1 MOISTURE TRANSPORT IN VENTILATED ATTIC SPACES FOR HOT HUMID CLIMATES By BARRETT L. MOONEY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERISTY OF FLORIDA 2012
2 2012 Barrett L. Mooney
3 To my loving wife
4 ACKNOWLEDGMENTS The author would like to thank Dr. Wendell Porter for his guidance and support which allowed this project to be completed. Additionally the author would like to thank Dr. Ray Bucklin, who served as a source of inspiration a nd encouragement to complete this research. The encouragement, advice and direction of the members of the super visory committee were also greatly appreciated. This work would not have been possible without the data collected by Dr. Wendell eam during 2005, 2006 and 2007. The author would also like to thank t he faculty, staff and student body of the Agricultural and Biological Engineering Department. Alexandra deserves special thanks for her emotional an d technical support throughout this research effort.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 6 LIST OF FIGURES ................................ ................................ ................................ .......... 7 LIST OF ABBREVIATION S ................................ ................................ ............................. 9 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ..... 12 2 LITERATURE R EVIEW ................................ ................................ ........................... 17 3 METHODS AND EQUIPMENT ................................ ................................ ................ 25 Data Collection ................................ ................................ ................................ ....... 25 Model Development ................................ ................................ ................................ 27 4 RESULTS AND ANALYSIS OF DATA ................................ ................................ ..... 33 Internal Microclimate ................................ ................................ .............................. 33 Seasonal Humidity Switch ................................ ................................ ...................... 37 Duct Model ................................ ................................ ................................ ............. 39 Discussion ................................ ................................ ................................ .............. 40 5 CONCLUSION ................................ ................................ ................................ ......... 70 APPENDIX A MODEL SOURCE CODE ................................ ................................ ........................ 73 B HEATIN G AND AIR CONDITIONI NG PARAMETERS ................................ ............ 79 LIST OF REFERENCES ................................ ................................ ............................... 81 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 84
6 LIST OF TABLES Table page 4 1 Compariso n of raw data between 2005 and 2006 ................................ ............. 67
7 LIST OF FIGURES Figure page 3 1 Plan view of test facility ................................ ................................ ...................... 31 3 2 Illustration of sensor positions in test facility attic space ................................ ..... 32 4 1 Time spent above 150F ( 65.6 o C ) for upper shingles ................................ ......... 47 4 2 Time sp ent above 100 o F (37.8C) for attic space ................................ ............... 48 4 3 Monthly averages of relative humidity at various points ................................ ..... 49 4 4 Time averaged daily relative humi dity conditions for the average day in January ................................ ................................ ................................ ............... 50 4 5 Time averaged daily relative humidity conditions for the average day in July .... 51 4 6 Average temperature difference of air travelling from soffit to ridge ................... 52 4 7 Monthly average temperature comparison for attic materials ............................. 53 4 8 Temperature comparison of attic elements to environmental conditions ............ 54 4 9 January moisture flux for the attic environment ................................ .................. 55 4 10 April moisture flux for the attic environment ................................ ........................ 56 4 11 July moisture flux for the attic environment ................................ ........................ 57 4 12 October moisture flux for the attic enviro nment ................................ .................. 58 4 13 Seasonal values for humidity ratio given for representative days of the year. .... 59 4 14 Seasonal values for humidity ratio comparing the two year data set. ................. 60 4 15 Time series variation in attic air changes per hour. ................................ ............ 61 4 16 Fluctuations in relative humidity as measured at various attic locations. ............ 62 4 17 Temperature comparison of various thicknesses with duct absorptivity of 0.1 in July. ................................ ................................ ................................ .................... 63 4 18 Temperature comparison of various thicknesses with duct absorptivity of 0.86 in July. ................................ ................................ ................................ .................... 64 4 19 Temperature comparison of various thicknesses with duct absorptivity of 0.86 In October. ................................ ................................ ................................ .............. 65
8 4 20 Temperature comparison of various thicknesses with duct absorptivity of 0.1 in October. ................................ ................................ ................................ .............. 66 4 21 Empirical ventilation effectiveness profile with respect to wind speed. ............... 68 4 22 Volumetric flow rate in the attic with respect to time of day averaged over years. ................................ ................................ ................................ ........................... 69
9 LIST OF ABBREVIATIONS ASHRAE American Society of Heating Refrigeration and Air Conditioning Engineers AtticSIM Attic Simulation Software by Oak Ridge National Laboratory CarrierHAP Carrier Hourly Analysis Program HVAC Heating Ventilation and Air Conditioning UF University of Florida US United States
10 Abstract of Dissertation Presented to t he Graduate School of the University o f Florida i n Partial Fulfillment o f t he Requirements f or the Degree o f Doctor o f Philosophy MOISTURE TRANSPORT IN VENTILATED ATTIC SPACES FOR HOT HUMID CLIMATES By Barrett L. Mooney December 2012 Chair: Wendell A. Porter Major: Agricultural and Biological Engineering Ventilated spaces in the built environment create unique and benef icial microclimates. While the current trends in building physics suggest sealing attics and crawlspaces, comprehensive research still supports the benefits of the ventilated microclimate. Data collected at the University of Florida Energy Park show the at tic environment of asphalt shingled roofs to be typically hotter than the outdoor conditions, but when properly ventilated sustains a much lower relative humidity. The hot, humid regions of the United States can utilize this internally convective air mass to provide stable moisture levels within attic spaces. Positioning the buildings primary boundary at the ceiling deck allows for utilization of a buffer climate to minimize moisture trapping in is investigation concludes that the conditions in a ventilated attic are stable through seasonal changes and promote cost effective, energy efficient climate control of unconditioned spaces in hot, humid regions. Hot and humid climates present a challenge to limit moisture entrainment and condensation within the building envelope. Ventilated attics perform the function of purging intruding moisture. Ventilated attics change with the seasons to make the best
11 use of ceiling insulation an d maintain building health. S offit and ridge vent configurations suppress excessive ventilation during mornings hours limiting the dense, moist air coming in contact with a cool duct surface. Using computer simulation and a mathematical model the data collected at the University of Florida Energy Park was used to investigate the interaction of mechanical ventilation ductwork placed in residential attics. Output data from this model show moisture can condense at the duct surface under certain c onditions Proper installation reduces the duct leakage and mitigate s the effects of condensation. Th us the data indicate crimped or damaged ductwork can sufficiently cool the surface to lead to condensation.
12 CHAPTER 1 INTRODUCTION The microclimatic conditions surrounding a building have a direct impact on the energy consumption necessary to provide indoor comfort. The hot, humid climate of Florida provides unique challenges for energy efficient building design. Silberstein and Hens show that insulation of the building envelope is greatly hindered by the intrusion and stagnation of absorbed moisture. With correctly ventilated spaces and properly positioned insulation, excessive moisture should not accumulate [1 ] Their study investigates the benefits of removing moisture continuously through ventilation. The ventilated attic structure serves the purpose of creating a climatic barrier enhancing the performance of the insulation. When a ventilated attic does not cool significantly over night in colder regions, a high rate of moisture removal has been reported  Their results indicated a reduction of soffit ice dams and increased insulation efficiency of the ceiling space. Although the previous research does not discuss all aspe cts of attic microclimates, their conclusions warrant further investigation of the effective ness of ventilated attics in warmer regions w ith increased nighttime temperatures. Alternatively, moisture can be transported by pressure differences moving vapor from a conditioned space to an unconditioned space which can lead to potential moisture pro blems such as mold and mildew growth. Lstiburek, et. al. describes the necessity to properly seal building cavities between conditioned and unconditioned spaces to prevent condensation of water vapor that is communicated between the se adjacent spaces  This research illustrates that ventilation can create pre ssure differences leading to moisture transport between spaces. The previous research did not investigate a particular attic for a full twelve month period. Thus, d uring certain
13 months pressure differences across adjacent spaces may fluct uate yielding a beneficial self regulating system of moisture movement. Research presented indicates v entilation of an attic space in hot, humid regions serves to purge hot, moist a ir providing an air mass with a lower humidity ratio to sit directly above the indoor conditioned space. This is a beneficial cycle that directly controls unwanted condensation and inhibits the growth of mold and mildew within attic insulation and on roof truss members. Therefore, the ventilated attic should not represent the effective building envelope, but instead be defined as a primary boundary to control the impact of weather on the interior of a building and the mechanical support syste ms. The condi tioned space below the attic, a separated and insulated space functions as the true building envelope where all other systems are operating to support the environmental conditions within that space. Humidity, temperature, and ventilation rates are importa nt factors in the heat and moisture transport of both conditioned and u nconditioned spaces that affect indoor comfort, energy efficiency, and air quality. A yearlong investigation exploring each factor of the climate through all seasons and taking into ac count varying outdoor conditions is necessary to quantify the benefits of a ventilated attic microclimate requires However, previous research has failed to present data over a yearlong period in a hot, humid region to evaluate the benefit s of passive vent ilation and its e ffects on moisture transport. For this study, yearlong measurements were taken during 2005 and 2006 at the University of Florida Energy Park The data presented in this paper provide a complete illustration of the climate developing within the unconditioned attic spaces of residential and light commercial construction. As weather patterns shift in different regions the attic
14 conditions are altered accordingly through temperature swings and wind speed. The attic conditions also vary based on roofing type and vent location. Building codes vary by state, but it is generally accepted by the building science community that residential construction should promote a minimum attic ventilation and position opposing vents at the soffits and ridge in standard residential construction. Different climatic regions across the United States (US) require different building strategies. A detailed overview of seasonal and temporal interactions between the building envelope and the climate is required to fully understand the system  Mathematical models have been developed by many researchers to predict attic ventilation rates and humidity levels required to preven t condensation on roofing surfaces. These models have identified key environmental characteristics effecting temperature and moisture movement in the building envelope. Assessing these complex elements requires detailed datasets to analyze and validate sim ulation results in the changing seasons. In humid regions attic ventilation is an area of strong focus. Various p hysical an d modeling experiments have investigat ed the benefits of attic ventilation including mitigating high temperatures, improving insulation performance, and energy efficient building envelopes. Moisture removal is not adequately addressed in many modeling experiments Thermal attic models require an analysis of heat transfer mechani sms and moisture balances in roofing structures. Stratified air transient models such as the one illustrated by Parker et. al. (1991) and Burch et. al. (1979) predict changes in attic temperature and explore the impact on condensation and air conditioning energy
15 consumption and performance. While these models have achieved success in prediction of appropriate ventilation rates and temperature conditions, validation datasets are not sufficiently large enough to validate diurnal moisture fluctuations and expl ore seasonal differences in attic environments. The data presented were collected using a variety of sensors over a two year period, and the data describe changes in attic humidity as a function of external and internal measured elements. The data clearly illustrate predictive seasonal shifts. Data collected and analyzed as part of this doctoral research show humidity fluctuations are regular oscillations and promote a healthy building en velope. Attic ventilation is also a seasonally beneficial element of the building envelope Evaluation of the building envelope and understanding seasonal microclimates is essential to achieving energy efficiency, and this data will aid in evaluating the i mpact of mechanical ductwork impact on condensation. Investigating the microclimate of attics in the southeast also requires analysis of heating ventilation and air conditioning (HVAC) duct work. Common construction practice places the ductwork and air han dler in unconditioned attic spaces. The test attics built at the University of Florida (UF) Energy Park did not contain ductwork, so analysis of the data required the use of a computer based simulation to examine the effects of conditioned ductwork. The digital simulation focuses on common installation procedures that could lead to condensation. Problems include crimping, compressing, and bending of flexible ducts that can create interstitial points of reduced insulati on. The model developed for this research adds to the thermal attic models published in the
16 literature and leverages the data collected from the UF Energy Park between 2005 and 2006.
