As Canada strives to curtail its greenhouse gas (GHG) emissions in line with the Paris Agreement, reductions in the building sector are necessary to avoid more dire degrees of global warming (Environment and Climate Change Canada 2019). This includes updating building codes to include net-zero energy ready mandates for new buildings by 2030 and creating building retrofit codes (Environment and Climate Change Canada 2017). The province of British Columbia (BC) has begun to adopt these approaches in performance-based codes for new construction and plans for retrofit-related action through a variety of policy instruments (Frappé-Sénéclauze et al. 2017).
In Vancouver, BC, a geographically constrained, transit-oriented, and rapidly densifying city, energy needs are to be provided by renewable resources before 2050 and GHG emissions reduced by 80% of 2007 levels (City of Vancouver 2015). To achieve these goals, the city has set several priorities for the building sector, which accounts for 55% of the city’s GHG emissions, 31% of which are from single-detached homes (SDHs) (Sustainability Group, City of Vancouver 2014). New construction in the city has been given a mandate for a zero-emission standard by 2030 through passive high-performance building techniques and electrification. However, in BC, new construction standards are expected to account for only 30% of the province’s proposed emissions reduction targets in the building sector (Frappé-Sénéclauze et al. 2017). A comprehensive retrofit strategy is necessary for decarbonization of the entire sector.
A wide variety of economic and non-economic drivers and barriers to the retrofit of SDHs have been documented globally (Achtnicht & Madlener 2014; Jakob 2007; Nair et al. 2010; Wilson et al. 2011). In North America there is recognition that deep retrofits of SDH building stocks remain difficult to incentivize (Palmer et al. 2013; Vergragt & Brown 2020; Wilson et al. 2015). Many of these retrofit studies, or similar studies that evaluate retrofit viability within a regional context, have relied on bottom-up energy and economic modelling of building performance and/or life-cycle costs. For example, in 2017, the City of Vancouver (CoV) contracted a local consultancy to provide scenario-modelling of technologies and energy reduction measures for carbon emissions mitigation through an energy–economy modelling tool developed previously by Rivers & Jaccard (2006) with the conclusion that it is feasible for 90% of the city’s building stock (by floor area) to achieve net zero emissions by 2050, with 30–40% (or 20–30 million m2 of floor space) being retrofitted buildings (City of Vancouver 2017).
These types of studies are important but may not be affordable or quickly executed and may fail to encapsulate the nuances of the local real estate market. It can be difficult, or costly, to methodologically reconcile complex building energy simulation models with equally comprehensive system models of a building stock’s overall economic behavior. The present study uses a low-cost method to obtain an initial, if not complete, assessment of retrofit viability based on an expert elicitation survey of local experts familiar with the local retrofit market, housing market, and building performance codes. Expert elicitation was used to develop knowledge about a variety of questions in a range of fields from risk assessment to species habitat modelling (Martin et al. 2005; Szwed et al. 2006). Its use as a methodology provides a structured and independent format for field experts to impart data on or guidance for the task at hand (Brown 1968).
In this study, a survey for expert elicitation was developed to evaluate the likelihood that representative SDHs may be retrofitted in the CoV. This study is used to extract answers to policy-relevant questions that would be equally sought after with a modelling study.
The paper is structured as follows. Next, the state of knowledge is reviewed with regard to barriers and drivers facing SDH retrofits in Vancouver, as well as internationally. Prior methods of expert elicitation are then considered. The following section presents the methodology for the current survey. Finally, the results are discussed and the outcomes presented.
Vancouver is a city of 630,000 residents located within Metro Vancouver (population 2.4 million), the largest metropolitan region of BC (population 4.6 million). Residential electricity is supplied by BC Hydro, from which the vast majority of annual electricity demand is met via hydroelectricity; considered near zero carbon, with a GHG emissions intensity of approximately 0.01 kg-CO2e/kWh (BC Ministry of Environment and Climate Change Strategy 2019). SDHs, freestanding residential buildings that may house one or several tenants, comprise 81% of Vancouver’s residential land area and are occupied by 35% of its households (Bergmann 2016). A total of 81% of SDHs use natural gas for space heating; the remaining 19% use electricity (Dahmen et al. 2018). A similar division of fuel share exists for domestic hot water. Space cooling currently remains rare among SDHs, accounting for less than 0.4% of annual energy use across Vancouver’s SDH stock. At the time of writing, retail electricity rates for residential consumers range from C$0.09 to C$0.14/kWh; retail residential rates for natural gas are approximately C$0.04/kWh (BC Hydro 2020).
In 2016, a forum of industry experts in BC drew the following statements with respect to the barriers and drivers facing retrofits of BC’s SDHs (Frappé-Sénéclauze et al. 2017):
Dahmen et al. (2018) identified the high land-to-building value ratio of the Vancouver real estate market as a barrier to retrofits. This incentivizes demolition and new construction, often with similar densities to what it replaces. This is a factor in the question of accelerating retrofit uptake that may not be unique to Vancouver and is perhaps understudied with respect to retrofit modelling.
In prior global reviews of retrofit barriers and drivers, commonly identified drivers include potential savings, thermal discomfort, values, high energy costs, and other renovation plans (Achtnicht & Madlener 2014; Jakob 2007; Nair et al. 2010; Wilson et al. 2011). Homeowners perceive financial constraints, a lack of meaningful cost savings, and a lack of necessity as barriers to retrofitting (Achtnicht & Madlener 2014; Hrovatin & Zorić 2018; Jakob 2007). The relationship between household characteristics and retrofit implementation has been extensively studied with inconclusive results. Kastner & Stern (2015) examined a variety of studies on households’ decisions to invest in domestic retrofits, grouping commonly studied household characteristics by demographics, housing characteristics, and decisionmaker dispositions. Of these, household size, duration of stay, income, and age distribution are among the most frequently included. Households with more members are less likely to retrofit (Gamtessa 2013; Wilson et al. 2011). Conversely, households who have lived in their home for longer are more likely to retrofit (Trotta 2018; Wilson et al. 2011). A variety of effects have been found for household income; some studies find households with higher incomes are more likely to retrofit, possibly reflecting their increased ability to pay (Achtnicht & Madlener 2014; Nair et al. 2010), while some find that household income has no effect (Hrovatin & Zorić 2018; Trotta 2018). Gamtessa (2013) found that income has a negative effect in their study of Canadian households, postulating that households with higher incomes are not as motivated by monetary savings from retrofitting. The effect of household age distribution is similarly uncertain; on the one hand, older households may have a shorter payback period to realize savings (Achtnicht & Madlener 2014; Hrovatin & Zorić 2018; Nair et al. 2010), but, on the other, older households may be free of time and financial constraints that would otherwise prevent retrofitting (Gamtessa 2013; Wilson et al. 2011). The uncertainty of the effect of household characteristics could be due to the range of countries studied, implying that the structural context plays an important role in retrofit decision-making. Indeed, several structural uncertainties, such as government policy change, climate change, and services change, affect retrofit activity (Ma et al. 2012). Throughout this literature review, succinctly presented in Table 1, real estate market forces are not directly cited as a key barrier to retrofits en masse, even the BC-based study.
