A set of three instructional modules is used to demonstrate BIM-based methods of building energy analysis and simulation. The workflow leverages existing data in the architectural model for energy analysis, as opposed to creating separate models within each analysis program. Autodesk Revit Architecture is used to create a building information model based on a residential dwelling designed by Frank Lloyd Wright. Relevant data from the building information model is exported to two separate analysis tools - Autodesk Ecotect and DesignBuilder. Ecotect is used to provide early feedback on weather, heating and cooling loads, and heat loss information, and DesignBuilder is used to evaluate the effect of adding various energy conservation measures (ECMs). The two programs are used in conjunction to analyze the effects of solar orientation and passive solar design elements.
The BIM data is able to successfully generate the two external energy analysis models. Some issues are encountered in the process of importing BIM data into the analysis tools; however, all issues are able to be resolved. The features of the BIM workflow found to add the most value are the ability to export building geometry and thermal zone information and the ability to use material take-offs for comparing the cost of design alternatives. Simulation results from Ecotect and DesignBuilder show a 15% difference in annual heating and cooling energy between the best case and worst case passive solar design strategies. Results from the DesignBuilder ECM study indicate that a total energy savings of 21% can be achieved cost-effectively compared to a baseline design, for a 10-year time horizon. iii
TABLE OF CONTENTS
CHAPTER 1: INTRODUCTION ................................................................................................................. 1
1.1 Background ......................................................................................................................................... 1
1.2 Purpose and Objectives ....................................................................................................................... 2
1.3 Organization ........................................................................................................................................ 2
CHAPTER 2: LITERATURE REVIEW ...................................................................................................... 4
2.1 Energy Consumed by the Built Environment ..................................................................................... 4
2.2 Building Information Modeling (BIM) ............................................................................................... 6
2.3 Building Energy Simulation.............................................................................................................. 11
CHAPTER 3: METHODOLOGY .............................................................................................................. 18
3.1 Software Selection ............................................................................................................................ 18
3.2 Building Selection ............................................................................................................................. 18
3.3 Baseline Energy Model Specification ............................................................................................... 18
CHAPTER 4: INSTRUCTIONAL MODULE CONTENT ........................................................................ 24
4.1 Summary ........................................................................................................................................... 24
4.2 Module 1: Building Model Creation ................................................................................................. 25
4.3 Module 2: Data Exchange and Energy Model Preparation ............................................................... 31
4.4 Module 3: Simulation and Data Analysis ......................................................................................... 46
CHAPTER 5: CONCLUSIONS AND RECOMMENDATIONS .............................................................. 65
5.1 Conclusions ....................................................................................................................................... 65
5.2 Recommendations ............................................................................................................................. 65
REFERENCES ........................................................................................................................................... 67
APPENDIX A ............................................................................................................................................. 70
APPENDIX B ............................................................................................................................................. 94 1
CHAPTER 1: INTRODUCTION
Residential and commercial buildings consume a significant amount of energy in the United States. In 2006, the built environment consumed 39% of all primary energy and 72% of all electricity (1), (2). In addition, buildings contributed 38% of all carbon dioxide emissions, 55% of sulfur dioxide, and 18% of nitrous oxides (3), (4).
Luckily, the opportunities to reduce the energy and environmental impacts of the built environment are equally as great. The U.S. Department of Energy's Building America program estimates that the energy consumption of new homes in the U.S. can be reduced by 40% or more with little to no impact on the cost of ownership (5). Similar programs for commercial buildings have shown energy savings of 50% for new construction and 30% for retrofits of existing building stock (6).
In 2008, a team of students and faculty from the University of Illinois at Urbana-Champaign participated in a study of energy efficient and net-zero energy buildings. Students and faculty were selected from both the engineering and architecture schools. The research study considered the possibility of an international prize competition for cost-effective, highly energy efficient homes. The competition would be similar to those organized by the well-known X-PRIZE Foundation (7). The students from both disciplines were asked to collaborate on many tasks associated with the project, including: assisting in the development of rules and performance metrics for the competition, assessing typical energy use and energy usage patterns in American homes, completing an energy audit of a local home in the Champaign-Urbana area, defining a typical "reference home" which could serve as a benchmark for energy analysis studies, and developing energy simulation models of the reference home and other more efficient design alternatives.
While many research studies have addressed specific components of building energy, few studies have provided a comprehensive demonstration of the tools and technologies that enable collaborative 2
design of energy efficient buildings. A deep multi-disciplinary understanding of building energy is critical for all who are involved in the field: professional engineers and architects who design the buildings that people live and work in, academic institutions that provide further improvement in energy simulation and decision-making tools, and students who are in building energy related disciplines and participate in competitions such as the Solar Decathlon (8).
1.2 Purpose and Objectives
This thesis addresses the limited availability of resources which provide a comprehensive discussion and demonstration of building energy modeling from a multi-disciplinary perspective. The overall objectives of this research are to: 1) take the fundamental approach of the aforementioned research study and apply it to this work, 2) discuss the capabilities of building information modeling (BIM) and building energy simulation programs in evaluating many alternatives early in the design process, 3) demonstrate a logical procedure for creating a building information model, analyzing its energy use through simulation, and evaluating its performance 4) develop a set of instructional modules that follow the procedure outlined in objective 3, and 5) make the modules freely available to all academic institutions to promote the collaboration of various disciplines in the design of energy efficient buildings.
This thesis is divided into five chapters, with additional appendices included at the end. Chapter 1 provides background information, as well as the purpose and motivation of the study.
Chapter 2 provides a review of building energy, BIM, and building energy simulation. Section 2.1 describes building energy consumed in the United States. Typical end-uses of building energy are discussed, as well as the distribution of energy end-uses common in residential and commercial buildings. Section 2.2 addresses BIM. A brief history of BIM is given along with a summary of programs that are currently available. The benefits of integrating BIM into the design process are addressed, as well as the challenges of BIM to achieve a more wide-spread adoption. A special emphasis is placed on the ability to 3
leverage the BIM model for the purpose of building energy simulation. Section 2.3 provides an overview of building energy simulation. The most popular computational methods used in building energy simulation programs are discussed. A summary of current simulation software is given, along with a comparison of software features and limitations.