17 CHAPTER 2 LITERATURE REVIEW Mechanical ve ntilation and refrigerated air condit ioning, in the hot and humid s outheastern US produce challenges to the building envelope Conditioned air must remain thermally disconnected from the humid environment to prevent condensation and moisture problems such as potentially hazardous mold and mildew growth. Additionally the structural integrity of a building can be challenged under circumstances of significant saturation of building materials Research efforts in the last thirty years have sought to investigate t he corresponding phenomena. Building codes used in the southern US come primarily from previous building experience in colder latitudes. These codes are often adapted to particular needs of the south such as high humidity. Ventilated attics as a construc tion technique originated in northern cli mates, primarily as a method for reducing ice dam formation at the eves. Samuelson, a Norwegian researcher concluded that ventilation serves little purpose during the winter when there is not a significant tempera ture gradient to allow the moving air to accumulate moisture. Alter natively ventilation in the summer environment is beneficial because it transports moisture In the Norwegian climate night time ventilat ion allows for condensation, and it is only during the summer months when there is adequate heat generation in the air space to purge the attic of stored moisture  research provides a useful understanding of ventilated attics applicable to the hot, humid regions similar to Florida. Florida has very few cold days and generally dry winter conditions can maintain a high rate of moisture removal. The air flow of a hot humid region can transport more moisture thereby increasing the
18 effectiveness of adequate ventilation. This research describes the development of a microclimate model useful in predicting and interpreting changes to ventilated attic structures. model investigate s the variability in vented spaces to provide an optimal minim um cavity size and to explore the use of increased insulation along the ceiling deck. These conclusions, while expected to return significantly different values, are exploring the same areas of building science that influence residential energy efficient c onstruction in the southeastern U S Blom draws similar concl usions in that moisture movement requires ventilation. The author addresses the ventilated attic climate in cold and moderate regions. It was concluded that ventilation in attic spaces should not be overlooked for modern roof construction. The ventilation gaps store enough heat to prevent significant ice formation, and remove sufficient moisture to protect insu lation and building materials  There is applicability in the conclusions of for hot, humid climates. T he ridge vent s produce little benefit given that ice formation is only a concern at the eves however their use is desirable for buildings in Florida where high heat moves more moisture through a thermal gradient This venting creates air flow and moisture removal which serv es as a barrier to the elements Therefore the ther mal barrier should be placed where the conditioned spaces meet the unconditioned spaces While increased temperatures which create snow melt in the north, are likely in vented spaces it is this heat gain coupled with moving air that reduces the humidity l evels of the attic spaces. The focus should be placed on insulati n g between unconditioned spaces and the interior Research from the northern climates and the problems in the hot, humid climates illustrate a need for reducing air intrusion. Lstiburek,
19 et al. found that building pressures created by mechanical ve ntilation retards air intrusion and can influence air flow which as a result carr ies moisture Case studies present evide nce that pressurization within building cavities can retard environmental intrusion and improve energy efficiency of mechanical equipment. Assessing the building envelope as a pressure boundary provides a unique view of environmental interactions influencing the microclimate created wit hin conditioned and unconditioned spaces. Lstiburek, et. al. utilize this difference to support claims of fo rced ventilation for crawlspaces and sealed attics. The cases presented give insight into the remediation of problem buildin gs, but fail to illustr ate other, simpler remediation techniques Additionally the authors economic feasibility o r cost of remediation measures  Lstiburek, et. al. presents a clear picture of a microclimate formation w ithin a building envelope between conditioned and unconditioned spaces. Pressure differences cannot be ignored and necessitate preventative measures to ensure energy efficiency, structural stability, and environmental quality. A pressure boundary exists fo r the conditioned space and improves the indoor air quality for the occupied space. Following this method for the unconditioned spaces will waste energy to improve efficiency for the mechanical systems. The semi conditioned space created from pressurizatio n will move moisture, but will also cost the building operator money to condition a space that will not be used by occupants. Previous research concludes that ventilation and pressurization limits air infiltration and red uces moisture in attics and crawlspaces However, the conclusion also works for other building envelope co nstruction techniques including above
20 sheathing v entilation Miller et. al. simulat ed and test ed comparisons between roofing materials and the benefits of the ve ntilation with respect to solar reflective materials. The simulation included a Computational Fluid Dynamics scheme, which laid the grou nd work for the development of the code package AtticSim. The study verified results on measured data taken from test at tics constructed to minimize extraneous parameters such as leakage of conditioned air. The previous study illustrates the benefits of ventilat ion to reduce heat effects and transport infiltrated moisture. Under normal conditions a roof with a tile or sha ke layer retains a thermal mass trapping heat that would enter the attic space with other roofing materials. In combination with ventilation beneath the roofing element a thermal ly buoyant, mixed convective region is created. Coupling these two elements reduces the heat flux through the roof decking. Additionally, when shake roofing of different reflection values w as u sed a reduction in heat flux was observed. It was concluded that darker roofing material produces more thermal buoyancy in the ventilated cavity offsetting the effects of the increased heat gain. The refore, the benefits of ventilation produce a more significant reduction in heat flux  The study by Miller, et al. illustrat ed the benefits of ventilation below sheathing in hot, humid regions Heating of a surface above a ventilated region creates a naturally regulated air flow Based on the ir conclusion air movement reduces heat flux, but also removes excess moisture Walker and Forest conducted research to investigate the changes in the attic environment with changes in ventilation rates. While the research focuse s on ventilation as a heat mitigation technique, data are needed to understand ventilation rates to
21 optim iz e moisture removal. Also, c old night s and wet winters of Alberta, Canada present a challenge for ventilation to remove moisture. T his experiment investigated ventilation rates measured through a tracer gas applied to the building envelope and attic space intrusion from both outdoor and interior areas. Additional moisture content on the roof sheathing was examined to determine a correlation between moisture and ventilation. The primary conclusion presented was a strong linear relationship between wind speed and attic ventilation. A ir changes in the attic increase with increasing wind speed allowing for more effective ventilation. Air changes increased from 2 per hour to 10 per hour when the attic space changed from sealed to vented. No conclusions were drawn linking moisture content of wood members to ventilation rates A ll ventilated spaces reported no significant moisture content  Research by Walker and Forest illustrates a need for linking the moisture removal in the attic with appropriate building code requirements that benefit the whole building envelope. Creating a solar heated and ventilated air space above the conditioned region heats the air allowing it t o store more moisture desorbed fr om the building products. B uoya nt and mixed convection will subsequently remove the moisture with constant air changes. This phenomenon was not proven by Walker and Forest but provides a strong case for the research undertaken at the University of Florida. The air movement in an unconditioned space relies on weather conditions outside of the structure. It has been shown to rely heavily on solar radiation, wind speed, wind direction, and surrounding obstructions. These factors make the modeling of the attic m oisture temperature interaction highly non linear, and the transient problem is identified using a microclimate approach.
