Cost Effective, Zero Energy Home Designs for Temperate and Tropical Climates in Australia

Cost effective, zero energy home designs for temperate and tropical climates in Australia

Sihong GONG

A thesis in fulfilment of the requirements for the degree of Doctor of Philosophy

School of Photovoltaic and Renewable Energy Engineering Faculty of Engineering University of New South Wales Sydney, Australia

January 2020

Thesis/Dissertation Sheet

Surname/Family Name : Gong Given Name/s : Sihong Abbreviation for degree as give in the University calendar : PhD Faculty : Engineering School : School of Photovoltaic and Renewable Energy Engineering Cost effective, zero energy home designs for temperate and tropical climates Thesis Title : in Australia

Abstract 350 words maximum: (PLEASE TYPE) Due to the renewed interest in Australia in improving the minimum energy performance in the building code and the declining price of a residential PV system, there is an opportunity to examine more closely the cost effectiveness of improvements to building thermal performance and PV systems. Moreover, reducing peak thermal loads is another important issue to be considered due to the increasing usage of air conditioning.

In this thesis, experimental measurements of a low energy dwelling in Perth, Western Australia, were compared with modelled results. The results indicated that the dwelling could perform as designed with little auxiliary heating and cooling. Using the validated model as a starting point, the design of the dwelling was optimised to minimise the cost of construction and operating energy consumption using EnergyPlus. Three locations in Australia (Sydney, Melbourne and Darwin) were investigated and a parametric study was conducted. In the absence of a PV system, the most cost-effective design achieved a NatHERS star rating of 7.6-star in Sydney, 7.0-star in Melbourne and 6.3-star in Darwin. When combined with a PV system, designed to offset the electricity costs of HVAC, the PV system was deemed cost-effective based on a benefit-cost ratios analysis. After being combined with a PV system, the 8-star design in Sydney was only marginally more expensive than the most cost-effective option of a 6-star design. In Melbourne and Darwin, the most cost-effective design option was a 7-star and 6-star design, respectively.

To reduce peak electricity loads, preconditioning strategies were investigated for a range of house designs integrated with a 5 kW PV system in the three cities. Weather data for the most extreme hot and cold periods over the past 10 years were selected. Two preconditioning scenarios were considered which consisted of preconditioning with continuous HVAC and constant set points and preconditioning with additional thermal conditioning powered by surplus PV generation. The results indicated that peak cooling loads were higher than the peak heating loads for all three cities. To reduce peak cooling loads, the latter preconditioning scenario achieved the best overall performance for reducing peak loads and imported electricity usage.

Declaration relating to disposition of project thesis/dissertation

I hereby grant to the University of New South Wales or its agents a non-exclusive licence to archive and to make available (including to members of the public) my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known. I acknowledge that I retain all intellectual property rights which subsist in my thesis or dissertation, such as copyright and patent rights, subject to applicable law. I also retain the right to use all or part of my thesis or dissertation in future works (such as articles or books).

…………………………………………………………… ……….……………………...…….… Signature Date The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years can be made when submitting the final copies of your thesis to the UNSW Library. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research.

ORIGINALITY STATEMENT

‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’

Signed ……………………………………………......

Date ……………………………………………......

‘I hereby grant the University of New South Wales or its agents a non-exclusive licence to archive and to make available (including to members of the public) my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known. I acknowledge that I retain all intellectual property rights which subsist in my thesis or dissertation, such as copyright and patent rights, subject to applicable law. I also retain the right to use all or part of my thesis or dissertation in future works (such as articles or books).’

‘For any substantial portions of copyright material used in this thesis, written permission for use has been obtained, or the copyright material is removed from the final public version of the thesis.’

Signed ……………………………………………......

Date ……………………………………………......

AUTHENTICITY STATEMENT ‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis.’

Signed ……………………………………………......

Date ……………………………………………......

INCLUSION OF PUBLICATIONS STATEMENT

UNSW is supportive of candidates publishing their research results during their candidature as detailed in the UNSW Thesis Examination Procedure.

Publications can be used in their thesis in lieu of a Chapter if: • The candidate contributed greater than 50% of the content in the publication and is the “primary author”, i.e. the candidate was responsible primarily for the planning, execution and preparation of the work for publication • The candidate has approval to include the publication in their thesis in lieu of a Chapter from their supervisor and Postgraduate Coordinator. • The publication is not subject to any obligations or contractual agreements with a third party that would constrain its inclusion in the thesis

Please indicate whether this thesis contains published material or not: ☐ This thesis contains no publications, either published or submitted for publication

Some of the work described in this thesis has been published and it ☒ has been documented in the relevant Chapters with acknowledgement

This thesis has publications (either published or submitted for ☐ publication) incorporated into it in lieu of a chapter and the details are presented below

CANDIDATE’S DECLARATION

I declare that:

• I have complied with the UNSW Thesis Examination Procedure • where I have used a publication in lieu of a Chapter, the listed publication(s) below meet(s) the requirements to be included in the thesis. Candidate’s Name Signature Date (dd/mm/yy) Sihong Gong 20/06/2020

Abstract

Due to the renewed interest in Australia in improving the minimum energy performance in the building code and the declining price of a residential PV system, there is an opportunity to examine more closely the cost effectiveness of improvements to building thermal performance and PV systems. Moreover, reducing peak thermal loads is another important issue to be considered due to the increasing usage of air conditioning.

Acknowledgements

I would like to express a heartfelt thank you to all the people whose assistance were a milestone in the completion of this Ph.D. They were very generous with their time and valuable advice in helping me get through the different periods of my Ph.D. study.

First, I wish to express my sincere gratitude to my supervisor, Professor Alistair Sproul, for his guidance, feedbacks, encouragement and financial support. He patiently guided and encouraged me through this period. Without his persistent help, the goal of this project would not have been realised. Next, I would like to thank my Joint-supervisor Dr. Jessie Copper. I would like to recognise the invaluable assistance that she has provided for me during my study. Thank you for sharing all the valuable ideas, giving constructive suggestions and providing support on building energy simulation software. I also want to thank John Blair for teaching me the methodology to present the research works as clearly as possible. In addition, I would like to say thank you to Kanyawee Keeratimahat, Baran Yildiz and Shaoyang Liu for giving invaluable guidance on the analysis of data.

I would also like to acknowledge the financial support of the University International Postgraduate Award to make this thesis possible.

I am extremely grateful for the support and great love of my family especially my husband, Dayang Wang; my parents and parents in law; Rengui Gong, Lanping Huang, De Wang and Lanying Wu; my brother, Wenkai Gong and my best friends, Qi Yangliang, Elina Tan and Nan Wang. They kept me going despite all the challenges and difficulties faced and this work would not have been possible without their input. Also, without their encouragements, my Ph.D. study would be much more insipid. I also would like to thank, Nemo and Trolley for all the happiness you brought into my life during my Ph.D. study.