|Achtnicht & Madlener (2014)||Germany||2009||Survey with a choice experiment||Retrofits are more likely when affordable, profitable, and favorable over existing conditions|
|Alberini et al. (2013)||Switzerland||2010||Survey with a choice experiment||Homeowners are responsive to upfront costs, rate of return, and expected thermal comfort improvement|
|Banfi et al. (2008)||Switzerland||2008||Choice experiment||Consumers are under-informed of the advantages of efficiency measures and not equipped to understand their economic implications|
|Haines & Mitchell (2014)||England, UK||2014||Persona development and expert elicitation||Owner-occupiers are best understood not as one homogeneous group|
|Hrovatin & Zorić (2018)||Slovenia||2010||Survey with a questionnaire||Retrofit decisions are linked to general improvements and quality-of-life considerations. The higher age and loan repayment timeline may inhibit action|
|Gamtessa (2013)||Canada||1998–2005||Analysis of home audit reports||Financial incentives play an important role; higher savings and rebates lead to a higher likelihood of action|
|Jakob (2007)||Switzerland||1986–2000||Survey with a questionnaire||Envelope renovation is triggered by general end-of-life renovation action|
|Judson & Maller (2014)||Victoria, Australia||2008–09, 2011||In-person interview and property inspection||Current and anticipated everyday activities determine homeowner action|
|Nair et al. (2010)||Sweden||2006–08||Survey with a questionnaire||Higher perceived cost of energy leads to a higher likelihood of action|
|Frappé-Sénéclauze et al. (2017)||British Columbia, Canada||2017||Summary of an expert forum||Utility programs in the absence of policy will not scale as desired. Economics and market inertia limit action|
|Tjørring & Gausset (2019)||Sonderborg and Middlefart, Denmark||2012–15||Participant observation with interviews||Home renovations should be viewed as investments in social relations|
|Trotta (2018)||England, UK||2011–14||In-person interview and property inspection||Dwelling characteristics are more important that sociodemographic characteristics|
|Wilson et al. (2011)||UK||2011||Survey with a questionnaire||Efficiency actions are more commonly done together with general improvements|
In cases where there are little data with which to evaluate or model the nature of a problem, particularly one relevant for policy decision-making, expert elicitation is one of several methods through which information can be gathered. Expert elicitation methodologies began to take shape in the 1960s and 1970s, notably with the RAND Corporation’s Delphi Method, which was a formalization of using expert judgment, traditionally developed in structured group discussions, for decision-making on matters of national security in the US (Brown 1968). Since then, these methodologies have been refined by thinking on cognitive heuristics and biases, particularly by Tversky & Kahneman (1974) and Kahneman & Tversky (1982), with the goal of contemporary expert elicitation to accurately represent experts’ knowledge and beliefs as a probability distribution for use in statistical modelling (Garthwaite et al. 2005).
A robust literature exists on the appropriate use of expert elicitation methods for obtaining subjective knowledge of a problem as well (Garthwaite et al. 2005; Kynn 2008; Morgan 2014). It has been used extensively in examining the future costs and performance of alternative energy generation technologies (Anadón et al. 2012; Baker et al. 2009; Bosetti et al. 2012; Wiser et al. 2016). In ecology, expert knowledge is used to inform ecological models in the absence of data, particularly in understanding the relationship between habitat conditions and the presence of certain species (Martin et al. 2005). Under highly uncertain future conditions, experts have been called upon to quantify the degree of uncertainty itself (Usher & Strachan 2013). Issues of data availability are especially salient in studies of future conditions when continuity with past events cannot be assumed, such as in risk assessment to understand the factors that influence unlikely but high impact events, such as shipping accidents (Szwed et al. 2006).
Expert elicitation has been used in climate change research to characterize uncertainty or improve modelling efforts (Morgan & Keith 1995; Dessai et al. 2018). Usher & Strachan (2013) conducted an elicitation of 25 experts to improve the robustness of uncertainty within integrated assessment models, stating that the formal approach from elicitation is similar to how modelers account for uncertainties in their models, but more transparent and robust due to a larger collection of statements. That is, in modelling climate change and energy and economic development modelers are expected to make judgements in place of data gaps, but this process is better supplemented through expert elicitation.
Several studies undertake expert elicitation with the goal of understanding retrofit drivers or modelling them. Addy et al. (2014) conducted principal component analysis on a questionnaire dataset received from 57 local architects when studying the perceived barriers and drivers to retrofits in Ghana. The questionnaire asked the architects to rate the importance of 18 barriers to retrofit identified in the literature. Their analysis was able to reduce the prior understanding of 18 barriers down to an understanding of five generalized barriers as the most important to accelerating retrofit uptake in the local region.
Duah et al. (2014) engaged 19 experts in a structured elicitation process to develop a computational retrofit decision-making tool. The experts were identified through industry knowledge, certifications, experience, professional ethics, and computer knowledge. A consensus-based method was used to formalize elicited knowledge from the experts through repeated discussion of the topic of whether a hypothetical home would be retrofitted. These consensus viewpoints were then used to develop a control-flow logic (e.g. if a dwelling has incandescent lighting, perform a lighting efficiency upgrade) that could be applied in a numerical modelling environment.
Elicitation can be conducted through a variety of methods, including interviews, surveys, or questionnaires, the choice of which is context dependent (Shadbolt 2005). In-person elicitation allows researchers to develop a relationship with the expert(s) to mitigate cognitive biases in their thinking and to ensure relevant information is not overlooked. Group elicitation efforts are also commonly used to arrive at a consensus, although this has been found to be the fixture of pressure for group conformity and a lack of anonymity and can lead to researchers discounting the importance of outliers (Woudenberg 1991; Mirzaee et al. 2019). In a study of several expert elicitation efforts for assessing the future cost of photovoltaics, Verdolini et al. (2015) found that an online survey can avoid many of the same biases while reaching a wider variety of experts more quickly, provided the survey is well-designed and offers the same access to dwelling and owner context available during in-person structured interviews or group forums.
Haines & Mitchell (2014) developed personas of homeowners to characterize a wide range of people’s motivation and preferences for estimating retrofit likelihood. They note that combining qualitative data about homeowners with quantitative market segmentation data can further enhance the personas and make them more useful for evaluating future policy interventions to support retrofits. Jakob (2007) noted the importance of the characteristics of the dwelling to informing retrofit potential. Bold (2012) describes a set of narrative-based research tools and methods that can synthesize multiple data streams to create coherent representations of real-world circumstances. Combined, these approaches may offer a way to contextualize dwellings in a succinct but rich manner to enhance on online elicitation effort.
At present, the research landscape of the available approaches can be broadly categorized into two types of methods: physics-based engineering simulation of building energy use; and econometric modelling of building owner/occupant behavior and retrofit investment decisions. Several recent retrofit modelling studies that have focused on BC’s building stock using these approaches are Feng et al. (2020), Prabatha et al. (2020), and Salter et al. (2020). No survey-based assessment of building retrofits in Vancouver or BC are known.
An expert elicitation survey is used to establish a relationship between owner-occupier metrics, building characteristics, regional economic conditions, and the likelihood of energy-efficient and/or carbon-reduction retrofits occurring across Vancouver’s SDHs in the future.
The underlying objective of the survey is to ask an expert to make an initial assessment of the retrofit viability of a synthetic household. A synthetic household comprises an owner, occupant(s), a location, a design, the existing physical characteristics of the home, and census-derived socioeconomic characteristics of the owner/occupants. The household is synthetic because it is not a real-world household with a fixed address, but an archetypal household parametrized from a sample of known statistical data and expert judgement. For this study 30 local experts were asked to review three of the 30 synthetic households that were created to represent the overarching characteristics of Vancouver’s SDH housing stock and owner/occupiers.
A synthetic household is presented to an expert in the form of a ‘representative construction.’ This is a fictional account, or narrative, of a dwelling and owner-occupier persona. Bold (2012) introduces and describes representative constructions at length, and the present paper draws heavily on Bold’s approach as it can combine many different data sources into an engaging format for experts to respond to. The prior work of Haines & Mitchell (2014) was also an influence on the present study. Their persona-based approach provides a more holistic representation of retrofit decision-maker(s) for the purpose of informing retrofit policies.
The representative constructions are written to replicate the outcome of a typical energy-auditing exercise of an SDH. On reading a representative construction, it is intended that the survey respondent would know, to the best extent possible, the following:
To create each representative construction, methods of segmentation and characterization were employed. This is a common technique applied in the modelling of building stocks (Mata et al. 2014; Pasichnyi et al. 2019). Segmentation employs high-level criteria to divide the building stock into mutually exclusive, but representative of the whole groups. In characterization each group is described through representative parameters.