22 Medina and Young identified several factors complicating the use of insulation at the roof decking to limit heat gain to an attic spa ce. Their research uses computer simulations to test the benefits of a radiant barrier system for the h ottest three months of the year for all climatic regions of the continental United States. A net benefit in the reduction of peak hour heat flux exists i n all climate zones. This investigation did not see an advantage for the winter periods, so pr oduced no data for those months  The model was a heat and mass transfer model taking into account the effects of solar radiation re radiation of roof decking, convective heat transfer, and ventilated moisture intrusion. This model describes the microclimatic effects of buoyant and mixed convective flows and highlights no overall benefit due directly to air movement. Heat flux redu ction was the primary indicator of successful use for the radiant barrier. It was determined that ventilation was not a heat flux reduction component, and varying the amount of ventilation had no effects. The greatest heat flux savings and best performing simulation for a hot, humid climate occurred in a sealed attic space. The model as constructed provides for an analysis of soffit/ridge vent combination and produces a larger heat flux savings over soffit/soffit or soffit/gable combinations due to the air stratification. Medina and Young give little explanation for how the model handles moisture in wooden attic members and the effect o f heat flux on moisture retention. Driving the internal climate of the attic is the solar load in combinatio n with the moisture infiltration and removal. The research justifies the evaporation of moisture in the attic a cooling element particularly for the attic members themselves. The moisture removal was not considered as a necessary element to maintain air q uality of the space or structural
23 integrity. This model illustrates the need to investigate the moisture compo nent and solve for equilibrium between the benefits of moisture removal and heat flux reduction. Shashua Bar, et. al. present research in the Int ernational Journal of Climatology describing a similar microclimate focusing on building orientation, proximity to other buildings, and albedo of surfaces. The simulations illustrate deviations of the built environmental climate from meteorologically expec ted data by a maximum of 6 degrees Kelvin The conclusions state that a clustered building group with narrow spacing decreases the surrounding temperature due to significant solar shading. This data illustrates the changes in the climate created by built structures, and predicting these effects will allow community planners to better engineer the urban landscape for energy efficiency and thermal comfort  The application of this research extends to the elements of the building that are enclosed but unconditioned. These areas are most affected by changes to the im mediate external microclimate. A s tructure s proximity and orientation affect the thermal perfor mance on internal boundary layers and should be considered when investigating the microclimate of the built environment. The expected thermal effect of building positions will change the interaction of ventilated attics and moisture condensation conditions especially in hot, humid regions. As previously determined by Walker, et al., the element driving ventilation i n an attic is wind incident on a building structure  Thermal effects of building proximity may disrupt temperature gradients necessary to move air in channels of heavily populated areas. The net effect of building microclimates both internally and externally must be evaluated together to assess the building impac t on the climate and residents
24 Moisture intrusion in hot, humid climates creates a unique design challenge for building envelopes. The literature illust rates the temperature moisture interaction must be investigated as a transient system. The structural integrity of a building can be challenged under circumstances of significant saturation, so envelope design should focus on building health. Refrigerated air conditioning adds complexity to the systems, but should be designed to cooperate with the existing attic conditions. Research efforts as pa rt of this dissertation are meant to model and analyze the interaction between the attic microclimate and refrigerated air ductwork.
25 CHAPTER 3 METHODS AND EQUIPMENT Data Collection The research took place at the Building Products Test Facility on the Universit y of Florida camp us in Gainesville, Florida The building was completed in 1998 and was wired to provide data related to the internal microclimate of structures, both ventilated and sealed in hot humid regions. Data ha ve been collected at that facility since 2001. For this study data collected during 2005 and 2006 were analyzed. Figure 3 1 illustrates the floor plan; the building used had 12 bays Bay 9 was chosen for this analysis due to the construction of a ventilated crawlspace and attic using wood framing members typical of Florida residential construction. The bay is also centrally located shielding the instr ume nts from edge effects. The conditioned space has dimensions of 10.5 ft (3.2 m) wide and 20.0 ft ( 6.1 m ) long with ceiling heights of 7.9 ft ( 2.4 m ). Above the conditioned space is a standard attic with floor dimensions of 9.8 ft ( 3.0 m ) long and 10.5 ft ( 3.2 m ) wide with a roof slope of 5:12. The attic also had a ventilation area of 1.91 ft 2 ( 0.18 m 2 ) for every 200.2 ft 2 ( 18.6 m 2 ) of attic floor space, or a ventilation ratio of 1:105. The building is oriented with the long axis running East West The pitc hed roof deck surfaces face North and South and are constructed with asphalt shingle roofing. The soffits and ridge were vented with materials from AirVent, Inc. The soffit vent provided a ventilation area of 9 in 2 / ft ( 190.5 cm 2 /m ) and the ridge vent provided 18 in 2 /ft ( 381 cm 2 /m ) Ceiling deck insulation to R 30 was applied alternating fiberglass and cellulose insulation. The interior space was conditioned with a single Packaged
26 Through the wall Air Conditioner (PTAC), and the spa ce was kept at 76 F +/ 1 F ( 24.4 C +/ 0.5 C ) in the summer and 68F +/ 1 F ( 20 C +/ 0.5 C ) in the winter. In addition to the building, a weather monitoring station from Campbell Scientific, Inc. was used and positioned 24.9 ft (7.6 m) northwest of the buildin g at a height of 9.8 ft (3.0 m) This monitoring station measure d the external conditions of temperature, relative humidity, barometric pressure, wind speed, and wind direction. Weather conditions were compared to measurements taken inside the build ing, constructing a relationship between the external conditions and the resulting internal attic climate. Sensors were placed throughout the attic to measure movement of air, temperature, and moisture. Combined temperature and relative humidity sensors were located in the center of the attic, ridge vent, and soffit vents. Additionally, thermocouples were attached to various roof joists at the lower and middle positions on both roof deck surfaces, and betwe en the plywood sheathing and shingles. Air flow was measured by bidirectional anemometers which were positioned in both soffits, in multiple positions at the ridge vent, and three other locations along the roof decking of the attic. Figure 3 2 illustrates t he placement of various sensors within the test facility Data were recorded every 15 minutes for the two year period. During the collection period, the system consistently collected data for 34,89 1 data points per year out of a possible 35,040 points for an accurate description of 99.6% of the year. The data were recorded and organized according to a digital timestamp. Times were consistently Eastern Standard Time without including changes for Dayl ight Savings Time. A n accurate conclusion was drawn from the collected data of both the average and instantaneous microclimatic conditions of ventilated attics.
27 Model Development The attic microclimate data were incorporated into the model to predict refrigerated air duct surface temperature for ductwork placed in an attic in Florida. Focusing on duct surface temperature was important for this study to assess condensation events. No data was collected for an HVAC duct placed in attics, so the computer simulat es a cylindrical flex duct positi oned parallel to the ridge line on the ceiling deck and centered in the attic. The model inputs can be manipulated to analyze many configurations of ducts and installation qualities. Varia ble parameters include duct inner and outer diameters insulation level and leakage which allow the user to analyze the two years of 15 minute data sets collected for Gainesville, F lorida and examine leaking or poorly installe d ductwork. Analysis of the duc t work in attics presented in this study examines a standard flex duct at optimal construction as per Residential Construction Manual D  Simulation models were completed using an iterative energy balance method with MATLAB software. Appendix A contains the MATLAB script used for the analysis. The model solves the energy balance in three steps. Conduction heat transfer from the conditioned air through the insulation is solved using a finite difference method. Nodes a re placed on the inner and outer surface of the duct. The air temperature inside the duct was considered well mixed as a simplifying assumption. Thermal storage of the duct insulation was calculated using Equation 3 1 ( 3 1 ) Where = Material density
28 = Specific heat or Heat capacity of material = Material volume = Surface temperature Heat transfer on the exterior of the duct is dominated by convection and radiation. Energy balance for convection was calculated by the standard algorithm in Equation 3 2 The convection coefficient was calculated at each time step and was based on the att ic configuration, forced air flow, natural air flow, and temperature. Equation 3 3 was used to compute the convection coefficient. This method is outlined in ASHRAE Fundamentals  ( 3 2 ) ( 3 3 ) Where = Raleigh number = Convective heat transfer coefficient = Reynolds number = Prandtl number The radiative algorithm is calculated for each surface in the attic in 15 minute intervals. For each surface in the calculat ion the temperature difference was considered using a fourth degree polynomial. This radiative component was calculated for all three participating surfaces: north roof deck, south roof deck, ceiling deck, and duct surface. Equation 3 4 illustrates the alg orithm adapted from the text by Modest 
29 ( 3 4 ) Where = Fraction of radiation attenuated in moist air A 1 = Surface area = Surface emissivity = Surface temperature F nm = Surface to surface view factor Air between the roof decks and the duct work was considered a participating medium. Relative humidity of the attic was used to calculate air moisture interactive effects. This ca lculation was simplified using lookup tables  Conditions inside the ductwork came from parametric simula tion of the heating and cooling demand of a conditioned space below the attic. The demand parameters were taken from CarrierHAP software simulations for this segment of the building. Data tables are presented in Appendix B. Since the collected data set di d not record the on and off cycling of the package terminal air conditioning unit, the CarrierHAP simulation wa s used to correlate the duct air flow rate to outside air temperature. Comparison of the data and the y intercept of the regression indicate tha t at an outdoor air temperature of less than 62 F (16.67 C) the space requires some heating demand and if greater than 72 F (22.22 C) some cooling is required. The model uses the outdoor signature to initiate airflow through the duct and air temperature within the duct. Since the attic microclimate data is only calculated at a resolution of 15 minutes, the initial mixing of air within the duct and maximum length of cycling was neglected. Internal duct temperatures of 85 F (29.44 C) were used for
30 heating and 55 F (12.78 C) for cooling. These values reflect the simulation analysis from Carrier H ourly A nalysis P rogram (CarrierHAP ) This program is often used in the industry for energy and HVAC analysis on residential and commercial buildings. Duct surface temperature data with varying the thickness of duct insulation and absorption coefficient of the duct surface w ere calculated for each month. While every month wa s investigated, the seasonal changes and the summer months are most applicable to the investigation of condensation
31 Figure 3 1 Plan view of test facility
32 Figure 3 2 Illustration of sensor positions in test facility attic space
33 CHAPTER 4 RESULTS AND ANALYSIS OF DATA Internal Microclimate Appl ying this data uniformly requires analysis of average s for hours, days, and months. The data collected in this study were characterized for 2005 and 2006 to develop an accurate average for the simulation model. Monthly averages define an operating and consist ent climate for the building envelope within the attic. On the exterior the shingles heat up and store heat, which is conducted to the roof decking and reradiated to the attic space. During the month of August, the shingle surface temperature peaks at 177F ( 80.6C ) and 163F ( 72.8C ), for the so uthern and northern facing roofs, respectively. These extreme temperatures are the primary driving forces that provide energy to create the attic microclimate. While the extreme data points are important to cons ider, it is the duration of these extreme temperatures that is significant. During May through August the shingles spend over 250 hours a month above 100 F ( 37.8C ), and in July through August over 80 hours a month above 150F ( 65.6C ). Figure 4 1 displays the time spent above 150F (65.6C) for the shingles on both the northern and southern roof faces. The sustained high temperatures of asphalt shingle s contribute to a hot attic environment. Temperature differences between north and south facing roof decks create convective currents of air inside the attic. The internal temperature of the attic space can peak as high as 121F (49.4C) during the month o f July, and spend over 200 hours above 100F (37.8 C) for the month of August. Figure 4 2 illustrates the temporal range for extended periods of high temperatures.
