Regional Utilities (GRU) was analy zed. The graph in Figure 6 10 compare s the electricity consumption data for 2009 2010. Figure 6 10 GRU electricity consumption data of test house during 2009 2010 0 200 400 600 800 1000 1200 1400 1600 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Electricity (kWh) 2009 2010

PAGE 69

69 It was found that in 2010 there is a 5 % reduction in the total electricity consumption as compared with 2009 The monthly electricity consumption data indicates that the electricity consumption for the months January through April is similar (maximum difference of about 20% ) for both 2009 and 2010. However, t he electricity consumption in June for 2010 was 18% higher than that in 2009 while in July for 2010 was over 37% higher than that in 2009. The reason for the increased energy consumption during June July 2010 was due to a malfunction in the 4 ton HVAC unit th at resulted in abnormally high electricity usage for June and July 2010 When the 4 ton unit was replaced by a new and more efficient 3.5 ton HVAC unit in August 9, 2010 t here was a 22% reduction in total electricity consumption from August to December, 2010 when compared with the same period in 2009. However, it was not clear if t his reduction in energy consumption was a result of the replacement of HVAC system or decrease in attic temperature of the test house due to foam insulation The Eneti cs recorder ( Figure 5 6 ) was used to record the electricity consumption of HVAC and other appliances in the test house. The time history of energy consumption (in kWh) of the HVAC in the experimental house at 15 minute intervals is shown in Figure 6 11 The peak energy consumption of the 4 ton unit increased from about 1.1 kWh to 1.95 kWh from June 8, 2010 due to technical malfunctioning. The existing 4 ton HVAC unit was replaced on August 9, 2010 by a 3.5 ton unit. The total daily energy consumption of the HVAC in the test house during the monitoring period is shown in Figure 6 12 The average daily electricity consumption of the 4 ton HVAC unit was 36 kWh (during May 19 August 8) while that of 3.5 ton unit was 14 kWh (August 9 September 30). The avera ge ambient temperature before the HVAC replac ement (May 19 August 8) is 81.1 0 F while the average temperature after the replacement (August 9 September 30) is about 79.3 0 F. Without a significant decrease in the ambient temperature, the

PAGE 70

70 reduction in the ele ctricity consumption seems to be primarily due to high efficiency of the new HVAC unit. Figure 6 11 HVAC energy consumption history (15 minute interval) Figure 6 12 HVAC electricity vs. ambient temperature during summer As the long term energy monitoring in the house did not give a clear idea of potential energy savings due to the installation of ccSPF only, a short term comparison of the average

PAGE 71

71 daily energy usage was made. The average daily electricity consumption by th e HVAC and house as a whole, for a week before and after the installation of ccSPF is tabulated in Table 6 3 Table 6 3 Average daily e lectricity consumption a week before and after ccSPF installation Before (05/1 9 05/24/2010) After (05/25 05/31/2010) Percentage Reduction (%) Premises (kWh) 42.2 30.8 27.0 HVAC (kWh) 22.5 16.5 26.7 The reduction in average daily electricity consumption by about 27% suggests the effectiveness ccSPF in possible energy savings besid es being a structural retrofit. T his short term observation is important as it gives an insight of the potential energy savings in the long run. A verage cumulative energy consumption of 4 ton HVAC unit, before and after ccSPF installation and after replacement by the 3.5 ton unit during a day is shown in Figure 6 1 3 The energy data recorded during June 8 August 8, 2010 (the period of HVAC malfunction) was removed from the average daily cumulative energy consumption calculations. The average daily cumulative energy consumption of the 4 ton HVAC unit was 23.4 kWh before the installation of ccSPF in the attic (during May 19 May 24, 2010) There was a reduction in the average daily energy consumption of HVAC after the installation of ccSPF by 6.4 kWh (27%) and it was 17 kWh during May 25 June 7 2010 (calculated after removing June 8 August 8, 2010 energy consumption dataset). The electricity consumed by the new 3.5 ton HVAC was 13.7 kWh during August 9 September 30, 2010. Further reduction of energy consumption by the 3.5 ton unit is attribute d by its lower capacity than the 4 ton unit thus reducing the total energy consumption of the house

PAGE 72

72 Figure 6 13 Average daily cumulative HVAC energy consumption To understand the variation of HVAC electricity consumption due to temperature effects, e lectricity consumption at every 15 minute s interval was averaged over a span of several days for the 4 ton HVAC unit and before and after ccSPF installation in the attic and also for the 3.5 ton HVAC unit ( Figure 6 14 Figure 6 15 and Figure 6 16 ) In addition to the energy profile, the average ambient house interior temperature difference for that period was also calculated and plotted at 15 minutes interval during the day. The temperature difference increases as the day progresses and peaks at aro und 3 pm every day. The peak difference in the attic house inter ior temperature reduced from 43 0 F to 27 0 F after the installation of ccSPF. At night, the ambient and attic temperatures are lower than the house interior temperature and hence a negative tempe rature difference can be observed in the temperature difference variation during a day. The minimum temperature difference between the ambient and house interior and attic and house interior occurs around 6 am every day.

PAGE 73

73 Figure 6 14 Average distribution of 4 ton HVAC unit energy consumption at 15 minute interval during May19 May 24, 2010 and ambient interior temperature difference Figure 6 15 Average distribution of 4 ton HVAC energy consumption at 15 min interval during May 25 June 8, 201 0, ambient interior and attic interior temperature difference

PAGE 74

74 The daily energy usage pattern shows a regular trend which is consistent before and after ccSPF installation and also after the replacement of 4 ton HVAC; gr eater energy consumption at night and during the day after 2 pm. The thermostat temperature is lowered to 75 0 F from 78 0 F by the homeowners at night, thus increasing the cooling load on HVAC which results in greater electricity consumption at night. As th e day progresses, the ambient to house interior temperature difference increases which in turn increases the cooling load on HVAC. Figure 6 16 Average distribution of 3.5 ton HVAC unit energy consumption at 15 min in terval during Aug 9 Sep 30, 2010, ambient interior and attic interior temperature difference The cooling energy use of the test house during the summer period was normalized to compare the HVAC energy usage pattern before and after the ccSPF installation. The cooling energy was normalized against three different measures of temperatures A) positive ambient interior temperature difference, B) peak ambient temperature and C) time integrated temperature difference. The three normalization techniques and the results of normalization are explained in the following paragraphs

PAGE 75

75 The calculation of cooling energy normalized against the positive ambient house interior temperature difference was done by taking the ratio of energy and temperature difference at every 15 minute interval Figure 6 17 shows the daily HVAC energy consumption of the 4 ton unit during May 19 May 24, 2010, for the hours of the day during which there was a positive ambient interior temperature difference. T here are no normalized values during 12 9:30 am and 9 pm to 12 am as the ambient temperature is lower than the house interior temperature, resulting in a nega tive temperature difference. Figure 6 17 Average HVAC energy consumption at 15 minute interval for positive ambient interior temperature difference for 4 ton unit during May19 May 24, 2010 The normalized cooling energy for the 4 ton and 3.5 ton is also calculated in the same way by taking the ratio of HVAC energy consumption to the respective temperature difference and the normalized values are plotted in Fig ure 6 18 The sum of normalized values at every 15 minute intervals gives the total normalized energy of HVAC during a day. The daily total normalized energy for the 4 ton HVAC before the installation of ccSPF is 4 kWh/ 0 F and 3