Each representative construction comprises an amalgamation of a representative building typology, the representative physical characteristics of the dwelling, and a representative persona of the owner/occupants. Section 2.1 details the datasets used for the development of the representative constructions. Sections 2.2–2.4 describe, respectively, how each representative construction was formulated from the assembled building typologies, physical characteristics, and personas. Section 2.5 describes how the building typologies and personas were assembled. Section 2.6 describes the process of selecting experts. Section 2.7 covers the distribution of the survey and questions asked. Section 2.8 describes the methods used in a correlation analysis of the responses and features of the representative constructions.
The survey addresses several questions that might be expected to be evaluated in an early-stage energy and economic modelling study of Vancouver SDH’s building stock:
There is no available spatial dataset providing the physical and socioeconomic characteristics of individual SDHs in Vancouver or BC. However, three datasets are relevant to this study: Statistics Canada’s 2016 Census Metropolitan Area (CMA) for Metro Vancouver (Statistics Canada 2017); the British Columbia Assessment (BCA) parcel inventory (British Columbia Assessment Authority 2017); and The Vancouver Heritage Foundation’s ‘House Styles Hub’ (The Vancouver Heritage Foundation 2018).
The CMA dataset is a sample of 1% of the private households in private dwellings in Canada, reporting 95 sociodemographic variables, and drawn from a questionnaire issued to 25% of private households in the country. The smallest geographic identifiers are major metropolitan regions across the country, such as Metro Vancouver.
The BCA dataset contains data points related to all buildings in BC, which can be subdivided by region and municipality, the boundary of the CoV BCA. In the BCA dataset, buildings are classified by type and contain, among many, features for occupancy, year constructed, gross floor area, and number of stories.
The Vancouver Heritage Foundation (2018) maintains an ‘encyclopedia’ of common dwelling types built in Vancouver since colonization, broadly characterized by age and architectural style. The BCA dataset does not provide a dwelling-type identity (i.e. style) to each parcel of land, but does provide the age of the residential property sitting on it. Both datasets were assessed for this study and five building typologies that appear to represent the majority of Vancouver’s SDH stock were extracted. Using terminology from the encyclopedia, four dwelling types were chosen that do not overlap in their typical ‘era of construction,’ covering the period 1910–2010. A fifth dwelling that represents the present era of construction in the SDH market was selected from real estate listings. These five building types each form one of the representative building typologies. Figure 1 shows the representative typologies selected from encyclopedia and local real estate listings that were used in the survey.
Dwellings built between 1910 and 1930 are considered heritage homes, typified by the Craftsman (CR) style. In Vancouver, homes can be added to an official, municipally managed heritage register upon review. The Craftsman homes and their variations are the most common home style in the register (The Vancouver Heritage Foundation 2018). The period around the Great Depression, considered here as 1930–35, was omitted due to a decline in housing starts. Housing built after the Great Depression through the post-war period, covering the period 1935–64, is represented by the single-storey, boxy home known as the Mid-Century Builder (MC), which was erected rapidly with little variation across the city. Homes constructed between 1965 and 1985 were commonly built from another builder’s template known as the Vancouver Special (VS), a two-storey home with a shallow-pitched roof. From 1986 through the last decade, larger homes with irregular facades and floorplans are common; these are termed the Millennium Builder (MB). Homes built from 2010 to the present are characterized by an improved baseline level of performance due to more stringent energy performance standards in the building code; they are termed here the West Coast Modern (WC).
Each typology was assigned to a parcel of land with an existing SDH that matches its estimated value derived the BCA dataset. This allowed the typology to be assigned a general neighborhood within Vancouver for the purposes of the enriching the narrative. Neighborhoods with distinct characteristics, particularly in terms of home prices and land values, were selected to represent the widest range of possible conditions.
Each of the five typologies is characterized by the 14 parameters shown in Table 2 with their accompanying text and numerical values. Text values were used to construct the written narratives, while the numerical values were used in the data analysis. For Vancouver, the only resource to ascribe age-relevant parameters to the typologies came through the building code review. By nature of their age and architectural traits, each of the SDH typologies implies a range of physical parameters, such as shell components and the efficiency of installed mechanical systems, that impact the energy efficiency of a building, parameters that are typically viewed to influence the energy consumption and GHG emissions of a building (Lam & Hui 1996), as well as having replacement timelines that are associated with retrofits, such as in the past with coal-fed furnaces being retired for natural gas.
|PARAMETER||TEXT VALUES||ENCODED VALUES|
|Construction era||[1910–30, 1935–64, 1965–84, 1985–2009, 2010–present]||[5, 4, 3, 2, 1]|
|House size (ft2)||[2100, 2600, 2800, 3000, 5400]||[1, 2, 3, 4, 5]|
|Lot size (ft2)||[3135, 4000, 8400]||[1, 2, 3]|
|Energy Step Code||[Step1, Step2, Step3, Step4, Step5]||[1, 2, 3, 4, 5]|
|Secondary suite||[No, Yes]||[0, 1]|
|Wall construction||[2 × 4 batt R13, 2 × 6 exterior R21]||[1, 2]|
|Wall age (years)||[0–5, 6–10, 11–15, 16–24, ≥ 25]||[1, 2, 3, 4, 5]|
|Roof construction||[blown R15, flat ext R36]||[1, 2]|
|Roof age (years)||[0–5, 6–10, 11–15, 16–24, ≥ 25]||[1, 2, 3, 4, 5]|
|Window type||[Single U = 0.48, Double U = 0.48, Double low-E U = 0.25]||[1, 2, 3]|
|Window age (years)||[0–5, 6–10, 11–15, 16–24, ≥ 25]||[1, 2, 3, 4, 5]|
|Space heating efficiency||[0.75, 0.85, 2.70]||[1, 2, 3]|
|Space heating age (years)||[0–5, 6–10, 11–15, 16–24, ≥ 25]||[1, 2, 3, 4, 5]|
|Space heating fuel||[Natural gas, Electric, Natural gas and electric]||[1, 2, 3]|
|Domestic hot water efficiency||[0.75, 0.85, 0.95]||[1, 2, 3]|
|Domestic hot water age (years)||[0–5, 6–10, 11–15, 16–24, ≥ 25]||[1, 2, 3, 4, 5]|
|Domestic hot water fuel||[Natural gas, Electric, Natural gas and electric]||[1, 2, 3]|
|Space cooling efficiency||[None, 2.7]||[0, 1]|
|Space cooling age (years)||[None, 1]||[0, 1]|
|Space cooling fuel||[None, Electric]||[0, 1]|
|Airtightness||[Very leaky, Leaky, Middle, Tight]||[1, 2, 3, 4]|
|Maintenance level||[Never, Poor, Average, Well maintained, Very well]||[1, 2, 3, 4, 5]|
|Presence of asbestos||[No, Yes]||[0, 1]|
Representative personas were developed to provide a more holistic model of the retrofit decision-maker(s), as suggested by Haines & Mitchell (2014). Each representative persona constructed for this study is indicative of sampled social and economic traits from census-derived information. The parameters and their text and numerical values are shown in Table 3. The CMA data were used as a primary data source, where a subset of CMA data was created for owner-occupied SDHs and segmented by household structure groups: ‘couple,’ ‘couple with children,’ ‘lone-parent family,’ ‘person living alone,’ and ‘other.’ Further characterization was done using eight additional parameters, determined to be relevant through the literature review.
|PARAMETER||TEXT VALUES||ENCODED VALUES|
|Homeowner age||[25–34, 35–44, 45–54, 55–64, 65–74, ≥ 75]||[1, 2, 3, 4, 5, 6]|
|Family type||[Couple, Couple + kids, Couple + kids + other, Single, Single + kids]||[1, 2, 3, 4, 5]|
|Years lived in home||[0–7, 8–14, 15–21, 22–28, ≥ 29]||[1, 2, 3, 4, 5]|
|Spending habits||[Frugal, Thrifty, Generous]||[1, 2, 3]|
|Mortgage percent of income||[0–8, 9–16, 17–24, 25–32, ≥ 33]||[1, 2, 3, 4, 5]|
|Environmental views||[None, Some, Intense]||[1, 2, 3]|
|Income level||[Below average, Average, Above Average, Retired]||[1, 2, 3, 4]|
|Credit rating||[Good, Fair, Excellent]||[1, 2, 3]|
|Plans for renovation||[No, Yes]||[0, 1]|
‘Homeowner age,’ ‘income level,’ and ‘mortgage percent of income’ were all assigned following distribution in the CMA data. Due to lack of open data within Vancouver and BC, values for ‘credit rating,’ ‘environmental views,’ ‘spending habits,’ ‘duration of stay,’ and ‘plans for renovation’ were assigned following the process for making each representative construction described in the following section. Each was found through the literature review to have a potential impact on retrofit decision-making (Jakob 2007; Trotta 2018). Of particular importance, environmental concern and pro-environmental behaviors were linked to investment in retrofits, but were also found to be moderated by social norms and high investment costs in some cases (Trotta 2018).
