University of Southern California Dissertations and Theses

Digital tree simulation for residential building energy savings: shading and evapotranspiration

(USC Thesis Other)

Digital tree simulation for residential building energy savings: shading and evapotranspiration

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content DIGITAL TREE SIMULATION FOR RESIDENTIAL BUILDING ENERGY SAVINGS: SHADING AND EVAPOTRANSPIRATION by Yi-Lun Cheng A Thesis Presented to the FACULTY OF THE USC SCHOOL OF ARCHITECTURE UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree MASTER OF BUILDING SCIENCE August 2012 Copyright 2012 Yi-Lun Cheng ii ACKNOWLEDGEMENTS I would like to express my gratitude to all those who gave me the possibility to complete this thesis. In the first place I would like to thank Professor Karen Kensek, my committee chair, for her stimulating suggestions and encouragement to help me through the time of research from the very early stage and writing of the thesis. Professor Marc Schiler, my committee member, for helping me shapes the research with his extensive knowledge. Also, his master thesis has been an important research material for my study. Professor Alexander Robinson, the thesis advisor, for his suggestions about landscape programs and sharing ideas and methods in all the presentations. In my daily work I have been blessed with a friendly and cheerful group of fellow classmates, the team Building Science. Special thanks go in particular to Carlita Broussard for reading through the whole book and checking the grammar and spelling for me. Bruce Carter has been, always, my joy and my guiding light, and I thank him. Last but not the least, I would like to thank my parents for supporting me spiritually throughout my life and taking great care of my grandparents while I was not at home to help. iii TABLE OF CONTENTS ACKNOWLEDGEMENTS.................................................................................................... ii LIST OF TABLES ............................................................................................................... vii LIST OF FIGURES ................................................................................................................x ABSTRACT:........................................................................................................................xiv HYPOTHESIS: ....................................................................................................................xv 1. CHAPTER 1 ....................................................................................................................1 1.1. Introduction .............................................................................................................1 1.2. Termiology.............................................................................................................. 3 1.2.1. Software............................................................................................................ 3 1.2.2. Terms relating to the tree .................................................................................5 1.2.3. Climate zones ..................................................................................................10 1.2.4. Terms for the different types of trees used in this thesis............................... 11 1.3. Tree benefits on energy saving ..............................................................................16 1.3.1. Tree shading ....................................................................................................16 1.3.2. Evapotranspiration .........................................................................................16 1.3.3. Others..............................................................................................................16 1.4. Background on tree shading.................................................................................. 17 1.4.1. Energy conserving tree shading design .......................................................... 17 1.4.2. Tree shading simulation .................................................................................18 1.4.3. A method by DOE-2....................................................................................... 20 1.4.4. A method by eQuest....................................................................................... 22 1.4.5. Previous developed tools ............................................................................... 24 1.4.6. Comparison.................................................................................................... 26 1.5. Background on tree evapotranspiration ...............................................................27 iv 1.5.1. Measurement of tree transpiration.................................................................27 1.5.2. Previous developed digital tools .....................................................................27 1.6. Simulation tools for landscape architects ............................................................ 28 1.6.1. CITYgreen....................................................................................................... 28 1.6.2. TownScope ..................................................................................................... 28 1.6.3. UFORE ............................................................................................................31 1.6.4. iTree ................................................................................................................31 1.6.5. Tree Benefit Estimator................................................................................... 33 1.7. Conclusion............................................................................................................. 36 2. CHAPTER 2..................................................................................................................37 2.1. Energy simulation software...................................................................................37 2.1.1. Simulation Engine.......................................................................................... 38 2.1.2. User interface................................................................................................. 43 2.1.3. Interoperability .............................................................................................. 44 2.2. Limitations ........................................................................................................... 44 2.2.1. Tree model ..................................................................................................... 44 2.2.2. Calculation on shading effect ........................................................................ 45 2.2.3. Transpiration rate ......................................................................................... 45 2.3. Methodology......................................................................................................... 46 3. CHAPTER 3..................................................................................................................47 3.1. eQuest and Vasari..................................................................................................47 3.1.1. Data exchange between eQuest and Vasari ....................................................47 3.1.2. INP file ........................................................................................................... 49 3.1.3. Bringing the tree from Vasari to eQuest.........................................................51 3.1.4. Workflow.........................................................................................................51 3.2. Problems............................................................................................................... 53 3.2.1. Building shade ............................................................................................... 53 3.2.2. Open .INP file in eQuest................................................................................ 54 3.2.3. Problem of building shade .INP file.............................................................. 54 3.2.4. Summary ....................................................................................................... 56 3.3. Surface normal ..................................................................................................... 56 3.4. Deciding on a shape for the tree ...........................................................................57 v 3.4.1. Decide a shape by calculating solar radiation in Vasari.................................57 3.4.2. Result............................................................................................................. 60 3.5. Energy Calculations of Windows versus Walls.....................................................61 3.5.1. Test model information ..................................................................................61 3.5.2. Window v.s. wall............................................................................................ 62 3.6. Discussion ............................................................................................................ 63 4. CHAPTER 4 ................................................................................................................ 64 4.1. Shadow Patterns of different tree models............................................................ 64 4.1.1. Tool: Autodesk Ecotect Analysis.................................................................... 64 4.1.2. Workflow........................................................................................................ 65 4.1.3. Results............................................................................................................ 65 4.2. Tilted plane............................................................................................................ 71 4.2.1. Definition of tilt .............................................................................................. 71 4.2.2. Limits of tilted plane tree model orientation.................................................72 4.3. Scheduling foliage .................................................................................................74 4.3.1. Evergreen and deciduous trees.......................................................................75 4.3.2. Leaf-on and leaf-off periods...........................................................................75 4.3.3. Winter and summer transmittance value ......................................................76 4.3.4. Schedule for energy simulation programs.....................................................77 4.3.5. Schedule deciduous trees in eQuest.............................................................. 78 4.3.6. Schedule workflow in eQuest .........................................................................79 4.4. Summary .............................................................................................................. 83 5. CHAPTER 5................................................................................................................. 86 5.1. Comparison to Tree Benefits Estimator............................................................... 86 5.1.1. Overview of Tree Benefit Estimator............................................................... 87 5.1.2. Energy simulation methods for Tree Benefit Estimator: Shading ............... 87 5.1.3. Input comparison........................................................................................... 89 5.1.4. Absolute energy saving output comparison ...................................................91 5.1.5. Conclusions regarding Tree Benefit Estimator and eQuest.......................... 94 5.2. Comparison to a previous method in DOE-2 ...................................................... 95 5.2.1. Summary of the compared study................................................................... 95 5.2.2. Building description ...................................................................................... 98 vi 5.2.3. Tilted plane model in eQuest ........................................................................ 99 5.2.4. Evapotranspiration....................................................................................... 101 5.2.5. Results........................................................................................................... 101 5.2.6. Conclusion regarding to rectangular box method and tilted plane method 104 5.3. Tree shading effects on daylighting in eQuest....................................................105 5.3.1. Reflection ......................................................................................................105 5.3.2. Simulation of the case study house ............................................................. 106 5.3.3. Result of the case study house..................................................................... 108 5.3.4. Simulation and results of a simple box room .............................................. 110 6. CHAPTER 6 ............................................................................................................... 113 6.1. Conclusion ........................................................................................................... 113 6.2. Future work suggestions:.................................................................................... 115 6.2.1. On-site experiment ....................................................................................... 115 6.2.2. Scale of simulation ....................................................................................... 116 6.2.3. Tool to calculate transmittance value of the tree canopy ............................ 116 6.2.4. Other potential programs............................................................................. 116 6.2.5. In depth studies on tree’s effect on daylighting........................................... 116 6.2.6. Reflectance of tree leaf ................................................................................. 117 BIBLIOGRAPHY:.............................................................................................................. 118 APPENDIX A-California climate zones............................................................................124 APPENDIX B- Comparision test result ............................................................................125 APPENDIX C- Case study simulation ..............................................................................126 APPENDIX D- Daylighting simulation output.................................................................129 APPENDIX E-Schedule “GndPrm Occ Sch” definition in eQuest ...................................132 APPENDIX F- Daylighting simulation result...................................................................134 vii LIST OF TABLES Table 1: 2D and 3D tree shapes ..........................................................................................12 Table 2: Trees change over time .........................................................................................15 Table 3: Comparisons between GLAS and SPS programs ................................................ 26 Table 4: List of simulation methods discussed in Chapter 1............................................. 36 Table 5: Simulation engines comparison chart..................................................................41 Table 6: User interface comparison chart ......................................................................... 44 Table 7: Import models from Vasari to eQuest ..................................................................55 Table 9: Average incident solar radiation on a vertical surface from different tree models (A). .............................................................................................................................. 59 Table 10: Average incident solar radiation on a vertical surface from different tree models (B)................................................................................................................... 59 Table 11: Percentage savings relative to the case with no tree, building shade transmittance value :0.3............................................................................................. 63 Table 12:Percentage savings relative to the case with no tree, building shade transmittance value:0.8 ............................................................................................. 63 Table 13 Shadow Patter Comparison in Ecotect (from 6AM to 8PM,time step value of 60) ..................................................................................................................................... 66 Table 14: Shadow pattern comparison in Ecotech (from 6AM to 8PM, time step value of 80)................................................................................................................................67 viii Table 15 Shadow pattern comparison in Ecotech (from 7AM to 5PM,time step value of 20)............................................................................................................................... 68 Table 16: 6 Types of schedule in eQuest............................................................................ 78 Table 17: .INP file for building shade schedule (From top: day-schedule, week-schedule, year-schedule, building shade schedule) ................................................................... 83 Table 18 Tree simulation parameters (McPherson, 1999)................................................ 88 Table 19 Reported transmittance value for Norway maple by different methods (McPherson, 1984) ..................................................................................................... 89 Table 20: Inputs for the tree model in eQuest .................................................................. 90 Table 21: Inputs for Tree Benefit Estimator...................................................................... 90 Table 22: Building description in eQuest. ..........................................................................91 Table 23: Monitoring schedule for the casestudy house T1 (Site C) in 1992.................... 96 Table 24: Methods used in compared paper to model shade trees (Akbari et al., 1997) . 96 Table 25: Building characteristic in eQuest based on the compared reported by Akbari et al.................................................................................................................................. 98 Table 26: Trees in experiment and the correlated models in eQuest ............................... 99 Table 27: Absolute daily energy savings from shade trees............................................... 101 Table 28: Percentage cooling energy savings, measure vs. simulation ...........................103 Table 29: Spaces in active zone........................................................................................ 106 Table 30: Lighting energy usage output in eQuest........................................................... 110 ix Table 31: Lighting energy usage output (small box room)............................................... 112 Table 32: California climate zone location .......................................................................124 Table 33: Estimator result: annual absolute energy saved (kWh)...................................125 Table 34: eQuest result: annual absolute energy saved (kWh)........................................125 Table 35: Simulation in eQuest result ..............................................................................128 Table 36: Illuminance level, no trees................................................................................134 Table 37: Illuminance level, trees (reflectance 0.5) .........................................................134 Table 38: Illuminance level, trees (reflectance 1.0) .........................................................135 x LIST OF FIGURES Figure 1: Reflection, absorption and transmission of a leaf (Moffat, Schiler, 1981) ...........5 Figure 2: Two methods to simulate foliage pattern. ..........................................................13 Figure 3: Generalized tree canopy shading as modeled in DOE-2 (Akbari et al., 1987) .. 20 Figure 4: (Upper) the model for one tree on the south side of the house; ........................21 Figure 5: Autonomy daylight plot for dense urban area (Saxena et al., 2011).................. 22 Figure 6: Shading model (Saxena et al., 2011) .................................................................. 23 Figure 7: TownScope example screen................................................................................ 29 Figure 8: Tree element screen in TownScope ................................................................... 30 Figure 9: iTree Street sample input screen ....................................................................... 32 Figure 10: i-Tree Street sample tree data input screen ..................................................... 32 Figure 11 Tree Benefits Estimator input window .............................................................. 34 Figure 12 Tree Benefits Estimator sample report ............................................................. 35 Figure 13: General data flow of simulation engines .......................................................... 39 Figure 14: Tree input in HEED .......................................................................................... 43 Figure 15: Research methodology...................................................................................... 46 xi Figure 16: Two possible building shade tree models created directly in eQuest(left: box- trees, right: cross-trees).............................................................................................. 48 Figure 17: Part of the DOE-2 input text file....................................................................... 49 Figure 18: Example .INP file eQuest model ...................................................................... 