PAGE 76

76 kWh/ 0 F after the installatio n of foam. This implies a reduction of 25% in cooling energy due to the installation of foam. The normalized energy consumption for the 3.5 ton HVAC is 1.1 kWh which is 63% less than the normalized energy of 4 ton unit after the installation of ccSPF. F igure 6 18 Normalized HVAC energy against ambient house interior positive temperature difference The cooling energy demand of the house was also normalized against the daily peak and time integrat ed ambient interior temperature differences to understand the effect of HVAC capacity on cooling energy requirements of the house due to changes in the att ic temperatures. The cooling energy was normalized by taking the ratio of total daily cooling energy to the daily peak temperature difference and time integrat ed temperature difference. The daily cumulative cooling energy of the house before and after the installation of ccSPF is shown in Figure 6 19. The time integrat ed temperature difference was calculated by integrating the ambient interior temperature differ ence over time as shown by the shaded area in Figure 6 20 Table 6 4 shows the values and calculation procedure to obtain the normalized cooling energy. Column A lists the average daily cooling energy requirements of the house, column B shows the peak temp erature

PAGE 77

77 difference and column C gives the time averaged temperature difference. The normalized cooling energy in columns D and E is calculated by taking the ratios A/B and A/C respectively Figure 6 19 Average daily cumulative cooling energy consumptio n before and after the installation of ccSPF Figure 6 20 Calculation of time integrat ed temperature difference by integrating the ambient interior temperature difference over time

PAGE 78

78 Table 6 4 Normalized cooling energy Avg. Daily Cooling Energy (kWh) Peak Temp Diff ( 0 F) Time Integrate d Temp Diff ( 0 F h) Normalized HVAC Cooling Energy ( kWh/ 0 F) ( kW/ 0 F) A B C D=A/B E=A/C 4 ton (May 19 24, 2010) 12.4 14.1 88.8 0.88 0.14 4 t on (May 25 June 8, 2010) 6.3 13.9 77.3 0.45 0.08 3.5 t on (Aug 9 Sep 30, 2010) 6.2 10.7 68.8 0.58 0.09 The average daily cooling energy consumption of the house was reduced from 12.4 kWh by 49% to 6.3 kWh after the installation of ccSPF. The peak temperature normalized cooling energy also reduced by 49% from 0.88 kWh/ 0 F to 0.45 kWh/ 0 F for the 4 ton HVAC unit There was a reduction of 43% in the time integrat ed temperature difference normalized cooling energy (from 0.14 kW/ 0 F to 0.04 kW/ 0 F) for the 4 ton HVAC unit after the installation of ccSPF. The average daily cooling energy consumption of the house was 0.1 kWh lower for the 3.5 ton HVAC unit than the 4 ton unit. However, the peak temperature difference normalized cooling energy for the 3.5 ton unit was 29% higher than the 4 ton unit after foam installation whereas the time integrated temperature difference normalized cooling energy for 3.5 ton HVAC was 12% higher than the 4 ton unit Gas Consumption The test house has a furnace and wat er heater that uses natural gas. During the test period the owners also installed a gas dryer on April 13, 2010. The monthly gas consumption of the test house for the years 2009 and 2010 is shown in Figure 6 2 1 The total gas consumption during 2009 and 20 10 were 477 and 621 Therms respectively. Forty seven percent higher gas consumption was recorded in the test house during January May, 2010 (before the installation of ccSPF) than in 2009. However, ccSPF installation seemed to have no effect on the gas

PAGE 79

79 con sumption of the test house as 159 and 154 Therms of gas units were consumed during June Decem ber 2009 and 2010 respectively. Figure 6 2 1 GRU gas consumption data of test house during 2009 2010 The consumption of gas is higher in the cold months of the year as it is used in the furnace for heating the house. This trend is opposite to that of electricity consumption, over 50% of which is used in cooling the house. Discussion of Experimental Result s T he monitoring of temperature/RH variation and energy consumption of the test house helps to better understand the thermal and energy performance of a vented residential roof retrofitted with closed cell polyurethane foam. The temperature monitoring in the attic before and after the installation of l evel II ccSPF retrofit on the underside of roof sheathing showed a reduction of peak attic te mperatures during summer by 13 20 0 F after the foam installation The research was not able to determine the cause of a mean 4.7 0 F difference in two attic temperature s before the installation of ccSPF except that 0 25 50 75 100 125 150 175 200 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Gas (Therms) 2009 2010

PAGE 80

80 the sensor showing the lower peak temperature was located close to the roof ridge, and the roof region direc tly above it is shaded by live oak trees. However, after the installation of ccSPF, the difference in attic te mperatures reduced to about 1.3 0 F. A temperature difference of about 8 10 0 F between living room and study room was recorded during December 2009 F ebruary 2010 This temperature difference is most likely due to insufficient attic insulation over the study room and air leakage underneath the window sill on the west wall of the room The a dditional fiber glass insulation blown on the attic floor on Dec ember 19, 2010 reduced the temperature difference between living room and study room reduced by about 4 0 F during December 2010 February 2011 House energy consumption data from Gainesville Regional Utilities and HVAC electricity consumption data as recorded by the installed energy meter were also analyzed. A total energy savings of about 22% was recorded during August December 2010 over last yea consumption during the same period. S hort term energy monitoring for a week before and after the foam installation showed up to 27% decrease in HVAC energy consumption. Due to technical malfunctioning, the existing 4 t on HVAC unit was repla ced by a new and energy efficient 3.5 t on unit. Hence, the total energy savings in heating and cooling the test house due to foam installation alone could not be exactly determined However, the analysis of HVAC electricity co nsumption data after removing the data set for the period of June 8 August 8, 2010 due to malfunctioning of existing 4 ton, showed a reduction of 49 % in cooling energy use after foam installation. A slight reduction (about 2%) in the daily cooling energy consumption of the 3.5 ton HVAC unit was also observed. The results of the three techniques of normalization of energy con sumption of HVAC against the ambient interior temperature difference showed about 27 % 49 % reduction in cooling energy consumption of the 4 ton HVAC unit after the ins tallation

PAGE 81

81 of ccSPF. T he average daily cooling energy consumption for the 3.5 ton unit is less than the 4 ton unit after ccSPF installation in the test house. However, the higher values of the peak and time integrated temperature difference normalized energy consumption for the 3.5 ton HVAC could not be explained.