Race and nationality were not factored into the data-collection process, and therefore the representative constructions used here do not reference or imply the race or nationality of a fictional household. By nature of giving names to household members—implying gender, and through the implied gender of a familial couple, implying sexual orientation. Thus, though the representative constructions here ultimately included heterosexual and same-sex couples, the distribution and make-up of gender and sexual orientation across the constructions were not rooted in the data. This is accepted as a gap in this study.
An iterative creative writing process, rooted in the continuous evaluation of work in-progress narratives against known data, was used to generate representative constructions from the established set of representative typologies, physical characteristics, and personas. Six representative constructions were written for each of the five representative typologies, each with a more or less perceived optimistic outlook on retrofit viability. Optimism was made more positive or negative through the association of parameters that were identified by the literature review and referenced in Table 1 as being associated with a higher or lower general likelihood of retrofit, such as holding feelings of environmental responsibility (Jakob 2007).
For the 30 representative constructions in their full narrative form, see the supplemental data online. As well, Tables S1–S3 (online) describe which demographic and building traits were assigned to each of the representative constructions. Adjudication of the overall relevancy and validity of the representative constructions came from ex-post analysis of the assignment of traits to each representative construction against CMA and BCA data. Table 4 provides a view of this analysis and indicates that the 30 representative constructions are consistent with, at least, 50% of Vancouver’s SDH stock.
|AGE SEGMENT (YEARS)||YOUNG (25–44 YEARS)||MIDDLE-AGED (45–64 YEARS)||OLDER (≥ 65 YEARS)||OTHER|
In the present study, the viewpoints of both Duah et al. (2014) and Haines & Mitchell (2014) are accounted for in the development of the representative constructions which provide the selected experts with a wide range of parameters necessary for making informed retrofit recommendations while also describing a sample of owner occupiers in a non-homogeneous manner.
As researchers in the building science profession in Vancouver, the authors are familiar with the region’s professional networks of practitioners and policymakers with experience undertaking building retrofit assessments, developing housing policy, and/or energy efficiency policy. This includes, but is not exclusive to, members of the International Building Performance Simulation Association (IBPSA) BC Chapter, members of the BC Energy Step Code Council (British Columbia 2019), and members of the University of British Columbia’s Steering Committee for its Green Building Action Plan.
A total of 60 prospective experts were identified in the early stages of this study. Of the prospective experts, three were selected to provide a trial run of the survey and provide feedback to the study authors. The testers included a university faculty member of high-performance building design, an energy manager for the University of British Columbia, and a municipal planner working with the CoV. Their feedback was incorporated into the final version of the survey, but their responses were excluded from this paper.
A total of 30 unique surveys were generated, each comprising a random selection of three representative constructions, and surveys were issued to a total of 38 known experts. A total of 22 completed surveys were returned, with a total response rate of 58%.
The experts were not human test subjects, and the individual characteristics and behavior of the experts was not evaluated as this was not a behavioral study of the experts. The purpose of the study was be to solicit the expert view of respondents on matters known to them. Ethics board approval for this project was not required by the authors’ institution.
In each survey a random sample of three representative constructions, each followed by the questions, was provided. The total estimated survey completion time was 20 minutes.
Surveys were distributed to experts using an anonymous link beginning in August 2019 and ending in December 2019. Table 5 describes the final count (n = 56) of each received, completed representative construction after filtering for completeness and expert quality. The surveys were created and distributed using Qualtrics, for which the host university of the study authors holds an institutional license (Qualtrics 2019).
|CRAFTSMAN (CR)||MID-CENTURY BUILDER (MC)||VANCOUVER SPECIAL (VS)||MILLENNIUM BUILDER (MB)||WEST COAST MODERN (WC)|
Upon beginning a survey, experts were asked to indicate their field of work, years of experience, and to rate on a five-point Likert scale their familiarity with the BC Energy Step Code, described below, and the Vancouver real estate market.
After reviewing each representative construction, experts were asked to respond to several questions through a Likert-based assessment of retrofit viability. The questions were asked relative to a common performance-based energy efficiency metric in BC, the Energy Step Code, which is currently in place as an optional code framework for new construction. The expected performance levels of each of the five steps are shown in Table 6 as they coordinate to expected airtightness, energy-use intensity, and improvement over a baseline building (British Columbia 2019). Currently, building energy-efficiency discussions in BC are driven by the Energy Step Code, but an equivalent code has yet to be adopted for existing buildings. As the Energy Step Code is not explicitly considerate of carbon emissions, the survey poses two questions that ask:
|STEP CODE||ACH AT 50 PASCALS||MEUI (kWh/m2/yr)||% REF||TEDI (kWh/m2/yr)|
Experts were then asked to write which parameters were most influential in their decision-making process. After considering the likelihood of retrofit-related actions, experts were asked to select from a list, or suggest, three policy instruments that would improve the chances of a retrofit occurring. They were then asked to evaluate their confidence in their answers on a five-point Likert-scale. Lastly, an optional write-in response space was provided for additional comments. Figure 2 provides a compiled view of the survey as it appeared to the experts through the Qualtrics interface. This includes an example of one of the representative constructions that would have appeared to the experts.
Correlation tests were conducted between features of the representative constructions and question responses using Kendall’s tau-b (Kendall 1948). Significance was evaluated using a two-tailed p-value test. Kendall’s tau-b is a non-parametric measure of the strength and direction of association that is recommended for use when calculating the correlation between two variables measured on an ordinal scale and in cases where the ordinal values being tested for correlation may have the same value, otherwise known as a tie (Puth et al. 2015). It is recommended to use a sample size of at least 46 to obtain a moderate-to-strong tau correlation of 0.5 or –0.5 with a significance (p < 0.05) under a 0.3 Fisher confidence interval (Bonett & Wright 2000; Akoglu 2018). Similar ranges of recommendations are found elsewhere (Zahiri & Elsharkawy 2018; Geekiyanage & Ramachandra 2020). While larger samples are ideal, it has been noted that decision-makers can be provided with results from elicitation based on sample sizes of three to five experts, and that larger numbers may lead to diminishing marginal returns on information with respect to the effort necessary to conduct the elicitation and analysis (Clemen & Winkler 1999).
Twenty-two surveys were returned for a total of 66 assessments of the representative constructions. Four of these were not completed entirely and were removed from the dataset. An additional six were removed from the dataset because they were from experts who indicated ‘low confidence’ in their answers and ‘less-than-five years’ of experience in their field. This provided a data set of 56 assessments. Figure 3 shows the distribution of experts’ responses to the characterization of profession, experience, and relevant knowledge. One respondent indicated a ‘very low’ familiarity with the BC Energy Step Code, but their response was kept due to their second-to-most familiar degree of familiarity with the ‘real estate’ market and ‘15–19’ years of experience in the ‘Government’ field. Appendix B in the supplemental data online contains the repository for the raw responses, as well as a Python notebook showing how the dataset was formatted for analysis.