34 Relative humidity measurements were taken throughout the attic and coupled with air flow measurements to calculate t he moisture transport through the attic vents. In addition wooden roof members were wired with thermocouples to assess the likelihood of condensation on attic members and absorption by the wood framing. Figure 4 3 displays the monthly relative humidity at inlets and outlets for air flow with respect to the attic interior and outdoor conditions. The data collected in this study offer a complete view of the average microclimatic conditions that can form within a ventilated attic under a variety of external conditions throughout all seasons. A ventilated attic possesses many advantages for a hot humid climate. While peak attic temperatures are high durin g summers, the relative humidity remains well below that of the external environment. Relative humidity is of specific interest in this study because the overall health of the attic depends on the temperature and moisture gradient around the wood framing members. Attic humidity ratios could be der ived from the data collected; however primar y focus was on with absorption and desorption of moisture. The exchange of moisture between the porous wood surfaces and the attic air is a function of rel ative humidity. As shown in Figure 4 3 the monthly average relative humidity for the attic is consistently lower than the exterior. This relationship is also true for the instantaneous data recorded at 15 minute intervals. Figure 4 4 and Figure 4 5 illustrate the temporally averaged daily conditions for January and July. The moist air entering the attic through the soffits gains heat and rises a vertical distance of 5.2 ft (1.6 m) before exiting the ridge vent. Figure 4 6 illustrates the temperature rise as air travels from soffit to ridge averaged over the course of each
35 month. As the temperature rises, the air can hold more moisture lowering the relative hum idity, decreasing the chance of condensation, and reabsorbing any moisture that precipitate s onto the insulation or attic members. This moist air is maintained at high temperatures well above the dew point which limits the instances where condensation can occur. Figure 4 7 displays the monthly average temperatures for the attic interior, roof joist surface and dew point calculated according to the ASH RAE method The roof joist member remains slightly higher in temperature than the attic interior air measurement due t o absorbed energy from the radiated heat of the roof deck. From this figure it can be concluded that condensation onto roofing members is rare and adequate ventilation does not cause accumulated moisture by inhibiting the radiative heat gain. Additionally, the radiation of heat from the roof decking is an important element in the heat balance of the attic space. Figure 4 8 compares the temperature of roofing materials and the attic temperatures with the outdoor conditions. Heat flux due to solar radiation drives the internal microclimate Both Figure 4 3 and Figure 4 8 illustrate th is hotter, drier microclimate with respect to outdoor conditions. A key element to this shift in microclimate is the effect on stored moisture. For each 15 minute increment the mass flow rate of moisture through the soffit and ridge was calculated using Eq uations 4 1 and 4 2 ( 4 1 ) ( 4 2 ) Where = M ass flow rate
36 = Air velocity A = Free area of ventilation = S volume of air = Moisture content in the air corrected for partial pressure T = Air temperature P = Atmospheric pressure = Gas constant for dry air based on the carbon 12 scale The moisture flow rate data were time averaged for each month creating an average day. Figure 4 9 through Figure 4 12 display 4 out of 12 months This illustrat es seasonal differences in attic moisture properties. These a verages r epresent inflow at the soffits and outflow at the ridge. taking in air and moisture at certain times and then releasing it later. The system cannot negate th e storage of moisture both within the attic space and the wooden attic members; ther e fore there is no observed equilibrium. A forward lagging of the system is evidence of the slight rise of moisture in the attic as it responds to incoming humid air. The overall trend in the graphs displays a consistent system, which is primar il y attributed to a ventilated attic. Looking closely at the data it is clear that the air flow consistently enters the attic through the soffits. The cross section al area of the soffit is smaller than the ridge vents. T hus, the ridge vent has a higher volumetric flow rate than the soffit vents and provides the only successful pathway to remove moisture. Therefore moisture flow is governed by the ridge vent.
37 Seasonal Humidity Switch The data were categorized and averaged to remove outliers by creating hourly averages from the collected 15 minute intervals The observation of seasonal trends was analyzed by using a monthly average day approach. Data collected at a given hour w ere then averaged daily during that month to construct a representative day. The representative day illustra tes typical attic conditions during the given month. Performing this data transformation allows for the interpretation of seasonal and monthl y data relations hips in the hot, humid southeast. The primary focus of this analysis will focus on data collected by sensors to measure mixed attic conditions as a result of seasonal climatic changes, including temperature, absolute humidity, and environmental characteri stics. Comparison of collected data to the ASHRAE design conditions illustrate s that the data ranges collected are normal for Gainesville, F lorida and the hot humid south. Additionally, the utilization of humidity ratios, air change rates, and average day analys es provide for a more detailed view of seasonal processes and microclimate interactions than limited data sets using single value average monthly conditions. Analys e s of collected data illustrate seasonal variety in the accumulated moisture of a ven tilated attic. The data confirm that ventilated attics retain more moisture during the summer months where the days are warmer and the outside humidity levels are high, but as the seasons shift the moisture level is naturally removed. Figure 4 13 and Figure 4 14 illustrate the seasonal changes in a box plot where the monthly avera ge days are combined into a single box and the monthly changes describe a consistent seasonal variation annually Measured data show that air chang es per hour fluctuate diurnally and seasonal fluctuations match
38 the changes in humidity ratio throughout the year. Figure 4 13 and Figure 4 14 illustrate changing attic humidity ratio over time. Daily fluctuations in the attic humidity ratio are a function of hot, humid air entering the attic through intended and untended penetrations in construction. The construction sealed the ceiling surface in conditioned space to eliminate movement of air through the ceiling plane. The airflow through soffit and ridge penetrations w as measured and converted to air changes per hour. Air changes were calculated using Equation 4 3 ( 4 3 ) Where = Air flow = Area = Length V = Volume of attic In Figure 4 15 a t ime series plot illustrates air change rates for the attic for each collection year. Figure 4 16 display s the stratification in relative humidity as measured at various attic locations. The figures show movement of moisture vertically with each air change. Data confirm the air stratification moves moisture from soffit to ridge. All data collected illustrate a useful picture of transient attic conditions through a variety of seasonal conditions.
39 Duct Model Microclimate characterization data were organized for input into the duct model. While the model calculates the energy balance for the attic, the duct surface temperature was the output analyzed for this study The surface temperature was analyzed for each month for given set of input conditions. T wo key variables were changed in the model runs to explore the controlling influence over duct condensation : the reflective or absorptive coefficient and the thickness of the duct insulation. The duct surface temperature changes in response to changes in the insulation thickness Figure 4 17 data come from input conditions with a reflective surface typical of most residential flexible ductwork. Surface absorbtivity coefficient of the duct modeled by Figure 4 17 was 0.1. To compare the results, a surface absorbtivity of 0.86 was compared in Figure 4 18 This value was chosen for a ductwork that might be painted or manufactured in a darker color to prevent condensation. For convergence of the radiation equation the duct surface must be of equal or lower absorbtivity coefficient with the roof deck. The two absorbtivity coefficient s chosen for this study 0.1 and 0.86, illustrate the system currently assembled in man y residential attics in Florida and a possible change to those systems to reduce building he alth issues respectively Figure 4 17 illustrates a much cooler duct surface than Figure 4 18 As duct surface temperature d ecreases, the likelihood of condensation increases. Figure 4 18 shows higher temperatures which increase the heat gained by conditioned air travelling to the living space. Either duct design can succeed during the summer months in hot humid climates, but both designs require proper installation. Crimping the duct or
40 compressing the installation by more than fifty percent significantly lowers the surface temperature allowing more heat transfer to conditioned air. Seasonal changes to temperature and humidi ty of the outside will affect the likelihood of condensation and any inconsistencies in installation will further exacerbate condensation problems. Figure 4 19 and Figure 4 20 illustrate the same duct surface resulting from average monthly conditions in October. The cooler attic in October shows that duct surface condensation is dependent on the temperature of the air inside the duct. Conditions were tested for several years, and Table 4 1 display s data for July 2005 and 2006. Discussion Data collected were used to analyze averages for months and seasons. The microclimate analysis points out problematic times and seasons where ductwork temperatures reach conditions of saturation on their surface as a result of moisture intrusion. Capillary movement of moisture through porous roofing materials and ceiling surfaces creates many potential intrusion points. Although this research discusses only shingled roofing, the most common roofing material in Florida, similar principles could be used to draw conclusions about various other roofing materials. Research conducted for this project investigated the mechanisms that mov e moisture once inside unconditioned spaces. Previously presented l iterature suggest ed sealing the space to prevent intrusion, but in hot, humid climatic zones sealing all intrusion points requires more thorough and precise methods. Research presented by Wendell Porter of the University of Florida showed how moisture travelled through a we ll shingled roof  Data presented as part of this
41 research show that transient attic conditions provide a natural convective cycle. Within the attic, diurnal cycles of heating cause thermal buoyancy to move air from the sof fits to the ridge. A ir change in the attic is naturally controlled by available solar radiation to heat the roof surface, the size and position of the attic penetrations. The air changes per hour of unconditioned space fluctuate in a consistent pattern that increases throughout the day. Minimum air exchange rates occur in the early morning during cooler months Air exchange rates decrease at the times when the humidity ratio of ambient air is highest, thus reducing the likelihood of moisture intrusion. T h e resulting microclimate maintains a resistance to saturation providing a solid barrier to the elements, and serves to remove moisture at a rate consistent with the prevention of mold and mildew growth. Under these circumstances the building envelope bou ndary of hot, humid regions should be considered conditioned space and be insulated accordingly. Overall this data supports increased insulation and materials with high thermal resistances to be positioned at the ceiling deck t hus reduc ing the heat flux in to conditioned space s S ealing the attic spaces entirely and insulating the structure would reduce the thermal heat transfer from an attic However, this would provide no outlet for moisture transported through porous roofing materials and the ceiling plane. When air enter s the attic, the model shows natural convective system mov ing moisture from soffit to ridge along the thermal gradient. In the event of moisture intrusion this movement alleviates damage to insulation and prevents seepage into the conditioned space.