PAGE 82

82 CHAPTER 7 ANALYTICAL MODELING IN WUFI PRO 4.2 Introduction to WUFI Pro 4.2 WUFI Pro 4.2 is a tr ansient, one dimensional heat and moisture transfer model ing software that enables predict ion of the hygrothermal behavior of building envelope systems It was jointly developed by the Fraunhofer Institute for Building Physics, Germany and Oak Ridge Nation al Laboratory, USA and has proved to be a powerful tool for advanced hygrothermal analysis of building envelope systems. The model is able to estimate the drying times of construction materials, temperature gradient across the components and even simulate the effect of wind driven rain on exterior building components using the moisture and energy transport equations shown in Equation 7 1 and 7 2 respectively The moisture transport equation is function of liquid conduction and vapor flux whereas the energy transport equation is a function of conductive heat and enthalpy flux by vapor diffusion The vapor flux is the rate of water vapor transfer across a surface t hrough vapor diffusion which in turn is dependent on the relative humidity and vapor pressure an d vapor permeability of the material The conductive heat flux is calculated as the product of thermal conductivity of the material and temperature gradient across it. The enthalpy heat flux by vapor diffusion is a measure of heat transfer through vapor di ffusion across the material and is calculated as a product of latent heat of phase change and vapor flux. ( 7 1) (7 2 ) W here,

PAGE 83

83 Analytical programs such as WUFI are useful for developing numerical models of various wall, roof and building envelope assemblies to compare the performance in various climatic conditions by comparing different design models. The use of WUFI for this purpose has been validated in previou s studies ( Karagiozis ( 2001 ) Finch et al. ( 2007 ) Straube et al. ( 2009 ) ) for residential building envelope systems. Model Inputs The models use a weather data file, generated for a specific area and climate condition to model the ambient condition outside of a building, and the set interior (conditioned) temperature and humidity conditions. These provide the initial boundary conditi ons for the models and variations of main parameters over time of the experiment. WUFI is able to very quickly simulate the daily fluctuations in multiple parameters that occur over years (i.e. a 10 year simulation may take only 2 3 minutes), and as such u sers can quickly develop multiple models for comparison and further study A dditional input s to the WUFI model are the standard material properties of the materials or components used in the building envelope cross section, such as material density, heat c apacity, permeability, porosity, and thermal conductivity. Note that this version of WUFI is

PAGE 84

84 only a 1 dimensional tool and therefore certain properties may not be accurately modeled (i.e. in a roof the effect of the roof framing depth versus the depth of t he wood sheathing would be different). The WUFI Pro 4.2 program has a built in database containing material properties for a range of common building construction materials, which can be selected and used in the models. Th e material database was created using the material properties from (ASHRAE 2002) The cross section of the surface can also be modeled with a specific orientation (to account for solar and wind impact), and the assembly of multiple layer of components must also be specified. In addition, the WUFI program includes a database of the exterior climate conditions for 64 cities in North America, providing hourly weather data. These standard climates can be used bu t the program also allows for a user created weather file to be used, where climactic i nformation is not available for a specific location. To create the weather file for a specific location, hourly temperature, relative humidity, rainfall, wind speed and direction, barometric pressure and solar radiation have to be defined ( WUFI Pro 4.2 Onl ine Help ) Calculation Procedure and Results The moisture and energy transport equations are solved numerically at each time step using a finite difference approach in which the transport equations are discretized with variable grid spacing across the cros s section of materials This approach uses the Thomas algorithm for tri diagonal matrices for solving the coupled equations which is a simplified form of Gaussian elimination Kuenzel et al. ( 2001 ) The transport and storage coefficients are continuously updated a t each time step until convergence criteria are achieved. WUFI gives users the option for monitoring the temperature and moisture results at user defined locations along the cross section of the model. The results ar e available as time histories

PAGE 85

85 of temperature and moisture or as their animated profiles over the cross section of building components. Application of WUFI to Predict Roof P erformance There is a growing trend in residential construction to maximize the int erior space by building compact steep sloped roof systems and eliminating vented attics. The elimination of attic ventilation may lead to moisture accumulation in case of incidental leakage and thus reduce the life of roofing components and materials. To u nderstand the effect of venting on the drying potential of roof sheathing, Nelson and der Ananian ( 2009 ) p erformed a comparative study of the performance of roof systems using WUFI by simulating an incidental water leakage event on the roof 24 roof assemblies we re tested for their long term performance due to difference s in insulation material ( glass fiber batt, op en cell a nd closed cell spray applied polyurethane foam), weather resistive barrie r (felt, and self adhering rubberized asphalt membrane ( SRAM) ) sheat hing type (plywood and oriented strand board ) and two attic configurations (vented and unvented) in a cold climate (Boston, MA) and hot humid climate (Miami, FL) The un vented and vented roof systems were modeled as shown in Figure 7 1 and Figure 7 2 respectively Roof models consist of asphalt shingles installed over felt or SRAM underlayment on in. OSB or plywood sheathing. Glass fiber batt, open cell or closed cell foam was installed underneath the roof sheathing in unvented roof assemblies. An ai r layer with five air changes per hour was inserted between the insulation and sheathing in vented roof configurations. The interior side of the roof had gypsum board installed to finish the roof system. For roof models in cold climate, a vapor retarder is placed on the interior side of the roof above the gypsum board to prevent the moisture drive towards the interior of house.

PAGE 86

86 Figure 7 1 Unvented a sphalt shingle roof model (after Nelson and Der Ananian 2009) Figure 7 2 Vented asphalt shingle roof model ( after Nelson and Der Ananian 2009)

PAGE 87

87 The material properties of the components are taken from the WUFI material database. The shingles were modeled as an external surface coating with a permeability of 0.5 perms. The exterior climate file for the two cities Miami, FL and Boston, MA is defined in the weather database and was used in these simulations. The interior conditions were set at a temperature of 70 2 0 F a nd relative humidity of 35 15 %. The roof assembly simulations were run for ten years with a two year initial pre conditioning period to minimize the effects of initial construction moisture content on the results. As the objective of this study was to compare the drying time of moisture sensitive roof elements, a single event leak of an average 0.038 in/h for a total duration of 8 hours water was introduced on the roof sheathing. The leak was introduced in the third summer of each ten year simulation. Short term decay in most building materials can be prevented if the relative humidit y levels are below 80% in the components (Sedlbauer 2002) The authors calculated the drying times of compone nts as the time for the relative humidity level in the material to go below the 80% threshold after the leak is introduced in the third summer of simulations This 80% relative humidity level corresponds to 4 lb/ft 3 water content in plywood and 5.2 lb/ft 3 water content in the OSB which is also the typical built in water content of plywood/ OSB at the time of construction (WUFI material database). Their results showed that vented roof assemblies were more tolerant and durable in case of incidental water leaka ge than the unvented roof assemblies in both hot and cold climates. A vented asphalt shingle ro of with felt underlayment, p lywood /OSB sheathing and ccSPF insulation had a drying time of less than 1.5 months whereas a similar roof with unvented attic had a drying time of over 7 months in case of an incidental leakage in a hot climate of Miami,