Contents

List of Figures ...... vii

List of Tables ...... xii

List of Abbreviations ...... xiv

Chapter 1 Introduction ...... 1

1.1 Background ...... 1

1.2 Motivation for and Objectives of the Research ...... 5

1.3 Thesis Outline ...... 7

Chapter 2 Literature Review ...... 10

2.1 Building Energy Simulation ...... 10

2.1.1 Building Energy Simulation Software - AccuRate ...... 11

2.1.2 Building Energy Simulation Software – EnergyPlus ...... 15

2.1.3 AccuRate versus EnergyPlus ...... 18

2.1.4 Summary ...... 21

2.2 Low Energy Houses ...... 22

2.2.1 Introduction ...... 22

2.2.2 Worldwide Zero Energy Building Strategies and Australia Building Standards ...... 22

2.2.3 The Building Envelope ...... 24

2.2.4 Cost Effective Low/Zero Energy Homes ...... 30

2.2.5 Summary ...... 35

2.3 Precooling/Preheating with a PV System ...... 36

2.3.1 Introduction ...... 36

2.3.2 Precooling and Preheating of Residential Buildings ...... 37

2.3.3 Coupling Solar PV with Precooling scenarios ...... 44

iii

2.3.4 Summary ...... 48

2.4 Conclusions ...... 48

Chapter 3 Case Study and Model Validation ...... 50

3.1 Introduction ...... 50

3.2 Case study ...... 51

3.2.1 House Description ...... 51

3.2.2 Construction ...... 53

3.2.3 Monitoring System ...... 56

3.2.4 Hilton Climate in 2015 ...... 57

3.2.5 Summary ...... 59

3.3 Modelling in AccuRate and EnergyPlus ...... 59

3.3.1 Model Structure ...... 59

3.3.2 Simulated Construction ...... 61

3.3.3 Other Assumptions ...... 63

3.3.4 Potential Uncertainties in Assumptions ...... 65

3.4 Validation Method ...... 66

3.5 Results ...... 68

3.5.1 Measured Performance in 2015 ...... 69

3.5.2 Simulated Star Rating ...... 71

3.5.3 Simulated Temperature versus Measured Temperature ...... 71

3.6 Josh’s House in Different Climate Zones ...... 74

3.7 Chapter Summary ...... 77

Chapter 4 Cost-Effective, Zero Energy Home Designs with Rooftop PV ...... 79

4.1 Simulation Tools and Other Assumptions ...... 79

4.1.1 Simulation Tools ...... 79

4.1.2 Other Assumptions ...... 82

4.2 Methodology ...... 85

4.2.1 Base Case Design ...... 85

4.2.2 Design Variables ...... 86

4.2.3 Net Present Costs ...... 89

4.2.4 Benefit-Cost Ratio ...... 94

4.3 Results ...... 95

4.3.1 Base Case Design ...... 95

4.3.2 Parametric Study ...... 96

4.3.3 Benefit-Cost Ratios – Construction ...... 101

4.3.4 PV system for the designs with lowest NPC...... 106

4.4 Chapter Summary ...... 109

Chapter 5 Precooling/Preheating with a PV system ...... 112

5.1 Methodology ...... 112

5.1.1 Selection of the Extremely Cold/Hot day ...... 116

5.1.2 General Load Profile Excluding HVAC Loads ...... 120

5.1.3 Surplus PV ...... 121

5.1.4 Equations ...... 124

5.2 Results ...... 125

5.2.1 Precondition Strategies ...... 126

5.2.2 Indoor Air Temperature for Different Scenarios ...... 127

5.2.3 Base case vs. base case with PV system ...... 132

5.2.4 Base Case vs. Precondition with Continuing HVAC ...... 136

5.2.5 Base Case vs. Precooling with Surplus PV ...... 140

5.2.6 Impacts of Different Star Rating Design on Peak Load Reduction ...... 142

5.2.7 Avoided Network Augmentation Costs ...... 145

5.2.8 Impacts of Different Design Variables on Peak Load Reduction ...... 147

5.3 Chapter Summary ...... 149

Chapter 6 Conclusions ...... 151

6.1 Original Contribution ...... 155

6.2 Limitations of the Research ...... 155

6.3 Future Work ...... 156

Appendix I Individual Results for Each Zone ...... 158

Appendix II Relative Humidity vs. Outdoor Air Temperature ...... 159

List of Publications ...... 161

References ...... 162

List of Figures

Figure 1-1. Observed GMST change and modelled future responses (Source: IPCC 2018 p. 6)...... 1 Figure 1-2. Proportions of residential energy used by householders in 2014 (Source: EnergyConsult 2015 p. 24)...... 2 Figure 2-1. Monitored (actual) vs. simulated cooling electricity consumption over a summer period of 72 hours based on: a) AccuRate’s default thermostat setting; b) actual air-conditioning operating temperatures (Source: Ren et al. 2018 p. 125)...... 13 Figure 2-2. Ground heat loss comparison between measured and simulated results from Chenath and HEAT3 for March 2003 (Source: Alterman et al. 2012 p. 14)...... 15 Figure 2-3. Measured and simulated hourly indoor temperatures in a living/kitchen area for occupied residence B from 25th to 31st May 2011 (Source: Daniel et al. 2013 p. 2713)...... 19 Figure 2-4. Comparison between hourly measured temperatures and simulated temperatures and a 5 day moving averages, replicating the AccuRate’s settings to EnergyPlus (Source: Copper 2012 p. 180)...... 20 Figure 2-5. Comparison between hourly measured temperatures and simulated temperatures and 5 day moving averages, using measured indoor temperatures as inputs for EnergyPlus (Source: Copper 2012 p. 217)...... 21 Figure 2-6. Life-cycle costs ($/m2) comparison of walls for different insulation thickness of residential dwellings in Canada (Source: Harvey 2006 p. 124)...... 26 Figure 2-7. Infiltration rate in different cities in Australia (Source: Ambrose & Syme 2015 p. 11)...... 29 Figure 2-8. Net present cost comparison for the base case and representative optimisation cases (Source: Bambrook et al. 2011 p.1708)...... 31 Figure 2-9. Recommended precooling scenarios and associated peak demand for a high- performance home in Sacramento (Source: German et al. 2014 p. 1-77)...... 39 Figure 2-10. The annual cumulative cooling load shapes for the deep precooling scenarios and the reference case for Miami. The red rectangle refers to the peak period (Source: Turner et al. 2015 p. 1064)...... 41