42 Figure 4 16 shows the stratification in humidity through various points in the attic. Natural convection purg es moisture through fresh air input from soffit to remov e moisture through the ridge and similarly to a switch can turn on and off depending on the time of day or season W ithout proper position ing of vents the thermal stratification will not allow for the removal of moisture. Looking closel y at the data it is clear that air flow consistently enters attic s through the soffits. The ridge vent, with a higher volumetric flow rate than the soffit vents, provides the bulk air flow out of the attic creating the only successful pathway to remove moisture. Therefore moisture flow is governed by the ridge vent. Th e building position plays a significant role in the stratification because although heating of the roof surface drives the system, strong wind entering the soffits caus es mixing in the unconditioned space. Since weather fluctuations in the hot, humid Sout h do not create a consistent wind direction and most conventional construction has obstructions like trees and /or other buildings, thermal stratification is the only air flow mechanism capable of providing consistent air movement. Roofing surface temperatu re s will create different gradients based on solar angle position; however if incident solar radiation is available after a moisture entrainment event the internal convection will still stratify the air and re move moisture. Condensation issues in Florid a occur on HVAC supply ductwork and insulation when the surface temperature of a less insulated duct reaches the dew point. Problems persist in early morning hours when space cooling demand may cause short cycles of HVAC units and the dew point of moist mo rning air is still high. Fluctuation s in time series data presented in Figure 4 15 illustrate how early morning ventilation is reduced,
43 restricting mo ist outsid e air from entering the attic, but subsequently increase s during the hotter parts of the day thus purging excess moisture. Ventilation rates illustrate that fluctuations vary seasonally and diurnally while maintain ing active airflow. Under conditions described in this study the mean effectiveness of this ventilation methodology is 0.13, based on Equation 4 4  ( 4 4 ) Where Q = Air flow rate ( ) E = Effectiveness of ventilation A = Free area of ventilation (m 2 ) V = Wind Velocity ( ) Wind driven ventilation effectiveness is 0.24 for a soffit only configuration and 0.095 for a ridge only configuration All measur es of effectiveness were calculated using free outlet area. This variation on ventilation effectiveness is necessary to illustrate removal hot moist air. Walker describes typical ventilation effectiveness to range between 0.25 0.35 for diagonal winds an d 0.5 0.6 for perpendicular winds. These values are derived from empirical data collected by Barre and Sammett (1950) for farm structures. Values in this study indicate a lower average effectiveness resulting from variation in exterior pressure patte rns on the roof, and a lack of predominant wind direction at the data collection site. Although positioned in an open suburban area, buildings and trees around the site also affect the wind driven effectiveness factor. The relative effectiveness between the soffit and ridge agree with the distribu tion of pressure along the roof and indicate a lower pressure at the ridge than at the soffit.
44 This also illustrates the necessity fo r multiple roof penetrations at multiple l ocations. Without the variation in effec tiveness and pressure under varying wind conditions the moist air would not have a gradient to follow and the moisture movement would not follow the path of greatest effectiveness in a transient system Figure 4 21 illustrates the change of ventilation effectiveness with wind speed. Ventilation effectiveness decreases quickly then remains fairly consistent overall f or the majority of wind speeds. Under the proposed ventilation conditions, soffit and ridge openings provide a balanced air movement configuration maintain ing a steady flow of air through the attic. Understanding the effectiveness of the ventilation openings ( Figure 4 21 ), Figure 4 22 illustrates the need for providing both ridge and soffit ventilation. The total air flow attributed to wind and thermal forces are outlined in Equations 4 4 and 4 5 respectively  ( 4 5 ) Where A = Free area of ventilation (m 2 ) = 0.40 + 0.0025|Ti To| H = Height difference between inlet and outlet (m) Ti = Attic Air Temperature (C) To = Outside Air Te mperature (C) The volumetric flow rate of air leaving the attic space through the net free area of installed o penings is controlled by both thermal and wind forces. It is the interaction of these forces which minimizes ventilation during cool, moist morni ngs wh en condensation on ductwork is most likely to occur.
45 T he duct model analyzed the surface temperature of the duct. Convective heat transfer from the roof decking moves air through the attic, and radiative heat transfer drive s temperature changes of t he duct surface. Insulation is intended to limit the energy transferred through the duct to the conditioned air. Manufactures produce ducts with a reflective surface to distribute the incident radiation away from the surfaces allowing the duct to maintain a lower surface temperature. Most attics in the hot, humid South contain at least a few feet of flexible duct. Th e round duct is installed in a variety of ways often hanging from rafters, draped over trusses, or bent to the boot. These installation procedures can lead to crimped, bent, or compressed insulation. As the insulation is compressed more heat is gained by the conditioned air. Additionally, these locations are the most li kely for condensation. As cool conditioned air reaches a junction of compressed insulation the surface temperature drops. The model presented in through this research investigates this transient condition with an iterative solution methodology solving for convergence between the two boundary conditions at the surface of the duct. Decreasing the thickness of insulation decrease s the temperature of the surface. Using average monthly conditions constructed for characterizing the ventilated microclimate, the h ighest likelihood of condensation occur s duri ng the swing months of Model outputs are presented in Figure 4 19 and Figure 4 20 Months w here cycling of the HVAC system in the cool early morning pose the highest risk of condensation even if the HVAC system is cycling for a short period of time
46 Duct surface absorbtivity can increase the amount of heat gained from roof deck radiation and decrease the risk of condensation. The model analyzed flexible duct s under normal reflective conditions with a n absorbtive coefficient of 0.1 and under a painted condition with an absorbtive coefficient of 0.86. Data presented for an average day in July illustrate the rise in surface temperature. A duct design with reflective or absorbtive surfaces can provide con ditioned air during the summer months i n hot, humid climates, but both designs require proper installation. Installation procedures for residential attic duct work are described in Manual D of the residential code  Duct l ayout and handling are a central focus, and if done properly will mitigate the building health issues associated with moisture intrusion caused by condensation. Moisture in the attic only condenses before being ventilated if the duct surface is sufficientl y cool ed Improperly stretched or bent ductwork creates compression areas which also limits the amount of insulation contacting the duct surface. Crimping the duct or compressing the installation by more than fifty percent significantly lowers the surface temperature when the HVAC system is running. Another influential component is surface air mixing caused by duct leakage. Gaps in the flexible duct due to poor joins or with improper sealing of the air handler will decrease film temperature at the duct sur face in the vicinity of the leak. The model data represents a system with a five percent duct leakage typical of residential installations. Leakage fractions are included in the initial Carrier HAP inputs and the Matlab model. The surface mixing was con sidered uniform for simplicity of the transient solution. Under average conditions a 5% average leakage did no t significantly increase condensation rates in the digital simulation.