PAGE 88

88 FL. The drying times for similar roofs were less than 2 months and 14 months respectively for a vented and unvented attic in a cold climate of Boston, MA Numerical Study Using WUFI Two numerical studies using WUFI Pro 4.2 were conducted for this research The first was a repeat of Nelson and der A nanian model discussed in the previous section to confirm the performance of WUFI. The second was the actual model of the experimental roof system and conditions. Model 1 The roof models developed by Nelson and der Ananian (discussed in previous section) w ere reproduced to validate the performance of WUFI and form a basis to develop roof models for this research Similar roof construction, material properties, weather data, amount of incidental leakage, attic ventilation rate were used to run roof simulatio ns in the hot humid climate of Miami, for a total duration of ten years including a two year preconditioning period. The drying time of roof sheathing was determined from the wate r content results obtained at the end of each simulation by calculating the t ime taken for relative humidity level to go below 80% after the occurrence of incidental leakage. The drying times of different roof models as determined by Nelson and der Ananian are compared against those obtained from roof models developed at UF and lis ted in Table 7 1 The simulation results of models developed at UF to replicate the findings of Nelson and der Ananian match up with fairly well with their results. The UF model of unvented closed cell polyurethane foam (ccSPF) retrofitted roof with felt underlayment predicted longer drying time of plywood sheathing by 3 months than that predicted by Nelson and der Anania n. Also, the vented roof model with closed cell polyurethane, SRAM and plywood predicted one month

PAGE 89

89 Table 7 1 Comparison of predicted drying times of sheathing in roof models developed at UF to tho se predicted by Nelson and der Ananian. Insulation Type Vented/ Unvented Weather Resistive Barrier Sheathing Type Drying time of sheathing after roof leak (months) Nelson and der Ananian R esults UF R esults Glass fiber Unvented Felt OSB >1 >1 batt Plywood >1 >1 SRAM OSB >1.5 >1.5 Plywood >1 >1 Vented Felt OSB <1.5 <1.5 Plywood <1.5 >1 SRAM OSB 1.5 1.5 Plywood 1.5 <1.5 Open cell Unvented Felt OSB >1.5 1.5 polyurethane Plywood >1.5 1.5 SRAM OSB >2 <2 Plywood <2 >1.5 Vented Felt OSB 1.5 <1.5 Plywood 1.5 1.5 SRAM OSB 1.5 1.5 Plywood >1 <1.5 Closed cell Unvented Felt OSB >7 8 polyurethane Plywood >7 10 SRAM OSB >12 12 Plywood >12 12 Vented Felt OSB <1.5 < 1.5 Plywood <1.5 >1 SRAM OSB 2 >1.5 Plywood 2 1 Th e comparison of drying times of sheathing and thus the validation of the 24 roof models helped us to model the experimental roof (Model 2) by using some of the unknown parameters like shingle permeability, air ventilation rate and amount of incidental water leakage. Model 2 An analytical model was developed in WUFI Pro 4.2 and used to recreate the test conditions in the experimental roof system The purpose of this model was to predict the hygrothermal performance of test roof using the experimentally collected temperature/ relative humidity and weather data. The analytical model is first validated by comparing the surface temperature data recorded by the EK H4 sensors at the plyw ood ccSPF interface and also on the

PAGE 90

90 underside of ccSPF layer exposed in the attic Once this model was validated, twelve different roof systems were m odeled to predict the long term performance (drying potential of roof components) of the roof in case of a n incidental water leakage. The roof assembly, model inputs, results and validation of the experimental model are discussed in the following sections. Roof Assembly As discussed in Chapter 4, the test roof is a vented asphalt shingle roof system with felt und erlayment and OSB sheathing. A l evel II ccSPF retrofit was installed on the underside of OSB sheathing in May 2010. Figure 7 3 shows the elevation and cross section of the vented retrofitted roof in test house with 1 in. closed cell sprayed polyurethane foam retrofit. A B Figure 7 3 Test r oof model A) elevation and B) section (A A) o f vented roof with retrofit The roof is modeled for a cross section between the roof trusses as WUFI Pro 4.2 is capable of one dimensional calculation only. Table 7 2 Roof components of the numerical model Component Thickness Asphalt Shingles 0.125 in Felt Underlayment 0.039 in OSB/ Plywood 0.500 in Sprayed Polyurethane Foam closed cell (2 .0 lb/ft 3 ) 1.000 in Vented Air Space (0.08 lb/ft 3 ) 2.000 in The simulated roof model is oriented in east direction with an inclination of 18 0 to the horizontal to closely resemble the experimental conditions. The components of roof along the Plywood ( 1/2 Air Layer Roof covering

PAGE 91

91 cross section have been listed below with their thickness and density as used from the WUFI material database and are shown in Table 7 2 Input and Simulation Parameters The attic air change rate was not monitored in the test house for the duration of experimental study. However, a constant external air change ra te of 5 changes/hour was added to the air layer in order to simulate attic ventilation bas ed on the roof models developed by (Nelson and der Ananian 2009) The exterior climate file for Gainesville, FL, was cr eated using the weather station data for 2010 from the test house. Hourly t emperature, relative humidity, wind speed and direction, rainfall intensity, air pressure and solar radiation a re the input s for the exterior climate file. In Figure 7 4 the scatte r in red shows the hourly distribution of ambient temperature whereas the red line gives the daily mean temperature. Hourly ambient relative humidity variation is shown by the dark blue scatter whereas the lighter blue line plots the average daily relative humidity. The annual average temperature for Gainesville is 67.2 0 F and the average relative humidity is about 75%. The maximum and minimum temperatures record ed during the year 2010 were 99 0 F and 23.2 0 F The interior climate for the roof model is taken as the experimentally measured attic interior temperature and relative humidity data recorded only for 2010 ( Figure 7 5 ). The roof simulations are modeled for a total duration of ten years with a two year in itial pre conditioning period. The preconditioning period provides a more accurate long term simulation results as the effects of initial conditions (temperature and relative humidity of each material layer) are minimized during the initial time period (Nelson and der Ananian 2009)

PAGE 92

92 Figure 7 4 Screenshot of e xterior climate file for Gainesville (2010) Figure 7 5 Screenshot of interior climate file (from recorded atti c temperature/relative humidity at the test house )

PAGE 93

93 Results of WUFI model The simulation results are shown graphically by plotting the temperature and relative humidity of each component through roof assembly on y axis vs. time on x axis. Figure 7 6 Figure 7 7 shows the variation of predicted temperature and humidity in plywood and ccSPF respectively, for a one year period (1/1/2010 to 1/1/2011). The red band in the graph is the hourly temperature variation while the green one shows the relative humidity of plywood The temperature of plywood from the annual simulation shows that the peak plywood surface temperatures may go up to about 145 0 F during the summer The predicte d temperature and relative humidity of ccSPF layer follow the trend of interior climate conditions as it forms the boundary condition of roof model on the inside. Figure 7 6 Predicted t emperature and relative humidity of plywood at the plywood foam interface in test house roof