vii

Figure 2-11. Ratio of the annual off-peak cooling loads increase to the annual peak cooling loads decrease (Source: Turner et al. 2015 p. 1066)...... 41 Figure 2-12. Precooling set-point strategies (Source: Fournier & Leduc 2014 p. 4)...... 43 Figure 2-13. The demand profile result for the “basic” precooling level (Source: Fournier & Leduc 2014 p. 11)...... 43 Figure 2-14. The demand profile result for the “smart” precooling level (Source: Fournier & Leduc 2014 p. 11)...... 43 Figure 2-15. The demand profile result for the “advanced” precooling level (Source: Fournier & Leduc 2014 p. 12)...... 44 Figure 2-16. Conventional air conditioning (AC) cooling (top) and smart air conditioning with precooling (bottom) (Source: O’Shaughnessy et al. 2018 p. 14) ...... 45 Figure 2-17. Customer electricity load profiles and PV profiles under stand-alone solar (top) and solar plus scenarios (bottom) for the Hawaii case study. Pink columns indicate the peak periods. BESS refers to battery energy storage system; Misc refers to miscellaneous loads; Net load is the total grid electricity usage (Source: O’Shaughnessy et al. 2018 p. 19)...... 46 Figure 2-18. Load profiles – sample home 1 (Source: Arababadi & Parrish 2017 p. 287) ...... 47 Figure 2-19. Load profiles – sample home 3 (Source: Arababadi & Parrish 2017 p. 288) ...... 47 Figure 3-1. Josh’s House plan (left is the back dwelling; right is the front dwelling) (Source: Low Carbon Living CRC, 2018)...... 52 Figure 3-2. The floorplan of Josh’s House with 5 conditioned zones (activity area, living/kitchen area, master suite, bedroom 2 and bedroom 3) (Source: Low Carbon Living CRC, 2018)...... 52 Figure 3-3. The front photo of the back dwelling (facing north) (Source: Low Carbon Living CRC, 2018)...... 52 Figure 3-4. The details of the wall construction in Josh’s House: a) double brick wall and reverse brick wall; b) stud wall (Source: Low Carbon Living CRC, 2018)...... 53 Figure 3-5. The details of the ceiling construction in Josh’s House (Source: Low Carbon Living CRC, 2018)...... 54

viii

Figure 3-6. The top view photo of dwelling and the red circle highlighted the removable shading above the large northern windows: a) in hot periods; b) in cold periods. (Source: Nearmap, 2019) ...... 55 Figure 3-7. Josh’s House monitoring system design for wall/slab in-situ temperature sensors (Source: Low Carbon Living CRC, 2018)...... 57 Figure 3-8. Average daily outdoor temperature in 2015 (Source: Bureau of Meteorology, 2015; National Centers for Environmental Information, 2017)...... 58 Figure 3-9. Comparison of measured average monthly air temperature in 2015 and average monthly air temperature in RMY...... 58 Figure 3-10. A screen shot of the project page of Josh’s House in the AccuRate ...... 60 Figure 3-11. The layout of Josh’s House modelled in SketchUp ...... 60 Figure 3-12. Measured average daily indoor temperatures in 2015 for the conditioned areas...... 69 Figure 3-13. Heating and cooling hours per day in a) the living/kitchen area; b) Bedroom 3...... 70 Figure 3-14. Hourly frequency distribution plots of whole building average hourly indoor air temperatures from measured data (sensors) and simulated data from AccuRate and EnergyPlus...... 72 Figure 3-15. Residuals for the whole building average hourly indoor air temperatures between measured data and simulated data from AccuRate and EnergyPlus...... 73 Figure 3-16. Thermal energy requirements and the corresponding star rating of Josh’s House in different BCA climate zones...... 75 Figure 4-1. 3D rendering of the northern façade of the case study 3-bedroom dwelling...... 84 Figure 4-2. Difference in net present costs vs. star rating for Sydney ...... 98 Figure 4-3. Difference in net present costs vs. star rating for Melbourne...... 98 Figure 4-4. Difference in net present costs vs. star rating for Darwin...... 99 Figure 4-5. Cost distribution for the best 6- to 10-star design options in Sydney: a) with grid electricity costs for HVAC load; b) with a PV system offsetting electricity costs for the HVAC load...... 107

Figure 4-6. Cost distribution for the best 6- to 9-star design options in Melbourne: a) with grid electricity costs for HVAC load; b) with a PV system offsetting electricity costs for the HVAC load...... 107 Figure 4-7. Cost distribution for the best 6- to 8-star design options in Darwin: a) with grid electricity costs for HVAC load; b) with a PV system offsetting electricity costs for the HVAC load...... 107 Figure 5-1. Outdoor air temperature on the days around the extremely hot day in a) Sydney; b) Melbourne; c) Darwin...... 119 Figure 5-2. Outdoor air temperature on the days around a) the extremely cold sunny day in Sydney; and b) the extremely cold cloudy day in Sydney...... 119 Figure 5-3. Outdoor air temperature on the days around the extremely cold cloudy day in Melbourne...... 120 Figure 5-4. The average daily general load profile excluding HVAC loads for a) summer; b) winter...... 121 Figure 5-5. Hourly PV generation and surplus PV after deducting the summer general load profile and excluding HVAC loads on the extremely hot day in a) Sydney; b) Melbourne; c) Darwin...... 123 Figure 5-6. Hourly PV generation and surplus PV after deducting the winter general load profile, on a) the extremely cold sunny day in Sydney; b) the extremely cold cloudy day in Sydney; c) the extremely cold cloudy day in Melbourne...... 124 Figure 5-7. Indoor air temperatures for the three scenarios on the extremely hot day in: a) Sydney; b) Melbourne; c) Darwin...... 129 Figure 5-8. Indoor air temperatures for the base case on cold days in the free running mode in a) Sydney under normal shading conditions; b) in Sydney under the worst-case shading condition...... 131 Figure 5-9. Indoor air temperatures for base case and preheating scenario with heating continuously on the extremely cold cloudy day in a) Sydney; and b) Melbourne...... 132 Figure 5-10. The imported/exported electricity for base case vs. the base case with PV system on the extremely hot day in a) Sydney; b) Melbourne; c) Darwin...... 134 Figure 5-11. The imported/exported electricity for base case on the extremely cold cloudy day under the normal shading and worst-case shading conditions in a) Sydney; b) Melbourne...... 135 x

Figure 5-12. The imported/exported electricity for base case vs. preconditioning scenario with cooling continuously on the extremely hot day in a) Sydney; b) Melbourne; c) Darwin ...... 137 Figure 5-13. Imported/exported electricity for the base case vs. the preconditioning scenario with heating continuously on the extremely cold cloudy day in a) Sydney under normal shading; b) Sydney under worst-case shading; c) Melbourne under normal shading; d) Melbourne under worst-case shading...... 139 Figure 5-14. Imported/exported electricity for the base case vs. the precondition scenario with additional thermal conditioning powered by surplus PV generation on the extremely hot day in a) Sydney; b) Melbourne; c) Darwin...... 141 Figure 5-15. The extremely hot, peak load during the peak periods in Sydney for different star rating designs at the basic cooling thermostat set point: a) 24°C and b) 22°C. .... 144 Figure 5-16. The extremely hot day, peak load during the peak periods in Melbourne for different star rating designs at the basic cooling thermostat set point: a) 24°C and b) 22°C...... 144 Figure 5-17. The extremely hot day, peak load during the peak periods in Darwin for different star rating designs at the basic cooling thermostat set point: a) 24°C and b) 22°C...... 144 Figure 5-18. The best 6- to 10-star design options in Sydney: a) Cost distribution of PV, construction, HVAC and peak reduction savings; b) Sum of the costs and savings. .... 146 Figure 5-19. The best 6- to 9-star design options in Melbourne: a) Cost distribution of PV, construction, HVAC and peak reduction savings; b) Sum of the costs and savings. .... 147 Figure 5-20. The best 6- to 8-star design options in Darwin: a) Cost distribution of PV, construction, HVAC and peak reduction savings; b) Sum of the costs and savings. .... 147 Figure 5-21. Peak load reduction (%) from the precooling scenarios with additional thermal conditioning powered by surplus PV generation. The peak load reduction is related to the peak cooling demand for the base case which is 3.22 kW for 22°C and 2.77 kW for 24°C...... 148 Figure II-1. The air temperatures and relative humidity in the selected hot periods in a) Sydney, b) Melbourne and c) Darwin...... 160