47 Figure 4 1 Time spent above 150 o F ( 65.6 o C ) for both north and south facing shingles 0 10 20 30 40 50 60 70 80 90 100 Time (hr) Months N Roof Sub Shingle Upper S Roof Sub Shingle Upper
48 Figure 4 2 Time spent above 100F ( 37.8C ) for attic spa ce 0 50 100 150 200 250 Time (hr) Months Attic Interior
49 Figure 4 3 Monthly averages of relative humidity at various points 0 10 20 30 40 50 60 70 80 90 100 Relative Humidity (%) Months Attic Interior Outside S Soffit Vent Attic Interior Ridge Vent
50 Figure 4 4 Time averaged daily relative humidity conditions for the average day in January 0 10 20 30 40 50 60 70 80 90 100 Relative Humidity (%) Time of Day (hr:min) Attic Outside
51 Figure 4 5 Time averaged daily relative humidity conditions for the average day in July 0 10 20 30 40 50 60 70 80 90 100 Relative Humidity (%) Time of Day (hr:min) Attic Outside
52 Figure 4 6 Average temperature difference of air travelling from soffit to ridge 0 1 2 3 4 5 6 7 8 9 10 F) Month
53 Figure 4 7 Monthly average temperature comparison for attic materials 0 10 20 30 40 50 60 70 80 90 100 Temperature ( F) Month N Attic Roof Joist Surface Attic Interior Dew Point ASHRAE method
54 Figure 4 8 Temperature comparison of attic elements to environmental conditions 0 20 40 60 80 100 120 Temperature ( F) Month Attic Interior Outside S Roof Sub Shingle Upper S Roof Sub Sheathing Middle
55 Figure 4 9 January moisture flux for the attic environment 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 Moisture Flow Rate ( lb/hr) Time of Day (hr:min) Soffit Ridge
56 Figure 4 10 April moisture flux for the attic environment 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 Moisture Flow Rate (lb/hr) Time of Day (hr:min) Soffit Ridge
57 Figure 4 11 July moisture flux for the attic environment 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 Moisture Flow Rate (lb/hr) Time of Day (hr:min) Soffit Ridge
58 Figure 4 12 October moisture flux for the attic environment 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045 Moisture Flow Rate (lb/hr) Time of Day (hr:min) Soffit Ridge
59 Figure 4 13 Monthly averaged s easonal values for the attic humidity ratio over 2005 and 2006 (1kg/kg = 1 lb/lb)
60 Figure 4 14 Monthly averaged s easonal values for the attic humidity ratio comparing the two year data set (1kg/kg = 1 lb/lb)
61 Figure 4 15 Time series variation in attic air changes per hour for 2005 and 2006
62 Figure 4 16 Fluctuations in relative humidity at various attic locations
63 Figure 4 17 Temperature comparison of various thicknesses with duct absorptivity of 0.1 in July 60 70 80 90 100 110 120 Temperature ( F) Time of Day Attic Temp Sheathing Temp 0.33 Duct (0.1) 0.25 Duct (0.1) 0.5 Duct (0.1) 1 Duct (0.1)
64 Figure 4 18 Temperature comparison of various thicknesses with duct absorptivity of 0.86 in July 60 70 80 90 100 110 120 Temperature ( F) Time of Day Attic Temp Sheathing Temp 0.33 Duct (0.86) 0.25 Duct (0.86) 0.5 Duct (0.86) 1 Duct (0.86)
65 Figure 4 19 Temperature comparison of various thicknesses with duct a bsorptivity of 0.86 I n October 60 70 80 90 100 110 120 Temperature ( F) Time of Day Attic Temp Sheathing Temp 0.25 Duct (0.1) 0.33 Duct(0.1) 0.5 Duct (0.1) 1 Duct (0.1)
66 Figure 4 20 Temperature comparison of various thicknesses with duct absorptivity of 0.1 in October 60 70 80 90 100 110 120 Temperature ( F) Time of Day Attic Temp Sheathing Temp 0.25 Duct (0.86) 0.33 Duct (0.86) 0.5 Duct (0.86) 1 Duct (0.86)
67 Table 4 1 Comparison of raw temperature data between 2005 and 2006 in F Time of Day July 2005 July 2006 0.25 0.33 0.5 1 0.25 0.33 0.5 1 100 74.07 75.97 76.76 77.18 77.26 79.73 80.74 81.29 200 73.22 74.91 75.62 76.01 76.31 78.63 79.58 80.09 300 72.62 74.15 74.84 75.21 75.29 77.52 78.42 78.90 400 71.93 73.52 74.17 74.53 74.40 76.49 77.35 77.81 500 71.40 72.94 73.58 73.92 73.78 75.75 76.55 76.98 600 70.93 72.53 73.18 73.53 73.30 75.17 75.92 76.32 700 70.79 72.27 72.89 73.22 72.94 74.66 75.36 75.73 800 70.37 72.08 72.77 73.15 72.44 74.49 75.33 75.79 900 73.12 75.39 76.25 76.72 74.93 77.87 78.97 79.57 1000 78.35 81.61 82.74 83.34 82.30 86.47 87.62 88.24 1100 85.13 88.64 90.04 90.79 91.94 95.93 97.19 97.87 1200 89.62 94.10 95.64 96.46 99.80 104.26 105.65 106.40 1300 92.31 96.72 98.40 99.30 104.75 109.06 110.54 111.33 1400 94.19 99.36 101.12 102.06 108.59 112.68 114.13 114.91 1500 94.89 99.47 101.32 102.25 110.28 114.42 115.92 116.72 1600 93.69 97.96 99.61 100.49 108.68 112.00 113.33 114.09 1700 91.88 94.73 96.24 97.05 105.95 108.24 109.55 110.25 1800 86.47 89.10 90.37 91.05 101.21 102.71 103.89 104.52 1900 83.51 86.17 87.38 88.03 95.66 97.65 98.78 99.38 2000 81.23 84.10 85.25 85.87 91.54 93.61 94.72 95.32 2100 79.66 81.60 82.66 83.23 86.70 89.36 90.53 91.15 2200 77.49 79.47 80.46 80.99 83.43 86.01 87.13 87.74 2300 75.79 78.00 78.94 79.45 80.72 83.34 84.53 85.17 2400 74.75 76.83 77.74 78.22 78.62 81.35 82.48 83.08
68 Figure 4 21 Empirical ventilation effectiveness profile with respect to wind speed
69 Figure 4 22 Volumetric flow rate in the attic with respect to time of day averaged over years
70 CHAPTER 5 CONCLUSION Passively cooling attic spaces in hot, humid climate s must be achieved in conjunction with ventilation rates that sustain moisture removal. Moisture movement through building materials will occur in hot humid regions. Capillary movement of moisture through shingles, siding, and venting penetrations will for ce water into plenum spaces. Sealing these spaces with an impermeable membrane will force the plywood sheathing and truss members to store more moisture increasing mold and mildew problems. Moving this moisture requires air movement. It is possible that se aling the attic spaces entirely and insulating the structure would reduce the thermal heat transfer from an attic, but this would provide no outlet for moisture transported through porous roofing materials or past the ceiling plane. The air movement in the attic serves to maintain a microclimate that is consistently dryer than the outdoor environment. Bulk air movement out of the ridge vent, contingent on radiated heat to the roof, is the result of natural thermal buoyancy and illustrates the necessity of r idge venting for proper microclimate development. Designing for a hot humid climate must balance undesirable heat storage while maintaining a moving air mass that can carry moisture through the ridge. A ventilated attic possesses many advantages for a ho t humid climate. While peak attic temperatures are high during summers, the relative humidity remains well below that of the external environment. Relative humidity is of specific interest in this study because the overall health of the attic depends on the temperature and moisture
71 gradient around the wood framing members. The exchange of moisture between the porous wood surfaces and the attic air is a function of relative humidity. The resulting attic climate provides high temperature gradients between t he attic and the conditioned spaces below. Under these circumstances the building envelope boundary of hot, humid regions should be considered conditioned space and insulated accordingly. Overall these data support increased insulation and materials with high thermal resistances positioned at the ceiling deck to reduce the heat flux into conditioned space. A dry attic climate creates efficient insulation and a healthy building. The built environment is served by ventilated attics that surround the air con ditioning ductwork. While traditional building codes allow for installation of this equipment in attic spaces, many cannot be effectively sealed and thereby lose energy in the transmission of conditioned air Best practices for building design would place ductwork in conditioned space s Since much of the existing housing still contains ductwork in attics and cannot easily be retrofitted ventilation and proper installation are necessary to gain efficiency and to reduce incidences of condensation. The model created for this research u se s data collected for attic parameters and creates a transient attic microclimate. Output data from this model show that the movement of moisture enter ing attic s is removed by the ventilation f rom n atural convection. If moistur e become s entrapped in the attic, it w ill become stagnant and condense at the duct surface or within ceiling insulation Proper installation reduces the duct leakage and mitigates the effects of co ndensation. These data indicate that ductwork with crimped or damaged parts present locations where conditioned air can
72 sufficient ly cool the surface and lead to condensation, especially during swing months when air conditioning may cycle during cool moist mornings. The methodology presented in this paper show s that ventilation will remove trapped moisture, mitigating the problems of poorly ins talled ductwork. Adding the necessary inputs for a sealed att ic and closing off the vents provide s insight into the likelihood of condensation Further analysis and long te rm studies on the effect of entrained moisture are necessary to draw conclusions needed to make policy updates to the state building code s This model can be used for future studies looking for insight into the mechanics of heat and moisture transfer in ventilated and un ventilated spaces.