50 Figure 19: Merging two .INP files ...................................................................................... 52 Figure 20: One-side surface normal (left) and twp-side surface normal (right)..............57 Figure 21: Surface normal test result..................................................................................57 Figure 22: Test model in Vasari......................................................................................... 58 Figure 23: Comparing the insolation value of different tree models................................ 59 Figure 24 「4-planes box and daily sunpath in December................................................ 60 Figure 25: Window versus wall test model in eQuest ....................................................... 62 Figure 26: Annual electricity use in 3 California climate zones........................................ 62 Figure 27 Test models in Ecotect....................................................................................... 65 Figure 28 Overlapped shadow pattern diagram ............................................................... 66 Figure 29 Timetable plot of Los Angeles area ................................................................... 69 Figure 30 Sun chart in Los Angeles area, left: summer; right: winter ............................. 70 Figure 31: Positioning building shade (LBNL, 2008) ........................................................ 71 xii Figure 32: Tilt degree definition.........................................................................................72 Figure 33: Front view of tilted plane and the building. .....................................................73 Figure 34: 4 different building and building shade plane layouts.....................................74 Figure 35: Leaf-on and leaf -off periods for some California deciduous trees. (Hammond, 1981).............................................................................................................................76 Figure 36: Winter and summer sunlight penetration for some California tree species (Hammond, 1981)........................................................................................................77 Figure 37: eQuest day schedules screenshot ..................................................................... 80 Figure 38: eQuest week schedules screenshot ...................................................................81 Figure 39: eQuest annual schedules screenshot ............................................................... 82 Figure 40: Workplace of eQuest(left) and Estimator(right)............................................. 86 Figure 41: Tree Benefit Estimator and eQuest energy saving outputs for 4 orientations 92 Figure 42: Measured vs. simulated (Akbari et. al, 1997)....................................................97 Figure 43: Building shade models in eQuest................................................................... 100 Figure 44: Comparison of daily cooling energy saved .....................................................102 Figure 45: Comparison on percentage energy saving between measured and simulated data ............................................................................................................................103 Figure 46: eQuest daylighting zone screen.......................................................................107 Figure 47: Average daylighting illuminance in SE-Perim space..................................... 109 xiii Figure 48: Average daylighting illuminance in NW-Perim space .................................. 109 Figure 49: A simple box room in eQuest ...........................................................................111 Figure 50: Average illuminance in the box room ..............................................................111 xiv ABSTRACT: Landscaping in general and trees specifically can be beneficial in helping to mitigate several environment problems such as carbon sequestration, urban hear island, reduced air quality due to pollution, and erosion. Yet simulation software programs are often weak in enabling designers to understand analytically and to specifically predict energy savings through the use of landscaping. The assessment done on existing energy simulation software shows that several programs could not model trees directly. Previous studies have looked at the effect of shade trees on energy use. Different strategies were used to try to model trees in these software programs, and simple case studies were undertaken to verify the results. Two critical potential energy saving features of trees were studied: direct shading on a wall and/or window and evapotranspiration. Shade trees can help in reducing solar gain and thus reduce energy consumption for cooling and should be taken seriously as a climate change adaptation initiative. The evapotranspiration of shade trees can also change both air temperature and relative humidity in the micro-climate. The evapotranspiration impact of trees has been simulated in DOE-2; the ambient temperatures were changed to estimate the indirect cooling effects of trees in reducing air-conditioning energy use. This methodology was cumbersome. In addition, several of the other software programs could not simulate this effect directly. Different methods had to be employed to achieve this capability. xv The final results show the limitations in simulation in practice. Unfortunately, the current energy simulation programs do not have the capability to simulate the tree’s effect on building energy usage. For shading, the proposed methodology can benefit in developing integrated tools for simulating the shading effects although there are still many issues unsolved. And there is a need to incorporate evapotranspiration. As a holistic building design tool, the direct and indirect effects of tree should be considered. HYPOTHESIS: It is possible to create an algorithm model of a tree that can be used within existing energy software programs to calculate its shading and evapotranspiration benefits. 1 1. CHAPTER 1 1.1. Introduction A U.S Department of Energy survey published in 2009 (U.S. Energy Information Administration, 2009) reports that transportation represents only 29% of total energy consumption by end-use sector. The same survey reports industrial consumption at 30%, commercial consumption at 19%, and residential consumption at 22%. Residential and commercial buildings account for roughly 40% of all U.S. energy use 1 . Energy efficient measures such as increasing daylight and adapting passive ventilation design to create high performance building are the quickest and most cost-effective practices to reduce energy usage. The U.S. Department of Energy estimates that heating and cooling account for 56% of energy use in a typical home. Technological solutions have been studied and developed for heating and cooling in buildings for many years, but passive design has been the most cost-effective way compared to mechanical system design (Moffat, Schiler, 1981). To reduce a building’s energy use, landscaping is one of the passive design strategies recognized by the earliest house builders. “Knowledge of and respect for the climate and a remarkable understanding of building materials and landscape contributed to the exceptional success of early architecture” (Moffat, Schiler, 1981). 1 Buildings Energy Databook, 2006. US Department of Energy and Annual Energy Review 2007. DOE/EIA- 0384 (2007). Energy Information Administration, U.S. Department of Energy. June 2008. 2 Landscaping to save energy is a powerful tool for conservation. Vegetation provides a form of passive cooling by two ways. First, shade from the trees reduces the conversion of radiant energy to heat, thereby reducing the surface temperatures of the shaded area. Increased planting of urban shade trees has been suggested as one way of conserving energy by reducing the demand for heating and cooling (Akbari, Kurn and Bretz 1997). Second, evapotranspiration results in cooling the leaf surface and adjacent air due to the exchange of latent heat (McPherson, 1984). On summer days, a tree can act as a natural cooler using up to 100 gallons of water a day and thus lowering surrounding temperature (Kramer and Kozlowski, 1960). The effect of evapotranspiration is considered minimal in winter because of the absence of leaves on deciduous and the lower ambient temperature (Akabari, 2002). Increased evapotranspiration can create cooler environments for buildings, thus consuming less cooling power and energy. In this study, six most commonly used software are assessed to determine whether they take the effect of exterior shades and evapotranspiration effects of trees into account since they are the main factors trees have on reducing a building’s energy use. 3 1.2. Termiology 1.2.1. Software DOE-2 2 : DOE-2 is a popular calculation engine for building energy simulation. It was developed by James J. Hirsch & Associates (JJH) in collaboration with Lawrence Berkeley National Laboratory (LBNL), with LBNL DOE-2 work performed mostly under funding from the United States Department of Energy (US DOE) and other work performed mostly under funding from a wide range of industry organizations (Hirsch, 2009). The raw DOE-2 is difficult to master because users are required to understand the input language for building description (BDL), loads description, system inputs, and economics input. It is detailed and widely recognized as the industry standard but requires a high level of user knowledge and can be complex to use. Many user interfaces have been developed to make it easier to work with DOE-2 3 . These shells provide multiple user-interface levels and generate both simple graphical results and detailed numerical results. E-Quest is the most up-to-date user interface. IES-<Virtual Environment> 4 (IES-VE): A suite of software modules to provide complex building performance simulation. This software has been accredited by some communities and local governments, so the results may be used to show compliance as part of an application for approval under building regulations (Gudeman et al, 2008). Apache Simulator, a dynamic thermal simulation tool to 2 http://doe2.com/. 3 http://doe2.com/ 4 http://www.iesve.com/ 4 model the heat transfer process in a building, is the calculation engine of IES- <VE>. It can be linked to MacroFlo for natural ventilation and infiltration analysis, to SunCast for detailed shading and solar penetration analysis, and to ApacheHVAC for component based system simulation. Vasari 5 : Project Vasari is a building design and analysis tool developed by Autodesk that focuses on the conceptual design phase. With built-in features such as energy analysis, solar radiation analysis, and a wind tunnel, a user can create, analyze, and redesign whole building models. The conceptual model created with Vasari can be used in Autodesk Revit when the project moves into production phase (Kendra Tupper et al. 2011, 29). The energy analysis is based on Green Building Studio, an open analysis tool based on the robust data exchange of the Green Building XML (gbXML) schema, which uses DOE-2 for its simulation engine (U.S. Department of Energy, 2011). HEED 6 : A whole building residential simulation tool. It is intended for use at the very beginning of the design process, when most of the decision are made that will affect the building eventually energy performance (Milne, 2005). It shows how much money you can save by making changes to your home. It also shows how much greenhouse gas (including CO2) it accounts for and its annual total energy consumption. HEED uses the Solar-5 engine, an hourly heat balance 5 http://labs.autodesk.com/utilities/vasari/ 6 http://www.energy-design-tools.aud.ucla.edu/heed/ 5 simulation program which calculates an hourly heat balance for 8760 hours in a year using the standard ASHRAE standard (James R 2002, 1067-1076). 1.2.2. Terms relating to the tree Transmittance. It is the fraction of incident solar radiation that is transmitted by the shading surface, which is not absorbed or reflected. In this study, there are two forms of transmittance. One is the transmittance of the tree silhouette; the other is the transmissivity of the tree leaf as a material. The transmittance of the tree silhouette is the visual density or overall occlusion, and it cannot represent the net transmittance value of a tree since leaf is actually not an opaque material. Leaves are the main shading surface of a tree; they reflect, absorb, and transmit the solar radiation, and the transmissivity varies depending on light wavelength and the property of the leaf. Figure 1: Reflection, absorption and transmission of a leaf (Moffat, Schiler, 1981) 6 To avoid confusion, the following equation expresses the relationship between net transmittance, occlusion, and the transmissivity of tree leaf. Net transmittance=(1-occlusion)+(occlusion × leaf transmittance) For deciduous trees, the value of transmittance has seasonal variation due to foliage development and amount of branches and twigs. This is one of the reasons that make tree shading different from the shading of a neighborhood structure to a building. Tree canopy transmissivity is accounted for using shade coefficients when the leaf transmissivity is considered, which typically ranges from 0.5 to 0.9 for leaf-off and from 0.1 to 0.3 for leaf-on periods 7 (McPherson, 1984). However, measuring the transmittance value of a tree is extremely complicated due to type of tree, season, and the age of tree. Different instruments and methodologies are also being used in different researches. For a mature English oak (Quercus robur), the transmittance is 0.19 in summer and 0.83 in winter (Schiler, 1979). This value is the visual density or overall occlusion of the tree, calculated by optical scanning of photography developed by Schiler (Schiler, 1979). In other studies, McPherson (McPherson, 1981) reported the transmittance value for London plane tree (Platanus acerifolia) 0.11 in summer and for the same species, Heisler reported 0.46 in winter (Heisler, 1982). These values are the net transmittance of the tree, in other words, the actual solar radiation transmitted through the leaf, branch, and the gap between leaves, which is measured by 7 McPherson, E. G.. 1984. Energy-conserving site design. Washington, DC: American Society of Landscape Architects: 144-146. 7 pyranometers. There is also no consistent geometric relationship between the different measurement locations and the location or subtended angle of the tree. Unfortunately, there is no consensus on procedure (McPherson, 1984). In DOE- 2, the default value of transmittance is 0.0, which means the surface is opaque. A value greater than 0.0 represents a device that passes some solar radiation, such as fabric or a tree canopy. 8 Figure 2: Glass Transmission Properties 9 Shading Coefficient In the Dictionary of Architecture and Construction (Harris, 2006): the term shading coefficient is used to describe the property of a glazing material that is defined as the total amount of solar energy that passes through a glass relative to a ⅛ inch (3 mm) thick clear glass under the same design conditions. This includes both solar energy transmitted directly plus any absorbed solar energy subsequently re-radiated or convected into a room (see 8 http://www.doe2.com/download/doe-22/DOE22Vol2-Dictionary_46.pdf 9 http://www.pilkington.com/resources/glasstransmissionproperties.jpg 8 Figure 2); it is expressed as a number without units between 0 and 1. Lower values indicate better performance in reducing summer heat gain and therefore air-conditioning loads 10 . In the book “Energy Conserving Site Design”(McPherson, 1984), shading coefficient is also used to describe the net transmittance of a tree. Evapotranspiration . There are many definitions. In general, evapotranspiration is the combined process of evaporation (from soil and plant surface to atmosphere) and transpiration (from plant tissue to atmosphere). In other words, evaporation is the whole process of water lost to the atmosphere from ground; transpiration is final part of this process. Transpirational cooling occurs when large amounts of energy are used to change water from a liquid to a vapor at the leaf surface (McPherson, 1984). From an energy conservation point of view, a tree can be regarded as an evaporative cooler, using up to 100 gallons of water a day, which translates into a cooling potential of 230,000 kcal/day (Huang et al, 1987) This cooling effect, observed in a study published in Geiger, is the primary cause of 5 deg C differences at noontime temperatures observed between forests as compared to open terrain (Geiger et al, 2003). Solar heat gain. This is the increase in temperature in a space, object, or structure that results from solar radiation. Dense shade reduces solar heat gain on windows or opaque wall surfaces of a shaded building, thereby reducing the surface temperature of the shaded building. 10 Harris, Cyril. 2006. Shading coefficient. In Dictionary of Architecture and Construction. 4th ed.,Revised ed. New York, NY, USA: McGraw-Hill. 9 Solar radiation. Solar radiation can be divided into direct and diffuse radiation. Direct solar radiation is sun-related light; diffuse radiation is sky-related light. Direct light consists of rays of light emanating from the sun. This directional and intense light creates shadows and burning rays, in other words, direct light is determined by the location of the sun. Diffuse light occurs naturally as the sun’s rays are scattered by the atmosphere, greenhouse gasses, or cloud cover. Diffused light seems to wrap around objects because it is not directional. It comes from all directions. The sun’s direct rays turn into diffused light as they pass through translucent material. The intensity of direct light is often ten times more than the intensity of diffuse light. Shades and building shades. “Shades” or “window-shades” are devices such as drapes, blinds, pull-down shades, etc., that are used for sun or glare control. They are distinguished from “building-shades” such as fins, overhangs, neighboring buildings and trees (Winkelmann, Frederick C. 1985). In e-Quest, a shade is defined as a device for sun control, attached shade; and external shades. In contrast, a building shade is defined as a shading surface outside the building that is not attached to the building. Shadow and shade. Shadow is the silhouette created by an object when light can't travel through it; a shadow is cast on the ground or another surface. Shade is the comparative darkness, the shape does not matter. 10 Urban forest. This term refers to all trees, both public and private, that are found growing in cities, towns, and communities. It plays an important role in ecology of human habitats in many ways. Urban trees provide benefits that are fundamental to our city's livability. Qualities such as clean air and water, cooler streets and homes, beauty, and wildlife habitat are essential elements to the health and comfort of any city. It is the term commonly used in the field of landscape and forestry. It describes an important ecosystem in a city, a neighborhood and a region as a whole, buildings are just one of the attributes of it. 1.2.3. Climate zones Three climate zones in California (as defined by Title 24) were chosen for the study as representative of climatic variation: 6, 12, and 16. Climate Zone 6 (Los Angeles) includes the beaches to the southern Californian foothills; it is hot during the summer, but has a very mild winter, which is when most of the rainfall occurs. Climate Zone 12 (Sacramento) has cool and damp winters with winter rains from November to April with a chance of frost on winter nights in lower areas; the summers are hot and dry. Climate Zone 16 (Mt. Shasta region – 5,000 feet elevation)) is located in a mountainous, semiarid. The winters are cold and snowy; the summers are warm and clear. 11 1.2.4. Terms for the different types of trees used in this thesis A range of digital tree configurations was established that would be potentially used for testing. The following parameters were chosen: geometry, opacity, and animation. Geometry / silhouette refers to the shape or the form of the tree; the character varies among species as much as leaf shapes or bark patterns. The tree silhouette models in this study include two dimensional (silhouette - plane, planes, lollipop), simplistic three dimensional (planes), and three dimensional (box, cone, pyramid, sphere, many planes). These shapes can correspond to real trees. For example, a pyramidal tree is similar to many evergreen trees such as white fir and bald cypress. A columnar tree has a very narrow and upright shape, which usually has just one trunk, e.g. lombardy poplar, leyland cypress. A lollipop tree has a round crown, for example, red oak or white ash. 12 Table 1: 2D and 3D tree shapes Opacity / foliage pattern refers to the transparency of the tree using transmittance values. These values will be based on the study done in 1981 13 (Moffat and Schiler). There are two ways to simulate the pattern of foliage. The first method is the combination of many opaque and transparent planes; the second method is to give a transmittance value range from 0 to 1 to one plane. The second method is used in this study. Figure 2: Two methods to simulate foliage pattern. Animation / changes over time: scheduling of opacity and making separate runs for summer and winter with different tree densities. This category includes changes during the day (for example, by wind), by season (especially important 14 for deciduous trees that lose their leaves in the winter), and by year (the growth of a tree over time). In this study, the seasonal changes are simulated. 15 Table 2: Trees change over time (Credit: Mark Boster / Los Angeles Times) 11 Days Stand tree Fallen tree Seasons Summer leaf-on tree Winter leaf-off tree Years Young tree Mature tree (over 20 years) 11 http://latimesblogs.latimes.com/lanow/2011/12/pasadena-neighborhood-trapped.html 16 1.3. Tree benefits on energy saving 1.3.1. Tree shading Shade cast by canopy reduces conversion of radiant energy to sensible heat, therefore reducing the surface temperatures of shaded objects. 