PAGE 94

94 Figure 7 7 Predicted surface temperature and relative humidity of ccSPF layer in test house roof A comparison of the predicted temperature of plywood ccSPF interface and the temperature of ccSPF on the underside (exposed to attic) with the experimentally measured values is shown in Figure 7 8 through Figure 7 10 The results of one dimensional simulation s capture the diurnal and weather changes effectively as is evident from th is graph An important obs ervation from Figure 7 8 Figure is th e difference between the predicted and measured daily peak temperatures of plywood The daily peak of the predicte d temperatures showed in blue in Figure 7 8 are 10 15 0 F higher than the experimentally measured plywood temperatures (red line) in the house. However, the daily minimum temperatures are predicted more accurately with a difference of about 2 5 0 F from the measured temperature values. A comparison of temperatures for a longer duration could not be done as experimental temperature data could not be recorded by the EK H4 surface temperature sensors

PAGE 95

95 Figure 7 8 Measured vs. p redicted surface t emp erature of plywood at the plywood foam interface in the attic of test house Figure 7 9 Measured t emperature (EK H4 Sensor 2) vs. p redicted surface t emperature of ccSPF exposed in t he attic of test house

PAGE 96

96 Figure 7 10 Measured t emperature (EK H4 Sensor 3) vs. p redicted surface t emperature of ccSPF exposed in the attic of test house The difference in daily maximum temperatures is most likely due to the effect of solar radiation discussed as follows: 1. The weather station recording the solar radiation is located in an open area not at all shaded by the trees, whereas, a large portion of the roof is shaded by oak trees. 2. The amount of radiation received on a horizontal surface (solar radiation sensor on t he weather station) is different from that on an inclined surface of the roof. As there is no solar radiation at night, the predicted values there is a better match of minimum temperatures at night and the predicted temperatures are within a difference of 5 0 F from the measured values. The predicted temperatures of ccSPF compared with the measured temperatures by two sensors are shown in Figure 7 9 and Figure 7 10 The predicted ccSPF surface temperature is up to 10 0 F lower than the measured ccSPF temperat ures at the two locations in the attic. A possible explanation for higher measured temperatures than the predicted temperatures at night is that the

PAGE 97

97 blown fiber insulation absorbs and stores the heat gained during the day and releasing the absorbed heat at night when the ambient temperature goes down This effect could not be incorporated in the one dimensional roof model. Also, the effect of framing members and irregularities in foam thickness cannot be taken into account in this roof model. WUFI Pro 4.2 s imulates the thermal movement across the roof components in a one dimensional building assembly. However, all the structures are three dimensional and the effect of several factors like framing members, additional blown insulation, varying air movement, et c. cannot be all taken into account in a simple one dimensional roof model. The WUFI simulations produced a fair agreement betwee n the measured and predicted plywood and ccSPF surface temperatures for the vented asphalt shingle roof system of the test hous e. Prediction of Long Term Hygrothermal P erformance of Different R oof C onfigurations As discussed in Chapter 3, three levels of ccSPF retrofit have been approved by the Florida Product Approvals for installation in residential houses. The three levels of ccSPF retrofit are a) 3 in. fillet at the truss/rafter and sheathing intersection, b) 3 in. fillet with 1 in. uniform layer on the underside of roof sheathing and c) 3 in. uniform layer. Homeowners have the discretion of installing the level of ccSPF prot ection they want for their roofs. In order to predict the long term hygrothermal performance of ccSPF retrofitted and non retrofitted roof s, six different roof configurations were selected for the numerical modeling. These roof s were analyzed for their thermal performance in a hot humid climate of Gainesville using the weather data collected during the course of this research The effect of ventilation and incidental leakage on the drying time of OSB sheathing was also analyzed Roof C onfigurations Twel ve asphalt shingle roof configurations were modeled to predict their long term performance by specifically calculating the drying times of these roofs The roofs are asphalt

PAGE 98

98 shingle roof systems with felt underlayment installed over in. thick OSB sheathi ng or plywood The construction of the twelve simulated roofs from exterior surface (exposed to environment) to interior surface (in attic) with the variation in level of insulation and ventilation is described as follows : R1 Asphalt shingles, 15 lb. felt underlayment, in. OSB, Vented R 2 Asphalt shingles, 15 lb. felt underlayment, in. OSB, Unvented R 3 Asphalt shingles, 15 lb. felt underlayment, in. OSB, 1 in. ccSPF insulation, Vented R 4 Asphalt shingles, 15 lb. felt underlayment, in. OSB, 1 in. ccSPF insulation, Unvented R 5 Asphalt shingles, 15 lb. felt underlayment, in. OSB, 3 in. ccSPF insulation, Vented R 6 Asphalt shingles, 15 lb. felt underlayment, in. OSB, 1 in. ccSPF insu lation, Unvented. R7 Asphalt shingles, 15 lb. felt underlayment, in. Plywood, Vented. R8 Asphalt shingles, 15 lb. felt underlayment, in. Plywood, Unvented. R9 Asphalt shingles, 15 lb. felt underlayment, in. Plywood, 1 in. ccSPF insulation, Vented. R10 Asphalt shingles, 15 lb. felt underlayment, in. Plywood, 1 in. ccSPF insulation, Unvented. R11 Asphalt shingles, 15 lb. felt underlayment, in. Plywood, 3 in. ccSPF insulation, Vented. R12 Asphalt shingles, 15 lb. felt underlayment, in. Plywoo d, 1 in. ccSPF insulation, Unvented. The material properties for the roof components are assigned from the WUFI material database except that for the permeability of asphalt shingles and felt underlayment which are taken as 0.5 perm and 5 perm respectively for all the roof models in numerical study

PAGE 99

99 Simulation Parameters The vented roof s were modeled in the same way as the test house roof model but the analy sis was run fo r a total duration of 10 years starting on January 1, 2009 and ending on December 31, 2018 A constant air ventilation of 5 air changes per hour was used to simulate attic ventilation in vented roof models Attic ventilation was simulated by defining an external air source applied unif ormly through the cross section of air layer. For the unvented roofs, there was no air layer in the roof models and hence no attic ventilation The incidental leakage in the event of a high wind event was modeled by introducing a leak of 0.038 in/hour for a total duration of 8 hours in the third summer on the surface of the sheathing A two year preconditioning period was also selected before the introduction of leakage on the sheathing surface to minimize the effect of initial moisture content on the simul ation results. The exterior climate of the test house as recorded by the installed weather station was selected for these simulations as well. The recorded data was at 15 minute interval whereas for the WUFI climate file, the data was modified to have data points at 1 hour interval. However, the interior climate was redefined in these twelve roof models as the experimentally collected attic temperature and relative humidity data corresponds to the condition of un retrofitted roof system during January to Ma y 2010 and then to that of a ccSPF retrofit ted roof system Hence, the interior conditions were set at a temperature of 70 2 0 F and relative humidity of 35 15% for all the roofs (Nelson and der Ananian 2009) Results The simulations calculate the water content of each roof component for the duration of analysis. As we are concerned with the moisture build up in the roof sheathing due to incidental water leakage, the drying times of sheathing is calculated as th e time for relative humidity to go below 80% which is equivalent to 4 lb/ft 3 water content in plywood and 5.2 lb/ft 3 in OSB. The