73 APPENDIX A MODEL SOURCE CODE %Attic physical Characterisitics (all in feet) R = 1.5; %Outer Radius of the duct r = 0.75; %Inner Radius of the duct atticWidth = 6.1/.3048; atticHeight = 1.6/.3048; atticLength = 3.162/.3048; sheathingLength = sqrt((atticWidth/2)^2+atticHeight^2); roofPitchDeg = atand(atticHeight/(atticWidth/2)); roofPitchRad = atan(atticHeight/(atticWidth/2)); charLength = atticWidth/2; %characterisitic length for sheathing surfaces A = sheat hingLength*atticLength; A1 = A; A2 = A1; A3 = 20; A4 =1; A345 = 22; a2 = sind(53.973)*(8 R); ductLength=1; Ad = 2*pi()*R*ductLength; %% Surface Area [Data,Variables] = xlsread( 'E:/Research/Attic 9 Data and Weather/Sep 2006.xls' 'Sheet1' ); numrows = size(Data,1); [radData,radVariables] = xlsread( 'E:/Research/Attic 9 Data and Weather/ViewFactors.xls' 'Sheet1' ); F12 = radData(1); F13= radData(2); F15= radData(3); F14= radData(4); F1345= radData(5); F1d= radData(6); F21= radData (7); F2345= radData(8); F2d= radData(9); F31= radData(10); F3451= radData(11); F32= radData(12); F3452= radData(13); F3d= radData(14); Fd1= radData(15); Fd2= radData(16); Fd3= radData(17); Fd4= radData(18);
74 Fd5= radData( 19); F23= radData(20); F25= radData(21); F24= radData(22); sigma = 0.1714E 8; epsr = 0.86; epsd = 0.1; rho = 1 epsd; RH = [10,50,75,100]; epsg =[0.1,0.19,0.22,0.47]; delta = [0,0.1,0.2,0.3,0.4,0.5,0.6,0.8,1.0,1.5,2,2.5,3,5]; Psi = [1,0.9157,0.8491,0.7934,0.7458,0.704,0.6672,0.6042,0.5532,0.4572,0.39,0.3401,0.301 6,0.2077]; %Guesses for iteration of duct surface temp Tduct = zeros(numrows,10); Troof1 = zeros(numrows,1); Troof1K = zeros(numrows,1); Troof2 = zeros(numrows,1); Troo f2K = zeros(numrows,1); Tceil = zeros(numrows,1); To = zeros(numrows,1); TceilK = zeros(numrows,1); ToK = zeros(numrows,1); aRH = zeros(numrows,1); atticTemp = zeros(numrows,1); airVelocity = zeros(numrows,1); exteriorTemp= zeros(numrows,1); bayTemp= zeros (numrows,1); time = zeros(numrows,1); tolerance = zeros(numrows,10); %Initial guess for exterior duct temp for h=1:10 for i=1:numrows Tduct(1,h) = Data(1,5); tolerance(i,h) = 1; end end %Tduct(1,1) = Data(2,5); %[ radData,radVariables] = xlsread5('F:/Research/Attic 9 Data and Weather/ViewFactors.xls','Sheet1'); for h=1:1 epsd = 0.1*h; rho = 1 epsd;
75 for i=2:numrows Tduct(i,h) = Data(i,5)+30; atticTemp(i,1) = Data(i,9); airVelocity(i,1)= abs(Data(i,14)); exteriorTemp(i,1) = Data(i,28); bayTemp(i,1) = Data(i,8); time(i,1) = Data(i,2); % Calculation of interior boundary conditions if atticTemp(i)>85 ductTemp = 55; ductAirVelocity = 2; elseif atticTemp(i)<62 ductTemp = 85; ductAirVelocity = 2; else ductTemp = bayTemp(i,1); ductAirVelocity = 1; end filmTemp = ductTemp; ductAirTK = (ductTemp+459.67)/1.8; ductAirK = (0.6325E 5*sqrt(ductAirTK)/((1+245.4*10^( 12/ductAirTK))/ductAirTK))*241.77; %thermal conductivity of Air (in english units I think) ductAirCp = (3.4763+(1.066E 4)*ductAirTK)*0.068559; ductAirMu = (145.8*ductAirTK*sqrt(ductAirTK)/(ductAirTK+110.4))*241.9E 7; ductAirRho = 22.0493/ductAirTK; ductAirKinVis = ductAirMu/ductAirRho; ductBeta= 1/(filmTemp+459.67); ductPr = 0.7880 2.63E 4*ductAirTK; duct Ra = 4.16975E8*ductBeta*ductAirRho*ductAirCp*ductTemp*charLength^3/ductAirKinVis/duct AirK; ductRe = ductAirVelocity*charLength/ductAirKinVis; if ductRe<5E5 ductforNu = 0.664*ductPr^(1/3)*sqrt(ductRe); else ductforNu = ductPr^(1/3)*(.037*ductRe^0.8 850); end %Calculate Convective Coefficients Interior forConCoeff = ductforNu*ductAirK/charLength; ductConvCoeff = forConCoeff;
76 % Calculation of Boundary Conditions for C onvective transfer atticT = atticTemp(i,1); filmTemp = atticT; atticAirVelocity = airVelocity(i,1)/60; atticAirTK = (atticT+459.67)/1.8; atticAirK = (0.6325E 5*sqrt(atticAirTK)/((1+245.4*10^( 12/atticAirTK))/atticAir TK))*241.77; %thermal conductivity of Air (in english units I think) atticAirCp = (3.4763+1.066E 4*atticAirTK)*0.068559; atticAirMu = (145.8*atticAirTK*sqrt(atticAirTK)/(atticAirTK+110.4))*241.9E 7; atticAirRho = 22.0493/atticAirTK ; atticAirKinVis = atticAirMu/atticAirRho; atticBeta= 1/(filmTemp+459.67); atticPr = 0.7880 (2.63E 4)*atticAirTK; atticRa = 4.16975E8*atticBeta*atticAirRho*atticAirCp*atticT*charLength^3/atticAirKinVis/atticAirK; a tticRe = atticAirVelocity*charLength/atticAirKinVis; atticnatNu = 0.56*(atticRa*cosd(90 roofPitchDeg))^0.25; if atticRe<5E5 atticforNu = 0.664*atticPr^(1/3)*sqrt(atticRe); else atticforNu = atticPr^(1/3)*(.0 37*atticRe^0.8 850); end %Calculate Exterior Convective Coefficients atticForConCoeff = atticforNu*atticAirK/charLength; atticNatConCoeff = atticnatNu*atticAirK/charLength; atticConvCoeff = nthroot (atticForConCoeff^3+atticNatConCoeff^3,3); % Defining Cyclindrical Mesh, Concentric circles R and Theta %deltaR=R r; % duct insulation thickness in ft. insTH = (R r)/12; % insulation thickness in inches. i nsK = 1/5; % Duct insulation value R 5 per inch Troof1(i,1) = Data(i,5); Tceil(i,1) = Data(i,16); To(i,1) = Data(i,28); aRH(i,1) = Data(i,11); Troof2(i,1) = Troof1(i,1); % Actually values in Rankine (all Kelvin temperature are Rankine) for % units consistancy Troof1K(i,1) = Troof1(i,1)+459.67; %*0.555556 + 255.37; Troof2K(i,1) = Troof2(i,1)+459.67; %*0.555556 + 255.37; TceilK(i,1) = Tceil(i,1)+459.67; %*0.555556 + 255.37; ToK(i,1) = To(i,1)+459.67; %*0.555556 + 255.37; Length = ductLength*atticLength; %length of duct in ft
77 Aduct = 2*pi()*r*Length + 2*pi()*R*Length + 2*(pi()*R^2 pi()*r^2); %2*pi()*R*Length; % area of duct in feet^2 Ad uctSurface = 2*pi()*R*Length; %Surface of duct area Vduct = pi()*Length*(R^2 r^2); %pi()*(deltaR)^2*Length; CpDuct = 0.23; %Btu/lbF rhoDuctSurface = 4.0; %lb/ft^3 j=0; while abs(tolerance(i,h))>0.7 j = j+1; % Conductive Heat Transfer Qcond = Aduct*insK*insTH*(ductTemp Tduct(i,h)); % Convective Heat Transfer Qconv = AductSurface*atticConvCoeff*(Tduct(i,h) atticTemp(i,1)); % Ma ss Capacitance of duct surface Qstored = rhoDuctSurface*Vduct*CpDuct*(Tduct(i,h) Tduct(i 1,h)); % Radiative Heat Transfer beta = interp1(RH,epsg,aRH(i)); deltaR = beta*a2; deltaCeil = beta*2*pi()*R; Psibr = interp1(delta,Psi,deltaR); Psibc = interp1(delta,Psi,deltaCeil); Q1S = Psibr*sigma*A1*(((Troof1K(i,1)^4 Troof2K(i,1)^4)/((1/epsr)+F1d*Fd2*rho*((1/epsr) 1))) + (Troof1K(i,1)^4 TceilK(i, 1)^4)/((1/epsr)+F1d*Fd3*rho*((1/epsr) 1))); Q2S = Psibr*sigma*A2*((Troof2K(i,1)^4 Troof1K(i,1)^4)/((1/epsr)+F2d*Fd1*rho*((1/epsr) 1)) + (Troof2K(i,1)^4 TceilK(i,1)^4)/((1/epsr)+F2d*Fd3*rho*((1/epsr) 1))); Q3S = Psibc*sigma*A3*(( TceilK(i,1)^4 Troof1K(i,1)^4)/((1/epsr)+F3d*Fd1*rho*((1/epsr) 1)) + (TceilK(i,1)^4 Troof2K(i,1)^4)/((1/epsr)+F3d*Fd2*rho*((1/epsr) 1))); TductK = Tduct(i,h)+459.67; %*0.555556 + 255.37; Q1g = Psibr*sigma*A1*(epsr*(Troof1K(i,1) ^4 ToK(i,1)^4) + ((Troof1K(i,1)^4 Troof2K(i,1)^4)/((1/epsr)+F12*((1/epsr) 1))) + (Troof1K(i,1)^4 TceilK(i,1)^4)/((1/epsr)+F13*((1/epsr) 1)) + epsr*(Troof1K(i,1)^4 ToK(i,1)^4)*F14 + epsr*(Troof1K(i,1)^4 ToK(i,1)^4)*F15 + (Troof1K(i,1)^4 TductK^4 )/((1/epsr)+F1d*((1/epsd) 1))); Q2g = Psibr*sigma*A2*(epsr*(Troof2K(i,1)^4 ToK(i,1)^4) + (Troof2K(i,1)^4 Troof1K(i,1)^4)/((1/epsr)+F21*((1/epsr) 1)) + (Troof2K(i,1)^4 TceilK(i,1)^4)/((1/epsr)+F23*((1/epsr) 1)) + epsr*(Troof2K(i,1)^4 ToK (i,1)^4)*F24 + epsr*(Troof2K(i,1)^4 ToK(i,1)^4)*F25 + (Troof2K(i,1)^4 TductK^4)/((1/epsr)+F2d*((1/epsd) 1)));
78 Q3g = Psibc*sigma*A3*((TceilK(i,1)^4 Troof1K(i,1)^4)/((1/epsr)+F31*((1/epsr) 1)) + (TceilK(i,1)^4 Troof2K(i,1)^4)/((1/epsr)+F3 2*((1/epsr) 1)) + (TceilK(i,1)^4 TductK^4)/((1/epsr)+F3d*((1/epsd) 1))); Qd = sigma*AductSurface*((TductK^4 Troof1K(i,1)^4)/((1/epsd)+Fd1*((1/epsr) 1)) + (TductK^4 Troof2K(i,1)^4)/((1/epsd)+Fd2*((1/epsr) 1)) + (TductK^4 TceilK(i,1)^4)/((1 /epsd)+Fd3*((1/epsr) 1)) + epsd*(TductK^4 ToK(i,1)^4)*Fd4 + epsd*(TductK^4 ToK(i,1)^4)*Fd5); Q1 = Q1S + Q1g; Q2 = Q2S + Q2g; Q3 = Q3S + Q3g; Qradiative = (Q1 + Q2 + Q3 + Qd); tolerance(i,h) = Qs tored (Qcond + Qconv + Qd); Tduct(i,h) = Tduct(i,h) 0.01; if j>10000 Tduct(i,h) = Tduct(i 1,h); tolerance(i,h)=0.5; end end Tduct(i,h) end end
79 APPENDIX B HEATING AND AIR CONDITIONING PARAMETERS Design Cooling Day Table B 1 System Data Component Location Dry Bulb Temp (F) Specific Humidity (lb/lb) Airflow (CFM) CO2 Level (ppm) Sensible Heat (BTU/hr) Latent Heat (BTU/hr) Ventilation Air Inlet 92.1 0.01618 36 400 535 1367 Vent Return Mixing Outlet 78.7 0.00836 1022 795 Supply Fan Outlet 79.4 0.00836 1022 795 757 Central Cooling Coil Outlet 54 0.00805 1022 795 27838 1486 Central Heating Coil Outlet 54 0.00805 1022 795 0 Cold Supply Duct Outlet 55 0.00805 971 795 Zone Air 79.5 0.00808 971 810 25511 120 Return Plenum Outlet 79.5 0.00808 971 810 0 Duct Leakage Air Outlet 54 0.00806 51 795 Return Duct Outlet 78.2 0.00808 1360 809 Air Density x Heat Capacity x Conversion Factor: At sea level = 1.080; At site altitude = 1.074 BTU/(hr CFM F) Air Density x Heat of Vaporization x Conversion Factor: At sea level = 4746.6; At site altitude = 4720.8 BTU/(hr CFM) Site Altitude = 151.0 ft Table B 2 Zone Data Zone Name Zone Sensible Load (BTU/hr) T stat Mode Zone Cond (BTU/hr) Zone Temp (F) Zone Airflow (CFM) CO2 Level (ppm) Terminal Heating Coil (BTU/hr) Zone Heating Unit (BTU/hr) Zone 1 27544 Cooling 25511 79.5 971 810 0 0
80 Design Heating Day Table B 3 System Data C omponent Location Dry Bulb Temp (F) Specific Humidity (lb/lb) Airflow (CFM) CO2 Level (ppm) Sensible Heat (BTU/hr) Latent Heat (BTU/hr) Ventilation Air Inlet 30 0.00173 41 400 1697 0 Vent Return Mixing Outlet 66.1 0.00173 658 543 Supply Fan Outlet 66.8 0.00173 658 543 488 Central Cooling Coil Outlet 66.8 0.00173 658 543 0 0 Central Heating Coil Outlet 95.7 0.00173 658 543 20435 Cold Supply Duct Outlet 95 0.00173 625 543 Zone Air 67.1 0.00173 625 552 18761 0 Return Plenum Outlet 67.1 0.00173 625 552 0 Duct Leakage Air Outlet 95.7 0.00173 33 543 Return Duct Outlet 68.5 0.00173 1360 552 Air Density x Heat Capacity x Conversion Factor: At sea level = 1.080; At site altitude = 1.074 BTU /(hr CFM F) Air Density x Heat of Vaporization x Conversion Factor: At sea level = 4746.6; At site altitude = 4720.8 BTU/(hr CFM) Site Altitude = 151.0 ft Table B 4 Zone Data Zone Name Zone Sensible Load (BTU/hr) T stat Mode Zone Cond (BTU/hr) Zone Temp (F) Zone Airflow (CFM) CO2 Level (ppm) Terminal Heating Coil (BTU/hr) Zone Heating Unit (BTU/hr) Zone 1 19211 Heating 18761 67.1 625 552 0 0
81 LIST OF REFERENCES  A. Silberstein and H. Hens, "Effects of Air Moisture Flows on the Thermal Performance of Insulations in Ventilated Roofs and Walls," Journal of Building Physics, vol. 19, pp. 367 385, 1996.  I. Samuelson, "Hygrothermal Performance of Attics," Journal of Building Physics, vol. 22, pp. 132 146, 1998.  J. Lstiburek, K. Pressnail and J. Timusk, "Air Pressure and Building Envelopes," Journal of Building Physics, vol. 26, pp. 53 91, 2002.  American Society of Agricultural and Biological Engineer s, Fundamentals 2007, Atlanta, GA: ASHRAE, 2007.  P. Blom, "Venting Attics and Pitched, Insulated Roofs," Journal of Building Physics, vol. 25, pp. 32 50, 2001.  W. Miller, M. Keyhani, T. Stovall and A. Youngquist, "Natural Convection Heat Transfer in Roofs with Above Sheathing Ventilation," American Society of Heating, Refrigeration, and Air Conditioning Engineers, Atlanta, GA, 2007.  I. S. Walker and T. W. Forest, "Field Measurements of Ventilation Rates in Attics," Building and Envir onment, vol. 30.3, pp. 333 347, 1995.  M. A. Medina and B. Young, "A perspective on the effects of climate and local environmental variables on the performance of attic radiant barriers in the United States," Building and Environment, vol. 41, pp. 1767 1778, 2006.  L. Shashua Bar, Y. Tzamir and M. Hoffman, "Thermal effects of building geometry and spacing on the urban canopy layer microclimate in a hot humid climate in summer," International Journal of Climatology, vol. 24, pp. 17 29 1742, 2004.  I. S. Walker, T. W. Forest and D. J. Wilson, "An attic interior infiltration and interzone transport model of a house," Building and Environment, vol. 40, pp. 701 718, 2005.  H. Rutkowski, Residential Duct Systems Manual D, Wa shington, D.C.: Air Conditioning Contractors of America, 1995.  M. F. Modest, Radiative Heat Transfer, Second Edition, Oxford, UK: Elsevier Science, 2003.  W. Porter, "Moisture transport in attic spaces located in hot humid climates, Ph.D. Disseration," University of Florida, Gainesville, FL, 2003.
82  J. W. Walker, Ventilation of Agricultural Structures, St.Josephs, MI: American Society of Agricutural Engineers, 1983.  H. Ruthkowski, Residential Duct Systems Manual D, Washington, DC: Air Conditioning Contractors of America, 1995.  Z. Pavlik and R. Cerny, "Experimental assessment of hygrothermal preformance of an interior thermal insulation system using a laboratory technique simulating on site conditions," Energy and Buildin gs, vol. 40, pp. 673 678, 2008.  A. Mochida, H. Yoshino, S. Miyauchi and T. Mitarmura, "Total analysis of cooling effects of cross ventilation affected by microclimate around a building," Solar Energy, vol. 80, pp. 371 382, 2006.  L. Adelard, H. Boyer, F. Garde and J. C. Gatina, "A detailed weather data generator for building simulations," Energy and Buildings, no. 31, pp. 75 88, 2000.  M. A. Brown and L. G. Berry, "Determinants of Program Effectiveness: Results of the National Weatherization Evaluation," Energy, vol. 20, no. 8, pp. 729 743, 1995.  H. C. Chiu, J. H. Jang and W. M. Yan, "Mixed convection heat transfer in horizontal rectangular ducts with radiation effects," International Journal of Heat and Mass Transfer, vol. 50, pp. 2974 2881, 2007.  F. S. De la Flor and A. D. Servando, "Mode ling microclimate in urban environments and assessing its influence on the performance of surrounding buildings," Energy and Buildings, vol. 36, pp. 403 413, 2004.  H. Janssen, B. Blocken and J. Carmeliet, "Conservative modeling of the moisture and heat transfer in building components under atmospheric excitation," International Journal of Heat and Mass Transfer, no. 50, pp. 1128 1140, 2007.  P. Norton, C. Christensen, E. Hancock, G. Barker and P. Reeves, "The NREL/Habitat for Humanity Zero En ergy Home: A Cold Climate Case Study for Affordable Zero Energy Homes," National Renewable Energy Laboratory, Golden, CO, 2008.  D. Parker, J. Sonne and J. Sherwin, "Comparitive Evaluation of the Impact of Roofing Systems on Residential Cooling Energ y Demand in Florida," in Proceedings of American Council for an Energy Efficient Economy 2002 Summer Study Washington, D.C., 2002.  D. S. Parker and M. D. Mazzara, "Monitored Energy Use Patterns in Low Income Housing in a Hot and Humid Climate," in Proceedings of the Tenth Symposium on Improving Building Systems in Hot and Humid Climates Fort Worth, TX, 1996.
83  B. B. Stankov, E. R. Westwater and E. E. Gossard, "Use of Wind Profiler Estimates of Significant Moisture Gradients to Improve Humidity Profile Retrieval," Journal of Atmospheric and Oceanic Technology, vol. 13, pp. 1285 1290, 1996.  P. Wemhoff, "GRU Trade Alliance Program: Constructing and Energy Efficient Home," Jacksonville, FL, 1995.
84 BIOGRAPHICAL SKETCH Barrett Mooney is interested in energy efficiency and in the residential and agricultural sectors. He began his studies as an undergraduate at the University of Florida in Nuclear Engineering where he earned his Bachelor of S cience degree While p ursuing his degree he was inducted to the American Nuclear Honor Society, Alpha Nu Sigma, and graduated Cum Laude. After graduation, Barrett worked with th e United States Nuclear Regulatory Commission inspecting mechanical systems for regulatory compliance. He returned to graduate school apply his background in physics to the study of temperature and moisture movement in agricultural structures. While at the D epartment of Agricultural and Biological Engineering, Barrett earned his Master of Engineering and worked on several non thesis projects including greenhouse modeling and animal housing designs. He studied weather modeling and moisture transport in pursu it of his doctorate, and was inducted into the Honor Society of Agriculture, Gamma Sigma Delta, and became an active member of the American Society of Agricultural and Biological Engineers. After graduation Barrett would like to publish his research and ad vance the field of transient thermal and moisture modeling in various systems including crop dynamics, animal housing, and greenhouse operations.