1.3.2. Evapotranspiration Evapotranspiration at the leaf surface results in cooling the leaf and adjacent air due to the exchange of latent heat (McPherson, 1984). Latent heat is the energy released or absorbed when the phase transitions happen without changes in temperature. 970 Btu of energy will be absorbed when 1lb of water vaporizes. A single mature, properly watered tree with a crown of 30 feet can "evapotranspire" up to 40 gallons (approximately 330 lb) of water in a day (LBL, 2011) which is like removing 320,000 Btu from the ambient air. This is almost the heat a room air-conditioner (12,000 Btu/hr) removes in 80 hours 12 . 1.3.3. Others There are other benefits associated with trees in energy saving. Trees act as windbreaks that lower the ambient wind speed, which may lower or raise a building’s cooling energy use (Akbari, 2002). Well-placed windbreak trees can reduce winter heating cost. 12 http://www.energystar.gov/ia/partners/manuf_res/downloads/2007RoomAC_prg.pdf?6f30-096b 17 1.4. Background on tree shading 1.4.1. Energy conserving tree shading design Trees help people create a nicer environment in which to live. In addition to helping make landscapes visually pleasing, they make our environment cleaner and more comfortable. Trees provide shade in the summer. Correctly placed shade trees, windbreaks, and foundation plantings can reduce heating and cooling costs by an estimated 25-30%( Akbari et al., 1997), with some estimates as high as 50% (Parker, 1981). The following discussion addresses various general tree shading approaches to save the energy use of the buildings. The closer a tree is located to a building, the higher the energy saving will be realized in certain climate zones. For example, trees located to shade east and west elevations will provide more daily shade for a longer period if they are close to the wall. Simpson and McPherson (Simpson and McPherson, 1998) used simulation modeling to evaluate the effect of trees on residential energy use in California. They found that trees shading the west side of the house have the biggest effect on cooling costs that reduces cooling cost by 10-50%. Window shade provided by arching tree canopy is preferred when shading windows. The canopy can effectively block solar radiation while still permitting a view to outside, even when the trees are in full leaf. For best results in locations with warm to hot summers, deciduous trees should be located to shade east- and west- facing windows. This will be demonstrated in chapter 3. 18 1.4.2. Tree shading simulation The relationship between tree shading and summertime home energy consumption has long been a topic of interest and has been the subject of many simulations and small-scale studies. Previous studies that have looked at the effect of the shade trees on energy use fall into two categories: (1) controlled experiments that examine the effects of trees on individual buildings and (2) large scale simulation modeling with some specific software. Akbari et al. (1997) quantified the effect of shade trees on the cooling costs of two similar houses in Sacramento, California. The experiments were carried out by a tree planting program of Sacramento Municipal Utility District (SMUD) and the Sacramento Tree foundation. Sixteen trees in pots (eight were 6 m tall, and eight were 2.4 m tall) were placed along the exterior south and west walls. Inside, occupants kept windows closed, thermostats at the same temperature, and used lighting in a similar manner. Results showed that the trees reduced seasonal cooling costs by between 26% and 47%. Finally, Akbari et al. modeled the effect of the trees on both houses using the DOE-2 simulation program. They concluded that the computer model underestimated the energy savings of the trees by as much as twofold. Akbari and Taha (1992) used simulation modeling to study the effect of trees on energy use in four Canadian cities. They used DOE-2.1 building analysis program for the energy simulations and used a weather processor that developed through heat island research at the Lawrence Berkeley Laboratory (Taha, 1990). The shadow cast by trees were simulated based on the assumptions that vegetation cover uniformly distributed on all orientations and the trees let in 19 70% of the sunlight in winter but only 10% in summer; the evapotranspiration was simulated based on the assumption that trees transpire only above ambient air temperatures of 10°C, and that evapotranspiration is insignificant between October and April. They concluded that increasing the vegetative cover of a neighborhood by 30% and increasing the albedo of houses by 20% would decrease heating costs by 10–20% and decrease cooling costs 30–100%. Simpson and McPherson (1996) used simulation modeling to evaluate the effect of trees on residential energy use in California. They found that trees shading the west side of houses had the biggest effect on cooling costs and that adding three shade trees to a house (two on the west side and one on the east side) reduced annual cooling costs by 10–50%. Other research done by the Lawrence Berkeley National Laboratory, heat island group, used simulation modeling to estimate the potential of urban trees and high albedo surfaces on offsetting the heat-island effect. Results suggest that existing trees and high-albedo surfaces can potentially reduce 20% of national energy use in air-conditioning. (Akbari et al. 2001). McPherson and Simpson estimate a planting program in California could reduce peak energy load by 10%(McPherson and Simpson, 2003). Donovan et al. estimated the effect of shade trees on the summertime electricity use of 460 single-family homes in Sacramento, California. Results show that trees on the west and south sides of a house reduce summertime electricity use, whereas trees on the north side of a house increase summertime electricity use (Donovan et al., 2009). 20 1.4.3. A method by DOE-2 The shading effects of trees can be simulated on the DOE-2 program as exterior building shade (Huang et al., 1987). Tree transmissivity values are based on previous research done by Thayer and McPherson (Thayer et al., 1985; McPherson, 1984). They have modeled the shading of typical canopy with an average height of 10m by a building shade uniformly distributed around a house at the height of 6.5m, see figure 1. This method made the assumption that the percent increases in the canopy density equivalent to the reduction in the transmissivity of the building shade. Diffuse light reflected from the sky was modeled by modifying the inputs for sky- and ground-form-factors and the ground reflectance of the surroundings (Akbari et al., 1987). Figure 3: Generalized tree canopy shading as modeled in DOE-2 (Akbari et al., 1987) 21 Figure 4: (Upper) the model for one tree on the south side of the house; (Lower) the model for three trees on the south and west side of the house (Huang et al., 1987) 22 1.4.4. A method by eQuest Another study (Saxena et al., 2011) used the building shade in eQuest to simulate tree shading by developing a set of obstruction models that were applied to the template to imitate either light or heavy shading. Exterior shading was categorized as none, light, or heavy shading in the table. A short modeling exercise was undertaken to determine an equivalent exterior shade to a light and heavy urban shading. In the process, an autonomy daylighting plot was created in a dense urban area; in Figure 5 the associated daylight plot is a metric for sufficient daylighting in a space. Then they found that among several shading models, a shading model shown in Figure 6 consisting of a vertical fin from the middle of the façade going outward with ceiling height and a parallel wall with the length of the façade had a daylighting plot that matched closely to the urban shading plot. Thus, a dense shading template is created. A light shading template was developed in the same method. Figure 5: Autonomy daylight plot for dense urban area (Saxena et al., 2011) 13 13 Mudit Saxena, Timothy Perry, Charlotte Bonneville, Lisa Heschong,Office Daylighting Potential Research Program,PIER report.2011 23 Figure 6: Shading model (Saxena et al., 2011) 24 Two key findings from Saxena’s paper (Saxena et al., 2011) are useful for this thesis. First, based on the site survey, about 60% of office buildings in California have some level of obstruction from trees and simulation models are typically modeled without exterior obstruction (Saxena et al., 2011). The building energy simulation practice often ignores the influence of exterior obstruction such as trees and other buildings. Second, the standard urban shading models made by Saxena were developed by comparing the daylight autonomy plots of urban area. The author thinks the shading models developed by Saxena still can not be a good representative model for trees for the reason that the lower angle sunlight will be blocked and the higher angle sun will get through. This exercise was demonstrated by matching the autonomy plots. This methodology has given author the idea for finding the representative model for a single tree. In this thesis, to develop a tree model, different building shade models in eQuest are made to compare with the incident solar radiation (Section 3.3) and to match the shadow pattern of a tree. (Section 4.1). 1.4.5. Previous developed tools Several early tools were developed to test the impacts of trees on building energy. The following are two methods that share the characteristic of using a graphic program and an energy analysis program to simulate tree-shading effect to the microclimate of the building. At Cornell University, an interactive computer graphics analysis system (GLAS) was developed to simplify and accelerate the input task for thermal analysis program. The visual input programs were used to define tree foliage in terms of 25 their silhouette or cross-section size, form and density, and shading values. A site plan could be drawn, and trees could be positioned with reference to the test building. It calculated the shading effects given the tree position, size, shape, and density. Then it calculated hour by hour energy requirements for weather condition, building materials, and other input in the GLAS program using a second program based on the National Bureau of Standards NBSLD program . (Schiler, 1979) Similar to GLAS, another method also applies more than one program to simulate the tree shading effects. The relative effects of various landscape tree configurations on space conditioning use were determined from computer simulation. Shading of buildings by trees was determined using the Shadow Pattern Simulator (SPS) program. SPS calculates hourly shading coefficient for each building surface including opaque and glazed area based on building and tree sizes and their relative orientations and distance from buildings. MICROPAS ver.4.01 14 uses building thermal characteristics and information related to occupant behavior (described subsequently) to provide hourly estimates of building energy use after the weather processor was modified to accept the solar coefficient simulated by SPS. Energy savings are determined by comparing predictions for identical unshaded (base case) and shaded buildings. (Simpson and McPherson, 1998). 14 http://micropas.com/ 26 1.4.6. Comparison The GLAS program (Schiler, 1979) method modeling the occlusion of direct radiation between sun, tree and window, the shading of the opaque wall and the effect of diffuse radiation were not included; the SPS program can calculate the shading on both opaque (wall) and glazed area. The preprocessor of GLAS method can identify the actual incremental shade when trees overlap because of the scanline algorithm. However because the SPS program method did not take tree shade overlapping into account the effects on energy use were considered underestimated (Simpson and McPherson, 1998). Both of the shading simulators are not being used in the energy simulation computer tools listed on the chart (Table 5) because they were research tools and might have been phased out because there is no related studies published in the recent 10 years, however, it is worthy to examine these tools developed before and to implement the current tools. Previous studies (Schiler, 1979) have shown the shading by the tree on the wall was a minimal part of the savings compared with shading on the windows; a related test has done in this thesis (Section 3.4.2) Table 3: Comparisons between GLAS and SPS programs GLAS (Schiler, 1979) SPS program (McPherson et al. 1985) Direct light √ √ Diffuse light x x Shading on wall x √ Shading on glazed area √ √ Tree leaf overlap √ x Current statue Phased-out (research tool) Phased-out (research tool) 27 1.5. Background on tree evapotranspiration 1.5.1. Measurement of tree transpiration Tree transpiration is pretty much an invisible process. Since the water is evaporating from the leaf surfaces, we cannot just go out and see the leaves "breathing." There are two methods being used to measure the amount of water vapor that the tree expels. The first quantitative measurements of transpiration were those made by Stephen Hales, prior to 1727 (Kramer & Kozlowski, 1960). He measured the rate of water loss by weighting the container at regular intervals from potted plants. This method is similar to those used today that are called “Gravimetric Methods;” investigators grow plants in a pot of soil, the soil mass has to be enclosed in the container from which water cannot evaporate. Then, by weighing the pot, the water loss can be measure. (Kramer & Kozlowski, 1960). Another long-used method of measuring water loss is to measure the volumes of water absorbed by a cut branch. The cut branch is immersed in a small tube of water; the changes in volume can be observed in the tube when the transpiration happens. This method is questionable because the transpiration value from a detach or attached branch may be not the same. 1.5.2. Previous developed digital tools Unfortunately, there was no computer tool developed to simulate the effects of evapotranspiration on energy saving in buildings. 28 1.6. Simulation tools for landscape architects Modeling environmental benefits provided by trees and other vegetation is an ongoing topic of research for landscape architects. The followings are simulation tools used by some landscape architects and urban planners. 1.6.1. CITYgreen CITYgreen is a GIS-based software developed by American Forests (2002). It is used to quantify the benefits that trees provide by allowing users to analyze the following: stormwater, summer energy savings, carbon storage and sequestration, and others. While cooling benefits of trees near larger structures are certainly realized, current research does not provide guidance for calculating these benefits (Jones and Stokes Associates, 1998). “Despite its frequent use, the inability to calculate energy savings for structures larger than one or two stories poses a substantial impediment to the analysis of densely populated urban neighborhoods and little research had been conducted to validate results of such tools. This is not a failing of CITYgreen, but represents the current state of research.” (Longcore et al, 2004) 1.6.2. TownScope Developed by University of Liege, Belgium, TownScope (see Figure 7) is a tool for sustainable urban design that provides the assessment of the impact of new developments on urban microclimate, landscape and energy. Thermal and visual 29 comfort, wind patterns, solar availability of urban open spaces can be assessed very quickly via TownScope. 15 Later in the study, an urban planning program was found that did contain a component for creating trees, TownScope. The input dialogue box showed a similar set of parameters on a tree that were used in eQuest, but slightly more sophisticated and easier to use: default volume opacity, trunk height, trunk width, canopy height, and canopy width (see Figure 8). Figure 7: TownScope example screen 15 http://www.townscope.com/index.php?page=home&lang=EN&theme=default 30 Figure 8: Tree element screen in TownScope 31 1.6.3. UFORE UFORE is an acronym for "Urban Forest Effects." This computer model (a.k.a. i- Tree ECO) was developed in the late 1990s by researchers at the United States Department of Agriculture (USDA) Forest Service to help managers and researchers quantify urban forest structure and its functions. The model calculates numerous attributes about the urban forest. Information on tree sizes, types, and distance and direction from two-story buildings are used to estimate tree effects on building energy use. The UFORE model uses published methods, based on a report by McPherson and Simpson (1999) to estimate existing tree effects in summer and winter space conditioning energy use 16 . 1.6.4. iTree iTREE is the packaging from the USDA Forest Service around various software tools designed to help communities better understand their urban forest and estimate its value. The package includes i-Tree Eco, i-Tree Streets, i-Tree Hydro (beta), i-Tree Vue, i-Tree Design (beta) and i-Tree Canopy. I-Tree Eco is the interface for UFORE. Figure 7 and 8 show the sample input screens of i-Tree Streets. i-Tree Streets focuses on the benefits provided by a municipality’s street trees. It makes use of a sample or complete inventory to quantify and put a dollar value on the street trees’ annual environmental benefits related to energy conservation, storm water control, carbon dioxide reduction and air quality 16 http://www.ufore.org/about/01-00.html 32 improvement. Because i-Tree Streets is the component that calculates the energy savings, further assessment has done in Section 5.2. Figure 9: iTree Street sample input screen Figure 10: i-Tree Street sample tree data input screen 33 1.6.5. Tree Benefit Estimator The Estimator was developed by Sacramento Municipal Utility District (SMUD) 17 based on SMUD’s Shade Tree Program and as part of American Public Power Association’s (APPA) Tree Power Program. The Shade Tree Program was initiated in 1991 with the objective of reducing summer air conditioning loads by planting trees to shade residential buildings. A secondary target of the program was to create an urban forest that would help mitigate heat island effect. (Hildebrandt and Sarkovich,1998) The program provided trees to homeowners, and the heating and cooling load impacts were monitored. By the end of 1994, 170,000 trees have been planted 18 . The tree benefit Estimator is a web-based program for estimating the tree planting benefits. Tree’s direct shading impact and indirect evapotranspiration impact on energy saving could be estimated by the simple input on the user- friendly interface (see Figure 11). The Estimator takes several parameters into account: 1) The average cost of electricity in summer and winter. 2) Tree species: 127 tree species are listed. 3) Tree age: All trees are grouped into three categories, small (up to 25ft in height) medium (up to 45ft in height), and large (at least 46ft in height). 4) Number of trees 17 https://www.smud.org/en/residential/environment/shade-trees/ 18 http://actrees.org/site/resources/research/shade_tree_program_impact_evaluation.php 34 5) Climate area 6) Tree orientation 7) Distance from the house The Estimator, based on i-Tree’s street tree assessment tool called STRATUM, calculated the energy saved in kWh, capacity kW saved, and the CO2 sequestration (lbs) resulting from mature trees planted (see Figure 12). Further evaluation on this program is in Chapter 5. Figure 11 Tree Benefits Estimator input window 35 Figure 12 Tree Benefits Estimator sample report 36 1.7. Conclusion We needed to do the assessment on current tools and methods for these tools and figure out the capabilities and obstacles for these tools to simulate the shading and evapotranspiration effects on the building. A list of the tools and methods discussed in this chapter are summarized in Table 4. Table 4: List of simulation methods discussed in Chapter 1 Type Tool Developer Description DOE-2 Akbari, 1987 Box building shades represent trees. Energy simulation programs eQuest Saxena, 2011 Standard shading models made of fins and building shades. GLAS Schiler, 1979 Require at least 2 processors. Research tools from 70’s to 80’s SPS McPherson et al., 1985 Require at least 2 processors. CITYgreen USDA Forest Service, 2002 TownScope University of Liege, Belgium, 2001 UFORE USDA Forest Service, 1990s iTree USDA Forest Service, 2006 Landscape tools Tree Benefit Estimator Sacramento Municipal Utility District (SMUD), 1991 These three tools are basically the same program. iTree is the current interface for UFORE; Estimator is a program developed base on iTree Streets. In the next chapter, more is discussed about common programs for architects and building engineers: eQuest (DOE-2), Vasari using Green Building Studio (DOE-2), HEED (Solar 5), IES <VE> (ApacheSim), and Design Builder (Energy Plus). First a specific “tree command” was looked for and then other methods that might be available in the software engine itself to simulate a tree. 37 2. CHAPTER 2 2.1. Energy simulation software Once a list of potential energy saving strategies is developed, an evaluation process must be developed to judge the feasibility of implementing each option. Computer simulation allows numerous options to be tested; however, input parameters and the model itself must be accurate for the results to provide a valid approximation of reality. There are several hundred building energy software tools available for performing building energy simulation. EERE tools directory lists 370 tools to perform energy simulation and related activities like insulation performance, lighting and daylighting calculations, sunpath plotting, etc. Although many of the programs are comprehensive, most of the available programs do not have the capacity for including specifically shadows trees in the energy analysis, only including the shading devices such as overhang or other window shading devices, which are attached to the building. None of the existing tools studied so far includes the effects of natural shading from foliage. In the studies done by Akbari ( Akbari, 2002) the evapotranspiration effect was simulated in existing energy simulating tools by reducing the ambient temperature of the building, not in a direct method of calculating water loss from a tree. Limitations almost apply to every available tools of this kind today; thus it is necessary to understand certain basic principles of energy simulation. Most of the programs consist of a calculation engine and the graphical user interface (Maile et al., 2007). The simulation engine is usually developed by research centers or academic institutions; the user interfaces are often implemented by 38 private software venders (Maile et al., 2007). An engine can usually be used by several user interfaces, however, most do not use the complete functionality of the related engine. The graphical user interface usually covers the whole process from input to output and enables simplified input files, starts the simulation run in the engine, and produces the output files in a graphical manner for a user to read. The user interfaces differ in their purposes; in general, interfaces can be categorized into three subjects, “whole building analysis”, for example HEED, “codes and standards,” for example EnergyPro, 19 and tools for other applications, for instance, Ecotect 20 for lighting systems, ApacheHVAC 21 for HVAC systems or EnergyGauge 22 for economic analysis (U.S. DOE 2007). 