Figure 4 shows the cumulative responses to each of the primary questions, depicting a general decline in likelihood of retrofit as higher retrofit performance levels are assessed. Experts generally viewed the likelihood of achieving high degrees of energy efficiency by 2035 as more unlikely than likely. They found it more likely than not that dwellings would perform at Step Code 2 levels in BC by 2035—10% more energy efficient than present-day code compliance, whereas new construction will be required to be 40% more energy efficient. The likelihood that any building would achieve high energy efficiency (40% more efficient) in the next 15 years was nearly split with three more responses on the unlikely side of the distribution.
Regarding the question on demolition, deep retrofit, or neither by 2050, 43% of responses expected dwellings to be demolished and replaced with new construction, and only 12% of responses indicated the favorability of deep retrofit—equivalent to net zero energy readiness. A total of 45% of responses expected the dwelling to be neither demolished nor retrofitted to the highest energy-efficiency standard. Considering their response to this question, experts were asked to evaluate whether the property would be carbon free by 2050. These mixed responses are traced in Figure 5, which shows how responses tracked from the question about whether the building would be demolished and replaced, retrofitted, or neither to the question about whether the building new or existing would reach carbon-free performance by 2050. Here, 67% of responses attributed a likely belief of a dwelling becoming carbon free by 2050.
Experts were then asked which three policy instruments would be useful in improving the likelihood that the representative construction in question would be retrofitted to a high-performance status in the future. Figure 6 shows these responses as a percentage of total responses and includes those submitted in write-in format under an ‘other’ category. In several cases these responses were categorized into one of the original instrument options.1
In Figure 7, the write-in questions, regarding influential factors and additional comments for the survey, are summarized to show which factors were perceived to be the most influential. Most of the responses were categorized, except for a few which were placed into an ‘Other’ category.2 Each category is represented as a percentage of the total responses. Of the written responses, the most common driver of a carbon-free property was the expectation that within the next several years to decade, Vancouver would have a policy in place requiring new construction not to use natural gas but instead the hydroelectric grid network. This was followed by remarks indicating that retrofit investment was unlikely due to competing financial interest or a lack of savings.
Figure 8 shows the resulting values and their significance from correlation tests between expert responses and features of the representative constructions. The color coding of individual correlations identifies Kendall’s tau correlation. The limits of the color scale (–0.5 and 0.5) correspond with indications of ‘moderate-to-strong’ correlations per the findings in Akoglu (2018). The size of individual relationship points corresponds with the relationship’s p-value.
Widespread, strong correlations were not found across the survey results. In only a few instances are moderate-to-strong correlations between several parameters observed, and these can be attributed to the general trend of survey responses considering the small sample size. For instance, few survey respondents felt that any residential dwelling would likely be retrofitted to levels equal or above Step Code 3. A total of 64 out of 224 survey responses felt that a retrofit to the candidate household was viable by 2035, but 59% of these suggested it was only likely the home would be retrofitted to Step Code 2 equivalent levels. Figure 8 reveals that some characteristic parameters of the household have a positive or a negative influence on whether a survey respondent would suggest a Step Code 2 retrofit is favorable. For instance, perceived pro-environmental views of the household are correlated positively with a retrofit to Step Code 2 levels. The age of the home’s envelope and mechanical equipment seem negatively correlated, suggesting that newer homes, or those recently renovated but not achieving high energy efficiency, are unlikely to be retrofitted significantly in the next 15 years. Such observations show significance with p < 0.05 and Kendall tau = 0.4–0.5. Figure 9 shows these values in the same format as Figure 8, but with the sample size excluding the West Cost Modern (WC) home typology, which is built to satisfy Step Code 2 performance levels at the time of construction.
Of the remaining correlations (p < 0.05, Kendall tau = 0.2–0.3) a positive relationship can be seen between dwellings with more ‘gross floor area,’ well maintained properties (‘maintenance’), and older constructions (‘typology’) and Step Code 5 performance, which is the highest performance level and would require energy efficiency improvements > 40%. Experts also perceive owner-occupiers with more feelings of responsibility to environmental protection as more likely to retrofit their homes.
Table S4 in the supplemental data online provides quotations from 26 individual write-in responses that experts provided in response to each representative construction, as well as the responses to the questions, details about their professional status derived from the surveys, and details about the representative construction being discussed in the quotation. Their answers help to contextualize issues and nuances around perceived retrofit viability, more so that can be drawn from only the responses to the survey questions.
Subjective evaluation of all write-in survey responses suggests four common themes shared across responses:
Through their direct responses to questions and their write-in responses, many participants indicated a lack of favorable pathways for SDHs to be retained in the long term. Ten comments (18% of responses) were received that contributed views in support of this. A further five comments (9% of responses) held opposite views, but only to properties with existing heritage value.
Comments that indicated a preference towards rebuilding, such as quotations (d, i, t) in Table S4 in the supplemental data online, appeared to reflect the current market tendency to demolish and build higher value properties on higher valued land. Comments such as that of quotation (v) suggested that even for the sake of carbon emissions, the construction of new SDH homes on existing properties provides an avenue to develop high-performance buildings that could achieve levels of performance that may be out of reach in retrofits due to their fixed form and orientation.
Although demolitions are currently a dominating factor of the Vancouver real estate market, there have been public calls and government action in support of preserving SDHs with heritage qualities (The Vancouver Heritage Foundation 2019). Experts shared views consistent with the awareness of this. For the heritage-valued Craftsman home, five responses (50% of the total responses for Craftsman-based narratives) indicated that the existing properties would likely remain standing in 2050. Preservation does not, however, translate into high performance.
Only two experts expected the existing property to perform at or greater than a ‘Step Code 4’-rated home by 2050. In some responses, quotations (g, n, r, u) in Table S4 in the supplemental data online, experts viewed that either the aesthetic impact of interventions or the economic conditions of the household would not favor deep retrofits. Some of the same experts, and others, identified that the trend toward installing air-source heat pumps (ASHPs) in Vancouver for space heating and cooling would suit the heritage property well—(quotations n, o, p)—indicating that ASHPs would be a more favorable solution to reducing the operating carbon emissions of these homes.
Quotation (p) suggests that due to Vancouver’s recent policies which incentivize retaining heritage homes, it is possible that the homeowners of the property would invest in constructing a small (about 70 m2) detached backyard rental unit on the property, which could be a source of additional revenue for the household, allowing for a deep retrofit investment.3 The expert still did not believe it likely that the property would achieve carbon neutrality by 2050.
As quotation (z) put it, ‘there is conflation of energy efficiency [and] GHG reduction.’ Thirteen responses (23% of the total), including quotations (t, x), suggested that switching the fuel type for space heating and domestic hot water needs, from natural gas to electricity, would be a viable way to decarbonize the energy consumption of the household. Similarly, seven responses (12%), including quotations (a, c), specifically identified the adoption of ASHPs as favorable or likely at the end of the service life of their existing heating systems. The relationship between ASHP adoption and future space cooling needs in Vancouver was also brought up by some experts, notably respondent 23 in quotations (b, c, n) where the respondent noted the importance of a lack of space cooling in the representative construction as significant in their perception of each representative constructions retrofit potential.
Some experts stated that the likelihood of retrofitting a property would necessitate a change in the personal values of the household and/or a change in how society judges those who do not act on their personal carbon emissions reductions (quotations l, m, x). The views of experts for these three cases were consistent with how the narratives they reviewed were established (VS1, VS3, and VS5, respectively). For these cases, owner-occupiers were presented as having little interest in environmental sustainability. However, narrative VS6 was considered to be occupied by environmentally conscious and financially secure owner-occupiers. However, neither of the two responses to VS6 perceived that the home would be retrofitted to its highest possible level of performance (quotation k).
Some experts raised concerns around the practical issues deep retrofits present with daily life. One poignant example explains that homes with children could create challenges for timing, staging, and realizing deep retrofits (quotation u). Deep retrofits can be disruptive (quotation y) and homes would likely combine or embed energy-saving retrofits within larger cosmetic renovation projects (quotation w).