PAGE 100

100 screenshot of water content of OSB sheathing for vented non retrofitted roof is shown in Figure 7 11 The plot shows the 80% relative humidity level in OSB and the time it takes for the water content to go below the 80% mark after the introduction of leak in the third summer of simulation period. The water content plots of roofs with OSB are shown Figure 7 12 while the water content of plywood roof systems is shown in Figure 7 13 In all the results, the water content of the sheathing increases to reach the peak water content and drops down sharply after the leak is stopped after 8 hours The water content in OSB or plywood sheathing then reduces slowly by the evaporation of water to the outside of roof (through the shingles) for ccSPF retrofitted roof systems and inside the at tic for the non retrofitted roofs. Figure 7 11 Pr edicted water content of OSB sheathing in vented, non retrofitted roof 80% RH Leak introduced in 3 rd summer (2012) Drying Time

PAGE 101

101 A B C D E F Figure 7 12 Water content of oriented strand board (OSB) sheathing in simulated asphalt shingle roof models A) non retrofitted, vented, B) non retrofitted, unvented, C) 1 in. ccSPF retrofit, vented, D) 1 in. ccSPF retrofit, unvented, E) 3 in. c cSPF retrofit, vented and F) 3 in. ccSPF retrofit, unvented 2009 2009 2009 2009 2009 2009

PAGE 102

102 A B C D E F Figure 7 13 Water content of plywood sheathing in simulated asphalt shingle roof models A) non retrofitted, vented, B) non retrofitted, unvented, C) 1 in. ccSPF retrofit, vented, D) 1 in. ccSPF retrofit, unvented, E) 3 in. ccSPF retrofit, vente d and F) 3 in. ccSPF retrofit, unvented The peak moisture and water content and the drying times of the OSB and plywood sheathing for the simulated ccSPF retrofitted and non retrofitted vented/unvented roof systems are listed in Table 7 3 2009 2009 2009 2009 2009 2009

PAGE 103

103 Table 7 3 Peak moisture content and drying times of oriented strand board ( OSB ) and plywood sheathing in the simulated vented and unvented roof models at UF Roof Sheathing Insulation Ventilation Peak Moisture Content (Water Content) Drying time R1 OSB None Vented 30% (12 lb/ft 3 ) 5 months R2 OSB None Unvented 15% (6 lb/ft 3 ) 3 months R3 OSB 1 in. ccSPF Vented 33% (13 lb/ft 3 ) 24 months R4 OSB 1 in. ccSPF Unvented 30% (12 lb/ft 3 ) 15 months R5 OSB 3 in. ccSPF Vented 58% (23 lb/ft 3 ) 6 years R6 OSB 3 in. ccSPF Unvented 28% (11 lb/ft 3 ) 7 years R7 Plywood None Vented 25% (7.4 lb/ft 3 ) 4 months R8 Plywood None Unvented 1 9 % (5.5 lb/ft 3 ) 4 months R9 Plywood 1 in. ccSPF Vented 28 % ( 8.2 lb/ft 3 ) 16 months R10 Plywood 1 in. ccSPF Unvented 25% (7.3 lb/ft 3 ) 24 months R11 Plywood 3 in. ccSPF Vented 63 % ( 18.3 lb/ft 3 ) 6 years R12 Plywood 3 in. ccSPF Unvented 41 % ( 11.8 lb/ft 3 ) 3 years The drying time and peak moisture content results show that the ccSPF retrofitted roofs are prone to moisture accumulation between the foam and sheathing. The presence of foam on the underside of sheathing prevents the drying of sheathing on the inside of attic. The only drying that occurs is on the exterior side of sheathing through the shingles. The drying time of OSB and plywood sheathing s in vented non retrofitted roofs are 5 and 4 months respectively whereas the drying time in retrofitted roofs varies from 15 months to up to 6 years The un vented roof s with OSB sheathing have reduced drying times than the vented attics for non retrofitted and 1 in. foam retrofitted roof but drying time of plywood is same in both vented and unvented roofs with no ccSPF The OSB sheathing of in vented roof with 3 in. foam has a year lower drying time than unvented roof whereas the plywood shows an opposite trend with three year lesser drying time in unvented roof over the vented one The results obtained by Nelson a nd Der Ananian showed lower drying times of roof sheathing in vented ccSPF retrofitted roofs than the unvented roofs in case of an incidental water leakage as the vented roof models had a ventilation layer right underneath the sheathing T he roof

PAGE 104

104 model test ed and validated in this research uses an air layer on the underside of ccSPF as the foam is applied directly to the wood sheathing without a ventilation gap in between the two layers.

PAGE 105

105 CHAPTER 8 CONCLUSIONS AND RECOMMENDATIO N S This chapter provides a summ ary of the experimental and analytical investigation undertaken and discusses the main conclusions that can be drawn from this research. The limitations of the work and r ecommendations for future work are also discussed. Summary The focus of this research was to evaluate the energy benefits of the ccSPF structural retrofit and also predict the thermal and moisture performance of foam retrofitted roof systems The work tested the hypothesis that a significant reduction in the attic temperature will result i n reduced energy usage to cool the home. The hypothesis was based on the assumption that the condition of the air ducts, particularly in older homes, there is a possibility that hot air from the attic can be drawn into the ductwork and thereby make the HVAC system work harder to maintain the set temperature. The roof of the test house was carefully inspected by a professional roofer prior to initiating this project. Having confirmed the roof was in structurally sound condition, the research ers proceeded to install the ccSPF structural retrofit. The l evel II protection used (1 in. layer plus a 3 in. fillet at framing members) was shown by previous testing to provide a reasonable increase in wind uplift capacity (about 2 5 time increase for this roof system ). It was found that the spray foam (as applied overhead) in this roof attic is not as uniform a fter application as was observed in previous laboratory testing. At this time, the effect of application method or uniformity of structural capacity is unknown, but is the focus of a related ongoing research study at the University of Florida. Long term temperature, relative humidity, weather and energy data were monitored through various instruments installed in the house. The recorded data was analyzed to under stand the thermal performance of the vented roof system and