2.1.1. Simulation Engine Different simulation methods are being used within the engine; the results are usually accurate within the domain of the software if the input is correct. In figure 13, the input includes building geometry, weather data, HVAC systems, internal loads and other components. Every simulation is based on certain equations and principles (Maile et al., 2007). The prediction of energy values of a simulation is rarely completely accurate because the calculation in the engines is usually based on many assumptions and sometimes simplifications -- for example a measured "typical" 20-year average weather files are often used instead of calculations from basic principles. Users need to be aware of these 19 http://www.energysoft.com/ 20 www.autodesk.com/Ecotect-Analysis 21 http://www.iesve.com/software/ve-pro/analysis-tools/hvac/apachehvac 22 http://www.energygauge.com/ 39 assumptions to make sure the result of the specific simulation is more reasonable, for example, the tree simulation in this study. Figure 13: General data flow of simulation engines 23 23 CIFE Working Paper #WP107, Stanford University, 2007 40 DOE-2 and EnergyPlus are the most widespread used engines both developed by Lawrence Berkeley National Laboratory. Proven through 30 years of development, DOE-2 is recognized as the industry standard. EnergyPlus is the successor of DOE-2 that incorporates the features of thermodynamic analysis of the BLAST system. In table 5, a chart is developed to compare the capabilities of several simulation engines; some of the descriptions in this chart imply the underlying difficulties for tree simulations. One of the limitations is that DOE-2 offers the possibilities for users to add case specific functions, but the process requires a great effort and understanding of DOE-2 language, only experienced user can use these functions and increase the accuracy of the result. It is difficult to have a clear language to compare the features of these tools in the table 5, and only general information and capability features that may relate to the use of digital tree simulation are provided in this chart. 41 Table 5: Simulation engines comparison chart DOE-2 Energy Plus Solar 5 (Milne et al, 1996) ApacheSim 2425 Developer James J. Hirsch & Associates (JJH) and LBNL LBNL,U.S. Army Construction Engineering Research Lab, OSU,DOE(Crawley et al. 2002) UCLA IES Ltd Cost Free Free Free Not free Description Most widely used simulation engine today. Heat load equipment loads, people loads, lighting load can be modeled and simulated Energy simulation, load calculation, building performance, simulation, energy performance, heat balance, mass balance. Calculate hourly energy performance for the whole, heat flow and indoor air temperature, daylighting, HVAC system, cost of electricity and heating fuel, and the corresponding amount of air pollution. ApacheSim is at the core of the IES suite of thermal analysis products. User Interface eQuest, RIUSKA Design Builder HEED Ies<virtual environment> Geometry Simplified geometry. Cannot exchange with CAD software. Can be created by using Green Building Studio. Simplified geometry Simplified geometry Through a CAD geometrical representation of the building in the IDM (integrated data model) Load Calculation Method Weighting factor method Thermodynamic equations Thermodynamic equations First-principles models of heat transfer processes Limitations Adding more component is cumbersome, errors in code is hard to resolve Missing graphical user interface that provides all EnergyPlus capabilities. Not intended for complex mechanical system design or equipment sizing Calculate diffuse radiation from global radiation Yes Yes Yes Yes User specific additions User definable functions limited code entry points Link to SPARK(user definable SPARK components) Flexible! No Yes 24 http://apps1.eere.energy.gov/buildings/tools_directory/software.cfm/ID=482/pagename=alpha_list 25 http://simcosm-india.com/pdf/ApacheSim.pdf 42 Table 5: Continued Input and output Text file Text file Input: text file Output: highly graphical 3D plots. Input:pre-built databases Output:The results are displayed by the <Virtual Environment> program Vista, shown in tabular and graphical form User-define coefficient Yes Yes Yes Yes Example text input BUILDING- LOCATION LATITUDE=41.98 LONGITUDE=87.90 ALTITUDE=673 TIME-ZONE=6 LOCATION Chicago Illinois USA, 41.98,87.90,205,-6 N/A (Program downloads no longer available after 2005) N/A 43 2.1.2. User interface A graphical user interface makes it easier to use the software engine. In table 6, several user interfaces are compared. eQuest is the most up-to-date user interface for DOE-2, while Design Builder is the most comprehensive user interface for EnergyPlus. HEED was the only tool of these that has a specific feature for trees, see Figure 14. However HEED cannot calculate the impact of tree shade or neighboring object. The input components of tree and neighbor are there for the future improvement. According to the program developer, the newer version (expected in 2012) will have to ability to calculate exterior shading from trees and neighbors. Besides the general information of each tool in table 2, the comparison items in table 2 imply some of the factors for developing potential methods for tree simulation in the future. Figure 14: Tree input in HEED 44 Table 6: User interface comparison chart eQuest Vasari HEED IES<VE> Design Builder Green Building Studio Developer Department of Energy Autodesk UCLA Glasgow-based Integrated Environmental Solutions (IES) Ltd. DesignBuilder Software Ltd GBS acquired by Autodesk in 2008 Cost Free Free Free VE Ware is free,VE-Pro,VE- Toolkit,VE-Gaia are not free not free Web-based Service, not free Engine DOE-2 GBL(DOE-2) Solar5 ApacheSim Energy Plus DOE-2 Weather file/ Editing BIN/Yes Autodesk Web Service/No TMY2 or EPW/Yes EPW/Yes EPW/Yes Feature for tree No No Yes No No No Obstruction modeling No Yes Yes No Yes No Estimate diffuse radiation from global radiation No No No Yes Yes Yes Schedule Obstruction Yes No No No No No Shading surface transmittance Yes No No Yes Yes No 2.1.3. Interoperability The calculation engine itself does not provide any data import or exchange with other tools (Maile et al., 2007). However, the user interfaces that have been developed sometimes provide the data exchange functionalities. For example, Vasari and eQuest are two different user interfaces that both use DOE-2 as calculation engines and allow transfer of data. 2.2. Limitations 2.2.1. Tree model Trees and neighboring buildings are common exterior obstructions that block direct solar radiation and effect both energy and daylight calculations (Saxena et al., 2011). It is common, however, to not include trees in energy simulations. Trees as one of the neighboring obstructions of the building that cannot be 45 modeled in most of the software. Thus, to create a tree outside the building is the first obstacle for this study. In 1.4, the previous studies did not actually model the tree, some indirect method were used to simulate trees, for example, creating building shade in DOE-2 program. One reason that energy modelers may not choose to include trees in their studies is that trees are not permanent; they might be cut down in the future. Instead the building is studied in isolation, it might be better to do simulations with both the trees present and absent. 2.2.2. Calculation on shading effect Some of the software engines can calculate the shading effect from neighboring objects, but none of the related user-interfaces have the feature to show specifically the shading effect from a tree from both the tree trunk and canopy. Canopy density can be directly simulated in terms of transmissivity but the transparency of leaf has not yet being simulated. In eQuest, only direct sunlight is calculated but the diffuse and reflected light is not. 2.2.3. Transpiration rate In most of the studies done before, the evapotranspiration effect is modeled by adjusting the ambient temperature of the model based on many assumptions. If a tree can function as an evaporative cooler, how to model that influence is important, but rarely is a feature in the software programs. 46 2.3. Methodology Figure 15: Research methodology This study begins with a preliminary research to investigate previous methods and tools developed. Current energy simulation programs will be compared and evaluated as well. Then different strategies will be undertaken to model trees in energy simulation programs. The capability of the proposed method will be examined with series of side tests. A simple case will be simulated with variables such as climate zones, orientation and quantity of trees. The final step is validation, to compare with historical data from measurement and computer simulation if available. 47 3. CHAPTER 3 3.1. eQuest and Vasari eQuest is used for the initial studies because it is the most up-to-date user- interface for DOE-2, and DOE-2 is the trustworthy energy simulation engine. Other software programs may also be studied depending on the results of the initial studies. Project Vasari by Autodesk uses web-based Green Building Studio software for simulation, and DOE-2 is the engine for Green Building Studio, therefore Vasari and eQuest both use the same engine, DOE-2. Although the results between Green Building Studio and eQuest could be studied, the primary use of Vasari in this study was to help generate tree forms that could be used in eQuest. 3.1.1. Data exchange between eQuest and Vasari Limitations in the user interface of eQuest make it difficult to create complex shapes. Figure 16 shows some of the possible tree models created directly in eQuest interface. The author assumed some of the limitations in eQuest, for example, incapable of creating complex geometry, can be solved by the data exchange function in eQuest -- the trees would be created in Vasari instead which is easier to use. 48 Figure 16: Two possible building shade tree models created directly in eQuest(left: box-trees, right: cross- trees) 49 3.1.2. INP file eQuest uses an INP file that is its text input file to DOE-2. In eQuest, an .inp file is created automatically in the project folder when a project is created; it can identify the data used to input into eQuest. The user can open the .inp file with Microsoft Word or Notepad or any other software that reads ASCII text. Figure 17 shows an example part of an INP file that represents a building shade named A, the description that will be used to create the tree forms. The example shown creates a horizontal rectangular shading surface (15ft in height, 20ft in width) near the building. It is located 25 ft south of the building and is 10 ft above the ground. X, Y, and Z give the position of the lower left hand corner when looking into the surface outward normal in the building coordinate system. Figure 18 shows the correspondent model in eQuest. Figure 17: Part of the DOE-2 input text file 50 Figure 18: Example .INP file eQuest model 51 Vasari can export INP files. After results are calculated (just run a test simulation; the results are not important, but getting access to the export command is), go to the result and compare tab, link to the web-based Green Building Studio to run the analysis, and then click on export button to save the result as DOE-2 file (. inp). 3.1.3. Bringing the tree from Vasari to eQuest In eQuest, a model can be generated by an interface wizard and “detail data edit” modes. Wizards are used to simplify data input through many default parameters; detail data edit mode can define all the parameters and changes according the definitions in DOE-2 engine (Maile et al., 2007). However, a building shade (used to simulate a tree) can only be created in detail data edit mode. In Vasari interface, there is no building shade component. However, the user can simply place a massing outside the building. In the .INP file exported from Vasari, these neighboring masses are interpreted as building shades. 3.1.4. Workflow First, create a project in eQuest with the simple test building (see 3.4.1); then create the neighboring model (tree) in Vasari. Second, open both .inp files, the one created by eQuest and the one created in Vasari, in text format. Copy and paste the Vasari building shade description to the .inp file of eQuest and save the new .inp file that is a combination of the eQuest building file and the Vasari tree 52 description. Initially, the new .inp file could not be opened in eQuest. It was discovered that a folder had to pre-exist of the correct name for the import to work. This workflow can ease the creation of complex building shades (and building forms) in eQuest. (Figure 19.) Note that it is also possible to simply type the description of the shading surface into the .inp file, using Word or Notepad or similar ascii editor. The rules for .inp and .bdl files are published by the Department of Energy (DOE) and can be used to create ascii input files, directly. Care must be exercised, however, not to overwrite the modified portions of the files by re-opening the files in eQuest. Figure 19: Merging two .INP files 53 3.2. Problems Immediately issues came up. eQuest does not model a tree – it allows the user only to model rectangular planes that are obstructions; these can be used to create neighboring buildings, or in this case, a very simple tree. 3.2.1. Building shade The building shade component in eQuest is the way to make a neighboring shading object that is not attached to the buildings, for example, adjacent buildings or trees. The user can specify the position, size, orientation and transmittance of a surface that casts shadows on the exterior of the building. Both the direct (from the sun itself) and diffuse (from the sky and ground) components of solar radiation are shaded by building-shades. The program does not account for reflection of solar radiation from building-shades onto the building, however the user can specify the visible radiation by editing the solar properties component under building shade. In the building shade component, there is another command option that is called “FIXED-SHADE”. This command has the same setup as building shade, however it is used only for stationary shading surfaces that are fixed with respect to the earth 26 , the shading surface does not rotate with the building if the user applies a building rotation. In this study, the building shade command is being used in eQuest to create trees. 26 http://doe2.com/download/DOE-22/DOE22Vol2-Dictionary_47b.pdf 54 3.2.2. Open .INP file in eQuest It was hoped that building information modeling (BIM) software would be able to provide some help in the creation of the tree. Several “trees” were made in Vasari, ran in Green Building Studio (a necessary step to get to the Export command), and then exported as an inp file. As mentioned before, initially, the inp file could not be opened in eQuest. It was discovered that a folder had to pre- exist of the correct name for the import to work. 3.2.3. Problem of building shade .INP file Fixing that, a second, more serious problem was evident. It appears that aside from shapes that are only created from bounded, rectangular planes, the translation in Vasari to an inp file does not work. This means that the 2d plane, 2d planes, 3d planes, and box tree were translated properly. However, cylinders, spheres, hexagons, etc. were not. The results are shown in Table 7. Correspondence was started with consultants at Autodesk to help resolve this issue, but the problem was never resolved. 55 Table 7: Import models from Vasari to eQuest Vasari eQuest Tube Hexagonal column Cone Sphere 56 3.2.4. Summary The problem of building shade (section 3.2.1) is subtler and has yet to be resolved; the problem of importing .INP file in eQuest (section 3.2.2) has been solved. To open an .INP file in eQuest, a new folder with the same file name as the .INP file has to pre-exist in the eQuest projects folder. The problem in section 3.2.3 is the most critical issue and was never solved, because of this, the idea of bringing a tree model from Vasari to eQuest failed. Some shapes could be imported, but the most interesting shapes could not. This conclusion is based solely on desirable geometries and is independent of how the calculations would actually model a tree. 3.3. Surface normal Initial energy runs were made with the building, and results compared between different shapes and rotations of the initial set of trees. Due to some unexpected results (section 3.2), two more side tests were conducted: one a quick check for the surface normal of a building shade, the other a more in depth study of the shape to use for the tree. In computer graphics, each surface has a surface normal, by definition perpendicular to the surface that sets whether or not one can “see” the surface (figure 20). In many rendering programs as a measure to save rendering time, surfaces are one-sided, and the direction they are placed is important. In some programs, all surfaces have “two sides,” and it does not matter which direction 57 they are placed. To check that the surface is “two-sided” in eQuest, a building shade was created, and the energy use calculated. Then, if the surface is rotated 180 degree and the energy output remains the same (Figure 21), it shows the building shade in eQuest is two sided. This turned out to be the case. Figure 20: One-side surface normal (left) and twp-side surface normal (right) Figure 21: Surface normal test result 3.4. Deciding on a shape for the tree 3.4.1. Decide a shape by calculating solar radiation in Vasari The major assumption for this part of the study was that that a spherical tree is a good enough representation for many types of trees. Only opaque trees were used at this stage. Given the known problem that one can only translate 58 rectangular planes in the inp file, the author used Vasari solar radiation tool to help determine which tree model to use by comparing shadow patterns and average insolation results. First, an R=10’ sphere was created, located on the south of a 50’x50’x30’ box at center (Figure 22). The location was set at Los Angeles, CA, USA. A one-year solar radiation analysis from 5:00AM to 7:00PM with one hour interval on the south face of the rectangular face was run. Figure 22: Test model in Vasari 59 Table 8: Average incident solar radiation on a vertical surface from different tree models (A). Table 9: Average incident solar radiation on a vertical surface from different tree models (B) Figure 23: Comparing the insolation value of different tree models 60 A box tree is created by 6 planes, in eQuest, the sunlight will pass through more than one surface, thus the surface transmittance value is not the actual value for some directions of light penetration. The same case applies to 3 boxes and 5 boxes stacked tree although the volume of these trees are most close to the sphere tree. To solve this issue, another tree is created which is an open box with only 4 planes. In this case, the sunlight will pass the surface only one time (see figure 24), but it depends on the orientation of planes. Figure 24 「4-planes box and daily sunpath in December 3.4.2. Result The results in Vasari (see table 8 and table 9) shows the 5 boxes stacked tree’s solar radiation is closest to the output of the sphere shape, the variation is about 0.1%, however the 45 degree tilt single plane is chosen to be the representative of sphere tree because it is easier to be modeled in eQuest and the variation to the solar radiation of sphere tree is still minor, which is about 1.4%. In addition, sunlight would pass through only one transparent or translucent layer, which 61 could be set to the tree density. Based on this finding, another comparison was made between a 45-degree tilted and 34 degree latitude (Los Angeles) tilted plane models. The percentage variation between the sphere model has dropped to 0.7% after the plane tilted from 45 degree to 34 degree. Although reasonable results were found for noon for south facing tilted plane tree, further studies (section 4.1) were done to check the shadow display range at different time of a day in different month. 3.5. Energy Calculations of Windows versus Walls 3.5.1. Test model information The test case building is a simple rectangular box with flat roof. The floor plan is 50’x50’. Floor to floor height is 10’. There was a 20’ wide by 5’ high double pane window on the south face. The first tree will be located 6 feet from the south wall, centered on the glazed area. The second tree will be located 6 feet from the south wall, centered on the non-glazed area. The height of tree canopy 20’ and the width is 20’, the height to the base of canopy is 5’. 62 Figure 25: Window versus wall test model in eQuest 3.5.2. Window v.s. wall Compare the effects when a tree is placed in front of the glazed area and a tree placed in front of the wall area. The results show a tree in front of a window saves 7 to 8 times more energy than the energy saved when there is a tree in front of the wall. Figure 26: Annual electricity use in 3 California climate zones 63 Table 10: Percentage savings relative to the case with no tree, building shade transmittance value :0.3 CZ12 CZ06 CZ16 No tree 0 0 0 1 tree (Window) 2.32% 2.53% 2.19% 1 tree (Wall) 0.31% 0.34% 0.27% Table 11:Percentage savings relative to the case with no tree, building shade transmittance value:0.8 CZ12 CZ06 CZ16 No tree 0 0 0 1 tree (Window) 0.67% 0.74% 0.64% 1 tree (Wall) 0.09% 0.10% 0.07% 3.6. Discussion Based on the initial studies in this paper on eQuest for tree shading, the author assumed the tilted building shade plane model could be a good representative of a tree model for the reason it could reach the accuracy of solar radiation on the surface when compared to the solar radiation from a sphere model. The author imagined that if the building shade plane model always faces perpendicularly to the sun, the accuracy of the simulation could improve. However, for author’s understanding, the rotation or movement cannot be scheduled in eQuest(DOE- 2). A compromised method might be used. Further discussion of this problem is in chapter 4. 64 4. CHAPTER 4 4.1. Shadow Patterns of different tree models Based on the solar radiation comparison test in Vasari, the author assumed the tilted plane model could be an ideal representative model for tree simulation because it had only 1.4% variation in solar radiation compared to the sphere tree model (rectangular box tree model had a variation of 17%). However, the author has done one more study to examine this discovery. The author used Ecotect in an attempt to compare the shadow pattern ranges of different tree models. Shadow range patterns, also known as butterfly diagrams, can be used to visualize the shadows that occur over a range of times for a given day. The shadow patterns of rectangular box and tilted plane were compared on 4 different days of a year in the Los Angeles area. 4.1.1. Tool: Autodesk Ecotect Analysis Autodesk Ecotect Analysis is a sustainable design analysis tool; it can be used at every stage of the design process, from defining the maximum site envelope or testing the conceptual model, to energy calculations and acoustics. It was chosen for this test for the following reasons: first, it has a highly visual and interactive display that presents analytical results directly within the context. Second, it can display the sun’s position and path relative to the model at any date, time, and location. It can also display the shadow pattern for a specific range of time. 65 4.1.2. Workflow The author followed the same methodology as in the previous test in the software, Vasari. A sphere (radius=10’) tree model was built as a baseline, along with that; a rectangular box (20’x20’x20’) and a tilted plane (20’x20’) were made. The tilted plane is 45 degree. See Figure 27. Four days in a year were calculated, March 21 st , June 21 st , September 21 st and December 21 st . The shadow pattern ranges from 6AM to 8PM. Time step value is set at 60 27 . The location is in Los Angeles. Figure 27 Test models in Ecotect 4.1.3. Results The results were drawn by graphical comparisons between Ecotect diagrams but not statistical analysis. For comparisons, the images were set at plan view. The visualized result is shown in Table 12. 27 Incremental step in minutes. This refers to the time increment between each shadow drawn in the diagram for the specified period. 66 Table 12 Shadow Patter Comparison in Ecotect (from 6AM to 8PM,time step value of 60) Sphere (Baseline) Rectangular Box Tilted Plane March June Sep Dec Figure 28 Overlapped shadow pattern diagram 67 Table 13: Shadow pattern comparison in Ecotech (from 6AM to 8PM, time step value of 80) Sphere (Baseline) Rectangular Box Tilted Plane March June Sep Dec By overlapping the shadow patterns of the baseline model the box model, the author found that the shadow pattern from box model is much bigger than the baseline shadow pattern. The box tree model in eQuest could over-estimate the energy saving. Compared to the baseline diagram, there is a lot of empty unshaded space in the shadow patterns of the tilted plane, especially during the early morning and the late afternoon. This suggested that a tilted plane tree model does not accurately simulate shading during early morning and late 68 afternoon if the plane is not rotated. Thus, a tilted plane building shade tree model in eQuest could under-estimate the energy savings. However, early morning and late afternoon is the time when the sunlight intensity is low. In an attempt to discover more on this, the author decreased the time step value from 60 to 20 in order to show more shadow intensity on the diagram and the shadow range period has been adjusted to find out the period when the tilted plane shadow pattern has the least spacing compared to baseline model’s shadow pattern. It was observed that when the time range adjusted to approximately 7AM to 5PM(table 14), the shadow pattern of tilted plane could almost have the similar pattern to the baseline model. Table 14 Shadow pattern comparison in Ecotech (from 7AM to 5PM,time step value of 20) Sphere (Baseline) Rectangular Box Tilted Plane March June Sep Dec 69 In Figure 29, the timetable plot showing that in Los Angeles, the over heating (75~100°F) could occur between 7AM to 5PM from July to September. In Figure 30, there is a sun chart showing that the shade is needed between 10AM to 3PM from June to September. Although the tilted plane model seems to be incapable of displaying the shadow in the early morning (before 7AM) and later afternoon (after 5PM), these periods are actually the time when shading is not needed. Based on the studies done on shadow pattern, the box tree model could overestimate the shading effect and the tilted plane tree model could underestimate the shading effect, however, the tilted plane model be able to display good tree shadow pattern between 7AM to 5PM. Figure 29 Timetable plot of Los Angeles area 28 28 Software used: Climate Consultant 4.0 70 Figure 30 Sun chart in Los Angeles area, left: summer; right: winter 29 29 Software used: Climate Consultant 4.0 71 4.2. Tilted plane 4.2.1. Definition of tilt In general, tilt is an inclination from vertical or horizontal surface. To avoid confusion, the tilt degree in this paper is based on the definition in the DOE-2 dictionary (LAWRENCE BERKELEY NATIONAL LABORATORY, 2008): Angle between the z-axis of the building coordinate system (vertical) and the surface outward normal. Outward is positive degree value; inward is negative degree value (Figure 31). Figure 31: Positioning building shade (LBNL, 2008) 72 The example in Figure 32 shows the 34-degree tilt plane. Figure 32: Tilt degree definition 4.2.2. Limits of tilted plane tree model orientation The shadow pattern test done in Ecotect was to discover the capability of the tilted plane model on displaying the shadow range of a day; the solar radiation test done in Vasari is to discover the tree model’s capability on getting the true value of solar radiation on a vertical surface (wall). To simplify the process, both of the tests have the assumption that the tilted plane is facing (perpendicular to) the source of sunlight or tilted with degree of location latitude and parallel to the vertical shaded surface (wall) of the rectangular box building (see Figure 33). For tree model placement, building coordinate system is applied. However, the tilted plane is 2D, so there will be some limitations when used in 3D environment. In 73 the real world, the layout of the building is not always a perfect rectangular, the tilted plane tree model is very hard to be placed if the tree is located outside more than one vertical surface, for example, HOUSE B in Figure 34. This is a major problem with the tilted tree type. Although it used fewer faces and hence the calculation time is faster, it is NOT a good representation for many cases, especially since it cannot be rotated in eQuest to always face towards the sun. Figure 33: Front view of tilted plane and the building. 74 Figure 34: 4 different building and building shade plane layouts 4.3. Scheduling foliage One of the difficulties in simulating trees is the scheduling issue. Compared to other types of shading devices, the tree is a living creature; some of the characteristics change over time. These are height, form, growth rate, density of branches and tree crown (Hammond, 1981). For tree shading simulation, scheduling the density of tree crowns in foliation and defoliation periods is the most critical. The user needs to decide the leaf-on and leaf-off periods first, then assign the seasonal transmittance value (density of the tree crown) to the tree model. Most trees grow in height fastest during the first twenty years, after maturity, the height of the trees unlikely change dramatically after that 75 (Hammond, 1981). In this study, to limit the variability, the author assumed that the trees were at their mature height. Future studies could show the change over time in energy savings (or lack of savings) as the trees increase in height. 4.3.1. Evergreen and deciduous trees Trees are not all of one type, and there are many ways to categorize them. By seed development, they can be grouped into gymnosperms and angiosperms. By internal structure, they can be categorized into hardwood and softwood. By the way they shed their leaves, trees can be divided into two major categories, deciduous and evergreen. Evergreen trees, for example, most of the conifers and palm trees, have leaves in all seasons; the foliage is persistent year round. To simulate it, a fixed transmittance value can be assigned to the tree crown model. Deciduous trees can also be called broadleaf trees, and come in a wide variety of sizes, shapes, forms, colors and textures. Generally, deciduous trees are those that lose their leaves each fall, go dormant for the winter, and leaf out again in spring. On the aspect of energy savings, deciduous trees might be the ideal choice for shading depending on the climate zone. A full leaf cover provides shade in hot seasons while admitting more sun in cold seasons (Hammond, 1981). 4.3.2. Leaf-on and leaf-off periods To simulate a deciduous tree, the user needs to know the leaf-on and leaf-off period of the tree for scheduling. The leaf period varies with species, growth rate and local factors such as weather, soil quality, availability of water, fertilizing, competition with other vegetation and landscaping practice (Hammond, 1981). In 76 California, most of the defoliation periods are from December to February (see Figure 35) Figure 35: Leaf-on and leaf -off periods for some California deciduous trees. (Hammond, 1981) 4.3.3. Winter and summer transmittance value After the foliation (and defoliation) period is determined, the overall transmittance value of the tree crown should be assigned. There is a large range of reported values from different studies; the transmittance value and the reductions in radiation cannot be predicted with certainty. Differences between these reported values resulted from a number of factors such as the number of trees sampled, different measurement techniques, and the different wavelength 77 range that was measured. Figure 36 is an example of sunlight penetration values used in a government consultant report (Hammond, 1981). Figure 36: Winter and summer sunlight penetration for some California tree species (Hammond, 1981) 4.3.4. Schedule for energy simulation programs In most of the energy simulation programs, the energy parameters such as occupancy, HVAC system operation, hourly thermostat, and lighting density can be scheduled to realistically predict the value of savings that an energy strategy would carry out. Most of the computer programs examined in this study do not have the tree components. Although some of the programs can model 78 obstructions, the real characteristics of this type of obstruction could not be modified and scheduled. 4.3.5. Schedule deciduous trees in eQuest Based on the tests done before, the author has done more research on the schedule component in eQuest and how to schedule building shades. To simulate the building shade, a custom schedule needed to be created in the detail-edit mode. There are 12 types of schedule in eQuest 30 that serve different purposes, in table 15, 6 most common types of schedule are explained: Table 15: 6 Types of schedule in eQuest Types Description Example On/off Accepts value 0 and 1 where 0 means off 1 means on. A value of 1 simply allows an item to operate and does not mean that the item is 100% on. Heating/cooling availability Fraction Accepts values between and including 0.0~1.0.In such a schedule, each hour has a fractional value of the design conditions. The fractional value is multiplied by the design value. Lights and internal loads. Multiplier Accepts values 0.0 and above. Lighting and people Temperature Accepts values represent temperature. It defines the setpoint for all 24 hours. Typically, it is necessary to have 2 annual schedules and 2 weekly schedules (1 heating, 1 cooling). However, it is typically necessary to have 4 daily temperature schedules (heating and cooling with occupied and unoccupied each) Zones, Thermostat schedule 30 http://doe2.com/download/DOE-22/DOE22Vol2-Dictionary_47b.pdf 79 Table 15: Continued Radiation Accepts a value that represents a radiative flux. Window On/off/temp Accepts value 0 and 1(similar to on/off), other value is also acceptable, and is assumed to represent a flag temperature. When a temperature, the meaning of the value and its action varies by the component referencing the schedule. Heating system Regardless of the schedule types, all schedules are divided into three schedule categories: daily, weekly, and annual. To access a schedule, select one of the modules on the top (project and site, building shell, internal loads, water side HVAC, air side HVAC, utility and economics), and scroll towards the bottom of the component tree, view in spreadsheet. 4.3.6. Schedule workflow in eQuest To explain the workflow of scheduling foliage in eQuest, consider this following example. A Norway Maple shading a building has a transmittance of 0.7 in winter and 0.1 in summer; the leaf-off period is from December 1 st to March 31 st . To open the schedule properties window, double click on any schedule type under building shell components tree. As with all things in eQUEST, when constructing schedules it is important to start at with the lowest level, in this case with daily schedules. First, create two new daily schedules named “Tree-0.7” and “Tree-0.1” by multiplier method (see Figure 37): hourly values for “Tree-0.7” are all o.7 80 (ratio), hourly values for “Tree-0.1” are all 0.1(ratio). Figure 37: eQuest day schedules screenshot 81 Second, create two new weekly schedules named “tree-0.1-week” and “tree-0.7- week” by using multiplier type (see Figure 38). Assign daily schedule “Tree- 0.1”created before to Monday to Sunday, holiday and heating and cooling design days for “tree-0.1-week.” Assign daily schedule “Tree-0.7”created before to Monday to Sunday, holiday and heating and cooling design days for “tree-0.7- week”. Figure 38: eQuest week schedules screenshot Annual schedules consist of weekly schedules or just one weekly schedule. Seasons can be constructed with multiple weekly schedules. For this case, create 82 one new annual schedule named “NorwayMaple-Tree-0.1/0.7” by multiplier method (see Figure 39). In weekly schedule assignments components, the costumed weekly schedules created in step 2 can be found at the bottom of the scroll. To enter the ending month and ending day, it is important to know that all weekly schedules have the default starting time with month 1, day 1 and ending time with month 12, day 31. Figure 39: eQuest annual schedules screenshot 83 The final step is to open the building shade property window and assign the annual schedule “NorwayMaple-Tree-0.1/0.7” to transmittance. The resultant inp file for this building shade schedule is shown in Table 16. Table 16: .INP file for building shade schedule (From top: day-schedule, week-schedule, year-schedule, building shade schedule) 4.4. Summary In this chapter, the limitations of tilted plane model have been discussed. An in- depth study on the tilted building shade model was done by comparing the shadow ranges for an entire day. The result has shown that the box tree model 84 could overestimate the shading effect and the tilted plane tree model could underestimate the shading effect, However, the tilted plane model is able to display good tree shadow pattern between 7AM to 5PM when it is facing south. The tilted plane also has the limitations in positioning in the 3D environment. In eQuest, an exterior obstruction can be scheduled. The workflow of scheduling foliage as an exterior shading in eQuest was explained in section 4.3. Among all the energy simulation programs examined in the thesis, eQuest is the one that has the capability to model and calculate exterior obstructions like trees. Unfortunately, a good representative form of tree model has not been found yet. A box tree model may overestimate the energy savings for the following reasons: One, 2 or more surfaces of the model will be hit when light goes through, thus, the information (transmittance value) assigned to the box model may be wrongly calculated. Two, in section 3.4.1, the incident solar radiation from a box model differs the most to the basecase sphere model. Three, a box model displays a larger range of daily shadow pattern, see section 4.1.3.The author assumes that a tilted plane which could rotate to face the sun always can be a better model, unfortunately, the tilt degree can not be scheduled in eQuest. It might be possible to schedule a sequence of tilted planes, one for each hour and then vary the transmissivity between 1.0 for 23 hours and 0.1 or 0.7 for the one hour that the sun is normal to that particular tilt. But the precalculation required is prohibitive. 85 In the next chapter, the eQuest tilted building shade tree model will be compared with a landscaping program named Tree Benefit Estimator. Although the tilted plane has many limitations and restrictions, a validation work will be done by comparing the measured data and simulated data reported in “Peak power and cooling energy savings of shade trees” (Akbari et al, 1997). 86 5. CHAPTER 5 5.1. Comparison to Tree Benefits Estimator In this section, the comparison between the web-based program Tree Benefit Estimator (see 1.6.5) and eQuest was done. Tree Benefit Estimator was designed to be used by homeowners who have no formal background on urban forestry, energy programs, or building science. It provides rough estimations of tree’s shading and evapotranspiration impact on residential building. In Estimator, there is no input for a building description; however, the case study house in section 4.2 was used in this study in order to run the calculation in eQuest. Figure 40: Workplace of eQuest(left) and Estimator(right) 87 5.1.1. Overview of Tree Benefit Estimator This program uses i-Tree as an engine for calculation. I-tree, formally called UFORE, was developed by the USDA Forest Department. 31 This engine, i-Tree, is frequently used by communities, non-profit organization and consultant 32 to quantify the structure of community trees and the environmental services that trees provide in urban forestry research and practice. The Estimator is freely accessible. 