List of Tables

Table 2-1. Pearson correlation coefficient between the residuals and different factors for the living/kitchen/dining zone in the Rose Bay, Australia dwelling (Source: Copper 2012 p. 215)...... 19 Table 2-2. Specifications for Case 1, Case 3 and Case 10 (Source: Bambrook et al. 2011 p.1708)...... 32 Table 2-3. Reference and precooling thermostat setpoints (Source: Turner et al. 2015 p. 1062)...... 40 Table 3-1. Summary of Josh’s House construction (Source: Low Carbon Living CRC, 2018) ...... 55 Table 3-2. Details of the simulated construction of Josh’s House in AccuRate and EnergyPlus...... 61 Table 3-3. The assigned zone types and corresponding conditioned period and assumptions (source: AccuRate Help File, 2019)...... 63 Table 3-4. NatHERS energy star and the corresponding maximum thermal energy loads in MJ/m2.annum for each star rating (Source: NatHERS, 2019a)...... 64

Table 3-5 Acceptable calibration tolerances for statistical indices of MBE and CV(RMSE). (Source: U.S. Department of Energy, 2015 p. 4-20) ...... 67 Table 3-6. Statistical metrics related the measured and simulated whole building average hourly indoor temperatures...... 74 Table 3-7. Performance of Josh’s House in different climate zones...... 75 Table 4-1. Design variables and costs for the parametric analysis (excluding GST)...... 86 Table 4-2. Assumed interest rates and inflation rates for Sydney and Melbourne...... 92

Table 4-3. Installed PV rooftop system capital price CPV ($AUD/kWp) in Sydney, Melbourne and Darwin without government incentives...... 94 Table 4-4. Base case designs for Sydney, Melbourne and Darwin...... 95 Table 4-5. Details of options with the lowest ∆NPC for 6 to 10-star ratings in Sydney100 Table 4-6. Details of designs with the lowest ∆NPC at 6 to 9-star ratings in Melbourne...... 100 Table 4-7. Details of designs with the lowest ∆NPC at 6 to 8-star ratings in Darwin. .. 101 Table 4-8. Benefit-Cost ratios for different design options in Sydney...... 102 xii

Table 4-9. Benefit-Cost ratios for different design options in Melbourne...... 103 Table 4-10. Benefit-Cost ratios for different design variables in Darwin...... 103 Table 5-1. The most cost-effective 6-star house design in Sydney, Darwin and Melbourne...... 114 Table 5-2. The selected extremely hot and cold days in Sydney, Melbourne and Darwin...... 117 Table 5-3. The hourly thermostat settings for the three scenarios for the 6-star design in Sydney on the extremely hot day...... 127 Table I-1. Statistical metrics related the measured and simulated (AccuRate) indoor temperatures for the five major zones using the 2015 weather file from the BoM. ... 158 Table I-2. Statistical metrics related the measured and simulated (EnergyPlus) indoor temperatures for the five major zones using the 2015 weather file from the BoM. ... 158

xiii

List of Abbreviations

ACH@50Pa Air changes per hour at 50 Pascal BASIX Building Sustainability Index BCA Building Code of Australia BCRs Benefit-cost ratios BEopt Building Energy Optimisation Tool BoM Bureau of Meteorology COP Coefficient of performance CSI Clear sky index

CV(RMSE) Coefficient of variance of the root mean square error DG Double glazed EER Energy efficiency ratio GMST Global mean surface temperature HVAC Heating, ventilation, and air conditioning LCLCRC Low Carbon Living Cooperative Research Centre M&V Guidelines Measurement and verification guidelines m2K/W Metres squared Kelvin per watt MBE Mean bias error NatHERS Nationwide House Energy Rating Scheme NPC Net present costs NSW New South Wales PV Photovoltaic RMSE Root mean squared error RMY Representative meteorological years SG Single Glazed SHGC Solar heat gain coefficient U.S. United States ZEB Zero energy buildings ZEH Zero energy house

xiv

Introduction

Chapter 1 Introduction

1.1 Background

According to the Intergovernmental Panel on Climate Change (IPCC, 2018), the observed global mean surface temperature (GMST) has increased approximately 0.87°C from 2006 and 2015 compared with the pre-industrial period from 1850 to 1900, as presented in Figure 1-1. At the current rate of increase, the GMST is likely to increase 1.5°C in the period from 2030 to 2052, which may cause long-lasting or irreversible impacts like loss of some ecosystems. Many other negative impacts related to global warming have already been observed, such as an increase in the number of extreme weather events, sea level rise and increasing risks for human health and the economy (European Environment Agency, 2017; NASA, 2019).

Figure 1-1. Observed GMST change and modelled future responses (Source: IPCC 2018 p. 6).

Since the mid-20th century, greenhouse gases from human activities have been the main cause of observed global warming (IPCC, 2013), and the electricity and heat production sector have been the primary source (25%) of global greenhouse gas emissions (IPCC, 1

Introduction

2014). In Australia, electricity generation is also the largest source of national greenhouse gas emissions (35%) as they are primarily from the combustion of black coal and brown coal (Commonwealth of Australia, 2016). Energy consumption in the residential sector accounts for 11% of total final energy consumption (Department of Industry and Science, 2015). The largest proportion of total energy used by households is for space conditioning, which accounted for 40% in 2014, as shown in Figure 1-2 (EnergyConsult, 2015).

Figure 1-2. Proportions of residential energy used by householders in 2014 (Source: EnergyConsult 2015 p. 24).

The design and construction of low-energy residential dwellings is one potential method to reduce electricity usage and greenhouse gas emissions within the Australian residential sector. Typically, low-energy residential dwellings have been achieved by either upgrading the energy efficiency of lighting and appliances or by modifying the building design (De Boeck et al., 2015). However, additional improvements to the thermal design of a residential dwelling will reduce the need for heating and cooling but will most likely deliver diminishing returns on investment. Hence, it is important to understand whether further increases to the energy star rating of Australian homes are cost-effective for the average homeowner, particularly in the light of new low-cost photovoltaics (PV).

Introduction

The installation of rooftop PV systems has been widely used as a method to achieve low- energy residential dwellings (Wells et al., 2018). The price of residential PV systems has dramatically decreased over the past decade with average cost reductions of 48% between December 2012 and November 2018 (Solar Choice, 2019a). With the trend of increasing electricity prices (Australian Competition and Consumer Commission, 2017; Parisot & Nidras, 2016) and decreasing costs for PV systems (IRENA, 2016; Solar Choice, 2019a), the cost of electricity generated from PV is lower than the average residential electricity rate of $0.29/kWh (Jäger-Waldau, 2018).

According to a report by Energy Action, in Australia, solar energy systems are a cost- effective option for commercial buildings, potentially more cost-effective than many high-performance building components (Harrington et al., 2018). Similarly, a study by Parker in 2009 (Parker, 2009) indicated that adding a PV system to houses trying to meet the Passivhaus standard, was a more cost-effective solution than over-investment in energy efficient measures, such as further increasing levels of air tightness or insulation. The price of PV systems is even lower today in comparison to Parker’s study in 2009, so it is worthwhile revisiting the comparison between PV systems and energy efficiency measures in Australia.

In addition to reducing annual energy requirements and greenhouse gas emissions of a residential dwelling, reducing the peak cooling/heating loads is another important issue to be considered due to the increasing usage of air conditioning. Heatwaves are the deadliest natural hazard in Australia (Coates et al., 2014). The intensity and frequency of heatwaves are increasing for the majority of Australia (Perkins-Kirkpatrick et al., 2016), making air conditioning increasingly essential. The proportion of householders with an air conditioning system increased from 48.6% in 2002 to 74% in 2014 (Australian Bureau of Statistics, 2008, 2014). The increased use of air conditioning not only raises residential electricity loads, but also raises utility peak loads. It is widely believed that residential air conditioning electricity use is a large portion of a utility’s peak loads, leading to significant investment for network augmentation1 to cope with the short term nature of peak loads (Smith et al., 2013). Based on the energy white paper from the Australian

1 An ageing grid infrastructure needs to be upgraded to cope with increasing peak demand when it is over current grid capacity. 3

Introduction

Government (Commonwealth of Australia, 2012), the installation of a 2 kW air conditioning unit costs approximately $1,500 but increases energy system costs up to $7,000 because of the additional peak demand.

Reducing or shifting peak demand is one of the most effective methods to address the load issues caused by air conditioning. Previous work has highlighted four major initiatives (Energy Supply Association of Australia, 2012):

i) Keeping consumers up to date with the latest changes to tariffs which allows them to modify their energy use period; ii) Attempting direct load control of major peak demand appliances, such as air conditioning, pool pumps and water heaters; iii) Applying small-scale distributed generation, such as PV, wind turbines, energy storage; iv) Improving energy efficiency standards, such as thermal performance of buildings and electrical appliance efficiency.

Direct load control of air conditioners can be achieved by precooling/preheating which is an operational strategy with no up-front costs (or small costs for programmable smart thermostats). Precooling/preheating the occupied area can occur in advance to avoid the peak demand period. Many studies have demonstrated that precooling/preheating a residential dwellings has the potential to effectively reduce the peak demand (Arababadi & Parrish, 2017; Fournier & Leduc, 2014; German et al., 2014; Rijksen et al., 2010).

Roof-top PV systems are another way of coping with energy demand. However, the mismatch between residential peak electricity usage and PV generation is a major challenge for the widespread deployment of residential PV systems (Boyle, 2012; Palminter et al., 2016; Schmalensee, 2015). PV generation depends on solar irradiance impinging on the PV modules. Over a day, solar irradiation peaks at solar noon. Thus, PV systems usually have excess generation during the middle of the day but have diminishing output during the late afternoon and evening when residential electricity loads tend to increase after people arrive home from work and other activities. Grid export is an economic solution, in which excess PV electricity generation is sold to the

Introduction

grid via different exporting policies, such as feed-in-tariffs in Australia, Europe and Asia, and net metering in United State (U.S.) (Couture & Gagnon, 2010; Yamamoto, 2012).

However, the oversupply of PV generated electricity to the grid can also cause grid stability issues. The amount of installed roof-top PV capacity has increased significantly over recent years such that the average daily peak generation at mid-day increased 25% from 2017 to 2018 in Australia (Australian Energy Market Operator, 2018). According to Energy Network Australia (Narayan, 2019), the networks are facing new operational challenges related to maintaining grid stability caused by the large amount of electricity generated by PV systems being fed back into the grid. In addition, grid export rates are decreasing in many of the major PV markets around the world (Dehamma et al., 2015; Jäger-Waldau, 2016). In Australia, most of the current feed-in-tariffs are around 6-16 c/kWh which is significantly less than the 20-60 c/kWh feed-in-tariffs on offer in 2011 (Martin, 2016; Solar Choice, 2019b). The decreased grid export rates encourage consumers to use the generated electricity on-site instead of exporting it. By integrating precooling/preheating scenarios using PV systems, this approach provides the opportunity to simultaneously flatten the export/import load profile, maximise economic benefits and optimise grid impacts (Arababadi & Parrish, 2017; German et al., 2014; U.S. Department of Energy, 2014a). The latter particularly defers network augmentation and increases grid reliability as the peak PV generation period usually matches the preconditioning period.

1.2 Motivation for and Objectives of the Research

The motivation for this research consists of the following three main points:

1. Without calibration, simulation models using default settings like weather files, occupancy schedules and electricity load profiles can have extremely low accuracy in predicting building energy usage. In addition, in Copper’s study (2012), a low energy house located in Sydney, Australia was simulated in AccuRate’s Chenath engine but Copper’s study (2012) indicated that there are potential issues in the engine. In this thesis, a case study of a high efficiency

Introduction

house located in Perth, Australia was accredited to a 10-star rating under NatHERS2 by also using the AccuRate Chenath engine with all the default settings. The aim was to test the real world performance of the high efficiency home and check whether it is able to perform as designed. In addition, the evaluation is necessary as the validated house model will be used for all the following research in this thesis; 2. Recently, there has been renewed interest in Australia in improving the minimum energy performance in the building code (Bannister et al., 2018) and many researchers are pushing for greater building efficiency. At the same time, the price of a residential PV system has declined by 8% per year for the past six years (Solar Choice, 2019a). This brings up the question of how far that could go given the diminishing return in terms of energy saved and costs incurred. Besides, adding a PV system is potentially a more cost-effective solution than over- investment in energy efficient measures. Hence, there is an opportunity to examine more closely the cost effectiveness of improving the thermal envelope of a home verses PV and heating, ventilation, and air conditioning (HVAC); and 3. Integrating precooling/preheating scenarios with PV systems provides an opportunity to simultaneously lower peak thermal loads and maximise the economic benefits (Arababadi & Parrish, 2017; German et al., 2014; U.S. Department of Energy, 2014a). However, only few studies have combined precooling/preheating with PV systems for reducing peak thermal loads worldwide and there is no relevant study that has been conducted in Australia.

Therefore, there are five objectives that this thesis addresses which will deal with the knowledge gaps:

1. To test whether the real performance of the case study, a 10-star rated dwelling under NatHERS, is able to operate as the initial design intended; 2. To develop a validated model that would be used in the rest of the thesis when testing the performance of the case study design;

2 The Nationwide House Energy Rating Scheme (NatHERS) is a star rating system determining house thermal comfort on a scale of energy star rating from zero to ten stars based on the area-adjusted thermal loads simulated from accredited software. The higher the star rating, the lower the energy usage and the more comfortable the house is (NatHERS, 2019b). 6

Introduction

3. To examine the cost effectiveness of various design variables, a PV and an HVAC system for zero energy houses by conducting a parametric study in Sydney, Melbourne and Darwin, Australia; 4. To determine whether the combination of the precooling and preheating scenarios and a PV system is able to reduce the peak thermal loads without increasing the total imported electricity in the three cities in Australia; and 5. To investigate how different relatively cost-effective design variables and increased star ratings enhance the peak load reduction when combined with precooling scenarios and a PV system in the three cities.

1.3 Thesis Outline

There are three major parts involved in this thesis. The first part is a case study, evaluating the thermal performance a low energy dwelling in Perth, Australia (Chapter 3). The second part explores a cost-effective zero energy house (ZEH) design in Sydney, Melbourne and Darwin, Australia (Chapter 4) and the final part investigates the combination of precooling/preheating with PV system on peak load reduction for the three locations (Chapter 5).

Chapter 2 is a literature review which covers the three major areas of investigation in this thesis. The first section covers building energy simulation with a focus on past studies that examine the comparison between measured and simulated performance using AccuRate and EnergyPlus. Also, a comparison of the AccuRate and EnergyPlus programs is also presented. The second section covers literature related to cost- effective ZEHs. This section summarises the worldwide zero energy building strategies and current Australian Building standards, building envelope elements, as well as low energy housing design optimisation. The third section provide a review of studies related to precooling/preheating with a PV system on peak load reduction.

Chapter 3 is related to the case study and model validation. This chapter presents the description of the case study in Perth, Australia, the simulation model, the validation method and the results of the comparison between measured and simulated

Introduction

performance. After the case study, the validated EnergyPlus model presented in Chapter 3 was used for the parametric study in Chapter 4. Chapter 4 provides the methodologies for conducting the parametric study to design ZEHs in Sydney, Melbourne and Darwin, followed by the results. The results include a parametric study for cost-effective low energy houses i) without a PV system and ii) with a PV system.

Chapter 5 investigates precooling/preheating scenarios with a PV system on the peak thermal load reduction based on the cost-effective house designs examined in the Chapter 4. There are three sets of results. The first set is related to peak load reduction from precooling/preheating scenarios combined with a PV system for 6-star house designs in Sydney, Melbourne and Darwin. The second set of results presents the influence of improving the star rating of the design on peak load reduction and the corresponding economic analyses for the three cities. The third set of results indicates how different relatively cost-effective design variables found in Chapter 4 impact the peak load reductions for Sydney.

Chapter 6 summarises the major results found in previous chapters and the main conclusions of the thesis. It also summarises the original contribution made by the research and presents limitations of the research and recommendations for future research work.

Methodology Outline

In this thesis, the methodologies are presented in each corresponding chapter. In order to test the real performance of the case study home, two simulation engines were used in this thesis for simulating the performance of the case study: AccuRate (NatHERS, 2018) and EnergyPlus (NREL, 2018a). The simulated performance of indoor temperatures was compared with the real-time measured indoor temperatures to test whether the case study home is able to operate as designed. For designing cost-effective ZEHs, a parametric study was conducted by considering different design variables including wall, ceiling and floor insulation; types of external wall, internal wall and window constructions; depth of eave and infiltration levels. Following that, the most cost- effective design for different NatHERS star ratings was combined with PV systems to offset the thermal loads to achieve ZEHs. Due to the regulatory nature of AccuRate’s

Introduction

Chenath engine, there is no parametric tool developed based on the Chenath engine. However, a range of parametric analysis tools, including jEPlus (Zhang, 2009; Zhang & Korolija, 2018) which was used in this thesis, have been developed based on EnergyPlus and have been widely used for optimisation and parametric studies. Therefore, the validated EnergyPlus model in Chapter 3 was used for the parametric study and the following preconditioning study. For the preconditioning study, two preconditioning scenarios were considered which are preconditioning with constant set points and preconditioning with additional thermal conditioning powered by surplus PV generation. The most cost-effective 6-star design defined in Chapter 4 was used for the preconditioning study. The two preconditioning scenarios were combined with a 5 kW PV system for extremely hot and cold days in the past decade in the three cities for testing the peak load reduction. Following that, different star rating designs and cost- effective design variables defined in Chapter 4, were also tested on how they enhance peak load reduction.

Literature Review

Chapter 2 Literature Review

This chapter presents the information in the literature relevant to the thesis’s aims. Three parts will be presented. The first part is related to the building energy simulation for validating the case study in the thesis. The second part is about low energy house optimisation and followed by the third part of precooling/preheating a dwelling on the peak thermal load reduction.

2.1 Building Energy Simulation

One of the aims of this thesis (Chapter 3) is to test whether a 10-star rated dwelling, known as Josh’s House, performs as designed. Josh’s House was rated by the building energy simulation software, AccuRate, under the NatHERS (Low Carbon Living CRC, 2018). Therefore, this section will summarise the literature of building energy simulation related to the comparison between the measured and simulated performance of dwellings by using AccuRate (which utilises the Chenath thermal engine). AccuRate is a rating software with many built-in assumptions, such as a default window construction library, and occupancy and appliance loads, which makes it less flexible than another well-known simulation engine, EnergyPlus. For the parametric study in Chapter 4, EnergyPlus was used as it has greater flexibility and more control over a wider range of parameters. Therefore, Josh’s house was also modelled in EnergyPlus, which will be used for the later optimisation in Chapter 4 and the precooling/preheating study in Chapter 5. Hence, a literature review on the EnergyPlus engine is presented followed by a comparison between the AccuRate and EnergyPlus programs as reported in the literature.

Literature Review

2.1.1 Building Energy Simulation Software - AccuRate

In Australia, all new residential buildings need to satisfy the minimum energy efficiency standard in the Building Code of Australia (BCA) under the National Construction Code (Australian Building Codes Board, 2019). Since 2011, the energy efficiency standard in the BCA requires residential buildings to have a minimum 6 stars energy rating (or equivalent design) under NatHERS3. There are three accredited software tools used to produce energy ratings under NatHERS which are AccuRate Version 2.3.3.13, BERS Professional Version 4.3.0.2c and FirstRate5 Version 5.2.11. Although the NatHERS accredited software tools have a different user interface, they all use the Chenath engine to perform the calculations and modelling of the building (NatHERS, 2018). The Chenath engine was developed in the 1990s and has continued to be improved by the Commonwealth Scientific and Industrial Research Organisation at the request of the Australian Government (2018).

In 2004, the validation study of the AccuRate Chenath engine was carried out using the building energy simulation test which was created by the International Energy Agency Solar Heating and Cooling Programme (Delsante, 2004). It worth noting that there was no lighting, occupancy or equipment loads modelled during the building energy simulation test process. So, only the ability of a building energy simulation program to model the dwelling thermal performance based on the external shell was tested in building energy simulation test. It was concluded that the Chenath engine showed good agreement with the reference programs with no significant discrepancies found, but the simulated cooling energy and peak thermal demands from AccuRate tended to be at the high end within the range of the results. One of the reasons mentioned for that is the different temperature calculations and control algorithms used in AccuRate. In 2006, Delsante (2006) showed the simulated data from AccuRate was similar to the measured data without any significant discrepancy for a mud brick house. The study suggested that

3 NatHERS is a star rating system determining house thermal comfort on a scale of energy star rating from zero to ten stars based on the area-adjusted thermal loads simulated from accredited software. The higher the star rating, the lower the energy usage and the more comfortable the house is (NatHERS, 2019b). 11

Literature Review

any discrepancies may be caused by the difference between actual occupant behaviour and assumed behaviour in AccuRate’s default settings.

The latest study comparing AccuRate’s simulation results with measured data (Ren et al., 2018) also highlighted the difference between the actual occupant behaviour and assumed behaviour in AccuRate’s default settings. In the study, the electricity consumption prediction was significantly improved after using the actual cooling thermostat setting and operation periods instead of the default thermostat setting in AccuRate Chenath engine. The study evaluated the capability of the Chenath engine for prediction of thermal performance by using the monitored data and actual thermostat settings from over forty dwellings located in three Australian cities (Adelaide, Brisbane and Melbourne). The study highlighted the importance of using the actual operation data (such as the usage pattern of appliances) and actual thermostat settings for accurate simulation results.

In addition, the Ren et al. study (2018) noted that the differences between actual weather and default weather inputs in the model will produce disagreement between simulated and actual energy consumption. So, in the study, all weather files were generated based on the observed data from the closest Bureau of Meteorology (BoM) weather station. The occupancy profile, sources of energy used and usage pattern of the major appliances as well as the actions taken to save energy were collected for each residential dwelling. Figure 2-1 presents the measured versus simulated cooling electricity consumption for the cases a) using AccuRate’s default thermostat setting; and b) using actual air-conditioning thermostat settings.

After replacing the default settings with actual operation data and actual thermostat settings throughout the living areas, it was found that the electricity consumption prediction was significantly improved. The statistical metric, R-squared, for the cases a) and b) was increased from 16.6% to 64.6% respectively. Although the difference between the simulated and measured thermal loads can be still large for some individual dwellings in Figure 2-1b), the estimation of electricity for space heating and cooling provided by the Chenath engine is generally reasonable when using the actual operation data and thermostat settings. The author suggested the major reasons for the differences between the simulated and measured data including the incomplete 12

Literature Review

operational data (such as the actual thermostat settings throughout the bedroom areas and the window and door operations) and unexpected standby power. In the last case, the occupant forgot to turn off the heaters in summer which were used for preventing the air conditioning from freezing in winter.

Figure 2-1. Monitored (actual) vs. simulated cooling electricity consumption over a summer period of 72 hours based on: a) AccuRate’s default thermostat setting; b) actual air-conditioning operating temperatures (Source: Ren et al. 2018 p. 125).

The above mentioned validation studies of the AccuRate Chenath engine showed that the simulated result from AccuRate is able to model the dwelling’ thermal performance reasonably well based on the external shell (with good agreement with the reference programs) and the engine is able to predict thermal loads reasonably accurately.

Literature Review

However, there is still a difference between measured and simulated data from AccuRate (Delsante, 2006; Ren et al., 2018). Despite the aforementioned human factors, such as different occupancy, appliance loads and natural ventilation, one of the potential issues related to AccuRate Chenath engine causing the difference is the ground coupling model (Dewsbury, 2011). The empirical validation study of the Chenath engine by Dewsbury (2011) showed that although AccuRate Chenath engine modelled energy flows well, there were constant differences between simulated and measured indoor temperatures for the test cells located in cool temperate climates. For the test cell with slab-on-ground construction, for over 3,500 (85%) hours the simulated temperature was lower than the measured temperature by at least 3°C. The simulated subfloor temperatures were consistently lower than the measured temperatures and there were monthly climatic trends for the mean difference between simulated and measured temperatures, indicating an error in the subfloor model and/or the ground model.

The ground modelling issue was also identified in Copper's study (2012) which indicated that the residuals4 between measured and simulated results were strongly correlated with the simulated ground temperature from AccuRate. Further, the ground coupling model of the Chenath engine Delsante's study (2004) was not tested as the floor construction was highly insulated. In contrast, Alterman et al (2012) validated the Chenath ground model of by comparing the simulated ground heat flow with the HEAT3 simulation results and measured ground heat flow for a slab-on-ground test house in Newcastle, Australia. The results indicated that the simulated heat losses from Chenath engine reasonably match the measured results along with the HEAT3 simulated results, as presented in Figure 2-2. However, although the averaged difference between measured and simulated results for each month (from March to November) are minimal or close to zero (up to 39 watts), the maximum hourly discrepancy reached up to 600 watts. Consequently, further investigations are needed in order to improve the Chenath ground model in future.

4 The residuals in Copper's study (2012) were calculated using the simulated temperatures minus the measured temperatures. In the following sections, the residual error in this thesis refers to the simulated temperatures minus the measured temperatures. 14

Literature Review

Figure 2-2. Ground heat loss comparison between measured and simulated results from Chenath and HEAT3 for March 2003 (Source: Alterman et al. 2012 p. 14).

Overall, AccuRate models a dwelling’s thermal performance and predicts thermal loads reasonably well. However, there is still potential for a difference between the measured and simulated data from AccuRate which may be caused by human factors such as different occupancy, appliance loads and natural ventilation levels as well as potential ground model issues in the Chenath engine. In this thesis, the simulated performance of a dwelling modelled using AccuRate will be compared with the real-time measured performance to investigate whether AccuRate can reasonably predict the performance of a dwelling.

2.1.2 Building Energy Simulation Software – EnergyPlus

Due to the potential ground model issues in AccuRate Chenath engine, a well-known whole building simulation engine, EnergyPlus, is also be used in this thesis. EnergyPlus was developed by the National Renewable Energy Laboratory with the support of the U.S. Department of Energy (NREL, 2018a). It was initially released in April 2001 (Crawley et al., 2001) with major updates twice a year, the latest version being Version 9.2.0. 15

Literature Review

EnergyPlus is a console-based simulation engine and the inputs and outputs are processed through text files without graphical interfaces. In addition, there are several more user-friendly interfaces for EnergyPlus which have been developed by third-party developers. One of the most popular cross-platform software packages, developed with the support of U.S. Department of Energy which utilises the EnergyPlus simulation engine, is OpenStudio (NREL, 2018b) and in this thesis, OpenStudio will be used for building an EnergyPlus model.

There are three majors tests and validations that have been conducted for EnergyPlus, documented on the official website, that is analytical tests, comparative tests and release and executable tests (EnergyPlus, 2015). The latest available testing report is for EnergyPlus Version 8.3.0. For the analytical tests (Henninger & Witte, 2015b), the ability of EnergyPlus to predict surface heat fluxes and temperatures and zone loads were tested. In those tests, building envelopes specified in the ASHRAE (American Society of Heating, Refrigerating and Air-Conditioning Engineers) 1052-RP report were modelled using EnergyPlus and the analytical results from EnergyPlus were very close to those reported in the ASHRAE Research Project 1052-RP Toolkit.

For the comparative tests (Henninger & Witte, 2015a), the ability of EnergyPlus to simulate thermal loads were tested. In those tests, different buildings specified in the ANSI/ASHRAE Standard 140-2011 were modelled using EnergyPlus. The results were compared to those simulated from 8 other whole building simulation software packages that participated in the 1995 International Energy Agency project. Over one hundred test cases were simulated and the comparisons showed that all results from EnergyPlus were within the range of those produced by the other 8 software packages except seven test cases related to window orientation, night ventilation and thermostat setback. The discrepancies of the seven cases were smaller than 5.5%. Hence one can say that EnergyPlus is a well-recognised building simulation engine that has been widely used for both office and residential building temperature and energy simulation (Ancrossed D Signelković et al., 2016; Boyano et al., 2013; Mateus et al., 2014; Mustafaraj et al., 2014; Yiqun Pan et al., 2010; Roberti et al., 2015; Shabunko et al., 2018). All these studies showed either the EnergyPlus simulated data was in a good agreement with

Literature Review

experimental and measured data or simulated outputs were verified against measured data.

Various optimisation and parametric analysis software packages, based on the EnergyPlus engine, have been developed such as: the Building Energy Optimisation Tool (BEopt) (NREL, 2019a); the Parametric Analysis Tool (NREL, 2013) and jEPlus (Zhang, 2009; Zhang & Korolija, 2018). BEopt is a cost-based optimisation and parametric tool that can be used to determine cost-optimal efficient design by selecting predefined options. The predefined options are designed for the U.S. situation, so it is somewhat less relevant to work conducted in Australia. The Parametric Analysis Tool in OpenStudio is a parametric tool that finds design alternatives by using EnergyPlus or OpenStudio measures5. Measures can be downloaded from an existing building component library or customised as needed. If a large set of custom measures is required, the parametric study becomes complicated. jEPlus is a simulation manager designed to set up parametric simulations with EnergyPlus. jEPlus allows users to set up parametric runs with an EnergyPlus model by defining a list of search tags and corresponding attributes of the design variables. All three packages, BEopt (Guerello et al., 2019; Wijesuriya et al., 2018), Parametric Analysis Tool (Liu & Cui, 2017; Polly et al., 2016) and jEPlus (Delgarm, Sajadi, Delgarm, et al., 2016; Delgarm, Sajadi, Kowsary, et al., 2016; Tokarik & Richman, 2016) are widely used for studies involving building energy optimisation and parametric analysis.

Overall, EnergyPlus is able to model dwelling thermal performance without large discrepancies. Further, there is a large set of parametric studies conducted using the EnergyPlus engine and a range of parametric analysis tools have been developed based on EnergyPlus. In this thesis, Chapter 4 presents a major parametric study of a dwelling in Australia by using the floor plan of the case study in Chapter 3, so jEPlus is used for the parametric study as it allows users to simultaneously run a large set of parametric runs directly with an existing EnergyPlus model.

5 A measure in OpenStudio is a set of programmatic instructions, like Excel Macro, which can be used to alter the existing model. For example, a measure of “add insulation to the roof” can be used to add different roof insulations to the model for a parametric study (NREL, 2013) 17

Literature Review

2.1.3 AccuRate versus EnergyPlus

There are a few studies that have validated the AccuRate model using the intermodal comparison between AccuRate and EnergyPlus. Daniel et al. (2013) indicated that there is a good correlation between predicting indoor temperatures using the AccuRate, EnergyPlus and EnerWin engine and measured temperatures. In that study, the validation of the Chenath engine was done by comparing the simulated results from two other simulation engines, and measured results from three occupied residential dwellings (named as residence A, B, and C) and two constructed test cells in a ‘non-rating’ mode6. The statistical analysis7 of coefficient of variance of the root mean square error

(CV(RMSE)) for the simulated and measured temperatures and graphical analysis of hourly indoor temperatures were used for comparison. The CV(RMSE) for residence B was in the range from 6.09% to 13.07%, which falls within the maximum acceptable

CV(RMSE) of 30%, defined in the measurement and verification guidelines (M&V Guidelines) (U.S. Department of Energy, 2015).

An example of the graphical analysis for residence B is presented in Figure 2-3. It can be seen that the simulated results from AccuRate and EnergyPlus have a similar pattern to the measured data, but the simulated peak temperature from AccuRate is higher than the measured temperatures. This is due to the AccuRate simulations applying the default natural ventilation schedule (which depends on indoor and outdoor temperatures), but the occupant typically opened and closed the windows in the morning and afternoon, regardless of outdoor temperatures. After modifying the ventilation schedule to reflect the real situation, the simulated peaks occurred at the same level as the measured peaks. The study concluded that AccuRate Chenath engine is able to adequately simulate buildings’ performance but many inputs, such as the natural ventilation schedule, need to be manually modified through the ‘Scratch’ file rather than the front interface of the program to achieve an accurate result.

6 There are two operation modes in AccuRate which are the rating mode and non-rating mode. The ‘non- rating’ mode is a free-running mode which estimates indoor temperatures without using HVAC. In the ‘rating’ mode, the star rating is estimated based on area adjusted thermal loads (MJ/m2). 7 The statistical analysis refers to using statistical indices related to the simulated and measured results (such as CV(RMSE)) to declare whether a simulation model is calibrated. 18

Literature Review

Figure 2-3. Measured and simulated hourly indoor temperatures in a living/kitchen area for occupied residence B from 25th to 31st May 2011 (Source: Daniel et al. 2013 p. 2713).

Despite the discrepancy caused by the in-built assumptions in AccuRate, Copper (2012) showed the residuals of the simulated results from AccuRate were strongly correlated to the simulated ground temperature (the ‘Ground Zone’ temperature in the AccuRate output) by comparing the measured and simulated indoor temperatures from AccuRate and EnergyPlus. In Copper’s study (2012), two residential dwellings at Rose Bay and North Balgowlah were investigated. Table 2-1 presents the Pearson correlation coefficient between the residuals of measured and simulated indoor temperatures and the different factors for the living/kitchen/dining zone in the Rose Bay dwelling. The Pearson correlation coefficients in Table 2-1 highlight that the ‘ground zone’ and outdoor temperatures were strongly related to the residual between measured and simulated indoor temperatures.

Table 2-1. Pearson correlation coefficient between the residuals and different factors for the living/kitchen/dining zone in the Rose Bay, Australia dwelling (Source: Copper 2012 p. 215).

Factors Living/kitchen/dining Outdoor temperature -0.41 Ground zone temperature -0.60 Wind direction