Chapter 3 provides the methodology of the study. A floor plan is selected for the building that is the basis of the study. A building information modeling program is selected, as well as two different external energy analysis tools. A "baseline" building is defined as a source of comparison for all other "alternative" design scenarios that include some sort of energy efficiency strategy. Most of the parameters for the baseline building are taken from the Building America Research Benchmark. Tables provide the baseline parameters that are specific to the site and building.
Chapter 4 provides content for instructional modules demonstrating building energy analysis using a building information model. The content is broken into three parts, or modules. Module 1 outlines the creation of the building model in a building information modeling program. In this module, the building envelope is defined as well as the layout of interior walls and spaces. Module 2 covers building energy simulation. The building from Module 1 is prepared for export by defining thermal zones and other export parameters. The building is then imported into two separate external energy analysis programs where the building's energy performance can be analyzed. The set-up procedures necessary for energy simulation are addressed, in addition to potential data quality issues that may be encountered in the import process. Module 3 describes the simulation process and how to evaluate the energy performance of the baseline and alternative designs. Energy performance is compared based on annual fuel consumption. The cost performance of design alternatives is judged based on the net present value of avoided energy costs compared to the baseline building.
Chapter 5 draws conclusions from the building study and offers recommendations for future work. Additional appendices provide module presentation slides and supplementary information. 4
TV and office equipment8%
Clothes washer/dryer and dishwasher6%
Other small electric devices4%
CHAPTER 2: LITERATURE REVIEW
2.1 Energy Consumed by the Built Environment
In 2005, there were 108 million residential households in the U.S. The majority of homes (71%) were single-family, detached dwellings with an average living space of 2,456 square feet. The remaining households were either multi-family units or mobile homes. Figure 1 shows a typical distribution of energy end-uses for a residential building. As the figure illustrates, the energy required by a home's heating, ventilation, and air conditioning (HVAC) system accounts for the greatest share (39.4%) of total home energy use, followed by hot water heating (12.5%), and lighting (11.6%). The remaining energy is consumed by major appliances (refrigerator, clothes washer and dryer, dishwasher, cooking appliances) and various electric plug loads (9).
Figure 1: Energy end-use splits for residential homes in the U.S. The adjustment factor accounts for discrepanices between data sources. 5
The U.S. commercial building stock is made up of 4.7 million buildings totaling 68.5 billion square feet of floor space. Office buildings account for the greatest percentage of commercial space (17%), followed by retail buildings and malls (16%), education facilities (14%), warehouse and storage facilities (14%), and others. Energy end-uses for commercial buildings look considerably different than residential dwellings. As shown in Figure 2, the largest end-use for commercial buildings is lighting (24.8%). Other major end-uses include space heating and cooling (25%), office equipment such as electronics and computers (11%), and ventilation (7%) (10).
Figure 2: Energy end-use splits for commercial buildings in the U.S. The adjustment factor accounts for discrepancies between data sources.
Understanding the difference between residential and commercial energy end-use profiles is critical for determining a strategy for energy savings. Typically, the internal heat gains (gains attributed to lighting, equipment, and building occupants) of commercial buildings tend to dominate heat gains through the building's envelope, or skin. Thus, commercial building energy savings are realized 6
primarily from the reduction of internal gains. For residential buildings, typically the opposite is true. Residential buildings tend to be skin-dominated, meaning that heat gains through the building skin are more significant than internal heat gains from lighting, equipment, and occupants (11). For this reason, residential energy savings are realized primarily through improvements to the building envelope. Although these trends are true in general, factors such as climate can significantly impact a building's energy end-use profile.
2.2 Building Information Modeling (BIM)
2.2.1 History of BIM
The construction industry traditionally involves bringing together multiple disciplines in one project. A successful project implementation requires a high level of collaboration and coordination among architecture, engineering, and construction (AEC) personnel. The highly fragmented nature of the design, build, and management process has resulted in a variety of information sources within each discipline (12). For this reason, a comprehensive representation of all necessary information has long been an objective for firms in the AEC sector (13).
Building information modeling applications evolved from the computer aided design (CAD) and computer aided manufacturing (CAM) systems that were first developed for the aeronautical and automotive industries (14). During the 1970's, CAD vendors released low-level data exchange applications targeted toward the construction industry. These early applications were written in proprietary file formats, so the integration of multiple applications and information sources was limited. The drawbacks of the proprietary formats led to the development of common formats, such as the Interim Graphics Exchange Specification (IGES) format and the Drawing eXchange Format (DXF). Perhaps the most influential common format was the Standard for the Exchange of Product Model Data (STEP) format developed in 1984 by the International Standards Organization (ISO). The STEP format was developed with the intention of improving communication of engineering information and enabling integration through the coordination of open standards for data exchange and data sharing (13). 7
Early CAD programs used explicit, two-dimensional (2D) coordinate-based geometries to create graphic entities. The 2D programs eventually evolved to three-dimensional (3D) surface and solid modeling programs. These programs were more intelligent and enabled the creation of more complex forms, however, the explicit, coordinate-based geometries made editing difficult. Parametric solid modeling was the breakthrough that enabled easier editing through the use of model parameters and constraints. The technology was adopted early on by the mechanical design world, but lagged in the area of building design due to issues of scaling (15). Eventually, parametric modeling was implemented for building models. By the mid-1990's, building model applications expanded the 3D model beyond vectorial data to include complementary data, such as physical characteristics, unit costs, quantity take-offs, and more. These innovations created what is now recognized as the BIM (14).
Only recently has BIM emerged in real world projects. Before 2000, most BIM use was limited to academic institutions. Since 2002, the term "building information model" has become widespread (16). A variety of BIM applications are currently available, including Graphisoft ArchiCAD, Bentley Architecture, and Autodesk Revit. A 2007 survey of industry professionals asked which BIM solutions they were currently using or evaluating. Revit was most popular, used by 67.1% of respondents, AchiCAD was second with 31.7%, and Bentley was third with 14.8% (17).
2.2.2 Capabilities and Benefits of BIM Applications
The modern building information model can be leveraged for a multitude of purposes throughout a building's entire lifecycle. In the early design phase, BIM applications can be used to create the building's form. Some applications allow the user to create a building using either a bottom-up or top-down design approach. For the bottom-up approach, a user creates the building on an element by element basis. Typically in a 2D plan view, walls are placed followed by floors, ceilings, etc. until the building's overall shape is defined (Figure 3). Top-down modeling takes an opposite approach. The building's overall shape is first defined, and then building elements, such as walls, floors, and walls, are applied to either interior or exterior surfaces (Figure 4). The use of top-down modeling at the early design stage can 8
be advantageous, especially for building energy simulation. A building's shape can impact its energy performance, thus a simple massing model allows the user to quickly compare the relative energy performance of different building forms without specifying the details required by a bottom-up design approach (18).
Figure 3: For a bottom-up design approach, the building's exterior and interior walls are defined in a 2D view (left image), followed by windows and doors (middle), and finally, floors, ceilings, and a roof (right).
Figure 4: For a top-down approach, the building's overall shape is defined in a 3D view (left image), followed by floors (middle), and finally, walls, windows, and a roof (right).
Building energy simulation is one of the major benefits afforded by the BIM, both in the early stages of design and later stages of analysis. The common green building Xtensible Markup Language (gbXML) format allows the export of BIM data to a building energy simulation program (19). The building energy simulation program can provide useful insight into changes that can be made for improved energy performance. The parametric nature of most BIM applications allows the building model to be quickly updated with the design changes suggested by the energy simulation. 9
The capabilities of BIM applications extend far beyond geometric modeling and energy simulation. The BIM can be used for a host of other purposes throughout a building's lifecycle, including:
ï‚· Visualization: The BIM can be used to create plan views, section views, and photo-realistic 3D renderings. Some applications allow additional visualization features such as virtual walk-throughs and simulated construction sequencing (20).
ï‚· Fabrication/shop drawings: Drawings for components requiring fabrication, such as HVAC ductwork, can be extracted from the BIM.
ï‚· Assembly instruction: Assembly notes can be attached to the pertinent elements in the BIM (14).
ï‚· Cost estimation: Quantitative information related to the building's elements can be linked to a cost database, which can produce cost estimates. Estimates are updated automatically if changes are made to the building.
ï‚· Conflict and collision detection: Architectural, structural, and HVAC sub-models can be combined into a composite model to check for the improper overlapping of building elements.
ï‚· Construction sequencing: The BIM can be used to create material ordering, fabrication, and delivery schedules for all building elements.
ï‚· Code reviews: The building model can be checked for compliance with fire and building codes.
ï‚· Forensic analysis: The BIM can be used to identify potential leaks and failures as well as possible evacuation plans.
ï‚· Facilities management: Building owners and operators can use the building model for renovations, space planning, and maintenance operations (20).
The benefits of BIM implementation across all disciplines of the AEC sector are well documented in the literature. The most direct benefit of a BIM is an accurate geometrical representation of the building in an integrated data environment. Other benefits can include faster and more effective processes, controlled whole-life costs, better production quality, automated assembly, better customer 10
service, and better design through rigorous analysis and simulation (20). A study of 32 building projects conducted by the Stanford University Center for Integrated Facilities Engineering (CIFE) found that the use of BIM could save both time and money. BIM projects eliminated 40% of unbudgeted costs and saved 10% of contract values through early collision detection. Cost estimation times were reduced by up to 80% and total project times reduced by up to 7% (21).
2.2.3 Implementation Challenges
188.8.131.52 Data standardization and platform integration
According to a survey of building practitioners, the integration (or lack of) of BIM software is seen as a barrier to its widespread adoption (22). While some vendors currently offer a suite of BIM tools that work well amongst themselves, it is still uncommon for applications from multiple vendors to have a high level of interoperability. This is due, in part, to issues of data format standardization that still exist in the industry. Over the past decade, organizations around the world have been established to further the development of neutral CAD data exchange formats (14). Currently, much of the effort in BIM standardization is focused on the Industry Foundation Classes (IFC) and the CIMSteel Integration Standards. Both standards are neutral data formats and are defined using the ISO STEP description methods (12).
184.108.40.206 Education and skill
Widespread BIM adoption is challenged by a deficiency of personnel with BIM skills. A survey of BIM use identified lack of adequate training as the most significant impediment to BIM adoption. Courses on BIM have been slow to enter engineering curricula. While 3D parametric solid modeling has become the main design medium in mechanical engineering and industrial design education, civil engineering programs have been less likely to include 3D modeling in the curriculum, especially at the undergraduate level (23). 11
2.3 Building Energy Simulation
Architects and engineers have performed building energy analyses since before the time of personal computers. At that time, analyses relied on rule-of-thumb methods and extrapolations in cases where design concepts were unconventional. This approach frequently led to incorrectly sized building equipment, and as a result, poor energy performance. Computer-based building energy simulations began in the 1960's, and became increasingly popular in the research community by the 1970's. In the late 70's and early 80's, advancements in desktop personal computing enabled the development of building simulation programs as they are known today. These programs attempted to imitate physical conditions by treating time as the independent variable and solving a series of equation sets in discrete steps. Due to the level of difficulty and high cost of early programs, most simulation projects were confined to research laboratories. It was not until the 1990's that building energy simulation was adopted into professional practices (24).