PAGE 106

106 also predict the moisture performance (drying potential) of roof systems by using WUFI Pro 4.2 in numerical simulations Conclu ding Remarks Temperature monitoring showed that ap plication of a l evel II Protection of ccSPF reduced the mean peak attic temperatures by 1 3 to 20 0 F. The research was not able to determine the cause of the 4.7 0 F difference in two attic temperature s before the installation of ccSPF except that the sensor showing the lower peak temperature was located close to the roof ridge, and the roof region directly above it is shaded by live oak trees. However, after the installation of ccSPF, the difference in attic temperatures reduced to ab out 1.3 0 F. As expected in a 37 year old house, air leaks were found in several places in the building envelope, which if sealed, could improve the energy efficiency of the home. The benefit of this research is that it will provide a baseline energy usage d ata o f the test house which can be compared in real time with the energy benefits of different retrofit strategies. The thermal inspection results showed that the house is under slight positive pressure relative to the attic, which could be caused by leaky returns in the attic or garage. The energy savings due to the installation of ccSPF alone could not be accurately quantified as th e older HVAC system had to be replaced in August 2010 due to technical malfunctioning. A total energy savings of about 22% wa s recorded during August December 2010 over the electricity consumption in 2009 during the same period. S hort term energy monitoring for a week before and after the foam installation showed up to 27% decrease in HVAC energy consumption. However, the total energy savings in heating and cooling the test house due to foam installation could not be exactly determined Three different approaches to normaliz e t he energy consumption of HVAC were developed to determine the changes in cooling energy usage pattern in the house due to foam installation alone. The results of energy normalization showed about 27% 49% reduction in

PAGE 107

107 cooling energy consumption of the 4 ton HVAC unit after the installation of ccSPF. The average daily cooling energy consumption for the 3.5 ton unit is less than the 4 ton unit after ccSPF installation in the test house. However, the higher values of the peak and time integrated temperature difference normalized energy consumpt ion for the 3.5 ton HVAC could not be explained. At this point, it cannot be determined which is the best method to normalize the cooling energy demand but all the three methods certainly show similar trends for energy consumption of the HVAC units. The re sults from the analysis of numerical model of the experimental roof in WUFI Pro 4.2 showed that the surface temperatures of roof components can be predict ed with fair accuracy. The predicted surface temperatures of plywood were up to 10 15 0 F higher than th e experimentally measured temperatures while those predicted from ccSPF were up to 10 0 F lower than the experimental values. The prediction of higher temperatures for plywood is most likely due to the effect of solar radiation on temperature calculations. T he roof is shaded by live oak trees whereas the solar radiation measurements were taken in an open area with no obstruction to sun light. Twelve different analytical roof models with different roof sheathing, insulation, and ventilation were compared for the long term moisture performance in case of an incidental leakage caused by a high velocity wind event. The foam retrofitted vented and unvented roofs were found to be prone to moisture accumulation in sheathing for long durations of times potentially le ading to mold growth on the wood. Non retrofitted and a 1 in. ccSPF retrofitted u nvented roof s models with OSB sheathing showed lower drying times than their vented counterparts in the event of an incidental leakage For non retrofitted, plywood roofs, the drying time is similar but with 1 in. of ccSPF, vented roofs had about 8 months lower drying time. It

PAGE 108

108 was also observed that the peak water content in all the simulated unvented roof systems was lower than the vented roofs. Limitations of this Study S ome limitations of the current study are discussed below. Long term monitoring of HVAC energy use before the installation of ccSPF could not be done in this study. Consequently, the long term comparison of HVAC energy consumption was not done. The existing 4 t on HVAC unit was replaced by 3.5 t on during mid summer due to technical malfunctioning As such, the total energy savings could not be separately determined due to HVAC change and ccSPF installation. The surface temperature monitoring devices (EK H4 se nsors) did not log the temperature and relative humidity continuously on the laptop installed at the test house resulting in gaps in data. The results of the numerical study are validated from the short term data experimentally collected surface temperature data. The material properties of roof components in the test house we re not experimentally measured. Approximate properties were assigned to the components from the materi al database of WUFI. The external weather data was recorded approximately 50 feet away from the roof section modeled in this research. As such, the actual solar radiation s on the inclined surface of roof section modeled were not measured. One dimensional roof model in WUFI does not account for surface temperature change due to attic ventilation As a result of this, both vented and unvented attic models resulted in similar predicted surface temperature of the components. Recommendations for Future Work The researchers caution strongly that experimental long term performance testing is still needed to verify that ccSPF retrofit of wood roofs do not lead to unforeseen premature deterioration due to water intrusion. As such, regular maintenance and inspections of the roofing system, flashing is recommended to increase likelihood that any leak location is identified and promptly repaired From our observations of the foam installation, careful training of technicians is required for better installation of ccSPF in the attics At best, residential attics are typically hot, uncomfortable places but during the installation the technicians must manipulate long hoses

PAGE 109

109 and direct a chemical spray that is at or near 130 140 0 F Quality control devices would be helpful to check the dimensions of the fillet and to confirm the minimum thicknesses of the foam layer should be provided. A similar study to determine the energy performance of an unvented attic retrofitted with a l evel II ccSPF protection is recommended to directly compare with the energy performance of a vented attic. The energ y performance of houses with a l evel III ccSPF protection in similar climatic conditions should also be experimenta lly determined to compare with l evel II ccSPF installation. The one dimensional numerical simulations in WUFI Pro 4.2 suggest that it has great potential to predict the long term hygrothermal performance of roof systems. Two dimensional numerical modeling studies should also be undertaken to incorporate the effect the roof framing members in the models. The numerical models for predicting the long term performance of building components should first validated by comparing with experimentally measur ed data at the point of interest to increase the accuracy of models Experimentally measured values vary with location and when used in numerical models, these variations may affect the results of the simulation.

PAGE 110

110 APPENDIX A BASELINE SURVEY REPO RT OF THE TEST HOUSE BY THERMO SCAN INSPECTIONS PRIOR TO CCSPF INSTA LLATION

PAGE 111

111

PAGE 112

112

PAGE 113

113

PAGE 114

114

PAGE 115

115

PAGE 116

116

PAGE 117

117 APPENDIX B BASELINE SURVEY REPO RT OF THE TEST HOUSE BY THERMO SCAN INSPECTIONS POST CCSPF INSTALLAT ION

PAGE 118

118

PAGE 119

119

PAGE 120

120

PAGE 121

121

PAGE 122

122 Position of infra red camera for taking the thermal images

PAGE 123

123 APPENDIX C WEATHER STATION DATA RECORDED AT THE TEST HOUSE Ambient temperature time history

PAGE 124

124 Ambient relative humidity

PAGE 125

125 Daily total rainfall

PAGE 126

126 Daily total solar radiation

PAGE 127

127 APPENDIX D MONTHLY ELECTRICITY AND NATURAL GAS CONS UMPTION DATA OBTAINE D FROM GAINEVILLE REGI ONAL UTILITIES (GRU) FOR 2009 AND 2010 Electric ity (kWh) Gas (therms) Month 2009 2010 2009 2010 Jan 530 577 62 162 Feb 550 426 136 101 Mar 480 468 73 138 Apr 420 452 27 49 May 640 541 20 17 Jun 862 1020 17 15 Jul 1133 1560 15 18 Aug 1184 1170 16 15 Sep 986 805 15 6 Oct 848 440 16 7 Nov 552 300 26 20 Dec 340 350 54 73 Total 8525 8109 477 621