5.1.2. Energy simulation methods for Tree Benefit Estimator: Shading The program estimates the effects of trees on building energy use and consequent carbon emission from power plants. Methods for these estimates are based on a report by McPherson and Simpson (McPherson and Simpson, 1999). The simulation in the report was made by Shadow Pattern Simulator program (SPS) and MICROPAS (see 1.4.5). SPS calculated the intensity of solar radiation by trees with the information of tree size, location, canopy density, and time. The canopy density value (expressed as the shade coefficient) used in this research is measured with pyanometer by McPherson (McPherson, 1981). A pyranometer is inherently non-directional and therefore yields values that are not accurate for specific shadows. However, this is the best data currently used in any of the published simulations. A grand total of 19,008 simulations (24 locations x 6 tree 31 http://www.itreetools.org/ 32 http://www.itreetools.org/resources/manuals/i-Tree%20Streets%20Users%20Manual.pdf 88 types x 3 prototypes x 11 climate regions x 4 ages) were done (see Table 17). To simplify these data, McPherson tabulated these mass simulation results into 3 building vintages and 6 tree types for a total of 18 values for each climate region. All simulations were done for a square 1 or 2-story house with a symmetrical window distribution based on a sample of 254 homes in Sacramento (McPherson, 1998). Default energy effects per tree were set for these parameters (Table 7) and integrated with the development of UFORE. However, the author could not find any validation or calibration report for this simulation method. The author assumed this tool was made for the home-users as a simple calculator, but not a design or study tool. Table 17 Tree simulation parameters (McPherson, 1999) 24 locations 8 tree azimuth 3 tree-to-building distance 6 tree types 3 sizes: large, medium, small Deciduous and evergreen 3 building prototypes Pre-1950, 1950-1980, post-1980 11 climate regions Southeast, South Central, Mountains, North Central, southwest, Gulf Coast/Hawaii, California Coast, Northern Tier, Mid-Atlantic, Dessert Southwest, Pacific Northwest 4 tree ages 5, 15, 25, 35 years 89 5.1.3. Input comparison Norway maple (Acer platanoides) is a medium-size deciduous tree species. It was chosen for this test for two reasons. First, this tree species is available in Tree Benefits Estimator’s tree type library. Second, the transmittance value of this tree species has been reported in different papers (see Table 18). Table 18 Reported transmittance value for Norway maple by different methods (McPherson, 1984) Summer Transmittance Value Winter Transmittance Value Reference Method 0.1 o.75 Schiler, 1979 Photograph and optical scanner 0.14 0.65 Hammond et al, 1980 Light meter 0.1 N/A McPherson, 1981 Pyranometer 0.14 N/A Heisler, 1982 Arrays of pyranometers Inputs in Tree Benefit Estimator are very limited. To do a comparison with eQuest, the author made assumptions on the building shade model (tree model). The author assumed the tree is 30ft in height with a canopy size of 25 ft x 25 ft x 25 ft with leaf evenly distributed for a medium mature Norway maple. In Table 8, different transmittance values of Norway maple were reported in different papers by different instrumentation. Unfortunately, there is no consensus about which is the best method for getting the true transmittance value of a tree. Furthermore, the transmittance probably varies, depending on angle and field of view. For this test, a transmittance value of 0.1 was used for summer; 0.7 for winter. “Leaf on” 90 period is from April to November, “leaf off” period is December and January to March (Hammond, 1981). Table 19: Inputs for the tree model in eQuest Size of tree model 25ft x 25ft Transmittance value of tree model 0.1 summer, 0.7 winter Location of the tree model Los Angeles, climate zone 6 Placement of the tree model 15ft from the building Table 20: Inputs for Tree Benefit Estimator kWh cost in dollars Summer:0.2, Winter:0.1 Tree species Norway Maple Tree Age 20 years Climate Area Los Angeles Tree distance to the house 0~15ft The first input in Tree Benefit Estimator is the average kWh cost in dollars; however, financial savings are not discussed in this study. Another issue for this comparison test is that Tree Benefit Estimator does not require any information about the building. Directional distribution of window and wall area can have a major impact on building shading, for example, a south, east or west oriented wall with larger glazed area will greatly affected by shade compared to the wall with little glazed area; neither wall would be greatly affected 91 if oriented to north. In the window-wall test done in chapter 3.4, the result has shown that the energy saving from a tree in front of a window is greater than the energy saving from a tree in front of a wall. Thus, the orientation of wall and window and the placement of the tree (in front of wall or in front of a window) are all critical to achieving energy efficiency. It is clear that the Estimator is not a tool for researcher or designer, to do the assessment on the Estimator and the comparisons with eQuest; a simple one-story residential building was created in eQuest. Table 21: Building description in eQuest. Floor area 1444 ft 2 (38ft x38ft) Exterior wall height 9 ft Roof Composite shake Wall Stucco Windows 6 windows (5’x13’) evenly distributed on each walls, 2-pane, aluminum frame. Glass type: double, U-value=3.16 Heater Gas, 0.042 Mbtu/hr Air Conditioner Central 5.1.4. Absolute energy saving output comparison The comparisons of the annual kWh saved for different quantities and orientation of trees was calculated in Tree Benefit Estimator and eQuest. These comparisons are shown in Figure 41 92 Figure 41: Tree Benefit Estimator and eQuest energy saving outputs for 4 orientations 93 Almost always, Estimator showed a larger benefit than eQuest. The annual electricity savings increased when more trees are placed. In Estimator, the incremental is regular; the graphs in 4 orientations (figure 40) are all straight line (linear growth). However, the graphs of eQuest are not straight line at all. In the east and west graphs, the marginal increment became smaller after a certain number of trees (west: 4 trees, east: 3 trees.) For one tree, Estimator and eQuest both have shown the largest annual electricity savings is when the tree placed in west; the least savings is when the tree placed in north; the savings from south and east are pretty close. However, the difference between eQuest and Estimator becomes larger when the number of trees increases. Due to a lack of on-site measured data, the comparison test done in this section is merely a sensitivity test but not validation work. The parameters in this test are restricted to tree quantity and orientation for two reasons: Tree Benefit Estimator has very limited inputs, and there is no input or modeling interface for the building information. This is critical -- a building could not be entered in Estimator and the underlying formulas and assumptions about energy savings in Estimator are not known. A building was input to for simulation in eQuest, but 94 comparisons cannot be considered accurate. The Estimator was developed based on the report published by McPherson and Simpson in 1999 and the real weather data of that time was used, However, eQuest uses CZ2 weather data supplied by California Energy Commission. The energy saving calculations from Estimator and eQuest might be incorrect because the same real weather data was not used. Besides the discoveries discussed, the results for Estimator does not make sense at all because savings should be a diminishing return until all solar gain is eliminate, but not a linear saving of infinite amount. Energy saving should be an asymptotic curve, and the output in eQuest seemed to be more reasonable in this case. 5.1.5. Conclusions regarding Tree Benefit Estimator and eQuest Tree Benefit Estimator seemed to be an easy tool to calculate a tree’s benefit in energy savings; however, the test done in this section has shown the limitations of this tool. It is impossible that the saving increases indefinitely when the quantity of trees increases, it should be getting close to a certain value that is the maximum energy savings of cutting off the direct gain to the building. The Estimator seems to be an unrealistic tool to use; in this case, a tilted plane representation in eQuest is better. 95 5.2. Comparison to a previous method in DOE-2 In the paper “Peak power and cooling energy savings of shade trees” (Akbari et al, 1997), the benefits of shade trees in energy savings were reported after experiments and computer simulation in DOE-2. In this section, the author tried to simulate one of the two case study houses in eQuest that were described in the paper and compared the results with the measured data and the DOE-2 simulation results from the same paper. When the house was shaded, the DOE-2 model over predicted cooling energy use by a large margin; if the eQuest method could decrease the margin then it could be a better method to use. 5.2.1. Summary of the compared study This study (Akbari et el, 1997) was supported by the California Institute for Energy Efficiency and the Sacramento Municipal Utility District (SMUD). It was a two-year project to monitor peak power and cooling energy savings from shade trees in some of the SMUD’s shade tree program buildings in Sacramento. The data was collected from June 1992 to October 1992. The case house 33 used in this comparison test is located in Sacramento; the monitoring periods for unshaded and shaded conditions are shown in table 22. 33 This case study house was named “Site T1” in the published paper (Akbari et al., 1997); named “Site C” in the LBL-34411 technical report (Akbari et al., 1993). 96 Table 22: Monitoring schedule for the casestudy house T1 (Site C) in 1992 Monitoring dates in 1992 Monitoring days Unshaded 6/8~8/3 28 Shaded 9/1~10/14 44 The base case (unshaded condition) was monitored during June 8 to August 3, and then sixteen trees in containers (eight tall and eight short trees) were placed at south and west facing walls. The case house for the shaded condition was monitored from September 1 to October 14. (Table 22) The measured data included air-conditioning electricity use, indoor and outdoor temperatures, humidity, and roof and ceiling surface temperatures (LBL-34411). The monitored buildings were simulated in DOE-2 with four methods of tree representation models (see Table 23). The sensitivity test for these four methods has shown that none of the methods could bring the simulation and measured data into agreement (Akbari et el., 1997). Table 23: Methods used in compared paper to model shade trees (Akbari et al., 1997) 1 Rectangular form, transmissivity value=0.1 2 Rectangular form, transmissivity value=0 (opaque) 3 Setting the window coefficient to 0 4 Eliminating insolation on both the windows and the walls. 97 Since the measured data were only collected from June 8 to October 15, the simulation was only performed for this period of time, and the tree model was not scheduled for foliage changes. The annual energy savings was not reported in the paper, only daily and seasonal savings were studied. The comparison showed the discrepancies between measured data and simulation estimates (see Figure 42). The DOE-2 simulation results under estimated the cooling energy saving by a factor of two. Figure 42: Measured vs. simulated (Akbari et. al, 1997) 98 5.2.2. Building description The building modeled in eQuest was based on the limited information recorded in the reports (Akbari et el, 1997), (LBL-34411). Assumptions had to made to have eQuest run the calculation. For example, it was assumed that the windows were distributed evenly on each wall because a detailed floor plan was not provided in the paper. The building description in Table 24 was from the compared report by Akbari et al., and the simulation done in eQuest used the same inputs. Table 24: Building characteristic in eQuest based on the compared reported by Akbari et al. eQuest model Floor area 1444 ft 2 Number of stories 1 Floor to floor height 8.5 ft Floor to ceiling height 8.5 ft Roof R-19, Ceiling Attic, Wall R-11, white color, stucco Window 2-pane, U value-3.16 Double 6mm/6mm, Shading coefficient:0. 57 Aluminum frame Thermostat setting Heating: 68°F Cooling: 78°F Heater Gas furnace, 47000 Btu/hr Air conditioner A/C, 36000 Btu/hr Airflow 1200 CFM 99 5.2.3. Tilted plane model in eQuest The tall tree was modeled as a 20ft x 20ft building shade surface, and the small tree was modeled as a 7ft x 7ft building shade surface in eQuest (Table 25). 4 tall trees and 4 small trees were place facing the facing south wall; 4 tall trees and 4 small trees were placed on the west. The planting location could not be found in the paper; thus, another assumption has made that the trees were 15ft ~ 20ft off the building (Figure 43). The annual scheduled was designed for the trees, however; only the output from June to September was studied and compared because the measurement was only made for this period of time. Table 25: Trees in experiment and the correlated models in eQuest Tree: quantity and type Building shade model in eQuest Tall tree 1 Chinese Hackberry 1 Chinese flametree 2 Raywood ashes 4 Tulip trees Quantity: 8 20ft x 20ft 34 degree tilted planes Transmittance value: 0.1-summer 0.7-winter Short tree 8 Eastern redbud Quantity: 8 7ft x 7ft 34 degree tilted planes Transmittance value: 0.1-summer 0.7-winter 100 Figure 43: Building shade models in eQuest 101 5.2.4. Evapotranspiration The effect of evapotranspiration proved difficult to isolate in onsite experiment (Akbari et el, 1997)The first attempt was to place a custom heating and cooling device to represent a tree outside the building in eQuest, unfortunately, there is no way to just place it outside. eQuest cannot calculate the non-solar microclimate changes from any exterior obstructions. It would be possible to modify the weather file to account for evapotranspiration effects. In one paper (Akbari and Konopacki, 2004), the standard WYEC2 (Weather Year for Energy Consumption) was modified to create modified ambient temperature (Tmodified=Tstandrd+ΔT) data for the DOE-2 to rerun the indirect effect of shade trees, but this was not done in this case. 5.2.5. Results The simulation was done by using the weather file CZ2 (California climate zones revision 2 supplied by the California Energy Commission and Sacramento is in California climate zone 12. CZ2 is a long-term average weather data (~30 years). Table 26: Absolute daily energy savings from shade trees Measured average data (Akbari et el, 1997) Simulated data in DOE-2 (Akbari et el, 1997) Simulated data in eQuest From June 8 to October 14 in 1992 From June 8 to October 14 in 1992 June July Aug Sep Oct Daily savings (kWh) 4.8±0.6 2.4 4.0 4.8 4.8 4.0 3.2 102 Figure 44: Comparison of daily cooling energy saved 103 Table 27: Percentage cooling energy savings, measure vs. simulation Cooling energy use (kWh) Period Case June July Aug Sep Oct 6/28 to 10/12(1992) Unshade N/A 484 512 274 75 Shaded N/A 361 390 160 39 Measure data (Akbari et al., 1997) Percent Savings N/A 25 24 42 48 All year Unshaded 880 1050 980 710 430 Shaded 760 900 830 590 330 Simulation data (eQuest) Percent savings 13.6 14.3 15.3 16.9 23.3 Figure 45: Comparison on percentage energy saving between measured and simulated data For absolute energy savings, the result has shown that the simulated data in eQuest is closer to the measured data when compared to the simulation data in Akbari et al.’s report. The method (scheduled tilted plane) used in this paper is 104 better than the 4 methods used (table 23) in the compared report (Akbati et al., 1997). For the percentage savings (table 27), the simulation done in eQuest under-estimated the cooling energy savings and the discrepancies are larger in September and October. The measured cooling energy usages are consistently lower than the eQuest simulation (table 27); a calibration work is suggested in future work. These results are preliminary. Although numerous attempts have been made to get more detailed information about the specifics of the case study, very little information was available about critical elements. 5.2.6. Conclusion regarding to rectangular box method and tilted plane method The comparison test done in this section has proved that the tilted plane model is better than the previous model (rectangular box) because the simulated output is closer to the measured data reported. Once again, these results are preliminary. Too many variables had to be "guessed" to make this an accurate representation of the true situation. In Akbari’s paper, it concluded the simulation failed for two reasons, the failure of tree representative model and the failure of the acquiring accurate building information. These two failures also applied to the simulation done in this study. Other reasons might be the following: inability to calculate the savings from evapotranspiration may have made the savings output lower than the measured data; the standard weather data cannot completely represent the weather 105 condition in 1992; the software does not do the calculations properly for the exact conditions specified; and/or too many assumptions were made that threw the simulations off. 5.3. Tree shading effects on daylighting in eQuest Based on the Building Energy Data Book published in 2011 34 , lighting accounts for 10% of overall energy consumption in residential buildings and 20% in commercial buildings. Optimal usage of daylight is expected to decrease lighting electricity consumption. Although this thesis focuses on residential energy usage, a preliminary daylighting simulation in eQuest was completed in this section to examine the effects of tree shading on daylight harvesting in order to gain a basic understanding of the tradeoffs between shading for energy savings and not shading for interior daylighting. 5.3.1. Reflection Programs based on Radiance such as DAYSIM, IES-VE, and Ecotect are common tools for daylighting simulation. eQuest also allows the user to determine the impact of daylighting on energy use (Winkelmann and Selkowitz, 1985). Its engine, DOE-2, does not account for reflection of solar radiation from building- shades (unattached shade) onto the building, only from the attached shading device. However the user can specify the reflection of visible and ground radiation by editing the solar properties component under “building shade.” In eQuest, “visible reflectance” is the reflectance from the shading surface, and the 34 http://buildingsdatabook.eren.doe.gov/default.aspx 106 “ground reflectance” is the reflectance from the ground in the vicinity of the shading surface. Input values range from 0 to 1; a black surface is zero reflectance. To simulate the luminous environment with the effect of the trees, the reflectance values of some tree leaves are. Leaf reflectance in the UV wavelength bands is approximately 0.05 for the trees studied in the paper: “Ultraviolet leaf reflectance of common urban trees and the prediction of reflectance from leaf surface characteristics”(Grant et al., 2003). 5.3.2. Simulation of the case study house The eQuest model from section 5.2 has been used (table 24 and table 25). The building was divided into three spaces, southeast perimeter space (conditioned), northwest perimeter space (conditioned), and under roof space (unconditioned). After activating the daylighting zones in the southeast perimeter space (SE Perim Spc) and the northwest perimeter space (NW Perim Spc) (see table 28), three scenarios were simulated: no trees, trees with the DOE-2 default reflectance 0.5, and trees with a reflectance of 1.0. The DOE-2 default daylighting schedule has been used. The lighting set point, the target light level that is to be maintained, is 50 foot-candle (see figure 47) and lighting control is on/off. The daylighting simulation done in this section is to examine the tree’s effect on daylight harvesting, and the program’s capability of simulating the trees with reflectance. It is a sensitivity test, and the real reflectance value for tree leaves was not used. Table 28: Spaces in active zone 107 SE Perim Spc (the grey area) NW Perim Spc (the grey area) Figure 46: eQuest daylighting zone screen 108 5.3.3. Result of the case study house The area lighting is usually uniformly distributed over the space. The energy results have shown that the trees will increase the area lighting energy usage (see table 29). Task lighting is concentrated where work takes place 35 and it remained the same (see table 29) after the introduction of the trees because it was scheduled as “GndPrm Occ Sch”(see Appendix E). It has shown that trees might have downside on daylighting and increase the lighting electricity usage. However, in the daylighting results (figure 47 and 48), it has shown that with or without trees, the light levels are well above the lighting set point, the light was sufficient, that explains to some extent that it does not remove enough light to change daylight harvesting by the introduction of trees. After adjusting the visible reflectance value of the trees from 0.1 to 1.0 (building shade models), the lighting energy usage and illuminance level remained the same (see table 29). The results of this test have indicated two issues related to daylighting. First, the tree’s effect on daylighting might increase the annual energy usage, but it was not shown in this test. A future study could carefully examine the trade-offs between tree shading, electrical savings due to cooling, and potential electrical increases due to use of more artificial lighting. Figures 47 and 48 seem to indicate that at least for the case study chosen (more case studies should be simulated in different climate zones), winter might be the major concern, and cooling loads are usually not predominant. This leads to another future study on the trade-offs 35 http://doe2.com/download/DOE-22/DOE22Vol3-Topics.pdf 109 between tree shading, potential increases in heating loads in the summer (one would need to study different types of trees), and potential electrical increases due to use of more artificial lighting. Figure 47: Average daylighting illuminance in SE-Perim space Figure 48: Average daylighting illuminance in NW-Perim space 110 Second, the reflectance values of the building shade is not all what was expected; why did the value of the tree reflectance not change the any of the other values? Based on the results of Table 29, another simple test has done in section 5.3.4. Table 29: Lighting energy usage output in eQuest Task lighting energy use (KWh) Area lighting energy use (KWh) No trees 175 1180 Trees, reflectance 0.1 175 1229 Trees, reflectance 0.2 175 1229 Trees, reflectance 0.3 175 1229 Trees, reflectance 0.4 175 1229 Trees, reflectance 0.5 175 1229 Trees, reflectance 0.6 175 1229 Trees, reflectance 0.7 175 1229 Trees, reflectance 0.8 175 1229 Trees, reflectance 0.9 175 1229 Trees, reflectance 1.0 175 1229 5.3.4. Simulation and results of a simple box room A 15ft x 20ft x 9ft south orientated box room with one 20ft x 20ft tilted plane tree located in front of the 6ft x 11ft single pane window on south wall. The simulation was done in CZ6 (California climate zone 6). The lighting set point, the target light level that is to be maintained, is 50 foot-candle and lighting control is on/off. Various scenarios were simulated: no tree, one tree in front of the window with reflectance value from 0.1 to 1.0. 