In the time since the first applications, hundreds of building energy simulation programs (BESPs) have been developed. The capabilities and intended uses of programs vary widely. The proper selection of tools is complicated by the fact that a common language for describing what tools can do does not exist (25). Furthermore, there appears to be little consensus on how BESPs should be classified and organized. Hong et. al. group simulation programs into two categories: design tools (DTs) and detailed simulation programs (DSPs). DTs are defined as purpose-specific, early design phase tools that require simple and relatively little input data. DSPs are considered to be tools that require more complex computational techniques (such as finite difference, finite element, state space, and transfer functions) for the calculation of building loads and energy use (24). Charron and Athientis categorize BESPs as either open architecture, "white box" tools or "black box" tools. White box tools are defined as those that allow users to examine and modify underlying source code to meet specific modeling needs, while black box tools are 12
defined as those programs that do not allow source code access to the user. According to Charron, black box programs tend to be more user friendly than white box programs (26).
There are BESPs available today that can simulate nearly every aspect of a building's energy performance. Some of the aspects that can be captured by current applications include (25):
ï‚· solar radiation, outside a building and within interior zones
ï‚· thermal zone loads and human comfort levels
ï‚· air flows due to infiltration and natural / mechanical ventilation
ï‚· lighting, equipment, and human loads
ï‚· renewable energy generation
ï‚· HVAC system behavior
ï‚· lighting controls and integrated shading devices
ï‚· environmental emissions
BESPs can also be used effectively in the early design phase to evaluate "what-if" scenarios in the search for optimal solutions. Such scenarios can investigate and assess the energy impact of:
ï‚· building shape and form
ï‚· solar orientation
ï‚· material construction of the envelope
ï‚· glazing type and placement
ï‚· HVAC system type and efficiency
2.3.2 Building Energy Simulation Programs (BESPs) Considered for this Study
DOE-2 is a BESP originally sponsored by the U.S. Department of Energy and developed by Lawrence Berkeley National Laboratory (LBNL) and James J. Hirsch & Associates (27). Since its first 13
release in 1979, DOE-2 has undergone a series of alterations and improvements. It is now used in more than 40 countries to design energy efficient buildings, develop energy conservation standards, and analyze the impact of new building technologies. It has been validated with measured data from actual buildings, and it has been used for the design and retrofit of well-known buildings such as the White House, the World Trade Center, and the Sears Tower. Over 20 interfaces have been created by private sector firms in order to make the program more user-friendly.
DOE-2 is an hourly, whole building simulation program. Whole building refers to the program's ability to model and simulate the entire building as opposed to just one building system. Hourly denotes the length of the time step at which the program's calculations are carried out. DOE-2 is internally made up of four subprograms: the Building Description Language (BDL) processor, LOADS, HVAC, and ECON. The BDL processor takes the building geometry inputted by the user and translates it into a form recognizable by the program. LOADS calculates hourly heating and cooling loads for the building's spaces. Heating and cooling loads are affected by a host of factors, including weather and solar conditions, infiltration, heat transfer through the envelope, and the scheduling of people, lighting, and equipment. HVAC sizes and simulates both primary and secondary HVAC equipment based on the demands calculated in LOADS. ECON is used to calculate cost performance of a building in terms of energy savings, design savings, or retrofit savings (28).
DOE-2's major advantage is its ability to analyze detailed, multi-zone buildings quickly and reliably. DOE-2 simulations run in a fraction of the time of other simulation programs. One disadvantage of the program is the high level of expertise required if a third party environment is not used (29).
EnergyPlus is a "new generation" BESP that combines the features of a few predecessor applications including DOE-2 and the Building Loads Analysis and System Thermodynamics (BLAST) 14
program developed by the University of Illinois at Urbana-Champaign. Like DOE-2, EnergyPlus is a whole-building energy simulation program. Unlike DOE-2, EnergyPlus has the ability to use sub-hourly time steps. This allows the program to more accurately predict space temperatures, which is crucial for system and plant sizing and human comfort calculations. EnergyPlus's simulation engine can simulate some systems and behaviors that DOE-2 and other programs cannot. These include realistic system controls, moisture adsorption and desorption in building elements, radiant heating and cooling systems, and inter-zone air flow. To calculate space temperatures, EnergyPlus uses the heat balance method. The heat balance method calculates radiation and convection processes at each time step. Although this method is more sophisticated than the methods used by other simulation packages, it is still a simplification of reality. The method assumes that air is well stirred and thus a uniform temperature throughout. A better alternative would be a computational fluid dynamics (CFD) based method; however this would add significant complexity and computational effort (30).
In addition to being able to simulate whole buildings in sub-hourly time steps, EnergyPlus has a few major advantages. Relative to other programs, it has been tested extensively and validated with actual data. Its modular structure allows for new components to be created and plugged in for simulation. Finally, it has one of the most extensive weather libraries with data available for 1250 locations worldwide. Similar to DOE-2, EnergyPlus has the disadvantage of requiring a high level of skill without the help of a third party interface and relatively long simulation times (29).
220.127.116.11 Autodesk Ecotect
Autodesk Ecotect is a flexible simulation tool that works as a stand-alone application and as a source of input for other energy simulation programs. Building geometry can be created with Ecotect's 3D modeler, or it can be imported using the gbXML or DXF formats. Ecotect provides solar, thermal, lighting, acoustic, and cost analyses that are ideal for early design stage feedback (Figure 5). For more detailed studies, building geometry and information can be exported to hourly programs such as eQuest or EnergyPlus (31). 15
Figure 5: A sun path analysis in Ecotect allows the user to see how building facades are shaded throughout the year.
EQuest is a widely used, time-proven simulation program that is freely available to the public. Its user base consists of building designers, operators, owners, energy and LEED consultants, regulatory professionals, universities and researchers. With an average of 10,000 downloads annually, it is one of the most popular simulation tools used today.
EQuest is an interface that uses DOE-2 as its simulation engine. A major advantage of the program is its ability to accommodate three different levels of user expertise. For beginners, the schematic design wizard allows a building to be created by using simple input data (Figure 6). The design development wizard allows a greater level of input for more intermediate users, and for advanced users, the detailed mode allows the manipulation of all building parameters. EQuest can accommodate single simulation runs, multiple runs, and parametric studies. Drawbacks of the program include simplified 16
models for ground coupling, infiltration, and natural ventilation. In addition, daylighting can be applied only to convex surfaces and cannot transmit through interior glazed surfaces (29).