PAGE 128

128 LIST OF REFERENCES ASHRAE. (2002). "RP 1018 A Thermal and Moisture Transport Database for Common Building and Insulating Materials." Beal, D., and Chandra, S. (1995). "The Measured Summer Performance of Tile Roof Systems and Attic Ventilation Strategies in Hot, Humid Climates." Thermal Perf ormance of Exterior Envelopes of Buildings VI, Clearwater Beach, FL. Cash, C. G. (2003). Roofing Failures Spon Press Inc., New York. FEMA. (2005). "Summary Report on Building Performance: 2004 Hurricane Season, FEMA 490." Federal Emergency Management Agency, Washington D.C. FEMA. (2006). "FEMA 548: Summary Report on Building Performance: Hurricane Katrina 2005." FEMA. (2010). "Wind Retrofit Guide for Residential Buildings, FEMA P 804." Federal Emergency Management Agency, 124. Finch, G., Straube, J., a nd Richmond, M. "Field Performance of Spray Polyurethane Foam: The Role of Vapour Diffusion Control." 11th Canadian Conference on Building Science and Technology Banff, Alberta. Graettinger, A. J., Van De Lindt, J. W., Gupta, R., Pryor, S. E., Skaggs, T. D., and Fridley, K. J. (2006). "Overview of wind damage to wood frame structure caused by Hurricane Katrina." ASCE Structures Congress 2006, American Society of Civil Engineers, Reston, VA, 20191 4400,United States, 57. Gurley, K., Davis, J. R. H., Ferre ra, S. P., Burton, J., Masters, F., Reinhold, T., and Abdullah, M. (2006). "Post 2004 Hurricane field survey An evaluation of the relative performance of the standard building code and the Florida building code." ASCE Structures Congress 2006, Reston, VA 20191 4400, United States. Hendron, R., Nagy, S. F., Anderson, R., Reeves, P., and Hancock, E. (2004). "Thermal Performance of Unvented Attics in Hot Dry Climates: Results from Building America." Transactions of the ASME 126, 6. IBHS. (2011). "Strengthen ing an Existing Roof Options Other than Re roofing." Jump, D. A., Walker, I. S., and Modera, M. P. "Field Measurements of Efficiency and Duct Retrofit Effectiveness in Residential Forced Air Distribution Systems." ACEEE Summer Study Asilomar, CA. Karagioz is, A. N. "Advanced Hygrothermal Models and Design Models." The Canadian Conference on Building Simulation Ottawa, Canada, 547 554.

PAGE 129

129 Kuenzel, H. M., Karagiozis, A. N., and Holm, A. H. (2001). "A Hygrothermal Design Tool for Architects and Engineers (WUFI O RNL/IBP)." Moisture Analysis and Condensation Control in Building Envelopes, American Society for Testing and Materials, West Conshohocken, PA. Lstiburek, J. W. (2006). "Understanding Attic Ventilation." McCarty, C., and Smith, S. K. (2005). "Florida's 200 4 Hurricane Season: Local Effects." Florida Focus 1(3), 13. Nelson, P., and der Ananian, J. (2009). "Compact Asphalt Shingle Roof Systems: Should They be Vented?" Journal of ASTM International 6(4), 18. Parker, D. (2005). "Literature Review of the Impact and Need for Attic Ventilation in Florida Homes." Parker, D., and Sherwin, J. (1998). "Comparative Summer Attic Thermal Performance of Six Roof Constructions." The 1998 ASHRAE Annual Meeting, Toronto, Canada. Parker, D., Sonne, J., and Sherwin, J. (2002). "Comparative Evaluation of the Impact of Roofing Systems on Residential Cooling Energy Demand in Florida." Residential Buildings: Technologies, Design, Performance Analysis, and Building Industry Trends, 1.229 1.234. Parker, D., Sonne, J., and Sherwin, J. (2005). "Flexible Roofing Facility: 2004 Summer Test Results." Florida Solar Energy Center/University of Central Florida, Cocoa, FL. Porter, W. A. (2003). "Moisture Transport Across Shingle Roof Attic Spaces in a Hot Humid Climate," University of Florida, Gainesville. Prevatt, D. O. (2007a). "Improving the Wind Uplift Capacity of Wood Roof Panels Retrofitted with ccSPF Adhesives." University of Florida, Gainesville, FL. Prevatt, D. O. (2007b). "Wind Uplift Behavior of Wood Roof Sheathing Panels Retrofitted with Spray applied Polyurethane Foam." University of Florida, Gainesville, FL. Rose, W. "Measured values of temperature and sheathing moisture content in residential attic assemblies." Thermal performance of the exterior envelopes of buildings V Atlanta, 379 390. Rudd, A., and Lstiburek, J. (1996). "Measurement of Attic Temperatures and Cooling Energy Use in Vented and Sealed Attics in Las Vegas, Nevada." Rudd, A. F., and Lstiburek, J. W. (1998). "Vented and Sealed Attics in Hot Climates." ASHRAE Transact ions: Symposia 1199 1210. Sedlbauer, K. (2002). "Prediction of Mould Growth by Hygrothermal Calculation." Journal of Thermal Envelope and Building Science 25(4), 6.

PAGE 130

130 Straube, J., and Finch, G. (2009). "Ventilated Wall Claddings: Review, Field Performance, and Hygrothermal Modeling." Straube, J., Smegal, J., and Smith, J. (2010). "Moisture Safe Unvented Wood Roof Systems." Building Science Corporation. TenWolde, A., and Rose, W. B. (1999). "Issues Related to Venting of Attics and Cathedral Ceilings." ASHRAE U.S. Census. (2003). "U.S. Census Bureau, Census of population and housing (2000). US Dept. of Commerce, Economics and Statistics Administration." Washington, D.C. U.S. Census. (2007). "2007 Population Estimates." U.S. Dept. of Commerce, Economics and St atistics Administration, Washington, D.C. U.S. Energy Information Administration. (2010). "Annual Energy Review 2009." Washington D.C.

PAGE 131

131 BIOGRAPHICAL SKETCH Sushmit Shreyans was born and brought up in Gaya, India, a place sanctified by both the Hindus and the Buddhists. He graduated with a Bachelor of Engineering (Hons.) degree in Civil Engineering from Birla Institute of Technology and Science (BITS), Pilani, India in June 2 009. He joined the University of Florida, Gainesville in August 2009 to pursue a Master of Science conduct research on the energy performance of residential struct ures in Florida. Sushmit received his Master of Science degree in August 201 1 and hopes to practice as a structural engineer He enjoys sketching, painting, photography, reading books, playing cricket racquetball and pool and hanging out with friends in h is leisure