Lastly multiple responses indicated that the family income could not support a deep retrofit (quotations f, r, s, v, w). Furthermore, quotations (h, t, y) indicate that any savings incurred from a deep retrofit would not be enough to support the initial investment. The point of developing additional rental space on the properties, through either basement suites or small detached dwelling units, was noted several times in the responses as a potential form of financing retrofit activity (quotations c, x). The expert who wrote quotation (x) was skeptical of the resulting property achieving high energy efficiency by 2050.
The most significant finding—and one not identified in prior retrofit studies concerning Vancouver or BC—is the potential negative impact of Vancouver’s real estate market on perceived retrofit uptake.
The survey results confirm that the overarching economics of Vancouver’s real-estate market is perceived to imbue a significant, downwards effect on perceived retrofit likelihood and uptake. For example, the survey experts were broadly aligned to the view that pre-2010 constructed, non-heritage homes in Vancouver will likely be demolished and rebuilt by 2050, given current market conditions and policies. Only 46% of responses to individual representative constructions (26 out of 56) felt the respective synthetic homes would remain standing by 2050 and receive some degree of retrofits in the years to come. Additionally, only 59% of responses to representative constructions (33 out of 56 responses) agreed that the SDH residing on the synthetic property, whether existing or a new construction, would be carbon free by 2050. This second point stems from a diverse set of survey experts, including several government officials across the Metro Vancouver region. It suggests there is more than just fringe cynicism or doubt with respect to Vancouver’s own 2050 decarbonization targets.4
This survey of experts identifies real estate market forces as a potential barrier to building retrofit. Previous studies modelling building physics and a life-cycle cost–benefit analysis did not account for real-estate market forces affecting the perceived cost–benefits of energy retrofits.
Three potential policy instruments emerged as the predominant retrofit incentives:
The first two options currently have traction in the provincial CleanBC: Clean British Columbia Climate Strategy (2018). The third has been investigated through research undertaken at the provincial level within the last few years, in preparation for the possible implementation of labelling in SDHs by 2022 (City Green Solutions 2018). However, other large-scale labelling schemes such as that in the European Union have not assisted countries in reaching their renovation targets and require coupling with other policy instruments (Gonzalez-Caceres et al. 2020). In some cases, labelling has proven effective, particularly when the label educates the homeowner on improvement options (Allcott & Rogers 2014), as echoed in the present study and mentioned in quotation (e) in Table S4 in the supplemental data online.
Through an analysis of the survey responses, the following points were drawn in the context of retrofit decision-making in Vancouver:
Additionally, experts noted that the lack of financial savings from deep retrofits made them difficult to undertake.
These points raise several questions for future research:
Representative constructions can be synthesized from a variety of data sources and used appropriately as a narrative for research in the absence of comprehensive real datasets (Bold 2012). Several questions arise when undertaking such work. These include whether a narrative represents its study population fairly and acknowledges the contextual conditions in which the data were gathered.
In this study, respondents were asked to characterize their confidence in whether they believed a particular representative construction and their responses to be plausible. For the expert, this determination of plausibility is based on their own knowledge of the subject matter and whether the data presented to them are legible and appear to be realistic. Quotation (j) in Table S4 in the supplemental data online is one example where the expert noted the similarity of the representative construction to their own life. Overall, the written responses do not mention that narratives being unrealistic.
The study sample size, in terms of both representative constructions and survey participants, is admittedly insufficient for the estimation of forecasts of retrofit viability in Vancouver in a statistically robust manner. However, this was not the aim. Our study sought to observe whether there was a general tendency of responses in favour of high retrofit likelihoods, or lower, and the alignment with the latter was resounding. This synthesis of the present study’s results in this manner aligns with Bold’s (2012) recommended best practices.
Some findings in this study are based on statistical metrics such as p-value and Kendall’s tau correlation, but the individual responses provided by the experts are potentially the most rewarding outcome of this study. This paper has not gone as far as establishing, based on the expert-elicited data, the prior conditional probability distributions of retrofit likelihood as a function of building/homeowner characteristics. As this is a frequent output of expert elicitation studies, the authors accept this as a limitation of the present study as much as it is a call for the future analysis of the generated dataset. Nevertheless, the authors accept the challenges this added step may face, based on the methods chosen for this study. Both Morgan (2014) and Tversky & Kahneman (1974) have advocated a high degree of caution against attempts to infer the probabilistic likelihood of a phenomena based only on experts’ qualitative view of what they perceive to be ‘likely’ or ‘unlikely.’ The present study is subject to these perils because it elected for a survey-based process that did not engage with respondents in order to calibrate their answers against a standard definition of likelihood. This issue must be considered in any future work.
An expert elicitation survey was used to gauge the professional perspectives of building retrofit experts in Vancouver about the viability of single-detached home (SDH) retrofits achieving high degrees of energy efficiency and low-carbon performance.
The responses were consistent with known barriers to residential building retrofits. Experts mentioned the practicality of deep retrofits in the context of busy family lives, the need to educate consumers about zero-carbon retrofit options, as well as the difficulty of financing a deep retrofit given prolonged payback periods due to low energy costs in Vancouver.
Some unexpected insights emerged. Among all barriers facing retrofits, experts felt the real estate market’s tendency towards a high rate of demolitions and the replacement of existing SDHs has the most negative impact. SDHs possessing heritage value were to an extent not affected by this phenomenon. However, even in these cases, maximizing dwelling energy efficiency through retrofitting, or achieving carbon neutrality, was not thought to be a likely outcome. There was pervasive doubt amongst experts that the majority of SDHs will achieve carbon neutrality by 2050, whether they are existing dwellings or new dwellings yet to be built.
Several recommended policy instruments did emerge from the survey responses, together suggesting there are viable pathways to accelerate retrofit uptake rates. Experts identified the need to establish retrofit codes and standards, energy label and performance disclosure programs, and innovative retrofit-financing mechanisms.
Regarding the overall method of undertaking an expert elicitation survey, the authors note several benefits and trade-offs. The survey is scalable, rapidly deployable, and able to provide insights into retrofit decision-making that would not have been easily obtained with a simulation-based study. Based on the experience of this study, undertaking a similar effort for a large metropolitan area may require further simplification of owners and building types.
4A total of 100% renewable (net zero carbon) resources by 2050, an 80% reduction of GHG emissions by 2050) (City of Vancouver 2015).
The authors are grateful to the group of professionals who anonymously submitted responses to this survey, as well as to those who assisted in its development.
The authors have no competing interests to declare.
This project was funded by the Pacific Institute for Climate Solutions (PICS) Energy Efficiency in the Built Environment Project (grant number PICS 36170-50280).
Achtnicht, M., & Madlener, R. (2014). Factors influencing German house owners’ preferences on energy retrofits. Energy Policy, 68, 254–263. DOI: https://doi.org/10.1016/j.enpol.2014.01.006
Addy, M. N., Adinyira, E., & Koranteng, C. (2014). Architect’s perception on the challenges of building energy efficiency in Ghana. Structural Survey, 32(5), 365–376. DOI: https://doi.org/10.1108/SS-03-2014-0014
Akoglu, H. (2018). User’s guide to correlation coefficients. Turkish Journal of Emergency Medicine, 18(3), 91–93. DOI: https://doi.org/10.1016/j.tjem.2018.08.001
Alberini, A., Banfi, S., & Ramseier, C. (2013). Energy efficiency investments in the home: Swiss homeowners and expectations about future energy prices. Energy Journal, 34(1), 49–86. www.jstor.org/stable/41969211. DOI: https://doi.org/10.5547/019565184.108.40.206
Allcott, H., & Rogers, T. (2014). The short-run and long-run effects of behavioral interventions: Experimental evidence from energy conservation. American Economic Review, 104(10), 3003–3037. DOI: https://doi.org/10.1257/aer.104.10.3003
Anadón, L. D., Bosetti, V., Bunn, M., Catenacci, M., & Lee, A. (2012). Expert judgments about R&D and the future of nuclear energy. Environmental Science & Technology, 46(21), 11497–11504. DOI: https://doi.org/10.1021/es300612c
Baker, E., Chon, H., & Keisler, J. (2009). Carbon capture and storage: Combining economic analysis with expert elicitations to inform climate policy. Climatic Change, 96(3), 379–408. DOI: https://doi.org/10.1007/s10584-009-9634-y
Banfi, S., Farsi, M., Filippini, M., & Jakob, M. (2008). Willingness to pay for energy-saving measures in residential buildings. Energy Economics, 30(2), 503–516. DOI: https://doi.org/10.1016/j.eneco.2006.06.001
BC Hydro. (2020, October). BC Hydro residential rates. https://www.bchydro.com/accounts-billing/rates-energy-use/electricity-rates/residential-rates.html
BC Ministry of Environment and Climate Change Strategy. (2019). Methodological guidance for quantifying greenhouse gas emissions. https://www2.gov.bc.ca/assets/gov/environment/climate-change/cng/methodology/2018-pso-methodology.pdf
Bergmann, J. V. (2016, June). SDH zoning and land use: How much land do single detached and duplex houses consume? https://doodles.mountainmath.ca/blog/2016/06/17/sdh-zoning-and-land-use/
Bold, C. (2012). Using narrative in research. SAGE. DOI: https://doi.org/10.4135/9781446288160
Bonett, D. G., & Wright, T. A. (2000). Sample size requirements for estimating Pearson, Spearman and Kendall correlations. Psychometrika, 65(1), 23–28. DOI: https://doi.org/10.1007/BF02294183
Bosetti, V., Catenacci, M., Fiorese, G., & Verdolini, E. (2012). The future prospect of PV and CSP solar technologies: An expert elicitation survey. Energy Policy, 49, 308–317. DOI: https://doi.org/10.1016/j.enpol.2012.06.024
British Columbia. (2019). BC energy step code revision 2. Building and Safety Standards Branch, BCBC2018, Division B (10.2.3). https://energystepcode.ca/
British Columbia Assessment Authority. (2017). Parcel inventory data [dataset]. https://www.bcassessment.ca/Property/AssessmentSearch
City Green Solutions. (2018). Put a label on it: The BC Energy Step Code & home energy labelling disclosure. http://energystepcode.ca/app/uploads/sites/257/2019/11/PutALabelOnIt_FINAL_V1.2.pdf
City of Vancouver. (2015). Renewable city strategy. https://vancouver.ca/files/cov/renewable-city-strategy-booklet-2015.pdf
City of Vancouver. (2017). Renewable city action plan economic modelling results. https://vancouver.ca/files/cov/energy-and-emissions-forecast-final-report.pdf
CleanBC. (2018). CleanBC: Clean British Columbia Climate Strategy (Technical Report). CleanBC. https://www2.gov.bc.ca/assets/gov/environment/climate-change/action/cleanbc/cleanbc_2018-bc-climate-strategy.pdf
Clemen, R. T., & Winkler, R. L. (1999). Combining probability distributions from experts in risk analysis. Risk Analysis, 19(2), 187–203. DOI: https://doi.org/10.1023/A:1006917509560
Connor Properties Listing. (2019). Listing. https://connorproperties.ca/recip.html/222926.search/listing.r2331638-2925-waterloo-street-vancouver-v6r-3j4.82343753
Dahmen, J., von Bergmann, J., & Das, M. (2018). Teardown index: Impact of property values on carbon dioxide emissions of single family housing in Vancouver. Energy and Buildings, 170, 95–106. DOI: https://doi.org/10.1016/j.enbuild.2018.03.012
Dessai, S., Bhave, A., Birch, C., Conway, D., Garcia-Carreras, L., Gosling, J. P., Mittal, N., & Stainforth, D. (2018). Building narratives to characterise uncertainty in regional climate change through expert elicitation. Environmental Research Letters, 13(7), 074005. DOI: https://doi.org/10.1088/1748-9326/aabcdd
Duah, D., Ford, K., & Syal, M. (2014). Expert knowledge elicitation for decision-making in home energy retrofits. Structural Survey, 32(5), 377–395. DOI: https://doi.org/10.1108/SS-01-2014-0004
Environment and Climate Change Canada. (2017). Canada’s nationally determined contribution to UNFCCC. UNFCCC. https://www4.unfccc.int/sites/ndcstaging/PublishedDocuments/Canada%20First/Canada%20First%20NDC-Revised%20submission%202017-05-11.pdf
Environment and Climate Change Canada. (2019). Canada’s official greenhouse gas inventory [dataset]. https://open.canada.ca/data/en/dataset/779c7bcf-4982-47eb-af1b-a33618a05e5b
Feng, H., Liyanage, D. R., Karunathilake, H., Sadiq, R., & Hewage, K. (2020). BIM-based life cycle environmental performance assessment of single-family houses: Renovation and reconstruction strategies for aging building stock in British Columbia. Journal of Cleaner Production, 250, 119543. DOI: https://doi.org/10.1016/j.jclepro.2019.119543
Frappé-Sénéclauze, T.-P., Heerema, D., & Wu, K. T. (2017). Deep emissions reduction in the existing building stock: Key elements of a retrofit strategy for BC. Results from the November 2016 Thought Leader Forum (Technical Report). Pembina Institute. https://www.pembina.org/reports/retrofit-strategy-bc-report-2017.pdf
Gamtessa, S. F. (2013). An explanation of residential energy-efficiency retrofit behavior in Canada. Energy and Buildings, 57, 155–164. DOI: https://doi.org/10.1016/j.enbuild.2012.11.006
Garthwaite, P. H., Kadane, J. B., & O’Hagan, A. (2005). Statistical methods for eliciting probability distributions. Journal of the American Statistical Association, 100(470), 680–701. DOI: https://doi.org/10.1198/016214505000000105
Geekiyanage, D., & Ramachandra, T. (2020). Nexus between running costs and building characteristics of commercial buildings: Hedonic regression modelling. Built Environment Project and Asset Management, 10(3), 389–406. DOI: https://doi.org/10.1108/BEPAM-12-2018-0156
Gonzalez-Caceres, A., Lassen, A. K., & Nielsen, T. R. (2020). Barriers and challenges of the recommendation list of measures under the EPBD scheme: A critical review. Energy and Buildings, 223, 110065. DOI: https://doi.org/10.1016/j.enbuild.2020.110065
Haines, V., & Mitchell, V. (2014). A persona-based approach to domestic energy retrofit. Building Research & Information, 42(4), 462–476. DOI: https://doi.org/10.1080/09613218.2014.893161
Hrovatin, N., & Zorić, J. (2018). Determinants of energy-efficient home retrofits in Slovenia: The role of information sources. Energy and Buildings, 180, 42–50. DOI: https://doi.org/10.1016/j.enbuild.2018.09.029
Jakob, M. (2007). The drivers of and barriers to energy efficiency in renovation decisions of single-family home-owners (Technical Report). ETH Zürich. https://cepe.ethz.ch/content/dam/ethz/special-interest/mtec/cepe/cepe-dam/documents/research/cepe-wp/CEPE_WP56.pdf
Judson, E. P., & Maller, C. (2014). Housing renovations and energy efficiency: Insights from homeowners’ practices. Building Research & Information, 42(4), 501–511. DOI: https://doi.org/10.1080/09613218.2014.894808
Kahneman, D., & Tversky, A. (1982). Variants of uncertainty. Cognition, 11(2), 143–157. DOI: https://doi.org/10.1016/0010-0277(82)90023-3
Kastner, I., & Stern, P. C. (2015). Examining the decision-making processes behind household energy investments: A review. Energy Research & Social Science, 10, 72–89. DOI: https://doi.org/10.1016/j.erss.2015.07.008
Kendall, M. (1948). Rank correlation methods. Griffin. https://psycnet.apa.org/record/1948-15040-000
Kynn, M. (2008). The ‘heuristics and biases’ bias in expert elicitation. Journal of the Royal Statistical Society. Series A (Statistics in Society), 171(1), 239–264. https://www.jstor.org/stable/30130739
Lam, J. C., & Hui, S. C. M. (1996). Sensitivity analysis of energy performance of office buildings. Building and Environment, 31(1), 27–39. DOI: https://doi.org/10.1016/0360-1323(95)00031-3
Ma, Z., Cooper, P., Daly, D., & Ledo, L. (2012). Existing building retrofits: Methodology and state-of-the-art. Energy and Buildings, 55, 889–902. DOI: https://doi.org/10.1016/j.enbuild.2012.08.018
Martin, T. G., Kuhnert, P. M., Mengersen, K., & Possingham, H. P. (2005). The power of expert opinion in ecological models using Bayesian methods: Impact of grazing on birds. Ecological Applications, 15(1), 266–280. https://www.jstor.org/stable/4543351. DOI: https://doi.org/10.1890/03-5400
Mata, E., Sasic Kalagasidis, A., & Johnsson, F. (2014). Building-stock aggregation through archetype buildings: France, Germany, Spain and the UK. Building and Environment, 81, 270–282. DOI: https://doi.org/10.1016/j.buildenv.2014.06.013
Mirzaee, S., Fannon, D., & Ruth, M. (2019). A comparison of preference elicitation methods for multi-criteria design decisions about resilient and sustainable buildings. Environment Systems and Decisions, 39(4), 439–453. DOI: https://doi.org/10.1007/s10669-019-09726-2
Morgan, M. G. (2014). Use (and abuse) of expert elicitation in support of decision making for public policy. Proceedings of the National Academy of Sciences, USA, 111(20), 7176–7184. DOI: https://doi.org/10.1073/pnas.1319946111
Morgan, M. G., & Keith, D. W. (1995). Subjective judgments by climate experts. Environmental Science & Technology, 29(10), 468A–476A. DOI: https://doi.org/10.1021/es00010a753
Nair, G., Gustavsson, L., & Mahapatra, K. (2010). Factors influencing energy efficiency investments in existing Swedish residential buildings. Energy Policy, 38(6), 2956–2963. DOI: https://doi.org/10.1016/j.enpol.2010.01.033
Palmer, K., Walls, M., Gordon, H., & Gerarden, T. (2013). Assessing the energy-efficiency information gap: Results from a survey of home energy auditors. Energy Efficiency, 6(2), 271–292. DOI: https://doi.org/10.1007/s12053-012-9178-2
Pasichnyi, O., Wallin, J., & Kordas, O. (2019). Data-driven building archetypes for urban building energy modelling. Energy, 181, 360–377. DOI: https://doi.org/10.1016/j.energy.2019.04.197
Prabatha, T., Hewage, K., Karunathilake, H., & Sadiq, R. (2020). To retrofit or not? Making energy retrofit decisions through life cycle thinking for Canadian residences. Energy and Buildings, 226, 110393. DOI: https://doi.org/10.1016/j.enbuild.2020.110393
Puth, M.-T., Neuhäuser, M., & Ruxton, G. D. (2015). Effective use of Spearman’s and Kendall’s correlation coefficients for association between two measured traits. Animal Behaviour, 102, 77–84. DOI: https://doi.org/10.1016/j.anbehav.2015.01.010
Qualtrics. (2019). Qualtrics survey software 2019. https://www.qualtrics.com/blog/citing-qualtrics/
Rivers, N., & Jaccard, M. (2006). Useful models for simulating policies to induce technological change. Energy Policy, 34(15), 2038–2047. DOI: https://doi.org/10.1016/j.enpol.2005.02.003
Salter, J., Lu, Y., Kim, J. C., Kellett, R., Girling, C., Inomata, F., & Krahn, A. (2020). Iterative ‘what-if’ neighborhood simulation: Energy and emissions impacts. Buildings and Cities, 1(1), 293–307. DOI: https://doi.org/10.5334/bc.51
Shadbolt, N. (2005). Eliciting expertise. In J. R. Wilson & N. Corlett (Eds.), Evaluation of human work, 3rd edn (pp. 185–218). CRC Press. https://eprints.soton.ac.uk/264563/1/Elciting_Expertise.pdf. DOI: https://doi.org/10.1201/9781420055948.ch8
Statistics Canada. (2017). 2016 Census metropolitan area of Vancouver [dataset]. https://www12.statcan.gc.ca/census-recensement/2016/as-sa/fogs-spg/Facts-cma-eng.cfm?LANG=Eng&GK=CMA&GC=933&TOPIC=7
Sustainability Group, City of Vancouver. (2014, June). Energy retrofit strategy for existing buildings. Report to the City of Vancouver Standing Committee on Planning, Transportation and Environment. https://council.vancouver.ca/20140625/documents/ptec1.pdf
Szwed, P., Dorp, J. R. V., Merrick, J., Mazzuchi, T., & Singh, A. (2006). A Bayesian paired comparison approach for relative accident probability assessment with covariate information. European Journal of Operational Research, 169(1), 157–177. DOI: https://doi.org/10.1016/j.ejor.2004.04.047
The Vancouver Heritage Foundation. (2018). Vancouver house styles hub. https://www.vancouverheritagefoundation.org/learn-with-us/discover-vancouvers-heritage/vancouver-house-styles/ (Library Catalog www.vancouverheritagefoundation.org).
The Vancouver Heritage Foundation. (2019). Heritage energy retrofit grant. https://www.vancouverheritagefoundation.org/get-a-grant/heritage-energy-retrofit-grant/
Tjørring, L., & Gausset, Q. (2019). Drivers for retrofit: A sociocultural approach to houses and inhabitants. Building Research & Information, 47(4), 394–403. DOI: https://doi.org/10.1080/09613218.2018.1423722
Trotta, G. (2018). The determinants of energy efficient retrofit investments in the English residential sector. Energy Policy, 120, 175–182. DOI: https://doi.org/10.1016/j.enpol.2018.05.024
Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185(4157), 1124–1131. DOI: https://doi.org/10.1126/science.185.4157.1124
Usher, W., & Strachan, N. (2013). An expert elicitation of climate, energy and economic uncertainties. Energy Policy, 61, 811–821. DOI: https://doi.org/10.1016/j.enpol.2013.06.110
Verdolini, E., Anadon, L. D., Lu, J., & Nemet, G. F. (2015). The effects of expert selection, elicitation design, and R&D assumptions on experts’ estimates of the future costs of photovoltaics. Energy Policy, 80, 233–243. DOI: https://doi.org/10.1016/j.enpol.2015.01.006
Vergragt, P. J., & Brown, H. S. (2020). The challenge of energy retrofitting the residential housing stock: Grassroots innovations and socio-technical system change in Worcester, MA. Technology Analysis & Strategic Management, 24(4), 407–420. DOI: https://doi.org/10.1080/09537325.2012.663964
Wilson, C., Chryssochoidis, G., & Pettifor, H. (2011). Understanding homeowners’ renovation decisions: Findings of the VERD Project (Technical Report). UK Energy Research Centre. https://tyndall.ac.uk/sites/default/files/verd_summary_report_oct13.pdf
Wilson, C., Crane, L., & Chryssochoidis, G. (2015). Why do homeowners renovate energy efficiently? Contrasting perspectives and implications for policy. Energy Research & Social Science, 7, 12–22. DOI: https://doi.org/10.1016/j.erss.2015.03.002
Wiser, R., Jenni, K., Seel, J., Baker, E., Hand, M., Lantz, E., & Smith, A. (2016). Expert elicitation survey on future wind energy costs. Nature Energy, 1(10), 1–8. DOI: https://doi.org/10.1038/nenergy.2016.135
Woudenberg, F. (1991). An evaluation of Delphi. Technological Forecasting and Social Change, 40(2), 131–150. DOI: https://doi.org/10.1016/0040-1625(91)90002-W
Zahiri, S., & Elsharkawy, H. (2018). Towards energy-efficient retrofit of council housing in London: Assessing the impact of occupancy and energy-use patterns on building performance. Energy and Buildings, 174, 672–681. DOI: https://doi.org/10.1016/j.enbuild.2018.07.010