Figure 6: An example of a two-story office building in eQuest.
DesignBuilder, similar to eQuest, is an interface combined with a powerful simulation engine, however instead of DOE-2, DesignBuilder uses an enhanced version of EnergyPlus. Building geometry can be created within DesignBuilder's OpenGL modeler, or it can be imported as a gbXML file from an architectural modeling program (Figure 7). Data inheritance allows changes to be made to the entire building, a building "block", or a single thermal zone (32). 17
Figure 7: An example residential building shown in DesignBuilder's rendering environment.
18.104.22.168 Autodesk Green Building Studio
Green Building Studio (GBS) is a web-based application that uses the DOE-2 engine for energy simulation (33). It provides information on a building's energy use, carbon footprint, water use, material costs, and more. Design alternatives can be compared side by side for quick design decisions. Building parameters can be changed in a web-based interface that is simpler to use than most other BESPs. The simple interface comes at the expense of being able to enter the level of detail afforded by other BESPs; however the GBS files can be exported to detailed hourly simulation programs. 18
CHAPTER 3: METHODOLOGY
3.1 Software Selection
Autodesk Revit Architecture 2011 is chosen as the software package for creating the building information model. Revit provides tools for both top-down and bottom-up build strategies, and accommodates the export of building data for the purpose of external energy analysis.
For energy analysis, Ecotect 2011 and DesignBuilder (version 2.2) are selected. Both programs support the ability to import building information model data for energy analysis. Ecotect is chosen for its ability to generate quick, early design studies as well as for purposes of visualization. DesignBuilder is chosen for its ability to execute detailed energy simulations using EnergyPlus.
3.2 Building Selection
The design and floor plan of the building used for this study are based on the First Jacobs House, a 1550 square foot residential building designed by Frank Lloyd Wright in 1936. The First Jacobs House was part of Wright's "Usonia" movement, which stressed the importance of home affordability and also energy efficiency. The building, located in Madison, Wisconsin, was chosen for two main reasons. One, its smaller size as a residential building compared to a commercial building made it more tractable for the purpose of a comprehensive study. Two, the analysis of a well-known design was thought to be of more interest than one created from scratch. The use of an existing design allows the interesting possibility of exploring both the energy performance of the building as designed, and the energy performance of the building with upgraded features.
3.3 Baseline Energy Model Specification
Depending on the analysis tool used and the particular study, a building energy model can require a host of parameters. These parameters may include:
ï‚· element material constructions and associated thermal properties
ï‚· HVAC and hot water system types and efficiencies
ï‚· lighting type, density, and efficiency
ï‚· building occupancy
ï‚· plug loads, such as appliances and electronic devices
ï‚· internal heat gains from plug loads and occupancy
ï‚· building natural infiltration rate (air leakage)
ï‚· natural ventilation (i.e. the opening and closing of windows)
ï‚· thermostat set-point temperatures
ï‚· operating schedules
Not all energy analysis programs require the same parameters or the same level of detail. This could reflect the fact that some programs are intended to be early-stage design tools, and thus require less detailed information. It may also reflect the lack of standardization in the industry.
The U.S. Department of Energy's Building America Research Benchmark Definition is used as a guiding standard for most baseline energy model parameters of this study (34). The goal of the Building America program is to reduce energy consumption of new homes by 40% or more with little to no impact on the cost of ownership. The Building America Research Benchmark (subsequently referred to as Research Benchmark or Benchmark) establishes baseline energy usage for residential buildings. It is generally consistent with construction practices from the mid-1990s. Using factors such as the building's type (e.g. single family attached, single family detached, multifamily), size, and location, the benchmark provides values for the energy model input parameters. These parameters include all of the ones in the list above. The number of heating degree days (HDD), base 65 degrees Fahrenheit, for the building's location is used in the calculation of certain parameters. HDDs are an index to reflect heating energy demand for a building and are equal to the summation of differences between the average daily temperature and a base temperature (typically 65Â°F) (35). For Madison, Wisconsin, 7499 HDD are 20 Building ElementU-Value (Btu/hÂ·ft2Â·Â°F)R-Value (hÂ·ft2Â·Â°F/Btu)Exterior wall0.05219.23Roof/ceiling0.02638.46Exterior doors0.2005.00Slab insulation0.1675.99Windows0.3602.78
assumed. Other parameters are based on the building's finished floor area (FFA), which in this case is 1550 square feet.
In certain cases, a standard other than the Benchmark is used for parameter definition. For example, energy usage figures for the refrigerator, clothes washer, and dishwasher are taken from Energy Star. In this case, this is done so that cost estimates for Energy Star appliances are consistent with their energy use. In other cases, a parameter specified by the Benchmark may be inconsistent with a parameter required by either Ecotect or DesignBuilder.
While the original Jacobs House was not built to the standard set by the Benchmark, it is assumed that the building is upgraded to meet modern standard practice. A summary of the major baseline energy model parameters is included below.
3.3.1 Envelope Parameters
The parameters of the baseline building envelope depend on the building's location, and thus on the number of HDD. Since a building is typically designed in the context of its environment, the location of the actual Jacobs House - Madison, WI - is chosen as the location for the energy model. Based on Madison's 7499 HDD, the Benchmark specifies the overall heat coefficients, or U-values, shown in Table 1. Also provided are the element R-values (a measure of thermal resistance) which are equal to the reciprocal of the U-values. An infiltration rate of 0.5 air changes per hour is assumed for all simulations.
Table 1: Envelope parameters for baseline energy model. 21
3.3.2 HVAC and Hot Water System Parameters
The efficiency of the baseline building's HVAC is dependent on the type of system used. Although the actual Jacobs House used radiant heating for occupant comfort, the Benchmark uses forced air systems for its standard. For that reason, the baseline building is assumed to have a natural gas furnace and electric air conditioning system. The furnace is assumed to have an Annual Fuel Utilization Efficiency (AFUE) of 78% and the air conditioner is assumed to have a Seasonal Energy Efficiency Ratio (SEER) of 10. Heating and cooling set points are maintained at 71Â°F and 76Â°F, respectively, with no setup or setback periods.
Hot water system efficiency also depends on system type. In this case, an electric tank hot water heater is assumed with an Energy Factor (EF) of 0.89 and hot water service of 50 gallons/day.
3.3.3 Appliance and Plug Load Parameters
Residential buildings contain many electric loads, both large and small. The Research Benchmark provides annual energy usages for typical loads. Energy usages for large loads, such as kitchen and laundry appliances, are given separately. Smaller loads, such as televisions, battery chargers, and toasters, are grouped into a larger miscellaneous load category. The magnitude of the loads is based on the number of bedrooms and finished floor area of the building under investigation. The appliance and plug loads of the example building are given in Table 2, based on the building having two bedrooms and approximately 1550 square feet of livable floor area. Sensible and latent heat gains associated with the electric loads should be included in the energy models as they affect the building's HVAC operation. Sensible heat gains affect air temperature while latent gains affect moisture in the air (36). 22 Appliance or Electric Load Electricity (kWh/yr) Sensible Load Fraction Latent Load Fraction Refrigerator 666 1.00 0.00 Clothes washer 333 0.80 0.00 Clothes dryer 696 0.15 0.05 Dishwasher 368 0.60 0.15 Range 504 0.40 0.30 Variable Miscellaneous 2218 0.83 0.02 Fixed Miscellaneous 556 0.12 0.23 Lighting Category Electricity (kWh/yr) Interior hard-wired 1305 Interior plug-in 326 Exterior 268
Table 2: Baseline energy model parameters for appliances and electric loads.
3.3.4 Lighting Parameters
The Benchmark provides a simple and a complex method for specifying baseline lighting loads. Here, the simple method is chosen. Annual energy usages can be calculated for interior hard-wired, interior plug-in, garage, and exterior lighting. Since the building used in this study does not have a garage, this fraction of the lighting is omitted. The annual energy usages for the other three categories is given in Table 3, based on the finished floor area of the building. The baseline energy usages assume that 86% of all lights are incandescent and 14% are compact fluorescent (CFL). The lighting loads are operated according to a schedule, which is included in the Benchmark and addressed in the following section.
Table 3: Baseline energy model lighting parameters.
3.3.5 Schedule Parameters
Many of an energy model's components operate according to a schedule. Energy model schedules are generally specified for a 24-hour period. Each hour, the component uses a certain fraction or percentage of, for example, its total daily energy use. This results in a schedule such as the one in Figure 8, showing the schedule for a cooking range. Note that the peaks of the graphed line correspond to 23
the times of the day when the range would most likely be used (e.g. during dinner time). Schedules can be specified differently based on whether the day is during the week, weekend, or holiday. For this study, however, schedules are assumed to be the same for every day of the year.
Figure 8: Daily energy usage profile for a cooking range. 24
CHAPTER 4: INSTRUCTIONAL MODULE CONTENT
In Module 1, the process of creating a building information model is covered. The process starts with a simple 2D sketch and ends with a 3D model. The sketch is drawn on a tablet PC using an existing floor plan for guidance. Using the sketch, 2D blocks are drawn in a building information modeling program and extruded to create a 3D massing model. Some building elements are applied by adding them directly to the mass surfaces, and others are defined independently of the massing model. A camera is placed in a 3D model view to generate a rendered image of the final building model.
In Module 2, building information model data is prepared for export. Thermal zones are defined in the building's spaces, and zone boundaries are checked in all three dimensions to ensure proper export. Export options are addressed, and a gbXML file of the building is generated. The gbXML data is imported into two programs - Ecotect, an early design analysis tool, and DesignBuilder, a detailed, sub-hourly simulation program. The import process for both programs is discussed, as well as data quality issues that may be encountered in the process. Analysis model parameters are defined using the Building America Research Benchmark for guidance.
In Module 3, Ecotect and DesignBuilder models are used to conduct energy analyses and simulations. First, early design studies are performed in Ecotect to indicate the relative impact of design factors on energy performance. A weather tool is used to investigate the type of environment in which the building exists. Heating and cooling load calculations are used to provide a relative comparison of the building's heating and cooling requirements. Heat gains and losses are graphed to indicate which factors contribute the most to HVAC energy use. Second, Ecotect and DesignBuilder are used in conjunction for solar orientation and passive solar design studies. Ecotect is used to visualize levels of solar radiation and solar ray penetration for varying building orientations and passive solar design strategies. DesignBuilder is used to quantify these design decisions by performing detailed annual simulations of each scenario. The individual and collective impact of passive design elements are explored. Lastly, DesignBuilder is 25
used to simulate the effect of adding various energy conservation measures (ECMs) to the baseline design, including increased envelope insulation, higher performance glazing, and higher efficiency lighting, appliances, and mechanical systems. An economic analysis is included to provide an insight into cost-effective design solutions.
4.2 Module 1: Building Model Creation
4.2.1 Digital Sketching
A floor plan of the Jacobs House, shown in Figure 9, was found online. The floor plan indicates the location of walls, windows, and doors, as well as the placement of rooms and the extent of roof overhangs. Photographs of the building reveal three distinct roofline elevations. As a result, the building's external walls are not a constant height. In order to visualize the location of the different wall heights, a tablet PC was used to bring the floor plan into a digital sketching program. Setting the floor plan to a background sketching layer, a second sketching layer was added on top of the original sketch to visualize the grouping of similar walls. The result is shown in Figure 10. The red outline indicates walls connecting to the building's lowest roof. These walls, at a height of 7Â½', are used in the private space of the building composed of the bedroom and study. The blue outline indicates the next level of walls, at 9Â½', which are found in the building's main public space, or living area. Finally, the kitchen and bathroom space, outlined in green, has the highest walls (11Â½') in the building.
A third sketching layer is added to the floor plan, consisting of the building's major dimensions, shown in Figure 11. These dimensions, listed in feet, will be used to create the building information model. To complete the digital concept sketch the floor plan background layer is hidden from view, leaving only the color-coded building blocks and block dimensions (Figure 12). 26
Figure 9: Floor plan of the Jacobs House.
Figure 10: Red (R), blue (B), and green (G) lines layered on top of the original floor plan to denote the location of the lowest, middle, and highest rooflines, respectively. 27
Figure 11: A third dimension layer is added to the floor plan to assist in the creation of the building information model.
Figure 12: The background floor plan layer is hidden from view, showing the final concept sketch to be used in creating the building information model. 28
4.2.2 Creation of the Building Information Model
The building's natural separation into three major blocks suggested that a top-down approach (Figure 4) would be ideal. The blocks can be created using conceptual mass elements within Revit Architecture. First, the three blocks can be sketched separately in a two-dimensional view. After each two-dimensional shape is created, the shape can be extruded to a given height, resulting in a three-dimensional volume. In this way, the bedroom, living area, and kitchen/bath sketches can be extruded to heights of 7Â½', 9Â½', and 11Â½', respectively. Once the three individual masses are defined, they can be joined to create a single conceptual mass. The mass surfaces can be edited to create additional details associated with the building's geometry. The finished massing model is shown in Figure 13. Red, blue, and green lines have been superimposed on the figure to show how the three-dimensional masses correspond to the original 2D concept sketch.
Figure 13: The three Revit massing objects with red (R), blue (B), and green (G) outlines superimposed.
Revit allows the addition of building elements such as floors, walls, and roofs directly to mass faces. In this example, the walls and floors are added to the building information model in this way. Figure 14 shows the massing model with the addition of brick and wooden walls to some of the mass faces. Although roof elements also can be added to mass faces as well, due to its large overhangs and atypical geometry the roof was easier to implement using a separate sketch feature. After the main envelope elements of the building are defined, additional elements such as windows, doors, and internal 29
walls can be added to the model. Figure 15 illustrates a progression of the build process. For a more detailed, step by step description of the build process, see Appendix A.
Figure 14: The Revit massing model is shown with external walls that have been added to mass faces.
When building elements are added to the model they are purposely left as generic. This means that the elements contain no material construction information. This is acceptable, because these properties will be varied as part of the energy simulation process later. 30
Figure 15: Progression of the build process. 31
In order to visualize how the building model would look in reality, a photo-realistic image can be rendered. Building surfaces can be edited to look like wood, brick, or any other material that can be created. A topography can be added to the building model, as well as site elements such as trees and shrubs. The rendering is created by placing a camera in one of the building model views. The camera takes a snapshot of the view, capturing a realistic, perspective view of the building. Figure 16 shows a rendered image of the building's completed geometry.
Figure 16: Rendered image of the building model.
4.3 Module 2: Data Exchange and Energy Model Preparation
4.3.1 Definition of Thermal Zones
Before the building information model can be leveraged for energy analysis downstream, it must first be separated into thermal zones. A thermal zone is a three dimensional conditioned space in a building, referring to a space that is heated and cooled by the building's HVAC system. Not all space in a building is necessarily conditioned, thus some parts of the building model may not be included in the thermal zone(s). It is possible to export unconditioned spaces as well. This can be useful if an unconditioned space shades part of a conditioned space. 32
In Revit, thermal zones are defined using Room objects. Although this may imply that every room is a separate thermal zone, this is not necessarily true. Since a thermal zone is a three-dimensional space, it must be ensured that not only the area of thermal zones is stored in the building information model, but volume data as well. In Revit, this is accomplished by making sure that Areas and Volumes is selected for the room computation method in the Rooms and Areas tab on the home menu. Thermal zone volumes can be computed at a wall finish or at a wall center. To avoid any gaps between zones in the export process, the volumes should be computed at the wall center.
One way of dividing a building into thermal zones is by room(s). In Revit, if a room is bounded by building elements on all sides - for example, a floor on the bottom, walls on all sides, and a ceiling or roof on top - then by default it is considered its own thermal zone. However, it is not necessary for all building elements to be thermally bounding, and in addition, it is not necessary to have a building element present in order to bound a thermal zone. In the former case, building elements can be defined as non-room bounding, and in the latter case, room separation lines can be used to bound a thermal zone in the absence of a physical bounding element. Figure 17 is provided to illustrate the possibilities for thermal zone placement. Each thermal zone is denoted by an X. The two bedroom zones are fully enclosed spaces, and thus require no additional constraints for thermal zone definition. However, the three remaining zones, composed of the living area, kitchen/bath, and hallway/laundry space, do require some additional constraints. Since there are no physical building elements to bound the hallway/laundry space from the living area space, a room separation line is used to bound the two spaces. The same technique is used to separate the kitchen from the living area. Because the kitchen and bathroom are relatively small spaces and both share the same vaulted ceiling, the wall separating the two is defined as non-room bounding to combine the two spaces into the same zone. 33
Non-room bounding wall
Room separation line
Room separation line
Figure 17: Floor plan view shown with five thermal zones defined.
It may be possible to simplify thermal zones even further. Figure 18 shows a scenario in which the building has been separated into two zones. The first zone encapsulates the private spaces of the building, including the bedrooms and a utility room. The second zone defines the public space: living area, kitchen and bathroom. There are a couple of reasons why further simplification of the building model's thermal zones may be beneficial at this time. First, simulation results for the two-zone and five-zone models may show that the additional zoning provides no additional accuracy. In this case, the two-zone model will save computation time. Second, the simplified two-zone model may prove more reliable when imported into an external analysis tool. Thus, when defining the thermal zones for a building information model that will be used for the purpose of energy modeling, there exists a trade-off among simulation accuracy, computation time, and import reliability. 34
Figure 18: Definition of thermal zones for the example building.
Since thermal zones are volumes, boundaries must be checked in all three dimensions. Figure 19 shows a section view of the building and the extent of the thermal zone boundaries. In the image on the left, the cantilevered section of the living area space is not being captured correctly by the current zone. This is due to the fact that Revit calculates room volumes at a specified height, and currently the calculation height is below the cantilever. In the image on the right, the calculation height is raised above the cantilever, resulting in a more accurate model. In addition, the zone on the left does not fully extend to the highest roof. This has been corrected in the zone on the right by properly defining the upper bound of the zone.
Depending on the BIM program, other information may be imparted to the model before export. Types of information may include building location or building type. Building location information could be used by an external analysis program to define a specific weather file for simulation. Building type information could be used to select default equipment and scheduling for a building given that it is a school, hospital, shopping mall, or single family home. 35
Figure 19: Section view of an incorrect thermal zone boundary (left) and a correct boundary (right).
4.3.2 Data Export
The building information model can be exported as a thermal analysis model using the Green Building Xtensible Markup Language (gbXML) file format. A gbXML file provides a number of options for export complexity. For example, one can choose whether or not to include features such as shading surfaces and window mullions. Since the roof of the example building has substantial overhangs, the decision is made to export the model as complex, with shading surfaces. Upon export, Revit provides a dialog box with a preview of the export model. At this point, one can check if thermal zones are correctly defined, and if applicable, shading surfaces as well. Figure 20 shows the export preview for the example building. 36
Figure 20: gbXML dialog box within Revit provides export settings and a preview of the analysis model.
4.3.3 Data Import
Both Ecotect and DesignBuilder allow the import of gbXML data, which make them ideal for this study. The import process for the two programs is different, so each is treated here separately.
22.214.171.124 Import Process for Ecotect
Ecotect allows the import of many types of building data - from simple geometric models to fully detailed models for thermal analysis. In this example, the building is imported as Model/Analysis Data, to which gbXML data belongs. After selecting the analysis file, the dialog box of Figure 21 is displayed. This dialog box is used to manage the proper mapping of gbXML data elements to Ecotect model elements. As the Model Element column in Figure 21 displays, most of the gbXML data elements are correctly identified. For example, the Basic Wall: Generic Woodâ€¦ data element is correctly identified as a wall model element. However, for other gbXML data elements, Ecotect is forced to guess as to which model elements they should represent. This is the case for a few of the data elements of the example building. In Revit, single-flush elements are door elements and casement elements are window elements. As Figure 21 shows, the single-flush element is not correctly identified as a door in Ecotect, and the casement element is not correctly identified as a window. These discrepancies can be resolved, however, 37
by changing the model element from guess to door or window, depending on the proper coversion. In this way, all elements should be checked for proper conversion before model import.
Figure 21: Ecotect dialog box for gbXML import.
Figure 22: A screenshot of the Ecotect analysis model just after gbXML import.
The imported gbXML model is shown in Figure 22. This is a view of the building shown in Ecotect's 3D Editor environment. As the name suggests, building geometry and properties can be edited in this environment. The building model geometry can also be represented with a rendered view, shown in Figure 23; however, building elements cannot be edited here. The different colors in Figure 22 are 38
used to identify different zones and shading surfaces in Ecotect. It can be seen that the five Ecotect zones correspond to the five thermal zones defined in Revit. The figure also indicates that the roof overhangs have been imported as shading surfaces.
Figure 23: Rendered view of Ecotect model just after import.
Although the imported building data looks generally complete and correct, the integrity of all model surfaces (both zone and shading surfaces) should still be confirmed. Errors in the building information model or the export/import process can result in analysis model inaccuracies. Such inaccuracies could include unintended voids or additional surfaces. Voids could cause thermal zones to be unenclosed, which is not permitted for thermal analysis. Additional surfaces could create a number of issues, including unwanted shading or overlapping building elements.
A close inspection of the Ecotect model reveals several gaps and additional surfaces in the building's geometry. One such inaccuracy is indicated in Figure 24, showing several unconnected surfaces located at the building's cantilevered wall. The source of this inaccuracy is unclear; however, it can be fixed in the 3D Editor environment.
Building elements in the 3D Editor are defined by lines and nodes. A collection of lines and nodes can be used to create a surface, and a surface can then be defined as a wall, door, or any other applicable modeling element. For example, a standard window element would consist of four lines with 39
four nodes at each corner. These lines and nodes can be manipulated in order to fix any model inaccuracies. Figure 25 shows a close-up view of the cantilevered wall in the 3D Editor. The surfaces and surface nodes associated with the inaccurate geometry are highlighted. A node can be moved by first selecting it (turning the node black), then left-clicking its current location, and finally left-clicking its desired location. Multiple nodes can be moved at once. The gap shown in Figure 25 is made up of four coplanar nodes. If all four nodes are selected, using the Shift key, and one of the nodes is moved, the remaining three nodes will follow.
Figure 24: A gap in the imported Ecotect geometry is indicated near the cantilevered section of the building. 40
Figure 25: The gap is shown in the 3D Editor environment with the appropriate nodes and surfaces highlighted.
126.96.36.199 Import Process for DesignBuilder
Like Ecotect, DesignBuilder allows the import of 3D building model data using the gbXML file format. Upon import, the user is given a preview of the imported geometry. Figure 26 shows the dialog box for inputting a gbXML file, as well as the preview box showing the import model. The preview box contains three options: import as building blocks, import shade surfaces, and merge coplanar surfaces. DesignBuilder uses building blocks as a way to organize different levels and/or sections of a building. Each building block is made up of a thermal zone or zones, and building elements, such as walls and windows, are assigned to each zone surface. This creates a data hierarchy in which modifications can be made at the building, block, zone, or element level. For example, if the material construction of walls is changed at the building level, then all walls in