Integration and Interoperability of Building Envelope Information and Energy Modeling

Home , Building envelope

The Pennsylvania State University

The Graduate School

College of Engineering

INTEGRATION AND INTEROPERABILITY OF BUILDING ENVELOPE INFORMATION AND ENERGY MODELING

A Dissertation in

Civil & Environmental Engineering

Mohammad Ehsan Kamel Hedayat Abad

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctoral of Philosophy

August 2017 The dissertation of Ehsan Kamel was reviewed and approved* by the following:

Ali Memari Professor and Hankin Chair of Residential Building Construction Department of Architectural Engineering and Department of Civil and Environmental Engineering

Dissertation Advisor Chair of Committee

James Freihaut Professor Department of Architectural Engineering

David Riley Professor Department of Architectural Engineering

José Duarte Professor of Architecture and Landscape Architecture Stuckeman Chair in Design Innovation and Director of the Stuckeman Center for Design Computing School of Architecture and Landscape Architecture

Farshad Rajabipour Associate Professor Department of Civil and Environmental Engineering

Patrick Fox Professor of Civil and Environmental Engineering Head of the Department of Civil and Environmental Engineering

*Signatures are on file in the Graduate School Abstract

Different methods can be adopted in order to reduce the energy consumption in buildings, which contribute to about 40% of total annual energy consumption in the U.S. The following three major approaches were considered in this study: evaluation of buildings during the design phase and pre-construction phase, energy retrofit of existing buildings, and energy monitoring of buildings.

Building envelope components are among the elements in a building that can be improved or monitored for better energy performance in all three approaches mentioned above.

Building envelope components considered in this study include both opaque and transparent components such as walls, roofs, floors, windows, and doors, which contribute to heat transfer in a building.

Building Energy Modeling (BEM) and analysis is one of the major tools contributing in energy conservation measures by providing users with data related to energy consumption of buildings. Therefore, improving different elements of BEM including the energy modeling process, obtaining outputs, and quality of outputs can be beneficial in energy conservation field.

This study is focused on improving these elements and the BEM process by expediting the whole process, making it more accurate by minimizing human interaction, and increasing the level of details of energy-related outputs. It can facilitate and increase the accuracy of

iii building’s energy evaluation during design phase, energy retrofit decision-making process, and energy monitoring.

A platform is developed to automate BEM, which provides users with detailed information on the amount of heat transfer through building envelope components. Building

Information Modeling (BIM) is adopted to facilitate the automation in modeling process.

The developed tool is capable of 1) reading a BIM file, 2) correcting some information within the file, 3) automatically convert the BIM file to a file format, which is suitable for energy simulation, and 4) automatically perform energy simulation and generate text files containing detailed heat transfer data through every single building envelope component.

The first task is limited to gbXML file format, followed by a corrective tool developed using Python, which receives the BIM file and performs some corrections on data related to building envelope components such as doors and floors. Next, a code is developed in

Ruby to use the pre-defined functions within OpenStudio source code in order to convert the gbXML file to IDF file. Since, this research is focused on building envelope, the issues and missing data related to other systems such as HVAC are resolved and added manually.

The final task is carried out using modified source code of EnergyPlus, which receives the generated IDF file and performs energy simulation to generate five text files for walls, floors, roofs, windows, and doors. These files contain the detailed and fine-grained information on the amount of heat transfer through each component as opposed to accumulative data for each thermal zone or whole-house energy consumption.

Table of Contents List of Figures ...... xi

List of Tables ...... xvii

Technical Abbreviations ...... xix

Acknowledgments...... xxiii

1) Chapter 1. Introduction ...... 1

1.1 Energy Smart Homes ...... 5

1.2 Building Envelope Energy Retrofit ...... 6

1.3 Application of BIM in Energy Simulation ...... 7

2) Chapter 2. Literature Review ...... 10

2.1 Energy Monitoring Systems ...... 11

2.2 Retrofit of Existing Homes ...... 24

2.3 Energy Modeling in Design Phase ...... 30

2.4 References ...... 40

3) Chapter 3. Methodology for Improving BEM Process and Outputs ...... 50

3.1 Introduction ...... 50

3.2 Objectives and Tasks Outline ...... 51

3.3 Further Explanation of the Tasks ...... 52

3.3.1 Tasks for Objective #1 ...... 52

3.3.2 Tasks for Objective #2 ...... 53

3.4 Research Process Summary ...... 56

4) Chapter 4. Energy Smart Homes ...... 60

State of the Art Review of Energy Smart Homes ...... 61

4.1 Abstract ...... 61

4.2 Introduction ...... 63

4.3 The Very First Projects and Examples ...... 67

4.4 Smart Homes Services ...... 69

4.5 Smart Homes Components ...... 72

4.6 Buildings, Projects, and Labs Focused on Smart Homes ...... 74

4.7 Challenges in Smart Home ...... 86

4.8 Design Different Types of Energy Smart Homes ...... 87

4.8.1 Energy Monitoring Systems ...... 95

4.8.2 Systems with Control Capabilities ...... 100

4.8.3 Systems with Advanced Data Processing Capabilities ...... 106

4.9 Summary and Concluding Remarks ...... 113

4.10 References ...... 117

5) Chapter 5. Energy Retrofit of Buildings ...... 166

Residential and Commercial Building Envelope Energy Retrofit: Innovative Measures and Example Projects ...... 167

5.1 Abstract ...... 167

5.2 Introduction ...... 168

5.3 Materials and Systems in Envelope Energy Retrofit ...... 172

5.3.1 Conventional retrofit measures ...... 180

5.3.2 New retrofit measures ...... 183

5.4 Example Retrofit Projects ...... 201

5.5 Multiple Criteria and Proper Tools on Choosing Retrofit Measures ...... 212

5.6 Numerical Study of Different Envelope Energy Retrofit ...... 216

5.6.1 Development of the Computer Model ...... 217

5.6.2 Results of the Computer Modeling and Discussion ...... 222

5.7 Summary and Conclusions ...... 230

5.8 References ...... 233

6) Chapter 6. Application of BIM in Energy Modeling ...... 244

Review of BIM’s application in energy simulation: tools, issues, and solutions ...... 245

6.1 Abstract ...... 245

6.2 Introduction ...... 247

6.3 BEM tool’s Graphical User Interface (GUI) and energy simulation engine.... 254

vii

6.4 Types of BIM file schemas and their properties ...... 260

6.5 Review of Identified Challenges and Issues in BBIP ...... 266

6.6 Review of Identified Solutions Adopted by Researchers Related to BBIP ..... 273

6.7 Review of three case studies in BBIP ...... 279

6.7.1 Modeling Process in Revit for Three Case Studies ...... 282

6.7.2 Results of the Case Studies, Issues Related to the BBIP, and Suggested

Solutions ...... 287

6.7.3 Developed gbXML corrective tool ...... 294

6.8 Summary and Conclusion ...... 298

6.9 References ...... 302

7) Chapter 7. Building Energy Performance Assessment Tool (BEPAT) ...... 312

Development of a Platform for Building Energy Performance Assessment Tool (BEPAT) for Energy Smart Homes and Design Optimization ...... 313

7.1 Abstract ...... 313

7.2 Introduction ...... 314

7.3 Methodology for Developing and Validating BEPAT ...... 320

7.4 Computer Model and Validation Method ...... 328

7.5 Results and Discussion ...... 333

7.6 Summary and Conclusions ...... 340

viii

7.7 References ...... 344

8) Chapter 8. Automated Building Energy Modeling and Assessment Tool (ABEMAT)

350

Automated Building Energy Modeling and Assessment Tool (ABEMAT) ...... 351

8.1 Abstract ...... 351

8.2 Introduction ...... 352

8.3 Methodology for Developing ABEMAT and Data Verification ...... 358

8.3.1 First component of ABEMAT: gbXML corrective tool ...... 358

8.3.2 Second component of ABEMAT: gbXML to IDF converter ...... 361

8.3.3 Third component of ABEMAT: IDF to fine-grained heat transfer outputs

364

8.3.4 Methodology for Data Validation ...... 368

8.4 Details of Modeling and Results of Data Validation ...... 370

8.4.1 Comparison between Existing BEM Methods ...... 371

8.4.2 Modeling a One-Story Residential Building for Data Verification ...... 375

8.4.3 The Outputs and Results of Data Verification ...... 378

8.5 Summary and Conclusions ...... 381

8.6 References ...... 384

9) Chapter 9. Discussion, Summary, and Conclusion ...... 391

9.1 Discussion ...... 391

9.2 Outcomes ...... 394

9.3 Research Contributions ...... 400

9.4 Potential Future Researches ...... 402

Appendix A: Codes ...... 404

A.1 Subroutines Added to EnergyPlus Source Code for Heat Transfer through

Windows ...... 404

Under the DataSurfaces module: ...... 404

Under the DataSurfaces header: ...... 404

Under OutputReportTabular module: ...... 404

A.2 Subroutines Used for Other Components (Wall, Roof, Ceiling, Doors, and

IntMass): ...... 407

Under the HeatBalanceSurfaceManager module: ...... 407

Under the DataSurfaces header and .cc: ...... 409

List of Figures

Figure 1-1. Three major areas that can benefit from the outcomes of this research ...... 1

Figure 1-2. Two major areas in BEM and focus of improvement in this research ...... 2

Figure 2-1. An example of IHD – Ewgeco IHD [5] ...... 16

Figure 2-2. Display of normative data in an IHD connected to smart meter [1] ...... 16

Figure 2-3. Overall system architecture for an energy monitoring and control system [9]

...... 17

Figure 2-4. GUI of a system used to monitor and control a smart home [10] ...... 18

Figure 2-5. Different layers in a smart space middleware [11] ...... 19

Figure 2-6. Major concepts in ThinkHome knowledge base system [12] ...... 21

Figure 2-7. Measuring and monitoring electricity consumption by using metering plugs

[14] ...... 22

Figure 2-8. Monitoring electricity consumption by using UbiLense mobile app [13] ..... 23

Figure 2-9. Operating strategies of the APSC&VU Tus [17]...... 26

Figure 2-10. Operating strategies of the ASP&EA Tus [18] ...... 27

Figure 2-11. ASTF operational mechanism as a wall component [19] ...... 27

Figure 2-12. Early-stage problem solving in design phase facilitates later stages of design activity [45] ...... 31

Figure 2-13. Higher modification cost through later stages of building life [15] ...... 31

Figure 2-14. Typical stages of design [45] ...... 32

Figure 2-15. Example of semi-automated BEM process to add additional options in material library [55] ...... 36

Figure 2-16. Example of semi-automated BEM process for adding missing information required for energy modeling [54] ...... 36

Figure 2-17. Example of semi-automated BEM process for using different tools and interfaces to add missing information required for energy modeling [56] ...... 37

Figure 2-18. Typical process for semi-automated and automated energy simulation [57]

...... 38

Figure 3-1. Two major BEM focused areas improved in this research ...... 58

Figure 4-1. Simplified Architecture of a Smart Home ...... 67

Figure 4-2. Intended Services in Smart Homes. Adapted from [84] ...... 71

Figure 4-3. Number of Papers Focsed on Different Areas in Smart Homes Services Since

2004. Adapted (with curves linearized) from [88]...... 71

Figure 4-4. Major Concepts in ThinkHome Knowledge Base System. Adapted from: [12]

...... 107

Figure 5-1. Fusion Temperature for different types of PCM [(data adapted from [225])].

...... 187

Figure 5-2. Latent heat for different types of PCM [(data adapted from [225])]...... 187

Figure 5-3. Pouches filled with PCM (left)(reproduced from [254], used with permission), PCM Macrocapsules (middle), and PCM Microcapsules (right)(reproduced from [259], used with permission) ...... 187

Figure 5-4. Overview of innovative energy retrofit methods ...... 188

xii

Figure 5-5. Transparent aerogel panel (reproduced from [226], used with permission) 191

Figure 5-6. Cross section of gas-filled panel (left) (reproduced from [263], used with permission) and installation of gas-filled panel (right) (reproduced from [262], used with permission) ...... 192

Figure 5-7. Details of single channel glazed photovoltaic thermal module (SCGPVTM)

(figure redrawn, adopted from [227]) ...... 193

Figure 5-8. XPS covered with final finish (left) and EPS coated with TRM (right)

(reproduced from [228], used with permission) ...... 194

Figure 5-9. Application of EPS coated with TRM (1) and installation process(2) for a social housing in Italy, (before retrofit (3) and after retrofit (4)) (reproduced from [265], used with permission) ...... 195

Figure 5-10. External Thermal Insulation Composite Systesm (ETICS) (figure redrawn, adopted from [229]) ...... 196

Figure 5-11. The structure of multi-functional energy efficient façade system (MEEFS)

Figure 5-12. Advanced Solar Protection & Energy Absorption Technological Unit

(ASP&EA TU) (top) and Advanced Passive Solar Collector & Ventilation Unit

Technological Unit (APSC&VU TU) (bottom) owned by Acciona and Tecnalia, respectively (reproduced from [17, 18], use with permission) ...... 199

Figure 5-13. Wall-based ASTF (figure redrawn, adopted from [19]) ...... 200

xiii

Figure 5-14. Annual heating and cooling site energy saving potential and material cost of different envelope energy retrofit measures in cold climate region ...... 229

Figure 5-15. Annual heating and cooling site energy saving potential and material cost of different envelope energy retrofit measures in hot-dry climate region ...... 229

Figure 6-1. Major contributions of BIM in building energy management ...... 249

Figure 6-2. Overview of BBIP ...... 252

Figure 6-3. Interoperation between energy simulation GUI and engine ...... 255

Figure 6-4. Example of energy analysis output generated by GBS ...... 256

Figure 6-5. Additional step added to integrate BIM in BEM process ...... 259

Figure 6-6. Partial presentation of a gbXML file containing information for energy simulation ...... 265

Figure 6-7. IFC standard’s data structure ...... 266

Figure 6-8. Overview of BBIP and the potential sources of issues ...... 268

Figure 6-9. An overview of the process adopted in case studies ...... 280

Figure 6-10. Process of three different approaches ...... 282

Figure 6-11. Plan view (Left) and 3D view (Right) of building modeled in Revit (Roof is not shown) ...... 284

Figure 6-12. Two methods in Revit for exporting data to gbXML file ...... 284

Figure 6-13. Advanced energy settings in Revit ...... 285

Figure 6-14. Major elements of gbXML file exported from Revit ...... 287

Figure 6-15. Errors generated in OpenStudio due to similar adjacent space names in gbXML file ...... 288

xiv

Figure 6-16. Corrected gbXML files using gbXML corrective tool developed using

Python (Top: Looks for duplications in floor’s adjacent spaces. Bottom: Add the missing data related to door’s construction data) ...... 296

Figure 7-1. Schematic illustration of potential applications and major contributions of

BEPAT ...... 319

Figure 7-2. IDF editor used for modifying EnergyPlus files in EPlaunch...... 321

Figure 7-3. Modules in EnergyPlus source code opened in VisualStudio ...... 323

Figure 7-4. Example of the sensible heat gain summary table generated by EnergyPlus

...... 326

Figure 7-5. Overview of the enegry analysis process used in this study ...... 328

Figure 7-6. Two methods used in this study for data verification ...... 330

Figure 7-7. Overview of two different methods used in the study ...... 331

Figure 7-8. Plan view (Left) and 3D view (Right) of building modeled in Revit ...... 331

Figure 7-9. Examples of name assigned to building envelope components in the spreadsheet outputs generated by EnergyPlus advanced outputs ...... 334

Figure 7-10. Automatically generated text files containing heat gain/loss data for each component and their thermal zones ...... 335

Figure 7-11. Building components tags ...... 338

Figure 7-12. Annual building sensible heat gain comopnents table, generated by E+ containing accumulated data for each thermal zone ...... 338

Figure 8-1. Different approaches for reducing energy consumption, and different tools involved...... 353

Figure 8-2. Different tools and components involved in development of ABEMAT .... 357

Figure 8-3. Errors generated in OpenStudio related to similar adjacent space name ..... 359

Figure 8-4. Missing door’s construction ID in gbXML file causes error during energy simulation ...... 359

Figure 8-5. Errors related to building envelope through BIM-to-BEM process ...... 360

Figure 8-6. IDF editor used for modifying IDF file to add missing data...... 363

Figure 8-7. Modified process of ABEMAT to add HVAC, schedules, loads, and thermostat set points data manually ...... 363

Figure 8-8. Some of the modules in EnergyPlus source code ...... 365

Figure 8-9. Example of ABEMAT’s outputs to illustrate the titles of each component and associated thermal zone ...... 367

Figure 8-10. Adopted process for validating ABEMAT’s outputs ...... 368

Figure 8-11. Enabling EnergyPlus to obtain advanced outputs using IDF editor ...... 369

Figure 8-12. OpenStudio GUI used for adding required data for energy simulation and converting gbXML to IDF ...... 371

Figure 8-13. Plan view (Left) and 3D view (Right) of building modeled in Revit (roof is not shown in this figure) ...... 375

Figure 8-14. Exporting the architectural model under “room/space volumes” using Revit

...... 377

Figure 8-15. Text files generated automatically by ABEMAT containing heat transfer data for each component and their thermal zones ...... 379

xvi

List of Tables

Table 4-1. Different Definitiosns of Smart Homes ...... 65

Table 4-2. Major Categories of Smart Home Services ...... 69

Table 4-3. Examples of Projects, Houses, and Living Labs Built or Equipped for

Research Purposes Related to Smart Homes ...... 80

Table 4-4. Definition of other concepts related to energy smart homes ...... 94

Table 4-5. Evaluation of Different Systems in Energy Smart Homes that are Proposed in

Literature or Available in Market ...... 140

Table 5-1. Summary of the energy retrofit methods used in the building envelope in research studies ...... 174

Table 5-2. Conventional types of insulation material ...... 182

Table 5-3. Forms of conventional insulation material ...... 182

Table 5-4. Summary of the energy retrofit measures in energy retrofit projects ...... 209

Table 5-5. The dominant properties of homes builtt between 1990 and 2000 in the U.S.

[281] ...... 218

Table 5-6. The building envelope properties of the benchmark house defined in BEopt

...... 219

Table 5-7. Other properties of the benchmark house defined in BEopt ...... 220

Table 5-8. Different energy retrofit methods used for computer modeling ...... 222

Table 5-9. Properties of the PCM materials used for PCM drywall and PCM finish .... 225

Table 5-10. Retrofit measures studied in computer modeling in details ...... 228

xvii

Table 6-1. Categories of energy simulation tools [314, 307, 315, 308, 289, 316, 310] . 258

Table 6-2. Comparison between gbXML and IFC based on literature review [322, 314,

331, 292, 303, 301, 308, 323] ...... 263

Table 6-3. Summary of research studies on using BIM in energy simulation ...... 277

Table 6-4. Properties of the building modeled in Revit ...... 285

Table 6-5. Data transfer issues and comparison between three case studies ...... 289

Table 6-6. Issues observed through the three case studies modeling process and adopted solutions ...... 290

Table 7-1. Properties of the building modeled in EnergyPlus ...... 332

Table 7-2. Number of different components in each thermal zone ...... 333

Table 7-3. Summary of energy simulation outputs obtained from standard and modified

EnergyPlus ...... 339

Table 8-1. Comparison between classic, semi-automated, and automated BEM using BIM

...... 372

Table 8-2. Properties of the building modeled in EnergyPlus ...... 376

Table 8-3. Summary of energy simulation outputs obtained from semi-automated BEM method and ABEMAT (Ext.=Exterior wall, Int.=Interior wall, Win.=Window, and

D.=Door) for data validation ...... 380

xviii

Technical Abbreviations

Active Solar Thermal Façade (ASTF)

Adaptive Control Of Home Environments (ACHE)

Advanced Passive Solar Collector & Ventilation Unit Technological Unit (APSC&VU TU)

Advanced Solar Protection & Energy Absorption Technological Unit (ASP&EA TU)

Architecture, Engineering, And Construction (AEC)

Automated Building Energy Modeling And Assessment Tool (ABEMAT)

Aware Home Research Initiative (AHRI)

Bim-To-Bem Interoperability Process (BBIP)

Building Automation Systems (BAS)

Building Automation Technology (BAT)

Building Energy Modeling (BEM)

Building Energy Performance Assessment Tool (BEPAT)

Building Heating Technology (BHT)

Building Information Modeling (BIM)

Building Integrated PV (BIPV)

Computer-Aided Design (CAD)

Concrete Masonry Unit (CMU)

Conditional Demand Analysis (CDA)

Deep Energy Retrofit (DER)

Demand Response (DR)

xix

Demand Side Management (DSM)

Distribute Energy Resources (DER)

Dynamic Load Priority (DLP)

Energy Conservation Measures (ECM)

Energy Consumption Indicators (ECI)

Energy Resource Management (ERM)

Energy Retrofit Measure (ERM)

Energy-Consumption Information System (ECOIS)

Exchange Requirements (ER)

Expanded Polystyrene (EPS)

Exterior Thermal & Moisture Management System (ETMMS)

External Thermal Insulation Composite System (ETICS)

Game Theory (GT)

Gas-Filled Panels (GFP)

Graphical User Interface (GUI)

Green Building Extensible Markup Language (gbXML)

Home Area Network (HAN)

Home Automation (HA)

Home Energy Displays (HED)

Home Energy Management (HEM)

Home Energy Management Systems (HEMS)

Human Computer Interaction (HCI)

Industry Foundation Class (IFC)

Information Delivery Manual (IDM)

In-Home Display (IHD)

Input Data File (IDF)

Integrated Wireless Technologies (IWT)

Interactive Room Operating System (iROS)

Life Cycle Assessment (LCA)

Load-Survey Meter (LSM)

Model Based Predictive (MPC)

Model View Definition (MVD)

Multi-Agent System Technology (MAST)

Multifunctional Energy Efficient Façade Systems (Meefs)

Multiple Criteria Complex Proportional Assessment (COPRAS)

National Association Of Home Builders (NAHB)

National Renewable Energy Laboratory (NREL)

Network Control Unit (NCU)

Neural Network (NN)

Nonintrusive Load Monitoring (NILM)

Oriented Strand Boards (OSB)

Pcm Thermal Shield (PCMTS)

Phase Change Materials (PCM)

Radio Frequency Identification (RFID)

Residential Energy Consumption Survey (RECS)

Service Hot Water (SHW)

xxi

Simple Object Access Protocol (SOAP)

Single Channel Glazed Photovoltaic Thermal Module (SCGPVTM)

Smart Energy Buildings (SEB)

Smart Energy Management System (SEMS)

Smart Energy Monitor (SEM)

Smart Home Energy Management Systems (SHEMS)

Smart Home Micro-Computers (SHMC)

Solar Heat Gain Coefficient (SHGC)

Solar Reflectance Index (SRI)

Supervisory Control And Data Acquisition (SCADA)

Textile Reinforced Mortar (TRM)

Vacuum Insulated Panel (VIP)

xxii

Acknowledgments

This work would not have been possible without supports of Dr. Ali Memari, Hankin

Chair of Residential Building Construction and Director of Pennsylvania Housing

Research Center (PHRC), who worked actively to provide me with the opportunity and support to pursue my academic goals and who has been supportive of my efforts to reach my career goals.

I am grateful to all of those with whom I have had the pleasure to work during this and other related projects. Each of the members of my Dissertation Committee – Prof. Jose

Duarte, Dr. David Riley, Dr. James Freihaut, and Dr. Farshad Rajabipour – have provided me extensive personal and professional guidance. I also want to sincerely thank the Pennsylvania Housing Research Center (PHRC) staff for their support over the past four years. I am also grateful to my peers in the Department of Civil Engineering and

Department of Architectural Engineering who have helped me by providing an inspiring and enjoyable atmosphere.

Finally, the pursuit of the Ph.D. degree and my research is greatly due to support and encouragement of my family. I would like to thank my parents, whose love and guidance are with me in whatever I pursue. They are the ultimate role models.

xxiii

1) Chapter 1. Introduction

Buildings consume about 40% of total energy consumption in the U.S. Different correction or improvement measures can be applied to reduce the energy consumption in buildings.

These measures can be applied during different phases of building life cycle such as design or use phase. For example, components such as building envelope can be designed more efficiently based on their energy performance. Moreover, energy retrofit of existing homes and energy consumption monitoring are two concepts related to use phase of buildings.

Energy retrofit software programs and energy smart homes are among the emerging tools dedicated to these areas. The major contributions of this research are dedicated to these three areas mentioned above and shown in Figure 1-1.

Figure 1-1. Three major areas that can benefit from the outcomes of this research

Building energy modeling (BEM) and analysis is the major component in all such measures to reduce the energy consumption. Improved BEM process and outputs can contribute to more energy efficient design, optimized decision-making process in energy retrofit of existing buildings, and detailed and high quality information during energy monitoring in smart homes. Therefore, improving the BEM process can be focused on two major areas including the modeling process and quality of outputs, which are the focus of this research shown in Figure 1-2.

Figure 1-2. Two major areas in BEM and focus of improvement in this research

The modeling process can be improved by reducing manual interaction and automating the

BEM process. An emerging tool that can contribute to this goal is Building Information

Modeling (BIM), which can eliminate the need for reentering data and expedite the BEM process, resulting in less error vulnerability and more accuracy. However, multiple issues and challenges occur during the BIM-to-BEM process, which requires special attention and

2 additional tools are required to resolve these issues. Existing tools still require manual interaction in many different stages of BIM-to-BEM process, since there are multiple components in this process including BIM tools, BIM files, BEM tools capable of reading

BIM files, and multiple data transfer and file conversions. The main goal in BIM-to-BEM process can be developing a fully automated process with minimum manual interaction after the architectural model is developed.

Quality of outputs can be improved by providing fine-grained outputs related to every single component within a building contributing to energy consumption as opposed to accumulated outputs. For example, existing BEM tools only provide accumulated energy consumption with regard to a thermal zone or the whole house; however, these outputs can include fine-grained and detailed information on the amount of heat transfer through each building envelope component such as windows and walls. All three areas illustrated in

Figure 1-1, can benefit from such information. Therefore, it can be beneficial if new tools are developed to provide such data in an easy, fast, and accurate way.

Developing a tool that can obtain a BIM file containing all the required data for energy simulation generated by a BIM tool, converting it to a file that can be read by BEM tools to perform energy analysis, and providing detailed outputs on the amount of heat transfer through building envelope components can be a great asset to automate the BEM process.

The resulting detailed outputs can then contribute to the quality of building design tools, building’s energy retrofit tools, and energy monitoring in energy smart homes.

Reducing energy consumption in buildings can involve several tools in AEC industry and can be fulfilled through using multiple assets; however, picking a certain methodology help boiling down these options to limited number of tools and aspects of a building. For example, this goal can be reached through instrumenting a house with energy meters or smart outlets to measure the energy consumption of certain spaces and appliances in a building. In addition, the focus can be on electrical components, mechanical system, or building envelope. However, this study is focused on building energy simulation computer tools and building envelope components.

The multiple areas of study and research for this work include BEM in general, BIM, application of BIM in BEM, building envelope energy retrofit, decision-making process for energy retrofit, Computer-aided design (CAD) tools, BEM tools, energy smart homes, and energy monitoring tools. Hence, the following topical areas are selected for detailed study in order to identify state-of-the-art advances and related shortcomings and issues.

- Energy Smart homes

- Building envelope energy retrofit

- Application of BIM in energy simulation

A short introduction related to each topic is provided in this section and detailed literature review presented in the next section.

1.1 Energy Smart Homes

Several definitions are suggested by researchers for smart homes and covered in detail in the literature review section. Various capabilities can be attributed to smart homes including communication network, automatic control, remote control, and monitoring systems. Smart homes can be studied with regard to different aspects including their intended services and their components. Considering the services and components reviewed in the literature related to smart homes, the following definition can be considered as a good representative:

“A dwelling in which data related to home environment and its residents are obtained from sensors, electric appliances, or home gateway and transferred through a network of communication tools to monitoring device or execution unit to help decide on or execute proper actions called services. These services are provided either automatically or directly through a remote or central control system in order to facilitate or improve the residents’ daily lives”

One of the services considered for smart homes is dedicated to energy conservation aspect of buildings, and the smart homes equipped as such are referred to as energy smart homes.

Energy smart homes are mainly focused on two aspects including control and monitoring capabilities. For example, control systems can automatically adjust the thermostat in a house to reach the comfort criteria. Monitoring capabilities, on the other hand, are dedicated to providing energy-related information for users, which can have incentive or decision-making effects in future energy consumption habits of residents. For example, 5 energy meters can save energy consumption data and compare it to the average numbers within the neighborhood for incentive purposes [1].

Several components can contribute to energy smart homes including the sensing unit, communication unit, processing unit, and control/execution unit. Review of multiple studies revealed that the processing unit, which collects information and sends proper outputs to either control or monitoring units, has room for improvements. This unit resembles the brain of the system and it can be equipped with new capabilities to perform more complicated computations such as energy simulation, in order to provide more detailed and accurate outputs.

This study is dedicated to similar aspect of energy smart homes and aims for developing a tool, which can equip the processing unit of energy smart homes with such capability to provide more detailed energy-related outputs in a fast, easy, and accurate way. Chapter 4, presents the results of the study on this area and explains the state-of-the-art technologies and shortcomings concerning energy smart homes.

1.2 Building Envelope Energy Retrofit

Energy retrofit is one of the major measures in reducing the existing building energy consumption, the majority of which can be considered to be due to mechanical systems, electrical components, and building envelope components. The effectiveness of different measures in energy retrofit has already been observed by researchers. Different measures can include replacing mechanical systems with more efficient products, replacing electrical

6 systems such as appliances and lighting systems with more efficient options, and adding more insulation to the existing wall system, for example.

Multiple tools and components can contribute to this goal. Energy simulation tools can help in decision-making process by providing users with energy performance data of a building and identifying the components, which can be improved. Energy retrofit materials, products, and methods are also among the important components of building energy retrofit. Several conventional and new materials, technologies, and methods are studied by researchers and their effectiveness in reducing energy consumption is evaluated.

This study is focused on energy retrofit building envelope components such as walls, windows, roof, and floor. In order to identify the existing and innovative technologies and materials, a thorough study is conducted in Chapter 5, which can lead to a better understanding of tools and technologies related to building energy retrofit and the shortcomings.

1.3 Application of BIM in Energy Simulation

It is more convenient and less costly to improve and make changes in building design during the design or pre-construction phase of buildings, where computer models have high contribution. Such models can help evaluate the design prior to construction and provide users with different information depending on the focus of evaluation. Energy performance can be an important focus of these tools and evaluation process, which can be referred to as BEM.

BEM tools can provide users with different types of information such as whole-house energy consumption, consumption of different fuel types, energy-related costs, and energy consumption of different systems such as HVAC and lighting. These types of information can be used by engineers in order to change more energy intensive components and optimize the design. Therefore, it is important to make these types of information as accurate and detailed as possible. Moreover, facilitating the building energy modeling and analysis process can also be of great importance. Improving the BEM process can be focused on different aspects such as speed, convenience, and accuracy.

One of the tools, which can help in fulfilling this goal, is BIM, which is used in different areas of building construction industry such as facility management, architectural and structural design, and construction management. It can help to avoid reentering all the data at different stages in pre-construction phase, especially when different tools are adopted for modeling a building for different types of evaluation. Avoiding reentering the data can make the process faster and improve the accuracy by minimizing the manual interaction and human errors.

The BIM-to-BEM process is composed of multiple steps and components including the

BIM authoring tools, BIM files, BEM tools, and converting BIM to BEM files. Many challenges and issues exist in these steps, and it is important to identify them to provide proper solutions. The BIM-to-BEM process is explained in Chapter 6, where the state-of- the-art tools and methods in this field are reviewed and also the challenges and shortcomings identified. A comprehensive categorization is also suggested for different

8 steps and components within the BIM-to-BEM interoperability process (BBIP). Also explained in Chapter 6 are the outcomes that can contribute to design of the computer tool developed in this study based on identification of the shortcomings and targeting their solution.

2) Chapter 2. Literature Review

In 2014, 41% of total U.S. energy consumption that is about 40 quadrillion British thermal units [2] was in residential and commercial buildings. Typical energy consumption per square meter over a 60-year life span of an office block is about 7.9, 2.3, 0.7, 15.7, 73.3, and 0.2% for material, manufacturing, transportation, recurring, use, and disposal, respectively [3]. As can be noted, the dominant section of energy consumption in a building’s life span is mostly related to the use phase of the building compared to other phases such as manufacturing phase or material preparation. The energy consumption during use phase can be affected by different parameters such as building envelope properties, mechanical system, appliances, electrical system, climate zone, energy consumption behavior of residents, and primary energy source. Moreover, some of these elements such as building envelope properties can be modified during the design phase and prior to construction. Reducing small percent of this rate of energy consumption in buildings can lead to significant financial and environmental impacts. Researchers have studied various measures to reduce the building energy consumption.

Review of literature shows the studies in this field can be divided into three major categories as follow:

1) Methods in energy monitoring, existing tools and technologies, and identifying

the shortcomings in this area

2) Methods in building energy retrofit and identifying associated issues,

challenges, and shortcomings

3) Methods in building energy modeling and identifying new methods in BEM,

the issues, and shortcomings

These methods lead to reduction in energy consumption in buildings during the use phase of buildings. The third method can be adopted during the design phase, which makes it more desirable, since it is more convenient to make changes during the design phase. Based on these three major approaches, more in-depth review of literature is necessary in order to investigate the advances in these fields and identify the areas that require further study.

2.1 Energy Monitoring Systems

Energy monitoring systems have evolved from simple energy meters to more complicated systems and technologies. Emerging technologies such as energy smart homes provide users with many capabilities such as remote monitoring and controlling systems. The major impacts of energy monitoring systems is the incentive effects on residents and active impacts by enabling users to control certain components within buildings, which are directly related to energy consumption such as thermostats.

Active and passive approaches require control and sensing capabilities, respectively. For example, automatic on/off lighting system that works based on sensing the movements, notification system that notify the occupants if a window is left open, smart meters that show the total energy consumption of the households, and automatic thermostat that adjust

11 the temperature based on the heating or cooling season are all among the activities that need control and sensing capabilities.

Three of the major components involved in energy smart home include sensors network, processing unit, and control unit. The sensors network consists of sensors, communication unit, and power supply. The processing unit can be a computer that receives all the data from sensors and take proper actions based on what is programmed. These actions can be as simple as analyzing these data and developing diagrams for better presentation to inform the residents on their real-time energy consumption, or sending signals to control unit.

Finally, the control unit can receive the outputs and based on that take a proper action. For example, it can be sending a notification to an electronic device (passive) or more complicated actions such as adjusting the thermostat’s set point temperature or turning off the HVAC system (active).

This control unit can be linked to any of the major components in the building that contributes to energy consumption including electrical systems, mechanical systems, and building envelope. Electrical systems mainly include the lighting and appliances, while mechanical systems of interest here include the whole HVAC system of the building.

However, building envelope is usually more static and cannot be adjusted or changed during the use phase of the building. Despite such a fact, some of the more innovative building envelope components such as dynamic façade, operable shading systems, and electrochromic windows can be adjusted, which means they have the potential to be integrated with control systems.

On the other hand, monitoring capabilities of energy smart homes can easily be connected to and focused on building envelope or thermal zones in their area of service by providing energy consumption or heat transfer data related to these components such as walls, windows, roofs, and certain thermal zones. Zonal energy consumption and the amount of heat transfer through certain surfaces can reveal more detailed energy performance of a building to users.

Most of the existing and available systems in energy smart home partially include the components of a smart home. A simple single-spot moisture sensor can sense the relative humidity, do a basic electric circuit process, and set off an alarm. More complicated systems can include a network of sensors; a computer as the central processing unit that is programmed to do simple analysis and create the ability of real-time monitoring of different parameters such as temperature, humidity, and energy consumption; and finally the control unit that is designed to send a notification to an electronic device such as PC, tablet, or smart phone.

Depending on the complexity of the data process in the central processing unit of a smart home, the monitoring systems and control units can be provided with more complex, accurate, and advanced outputs. A basic data analysis includes illustration of measurements such as temperature, energy consumption, or humidity in a real-time time vs. parameter diagram. However, there are more advanced information that can help tenants to improve energy consumption of homes. For example, monitoring the energy consumption in a certain thermal zone of the house such as a bedroom, attic space, kitchen, or living room

13 can be much of an interest to locate the possible energy loss within a building. In addition to that, the residents may be interested to see how much of the energy supplied for living room is transferred through walls, roof, and windows. Therefore, the central processing unit should be able to evaluate the energy performance of a building in more details.

To fulfil this goal, more than a simple data analysis is required. An energy smart home requires being equipped with an energy simulation tool to be able to receive all of the information including building information, schedules, data related to HVAC, and loads to perform a high-level analysis. The outcome can include energy consumption of a certain area in a building, sending proper signals to the control unit of the smart home to control a dynamic façade, or simply informing residents about the amount of heat transfer through certain components such as windows and walls.

In order to upgrade the existing smart home systems into an advanced system that can monitor building energy performance in higher level of details, central processing unit needs be equipped with an energy simulation tool and it needs to be provided with required information for energy analysis such as building geometry and materials properties.

Building information can include floor square footage, number of the bedrooms, windows square footage, types of cavity insulation materials, their thermal properties, and solar heat gain coefficient (SHGC) of the glazing system.

Review of academic publications on state-of-the-art energy monitoring systems reveals that energy smart homes are at the heart of recent advances in this area. They can be categorized under three groups including 1) the houses equipped with only energy

14 monitoring capabilities, 2) control systems, and 3) advanced data processing capabilities, which the combination of the first and the third is the focus of this research.

Energy smart homes equipped with monitoring capabilities can track environmental conditions including human activities, temperature, and energy consumption, which can provide residents with comfort and improve energy consumption habits by providing energy-related information [4]. For example, smart homes can be equipped with smart meters that can monitor and keep the record of energy consumption at different time intervals or even real-time. For example, energy use visualizers or in-home display (IHD) show the total energy consumption. Compared with homes without IHD, reduction in gas and electricity usage has been notices by 20% and 7%, receptively [5]. It shows providing household with more information about their energy consumption can help them reduce their energy consumption. Application of IHD requires low-level analysis of data acquired from the grid for gas and electricity. The data shown in display are obtained by performing simple data analysis that is basically uses graphical interfaces to show the acquired data from smart meter with no further analysis required. Figure 2-1 shows an example of these

IHD. It is also observed that the households that get normalized data about their energy consumption tend to use less energy compared with houses that are provided with absolute energy consumption [1]. Figure 2-2 shows how an IHD connected to a smart meter can show the current energy consumption that can be compared with similar households.

Figure 2-1. An example of IHD – Ewgeco IHD [5]

Figure 2-2. Display of normative data in an IHD connected to smart meter [1]

Energy smart homes with control capabilities are also another major category studied by researchers. For example, Han and Lim [6] studied application of two different communication standards including ZigBee and IEEE 802.15.4 that help interoperability between different electrical equipment, meters, and smart energy enabling products. The system is capable of either turning on and off or dimming the lights and it can also use the same communication system to receive data from meters [6]. ZigBee is a wireless sensor network that is secure and more importantly it is low power consumption and fast reaction and that is why it is used in automatic control, energy monitor, light control, home security,

16 and remote control [7]. Similar applications for energy smart homes with control capabilities are reported, which use sensors to detect the presence of a person in a room to control the systems such as lighting and HVAC to reduce energy consumption [8, 9]. An example of the overall system architecture is illustrated in Figure 2-3. The data processing module will receive the acquired data from sensors in terms of electricity consumption and the web server designed and implemented in the central processing module can be designed in a way that provide the user with energy consumption, price comparison, and statistical analysis. The results will be sent to a PC or mobile device, for example, and there is an application installed on the device that can be used as the GUI. Depending on the results, the user can decide to turn a device on/off and send a command through the electronic device and the GUI [9]. It should be pointed out that the data analysis is low-level and does not need any complicated whole-house modeling to obtain the desired results such as total energy consumption or price estimation.

Figure 2-3. Overall system architecture for an energy monitoring and control system [9]

Another example of such systems is the application of multiple sensor-based autonomous monitoring and control system that is basically a combination of temperature and motion sensors to optimize the activity of appliances or mechanical systems such as HVAC based on the temperature and human activities in a building to save more energy. The GUI developed for this system is illustrated in Figure 2-4. De Silva et al. (2010) found that the application of this system can reduce the energy consumption from 35 kWh to 15 kWh compared to a normal home [10]. These semi-advanced systems typically have a graphical user interface (GUI) that help the user to monitor and control the smart home.

Figure 2-4. GUI of a system used to monitor and control a smart home [10]

A study conducted in University of Florida is another example of this type of system. The research is focused on a programmable system that consist of different layers such as sensors, service layer, and application layer that are basically the sensing, processing, and control (command) sections, respectively [11]. More details and the architecture of the middleware are illustrated in Figure 2-5. Two more innovative systems used in this house

(Gator Tech Smart House) are smart plugs and floor. The smart plug that is equipped with

18 a RFID (Radio Frequency Identification) sensor can automatically detect, read, and connect a plug to the main computer as soon as an electrical device is installed. The smart floor employs sensors connected to the bottom of the flooring tiles in order to detect the location of the user. It can be observed that this system also, does not need any high-level analysis such as energy modeling in order to be able to function. All the commands are based on a less complicated analysis executed in the programmed processing unit and the actions selected by the user on a GUI installed on a computer.

Figure 2-5. Different layers in a smart space middleware [11]

Although these systems can be equipped with advanced control capabilities, the processing unit does not perform any high-level and complicated analysis and all the calculations are based on comparing the acquired data with predefined limits or simple calculations such as obtaining the average or developing graphs. 19

On the other hand, the third major category in energy smart homes is the system equipped with advanced data analysis capabilities. In such systems, a high-level data analysis is required such as whole house energy modeling, which is the same idea this research is focused on. The same concept is used in some other studies such as ThinkHome project

[12]. In this project, the researchers created a comprehensive knowledge-based system in order to reach higher energy efficiency and intelligent control system in a smart home that is equipped with building automation systems (BAS). Different types of information were required to develop this knowledge-based system illustrated in Figure 4-4. One of the components is building information that refers to data such as wall thickness, square footage, spaces, and materials. Such information is required in order to be able to perform more optimized control strategies toward a smart home with higher energy efficiency.

Accordingly, the data stored in a building information model (BIM) and Green Building

XML (gbXML) was selected as the open format of BIM. As discussed earlier, in an advanced system it is also required to perform a high-level energy simulation as opposed to a simple data analysis explained under systems with monitoring and control capabilities.

In this system, the simulation was performed using MATLAB/Simulink and the HAMLab tools (HAMBase), where the latter is used for modeling heat and vapor flows in buildings.

Moreover, ELAN is used as a computer model for building energy design in order to provide the physics of the HAMBase model. Energy usage and the user behavior are optimized by obtaining some patterns for long time spans. These patterns and profiles are obtained from daily data collections from smart metering and sensors. Eventually based on

20 the outputs, proper action will be taken by controlling unit such as adjusting the set point temperature [12].

Figure 2-6. Major concepts in ThinkHome knowledge base system [12]

The concept of monitoring more detailed energy-related data is also considered in other studies. Jahn et al. (2010) studied interconnecting electric devices in a house by using wireless power metering plugs to obtain the energy consumption of different appliances.

Hydra (the name changed to LinkSmart after 2010) is a middleware framework that is used in this research as an intelligent communication tool to create the smart home environment by enabling the user to control any type of physical device regardless of its network technology such as Ethernet, Bluetooth, RF, ZigBee, and RFID. The main application of such system is to develop an energy consumption profile for each device, which allows the user to identify the devices that use high energy [13]. The wireless power metering plugs used in this research are called Plogg that are designed to measure and monitor the electricity used and the data can be collected in a gateway and monitored in a mobile

21 interface, for example, as it is illustrated in Figure 2-7. The other feature of the system designed in this research is that users can use their mobile phones to view energy consumption of their appliances by using UbiLense that sends picture of the device to an image recognition server and the server has the information related to the Plogg connected to that device and returns its energy consumption. Figure 2-8 shows an example of such an application [13].

Figure 2-7. Measuring and monitoring electricity consumption by using metering plugs [14]

Figure 2-8. Monitoring electricity consumption by using UbiLense mobile app [13]

It should be pointed out that most of these systems in energy smart homes adopted for improving resident’s energy consumption habits or to actively reducing the energy consumption, are mainly dedicated to electric and mechanical systems. Moreover, the central processing unit, which works as the brain of the system, is mainly performing simple calculations and as a result, the outputs are not detailed. Developing an energy simulation tool, which is capable of performing high-level analysis using building information to provide users with detailed energy-related data, seems to be an area not investigated as much as other fields in energy smart homes and monitoring systems.

Performing a thorough research on the state-of-the-art technologies in energy smart homes and identifying the BEM tools capable of being integrated with such systems, seems to be the first step in order to identify the shortcomings in this area in more details. Since there are multiple outputs and tools that can be provided and used, the outcomes of this initial

23 research can be used to develop a tool directly targeting these areas, which provide useful detailed outputs using a proper energy simulation tool. Chapter 4 is dedicated to this initial study.

2.2 Retrofit of Existing Homes

The second important approach observed in the literature for reducing energy consumption is energy retrofit of existing homes. It was decided to limit this research to building envelope components such as walls, windows, and roofs as opposed to mechanical and electrical systems. The review here considers several types of materials, systems, and technologies suggested by researchers, and applied by engineers in retrofit projects.

Energy retrofit of buildings is composed of multiple steps and components. Decision- making process can be considered as the first and one of the most important steps in this process. This step is tied to application of computer tools, which performs energy simulation and provides users with information on energy performance of buildings to inform them how a new component or system can contribute to the current condition of their buildings, in terms of energy consumption.

Several conventional and innovative materials and systems are studied by researchers related to building energy retrofit. These materials and systems include conventional insulation materials such as Extruded Polystyrene (XPS), Polyisocynurate spray foam, and fiberglass batt insulation, which can be applied directly on building surface such as walls and floors. More innovative materials and systems include phase change material (PCM),

24 aerogel, dynamic façade, and precast elements. All of these options can be applied on opaque components such as walls, roof, and floors. However, envelope energy retrofit is not limited to opaque surfaces and insulation materials, while it can be extended to transparent components such as windows and adding reflective coatings with different solar heat gain coefficient (SHGC), replacing windows with more energy efficient options such as double or triple pane windows.

Researchers have performed comprehensive reviews on existing and more advanced energy retrofit options and materials [15, 16]; however, existing and under-development innovative methods and materials such as aerogel, PCM, dynamic façade, double-skin façade, and pre-cast components are not investigated and categorized as much as conventional options such as rigid insulation and more energy efficient window systems.

Performing such studies can contribute to identifying new needs for more innovative retrofit systems. Examples of such systems are presented in other studies such as Advanced

Passive Solar Collector & Ventilation Unit Technological Unit (APSC&VU TU),

Advanced Solar Protection & Energy Absorption Technological Unit (ASP&EA TU), and

Active Solar Thermal Façades (ASTFs) shown in Figure 2-9, Figure 2-10, and Figure 2-11, respectively [17, 18, 19]. The APSC&VU TU shown in Figure 2-9 is operable and can switch between the thermal insulation with cool coating and PCM material with warm coating, which absorbs less and more solar energy, respectively [17]. For example, the heat can be absorbed during the day and then the warm side would be switched toward the interior space to warm up the space. The ASP&EA TU, however, works based on both thermal storage and air movement [18]. Solar radiation can go through the semi-transparent

25 glazing component and heat will be absorbed by thermal storage behind it. Operable louvers on top and bottom of the panel can be opened and closed if needed and the air will be able to move over the thermal storage to exchange heat. Both of these panels can be installed on the facade as precast elements to speed up the construction and their electric control capabilities would be a great feature in homes equipped with energy management and automation systems. ASTF works based on solar heat gain and the space between two layers acts as an insulation layer that is heated up by solar radiation, and it can be ventilated if over-heated. In order to control and optimize the solar heat gain in double-skin façade, the effect of installing shading device in the cavity of this system was has been suggested as an option [20].

Figure 2-9. Operating strategies of the APSC&VU Tus [17].

Figure 2-10. Operating strategies of the ASP&EA Tus [18]

Figure 2-11. ASTF operational mechanism as a wall component [19]

Two important points of interest are noted here by studying these innovative retrofit systems. First, these systems are advanced, more expensive, and more effective than conventional retrofit methods; therefore, their design and application needs to be more optimized. Second, some of new technologies are capable of being integrated with electric control systems, which facilitate their application in emerging technologies such as energy smart homes. Therefore, it can be concluded that energy retrofit decision-making process for such systems needs to be based on detailed data for more accurate and optimized decisions and accumulated energy data do not provide proper level of details.

Various BEM tools are used and studied in literature related to building energy retrofit, which are capable of providing energy-related data discussed above. Review of academic

27 publications shows several computer tools can contribute to this area. To provide a better categorizing, building energy modeling graphical user interfaces (GUIs) should be distinguished from energy simulation engines. Tools such as EnergyPlus and DOE-2 are the energy simulation engines, which work in the background within energy simulation

GUIs such as OpenStudio, DesignBuilder, BEopt, eQuest, and Green Building Studio

(GBS). Energy simulation GUIs facilitate data input process and work as user-machine interface. Researchers used tools such as TRNSYS [21, 22], EnergyPlus [23, 24, 25, 26,

27], DesignBuilder [28, 29, 30], eQuest [31], and OpenStudio [32, 33] in their studies dedicated to energy retrofit.

The energy retrofit methods used in these studies with different BEM tools vary between application of both conventional and innovative materials and methods such as replacing existing window systems with double glazed filled with argon and krypton [21], installing exterior rigid insulation on walls [24], adding window shade [34], improving roof insulation by adding rigid insulation [24], application of aerogel as blankets over the floor and walls [28], application of PCM mixed with plaster [25], and installing APSC&VU TU

[27]. However, the reported energy saving potentials based on computer modeling are based on considering all the retrofit measures including the approaches focused on mechanical and electrical systems. Therefore, neither the modeling phase nor the evaluation phase can provide detailed information on how improving each envelope component is contributing to energy consumption of building, separately.

For example, reported energy conservation potential in most of these studies is according to the savings in annual heating and cooling site energy use, while most of the energy saving could be due to the improvement of HVAC system. This issue can be resolved during the decision-making process by performing optimization studies and evaluating different retrofit scenarios and combinations. Different optimization methods are applied in research studies such as multiple criteria complex proportional assessment (COPRAS) method in order to obtain the best retrofit scenarios [35, 36, 37, 38, 39]. Evaluation of all the combinations and possibilities can be time-consuming, while identifying the envelope components with highest contribution in energy consumption can help eliminating some of the options and focus the energy retrofit scenarios on components with higher contribution. Based on the observations from academic publications, existing computer tools are not capable of providing such detailed outputs for every component, separately.

In addition, several projects show the difference between the actual energy conservation and predicted one up to 42% [40, 41, 42, 43], which reveals possible inaccuracy in computer tools adopted in energy retrofit process. Some researchers tried to resolve such issues by calibrating models through using data obtained through measurements such as plug loads, unregulated loads such as elevators and security equipment, and updating the weather data. The results showed the difference between predicted energy consumption during the design phase and the actual energy consumption can be reduced from 36% to

7% by increasing the accuracy of energy modeling [44].

It shows improving the accuracy and level of details in outputs can be an area for conducting further studies. In addition, as noted earlier, several existing or under- development innovative retrofit systems show that dynamic components such as dynamic façade and operable shading systems can be integrated with buildings and control systems, which will be capable of controlling building envelope components besides other components such as HVAC and electrical systems. Therefore, it is important to design future tools to provide users more detailed data on the amount of heat transfer through envelope components such as exterior walls and windows. This study is dedicated to this aspect and contribution of energy simulation tools.

Proposing a categorization for different existing or under-development retrofit materials, systems, and technologies seems to be necessary in order to customize the future tools related to this field according to the direction these technologies and methods are heading.

Moreover, it seems all the tools available for energy retrofit purposes are focused on accumulated energy-related data as opposed to detailed information. Review of existing tools in this field can help in identifying the shortcomings and deciding on capabilities of future tools.

2.3 Energy Modeling in Design Phase

Energy evaluation of design prior to construction can help in modifying building components, which is more convenient and economical than performing any changes after construction. Figure 2-12 shows how design process would face less struggle if early stages were dedicated to solving potential problems. Buildings can benefit from the same concept 30 in terms of energy performance if proper energy evaluation is performed during the design phase. Similar concept is pointed out by Al-Homoud [15] shown in Figure 2-13, which reveals higher cost of modification through later stages of a project.

Figure 2-12. Early-stage problem solving in design phase facilitates later stages of design activity [45]

Figure 2-13. Higher modification cost through later stages of building life [15]

Figure 2-14 shows typical stages of design of a product. Not all of these steps necessarily exist in designing of a building. Conceptual and preliminary designs are the major stages in design of the building, before evaluation. A CAD tool such as AutoCAD or Revit is typically used to develop construction drawings and after each evaluation, depending on the area of focus, components can be modified and improved, which in this case is in terms of energy performance. For example, Granadeiro et al. studied integration of architectural design system and energy simulation tool in order to find an optimum building envelope shape in terms of energy consumption [46]. They used EnergyPlus as the energy simulation tool integrated to a base text file they developed with constant building properties. The variations in building envelope were applied to the energy model by changing the desired properties and compiling a new file to the text file and the outputs including heating and cooling loads were obtained and compared to identify the best option in terms of minimum energy consumption [46].

Figure 2-14. Typical stages of design [45]

The energy evaluation, which is composed of energy modeling and analysis, can be performed using BEM tools. A brief review of existing tools was discussed in previous section related to building energy retrofit and other researchers have also performed comprehensive studies on such tools [47, 48]. Development of new BEM tools is growing and about 60 different BEM tool with whole-building energy simulation capability for all

32 platforms are identified by IBPSA-USA [49]. Researchers have constantly been pursuing improvements in these BEM tools in terms of accuracy, speed, and convenience of application, where such goals have also been pointed out in other studies [50, 15]. For example, to simplify the data input related to hourly schedules, Fumo et al. [51] suggested a conversion coefficient to convert the monthly use to hourly energy usage to simplify the process for energy modeling.

The BEM process and related computer tools are the vital components in energy evaluation of buildings. Multiple tools and approaches are studied in the literature for the BEM process. Among approaches used with higher frequency, some innovative methods are suggested by researchers for prediction of building’s energy consumption, which mostly work based on machine learning approach. For example, application of neuro-fuzzy method using main parameters such as material thickness, thermal properties of material has been studied based on root-mean-square error for predicted outputs with satisfactory results [52]. Other innovative methods such as conditional demand analysis (CDA) for modeling residential buildings energy end-use are studied by researchers and compared with neural-network (NN) methods. Methods such as CDA are regression-based, which means they need great amount of data and samples to be able to predict the energy consumption [53]. Beside such innovative BEM approaches, there are other methods with higher use frequency, which are categorized under three major groups in this study including conventional, semi-automated, and automated BEM methods.

The conventional approach involves reentering all the required data for energy simulation including data related to building geometry, material properties, and HVAC system. An energy simulation GUI need to be used to develop the geometry of building and defining all the components. Then energy performance can be performed within the GUI and outputs will be limited to what is defined for that specific tool.

Semi-automated and automated BEM process, on the other hand, benefit from emerging tools such as BIM. Efforts are on-going to facilitate BEM process, which can include the modeling phase, simulation or analysis phase, and presenting outputs. Improving and facilitating BEM process can include expediting any of these phases and making them less error-prone by minimizing human interaction. For example, a graphical model can be created in DesignBuilder based on building geometry and other data such as schedules, loads, thermostat set points, and HVAC system will be defined. Energy analysis will be performed and energy-related data such as thermal zone and whole-house energy consumption will be provided.

Semi-automated approach can benefit from adoption of BIM, an authoring CAD tool, which develops a BIM file containing all the required information for energy simulation.

The user needs to use an energy modeling tool to import the file and add some of the data that might not be transferred properly and perform the energy simulation, which starts by converting the BIM file to a readable file for BEM tool. Outputs will be limited to what the certain BEM tool is capable of presenting. For example, a BIM file is generated by a BIM authoring tool such as Revit, and will be imported in OpenStudio, manually, and additional

34 data related to other components such as HVAC and lighting system will be added. Energy simulation will be performed and the energy consumption in each thermal zone will be calculated. Figure 2-16 shows an example of such process used for air flow simulation. It can be observed that a BIM file (IFC) is generated by Revit and exported. It includes building information such as geometry and material, which helps avoiding reentering such data; however, other data such as HVAC system data need to be added manually. Python script converts the file to proper format for CONTAM for airflow analysis [54]. Manual interaction is not limited to adding required information for energy modeling. There are examples of adding additional data related to material properties to the BIM-to-BEM process. For example, Kim et al. [55] developed an interface, which modifies the BIM file to add additional options for materials. The adopted process is shown in Figure 2-15. A

BIM file containing building information that is in IFC format in their study is modified based on new user’s inputs and an updated data input file (INP) is used for energy simulation in DOE-2.2. It was observed that the developed process is capable of enabling users to compare performance of new building assemblies instead of relying on existing libraries [55]. Multiple tools and interfaces might be required in order to add the missing information in a semi-automated approach, since the BIM file might not be able to transfer all the data. Moreover, the BEM tool might not be able to import all the required data; therefore, other tools need to be adopted, which is shown in Figure 2-17.

Figure 2-15. Example of semi-automated BEM process to add additional options in material library [55]

Figure 2-16. Example of semi-automated BEM process for adding missing information required for energy

modeling [54]

Figure 2-17. Example of semi-automated BEM process for using different tools and interfaces to add

missing information required for energy modeling [56]

An automated approach, on the other hand, does not need human interaction for converting the BIM file and adding required information for energy analysis. Tools such as GBS can import the architectural model containing all the required information for energy modeling directly from an authoring tool such as Revit and perform the file conversion and energy simulation. In this case, DOE-2 works as the simulation engine and provides users with predefined outputs, which in this case is too general and only includes whole-house energy consumption and share of each sector in total energy consumption such as HVAC, lighting, and equipment. For example, Abanda and Byers [57] used both semi-automated and

37 automated approaches in their study using GBS and Ecotect. It should be noted that Ecotect is discontinued and new versions are not provided anymore and alternatives such as GBS are available for energy simulation. Figure 2-18 shows the approach used in their study, which is focused on impact of building orientation on energy consumption. The BIM files that is in gbXML format is generated by Revit and can be used directly by GBS or manually imported in a BEM tool such as Ecotect. GBS uses the data defined in Revit, but in Ecotect, it needs manual interaction and energy simulation will not be performed directly, which represents automated and semi-automated BEM process, respectively. It should also be pointed out that GBS that represents automated BEM process using BIM leads to accurate results, which is compatible with actual data based on utility bills [57].

Figure 2-18. Typical process for semi-automated and automated energy simulation [57] 38

Different issues, challenges, and shortcomings are reported in the literature addressing the data transfer issues in BIM files and interoperability issues between BIM file and BEM tools. Shortcomings in the link between architecture, engineering, and construction (AEC) industry and BIM are also pointed out in some works [58], while there are also studies that identify the challenges of application of BIM in existing building in general [59].

Furthermore, automation of data capture and BIM creation is identified as one of the major challenges in field of application BIM for existing buildings [59].

The literature shows that the BEM process is moving toward automation, and BIM seems to be the major tool used in some studies and identified as an idea tool for optimizing buildings design [60, 61]. However, challenges and shortcomings are also identified, which need to be resolved to get to the ideal point of fully automated building energy modeling.

Resolving these issues can be obtained by developing and adopting new tools such as middleware tools, which works between different components in BBIP.

It is highly important to perform a comprehensive investigation for similar issues mentioned above, and an organized categorization seems to be required concerning the challenges and issues in BBIP. In addition, it can help defining the major focus of tools, which can be designed for automating BEM process. BIM seems to be one of the major tools used in the BEM process in several studies, which influenced the decision to use the same approach in this study as well in order to automate the BEM process.

The major objectives that are explained in next chapter are developed based on the literature review presented in this chapter. Such objectives are focused on automating and

39 facilitating energy modeling process and providing detailed outputs on the amount of heat transfer through building envelope components.

2.4 References

[1] P. Schultz, M. Estrada, J. Schmitt, R. Sokoloski and N. Silva-Send, "Using in-home displays to provide smart meter feedback about household electricity consumption: A randomized control trial comparing kilowatts, cost, and social norms," Energy, vol. 90, pp.

351-358, 2015.

[2] EIA, "Annual Energy Outlook 2016," Energy Information Administration (EIA),

Washington, DC, 2016.

[3] M. Ashby, Materials and the Environment, Elsevier Inc. Second Edition, 2013.

[4] V. Ricquebourg, D. Maenga, D. Durand, B. Marhic, L. Delahoche and C. Loge,

"The Smart Home Concept: our immediate future," Hammamet, 2006.

[5] J. Stinson, A. Willis, J. Williamson, J. Currie and R. Smith, "Visualising energy use for smart homes and informed users," Energy Procedia, vol. 78, p. 579 – 584, 2015.

[6] D. Han and J. Lim, "Smart Home Energy Management System using IEEE

802.15.4 and ZigBee," IEEE Transactions on Consumer Electronics, vol. 56, no. 3, pp.

1417-1425, 2010.

[7] Y. Tsou, J. Hsieh, C. Lin and C. Chen, "Building a Remote Supervisory Control

Network System for Smart Home Applications," Taipei, Taiwan, 2006.

[8] L. De Silva, C. Morikawa and I. Petra, "State of the art of Smart homes,"

Engineering Applications of Artificial Intelligence, vol. 25, pp. 1313-1321, 2012.

[9] W. Kim, S. Lee and J. Hwang, "Real-time Energy Monitoring and Controlling

System based on ZigBee Sensor Networks," Procedia Computer Science, vol. 5, p. 794–

797, 2011.

[10] L. De Silva, T. Dewana, M. Petra and G. Punchihewa, "Multiple Sensor Based

Autonomous Monitoring and Control for Energy Efficiency," 2010.

[11] S. Helal, W. Mann, H. El-Zabadani, J. King, Y. Kaddoura and E. Jansen, "The

Gator Tech Smart House: A Programmable Pervasive Space," IEEE Computer Society, vol. 38, no. 3, pp. 50-60, 2005.

[12] C. Reinisch, M. Kofler, F. Iglesias and W. Kastner, "ThinkHome Energy Efficiency in Future Smart Homes," EURASIP Journal on Embedded Systems, vol. 2011, pp. 1-18,

2011.

[13] M. Jahn, M. Jentsch, C. Prause, F. Pramudianto, A. Al-Akkad and R. Reiners, "The

Energy Aware Smart Home," Busan, South Korea, 2010.

[14] ETH, "EnergieVisible," ETH, [Online]. Available: https://www.vs.inf.ethz.ch/res/show.html?what=energie-visible. [Accessed 22 April

2016].

[15] M. Al-Homoud, "Performance characteristics and practical applications of common building thermal insulation materials," Building and Environment, no. 40, p. 353–

366, 2005.

[16] B. P. Jelle, "Traditional, state-of-the-art and future thermal building insulation materials and solutions – Properties, requirements and possibilities," Energy and Buildings, no. 43, pp. 2549-2563, 2011.

[17] M. Rozanska, "2nd Reporting Period Publishable Summary on Progress,

Multifunctional Energy Efficient Façade System for Building Retrofitting across Europe,"

2015.

[18] M. Rozanska, "1st Year Progress Report Publishable Summary on Progress,

Multifunctional Energy Efficient Façade System for Building Retrofitting across Europe,"

2013.

[19] X. Zhang, J. Shen, Y. Lu, W. He, P. Xu, X. Zhao, Z. Qiu, Z. Zhu, J. Zhou and X.

Dong, "Active Solar Thermal Facades (ASTFs): From concept, application to research questions," Renewable and Sustainable Energy Reviews, vol. 50, pp. 32-63, 2015.

[20] G. Kim, H. S. Lim and J. T. Kim, "Sustainable lighting performance of refurbished glazed walls for oldresidential buildings," Energy and Buildings, no. 91, pp. 163-169,

2015.

[21] C. Guattari, G. Battista, L. Evangelisti, A. Fanchiotti and L. Santarpia, "Energy

Retrofit of Historical Buildings Based on Windowed Elements," Applied Mechanics and

Materials, vol. 737, pp. 154-158, 2015.

[22] D. I. Kolaitis, E. Malliotakis, D. A. Kontogeorgos, I. Mandilaras, D. I. Katsourinis and M. A. Founti, "Comparative assessment of internal and external thermal insulationsystems for energy efficient retrofitting of residential buildings," Energy and

Buildings, no. 64, pp. 123-131, 2013.

[23] M. Pertosa, A. Laura Pisello, V. Lucia Castaldo and F. Cotana, "Environmental sustainability concept applied to historic buildings: the experience of LEED international protocol in the stable of Sant’Apollinare fortress in Perugia," in 14th CIRIAF National

Congress. Energy, Environment and Sustainable Development, Perugia, Italy, 2014.

[24] "Advanced Energy Retrofit Guide, Practical Ways to Improve Energy

Performance, K-12 Schools," National Renewable Energy Laboratory for U.S. Deparment of Energy, 2013.

[25] F. Ascione, N. Bianco, R. F. De Masi and F. de’Rossi, "Energy retrofit of an educational building in the ancient center ofBenevento. Feasibility study of energy savings and respect of the historical value," Energy and Buildings, no. 95, pp. 172-183, 2015.

[26] P. Marin, M. Saffari, A. de Gracia, X. Zhu, M. M. Farid, L. F. Cabeza and S. Ushak,

"Energy savings due to the use of PCM for relocatable lightweightbuildings passive heating and cooling in different weather conditions," Energy and Buildings, vol. 129, pp. 274-293,

2016.

[27] S. Paiho, I. P. Seppä and C. Jimenez, "An energetic analysis of a multifunctional fac¸ ade system for energy efficient retrofitting of residential buildings in cold climates of

Finland and Russia," Sustainable Cities and Society, no. 15, pp. 75-85, 2015.

[28] S. S. Filate, "Investigation of an energy refurbishment concept for office building using Nanogel®Aerogel insulation plaster and replaced windows by building simulation,"

UPPSALA UNIVERSITET, Master Programme in Energy Technology, 2014.

[29] E. Lo Cascio, Z. Ma, D. Borelli and C. Schenone, "Residential building retrofit through numerical simulation: a case study," Energy Procedia, vol. 111, pp. 91-100, 2017.

[30] G. EVOLA and G. MARGANI, "ENERGY RETROFIT TOWARDS NET ZEB,

Application of BIPV on an Apartment Block in Southern Italy," in Architectural Research through to Practice: 48th International Conference of the Architectural Science

Association, Genoa, 2014.

[31] A. F. Setiawan, T.-L. Huang, C.-T. Tzeng and C.-M. Lai, "The Effects of Envelope

Design Alternatives on the Energy Consumption of Residential Houses in Indonesia,"

Energies, vol. 8, pp. 2788-2802, 2015.

[32] A. F. Radwan, A. A. Hanafy, M. Elhelw and A. E.-H. A. El-Sayed, "Retrofitting of existing buildings to achieve better energy-efficiency in commercial building case study:

Hospital in Egypt," Alexandria Engineering Journal, p. In Press, 2016.

[33] U. Eicker, E. Demir and D. Gürlich, "Strategies for cost efficient refurbishment and solar energyintegration in European Case Study buildings," Energy and Buildings, no. 102, pp. 237-249, 2015.

[34] "Advanced Energy Retrofit Guide, Practical Ways to Improve Energy

Performance, Office Buildings," Pacific Northwest National Laboratory for U.S.

Department of Energy , 2011.

[35] A. Mohamed, M. Hamdy, A. Hasan and K. Sirén, "The performance of small scale multi-generation technologies in achieving cost-optimal and zero-energy office building solutions," Applied Energy, no. 152, pp. 94-108, 2015.

[36] P. Penna, A. Prada, F. Cappelletti and A. Gasparella, "Enhancing the Energy and

Non-Energy Performance of Existing Buildings: a Multi-Objective Approach".

[37] F. Ascione, N. Bianco, C. De Stasio, G. Maria Mauro and G. P. Vanoli, "A new methodology for cost-optimal analysis by means of themulti-objective optimization of building energy performance," Energy and Buildings, no. 88, pp. 78-90, 2015.

[38] A. Kaklauskas, E. K. Zavadskas, S. Raslanas, R. Ginevicius, A. Komka and P.

Malinauskas, "Selection of low-e windows in retrofit of public buildings by applying

45 multiple criteria method COPRAS: A Lithuanian case," Energy and Buildings, no. 38, pp.

454-462, 2006.

[39] "15251–2007:Indoor environmental input parameters for design and assessment of energy performance of buildings addressing indoor air quality, thermal environment, lighting and acoustics," Comité Européen de Normalisation (CEN) Standard EN, 2007.

[40] M. D. Bianco and J. Wiehagen, "Using Retrofit Nail Base Panels to Expand the

Market for Wall Upgrades," U.S. Department of Energy’s Building America Program,

Denver, 2016.

[41] T. Häkkinen, M. Ala-Juusela and J. Shemeikka, "Usability of energy performance assessment tools for different usepurposes with the focus on refurbishment projects,"

Energy and Buildings, vol. 127, pp. 217-228, 2016.

[42] J. Khoury, H. Pierre and L. Bernard Marie, "Energy performance gap in building retrofit : characterization and effect on the energy saving potential," in Building and

Renewable Energies Network of Technology, Zürich, 2016.

[43] M. Alam, P. X.W. Zou, J. Sanjayan, R. Stewart, O. Sahin, E. Bertone and J. Wilson,

"Guidelines for Building Energy Efficiency Retrofitting," in Sustainability in Public Works

Conference, Melbourne, Victoria, 24 – 26 August 2016.

[44] H. W. Samuelson, A. Ghorayshi and C. F. Reinhart, "Analysis of a simplified calibration procedure for 18 design-phase building energy models," Building Performance

Simulation, vol. 9, no. 1, pp. 17-29, 2014. 46

[45] N. Slack , A. Brandon-Jones and R. Johnston, Operations Management, Harlow,

England: Pearson, 2013.

[46] V. Granadeiro, J. P. Duarte, J. R. Correia and V. M.S. Leal, "Building envelope shape design in early stages of the design process: Integrating architectural design systems and energy simulation," Automation in Construction, vol. 32, p. 196–209, 2013.

[47] T. Hong, S. Chon and T. Bong, "Building simulation: an overview of developments and information sources," Building and Environment, vol. 35, pp. 347-361, 2000.

[48] D. B. Crawley, J. W. Hand, M. l. Kummert and B. T. Griffith, "Contrasting the capabilities of building energy performance simulation programs," Building and

Environment, vol. 43, pp. 661-673, 2008.

[49] IBPSA-USA, "Software Listing," IBPSA-USA, [Online]. Available: http://www.buildingenergysoftwaretools.com/software- listing?combine=&field_catagory_tid%5B%5D=201&field_platform_tid=All&field_pric e_tid=All&field_last_updated_value%5Bvalue%5D%5Byear%5D=&field_language_tid

=All&sort_by=field_rating_rating&sort_order=DESC&. [Accessed 1 May 2017].

[50] H. Wang and Z. Zhai, "Advances in building simulation and computational techniques: Areview between 1987 and 2014," Energy and Buildings, vol. 128, pp. 319-

335, 2016.

[51] N. Fumo, P. Mago and R. Luck, "Methodology to estimate building energy consumption using EnergyPlus Benchmark Models," Energy and Buildings, vol. 42, pp.

2331-2337, 2010.

[52] S. Naji, S. Shamshirband, H. Basser, A. Keivani, U. J. Alengaram, M. Z. Jumaat and D. Petkovic, "Application of adaptive neuro-fuzzy methodology for estimating building energy consumption," RenewableandSustainableEnergyReviews, vol. 53, pp.

1520-1528, 2016.

[53] M. Aydinalp-Koksal and V. I. Ugursal, "Comparison of neural network, conditional demand analysis, and engineering approaches for modeling end-use energy consumption in the residential sector," Applied Energy, vol. 85, pp. 271-296, 2008.

[54] N. Yu, "INFORMATION INTEROPERABILITY BETWEEN BUILDING

INFORMATION MODELING AUTHORING TOOLS AND SIMULATION TOOLS TO

SUPPORT ENERGY EFFICIENT BUILDING DESIGN," Submitted in Partial

Fulfillment of the Requirements for the Degree of Master of Science, Information Sciences and Technology, The Pennsylvania State University, University Park, 2014.

[55] H. Kim, Z. Shen, I. Kim, K. Kim, A. Stumpf and J. Yu, "BIM IFC information mapping to building energy analysis (BEA) model with manually extended material information," Automation in Construction, vol. 68, pp. 183-193, 2016.

[56] V. Bazjanac, "IFC BIM-Based Methodology for Semi-Automated Building Energy

Performance Simulation," in 25th International Conference on Information Technology in

Construction, Santiago, Chile, 2008.

[57] F. Abanda and L. Byers, "An investigation of the impact of building orientation on energy consumption in a domestic building using emerging BIM (Building Information

Modelling)," Energy, vol. 97, pp. 517-527, 2016.

[58] T. Gerrish, K. Ruikar, M. Cook, M. Johnson, M. Phillip and C. Lowry, "BIM application to building energy performance visualisation andmanagement: Challenges and potential," Energy and Buildings, vol. 144, pp. 218-228, 2017.

[59] R. Volk, J. Stengel and F. Schultmann, "Building Information Modeling (BIM) for existing buildings — Literature review and future needs," Automation in Construction, vol.

38, pp. 109-127, 2014.

[60] G. Gourlis and I. Kovacic, "Building Information Modelling for analysis of energy efficient industrial buildings – A casestudy," Renewable and Sustainable Energy Reviews, vol. 68, pp. 953-963, 2017.

3) Chapter 3. Methodology for Improving BEM Process

and Outputs

3.1 Introduction

The literature review presented in the previous chapter revealed certain areas in BEM process, which can be improved in order to contribute to energy evaluation and monitoring tools and systems. Three major areas related to energy conservation in buildings are focused in this research including the energy monitoring systems in energy smart homes, energy evaluation prior to energy retrofit, and energy evaluation during the design phase.

It was shown that all three areas could benefit from an easy, fast, and accurate BEM process with detailed outputs. BIM was identified as one of the major tools being used and studied by researchers in order to facilitate and expedite the BEM process, which led to adoption of BIM in this research to automate modeling and analysis process. The automated BEM process in this research refers to the process, in which the BIM file is directly converted to a proper file format for energy simulation without the need to use another tool or GUI; the energy simulation is performed automatically using the converted file. The second major focus of this research is developing a capability for BEM tools to provide detailed and accurate outputs for the amount of heat transfer through every single building envelope components as opposed to accumulated data for each thermal zone or the whole-house energy consumption. It should be noted that all the calculations and process is focused on

50 building envelope components such as walls, windows, roof, floor, and doors. The major objectives, required tasks, and the methodology in general are explained in this chapter.

3.2 Objectives and Tasks Outline

In order to simplify the research process and determine the focus of study in more details, it was decided to break down the major objectives and required tasks in smaller areas of focus and explain each task in more details. Based on the review of related academic publications, the following two objectives are determined to be the major objectives pursued in this research:

Objective #1: “Improve the outputs of BEM tools”

This objective is focused on “improving” the quality of outputs of BEM tools to contribute to energy evaluation during the design phase, decision-making process of building’s energy retrofit, and energy monitoring systems in energy smart homes. The intention is to ensure that the adopted method leads to a tool compatible with others involved in design phase, energy retrofit, and energy monitoring phases of a building. The tasks defined for this objective are designed and selected to fulfil this objective.

The word “improve” used above refers to the output details. The objective is to provide users with fine-grained information on energy-related data, which consists of separate heat transfer values for every surface of buildings corresponding to walls, windows, doors, roof, and floors. Other surfaces and components might be added to this list depending on the

51 outcomes of the tasks defined for this objective. The required tasks for this objective are explained in the following sections.

Objective #2: “Develop a platform for using BIM to minimize the human-machine interaction in BEM process”

This objective is focused on facilitating and improving the BEM process using BIM to contribute to energy evaluation during the design phase, decision-making process of building’s energy retrofit, and energy monitoring systems in energy smart homes. BIM was shown to be a robust tool, which can eliminate the need for reentering data for BEM process and can be later used for convenient visual presentation of outputs.

Improving the process refers to different aspects of BEM process including speed, convenience, and accuracy. Minimizing manual or human-machine interaction is the major focus of this objective. The required tasks for his objective are explained in the following sections.

3.3 Further Explanation of the Tasks

To accomplish the two major objectives of this research several detailed tasks need to be undertaken. These steps are decided based on previous studies on similar issues and the lessons learned from those works reviewed in Chapter 2.

3.3.1 Tasks for Objective #1

To “Improve the outputs of BEM tools”, the following tasks are defined: 52

Task (a). Perform a comprehensive study on building energy retrofit measures and tools

This task helps to identify the trends in building envelope energy retrofit options and the computer tools involved in this area. The task helps to evaluate the capabilities of existing tools, especially the open-source energy simulation tools, which can be used for fulfilling

Objective #1. Moreover, the results of this study can help to identify other components than walls, windows, doors, floors, and roof, which might be the focus of recent advancements in building energy retrofit. In addition, proper file format for fine-grained outputs can be decided based on the selected open-source simulation tool.

Task (b). Modify an open-source energy simulation tool

This task consists of major efforts to fulfil Objective #1. To facilitate the process and based on the comprehensive study performed under Task (a), EnergyPlus was selected as an open-source energy simulation tool. The source code of this energy simulation tool was modified in order to provide fine-grained outputs related to walls, windows, doors, floors, and roof. The tool developed under this task will be referred to as BEPAT (Building Energy

Performance Assessment Tool). It is important to make the generation process of fine- grained outputs automated to minimize the human-machine interaction.

3.3.2 Tasks for Objective #2

To accomplish the objective “Develop a platform using BIM to minimize the human- machine interaction in BEM process”, the following tasks were defined:

Task (a). Perform a comprehensive study on energy smart home

This task aims at identifying the required inputs and outputs in an energy smart home equipped with a simulation tool for energy monitoring purposes. To do so, the state-of-the- art systems in energy smart homes was reviewed and the outputs of the developed platform adjusted based on the need and gaps in existing tools. Moreover, review of challenges and components in energy smart homes helped designing a platform compatible with existing systems in terms of other computer tools and electronic systems. The platform’s outputs are saved in text files to enable using them in other tools or GUI’s in follow-up studies.

Task (b). Perform a comprehensive study on BIM-to-BEM process

This task helped preparing a comprehensive layout for issues and challenges in BBIP, as it seemed to be lacking in literature. Moreover, the task led to selecting gbXML as the proper

BIM file format, which works as the input file for this platform. In addition, such a comprehensive study facilitated the most important aspect of this platform, which is automated conversion of BIM file to a proper format for BEM.

A comparative study was performed to evaluate the issues and differences between automated and semi-automated BEM methods and outputs that were used to optimize the design of the tools resulting from Tasks (c) and (d). The BEM tools used for this comparative study are selected based on this comprehensive study.

Task (c). Develop a corrective tool for BIM files to resolve the related issues in BBIP

Based on the challenges and issues identified in Task (b), a proper corrective tool had to be developed in order to solve the interoperability issues concerning data transfer related to building envelope, which was the focus of this research. In follow-up studies, this tool can be further enhanced to other components such as HVAC and lighting systems. The corrective tool developed under this task was used as a middleware between the BIM authoring tool, which is a CAD tool in this case, and the BEM tool before being converted to a new file format for BEM. The programing language used for developing this corrective tool need to be decided based on the file formats and capability of available programming languages.

Task (d). Develop an automated method for converting BIM to proper BEM file

Tasks (b) led to finding available open-source tools capable of performing BIM file to

BEM file conversion such as OpenStudio. The comprehensive study performed under Task

(b) helped to back-up selecting the tool, which is modified and tailored for the specific purpose of this platform.

The output of this task was a set of commands in a certain programming language, which calls predefined functions within an open-source tool capable of converting BIM file to

BEM file. The programing language used for writing these commands was decided based on the selected tool for file conversion.

Task (e). Validation of all developed tools under a single platform

The combination of the corrective tool developed under Task (c) for Objective #2, the BIM file to BEM file converter developed under Task (d) for Objective #2, and the developed energy simulator capable of providing fine-grained energy-related outputs developed under

Task (b) for Objective #1 were tested as a single platform as the main outcome of this research. This is referred to as ABEMAT (Automated Building Energy Modeling and

Assessment Tool). All the data transfer and output generation process related to building envelope are automated and the data for other systems can be added manually.

A model of a house had to be developed in a BIM authoring tool such as Revit or AchiCAD, where in this study Revit was used. The house model was simplified to make sure the number of surfaces are limited as much as possible to make it viable for visual inspection of some data transferred between different tools. The outputs were verified with similar detailed outputs from EnergyPlus.

A summary of all the objectives and their corresponding tasks is provided in next section.

3.4 Research Process Summary

Based on the objectives and their corresponding tasks, major steps and phases of this research undertaken can be shown as Figure 3-1. As the figure shows, the objectives and some tasks are intertwined and their functionality relies on other tasks or objectives.

Therefore, it is important to perform the process under separate research focus, which led

56 to the suitability for the results of this research to be presented in the form of multiple research articles on different aspects of the work.

The developed energy simulation tool developed under Objective #1 and capable of providing fine-grained outputs had to be used as a component in a bigger platform. The comprehensive study on building energy retrofit systems and computer tools contributed to selecting a proper open-source energy simulation tool, and proper building components were considered for fine-grained outputs based on the trend of energy retrofit systems.

Objective #2 led to developing other major components of this platform including a corrective tool for resolving issues with BIM files and a file converter, which converts a

BIM file to BEM file suitable for the tool developed under Objective #1. Two comprehensive studies were performed on energy smart homes and BIM-to-BEM process to contribute to Objective #2.

Review of energy smart homes can help in identifying the types of inputs and outputs in a platform, which is going to serve as the brain of energy smart home tool for monitoring purposes. Moreover, the challenges and components in energy smart homes can affect the selection process for open-source tools to make sure they are compatible with other components such as GUIs and control systems being used in energy smart homes. In addition, a comparative study was performed to evaluate and compare the performance of innovative and conventional energy retrofit systems, which can be helpful for other researchers in future to be focused on certain retrofit methods and develop more customized tools.

Figure 3-1. Two major BEM focused areas improved in this research

Review of BBIP and can contribute to identifying the challenges, shortcomings, and capabilities of existing BEM tools and also in selecting a proper format for BIM files in this platform. Other formats can also be added in future versions of this platform. The open- source tool, which is going to be used as the BIM file to BEM file converter was selected based on this comprehensive study.

Finally, all the aforementioned tools were used as a single tool or platform and a validation study performed to verify the outputs and confirm applicability of such tools in BEM industry and fields related to that.

4) Chapter 4. Energy Smart Homes

This chapter has been written as a journal paper and is already submitted for review.

State of the Art Review of Energy Smart Homes

1 2 Ehsan Kamel , Ali M. Memari

1Ph.D. Candidate, Department of Civil Engineering, Structural Engineering, Penn State University, 321 Sackett Building, University Park, PA 16802 (corresponding author). E-mail: [email protected] 2Professor, Hankin Chair in Residential Building Construction, and Director, Pennsylvania Housing Research Center, Dept. of Architectural Engineering and Dept. of Civil and Environmental Engineering, Penn State Univ., University Park, PA 16802. E-mail: [email protected]

4.1 Abstract

With advancement in sensor, control, and communication technologies, homes can now be equipped with more intelligent lighting, HVAC, entertainment, and safety/security systems as well as appliances with many advanced features, including high energy efficiency. In particular, with computers becoming an integrated part of homes and smart phones offering remote communication with home systems, we are entering a new era in home industry, an era of “smart homes”. A main objective of this paper is to present a categorization of smart homes based on review of the literature. Different types of smart homes are introduced starting with homes that are equipped with technologies for information and communication, security, health, environmental, home entertainment, and domestic appliances. Next, different types of “energy smart homes” are reviewed in more details under three major groups including the houses with energy monitoring systems, systems with control capabilities, and systems with advanced data processing capabilities. Smart homes with energy monitoring systems merely provide the total or “fine- grained” energy consumption of the house by using equipment such as in-home displays, while

61 systems with control capabilities also include a control unit that can send proper signals for either passive or active measures such as appliances on/off command. On the other hand, systems with advanced data processing capabilities also include an advanced central processing unit that can provide more complicated analysis results such as systems that are equipped with optimization algorithm to optimize the temperature or appliances schedule based on the residents’ comfort level and energy cost. Different components of smart homes and also the challenges in smart homes’ design are also discussed in this paper.

Author Keywords: Smart Homes, Energy Smart Home, Energy Monitoring, Control Systems

4.2 Introduction

The desire for higher level of comfort, convenience, utility and efficiency in homes has been driving most of the technologies built around such goals. Technologies inside a house are designed in a way to facilitate the daily life and improve the life quality by reducing the need for direct human interaction. Automation, remote access, and monitoring of every-day’s life activities can be among the top priorities for home residents that can contribute to their well-being. Recent advancements in the technologies such as sensors, control units, communications technologies, and also software and hardware advancements have made it much easier for homes to be more intelligent and smart to explain themselves to the user or decide and execute actions to improve the residents’ life quality. Smart homes are defined differently in the literature. Table 4-1 shows a summary of some of these definitions. Some of the keywords commonly used in the literature on smart homes are the following: ambient intelligence, using sensors and actuators, home networking, transporting data from smart objects to residential gateway, connecting the house to the internet, communication of house components through a local network, remote or central controlled functionalities and services, allowing electrical appliances and services to be remotely controlled and monitored, automatic control, adjusting house functions to residents’ needs, response to the behavior of residents, facilitating remote home control, assisting in performing the activities of daily life, and promoting the residents comfort, convenience, security, and entertainment.

Considering various definitions available in literature, the following is suggested here as an encompassing definition for smart homes: “A dwelling in which data related to home environment and its residents are obtained from sensors, electric appliances, or home gateway and transferred

63 through a network of communication tools to monitoring device or execution unit to help decide on or execute proper actions called services. These services are provided either automatically or directly through a remote or central control system in order to facilitate or improve the residents’ daily lives”.

As it can be observed, the definition of smart home covers three main areas. First, a smart home is intended to facilitate the residents’ lives and enhance comfort and satisfaction. Depending on the area of focus of smart home, it can facilitate or improve security, health, entertainment, and energy efficiency, among other aspects. This can be done by enabling a remote, central, or automatic control/access to different types of services such as turning on/off a fan, the security system, or the

HVAC system. The automatic control of services and appliances is referred to as the “home automation”. Although home automation should lead to residents’ convenience, there are studies that encourage researchers not to focus only on total home automation; instead, the interface technologies should be improved in order to include user’s intentions [62]. Second, there should be a network to facilitate the communication among different components. This network basically receives data from objects and residents and transports it to other components such as residential gateway to transfer it to the desirable destination that could be a processing unit to make further decisions or directly to the residents for monitoring purposes. Third, there should be an environment related to home that includes either objects or residents. The objects within this environment should be able to send and/or receive data or they should be capable of being controlled remotely or directly through a central system. These data are required for deciding on proper actions in order to facilitate the residents’ daily lives. Based on these major areas a simplified architecture of smart homes is shown in Figure 4-1. Although there are no standard or unified published guidelines for designing or defining a smart home, there are guidelines focused

64 on certain intended services of these homes such as “Design Guideline on Smart Homes” , which is an action by the European Cooperation in Science and Technology focused on elderly and disabled people [63].

Table 4-1. Different Definitiosns of Smart Homes

Definition Reference

“A home equipped with lighting, heating, and electronic devices that can be controlled remotely by [64]

smartphone or computer”

“When you assemble a collection of smart devices under one roof, and you enable them to connect [65]

to and communicate with one another, what you end up with is typically called the smart home.”

“A smart home is a home- like environment that possesses ambient intelligence and automatic [8]

control, which allow it to respond to the behavior of residents and provide them with various

facilities”

“A smart home is an application of ubiquitous computing in which the home environment is [66]

monitored by ambient intelligence to provide context-aware services and facilitate remote home

control.”

“The integration of technology and services through home networking for a better quality of living.” [67]

“Smart Home uses different devices including sensors and actuators to provide a diversity of [68]

functions that assist in performing the activities of daily living.”

“The term ‘smart home’ is used for a residence equipped with technology that allows monitoring of [69]

its inhabitants and/or encourages independence and the maintenance of good health.”

Definition Reference

“In a schematic way, a smart home can be described by a house which is equipped with smart [4] objects, a home network make it possible to transport information between objects and a residential gateway to connect the smart home to the outside Internet world.”

“Smart homes are homes provided with some integrated technological system offering remote or [70] central controlled functionalities and services.”

“Smart Home Technology is a collective term for information and communication-technology in [71] homes, where the components are communicating through a local network.”

“The smart home adjusts its functions to the inhabitants’ needs according to the information it [72] collects from the inhabitants, the computational system, and the context.”

“A residence equipped with computing and information technology which anticipates and responds [73] to the needs of the occupants, working to promote their comfort, convenience, security and entertainment through the management of technology within the home and connections to the world beyond.”

“A dwelling incorporating a communications network that connects the key electrical appliances [74] and services, and allows them to be remotely controlled, monitored or accessed.”

“A smart home is a residence in which computing and information technology apply to expect and [75] respond to the occupants' needs and can be used to enhance the everyday life at home.”

“one which provides a productive and cost-effective environment through optimization of its four [76] basic elements including structures, systems, services and management and the interrelationships between them”

Definition Reference

“A smart home is a harmonious home, a conglomeration of devices and capabilities working [77] according to the Zen of Home Networking.”

“smart home technology is the integration of services and technologies, applied to homes, flats, [63] apartments, houses and small buildings with the purpose of automating them and obtaining and increase safety & security, comfort, communication, and technical management”

Figure 4-1. Simplified Architecture of a Smart Home

4.3 The Very First Projects and Examples

Most of the early examples of smart houses were constructed as a research project or a showcase house mainly focused on simple activities such as switching on/off the lights or control of other appliances since 1933. “Wired” homes built around 1960s can be considered as the origin of smart homes; however, the term “smart houses” was used by National Association of Home Builders

(NAHB) around 1984 [73]. Miller (2015) talks about houses featured in 1933 Chicago World’s

Fair and 1962 Seattle World’s Fair that can also be considered as the initial prototypes of smart homes. The latter was equipped with pools with automatic vacuum system and a computer that could program for reminding meals plan, and calling the library and the grocery store [65].

According to Miller (2015), a Scottish company started a home automation project in 1975 called

X10, which used existing electrical wiring to send/receive signals to/from lights and appliances

[65]. According to Austin (1985), “The Smart House concept scraps the current system of separate wiring for power and communications in a home and utilizes an integrated wiring and central control system that, combined with updated appliances and electronic devices, can virtually think for itself.” Austin explains the automatic lights as one of the features of the house that can be lighted as you walk in the house [78]. In similar articles, Kerch (1985, 1986) talks about this house that was under development in those years by NAHB Research Foundation at Rockville, MD [79,

80]. Youngblood (1986) predicted that smart houses will become more common in 1990s and that such homes will be convenient for people with disabilities and elderly [81]. Allison (1992) talks about a showcase smart house located at Fieldcrest Farms in northwest suburban Algonquin that was built by Burnside Construction Co. and was sponsored by the housing foundation and

Northern Illinois Gas Co. [82]. The design of smart homes for people with disabilities and elderly has been considered as an important goal since early developments in this area. The features provided in these initial examples of smart homes are mainly focused on simple tasks such as light control and the communication systems were utilizing the wirings within the house.

It can be observed that most of the initial examples of smart homes are mainly dedicated to home automation and target different areas such as residents’ well-being, health, and entertainment. This brings up another major aspect in smart homes, which is the various services they can provide for users. That aspect is discussed in the next section.

Table 4-2. Major Categories of Smart Home Services

Reference Services

[83] Energy saving Convenience Security - - -

[65] Efficiency Convenience Security - - -

Energy

consumption Lifestyle [84] Safety - - - and support

management

[66] Healthcare Comfort Security - - -

[4] Energy Saving Comfort Security - - -

Home Domestic Information and [74] Environmental Health Security entertainment appliances Communication

Technical Safety & [63] Comfort Communication - - Management Security

Audiovisual Security and Energy Comfort Communication - [85] entertainme Safety Management Control Services nt

4.4 Smart Homes Services

Different goals or services with the target of facilitating the residents’ daily lives can be considered

in smart home design process. Various researchers have categorized the area of focus and intended

features differently. Based on the experience of different smart home projects, researchers have

categorized the intended services of smart homes into categories such as energy saving, comfort,

and security [83, 4]. There are also other categories such as (a) comfort, healthcare, and security

[66], (b) convenience, security, and efficiency [65], and (c) lifestyle support, safety, and energy

consumption and management [86]. There are also some subcategories suggested that include

69 communications, entertainment, assisted living, e-health, convenience and comfort, security, and energy efficiency. Figure 4-2 shows an example of how these areas cover the smart home services

[84]. According to King (2003), Housing Learning & Improvement Network published a document on the definition of smart homes that categorizes the services in six main areas including environmental, security, home entertainment, domestic appliances, information and communication, and health [74]. However, there are other studies that identify smart homes by highlighting other aspects such as automation, multi-functionality, adaptability, interactivity, and efficiency [87]. Labonnote and Hoyland (2015) considered six different purposes for smart homes including health, safety, security, wellness, entertainment, and energy management. Their study is focused more on health services, while energy management and entertainment are the least studied topics in smart homes. Figure 4-3 shows the distribution of total number of research efforts in different areas since 2003 [88]. There are many examples of smart homes oriented around health improvement and home control applications for elderly and people with disabilities [69, 89, 90].

For example, Bonino et al. (2011) studied a control system for a smart home that is controlled by eye motions coded in C#, which is called DOGeye. Control of lights, alarms, and temperature is among the functions of this system [90]. Table 4-2 summarizes some of these intended services or features that smart homes can be equipped with. Although there are various categories in smart homes intended services, it can be noted that they are mostly focused on efficiency, convenience, and security. Energy saving and efficiency is one of the major features of smart homes that is of interest for discussion in this paper. Other researchers have categorized the functionalities of smart homes in a different way; for example, Bejarano et al. (2016) identified different systems such as speech command, visual tracking and activity identification, context aware, environmental

70 monitoring, home monitoring, user monitoring, home control, user authentication, and scheduling

[91], which can be considered as subcategories for services presented in Table 4-2.

Figure 4-2. Intended Services in Smart Homes. Adapted from [84]

Figure 4-3. Number of Papers Focsed on Different Areas in Smart Homes Services Since 2004. Adapted (with

curves linearized) from [88]

4.5 Smart Homes Components

Design of smart homes relies on having different components to accomplish the intended services.

It could be a network of sensors to obtain environmental data, a communication system to receive and send data from appliances to the processing unit, and a control unit that performs the required actions based on the acquired data. Dingli and Seychell (2015) categorize the main technologies in smart homes into two major groups including the networks and devices [83]. The available technologies to be used in a network can be Bluetooth, Firewire, USB, X10, BACnet, EHS,

ZigBee, Ethernet, and Wifi, but of course, wireless systems are more popular and in higher demand. According to Anju et al. (2016), the communication systems in home automation can be categorized into different systems such as global system for mobile communication (GSM),

Bluetooth based, phone based, ZigBee based, wireless based, and Internet or “internet of things”

(IoT) [92]. Furthermore, the analysis of different wireless protocols for Home Area Network

(HAN) in smart homes is studied by researchers such as Mendes et al. (2015) [93]. The devices used in smart home systems can be also categorized into three major groups including the sensors, actuators, and controllers. Sensors are the most basic components in this system that retrieve data from the environment and send to an actuator, which can perform an action and modify the environment such as a light switch. Controllers are relatively the most complicated part of the system, which can be a hardware or software component and can perform a process on data and decide on actions such as adjustment of thermostat in HVAC system, control of lights, audio visual, gardening, and security that need to be taken by the home automation system [83]. Toschi et al. summarized the technologies in Home Area Networks (HAM) and communication systems that can contribute to smart homes under three groups including the standards (e.g. Universal Plug

& Play (UPnP), Digital Living Network Alliance (DLNA), Konnex (KNX), LonWorks, ZigBee,

72 and X-10), architectures (e.g. DomoNet, and Jini), and initiatives (e.g. Project Hydra, Amigo, and

Open Building Information Exchange group) and studied the challenges and future directions.

None of the technologies and standards presented in the paper is yet adopted as the major standard for home automation and smart homes, unifying this industry by adopting a single standard can guide the market trend [94]. However, there are protocols for the components of smart homes as opposed to the whole system such as the standard for wired or wireless lighting control systems that unifies the communication protocols [95].

Madakam and Ramaswamy define five main components for smart homes including actuators, controllers, central unit, networks, and interface. They define the control unit as a preprogrammed component that can make a decision for the system based on the inputs and a central unit that can render possible programs in the system [96]. Similar smart home elements are defined in other studies including network, sensors and actuators, smart devices, and interfaces & controllers [97] or sensors, controllers, actuators, buses, interfaces, and networks [65]. It should be noted that when it comes to heterogeneous systems with multiple components such as home entertainment, surveillance and access control, energy management, home automation, and assistive computing and healthcare that need to be connected together through a system, it is not easy to make them interoperate as a single system because there might be different operating platforms for each component. Therefore, there are studies that suggest solutions for better interoperability such as using Simple Object Access Protocol (SOAP) that is lightweight protocol targeted for exchanging structured information in a distributed environment with various operating platforms [98]. Based on the literature review discussed by Bejarano et al. (2016), different components in smart homes are identified as remote access, home gateway, home access, desktop and mobile devices, user interface device, system display and main controller, device manager, and sub-networks and

73 connected devices that are integrated in a design based on the user’s need [91]. Therefore, depending on the intended services and complexity of the design, smart homes may need to have different components. For example, the required services for a smart home can be limited to simple moisture sensors that go off when there is a water leakage or it can be as complicated as a network of sensors or metering devices that sends data through a communication system to the preprogrammed central processing unit, which sends commands to the control unit to dim the lights or adjust the thermostat.

4.6 Buildings, Projects, and Labs Focused on Smart Homes

In order to study the performance of different smart home systems, it is important to be able to install and use the system in an environment that is close to the real-life conditions, especially when the intention is to study human interactions with these systems. Indeed, there are projects, buildings, and labs in different countries and research institutes that are constructed in order to study these systems. The literature suggests there is an abundance of academic projects or smart houses to investigate the feasibility of application of different technologies in smart homes, including the following references: Ricquebourg et al. (2006), Harper (2003), Johanson et al.

(2004), Mozer (2015), Laberg (2004), Berlo (1999), and Madakam (2014). Research programs such as Aware Home Research Initiative (AHRI) developed at Georgia Institution of Technology

[99] are focused on health and well-being, digital media and entertainment, and sustainability. This program consists of four different lab areas, one being the Aware Home, which is a 3-story, 468 square meter (5040 square foot) building with two identical floors. The technologies incorporated in this “smart” building are based on Human Computer Interaction (HCI) and are feasible to be applied in real homes [99].

A” living laboratory residential home” called the PlaceLab is developed by the MIT University that is part of the House_n research project to investigate application of new technologies to meet people’s needs in intelligent homes. Using sensors installed in the house, researchers would be able to monitor the location and activities of residents. Sensors can sense the temperature, humidity, audio, and CO and CO2 levels. There are also other radio frequency devices to help locating people, video capturing devices, wearable biometric, and motion sensors. Controlling the environmental qualities such as air and temperature is also another feature of this house [100].

“Intelligent Room” is another project developed at the MIT University to study human-computer interaction (HCI) by equipping the house with computer vision, speech, and gesture recognition to follow the residents’ behavior [101]. Another example of a smart building that is focused on healthcare of residents is a project in which smart home technology was installed in residential flats in TΦnsberg, Norway in 1995 as part of a bigger project called BESTA-project that started in 1994. These smart homes were designed to increase both security and convenience of residents with dementia, more details available in a documentary called “Smarthus” [102]. The house was equipped with fire alarms connected to pagers and exit doors, automatic lighting at night, magnet sensors at doors that send a message to staff, automatic turn off system for cooker, and fall detection system [63, 71].

There are also smart homes more specifically intended for senior living. Casattenta project is another example of Aware Homes built in Italy and designed for improving the life quality of elderly people. ZigBee wireless technology is used for communication system within the house and acquires data from both fixed and wearable sensors. The collected data such as light, temperature, humidity, and human presence are used only for monitoring purposes [103].

Improving seniors’ life quality is also studied in other research projects such as Smart Home in a

Box (SHiB) project conducted by Washington State University that is a ready-to-install kit of required tools in a smart home (Hu et al., 2016). A prototype of a robot-assisted house is built and studied in research building of Korea Institute of Industrial Technology (KITECH) to evaluate the feasibility of such technologies. Smart objects with radio frequency identification (RFID), a server, and service robots are the main components of this smart home being connected through ZigBee communication technology, and the functionality of house is evaluated by defining three main scenarios including cleaning up, executing errands, and security service. For example, robots are supposed to clean up the space when the residents are out for work and the smart objects such smart table, shelf, and camera will facilitate these tasks. Smart objects are equipped with RFID modules, which can help, for example, the smart table to recognize the objects on top of it [104].

There are also smart homes built for research purposes that are focused on decreasing and optimizing the energy consumption within a house such as Adaptive House [105] and MavHome

[106] that are explained in more details in the section on “Systems with advanced data processing capabilities”. There are also examples of application of monitoring systems in college environments such as building dashboard program in Oberlin College that encourages reduction of water and electricity usage across campus; however, technical data about this program is not readily available in open literature. This program is limited to monitoring system and providing statistical data and does not include any controlling or high-level data processing unit. Another example of energy consumption monitoring in a live-in research lab is the 557.5 square meter

(6000 square foot) smart dorm developed at Duke University called Home Depot Smart Home.

This building along with two other student residences were used for a research study about energy consumption behavior of students by collecting related data. The monitoring system is called eGauge, and the data were monitored by a web-based platform and stored in a computer system

[107]. This research that is also limited to energy consumption monitoring does not include any control or high-level data processing unit and is aimed for raising student awareness. Raising awareness about energy consumptions and environmental consequences could have a huge impact on reducing CO2 emissions considering that 40% of the total CO2 emissions in U.S. is due to the electricity generation [108]. The effectiveness of raising awareness of the residents is also studied at the University of California, Los Angeles in a large-scale experiment involving 120 apartments in Los Angeles [108]. The apartments are equipped with outlet panels that use different outlets for appliances, heating and cooling systems, and lightings. Therefore, by using energy metering systems, the real-time energy consumption of the apartments in appliances level can be monitored.

They also used such data to inform the residents about their energy consumption and CO2 emissions, including air quality impacts. The effect of raising awareness was noticed to be a 6% reduction in overall energy consumption [108].

Communication systems and remote access to control units in a smart home is important enough that mobile network companies are also interested in performing research in this area in order to expand the area of their services’ application. The “Orange at Home” project is one of the projects focused on smart homes that was executed in 2001 by a UK mobile network company called

Orange. The outcomes of that research remain confidential except for a portion of the work that is part of a book published by Springer-Verlag UK in 2003 [73, 109]. The system used in this case study is a server that can control almost all the functions of a house such as lighting, heating, security, audio-visual systems, curtains, bath, appliances, and health-monitoring systems that can be controlled by different means such as wireless systems or SMS. Harper (2003) believes that a lesson that was learned from this project is that “technologies that succeed in work environments sometimes fail in home settings“. Another issue identified was deciding who should control the

77 systems. Some of the technologies cannot be used in residential setting; for example, some of the functions that might be common in office buildings, such as wall panel units to control the lighting, might be too sophisticated for users in residential houses. Different services were available in this house, including wall-mounted control panels, Compaq TP/IP devices, mobile phones, entertainment media, kitchen equipment, baby monitoring, computer networks, security systems, health monitoring services, and internet shopping [73]. Similar European projects are dedicated to study the influence of users’ activities on the performance of sustainable Product and Service

Systems (PSS) by using living labs such as SusLabNWE projects that is a joint project between four European countries and studies on smart homes (e.g., application of sensor networks) are conducted in real living environments [110, 111]

Another example of a project that provides the opportunity to work in a real building, is GridSTAR

Center (Grid Smart Training Application & Resource Center) that is an education and research center funded by Department of Energy (DOE) and constructed by the Pennsylvania State

University. It is located at The Navy Yard in Philadelphia and designed for the plug-and-play testing of smart grid components and system configurations. Most of the facilities in this center are currently devoted to educational purposes. Some of the functionalities defined for this project are related to improvement in energy supply and energy use. The components within the building related to energy smart homes are smart meters, power quality meters, and remote displays to monitor conditions and archive data [112]. The Pennsylvania State University has also built a solar powered house called MorningStar that was used for participation in Solar Decathlon Competition in 2007 and is currently located on the main campus [113, 114]. Although the main purpose of this project was to show potential uses of solar energy in homes, installation of temperature and humidity sensors enables the researchers to perform studies related to smart homes as well.

A testbed for grid technologies developed in Wunsiedel, Germany is another example that is conducted by University of Bayreuth as a research project. A cellular approach is taken that considers computational capabilities for each component in energy generation and consumption unit in a microgrid system. As the long-term goal, the microgrid would be able to control the supply side based on the demand and it will not be limited to shaping the demand load profile. The

Supervisory Control and Data Acquisition (SCADA) system operates as an ICT (Information and

Communications Technology) that collects the power need information and provides forecast services and diagnostic functionality [115].

Some manufacturers have also responded to the need for smart homes and have started developing some products to be used in smart homes; examples include Philips Electronics Company (uWand) and Intel Smart Homes with Intel Internet of Things, devoting some efforts to produce smart homes technologies.

The summary of the smart homes constructed for research purposes or labs and projects that have studied the feasibility and performance of systems in smart homes are presented in Table 4-3, some of which have been reviewed by [116], including Internet Home, EasyLiving, Home of the Near

Future, Smart Homes in Portsmouth, CUSTODIAN, Smart Home, LIVEFutura, Sussex, Intel

Architecture Labs, and X10. It can be noticed that these projects are focused on health, entertainment, and energy efficiency of the buildings and residents. Different types of sensors or metering tools are installed in these houses and the obtained data are either used for monitoring or controlling purposes such as detecting the opening or closing state of doors or windows, energy- related data, human computer interaction, voice control systems, and interoperability of smart homes and smart grids. There are always challenges and issues concerning application of these

79 systems in smart homes with regards to communication systems, processing tools, human interaction, and security, which is discussed in more details in next section.

Table 4-3. Examples of Projects, Houses, and Living Labs Built or Equipped for Research Purposes Related to

Smart Homes

Title Executer/Location Description

Control systems related to turning heat on/off, providing hot water

based on statistical patterns of water usage, and controlling lighting

systems work based on neural network and resident’s behavior Adaptive University of Colorado prediction techniques. Sensors monitor temperature, light, sound, House motions, outdoor environmental conditions, boiler temperature, and

water usage in order to control the lighting, ventilation, and air and

water heating [101, 105].

This NREL project is intended to reduce energy consumption of

homeowners and improve their life quality. The automated home energy

management (AHEM) system suggested for use in this lab aims to

improve the whole-house energy management based on the data AHEM Lab NREL acquired from residents, sensors, and other measuring tools. The

appliances in this house are demand response enabled and other

components are also capable of communicating with the utility

company [117].

Aware Home

Research Aware Home is a 3-story, 5040 ft2 facility designed to facilitate research Georgia Tech Initiative similar to a house environment [99, 118].

(AHRI)

Title Executer/Location Description

Real-time feedback of water and electricity usage. Comparison of

Building different campus usage rates motivates students to improve their Oberlin College Dashboard consumption habits.

[119]

This project is dedicated to investigating the effects of sensor The Interactive ComHOME technologies and voice control systems in home activities such as Institute, Sweden communication, distant work, and social activities [120].

The project is focused on improving the life quality by providing an Funded by European easy access to technology and services for people with disabilities and Custodian Commission's the elderly through the use of information, assistive, and Telematics communication technologies [121].

Casattenta project is an aware home built in Italy to study the smart

home technologies that can contribute to the life quality of the elderly. Casattenta - Data such as temperature, light, humidity and human presence is project obtained and monitored through ZigBee communication technology

[103].

This project that is mainly focused on the entertainment aspect of smart

homes involves four major components including middleware,

geometric world modelling, perception, and service abstraction. The

EasyLiving Microsoft Research sensing component recognizes human activities, while the sensor

network of the system tracks activities using different sensing

equipment and the service abstraction controls media units such as TV

or DVD player [122].

Located at The Navy Yard in Philadelphia, the GridSTAR home is Pennsylvania State GridSTAR designed for plug-and-play testing of smart grid components and system University configurations. The energy smart home components within the building

Title Executer/Location Description

are smart meters, power quality meters, and remote displays to monitor

conditions and archive data [112].

Energy monitoring and feedback project in Duke student residences

uses energy gauges to measure total energy consumption. Due to high Home Depot Duke University costs of monitoring and feedback systems in large dorms, this project Smart Home served as a case study for using similar technologies in larger residences

[107].

House_n is a multi-disciplinary project that started in the Department of

Architecture at MIT. Homes are built incorporating many sensing

House_n MIT University components in order to enable people to “easily control their

environment, save resources, remain mentally and physically active,

and stay healthy” [123, 124, 125].

This house is instrumented to study the adaptive decision making École des Mines de method of Markov Decision Process (MDP) based on data obtained Living Lab Douai from sensors and transmitted to the central processing unit through

wireless connection in a ZigBee network [126].

To experiment Human Computer Interaction (HCI), this Intelligent

MIT Room is equipped with computer vision, speech and gesture recognition

Intelligent MIT University systems that connect it to what the residents are doing and saying to

Room allow computers to participate in human-level activities and to allow

people to interact with computational processes [101].

This home monitors and records the residents’ behavior and also

controls electrical devices such as automated mini-blinds based on such University of Texas MavHome behavior and predictions techniques (Cook et al. 2003). Sequence Arlington mining algorithm and hierarchical Markov model is used to predict and

learn the residents’ actions [101, 127].

Title Executer/Location Description

Medical

Automation This building consists of in-home, health-status monitoring system that

Research University of Virginia works using gait monitoring technologies to assess gait characteristics

Center without the need to wear any devices [128].

Although it was built for Solar Decathlon competition, the MorningStar

Pennsylvania State house is also used for research on energy efficiency, renewable energy, MorningStar University and smart grid systems by equipping the house with temperature and

humidity sensor [114, 113].

Different services are available in this Orange house including wall-

mounted control panels, Compaq TP/IP devices, mobile phones,

Orange at entertainment media, kitchen equipment, baby monitoring, computer Orange UK Home networks, security systems, health monitoring services, and internet

shopping [73, 109].

PREDIS smart building is a platform installed in G2elab laboratory

dedicated to multi-sensor monitoring, user activities and their energy

impact analysis, multi-physical modelling, handling measurement and

sensitivity analysis, and optimal control strategies development. It has Grenoble Electrical been renovated by adding 14 cm cellulose wadding insulation to the PREDIS Engineering walls and improving HVAC dual flow so, the building’s maximal Laboratory, France energy consumption is about 50kWh/year.m². There is a computer room

for training courses of Grenoble University and equipped with laptops

connected to the electrical grid and photovoltaic generators and also a

researcher office with measurement tools and control systems [129].

Title Executer/Location Description

Washington State SHiB system provides a self-installed kit for smart homes for senior SHiB University population [130].

This smart home project is focused on building energy efficiency and

also users’ ease of access and control of HVAC or other building

components. It includes multiple products such as door and window Smart Home Siemens contact sensors to detect opening and closing of fenestrations, different

actuators to control the mechanical systems within the house, and lights

and blinds control systems [131]

This system is mainly dedicated to studying the interaction design of

Stanford shared computer-augmented spaces, flexible infrastructure for

Interactive Stanford University integration, and empirical studies of collaborative work. The

Worspaces technologies studied in this research are based on iROS (Interactive

Room Operating System) developed by the research team [132].

SusLabNWE A living lab project started in Europe by 12 partners from four

(Sustainable European countries including United Kingdom, Germany, Netherlands,

Lab North European Project and Sweden. Multiple projects conducted so far in these countries to

Western assess the sustainable Products and Service Systems (PSS) and observe

Europe) the effects of user’s behavior on their performance [110].

Vallgossen The two projects include buildings equipped with different systems that JM and Skanska and can be monitored and controlled by computer including the room construction Ringblomman temperature, gas or electric stove shut off system, ventilation system, companies buildings and the security system [133].

- A field test laboratory in Wunsiedel, Germany provides a testbed to

other utilities and grid technologies developers for assessment of University of Bayreuth different smart grid systems and smart homes. The Supervisory Control

and Data Acquisition (SCADA) system collects data as an Information

Title Executer/Location Description

and Communication Technologies (ICT) service provider and in return

forecast services and provide diagnostic functionality. The whole

system takes a cellular approach in which all the components have

computational capabilities [115].

120 Apartments were equipped with detailed outlet electricity

University of consumption level in appliances, and the residents were informed about

- California, Los their energy consumption in different ways, which helped

Angeles understanding the effectiveness of raising awareness in this project

[108].

This smart home is designed to evaluate the feasibility of using service

Korea Institute of robot, which rely on RFID to identify objects. Different service - Industrial Technology scenarios are studied in order to assess the functionality of service

robots in smart homes [104].

This project involves understanding the effect of using thermal Merthyr Tydfil (South - feedback devices and a smart heating control system called Wattbox in Wales) seven different households [134].

- This research involved a residential home set up for 5 people with

dementia in Norway in 1995. The house was equipped with fire alarms

TΦnsberg (Norway) connected to pagers and exit doors, automatic lighting at night, magnet

sensors at doors that send a message to staff, automatic turn off system

for cooker, and fall detection system [71, 63].

4.7 Challenges in Smart Home

Components of smart homes and the connections among them could deal with different kind of issues and challenges. Areas such as data acquisition, communication, algorithms used in data processing unit, control unit, and cost can all be susceptible to some challenges in different stages of design and implementation of smart homes systems. For example, the communication systems may experience inconsistencies between different communication protocols in different devices or the security issues in transferring data. Samuel (2016) reviewed some of the challenges in communication systems in smart homes and also the most common wireless standards [135].

Jacobsson et al. (2016) studied 32 different projects on smart home automation systems and classified nine of them as systems with low risk and four with high risks and the rest are considered as systems with moderate risks. Most of the risks are related to either human factors or software components. The risks considered in smart home automation systems considered in this study include hardware theft, data manipulation, damaging devices and servers, inadequate access control configuration in the gateway, and inadequate authentication and confidentiality settings

[136]. There are also other challenges identified in literature including poor manageability (e.g., difficulties in using systems for users), government regulations, and cost. The average cost of components in a smart home is about $25,000, aside from installation and support expenses [137].

Examples of other issues associated with components used in smart homes include reliability, interoperability, manageability in developing technologies, security and privacy concerns in designing technologies [86, 138], and application of multi-attribute trust metric (MATM) [139].

Fit to current life style, administration barriers such as required knowledge for operation, interoperability barrier such as communicating with other appliances, reliability, security, trust barriers of consumers on utility companies, and cost barriers are among additional challenges

86 identified by some researchers [84]. Rathnayak et al. (2012) discussed the challenges that come with energy resource management (ERM) systems and identified barriers such as lack of intelligence to handle uncertainty, passive decisions in dynamic environment, fixed priorities for all users, not considering the optimum point between user’s comfort and energy consumption, and decision making based on single input while there are multiple factors [140]. Saturation of the communication channels is another issue that is identified by Stojkoska & Trivodaliev (2017) among other challenges such as adaptability of new and older types of networking protocols, interoperability of devices, and security and privacy. Three methods of data compression, data prediction, and in-network processing are suggested as solutions in order to avoid unnecessary data transmission through the communication network and overloading the communication channels [141]. Besides the technological challenges, there are also concerns among users with regard to using smart homes. Top three users’ barriers identified by Bulut (2016) are the high investment cost, low energy prices, and regulatory framework [142].

The identified challenges do not remove or reduce the feasibility of application of smart homes systems. It can also be observed that there are solutions for these challenges and issues, which only need to be considered in order to improve the performance and also make the users more enthusiastic about using these systems.

4.8 Design Different Types of Energy Smart Homes

One of the important intended services of smart homes identified in multiple research studies is energy efficiency. Besides decreasing the energy consumption, it can also decrease the energy- related costs by integrating other concepts such as Demand Response (DR), which basically shifts the energy demand to the times that energy tariffs are lower. The control systems in energy smart

87 homes can be as simple as switching on/off the lights remotely, for example, and application of sensors and actuators can help the residents to make sure energy is used in the most efficient way

[143, 83]. Aman et al. (2013) define eight different abilities for energy smart homes including monitoring, disaggregation, availability and accessibility, information integration, affordability, control, cyber-security and privacy, and intelligence and analytics. However, the currently available systems in market or those proposed in the literature do not necessarily include all of these features [144]. Wilson et al. (2014) recognize three major categories in smart homes including functional, instrumental, and socio-technical. The instrumental functionality of smart home is devoted to managing and reducing the energy consumption and carbon production, which is basically the definition of energy smart homes. They also discussed the challenges and useful approaches in each category [138]. Smart homes could be also categorized under other categories based on their service, technology, finance, and organization and the statistical study of over 154 papers conducted by Solaimani et al. (2016) shows that most of the literature published between

2003 and 2013 are focused on technology area rather than three other categories including service, organization, and finance [145]. Rathnayak et al. (2012) identified two major areas in literature that are focused on energy resource management (ERM) systems including decision making and resident behavior modelling area and introduced four major sections for ERM frameworks including device and interconnection, information layer, decision layer, and interface [140].

Lobaccaro et al. (2016) reviewed different smart home technologies intended to help users to communicate with appliances and enable them to monitor or remotely control systems under the four categories of integrated wireless technologies (IWT), home energy management systems

(HEMS), smart home micro-computers (SHMC), and home automation (HA) technologies. Some of these technologies and products are related to energy smart homes [146]. They studied the

88 available products, which are mostly limited to energy monitoring systems. There are not many research papers available about these commercial products however, the systems that are based on a software or hardware either in processing or control unit of the smart home that is developed in a research study with available documents and their features are also reviewed in our paper.

Zipperer et al. (2013) reviewed different aspects of electric energy management in smart homes and categorized the loads involved in energy smart homes into three main groups including thermal loads such as cooling and heating loads, electric vehicles that are expected to contribute to 3% of the U.S. light duty vehicles by 2025 [147], and smart appliances that can be controlled by intelligent controllers. They discuss two major thoughts on control systems: a) the whole system should be fully automated, and b) for well-informed users, control strategies should be decided based on the outputs of energy smart systems. The major issues in control systems are also categorized into five major groups including machine learning, rule-based, multi-agent, and decision making systems. Another category in energy smart home in their study is utility-side and costumer-side technologies in energy smart homes. The smart meters and demand response (DR) concept contribute to the utility-side. The latter contributes by shifting the energy consumption of the house from peak-load time to reduce the cost. Commercial web portals that provide detailed information of energy usage are categorized under customer-side technologies [148]. Home energy management systems need different tools in order to achieve the main goal, which is reducing or optimizing the energy consumption. Vega et al. (2014) referred to these tools as models and reviewed multiple studies from 1970 to 2014 and showed that among all the literature reviewed in their study, topics such as hardware components, management strategies, software techniques, communication protocols, and relationship with users have been the focus of most discussions, respectively, with the highest to lowest frequency [149].

Building on the literature reviewed, in this paper, the systems and products identified in the literature are categorized based on the control capability and also the intelligence and analytics ability, which is the level of process that is performed on data obtained from the house environment through sensors or any other measuring tools. Accordingly, the following categories are suggested:

Energy monitoring systems: Data such as total energy consumption of a house or comparison of the total energy consumption with other households do not need any advanced analysis on the retrieved data from sensors; such situations are typically limited to monitoring systems as opposed to systems with controlling capabilities. In this paper then, any research or project reviewed that involves only monitoring capabilities falls under this category.

Systems with control capabilities: Similar components in the first category, but the components also have a controlling or execution unit. It means that data are not retrieved only for monitoring purposes, and some decisions will be made based on these data for controlling units. Studies that have any type of controlling systems including active or passive controls are categorized under this group.

Systems with advanced data processing capabilities: Energy smart homes capable of performing more advanced and high-level analysis on the retrieved data fall under this category, for example systems with energy analysis tools or systems with central processing unit that can perform optimization process to optimize the energy consumption based on the residents’ comfort level and energy cost. This category is more focused on the data manipulation rather than the decisions that are made based on these data or the control units.

There are also other categories concerning energy management devices that could be used in energy smart homes. Abubakar et al. (2017) categorized all the systems under energy monitoring

90 devices within six smaller groups including measuring devices, optimization tools, communication devices, recognition devices, control devices, and display devices. They also identified two different approaches in energy monitoring including Intrusive Load Monitoring (ILM) (e.g., smart plugs and smart appliances) and Non-Intrusive Load Monitoring (NILM) [150].

There are also other concepts and systems that are somehow related to the smart home such as demand response (DR), demand side management (DSM), distributed energy resources (DER), microgrids, and smart grids. Energy smart homes can somehow contribute to any of these concepts and that’s why many of the research studies on energy smart homes are focused on these topics.

Table 4-4 provides basic definition and explanation of each concept. It should be noted that some energy smart home systems that integrate these concepts do not necessarily contribute to reduction of energy consumption, while concepts such as DR are more geared toward minimizing energy- related costs as relates to both generation and consumption.

One of the features of an energy smart home could be the learning or adaptability capabilities. For example, energy smart homes can simply control thermostat by learning when and how the temperature is usually adjusted during the day by residents and then it can be controlled automatically based on the data acquired during learning process [65], which leads to what can be called an adaptive smart home. These energy smart homes are designed to include a decision making capability within the data processing unit based on new every day data. The general architecture of such system is presented by Karami et al. (2016) where a case study project is explained that uses a java-based software as a part of the central processing unit and also adopts

Markov Decision Process (MDP) for adaptive decision making [126]. Zhou et al. (2016) discuss different aspects of home energy management systems (HEMS) and define five different functionalities for smart HEMS including monitoring, logging, control, alarm, and management.

They also define four different components for HEMS including communication and networking system, smart meters, smart HEMS center, and home appliances. Smart HEMS center is similar to the central processing unit of the system that enables the system to collect data for monitoring and to provide a user friendly interface. It can also manage the control unit within a house and perform analysis based on the acquired data [151]. The contribution of Smart HEMS is not limited to energy management within a house; it can also be beneficial when smart grids are part of the system. Two- way communication between the energy provider and consumers is necessary in order to optimize the energy consumption. As mentioned earlier, Demand Response (DR) is one of the energy scheduling strategies that can be optimized by using different methods within smart HEMS such as neural network, fuzzy logic, and genetic algorithms. Zhoue et al. (2016) also introduced two different approaches in DR including price-based and incentive-based DR. The first approach is more about reducing the energy cost, while the latter is more about change in users’ energy consumption behavior [151]. Different methods can be used in order to use mathematical models to predict the load or supply side of the DR such as multi-agent system (MAS) to reduce the peak demand. Although it does not lead to decrease in energy consumption, it can lower the energy related costs. Wang & Paranjape (2016) report about 24% decrease in peak load after applying an energy management system model in a MAS [152]. The reported cost and peak load reduction varies in literature from 4 to 80% and 7 to 90%, respectively. Most of the devices that benefit from

DR system in residential buildings include the HVAC, water heating, and refrigeration system with different objectives mostly dedicated to minimizing the cost, consumption, and improving well-being [153]. Beaudin & Zareipour (2015) also summarize the methods such as artificial neural network, linear regression, and auto-regressive forecasting that are used in DR systems to predict the electricity use and demand in advance [153].

All of these tools, systems, and concepts can be used in energy smart homes and to provide the intended services and based on the complexity and the components in the energy smart home they can be categorized under three major groups of energy monitoring smart home, systems with control capabilities, and systems with advanced data processing capabilities.

Table 4-4. Definition of other concepts related to energy smart homes

Title Definition and Explanation

Demand Response (DR) Demand Response is a tariff or program established to motivate changes in

electric consumption by end-use customers in response to changes in the

price of electricity over time [154].

Smart Grid A transmission grid that uses sensing, monitoring, and power engineering

technologies to provide advantages such as reducing energy loss, providing

capability of connecting new (renewable) energy resources, and enabling

users to sell energy back to the grid [155].

Microgrid An electricity subsystem detached from national grids such that unlike the

conventional centralized systems, power generation is distributed with

medium- or low-voltage networks. Besides distributed power generation or

energy resources, there are components such as local storage where

consumption can also be distributed [156, 130], with microgrid

coordinating operation of all components [157]. The advantages are

increased use of local renewable energy, enhanced use of heat generated by

combined heat and power (CHP) generator, stabled electricity supply,

reduced transmission loss, and decreased carbon dioxide emission from

centralized fossil fuel power plants [130].

Distributed Energy Resources Generates power through different technologies that can work separately

(DER) from the main grid such as internal combustion (IC) engines, gas turbines,

microturbines, photovoltaic, fuel cells and wind-power [158].

Demand Side Management (DSM) “Demand-side management is the planning, implementation, and

monitoring of those utility activities designed to influence costumer use of

electricity in ways that will produce desired changes in the utility’s load

shape, i.e., changes in the time pattern and magnitude of a utility’s load”

[159].

4.8.1 Energy Monitoring Systems

Smart homes can be equipped with smart meters that can monitor and keep the record of energy consumption in different time intervals or even real-time. They can also transmit data to users including the residents of the house or energy distributor companies. Use of smart meters can also be beneficial when it comes to smart grids and energy generation inside the house [160, 97]. Smart meters are becoming highly recognized for their role in reducing energy consumption as evidenced by a national rollout plan in England, Wales, and Scotland to have all homes equipped with smart meters [134].

For example, energy use visualizers or in-home displays (IHD) show the total energy consumption, and according to Stinson et al. (2015), compared to homes without IHD, gas and electricity usage has been reduced by 20% and 7%, receptively [5]. It shows the household more information about their energy consumption and can help them reduce it.

The data shown in the IHD are obtained by performing a low-level data analysis that basically uses illustrative tools to show the acquired data from smart meter with no further analysis required. It is also observed that the households that get normalized data about their energy consumption tend to use less energy compared with houses that are provided with absolute energy consumption [1]. Zheng et al. (2013) divided smart meters into two main groups including radio frequency (RF) and power line carrier (PLC). PLC basically uses utility power lines to transmit data, while radio frequency is a wireless method. Both methods have advantages and disadvantages such as lower cost of the PLC; however, this imposes longer data transmission time [160].

The displays that show energy consumption of households are referred to by different names such as In-Home Display (IHD), Energy Consumption Indicators (ECI), Smart

Energy Monitor (SEM), or Home Energy Displays (HED). Wood and Newborough (2003) investigated the effectiveness of providing basic energy consumption data related to cooking to 44 households by using ECI, and it was noticed that it affects the behavior of people, which led to 10% to 20% reduction in energy consumption. The ECI used in this study is a display indicating the current, this week, last week, today, and yesterday energy consumption. To calculate the energy, it was assumed the mean voltage is equal to 240 V although it is not constant and varies depending on the location and time [161]. Abrahamse et al. (2005) conducted a thorough research on the effect of providing energy consumption data to users to encourage reduction, where the related literature can be categorized as intervention studies focused on energy consumption. Both short-term and long-term effects were studied where the results indicated up to about 20% reduction in either electricity or gas use [162]. The effect of presenting energy information to users is a topic studied by other researchers too [163, 164], some noting that at a certain point the users understand the limit of energy saving potential of their house, beyond which point it might not be as effective as before. Van Dam et al. (2010) also studied the effectiveness of home energy management systems (HEMS), where the case study involved energy monitoring a home for 15 months in Netherlands. The monitoring system includes sensors, a sending unit, and a display that shows the real-time energy consumption. Results show that in short-terms, the energy consumption can decrease about 8%; however, it cannot be sustained for long- terms [165]. Among other researchers who investigated the effectiveness of using HEMS

96 is Ueno et al. (2006) who studied the effectiveness of an energy-consumption information system (ECOIS) on energy saving. The system includes a monitoring and distribution units.

The monitoring unit includes a load-survey meter (LSM) to measure the whole-house energy consumption and also an end-use meter for energy consumption of individual home appliances. The network control unit (NCU) sends data through telephone line to the central computer [166].

The concept of monitoring more detailed energy consumption data is also considered in other studies, including Jahn et al. (2010) who studied interconnecting electric devices in a house by using wireless power metering plugs to obtain the energy consumption of different appliances. Hydra (the name changed to LinkSmart after 2010) is a middleware framework that is used in this research as an intelligent communication tool to make the smart home environment by enabling the user to control any type of physical device regardless of its network technology such as Ethernet, Bluetooth, RF, ZigBee, and RFID.

The main application of such systems is to develop an energy consumption profile for each device, which allows the user to identify the devices that use high energy [13]. The wireless power metering plugs used in this research are called Plogg, which are designed to measure and monitor the electricity used, where the data can be collected in a gateway and monitored in a mobile interface, for example. The other feature of the system designed in this research is that users can use their mobile phones to view energy consumption of their appliances by using UbiLense that sends picture of the device to an image recognition server, which has the information related to the Plogg connected to that device and returns its energy consumption [13].

Energy monitoring systems might be limited to a single energy meter for the whole house, but there are methods such as nonintrusive load monitoring (NILM) that first appeared in

1980s to enable the user to have more detailed information on the energy consumption of appliances by determining the jumps in energy consumption and relating them to specific appliances with the same level of required power. This method replaces the detailed energy consumption of separate appliances, which is cost intensive. Marchiori et al. (2011) presented two different approaches in NILM named Heuristic and Bayesian approaches to solve a common issue that comes with NILM, that is, appliances with low power loads and multiple states that make them hard to detect and distinguish in NILM system. The first approach is based on probability estimates and depends on training data, while the second approach uses a Bayes classifier to calculate the highest probable state for each device based on measured total consumption and also detected changes in working state of the device (e.g., on or off) [167]. In order to monitor the energy consumption of appliances in a home, Kim et al. (2009) proposed an indirect and non-intrusive sensing tool called

ViridiScope to eliminate the need for meters connected between the appliance and AC plug. The non-intrusive methods, known as Non-Intrusive Appliance Load Monitoring

(NIALM), are studied in other research works as well [168]. ViridiScope works with measuring magnetic, acoustic, and light sensors that are installed close to the appliances.

This system is also equipped with a sensor calibration framework that works by formulating a model based regression problem that helps the sensors to auto-calibrate themselves [169]. This system also only enables the user to measure and monitor energy consumption.

LaMarche et al. (2011) reviewed available products that can be categorized under Home

Energy Management (HEM) systems including energy displays, control systems, or phone apps. Reviewed products are categorized under three major categories including control devices, user interfaces, and enabling technologies. The control systems considered in this study include both device-level and centralized systems, with the latter controlling multiple devices. Systems reviewed LaMarche et al. (2011) related to energy monitoring are categorized into two groups including the system that provides raw and processed data.

However, the processed data are limited to information such as appliances breakdown data and social comparison; therefore, it is still considered an energy monitoring systems [170].

The energy monitoring systems can be equipped with software systems that control collecting, storing, analyzing, and visualizing the data or inputs. Brewer and Johnson

(2010) developed an open-source software that can be used to illustrate the energy consumption of a user in a specific location that is equipped with measuring tools. The outputs are obtained by executing a low-level analysis, which means the data processing is limited to statistical analysis and developing charts and tables about energy consumptions.

Another advantage of using a monitoring system that collects data from different resources and is capable of storing and analyzing them is that it enables the energy provider companies to assess the level of energy production and also the carbon intensity of different plants in different areas that can be helpful in comparing the simulated current energy generation level in these plants [171].

Some of the data collection and analysis software that can be used for smart homes are only focused on heating systems such as what is studied by Vanus et al. (2016), where they try to connect building heating technology (BHT) with building automation technology

(BAT) through PI system software tool. The data such as temperature were obtained from

Moravian-Silesian Wood Cluster building that is built in the Faculty of Civil Engineering,

VSB-Technical University of Ostrava and the system enables the user to perform a real- time monitoring, while the PI system provides services such as visualizing, storing, and other statistical analysis [172].

4.8.2 Systems with Control Capabilities

Energy smart homes with control capabilities use the data acquired from environment to process and send proper commands to control unit, which is designed based on the intended service of the smart house such as controlling the lights, blinds, appliances, HVAC system.

However, these commands could be based on either simple or complex process in the processing unit and the systems reviewed in this chapter fall into the first category, which includes systems with simple processing capabilities.

Man Li et al. (2016) categorized different generations of smart homes technologies under three groups of Bluetooth and ZigBee enabled, smart homes with artificial intelligence, and smart homes adopting robot technologies. They also summarized the environmental, economical, and social contribution of different products for smart homes. The features of these products mostly fall under the smart home systems with control capabilities such as

100 automation of light control, auto delaying the washing and drying systems, and solar tracking systems [173].

Communication between the sensing and control unit plays an important role in energy smart home. Han & Lim (2010) studied application of two different communication standards including ZigBee and IEEE 802.15.4 that helps interoperability among different electrical equipment, meters, and smart energy enabling products. The system is capable of either turning on/off or dimming the lights and it can also use the same communication system to receive data from meters [6]. ZigBee is a wireless sensor network that is secure and more importantly it has low power consumption and fast reaction, and that is why it is used in automatic control, energy monitor, light control, home security, and remote control

[7].

Energy smart homes can be designed to focus on different areas that contribute to the energy consumption of the house, for example, lighting system and other electrical units.

However, non-electrical systems such as window systems can also be designed as part of smart homes by using various advanced glazing technologies that can be adjusted based on the residents’ needs. One type of smart glass that is called dynamic or switchable glass, which can change from clear to tint to block light or heat, is fabricated by applying a layer of ceramic coating or tin oxide nanocrystals on the glass. The ceramic coating, for example, darkens when low-voltage electricity is applied to it [65].

Another example of systems with control capabilities in energy smart homes is the application of multiple sensor-based autonomous monitoring and control system that is

101 basically a combination of temperature and motion sensors to optimize the activity of appliances or mechanical systems such as HVAC based on the temperature and human activities in a building to save more energy. The application of this system can reduce the energy consumption from 35 kWh to 15 kWh compared with a normal home [10]. These systems with control capabilities typically have a graphical user interface (GUI) that helps the user to monitor and control the operation/performance of different units in smart homes.

Such control systems can be as simple as smart sockets that can measure energy consumption, transmit data to web server for users, and be equipped with a relay that can turn on/off the appliances connected to that socket [174]. Furthermore, the hardware that controls the appliances can be as simple as an Arduino kit that is connected to a fan or any other appliance in a house [175].

A study conducted at University of Florida by Helal et al. (2005) is focused on a programmable system that consists of different layers such as sensors, service layer, and application layer that are basically the sensing, processing, and control (command) sections, respectively. When the platform is switched on, the required data such as behavioral information that are required to interact with components such as devices, and sensors will be read from a memory that is a EEPROM store and will be translated to either human-readable or machine-readable specifications, and from that point on, proper commands will be sent to an end-system. The processing module is also an 8-bit Atmel

ATmega 128 processor that can be programmed according to the properties of the components attached to it, e.g., sensors, and the devices that need to be controlled [11]).

Two more innovative systems used in this smart house, called Gator Tech Smart House,

102 are smart plugs and floor. The smart plug that is equipped with a RFID (Radio Frequency

Identification) sensor can automatically detect, read, and connect a plug to the main computer as soon as an electrical device is installed. Smart floor employs sensors connected to the bottom of the flooring tiles in order to detect the location of the user. All commands and decisions made in this smart system are based on an analysis executed in the programmed processing unit and the actions selected by the user on a GUI installed on a computer without the need for any energy simulation or advanced computations.

The controlling system can function as simple as a remote on/off service [176, 177]; however, according to Fomin and Orlov (2015), even switching on/off can be based on more complicated decision making process based on residents’ posture (e.g., lying, sitting, and standing) [178]. Kim et al. (2011) proposed a real-time energy monitoring and controlling system that employs ZigBee sensor networks. The data processing module will receive the acquired data from sensors in terms of electricity consumption, and the web server in the central processing module can be designed in a way that provides the user with energy consumption, price comparison, and statistical analysis. The results will be sent to a PC or mobile device with an application installed that can be used as the GUI.

Depending on the results, the user can decide to turn a device on/off and send a command through the electronic device and the GUI [9]; accordingly, the data analysis is relatively simple and does not need any complicated whole house modeling to obtain the desired results such as total energy consumption or price estimation. Control of appliances by mobile user interface is a technology that is readily available, and there are studies focused on off-the-shelf products that can help the home owner to monitor, control, and compare

103 the energy consumption of their devices. One example is the research presented by Weis and Guinard (2010) who studied application of smart outlets and using web server for receiving and sending data to control the appliances [179]. LaMarche et al. (2011) reviewed products contributing to Home Energy Management (HEM) and reviewed the products with control capabilities, including smart plugs with remote access and timer, smart thermostats that control the temperature, and systems that can automatically stop appliances use during peak hours (e.g. systems with DR capabilities) [170].

The control systems can be programmed based on simple systems such as sensors inputs or timers. Park et al. (2013) studied the functionality and effectiveness of transforming a common power strip to a smart energy management system (SEMS) by using motion sensors and also applying time and power limits. An 8-bit low power microcontroller unit

(MCU), energy metering IC, Zigbee communication module, power relay, and a LCD are the main components of this system that led to a lower energy consumption [180].

Bluetooth low energy (BLE) systems are also discussed in the literature as another option for a remote home automation monitoring and control systems [181]. These simple on/off

Bluetooth systems not only can contribute to energy smart homes, they can also be beneficial to the elderly or people with disabilities [182].

It can be noticed that most of the systems with control capabilities that works based on simple processing capabilities are either focused on simple command signals such as turning on/off the lights, appliances, or the HVAC systems or they are dedicated and

104 contributing to the communication technologies such as wireless systems (e.g., ZigBee and

Bluetooth).

Shaikh et al. (2014) studied smart energy buildings (SEB) in different countries with a focus on intelligent control of energy and comfort of the residents. Their statistical study shows that top five controlling systems used in different studies are mostly focused on

MBPC (model based predictive control), MAST (multi-agent system technology), Fuzzy,

On/Off, and Scheduling. Moreover, the top five optimization methods are focused on GA

(genetic algorithm), MOPSO (Multi-objective Particle Swarm Optimization), scheduling optimization, decision process, and PSO (particle swarm optimization) [183]. Beaudin &

Zareipour (2015) also studied the multi-objective optimization methods in more details and identified the three most common methods as weighted sum, bounded objectives, and physical programming [153]. Shaikh et al. (2014) identified the top five simulation tools as MATLAB, EnergyPlus, Simulink, TRNSYS, and GenOpt. As for the types of buildings researchers have studied for smart features, of the literature reviewed by Shaikh et al.

(2014) and Labeodan et al. (2015) it seems that residential and commercial buildings are the focus of discussion each for 30% of the literature, while office buildings have been of interest to 40% of the researchers. Most of the software related to energy smart home in various studies are energy simulation tools; however, there are also other tools such as

LabVIEW that can specifically contribute to data acquisition and control systems in energy smart homes [183, 184]. Hamed (2012) studied the implementation of a control system in a smart home using this tool in order to control lighting, fire and burglar alarm, and temperature, noting that the latter can contribute to the energy efficiency of the house [185].

105

Although all of these systems contribute to the control unit, the command signals are based on complicated processes using modeling software and optimization methods, which could be placed under the third category, i.e., systems with advanced data processing capabilities.

4.8.3 Systems with Advanced Data Processing Capabilities

In the category of systems with Advanced Data Processing Capabilities in energy smart homes a high-level data analysis such as whole house energy modeling or applying an optimization method to optimize the energy consumption based on the energy tariff and residents’ comfort is required. Such is the case in the ThinkHome project [12] where researchers created a comprehensive knowledge-based system in order to reach higher energy efficiency and intelligent control system in a smart home that is equipped with building automation systems (BAS). Figure 4-4 illustrates the types of information required to develop this knowledge-based system, e.g., building information that refers to data such as wall thickness, square footage, spaces, and materials. Such information is required in order to be able to perform more optimized control strategies toward a smart home with higher energy efficiency. Accordingly, the data stored in a building information model

(BIM) and Green Building XML (gbXML) was selected as the open format of BIM. This is an example of more complicated data analysis in smart home system, where the simulation was executed using MATLAB/Simulink and the HAMLab tools (HAMBase), the latter being used for modeling heat and vapor flows in buildings. For the model physics required in HAMBase, a computer model referred to as ELAN was used for building energy design. In this system, the energy usage and the user behavior are optimized by

106 obtaining some patterns for long time spans. These patterns and profiles are obtained from daily data collections from smart metering and sensors. Eventually based on the outputs, proper action will be taken by controlling unit such as adjusting the thermostat [12].

Figure 4-4. Major Concepts in ThinkHome Knowledge Base System. Adapted from: [12]

Another example of equipping a smart house with a decision making tool is MavHome

(Managing An Intelligent Versatile Home) studied by Cook et al. (2003). The main purpose of this research was to equip the house with an algorithm to predict, reason about, and adapt to its residents with whom it can interact and record their behavior. This system works based on a combination of collection of activities in a database, prediction of resident actions, identification of residents from observed activities, mobility prediction, robotic assistants, multimedia adaptability, and intelligent control and visualization of home activities through the ResiSim 3D simulator. These components are connected using

CORBA, which is an interface between software component and electric devices; therefore, off-the-shelf devices such as automated mini-blinds can be controlled by this system. There are several prediction algorithms integrated within this system that helps predicting resident’s activities [106]. Adaptability to user’s behavior and schedule changes is an important feature of systems with processing capabilities and Keshtkar et al. (2016) proposed an algorithm in their work that adjust the Programmable Communicating 107

Thermostat (PCT) or Programmable Thermostat (PT) based on the outdoor temperature, residents’ presence, and DR programs. The Matlab GUI is used to perform the energy simulation and the prototype of the smart AC is also developed in their work using ZigBee technology and it was shown that applying the adaptive program can lead to about 27-38% energy conservation [186].

Yang and Li (2010) proposed a framework of Pervasive Service-Oriented Networks

(PERSON) to study the Energy Management System (EMS) on the costumer side in a

Smart Grid (SG). They recognized that the existing EMSs are limited to monitoring and controlling systems, which means they might not be intelligent enough and adaptable to the needs of SG such as demand response (DR) features, making energy management decisions based on intelligent algorithms, and ability to achieve optimized performance.

Yang and Li (2010) defined the following three major challenges for the proposed framework: heterogeneity (related to communication protocols or media), distributed system (related to appliances spread throughout the house), and dynamicity (related to energy flow). To address the challenges, the proposed framework is used in a demand response (DR) application as a ZigBee-based system to demonstrate its effectiveness when it is used for dimmable lights. The processing unit considers the supply limit, calculates the related load target, and determines whether the total power load of the house is to be increased or decreased depending on the load state (e.g. peak or non-peak) [187].

Francillette et al. (2016) also defined the dynamic nature of contexts within a smart home as a challenge for smart home design, and proposed the idea of “context-aware design” as

108 a solution. The context-aware system adjusts the decisions or responses of the smart system based on the current conditions without developing predefined responses as outputs [188].

There are examples of research projects dedicated to overcome the challenges related to the interoperability of multiple systems in smart homes such as interoperability of sensors and actuators from different vendors. SmartSEAL interconnection system is developed for the SEAL project focused on Home Automation (HA) in energy smart homes. Robot

Operating System (ROS) is the name of the middleware used in this system and tested in two real environments including “Palazzo Fogazzaro in Schio” and ”Le Case” childhood school in Malo. There is also a habit learning module suggested in this research that uses

Convolutional Neural Network (CNN) to recognize residents’ behavior and objects in the building using a smart camera. The actuators control the boiler with energy consumption monitoring system, automated gate actuators, smart light, and heating plant [189].

University of Colorado developed a smart house known as Adaptive House in Marshal,

Colorado, outside of Boulder that is equipped with over 75 sensors monitoring temperature, ambient light level, sound, motion, and door and window opening [105]. The actuators inside the house control the furnace, space heaters, water heater, lighting units, and ceiling fans through neural network reinforcement learning, prediction techniques, and Q learning that is optimization method. For example, it can predict the arrival time of the residents to turn on the heater or the statistical pattern of using hot water in order to provide hot water whenever it is needed to increase the comfort and energy saving aspects of the house. It has been observed that using statistical regularities and Adaptive Control of Home

109

Environments (ACHE) has been effective to optimize the energy consumption. Mozer

(2015) also reports that residents make better decisions after energy related information is provided to them. Application of neural network for smart homes and energy efficiency is also used in other studies for purposes such as predicting the residents’ presences and behavior to optimize the active time of heating systems [190].

One of the systems studied by Tascikaraoglu et al. (2014) is demand side management

(DSM) system that adjusts the activity of appliances and energy consumption based on a prediction algorithm for residential renewable sources such as wind turbines, solar PV, or battery bank. This system uses artificial neural network (ANN) algorithms to predict the supply of energy in order to enable better management of appliances and energy consumption, for example, by turning off an appliance during the peak-time and leave its activity for off-peak time, which is the same concept as demand response (DR) in which the load shifting and re-scheduling operations are conducted using fuzzy logic [191]. DSM is the focus of other studies such as that proposed by Iqbal et al. (2016) regarding an ECG optimization method used for managing the energy consumption and costs. In such optimization process, both appliances and microgrid source including PV panels and local diesel generator, the optimization is focused on minimizing the costs such as fuel cost, startup cost, and maintenance [192]. Basu et al. (2013) also studied the effectiveness and accuracy of using a prediction system for home appliances. The prediction system contributes to the DSM to organize the energy consumption and production and is capable of learning and being adjusted based on the residents’ behavior; and it needs at least 24

110 hours to have enough data to modify the system [193]. Developing a prediction method for residents’ behavior using stochastic methods is also studied in other papers [194].

Supervisory control and data acquisition (SCADA) is a system that is used for remote monitoring and control systems. SCADA House Intelligent Management (SHIM) system was proposed by Fernandes et al. (2013) that includes a set of optimization algorithms to optimize the power consumption, the micro generation system, the charge and discharge of the electric or plug-in hybrid vehicles, and the participation in Demand Response (DR) programs [195]. The main goal of this system is to prioritize the active time of devices. In the general idea of this energy smart system, data are obtained from data acquisition unit, and the outputs of the processing unit that are obtained based on the optimization algorithm can be transmitted to the control units including the actuators and motor drives. Fernandes et al. (2014) used this system in a demand response context in order to define the priorities of various loads such as washing machine and lighting. Use of the optimization algorithm led to defining the minimum time of functioning of an appliance during the demand response event and using dynamic load priority (DLP) method led to more efficient load management [196]. The same optimization algorithm is used in another research conducted by Fernandes et al. (2016) for energy management of an office building, and the algorithm is implemented in the processing unit of the smart system [197]. Application of optimization algorithm in smart home systems also contributes to distribute energy resources (DER) optimization. DER is basically a combination of small energy resources that provide the total required energy of a house. Pedrasa et al. (2010) studied application of particle swarm optimization (PSO) to optimize the energy services in a smart home case

111 study [198]. Optimization for distributed systems is not limited to the energy resources and it can be also dedicated to load profiles such as distributed integrated energy management

(diEM) system that is studied by Honold et al. (2016) to organize the load profiles separately and eliminate the need for a central decision-making unit [199]. Optimizing the energy consumption based on the energy tariff and residents’ comfort level has been the focus of different studies, including using Mixed Integer Linear Programming (MILP)

[200, 201], Multi-Objective Genetic Algorithm (MOGA) [202], Nelder-Mead geometrical optimization method [203], mixed objective function [204], general algebraic modeling system (GAMS) [205], or a similar study conducted by Reka and Ramesh (2016) by using

MATLAB to apply PSO for creating minimum load schedule during peak load [206]. There are also online open source applications such as ThingSpeak application that can collect environmental data and perform basic processing using MATLAB or MathWorks. An example for application of such open source cloud-based systems is a prototype design that collects data through meters and adopt a genetic algorithm to adjust the energy consumption of the house based on the priorities defined by residents through a mobile application [207].

Microgrids can also be an important component in smart homes, but if there are multiple houses using the same microgrid, then a fair distribution and optimization of energy among these houses becomes more complex. However, Zhang et al. (2014) shows that lexicographic minimax method that uses a mixed integer linear programming (MILP) approach can be beneficial with microgrids that involve multiple distributed energy resources (DER) [208]. Another tool that can be helpful in considering real-time events

112 such as number of people in a certain area or natural lighting is fuzzy logic based controllers studied by Ghadi et al. (2014) [209]. Application of Layered hidden Markov models

(LHMM) was also studied [210] for recognition of desk activities and people count in an office building equipped with plug-in power meters, passive infrared sensors, and ultrasound range (USR) finders to control the lighting system.

It is noted that the current energy smart homes with advanced data processing capabilities are mostly dedicated to systems with optimization capabilities that contribute to the DR systems, houses connected to microgrids, and houses connected to DER. Also, there are some works focused on the energy smart home systems with more complex processing capabilities such as systems equipped with energy simulation tools connected to other instrumentations of smart homes.

4.9 Summary and Concluding Remarks

The definition of smart home covers different areas such as ambient intelligence, sensors and actuators, home networking, connecting the house to outside internet world, remote or central monitoring and controlling functionalities and services, automatic control, adjusting house functions to residents’ needs, and assisting in performing the activities of daily life. Some of the current characteristics of smart homes have existed even since the very beginning of the idea back in the 1930’s. This paper has reviewed different definitions for smart homes or smart dwellings as summarized in Table 4-1.

113

The focus or intended service of smart homes could be quite different, and researchers have identified various services of smart homes that are mainly focused on four major aspects including improving the comfort, security, health, and energy saving. The proposed services discussed in the literature are reviewed and summarized in Table 4-2, noting that energy efficiency or saving has been one of the major goals pursued in smart homes.

Several research projects that focused on various full-scale buildings such as living-labs built by research institutes or universities are reviewed and summarized in Table 4-3. Most of these prototypes or actual buildings are focused on facilitating life for the elderly or people with disabilities, energy saving, and residents’ entertainment, and provide ample opportunities to investigate the feasibility and performance of different systems in smart homes and the interaction of people in such environments.

Some of the reviewed studies have identified different types of issues or challenges in different components of smart homes. Such challenges include reliability, security, ease of human interaction, cost, and have a dynamic decision making process that adjusts itself based on other dynamic properties in smart home environment. However, these challenges do not hinder the use of smart home systems; rather, they are issues that need to be taken into account in future developments in order to improve the performance of such systems.

The main focus of this state-of-the-art review paper was, however, on energy smart homes, which is basically a smart home that contributes to the energy management. Relevant literature contributing to the energy smart homes was reviewed and summarized in Table

4-5 and categorized under three major headings including energy monitoring systems,

114 systems with control capabilities, and systems with advanced data processing capabilities.

The monitoring systems are limited to monitoring the energy-related data and the obtained data are processed only for statistical evaluation. However, systems with control capabilities can go one step further and perform some actions such as turning on/off the heating system or activate a control unit to turn on/off lights based on the acquired data from sensors. On the other hand, systems with advanced data processing capabilities are defined to be able to perform more complicated operations on the acquired data from sensors by using different tools such as optimization algorithms or simulation software.

Not all systems in energy smart homes are focused on reducing energy consumption. There are multiple studies contributing to other aspects of energy management such as integration of homes with microgrids, smart grids, and demand response systems. Concepts such as demand response contributes to reduction in energy-related costs rather than energy saving.

These systems are mostly categorized under systems with advanced processing capabilities because most use more complicated decision making algorithms such as optimization algorithms rather than basic statistical calculations used to obtain the average or total energy consumption. The literature that is summarized in Table 4-5 was reviewed based on different properties of these systems such as technologies, algorithms, software, input/output data, and the effect of the system in terms of energy saving. The review shows that most of the systems with advanced processing units are equipped with either the optimization algorithms or energy simulation systems in order to obtain more detailed information about the environment they are installed in. The following concluding remarks are also offered:

115

 Health, security, reducing the living costs, and welfare in general are the main

motives in development of smart homes with different intended services.

 At the current state, users are more interested in automation aspect of smart homes

focused on health, security, and comfort. Users are more inclined toward easy to

use energy smart home systems such as automated temperature control or demand

response systems. However, more advanced systems could become more prevalent

after emerging energy supply systems such as smart grids or microgrids. Moreover,

higher utility cost can also play an important role in motivating users in adopting

energy smart home systems that are not common at this point.

 The current monitoring, measuring, communication, and controlling technologies

and also simulation and mathematical tools seem to be adequate for the current

demand and intended services of smart homes; however, it seems the growth rate

in terms of technologies is higher in communication (e.g., WiFi and ZigBee) and

controlling tools (e.g. relays and electronic switches) as opposed to monitoring,

measuring (e.g., sensors), and simulation tools.

 The market trend and existing technologies related to energy smart homes are

mainly dedicated to energy monitoring systems that work based on measuring the

whole house energy consumption.

 The research trend, however, is more inclined toward integration of energy smart

homes with smart grids and microgrids. Therefore, most of the more advanced

116

systems use optimization algorithm and are dedicated to systems with demand

response or distributed energy resources capabilities.

 There seems to be no guidelines to define the requirements in design and use phase

of smart buildings. Development of a unified set of guidelines can be a viable and

necessary goal in future and it can also be beneficial to overcome challenges such

as adaptability of different systems, ease of human interaction, and security in terms

of data transmission.

4.10 References

[61] S.-M. Oh, Y.-J. Kim, C.-S. Park and I.-H. Kim, "PROCESS-DRIVEN BIM-

BASED OPTIMAL DESIGN USING INTEGRATION OF ENERGYPLUS, GENETIC

ALGORITHM, AND PARETO OPTIMALITY," in 12th Conference of International

Building Performance Simulation Association, Sydney, 14-16 November, 2011.

[62] T. Yamazaki, "Beyond the Smart Home," 2006.

[63] A. Berlo, B. Allen, J. Ekberg, K. Fellbaum, M. Hampicke and C. Willems, "Design

Guidelines on Smart Homes," A COST 219bis Publication, 1999.

[64] Oxford, "Oxford," 2016. [Online]. Available: https://en.oxforddictionaries.com/definition/smart_home. [Accessed 11 November 2016].

117

[65] M. Miller, The Internet of Things: How Smart TVs, Smart Cars, Smart Homes, and

Smart Cities Are Changing the World, 1st ed., Que, Pearson Education: Pearson Education,

2015.

[66] M. Alam, M. Reaz and M. Ali, "A Review of Smart Homes – Past, Present, and

Future," Transactions on Systems, Man, and Cybernetics—Part C: Applications and

Reviews, vol. 42, no. 6, 2012.

[67] R. Robles and T. Kim, "Applications, Systems and Methods in Smart Home

Technology: A Review," International Journal of Advanced Science and Technology, vol.

15, pp. 37-48, 2010.

[68] C. Lo, D. Chen and K. Chao, "Dynamic Data Driven Smart Home System based on a Service Component Architecture," 2010.

[69] M. Chan, E. Campo, D. Esteve and J. Fourniols, "Smart homes — Current features and future perspectives," Maturitas, vol. 64, pp. 90-97, 2009.

[70] G. Sandstrom, S. Gustavsson, S. Lundberg, U. Keijer and S. Junestrand, "Long- term viability of smart home systems–Business modeling and conceptual requirements on technology," York, 2005.

[71] T. Laberg, "Smart Home Technology; Technology supporting independent living - does it have an impact on health?," Tromso, Norway, 2005.

118

[72] T. Koskela and K. Vainio-Mattila, "Evolution towards smart home environments: empirical evaluation of three user interfaces," Personal and Ubiquitous Computing, vol. 8, pp. 234-240, 2004.

[73] R. Harper, Inside the Smart Home, 1st ed., Springer-Verlag, 2003.

[74] N. King, "Smart Home – A Definition," September 2003. [Online]. Available: http://www.housingcare.org/downloads/kbase/2545.pdf. [Accessed 11 November 2016].

[75] F. Aldrich, "Smart Homes: Past, Present and Future," in Inside the Smart Home,

UK, Springer-Verlag , 2003, pp. 17-39.

[76] M. Wigginton and J. Harris, Intelligent Skins, Routledge, 2002.

[77] D. Briere and P. Hurley, Smart Homes for Dummies, Wiley Publishing, 1999.

[78] G. Austin, "Safer, More Convenient Home Wiring Developed," 1985. [Online].

Available: http://articles.chicagotribune.com/1985-01-11/news/8501030134_1_smart- house-ken-geremia-wiring. [Accessed 11 November 2016].

[79] S. Kerch, "Smart House Borrowing from NASA to Create a Fantasy Home of

Future," 1985. [Online]. Available: http://articles.chicagotribune.com/1985-02-

16/news/8501100121_1_smart-house-new-wiring-system-home-builders-research- foundation. [Accessed 11 November 2016].

[80] S. Kerch, "Smart House Opens Door to Electronic Marvels of the Future," 1986.

[Online]. Available: http://articles.chicagotribune.com/1986-03-

119

16/business/8601200336_1_smart-house-wiring-system-housing-industry. [Accessed 11

November 2016].

[81] R. Youngblood, "Smart House` May Take Over In The 1990s," 1986. [Online].

Available: http://articles.chicagotribune.com/1986-05-03/news/8602010304_1_wiring- system-smart-house-nahb. [Accessed 11 November 2016].

[82] E. Allison, "`Smart House` About To End Its Run," 1992. [Online]. Available: http://articles.chicagotribune.com/1992-10-17/news/9204040052_1_smart-house-bright- room. [Accessed 11 November 2016].

[83] A. Dingli and D. Seychell, The New Digital Natives, Cutting the Chord, 1st ed.,

Springer-Verlag Berlin Heidelberg, 2015.

[84] N. Balta-Ozkan, R. Davidson, M. Bicket and L. Whitmarsh, "The development of smart homes market in the UK," Energy, vol. 60, pp. 361-372, 2013.

[85] R. Lutolf, "SMART HOME CONCEPT AND THE INTEGRATION OF ENERGY

METERS INTO A HOME BASED SYSTEM," Glasgow, 1992.

[86] N. Balta-Ozkan, B. Boteler and O. Amerighi, "European smart home market development: Public views on technical and economic aspects across the United Kingdom,

Germany and Italy," Energy Research & Social Science, vol. 3, pp. 65-77, 2014.

[87] Q. Le, H. Nguyen and T. Barnett, "Smart homes for older people: Positive aging in a digital world," Future Internet, vol. 4, pp. 607-617, 2012.

120

[88] N. Labonnote and K. Hoyland, "Smart home technologies that support independent living: challenges and opportunities for the building industry – a systematic mapping study," Intelligent Buildings International, Taylor & Francis, 2015.

[89] L. Liu, E. Stroulia, I. Nikolaidis, A. Miguel-Cruz and A. Rincon, "Smart homes and home health monitoring technologies for older adults: A systematic review," International

Journal of Medical Informatics, vol. 91, pp. 44-59, 2016.

[90] D. Bonino, E. Castellina, F. Corno and L. De Russis, "DOGeye: Controlling your home with eye interaction," Interacting with Computers, vol. 23, pp. 484-498, 2011.

[91] A. Bejarano, A. Fernández, M. Jimeno, A. Salazar and P. Wightman, "Towards the

Evolution of Smart Home Environments: A Survey," International Journal of Automation and Smart Technology, vol. 6, no. 3, pp. 105-136, 2016.

[92] A. N. Gupta and A. K. Singh, "Home Automation Systems," International Journal of Engineering Science and Computing, vol. 6, no. 5, pp. 4828-4833, 2016.

[93] T. Mendes, R. Godina, E. Rodrigues, J. Matias and J. Catalao, "Smart Home

Communication Technologies and Applications: Wireless Protocol Assessment for Home

Area Network Resources," Energies, vol. 8, pp. 7279-7311, 2015.

[94] G. M. Toschi, L. B. Campos and C. E. Cugnasca, "Home automation networks: A survey," Computer Standards & Interfaces, vol. 50, pp. 42-54, 2017.

121

[95] M. Patel and S. Mukherjee, Handbook of Advanced Lighting Technology (Lighting

Control Protocols and Standards), Springer International Publishing Switzerland, 2016.

[96] S. Madakam and R. Ramaswamy, "Smart Homes (Conceptual Views)," New Delhi,

India, 2014.

[97] H. Altan, L. Han, K. Wagner, J. Ortega and A. Deep, "Energy Use in Housing," in

ZEMCH: Toward the Delivery of Zero Energy Mass Custom Homes, Springer Tracts in

Civil Engineering, 2016, pp. 175-207.

[98] T. Perumal, A. Ramli, C. Leong and S. Mansor, "Interoperability among

Heterogeneous Systems in Smart Home Environment," Bali, 2008.

[99] J. Kientz, E. Price, S. Patel, E. Myantt, B. Jones and G. Abowd, "The Georgia Tech

Aware Home," Florence, Italy, 2008.

[100] K. Larson and R. Topping, "PlaceLab: A House_n + TIAX Initiative," 2003.

[Online]. Available: http://web.mit.edu/cron/group/house_n/documents/PlaceLab.pdf.

[Accessed 11 November 2016].

[101] S. Madakam, "Smart Homes (Conceptual Views)," 2014.

[102] Smarthus, "Smarthome (Smarthus)," 2011. [Online]. Available: https://www.facebook.com/198380953533766/videos/10150215384429973/. [Accessed

10 November 2016].

122

[103] E. Farella, M. Falavigna and B. Ricco, "Aware and smart environments: The

Casattenta project," Bari, Italy, 2009.

[104] S. Baeg, J. Park, J. Koh, K. Park and M. Baeg, "Building a Smart Home

Environment for Service Robots Based on RFID and Sensor Networks," Seoul, Korea,

2007.

[105] M. Mozer, "Lessons From An Adaptive House," 2015. [Online]. Available: https://www.researchgate.net/publication/266594221_Lessons_From_An_Adaptive_Hou se. [Accessed 11 November 2016].

[106] D. Cook, M. Youngblood, E. Heierman, K. Gopalratnam, K. Rao, A. Litvin and F.

Khawaj, "MavHome: An Agent-Based Smart Home," 2003.

[107] C. Joseph and D. Schaad, "Energy Monitoring and Feedback in Duke Student

Residences," 2011. [Online]. Available: https://smarthome.duke.edu/fellow- projects/energy-monitoring. [Accessed 11 November 2016].

[108] M. Delmas and W. Kaiser, "Behavioral responses to real-time individual energy usage information: A large scale experiment," ARB Contract 10-332, University of

California, Los Angeles, Prepared for the California Air Resources Board and the

California Environmental Protection Agency, 2014.

[109] Orange Smart House, "Orange Smart House," 2016. [Online]. Available: http://www.surrey.ac.uk/dwrc/archived-projects/orange-smart-house/. [Accessed 11

November 2016]. 123

[110] C. Baedeker, K. Greiff, V. Grinewitschus, M. Hasselkuß, D. Keyson, J. Knutsson,

C. Liedtke, D. Lockton, T. Lovric, G. Morrison, M. v. Rijn, H. Rohn, S. Silvester, W. v.

Harinxma and L. Virdee, "Transition through sustainable Product and Service Innovations in Sustainable Living Labs: application of user-centred research methodology within four

Living Labs in Northern Europe," Utrecht, The Netherlands, 2014.

[111] D. V. Keyson, D. Lockton and O. Guerra-Santin, Living Labs Design and

Assessment of Sustainable Living, 1st ed., Springer International Publishing, 2017.

[112] GridSTAR, "GridSTAR," 2016. [Online]. Available: https://smartenergyacademy.psu.edu/gridstar/smart-grid-experience-center. [Accessed 11

November 2016].

[113] E. Kamel and A. Memari, "Energy Efficiency and Sustainability Attributes of the

Solar Decathlon Projects," 2015.

[114] MorningStar, "MorningStar Solar Home," 2016. [Online]. Available: http://sustainability.psu.edu/morningstar-solar-home. [Accessed 10 November 2016].

[115] G. Kleineidam, M. Krasser and M. Reischböck, "The cellular approach: smart energy region Wunsiedel. Testbed for smart grid, smart metering and smart home solutions," Electr Eng, pp. 1-6, 2016.

[116] L. Jiang, D. Liu and B. Yang, "SMART HOME RESEARCH," Shanghai, 2004.

124

[117] NREL, "AHEM Lab Opens Doors to New Technology Test Bed at NREL," 2016.

[Online]. Available: http://www.nrel.gov/docs/fy13osti/55578.pdf. [Accessed 11

November 2016].

[118] Aware Home, "Aware Home," 2016. [Online]. Available: http://awarehome.imtc.gatech.edu/. [Accessed 11 November 2016].

[119] Oberlin, "buildingdashboard," 2016. [Online]. Available: http://buildingdashboard.net/oberlin/#/oberlin/ajlc. [Accessed 11 November 2016].

[120] S. Junestrand, S. Lenman, B. Thuresson and K. Tollmar, "Private and Public Spaces

– Video Mediated Communication in a Future Home Environment," CID-61, Stockholm,

Sweden, 1999.

[121] Custodian, "Custodian," 2016. [Online]. Available: http://www4.rgu.ac.uk/sss/research/page.cfm?pge=33336. [Accessed 11 November 2016].

[122] B. Brumitt, B. Meyers, J. Krumm, A. Kern and S. Shafer, "EasyLiving:

Technologies for Intelligent Environments," Bristol, UK, 2000.

[123] PlaceLab, "PlaceLab," 2016. [Online]. Available: http://web.mit.edu/cron/group/house_n/placelab.html. [Accessed 11 November 2016].

[124] S. Intille, "Designing a Home of the Future," IEEE Pervasive Computing, vol. 1, no. 2, pp. 76-82, 2002.

125

[125] S. Intille, K. Larson, E. Tapia, J. Beaudin, P. Kaushik, J. Nawyn and R. Rockinson,

"Using a live-in laboratory for ubiquitous computing research," in Pervasive Computing,

349-365, Springer Berlin Heidelberg, 2006.

[126] A. Karami, A. Fleury, J. Boonaert and S. Lecoeuche, "User in the Loop: Adaptive

Smart Homes Exploiting User Feedback—State of the Art and Future Directions,"

Information, vol. 7, no. 35, pp. 1-18, 2016.

[127] S. Das, D. Cook, A. Bhattacharya, E. Heierman and T. Lin, "The role of prediction algorithms in the MavHome smart home architecture," IEEE wireless communications, pp.

77-84, 2002.

[128] U. o. Virginia, "UVa's Medical Automation Research Center to Demo New

Effective Monitors at the White House Conference on Aging," 2005. [Online]. Available: http://www.prnewswire.com/news-releases/uvas-medical-automation-research-center-to- demo-new-effective-monitors-at-the-white-house-conference-on-aging-55480052.html.

[Accessed 11 November 2016].

[129] H.-A. Dang, S. Gaalou, B. Delinchant and F. Wurtz, "BUILDING SIMULATION

OF ENERGY CONSUMPTION AND AMBIENT TEMPERATURE : APPLICATION

TO THE PREDIS PLATFORM," Chambéry, 2013.

[130] M. Hu, S. Lu and Y. C, "Stochastic programming and market equilibrium analysis of microgrids energy management systems," Energy, vol. 113, pp. 662-670, 2016.

126

[131] Siemens, 2016. [Online]. Available: http://www.buildingtechnologies.siemens.com/bt/global/en/buildingautomation- hvac/hvac-products/home-automation-system-synco- living/Documents/synco_living/all/en/index.html. [Accessed 08 11 2016].

[132] B. Johanson, "The Stanford Interactive Workspaces Project," Stanford, CS technical reports. Palo Alto, Computer Science, 2004.

[133] G. Sandstrom, Smart Homes and User Values - Long-term evaluation of IT-services in Residential and Single Family Dwellings, Stockholm, Sweden: Doctoral Thesis Royal

Institute of Technology, School of Architecture and the Built Environment, 2009.

[134] K. Bickerstaff, E. Hinton and H. Bulkeley, "Decarbonisation at home: The contingent politics of experimental domestic energy technologies," Environment and

Planning A, pp. 1-20, 2016.

[135] S. Samuel, "A Review of Connectivity Challenges in IoT-Smart Home," Muscat,

2016.

[136] A. Jacobsson, M. Boldt and B. Carlsson, "A risk analysis of a smart home automation system," Future Generation Computer Systems, vol. 56, pp. 719-733, 2016.

[137] D. Khedekar, A. Truco, D. Oteyza and G. Huertas, "Home Automation—A Fast -

Expanding Market," Thunderbird Int'l Bus Rev., pp. 1-13, 2016.

127

[138] C. Wilson, T. Hargreaves and R. Hauxwell-Baldwin, "Smart homes and their users: a systematic analysis and key challenges," Pers Ubiquit Comput, vol. 19, pp. 463-476,

2014.

[139] U. Premarathne, I. Khalil and M. Atiquzzaman, "Trust based reliable transmissions strategies for smart home energy consumption management in cognitive radio based smart grid," Ad Hoc Networks, vol. 41, pp. 15-29, 2016.

[140] A. Rathnayaka, V. Potdar and S. Kuruppu, "Energy Resource Management in

Smart Home: State of the Art and Challenges Ahead," Sustainability in Energy and

Buildings, pp. 403-411, 2012.

[141] B. L. R. Stojkoska and K. V. Trivodaliev, "A review of Internet of Things for smart home: Challenges and solutions," Journal of Cleaner Production, vol. 140, p. 1454e1464,

2017.

[142] M. B. Bulut, BUILDING AS ACTIVE ELEMENTS OF ENERGY SYSTEMS,

Västerås, Sweden: School of Business, Society and Engineering, 2016.

[143] M. LaMonica, "Nest Thermostat Slays Peak Power," 22 April 2013. [Online].

Available: https://www.technologyreview.com/s/513961/nest-thermostat-slays-peak- power/. [Accessed 11 November 2016].

[144] S. Aman, Y. Simmhan and V. Prasanna, "Energy Management Systems: State of the Art and Emerging Trends," IEEE Commun. Mag., vol. 51, no. 1, pp. 114-119, 2013.

128

[145] S. Solaimani, W. Keijzer-Broers and H. Bouwman, "What we do - And don't -

Know about the Smart Home: An analysis of the Smart Home literature," Indoor and Built

Environment, pp. 1-14, 2016.

[146] G. Lobacco, S. Carlucci and E. Lofstrom, "A Review of Systems and Technologies for Smart Homes and Smart Grids," Energies, vol. 9, no. 5, pp. 1-33, 2016.

[147] EIA, "Annual Energy Outlook 2012," U.S. Energy Information Administration,

2013.

[148] A. Zipperer, P. Aloise-Young, R. Roche, L. Earle, D. Christensen and P. Bauleo,

"Electric Energy Management in the Smart Home: Perspectives on Enabling Technologies and Consumer Behavior," NREL/JA-5500-57586, Contract No. DE-AC36-08GO28308,

2013.

[149] A. Vega, F. Santamaria and E. Rivas, "Modeling for home electric energy management: A review," Renewable and Sustainable Energy Reviews, vol. 52, p. 948–

959, 2014.

[150] I. Abubakar, S. Khalid, M. Mustafa, H. Shareef and M. Mustapha, "Application of load monitoring in appliances’ energy management – A review," Renewable and

Sustainable Energy Reviews, vol. 67, pp. 235-245, 2017.

[151] B. Zhou, W. Li, K. Chan, Y. Cao, Y. Kuang and X. Liu, "Smart home energy management systems: Concept, configurations, and scheduling strategies," Renewable and

Sustainable Energy Reviews, vol. 61, pp. 30-40, 2016. 129

[152] Z. Wang and R. Paranjabe, "A Multi-Agent System of Evaluating Residential

Demand Response," NY, USA, 2016.

[153] M. Beaudin and H. Zareipour, "Home energymanagementsystems:Areviewofmodelling and complexity,"

RenewableandSustainableEnergyReviews, vol. 45, pp. 318-335, 2015.

[154] M. Parvania and M. Fotuhi-Firuzabad, "Demand Response Scheduling by

Stochastic SCUC," IEEE TRANSACTIONS ON SMART GRID, vol. 1, no. 1, pp. 89-98,

2010.

[155] A. Graab, "The Smart Grid: A Smart Solution to a Complicated Problem," 52 Wm.

& Mary L. Rev., pp. 2051-2076, 2011.

[156] Z. Akhtar and M. Saqib, "Microgrids formed by renewable energy integration into power grids pose electrical protection challenges," Renewable Energy , vol. 99, pp. 148-

157, 2016.

[157] H. Jiyai, J. Chuanwen and X. Rong, "A review on distributed energy resources and

MicroGrid," Renewable and Sustainable Energy Reviews , vol. 12, p. 2472–2483, 2008.

[158] R. Lasseter and P. Paigi, "Microgrid: a conceptual solution," 2004.

[159] C. Gellings, "The concept of demand-side management for electric utilities,"

Proceedings of the IEEE, vol. 73, no. 10, p. 1468 – 1470, 1985.

[160] J. Zheng, D. Gao and L. Lin, "Smart Meters in Smart Grid: An Overview," 2013.

130

[161] G. Wood and M. Newborough, "Dynamic energy-consumption indicators for domestic appliances: environment, behaviour and design," Energy and Buildings, vol. 35, pp. 821-841, 2003.

[162] W. Abrahamse, L. Steg, C. Vlek and T. Rothengatter, "A review of intervention studies aimed at household energy conservation," Journal of Environmental Psychology, vol. 25, pp. 273-291, 2005.

[163] L. McCalley, C. Midden and K. Haagdorens, "Computing systems for household energy conservation: Consumer response and social ecological considerations," 2005,

2005.

[164] T. Hargreaves, M. Nye and J. Burgees, "Keeping energy visible? Exploring how householders interact with feedback from smart energy monitors in the longer term,"

Energy Policy, vol. 52, pp. 126-134, 2013.

[165] S. Van Dam, C. Bakker and J. Van Hal, "Home energy monitors: impact over the medium-term," BUILDING RESEARCH &INFORMATION, vol. 38, no. 5, pp. 458-469,

2010.

[166] T. Ueno, F. Sano, O. Saeki and K. Tsuji, "Effectiveness of an energy-consumption information system on energy savings in residential houses based on monitored data,"

Applied Energy , vol. 83, pp. 166-183, 2006.

131

[167] A. Marchiori, D. Hakkarinen, Q. Han and L. Earle, "Circuit-Level Load Monitoring for Household Energy Management," 2011. [Online]. Available: http://pecs.mines.edu/docs/ieeepc11.pdf. [Accessed 11 November 2016].

[168] S. Hallé, S. Gaboury and B. Bouchard, "Towards User Activity Recognition

Through Energy Usage Analysis And Complex Event Processing," Corfu Island, Greece,

2016.

[169] Y. Kim, T. Schmid, Z. Charbiwala and B. Srivastava, "ViridiScope: Design and

Implementation of a Fine Grained Power Monitoring System for Homes," Orlando,

Florida, 2009.

[170] J. LaMarche, K. Cheney, S. Christian and K. Roth, "Home Energy Management

Products & Trends," 2011. [Online]. Available: http://cdn2.hubspot.net/hub/55819/docs/fhcse-hem-products.pdf. [Accessed 11 November

2016].

[171] R. Brewer and P. Johnson, "WattDepot: An open source software ecosystem for enterprise-scale energy data collection, storage, analysis, and visualization," 2010.

[172] J. Vanus, R. Martinek, M. Kelnar, Z. Machacek, P. Bilik, J. Koziorek and J. Zidek,

"Building heating technology in Smart Homeusing PI System management tools,"

Perspectives in Science, vol. 7, p. 114—121, 2016.

132

[173] R. Y. Man Li, H. C. Yu Li, C. K. Mak and T. B. Tang, "Sustainable Smart Home and Home Automation: Big Data Analytics Approach," Interntional Journal of Smart

Home, vol. 10, no. 8, pp. 177-198, 2016.

[174] O. Omole, D. Akpobasah and A. Atayero, "Development of a Smart, Low-cost and

IoT-enabled System for Energy Management," 2016.

[175] D. Nikam, S. Yavalkar, V. Shinde, N. Mate and A. Dhumane, "Smart Home

Automation System," International Journal of Engineering Science and Computing, vol. 6, no. 5, pp. 4897-4900, 2016.

[176] G. Garg, D. Goyal and H. Aggarwal, "Controlling Home Appliances in IOT

Environment," International Journal of Smart Home, vol. 10, no. 8, pp. 11-18, 2016.

[177] R. Chasta, R. Singh, A. Gehlot and R. G. Mishra, "A Smart Building Automation

System," Internation Journal of Smart Home, vol. 10, no. 8, pp. 91-98, 2016.

[178] S. Fomin and A. Orlov, "Simulation modeling of “Smart Home” operations using the data on a man’s position," Procedia Engineering, vol. 129, p. 652 – 656, 2015.

[179] M. Weis and D. Guinard, "Increasing Energy Awareness Through Web-enabled

Power Outlets," Limassol, Cyprus, 2010.

[180] S. Park, M. Choi, B. Kang and S. Park, "Design and Implementation of Smart

Energy Management System for Reducing Power Consumption using ZigBee Wireless

Communication Module," Procedia Computer Science, vol. 19, p. 662 – 668, 2013.

133

[181] J. Sivakrishnan, E. Esakki Vigneswaran and R. Sakthi Vishnu, "Home Automation

Control and Monitoring System Using BLE Device," Middle-East Journal of Scientific

Research, vol. 24, pp. 78-82, 2016.

[182] D. Bajaj, A. Kumar, J. Sahu, N. Charaya and V. Prakash, "Home Automation by

Android Application Based Remote Control," International Journal of Engineering Science and Computing, vol. 6, no. 5, pp. 6053-6056, 2016.

[183] P. Shaikh, N. Mohd Nor, P. Nallagownden, I. Elamvazuthi and I. Ibrahim, "A review on optimized control systems for building energy and comfort management of smart sustainable buildings," Renewable and Sustainable Energy Reviews, vol. 34, p. 409–429,

2014.

[184] T. Labeodan, K. Aduda, G. Boxem and W. Zeiler, "On the application of multi- agent systems in buildings for improved building operations, performance and smart grid interaction – A survey," Renewable and Sustainable Energy Reviews, vol. 50, p. 1405–

1414, 2015.

[185] B. Hamed, "Design & Implementation of Smart House Control Using LabVIEW,"

International Journal of Soft Computing and Engineering (IJSCE), vol. 1, no. 6, pp. 98-

106, 2012.

[186] A. Keshtkar, S. Arzanpour and F. Keshtkar, "Adaptive residential demand-side management using rule-basedtechniques in smart grid environments," Energy and

Buildings, vol. 133, pp. 281-294, 2016.

134

[187] G. Yang and V. Li, "Energy Management System and Pervasive Service-Oriented

Networks," 2010.

[188] Y. Francillette, S. Gaboury, A. Bouzouane and B. Bouchard, "Towards an

Adaptation Model for Smart Homes," Springer International Publishing Switzerland, pp.

83-94, 2016.

[189] E. Bellocchio, G. Costante, S. Cascianelli, P. Valigi and T. A. Ciarfuglia,

"SmartSEAL: A ROS based home automation framework for heterogeneous devices interconnection in smart buildings," Trento, Italy, 2016.

[190] T. Teich, F. Roessler, D. Kretz and S. Franke, "Design of a Prototype Neural

Network for Smart Homes and Energy Efficiency," Procedia Engineering, vol. 69, pp. 603-

608, 2014.

[191] A. Taschikaraoglu, A. Boynuegri and M. Uzunoglu, "A demand side management strategy based on forecasting of residential renewable sources: A smart home system in

Turkey," Energy and Buildings, vol. 80, p. 309–320, 2014.

[192] Z. Iqbal, N. Javaid, M. Riaz Khan, F. Ali Khan, Z. Ali Khan and U. Qasim, "A

Smart Home Energy Management Strategy based on Demand Side Management," Crans-

Montana, Switzerland, 2016.

[193] K. Basu, L. Hawarah, N. Arghira, H. Joumaa and S. Ploix, "prediction system for home appliance usage," Energy and Buildings, vol. 67, p. 668–679, 2013.

135

[194] N. Arghira, L. Hawarah, S. Ploix and M. Jacomino, "Prediction of appliances energy use in smart homes," Energy, vol. 48, pp. 128-134, 2012.

[195] F. Fernandes, H. Morais, P. Faria, Z. Vale and C. Ramos, "SCADA house intelligent management for energy efficiency analysis in domestic consumers," Sao Paulo,

Brazil, 2013.

[196] F. Fernandes, H. Morais, P. Faria, Z. Vale and C. Ramos, "Dynamic load management in a smart home to participate in demand response events," Energy and

Buildings, vol. 82, p. 592–606, 2014.

[197] F. Fernandes, L. Gomes, H. Morais, M. Silva, Z. Vale and J. Corchado, "Dynamic

Energy Management Method with Demand Response Interaction Applied in an Office

Building," 2016.

[198] M. Pedrasa, T. Spooner and I. MacGill, "Coordinated Scheduling of Residential

Distributed Energy Resources to Optimize Smart Home Energy Services," IEEE

TRANSACTIONS ON SMART GRID, vol. 1, no. 2, pp. 134-143, 2010.

[199] J. Honold, C. Kandler, P. Wimmer, B. Schropp, R. Reichle, M. Gröne, M.

Bünemann, J. Klein and M. Kufner, "DISTRIBUTED INTEGRATED ENERGY

MANAGEMENT SYSTEMS IN RESIDENTIAL BUILDINGS," Applied Thermal

Engineering, 2016.

136

[200] M.-H. Le and S. Ploix, "Robust anticipative energy management system:

Application of a smart building platform," Building Services Engineering Research &

Technology, pp. 1-23, 2016.

[201] D. Zhang, S. Liu and L. G. Papageorgiou, "Energy Management of Smart Homes with Microgrid," Advances in Energy Systems Engineering, pp. 507-533, 2016.

[202] P. Shaikh, N. Mohd Nor, P. Nallagownden and I. Elamvazuthi, "Intelligent

Optimized Control System for Energy and Comfort Management in Efficient and

Sustainable Buildings," ”, Procedia Technology, vol. 11, pp. 99-106, 2013.

[203] R. Missaoui, H. Joumaa, S. Ploix and S. Bacha, "Managing energy Smart Homes according to energy prices: Analysisof a Building Energy Management System," Energy and Buildings, vol. 71, pp. 155-167, 2014.

[204] A. Anvari-Moghaddam, H. Monsef and A. Rahimi-Kian, "Optimal Smart Home

Energy Management Considering Energy Saving and a Comfortable Lifestyle," IEEE

TRANSACTIONS ON SMART GRID, vol. 6, no. 1, pp. 324-332, 2015.

[205] M. Aftab, Towards Automated Demand-Side Management for Commercial

Enterprises, Masdar: A Thesis Presented to the Masdar Institute of Science and Technology in Partial Fulfillment of the Requirements for the Degree of Master of Science in

Engineering Systems and Management, 2012.

[206] S. Reka and V. Ramesh, "Demand Side Response Modeling with Controller Design

Using Aggregate Air Conditioning Loads and Particle Swarm Optimization," Advanced 137

Computing and Communication Technologies, Advances in Intelligent Systems and

Computing, vol. 452, pp. 573-581, 2016.

[207] K. U. Rao, P. K, M. Moiz, S. P. Hemagiri and M. B. T, "An Efficient Power

Management Scheme for Making Indian Grids Smart," International Journal of Electrical

Electronics & Computer Science Engineering, pp. 47-50, 2016.

[208] D. Zhang, S. Liu and L. Papageorgiou, "Fair cost distribution among smart homes with microgrid," Energy Conversion and Management, vol. 80, p. 498–508, 2014.

[209] Y. Ghadi, M. Rasul and M. Khan, "Potential of Saving Energy Using Advanced

Fuzzy Logic Controllers in Smart Buildings in Subtropical Climates in Australia," Energy

Procedia, vol. 61, pp. 290-293, 2014.

[210] M. Milenkovic and O. Amft, "Recognizing energy-related activities using sensors commonly installed in office buildings," Procedia Computer Science, vol. 19, p. 669 – 677,

2013.

[211] P. Petromil, K. Felix, F. Marcus and K. Helmut, "Motivating Domestic Energy

Conservation through Comparative, Community-Based Feedback in Mobile and Social

Media," Brisbane, 2011.

[212] V. Sundramoorthy, Q. Liu, G. Cooper, N. Linge and J. Cooper, "DEHEMS: A user- driven domestic energy monitoring system," 2010.

138

[213] J. Rodgers and L. Bartram, "ALIS: An interactive ecosystem for sustainable living,"

Copenhagen, Denmark, 2010.

[214] C. S. Okinda, I. NYALALA, Z. Chi and S. Mingxia, "Intelligent Energy Saving

System," International Journal of Emerging Technology and Advanced Engineering, vol.

5, no. 12, pp. 173-179, 2015.

139

Table 4-5. Evaluation of Different Systems in Energy Smart Homes that are Proposed in Literature or Available in Market

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category [1] Location: In-home - - Household Total About 7% A case study on the effect of

USA display electricity energy less informing households about their

(IHD) consumption consumptio energy total energy consumption and

n and cost consumpti comparative data about household

on. in the neighborhood using in-

home displays (IHDs) for a period

of 3 months that included 431

single-family households.

[5] Location: in-home - - Real-time electricity consumption Reduced A case study to review the

Scotland, UK display gas and changes in energy consumption

(IHD) electricity habits for users with in home

by 20% energy displays over a period of

and 7%, six months.

respective

Energy Monitoring System

140

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category [211 Title: Sensor EnergyWiz mobile Energy –related Real-time Energy monitoring system with

] EnergyWiz clamp in the application, RESTful data energy comparative energy-related

household’s API (application consumptio feedbacks through a mobile app

power box. program interface), n, CO2, and called EnergyWiz.

Java application, and cost.

MySQL

[212 Title: Electricity - - Temperature and Energy- Save 8% Experimental study on energy

] DEHEMS meter sensor energy related data in energy monitoring systems and

(clamp), and consumption data. such as consumpti environmental data such as

electricity total and on temperature for both whole-house

meter appliances- and detailed energy consumption

level using sensors and energy meters.

energy The data are presented through a

consumptio dashboard in PC and it was noted

n that users tend to change their

energy consumption behavior

141

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category when these data are provided to

them.

[213 Title: Aware - - Ruby on - Detailed - Application of energy

] Living Rails, energy consumption awareness systems

Interface Python, and consumptio in two buildings to inform users

System MySQL. n and also about detailed energy-related data

(ALIS) controlling and enabling the users to control

systems. lights or shades using PC, mobile

app, or house dashboard.

[171 Title: - - RESTful Energy Gather - Study of an Internet-based,

] WattDepot API consumption energy- service-oriented framework for

Location: (application obtained from related processing the energy-related data

USA program energy meters data, for use in smart grids.

interface) analyze,

and

Energy Monitoring System

142

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category visualized

the outputs.

[13] Title: The LinkSmart - UbiLense Fine-grained User can - Interconnecting electric devices in

Energy Aware middleware mobile appliances energy access the a house by using wireless power

Smart Home and Plogg application consumption detailed metering plugs to obtain the

Location: wireless energy energy consumption of different

Germany plugs consumptio appliances and provide the user

n through with detailed energy consumption

mobile app using a mobile app and also

provide the opportunity for

controlling the appliances.

[165 Location: Home - - Electricity Real-time Different A case study to understand the

] Netherland Energy consumption energy families effect of interventions with

Monitor consumptio reacted energy monitoring on energy

n differently saving. Houses equipped with

to energy

Energy Monitoring System

143

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category monitorin sensors, sending unit, and a

g data. On display are studied for 15 months.

average

they

saved

about 8%

[166 Title: ECOIS Energy - - Whole-house and end-use 9% Study of the effect of awareness

] Location: Monitoring electric-power consumption at reduction of residents about energy

Osaka system interval of 30 minutes in power consumption on energy saving

University through PC consumpti using on-line Energy

on Consumption Information System

(ECOIS).

[161 - Home - - Real-time electricity consumption About Study of the effectiveness of

] Energy and CO2 emissions. 15% using energy consumption

Display indicator (ECI) in 44 households

for a period of 1 year and it was

144

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category compared with paper-based

energy consumption information.

[176 Location: Raspberry - Python, - Switching - A Raspberry Pi is programmed by

] India Pi, IC- PHP, SQL the Python connects to the appliances

L293D appliances to switch them on/off and PHP

on/off. script is used to create a web

portal to enable the user to control

the appliances.

[200 Location: - Mixed - Available power Scheduling - A case study project conducted in

] France Integer and price of the of the PREDIS, which is a living

Linear electrical resource appliances laboratory in France to study

Progra at time optimizing the energy

mming consumption based on energy

(MILP) cost and residents’ comfort to

with Control Capabilities and improve the interconnectivity

Equival between smart grid and smart

Systems

145

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category ent RC home. Equivalent RC circuit is

circuit used to perform the thermal

modeling.

[201 - - Lexicog - Heat demand, Minimize About 19 The appliances scheduling is

] raphic equipment daily total % cost adjusted based on the

minima capacities and cost, saving optimization process with three

x efficiencies, distribute objectives including cost

approac maintenance cost, optimal minimization, fair cost

h in heat-to-power ratio cost, distribution, and cost versus CO2

Mixed of CHP optimal emissions using Lexicographic

Integer generator, charge solution minimax approach in Mixed

Linear and discharge limit considering Integer Linear Programming

Progra rates for both (MILP). The controllers will

mming thermal/electrical economic affect the resources in DER

(MILP) storage, gas and system.

146

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category price, real-time environmen

electricity prices tal aspects.

from grid, task

duration, CO2

emission intensity

[182 Location: AT89s52Mi - Embedded C - Switching - Android application that works

] India crocontroller or Assembly the with Bluetooth to control the

Language appliances appliances remotely.

on/off.

[172 Location: Dynami PI system Temperatures, Real-time - A case study for using PI system

] Center of c Time developed relative humidity, environmen software to connect building

Moravian- Warpin by OSIsoft CO2 tal data automation technology (BAT)

Silesian Wood g concentrations, monitoring and building heating technology

Cluste, Czech (DTW) pressures, heat such as (BHT) to perform real-time

Republic consumptions, temperature monitoring of temperature and

and also

Systems with Control Capabilities

147

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category electricity controlling controlling the HVAC system

consumptions. the HVAC remotely.

system

remotely.

[175 Location: Arduino uno - - Remotely switching on/off the - A prototype of an automation

] India kit appliances connected to the board. system using an android system

and Arduino kit to control

appliances remotely through a

mobile device.

[174 Location: Arduino - Thingspeak Energy meters Fine- - Smart sockets that transmit

] Nigeria Uno R3 © service within the socket grained energy consumption data and

energy enable the user to switch on/off

consumptio the appliances connected to it

n. Control through a graphical user interface

signals to (GUI). Also, the acquired data

control the through the smart socket are kept

Systems with Control Capabilities

148

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category appliances in a database as costumer energy

activities. usage.

[181 Location: Bluetooth - - Read current Appliances - A small scale prototype to study

] India low energy energy will be the application of Bluetooth low

(BLE) and consumption switched energy (BLE) system for

Adruino on/off by monitoring and controlling

boards the remote. appliances in home automation.

[204 Location: Iran - GAMS and Schedule of the Minimized - Application of multi-objective

] Cplex/Dicopt solvers home appliances energy function to optimize the energy

for optimizing the consumptio consumption and user’s comfort

multi-objective function n load in smart homes connected to

pattern with smart microgrid.

maximized

comfort

level

149

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category [214 Location: Human zero - Human presence, Adjustment Saved A voltage optimization system by

] China Presence crossing lighting level, and of lighting about auto-phase angle control for

Detector detectio temperature system and 30% in lighting by zero crossing

Sensor, n (ZCD) fan. energy detection (ZCD). Three sensors

Light consumpti are used including Human

Dependent on Presence Detector Sensor, Light

Resistor Dependent Resistor (LDR), and a

(LDR), and temperature sensor. A

a microcontroller used for the

temperature control unit to control the

sensor actuators such as lamps, fans, and

optoisolator and the system is

tested in a library.

[178 Location: - - - Residents’ posture Switching - of the system controls the lights

] Russia on/off the by using residents’ posture such

lights as lying, sitting, and standing.

Systems with Control Capabilities

150

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category [180 Location: Smart power strips consist of 8-bit - Switch - A five-day experimental study of

] Korea low-power MCU, energy metering IC, on/off a smart energy management

ZigBee communication module, appliance system (SEMS) by equipping a

power relay and LCD s power strips with motion sensors,

connecte time limit, and power limit

d to the controlling with microcontroller

power unit controlling them using

strip ZigBee.

based on

motion

sensor.

[9] Location: A wireless sensor consist of a Real-time electricity consumption - An energy management system

Germany Microcontroller Unit (MCU) with an based on wireless sensor

Analog-to-Digital Converter (ADC), a networks to sense and transmit

sensing unit, and a ZigBee transceiver. electricity-related data to monitor

and control the appliances.

151

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category [6] Location: CC2430 Kruskal' Sensor Data acquired from Automatic - Experimental study of a smart

Korea System-on- s Network motion sensors. remote home equipped with sensing

Chip, 8051 algorith Analyzer control of device control, pricing, and

MCU, m (SNA) lights. demand response capabilities to

passive remotely control the lights, AC,

infrared and security system.

(PIR)

sensors

[179 Location: - - RESTful Energy Switching - Experimental study on use of

] Cudrefin02 in API consumption in devices smart outlets to provide fine-

Swiss (application devices/appliances remotely grained energy consumption and

program level on/off a web server for data collection

interface) through a and controlling the appliances

mobile through a mobile interface for a

interface 12-month period.

Systems with Control Capabilities

152

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category [11] Title: The Multiple smart Open Services Residents presences Control - Experimental study of a

Gator Tech appliances and Gateway or temperature of programmable smart environment

Smart House components such as Initiative detected by motion actuators that is also adaptable to the new

Location: microwave oven, (OSGi) sensors and other in technologies that might be

USA and displays, smart framework types of sensing HVAC hooked to the system later in its

plugs, smart floors, equipment. system life span and it is designed for the

and HVAC system. and other elderly and disabled with remote

Also electrically smart monitoring capabilities. It is

erasable appliance equipped with sensors and reacts

programmable read- s. based on the acquired data. The

only memory physical layer of this system also

(EEPROM), 8-bit includes actuators such as AC or

and Atmel ATmega heating thermostats and security

128 processor as the system.

processing unit

153

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category [206 Location: - Particle MATLAB/S The desirable range Minimized - Numerical study to investigate the

] Singapore swarm IMULINK of temperature and peak load use of particle swarm

optimiz energy demand. optimization method developed

ation consumption. by MATLAB to model a demand

techniq response system to be used for

ing Capabilities ue controlling the load demand by

adjusting different units in the air

conditioner (AC) such as

compressor, condenser, expansion

device, and evaporator.

Systems with Data Process [189 Title: Matrix Convol - Residents’ Controlling - SmartSEAL interconnection

] SmartSEAL Vision utional presence and the boiler system aims to facilitate the

Location: Italy BlueFox3 Neural behavior. with energy interconnectivity of various

Camera, Networ consumptio sensors and actuators from

YOLO k n different vendors. The system is

sensor (CNN) monitoring tested in two buildings in Italy.

Systems with Data Processing Capabilities

154

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category object system, There is also a habit learning

detector automated module suggested in this research

gate that uses Convolutional Neural

actuators, Network (CNN) to recognize

smart light, residents’ behavior and objects in

and heating the building using a smart

plant. camera. The actuators control the

boiler with energy consumption

monitoring system, automated

gate actuators, smart light, and

heating plant to manage the

energy consumption.

[186 Country: ZigBee and - Matlab GUI Users’ preferences, The AC About The study compares the regular

] Canada Arduino schedules, and temperature 27% and Programmable Communicating

presence, DR and fan 38% Thermostats (PCTs) and

speed. compared Programmable Thermostats (PTs)

155

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category programs, and to regular with more advanced PCTs with

energy tariffs PCTs and adaptive learning algorithm

PTs, capabilities. Matlab GUI is used

respective to perform the energy modeling

ly. and an Arduino board with

ZigBee is installed on a regular

AC system that is programmed

using the adaptive algorithm.

[152 Title: Multi- - Probabil - - Controlling Reduce Developed a multi-agent

] agent system ity the washing peak mathematical model for the

function machine, demand demand response system to

s to clothes and control the appliances and

model dryer, and generation activate them during the off-peak

resident dish cost by period.

ial load, washer. about

dynami 24% and

Systems with Data Processing Capabilities

156

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category c price 31%,

load, respective

and ly.

energy

manage

ment

system.

[192 - - Electrocardiogram The mathematical Minimized - A numerical study to investigate

] (ECG) optimization formulation of energy and application of Electrocardiogram

method and binary appliances’ energy the cost of (ECG) optimization method for

multiple knapsack consumption, cost generation minimizing the energy

problem (MNKP) models, and energy and consumption and energy

generation model. consumptio consumption & generation cost

n. based on a demand side

management (DSM).

157

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category [209 Location: - Fuzzy - Real-time dynamic events such as - Numerical study to evaluate the

] Australia logic number of residents are also use of fuzzy logic control systems

integrated in the decision making to be able to consider the real-

process related to the energy time events such as number of

management systems. residents, indoor air quality

(IAQ), and natural light.

[191 Location: - Artificia - Real-measured Wind and 3% Numerical study on application of

] Turkey l Neural load profiles solar power reduction artificial neural network-based

Networ predictions in cost of prediction system for power

k electricity resources such as wind and solar

power in a demand side

management (DSM) system.

[203 Location: - Nelder- MATLAB/S The load demand for appliances Reduced A numerical case study for

] France Mead imulink such as heating system, washing electricity evaluation of Global Model

geometr machine, and dish washer is bill about Based Anticipative Building

ical optimized. 20%. Energy Management System

Systems with Data Processing Capabilities

158

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category optimiz (GMBA-BEMS) using Nelder-

ation Mead optimization method to

algorith optimize both energy

m consumption and comfort level of

users.

[190 Location: - Artificia - Outside and room Required - Apply neural network as a

] Germany l Neural temperature, wind time for numerical study to predict

Networ speed, and heating and residents’ presence and behavior

k residents presence. target to optimize heating system by

temperature adjusting the room temperature.

[208 - lexicographi - Heat demand of Plan for 24-30% Computational study of

] c minimax each house, energy save in application of lexicographic

method - capacity, production, cost minimax method as a

mixed efficiency, and equipment mathematical programming

integer maintenance cost output formulation using a mixed integer

Systems with Data Processing Capabilities

159

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category linear of equipment, heat- sharing, linear programming (MILP)

programmin to-power ratio of thermal/ele approach to optimize energy

g (MILP) CHP generator, and ctrical, and distribution and fair cost among

gas price. electricity multiple houses equipped with

bought smart grid that are using a shared

from and microgrid.

sold to the

grid.

[195 Location: - An - Desired range of Adjust the - SCADA House Intelligent

] Portugal optimiz energy room and Management (SHIM) system that

ation consumption and water includes set of optimization

method comfort temperature algorithms to optimize the power

and room consumption, the micro

luminance generation system, the charge and

based on discharge of the electric or plug-

in hybrid vehicles, and the

160

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category comfort participation in Demand

level Response (DR) programs to

prioritize the active time of

devices.

[210 - passive - - Presence of the Control of - Experimental study on application

] infrared peoples in office computer of Layered hidden Markov

(PIR) screens and models (LHMM) in an

sensors and dimming instrumented office building to

power plug lights recognize activities and control

meters the lighting system and computer

screens.

[202 - - Multi- - The difference in Balanced - Uses Multi-Objective Generic

] objectiv the sensed value environmen Algorithm (MOGA) to optimize

e and desired t the energy consumption and

Genetic temperature. Also, temperature comfort level of users. It balances

Algorith and power the total power consumption and

Systems with Data Processing Capabilities

161

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category m the power consumptio customer comfort based on a

(MOGA consumption. n. range of comfortable temperature

) to to improve the building

generate automation intelligence.

Pareto

front

[205 Location: - general MATLAB/S Electricity price Controlling Potential Numerical study for evaluation of

] UAE algebrai imulink and and weather file for the for 12% using general algebraic modeling

c EnergyPlus simulation. demands reduction system (GAMS) for optimizing

modelin side for in average the energy consumption as part of

g HVAC, daily the demand-side energy

system lighting, energy management program. The

(GAMS and consumpti control system is focused on

) refrigeratio on HVAC, lighting and refrigeration.

162

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category [194 Location: - Stochast - - The - Use stochastic methods to predict

] France ic-based probability residents’ behavior and energy

predicto of the consumption of appliances to be

r appliance used in smart grids and

to consume segmentation and aggregation of

on an data are also used to improve the

hourly prediction quality.

base.

[12] ThinkHom - SPARQ MATLAB/S The energy usage Comfort - A simulation study on application

L imulink and by smart temperature of extensive knowledge based

languag the metering and user profile and system for smart homes in order

e, HAMLab behavior by means occupancy to reduce the energy consumption

Extensi tools of the information profile to and keep the comfort level of

ble collected from be used for residents. Extensive building

Styleshe sensors. optimizatio information is obtained from BIM

et n. files such as gbXML to be used

Systems with Data Processing Capabilities

163

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category Langua for other evaluations and decision

ge making. Artificial intelligence

Transfo (AI) is integrated with Multi

rmation agent system (MAS) concept in

(XSLT), order to realize optimized control

strategies.

[198 Australia - - MATLAB DER operation schedules that - Application of particle swarm

] maximize the net benefits. optimization (PSO) to optimize

the acquisition of different energy

services in a distributed energy

resources (DER) system.

[187 Title: ZigBee-based - Environmental data Energy- - Suggests a framework of

] PERSON wireless sensor such as temperature related Pervasive Service-Oriented

network, smart and humidity and data and Networks (PERSON) as an

power outlet, energy consumption control energy management system

sensors for data. signals (EMS) and its application for a

164

Project/Prod Mathe Software/ Output uct/ System Input/measured Energy/C Technology matical Programmi data/action Description and findings

Title and data type ost saving

Reference Method ng language type Location

Category environmental data, such as demand response system is

and on/off switches. on/off studied.

controls.

[169 Title: Energy meter CVX toolbox - Data obtained from Estimate - A prototype installation of

] ViridiScope monitor, light MATLAB sensors d energy sensors to perform a non-intrusive

Location: intensity sensors, consump measurement of the energy

USA magnetometers, and tion by consumption of appliances and

microphones appliance managed to estimate the energy

s consumption with 10% error.

Systems with Data Processing Capabilities

165

5) Chapter 5. Energy Retrofit of Buildings

This chapter has been written as a journal paper and is already submitted for review.

166

Residential and Commercial Building Envelope Energy

Retrofit: Innovative Measures and Example Projects

1 2 Ehsan Kamel , Ali M. Memari

1Ph.D. Candidate, Department of Civil Engineering, Structural Engineering, Penn State University, 321 Sackett Building, University Park, PA 16802 (corresponding author). E- mail: [email protected] 2Professor, Hankin Chair in Residential Building Construction, and Director, Pennsylvania Housing Research Center, Dept. of Architectural Engineering and Dept. of Civil and Environmental Engineering, Penn State Univ., University Park, PA 16802. E- mail: [email protected]

5.1 Abstract

The energy consumed in buildings can be reduced through using more energy efficient appliances as well as reducing the energy loss through building envelope. Energy retrofit of buildings considering the three major nonstructural component categories of building envelope, mechanical systems, and electrical systems can contribute to these objectives.

This paper is mainly focused on the first category and discusses different methods of building envelope energy retrofit that could be carried out on opaque and transparent components of existing buildings such as improving the insulation of wall systems or enhancing thermal performance of window systems. Representative projects that have used these retrofit measures are also reviewed in this paper. The literature materials and systems considered for review are categorized into two main groups of conventional and innovative

167 measures. The paper also compares the effectiveness of some conventional energy retrofit methods with the application of some innovative materials such as aerogel and Phase

Change Materials (PCM) by developing a computer model in BEopt for a typical residential building in U.S. that serves as the benchmark house. Results show that regardless of the retrofit methods, improving the thermal performance of opaque components such as wall and roof systems is more beneficial compared to floors, roofing material, and glazing system in a typical residential building and it can save up to about

24% and 31% in annual heating and cooling loads for cold (5A) and hot-dry (3B) climate, respectively.

Keywords: Building Envelope, Energy Retrofit, Residential Buildings, BEopt

5.2 Introduction

Growing world energy use has raised concerns about supply, exhaustion of resources, and environmental impacts. In 2013, the world total energy consumption was about 558.6 quadrillion British thermal unit (Btu) and the U.S. total primary energy use was about 97 quadrillion Btu, which is about 17% of world total annual energy consumption [215].

Residential and commercial sectors approximately contribute to 21% (20370 Trillion Btu) and 19% (18430 Trillion Btu) of total annual energy consumption, respectively. Based on the most recent report published by Residential Energy Consumption Survey (RECS) based on the data until 2009 regarding the energy consumption in the U.S., space conditioning, domestic hot water, and refrigerators have consumed, respectively, about

48%, 17%, and 5% in residential sector. The remaining 30% is consumed by appliances, 168 electronics, and lighting [216]. Energy consumption due to space conditioning could be directly affected by building envelope properties and climate zone. In 2009, the energy consumption of residential buildings due to space conditioning was approximately 38% and 55% for Marine and Very Cold climate zones, respectively [216]. This number is about

50% of the total energy consumption in buildings in developed countries [217]. Residential buildings constructed before 1940 use about 54% of the energy for space conditioning, while consumption is about 43% for houses built from 2000 to 2009 [216]. Reducing even a small amount of this energy can lead to desirable environmental effects and financial benefits, which shows how important the energy retrofit of buildings could be. Identifying conventional and new energy retrofit methods in building energy retrofit and studying the example projects and research studies in literature to evaluate the effectiveness of these measures could contribute to this goal. Therefore, the references addressing these areas are targeted and all the site energy savings available in these references due to different retrofit measures are reported.

Measures contributing to building envelope energy retrofit can lead to reduction of energy consumption in residential and commercial buildings. This paper focuses on reviewing different energy retrofit materials, techniques, and systems that are commonly used, new to the market, or innovative retrofit methods still under development and not available in the market. Research studies on conventional insulation materials such as fiberglass, rockwool, cellulose, expanded polystyrene, polyurethane, and polyisocynurate in different forms such as rigid boards, blankets, loose-fill, foamed-in-place, and sprayed-in are reviewed [15, 218, 16, 219, 21, 220, 221, 222, 223, 34]. These materials are typically

169 applied in different forms on opaque components. In addition, the concept of using color coatings and materials with higher/lower heat absorptivity is also used in energy retrofit such as the concept of “cool roof”, which works based on lower heat absorption and could be applied on the roof system. Transparent components such as window systems are also among the components that could be retrofitted. There are measures for their energy retrofit such as adding new coatings with different Solar Heat Gain Coefficient (SHGC) or emissivity to the existing window system and installing overhangs or shadings to the existing window systems. As it was noted, these retrofit measures could be applied to different building envelope components such as above and below grade walls; roof or attics; fenestrations such as windows and doors, floors, and façade.

In addition, one of the important contributions of this paper is reviewing more innovative retrofit measures and materials. It includes Nanogel, Phase Change Material (PCM) capsules mixed with conventional construction materials, combination of conventional and innovative materials such as prefabricated elements, dynamic façade systems, building integrated photovoltaic systems, active thermal solar facades, vacuum insulated panels, and multi-functional energy efficient façade system. These innovative systems and techniques could be also applied on walls, windows, roof, and slabs. However, retrofitting existing homes façade systems is shown to be of great interest [25, 224, 225, 226, 227, 228, 229,

30, 230, 231]. Of course, as discussed subsequently, innovative materials such as aerogel or PCM may not be able to compete with conventional materials in terms of cost; however, they can improve the thermal performance of buildings and could be beneficial options, especially when there are limitations in terms of the thickness of insulation material.

170

Some representative building retrofit projects are also reviewed in this paper to see how the energy retrofit measures would be applied and perform in real-world projects. Retrofit measures used in these projects vary depending on the climate zone, initial condition of buildings, and the budget considered for the project. Most of the methods used in these projects are among conventional measures studied in this paper such as installing window films, rigid insulation on the exterior wall, batt insulation in cavities, triple pane glazing systems, cool roof, and improving air tightness [232, 40, 233, 234, 235, 236, 237, 238,

239]. There are also retrofit example projects, which adapted innovative measures such as

Nanogel products, vacuum panels, and dynamic façade systems that could be good examples to show feasibility in application of such systems [240, 241, 235, 242, 243].

Multiple criteria could be considered in choosing suitable retrofit material, system, or technique including type of building, climate zone, financial restrictions, and Life Cycle

Assessment (LCA) results. A short review of proper tools for considering these criteria is presented to have an overview on computer modeling and optimization methods that could contribute to energy retrofit of buildings. Current energy simulation tools being used in energy retrofit of buildings could be great assets, when it comes to considering multiple criteria in choosing suitable energy retrofit scenarios [31, 244, 245, 246, 247, 33].

Optimization methods such as multiple criteria complex proportional assessment

(COPRAS) method could also be implemented in simulation tools, when there are multiple retrofit scenarios in order to reach the best and optimized option [35, 36, 37, 38, 39].

171

Discussed is also development of a computer model in BEopt to help study the effects of multiple conventional and innovative envelope energy retrofit measures in comparison with a benchmark house. The benchmark house is designed based on common characteristics of residential buildings built from 1990 to 2000 in the U.S., representing under insulated homes. Different energy retrofit methods are considered, including adding common insulation materials to exterior walls, slabs, or roofs; using cool roof, adding window film, application of PCM coating, and utilizing aerogel blankets. Energy analysis results show that the methods reviewed can potentially contribute up to about 24% and

31% reduction in annual heating and cooling loads in the benchmark house for cold (5A) and hot-dry (3B) climate, respectively. However, some of these methods such as using aerogel or PCM might not be economical. It should be noted that this study was not intended to find the best retrofit scenario; rather it was meant to be an attempt to compare and evaluate the effectiveness of retrofit methods, in particular the more innovative systems.

5.3 Materials and Systems in Envelope Energy Retrofit

Energy retrofit of buildings may be addressed under different categories including mechanical, electrical, and envelope systems. This paper, however, is focused on building envelope energy retrofit that can improve thermal and energy performance of building envelope components including above and below grade walls, roof, fenestrations, and façade [248, 249]. The choice of building envelope components for energy retrofit may depend on factors such as building location, existing envelope systems, climate zone, and

172 financial restrictions of the project. For example, energy retrofit of below grade walls in colder regions can be of great importance since the basement can account for up to 25% of the total energy consumption in a house [250]. Besides these opaque or transparent components, the envelope energy retrofit could also aim for improving the air tightness of the building that has a huge impact on energy performance.

Building envelope energy retrofit may be influenced by three sources of heat gain/loss including conduction, convection, and radiation. Most retrofit methods applicable to transparent or translucent components, which are developed to deal with radiation, work by either increasing or decreasing the heat absorbed for example by applying window films over the existing glass. To improve airtightness, however, the amount of heat transfer through convection should be limited. Although convection and air infiltration can have a huge impact on energy loss, most existing methods of envelope energy retrofit tend to limit energy loss through conduction. Application of insulation materials with low thermal conductivity is the basis for such retrofit methods, although different insulation material might also benefit from limiting multiple heat transfer forms simultaneously. Moreover, retrofit measures might benefit from storing heat energy by using materials with high thermal inertia [249].

Building energy retrofit could target a limited number of components in mechanical systems, lighting, or envelope; however, for what is referred to as “deep energy retrofit” that leads to higher levels of energy saving, the retrofit design needs to target multiple areas and suggest significant changes. National Renewable Energy Laboratory (NREL) has

173 developed a series of guides for energy retrofit of different types of buildings such as healthcare facilities, schools, office, and retail buildings, which has allowed various types of energy retrofit methods to be applied not only to building envelope systems, but also to lighting, HVAC, and service hot water (SHW) [223, 24, 222, 34]. Among the most common retrofit methods used by some energy saving companies (ESCO) for healthcare facilities, building envelope retrofit has a rather small share (about 10%), presumably due to long payback period compared to other retrofit options such as HVAC systems [223].

Review of the retrofit projects in DOE guides shows that building envelope retrofit options that can be categorized under the conventional systems and materials are mainly focused on openings such as windows and doors; insulation of walls, roof and slabs; as well as measures such as installation of cool roof (in hot climate regions); reducing air leakage using air sealing materials; and high R-value roll-up receiving doors. [223, 24, 222, 34].

The retrofit material and systems reviewed in this paper are categorized into two groups of conventional and innovative technologies. Systems reviewed under either of these categories, could be applied to aforementioned building envelope components. Different materials and systems reviewed in this paper are summarized in Table 5-1.

Table 5-1. Summary of the energy retrofit methods used in the building envelope in research studies

Objective,

building type, Energy Simulation Reference Proposed envelope retrofit measures Energy Saving or location of Tool

the study

Guattari et Historical Replacing single glass with About 20-28% and TRNSYS al. [21] Building different options including 6.6-26% decrease in

174

Objective,

building type, Energy Simulation Reference Proposed envelope retrofit measures Energy Saving or location of Tool

the study

double glazing filled with air or annual heating and

argon and triple glazing filled cooling loads,

with krypton respectively.

Application of insulation layer Pertosa et Historical Decrease in energy made of reed and innovative tile EnergyPlus al. [251] building consumption system

Decrease in ice dam Ojczyk Exterior Thermal & Moisture - - formation and [221] Management System (ETMMS) energy consumption

A residential Reduction in Pisello et building in Application of cool-green roof - overheating hours al. [252] Italy by 98%

Application of innovative tile Boarin et Historical system with higher reflectance - - al. [253] building rate

Pacific Up to 32% Adding exterior insulating finish Northwest reduction in annual system (EIFS), replacing National School EnergyPlus energy consumption windows, rigid insulation on roof, Laboratory for whole-house and slab insulation. [24] retrofit packages.

Pacific Decrease in energy Office Exterior window film, exterior Northwest EnergyPlus consumption. It is Building window shading, add wall National concluded that up to

175

Objective,

building type, Energy Simulation Reference Proposed envelope retrofit measures Energy Saving or location of Tool

the study

Laboratory insulation, roof insulation, and 25% reduction in

[34] cool roof annual energy

consumption is

feasible and up to

50% is also

achievable after a

deep retrofit.

Application of different products

containing nano material such as Materials and aerogel. These products include systems Casini et aerogel underfloor mats, panels, containing - - al. [225] and pre-coupled gypsum boards PCM and with aerogel. aerogel Application of Micronal PCM to

plaster, Trombe walls, and etc.

Ascione et Application of PCM as a plaster Reduction in annual al., 2014 over the ETICS, replacing EnergyPlus energy consumption

[25] windows, new roof insulation by 38%

From 1% to 36% Marin et Transportable Application of PCM gypsum EnergyPlus reduction in annual al. [26] buildings board energy demand

Lee et al. Residential Application of pouches Depending on the - [254] building – containing hydrated salt-based location of the wall,

176

Objective,

building type, Energy Simulation Reference Proposed envelope retrofit measures Energy Saving or location of Tool

the study

experimental PCM covered by aluminum foil, the heat flux was

study referred to as PCM thermal shield reduced between

(PCMTS) on the exterior walls. 3.6% and 51.3%.

Up to 80% decrease

Different thickness of aerogel and in annual heating

Application percentage of windows coverage and cooling load is

Berardi of aerogel in is studied to evaluate a THERM and observed after

[226] glazing transparent insulation such as EnergyPlus covering 100% of

system aerogel integrated with glazing windows with 5cm

systems. double pane

aerogel.

About 50% increase Air and argon-filled panels One- in R-value in winter Application installed on top of conventional Baetens et dimensional condition comparted of gas-filled fiberglass batts between beams of al. [255] analytical to the system with panels a roof system in a residential model only fiberglass as an house. insulation

Application of Nanogel®

Aerogel insulation plaster, 71% reduction in Filate et Office ThermablokSP board, and DesignBuilder energy loss through al. [28] building SLENTIT aerogel insulation walls

board.

177

Objective,

building type, Energy Simulation Reference Proposed envelope retrofit measures Energy Saving or location of Tool

the study

Residential

building – Application of 20 mm thick Cuce and The heat flux is experimental fibre–silica opaque aerogel Statistical Mert Cuce reduced from 5.86 and statistical blanket in energy retrofit of model [256] to 0.66 W/m2.prepar research on a uninsulated separating walls

test bedroom

Kolaitis et Thermal insulation composite TRNSYS Better energy al. [22] panels model performance by 5%

Numerical study to evaluate

different Energy Conservation

Measures (ECM) including Energy saving from Residential Lo Cascio adding roof insulation (13cm of DesignBuilder 2% to 8% building in et al. [29] EPS), low emissivitiy single Model depending on the Australia glaze windows, and reducing air retrofit method.

leakage (from 7 ACH50 to about

4 ACH50).

Single layer XPS with surface Development Masera et finish & composite EPS panel of innovative - - al. [228] coated on both sides with textile systems reinforced mortar

Evola and Application of external thermal Decrease in heat Apartment in DesignBuilder Margani insulation composite system loss through walls Italy model [30] (ETICS) containing stone wool by 85%

178

Objective,

building type, Energy Simulation Reference Proposed envelope retrofit measures Energy Saving or location of Tool

the study

and building integrated PV

(BIPV)

Application of two different

multifunctional energy efficient

façade systems (Meefs) including

Advanced Passive Solar Total heat

Cold climate Collector & Ventilation Unit consumption

Paiho et region of Technological Unit (APSC&VU reduced from 219 EnergyPlus al. [27] Finland and TU) and Advanced Solar kWh/m2 to 79

Russia Protection & Energy Absorption kWh/m2 after using

Technological Unit (ASP&EA Meefs system.

TU) that work based on thermal

storage and phase change

material.

Multiple case

Classification Application of Active Solar studies are reviewed Zhang et of different Thermal Façade (ASTFs) as a - and the energy al. [19] ASTFs building envelope component. saving of each case

is reported.

Application of double-skin façade Radiance (To

Kim et al. in old residential buildings. A analyze the - - [20] performance of a 90cm cavity daylighting

with 20cm slats is evaluated. performance)

179

Objective,

building type, Energy Simulation Reference Proposed envelope retrofit measures Energy Saving or location of Tool

the study

Ascione et Reduction in annual al., 2014 Application of PCM wallboard EnergyPlus energy consumption

[37] by 6%

5.3.1 Conventional retrofit measures

Conventional insulation materials such as fiberglass, polyurethane, rockwool, and cellulose can be used for energy retrofit of a building envelope. The thermal conductivity of these materials range between 0.030 and 0.054 W/m-K [15]. Besides thermal resistance of conventional insulation materials, other properties such as density, fire resistance, service temperature, vapor resistive properties, durability, potential health risks, and their composition in terms of being organic, inorganic, or combined have also been of interest

[15, 218, 16]. Table 5-2 lists different conventional insulation materials and their thermal conductivities. These materials can be produced and used in different forms such as sprayed-in-place, batts, rolls, loose-fill, and rigid board, as summarized in Table 5-3.

Depending on the type and form of the material to be used for retrofit, the construction and installation of the retrofit system will vary. One of the most commonly used envelope energy retrofit methods is adding insulation to the exterior walls, which requires a rigid form of insulation material to be attached to the existing wall using fasteners. In some cases, a gap may be required between the new exterior insulation and the existing envelope,

180 which provides a drainage system or rain screen. It is also convenient to install different types of siding over a new exterior rigid insulation added for retrofit purposes [220].

Providing a drainage surface or ventilation gap is also part of the retrofit design in other systems such as exterior thermal and moisture management system (ETMMS), referred to as “overcoat” system, which allows air and moisture movement. This system further incorporates another membrane installed adjacent to the insulation material to control the air, water, or vapor, and such additions seem to have made the system to be an expensive option [221].

The aforementioned materials in different forms are typically applied to the opaque components such as walls, roof, and floors. In addition to insulation materials, color coatings or materials with higher or lower heat absorptivity could be applied on opaque components of buildings. The concept of reducing or increasing the solar heat gain through building envelope surfaces is used in approaches specific to roofs, one being “cool roof”, which basically reflects the sunlight. This requires the surface to have high reflectance and thermal emittance, which is more useful for warm climates as it helps to absorb less heat

[257]. The difference in temperature before and after applying a light color on the roof could be up to 24oC [232]. The combination of green roof and cool roof concept using the

Helichrysum Italicum plant has shown that it can reflect about 44% of the solar radiation, which is about 4% more than a common concrete roof [252]. Reflective tiles can also be used in order to reduce the solar heat gain. To measure the performance of such materials,

Solar Reflectance Index (SRI) with extreme values of 0 and 100, respectively, for a standard black and white surface can be used. As an example of the use of SRI to evaluate

181 the performance of a certain clay tile, Boarin et al. [253] reports on application of an innovative reflective clay tile with about 15% higher reflectance compared to conventional tiles and with a SRI value of 67 used in GBC Historical Building for roof and external pavement that led to a decrease in energy consumption [253].

Table 5-2. Conventional types of insulation material

General Category Material Thermal conductivity (W/m-K)

Fiberglass 0.030-0.040

Natural Rockwool 0.037-0.40

Cellulose 0.046-0.054

Natural Lightweight Perlite 0.04-0.06

Aggregates Vermiculite 0.063-0.068

Polyethylene 0.041

Polymers Polystyrene (expanded or extruded) 0.030-0.038

Polyurethane and Polyisocynurate 0.023

Aluminized Sheets - Reflective Systems Ceramic Coatings -

Table 5-3. Forms of conventional insulation material

Form

Blankets (batts or rolls)

Loose-fill (blown-in or poured-in)

Rigid board

Sprayed-in-place

Foamed-in-place

Reflective Systems

182

Another area of focus for energy retrofit could be transparent components such as window systems, which in a general term refers to both glazing part (transparent) and framings

(opaque). There are different measures for energy retrofit of windows reviewed in literature such as adding coatings with different Solar Heat Gain Coefficients (SHGC); replacing windows with more efficient window systems such as double or triple pane windows; and adding overhangs or shadings. For example, Setiawan et al. [31] used eQuest to evaluate the effect of using single-layer low-e glass, double-layer low-e glass, horizontal shading, vertical shading, and box shading. It was shown that they can lead, respectively, to about

7%, 8%, 5%, 11%, and 15% reduction in annual energy consumption of a 163 m2 building in Indonesia. Depending on the building type, transparent components such as glazing parts might be more important over other options. For example, commercial buildings with high window to wall ratio could be more focused on improving the thermal properties of the glazing system. Guattari et al. [21] compared the performance of double pane glazing filled with air or argon and triple pane glazing filled with krypton to single glass using TRNSYS and observed about 20-28% and 6.6-26% decrease in annual heating and cooling loads, respectively [21].

5.3.2 New retrofit measures

New retrofit measures could consist of conventional materials and systems being used in an innovative way; innovative materials and technologies; or a combination of conventional and innovative methods and materials used in a new way such as dynamic façade and prefab elements. In order to have a better understanding of new retrofit systems,

183 first, a short review of innovative materials including Phase Change Material (PCM) and

Nanogel will be presented. This short review will be followed by studying the innovative systems, which could be a combination of different conventional and new components and materials. Figure 5-4 shows an overview of innovative measures in energy retrofit studied in this paper and illustrate how different components could combine to provide a new system, which could contribute to the energy performance of buildings.

Phase Change Material (PCM)

PCMs that can be categorized under materials with thermal inertia are activated when the temperature reaches a certain level, typically between 23 and 26o C, whereby the PCM undergoes a phase transition by absorbing heat. Phase transition due to heat absorption could be from solid to solid, solid to gas, solid to liquid, or liquid to gas. The opposite phase transition, i.e., release of heat by the PCM, occurs when the ambient temperature reaches the set point at night.

These materials can be used in walls, floors, and ceilings with operating temperature range of 20 to 35o C [225]. Commonly used PCMs with different latent heat and fusion point include paraffin, non-paraffin, fatty acid, salt hydrates, and eutectics, as presented in Figure

5-1 and Figure 5-2, which has the capability to be used in energy retrofit. Depending on the climate zone and outdoor/indoor temperature, proper PCM material could be used in an energy retrofit project. There is a lower and upper bound for the latent heat and fusion point for these materials. Figure 5-1 compares these materials with water (Material # 6) and it can be seen that the lower and upper bound of fusion temperature ranges between 6-

184

17oC and 75-127 oC, respectively. The same comparison is shown in Figure 5-2 for latent heat and the lower and upper bound are between 70-170 KJ/kg and 220-300 KJ/kg, respectively, compared to water with 333 KJ/kg of latent heat.

Application of PCM to improve the thermal properties of building envelope components can be in different forms such as micro and macro encapsulation. Micro-capsules are less than 1mm in diameter, which makes it easier to be placed within polymer films. Macro- capsules, however, can be larger tubes, spheres, or panels containing PCM. Examples for application of both macro and microcapsules are available in literature [25, 37, 26, 225,

258, 259]. Review of cooling applications of PCM shows it can be used in two ways including the passive and active methods. Active methods rely on ventilation of air moving over the PCM and need other mechanical components; therefore, it will be reviewed in other sections under the combinatory retrofit systems. Passive methods include direct application of PCM such as mixing PCM with conventional construction materials or installing pouches containing PCM over the walls [260]. Figure 5-3 shows PCM pouches, which could be used directly on building envelope surfaces such as walls and micro or macro capsules, which could be mixed with other materials such as plasters and cementitious materials, where it could be used on other surfaces such as walls or ceilings

[254, 259]. As an example, Lee et al. (2015) studied the effect of installing PCM thermal shield (PCMTS), which consist of pouches containing hydrated salt-based PCM covered by aluminum foil on both sides, on wall system. It was noticed that depending on the location of the wall, the heat transfer reduction could be different and it varies between

3.6% and 51.3% [254].

185

The form of the PCM macro-capsules allows them to be mixed with common building products such as plaster, screed, concrete, gypsum, acrylic paints, and wood products such as MDF and OSB. Such products could be used in energy retrofit of buildings. Other examples of such application could be Salt hydrate PCM, which could be placed inside insulating glazing system that allows most of the light through, while enhancing the thermal performance [225]. PCM shows better performance in an environment with higher fluctuations in temperature. Transportable houses are typically subjected to such a condition. Marin et al. [26] studied the effect of PCM integrated with gypsum boards in these buildings. The thermal conductivity, latent heat capacity, and peak melting temperature of the PCM used in this study is 0.23 W/m K, 200 kJ/m2, and 25oC, respectively. Some other researchers also studied the effect of the PCM gypsum board in different climate zones using EnergyPlus and results showed that it can reduce the annual energy demand from 1% to 36%, depending on the climate zone [26].

There are also examples of application of micro-capsules embedded in common construction materials. For example, in a research study on an educational building in Italy, a 3 cm wallboard that included PCM with melting point of 27oC was installed on the inner face of the wall, while plaster containing PCM with melting point of 32oC was applied on the exterior face of the wall. Combination of this method along with replacing deficient windows with low-e glass as well as new roof insulation materials resulted in up to 38% reduction in annual energy consumption [25]. The benefit of applying two layers of PCM plaster with different melting point is that individual layers can be activated in either summer or winter [25]. Similar application is studied by Kusama and Ishidoya (2017) in

186

Japan where microencapsulated PCM with melting point of 25oc is mixed with gypsum plaster applied on the interior walls and ceilings [258].

140 Lower Bound Upper Bound 120 100 80 60 40 20 0 Fusion Temperature (oC) Temperature Fusion 0 1 Salt2 Fatty3 4 5 6 7 Non Eutectics Paraffin Water Hydrates Acids PCM Material type

Figure 5-1. Fusion Temperature for different types of PCM [(data adapted from [225])].

350 Lower Bound Upper Bound 300 250 200 150 100

50 Latent Heat (KJ/kg) Heat Latent 0 0 Water1 Salt2 Paraffin3 Non4 Fatty5 Eutectics6 7 Hydrates Paraffin Acids PCM Material type

Figure 5-2. Latent heat for different types of PCM [(data adapted from [225])].

Figure 5-3. Pouches filled with PCM (left)(reproduced from [254], used with permission), PCM

Macrocapsules (middle), and PCM Microcapsules (right)(reproduced from [259], used with permission)

187

Figure 5-4. Overview of innovative energy retrofit methods

188

Nanogel

Materials on the size scale of between 1 and 100 nanometer (nm) are made through nanotechnology that can be used for insulation applications as well. Aerogel is an example of such a material, which is a “solid nanoporous material with ultra-low density obtained by the dehydration of a gel by replacing liquid component with a gaseous one” and can be obtained from different base material such as silicon, aluminum, chromium, tin, or carbon, but the most used material is silica-based [225]. The thermal conductivity of such a material made of silica at 25°C can be about 0.016-0.03 W/mk, which shows better thermal resistance compared with common insulation materials such as Polyisocyanurate foam

(0.023 W/mk), expanded polystyrene (EPS) (0.037-0.038 W/mk), and extruded polystyrene

(XPS) (0.030-0.032 W/mk). Thermal properties are not the only advantage of aerogel, as their sound insulation and vapor resistance properties can also be higher than conventional insulation materials used in walls mentioned above [225].

Aerogel can be used in different forms for energy retrofit such as rolls, semi-rigid panels, and pre-coupled gypsum boards specially when there are limitations on the thickness of insulation materials that may need to be added for retrofit purposes to existing walls, roof, or floors in a historic or any other type of buildings. An example for limitations on the thickness of the insulation material is a case study project conducted in the U.K. to refurbish the sea containers for residential use and due to the limited interior space, a minimum thickness of the insulation material was much more desirable [224].

189

There are also other examples for application of aerogel in energy retrofit of residential buildings using reinforced flexible aerogel insulation blankets. Polyester, glass, and carbon are more common as fiber reinforcement and the blankets can be installed on the wall, façade, ceiling, framing, and floors [261]. Cuce and Mert Cuce (2016) studied the effect of applying aerogel blankets on uninsulated separating walls and using fluxmeters noticed the heat loss could be reduced from 5.86 to 0.66 W/m2, which is about 88% lower [256].

Application of aerogel in a retrofit case study project of an old building in Germany discussed by Filate [28] is another example, where aerogel (with conductivity of 16 mW/m.k) is used in external walls with total U-Value of 0.26 W/(m2.k) and flooring panels that consist of aerogel boards covered by rigid Magnesium Silicate with total heat transfer coefficient of 0.42 W/(m2.K). Aerogel is also beneficial when a transparent insulation material is desirable and it can be integrated with glazing system. Berardi [226] studied the effect of different thickness and percentage of aerogel covering the window that was integrated with glazing system shown in Figure 5-5, and it was observed that it can lead to about 80% decrease in annual heating and cooling loads. THERM was used to evaluate the

R-value of window system and EnergyPlus was used for evaluating the energy performance of the modeled building [226]. Currently, these materials are still about 8-10 times more expensive compared with more common insulation materials [225].

190

Figure 5-5. Transparent aerogel panel (reproduced from [226], used with permission)

Combined systems

With reference to Figure 5-4, innovative energy retrofit systems could be a combination of multiple materials, technologies, and systems. These measures could be focused on the façade of buildings such as dynamic or double façade or precast elements, which could be used in other building envelope components.

Gas-filled panels (GFPs) could be used as an insulation material in cavities and need to be sealed in order to avoid letting the gas inside it out. It can be filled with different gas types such as air, argon, and xenon, with reported conductivity ranges of 28-40 mW/mK, 20-46 mW/mK, and 55 mW/mK, respectively [255, 262]. There are not many studies on its application for improving building energy performance; however, the existing research shows that the combination of fiberglass insulation and air filled and argon-filled panels could lead to about 50% increase in overall R-value of a roof system during winter condition compared to using only fiberglass [255]. Figure 5-6 shows a cross section of gas- filled panels and how it could be installed inside cavities.

191

Figure 5-6. Cross section of gas-filled panel (left) (reproduced from [263], used with permission) and

installation of gas-filled panel (right) (reproduced from [262], used with permission)

Innovative methods could also consist of conventional materials that are used as precast components that are basically a combination of different layers including an insulation layer made of conventional material or even more innovative materials such as PCM or aerogel. The important benefits of ease and speed of installation have made the precast elements important part of solutions for retrofit construction with the added benefit of allowing the use of building science guidelines and understanding such as implementing moisture control provisions in the wall system before installation. An example of such precast elements in energy retrofit is the single channel glazed photovoltaic thermal module

(SCGPVTM) shown in Figure 5-7 that has a ventilation channel underneath it to make sure it does not heat up, as heat can decrease the efficiency of the photovoltaic (PV) panels

[227].

192

Figure 5-7. Details of single channel glazed photovoltaic thermal module (SCGPVTM) (figure redrawn,

adopted from [227])

Precast elements can also be beneficial in terms of installation costs because the total retrofit cost as expected is a function of the materials used in the system and also the installation method and the location. The software models developed by TRNSYS, for example, show that using external insulation has 8% better energy performance, while the internal thermal insulation leads to 50% less investment cost and lower payback period

[22]; therefore, it is important to evaluate different precast components with various materials and layouts. For example, a study by Carbonaro et al. [264] on application of a mixture of common existing materials such as natural hydraulic lime, Portland cement, expanded perlite, granulated corncob, and granulated wheat straw as a vegetal based thermal plaster showed that the conductivity of different mixtures of such plasters can vary between 0.086 and 0.115 W/mK [264], which could be used as an insulation material.

193

There are research programs such as Envelope Approach to improve Sustainability and

Energy efficiency in Existing multi-story, multi-owner residential buildings (EASEE) devoted to find innovative and preferably modular solutions such as prefabricated shapeable retrofitting panel systems, shown in Figure 5-8, which include 1) a single layer of XPS with surface finishing and 2) composite EPS panel coated on both sides with textile reinforced mortar (TRM). The FRP (fiber reinforced polymers) reinforcements can be glass, carbon or aramid fibers. The tests for assessment of these products include thermal evaluation, as well as mechanical and hygrothermal performance. An example application and case study related to such products is shown in Figure 5-9 [228, 265].

Figure 5-8. XPS covered with final finish (left) and EPS coated with TRM (right) (reproduced from [228],

used with permission)

194

Figure 5-9. Application of EPS coated with TRM (1) and installation process(2) for a social housing in

Italy, (before retrofit (3) and after retrofit (4)) (reproduced from [265], used with permission)

The energy retrofit design for exterior walls can also benefit from combined systems such as external thermal insulation composite system (ETICS) shown in Figure 5-10, which is basically a layer of insulation board attached to the existing layer and covered with the final finish. Different insulation materials such as stone wool, EPS, and dense glass wool can be used. Different components of such a system can be adhesive, thermal insulation material, anchors, base coat, reinforcement that is usually glass fiber mesh, and finishing layer [229]. An apartment built in 1983 located in a social housing compound in southern

195

Italy is an example of using this system. The roof area was not sufficient to install PV panels, therefore, a combined system approach including the use of building integrated PV

(BIPV) system, improvement of the mechanical systems, and application of ETICS were considered. The insulation material in this ETICS is stone wool. Energy performance of the building was studied using DesignBuilder, which showed that the heat loss through walls was reduced by 85% [30]. BIPV along with other new methods such as movable structures that rotate around their axis, double glazed façade systems, and solar thermal vacuum tube collectors can be considered as examples of innovative energy retrofit methods [230].

Figure 5-10. External Thermal Insulation Composite Systesm (ETICS) (figure redrawn, adopted from

[229])

Composite systems such as ETICS can be designed in different ways in terms of the type of the material, number of layers, thickness, and the order of layers. When it comes to 196 deciding on the number of the insulation layers to install in a wall system with multiple layers, it is important to have an optimum design in terms of thermal properties, cost, temperature radiant within the wall, thickness, and order of layers. There is a study that shows in a two-layer wall system, it is more effective to put the insulation layer closer to the exterior side and it was also observed that in a wall system with multiple layers, there is an optimum number of layers and increasing the layers doesn’t necessarily result in a better performance [231].

While prefabricated new combined retrofit elements could be used in different envelope components such as walls or roofs, some of these combined systems are focused only on façade systems, multi-functional energy efficient façade system (MEEFS) that is still under development being such an example. As shown in Figure 5-11, basically, MEEFS is a modular system consisting of precast units that could be installed on a frame attached to the façade of the existing building [266]. Ease and speed of the installation could be the major benefits of this system. In addition, the panels, which could be installed on these units, are mostly innovative products that could be integrated with automated control system within a building such as energy smart homes.

197

Figure 5-11. The structure of multi-functional energy efficient façade system (MEEFS) (reproduced from

Examples of these panels are Advanced Passive Solar Collector & Ventilation Unit

Technological Unit (APSC&VU TU) and Advanced Solar Protection & Energy Absorption

Technological Unit (ASP&EA TU), which are both suggested and under development by

European Commission’s 7th Framework Programme for Research and Technological

Innovation. These two systems work based on thermal storage and phase change material concept [27, 266]. The APSC&VU TU shown in Figure 5-12 is operable and can switch between the thermal insulation with cool coating and PCM material with warm coating, which absorbs less and more solar energy, respectively [17]. For example, the heat can be absorbed during the day and then the warm side would be switched toward the interior space to warm up the space. The ASP&EA TU, however, works based on both thermal storage and air movement [18]. Solar radiation can go through the semi-transparent glazing component and heat will be absorbed by thermal storage behind it. Operable louvers on top

198 and bottom of the panel can be opened and closed if needed and the air will be able to move over the thermal storage to exchange heat. Both of these panels could be installed on the frames shown Figure 5-11 as precast elements to speed up the construction and their electric control capabilities would be a great feature in homes equipped with energy management and automation systems.

Figure 5-12. Advanced Solar Protection & Energy Absorption Technological Unit (ASP&EA TU) (top) and

Advanced Passive Solar Collector & Ventilation Unit Technological Unit (APSC&VU TU) (bottom) owned

by Acciona and Tecnalia, respectively (reproduced from [17, 18], use with permission)

Other façade system such as Active Solar Thermal Façades (ASTFs) can function as both a building envelope and solar collector component. The basic functions of an ASTF include solar absorption and heat gain by the thermal absorber and the amount of heat transfer

199 through convection, conduction, and radiation. Zhang et al. [19] classified different ASTFs that can be used as part of walls, windows, balcony, sunshield, or roof. Another type of innovative system for façade energy retrofit is double-skin façade, where the second layer is basically an additional façade over the existing transparent façade. An example of a wall- based ASTF is illustrated in Figure 5-13, which could be used as an opaque component in building envelope and the same concept is used for other components such as roof. The space between these two layers acts as an insulation layer that is heated up by solar radiation, and it can be ventilated if over-heated. In order to control and optimize the solar heat gain in double-skin façade, the effect of installing shading device in the cavity of this system was has been suggested as an option [20].

Figure 5-13. Wall-based ASTF (figure redrawn, adopted from [19])

200

As it was shown in Figure 5-4, the new energy retrofit methods could be made of innovative materials such as PCM or Nanogel; new technologies such as BIPV and SCGPVTM; or it could be a combination of conventional and new materials and technologies such as prefabricated elements or dynamic facades. The following section will have an overview of energy retrofit projects performed in buildings using both conventional and new retrofit methods reviewed already.

5.4 Example Retrofit Projects

Most of the energy retrofit projects target multiple retrofit measures and they are focused on building envelope, mechanical, or electrical systems. Reviewing these projects reveals that there could be a difference between the actual energy savings and results obtained based on computer modeling, which could be up to 42% [40, 41, 42, 43]. It could also bring up another field of study for researchers as risk assessment that investigates the inaccurate predictions in energy saving potentials after a retrofit project, which should be considered and included in energy retrofit guidelines [43]. The energy retrofit case studies reviewed in this section and summarized in Table 5-4 are not performed for research purposes.

However, these projects aim for improving the energy performance of existing buildings, which is the difference between the cases reviewed in Table 5-1 and Table 5-4.

These projects have used both conventional and innovative materials such as adding spray foam, rigid insulation, aerogel, PCM, improving window system, and double layer façade.

The main goal in many retrofit projects seems to be determining retrofit options with the

201 highest impact on reduction of energy consumption at the lowest cost, which will then have the combined effect of the shortest payback period.

An example project for using multiple retrofit measures is related to Aspinall Courthouse

[233] where alongside improving the mechanical systems, different approaches such as adding R-10 spray foam to walls, R-35 rigid insulation over the roof, and cool roof, were used in addition to improving the glazing components by installing storm panels with low

U-value and solar heat gain coefficient (SHGC). Another example of using multiple measures but with the main focus on the glazing system is the retrofit of the Empire State

Building [267] that consisted of replacing windows for 2,700,000 square feet of building envelope, which along with other retrofit measures led to decrease in annual energy use from 88 kBtu/sf to 60 kBtu/sf. In order to improve the energy performance of 6,514 windows, the existing window systems were re-manufactured onsite by using suspended coated film and gas filling the cavity between glass panes. Another measure was the addition of reflective barriers behind radiators that were close to the exterior walls [267].

Improving window systems could also include the framing. A retrofit case study of a hospital in Alexandria, Egypt is an example of improving the window system by adding thermal breaks to aluminum frames and replacing the clear glasses with double low-e with air gap glazing system. Batt insulation was also added to the current wall system to increase the R-value up to about 11. The final energy saving was obtained through computing modeling using EnergyPlus and OpenStudio as the Graphical User Interface (GUI) and was reported to be about 7,068,178 kW h/year [32].

202

There are also architecture-driven energy retrofit case studies such as California

Department of Motor Vehicles [240] that was focused on façade besides other measures such as lighting and HVAC improvement. Façade retrofit included a double layer skin that allowed using a ventilation layer with operable vents within it. During the hot months, natural ventilation helps cooling down the interior area, while during the cold months the trapped air helps to keep the interior warm [240]. Installing R-15 batt insulation behind masonry walls alongside the lightweight external sunshade and curtain wall and external shading with operable louvres above roof-level have also been reported in other projects mainly focused on the façade retrofit as in Stanford Medicine Outpatient Center, and

UCLA center for Health Sciences [240]. Another example is a project involving a 1973 built building in Portland, Oregon that was retrofitted to improve the HVAC, envelope, lighting, daylighting, and monitoring systems [241] where the energy retrofit led to 30% reduction in annual energy consumption. The envelope retrofit included the addition of 2 inches polyisocyanurate insulation with a white asphalt cap on the roof as well as the use of translucent cloth shades on the single-pane window to reduce heat gain and infiltration.

More advanced façade retrofit techniques were used in the 1998 built ERGO building that is the Italian headquarters of a major insurance company in Milan, Italy [235]. The initial computer model developed in DesignBuilder showed that the annual energy consumption was about 413 kWh/m2/year, with the breakdown of heating, cooling, artificial lighting, and electric purposes contributing, respectively, to 90, 76, 84, and 163 kWh/m2/year of the energy use. Different energy retrofit measures (ERMs) that led to 40% decrease in annual energy consumption included replacing the glazing system with argon filled double- pane

203 glass, installing shading system, installing ventilated façade consisting of an air gap behind the stone façade, installing stone façade to work as thermal mass and glass wool as insulation, installing dynamic double skin façade on the south of the building consisting of computer-controlled blinds and photovoltaic (PV) panels, and improved roof insulation by adding cellular glass [235]. Modular prefabricated products such as EPS panels covered with TRM could also be categorized under the innovative energy retrofit methods installed on the façade. Such a retrofit project was performed in Italy on a social housing building.

Installation of prefabricated panels made of 100mm EPS covered with 125mm polymer fiber-reinforced concrete over a façade of a building in Italy showed about 30% reduction in primary energy demand [265].

Deep energy retrofit has also been of interest in some projects including Alliance for

Sustainable Colorado, which is an example of deep energy retrofit of a 100-year old building renovated in 2006 [242]. The energy consumption of the building after energy retrofit is 42 kBtu/sf/yr that is about 55% better than the average for U.S. office building.

The energy retrofit effort was focused on HVAC, envelope, lighting, daylighting, controls, retro-commissioning, photovoltaics, and monitoring systems. A Mylar film was applied on the interior of curtain walls to reflect up to 60% of the heat during sunny days and reduce the internal heat loss in the winter [242]. Other examples of retrofitting office buildings include Beardmore in Priest River, Idaho that involved adding extensive insulation to the exterior walls and R-50 insulation in roof cavities in addition to converting the roof to a cool roof by improving the roofing material to a high solar reflectance material [236]. In order to keep the exterior window systems intact, it was decided to add insulated low-e

204 glazing system on the interior side [236]. Similar approaches have been taken in other projects such as Home on the Range in Montana that includes envelope retrofit by addition of exterior insulation over concrete block walls and replacing the windows with low-e glazing system [237]. The exterior walls and roof were also painted a light color and louvers, awnings, and trellises were used over the south windows to reduce solar heat gain

[237]. However, in retrofit projects such as Aventine office building in La Jolla, California, installing an EPA cool roof was the only retrofit measure because it was observed that the envelope already had enough R-value provided by thermal mass of concrete walls [268].

There are also other examples of using the cool roof strategy that can be more effective in hot and arid climate such as Abu Dhabi, UAE. A study of 10 villas in Abu Dhabi for a period of one year that were retrofitted by using windows with higher shading coefficient, cool roof, more efficient electrical and mechanical systems, and reducing the airtightness of the buildings to 5 ACH50 shows that the annual energy consumption can be reduced by something between 14.4% and 47.6%, depending on the behavior of the occupants [232].

Application of precast retrofit insulated panels are also studied in literature as the first step in deep energy retrofit (DER). The panels studied by Bianco and Wiehagen [40] consist of expanded polystyrene (EPS) covered by oriented strand boards (OSB), which are referred to as nail base panels. The case study residential house is a two-story building in climate zone 5, Albany, New York and the energy simulation is conducted by BEopt. The initial building is constructed by concrete masonry unit (CMU). In certain locations such the area between below the door and stairs landing that doesn’t have enough room, it was decided to use 0.4-inch aerogel fabric with R 4.8 thermal resistance, which is also water and fire

205 resistance. Results show about 21% and 16% reduction in gas and electricity annual use, respectively [40].

Passive House performance criteria have also been of interest as retrofit goals for some projects including the Glasswood building in Portland, Oregon [239]. The retrofit was designed to meet an airtightness of equal or better than 0.6 air changes per hour at 50 Pascal of pressure, annual heating and cooling energy consumption of less than or equal to 4.75 kBtu/sf/yr, and Less than or equal to 38.1 kBtu/sf/yr for primary energy demand. In order to provide the air tightness, OSB with taped joints was used. It was decided to keep the main structure of the wall system and improve the existing 2x4 wood frame by adding high density cellulose to the existing wall cavity and a layer of expanded polystyrene (EPS) added on the exterior face behind the rain screen [239]. The 1,658 ft2 Sunnyvale residential building is another example for Passive House approach and also the Building America program [238]. The envelope energy retrofit in this project included the addition of R-13 dense-pack cellulose in wall cavity, R-12 exterior foam, R-24 polyiso foam in cavity, spray foam on girders and rim joist, R-12 polyiso over the slab, R-38 polyiso over the roof deck, and replacement of windows with triple-pane glazing. The energy simulations were conducted by BEopt and finally, it was shown that the deep energy retrofit led to 40% energy saving [238].

There are also other target performance criteria in energy retrofit such as net zero energy house, an example being an old Danish multi-family house built in 1896 that was retrofitted to become a nearly-zero energy building by using more innovative materials and

206 techniques [243]. Two types of insulation materials/technologies consisting of aerogel- stone wool mixture and vacuum insulated panels were installed on the interior face of the walls. Also, thermal properties of windows were improved by installing a secondary frame, a sash mounted on the frame or to coupled frames. It has been shown that all retrofit measures that were not limited to building envelope led to about 68% saving in energy use from 162.5 kWh/(m2 year) to 51.5 kWh/(m2 year). The new layer of glazing installed over the existing one had low-e coating towards the exterior. Aerowolle or the Vacupor NT products are used in this project to retrofit the walls. The thermal conductivity of Aerowolle that is made of aerogel and stone wool fibers is 0.019 W/(m2 K). The second product used in this project is Vacupor NT, which is a vacuum insulated panel (VIP) with thermal conductivity of VIPthat varies under different ambient pressures. The product used in this

2 project had a thermal conductivity of 0.005 W/(m K) for a thickness of 20 mm under 1 mbar pressure measured as a center value, while at atmospheric pressure the thermal

2 conductivity of the Vacupor NT product is 0.019 W/(m K) [243]. The “three-liter” BASF house in Germany retrofit project also used innovative materials and techniques. The house was designed to keep the indoor temperature in the comfort zone with only three liters of heating fuel consumption per square meter per year. After the retrofit, it was reduced to 2.6 liter of heating oil per square meter. Different ERMs were used in this project, including high R-value exterior foam sheathing, triple-glazed windows, passive solar heating from glazed sunspaces, a controlled ventilation system with 85% thermal recovery, efficient heat and electricity generation by a new miniature power plant in the basement, and PCM- enhanced gypsum boards or internal plasters [269].

207

As it was explained in new retrofit measures section, MEEFS are a new area of focus studied by European Commission’s 7th Framework Programme for Research and

Technological Innovation. In order to evaluate some of these products, there is an ongoing project targeting a real building in Spain to investigate the performance some of the innovative measures such as green façade, ventilated façade, Advanced Passive Solar

Collector & Ventilation Unit Technological Unit (APSC&VU TU), and Advanced Solar

Protection & Energy Absorption Technological Unit (ASP&EA TU). These systems will be connected to an energy management system to evaluate the energy performance and monitor their functioning [270].

More common envelope retrofit techniques such as adding exterior insulation to walls and roof can be combined with more innovative materials and systems such as VIP, aerogel,

PCM, double-layer skin façade equipped with operable shading systems. While most retrofit measures apply to office, residential, and commercial buildings, the focus of much of the review presented was on residential buildings, including Passive House and net zero house that have more challenging retrofit goals compared to simpler goal of just decreasing the energy consumption. Table 5-4 summarizes the features of these example projects and the measures and results of the energy retrofit. Most of the example projects are not limited to envelope energy retrofit, and the final reported energy saving is due to the combination of improvement in envelope, mechanical, and electrical systems. It can be observed that the energy saving could be up to 40% depending on the methods used, however, it should be noticed that some of the reported numbers are energy saving potential or expected performance and the actual measurement is not conducted after the energy retrofit.

208

Table 5-4. Summary of the energy retrofit measures in energy retrofit projects

Project Title/Location Retrofit Measures in Building Enclosure Energy Saving

From 14.4% to Windows with higher shading coefficient and cool 47.6% is saved in Abu Dhabi, UAE [232] roof are used. Also, the airtightness of the buildings annual energy is reduced to 5 ACH50. consumption.

Nail base precast insulation panels consist of About 21% and expanded polystyrene (EPS) covered by oriented 16% reduction in strand boards (OSB) were installed over the existing Albany, New York [40] annual gas and concrete masonry units (CMU) and aerogel fabrics electricity use, are used, where there was not enough space for nail respectively base panels.

R-10 spray foam on walls, R-35 rigid insulation on

Aspinal Courthouse [233] the roof, cool roof, and storm panels with low U- -

Value and SHGC were applied

The existing windows were remanufactured onsite Annual energy

Empire State Building by using suspended coated film and gas filling the use reduced from

[267] cavity and reflective barriers were installed behind 88kBtu/sf to

the radiators 60kBtu/sf (32%)

Thermal breaks added to aluminum frames and the The annual energy

clear glasses replaced with double low-e with air gap consumption was

Alexandria, Egypt [32] glazing system. Batt insulation was also added to the reduced by

current wall system to increase the R-value up to 7,068,178 kW

about 11 h/year

California Department of A double layer skin façade with operable vents were - Motor Vehicles [241] installed

209

Project Title/Location Retrofit Measures in Building Enclosure Energy Saving

Stanford Medicine R-15 batt insulation behind masonry wall, external Outpatient and UCLA sunshade, curtain wall, and external shading with - center for Health Sciences operable louvres were added [241]

Two inches of polyisocyanurate with white asphalt 30% reduction in

Portland, Oregon [235] on the roof, and translucent cloth shades on single- annual energy

pane window consumption

Argon filled double-pane glass, shading system,

ERGO, Italian headquarters ventilated façade with an air gap behind the stone 40% reduction in of a major insurance façade, stone façade as thermal mass, glass wool as annual energy company [235] insulation, and dynamic double skin façade with consumption

computer controlled blinds were used.

Mylar film applied on the interior of curtain walls to Reached new 43 Alliance for Sustainable reflect up to 60% of heat during cooling seasons and kBtu/sf/yr energy Colorado [242] reduce internal heat loss in heating season consumption

Extra R-50 exterior wall insulation, high solar Beardmore, Priest River, reflective material for roofing materials, and - Idaho [236] insulated low-e glazing system were used

Exterior insulation over concrete block walls, low-e

Home on the Range, glazing systems, light color over the exterior walls - Montana [237] and roof, awnings, and trellises to reduce the solar

heat gain were used.

Annual Glasswood building, Airtightness reduced to 0.6 air changes per hour at consumption Portland, Oregon [239] 50 Pascal of pressure by taping the OSB sheathings. reduced to less

210

Project Title/Location Retrofit Measures in Building Enclosure Energy Saving

Also, high density cellulose and EPS were added to than 38.1

the cavity and behind the rain screen, respectively. kBtu/sf/yr

R-13 dense-pack cellulose in wall cavity, R-12

exterior foam, R-24 polyiso foam in cavity, spray 40% reduction in Sunnyvale residential foam on rim joist and girdirs, R-12 on slabs, R-38 annual energy building [238] polyiso over the roof deck, and triple-pane glazing consumption

system.

After applying

both envelope and

mechanical

Insulation materials containing aerogel-stone wool systems energy

mixture (thermal conductivity of 0.019 W/(m2 K)) is retrofit, energy

Old Danish multi-family used and vacuum panels installed on the interior face consumption house [243] of the wall. Windows thermal properties improved reduced from

by installing secondary framing over the existing 162.5

glazing system. kWh/m2/year to

51.5

kWh/m2/year

(68% decrease)

Primary energy

Installing 128 panels made of 100mm EPS covered demand was

EASEE, Italy [265] with 125mm polymer fiber-reinforced concrete over estimated to be

the façade of residential building. reduced by about

30%.

211

Project Title/Location Retrofit Measures in Building Enclosure Energy Saving

Evaluation of some innovative retrofit methods such

as green façade, ventilated façade, solar protection,

building-integrated photovoltaics (BIPV), an Results are not MeeFS, Spain [270] advanced passive solar protector/energy absorption published yet. auto mobile unit, and an advanced passive solar

collector/ventilation module on a real building in

Spain.

The required

High R-value exterior foam sheathing and triple energy for heating Three-liter BASF house, glazed windows were used alongside other is reduced to 2.6 Germany [271] mechanical systems retrofit. liter of heating oil

per square meter.

5.5 Multiple Criteria and Proper Tools on Choosing Retrofit

Measures

Multiple research studies exist using a combination of retrofit measures focused on both opaque and transparent components. However, depending on type of the building, climate zone, financial restrictions, and Life Cycle Assessment (LCA) results different retrofit measures could be chosen. Because of availability of multiple options, finding the optimum retrofit scenario has always been of great interest. For example, in a study with two different target objectives including cost-optimal and net-zero energy retrofit of a building, adding expanded polystyrene (EPS) insulation to the exterior wall, extruded polystyrene

212

(XPS) to roof and floor, and replacing windows and their frames with double glazed and

PVC frames, respectively, are considered as viable energy retrofit scenarios [272].

Therefore, it is important to review different criteria on choosing retrofit measures.

Building type can influence energy retrofit scenario. For example, the energy retrofit of historical buildings could be more challenging compared to retrofit of residential or commercial buildings, because of the need to preserve the historical identity of the building. Schwartz [273] discusses the pros and cons of adding batt insulation, dense-pack cellulose or spray foam, and adding an exterior layer of rigid foam insulation on such historical buildings. Accordingly, adding batt insulation could be easy to install or remove; however, it does not lead to the most energy efficient scenario. Moreover, the issue with more energy efficient options such as dense-pack cellulose is explained to be the difficulty of removing, while adding exterior insulation could damage the surface of existing roof or wall of the historical building [273].

Climate zone could also affect the decision-making on retrofit measures [246]. For example, there is a study focused on comparison of retrofit options in two different climate zones such as cold climate and mild Mediterranean or hot and arid climates such as Israel, where residential buildings consume about 30% of the total energy of this country [247].

The EnergyPlus analysis results showed that retrofit methods such as painting the roof white, adding roof insulation, adding insulation to exterior and interior face of the wall, and adding shading over windows have smaller impact on energy saving compared to cold climate region that is due to the current electricity cost and also the hot and arid climate

213 type in this area [247]. Research studies for different climate zones in Europe also shows that envelope retrofit might be more or less effective compared with renewable energy retrofits such as using solar systems. For example, Eicker et al. (2015) used OpenStudio that relies on EnergyPlus as the core simulation engine, and their study shows that in southern dry climates the envelope retrofit has slight influence, while in Mediterranean and

Oceanic climate regions it has higher influence compared with buildings retrofitted by incorporating solar systems [33].

Financial restrictions is always one of the most important criteria in choosing a material for retrofit; therefore, some studies evaluate a combination of multiple retrofit scenarios to find the most economical option. For example, studying 25 combinations of different retrofit options including PVC double glazed windows, and adding EPS, stone wool, insulated plaster, polyurethane foam, and mineral wool on the exterior wall, basement, and terrace floor shows that the cost difference between the most expensive and the optimal set of measures is about 2.7%; however, the energy saving could be up to 37% [219].

Life Cycle Assessment (LCA) and the energy associated with the materials production used in retrofit or construction can also be important in decision-making process of energy retrofit [274, 264, 251, 275]. Beccali et al. [274] discusses the life cycle assessment (LCA) of a few conventional retrofit materials for single-family homes by considering different material life stages such as manufacturing embodied energy, operation energy, and demolition energy. The retrofitting options considered by Beccali et al. [274] include installing EPS on exterior walls, adding rock wool over the roof, adding a layer of XPS

214 over the ground floor, installing PV panels, and replacing the boiler. Their study shows that the materials with higher embodied energy lead to lower operation energy [274].

Application of more environmental friendly retrofit materials such as reed is also reported as a more sustainable material to reduce the manufacturing energy and emissions [251]. A similar study [13] looked at life cycle impacts of 21 different scenarios of energy retrofit under Mediterranean climate in Portugal for an attic space. The study [13] discusses three different frame materials including wood, light steel, and lightweight concrete, three different insulation materials including rock wool, XPS, and polyurethane foam, and three thickness of insulation including 40, 80, and 120 mm. The polyurethane foam had the lowest lifecycle impact in most cases, while rock wool had the lowest primary energy impact in construction phase. The study results also showed that adding 40 mm of insulation led to 30% reduction in annual operational energy of the second floor, but a thicker insulation did not lead to more significant change, and that the operation phase accounted for 40-70% of the life cycle impacts [275].

Energy simulation tools and optimization methods could contribute in decision-making process based on the criteria discussed above and considered in a retrofit project. Current energy simulation tools being used in energy retrofit of buildings could be great assets when it comes to quantifying the effect of different measures, simultaneous evaluation of multiple measures to consider their interaction, long-term evaluation of different measures, and assessment of measures in different climate zones [31, 244, 245, 246, 247, 33]. For example, Crawley et al. [276] compared the capabilities of 20 different energy modeling software tools such as EnergyPlus, BEopt, eQuest, TRNSYS, BLAST, DOE-2, and HAP.

215

Most of the energy simulation tools only offer graphical user interface (GUI) capability such as OpenStudio, DesignBuilder, and eQuest, which is interconnected with a simulation engine that actually performs the energy simulation, where in most cases either EnergyPlus or DOE-2 is used as the energy simulation engine. However, there are examples of independent energy simulation tools such as DeST developed in China [277]. Optimization methods such as multiple criteria complex proportional assessment (COPRAS) method could be also implemented in simulation tools, when there are multiple retrofit scenarios in order to reach the best and optimized option [35, 36, 37, 38, 39]. Assessment of energy retrofit measures is not always conducted by direct use of energy simulation tools. Lee et al. [278] reviewed the application of a database that uses EnergyPlus pre-run simulations for seven building types in different climate zones as part of a project titled database of energy efficiency performance (DEEP), which is a SQL database, to replace the energy audit process that could be expensive and time-consuming [278].

5.6 Numerical Study of Different Envelope Energy Retrofit

In this section, the effect of certain conventional and new energy retrofit measures for building envelope systems is studied by quantifying the influence of each method and comparing their performances in cold and hot-dry climates. To do so, a BEopt computer model was developed for a residential building retrofitted with different retrofit options including adding exterior rigid insulation to the walls, adding window film, roof insulation,

PCM materials, materials with low/high reflectance on the roof (warm/cold roof), and materials containing aerogel. Considering the energy saving potential of each retrofit

216 option, the results are compared to the benchmark building that represents a typical two- story residential building in the U.S.

BEopt could model a house under two different circumstances including a new construction or retrofit case. The major difference between a new construction and retrofit options in BEopt is related to defining the age of appliances and mechanical system’s components, relevant to retrofit cases. This option helps calculating the current value of different components of building. Since this study is only focused on the retrofit measures on building envelope (not appliances), there would not be any differences between these two options; therefore, it is assumed as a new construction, with each measure evaluated in two different climate zones to assess their performance under different heating and cooling loads. Especially because materials such as PCM show different performance under different temperature variance, it would be beneficial to consider two climate zones with different characteristics. Based on the DOE’s Building Technologies Office guide, the weather files for Pittsburgh, PA and Los Angeles, CA, which represents cold (5A) and hot- dry (3B) climate, respectively were chosen [279].

5.6.1 Development of the Computer Model

The benchmark house was modeled in BEopt, followed by applying all envelope retrofit options to the benchmark house one at a time and comparing the results. The benchmark house should be a good representative of the existing houses; therefore, it might need different properties depending on the location of the house [271, 280]. There are also codes in other countries such as Germany that present certain requirements for energy efficient

217 reference buildings (benchmark building) that recommends minimum values for different properties of the house [280].

Data from United States Census Bureau are used to study the dominant properties of residential buildings in the U.S. and select the building envelope and architectural components in benchmark house design for this study. Houses built in the U.S. from 1990 to 2000, with the living area ranging from 2105 to 2435 square feet, are considered representative of buildings that need some energy retrofit. The dominant properties of these houses with regards to the number of bathrooms, bedrooms, method of construction, foundation type, framing, wall material and number of the stories are also studied and used to design the benchmark house to ensure the model is a good representative of existing single-family houses in the U.S. These are presented in Table 5-5, noting that properties such as framing type and number of stories are only available for the more recent constructions according to which the benchmark house is designed [281].

Table 5-5. The dominant properties of homes builtt between 1990 and 2000 in the U.S. [281]

Building Properties Values/Types

Square feet (ft^2) 2232 # of Bathroom 2.5 # of Bedroom 3 Construction Method Site-built Exterior Wall Material Vinyl siding Foundation Type Full/Partial basement Framing Type Wood Number of Stories 2

218

The building properties related to the building envelope of the benchmark house in BEopt are presented in Table 5-6, while other properties such as mechanical, electrical, and appliances are presented in Table 5-7. The mechanical and electrical components are not the focus of this study; therefore, the properties related to these two categories are selected based on the Building America Benchmark house available in BEopt.

Table 5-6. The building envelope properties of the benchmark house defined in BEopt

R-value Component/System (h.ft^2.R/Btu)

R13 Fiberglass batt, Wall Framing Factor = 0.25 11.4 2×4, 16 in O.C. Wall Sheathing OSB Exterior Finish Vinyl, Light Solar Absorptivity = 0.3 0.6 Ceiling R-38 Roof Taper Factor = 0.99, 1, 1 Unfinished Attic 31.3 Cellulose, Vented Framing Factor = 0.07 Asphalt Shingles, Color = medium Absorptivity = Roof Material Medium 0.85 Emissivity = 0.91 Slab Uninsulated Carpet 80% Carpet 2.08

Exterior Wall Sensible Capacity (Btu/F.ft^2) = 0.5 in Drywall Mass/Partition Wall 0.42 Mass/Ceiling Mass

Window to Wall Ratio 15% Perimeter/Area Ratio = 1.41 (B, F, L, R) Medium-Gain and Non-metal Frame Windows 2.86 Low-E Double-Pane and Argon-Fill

Interior Shading Heating and Cooling Multiplier = 0.7

Door Area 20 ft^2 Doors Fiberglass - Swinging Eaves 2 ft

219

Table 5-7. Other properties of the benchmark house defined in BEopt

Component/System

Air Leakage 7 ACH50 Mechanical Ventilation Exhaust, 2010 ASHRAE 62.2 Natural Ventilation Year-Round, 3 days/week Central Air Conditioner SEER 13 Gas, 78% AFUE Furnace Max Supply Temp = 120 F 15% Leakage Ducts Insulation R-Value = 8 (h.ft^2.R/Btu) Ceiling Fan Cooling Set Point 76 F Heating Set Point 71 F Energy Factor = 0.59 Water Heater Tank Volume = 40 gal Uninsulated, TrunkBranch, Distribution Copper 34% CFL Hardwired, 34% Lighting CFL Plugin Top freezer, EF=17.6 Refrigerator Elec Use = 434 (kWh/yr) Volume = 18 (ft^3) Cooking Range Electric Dishwasher 318 Rated kWh Clothes Washer Standard Clothes Dryer Electric Plug Loads 2317 (kWh/unit/yr)

Based on Table 5-5, the number of bedrooms and living area footage are selected as 3 and

2232 square feet, respectively. Other properties include full basement foundation, 2.5 bathrooms, and two stories, each 1116 square feet (31 by 36 feet). The Full basement is not included in the square footage; however, it was decided to have the same area as each of

220 the two stories. Based on the main energy retrofit methods reviewed in this study, seven different major retrofit scenarios are considered in this simulation study listed in Table 5-8, which is explained in details in Table 5-10. BEopt uses EnergyPlus as the simulation engine; therefore, a weather file needs to be defined. For this study, Pittsburgh, PA and Los

Angeles, CA represent cold (5A) and hot-dry (3B) climate regions, respectively.

Orientation of the house can be another parameter affecting the energy consumption; however, it was decided to keep the direction toward north for all the simulations. There are studies that show different orientations can lead up to 0.46% change in annual energy consumption [282]; however, it depends on different properties such as building type, location, and window to wall ratio. The envelope of benchmark house with properties presented in Table 5-6 is representative of houses in the U.S. with insufficient insulation.

Based on the model explained in this section, the results of the computer modeling are presented in the next section.

221

Table 5-8. Different energy retrofit methods used for computer modeling

Energy Retrofit Method Method of modeling in BEopt

Change in wall R-value by adding Exterior wall insulation exterior insulation

Exterior window film Change in window properties

Roof insulation Change in roof R-value

Change in roof reflectance Cool roof properties

Slab insulation Change in slab R-value

PCM as plaster or Change in thermal mass

wallboard properties

Mats, panels, and gypsum

boards containing or Change in wall R-value

coated by aerogel

5.6.2 Results of the Computer Modeling and Discussion

In order to compare different retrofit methods, energy saving related to heating and cooling energy is compared with the material cost for each measure. Because the effects of mechanical and electrical components are not the focus of this study, only the changes in the heating plus cooling loads directly related to the building envelope properties are reported. Moreover, in order to make the outputs scalable for other climate zones and square footage, the ratio of annual heating and cooling loads reduction over the heating and cooling load of the benchmark house as opposed to the actual annual energy consumption are reported. The Y-axis in Figure 5-14 and Figure 5-15 represent these values.

222

Moreover, both energy saving and material cost are normalized relative to the benchmark house. The material and installation costs considered in this study are mostly based on default values available in BEopt. In some cases such as Nanogel that is not a default material available in BEopt the values available in literature are used. This study is not focused on the actual cost and sole economical aspect of the energy retrofit; therefore, all the material and construction costs are normalized based on the total cost obtained from

BEopt to reduce the effects of inaccuracies in costs available in BEopt. The X-axis in

Figure 5-14 and Figure 5-15 represent these values. However, the costs considered for aerogel and PCM are explained in more details, since they are not among conventional retrofit materials defined in BEopt and need more in depth explanation.

Figure 5-14 and Figure 5-15 show bubble charts summarizing BEopt analysis results in two different climate zones that helps understanding the effectiveness of different envelope retrofit measures reviewed in this study. Multiple scenarios can be considered for each measure as summarized in Table 5-10. Different measures considered and modeled are explained and results are evaluated in more details.

Exterior/Interior Wall Insulation

To study the effectiveness of adding exterior insulation to a residential house similar to the benchmark house considered in this study, the insulation materials presented in Table 5-10 consisting of R-10 XPS, R-15 XPS, R-6 Polyiso, and R-12 Polyiso were considered and compared with the benchmark house, which does not have any exterior insulation. Results show that different insulation materials added to the exterior walls can lead to about 15-

223

24% and 24-30% decrease in annual heating and cooling loads combined in cold and hot- dry climate region, respectively. The highest increase in material cost is up to 6% compared to benchmark house, which is correspondent to the R-12 Polyiso insulation. The effects of different energy retrofit measures depends on the building type and energy consumption pattern prior to the retrofit and outcomes could be significantly different [32]. Besides these rigid insulation types, aerogel and PCM were also considered for this retrofit measure as innovative materials. For aerogel, it is assumed that the retrofit budget is the same as conventional material (R-15 XPS and R-12 Polyiso); therefore, the aerogel added to wall has either R-3 or R-2.3. Based on recent studies on aerogel [225, 28], the material can have thermal conductivity of 0.016 W/(m.K), which means aerogel with one inch (2.54 cm) thickness would be equivalent to an R-value of about R-10. The approximate cost of 10- mm thick aerogel blanket is $5.5/ft^2, which means that for an inch of aerogel, it will be about $14/ft^2 [283, 234]. For PCM, there are two default scenarios defined in BEopt used for this study. Application of PCM could be in form of micro capsules within a drywall or a coating applied over the drywall. The properties of the PCM considered in this study are presented in Table 5-9. Based on previous studies it was decided not to consider partitions among the retrofitted components because the effectiveness was found to be negligible

[284]. The scenarios considered for wall retrofit using PCM, and aerogel resulted up to about 2% and 11% reduction in annual heating and cooling loads, respectively, for cold climate region. Both PCM and aerogel show higher relative effectiveness in hot-dry climate regions by 18% and 17% reduction in annual heating and cooling loads combined. The cost

224

difference could be between 5-20% and 5-6% higher compared to the initial construction

cost of the benchmark house for PCM and aerogel, respectively.

Table 5-9. Properties of the PCM materials used for PCM drywall and PCM finish

Sensible Latent Specific Melting Latent Capacity Capacity Thickness Density Heat Cost Temp. Heat (Btu/F- (Btu/F- (in) (lb/ft^3) (Btu/lb- ($/ft^2) (F) (Btu/lb) ft^2) ft^2) F) PCM 0.69 21.87 0.5 50 0.33 73 10.5 8.63 Drywall PCM Drywall 0.42 0.5 53.7 0.2 coated Coating 0.3 51.19 0.133 53.7 0.5 73 86 2.57 Drywall

Windows

To study the range of effectiveness of retrofitting windows on heating and cooling loads,

four scenarios were considered in this study including low emissivity, double/triple pane,

non-metal/insulated framing, air filled, and high SHGC. As it is illustrated in Figure 5-14

and Figure 5-15, retrofit of windows considered in this study can save up 11% and 6% in

annual heating and cooling loads in cold and hot-dry climate regions, respectively;

however, the related costs could be 1-20% higher compared to the benchmark house. Triple

pane, low-e, and air filled glazing system is the most expensive option considered in this

study. Similar numerical studies using DesignBuilder shows about 2% saving in total

energy consumption due to replacing the existing windows with low emissivity single glaze

[29].

225

Roofing Materials/Insulation

Two major methods can be considered in roof retrofit, including adding insulation or application of cool/warm roof, where warm roof has roofing materials with lower reflectance, which is more desirable in cold climate regions. To increase the thermal resistance of roof, R-38 and R-60 cellulose and open-cell spray foam were considered as retrofit options. Based on the simulation results shown in Figure 5-14 and Figure 5-15, for residential buildings with small roof area, the warm/cool roof cannot be as effective as other retrofit measures, and the range of effectiveness is not significant. On the other hand, as it can be observed in Figure 5-14 and Figure 5-15, the roof retrofit by adding roof insulation led up to 5% and 8.5% decrease in annual heating and cooling loads in cold and hot-dry climate zones, respectively, and the material cost can be about 6% higher compared to the benchmark house construction costs.

Slab Insulation

The results show that improving slab insulation by adding R-5 and R-7 fiberglass batt insulation for such a two-story residential building is not as effective as other retrofit measures considered in this study. Figure 5-14 shows that the slab retrofit does not indicate a significant change in heating and cooling load; however, it can cost up to 5% more than the initial construction cost of the benchmark house. With respect to hot-dry climate region,

Figure 5-15 shows a negative effect of slab retrofit up to 5% on annual energy consumption on heating and cooling, which may be explained by the positive effect of unfinished basement on cooling of the building. In hot-dry climate, the heating load is lower and the

226 impact of the cooling load is higher. Therefore, retrofitting of slabs diminishes the positive effect of the unfinished basement, which lead to negative effect in annual cooling and heating load.

In general, as it can be observed in Figure 5-14 and Figure 5-15, which shows the retrofit options with material costs and final heating and cooling loads, options such as adding insulation to the slab does not show any major impact on energy performance. Warm roof and adding PCM to drywalls could be expensive and does not affect the heating and cooling load significantly compared to more common insulation materials; however, in hot-dry climate materials such as PCM shows better performance. Methods such as adding aerogel or conventional insulation materials to the walls, improving the glazing system, and adding insulation to the roof could be more economical and effective in terms of lowering the heating and cooling loads.

227

Table 5-10. Retrofit measures studied in computer modeling in details

Retrofit measure Exterior/Interior Insulation No exterior insulation (Benchmark) R-10 XPS R-15 XPS R-6 Polyiso R-12 Polyiso

Aerogel R-3

Aerogel Aerogel R-2.3

PCM Drywall

PCM Drywall/PCM Mat Windows Low E, Double Pane, Non-metal, Arg, M-Gain (Bechmark) Low E, Double Pane, Non-metal, Air, H-Gain Low E, Double Pane, Insulated, Air, H-Gain Low E, Triple Pane, Non-metal, Air, H-Gain Low E, Triple Pane, Insulated, Air, H-Gain Roofing Materials/Insulation R-30, Cellulose (Benchmark) R-38, Cellulose R-60, Cellulose R-38, Open Cell Spray Foam R-60, Open Cell Spray Foam Tile, Dark Tile, White

Roof Metal, Dark

Warm/Cold Metal, White Slab Insulation No insulation-80% Carpet (Benchmark) R-5 Fiberglass batt R-7 Fiberglass batt

228

Figure 5-14. Annual heating and cooling site energy saving potential and material cost of different

envelope energy retrofit measures in cold climate region

Figure 5-15. Annual heating and cooling site energy saving potential and material cost of different

envelope energy retrofit measures in hot-dry climate region 229

5.7 Summary and Conclusions

Among different sectors that contribute to energy consumption in the U.S., residential and commercial sectors approximately contribute to 21% and 19% of total annual energy consumption. About 42% of residential building’s energy use is considered resulting from heating and cooling. Energy retrofit of existing buildings is one of the most important and effective approaches in reducing energy consumption. Building envelope is one of the areas of focus in energy retrofit projects and there are several research studies and projects on actual buildings focused on that.

Materials and systems used for building envelope energy retrofit could be categorized under conventional and innovative measures. The former includes windows with multiple panes; coatings with different SHGC and emissivity for glazing; and insulation materials such as EPS, XPS, and polyisocynurate in form of either rigid boards or sprays. Innovative measures could include systems consisting of different layers of conventional materials, new materials such as PCM and aerogel, a combination of both, and innovative combined systems such as dynamic facades, SCGPVTM, ETICS, BIPV, and MEEFS. In order to study the effectiveness of these measures in real-world projects, multiple building energy retrofit project cases are also reviewed in this paper. Most of the research studies and real- world projects are focused on multiple areas as opposed to solely envelope retrofit. The reported results on total energy consumption reduction vary between 6% and 48% considering that various retrofit methods are studied.

230

Different criteria such as type of the building, climate zone, financial restrictions, and Life

Cycle Assessment (LCA) results could lead to different selecting different retrofit measures, which need to be considered during the decision-making process. In addition, tools that could contribute in decision making process such as energy simulation GUIs such as OpenStudio, BuildingDesign, BEopt, and eQuest; energy simulation engines such as

DOE-2, EnergyPlus, and TRNSYS; and optimization methods such as COPRAS could be used. Some of the research studies that identified these criteria and used these tools are reviewed in this paper and results show that computer modeling based on these criteria prior to performing the actual retrofit; however, the accuracy of the energy simulation tools could still be an issue.

In order to compare certain conventional retrofit measures with some new innovative systems and materials, BEopt software capability was used to model a typical residential house in the U.S. in two different climate zone of Pittsburgh, PA and Los Angeles, CA, which represents cold (5A) and hot-dry (3B) climate, respectively. The dominant structural, architectural, and building envelope properties of residential houses from 1990 to 2000 obtained and used in the computer model. Measures used in this study include adding insulation to exterior walls, roof, and slabs; improving the windows by installing air-filled double and triple pane windows with different SHGC coefficients; installing roofing materials with darker or lighter colors; and mixing PCM with plasters or being used within gypsum boards on the interior walls. The insulation options used include PCM, aerogel, XPS, Polyisocynurate, cellulose, and fiberglass. The results are reported as the ratio of annual heating and cooling loads reduction over the heating and cooling load of

231 the benchmark house versus the percentage of increase in material and installation costs.

All the costs are based on default values in BEopt, except the aerogel, which is assumed to be $14/ft^2 for an inch thick blanket of aerogel.

Results as summarized in Figure 5-14 and Figure 5-15 show that adding insulation to walls could lead to highest reduction in heating and cooling loads in both cold and hot-dry climates by about 15-23% and 23-30%, respectively, with an increase in the initial costs up to about 6%. Adding aerogel to walls could be more effective compared to adding conventional insulation to the roof, however, it is more expensive. It can increase the initial cost by 3%, while adding roof insulation increase it up to 7%. PCM added to walls shows better performance in hot-dry climate region. In cold and hot-dry climates, it leads to about

2.5% and 18% decrease in heating and cooling loads, respectively, and increase the cost by about 5% when it is used within gypsum boards. Insulating slabs and changing roofing materials to apply the warm/cold roof concept do not show any significant change in heating and cooling loads in a residential building modeled in this study, since it is only a two-story building and the surface area of the roof is not significant enough.

In general, it can be observed that new and emerging energy retrofit systems can compete with conventional systems; however, the initial costs might be still an issue in some cases.

A focus on the façade of buildings could be noted and the capability of adding dynamic components to these systems is a good option, which makes them more interesting especially with emergence of new automation systems within buildings. Computer tools and modeling still need to be improved in terms of accuracy and they also need to be

232 modified and improved to be able to integrate the capability of modeling these new retrofit measures, especially the ones focused on the façade of the building.

5.8 References

[215] 22 November 2016. [Online]. Available: http://www.eia.gov/totalenergy/data/monthly/#consumption. [Accessed 26 August 2016].

[216] "2009 RECS Survey Data," EIA, 2009. [Online]. Available: http://www.eia.gov/consumption/residential/data/2009/. [Accessed 9 September 2016].

[217] L. Pe´rez-Lombard, J. Ortiz and C. Pout, "A review on buildings energy consumption information," Energy and Buildings, no. 40, pp. 394-398, 2008.

[218] A. Papadopoulos, "State of the art in thermal insulation materials and aims for future developments," Energy and Buildings, no. 37, pp. 77-86, 2005.

[219] D. Rusu, D. Popescu and D. Popescu, "TECHNICAL AND ECONOMIC

CONSIDERATIONS ON ENERGY EFFICIENCY OF EXISTING BUILDINGS

THROUGH REFURBISHMENT," in TIIM Joint International Conference, Bari, Italy,

2015.

[220] M. Holladay, "Retrofitting Exterior Foam on Existing Walls," Building Green, 09

May 2009. [Online]. Available: https://www.buildinggreen.com/product- review/retrofitting-exterior-foam-existing-walls. [Accessed 10 September 2016].

233

[221] C. Ojczyk, "Cost Analysis of Roof-Only Air Sealing and Insulation Strategies on 1

½-Story Homes in Cold Climates," U.S. Department of Energy, 2014.

[222] "Advanced Energy, Practical Ways to Improve Energy Performance, Retail

Buildings," Pacific Northwest National Laboratory for U.S. Department of Energy, 2011.

[223] "Advanced Energy Retrofit Guide, Practical Ways to Improve Energy

Performance, Healthcare Facilities," National Renewable Energy Laboratory for U.S.

Department of Energy, 2013.

[224] Aspen, "Aspen," 2016. [Online]. Available: http://www.aerogel.com/resources/case-studies/. [Accessed 14 December 2016].

[225] M. Casini, "Smart materials and nanotechnology for energy retrofit of historic buildings," International Journal of Civil and Structural Engineering– IJCSE, vol. 1, no. 3, pp. 88-97, 2014.

[226] U. Berardi, "The development of a monolithic aerogel glazed window for an energy retrofitting project," Applied Energy, vol. 154, pp. 603-615, 2015.

[227] S. Singh, S. Agarwal, G. Tiwari and D. Chauhan, "Application of genetic algorithm with multi-objective function to improve the efficiency of glazed photovoltaic thermal system for New Delhi (India) climatic condition," Solar Energy, no. 117, pp. 153-166,

2015.

234

[228] G. Masera, G. Iannaccone and G. Salvalai, "Retrofitting the Existing Envelope of

Residential Buildings: Innovative Technologies, Performance Assessment and Design

Methods," in Advanced Building Skins - 9th Energy Forum, Bressanone, 2014.

[229] "ETICS are quality systems," Europian Association for External Thermal

Insulation Composite Systems, [Online]. Available: http://s278536490.online.de/views/etics/about-etics.html. [Accessed 11 September 2016].

[230] D. A. Chwieduk, "Towards modern options of energy conservation in buildings,"

Renewable Energy, vol. 101, pp. 1194-1202, 2017.

[231] P. Gori, C. Guattari, L. Evangelisti and F. Asdrubali, "Design criteria for improving insulation effectiveness of multilayer walls," International Journal of Heat and Mass

Transfer, vol. 103, pp. 349-359, 2016.

[232] F. AlFaris, A. Juaidi and F. Manzano-Agugliaro, "Energy retrofit strategies for housing sector in the arid climate," Energy and Buildings, vol. 131, pp. 158-171, 2016.

[233] R. Chang, S. Hayter, E. Hotchkiss, S. Pless, J. Sielcken and C. Smith-Larney,

"Aspinall Courthouse: GSA’s Historic Preservation and Net-Zero Renovation," U.S.

Department of Energy, 2014.

[234] BuyAerogel.com, [Online]. Available: http://www.buyaerogel.com/. [Accessed 19

September 2016].

235

[235] N. Aste and C. Del Pero, "Energy retrofit of commercial buildings: case study and applied methodology," Energy Efficiency, vol. 6, pp. 407-423, 2013.

[236] "Beardmore Building Case Study," 27 April 2012. [Online]. Available: http://newbuildings.org/resource/beardmore-building-case-study/. [Accessed 14

September 2016].

[237] "Home On The Range Case Study," 27 April 2012. [Online]. Available: http://newbuildings.org/resource/home-range-case-study/. [Accessed 14 September 2016].

[238] A. German, A. Siddiqui and B. Dakin, "Sunnyvale Marine Climate Deep Retrofit,"

U.S. Department of Energy, Davis, 2014.

[239] S. Hagerman, "Construction Methods and Approach to the Glasswood Commercial

Passive House Retrofit," Journal of Green Building, vol. 9, no. 3, pp. 23-36, 2014.

[240] "Deep Energy Retrofits: An Emerging Opportunity," The American Institute of

Architects and Rocky Mountain Institute, 2013.

[241] "200 Market Building Case Study," 27 April 2012. [Online]. Available: http://newbuildings.org/resource/200-market-building-case-study/. [Accessed 14

September 2016].

[242] "Alliance Center Case Study," 27 April 2012. [Online]. Available: http://newbuildings.org/resource/alliance-center-case-study/. [Accessed 14 September

2016].

236

[243] M. Morelli, L. Rønby, S. E. Mikkelsen, M. G. Minzari, T. Kildemoes and H. M.

Tommerup, "Energy retrofitting of a typical old Danish multi-family building to a “nearly- zero” energy building based on experiences from a test apartment," Energy and Buildings, vol. 54, pp. 395-406, 2012.

[244] S. Chidiac, E. Catania, E. Morofsky and S. Foo, "Energy," Effectiveness of single and multiple energy retrofit measures on the energy consumption of office buildings, no.

36, pp. 5037-5052, 2011.

[245] P. Xu, Y. Shen, L. Chen and J. Mao, "Assessment of energy-saving technologies retrofitted to existing public buildings in China," Energy Efficiency, vol. 9, no. 1, pp. 67-

94, 2016.

[246] V. M. Nik, É. Mata and A. S. Kalagasidis, "A statistical method for assessing retrofitting measures of buildingsand ranking their robustness against climate change,"

Energy and Buildings, no. 88, pp. 262-275, 2015.

[247] C. Friedman, N. Becker and E. Erell, "Energy retrofit of residential building envelopes in Israel: A cost benefit analysis," Energy, no. 77, pp. 1-11, 2014.

[248] J. Ye, "EFFECTIVENESS OF AIR SEALING MEASURES AND OTHER

ENERGY RETROFIT MEASURES IN LAFAYETTE ENERGY IMPROVEMENT

PROGRAM," PURDUE UNIVERSITY GRADUATE SCHOOL, 2014.

[249] E. Kamel and A. Memari, "Energy Retrofit Methods for Building Envelope

Systems," Pennsylvania Housing Research Center (PHRC), State College, 2016. 237

[250] J. Lstiburek and K. Ueno, "Hybrid Foundation Insulation Retrofits: Measure

Guideline," Building America for U.S. Department of Energy, 2012.

[251] M. Pertosa, A. Laura Pisello, V. Lucia Castaldo and F. Cotana, "Environmental sustainability concept applied to historic buildings: the experience of LEED international protocol in the stable of Sant’Apollinare fortress in Perugia," in 14th CIRIAF National

Congress. Energy, Environment and Sustainable Development, Perugia, Italy, 2014.

[252] A. L. Pisello, C. Piselli and F. Cotana, "Thermal-physics and energy performance of an innovative green roof system: The Cool-Green Roof," Solar Energy, no. 116, pp. 337-

356, 2015.

[253] P. Boarin, D. Guglielmino, A. L. Pisello and F. Cotana, "Sustainability assessment of historic buildings: lesson learnt from an Italian case study through LEED® rating system," Energy Procedia, no. 61, pp. 1029-1032, 2014.

[254] K. Ok Lee, M. A. Medina, E. Raith and X. Sun, "Assessing the integration of a thin phase change material (PCM) layer in a residential building wall for heat transfer reduction and management," Applied Energy, vol. 137, pp. 699-706, 2015.

[255] R. Baetens, B. Petter Jelle, A. Gustavsen and S. Grynning, "Gas-filled panels for building applications: A state-of-the-art review," Energy and Buildings, vol. 42, pp. 1969-

1975, 2010.

238

[256] E. Cuce and P. Mert Cuce, "The impact of internal aerogel retrofitting on the thermal bridges ofresidential buildings: An experimental and statistical research," Energy and Buildings, vol. 116, pp. 449-454, 2016.

[257] R. Levinson, Cool Roof Q & A, Lawrence Berkeley National Laboratory, 2009.

[258] Y. Kusama and Y. Ishidoya, "Thermal effects of a novel phase change material

(PCM) plaster underdifferent insulation and heating scenarios," Energy and Buildings, vol.

141, pp. 226-237, 2017.

[259] P. Han, L. Lu, X. Qiu, Y. Tang and J. Wang, "Preparation and characterization of macrocapsules containing microencapsulated PCMs (phase change materials) for thermal energy storage," Energy, vol. 91, pp. 531-539, 2015.

[260] F. Souayfane, F. Fardoun and P.-H. Biwole, "Phase change materials (PCM) for cooling applications in buildings: A review," Energy and Buildings, vol. 129, pp. 396-431,

2016.

[261] S. Harasim, "Aspen," March 2011. [Online]. Available: http://apps1.eere.energy.gov/buildings/publications/pdfs/building_america/ns/eemtg0320

11_c19_aerogel_insul.pdf. [Accessed 14 December 2016].

[262] B. Griffith, D. Arasteh and D. Türler, "Gas-Filled Panels: An Update on

Applications in the Building Thermal Envelope," in Presented at the BETEC Fall

Symposium, Washington, D.C., 1995.

239

[263] A. Chen, "sciencebeat," Lawrence Berkeley National Laboratory, 19 June 2000.

[Online]. Available: http://www2.lbl.gov/Science-Articles/Archive/airliner- packaging.html. [Accessed 15 April 2017].

[264] C. Carbonaro, S. Tedesco, F. Thiebat, S. Fantucci, V. Serra and M. Dutto, "An integrated design approach to the development of a vegetal-basedthermal plaster for the energy retrofit of buildings," Energy and Buildings, vol. 124, pp. 46-59, 2016.

[265] G. Salvalai, E. Pizzi, G. Iannaccone and M. Maria Sesana, "HE BUILDING

ENVELOPE RENOVATION TROUGHT COMPOSITE INSULATION SYSTEM.

INSTALLATION AND MEASUREMENT OF PERFORMANCE," in Proceedings of the

Colloqui.AT.e 2016, Matera (ITALY), October 13-14, 2016, ISBN 978-88-492-3311-7,

Matera, Italy, 2016.

[266] "Multifunctional Energy Efficient Facade System for Building Retrofit," European

Commission’s 7th Framework Programme for Research, [Online]. Available: http://www.meefs-retrofitting.eu/project/retrofitting-facade-concept.html. [Accessed 12

September 2016].

[267] "Case Studies, Empire State Building, New York," [Online]. Available: http://www.rmi.org/retrofit_depot_get_connected_true_retrofit_stories##empire.

[Accessed 14 September 2016].

[268] "Aventine Case Study," 27 April 2012. [Online]. Available: http://newbuildings.org/resource/aventine-case-study/. [Accessed 14 September 2016].

240

[269] J. Kośny, E. Kossecka, A. Fallahi and D. Yarbrough, "Use of Thermal Inertia for

Reduction of HVAC Energy Consumption in Cooling Dominated and Mixed Climates," in

BEST4 Technical Committee, National Institute of Building Sciences, 2015.

[270] MeeFS, "European Commission’s 7th Framework Programme for Research,"

MeeFS, 2017. [Online]. Available: http://www.meefs-retrofitting.eu/demonstration.html.

[Accessed 12 April 2017].

[271] M. MacDonald and S. Livengood, "Benchmarking Residential Energy Use," 2000

ACEEE Summer Study on Energy Efficiency in Buildings, Washington, D.C. : United

States. Dept. of Energy ; Oak Ridge, Tenn, 2000.

[272] M. Ferreira, M. Almeida, A. Rodrigues and S. Monteiro Silva, "Comparing cost- optimal and net-zero energy targets in building retrofit," Building Research & Information, pp. 1-14, 2014.

[273] S. M. SCHWARTZ, "HISTORIC ATTICS: HOW TO RETROFIT WITH

ENERGY & INTEGRITY IN MIND," BALL STATE UNIVERSITY, INDIANA,

MUNCIE, 2015.

[274] M. Beccali, M. Cellura, M. Fontana, S. Longo and M. Mistretta, "Energy retrofit of a single-family house: Life cycle net energy saving and environmental benefits,"

RenewableandSustainableEnergyReviews, no. 27, pp. 283-293, 2013.

241

[275] C. Rodrigues and F. Freire, "Integrated life-cycle assessment and thermal dynamic simulation of alternative scenarios for the roof retrofit of a house," Building and

Environment, no. 81, pp. 204-215, 2014.

[276] D. B. Crawley, J. W. Hand, M. l. Kummert and B. T. Griffith, "Contrasting the capabilities of building energy performance simulation programs," Building and

Environment, no. 43, pp. 661-673, 2008.

[277] D. Yan, J. Xia, W. Tang, F. Song, X. Zhang and Y. Jiang, "DeST—An Integrated

Building Simulation Toolkit Part Ⅰ: Fundamentals," Build Simul, vol. 1, pp. 95-110, 2008.

[278] S. H. Lee, T. Hong, M. A. Piette, G. Sawaya, Y. Chen and S. C. Taylor-Lange,

"Accelerating the energy retrofit of commercial buildings using a database of energy efficiency performance," Energy, vol. 90, pp. 738-747, 2015.

[279] M. C. Baechler, T. L. Gilbride, P. C. Cole, M. G. Hefty and K. Ruiz, "Guide to

Determining Climate Regions by County," Pacific Northwest National Laboratory, 2015.

[280] F. P. Torga, M. Mistretta, A. Kaklauskas and C. G. Granqvist, Nearly Zero Energy

Building Refurbishment, London: Springer-Verlag, 2013.

[281] "United States Census Bureau," 2015. [Online]. Available: https://www.census.gov/construction/chars/pdf/squarefeet.pdf. [Accessed 18 September

2016].

242

[282] E. Bostancioglu and B. Telatar, "Effect of Window Size on Residential Buildings’

Energy Costs," World Applied Sciences Journal, vol. 21, no. 4, pp. 631-640, 2013.

[283] N. Shukla,, A. Fallahi and J. Kosny, "Aerogel for Thermal Insulation of Interior

Wall Retrofit in Cold Climates," 4 April 2012. [Online]. Available: https://c.ymcdn.com/sites/www.nibs.org/resource/resmgr/BEST/best3_shukla.2.11.pdf.

[Accessed 19 September 2016].

[284] E. Kamel and A. Memari, "Different Methods in Building Envelope Energy

Retrofit," in Proceedings of the 3rd Residential Building Design and Construction

Conference, State College, 2016.

243

6) Chapter 6. Application of BIM in Energy Modeling

This chapter has been written as a journal paper and is already submitted for review.

244

Review of BIM’s application in energy simulation: tools,

issues, and solutions

1 2 Ehsan Kamel , Ali M. Memari

1Ph.D. Candidate, Department of Civil Engineering, Structural Engineering, Penn State University, 321 Sackett Building, University Park, PA 16802 (corresponding author). E- mail: [email protected] 2Professor, Hankin Chair in Residential Building Construction, and Director, Pennsylvania Housing Research Center, Dept. of Architectural Engineering and Dept. of Civil and Environmental Engineering, Penn State Univ., University Park, PA 16802. E-mail: [email protected] 6.1 Abstract

Building information modeling (BIM) initially emerged as a capability to transfer information and allow interaction or interoperability of various software tools used for architectural, structural, mechanical, and electrical design and building construction. Its utility from developing 3D models, structural analysis, cost estimation, and mechanical analysis has now expanded to other applications such as energy simulation. Multiple computer-aided design (CAD) tools exist that can be used as BIM tools to generate BIM files in different formats containing various types of building information. In addition, there are various building energy modeling (BEM) tools capable of importing these BIM files to perform energy simulation. However, such tools have various capabilities and limitations and need to be investigated and categorized in order to facilitate choosing a proper tool for design professionals in different phases of project and purposes. In addition, interoperability and data exchange issues between BIM and BEM tools should be

245 understood in order to find solutions such as developing proper corrective middleware tools to rectify them.

This paper reviews the challenges, issues, and shortcomings in BIM-to-BEM interoperability process (BBIP) by proposing a detailed classification for these issues and studying the available solutions. The paper also proposes a corrective middleware, which is developed using Python and modifies gbXML files prior to adoption in energy simulation to resolve the issues related to building envelope in BBIP. To do so, initially a review is presented on research studies focused on different types of BIM schemas such as

IFC and gbXML and energy simulation tools capable of reading these files such as Green

Building Studio (GBS), Ecotect, Integrated Environmental Solutions-Virtual Environment

(IES), and OpenStudio. In addition, some of the challenges in application of BIM for energy simulation such as interoperability issues, lack of standards, and lack of easy solutions for extending existing BIM schemas and available corresponding solutions are also reviewed. With the focus on building envelope in mind, three case studies are discussed to observe the challenges and issues with respect to BBIP using Revit, GBS, and

OpenStudio. Moreover, these case studies provide an opportunity to investigate the performance of the developed BIM file corrective tool using Python.

Keywords: Building Information Modeling, BIM, gbXML, Building Energy Simulation,

BEM, Revit, OpenStudio, Green Building Studio, GBS, EnergyPlus

246

6.2 Introduction

In order to study the shortcomings in BIM-to-BEM interoperability process (BBIP), it is important to understand the components contributing to the process. BIM tools, BIM files, and BEM tools are among the major components in the process. This section introduces these components, reviews different applications of BIM in engineering, discusses the benefits of application of BIM in BEM, and initiates the important topics discussed in each section of the paper.

Definition of Building Information Modeling (BIM) could be different depending on what is included in the model. For example, an information model could include information about the building geometry, envelope components, materials, costs, HVAC system, electrical systems, and thermal properties of materials. The U.S. national BIM standard

(NBIMS-US) defines BIM as “the act of creating an electronic model of a facility for the purpose of visualization, engineering analysis, conflict analysis, code criteria checking, cost engineering, as-built product, budgeting, and many other purposes.” [285]. Krygiel and Nies (2008) define BIM as “information about the entire building and a complete set of design documents stored in an integrated database” [286]. BIM is also defined by Smith and Tardif (2009) as “a mechanism to transfer from data into information to gain the knowledge that allows us to act with wisdom” [287].

Application of BIM in different area is expanding, as researchers realize the potential values BIM can offer. While application of BIM for Structural and energy analysis is reported with a frequently of 27% and 25%, respectively, its major use still seems to be for 247 faster development of 3D geometric models and 3D coordination with a use frequency of

60% [288]. Application of BIM is not limited to architects and engineers. There are also motives for homeowners, facility managers, contractors, and fabricators to use BIM [289].

Most of the important factors that lead to adopting BIM in a project are focused on automation in modeling process, improving the accuracy of construction documents, improving the communication among parties in design and construction process, automatic reflection of changes in all views after modifying one view, and reducing the field coordination problems [290, 291, 289, 292]. While most of the BIM application cases are dedicated to building design, equivalent emphasis has not been placed on some other areas such as use of BIM in energy modeling, which can be referred to as BBIP. A relatively comprehensive literature review on BIM and its applications shows that most resources are mainly focused on planning, design, construction, operation, and energy consumption, with archival publication focusing more on design and energy consumption [293]. A similar trend is also noted by researchers from different countries, meaning that although the construction companies are already using energy simulation tools, there is a void in the integration of BIM and BEM in a single tool, which can help avoid reentering of all the data that already exist in other models developed for the project [294, 295].

Application of BIM in building energy modeling offers multiple benefits as illustrated in

Figure 6-1. Contributions of BIM in energy-related modeling, simulations, and information is mainly about easier handling of data, which can lead to automation in energy modeling, better presentation of outputs, capability of storing and organizing new building data especially the real-time information to have an up-to-date energy model, and enhancing

248 existing libraries by adding new attributes to the normal energy simulation process. The explanation of these contributions are provided here in more details.

Automation of energy modeling

Storing & organizing Better presentation building's data(e.g., BIM of energy-related real-time data) outputs

Enhancing exisitng libriaries (e.g., adding extra attributes to materials)

Figure 6-1. Major contributions of BIM in building energy management

Automation of modeling process is one of the major benefits of application of BIM in energy simulation of buildings [292, 296] that could save time, lower cost, and reduce human error compared to the conventional energy modeling process, which includes developing a graphical model in a BEM tool using data related to geometry, material properties, equipment, and schedules. For example, adding an add-on to an energy simulation tool such as OpenStudio offers such an opportunity where it connects a

BIMserver to energy simulation tool to import the data related to the geometry, materials, window types, and thermal properties from IFC file [292].

The second major contribution of BIM in energy modeling is facilitating the output presentation in energy management systems [297, 298], especially when it comes to

249 computer tools without a GUI. For example, Jen and Vernatha (2016) studied a conceptual framework for a BIM-based Energy Management Support System (BIM-EMSS), which uses BIM models developed by Revit to perform a real-time energy simulation using eQuest by adoption of sensors and smart meters. The outputs can be visualized using the geometric data available in the BIM model, which allows the user to monitor real-time energy performance of different zones in a building [297].

Another major benefit for application of BIM is storing and organizing the energy-related building information. For example, real-time energy monitoring systems generate information with regards to the home energy consumption where such information need to be stored in an organized way under proper thermal zone, equipment, and building components. Energy-related data that can be recorded in a BIM model includes but not limited to energy consumption, temperature, and occupancy. An example of adoption of

BIM models in real-time energy monitoring systems is discussed by Alahmad et al. (2010) proposing a system that is a combination of a hardware component called Real-Time Power

Monitoring (RTPM) system and a software system called Real-Time Electrical BIM Model

(RE-BIM Model). The hardware components can facilitate fine-grained energy measurement of different appliances and sources within a house and transmit the data to the software component, which is a BIM model that can save the real-time energy-related data such as current energy consumption for a particular zone and load [299]. Integration of a BIM within a building energy model could also contribute in real-time monitoring of changes in occupancy, facility upgrades, and energy management strategies [290], to store their up-to-date and real-time data. For example, Woo et al. (2011) report application of

250

BIM in a building equipped with sensors whose data are linked to a BIM model using

SensorML, which is a standard schema to facilitate processing the data related to sensors and actuators [290, 300]. This way, the energy model is always updated based on the real- time data obtained from sensors, where the updated data can also be linked to an energy simulation graphical user interface or directly to the energy simulation engine in order to extract updated outputs.

BIM could also be helpful in enhancing the existing libraries concerning material properties, which is used in energy modeling. For example, existing libraries could provide thermal properties such as thermal conductivity of a material. However, certain projects might need further investigations about Life Cycle Assessment (LCA) of a building. BIM could be incorporated for use as a middleware in between the CAD tool and energy simulation tool to add extra attributes required for LCA, which might not be available in libraries of either the CAD tool or energy simulation tool. For example, the materials and systems that can be used in building envelope or sizing the HVAC system can be evaluated in advance to optimize the design and facilitate the process of assessing different design alternatives [301].

Different components contribute to BBIP and studying each component separately could lead to a better understanding of challenges, issues, and corresponding solutions. The overview of BBIP illustrated in Figure 6-2 shows three major components involved in the process. The discussion presented here is focused more on the second and third links of the chain shown in Figure 6-2, illustrated as BIM files and BEM tools.

251

Figure 6-2. Overview of BBIP

BEM tools could be comprised of two components including a Graphical User Interface

(GUI) and a simulation engine. GUIs such as OpenStudio, BEopt, DesignBuilder, and eQuest facilitate the energy modeling process by providing graphical interface for users.

However, the second component, that is the energy simulation engines such as EnergyPlus and DOE-2, works behind these tools. These two components as well as reported tools that have used such tools in a BBIP are reviewed and summarized in the next section. Mapping building information to energy simulation tool and the reverse process of mapping the results back to the GUI are among the steps where some of the issues related to BBIP occur.

BIM file schemas are the second link of the chain shown in Figure 6-2. Two of the most prevalent types are IFC and gbXML files, which could include different types of information about buildings. Data structures used in IFC and gbXML are referred to as bottom-up and top-down structure, respectively. It is important to study the types of information these BIM file standards are capable to transfer and store in order to have a better understanding of issues and challenges related to BBIP since the BIM files are directly involved in the process. BIM file formats and their properties are reviewed in more details in the second section of this paper followed by a section dedicated to a review of issues and challenges in BBIP.

252

One of the major objectives of the discussion presented here is proposing a detailed classification of issues and challenges in BBIP that helps to guiding future research works in this field. Multiple steps and components in BBIP are susceptible to experience issues causing errors or inaccuracies such as writing data from a CAD tool to BIM file, mapping data from BIM file to a readable file for a BEM tool, Mapping data from a BEM GUI to a readable file for a simulation engine, and mapping data from simulation engine to BEM

GUI. Different studies experienced different issues and challenges occurring at different steps of BBIP [302, 303, 304, 305, 54]. The discussion presented also includes a proposed detailed classification for these challenges and issues based on review of related research studies followed by identification of existing solutions adopted by other researchers.

Review of research studies on BBIP reveals that there is a need for developing middleware, which works between BIM and BEM tools in order to resolve the issues related to BIM files occurring at this stage and it is also adopted by other researchers [305, 306, 54].

Therefore, three case studies are performed in this paper using prevalent tools such as

Revit, OpenStudio, EnergyPlus, Green Building Studio (GBS), and gbXML file format in order to provide the opportunity to evaluate the performance of a suggested corrective tool developed in Python, which is focused on the issues related to information on building envelope. It can also provide a good opportunity to observe the potential issues in BBIP related to the prevalent tools used in these case studies.

253

6.3 BEM tool’s Graphical User Interface (GUI) and energy

simulation engine

There is a difference between energy simulation GUI and energy simulation engine. As shown in Figure 6-3, Energy simulation interfaces only facilitate the energy modeling process by providing more graphical interfaces for inputs and outputs, including developing the geometric model of the building or detailed graphs on energy consumption data. These interfaces such as OpenStudio, DesignBuilder, BEopt, or eQuest do not perform the simulation inherently and there is the need for integration of energy simulation engines such as EnergyPlus or DOE-2, which perform the analysis based on different mathematical tools and thermodynamic equations. Other tools such as eQuest or TRNSYS are also categorized under energy simulation engines. Currently, none of the major energy simulation engines such as EnergyPlus and DOE-2 is capable of direct import of BIM files such as gbXML and IFC directly from CAD tools. However, there are energy simulation

GUIs that have the capability of using BIM files such as OpenStudio, IES-VE, GBS, IDA

ICE, RIUSKA, and Ecotect, which the latter is discontinued. There are also file generators such as Space Boundary Tool (SBT) that imports a BIM file and generates a text file, which is a suitable format for energy simulation engines such as EnergyPlus [307, 308]. GUIs have built-in functions that could be coded in different languages such as C++ and Ruby, which perform the file conversion from BIM files to other file formats suitable for energy simulation. Table 6-1 shows how common energy simulation tools are aggregated under either energy simulation GUI or simulation engine.

254

Figure 6-3. Interoperation between energy simulation GUI and engine

Multiple energy simulation tools discusses subsequently to various depth are shown in

Table 6-1. GBS that only needs costless activation of Autodesk ID is a cloud based energy simulation tool integrated with Revit through Autodesk to facilitate energy modeling in this CAD tool [309]. GBS uses DOE-2 for energy analysis and the outputs can be detailed to present annual the amount of heat transfer through building components such as walls and roofs (Figure 6-4). The energy model can be based on conceptual mass or building element depending on available design details, so the energy analysis results can be only a rough evaluation and not detailed results. Moreover, depending on type of the model

(architectural model or MEP model) the export category can be set to rooms or spaces, respectively [309]. An example for application of such energy simulation tools is a case study reported by Krygiel and Nies (2008), which shows application of gbXML file generated by Revit in performing an energy analysis to evaluate two façade systems. Four

255 different energy simulation tools are reviewed in their study, including Ecotect, IES-VE, eQuest, and Green Building Studio (GBS), which was used to calculate the energy loads and assess different designs [286].

Figure 6-4. Example of energy analysis output generated by GBS

Considered here as the second tool, Modelica is another language that can be used for building thermal analysis capable of being integrated with BIM for energy simulation and importing required data. Researchers have developed libraries called ModelicaBIM library for use in Modelica for BIM-based energy analysis [310, 311]. Kim et al. (2015) developed a library for Modelica to facilitate the BBIP, and its performance is compared with LBNL

Modelica buildings library developed by Lawrence Berkeley National Laboratory. The

BIM library used in their study could also perform a component-level energy simulation to present the amount of heat transfer through envelope components such as walls, roof, and floor. The outcomes show the feasibility and capability of using BIM and Modelica in automatic BEM using BIM, especially in design stage of buildings [310]. 256

Understanding the process of energy simulation and different components in inputs or outputs could contribute to a better understanding of what is needed in a BIM file for faster, more accurate, and flawless energy simulation. In general, an energy simulation tool requires information on building structure (e.g., geometry, construction materials, and thermal zones), HVAC systems (e.g., heating, cooling, and ventilation system), weather data, and simulation properties (e.g., run period). Other studies use different categories for energy simulation including loads, meters, pumps, circulation loops, plant equipment, air systems, zones, additional systems and zones, economic, and reports [312]. Other major categories including building geometry, weather condition, HVAC system, internal loads, and operating strategies and schedules are also introduced by other researchers [307], which suggest the general need for BIM models to consider addressing all such data in future efforts. The main goal in common energy simulation is to evaluate the energy consumption due to different components such HVAC, lighting, and appliances. The energy consumption due to lighting and appliances could be obtained based on schedules set in the program, which is a much easier process compared to the load calculation in

HVAC system. The energy consumption due to HVAC system is acquired based on the desirable temperature within the house and the energy needed to maintain that temperature.

Different energy simulation engines use different approaches to calculate the space load.

For example, DOE-2 and EnergyPlus use weight factor and heat-balance-based methods, respectively [307]. In heat-balanced-based approach, which is based on the first law of thermodynamic, for each building surface such as walls, roof, floors, and windows, a control volume will be considered for outside surface, inside surface, and inside air, for

257 which the heat balance will be calculated. The final goal in these calculations is to find the temperature at which the heat balance is satisfied [313]. Therefore, it is important to make sure all the related data for these calculations transferred through a BIM file to a BEM tool are valid and adequate.

Table 6-1. Categories of energy simulation tools [314, 307, 315, 308, 289, 316, 310]

Energy Simulation Independent Energy Simulation Energy Simulation Engine GUI only Engine with Graphical Interface OpenStudio Ecotect DesignBuilder TRNSYS Hevacomp EnergyPlus IDA ICE Simergy ApacheSim (used in IES VE) BEopt EDSL Tas GBS Modelica language (can be used with eQuest DOE-2 graphical interfaces such as Dymola) RIUSKA

There are alternative methods used in BBIP and a CAD tool is not the only option adopted by researchers. Other BIM generating tools or middleware have also been developed that interoperate between CAD and BEM tools by converting a BIM file in a format such as

Industry Foundation Class (IFC) exported from ArchiCAD 13 to Input Data File (IDF), which is an input file for EnergyPlus in order to automate the energy modeling process

[317]. One may use more innovative methods to generate a BIM file such as image processing or laser scanning point clouds, which are automatically textured with thermographic and RGB images [318] to generate a thermographically textured as-built

Green Building XML (gbXML) BIM file automatically. The 3-D thermography-based method can also be adopted for updating the thermal properties of building components in order to use the as-is properties such as R-value of insulation material, which could be

258 changed due to degradation over time. The results can be mapped into the BIM file (e.g., gbXML file) to update the old attribute within the file [319].

As shown in Figure 6-5 that expands on Figure 6-3, the difference between a regular BEM tool and a BEM tool with the capability of reading a BIM file is the extra step shown as mapping BIM file data to a readable file for BEM tool. There are built-in codes in BEM tools that perform mapping data from BIM file into a proper format. For example, in

OpenStudio source code, there is a module developed in C++ dedicated to translating BIM files to OSM file, which is the format for OpenStudio files. These functions should perform properly in order to be able to read all the required data from BIM files and map them properly to the BEM tool file format. Next section will discuss the properties and attributes within BIM files to have a better understanding of the type of data provided by BIM files.

Figure 6-5. Additional step added to integrate BIM in BEM process

259

The information provided in this section revealed important steps in BBIP, which could be sources of errors and inaccuracies. Mapping data from BIM file to a readable file for BEM

GUI, mapping data from GUI to energy simulation engine, and mapping data back to the

GUI are such steps. BIM files generated by BIM tools such as CAD tool, initiate this process (as shown in Figure 6-5), which will be discussed in more detail subsequently.

6.4 Types of BIM file schemas and their properties

Different BIM file schemas have been developed for use by BIM authoring tools such as

CAD tools. However, two prevalent third generation models are actually used more often for energy simulation purposes. The Industry Foundation Class (IFC) and Green Building

XML (gbXML) deployed by BuildingSmart and Green Building Studio Inc., respectively, are two of the most comprehensive BIM file formats, each with its own features. Both of these schemas are capable of providing parts of the required information needed for energy simulation. The IFC is developed by an international organization formerly known as the

International Alliance for Interoperability (IAI) that was renamed to BuildingSmart in

2005. The current version IFC4 supersedes an earlier version referred to as IFC2x4 released by BuildingSmart in 2013. Two additional addendums were released in July 2015 and 2016 with improvement in multiple areas. One of the improvements is related to energy and performance analysis focused on space boundaries, adding spatial zones, external spaces, and shading devices [320, 321]. These third generation models can provide information during design, construction, and operation time span of a building as opposed to the second and first generation models that could only provide data on geometry or a limited domain

260 such as lighting or building’s LCA [322]. The gbXML format can include information on building, zones, surfaces, construction types, fenestrations, and the environment around the building, which is a comprehensive coverage of information [323]. However, it has been noted that not many simulation tools can comprehend the data related to HVAC systems provided by gbXML [324].

The differences between capabilities of two formats of major BIM schemas, i.e., IFC and gbXML, in terms of both general properties and data related to energy simulation are summarized in Table 6-2. The current versions of both file formats use XML but with different approaches. Top-down and bottom-up data structure approaches are used for gbXML and IFC, respectively. Both provide material properties, limited data for HVAC systems, and thermal zone data; however, the data related to locations are only provided in gbXML file format. One of the features of IFC file format is the capability of defining

Information Delivery Manual (IDM) and Model View Definition (MVD) as subsets of IFC file targeting specific areas of a project such as structural design or energy analysis. The case studies discussed in this paper will investigate the data structure and attributes within gbXML file in more details. The shortcomings of data exchange using gbXML file in energy simulation will also be discussed later through case studies.

For different software or particular stages of a project, there are certain sets of data within a BIM model that need to be extracted as opposed to the whole BIM file data. The subset could include the required information for energy simulation, including HVAC system, operation schedules, and control parameters (e.g., supply air temperature and air volume

261 flow rate). IFC BIM schema has this capability to extract a subset of information for a particular application by using IDM and MVD [308]. There are also standards for minimum content of a model in BIM such as model view definition (MVD) standards for

IFC file that defines the minimum required information for different uses [325].

Application of IDM is needed to define the required types and data format for energy analysis and appropriate level of details of information [326]. While the current BIM file formats cover vast amount of information related to a project, there are still missing pieces of information for particular simulations that should be added to the existing standard for their extension; this can be done, for example, by using IDM and MVD for IFC BIM standard [323]. The International Energy Agency in Buildings and Communities programme (IEA-EBC) Annex 60 project reports on an MVD proposed by Pinheiro et al.

(2016) to facilitate integration of BIM models and energy simulation. The energy simulation tool used in that project is Modelica, which is an object-oriented language to model systems with multiple components such as mechanical, electrical, thermal, and control systems [308, 327]. The development of MVDs focused on energy simulation is an international effort led by initiatives such as Architectural Design to Building Energy

Analysis (in the United States) and Nordic Energy Analysis (in Norway) [328]. The former is developed by multiple organizations, including US General Services Administration and

Autodesk. The Exchange Requirements (ER) identified in this MVD are categorized into two major input groups (including the data related to architecture and mechanical systems) and a group of output requirements [329]. In addition, an IDM for BIM based energy analysis at concept design phase is developed by BuildingSMART Alliance (BSA) [330].

262

Table 6-2. Comparison between gbXML and IFC based on literature review [322, 314, 331, 292, 303, 301,

308, 323]

Characteristics gbXML IFC Presentation of Building’s Only rectangular geometry Any geometry Geometry

XML (Extensible Markup IFC, PKZIP, and XML (Extensible Data structure Language) Markup Language)

Top-down approach with Bottom-up approach with Data structure approach relatively more complex relatively more straight forward representation representation

Mostly energy simulation Different domains such as building Domain of application domain construction to building operation

Capability of defining thermal Yes Yes zones

Location Yes No

Standard for minimum content Yes – There is MVD standard for for a certain type of model and No IFC and IDM capabilities using subsets

Material Thickness Yes Yes

Limited data related to HVAC Yes Yes system

The two main data structure approaches used in prevalent BIM schemas are explained as top-down and bottom-up approaches for IFC and gbXML by Jalaei and Jrade (2014). The

“top-down” approach is defined as a relative complex schema with large file size that is hard to be programmed and used in software application; however, it can trace back the semantic changes in a value within the schema. On the other hand, the “bottom-up” approach is flexible, open source, and has a relatively straightforward data schema [301].

Jalaei and Jrade (2014) also note that some of data might not be transferred through an IFC

BIM model such as location, construction assignments, assumptions for lighting,

263 equipment and people loads, airflow data, and units. It could be explained due to the limitations in tools, which read the BIM file as opposed to shortcomings in BIM files.

Otherwise, the IFC schema might have the capabilities of transferring these types of data

[301, 326]. In their particular research, the gbXML could transfer data related to shape, areas, volumes, location, construction assignments, building type (e.g., residential or commercial), and building services [301].

Besides the most prevalent BIM standards discussed above, researchers can employ other

BIM exchange data formats. El Asmi et al. (2015) categorized the available data formats into two major groups of open and proprietary. EXPRESS data language and eXtensible

Markup Language (XML) can be used to develop these formats. Data formats such as

STEP, CIS/2, and earlier versions of IFC are EXPRESS-based. On the other hand,

CityGML, landXML, gbXML, aecXML, and ifcXML are among the XML-based data formats. Examples of proprietary BIM data formats could be RVT (Revit), DWG

(AutoCAD), DGN (Bentley Systems), and DXF [323].

The BIM file schemas used in the BBIP may be evaluated based on their capability to save and report all the attributes and required information for energy simulation. In other words, the BIM file standard should be capable of defining proper attributes for each required data such as schedules, material properties, and building’s geometry. Then, the generated BIM file shown in Figure 6-5 could be used directly in the BBIP. Figure 6-6 and Figure 6-7 show how different attributes are defined, respectively, in gbXML and IFC file standards.

Figure 6-6 partially shows an example of gbXML file containing information concerning

264 units, location, building types, and different surface properties. Figure 6-7 shows the data structure in an IFC file and different glass and wood materials are defined for a door component. Based on the studies reviewed in this paper, it seems there is still a need for comprehensive research on capabilities of these BIM file standards and their interoperability properties with different BIM and BEM tools in order to identify all the potential shortcomings.

Figure 6-6. Partial presentation of a gbXML file containing information for energy simulation

265

Figure 6-7. IFC standard’s data structure

To summarize, this section explained two of the most prevalent BIM file standards and their properties and all the components and steps within BBIP. The next section proposes a detailed classification for issues and challenges that might be experienced through BBIP based on different steps and components in BBIP followed by a review of related research studies and adopted solutions.

6.5 Review of Identified Challenges and Issues in BBIP

In order to have a better understanding of challenges and issues in the BBIP, a classification of different components in this process is suggested as shown in Figure 6-8. This classification can also be used in order to target a certain area of BBIP. The overview of this classification was already presented in Figure 6-5 showing major components including BIM tools, BIM files, and BEM tools. Corresponding components in Figure 6-8 are numbered as 1, 3, and 4. The interaction between these components could also

266 contribute to issues and challenges, as well as potential sources, all numbered in the figure and explained as follows: #1: BIM tools such as Revit and ArchiCAD; #2: the process of mapping building information to a BIM file; #3: BIM file under a certain file standard such as gbXML or IFC; #4: GUI in the BEM tool such as OpenStudio and DesignBuilder; #5: the process of mapping data from BIM file to a readable file for BEM tool; and #6: the process of mapping data from GUI to a readable file for simulation engine such as

EnergyPlus or DOE2. The interoperability issues are typically due to the components numbered as 2, 5, and 6, which are also identified by other researchers [307, 314, 48, 332,

315, 293].

General challenges in broader point of view that could be related to any of the components identified in Figure 6-8 in BIM exchange including managing, provenance, required quality, detail, reliability, transparency, security, and accessibility of information are reviewed by other researchers such as Smith and Tardif (2009). Two of the BIM exchange challenges reviewed in their book is more common in other studies, which are satisfying different views of stakeholders and interoperability between different tools [287].

Although lack of a standard for data exchange process is identified by some researchers

[308], nonetheless there are examples of building information exchange standards such as agcXML or Construction Operations Building Information Exchange (COBIE). The outcome of agcXML project is used for information exchange in different phases of a project such as design and construction [287].

267

The reviewed literature mostly includes all three major components illustrated in Figure

6-2. While BEM tools and BIM files are reviewed in this paper, multiple CAD tools with

BIM authoring capabilities in BIM to BEM process are reviewed in literature. For example,

Epstein (2012) reviewed some of these tools, which are referred to as BIM CAD programs and illustrated as component #1 in Figure 6-8, such as Allplan (by Nemetchek), ArchiCAD

(by Graphisoft), Microstation (by Bentley Systems), Revit (by Autodesk), and

Vectorworks (by Nemetchek). Epstein also reviewed a case study in Greece adopting

ArchiCAD to use BIM for energy modeling using IFC and EnergyBuild as the energy simulation engine [333].

Figure 6-8. Overview of BBIP and the potential sources of issues

268

Two of the most prevalent issues associated with components #2 and #3 shown in Figure

6-8 could be categorized as file-related issues. According to [334], the file-related BIM interoperability issues can be studied under three different categories including file and syntax level, visualization level, and semantic level. Integration of BIM and energy simulation tools can include all three levels of interoperability issues. For example, in a

CAD tool, it is important to visualize and illustrate the thickness of a wall, while the energy simulation tool needs the thermal properties of a wall component and only use the centerline, which can cause a gap between two different components at intersections [334].

It explains one of the issues identified by researchers, which is the difference between the structural volume and the analytical volume need for energy simulation. The analysis performed within an energy simulation tool might require different information from the model developed in the CAD tool, which develops the structural volume. For example,

IES-VE requires two different volumes of the building in order to perform energy simulation and air circulation computations. For air circulation, IES uses inner volume while the energy and thermal evaluation of the building uses analytical volume, which is bounded by the center plane of building envelope components [303]. The information provided in BIM schemas can be highly comprehensive and make the energy or structural modeling a rigorous process, which could be considered as another challenge in working with BIM files (#3).

Besides energy simulation engine, there are built-in codes within energy simulation tools that work behind the GUI and are responsible for interpreting the imported BIM files discussed in previous chapters. This component is illustrated by #4 in Figure 6-8, but it

269 could be another source of issues and also include both processes #5 and #6. The ability of an energy simulation tool in importing a BIM file could be a challenge since some attributes in a single BIM file might not be supported by a certain energy simulation tool. For example, Attributes such as material and window type are available in gbXML format; however, unlike the material attribute, the latter (window type) might not be compatible with EnergyPlus [332] or IES-VE, which does not import the construction material or mechanical data exported by Revit [326]. These issues are associated with the process #5 illustrated in Figure 6-8.

Energy simulation tools shown in general as #4 in Figure 6-8, could be a separate source of issues and inaccuracies. Different techniques and methods in various energy simulation tools can lead to different major outputs such as heating and cooling loads. For example, the heating and cooling loads obtained from energy simulation tools using BIM models could be different due to the difference in calculation methods [312]; for example, Ecotect uses worst design annual load, while IES uses the worst monthly cases [301]. These differences could exist in component #4 shown in Figure 6-8 and could cause errors in the

BIM-to–BEM interoperability process, thus it is important to understand the features of a

BEM tool used in the process.

Losing data through information exchange between tools is one of the most reported issues

[335, 336]. Accordingly, not all the required data can be transferred through BIM model to the energy simulation tools; therefore, missing data should either be added manually or be automatically generated by simulation tools used. It could be caused by different

270 components shown in Figure 6-8 including the BIM tool capabilities in exporting building information (#1). It could also be due to the issues in mapping process (#2), which means the data might be available but not transferred to the BIM file properly. Moreover, the BIM file might not be able to save the data (#3) nor define any attribute for a specific piece of information. The same issue could also happen during mapping data within BEM tool, illustrated as #5 and #6. Therefore, almost all the components in the BIM-to–BEM interoperability process could contribute to data loss. The studies on interoperability between other BIM models such as IFC and energy simulations tools such as DOE-2 show some lack of information such as the desired run period and building schedule [312].

Lack of required data is another challenge that is similar to data loss; however, it is easier to handle as long as all the required data are identified and provided by the initial BIM tool.

There are also other solutions such as linking the model to a BIM tool as a hub, which is identified as multi-model framework as opposed to extending a single BIM schema such as IFC. The model suggested in a study that covers data related to multiple areas is called energy-extended BIM (eeBIM), which could be a link between a simple CAD file and an energy simulation engine such as EnergyPlus [337] to add the required data, which are either missing or have not been developed in the first place such as schedules.

Even after performing the energy simulation using BIM, there will still be challenges concerning storing and presenting the data to benefit from BIM. These include the need for storing the generated data in the BIM file and updating them after conducting a simulation

(e.g., energy simulation), and the desirability for development of easy and understandable

271 graphical presentation of the outputs for various users by using BIM [299, 338]. Sanguinetti et al. (2014) reviewed a case study using IFC and IDF files, where energy performance outputs are visualized using automated generation of thermal zones in Revit, which eventually provides detailed thermal flows of building envelope components. They identified lack of data for energy simulation at the initial stages of design and quantifying the performance of design and providing feedbacks as two challenges [339].

The summary of the reviewed research studies on application of BIM in energy simulation is presented in Table 6-3. It can be observed that ArchiCAD and Revit are the most common BIM authoring tools used in these studies, and depending on the research topic, either IFC or gbXML BIM file schemas are adopted. Energy tools vary between conventional tools such as OpenStudio, EnergyPlus, GBS, and eQuest and other tools such as Modelica, COMETH, and EnergyBuild, some of which are international tools not necessarily commonly used in the U.S.A. It is also observed that most of the identified challenges and issues occur during the mapping process, which could be categorized under interoperability issues. As examples of such challenges, the following may be noted: the

BIM tool might not transfer all the information in a model, the BIM file might not be able to save all the information properly, the BEM tool might not be able to read all the information from BIM file, and finally the information might not be mapped and transferred properly to the BEM and energy simulation engine’s file format. However, mapping data to energy simulation engine shown as #6 in Figure 6-8 is not discussed properly in the literature and can be an area in need of more research.

272

6.6 Review of Identified Solutions Adopted by Researchers

Related to BBIP

Besides multiple sources and types of challenges and issues related to BBIP discussed, one needs also explore solutions such as developing middleware works between BIM and BEM tool, manually adding missing data, extending the model, and semantic enrichment of BIM schema. This section reviews literature that suggests some solutions as summarized in

Table 6-3.

To address the challenges related to mapping data to a certain BIM file format (#2) and other issues related to the BIM file itself (#3), methods have been suggested [304] to enable revising such information prior to end use, an example being Interpreted Information

Exchange (IIE) that is focused on revising IFC models for structural analysis tool. In general, it is important to keep the original and initial information intact through transferring data back and forth between different software packages while using BIM, which has led to the emergence of the concept of “seamless data exchange” during the data transfer [303]. Developing a middleware that works between the BIM tool and BIM file is another solution to resolve the issues illustrated as #2 in Figure 6-8. There are ongoing efforts toward interoperability of different CAD tools and energy simulation tools such as

Revit and OpenStudio. In order to overcome the existing issues of the process, a middleware has been developed to read the IFC file from Revit and use the BIMserver and

Query Generator to retrieve the data and extract the required information, respectively. The main purpose is to resolve the issues with building’s geometry data and provide a corrected

273 file for energy simulation in OpenStudio [54]. As an example for the BBIP, Salakij et al.

(2016) developed an energy simulation tool using Matlab called Building Energy Analysis

Model (BEAM) capable if reading gbXML BIM file; it is reported that BEAM could lead to less data loss compared to IDF file used for simulation tools such as EnergyPlus [305].

Uncommon challenges such as modeling uncommon surfaces such as curved surfaces and components in a building and transferring related information through a BIM model is another challenge in application of BIM in energy modeling. Therefore, some studies have suggested using middleware tools, which use methods such as mesh planarization algorithms to transfer information from a CAD tool to EnergyPlus [306].

To resolve the issues related to BIM files (#3), a general solution is addressed by El Asmi et al. (2015) to address the shortcomings of BIM files in terms of lacking required attributes for energy simulation. They identified two major methods for this purpose in the IFC data format including model extension and semantic enrichment. Model extension could be burdensome because new concepts, attributes, and relations need to be added to the model to be able to add a new element. However, using prebuilt extension capabilities in subclasses such as IfcPropertySet, IfcProxy, IfcRelationship seems to offer a less effort intensive easier approach for model extension. On the other hand, it should be noted that these prebuilt capabilities do not add semantic information to the IFC model. As an alternative approach, one may use semantic enrichment such as International Framework for Dictionaries (IDF) and Web Ontology Language (OWL). IDF helps the interpretation of information within an IFC model and OWL, which is a set of languages for authoring ontologies, helps the semantic enrichment of IFC. There are also already developed open

274 source converters, which convert XML schema to RDF and OWL [323]. Other examples are available in the literature that show the data related to HVAC systems and internal loads are not provided in detail in a BIM file format such as IFC, which is an important challenge in automating of energy modeling [302]. Accordingly, extending the current IFC schema in order to include detailed information about the required data for an energy simulation might be a solution to this issue.

In order to reduce the risk for issues associated with data exchange between BIM files and energy simulation tools, which typically happen during the process #5 shown in Figure 6-8, researchers propose BIM API method instead of using conventional BIM schemas. The

BIM API method, which could be a rigorous process, can be defined as developing an independent BIM file directly from BIM tools such as Revit, ArchiCAD, or Bentley using

Application Program Interface (API) [310, 311]. This solution is provided in order to resolve the issues concerning BIM files shown in Figure 6-8 as #3.

Missing data could occur at any step through the BBIP shown in Figure 6-8. Different measures could be applied as a solution for missing data issue. A semi-automated energy modeling process suggested by Lawrence Berkley National Laboratory (LBNL) is an example of adopting BIM IFC schema to develop a proper output for EnergyPlus.

However, it is not fully automated because it does not include all the required information such as data related to HVAC, which need to be added manually or using other tools.

ArchiCAD is used to develop the IFC file. The IDF file for EnergyPlus is converted to

DXF format only for prior inspection, and then the building energy modeling is performed

275 based on the model that was obtained from tools such as BIM for providing the required data related to other areas such as mechanical systems and schedules [340, 56]. Other solutions are adopted to resolve the issue concerning missing HVAC system data in a BIM model. For example, El Asmi et al (2015) developed a middleware between IFC BIM model and COMETH simulation engine, which is a tool developed by the French Scientific and Technical Center for Building. El Asmi et al (2015) used the Model View Definition

(MVD) approach to extract the required data from IFC model, and to provide the missing data related to the HVAC system, they used IfcPropertySet, which is the prebuilt extension capability of IFC model [323]. Different tools in the BIM–to-BEM interoperability process might develop proper error messages according to the corresponding missing piece of information; but that is not the case in all simulation tools and proper warnings and error messages might not be developed [54].

To summarize, this section reviewed the adopted and proposed solutions by other researchers related to BBIP issues and challenges, as reflected in Table 6-3. Methods discussed include developing a middleware, manually adding the missing information, extending the existing models to include the missing data, and semantic enrichment of BIM schemas. A new middleware focused on issues related to building envelope is also adopted and discussed, followed by evaluation of the developed tools through three case studies.

276

Table 6-3. Summary of research studies on using BIM in energy simulation

BIM BIM Reference file Energy Tool Description of the study and adopted solutions CAD tool schema

Epstein (2012) A case study in Greece adopted BIM to BEM process ArchiCAD IFC EnergyBuild [333] as part of the project

Development of an add-on to import the data related Ramaji et al. - IFC OpenStudio to geometry, materials, windows types, and thermal (2016) [292] properties from IFC file.

Development of a middleware, which use BIMserver

and Query Generator to extract required data from Yu (2014) Revit IFC OpenStudio IFC file. Users add missing data manually and a [54] python script convert the file to a proper format for

CONTAM for air distribution analysis.

Building

Energy

Analysis Developed an energy simulation tool using Matlab, Salakij et al. - gbXML Model which is capable of reading gbXML file to perform (2016) [305] (BEAM) the energy analysis.

developed by

Matlab

Krygiel and Energy analysis using BIM is performed to evaluate Nies (2008) Revit gbXML GBS two façade systems. [286]

277

BIM BIM Reference file Energy Tool Description of the study and adopted solutions CAD tool schema

Kim et al. The study is based on integration of Modelica to

(2015) & perform energy analysis, ModelicaBIM library to

Jeong et al. - - Modelica provide required data needed from BIM file, and

(2014) [310] using BIM API to retrieve data from conventional

[311] BIM tools such as Revit and ArchiCAD.

O’Donnell et A semi-automated method is adopted to add the al. (2013) & required data for energy simulation to the IDF file Bazjanac ArchiCAD IFC EnergyPlus generated from IFC file, prior to the energy analysis (2008) [324] in EnergyPlus. [56]

The study is focused on application of MVD in

obtaining the required data for energy simulation El Asmi et al. - IFC COMETH (e.g., HVAC system data), which are missing through (2015) [323] the process and using IfcPropertySet to add the data to

the IFC file

Adopting mesh planarization algorithm to divide the Santos et al. - - EnergyPlus curved surfaces into flat panels and exporting the (2017) [306] required data from CAD tool to EnergyPlus.

Karen and Generation of thermal zones is automated using the

Douglas Revit IFC EnergyPlus data obtained from Revit, provided in IFC file and the

(2014) [339] outputs obtained from EnergyPlus are visualized.

278

BIM BIM Reference file Energy Tool Description of the study and adopted solutions CAD tool schema

Dimitriou et Development of a gbXML editing tool to provide the

al. (2016) Revit gbXML EnergyPlus data, which lacks for energy simulation in EnergyPlus

[341] prior to generating the IDF file.

Garcia and Development of a corrective tool for modifying the

Zhu (2015) - gbXML eQuest gbXML file and converting it to DOE-INP file for use

[342] in eQuest.

A case study to review the feasibility of optimization

in design by linking BIM to energy simulation tool.

Egwunatum et Faster, more accurate, and detailed outputs - IFC IES VE al. [298] concerning energy consumption, airflow analysis,

visualization, and daylight analysis were among the

benefits observed in the study.

Automation of Building Performance Simulation

(BPS) is studied using different tools in order to Somboonwit Revit & GBS, DOE-2, gbXML investigate the interoperability between them. Kinetic et al. [296] Dynamo eQuest PV façade (KPVF) is modeled and issues such as

misplaced and distorted geometry is observed.

6.7 Review of three case studies in BBIP

The state-of-the-art review related to the BBIP revealed different tools that contribute to this process, their capabilities, and the known issues, challenges, and available solutions.

For a better understanding of BIM files and BEM tools reviewed, their file formats, and to

279 identify the interoperability issues in more details, it was decided to model a residential building in Revit and obtain the energy analysis results through three different ways. Such file-related issues are identified and shown as #2 in Figure 6-8. Moreover, these case studies can provide an opportunity to evaluate the performance of a middleware corrective tool developed using Python, which resolves the issues related to building envelope in gbXML file format. In addition, multiple areas in the BBIP shown in Figure 6-8 have the potential for occurrence of issues and challenges. However, not all of the components such as the stage of mapping data to simulation engine, which is illustrated in Figure 6-8 as #6, are studied sufficiently. These case studies provide an opportunity to review this component and evaluate potential issues and challenges. The overview of the whole process used in these case studies is shown in Figure 6-9.

Figure 6-9. An overview of the process adopted in case studies

Figure 6-10 shows the schematic process of these three case studies separately, which provides a good opportunity to observe data exchange between a CAD tool and multiple energy simulation tools to study the interoperability issues and recommend solutions for each case. Revit, GBS, OpenStudio, EnergyPlus, and gbXML files are among the components in the BBIP reviewed in the literature, as reflected in Table 6-3, which is used in these case studies to identify the common issues in the BBIP. Energy simulation tools such as GBS and OpenStudio are also responsible for some additional tasks including performing gbXML to IDF file conversion. Such file conversions are used to map the data

280 within gbXML file (generated by Revit) to IDF file. IDF includes all the required information needed for energy simulation in EnergyPlus. Other BIM file formats such as

IFC have also been studied by researchers in relation to the development of IFC to IDF convertors in order to resolve the interoperability issues and improve the BIM-to-BEM automated interoperability process [343]. Some of these research studies are not limited to

IFC file conversion. They also integrate the convertor with the energy simulator (e.g.,

EnergyPlus) and use the outputs to visualize the results [344].

The first case study uses GBS within Revit. GBS gives rough results for building’s energy consumption and requires activating a free Autodesk account. This energy simulation returns a rough estimation of the whole building energy consumption using DOE-2, and it is more beneficial in design phase of the building in order to come up with efficient design and construction methods and materials in terms of energy consumption. Therefore, the needed input data are not as detailed as other energy simulation GUIs such as OpenStudio.

In second case study, the gbXML file is exported from Revit (first file conversion) to be used in OpenStudio, which is then used as an energy simulation tool GUI to read the BIM file and convert it to IDF (second conversion). The final energy simulation is performed by EnergyPlus as the energy simulation engine within OpenStudio.

In the third case study, the IDF file is generated using GBS to be used directly in

EnergyPlus. Case study #3 is based on three different tools including a BIM tool (Revit) and two energy simulation tools (GBS and EnergyPlus), as illustrated in Figure 6-10. Revit produces an energy model (first conversion) for GBS and a gbXML file (second

281 conversion), and then converts it to an IDF file for EnergyPlus (third conversion). The three conversions in file formats provide a good opportunity to study interoperability issues in the BBIP. The third file conversion is referred to as #6 in Figure 6-8, which is not so well covered in literature.

Figure 6-10. Process of three different approaches

6.7.1 Modeling Process in Revit for Three Case Studies

Figure 6-11 illustrates the plan and the 3D view of the Revit building model that was simplified by removing the bathroom and mechanical room and reducing the number of windows. The properties of the modeled house in Revit are presented in Table 6-4. The construction properties and envelope components are selected from Revit library, and four separate spaces are assigned to the living room and three bedrooms. In order to have separate thermal zones, four zones are also assigned to these areas, separately. Space is defined as any enclosed area bounded in at least six surfaces. OpenStudio and accordingly,

EnergyPlus will be looking for six surfaces for each space and will generate a warning if there are less than six surfaces. Unlike severe and fatal errors, this warning does not stop the simulation process; however, it shows there is an issue with reading the BIM file and

282 surfaces defined in it. On the other hand, zones are related to the HVAC system and define how the mechanical system is distributed and whether or not the area is heated/cooled.

Therefore, both zones and spaces are assigned to these areas to make sure the outputs are as detailed as possible. Spaces in Revit also include different attributes, which are needed for energy simulation such as number of people, lighting load, and occupancy schedule.

Other than architectural components and properties related to each space, it is needed to define other components such as HVAC system [345]. Two different approaches exist in

Revit for exporting the gbXML file, which affect the approach for defining systems such

HVAC systems.

Two methods are available in Revit for exporting a gbXML file shown in Figure 6-12. It can be exported based on “energy settings” or “room/space volumes”. For case studies #1 and #3, the advanced energy settings shown in Figure 6-13 and partially in Table 6-4 are selected. These properties can only be transferred to gbXML file or other tools if the export is based on the “energy settings” shown in Figure 6-13. For case study #2, the gbXML file is exported based on “room/space volumes”. The differences between these two options, their limitations, and possible issues with each option is explained in more details in next section.

283

Figure 6-11. Plan view (Left) and 3D view (Right) of building modeled in Revit (Roof is not shown)

Figure 6-12. Two methods in Revit for exporting data to gbXML file

284

Figure 6-13. Advanced energy settings in Revit

Table 6-4. Properties of the building modeled in Revit

Building Component Properties Bedrooms 3 Living Room 1 Square Footage 1098 SF Brick/Air/Rigid Insulation/Vapor

Exterior Wall barrier/CMU/Metal Furring/Gypsum board (R-32) Gypsum board/Metal Stud/Gypsum board (R- Interior Wall 21) Roof Asphalt shingle/Plywood/Wood joist (R-58) Construction Ceiling Acoustic Ceiling Tile (R-1.6) Double Hung with Trim Windows (36”×48”)(SHGC=0.78)(R-1.5) Spaces 4 Spaces Zones 4 Thermal Zones

Location Boston, MA Building Type Single Family Building Operating

Energy Energy

Settings Default Schedule

285

Residential 17 SEER/9.6 HSPF Split HP <5.5 HVAC System ton Outdoor Air per Person 15 CFM Export Category Spaces Project Phase New Construction Building Service VAV-Single Duct Building Infiltration Medium Class Export Default Values Yes

The BIM file schema used in these case studies is gbXML. For example, for case study #2, the gbXML file exported from Revit under “room/space volumes” is version 0.37 and includes elements and attributes demonstrated in Figure 6-14. Space element includes all four spaces defined in Revit and consists of multiple components such as number of people per unit of area, people heat gain, area, volume, and coordination data. Surface provides the information for all the surfaces such as exterior walls, interior walls, roof, and floor.

Construction defines multiple building component systems (e.g., wall system, which is brick on CMU) and their thermal properties such as U-value and absorptance. A layer is assigned to each envelope component and the materials used in each layer are defined under layer element. Material element includes all the required properties such as R-value, thickness, conductivity, specific heat, and density. Schedule elements include the schedule- related data that directly affect the energy consumption outputs. The top-down structure of gbXML schema could be seen in this figure. The properties are assigned to the biggest components such as spaces and it goes down to surfaces, layers, and materials. Each attribute is linked to the higher component in this hierarchy as a parent-child relationship using digit codes (e.g., aim####, in which the # is a single digit), which was shown earlier in Figure 6-6.

286

Figure 6-14. Major elements of gbXML file exported from Revit

6.7.2 Results of the Case Studies, Issues Related to the BBIP, and

Category

The data related to HVAC, and loads are not Data are added in exported to gbXML file from Revit due to Missing Data HVAC OpenStudio manually before selection room/spaces option during exporting a energy simulation. gbXML file from Revit.

The same issue mentioned above remains when OpenStudio export the IDF file to EnergyPlus. Data are automatically

related Missing Data HVAC For example, warnings with respect to outdoor generated based on default

- air per person is generated and default values are values in EnergyPlus. File used, although the value was defined in Revit.

A program is developed to Floors in gbXML file have two similar adjacent look for similarity in Redundant Envelope space, which causes warning when imported in adjacent spaces and remove Data OpenStudio. the redundant data automatically.

The data related to location are not read automatically from gbXML in OpenStudio. Data added manually in Missing Data Location Moreover, the weather file and design days (.ddy OpenStudio. file) need to be provided manually for OpenStuido.

A program is developed to Data look for doors within OpenStudio does not recognize the doors Recognition Envelope gbXML file and assign their construction data in the imported gbXML file. Issue corresponding construction properties automatically.

Data OpenStudio does not recognize the windows The source code of the latest Recognition Envelope construction data in the imported gbXML file. version of OpenStudio Issue (2.0.1) was used to build the OpenStudio app. This issue OpenStudio does not keep the name of materials is resolved in the latest Data Transfer Interoperability General and surfaces, instead uses the codes within version. The current verified Issue gbXML file (e.g., aim####) version of OS is 1.14.0.

The IDF file generated by GBS generates the missing data based on defaults data and they are Inconsistency not consistent with defaults values generated by in Generated General - OpenStudio or EnergyPlus in the case study #2. Data This is a potential source of error in projects, which use multiple simulation tools

The number of the thermal zones generated by Unwanted GBS in gbXML file are more than four, which is Envelope - Generated Data based on the analytical model developed in Revit that divides each surface into subsurface.

290

Issues such as generation of extra subsurface, missing data, undesirable merging of thermal zones, issues with data exchange related to HVAC systems due to its complexity, generation of redundant data such as shading elements, inconsistency in area calculations, and lack of layer information are all observed through the BBIP in different BIM and energy simulation tools [346, 347]. Similar issues are observed and discussed in this paper.

The first challenge occurs in exporting the gbXML file from Revit. Two approaches for generating gbXML file discussed in previous section including the export based on “energy settings” and “room/space volume” shown in Figure 6-12. Exporting based on “energy settings” requires developing an energy model in Revit and it exports all the data defined in the model related to energy modeling including the HVAC data, lighting loads, and number of people defined in spaces. On the other hand, exporting based on “room/space volumes” does not transfer anything other than architectural components, zones, and schedules. There is an issue about exporting based on “energy settings” related to number of surfaces and spaces. The energy model generated by Revit, divide surfaces and spaces to sub-surfaces and sub-spaces, respectively. Therefore, when imported by other tools such as OpenStudio, there will be more than four spaces, which can be confusing and causes errors. For example, in case study #2, OpenStudio reads all the spaces defined in gbXML file and the house modeled in these case studies shown in Figure 6-11, can be imported as a house with up to eight spaces. Therefore, in this study, it was decided to use the

“room/space volumes” export option for case study #2 to avoid such an issue and, instead, add the missing data manually. Because “room/space volume” only export data such as architectural properties, schedules, and thermal zones shown in Figure 6-14.

291

Because of incapability of Revit in generating the gbXML file including all the information, while surfaces and spaces are kept the same as the original model, there will be the issue of missing data, which is also shown in Table 6-5 as “missing data”. It should be noted that according to the case studies reviewed in this paper, gbXML file schema is capable of transferring these information; however, the tools generating or reading the BIM files sometimes are not capable of mapping or reading them properly. For example, data related to schedules are properly mapped to the gbXML file; however, OpenStudio could not read these data properly and its GUI does not show this information. This is a good example of steps #2 and #5, respectively, shown in Figure 6-8. In this study, the missing data such as HVAC system, schedules, and loads are added manually to OpenStudio in case study #2 as it is presented in Table 6-6. Such issues are also identified by other researchers and needs to be addressed in future versions of BIM tools [335, 336, 324, 56,

337].

Another challenge experienced in these case studies is related to defaults values and assumptions in different tools. As it can be seen in Table 6-5, some of the transferred data in BBIP are based on the template and predefined data within different tools. For example,

GBS generates missing data related to HVAC system based on built-in ASHRAE standards, if it is not defined by user. These assumptions are not necessarily similar to what is automatically generated by other tools. EnergyPlus also put default values in case there is no value assigned to certain components by users such as thermostat set points related to thermal zones, which is shown in . Although, this capability is defined to facilitate energy simulation, users should make sure these values are compatible between different tools. In

292 case study #2, these default values such as ground temperature are generated by EnergyPlus and added to the IDF. In addition, in case study #3, many of the required data for energy simulation are not defined in Revit; therefore, GBS generates them based on its internal default values. These default values are not necessarily similar to what is generated by

EnergyPlus in case study #2. Full attention is required, since such differences could cause errors in projects using multiple design and analysis tools. Other researchers have addressed similar issues that involve different tools such as DesignBuilder, Ecotect,

EnergyPlus, and eQuest and two BIM file formats including gbXML and IFC [346, 347].

Therefore, at this state it is recommended to double-check the information in BIM files with exact modeling data to make sure they are accurate and default values are not used without inspection.

The result of these issues and inaccuracies could also be reflected in the outputs. In case study #1, the rough estimation predicts 70 kBtu/sf/yr energy consumption for 1098 square feet of floor area. It was noted that the data related to HVAC, schedules, and loads were not presented in detail in outputs, which was expected, since the GBS is not meant to perform a detailed energy analysis. Case study #2, on the other hand, adopts OpenStudio, which is capable of performing detailed energy analysis and the outputs show about 103 kBtu/sf/yr of energy consumption and 1120 square feet of conditioned area. The difference between case study #1 and #2 is due to multiple factors including the difference between the data manually added such as HVAC system, schedules, loads, and thermostat data.

Case study # 3 shows about 99 kBtu/sf/yr of energy consumption and 1097 square feet of conditioned area, which results in approximately close energy consumption compared to

293 case study #2. Although the outputs might be close between case studies #2 and #3, the automatically generated data based on template values according to ASHRAE standards concerning the HVAC system, loads, and thermostat set points are not similar and could be different from the original model, necessitating verification of the information at each stage. Moreover, there are still missing data especially on HVAC system that need to be added manually unless there is already a middleware developed. There are also differences in conditioned floor area that may be explained by the difference between the analytical model and architectural model used for energy simulation and construction drawings, respectively. Either the centerline or outer surface of wall system can be considered as the borders of the energy analytical model, which can cause a difference in final calculated floor area.

It is shown in Table 6-5 and Table 6-6 that different issues can happen during the BBIP and how different actions can be taken to resolve them including the manual data input or modifying the source code of BIM authoring tools or BEM tools, which is responsible for generating and reading BIM files, respectively. In addition to such options, this paper proposes a new developed corrective tool targeting certain issues in gbXML file related to building envelope, which is explained in details in next section.

6.7.3 Developed gbXML corrective tool

This study has resulted in development of a corrective tool using Python in order to solve the issues related to building envelope data in BIM file prior to importing it in OpenStudio in case study #2. This tool is an executable file, which receives the gbXML file in its root

294 folder and generate a corrected file within the same folder. Such tools can be adopted after building information are mapped to a BIM file, which is the process shown as #2 in Figure

6-8. Other studies have also suggested applicability of adding similar corrective tools to the workflow in automating and improving the BBIP and decision-making process of building’s energy retrofit [341]. An example of such process can be found in the literature devoted to Design4Energy project, which is a European retrofit project that is trying to utilize BIM to BEM process [341]. Dimitriou et al. (2016) developed a gbXML editing tool and gbXML to IDF conversion tool in order to overcome similar interoperability issues such as missing data and automating the BBIP [341]. The editing tool used in their research includes an iterative algorithm, by which the effects of change in different envelope components and adding energy retrofit measures such as adding or making changes in internal mass, windows, heating loads, and air infiltration is evaluated [341]. Application of such corrective tools for gbXML files is also tested on other energy simulation tools such as eQuest, which use Extensible Style Sheet Language Transformation (XSLT) to modify the gbXML file and convert it DOE-INP file for eQuest to rectify the issues with geometry and material properties, when BIM file is directly imported in eQuest [342].

The developed corrective tool in this article resolves two major issues observed during the case study #2, which are related to building envelope, and data exported to the BIM file including 1) duplicate floor’s adjacent spaces and 2) missing data concerning construction material of doors within gbXML file. Figure 6-16 shows the corresponding sections in initial and corrected gbXML file, which occurs in second case study when Revit exports gbXML file under “room/spaces volume” option. Such errors cannot be tracked in case

295 study #1 and #3, since the whole process occurs between Revit and GBS and their file transfer process is not accessible. Such errors could be simply resolved manually; however, the operator need to be familiar with the BIM file structure, which is more complicated when the model include multiple floors, openings, and other building envelope components. The warnings and errors related to these issues were eliminated after using the corrective tool developed for this study. Other issues addressed in Table 6-5, either occur during file conversions through GBS, which is not an open-source tool, or they are data related to HVAC system. Another approach for dealing with these issues is modifying the source code of the built-in codes in tools such as Revit, GBS, OpenStudio, or

EnergyPlus. Revit and GBS are not open-source tools; therefore, the BIM file generator codes could not be modified. OpenStudio, on the other hand, is an open source tool and built-in functions in OpenStudio, which are responsible for converting BIM file to IDF, could be modified. Modifying the existing built-in functions could be an important area of focus for follow-up studies in BBIP.

Figure 6-16. Corrected gbXML files using gbXML corrective tool developed using Python (Top: Looks for duplications in floor’s adjacent spaces. Bottom: Add the missing data related to door’s construction data)

296

The code developed in Python reads and generates a corrected XML file. This tool consists of four functions including functions focused on finding duplications, replacing function, finding doors, and finding the construction data related to doors. The last three functions work within each other, look for the data related to doors, extract the door’s construction reference ID, and replace the line illustrated in Figure 6-16, with a new line that includes the construction ID. The first function that looks for duplications spots the repeated lines related to adjacent spaces and removes one of them to eliminate the warnings illustrated in

Figure 6-15. As it was shown in Figure 6-16, this tool was able to perform the desired tasks including removing the duplicate adjacent spaces and adding the missing information about the door’s construction IDs within the BIM file to prove the feasibility of adopting this tool and similar middleware tools in BBIP.

In general, it could be observed how issues related to BBIP could affect the energy simulations outputs and causes difficulties, inaccuracies, and necessitate adopting middleware tools to resolve some of these issues. Even though, BIM files schemas such as gbXML might be capable of transferring all the required data for energy simulation, the tools generating or reading these files might not perform properly and some information could be missed due to interoperability issues. Moreover, if some of the required data for energy simulation are not defined by user, some tools use the template and default values, which need to be carefully inspected, especially if there are multiple tools used in a project adopting BBIP, which have different resources for their defaults.

297

6.8 Summary and Conclusion

This paper has provided a state-of-the-art review of integration of building energy simulation tools with BIM authoring tools by conducting a thorough review of the literature related to the BBIP and performing three case studies related to this topic, which resulted in developing a gbXML corrective tool focused on building envelope issues. Moreover, a detailed classification is proposed for issues and challenges related to BBIP based on the tools and processes involved in BBIP, which could be beneficial for categorizing existing studies in this field and future research studies. The three major components within BBIP include the BIM tools, BIM file, and BEM tools.

The BIM tools mentioned are CAD tools with capabilities of generating BIM files, while the file formats are mostly in IFC and gbXML. Both gbXML and IFC BIM file schemas use XML language in their newer versions with two different approaches for data presentation, which are top-down and bottom-up structure, respectively. The gbXML file format is designed to facilitate energy modeling, while IFC is more comprehensive and includes other types of data that might not be needed for energy simulation. Therefore,

IDMs and MVDs are developed to extract the data related to energy simulation. Although both schemas are considered as comprehensive and capable BIM file formats, BIM authoring tools might not be able to transfer required data to BIM files, properly. Moreover, in case the BIM files carry required information, the BEM tools reading such files might not be capable of retrieving these data, properly. The BEM tools are also reviewed briefly

298 in this article including OpenStudio, Ecotect, GBS, DesignBuild, and eQuest, while distinguishing GUIs and energy simulation engines such as EnergyPlus and DOE-2.

Different issues and challenges in the BBIP were reviewed based on a detailed classification proposed. The summary of important issues and challenges observed in other studies and the case studies in this paper are as follow:

 BEM tools need to support all the attributes provided by BIM files.

 Difference between the inner volume and analytical volume, that is the space

enclosed between center planes, should be considered.

 Application of middleware tools or manual process is needed to avoid data loss

between two tools during data exchange. An example of such tools is developed

using Python and introduced in this paper.

 BIM file schema need to define all the required information for energy simulation.

 Although BIM file schemas such as gbXML could be capable of transferring all the

required data for energy simulation, tools responsible for generating or reading

these might not perform properly and causes missing information within the

interoperability process.

 BIM authoring tools such as Revit create sub-surfaces and smaller spaces during

exporting a gbXML file, which might causes confusion and error during the

process, especially during the presentation of outputs if the number of spaces and

elements are different from what is modeled in the BIM authoring tool.

 Standards for data exchange process need to be developed.

299

 Additional tools might need to be interconnected to compensate for lack of data,

especially when one tool works as a source of building information that might not

include all the required data.

 Proper error or warning messages need to be generated if the BIM file does not

include the required information.

 BIM files could add new attributes for more exceptional cases such as curved

surfaces.

 Initial stages of design might not include all the required data for energy simulation,

which might affect the results.

 BIM files could be updated using the new outputs after performing an energy

simulation.

 Corrective tools are needed to resolve issues such as development of redundant data

related to building envelope as subsurface.

 Some of the required data for energy simulation are not defined by user; therefore,

some tools use the template and default values based on different standards such as

ASHRAE, which need to be carefully inspected, especially if there are multiple

tools used in a project adopting BBIP.

Finally, three case studies are investigated in this paper to study the different issues and challenges including interoperability issues in the BBIP using Revit, OpenStudio, GBS,

EnergyPlus, and gbXML file format in more details. The results show that not all the required data for energy simulation such as HVAC system data, schedules, and loads are exported properly to the gbXML file using Revit and data needed to be added manually. 300

Different energy consumption and floor area are calculated in three different case studies for the same building. The case study #1 is based on a rough estimation using GBS and the energy modeling is not as detailed as the other two case studies. The results of energy consumption and floor area calculations are close in case studies #2 and #3, despite of the differences in data generated automatically based on standards such as ASHRAE or other default values within different tools. The generated data are related to HVAC, schedules, and loads that are due to loss of data through data exchange between Revit, GBS, and

OpenStudio.

Although, all the issues identified in these case studies could be corrected manually, the process might not be easy for an operator who is not familiar with the structure of BIM files. More importantly, the models that have more floors, windows, and building envelope components cannot be easily modified for each component separately; therefore, the corrective tool reviewed was developed using Python and the results of three case studies show the applicability of this tool, which could be a good sample of corrective middleware tools in BBIP. The gbXML corrective tool is focused on building envelope data such as missing construction materials ID in gbXML file. The results show that the corrective tool is working properly and energy simulation is conducted without sever errors or warnings concerning building envelope. It shows such tools and middleware could be integrated with the BBIP to mitigate the issues in the process.

In conclusion, it can be observed that there is still a long way to get to reach the objective of performing the BBIP by click of a button. There are many interoperability issues

301 identified and reviewed in this paper, which need to be addressed in future versions of BIM tools, BEM tools, and BIM schemas. Most of these issues slow down the automation in the

BBIP, and there are many areas, which still require manual interaction as opposed to automatic data input. It seems that the current BEM tools are capable of reading the BIM files with minimum errors as long as the required data are provided within the file.

However, the majority of issues occur either when the BIM files are generated by CAD tools considered here as BIM tools such as Revit or during reading the BIM files by energy simulation tools, which could be the focus of future research areas in this field.

6.9 References

[285] NBIMS-US, "National BIM Standard-United States® Version 3," National

Institute of Building Sciences, Washington, DC, 2013.

[286] E. Krygiel and B. Nies, Green BIM: Successful Sustainable Design with Building

Information Modeling, Indianapolis, Indiana: Wiley, 2008.

[287] D. K. Smith and M. Tardif, Building Information Modeling - A Strategic

Implementation Guide, Hoboken, New Jersey: John Wiley & Sons, 2008.

[288] R. Kreider, J. Messner and C. Dubler, "DETERMINING THE FREQUENCY AND

IMPACT OF APPLYING BIM FOR DIFFERENT PURPOSES ON PROJECTS," in

Proc., 6th Int. Conf. on Innovation in Architecture, Engineering and Construction (AEC),

Pennsylvania State Univ., University Park, PA, USA, 2010.

302

[289] C. Eastman, P. Teicholz, R. Sacks and K. Liston, BIM Handbook, A Guide to

Building Information Modeling for Owners, Managers, Designers, Engineers, and

Contractors, New Jersey: John Wiley & Sons, 2011.

[290] J.-H. Woo, C. Diggelman and B. Abushakra, "BIM-BASED ENERGY

MORNITORING WITH XML PARSING ENGINE," in Proceeding of the 28th ISARC,

Seoul, Korea, 544-545, 2011.

[291] N. W. Young Jr., S. A. Jones, H. M. Bernstein and J. E. Gudgel, "THE BUSINESS

VALUE OF BIM," McGraw Hill, New York, 2009.

[292] I. J. Ramaji, J. I. Messner and R. M. Leicht, "Leveraging Building Information

Models in IFC to Perform Energy Analysis in OpenStudio," in ASHRAE and IBPSA-USA

SimBuild 2016, Salt Lake City, 2016.

[293] H.-Y. Chong, C.-Y. Lee and X. Wang, "A mixed review of the adoption of Building

Information Modelling (BIM) for sustainability," Journal of Cleaner Production, vol. 142, p. 4114e4126, 2017.

[294] L. Mohamed Khodeir and A. Ali Nessim, "BIM2BEM integrated approach:

Examining status of the adoption of building information modelling and building energy models in Egyptian architectural firms," Ain Shams Engineering Journal, p. In Press, 2017.

[295] C.-Y. Lee, H.-Y. Chong and X. Wang, "Streamlining Digital Modeling and

Building Information Modelling (BIM) Uses for the Oil and Gas Projects," Arch Computat

Methods Eng, p. 1–48, 2016. 303

[296] N. Somboonwit, A. Boontore and Y. Rugwongwan, "Obstacles to the Automation of Building Performance Simulation: Adaptive Building Integrated Photovoltaic (BIPV) design," in 5th AMER International Conference on Quality of Life, 25-27 February 2017,

Bangkok, Thailand, 2017.

[297] K.-J. Tu and D. Vernatha, "Application of Building Information Modeling in

Energy Management of Individual Departments Occupying University Facilities,"

International Journal of Civil, Environmental, Structural, Construction and Architectural

Engineering, vol. 10, no. 2, pp. 225-231, 2016.

[298] E. I.Samuel, E. Joseph-Akwara and A. Richard, "Assessment of energy utilization and leakages in buildings with building information model energy," Frontiers ofArchitecturalResearch, vol. 6, pp. 29-41, 2017.

[299] M. Alahmad, W. Nader, J. Neal, J. Shi, C. Berryman, Y. Cho, S.-K. Lau, H. Li, A.

Schwer, Z. Shen, J. Stansbury and T. Zhang, "Real Time Power Monitoring & Integration with BIM," in IECON 2010 - 36th Annual Conference on IEEE Industrial Electronics

Society, Glendale, AZ, USA, 2010.

[300] OGC, "Sensor Model Language (SensorML)," 2017. [Online]. Available: http://www.opengeospatial.org/standards/sensorml. [Accessed 18 01 2017].

[301] F. Jalaei and A. Jrade, "INTEGRATING BUILDING INFORMATION

MODELING (BIM) AND ENERGY ANALYSIS TOOLS WITH GREEN BUILDING

CERTIFICATION SYSTEM TO CONCEPTUALLY DESIGN SUSTAINABLE

304

BUILDINGS," Journal of Information Technology in Construction, vol. 19, pp. 494-519,

2014.

[302] A. Bres and G. Suter, Automated building energy performance simulation: a review of approaches and outline of research opportunities, Krakow, Poland: 23nd Workshop of the European Group for Intelligent Computing in Engineering, 2016.

[303] S. Kumar, "INTEROPERABILITY BETWEEN BUILDING INFORMATION

MODELS (BIM) AND ENERGY ANALYSIS PROGRAMS," SCHOOL OF

ARCHITECTURE, UNIVERSITY OF SOUTHERN CALIFORNIA, Los Angeles, 2008.

[304] I. J. Ramaji and A. M. Memari, "Interpreted Information Exchange: Systematic

Approach for BIM to Engineering Analysis Information Transformations," Jurnal of

Computing in Civil, pp. 1-17, 2016.

[305] S. Salakij, N. Yu, S. Paolucci and P. Antsaklis, "Model-Based Predictive Control for building energy management. I:Energy modeling and optimal control," Energy and

Buildings, vol. 133, p. 345–358, 2016.

[306] L. Santos, S. Schleicher and L. Caldas, "Automation of CAD models to BEM models for performance based goal-oriented design methods," Building and Environment, vol. 112, pp. 144-158, 2017.

[307] T. Maile, M. Fischer and V. Bazjanac, "Building Energy Performance Simulation

Tools—A Life-Cycle and Interoperable Perspective;," Stanford University: Stanford, CA,

USA, 2007. 305

[308] S. Pinheiro, J. Donnell, R. Wimmer, V. Bazjanac, S. Muhic, T. Maile, J. Frisch and

C. van Treeck, "MODEL VIEW DEFINITION FOR ADVANCED BUILDING ENERGY

PERFORMANCE SIMULATION," in CESBP/BauSIM 2016 Conference, Dresden,

Berlin, 2016.

[309] N. O. Nawari and M. Kuenstle, Building Information Modeling - Framework for

Structural Design, Taylor & Francis, 2015 .

[310] J. B. Kim, W. Jeong, M. J. Clayton, J. S. Haberl and W. Yan, "Developing a physical BIM library for building thermal energy simulation," Automation in Construction, vol. 50, pp. 16-28, 2015.

[311] W. Jeong, J. B. Kim, M. J. Clayton, J. S. Haberl and W. Yan, "Translating Building

Information Modeling to Building Energy Modeling Using Model View Definition,"

Hindawi Publishing Corporation, pp. 1-21, 2014.

[312] H. Kim and K. Anderson, "Energy Modeling System Using Building Information

Modeling Open Standards," JOURNAL OF COMPUTING IN CIVIL ENGINEERING, vol. 27, no. 3, pp. 203-211, 2013.

[313] R. K. Strand and C. O. Pedersen, "Modeling Radiant Systems in an Integrated Heat

Balance Based Energy Simulation Program," ASHRAE Transactions, vol. 108, no. 2, pp.

1-9, 2002.

306

[314] Y. N. Bahar, C. Pere, J. Landrieu and C. Nicolle, "A Thermal Simulation Tool for

Building and Its Interoperability through the Building Information Modeling (BIM)

Platform," Buildings, vol. 3, pp. 380-398, 2013.

[315] O. Granlund Oy, M. Korhonen and T. Laine, "Energy Analysis Software Evaluation

BIM Interface and Interoperability," Granlund, Eskilstuna, Sweden, 2008.

[316] Modelica, "Modelica and the Modelica Association," Modelica, 2017. [Online].

Available: https://www.modelica.org/. [Accessed 12 February 2017].

[317] K.-U. Ahn, Y.-J. Kim, C.-S. Park, I. Kim and K. Lee, "BIM interface for full vs. semi-automated building energy simulation," Energy and Buildings, vol. 68, pp. 671-678,

2014.

[318] S. Lagüela, L. Díaz-Vilariño, L. Martínez and L. Armesto, "Automatic thermographic and RGB texture of as-built BIM for energy rehabilitation purposes,"

Automation in Construction, vol. 31, pp. 230-240, 2013.

[319] Y. Ham and M. Golparvar-Fard, "Mapping actual thermal properties to building elements in gbXML-based BIM for reliable building energy performance modeling,"

Automation in Construction, vol. 49, pp. 214-244, 2015.

[320] buildingSmart, "Summary of IFC Releases," buildingSmart, 2017. [Online].

Available: http://www.buildingsmart-tech.org/specifications/ifc-releases. [Accessed 23

January 2017].

307

[321] T. Liebich, "IFC4 – the new buildingSMART Standard," 2013. [Online]. Available: www.buildingsmart-tech.org/.../ifc.../ifc4.../buildingSMART_IFC4_Whatisnew.pdf.

[Accessed 23 January 2017].

[322] B. Dong, K. P. Lam, Y. C. Huang and G. M. Dobbs, "A comparative study of the

IFC and gbXML informational infrastructures for data exchange in computational design support environments," in Proceedings: Building Simulation 2007, 2007.

[323] E. El Asmi, S. Robert, B. Haas and K. Zreik, "A standardized approach to BIM and energy simulation connection," International Journal of Design Sciences and Technology, vol. 21, no. 1, pp. 1630-7267, 2015.

[324] J. O'Donnell, R. See, C. Rose, T. Maile, V. Bazjanac and P. Haves, "SIMMODEL:

A DOMAIN DATA MODEL FOR WHOLE BUILDING ENERGY SIMULATION," in

12th Conference of International Building Performance Simulation Association, Sydney,

2011.

[325] M. Venugopal, C. Eastman, R. Sacks, I. Panushev and S. Aram, "SEMANTIC

EXCHANGE MODULES: A NEW METHOD FOR SPECIFYING AND

IMPLEMENTING MODEL VIEWS," Georgia Institute of Technology, 2011.

[326] S. Hyun, L. Marjanovic-Halburd and R. Raslan, "Investigation into informational compatibility of Building Information Modelling and Building Performance Analysis software solutions," WIT Transactions on The Built Environment, vol. 149, pp. 543-553,

2015.

308

[327] EBC, "IEA EBC Annex 60," EBC, 2017. [Online]. Available: http://www.iea- annex60.org/about.html. [Accessed 23 January 2017].

[328] BuildingSMART, "IFC Solutions Factory, The Model View Definition site," BLIS

Project, 12 February 2017. [Online]. Available: http://www.blis-project.org/IAI-MVD/.

[Accessed 12 February 2017].

[329] BLIS-Project, "BLIS-Project," 2017. [Online]. Available: https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact

=8&ved=0ahUKEwing9KEzYvSAhVCCiwKHV0-

DhgQFggaMAA&url=http%3A%2F%2Fwww.blis-project.org%2FIAI-

MVD%2FIDM%2FGSA-003%2FER_GSA-003.pdf&usg=AFQjCNFJ2y- zQhXTaYz33sts3db9aWWzng&sig2=X3wxGv. [Accessed 12 February 2017].

[330] BLIS-Project, "BLIS-Project," 2009. [Online]. Available: www.blis- project.org/IAI-MVD/IDM/BSA-002/PM_BSA-002.pdf. [Accessed 12 February 2017].

[331] "gbXML," [Online]. Available: http://www.gbxml.org/. [Accessed 23 September

2016].

[332] H. J. Moon, M. S. Choi, S. K. Kim and S. H. Ryu, "CASE STUDIES FOR THE

EVALUATION OF INTEROPERABILITY BETWEEN A BIM BASED

ARCHITECTURAL MODEL AND BUILDING PERFORMANCE ANALYSIS

PROGRAMS," in 12th Conference of International Building Performance Simulation

Association, Sydney, 2011.

309

[333] E. Epstein, Implementing Successful Building Information Modeling, Artech

House, 2012.

[334] J. Steel, R. Drogemuller and B. Toth, "Model interoperability in building information modelling," Softw Syst Model, vol. 11, pp. 99-109, 2012.

[335] L. Che, Z. Gao, D. Chen and T. H. Nguyen, "Using building information modeling for measuring the efficiency of building energy performance," in Proceedings of the international Conference on Computing in Civil and Building Engineering, Nottingham,

2010.

[336] F. Jalaei and A. Jrade, "An Automated BIM Model to Conceptually Design,

Analyze, Simulate, and Assess Sustainable Building Projects," Journal of Construction

Engineering, pp. 1-22, 2014.

[337] P. Katranuschkov, R. Scherer and C. Balaras, "BIM-BASED MULTI-MODEL

FRAMEWORK FOR ENERGY EFFICIENT BUILDING DESIGN," in 41st IAHS

WORLD CONGRESS, Sustainability and Innovation for the Future, 13-16th September,

Albufeira, Algarve, Portugal, 2016.

[338] E. Turkyılmaz, "A proposal for energy efficient design: an IFC based design decision system and its application," WIT Transactions on Ecology and The Environment, vol. 179, pp. 663-670, 2013.

310

[339] K. Karen and N. Douglas, "Automated Energy Performance Visualization for

BIM," in Building Information Modeling : BIM in Current and Future Practice, Wiley ,

2014, pp. 119-128.

[340] J. T. O’Donnell, T. Maile, C. Rose, N. Mrazović, E. Morrissey, C. Regnier, K.

Parrish and V. Bazjanac, "Transforming BIM to BEM: Generation of Building Geometry for the NASA Ames Sustainability Base BIM," Ernest Orlando Lawrence Berkeley

National Laboratory, Berkeley, 2013.

[341] V. Dimitriou, S. K. Firth, T. M. Hassan and F. Fouchal, "BIM enabled building energy modelling: development and verification of a GBXML to IDF conversion method," in Proceedings of the 3rd IBPSA-England Conference BSO 2016, Newcastle, 12th-14th

September 2016, 2016.

[342] E. Guzmán Garcia and Z. Zhu, "Interoperability from building design to building energy modeling," Journal of Building Engineering, vol. 1, pp. 33-41, 2015.

311

7) Chapter 7. Building Energy Performance Assessment

Tool (BEPAT)

This chapter has been written as a journal paper and is already submitted for review.

312

Development of a Platform for Building Energy

Performance Assessment Tool (BEPAT) for Energy

Smart Homes and Design Optimization

1 2 Ehsan Kamel , Ali M. Memari

1Ph.D. Candidate, Department of Civil Engineering, Structural Engineering, Penn State University, 321 Sackett Building, University Park, PA 16802 (corresponding author). E- mail: [email protected] 2Professor, Hankin Chair in Residential Building Construction, and Director, Pennsylvania Housing Research Center, Dept. of Architectural Engineering and Dept. of Civil and Environmental Engineering, Penn State Univ., University Park, PA 16802. E-mail: [email protected] 7.1 Abstract

Buildings in the U.S. consume about 40% of the total energy; therefore, emerging tools and approaches such as home energy monitoring and management systems, optimizing the pre-construction design of a building, and energy retrofit of existing homes can have a significant contribution toward reduction in energy related costs and CO2 emissions. These tools and approaches require proper inputs. Fine-grained energy consumption data can be used to identify the envelope components that can be thermally improved through energy retrofit or during the design phase. In this paper, fine-grained data refers to the amount of heat transfer through every single wall, window, door, roof, and floor in a building.

The Building Energy Performance Assessment Tool (BEPAT) explained in this paper is a platform, which can be further enhanced to be adopted in energy smart home, building

313 design tools, and energy retrofit tools as a supplementary software tool. It provides users with fine-grained information through an automated process. It is developed by modifying

EnergyPlus source code as the energy simulation engine using C++. A computer model is developed in Revit to verify the BEPAT results. Revit is used to export the architectural model to a BIM file and OpenStudio converts it to Input Data File (IDF) to be used in

BEPAT. Validating BEPAT’s outputs with EnergyPlus advanced outputs shows the outputs are similar and prove the capability of this tool to facilitate providing detailed outputs on the amount of heat transfer through walls, fenestrations, roofs, and floors.

Author Keywords: Energy Retrofit, Energy Smart Home, Energy Monitoring,

EnergyPlus, Building Energy Simulation

7.2 Introduction

Three of the major emerging technologies and tools such as energy smart homes that are developed for building energy management and tools that are being adopted during design and energy retrofit decision-making process require proper input in order to provide accurate and useful output more efficiently to contribute to reduction in energy consumption. The types of input studied in this article are focused on building envelope components such as walls, windows, and roof. Detailed heat transfer data through these components could contribute to three major areas including monitoring systems in energy smart homes, decision-making process for energy retrofit of buildings, and evaluation process in pre-construction or design phase of buildings. Conventional output from existing energy simulation tools only includes the accumulated energy-related output such as 314 whole-house energy consumption. More detailed output such as the amount of heat transfer through a single window in a specific thermal zone could be also acquired by existing energy simulation tools such as EnergyPlus; however, the process needs software-related skills, is time-consuming, and is error-prone. Therefore, it is required to equip these systems and tools in the three major areas mentioned above with new computer tools.

First tool is energy smart home. Alam et al. [43] defined smart homes as “…an application of ubiquitous computing in which the home environment is monitored by ambient intelligence to provide context-aware services and facilitate remote home control.”

Different services exist for smart homes and improving energy efficiency of a building is one of the most important services identified by researchers. Adopting sensors, actuators, and monitoring energy related data could contribute to energy smart homes [143, 348].

Energy monitoring systems are mainly limited to either, whole house or appliances energy meters [146]. Energy smart homes can also be equipped with energy data processing capabilities such as energy simulation tools to perform more in-depth energy analysis based on the acquired data. The concept considered in this paper is based on equipping the building with an energy simulation tool, which can provide detailed information on the amount of heat transfer through building envelope components for monitoring purposes.

Energy retrofit is the second approach in reducing building energy consumption. The detailed outputs can also contribute to evaluation and decision-making process of energy retrofit of a building to make more cost effective decisions by identifying the components that contribute more to energy loss as opposed to whole-house energy retrofit. There are

315 many research and case studies only focused on building envelope energy retrofit in contrast with retrofit of HVAC or lighting systems [232, 236, 40, 267, 233], which shows the importance of this area and contribution of this tool in improving the process of building energy retrofit.

Optimizing building components during the design phase is the third approach in reducing building energy consumption. Decisions made at the design-phase and early stages of building design could be more effective in terms of energy efficiency compared to measures adopted later during the use phase [349, 350, 46]. Similar studies have also emphasized the importance of design phase for implementing energy efficiency measures

[350] with different areas of focus such as building envelope shape [46] or suggesting indicators that correlate between design variations and energy consumptions [351].

Therefore, detailed information during the pre-construction phase could contribute to design tools and be used as input for optimization methods during the decision-making process. Detailed information as opposed to accumulated energy consumption information could increase the accuracy of outputs from optimization tools and improve the decisions during the design phase.

Accumulated energy consumption data should be distinguished from detailed or fine- grained information on energy consumption. The current energy analysis tools present the accumulative energy-related data such as whole-house energy consumption, heating or cooling total energy consumption, and energy consumption due to lighting systems [352].

However, the concept studied in this paper is focused on detailed energy consumption data.

316

It can be defined as fine-grained information about the heat transfer or energy consumption through/of smaller components of buildings such as appliances, each of the lights installed in a room, or each of the building envelope components such as a certain exterior wall, window, or door. Current tools provide the zone energy consumption by summing up the heat loss/gain through all the walls, roof, floor, and windows within that zone, while the detailed energy-related data is focused on each component, separately. This type of information could contribute to better energy management and more efficient energy retrofit. For example, in energy retrofit of a building it is desirable to understand detailed energy loss through each component such as a wall or a window [353]. Moreover, with current advancements in smart home monitoring systems, these types of data are very useful in order to inform the users on their energy consumption, since they could be integrated with monitoring system’s Graphical User Interface (GUI).

EnergyPlus used in this study as the energy simulation engine is an open-source energy simulation software, which works as the simulation engine in other energy simulation GUIs such as OpenStudio, DesignBuilder, and BEopt and it is capable of performing a comprehensive building energy analysis used in multiple research studies [23, 222, 223,

24, 34, 306]. EnergyPlus provides two types of output, including standard and advanced outputs. Standard outputs include the common and accumulative energy-related outputs such as annual energy consumption, heating and cooling loads, energy loss through air infiltration, and accumulative energy loss through windows. Advanced outputs include average heat loss/gain through different building envelope components such as walls, roof, floors, and doors. Although these data are available in EnergyPlus outputs, none of the

317 energy analysis GUIs that use EnergyPlus as the energy simulation engine provides such data. Therefore, to access such data, users should be familiar with the definition of these advanced outputs and acquire them through a process, which is not straightforward and automated. Moreover, using such data in smart homes requires more ease of access and faster calculation process especially if it is semi-real-time monitoring system. Therefore, it is important to facilitate and automate obtaining detailed data related to the energy consumption of a building.

BEPAT provides fine-grained energy consumption data about heat loss/gain through each wall, roof, floor, and fenestrations, i.e., windows and doors. This tool works based on modifying the source code of EnergyPlus. Outputs of this tool are saved in separate text files for ease of access and automation of data acquisition, which could be further enhanced by using these data in other GUIs or tools dedicated to building design, energy retrofit, and energy monitoring systems. There are examples of GUIs designed for energy smart homes using energy simulation tools, which shows the applicability of such tools in improving the energy performance of buildings [174, 11, 9]. Figure 7-1 shows the application and major contributions of this tool and demonstrates how these outputs can be used in other GUIs in three different areas. For example, it allows the user to select a specific building envelope component or thermal zone and it provides the user with detailed energy loss/consumption through/within that component/zone and within a certain period. Hence, it can help decision making on possible energy retrofit scenarios and the effect of partial energy retrofit of buildings on total energy consumption [41].

318

The adopted methodology for developing the tool is explained in the next section.

Development of a computer model for verification purposes is also covered subsequently, and it is followed by results of the computer model and comparison between the BEPAT’s outputs and two other methods to obtain either detailed or accumulative energy-related outputs. Finally, the results are discussed in more details and the summary and conclusion is provided.

Figure 7-1. Schematic illustration of potential applications and major contributions of BEPAT

319

7.3 Methodology for Developing and Validating BEPAT

The adopted approach for developing BEPAT is modifying the EnergyPlus source code, an existing open-source energy simulation software developed by National Renewable

Energy Laboratory (NREL) and funded by the U.S. Department of Energy (DOE) [354].

Inputs are fed to the software as text files known as IDF (Input Data File). The basic IDF editor available for EnergyPlus is known as EPlaunch illustrated in Figure 7-2. To work with such an editor or directly modify the IDF file, the user would need an in-depth understanding of the IDF structure and different attributes related to envelope, HVAC, electrical, or schedules within the IDF file. In addition, many Graphical User Interfaces

(GUI) exist, which are developed for energy simulation and analysis that use EnergyPlus as their core simulation engine and facilitate the energy modeling by providing graphical interfaces such as OpenStudio, BEopt, and DesignBuilder [307, 308]. Although application of these tools are more straightforward, the whole process for modeling and obtaining outputs is not automated and still needs manual process, which could be time taking, costly, and error prone. Moreover, all of the aforementioned tools only provide accumulated heat transfer for windows, walls, roofs, and floors in each thermal zone as opposed to detailed information for each envelope component.

320

Figure 7-2. IDF editor used for modifying EnergyPlus files in EPlaunch

The availability of EnergyPlus source code helps users to modify this simulation engine in order to integrate it with other tools or obtaining customized outputs. The latter function has been quite beneficial to this study, part of which involves developing a tool based on modification of EnergyPlus source code to develop BEPAT. The source code is free and available for users on GitHub, and the version used for this research is EnergyPlus 8.6.0, which is also the latest verified version of this software at the time this study was conducted. The instructions for building the executable file for EnergyPlus is also available

321 on GitHub website [355]. The source code can be viewed and modified by any Integrated

Development Environment (IDE) such as VisualStudio, which is the IDE used in this study

(version 14.0 Update 3, 2015). The old version of EnergyPlus is developed in FORTRAN; however, the recent versions use C++ as the main programming language, and the previous modules in FORTRAN are converted to corresponding codes in C++.

Open-source software tools consist of multiple modules. Each module could include hundreds of lines of a certain programing language, which could be boiled down to several subroutines and functions responsible for a certain task or output. For example, EnergyPlus consists of about 275 modules. Figure 7-3 illustrates some of the modules such as Boilers,

CoolTower, and BaseboardElectric modules, which could be viewed and modified in

VisualStudio. Each module is focused on different parts of the whole-house energy consumption calculations such as the amount of heat transfer through conduction, convection, radiation, different components of the HVAC system, lighting, shadings, and appliances. These modules are interconnected and data are transferred back and forth between them until the final outputs are obtained.

322

Figure 7-3. Modules in EnergyPlus source code opened in VisualStudio

This research study is focused on building envelope components and is intended to obtain detailed data on heat gain or loss through certain components within a certain period. The components considered in this study to develop BEPAT include both opaque and transparent components such as walls, roof, floors, windows, and doors. Therefore, the first step is to identify the source code modules in which the related calculations and outputs are provided. In this case, two main modules are modified that contribute to the heat balance and data related to surfaces in buildings. The code developed in C++ is added in these two modules and the corresponding required changes are applied in the header files

(.hh files).

323

The new C++ commands added to the existing source code modules perform the following tasks:

- Define a new variable that collects the detailed data as opposed to accumulated data

- Collect the data before EnergyPlus sums them up for each thermal zone.

- Add them up for each time step of the analysis in order to obtain the total heat

gain/loss through the simulation run period.

- Generate a text file in the root directory of the modified EnergyPlus executable file.

- Write the total heat gain/loss for each opaque and transparent component in each

thermal zone, separately.

- Assign a thermal zone and proper name to each component in order to facilitate

identifying these components and their corresponding thermal zone to be used in

other tools in future by enhancing this platform.

In order to obtain the heat transfer data through building envelope components, it should be noted that there are three major heat transfer methods considered in EnergyPlus incorporated in the source code including radiation, convection, and conduction. The same concept is used in other energy simulation tools; however, the codes used in different tools might define different functions and modules to model the heat transfer and perform the process. The subroutine used in one of the EnergyPlus’s modules adopts the first law of thermodynamic to obtain the heat flow rate using Equation 1 in order to calculate the required energy that satisfies equilibrium. Depending on the HVAC system, the fluid considered in Equation 1 could be different. In this study, the HVAC uses air, where the

324 heat capacity 푐푝constant would be 푐푝,푎푖푟. The air flow rate (푚̇ 푠푦푠) is calculated based on blown air (푇푖푛) , zone mean air temperature (푇푧 ) obtained from another module based on well-stirred model. 푄푠푦푠 is also set equal to the zone demand. To obtain this temperature, it is required to calculate the temperature on all the surfaces within a thermal zone, which is obtained based on the amount of heat transfer through these surfaces. To calculate the conduction share of total heat transfer, for example, it requires the thermal properties of building envelope components such as U-value (푢), exterior temperature to calculate the

∆푇, and the surface area of the component (퐴), all shown in Equation 2, which is based on the heat flux through a component (푄). In this study, it is assumed that the heat loss through convection or air infiltration is zero to simplify the process. Using Equation 1 and Equation

2, the required thermal energy to keep the temperature of thermal zone within a certain range is calculated and reported in EnergyPlus.

Equation 1. 푚̇ 푠푦푠 = 푄푠푦푠/(퐶푝, 푎푖푟 ⋅ (푇푖푛 − 푇푧))

Equation 2. 푄 = 푢. 퐴. ∆푇

From the multiple outputs and tables produced by EnergyPlus after simulation, the output components of interest in this study are included in the Annual Building Sensible Heat Gain

Components table. Furthermore, outputs such as HVAC terminal unit sensible air cooling/heating, people sensible heat addition, equipment sensible heat addition, infiltration heat addition, opaque surface conduction and other heat removal/addition, and window heat removal/addition for each thermal zone are also provided in this table. An

325 example of such a table is presented in Figure 7-4. However, all the data are accumulated for each zone. For example, the window heat addition and opaque surface conduction provide the heat gain through all the windows together and all the opaque components together (such as walls, roof, and floor) within a certain thermal zone, respectively. All the available energy simulation GUIs that use EnergyPlus or even other simulation engines such as DOE-2 only provide these standard outputs as the accumulated results.

Zone Number

Figure 7-4. Example of the sensible heat gain summary table generated by EnergyPlus

None of the available GUIs (with EnergyPlus as energy simulation engine) provides detailed output by default; however, another method for obtaining the detailed heat transfer information in EnergyPlus is to use the advanced outputs option in the IDF editor, i.e.,

EPlaunch directly by user, which is explained in the next section as one of the methods used for data verification, which is referred to as Method #1 in this study. The issue with this method is that it is not convenient for users to obtain such data and it adds complicated actions to the process, which could be labor intensive, time-taking, and error prone; therefore, allowing direct and access to these data can reduce the risk of human errors and save time. Automation is one of the most important aspects of newly emerged technologies

326 such as smart homes [87], and one of the major contributions of this tool is facilitating data acquisition by automating the process by eliminating the need for special expertise for using the software. It minimizes the need for complex process to obtain detailed heat transfer data through each building envelope component.

Automation could be expanded to other areas such as energy modeling, where the use of

Building Information Modeling (BIM) can significantly help the process. Moreover, for future applications it can facilitate visualizing the outputs of energy simulation as it is shown schematically in Figure 7-1 by storing and organizing the outputs in standard BIM formats such as gbXML. Therefore, the energy modeling for validating BEPAT’s outputs is performed using BIM. Revit is used in this study to prepare the architectural model, which is later exported to a gbXML BIM file format. OpenStudio is used as a middleware to add the required data for energy modeling such as data related to the HVAC system and schedules, and to convert the BIM file to IDF, which is the input file for EnergyPlus. An overview of the energy analysis process used in this paper is presented in Figure 7-5. There are still deficiencies with using BIM in energy simulation. For example, not all the required data related to HVAC systems and schedules are exported through gbXML file, which itself is exported from Revit. Therefore, additional data are defined in OpenStudio. Other researchers have also identified these deficiencies associated with using BIM in energy modeling, and missing data are added either manually or with other semi-automated methods such as application of middleware tools [324, 56, 341, 342].

327

Figure 7-5. Overview of the enegry analysis process used in this study

Methods explained above are adopted in order to perform an energy modeling and analysis in an automated, more convenient, less error-prone, faster, and more reliable way. In order to make sure the changes in the source code of EnergyPlus are working properly and

BEPAT’s outputs are valid, a data validation need to be performed, which is explained in more detail in the next section.

7.4 Computer Model and Validation Method

In order to verify the outputs generated by BEPAT, a simplified one-story house with four thermal zone is modeled in Revit. Two methods based on EnergyPlus used in the study for output verification are illustrated in Figure 7-6. The whole process is also depicted in

Figure 7-7 in order to show how these methods relate to each other and illustrate the process within EnergyPlus (E+). The detailed outputs of BEPAT (Method #2) are obtained and compared to the outputs obtained from advanced outputs in EnergyPlus (Method #1).

Based on the overview shown in Figure 7-7, it can be observed that the major difference between Methods #1 and #2, which provide similar outputs, is ease of access. BEPAT fetches these outputs and saves them in series of text files, which enables the user or software developers to monitor the data or integrate them with other software tools in

328 easier, faster, more reliable, and automated way compared to regular energy simulation and output acquisition methods. BEPAT only fetches, organizes, and saves the data as opposed to performing any calculation; therefore, its outputs could be easily validated as long as similar outputs are obtained by EnergyPlus (E+), since the source of these outputs are the same, and they are only obtained through a different path, as it is shown in Figure 7-7. The steps, challenges and issues with Methods #1 will be explained in more detail in the next section.

The energy modeling performed for validating BEPAT’s outputs in this study is also semi- automated using BIM. Revit is used in this study to prepare the model, which is later exported to a gbXML BIM file format, while OpenStudio GUI is used as a middleware to add required data for energy modeling such as data related to HVAC system and schedules and convert the BIM file to IDF. An overview of the energy modeling and analysis process is presented in Figure 7-5. A rooftop air conditioning system is defined as the HVAC system using OpenStudio GUI and connected to all four thermal zones. Schedules, loads, and thermostat set points are also assigned to each thermal zone using OpenStudio GUI.

Finally, OpenStudio is used to generate the IDF file needed for both methods illustrated in

Figure 7-6. Both IDF file and a weather file required for energy simulation should be provided for BEPAT and it generates the text files containing the detailed outputs. BEPAT works by running an executable file, which is generated based on the modified source code of EnergyPlus.

329

The properties of the building modeled in this study are presented in Table 7-1. Figure 6-11 also shows the plan and 3D view of the building modeled in Revit.

Table 7-2 summarizes the number of interior and exterior walls and openings for each zone.

The outputs discussed in this paper will be based on these components, and detailed the amount of heat transfer through these components is the focus of BEPAT. The outputs are based on a one-year simulation and Boston weather file as it is presented in Table 7-1.

Figure 7-6. Two methods used in this study for data verification

330

Figure 7-7. Overview of two different methods used in the study

Figure 7-8. Plan view (Left) and 3D view (Right) of building modeled in Revit

331

Table 7-1. Properties of the building modeled in EnergyPlus

Building Component Properties

Bedrooms 3

Living Room 1

Square Footage 1098 SF

Exterior Wall Brick/Air/Rigid Insulation/Vapor barrier/CMU/Metal

Furring/Gypsum board (R-32)

Interior Wall Gypsum board/Metal Stud/Gypsum board (R-21)

Roof Asphalt shingle/Plywood/Wood joist (R-58)

Ceiling Acoustic Ceiling Tile (R-1.6)

Windows Double Hung with Trim (36”×48”)(SHGC=0.78)(R-

1.5)

Spaces 4 Spaces

Zones 4 Thermal Zones

Construction Location Boston, MA

Building Type Single Family

Building Operating Default

Schedule

HVAC System Residential 17 SEER/9.6 HSPF Split HP <5.5 ton

Outdoor Air per Person 15 CFM

Export Category Spaces

Energy Settings Project Phase New Construction

Building Service VAV-Single Duct

Building Infiltration Class Medium

Export Default Values Yes

332

Table 7-2. Number of different components in each thermal zone

Number of Number of Number of Number of Zone # Windows Doors Exterior Walls Interior Walls

1 2 1 3 3

2 1 1 2 2

3 0 1 1 3

4 2 1 2 2

7.5 Results and Discussion

As it is shown in Figure 7-6, two methods are used for obtaining outputs and performing data verification. The data obtained from each method are presented in Table 7-3. The details of each method and their outputs are discussed in this section.

Method #1 outputs are the advanced EnergyPlus outputs including the detailed energy consumption of each component in each thermal zone, separately. These data are stored in a spreadsheet within output folders after running EnergyPlus. The spreadsheet file should be opened manually and data need to be extracted, assuming the user knows the titles used for each component because the names assigned to the building envelope components do not represent type of the building component and its corresponding thermal zone. Figure

7-9 illustrates some of the titles used for each building envelope component within the spreadsheet generated by Revit during exporting the architectural model to gbXML file.

These names are based on the titles assigned to each thermal zone in Revit. If proper names are assigned, then it would be easy to interpret and organize the E+ advanced outputs; however, in case the titles are not clear that which thermal zone these components belong

333 to, BEPAT resolves this issues by generating and saving additional data showing their corresponding thermal zone based on the data available in EnergyPlus. For example, in this study the model developed in Revit uses 12, 13, 14, and 15 to indicate different zones, which are corresponding to thermal zone number 2, 4, 3, and 1 shown in Figure 6-11. This could be confusing and causes errors in output interpretation; therefore, as it is shown in

Figure 7-10, BEPAT assign a simple number from 1 to 4 to each wall component. Outputs related to Method #1 are presented in Table 7-3 to be compared with BEPAT outputs, which are supposed to be the same numbers but obtained directly from EnergyPlus and stored in a different path into series of text files as shown in Figure 7-7. Figure 7-11 shows the tags used for each component that can be used to facilitate interpreting data in Table

7-3.

Figure 7-9. Examples of name assigned to building envelope components in the spreadsheet outputs

generated by EnergyPlus advanced outputs

Method #2 outputs are the BEPAT’s detailed outputs, which include similar data to the first methods except that data acquisition is semi-automated and the intended data will be stored in a series of text files for each building envelope component by an easier, faster, and more accurate method. Figure 7-10 shows the automatically generated text files named after each

334 category of building envelope component, which contain the heat transfer data for each component and its corresponding thermal zones. The text file containing the data related to walls heat gain/loss is also illustrated in Figure 7-10 as an example. Comparison of these values with data adopted from EnergyPlus presented in Table 7-3, which shows BEPAT also obtain similar data and proves the functionality of this tool.

Figure 7-10. Automatically generated text files containing heat gain/loss data for each component and

their thermal zones

For a better understanding of what BEPAT does, the procedure to obtain the outputs in each method could be reviewed. The following steps explain the process to get the detailed values under “advanced E+ outputs” column in Table 7-3 from EnergyPlus using IDF file, which is method #1:

335

1) Open the IDF file in EPlaunch, which is an IDF editor tool.

2) Pick the advanced outputs option to enable accessing detailed outputs by defining

a new object in Output:Variable.

3) Pick the related outputs such as Surface Window Heat Loss Energy, Surface

Window Heat Gain Energy, and Surface Average Face Conduction Heat Transfer

Energy among about 140 advanced outputs that need deep understanding of each

item.

4) Run EnergyPlus and open the related spreadsheet output file.

5) Check the name of each component in the IDF file and determine the corresponding

wall number in order to find other properties such as thermal zone.

On the other hand, as metheod #1, BEPAT is used to obtain the values under the “Direct outputs of modified E+” column in Table 7-3. It obtains the detailed values automatically and indicates the thermal zone of each component in separate text files. The process is faster and reduces the risk of human errors. The breakdown of the process adopted in using

BEPAT could be as follow:

1) Copy the IDF and weather file under the tools installation folder.

2) Run the executable file.

3) Open the generated text files related to each category of building envelope

component such as walls or windows to check out the amount of heat transfer

through each of them.

336

It turns out that the alternative process used in Method #1 could be complex, excessively time consuming, and error prone. In addition, it requires certain software related skills and understanding of components in energy modeling. Advanced outputs need to be selected manually and the detailed outputs are not easy to comprehend because the titles used for each component do not necessarily represent the thermal zones. BEPAT, on the other hand, identifies the thermal zone of each building envelope component and also the detailed heat transfer.

Given that standard energy simulation tools only provide accumulated energy consumption data such as whole-house energy consumption, or zone energy performance, standard outputs of EnergyPlus includes the accumulated energy consumption of components such as walls and windows for each thermal zone. These outputs are the standard outputs presented in all the GUIs using EnergyPlus as the energy simulation engine such as

OpenStudio, BEopt, and DesignBuilder. If EnergyPlus is used without any GUI, these data can be found in the Annual Building Sensible Heat Gain Components table in EnergyPlus outputs shown in Figure 7-12 as an example where all the data are accumulated for each thermal zone without any detailed data for each envelope component. Moreover, the energy simulation in this study is simplified in the sense that additional inputs such as infiltration, lights, people’s sensible heat, and heated surfaces are not considered in energy modeling, i.e., their effect neglected since the outputs are only focused on energy gain/loss through building envelope components.

337

Figure 7-11. Building components tags

Figure 7-12. Annual building sensible heat gain comopnents table, generated by E+ containing

accumulated data for each thermal zone

338

Table 7-3. Summary of energy simulation outputs obtained from standard and modified EnergyPlus

Advanced outputs Direct outputs of Zones and of E+ (GJ) BEPAT (GJ) components (Method #1) (Method #2)

Gain Loss Gain Loss Ext. 1 -0.617 -0.629 Ext. 2 -1.469 -1.498 Ext. 3 -0.585 -0.595 Int. 1 ±0.024 ±0.024 Int. 2 ±0.015 ±0.015 Int. 3 ±0.017 ±0.017

Zone 1 Flr. 1 -0.592 -0.609 Rf. 1 -1.444 -1.470 Win. 1 2.033 -0.934 2.033 -0.934 Win. 2 0.766 -0.958 0.766 -0.958 D. 1 -1.220 -1.243 Ext. 4 -0.585 -0.595 Ext. 5 -0.433 -0.443

Int. 1 ±0.024 ±0.024 Int. 4 ±0.002 ±0.003 Flr. 2 -0.248 -0.254 Zone 2 Rf. 2 -0.530 -0.539 Win. 3 2.598 -0.958 2.598 -0.958 D. 2 ±0.037 ±0.037 Ext. 6 -0.343 -0.351 Int. 2 ±0.015 ±0.015 Int. 4 ±0.002 ±0.003 Int. 5 ±0.004 ±0.004

Zone 3 Flr. 3 -0.167 -0.170 Rf. 3 -0.373 -0.379 D. 3 ±0.034 ±0.034 Ext. 7 -0.299 -0.306 Ext. 8 -0.539 -0.549 Int. 3 ±0.017 ±0.017 Int. 5 ±0.004 ±0.004 Flr. 4 -0.188 -0.193

Zone 4 Rf. 4 -0.386 -0.393 Win. 4 2.574 -0.928 2.574 -0.928 Win. 5 1.724 -0.949 1.724 -0.949 D. 4 ±0.039 ±0.039

Comparison of EnergyPlus advanced outputs related to average heat conduction and energy loss/gain through opaque and transparent building envelope with BEPAT’s outputs helps validate BEPAT’s results. For example, exterior walls for zone 2, which includes Ext.4 and

Ext.5 walls, are calculated by EnergyPlus advanced outputs to let 0.585 and 0.433 GJ of

339 heat energy out, respectively. As it is presented in Table 7-3, BEPAT’s outputs shows

0.595 and 0.443 GJ of heat loss through these walls, respectively. Obtaining detailed outputs using BEPAT, which can be obtained through a straightforward and accurate process, eliminates the need for sophisticated conventional energy simulation process.

7.6 Summary and Conclusions

This paper has discussed development of software capability Building Energy Performance

Assessment Tool (BEPAT), a tool for monitoring and evaluating the amount of heat transfer through building envelope components in an automated, more reliable, and faster process compared to conventional energy modeling and simulation process using energy simulation tools such as EnergyPlus. It can contribute to 1) energy smart homes, which target improving the energy performance of a building and automation of data acquisition and monitoring; 2) design phase of buildings by providing detailed information on the amount of heat transfer through each component for optimization methods; and 3) decision-making process of building energy retrofit. The approach used in developing

BEPAT is editing the source code of EnergyPlus to generate an executable file. This tool can provide detailed energy-related information as opposed to accumulated heat transfer.

It is important to be able to monitor and assess each component (e.g., an exterior wall) separately compared to total energy loss through the exterior walls all together. The detailed data on heat gain/loss of building envelope components can help improving and facilitating the decision making process for energy retrofit of a building. It would be more

340 cost effective to be able to recognize specific envelope components of the building, which contribute to energy loss more than other components.

The BEPAT tool requires an IDF and weather file to generate five separate text files containing detailed data on heat gain/loss through building envelope components, including windows, doors, walls, roofs, and floors. Moreover, the title of the thermal zone of each component is also assigned to it in order to make it easier to interpret and use the outputs in other GUIs for monitoring, optimization, and evaluation purposes. These data could be used directly from text files or could be linked to another GUI specifically designed for smart homes or building energy retrofit in order to visualize the outputs. Users can monitor zone energy consumption and the amount of heat transfer through each component in building envelope and check out the energy gain/loss. Moreover, the generated detailed data could be used for further statistical analysis and provide useful information such as the percent contribution of each component in total energy consumption. These pieces of information could be beneficial for decision-making process during the design phase of the building or energy retrofit of existing homes.

In order to validate BEPAT, a building with four thermal zones is modeled in Revit. The concept of automation is an important feature in newly emerging technologies such as smart homes and design process. BIM was thus used to generate the IDF file automatically and perform building energy simulation as opposed to performing a conventional energy modeling process. Revit is used as the BIM authoring tool to generate a gbXML file.

OpenStudio is used to read the BIM file and add additional data concerning HVAC system

341 and schedules, which are required for energy analysis. The IDF file is generated using

OpenStudio to be used in two different ways including using BEPAT and obtaining advanced outputs of EnergyPlus to verify BEPAT’s outputs. Standard outputs of

EnergyPlus only provide accumulated data while advanced outputs could provide detailed data about the amount of heat transfer through each building envelope component.

However, the process for obtaining these data through EnergyPlus is complex, time consuming, and error prone. In addition, the outputs are not easy to comprehend because the thermal zone of each component is not provided. On the other hand, BEPAT provides detailed information automatically and saves them in text files with additional information about their thermal zone, which makes it easy to interpret.

Based on the results of the study, the following concluding remarks are offered:

 Automation in building energy modeling and providing detailed data related to

energy consumption for monitoring purposes can contribute to efforts toward

optimizing the building design, energy retrofit decision-making process, and

development of energy smart homes. For example, it can contribute to identifying

the thermal zones or building envelope components, which has higher contribution

in building energy consumption and could be retrofitted.

 Current energy simulation tools cannot provide detailed energy consumption data

in an easy and straightforward process. Facilitating the data acquisition and making

the process faster and more accurate could contribute to both energy simulation

342

tools, energy smart homes, improving the decision-making process of building

energy retrofit, and optimizing the design during the design-phase.

 With regard to EnergyPlus outputs and data related to the amount of heat transfer

through building envelope components, they are not easily interpreted since the

thermal zones and other properties are not associated with them. BEPAT, on the

other hand, appends additional information to these outputs to make it easier for

interpretation.

 Application of BIM in energy simulation can facilitate the process; however, there

are still deficiencies, as it is not fully automated for functions such as transferring

data related to mechanical systems between different drawing and energy

simulation tools.

 It was noted that names assigned to each building envelope components in

EnergyPlus and BEPAT outputs are based on the names assigned to these

components in Revit to gbXML file export. It could be easier for future applications

if the names used in the BIM file are kept the same through the process to make the

outputs easier to interpret, which might need development of another middleware

to perform such a task.

 Multiple energy-related information are calculated within EnergyPlus such as

detailed information on different types of heat transfer contributed by radiation,

conduction, and convention; however, not all of them are presented and available

in energy simulation tools and GUIs. It is recommended that similar studies be

343

performed and similar tools be developed focused on such information, which

could make acquiring such data easier and more accurate.

BEPAT can provide detailed data about the amount of heat transfer through building envelope components by conduction. The provided accurate and easy to interpret data can be used in GUIs designed for energy smart homes and could be used for building energy retrofit purposes or design process of buildings.

7.7 References

[9] W. Kim, S. Lee and J. Hwang, "Real-time Energy Monitoring and Controlling

System based on ZigBee Sensor Networks," Procedia Computer Science, vol. 5, p. 794–

797, 2011.

[11] S. Helal, W. Mann, H. El-Zabadani, J. King, Y. Kaddoura and E. Jansen, "The

Gator Tech Smart House: A Programmable Pervasive Space," IEEE Computer Society, vol. 38, no. 3, pp. 50-60, 2005.

Congress. Energy, Environment and Sustainable Development, Perugia, Italy, 2014.

344

[24] "Advanced Energy Retrofit Guide, Practical Ways to Improve Energy

Performance, K-12 Schools," National Renewable Energy Laboratory for U.S. Deparment of Energy, 2013.

[34] "Advanced Energy Retrofit Guide, Practical Ways to Improve Energy

Performance, Office Buildings," Pacific Northwest National Laboratory for U.S.

Department of Energy , 2011.

[40] M. D. Bianco and J. Wiehagen, "Using Retrofit Nail Base Panels to Expand the

Market for Wall Upgrades," U.S. Department of Energy’s Building America Program,

Denver, 2016.

[41] T. Häkkinen, M. Ala-Juusela and J. Shemeikka, "Usability of energy performance assessment tools for different usepurposes with the focus on refurbishment projects,"

Energy and Buildings, vol. 127, pp. 217-228, 2016.

[43] M. Alam, P. X.W. Zou, J. Sanjayan, R. Stewart, O. Sahin, E. Bertone and J. Wilson,

"Guidelines for Building Energy Efficiency Retrofitting," in Sustainability in Public Works

Conference, Melbourne, Victoria, 24 – 26 August 2016.

345

[56] V. Bazjanac, "IFC BIM-Based Methodology for Semi-Automated Building Energy

Performance Simulation," in 25th International Conference on Information Technology in

Construction, Santiago, Chile, 2008.

[87] Q. Le, H. Nguyen and T. Barnett, "Smart homes for older people: Positive aging in a digital world," Future Internet, vol. 4, pp. 607-617, 2012.

[143] M. LaMonica, "Nest Thermostat Slays Peak Power," 22 April 2013. [Online].

Available: https://www.technologyreview.com/s/513961/nest-thermostat-slays-peak- power/. [Accessed 11 November 2016].

[146] G. Lobacco, S. Carlucci and E. Lofstrom, "A Review of Systems and Technologies for Smart Homes and Smart Grids," Energies, vol. 9, no. 5, pp. 1-33, 2016.

[174] O. Omole, D. Akpobasah and A. Atayero, "Development of a Smart, Low-cost and

IoT-enabled System for Energy Management," 2016.

[222] "Advanced Energy, Practical Ways to Improve Energy Performance, Retail

Buildings," Pacific Northwest National Laboratory for U.S. Department of Energy, 2011.

[223] "Advanced Energy Retrofit Guide, Practical Ways to Improve Energy

Performance, Healthcare Facilities," National Renewable Energy Laboratory for U.S.

Department of Energy, 2013.

[232] F. AlFaris, A. Juaidi and F. Manzano-Agugliaro, "Energy retrofit strategies for housing sector in the arid climate," Energy and Buildings, vol. 131, pp. 158-171, 2016.

346

[233] R. Chang, S. Hayter, E. Hotchkiss, S. Pless, J. Sielcken and C. Smith-Larney,

"Aspinall Courthouse: GSA’s Historic Preservation and Net-Zero Renovation," U.S.

Department of Energy, 2014.

[236] "Beardmore Building Case Study," 27 April 2012. [Online]. Available: http://newbuildings.org/resource/beardmore-building-case-study/. [Accessed 14

September 2016].

[267] "Case Studies, Empire State Building, New York," [Online]. Available: http://www.rmi.org/retrofit_depot_get_connected_true_retrofit_stories##empire.

[Accessed 14 September 2016].

[306] L. Santos, S. Schleicher and L. Caldas, "Automation of CAD models to BEM models for performance based goal-oriented design methods," Building and Environment, vol. 112, pp. 144-158, 2017.

[307] T. Maile, M. Fischer and V. Bazjanac, "Building Energy Performance Simulation

Tools—A Life-Cycle and Interoperable Perspective;," Stanford University: Stanford, CA,

USA, 2007.

[308] S. Pinheiro, J. Donnell, R. Wimmer, V. Bazjanac, S. Muhic, T. Maile, J. Frisch and

C. van Treeck, "MODEL VIEW DEFINITION FOR ADVANCED BUILDING ENERGY

PERFORMANCE SIMULATION," in CESBP/BauSIM 2016 Conference, Dresden,

Berlin, 2016.

347

[324] J. O'Donnell, R. See, C. Rose, T. Maile, V. Bazjanac and P. Haves, "SIMMODEL:

A DOMAIN DATA MODEL FOR WHOLE BUILDING ENERGY SIMULATION," in

12th Conference of International Building Performance Simulation Association, Sydney,

2011.

September 2016, 2016.

[342] E. Guzmán Garcia and Z. Zhu, "Interoperability from building design to building energy modeling," Journal of Building Engineering, vol. 1, pp. 33-41, 2015.

[348] S. Aman, Y. Simmhan and V. Prasanna, "Energy Management Systems: State of the Art and Emerging Trends," IEEE Commun. Mag., vol. 51, no. 1, pp. 114-119, 2013.

[349] Z. Tian, W. Chen, P. Tang, J. Wang and X. Shi, "Building Energy Optimization

Tools and Their Applicability in Architectural Conceptual Design Stage," Energy Procedia, vol. 78, p. 2572 – 2577, 2015

[350] R. Pacheco-Torres, M. López-Alonso, G. Martínez and J. Ordóñez, "Efficient design of residential buildings geometry to optimize photovoltaic energy generation and energy demand in a warm Mediterranean climate," Energy Efficiency, vol. 8, p. 65–84,

2015.

348

[351] V. Granadeiro, J. R. Correia, V. M.S. Leal and J. P. Duarte, "Envelope-related energy demand: A design indicator of energy performance for residential buildings in early design stages," Energy and Buildings, vol. 61, p. 215–223, 2013.

[352] D. B. Crawley, J. W. Hand, M. Kummert and B. T. Griffith, "Contrasting the capabilities of building energy performance simulation programs," Building and

Environment, vol. 1, pp. 95-110, 2008.

[353] D. Denis, "Bottom-up modelling of energy demand and technical energy savings potential in the Irish residential sector," National University of Ireland, Thesis submitted for the degree of Doctor of Philosophy, Cork, 2014.

[354] "EnergyPlus," the U.S. Department of Energy’s (DOE) Building Technologies

Office (BTO), 2017. [Online]. Available: https://energyplus.net/. [Accessed 16 February

2017].

[355] GitHub, "NREL/EnergyPlus," 2017. [Online]. Available: https://github.com/NREL/EnergyPlus. [Accessed 7 February 2017].

349

8) Chapter 8. Automated Building Energy Modeling and

Assessment Tool (ABEMAT)

This chapter has been written as a journal paper and is already submitted for review.

350

Automated Building Energy Modeling and Assessment

Tool (ABEMAT)

1 2 Ehsan Kamel , Ali M. Memari

1Ph.D. Candidate, Department of Civil Engineering, Structural Engineering, Penn State University, 321 Sackett Building, University Park, PA 16802 (corresponding author). E- mail: [email protected] 2Professor, Hankin Chair in Residential Building Construction, and Director, Pennsylvania Housing Research Center, Dept. of Architectural Engineering and Dept. of Civil and Environmental Engineering, Penn State Univ., University Park, PA 16802. E-mail: [email protected] 8.1 Abstract

Saving even a small percentage of total building’s energy consumption in the U.S., which contributes to about 40% of total annual energy consumption, can lead to significant economic and environmental impacts. Building Energy Modeling (BEM) tools play an important role in reducing building’s energy consumption. Most of the available tools only provide total energy consumption of each sector in a building such as HVAC system, lighting, hot water, and appliances. While fine-grained data may be available in some computer tools, the data acquisition process is time-consuming, error-prone, and requires software-related skills.

This paper explains a developed tool that contributes in automation of building energy modeling, simulation, and providing fine-grained outputs by using Building Information

Modeling (BIM) and modifying the source code of existing energy simulation tools such as EnergyPlus and OpenStudio. This tool, which is referred to as ABEMAT (Automated 351

Building Energy Modeling and Assessment Tool), receives gbXML file and provides users with the amount of heat transfer through every single building envelope component such as windows, doors, walls, roof, and floors through an automated process. Comparing

ABEMAT’s outputs with EnergyPlus advanced outputs, showed similar results indicating the functionality of ABEMAT in energy modeling and providing fine-grained outputs.

Keywords: Building Information Modeling (BIM), Building Energy Modeling (BEM),

Smart Homes, gbXML, OpenStudio, EnergyPlus

8.2 Introduction

Buildings consume about 40% of the total energy in the U.S., and to reduce such high level of energy consumption, multiple measures are needed. Building’s design can be made to be more energy efficient by modifying building properties such as orientation, dimensions, and building envelope materials/details. In addition, the energy consumed during the use phase of buildings can be decreased by either monitoring the energy consumption and its incentive effects or by energy retrofit of buildings, which could be benefitted by emerging tools such as energy smart homes and BEM tools. These measures and all the tools and technologies involved in their development are illustrated in Figure 8-1, where all of these tools are shown to be interconnected and working together. For example, to evaluate the design of a building developed in a CAD tool in terms of energy performance, an energy simulation tool is required. In addition, energy smart homes can also be equipped with energy simulation tools in order to monitor and optimize the energy consumption. A brief

352 introduction about the areas and tools involved in this research including BIM files, application of BIM in BEM, and building’s energy retrofit is presented in this section.

Figure 8-1. Different approaches for reducing energy consumption, and different tools involved

BIM files may include data related to multiple aspects of buildings such as mechanical systems, electrical systems, schedules, structural, and architectural components. Two of the most prevalent BIM file schemas are Industry Foundation Classes (IFC) and Green

Building Extensible Markup Language (gbXML). Application of BIM in different areas such as developing the 3D model of buildings, structural analysis, and energy modeling has been of interest and subject of some past studies. Information related to the geometry, materials, HVAC system, and costs can be stored and transferred through these files, which makes the modeling process faster, less error-prone, and more toward automated process, while making it easier for different parties in design and construction process to

353 communicate and facilitate visualizing data [290, 291, 289, 356, 357]. Both of these schemas are capable of storing and transferring most of the data required for energy simulation such as defining thermal zones, material thermal properties, material thickness, and HVAC system properties. However, there are some shortcomings in these file schemas related to the geometry of buildings, location, and domain of the application. For example, gbXML is only capable of transporting the data related to rectangular building geometry, while IFC works with other geometry types [322, 314, 331, 356, 303, 301, 308, 323].

Although these file schemas are capable of handling the data related to energy simulation, some of the tools being used for developing or importing these files are not capable to perform the process flawless and properly. Application of BIM in BEM can be explained in more details, where shortcomings can be identified and appropriate solutions developed.

Applications of BIM in building energy modeling (BEM) consists of multiple components such as BIM files and conversion of BIM file to a readable file for energy simulation tools.

Each of these components can experience shortcomings and issues during the process. The main core of the process includes using a BIM file for Building Energy Modeling (BEM) shown in Figure 8-2, which can be referred to as BIM-to-BEM interoperability process

(BBIP) or simply BIM-to-BEM. Some studies have addressed adoption of BIM-to-BEM approach in design phase of buildings to optimize it [358, 359, 360]. The challenges and issues in BBIP are identified in some studies to discuss the shortcomings related to CAD tools and mapping information to a BIM file, the process of mapping data from a BIM file to a readable file for BEM tools, and the interoperability issues during translating data for an energy simulation engine [307, 314, 48, 332, 315, 293]. On the other hand, there are

354 solutions suggested and used by some researchers to resolve these issues such as using middleware tools, which work between BIM and BEM tools, adding the missing information manually, and linking the files to database servers to add the required information missing during the process [56, 341, 311, 310, 305, 54]. Adopting a middleware corrective tool is the same approach used in this study using Python, which will be explained in more details subsequently.

As shown in Figure 8-1, energy retrofit can be another approach in reducing the energy consumption of existing buildings [361]. Energy retrofit can target different components of buildings such as HVAC system, electrical system, appliances, and envelope. ABEMAT and this study are focused on building envelope including both opaque and transparent components such as walls and windows, respectively. Different computer tools can be adopted prior to performing energy retrofit, and the data provided by these energy simulation tools can contribute to evaluating the existing conditions of buildings in order to improve the decision making process of energy retrofit. Current energy simulation tools such as OpenStudio, DesignBuilder, and BEopt, which are used in energy retrofit of buildings, allow consideration of multiple criteria in choosing suitable energy retrofit scenarios [31, 362, 245, 246, 247, 33]. Most of these tools only provide accumulated energy-related data, for example the whole-house energy consumption or the total heat transfer through all the components in one thermal zone. Even if the detailed outputs are provided by an energy simulation tool, they are not presented in existing commercial computer tools or they are not easy to access. The process to obtain such data can be time consuming, error-prone, and requires software-related skills. It would be of great interest

355 to have detailed energy consumption of each building envelope component such as the amount of heat transfer through a specific window or exterior wall as opposed to the total heat transfer through all windows and walls in a fast, easy, and more accurate approach.

ABEMAT can provide such data by saving the amount of heat (gain or loss) transfer, through each building envelope component and floors in separate text files for walls, windows, floors, roofs, and doors.

As shown in Figure 8-1, energy smart homes can also contribute to reducing energy consumption in buildings, for example through smart control of appliances and energy monitoring and its incentive impacts. Shortcomings also exists in energy smart homes, as these homes are neither equipped with energy simulation tools within their processing units, nor are the adopted simulation tools capable of providing fine-grained or detailed energy consumption data. Therefore, application of tools such as ABEMAT can contribute to providing fine-grained data on heat transfer through building envelope components to users in an automated and fast way, which is more compatible with emerging technologies and tools such as energy smart homes.

The tool introduced in this paper, which is focused on building envelope components and floors, automates the whole process of building energy modeling, simulation, and providing fine-grained outputs by using BIM and modifying or using the source code of existing tools in energy simulation such as EnergyPlus and OpenStudio. As a result, it is referred to as Automated Building Energy Modeling and Assessment Tool (ABEMAT).

The overview of its process and contributing components are presented in Figure 8-2,

356 which indicates that several tools and components are involved in its process. The process within ABEMAT includes the following major steps:

1) Receive a gbXML file, which can be generated by any BIM tool such as CAD tools

including Revit.

2) Use a corrective tool developed based on using Python to resolve some of the issues

occurring during file conversion.

3) Call the subroutines and functions defined in OpenStudio using Ruby to convert

the gbXML file to Input Data File (IDF) for EnergyPlus (E+)

4) A modified EnergyPlus using C++ performs the energy analysis and provides

detailed information on heat transfer through building envelope components such

as windows and walls and saves them in separate text files for further use.

Figure 8-2. Different tools and components involved in development of ABEMAT

The methods used in order to develop each component in ABEMAT are explained in details under the methodology section. Different programming languages including C++,

Python, and Ruby are used to develop a new tool or modify an existing tool such as

EnergyPlus or OpenStudio. ABEMAT is only capable of reading gbXML files at this point, and its outputs are saved in text files. In order to verify the outputs and process within

357

ABEMAT, a computer model is also developed using Revit with EnergyPlus used as the energy simulator. The process and results are explained further subsequently.

8.3 Methodology for Developing ABEMAT and Data

Verification

As shown in Figure 8-2, three major tools are involved in ABEMAT including a corrective tool for gbXML files, certain modules and subroutines in OpenStudio source code adopted by using Ruby, and modified EnergyPlus using C++. As discussed earlier, data transfer using BIM for energy modeling faces some challenges, which requires adoption of different approaches to resolve them. However, at this stage, ABEMAT is focused on building envelope components, and as a result, issues such as missing information are resolved by adding the data manually. Hence, the automated process of ABEMAT does not cover areas related to HVAC system, electrical system, appliances, or schedules.

Moreover, the Integrated Development Environment (IDE) used in this study is

VisualStudio (version 14.0 Update 3, 2015), when programming is performed in C++ and

Python. The following sections explain each of the three major components of ABEMAT in details.

8.3.1 First component of ABEMAT: gbXML corrective tool

The first component of ABEMAT is the corrective tool for BIM files and specifically for gbXML file schema. It receives the exported file from a CAD tool such as Revit and resolves some of the issues, which occur during the BBIP related to building envelope 358 components. Prior to developing this tool, a pilot study was performed in order to identify the potential issues, which can occur related to building envelope data through the BBIP.

Accordingly, a simple one-story building was modeled in Revit, which was further enhanced for the main case study in this paper. A gbXML file was exported from Revit to be used by OpenStudio GUI for energy analysis. Two major issues were noticed through the BBIP including the error shown in Figure 8-3 is generated by OpenStudio related to duplicate adjacent spaces for floors, which causes errors during gbXML to IDF file conversion within OpenStudio. The second issue shown in Figure 8-4 occurs during the energy analysis process while EnergyPlus is reading the IDF file, which is related to missing construction ID for doors.

Figure 8-3. Errors generated in OpenStudio related to similar adjacent space name

Figure 8-4. Missing door’s construction ID in gbXML file causes error during energy simulation 359

In order to resolve these issues, a program is developed using Python to read and generate a corrected XML file. It consists of four functions with different tasks including finding duplications, replacing incorrect lines with corrected ones, finding lines related to doors, and finding the construction data related to doors. The last three functions work interactively to look for the data related to doors, extract the door’s construction reference

ID, and replace the line with a new one that includes the construction ID. The first function that looks for duplication spots the repeated lines related to adjacent spaces and removes one of them to eliminate the warnings illustrated in Figure 8-5.

Figure 8-5. Errors related to building envelope through BIM-to-BEM process

In ABEMAT, this tool starts the process by reading the gbXML file and correcting it. The pilot study shows that the corrective tool is able to perform the intended tasks including removal of the duplicate adjacent spaces and addition of the missing information about the door’s construction IDs within the BIM file as it is shown in Figure 8-4. Its functionality is further verified as a component within ABEMAT working together with other components, which will be covered in the follow-up section.

360

8.3.2 Second component of ABEMAT: gbXML to IDF converter

The second component in ABEMAT is a file converter, which converts a gbXML file to

IDF. While some researchers have used MATLAB [363] to convert gbXML files to IDF, this study performs this task by using existing modules within OpenStudio, which is an open-source graphical user interface (GUI) for energy modeling and simulation. One of the

OpenStudio capabilities is to import a BIM file in either gbXML file format or Industry

Foundation Classes (IFC) format to perform an energy simulation for which the engine used by OpenStudio is EnergyPlus; therefore, the BIM files imported to OpenStudio need to be converted to IDF, which is the input file format for EnergyPlus. The process of importing a BIM file and performing energy simulation in OpenStudio is typically performed through its GUI, which means the process cannot be automated and requires human interaction. However, OpenStudio provides Software Development Kit (SDK) and proper online documentations for users. It helps software developers to access the source code of OpenStudio using Ruby bindings, for example, to perform similar tasks as the GUI such as converting a gbXML file to IDF. This then led the current study to develop a measure using Ruby bindings and OpenStudio source code, which performs the following tasks:

1) Look for the gbXML file in a predefined location on the computer

2) Convert the gbXML file to .OSM file, which is the OpenStudio file using

ReverseTranslator class under the gbXML objects

361

3) Convert the .OSM file to IDF using ForwardTranslator class under the EnegryPlus

objects

4) Save the final IDF file in a predefined location on the computer

All the aforementioned tasks are performed using Ruby commands, which are saved in a single text file, so it can be run by Ruby automatically when needed. The input and output of this measure are the gbXML file and IDF file, respectively.

The generated IDF file needs to be fed to the modified EnergyPlus as the input. One of the shortcomings of BBIP is the issue of missing data related to different components (e.g.,

HVAC systems or schedules) during mapping data from CAD tool to BIM file. In this study, Python is used as the corrective tool to resolve the issues associated with building envelope systems; however, issues related to other aspects are resolved by adding data manually. Accordingly, the properties of the HVAC system, schedules, and thermostat set points, which are not exported properly from Revit, are added manually to the IDF file using IDF editor shown in Figure 8-6. The properties of mechanical systems will be explained in more details in the next section. As a result, the actual process used in this paper shown in Figure 8-2 can be modified as shown in Figure 8-7. Manual data input approach is used, which is also adopted by other researchers in similar studies [340, 56].

362

Figure 8-6. IDF editor used for modifying IDF file to add missing data

Figure 8-7. Modified process of ABEMAT to add HVAC, schedules, loads, and thermostat set points data

manually

363

8.3.3 Third component of ABEMAT: IDF to fine-grained heat

transfer outputs

The modified EnergyPlus is the third and last tool used in ABEMAT. The EnergyPlus source code is an existing open-source energy simulation software developed by National

Renewable Energy Laboratory (NREL) [354], which makes it possible to modify the source code based on different needs. The main contribution of this component is automating the saving and acquisition process for detailed outputs concerning the amount of heat transfer through building envelope components and floors. The shortcomings related to reporting the outputs in existing energy simulation tools include lack of detailed or fine-grained information as opposed to accumulated data, issues with access to detailed data, and error-prone process of data acquisition because of manual process in the absence of software-related skills. The modified EnergyPlus retrieves the IDF generated in the previous step and performs the needed energy simulation. This tool is an executable file, which can be run automatically after the proper IDF file is generated. The outputs consist of six separate text files for doors, windows, walls, floors, and roof containing the amount of heat transfer through every single component in a building. This tool is developed by modifying EnergyPlus 8.6.0 source code.

The source code of EnergyPlus consists of about 275 modules in C++, where Figure 8-8 illustrates some of the modules such as AirflowNetworkSolver, DataAirLoop, and

ChillerAbsorption modules, which can be viewed and modified in VisualStudio. Modules are interconnected in a certain pattern, with each module being focused on different parts

364 of the whole-house energy consumption calculations such as conduction, convection, or radiation modes of heat transfer through different components of the HVAC system, lighting, shadings, and appliances. In this study, it is important to identify the modules related building envelope energy-related calculations in order to modify them.

Figure 8-8. Some of the modules in EnergyPlus source code

The building envelope components targeted by ABEMAT include both opaque and transparent components, and as the first step, it was required to identify the modules contributing to heat transfer calculations related to such components. Two of the modules contributing to heat balance calculations and other data related to different surfaces within the modeled building include HeatBalanceSurfaceManager and DataSurfaces, which are modified for use in ABEMAT. In addition, another module

OutputReportTabular includes the final outputs and can be modified in order to retrieve

365 some of the required data in ABEMAT. The code developed in C++ is added in these modules and additional required changes are applied in the header files (.hh files), accordingly. The modifications applied in C++ perform the following tasks:

- Define a new variable that collects the detailed data as opposed to accumulated

data.

- Collect the data before EnergyPlus sums them up for each thermal zone.

- Add them up for each time step of the analysis in order to obtain the total heat

gain/loss through the simulation run period.

- Generate a text file in the root directory of the modified EnergyPlus executable file.

- Write the total heat gain/loss for each opaque and transparent component in each

thermal zone, separately.

- Assign a thermal zone and proper name to each component in order to facilitate

identifying these components and their corresponding thermal zones to be used in

other tools in the future.

The last task indicates another important aspect of reporting fine-grained outputs. Different titles are used for each surface of building by EnergyPlus. However, it is important to keep the same names assigned to each surface by the BIM tool, which is a CAD tool (Revit) in this study. It makes data handling easier for follow-up applications and facilitates interpretation of outputs. The third component of ABEMAT that performs the energy analysis saves the amount of heat transfer for each surface under similar names generated by BIM tool. In addition, it is important to identify the thermal zone each building envelope

366 component is associated with; therefore, ABEMAT also saves the corresponding thermal zone. Figure 8-9 shows an example of ABEMAT’s outputs and illustrates how a thermal zone is addressed and associated with a proper title for each building envelope component.

For example, in this study the model developed in Revit uses 12, 13, 14, and 15 to indicate different thermal zones, which corresponding to thermal zone numbers 2, 4, 3, and 1 shown in Figure 8-13. This can be confusing and may cause errors in output interpretation; therefore, as it is shown in Figure 8-9, ABEMAT assigns a simple number from 1 to 4 to each component.

Figure 8-9. Example of ABEMAT’s outputs to illustrate the titles of each component and associated thermal zone

These three major components of ABEMAT cover the BIM-to-BEM process, automate the whole process related to building envelope components, and facilitate it by speeding it up and improving the accuracy. ABEMAT bundles up all three components and provides fine-

367 grained outputs using an input gbXML file, which needs to include all the required data for energy simulation. However, in this study, pieces of information related to building components other than the envelope, are provided manually. Studying the capability of existing tools in transferring other types of information such as HVAC, schedules, and loads can be the focus of follow-up studies.

8.3.4 Methodology for Data Validation

A model of a one-story building with four thermal zones was developed in Revit for use in data validation process, which is explained in more details in the next section. ABEMAT’s fine-grained outputs cannot be obtained from any other energy simulation GUI. However,

EnergyPlus can provide similar data through a more complicated process compared to adopting ABEMAT. In this study, the method suggested in Figure 8-10 is used for data validation to prove its functionality and valid outputs as an alternative for ABEMAT even though, compared to ABEMAT, the process is time consuming and complicated.

Figure 8-10. Adopted process for validating ABEMAT’s outputs

A BIM file under gbXML format is generated by Revit and is fed to OpenStudio GUI.

Other required data for energy simulation such as HVAC components, schedules, and thermostat set points, which might be among the missing information through

368 interoperability process, are added manually using OpenStudio GUI. Finally, it will be exported to an IDF file for the final analysis by standard EnergyPlus.

In order to obtain fine-grained outputs from standard EnergyPlus, first, this capability should be enabled by either modifying the IDF text file or using the IDF editor and adding new items under Output: Variable object shown in Figure 8-11. It can be observed there are multiple options available as advanced outputs of EnergyPlus, which require software- related skills to be able to pick the proper one to obtain fine-grained outputs as opposed to

ABEMAT, which performs this task automatically, fetches, and saves detailed outputs in separate text files. Eventually, detailed outputs will be saved in a spreadsheet file to be compared with ABEMAT’s outputs. The advanced EnergyPlus outputs used in this study for data validation include heat gain or loss through window and average conduction heat transfer energy.

Figure 8-11. Enabling EnergyPlus to obtain advanced outputs using IDF editor

369

8.4 Details of Modeling and Results of Data Validation

To validate the outputs of ABEMAT, a model of a one-story building with four thermal zones is developed in Revit, which can export the model into two of the most prevalent

BIM file types including IFC and gbXML. In this study, the architectural model is exported to a gbXML file to be used in two different methods shown in Figure 8-7 and Figure 8-10, which explain ABEMAT and data validation process, respectively.

The BIM file directly exported from Revit does not include certain data such as HVAC system, loads, and thermostat set points, which are needed for comprehensive energy simulation. Therefore, such missing data need to be added to the IDF file. ABEMAT resolves these issues using two approaches. Issues related to building envelope are resolved using the gbXML corrective tool shown in Figure 8-7 and explained as the first component of ABEMAT. However, other data such as HVAC and thermostat set points need to be added manually to the IDF file, since they are not currently the focus of ABEMAT. The method used for data validation shown in Figure 8-10 resolves these issues using two different approaches. First, the issues related to the building envelope components within

BIM file are resolved by modifying the gbXML file manually. Then, the rest of the required data related to HVAC and schedules are added manually using OpenStudio GUI shown in

Figure 8-12. The corresponding IDF file will be exported to be used by standard

EnergyPlus and its advanced outputs will be used for data validation.

370

Figure 8-12. OpenStudio GUI used for adding required data for energy simulation and converting gbXML

to IDF

8.4.1 Comparison between Existing BEM Methods

To compare existing methods for BEM and also elaborate the data validation process,

Table 8-1 presents four different approaches and examples of tools, which can be involved in BIM to BEM process. Steps related to conventional, semi-automated, and automated

BEM processes are explained in this table. Application of ABEMAT can be categorized under the automated BEM process, although it still requires manual data input such as

HVAC system, schedules, and thermostat set points. The method used for data validation in this paper shown in Figure 8-10 is an example of semi-automated BEM process, since it still requires manual interaction for gbXML to IDF conversion.

371

Table 8-1. Comparison between classic, semi-automated, and automated BEM using BIM

Semi-automated BEM Automated Automated BEM Classic BEM Method Using BIM BEM Method Method Using BIM Method (used for validating Using BIM (e.g., (e.g., ABEMAT) ABEMAT) GBS) Develop architectural Develop architectural Develop Develop architectural model model architectural model model Export the Reenter data and Export the architectural architectural model develop a new model model under a certain Define energy under a certain BIM in a GUI such as BIM file schema such as analysis parameters file schema such as SketchUp IFC or gbXML such as HVAC, IFC or gbXML Identify issues with BIM schedules, space Import the model in file, which can cause types, and thermal an energy modeling errors during the energy zones tool such as modeling process and OpenStudio resolve them manually Define other required Import the file in an data such as HVAC, Run ABEMAT energy modeling tool schedules, run period, such as OpenStudio Create the energy and location analytical model Manually add missing and run GBS Run the energy information related to within Revit analysis other components such as HVAC or schedules

(some data such as Obtain accumulated Obtain HVAC and schedules data (not fine- Run the energy analysis accumulated data data still need to be grained) (no fine-grained) added manually) Obtain accumulated data Obtain fine-grained

(not fine-grained) data

372

Conventional BEM method starts with developing an architectural model in a CAD tool such as AutoCAD and Revit, redrawing the model in a GUI such as DesignBuilder and reentering information related to construction materials, and performing energy simulation to obtain the energy-related outputs. This process requires reentering data, which can be time consuming and error prone. The benefit of such process is that there will not be any interoperability issues between different tools, since data are manually added in each step.

Semi-automated BEM process can be achieved by eliminating the need for reentering data, which can be skipped by adopting BIM that is the basis of both semi-automated and automated BEM process. The generated BIM file needs to be converted to a proper file to be read by a BEM tool. If importing the BIM file and exporting it to another file format requires manual interaction and using other GUIs such as OpenStudio, it can be referred to as semi-automated BEM process. For example, gbXML file is fed to the OpenStudio GUI for energy analysis. The OpenStudio GUI shown in Figure 8-12 is capable of reading the gbXML file and adding other information with regard to the HVAC system, schedules, and loads to evaluate the energy performance of the building. Not all the data need to be reentered manually, and it only requires users to perform the process of importing, initiating the file conversion, and running the analysis.

On the other hand, the BEM process can be automated using existing tools such as Green

Building Studio (GBS), which is a cloud-based energy simulation tool or developed tools such as ABEMAT, which contribute to automation of the BEM process by eliminating the need for reentering data and manual conversion of the BIM file. Additional services and

373 benefits of ABEMAT include open-source components and fine-grained outputs as opposed to tools such as GBS with no openly accessible source code and outputs limited to what is provided by DOE-2, its energy simulation engine. However, Revit is more compatible with GBS, and all the data defined in Revit including HVAC system, schedules, and loads are transferred properly to GBS, which eliminates the need for manual interaction for data input.

Details of the steps explained above are listed in Table 8-1, which shows that several extra steps are needed in conventional or semi-automated BEM approaches compared to automated method represented by ABEMAT or GBS. These steps can slow down the process, making the process more error-prone, and make it more challenging to access detailed outputs.

The next section is dedicated to comparison of ABEMAT’s outputs and the adopted method for data validation shown in Figure 8-10. The process of modeling the house and outputs are further explained in here.

374

8.4.2 Modeling a One-Story Residential Building for Data

Verification

The house modeled in Revit for data verification identifies four thermal zones in this building. As shown in Figure 8-13, the model is simplified to minimize the number of surfaces to facilitate comparison of the results. The fine-grained outputs of ABEMAT are focused on each component in each thermal zone, separately, and identify the amount of heat transfer through each of them. The properties related to building envelope and other systems considered for this model are presented in Figure 8-2. Energy simulation is performed based on a one-year simulation under Boston weather condition. Because of the issues occurring during data transfer from Revit to OpenStudio, there are missing data that need to be resolved. The semi-automated method used for validating ABEMAT requires software-related skills in OpenStudio to define missing data such as HVAC data, schedules, location, and run period. All of these data are added using OpenStudio GUI shown in Figure 8-12.

Figure 8-13. Plan view (Left) and 3D view (Right) of building modeled in Revit (roof is not shown in this

figure) 375

Table 8-2. Properties of the building modeled in EnergyPlus

Building Component Properties

Bedrooms 3

Living Room 1

Square Footage 1098 SF

Exterior Wall Brick/Air/Rigid Insulation/Vapor barrier/CMU/Metal

Furring/Gypsum board (R-32)

Interior Wall Gypsum board/Metal Stud/Gypsum board (R-21)

Roof Asphalt shingle/Plywood/Wood joist (R-58)

Ceiling Acoustic Ceiling Tile (R-1.6)

Windows Double Hung with Trim (36”×48”)(SHGC=0.78)(R-

1.5)

Spaces 4 Spaces

Zones 4 Thermal Zones

Construction Location Boston, MA

Building Type Single Family

Building Operating Schedule Default

HVAC System Residential 17 SEER/9.6 HSPF Split HP <5.5 ton

Outdoor Air per Person 15 CFM

Export Category Spaces

Energy Settings Project Phase New Construction

Building Service VAV-Single Duct

Building Infiltration Class Medium

Export Default Values Yes

376

Either “energy setting” or “space volumes” method can be adopted to generate the gbXML file in Revit. Exporting under “energy setting” leads to getting all the data defined in Revit including schedules, loads, and HVAC data; however, it also generates additional spaces because the energy model developed in Revit breaks down a space into sub-spaces for easier energy analysis. Moreover, not all the data within gbXML file can be imported in

OpenStudio properly and it still requires manual data input. Therefore, it was decided to use the other approach, which is exporting under “room/space volumes”. As shown in

Figure 8-14, no error is identified by Revit prior to generating the gbXML file for use by

ABEMAT. This method does not transfer data such as those related to HVAC system and loads since these data will be added manually. The other method shown in Figure 8-10 for data validation and outputs are explained in the next section.

Figure 8-14. Exporting the architectural model under “room/space volumes” using Revit

377

8.4.3 The Outputs and Results of Data Verification

After performing energy modeling using ABEMAT and semi-automated BEM method for data validation, the outputs obtained are presented in Table 8-3. The data in Table 8-3 correspond to each thermal zone shown in Figure 8-13. For example, thermal zone #1 shown in Figure 8-13 consists of three exterior walls, three interior walls, one roof, one floor, two windows, and one door. Table 8-3 shows all of these components recognized by

ABEMAT and the amount of heat transfer through each component that is provided in separate text files for windows, walls, floors, roofs, and doors.

The text files generated by ABEMAT are shown in Figure 8-15, where the content of the text file related to walls is also shown as an example. To facilitate data interpretation, the titles for components are simplified and shorter names are used in Table 8-1 compared to what is generated by ABEMAT shown in Figure 8-15. Data related to interior components such as walls and doors can be either positive or negative, which represents heat gain or loss, respectively. The amount of heat loss and heat gain for interior walls and doors are equal in this case study, because the thermostat set point temperature for all thermal zones are set to be equal, which means there wouldn’t be any colder or warmer thermal zone in this house. Similar observation is also noted in Figure 8-15, where the interior components have corresponding component with a “REVERSED” title added to their name. It represents the same surface with reversed heat transfer.

Yielding of similar results by both methods is indicative that ABEMAT works accurately while fetching and saving the detailed data. Text files can be easily used by software

378 developers to be used on GUIs in follow-up studies, since they are easy to interpret and transferred electronically to other file formats.

Figure 8-15. Text files generated automatically by ABEMAT containing heat transfer data for each

component and their thermal zones

379

Table 8-3. Summary of energy simulation outputs obtained from semi-automated BEM method and

ABEMAT (Ext.=Exterior wall, Int.=Interior wall, Win.=Window, and D.=Door) for data validation

Zones and Advanced outputs Direct outputs of components of E+ (GJ) ABEMAT (GJ)

Gain Loss Gain Loss Ext. 1 -0.617 -0.629 Ext. 2 -1.469 -1.498 Ext. 3 -0.585 -0.595 Int. 1 ±0.024 ±0.024 Int. 2 ±0.015 ±0.015 Int. 3 ±0.017 ±0.017

Zone 1 Flr. 1 -0.592 -0.609 Rf. 1 -1.444 -1.470 Win. 1 2.033 -0.934 2.033 -0.934 Win. 2 0.766 -0.958 0.766 -0.958 D. 1 -1.220 -1.243 Ext. 4 -0.585 -0.595 Ext. 5 -0.433 -0.443

Zone 3 Flr. 3 -0.167 -0.170 Rf. 3 -0.373 -0.379 D. 3 ±0.034 ±0.034 Ext. 7 -0.299 -0.306 Ext. 8 -0.539 -0.549 Int. 3 ±0.017 ±0.017 Int. 5 ±0.004 ±0.004 Flr. 4 -0.188 -0.193

Zone 4 Rf. 4 -0.386 -0.393 Win. 4 2.574 -0.928 2.574 -0.928 Win. 5 1.724 -0.949 1.724 -0.949 D. 4 ±0.039 ±0.039

380

8.5 Summary and Conclusions

Adopting tools such as ABEMAT can contribute to multiple areas related to reducing energy consumption in buildings. Evaluating buildings during design and use phases can lead to optimized design and identification of thermal zones and components with higher contribution to energy consumption, which can optimize decision-making process in energy retrofit. Facilitating and improving the accuracy of BEM process can contribute to these areas. In this paper, BIM is adopted as a tool to automate and improve the energy modeling process besides other capabilities such as adopting corrective tools for resolving interoperability issues and providing fine-grained outputs by modifying existing open- source tools such as EnergyPlus.

The current version of ABEMAT is focused only on building envelope systems, and other tasks such as adding data related to HVAC system, schedules, and initial information needed for energy simulation such as run period are added manually. Therefore, there is room for improvement of ABEMAT in follow-up studies. The whole BIM-to-BEM process cannot be fully automated unless the origin of BIM file, which is the BIM tool or CAD tool in this case, can include all the required information. Even if a GUI is added to such tools for defining the missing required data such as HVAC and schedules data, it still requires human interaction within the BIM-to-BEM process. Ideally, the best scenario for the BIM- to-BEM process would be if all the required information are entered only once at the beginning of the BIM-to-BEM process, before generating the BIM file. Other middleware tools with GUI can be further developed and adopted only to enable users for minor

381 changes such as changing the run period of energy analysis or changing the schedules.

Similar tools can be developed using the same approach by adopting existing open-source software related to energy modeling. The existing BIM or CAD tools seem to be capable of adopting these tools as add-ins to add the capability of performing fast and accurate energy analysis in a user-friendly manner.

The summary of important observations in this study are as follow:

 Reduction in energy consumption in buildings is possible through multiple

measures and at different stages of their life cycle including considering this as a

design objective during the design phase and pre-construction, energy retrofit of

existing buildings during use phase, and monitoring the energy consumption during

the use phase using the emerging systems and technologies such as smart homes

 Data related to energy consumption can be either accumulated such as total energy

consumption in a thermal zone (e.g., living room) or it can be fine-grained by

providing the amount of heat transfer through each component separately such as

each exterior wall, windows, and doors.

 The existing energy simulation tools provide either only the accumulated data

concerning the energy consumption of a thermal zone or the fine-grained data

cannot be obtained through an easy, accurate, and fast process.

 Application of BIM can facilitate the energy modeling, simulation, and visualizing

process and make the process faster and more accurate by avoiding reentering all

the data in different stages of modeling.

382

 Multiple shortcomings and issues exist in BBIP, which need to be addressed in

studies focused in this area.

 ABEMAT is a tool developed by compiling different tools, which can automate the

energy modeling and simulation process. It is comprised of three major tools

developed by modifying the source code of existing tools.

 The first tool in ABEMAT receives the gbXML file and performs some corrective

actions on the file to eliminate the issues related to building envelope, which occur

in BIM-to-BEM process.

 The second tool that uses existing modules within OpenStudio source code receives

the corrected gbXML file and converts it to an IDF file for EnergyPlus. This file

converter uses the existing modules in OpenStudio designed for this purpose.

 The third tool that is modified EnergyPlus receives the IDF file, performs the

energy modeling, and generates text files containing fine-grained data on the

amount of heat transfer through every single building envelope component

including the windows, doors, walls, floors, and roof.

The case study used in this article to validate the results of ABEMAT shows how this tool can automate the energy modeling process, make the process more reliable and accurate, eliminate the need for software-related skills at multiple stages of process, make it faster, and provide fine-grained outputs to be used in future tools. These tools can use these fine- grained data for designing more energy efficient houses, performing better evaluation of existing homes for energy retrofit, and monitoring purposes in GUIs related to energy smart homes.

383

8.6 References

[31] A. F. Setiawan, T.-L. Huang, C.-T. Tzeng and C.-M. Lai, "The Effects of Envelope

Design Alternatives on the Energy Consumption of Residential Houses in Indonesia,"

Energies, vol. 8, pp. 2788-2802, 2015.

[33] U. Eicker, E. Demir and D. Gürlich, "Strategies for cost efficient refurbishment and solar energyintegration in European Case Study buildings," Energy and Buildings, no. 102, pp. 237-249, 2015.

[48] D. B. Crawley, J. W. Hand, M. l. Kummert and B. T. Griffith, "Contrasting the capabilities of building energy performance simulation programs," Building and

Environment, vol. 43, pp. 661-673, 2008.

[54] N. Yu, "INFORMATION INTEROPERABILITY BETWEEN BUILDING

INFORMATION MODELING AUTHORING TOOLS AND SIMULATION TOOLS TO

SUPPORT ENERGY EFFICIENT BUILDING DESIGN," Submitted in Partial

Fulfillment of the Requirements for the Degree of Master of Science, Information Sciences and Technology, The Pennsylvania State University, University Park, 2014.

[56] V. Bazjanac, "IFC BIM-Based Methodology for Semi-Automated Building Energy

Performance Simulation," in 25th International Conference on Information Technology in

Construction, Santiago, Chile, 2008.

384

[245] P. Xu, Y. Shen, L. Chen and J. Mao, "Assessment of energy-saving technologies retrofitted to existing public buildings in China," Energy Efficiency, vol. 9, no. 1, pp. 67-

94, 2016.

[246] V. M. Nik, É. Mata and A. S. Kalagasidis, "A statistical method for assessing retrofitting measures of buildingsand ranking their robustness against climate change,"

Energy and Buildings, no. 88, pp. 262-275, 2015.

[247] C. Friedman, N. Becker and E. Erell, "Energy retrofit of residential building envelopes in Israel: A cost benefit analysis," Energy, no. 77, pp. 1-11, 2014.

[289] C. Eastman, P. Teicholz, R. Sacks and K. Liston, BIM Handbook, A Guide to

Building Information Modeling for Owners, Managers, Designers, Engineers, and

Contractors, New Jersey: John Wiley & Sons, 2011.

[290] J.-H. Woo, C. Diggelman and B. Abushakra, "BIM-BASED ENERGY

MORNITORING WITH XML PARSING ENGINE," in Proceeding of the 28th ISARC,

Seoul, Korea, 544-545, 2011.

[291] N. W. Young Jr., S. A. Jones, H. M. Bernstein and J. E. Gudgel, "THE BUSINESS

VALUE OF BIM," McGraw Hill, New York, 2009.

[293] H.-Y. Chong, C.-Y. Lee and X. Wang, "A mixed review of the adoption of Building

Information Modelling (BIM) for sustainability," Journal of Cleaner Production, vol. 142, p. 4114e4126, 2017.

385

[301] F. Jalaei and A. Jrade, "INTEGRATING BUILDING INFORMATION

MODELING (BIM) AND ENERGY ANALYSIS TOOLS WITH GREEN BUILDING

CERTIFICATION SYSTEM TO CONCEPTUALLY DESIGN SUSTAINABLE

BUILDINGS," Journal of Information Technology in Construction, vol. 19, pp. 494-519,

2014.

[303] S. Kumar, "INTEROPERABILITY BETWEEN BUILDING INFORMATION

MODELS (BIM) AND ENERGY ANALYSIS PROGRAMS," SCHOOL OF

ARCHITECTURE, UNIVERSITY OF SOUTHERN CALIFORNIA, Los Angeles, 2008.

[305] S. Salakij, N. Yu, S. Paolucci and P. Antsaklis, "Model-Based Predictive Control for building energy management. I:Energy modeling and optimal control," Energy and

Buildings, vol. 133, p. 345–358, 2016.

[307] T. Maile, M. Fischer and V. Bazjanac, "Building Energy Performance Simulation

Tools—A Life-Cycle and Interoperable Perspective;," Stanford University: Stanford, CA,

USA, 2007.

[308] S. Pinheiro, J. Donnell, R. Wimmer, V. Bazjanac, S. Muhic, T. Maile, J. Frisch and

C. van Treeck, "MODEL VIEW DEFINITION FOR ADVANCED BUILDING ENERGY

PERFORMANCE SIMULATION," in CESBP/BauSIM 2016 Conference, Dresden,

Berlin, 2016.

386

[310] J. B. Kim, W. Jeong, M. J. Clayton, J. S. Haberl and W. Yan, "Developing a physical BIM library for building thermal energy simulation," Automation in Construction, vol. 50, pp. 16-28, 2015.

[311] W. Jeong, J. B. Kim, M. J. Clayton, J. S. Haberl and W. Yan, "Translating Building

Information Modeling to Building Energy Modeling Using Model View Definition,"

Hindawi Publishing Corporation, pp. 1-21, 2014.

[314] Y. N. Bahar, C. Pere, J. Landrieu and C. Nicolle, "A Thermal Simulation Tool for

Building and Its Interoperability through the Building Information Modeling (BIM)

Platform," Buildings, vol. 3, pp. 380-398, 2013.

[315] O. Granlund Oy, M. Korhonen and T. Laine, "Energy Analysis Software Evaluation

BIM Interface and Interoperability," Granlund, Eskilstuna, Sweden, 2008.

[322] B. Dong, K. P. Lam, Y. C. Huang and G. M. Dobbs, "A comparative study of the

IFC and gbXML informational infrastructures for data exchange in computational design support environments," in Proceedings: Building Simulation 2007, 2007.

[331] "gbXML," [Online]. Available: http://www.gbxml.org/. [Accessed 23 September

2016].

387

[332] H. J. Moon, M. S. Choi, S. K. Kim and S. H. Ryu, "CASE STUDIES FOR THE

EVALUATION OF INTEROPERABILITY BETWEEN A BIM BASED

ARCHITECTURAL MODEL AND BUILDING PERFORMANCE ANALYSIS

PROGRAMS," in 12th Conference of International Building Performance Simulation

Association, Sydney, 2011.

[340] J. T. O’Donnell, T. Maile, C. Rose, N. Mrazović, E. Morrissey, C. Regnier, K.

Parrish and V. Bazjanac, "Transforming BIM to BEM: Generation of Building Geometry for the NASA Ames Sustainability Base BIM," Ernest Orlando Lawrence Berkeley

National Laboratory, Berkeley, 2013.

September 2016, 2016.

[354] "EnergyPlus," the U.S. Department of Energy’s (DOE) Building Technologies

Office (BTO), 2017. [Online]. Available: https://energyplus.net/. [Accessed 16 February

2017].

[356] I. J. Ramaji, J. I. Messner and R. M. Leicht, "Building Performance Modeling

Conference," in ASHRAE and IBPSA-USA SimBuild 2016, Salt Lake City, 2016.

388

[357] T. Gerrish, K. Ruikar, M. Cook, M. Johnson, M. Phillip and C. Lowry, "BIM application to building energy performance visualisation andmanagement: Challenges and potential," Energy and Buildings, vol. 144, p. 218–228, 2017.

[358] G. Gourlis and I. Kovacic, "Building

InformationModellingforanalysisofenergyefficient industrial buildings – A casestudy,"

Renewable and Sustainable Energy Reviews, vol. 68, pp. 953-963, 2017.

[359] H. Kim, Z. Shen, I. Kim, K. Kim, A. Stumpf and J. Yu, "BIM IFC information mapping to building energy analysis (BEA) model with manually extended material information," Automation in Construction, vol. 68, pp. 183-193, 2016.

[360] F. Abanda and L. Byers, "An investigation of the impact of building orientation on energy consumption in a domestic building using emerging BIM (Building Information

Modelling)," Energy, vol. 97, pp. 517-527, 2016.

[361] R. Volk, J. Stengel and F. Schultmann, "Building Information Modeling (BIM) for existing buildings — Literature review and future needs," Automation in Construction, vol.

38, pp. 109-127, 2014.

[362] S. Chidiac, E. Catania, E. Morofsky and S. Foo, "Effectiveness of single and multiple energy retrofit measures on the energy consumption of office buildings," Energy, no. 36, pp. 5037-5052, 2011.

[363] S.-M. Oh, Y.-J. Kim, C.-S. Park and I.-H. Kim, "PROCESS-DRIVEN BIM-

BASED OPTIMAL DESIGN USING INTEGRATION OF ENERGYPLUS, GENETIC 389

ALGORITHM, AND PARETO OPTIMALITY," in 12th Conference of International

Building Performance Simulation Association, Sydney, 14-16 November, 2011.

390

9) Chapter 9. Discussion, Summary, and Conclusion

9.1 Discussion

Saving even a small percentage of total building’s energy consumption in the U.S. can lead to significant economic and environmental impacts. In order to reduce energy consumption in buildings, energy conservation methods such as energy retrofit of existing buildings can be adopted. In addition, several computer tools and components can help contribute to the field of building energy conservation. In particular, some computer tools can be improved and enhanced in order to perform better and have higher contribution to the main goal of saving energy in buildings. These methods and tools can target different building component types such as mechanical systems, electrical components, and building envelope systems for improvement of building energy performance.

A logical strategy for energy retrofit design would be to identify and focus on areas and components in a building that have the highest contribution to energy consumption. The main emphasis and major goal of this research is to facilitate the process of identifying the areas and components in a building, which have the highest contribution to energy consumption.

While different building components types and several optional approaches are available to contribute toward this goal, this study focused on energy simulation computer tools and building envelope components of buildings. Therefore, the specific goal of the study in

391 facilitating the process of identifying the areas and components in a building with highest contribution to energy consumption can be reached by improving or developing building energy simulation tools that can help acquiring building envelope components energy- related data.

A computer tool that can help determining the amount of heat transferred through building envelope and floors can be great interest to identify the areas and components with highest contribution to energy consumption. Existing computer tools related to this area have experienced significant improvements and advancements since the last decade; however, there is more room for further improvement.

The state-of-the-art computer tools available related to building energy simulation can be evaluated with respect to two aspects. First, the process it takes a user to get energy-related information, and second, the quality of outputs provided by these tools. The state of the art review of such tools can help answer the following two questions: 1) the methods or approaches that these tools obtain energy-related information and 2) the kind of outputs these tools provide.

The process of building energy simulation is the focus of the first question. The process of obtaining energy-related data about building envelope has improved significantly, and existing tools are capable of performing energy simulation in an automated way using emerging tools such as BIM. Although there are still challenges, issues, and shortcomings in this field, the process is much more developed and faster compared to what was available

392 a decade ago. However, the question regarding the kind of outputs these tools generate can have different answers and there is still room for additional answers.

The second aspect is then focused on outputs. Most of the available tools only provide total energy consumption of each sector in a building such as HVAC system, lighting, hot water, and appliances. On the other hand, there are tools such as EnergyPlus that are capable of providing high quality outputs, but the process to obtain them is not straightforward.

The major part of this research is dedicated to improving these two aspects mentioned above to reach high quality energy simulation 1) process and 2) outputs.

A high quality process refers to a BEM process, which is fast, accurate, and preferably automated with minimum human interaction. For example, one can compare the conventional energy modeling process using a BEM tool GUI to develop an energy model and reentering all the data and run the analysis versus an ideal “click-of-a-button” automated process, which only requires the architectural model developed in a typical CAD tool such as Revit.

High quality outputs refer to the energy-related outputs, which are detailed as opposed to accumulative energy-related information. For example, one can consider the accumulated data with regard to the whole building energy consumption versus the value of heat transfer through every single wall, window, door, floor, and roof. Combining these two features in one tool can be an important contribution in fulfilling the main stated goal, which is performing building energy modeling to identify the areas and building envelope

393 components in a building with highest contribution to energy consumption in an easy, accurate way with minimum human interaction. This research has developed a platform consisting of different tools to perform building energy modeling and analysis using BIM and provide fine-grained outputs related to building envelope components with minimum human interaction.

This platform can contribute to energy monitoring systems in energy smart homes, BEM tools in decision-making process for energy retrofit of buildings, and BEM tools in energy performance evaluation during design phase of buildings.

Different software tools were used to develop this platform including EnergyPlus source code, OpenStudio source code, C++, Python, and Ruby programing languages. The platform consists of three major components including 1) a corrective tool to modify gbXML files in order to resolve issues related to building envelope data, 2) a file conversion unit to convert gbXML to IDF, and 3) modified EnergyPlus as energy simulator to save fine-grained data on heat transfer through walls, windows, floors, roof, and doors in text files. The outcomes of this research are summarized and outlined in this chapter.

9.2 Outcomes

Based on the main goal explained in the previous section, two major objectives were defined and explained in Chapter 3. Objective #1 and #2 were defined as improving the quality of outputs and developing a platform to perform a high quality BEM process using

BIM. In this research, high quality outputs and BEM process refer to fine-grained energy-

394 related outputs and fast and easy with minimum manual interaction BEM process, respectively.

Several tasks were also defined to fulfil these objectives in a systematic approach.

Objective #1 required performing a comprehensive study on building energy retrofit measures and relevant computer tools. It also required modifying the EnergyPlus source code, so it can generate fine-grained outputs. Objective #2 initially required two comprehensive study on smart homes and BIM-to-BEM process in order to have better understanding of required inputs for energy simulation tools and help in picking the proper

BIM and energy simulation file format. In addition, a corrective tool had to be developed to rectify the issues related to BIM file prior to use for energy simulation.

Based on the objectives and tasks defined for this research study presented in Chapter 3, the major outcomes of this research study can be summarized as follow. The follow-up section will provide the contributions of these outcomes, and it will be explained how these findings can contribute to this study and other areas related to it.

 Review of Smart Homes

A comprehensive study on smart homes and, in particular, energy smart homes was conducted, which presents state-of-the-art systems and services in energy smart homes as explained in Chapter 4. Components in energy smart homes are reviewed and a new category is proposed for energy smart homes based on these components dividing them into the homes with 1) energy monitoring capabilities, 2) control capabilities, and 3)

395 advanced data processing capabilities. Buildings, projects, and labs focused on smart homes are also reviewed in this study and summary of challenges in adopting technologies related to smart homes are also summarized and presented in this state-of-the-art review study. The contributions of this study are summarized in next section.

Review of literature showed that the control and automation aspect of energy smart homes is more desirable among users such as incorporating temperature control, HVAC and lighting control, and DR systems. In addition, there has been more advancements in control and communication systems compared to other components such as monitoring and processing units.

 Review of Building Energy Retrofit

In order to study the trend of advances in building envelope energy retrofit systems and techniques, an in-depth review of the literature was carried out, as explained in Chapter 5.

The review focused on conventional and innovative building’s energy retrofit materials and systems such as application of aerogel, PCM, and dynamic façade. Different retrofit options that are currently being used or are under development were reviewed in this study, and example real-world projects using these measures were investigated and summarized.

The review of literature showed that new and emerging energy retrofit systems can compete with conventional systems; however, the initial costs may still be an issue in some cases. The review also showed a focus on the façade of buildings and the capability of adding dynamic components to these systems as a desirable option, in particular

396 considering the emergence of new automation systems within buildings. Computer tools and modeling still need to be improved in terms of accuracy and they also need to be modified and improved to be able to integrate the capability of modeling these new retrofit measures, especially the ones focused on the façade of the building.

Finally, a comparative study was performed on a typical residential building in the U.S. using BEopt in two different climate zones to evaluate and compare different conventional and innovative retrofit materials and techniques, including adding XPS or PCM to exterior walls, replacing windows with double or triple pane windows, adding aerogel to floors and walls, using roofing materials with different colors as cool/warm roof, and adding insulation to slabs. Climate zones considered in this study include cold and hot-dry climate zones in Pittsburgh, and Los Angeles. Results show that innovative options such as aerogel and PCM are effective in terms of improving the thermal performance of building; however, their costs still need to be lowered to be able to compete with conventional materials such as XPS, fiberglass, and polyisocynurate spray foam. In cold and hot-dry climates, PCM led to about 2.5% and 18% decrease in heating and cooling loads, respectively, and an increase in the cost by about 5% when used within gypsum boards.

Moreover, adding wall insulation, roof insulation, and using more energy efficient windows were found to be effective options in both climate zones, which can save up to

30% and 23% in annual heating and cooling site energy consumption in hot-dry and cold climates, respectively. The summary of major contributions are presented in next section.

397

 Review of Application of BIM in BEM

The third part of the study explained in Chapter 6 is dedicated to review of application of

BIM in BEM to investigate different tools involved in the process, issues related to the process, and provided solutions. A detailed categorization for issues and challenges based on different steps and components in BBIP is presented and six different groups are provided. All the issues, challenges, and shortcomings happen in or during one of these components or steps including 1) the BIM authoring tools such as Revit or AchiCAD, 2)

The process of mapping data from BIM tool to a BIM file, 3) in a BIM file under different formats such as gbXML or IFC, 4) within a BEM’s GUI, 5) during mapping data from GUI to a BEM file, and 6) during reading data by simulation engine from BEM file. Different solutions for issues related to any of these steps suggested by researchers are also reviewed in this study such as adding manual data to make up for the missing data or developing middleware tools to correct issues occurring during files transfer.

Based on the issues identified , a corrective tool was developed using Python, which reads gbXML files and resolves certain issues related to building envelope components such as missing data related to doors or redundant data generated to during file transfer. To investigate the issues occurring during BBIP in details and assess the performance of this corrective tool, three case studies were undertaken using Revit to model a house and three energy modeling and analysis tools including GBS, OpenStudio, and EnergyPlus. The outcomes show applicability of such corrective middleware tools in BBIP. The major contributions and impacts of this study are explained in next section.

398

 Energy Simulation Tool with Fine-Grained Outputs

An energy simulator is developed by modifying EnergyPlus source code in order to provide detailed outputs related to heat transfer through every single walls, windows, floors, roof, and doors. This tool is referred to as BEPAT (Building Energy Performance Assessment

Tool) and was developed to be used as a component within ABEMAT (Automated

Building Energy Modeling and Assessment Tool), which is the main platform developed for this research. The need for improving the quality of BEM tool’s outputs by providing fine-grained data was noted after review of literature and to keep this tool compatible with other tools, which can be developed in follow-up studies, outputs are saved in text format, which makes it more convenient for software developers to acquire and use the data for other tools or GUIs. The source code of EnergyPlus is modified to fetch the desirable data and save them in separate text files under walls, windows, doors, roof, and floors names.

In order to validate the performance of BEPAT, a one-story residential building with four thermal zone is developed in Revit and fine-grained outputs of energy simulation using

BEPAT are compared to advanced outputs of EnergyPlus, which are similar to BEPAT’s outputs. Comparison of results show similar outputs in both methods. BEPAT’s outputs are obtained through a single-click action and does not need software-related skills.

Detailed contributions of developing such a tool are explained in next section.

399

 A Platform for Automated Energy Evaluation

The platform developed in this study (ABEMAT) is the major outcome and consists of all the tools developed in previous sections, in addition to a file converter that transforms gbXML files to IDF files. To add this component to ABEMAT, it was decided to use

OpenStudio and call predefined functions within its source code, which are responsible for this file conversions. Ruby script is developed to perform this action by calling different functions to convert gbXML to OpenStudio file and convert the OpenStudio file to an IDF file.

The three major components of ABEMAT are 1) gbXML corrective tool, 2) gbXML to

IDF file converter, and 3) BEPAT, which can work by perform the whole BEM process and generating detailed outputs in s user-friendly manner. However, this platform is focused on building envelope, and other data required for energy simulation are to be added manually. With respect to building envelope components, this platform can be considered as a fully automated approach for energy evaluation of buildings. The major contributions of such a tool are explained in next sections.

9.3 Research Contributions

Considering the outcome of the main tasks undertaken in this research, the following conclusions can be offered and considered useful for the AEC industry and in particular the field of BEM.

400

 The study has led to design of an approach to interoperability of building envelope

information and energy modeling.

 The proposed process demonstrated successfully can be considered a contribution

to the field of BEM.

 The demonstration of the functionality of BIM in improving the interoperability of

building envelope information can be considered a contribution to integrating other

types of information in BEM using BIM.

 A platform developed facilitates acquiring values for heat transfer through building

envelope systems.

 The automated BEM process through ABEMAT can lead to saving time, improving

accuracy, and making the process more convenient.

 ABEMAT’s fine-grained outputs can be benefitial to tools performing building

energy evaluation during design phase, during energy-retrofit of existing homes,

and energy monitoring systems in energy smart homes.

 The corrective tool developed in this research as the gbXML file corrector can be

used (in other tools) to resolve the issues related to redundant data generated during

the file transfer and missing construction IDs for doors, which occur during data

transfer from Revit to gbXML.

401

 The script developed in Ruby for gbXML to IDF conversion can be beneficial to

other tools because it uses open-source OpenStudio, which makes it more practical,

since it can be later modified to resolve potential issues in file transfer.

9.4 Potential Future Researches

Review of literature performed for this study highlighted some other areas related to the

BEM process, which can be of great importance and contribute to several areas including

BEM, BIM, energy retrofit, BIM to BEM, and energy smart homes.

The potential follow-up research needs are suggested as follow:

 Energy smart homes need to move toward advanced processing units with energy

analysis capabilities. Integrating these simulation tools with real-time data obtained

from sensors can be one of the important steps.

 Integrating real-time data acquisition using sensors with energy simulation tools

similar to ABEMAT can be undertaken with focus on other components such as

appliances, HVAC, or lighting system in order to incorporate it with microgrids,

smart grids, and DR systems.

 The new layout proposed for BIM to BEM process and categorizing components

and steps in this process can lead to more organized researches in this field in

follow-up studies. Moreover, future corrective tools can be focused on certain areas

identified in this research as more error-prone areas.

402

 Integrating energy simulation tools with control systems in energy smart homes can

lead to optimized performance of such tools. These control systems can be

dedicated to lighting system, equipment, and HVAC systems.

 Innovative energy retrofit systems such as dynamic façade and the precast

components can be linked to control systems, which provide this opportunity to

involve building envelope in control systems.

 BIM to BEM process can be improved by developing middleware tools targeting

the step between BIM authoring tools and BEM tools in order to add typical missing

data such as HVAC system data, thermostat set points, and loads.

 Implementing a GUI to ABEMAT or similar tools can contribute to make their

applications more convenient.

 Source codes in similar tools used in this study such as OpenStudio can be of high

importance, since they can be modified in order to resolve issues related to BIM

file to BEM file conversion process.

 Optimization methods can be integrated with this tool as a complimentary side tool

to convert it to an energy retrofit design tool in order to evaluate multiple design

options simultaneously and provide the best option.

403

Appendix A: Codes

A.1 Subroutines Added to EnergyPlus Source Code for Heat

Transfer through Windows

Under the DataSurfaces module:

//Ehsan float AccWinHeatGain[80]; float AccWinHeatLoss[80]; //End Ehsan

Under the DataSurfaces header:

//Ehsan extern float AccWinHeatGain[80]; extern float AccWinHeatLoss[80]; //End Ehsan Under OutputReportTabular module:

//Ehsan

using DataSurfaces::Surface; using DataSurfaces::TotSurfaces; using DataSurfaces::WinHeatGain; using DataSurfaces::WinHeatGain; using DataSurfaces::WinHeatGainRepEnergy; using DataSurfaces::WinHeatLossRepEnergy; using DataSurfaces::AccWinHeatGain; using DataSurfaces::AccWinHeatLoss; static int iSurf(0); static int ZoneNum(0);

404

for (iSurf = 1; iSurf <= TotSurfaces; ++iSurf) { AccWinHeatGain [iSurf] += WinHeatGainRepEnergy(iSurf) * timeStepRatio; AccWinHeatLoss [iSurf] -= WinHeatLossRepEnergy(iSurf) * timeStepRatio; } //End Ehsan ______

// SUBROUTINE INFORMATION: // AUTHOR Ehsan Kamel // DATE WRITTEN November 2016 // DATE MODIFIED December 2016

// PURPOSE OF THIS SUBROUTINE: // This subroutine collects and saves the heat gain/loss through different windows in // different thermal zones, separately (as oppose to the totall heat gain/loss that is //reported in Seinsible Heat Gain Summary Outputs" is the outputs report tables.

// METHODOLOGY EMPLOYED: // A new text file is created on Desktop and the WinHeatGain (i) will be read // and saved there. The heat gain and loss will be distinguished by their sign // that means negative numbers will be considered as heat loss.

using DataGlobals::NumOfTimeStepInHour; using DataGlobals::CompLoadReportIsReq; using DataGlobals::isPulseZoneSizing; using DataSizing::CurOverallSimDay; using DataSurfaces::Surface; using DataSurfaces::TotSurfaces; using DataSurfaces::WinGainConvGlazToZoneRep; using DataSurfaces::WinGainConvGlazShadGapToZoneRep; using DataSurfaces::WinGainConvShadeToZoneRep; using DataSurfaces::WinGainFrameDividerToZoneRep; using DataSurfaces::SurfaceClass_Window; using DataZoneEquipment::ZoneEquipConfig; using DataSurfaces::WinHeatGain; using DataSurfaces::WinHeatGain; using DataSurfaces::WinHeatGainRepEnergy;

405

using DataSurfaces::WinHeatLossRepEnergy; using DataSurfaces::AccWinHeatGain; using DataSurfaces::AccWinHeatLoss;

static int iSurf(0); static int ZoneNum(0); static int TimeStepInDay(0);

#include #include using namespace std; const char *path = "C:/Users/MZK221/Desktop/Windows Heat Gain or Loss.txt"; const char *path2 = "C:/Users/MZK221/Desktop/Surfaces and Zones.txt";

static Real64 timeStepRatio(0.0); using DataHVACGlobals::TimeStepSys; using DataGlobals::TimeStepZone; timeStepRatio = TimeStepSys / TimeStepZone;

for (iSurf = 1; iSurf <= TotSurfaces; ++iSurf) { if (!Surface(iSurf).ExtSolar) continue; // WindowManager's definition of ZoneWinHeatGain/Loss if (Surface(iSurf).Class != SurfaceClass_Window) continue; ZoneNum = Surface(iSurf).Zone; if (ZoneNum == 0) continue; //In here, ZoneNum only helps to detect if it is in warmup period or not. ofstream myfile(path, std::ios_base::app); myfile << "Surface # is " << iSurf << ", in zone " << ZoneNum-1 << ", energy gain is " << AccWinHeatGain[iSurf] * convertJtoGJ << ", and energy loss is " << AccWinHeatLoss[iSurf] * convertJtoGJ << ".\n"; myfile.close(); } for (iZone = 1; iZone <= NumOfZones; ++iZone) { ofstream myfile2(path2, std::ios_base::app); myfile2 << "Zone # is " << iZone-1 << ", energy gain is " << ZonePreDefRep(iZone).SHGSAnWindAdd * convertJtoGJ << ", and energy loss is " << ZonePreDefRep(iZone).SHGSAnWindRem * convertJtoGJ << ".\n"; myfile2.close(); } // End of subroutine (Ehsan Kamel)

406

A.2 Subroutines Used for Other Components (Wall, Roof,

Ceiling, Doors, and IntMass):

Under the HeatBalanceSurfaceManager module:

//Ehsan using namespace DataSizing; using DataSurfaces::AccCompEnergyCond; //End Ehsan ______//Ehsan if ((!WarmupFlag) && (ZoneSizingRunDone) && (SysSizingRunDone)) { AccCompEnergyCond[SurfNum] += OpaqSurfAvgFaceConductionEnergy(SurfNum); } //End Ehsan ______

//Ehsan using DataZoneEquipment::ZoneEquipConfig; using DataSurfaces::SurfaceClass_Wall; using DataSurfaces::SurfaceClass_Roof; using DataSurfaces::SurfaceClass_Floor; using DataSurfaces::SurfaceClass_Door; using DataSurfaces::SurfaceClass_IntMass; using DataSurfaces::AccCompEnergyCond; using namespace DataSurfaces; using namespace DataEnvironment; #include #include using namespace std; const char *path3 = "C:/Users/MZK221/Desktop/Heat Gain or Loss for Walls.txt"; const char *path4 = "C:/Users/MZK221/Desktop/Heat Gain or Loss for Roofs.txt"; const char *path5 = "C:/Users/MZK221/Desktop/Heat Gain or Loss for Floors.txt"; const char *path6 = "C:/Users/MZK221/Desktop/Heat Gain or Loss for Doors.txt"; const char *path7 = "C:/Users/MZK221/Desktop/Heat Gain or Loss for IntMass.txt"; if ((! WarmupFlag) && (ZoneSizingRunDone) && (SysSizingRunDone) && (EndDesignDayEnvrnsFlag)) { for (SurfNum = 1; SurfNum <= TotSurfaces; ++SurfNum) { 407

ZoneNum = Surface(SurfNum).Zone; if ((ZoneNum - 1) != 0) { if (Surface(SurfNum).Class == SurfaceClass_Wall) { ofstream myfile(path3, std::ios_base::app); myfile << "Surface # is " << SurfNum << ", in zone " << ZoneNum - 1 << ", heat conduction is " << AccCompEnergyCond[SurfNum] / 1000000000 << ".\n"; myfile.close(); } if (Surface(SurfNum).Class == SurfaceClass_Roof) { ofstream myfile(path4, std::ios_base::app); myfile << "Surface # is " << SurfNum << ", in zone " << ZoneNum - 1 << ", heat conduction is " << AccCompEnergyCond[SurfNum] / 1000000000 << ".\n"; myfile.close(); } if (Surface(SurfNum).Class == SurfaceClass_Floor) { ofstream myfile(path5, std::ios_base::app); myfile << "Surface # is " << SurfNum << ", in zone " << ZoneNum - 1 << ", heat conduction is " << AccCompEnergyCond[SurfNum] / 1000000000 << ".\n"; myfile.close(); } if (Surface(SurfNum).Class == SurfaceClass_Door) { ofstream myfile(path6, std::ios_base::app); myfile << "Surface # is " << SurfNum << ", in zone " << ZoneNum - 1 << ", heat conduction is " << AccCompEnergyCond[SurfNum] / 1000000000 << ".\n"; myfile.close(); } if (Surface(SurfNum).Class == SurfaceClass_IntMass) { ofstream myfile(path7, std::ios_base::app); myfile << "Surface # is " << SurfNum << ", in zone " << ZoneNum - 1 << ", heat conduction is " << AccCompEnergyCond[SurfNum] / 1000000000 << ".\n"; myfile.close(); } } } } //End Ehsan

408

Under the DataSurfaces header and .cc:

//Ehsan float AccWinHeatGain[80]; float AccWinHeatLoss[80]; float AccCompEnergyCond[80]; //End Ehsan

409

Ehsan Kamel Phone: 814-933-5111 Email: [email protected] www.linkedin.com/in/ehsan-kamel-727607129

Education  PhD in Civil Engineering, The Pennsylvania State University, University Park, PA (GPA = 3.82/4.00) Thesis title: “Development of a Platform for Modeling and Improving Energy Efficiency in Smart Homes"  M.Sc. in Civil Engineering, Tehran University (in top three universities in Iran), 2010-13  B.Sc. in Civil Engineering, Amirkabir University of Technology (Tehran Polytechnic), (in top three universities in Iran), 2006-10 Areas of Focus

Building Enclosure and Energy Retrofit Structural

 Energy efficient buildings  Experimental study and computer modeling of  Building enclosure and energy retrofit structural and seismic behavior of buildings.  Energy modeling and application of BIM Mechanical properties of normal and lightweight  Smart homes and application of sensors concrete (structural and non-structural). Research Experiences

 Design Engineer and Leader U.S. Department of Energy Race to Zero & Challenge Home Student Design Competition, 2014 and 2015. o Led the Penn State building enclosure and energy design teams. o Managed group of 10 students in two teams to design a net-zero energy and sustainable house and won Design Excellence and Systems Integration Excellence awards. o Designed the building enclosure for Penn State team and won the Best Technical Integration award.  Researcher and Author Authored three technical reports for Pennsylvania Housing Research Center (PHRC) as a research assistant, The Pennsylvania State University. It led to winning the Roger Glunt Graduate Fellowship in Residential Building Construction twice in 2015 and 2016. Report’s topics are as follow: o Building Envelope Energy Retrofit (2015) o Moisture Management in Residential Buildings (2014) o Experiences Learnt from U.S. Solar Decathlon Competitions (2013)  Instructor and Teacher Assistant o Special lecturer in “Building Enclosure Science and Design” on the topics of energy modeling and analysis, THERM, and Window. The Pennsylvania State University, 2014-16. o Instructed “Structural Analysis”, The Pennsylvania State University, 2014. o Teaching Assistant in “Design of Concrete Structures”, “Design of Steel Structures”, and “Structural Analysis by Matrix Methods”, The Pennsylvania State University, 2015-16. o Teaching Assistant in “Statics” and “Engineering Economics” courses, Amirkabir University of Technology, Iran, 2007-12. o Teaching Assistant in “Engineering Economics”, Amirkabir University of Technology, Iran, 2007-10. Computer and Language Skills Proficient in: BEopt OpenStudio THERM WINDOW WUFI EnergyPlus Energy-10 3D Studio SAP SAFE ETABS AutoCAD Revit COMFAR Max Familiar Python Visual Basic MATLAB Minitab Artificial Neural Network (ANN) with: Persian: Mother Tongue English: Fluent German: Basic Knowledge