111 Figure 49: A simple box room in eQuest In figure 50, the illuminance is well above the design set point when there is no tree; the illuminance is below the light set point with the introduction of the tree (Appendix F), in other words, the usage of artificial lighting will increase because of the tree’s impact on daylight harvesting. In this case, 237.2 kWh was saved by the cooling and 98.3 kWh was used to turn on the lights annually. Figure 50: Average illuminance in the box room 112 In table 30, the area lighting energy usage will increase when there is a tree placed outside the window. However, the reflectance value of the tilted building shade (tree model) still seems to be unclear – the values are the same regardless of the reflectance value; it is slightly different from not having any tree at all. The reason for this is unknown. This also needs further study, but does illustrate another difficulty involved in modeling trees in eQuest. Table 30: Lighting energy usage output (small box room) Task lighting energy use (KWh) Area lighting energy use (KWh) No trees 85.0 2204.6 Tree, reflectance 0.1 175 2302.9 Tree, reflectance 0.2 175 2302.9 Tree, reflectance 0.3 175 2302.9 Tree, reflectance 0.4 175 2302.9 Tree, reflectance 0.5 175 2302.9 Tree, reflectance 0.6 175 2302.9 Tree, reflectance 0.7 175 2302.9 Tree, reflectance 0.8 175 2302.9 Tree, reflectance 0.9 175 2302.9 Tree, reflectance 1.0 175 2302.9 113 6. CHAPTER 6 A tree is a unique type of exterior obstruction for buildings compared to a shading device such as a fin or shade. It is also a living organism that is difficult to simulate in computer programs like eQuest that are used in the architecture profession. 6.1. Conclusion The hypothesis “it is possible to create an algorithm model of a tree that can be used within existing energy software programs to calculate its shading and evapotranspiration benefits” has been disproved. Current energy simulation tools do not seem to accurately predict the energy savings on residential building from shade trees and the simulation for evapotranspiration has failed. The two comparisons done in chapter 5 have shown that the suggested tilted plane in eQuest model is better than the landscaping program Tree Benefit Estimator (iTree) and the rectangular box in DOE-2. Tree Benefit Estimator is unrealistic to use for tree benefit simulation because inputs are limited. This thesis involved a series of comparisons in order to discover a better method for simulating the effects of trees on energy saving. The scope of this study is limited to sun shading and evapotranspiration and the scale of residential buildings. Most of the previous research has been done to prove the tree’s benefit at the municipal level and tried to adapt as many factors, such as utility billings, 114 and carbon benefits. However, those research projects need to have more in- depth examination and discussion on the tool selection and methodologies. On software assessments (Chapter 2), one of the biggest difficulties is the ability to model obstructions and furthermore, to adjust the properties of the obstruction to have it function as a tree. Unfortunately, most of the energy modeling software does not have such features for tree simulation. eQuest was chosen for several reasons (Chapter 2), one is that most of the previous research were done in DOE-2, and there was more historical data to compare with. Thus, eQuest, the most up to date interface of DOE-2 was chosen. With the side tests done in this study, several limitations of eQuest were discovered. First, data-exchange could not be performed well in DOE-2 engine based interfaces such as eQuest. Autodesk Vasari cannot create the correct INP file for eQuest. After researching more carefully the limitations of building shade geometry in eQuest, different shapes of trees were compared for the solar radiation values and daily shadow pattern range. A tilted plane building shade was eventually chosen as it compared well to a hypothetical spherical tree and only had one face that could more easily be scheduled for transmittance. An idealized model is to have the building shade surface (tree) always facing the light source. This could not easily be accomplished in eQuest. The second limitation is scheduling. Seasonal leaf on or off can be scheduled by the multiplier method in eQuest. However, there is a no consensus of an accurate 115 method to measure net transmittance value of tree canopy. The historical reported transmittance value for different seasons and tree species are measured and calculated by different methodologies, thus, it is not known which version, if any, would be realistic for simulation use. For building shade component in eQuest, the scheduling function is restricted to the value of transmittance and the tilted degree cannot be scheduled. After the side tests, simulation was done in 3 different climate zones. It was shown in the eQuest simulation that the tree’s benefit on energy saving can be much more significant (7 to 9 times) if it is placed in front of the glazed area than placed in front of the wall. This had been also mentioned in several papers. Another finding is that simulation output showed a greater energy saving from trees planted on the west than trees planted on south of the building. 6.2. Future work suggestions: 6.2.1. On-site experiment The biggest obstacle in such studies is the lack of good validation data; the compared data used in this thesis is only available from June to October in 1992. A long-term monitoring on-site test should be done to help the development of shade tree simulation program. This is critical. Real data is critical for validation both for the energy shading studies and evapotranspiration research. It is strongly recommended that this be the next phase of the research. 116 6.2.2. Scale of simulation In this study, simulation was done for up to 16 trees, 3 climate zones and four orientations. Simulation can be done with different types and numbers of trees, more climate zones and different orientation such as south-east, south-west. Expand the geographical boundary and to compare with more landscaping programs such as TownScope. 6.2.3. Tool to calculate transmittance value of the tree canopy Foliage transmittance values need to be collected. One method might be to create a database of tree types for different seasons based on photographs analyzed in Photoshop. Others have made studies into a method similar to this (Schiler et al., 1979). 6.2.4. Other potential programs Software such as Design Builder and IES were suggested that might be a potential tool for tree simulation. A deeper examination of these software programs can be done in the future. 6.2.5. In depth studies on tree’s effect on daylighting The tests done in section 5.3 are a preliminary study on tree’s effect on daylighting. In some cases, trees may have substantial effect on lighting energy usage. More case studies in different area, location, orientation, window sizes…. etc. can be simulated. 117 6.2.6. Reflectance of tree leaf In section 5.3, the reflectance value input was not doing anything to the daylighting and energy simulations. The reason is currently unknown; it could be user error on the program or the limitation of the program itself. Further study on eQuest’s capability of calculating the reflectance from the unattached obstruction (trees, neighboring building …etc) of the building is suggested. 118 BIBLIOGRAPHY: “Applicability of citygreen urban ecosystem analysis software to a densely built urban neighborhood”. (2004) 25: pp. 173-186. Akbari, H., Konopacki, S. (2004) Energy effects of heat-island reduction strategies in Toronto, Canada. Energy 29: pp. 191-210. Akbari, H. (2002) Shade trees reduce building energy use and CO2 emissions from power plants. Environmental Pollution 116: pp. 119-26. Akbari, H. (2001) Cool surfaces and shade trees to reduce energy use and improve air quality in urban areas. Solar Energy 70 (3): pp. 295. Akbari, H., Bretz, S., Kurn, D. M.,Hanford, J. (1997) Peak power and cooling energy savings of high-albedo roofs. Energy and Buildings 25 (2): pp. 117-26. Akbari, H., Bretz,S.E., Hanford, J.W., Kurn ,D.M. , Fishman, B.L., Taha, H.G. (1993). Monitoring peak power and cooling energy savings of shade trees and white surfaces in the sacramento municipal utility district (SMUD) service area: Data analysis, simulations, and results. (LBL 34411). University of California Berkeley, California: Energy and Environment Division Lawrence Berkeley Laboratory. Akbari, H., Taha, H. (1992) The impact of trees and white surfaces on residential heating and cooling energy use in four Canadian cities. Energy 17 (2) (2): pp. 141-149. Al-Sallal, A. K. (2009) Practical method to model trees for daylighting simulation using hemispherical photography. Paper presented at Eleventh International IBPSA Conference, Glasgow, Scotland. Carmody, J. ed. (2004). Window systems for high-performance buildings. New York: W.W. Norton & Company. 119 Crawley, D. B., Hand, J. W., Kummert, M., Griffith, B. T. (2008) Contrasting the capabilities of building energy performance simulation programs. Building and Environment 43 (4) (4): pp. 661-73. Deussen, O. (2005) Digital Design of Nature: Computer Generated Plants and Organics: pp. 149-80. Berlin: Springler Donovan, G. H., Butry, D. T. (2009) The value of shade: Estimating the effect of urban trees on summertime electricity use. Energy and Buildings 41 (6) (6): pp. 662-668. Gates, D. M., Tantraporn, W. (1952) The reflectivity of deciduous trees and herbaceous plants in the infrared to 25 microns. Science 115 (2997) (Jun. 6): pp. 613-616. Geiger, R., 1894, Aron, R. H., & Todhunter, P., 1954. (2003). The climate near the ground. Sixth edition. Lanham, Md: Rowman & Littlefield. Grant, R. H., Heisler, G. M., Gao, W., Jenks, M. (2003) Ultraviolet leaf reflectance of common urban trees and the prediction of reflectance from leaf surface characteristics. Agricultural and Forest Meteorology 120: pp. 127-39. Gudeman, S., (2008) Product reviews: IES VE building design software.http://www.green-buildings.com/content/781795-ies-ve-building- design-software (accessed 2011,11,12). Hammond, J. (1980) Planning solar neighborhoods, Sacramento: The California Energy Commission (accessed 3/26/2012 ). Harris, C. (2006). Shading coefficient. In Dictionary of architecture and construction. 4th ed., Revised ed. New York, NY, USA: McGraw-Hill. Heilman, J. (1991). Effects of turfgrass evaporation on external temperatures of buildings. Theoretical and Applied Climatology 43 (4) (-01-01): pp.185-94. Heisler, G. M. (1986). Energy savings with trees. Journal of Aboriculture 12 (5): pp.113-125. 120 Heisler, G. M. (1986) Effects of individual trees on the solar radiation climate of small buildings. Urban Ecology 9: pp.337-59. Hildebrandt, E. W., and Sarkovich, M. (1998). Assessing the cost-effectiveness of SMUD'S shade tree program. Atmospheric Environment 32 (1): pp. 85-94. Hirsch, J. J. (2009). http://doe2.com/ (accessed 2012/12/25). Huang, Y. J., Akbari, H., Taha, H., Rosenfeld, A. H. (1987). The potential of vegetation in reducing summer cooling loads in residential buildings. Journal of Applied Meteorology 26: pp. 1103-16. Jones and Stokes Associates. (1998). Product specification for the transagency resources for economic and environmental sustainability project. Sacramento, CA: Prepared for TreePoeple,Beverly Hills, CA. Kramer, P. J., Kozlowski, T. T. (1960). Physiology of trees, New York: McGraw- Hill. Laurence Berkeley National Laboratory, James J. Hirsch & Associates (2008). DOE-2.2 volume 2:Dictionary. ,http://www.doe2.com/download/doe- 22/DOE22Vol2-Dictionary_46.pdf. Longcore, T., Li, C., Wilson, J. P. (2004). Applicability of CITYgreen urban ecosystem analysis software to a densely built urban neighborhood. Urban Geography 25 (2): pp.173-186. Maile, T.,Fischer, M., Bazjanac, V. (2007). Building energy performance simulation tools - a life-cycle and interoperable perspective. CIFE Working Paper #WP107. Stanford University: Center for Integrated Facility Engineering. McClure, S. (1998). Landscaping to conserve energy. Chicago Sun-Times, December 6, 1998. 121 McPherson, E. G. (2004). Carbon dioxide reduction through urban forestry: guideline for professional and volunteer tree planters. Arboriculture & Urban Forestry 30 (6): pp. 380 (accessed 3/5/2012). McPherson, E. G., Simpson, J. R. (2003). Potential energy savings in buildings by an urban tree planting programme in california. Urban Forestry & Urban Greening 2 (2): pp. 73-86. McPherson, E. G. (1998). Structure and sustainability of Sacramento's urban forest. Journal of Arboriculture 24 (4): pp. 174-190. McPherson, E. G. (1994). Chicago's urban forest ecosystem: Results of the chicago urban forest climate project. Radnor, PA: U.S. Department of Agriculture, Forest Service, Northeastern Forest Experiment Station, Rep. NE-186. McPherson, E. G., Herrington, L. P., Heisler, G. M. (1988). Impacts of vegetation on residential heating and cooling. Energy and Buildings 12 (1): pp. 41-51. McPherson, E. G., Brown, R., Rowntree, R.A., (1985). Simulating tree shadow patterns for building energy analysis. Paper presented at National Passive Solar Conference, Boulder, CO. McPherson, E. G. (1984). Planting Design for Solar Control, Chapter 8 in Energy- conserving site design. Washington, DC: American Society of Landscape Architects, pp. 141-64. McPherson, E. G. (1981). Methodology for locating and selecting trees for solar control in utah. United States: Utah State Univ., Logan, CONF-810509- (Vol.1). Milne, M., Gomez, C. F., Epailly, G. (1996). Picturing the hidden environmental benefits of passive building design decisions. Paper presented at ASES-96: Proceedings of the 1996 American Solar Energy Society Conference, Ashville, North Carolina. Moffat, A. S., and Schiler, M. (1981). Landscape design that saves energy. New York: Willam Morrow and Company, INC. 122 Osborne, A. M. (2009) Landscape & building solar loads: Development of a computer-based tool to aid in the design of landscape to reduce solar gain and energy consumption in low-rise residential buildings. Los Angeles: University of Southern California. https://libproxy.usc.edu/login?url=http://search.proquest.com/docview/30 4997006?accountid=14749. Parker, J. H. (1983). Landscaping to reduce the energy used in cooling buildings. Journal of Forestry 81 (2): pp. 82. Saxena, M. et al. (Heschong Mahone Group). (2011). Office daylighting potential.Public Interest Energy Research Program final project report. Prepared for California Energy Commission. Schiler, M., Greenberg, P. D. (1979). Computer simulation of foliage shading in building energy loads. Paper presented at DAC '79 Proceedings of the 16th Design Automation Conference. Simpson, J. R. (2002). Improved estimates of tree-shade effects on residential energy use. Energy and Buildings 34 (10) (11): pp. 1067-76. Simpson, J. R., McPherson, E. G. (1998). Simulation of tree shade impacts on residential energy use for space conditioning in Sacramento. Atmospheric Environment 32 (1) (1): pp. 69-74. Taha, H. G. (1990) An urban microclimate model for site-specific building energy simulation: Boundary layers, urban canyon, and building conditions. University of California, Berkeley, https://libproxy.usc.edu/login?url=http://search.proquest.com.libproxy.usc. edu/docview/303819993?accountid=14749. Thayler, R., Maeda, B. (1985) Measuring street tree impact on solar performance: a five-climate computer modeling study. Journal of Arboric 11: pp. 1-12 Tupper, K., Franconi, E., Chan, C., Fluhrer, C., Jenkins, M., Hodgin, S. (2011). Pre-read for BEM innovation summit. Rocky Mountain Institute, http://rmi.org/Content/Files/Summit_PreRead_Apr-19-2011(2).pdf (accessed september 21, 2011). 123 U.S. Department of Energy (2011) Building energy software tools directory. in U.S. Department of Energy http://apps1.eere.energy.gov/buildings/tools_directory/software.cfm/ID=44 0/pagename=alpha_list. U.S. Energy Information Administration (2009) Annual energy review. in U.S. Department of Energy, www.eia.gov/aer. Winkelmann, F. C., and Selkowitz, S. (1985). Daylighting simulation in the DOE-2 building energy analysis program. Energy and Buildings 8 (4) (12): pp. 271- 86 (accessed 5/4/2012 ). . 124 APPENDIX A-California climate zones Table 31: California climate zone location 125 APPENDIX B- Comparision test result The Estimator and eQuest simulation result for section 5.1: Comparison to Tree Benefit Estimator Table 32: Estimator result: annual absolute energy saved (kWh) Orientation Number of the trees South North East West 0 0 0 0 0 1 77 5 73 113 2 155 9 147 225 3 232 14 220 338 4 309 18 293 451 5 386 23 366 564 6 464 27 440 676 Table 33: eQuest result: annual absolute energy saved (kWh) Orientation Number of the trees South North East West 0 0 0 0 0 1 47 3 48 125 2 86 4 184 177 3 100 8 280 280 4 144 9 289 361 5 163 13 306 363 6 172 13 311 365 126 APPENDIX C- Case study simulation The eQuest simulation result for section 5.2: Comparison to a previous method in DOE-2 Unshaded scenario eQuest report: 127 Shaded scenario eQuest report: 128 Unshaded and shaded result comparision: Table 34: Simulation in eQuest result June July August Sep Oct Days in the month 30 31 31 30 31 Unshade Monthly cooling energy (kWh) 880 1050 980 710 430 Average daily use (kWh) 29.33 33.87 31.61 23.67 13.87 Shade Monthly cooling energy (kWh) 760 900 830 590 330 Average daily use (kWh) 25.33 29.03 26.77 19.67 10.65 Cooling energy saved (kWh) 4 4.84 4.84 4 3.23 Percentage energy saving (%) 13.64 14.23 15.31 16.90 23.23 129 APPENDIX D- Daylighting simulation output No tree: Pro je c t / Run: 20 1 2 - c a s e s tu d y 1 - Ba s e l i n e De s i g n R u n Da te / Ti m e: 06 / 27 / 12 @ 13 : 53 eQU E ST 3 . 64 . 71 3 0 A n n u a l E n e r g y C o n s u m p ti o n b y E n d u s e Pag e 1 E l e c tr i c i ty 9% 33% 2% 18% 37% N a tur a l G a s 28% 72% A re a L i g ht i ng T a s k L i g ht i ng M i s c . E q ui p m e nt E xt e ri o r U s a g e P um p s & A ux. V e nt i l a t i o n F a ns W a t e r H e a t i ng H t P um p S up p . S p a c e H e a t i ng R e f ri g e ra t i o n H e a t R e je c t i o n S p a c e C o o l i ng A nnua l E ne r gy C onsum pt i on by E nduse El e ct r i ci t y Natu r al G as Ste am Ch i l l e d W a te r kW h Btu ( x 000 ) Btu Btu S p a c e C o o l 4 , 771 - - - H e a t R e j e c t . - - - - R e fr i g e r a t i o n - - - - S p a c e H e a t - 15 , 670 - - H P S u p p . - - - - H o t Wa t e r - 6 , 066 - - V e n t . Fan s 2 , 277 - - - Pu mp s & Aux . 250 - - - E x t . Usage - - - - M i s c . Equ i p . 4 , 261 - - - T a s k L i g h t s 175 - - - A r e a L i g h t s 1 , 181 - - - T ot a l 12 , 916 21 , 737 - - 130 Trees with reflectance value 0.5: Pro je c t / Run: 20 1 2 - c a s e s tu d y 1 - Ba s e l i n e De s i g n R u n Da te / Ti m e: 06 / 27 / 12 @ 14 : 02 eQU E ST 3 . 64 . 71 3 0 A n n u a l E n e r g y C o n s u m p ti o n b y E n d u s e Pag e 1 E l e c tr i c i ty 11% 2% 39% 2% 15% 32% N a tur a l G a s 22% 78% A re a L i g ht i ng T a s k L i g ht i ng M i s c . E q ui p m e nt E xt e ri o r U s a g e P um p s & A ux. V e nt i l a t i o n F a ns W a t e r H e a t i ng H t P um p S up p . S p a c e H e a t i ng R e f ri g e ra t i o n H e a t R e je c t i o n S p a c e C o o l i ng A nnua l E ne r gy C onsum pt i on by E nduse El e ct r i ci t y Natu r al G as Ste am Ch i l l e d W a te r kW h Btu ( x 000 ) Btu Btu S p a c e C o o l 3 , 503 - - - H e a t R e j e c t . - - - - R e fr i g e r a t i o n - - - - S p a c e H e a t - 22 , 117 - - H P S u p p . - - - - H o t Wa t e r - 6 , 083 - - V e n t . Fan s 1 , 599 - - - Pu mp s & Aux . 250 - - - E x t . Usage - - - - M i s c . Equ i p . 4 , 261 - - - T a s k L i g h t s 175 - - - A r e a L i g h t s 1 , 229 - - - T ot a l 11 , 018 28 , 200 - - 131 Trees with reflectance value 1.0: Pro je c t / Run: 20 1 2 - c a s e s tu d y 1 - Ba s e l i n e De s i g n R u n Da te / Ti m e: 06 / 27 / 12 @ 14 : 05 eQU E ST 3 . 64 . 71 3 0 A n n u a l E n e r g y C o n s u m p ti o n b y E n d u s e Pag e 1 E l e c tr i c i ty 11% 2% 39% 2% 15% 32% N a tur a l G a s 22% 78% A re a L i g ht i ng T a s k L i g ht i ng M i s c . E q ui p m e nt E xt e ri o r U s a g e P um p s & A ux. V e nt i l a t i o n F a ns W a t e r H e a t i ng H t P um p S up p . S p a c e H e a t i ng R e f ri g e ra t i o n H e a t R e je c t i o n S p a c e C o o l i ng A nnua l E ne r gy C onsum pt i on by E nduse El e ct r i ci t y Natu r al G as Ste am Ch i l l e d W a te r kW h Btu ( x 000 ) Btu Btu S p a c e C o o l 3 , 503 - - - H e a t R e j e c t . - - - - R e fr i g e r a t i o n - - - - S p a c e H e a t - 22 , 117 - - H P S u p p . - - - - H o t Wa t e r - 6 , 083 - - V e n t . Fan s 1 , 599 - - - Pu mp s & Aux . 250 - - - E x t . Usage - - - - M i s c . Equ i p . 4 , 261 - - - T a s k L i g h t s 175 - - - A r e a L i g h t s 1 , 229 - - - T ot a l 11 , 018 28 , 200 - - 132 APPENDIX E-Schedule “GndPrm Occ Sch” definition in eQuest Monday to Friday schedule: Saturday, Sunday and Holiday schedule: 133 Annual schedule: 134 APPENDIX F- Daylighting simulation result Casestudy house: Table 35: Illuminance level, no trees No trees SE Perim Spc NW Perim Spc Month Average daylighting illuminance (footcandles) Percent hours daylighting illuminance above set point Average daylighting illuminance (footcandles) Percent hours daylighting illuminance above set point July 111.6 73.3 115.7 86.7 Aug 120.0 76.6 112.1 79.1 Sep 135.9 77.4 100.3 76.2 Oct 159.2 78.0 87.0 67.5 Nov 163.5 79.4 72.9 57.7 Dec 23.1 0.0 27.8 0.0 Table 36: Illuminance level, trees (reflectance 0.5) Trees (Reflectance 0.5) SE Perim Spc NW Perim Spc Month Average daylighting illuminance (footcandles) Percent hours daylighting illuminance above set point Average daylighting illuminance (footcandles) Percent hours daylighting illuminance above set point July 108.3 73.3 82.2 80.0 Aug 114.3 76.6 77.3 78.9 Sep 119 76.9 64.8 73.9 Oct 109.7 75.5 53.4 64.0 Nov 91.3 77.8 44.7 57.7 Dec 23.1 0.0 27.8 0.0 135 Table 37: Illuminance level, trees (reflectance 1.0) Trees (Reflectance 1.0) SE Perim Spc NW Perim Spc Month Average daylighting illuminance (footcandles) Percent hours daylighting illuminance above set point Average daylighting illuminance (footcandles) Percent hours daylighting illuminance above set point July 108.3 73.3 82.2 80.0 Aug 114.3 76.6 77.3 78.9 Sep 119 76.9 64.8 73.9 Oct 109.7 75.5 53.4 64.0 Nov 91.3 77.8 44.7 57.7 Dec 23.1 0.0 27.8 0.0 Simple box room: No Trees Trees (reflectance 0.5) Trees (reflectance 1.0) Month Average daylighting illuminance (footcandles) Average daylighting illuminance (footcandles) Average daylighting illuminance (footcandles) July 49.6 11.9 11.9 Aug 53.5 12.6 12.6 Sep 58.1 13.5 13.5 Oct 76 14.4 14.4 Nov 129.4 14.0 14.0 Dec 27.6 4.1 4.1

Abstract (if available)

Abstract Landscaping in general and trees specifically can be beneficial in helping to mitigate several environment problems such as carbon sequestration, urban hear island, reduced air quality due to pollution, and erosion. Yet simulation software programs are often weak in enabling designers to understand analytically and to specifically predict energy savings through the use of landscaping. ❧ The assessment done on existing energy simulation software shows that several programs could not model trees directly. Previous studies have looked at the effect of shade trees on energy use. Different strategies were used to try to model trees in these software programs, and simple case studies were undertaken to verify the results. Two critical potential energy saving features of trees were studied: direct shading on a wall and/or window and evapotranspiration. Shade trees can help in reducing solar gain and thus reduce energy consumption for cooling and should be taken seriously as a climate change adaptation initiative. ❧ The evapotranspiration of shade trees can also change both air temperature and relative humidity in the micro-climate. The evapotranspiration impact of trees has been simulated in DOE-2

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Energy simulation in existing buildings: calibrating the model for retrofit studies

PDF

Energy efficient buildings: a method of probabilistic risk assessment using building energy simulation

PDF

Microclimate and building energy performance

PDF

Landscape and building solar loads: development of a computer-based tool to aid in the design of landscape to reduce solar gain and energy consumption in low-rise residential buildings

PDF

The effectiveness of enviro-materially actuated kinetic facades: evaluating the thermal performance of thermo-bimetal shading component geometries

PDF

Net zero energy building: the integration of design strategies and PVs for zero-energy consumption

PDF

Daylight prediction: an evaluation of daylighting simulation software for four cases

PDF

Zero net energy institutional building

PDF

Energy savings by using dynamic environmental controls in the cavity of double skin facades

PDF

Harvesting energy with solar panels and adaptive shading for building skins: A case study of an office building in Saudi Arabia

PDF

District energy systems: Studying building types at an urban scale to understand building energy consumption and waste energy generation

PDF

A simplified building energy simulation tool: material and environmental properties effects on HVAC performance

PDF

Occupant-aware energy management: energy saving and comfort outcomes achievable through application of cooling setpoint adjustments

PDF

Energy saving integrated facade: design and analysis using computer simulation

PDF

Simulation-based electric lighting control algorithm: integrating daylight simulation with daylight-linked lighting control

PDF

Post occupancy energy evaluation of Ronald Tutor Hall using eQUEST; computer based simulation of existing building and comparison of data

PDF

Facade retrofit: enhancing energy performance in existing buildings

PDF

Studies in preliminary design of fenestration: balancing daylight harvesting and energy consumption

PDF

Comparing visual comfort metrics for fourteen spaces using simulation-based luminance mapping

PDF

Energy use intensity estimation method based on building façade features by using regression models

Asset Metadata

Creator Cheng, Yi-Lun (author)

Core Title Digital tree simulation for residential building energy savings: shading and evapotranspiration

School School of Architecture

Degree Master of Building Science

Degree Program Building Science

Publication Date 07/31/2012

Defense Date 07/02/2012

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag building shade,DOE-2,energy saving,energy simulation,eQuest,evapotranspiration,Foliage,Landscape,OAI-PMH Harvest,shading,Tree

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Kensek, Karen M. (committee chair), Carlson, Anders (committee member), Schiler, Marc (committee member)

Creator Email yilunc@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-77812

Unique identifier UC11288772

Identifier usctheses-c3-77812 (legacy record id)

Legacy Identifier etd-ChengYiLun-1073.pdf

Dmrecord 77812

Document Type Thesis

Rights Cheng, Yi-Lun

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA