key: cord-0690893-eitunn6h authors: Rana, Anber; Perera, Piyaruwan; Ruparathna, Rajeev; Karunathilake, Hirushie; Hewage, Kasun; Alam, M. Shahria; Sadiq, Rehan title: Occupant-based energy upgrades selection for Canadian residential buildings based on field energy data and calibrated simulations date: 2020-06-16 journal: J Clean Prod DOI: 10.1016/j.jclepro.2020.122430 sha: b4f2666933cc2859f6403daf051e731c94e9d1aa doc_id: 690893 cord_uid: eitunn6h Occupant behavior in residential buildings has a direct impact on the effectiveness of energy-saving measures. In order to realize a buildings’ carbon mitigation targets, the impact of individual occupancy profiles needs to be integrated with building simulation models. This paper introduces a decision support framework as a potential solution to make energy performance upgrade choices based on different occupancy profiles. This framework has been demonstrated through a case study of two single-family detached homes in Canada, which were highly instrumented with sensors for monitoring energy input and output. The case studies represented two common occupancy profiles-(1) a family of four (consisting of 2 working adults and 2 teenagers); and (2) a retired couple. Firstly, calibrated energy models were developed by using one-year energy use data collected through an intrusive load monitoring technique. Secondly, energy upgrade combinations were considered for each profile and tested for additional investment, payback period and greenhouse gas (GHG) emissions. Lastly, the most suitable combination of energy upgrade for each profile was ranked using a multi-criteria decision-making method (e.g., TOPSIS). Results indicated that the retired couple used more energy than the family of four and required energy upgrades with usually higher payback periods to achieve the same level of GHG emission reduction. The results of this research are timely for energy policymaking and developing best management practices, which need to be implemented along with the deployment of more stringent building standards and codes. Detailed monitoring studies further affirm the significance of OB for obtaining realistic BES 7 results. Bahaj and James (2007) performed a comparative study on nine identical homes using 8 monitored data of electricity generation and consumption from photovoltaics. Influence of 9 occupants' awareness regarding energy use behavior, energy tariffs, subsidies and time of usage 10 of appliance were considered for generating energy load profiles. Their results show that 11 occupants' behavior is responsible for energy consumption differences which can be as high as going children) yield notable variation in energy consumption. Occupancy profiles affected all 23 three energy loads but the highest variation of 27% was observed for the heating loads. calibrated model in eQuest tool for low-income housing through 2 year utility energy use data, 10 energy auditing and indoor temperature monitoring. The results from calibrated model were used 11 to select retrofits for the housing units. They used MBE and RMSE thresholds for monthly data. 12 Same metrics were used by Rakhshan and Friess (2017) for calibrating residential buildings in 13 Dubai. They used one-year utility data and results of BES model constructed with DesignBuilder 14 tool. These and similar studies indicate necessity of having more monitored studies that show the 15 influence of occupant behavior on energy use and consequent impact on EPUs performance. 16 17 The main objective of this study is to investigate the possibility of catering to the energy demand 18 of a new residential building according to expected occupant behavior (specifically occupancy 19 profile) and compare energy performance upgrades that can reduce this energy demand. To meet 20 this objective, a decision support framework is developed that ranks the best energy performance 21 upgrades for a specific occupancy profile. The framework is demonstrated for two single-family 22 residential buildings as a case study in Canada. In particular, this paper (1) develops calibrated 23 models using one-year monitored data for predicting energy use associated with individual This study provides an important opportunity to advance the understanding of the impacts of 1 occupancy profiles through the use of detailed monitored energy use data that can help make 2 informed decisions for energy improvements in residential buildings. The relevant results 3 obtained from this study will be useful for proposing suitable EPUs in homes with similar 4 occupancy profiles. This framework can facilitate informed decision making for the home-5 owners, home-occupiers, and developers. In particular, the energy upgrades that this study 6 recommends for age groups and the size of a household will be useful for policymakers and 7 utility companies in improving energy conservation policies and associated financial incentives. 8 The framework also provides flexibility to choose between pro-environmental and pro-economic 9 choices of energy upgrades with different occupancy profiles, therefore, accounting for 10 perspectives of different types of investors. In this study, a framework is developed to identify the most appropriate energy performance 13 upgrades for new residential buildings with expected occupancy profiles. The framework shown 14 in Fig. 1 has two main phases: (1) Data Collection and (2) Data analysis. 15 2.1.Data collection and processing 16 The two key sources of data include: (1) literature-based data that was used to identify potential 17 energy upgrades; and (2) data captured from the field monitoring system to create calibrated 18 energy models. 19 The main sources used for literature were peer-reviewed articles and technical reports related to 20 residential building energy and cost modelling. A thorough literature review was performed for 21 current building energy standards for residential buildings, energy and cost modelling methods, 22 energy monitoring techniques, occupancy profiles, and energy performance upgrades possible 23 for residential buildings. Energy performance upgrades can be categorized into three categories: (1) passive, (2) active and (3) renewable energy upgrades. Passive energy upgrades make use of and D) to help in energy conservation projects (Webster et al., 2015) . Among the four 8 methodologies Option D, is suitable for assessing energy savings for whole building through 9 calibration process. However, the assessment of energy savings require monitored data before 10 and after application of the EPUs in building. In absence, of complete data only calibration 11 process can be performed. MBE is a non-dimensional bias measure (i.e., sum of errors), between measured and simulated 17 data for each hour. It captures the mean difference between measured and simulated data points 18 and is considered a good indicator of the overall bias in the model. RMSE index assesses data by 19 capturing offsetting errors between measured and simulated data and does not suffer from the 20 cancellation effect that can come in MBE where the positive bias compensates the negative bias 21 and can result in incorrect validation. Acceptable tolerance limits for the two benchmarks vary 22 according to the type of calibration hourly or monthly and are provided in Appendix A, Table 23 A1. The following steps are involved in performing the calibration process: local weather data, occupancy, equipment specifications, and indoor temperatures. 2) Adjusting constructed BES model with data obtained from monitoring and making 1 suitable assumptions for unknown parameters. (1-b) 6 Where, m k and s k are the respective measured and simulated data points for each model . Calibrated energy models were tested for variation in energy consumption due to changes in 13 energy upgrades. The energy consumptions under upgrades combinations were used to find the 14 change in the carbon footprint of the house and operational costs. 1 For this study, the Net Present Value (NPV) method was used to determine the lifecycle cost 2 analysis (LCC) for assessing building upgrades. LCC is an economic evaluation technique that 3 determines the costs of the entire life span of a product or process (Warren, 1994 and payback period (Years). Weighting matrix is generated by defining five weighting 5 schemes (Table 1) representative of investors' preference for pro-environmental or pro-6 economic investments. In pro-environmental option, more weight is given to the 7 reduction of GHG emissions while for a pro-economic option lowest LCC is given 8 maximum weight. ; for i= 1, n to j=1, m (4) 16 Where, "z -. " is an entry in the decision matrix and m represents the total number of rows 17 and n represents the total number of columns. The distance of alternative from the positive ideal was found by using Eq. (5-a) and for 4 negative ideal using Eq. (5-b). (5-b) 7 The relative closeness to the ideal solution was found by Eq. (5-c) M (5-c) 9 6) Ranking the alternatives based on relative closeness to the ideal solution. A case study approach is taken to apply and demonstrate the framework using field data obtained Okanagan region construction practices. The two houses are located in one of the fastest growing areas in Canada, the Okanagan Valley. 19 The case study homes were constructed in 2016 and occupied in 2017. These mid-size residences 20 are identical in their architectural design, location, and orientation and are exposed to same external features. However, the two homes vary with respect to the building envelope, heating 1 and cooling systems, and occupancies. The household characteristics for the homes are provided 2 in Table 2 . The construction materials and thermal characteristics obtained from the industrial 3 partners were used to develop the energy model. The geometry and other relevant information of 4 this house were extracted from the relevant drawings and bills of quantities. The thermal 5 characteristics of specific envelope components and equipment such as HVAC, lighting, and 6 appliances were collected from the manufacturers' specifications. forms-ICF blocks), better performance windows (Vinyl triple glazed windows c/w 366 Low-E), 20 an HVAC system consisting of a geothermal heat source pump and energy star rated appliances 21 the ATH is supplied with renewable energy from a solar (PV) system installed on the roof. representation of occupants behavior and savings from energy upgrades. In order to make a 9 comparison between the influence of two occupancy profiles on energy and GHG reduction 10 potential of various EPSs, STH was upgraded with energy upgrades present for ATH while ATH 11 was downgraded with the standard materials, equipment and appliances present in STH. The calibrated models were then run for different EPU combinations to determine the variation 10 in annual energy use. A total of 514 simulation runs were performed on the nine EPU options 11 (present in Table 2 ). In order to increase the simulation time interaction impact between EPUs upgrades. Costs relating to home insurance, landscaping, furniture, mortgage payment, government incentives for green upgrades were not considered. in Table 6 . The discount rate of 3% and an inflation rate of 5% were assumed based on the In order to rank the energy upgrade for each profile decision matrix composed of initial cost 19 investment, the payback period and GHG emissions were constructed. Fig. 8 and Table 7 show 20 results of the first ten highest ranked energy upgrades. It is seen for among the given scenarios family of four is likely to spend more hot water in showering and bathing and hence are able to 1 get more benefits by using a more efficient water heating system. of GHG reduction targets. Highest reduction potential for Occupancy Profile 2 for the case study 26 was found for the combination "RF WL LED HVAC DHW PV". It is observed that for this 27 profile higher investment of up to CA$ 11,000 is not significantly decreasing GHG emissions. Therefore, for Occupancy Profile 2 in order to achieve high reduction targets needed to achieve sustainable buildings will require external help from the government and other organizations. 1 This is also significant since the majority of the studies have shown senior residents often own 2 bigger houses and usually have a low-income source (Yohanis et al., 2008) . Interestingly, the WN upgrade forms part of the top ten choices of energy upgrades. These 22 findings imply that a larger expenditure is needed for achieving meaningful GHG reduction. Some studies have also shown that the older occupants belong to the low-income group. This In current framework, one-year monitored data was used for calibrating the energy models. In 21 the absence of monitored data, energy models can be calibrated using house characteristics, annual data on monthly utility bills coupled with a questionnaire survey for more accuracy. When only utility data is used for calibration at least three-year monthly utility data is 24 recommended to be used to ensure average energy use per month is correct (Hubler et al., 2010) . The results of calibration from either monitored data or utility bills can be reasonably 26 representative of the occupancy profile energy use. However, model generated using utility data 27 is unable to predict performance accuracy for individual systems and equipment. It is therefore 28 recommended to assess the occupant behavior, occupancy times on weekdays and weekends, indoor temperature settings in order to remove possible errors. Hence, the proposed framework 1 can be modified to include energy models calibrated through a survey of occupants and the 2 monthly energy bills to determine suitable EPUs. This study did not find a significant difference between the energy upgrade for the pro-economic 5 preference for the two profiles. However, changes in both ranking order and the type and number 6 of energy upgrades vary with the pro-environment preference. This finding has important 7 implications for developing tailored energy audit and energy saving incentives programs. Therefore, the GHG emission reduction potential for energy upgrades will be higher for regions 5 with energy generation based on fossil fuels. Another factor effecting ranking of energy upgrades conditions which also influence energy use and hence, the impact of energy upgrades. 8 As the next phase of this study, data collected through the load monitoring method will be 9 further analyzed to extract the occupancy patterns for various daily and weekly activities (e.g. models. Calibrated models were tested impacts of EPUs on GHG emissions and payback period. The results showed that a single-family detached house family of four is capable of reducing 6 GHG emissions at a higher rate as compared to the application of the same upgrades in the house 7 of a retired couple. Prioritization for family of four showed that for a pro-environment preference 8 the best option was a combination of "RF WL HVAC DHW PV" while for pro-economic the 9 best option was efficient lighting upgrades. For the retired couple, the best option for pro- Considering user 11 profiles and occupants' behaviour on a zero energy renovation strategy for multi-family Analysis of Measurement and Verification Methods for Energy Retrofits Applied to demand of the energy efficient house Multi-unit residential building energy and peak demand study A comparative study of passive solar building simulation using HOT2000 Canada's Renewable Power Landscape 2016 -Energy Market Analysis Directory of Energy Efficiency and Alternative Energy Programs in Canada. National Energy Use 20 Database Generation of ambient temperature bin data of 26 cities in 36 Scenario-based economic and environmental 38 analysis of clean energy incentives for households in Canada: Multi criteria decision making approach Domestic electricity use: A high-resolution energy 7 demand model A Survey on Intrusive Load Monitoring for Appliance Recognition A unified probabilistic model 12 for predicting occupancy, domestic hot water use and electricity use in residential buildings Rethinking investment planning and optimizing net zero emission 20 buildings Summary for Decision-Makers 1-62 M&V guidelines: measurement and verification for federal energy projects version 3.0 Multi-criteria building energy performance 9 benchmarking through variable clustering based compromise TOPSIS with objective entropy weighting Development of an integrated life-cycle cost assessment model The impacts of household retrofit and domestic energy efficiency M&V Guidelines: Measurement and Verification for Performance-Based Contracts_Version 18 4 IEA EBC Annex 66: Definition 20 and simulation of occupant behavior in buildings a Appliances include kitchen (dishwasher, exhaust hood, stove and oven) and laundry appliances (dryer and clothes washer) b HVAC systems upgrade was considered to be ground source heat pump that serves part of both heating and cooling systems of the house and involves use of electricity for running pumps and renewable energy c Maintenance and replacement of Wall-insulation, Ceiling insulation and ICF foundation is beyond the study period of 30-year • A scenario-based framework was developed for residential homes energy upgrades.• It considers occupant behaviour and preference in selecting energy upgrades.• Field data is used to represent energy use of a family of four and a stay-at-home retired couple.• Energy upgrades are same under pro-economic choices for two profiles.• Energy upgrades vary under pro-environmental scenarios for two profiles. ☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: