Analyzing solar installations: a catalyst or barrier to subsequent residential retrofits

Analyzing solar installations: a catalyst or barrier to subsequent residential retrofits | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Analyzing solar installations: a catalyst or barrier to subsequent residential retrofits Marek Bohacek This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6330919/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 07 Jan, 2026 Read the published version in Energy Efficiency → Version 1 posted You are reading this latest preprint version Abstract This study investigates the impact of solar photovoltaic (solar PV) system installations on subsequent energy retrofitting behaviour among residential dwellings in Ireland, particularly within the context of the nation's commitment to sustainability and renewable energy as outlined in the National Development Plan 2021–2030. Despite over 65,000 solar PV installations facilitated by the Sustainable Energy Authority of Ireland (SEAI), findings suggest that the installation of solar PV does not significantly encourage additional retrofitting actions when compared to alternative measures, such as attic insulation and heating controls. Using statistical methodologies, including chi-square tests and survival analysis, the study reveals that homeowners who adopt solar PV systems exhibit a minimal propensity to pursue further retrofits, with an average additional investment of only €200 per installation. While the results indicate a nuanced relationship influenced by dwelling type and energy building rating (BER), the hypothesis that solar PV serves as a "gateway" to broader energy efficiency improvements is unsupported. The findings emphasize the need for integrated energy strategies that encompass both renewable technology adoption and comprehensive energy performance upgrades to achieve Ireland's ambitious climate targets. This research underscores the importance of fostering holistic approaches to retrofitting that encourage homeowners to invest not just in renewable systems such as solar PV but also in measures that enhance overall energy efficiency such as dwelling insulation. Photovoltaics residential energy retrofit energy efficiency Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction The urgent need to address climate change has driven a global transition toward renewable energy, with solar photovoltaics (PV) emerging as a key solution. In Ireland, the National Development Plan 2021–2030 (Department of Public Expenditure and Reform, 2021 ) highlights the importance of solar PV in achieving the ambitious target of generating 70% of electricity from renewable sources by 2030, making solar energy a cornerstone of the country's sustainability strategy. Supportive policies, including feed-in tariffs and grants, have facilitated residential solar PV installations. A study in the UK found that feed-in tariffs significantly boosted the popularity of residential PV systems (Frances and Stevenson, 2019 ), a trend mirrored in Ireland with rising application rates (SEAI, 2024 ). Additionally, a social contagion effect may drive the adoption of solar PV, as individuals and communities influence one another, motivated by social norms, visibility, and peer effects (Graziano and Gillingham, 2014 ). Despite the increasing interest in solar PV, several critical issues remain. Notably, the timing of solar electricity generation often does not align with residential energy demand, which typically peaks in the evening when solar output is low. While solar PV can reduce electricity bills and generate income through feed-in tariffs, it does not inherently improve a home's overall energy efficiency. Ideally, solar PV should be the final component in a series of energy efficiency measures (Haas et al., 1999 ), a consideration often overlooked in Ireland. In the UK, a dwelling qualifies for a solar PV grant only after achieving sufficient insulation in its envelope (Frances and Stevenson, 2019 ). This raises concerns about the opportunity cost of investing in solar PV instead of more impactful energy efficiency measures, such as insulation and heating system retrofits. Financial returns on residential solar PV investments in Ireland and the UK are estimated to extend beyond 20 years (Bergman et al., 2009 ; Ayompe, 2011 ). Even a recent Irish study, which accounts for feed-in tariffs and Sustainable Energy Authority of Ireland (SEAI) grants under ideal conditions (e.g., south-facing installation, no shading, clean panels), suggests a return on investment of at least 10 years (Ryan, Wheatley, and Saba, 2023 ). Furthermore, the demographic disparity in solar PV adoption is noteworthy, as installations are primarily undertaken by wealthier households (Lukanov and Krieger, 2020; Barbose et al., 2018 ), likely due to the high initial costs. Research into the secondary benefits of solar PV indicates that homeowners often report lower overall electricity consumption as they become more efficient energy users (Frances and Stevenson, 2019 ) (DECC, 2010a). However, Haas et al. ( 1999 ) identified a more complex relationship: while larger energy consumers reduced their consumption after installing PV, smaller consumers tended to increase it. A Swedish study examining electricity consumption one year before and after solar PV installation found no significant change (Palm, Eidenskog, and Luthander, 2018), while Bepler, Matisoff, and Oliver (2018) reported a 28.5% increase in consumption following installation. Bergman et al. ( 2009 ) attributes these variations to the differences between habitual and deliberative behaviours, suggesting that changes in electricity usage often fall under habitual behaviour, which is challenging to modify. Despite these concerns, proponents argue that, due to its relative ease of installation, solar PV may serve as a "gateway to retrofit", encouraging homeowners to pursue further energy efficiency measures. Those who install solar PV often report increased environmental motivation (Kastner and Stern, 2015 ), potentially leading to changes in deliberative behaviours, such as implementing additional energy efficiency measures (Bergman et al., 2009 ). Dobbyn and Thomas ( 2005 ) provide evidence of shifts in awareness, behaviour, and attitudes among individuals that gained microgeneration experience. This study aims to empirically evaluate whether the installation of solar PV encourages further energy retrofitting, potentially establishing solar PV as a "gateway to retrofit". Methodology Data sources Data on the installation of solar PV and other energy efficiency measures were collated from four active SEAI grant programmes (Better Energy Homes, Better Energy Communities, One Stop Shop and Solar PV) as well as two older programmes (Deep Retrofit and One Stop Shop Pilot). In addition, data on demographics, socioeconomic conditions and indicators related to community wellbeing were sourced from Central Statistical Office. This resulted in a dataset providing information on the retrofitting journey of Irish dwellings over time, where dwellings with multiple retrofits could be easily identified. Analysis approach The data described above were used to categorise dwellings that have solar PV installed into the following four categories: Solar PV followed by retrofit . This category represents dwellings that first installed solar PV and later installed additional retrofit measures. Retrofit followed by solar PV. This category represents dwellings fitting the opposite pattern whereby another energy efficiency measure was installed prior to solar PV installation. Solar PV and retrofit at the same time. This category represents dwellings that installed solar PV and other energy efficiency measure/s at roughly the same time – the second retrofit measure was applied for between first measure’s application date and completion date. [1] Solar PV only. This category represents dwellings that only installed solar PV and did not apply for grants for any other energy efficiency measures. [2] The above classification allows us to observe the frequency with which solar PV is followed by other retrofits but does not compare this against a counterfactual. For this reason, two counterfactual measures of attic insulation/roof insulation and heating controls were also chosen for comparison. The reasons for choosing attic insulation and heating controls as counterfactuals are threefold. First, they are common retrofit measures, giving enough data points for statistical analysis. Second, both of these measures are suitable for most of Irish dwellings. Third, they are relatively inexpensive to install compared with solar PV and can be considered “entry level” retrofits (see Table 1). Table 1: Mean and median installation cost for the most common measures over the last two years. Measure Mean cost of works Median cost of works Attic insulation/roof insulation 2,623 € 2,150 € Solar PV 13,299 € 11,100 € Heating controls 4,363 € 3,750 € Both attic insulation and heating controls were grouped into four categories in the same fashion as solar PV as described above. This allows for direct comparison of the likelihood solar PV installation is followed by further measures against the counterfactuals of attic insulation or heating controls being followed by retrofit. After data cleaning and categorization, statistical analyses were conducted using multiple methods. Initial chi-square tests provided a straightforward interpretation of associations. Subsequently, survival analysis modelling was employed, focusing on time-to-event data to understand the influence of covariates on the timing of events, such as applications for additional retrofits. The Kaplan-Meier estimator (Kaplan and Meier, 1958) was utilized for non-parametric estimation of the time-to-event function, allowing for visual representation of event probabilities over time and group comparisons via log-rank tests. The Cox proportional hazards model (Cox, 1972) further examined the relationship between covariates and the hazard rate, providing hazard ratios to quantify the effects of predictors like dwelling size and deprivation index. Additionally, random survival forests, as proposed by Ishwaran et al. (2008), leveraged an ensemble of decision trees to capture complex interactions and nonlinear relationships in censored time-to-event data. Results Descriptive statistics A total of 65,110 solar PV installations have been completed since the beginning of SEAI’s grant programme of which 1,361 were later followed by additional retrofit projects, totalling 1,839 additional measures (see Table 2 ). The cumulative investment in these additional retrofits has reached approximately €13.0 million, indicating an average of €200 in further retrofits per solar PV installation. As the SEAI PV grant has been present for 6 years, only the last 6 years of attic insulation and heating controls data are reported below for comparability. A total of 33,935 attic insulation installations were finalized during the same period. Among these, 1,063 projects were followed by subsequent retrofits, totalling 1,196 measures and amounting to a total cost of €11.9 million or €350 in additional retrofits per attic insulation installation Lastly, there were 24,225 installations of heating controls completed. In total 1,866 of these heating controls resulted in further retrofitting activities amounting to 2,393 measures, generating a total investment of €18.0 million. On average, each heating controls installation has led to an additional €740 in retrofits. For detail breakdown of measures following Heating controls installations see Table 2 , Heating controls column. Table 2 Breakdown of number of measures installed after initial retrofit shown for solar PV, attic insulation and heating controls over the last 6 years. Initial retrofit Type of measure installed at future date following the initial retrofit solar PV (65,110) Attic insulation (33,935) Heating controls (24,225) External wall insulation 178 124 123 Internal wall insulation 26 35 23 Cavity wall insulation 487 176 464 Heat pump 320 68 28 Doors 39 11 15 Windows 37 11 14 Other measures 21 2 3 Solar PV - 664 1004 Attic insulation 641 - 719 Heating controls 90 105 - Total 1839 1196 2393 The descriptive statistics above show that solar PV was followed by additional retrofit the least when compared with attic insulation and especially with heating controls. However, even though there have been a large number of solar PV, attic insulations and heating controls installed over the last 6 years, the level of subsequent retrofit is very small (Table 2 ). The amount of additional retrofitting happening at the same time is substantially larger, especially in the case of attic insulation (Table 3 ). Nonetheless, there are number of issues when looking at the descriptive data in this way. Firstly, it is unclear if solar PV is more or less likely to be followed by a retrofit at future date than attic insulation or heating controls. Secondly, those descriptives look at all the solar PV, attic insulation and heating controls installs in the last 6 years and for many of those, especially the recent ones, there hasn’t been enough time to observe additional retrofit that might happen sometime in the future. This is especially problematic for solar PV which saw large increase in applications in the last 3 years (Fig. 1 ). To address those shortcomings a series of chi-square tests was conducted next. Table 3 Breakdown of number of measures installed together (at the “same time”) with initial retrofit shown for solar PV, Attic insulation and Heating controls during last 6 years. Initial retrofit Number of Measures installed with initial retrofit (solar PV/attic insulation/heating controls) solar PV (65,110) Attic insulation (33,935) Heating controls (24,225) External wall insulation 751 2,993 507 Internal wall insulation 209 1,866 356 Cavity wall insulation 1,179 13,817 395 Heat pump 1,722 3,266 - Doors 863 1,747 43 Windows 808 1,712 43 Other measures 211 693 6 solar PV - 1,740 219 Attic insulation 2,030 - 1,246 Heating controls 508 1,933 - Total 8,281 29,767 3,084 Chi-square test results To address the issue of unobserved future retrofits, the data were coded based on the number of retrofits occurring within a two-year period following the completion of a solar PV installation, attic insulation, or heating controls (collectively referred to as "Measure"). Data from the most recent two years were excluded from the analysis, as it remains uncertain whether additional retrofits will occur following installation made during that time. A two-year period was chosen to balance having enough time between the completion of the measure and any subsequent retrofits, while minimizing the exclusion of relevant data. This adjustment effectively reduces the available data from six years to four. The category "retrofit followed by measure" follows the same two-year rule, but to avoid excessive data loss, retrofits from as far back as eight years are considered. The following rules apply: A retrofit is counted as being followed by a Measure only if it occurred within two years after the Measure's completion and no retrofit occurred in the two years preceding the Measure. A Measure is counted as being followed by a retrofit only if it occurred within two years after the completion of the retrofit. Three sets of chi-square tests were conducted: The first set of tests compared all four categories: “Measure is followed by retrofit,” “Retrofit is followed by Measure,” “Measure and retrofit at the same time,” and “Measure only.” The findings indicate that both solar PV vs attic insulation (X-squared = 16501, df = 3, p-value < 2.2e-16, Cramer’s V = 0.38) and solar PV vs heating controls (X-squared = 261, df = 3, p-value < 2.2e-16, Cramer’s V = 0.044) were significantly different in four categories (Table 4). However, it appears that the difference is mostly driven by “Measure and retrofit at the same time” category. The second set of tests recoded the data into two categories: “Measure is followed by retrofit” and “Measure is not followed by retrofit,” the latter of which includes “Retrofit is followed by Measure,” “Measure and retrofit at the same time,” and “Measure only”. This test directly assesses whether a difference in the probability of additional retrofits exists between solar PV and attic insulation/heating controls. The findings found solar PV vs attic insulation (X-squared = 1.54, df = 1, p-value < 0.21, Cramer’s V = 0.006) and solar PV vs heating controls (X-squared = 0.75, df = 1, p-value = 0.39, Cramer’s V = 0.004) were not significant. The third set of tests excluded the categories “Retrofit is followed by Measure” and “Measure and retrofit at the same time,” resulting in a dataset containing only “Measure is followed by retrofit” and “Measure only”. It was found that solar PV vs Attic insulation (X-squared = 627.94, df = 1, p-value < 2.2e-16, Cramer’s V = 0.091) was significant indicating that Attic insulation was more likely to be followed by retrofit than solar PV when multi-measure applications and “Retrofit is followed by Measure” of solar PV/Attic insulation are excluded. Solar PV vs Heating controls (X-squared = 0.65, df = 1, p-value = 0.42, Cramer’s V = 0.004) found no significant difference indicating that solar PV was no more likely to be followed by retrofit than Heating controls. Table 4: Probabilities of a measure (solar PV/attic insulation/heating controls) being followed by retrofit in next 2 years displayed by 4 categories. Last 2 years of data are excluded to allow for two-year window to see if additional retrofit materializes. Initial Measure Probabilities of: Measure only Measure and retrofit at the same time Retrofit is followed by Measure Measure is followed by retrofit Solar PV 87.2% 5.3% 4.3% 3.2% Attic insulation 28.0% 65.0% 4.0% 3.0% Heating controls 87.1% 7.6% 2.2% 3.1% The findings from the chi-square tests indicate that solar PV is no more likely to be followed by retrofit than attic insulation or heating controls. If anything, Attic insulation might be more likely followed by retrofit than solar PV when looking only at first-time standalone Measures. However, there are number of limitations. The last two years of data were excluded to allow enough time to observe whether additional retrofit materializes or not. This is of particular concern for solar PV as majority of solar PV applications were only completed in last two years (see Figure 1) and hence many datapoints from solar PV were excluded. Also, this two-year window is limiting as it does not consider retrofits that happened more than two years after the initial measure. Moreover, the chi-square test does not allow to control for covariates such as dwelling type, dwelling size, deprivation index, Building energy rating, etc. To address these limitations survival analysis was explored next. Survival analysis model results This section presents the results of the survival analysis, which utilized a comprehensive dataset that includes 6 years of data for solar PV and 16 years for attic insulation and heating controls. The analysis employed two sets of survival models to assess the likelihood of additional retrofits following initial measures: Using the full dataset, containing 160k attic insulations, 101k heating controls, and 65k solar PV installations. Using a partial dataset from which the “Measure and retrofit at the same time” and “Retrofit is followed by Measure” categories were excluded, resulting in 15.7k attic insulations, 71.4k heating controls, and 44.4k solar PV installations. The purpose of this was to investigate whether solar PV installation is likely to be followed by subsequent retrofitting where it is a first-time, standalone measure. Kaplan-Meier plots Full dataset Attic insulation is least likely to be followed by retrofit compared to solar PV and heating controls (Figure 2). Attic insulation shows a rapid initial uptake with minimal follow-up retrofits after two years, while solar PV and heating controls see a steady increase in follow-up retrofits showing almost a linear trend. However, the level of additional retrofit is limited to about 1% per year for solar PV and heating controls. Partial dataset When considering only first-time standalone measures, attic insulation and heating controls are both more likely to be followed by retrofit compared to solar PV (Figure 3). There is a dramatic increase in the likelihood of attic insulation being followed by additional retrofits, resulting from the exclusion of a large number (about 80%) of attic insulation installations for this partial dataset. Attic insulation is primarily part of more extensive retrofits including multiple measures. The level of additional retrofit increased slightly but is still limited to about 1.5% per year for solar PV and heating controls. Cox Model Full dataset The findings from the Cox model analysis indicate that all covariates are statistically significant due to the large size of the dataset (Table 5). It shows that solar PV is more likely to be followed by retrofit than attic insulation but less likely than heating controls, which aligns with the Kaplan-Meier plot. Moreover, a better BER after the initial measure increases the likelihood of applying for additional retrofits. Additionally, apartments and mid-terraced dwellings are less likely to seek additional retrofits than detached and semi-detached houses. The analysis also suggests that more affluent neighbourhoods and larger homes are more inclined to apply for subsequent retrofits, although the effect sizes are small as indicated by hazard ratios (exp(coef)). Partial dataset When considering only first-time standalone measures, attic insulation and heating controls both have a higher probability of additional retrofit when compared to solar PV (Table 6), controlling for covariates. Dwelling size and the deprivation index have a smaller influence when compared to full dataset, but other covariates follow the same pattern. Table 5: Output from Cox model (All Data). Total number of datapoints – 330,680 with 13,726 of events (initial Measure being followed by a retrofit). Covariate coefficient exp(coef) se(coef) z Pr(>|z|) Attic insulation -0.8848 0.4127 0.0402 -21.983 < 2e-16 *** Heating Controls 1.0638 2.8975 0.0356 29.872 < 2e-16 *** Deprivation index 0.0236 1.0239 0.0029 7.974 1.54e-15 *** Dwelling size 0.1526 1.1648 0.0276 5.516 3.46e-08 *** Semi-detached 0.1082 1.1143 0.0209 5.156 2.52e-13 *** Mid-terraced -0.2678 0.7650 0.0372 -7.192 6.40e-13 *** Apartment -0.5259 0.5909 0.0710 -7.408 1.29e-13 *** Unknown dwelling -1.8973 0.1499 0.0650 -29.160 < 2e-16 *** BER A 1.0214 2.7772 0.0462 22.076 < 2e-16 *** BER B 0.6277 1.8734 0.0214 29.314 < 2e-16 *** BER D -0.2300 0.7944 0.0303 -7.571 3.71e-13 *** BER EFG -0.2038 0.8156 0.0498 -4.086 4.39e-13 *** BER Unknown 0.2087 1.2321 0.0313 6.657 2.79e-13 *** Table 6: Output from Cox model (Partial Data). Total number of datapoints – 131,484 with 13,726 of events (initial Measure being followed by a retrofit). Covariate coefficient exp(coef) se(coef) z Pr(>|z|) Attic insulation 2.1863 8.9026 0.0403 54.202 < 2e-16 *** Heating Controls 1.2125 3.3619 0.0372 32.588 < 2e-16 *** Deprivation index 0.0059 1.0059 0.0029 1.991 0.0464 * Dwelling size 0.0774 1.0805 0.0282 2.746 0.0060 ** Semi-detached -0.1748 0.8396 0.0212 -8.229 < 2e-16 *** Mid-terraced -0.6613 0.5161 0.0374 -17.646 < 2e-16 *** Apartment -1.1039 0.3315 0.0711 -15.506 < 2e-16 *** Unknown dwelling -2.1529 0.1161 0.0651 -33.061 < 2e-16 *** BER A 1.3854 3.9964 0.0477 28.993 < 2e-16 *** BER B 0.9165 2.5007 0.0216 42.337 < 2e-16 *** BER D -0.4190 0.6576 0.0304 -13.757 < 2e-16 *** BER EFG -0.7156 0.4888 0.0502 -14.244 < 2e-16 *** BER Unknown 0.2110 1.2350 0.0323 6.521 < 2e-16 *** Random Forest Model Full dataset The random forest model does not generate regression outputs or dummy variables for categorical covariates, but it provides variable importance metrics (Table 7) that align closely with Cox model findings. A comparison of models suggest that the Kaplan-Meier and random forest models yield similar results, while the Cox model consistently shows a somewhat lower probability of additional retrofit. Table 7: Variable importance as produced by random forest model (all dataset). Variable Importance Type of measure 0.0935 BER following measure 0.0207 Dwelling type 0.0196 Dwelling size 0.0152 Deprivation index 0.0039 Partial dataset The pattern of variable importance metrics (Table 8) for the partial dataset is comparable to full dataset findings. A comparison of models show that all three models are predicting very similarly for the first five years since following installation of the initial measure, diverging later. Table 8: Variable importance as produced by random forest model (Partial data). Variable Importance Type of measure 0.0679 BER following measure 0.0549 Dwelling type 0.0347 Dwelling size 0.0085 Pobal deprivation index 0.0045 The survival analysis shows agreement across all three models suggesting that most important covariate affecting an increased probability of future retrofit is a dwelling’s BER. Dwellings with a better BER are more likely to retrofit again compared to dwellings with a poorer BER. The impacts of the deprivation index and size of dwelling are least important. The probability of additional retrofit in a given year following the initial measure for solar PV only is between 1% (Figure 4) based on full dataset analysis and 1.5% (Figure 5) as given by the partial dataset analysis. Discussion This research explores the potential for solar photovoltaic (PV) installation to act as a catalyst for subsequent home retrofitting. The findings provide valuable insights into how the adoption of renewable energy technologies may influence broader residential energy retrofitting behaviours, particularly in Ireland, where the urgency around climate action necessitates transformative changes in home energy consumption patterns (DECC, 2024). The data indicate that subsequent retrofitting activities are minimal compared to the extensive installation of solar PV systems. Of the 65,000 solar PV installations, only 1,839 have applied for further retrofits. Findings from survival analysis reveal that only 1-1.5% of solar PV installations seek additional retrofits each year. Similar trends are observed in the cases of attic insulation and heating controls, where additional retrofits are also relatively uncommon albeit to lesser degree. Furthermore, the results of the survival analysis demonstrate almost linear hazard functions (indicating the probabilities of additional retrofits) over time, suggesting that these retrofits occur at nearly constant rates. The results from both chi-square tests and survival analysis indicate that solar PV does not inherently catalyse additional retrofitting to the same extent as other energy efficiency measures, such as attic insulation and heating controls. The tests reveal that heating controls are more likely to be followed by additional retrofits than solar PV, while attic insulation is more frequently installed as part of a package of measures. Reported increases in environmental motivations (Kastner and Stern, 2015 ) or shifts in awareness, behaviour, and attitudes (Bergman et al., 2009 ) following solar PV installation did not translate into additional retrofits. A meta-analysis by Kormos and Gifford ( 2014 ) may provide an explanation, as it found significant discrepancies between self-reported pro-environmental behaviour and actual behaviour. Importantly, solar PV installations could be counterproductive, diverting resources from more effective measures such as insulation or heating retrofits. However, it remains unclear whether solar PV competes directly with these other measures. It is uncertain if homeowners who choose to install solar PV ever consider other, more effective retrofits. For instance, solar PV might only be considered against other types of expenditures (e.g., vacations), making it a more desirable option than forgoing improvements altogether. Further research is needed to explore this possibility. Additionally, the research reveals that not only does the probability of an additional retrofit following a solar PV installation remain low, but significant differences are observed when controlling for variables such as dwelling type, size, and energy performance rating (BER). Detached and semi-detached dwellings with better BERs in more affluent neighbourhoods are correlated with higher additional retrofit probabilities, suggesting that energy efficiency investments may be more effectively pursued within specific dwelling contexts. One limitation of the study is the lack of information regarding retrofitting measures undertaken outside of the SEAI programs. This potentially underestimates the overall impact of solar PV installations on home energy upgrades by failing to account for homeowners who may have pursued concurrent or subsequent retrofits without governmental grants. It is also important to note that a decision for additional retrofitting cannot be directly attributed to initial PV installation, as it is unknown when that decision was made and what factors influenced it. For example, financial constraints may lead homeowners to stagger their energy efficiency investments rather than pursue multiple upgrades simultaneously. This paper highlights the dual nature of energy investments, emphasizing the need for a more comprehensive retrofit strategy that encompasses not only the adoption of renewable technology but also the necessity of improving the energy performance of dwellings through integrated energy efficiency measures. Without addressing the interplay between renewable energy adoption and home energy efficiency upgrades, the potential gains from solar technology may not be fully realized in reducing overall energy consumption and greenhouse gas emissions. Conclusion While this study provides significant data and analysis regarding the potential role of solar PV in stimulating retrofit behaviours, it also underscores the complexity of transforming residential energy practices. The findings indicate that simply installing solar panels is insufficient for a substantial shift in energy efficiency uptake. For policymakers, these insights emphasize the importance of fostering an environment that encourages holistic approaches to energy retrofitting. This involves integrating financial incentives, public education, and targeted outreach to guide homeowners through the lifecycle of energy efficiency, from initial installations to meaningful retrofitting across the home. Such a multi-faceted approach will be essential for achieving Ireland's ambitious targets for renewable energy adoption and emission reductions in the coming years, ultimately contributing to a more sustainable future. Declarations Author Contribution M.B. did all Acknowledgement I would like to thank Hanna Julianne, programme manager of behavioural economics unit at Sustainable Energy Authority of Ireland (SEAI) for her comments and support in writing this paper. I would also like to thank SEAI for providing me with access to the raw data to perform this research.This research was conducted as part of my employment at Sustainable Energy Authority of Ireland, where I received funding and support for this work. Data Availability The data can be provided on request References Ayompe, L. (2011) Performance and Policy Evaluation of Solar Energy Technologies for Domestic Application in Ireland. Doctoral Thesis. Technological University Dublin. https://doi.org/10.21427/D7SW4T Barbose, G. L., Darghouth, N. R., Hoen, B., & Wiser, R. H. (2018). Income Trends of Residential PV Adopters: An analysis of household-level income estimates . https://doi.org/10.2172/1433126 Beppler, R. C., Matisoff, D. C., & Oliver, M. E. (2021). Electricity consumption changes following solar adoption: Testing for a solar rebound. Economic Inquiry , 61 (1), 58–81. https://doi.org/10.1111/ecin.13031 Bergman, N., Hawkes, A., Brett, D. J. L., Baker, P., Barton, J., Blanchard, R., Brandon, N. P., Infield, D., Jardine, C., Kelly, N., Leach, M., Matian, M., Peacock, A. D., Staffell, I., Sudtharalingam, S., & Woodman, B. (2009). UK microgeneration. Part I: policy and behavioural aspects. Proceedings of the Institution of Civil Engineers - Energy , 162 (1), 23–36. https://doi.org/10.1680/ener.2009.162.1.23 Department of the Environment, Climate and Communications DECC. (2024). Climate action plan 2024. Government of Ireland. Retrieved March 10, 2025, from https://assets.gov.ie/296414/7a06bae1-4c1c-4cdc-ac36-978e3119362e.pdf Cox, D. R. (1972). Regression models and Life-Tables. Journal of the Royal Statistical Society Series B (Statistical Methodology) , 34 (2), 187–202. https://doi.org/10.1111/j.2517-6161.1972.tb00899.x Department of Energy and Climate Change (DECC) (2010a). Feed-in Tariffs Government’s Response to the Summer 2009. Consultation. Tech. Rep. Building Britain’s Future; Act on CO2. Retrieved March 10, 2025, from https://www.fitariffs.co.uk/library/regulation/100201FinalDesign.pdf Department of Public Expenditure and Reform. (2021). National development plan 2021-2030. Irish Government. Retrieved March 3, 2025, from https://assets.gov.ie/200358/a36dd274-736c-4d04-8879-b158e8b95029.pdf Dobbyn, J., & Thomas, G. (2005). Seeing the light: The impact of micro-generation on our use of energy. Sustainable Consumption Roundtable . London. Frances, Z., & Stevenson, F. (2019). A relational approach to understanding inhabitants’ engagement with Photovoltaic (PV) technology in homes. Architectural Science Review , 1–13. https://doi.org/10.1080/00038628.2019.1682962 Graziano, M., & Gillingham, K. (2014). Spatial patterns of solar photovoltaic system adoption: The influence of neighbors and the built environment. Journal of Economic Geography , 15 (4), 815–839. https://doi.org/10.1093/jeg/lbu036 Haas, R., Ornetzeder, M., Hametner, K., Wroblewski, A., & Hübner, M. (1999). SOCIO-ECONOMIC ASPECTS OF THE AUSTRIAN 200 kWp-PHOTOVOLTAIC-ROOFTOP PROGRAMME. Solar Energy , 66 (3), 183–191. https://doi.org/10.1016/s0038-092x(99)00019-5 Ishwaran, H., Kogalur, U. B., Blackstone, E. H., & Lauer, M. S. (2008). Random survival forests. The Annals of Applied Statistics , 2 (3). https://doi.org/10.1214/08-aoas169 Kaplan, E. L., & Meier, P. (1958). Nonparametric Estimation from Incomplete Observations. Journal of the American Statistical Association , 53 (282), 457–481. https://doi.org/10.1080/01621459.1958.10501452 Kastner, I., & Stern, P. C. (2015). Examining the decision-making processes behind household energy investments: A review. Energy Research & Social Science , 10, 72–89. https://doi.org/10.1016/j.erss.2015.07.008 Kormos, C., & Gifford, R. (2014). The validity of self-report measures of proenvironmental behavior: A meta-analytic review. Journal of Environmental Psychology, 40, 359–371. https://doi.org/10.1016/j.jenvp.2014.09.003 Lukanov, B. R., & Krieger, E. M. (2019). Distributed solar and environmental justice: Exploring the demographic and socio-economic trends of residential PV adoption in California. Energy Policy , 134 , 110935. https://doi.org/10.1016/j.enpol.2019.110935 Palm, J., Eidenskog, M., & Luthander, R. (2017). Sufficiency, change, and flexibility: Critically examining the energy consumption profiles of solar PV prosumers in Sweden. Energy Research & Social Science , 39 , 12–18. https://doi.org/10.1016/j.erss.2017.10.006 Ryan, L., Wheatley, J., & Saba, N. (2023). A review of policies for the rollout of rooftop solar PV in Ireland . Climate Change Advisory Council. Retrieved March 3, 2025, from https://www.climatecouncil.ie/councilpublications/councilworkingpaperseries/WORKIN~2.PDF SEAI. (2024). Energy in Ireland 2024 Report. Sustainable Energy Authority of Ireland. Retrieved March 7, 2025, from https://www.seai.ie/sites/default/files/publications/energy-in-ireland-2024.pdf Footnotes Note that not all SEAI programs record completion date. In these cases, grant payment date was used instead. Note that homeowners could have installed energy efficient measures outside of SEAI programs, but these data are not available. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 07 Jan, 2026 Read the published version in Energy Efficiency → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6330919","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":456691988,"identity":"867aafd1-858f-4c72-bdf5-690c5c895b8a","order_by":0,"name":"Marek Bohacek","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA30lEQVRIiWNgGAWjYHACZgbGBoYEMOMDkGQDcYnWwjiDZC3MPMS4Srf97GMDxh02efzthw8b2/yxkedjYG58gE+L2Zl04wTGM2nFEmfSkpNz29IM2xgYmw3wajmQxnyAse1wYsMNHuPDuQ2HGYFa2iTwajn/DKJl/g3+z4ct/vy3B2pp/4FXy4005gSQlg03eJiTGdgOJIJswacDqOUZs0HimbTEjWfSjA1725KT25gZmwk4LI1Z4uMOm8R5xw8/lvjxx852fnv7ww94rQGBBBQeM0H1o2AUjIJRMAoIAQBNDkpzSV2HGQAAAABJRU5ErkJggg==","orcid":"","institution":"Sustainable Energy Authority of Ireland","correspondingAuthor":true,"prefix":"","firstName":"Marek","middleName":"","lastName":"Bohacek","suffix":""}],"badges":[],"createdAt":"2025-03-28 22:08:09","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6330919/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6330919/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s12053-025-10402-5","type":"published","date":"2026-01-07T15:58:57+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":82916487,"identity":"7d0796b1-de2c-413b-afb4-433274ca4d9f","added_by":"auto","created_at":"2025-05-16 16:22:25","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":61621,"visible":true,"origin":"","legend":"\u003cp\u003eGraph of monthly completion of solar PV installs since inception of the solar PV programme.\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6330919/v1/0dd960814da38fde831bde43.jpg"},{"id":82917171,"identity":"dfbae338-4c04-4f09-b96a-486580deb16e","added_by":"auto","created_at":"2025-05-16 16:30:25","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":85362,"visible":true,"origin":"","legend":"\u003cp\u003eKaplan Meyer plot showing probability of Measure being followed by additional retrofit over time (All Data).\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6330919/v1/5e9b1fd1ea8efbf26c706e43.jpg"},{"id":82916489,"identity":"b9e20087-08a2-4168-82a0-f9d659ce105d","added_by":"auto","created_at":"2025-05-16 16:22:25","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":108518,"visible":true,"origin":"","legend":"\u003cp\u003eKaplan Meyer plot showing probability of Measure being followed by additional retrofit over time (Partial Data).\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6330919/v1/937ff560ac3df1d64413e708.jpg"},{"id":82917172,"identity":"4597e498-af3e-4293-b1dc-9091e7cadb90","added_by":"auto","created_at":"2025-05-16 16:30:25","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":61472,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of predicted probabilities for solar PV by Kaplan Meyer plot, Cox model and random forest model (All data).\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6330919/v1/20921655abb496bab8394bc1.jpg"},{"id":82916494,"identity":"f7f0c213-272a-4bbc-acd1-75e8033d807c","added_by":"auto","created_at":"2025-05-16 16:22:25","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":65564,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of predicted probabilities for solar PV by Kaplan Meyer plot, Cox model and random forest model (Partial data).\u003c/p\u003e","description":"","filename":"5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-6330919/v1/1866aa0637f4e6f567598783.jpg"},{"id":100071107,"identity":"7e08d122-4881-4977-ba91-139c1bbee0a2","added_by":"auto","created_at":"2026-01-12 16:19:06","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1106504,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6330919/v1/1c57c8df-529e-49ee-9868-2314772b048a.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eAnalyzing solar installations: a catalyst or barrier to subsequent residential retrofits\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe urgent need to address climate change has driven a global transition toward renewable energy, with solar photovoltaics (PV) emerging as a key solution. In Ireland, the National Development Plan 2021\u0026ndash;2030 (Department of Public Expenditure and Reform, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) highlights the importance of solar PV in achieving the ambitious target of generating 70% of electricity from renewable sources by 2030, making solar energy a cornerstone of the country's sustainability strategy. Supportive policies, including feed-in tariffs and grants, have facilitated residential solar PV installations. A study in the UK found that feed-in tariffs significantly boosted the popularity of residential PV systems (Frances and Stevenson, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), a trend mirrored in Ireland with rising application rates (SEAI, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Additionally, a social contagion effect may drive the adoption of solar PV, as individuals and communities influence one another, motivated by social norms, visibility, and peer effects (Graziano and Gillingham, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDespite the increasing interest in solar PV, several critical issues remain. Notably, the timing of solar electricity generation often does not align with residential energy demand, which typically peaks in the evening when solar output is low. While solar PV can reduce electricity bills and generate income through feed-in tariffs, it does not inherently improve a home's overall energy efficiency. Ideally, solar PV should be the final component in a series of energy efficiency measures (Haas et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e1999\u003c/span\u003e), a consideration often overlooked in Ireland. In the UK, a dwelling qualifies for a solar PV grant only after achieving sufficient insulation in its envelope (Frances and Stevenson, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). This raises concerns about the opportunity cost of investing in solar PV instead of more impactful energy efficiency measures, such as insulation and heating system retrofits. Financial returns on residential solar PV investments in Ireland and the UK are estimated to extend beyond 20 years (Bergman et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Ayompe, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Even a recent Irish study, which accounts for feed-in tariffs and Sustainable Energy Authority of Ireland (SEAI) grants under ideal conditions (e.g., south-facing installation, no shading, clean panels), suggests a return on investment of at least 10 years (Ryan, Wheatley, and Saba, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Furthermore, the demographic disparity in solar PV adoption is noteworthy, as installations are primarily undertaken by wealthier households (Lukanov and Krieger, 2020; Barbose et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), likely due to the high initial costs.\u003c/p\u003e \u003cp\u003eResearch into the secondary benefits of solar PV indicates that homeowners often report lower overall electricity consumption as they become more efficient energy users (Frances and Stevenson, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) (DECC, 2010a). However, Haas et al. (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e1999\u003c/span\u003e) identified a more complex relationship: while larger energy consumers reduced their consumption after installing PV, smaller consumers tended to increase it. A Swedish study examining electricity consumption one year before and after solar PV installation found no significant change (Palm, Eidenskog, and Luthander, 2018), while Bepler, Matisoff, and Oliver (2018) reported a 28.5% increase in consumption following installation. Bergman et al. (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2009\u003c/span\u003e) attributes these variations to the differences between habitual and deliberative behaviours, suggesting that changes in electricity usage often fall under habitual behaviour, which is challenging to modify.\u003c/p\u003e \u003cp\u003eDespite these concerns, proponents argue that, due to its relative ease of installation, solar PV may serve as a \"gateway to retrofit\", encouraging homeowners to pursue further energy efficiency measures. Those who install solar PV often report increased environmental motivation (Kastner and Stern, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), potentially leading to changes in deliberative behaviours, such as implementing additional energy efficiency measures (Bergman et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). Dobbyn and Thomas (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2005\u003c/span\u003e) provide evidence of shifts in awareness, behaviour, and attitudes among individuals that gained microgeneration experience.\u003c/p\u003e \u003cp\u003eThis study aims to empirically evaluate whether the installation of solar PV encourages further energy retrofitting, potentially establishing solar PV as a \"gateway to retrofit\".\u003c/p\u003e"},{"header":"Methodology","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eData sources\u003c/h2\u003e \u003cp\u003eData on the installation of solar PV and other energy efficiency measures were collated from four active SEAI grant programmes (Better Energy Homes, Better Energy Communities, One Stop Shop and Solar PV) as well as two older programmes (Deep Retrofit and One Stop Shop Pilot). In addition, data on demographics, socioeconomic conditions and indicators related to community wellbeing were sourced from Central Statistical Office. This resulted in a dataset providing information on the retrofitting journey of Irish dwellings over time, where dwellings with multiple retrofits could be easily identified.\u003c/p\u003e \u003c/div\u003e\u003ch2\u003eAnalysis approach\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eThe data described above were used to categorise dwellings that have solar PV installed into the following four categories:\u0026nbsp;\u003c/p\u003e\n\u003col start=\"1\" type=\"1\"\u003e\n \u003cli\u003e\u003cstrong\u003eSolar PV followed by retrofit\u003c/strong\u003e. This category represents dwellings that first installed solar PV and later installed additional retrofit measures.\u0026nbsp;\u003c/li\u003e\n\u003c/ol\u003e\n\u003col start=\"2\" type=\"1\"\u003e\n \u003cli\u003e\u003cstrong\u003eRetrofit followed by solar PV.\u003c/strong\u003e This category represents dwellings fitting the opposite pattern whereby another energy efficiency measure was installed prior to solar PV installation.\u0026nbsp;\u003c/li\u003e\n\u003c/ol\u003e\n\u003col start=\"3\" type=\"1\"\u003e\n \u003cli\u003e\u003cstrong\u003eSolar PV and retrofit at the same time.\u003c/strong\u003e This category represents dwellings that installed solar PV and other energy efficiency measure/s at roughly the same time \u0026ndash; the second retrofit measure was applied for between first measure\u0026rsquo;s application date and completion date.\u003csup\u003e[1]\u0026nbsp;\u003c/sup\u003e\u003c/li\u003e\n\u003c/ol\u003e\n\u003col start=\"4\" type=\"1\"\u003e\n \u003cli\u003e\u003cstrong\u003eSolar PV only.\u003c/strong\u003e This category represents dwellings that only installed solar PV and did not apply for grants for any other energy efficiency measures.\u003csup\u003e[2]\u003c/sup\u003e\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eThe above classification allows us to observe the frequency with which solar PV is followed by other retrofits but does not compare this against a counterfactual. For this reason, two counterfactual measures of attic insulation/roof insulation and heating controls were also chosen for comparison. The reasons for choosing attic insulation and heating controls as counterfactuals are threefold. First, they are common retrofit measures, giving enough data points for statistical analysis. Second, both of these measures are suitable for most of Irish dwellings. Third, they are relatively inexpensive to install compared with solar PV and can be considered \u0026ldquo;entry level\u0026rdquo; retrofits (see Table 1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 1: Mean and median installation cost for the most common measures over the last two years.\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"538\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 38.4758%;\"\u003e\n \u003cp\u003eMeasure\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 29.9257%;\"\u003e\n \u003cp\u003eMean cost of works\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31.5985%;\"\u003e\n \u003cp\u003eMedian cost of works\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 38.4758%;\"\u003e\n \u003cp\u003eAttic insulation/roof insulation\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 29.9257%;\"\u003e\n \u003cp\u003e2,623 \u0026euro;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31.5985%;\"\u003e\n \u003cp\u003e2,150 \u0026euro;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 38.4758%;\"\u003e\n \u003cp\u003eSolar PV\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 29.9257%;\"\u003e\n \u003cp\u003e13,299 \u0026euro;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31.5985%;\"\u003e\n \u003cp\u003e11,100 \u0026euro;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 38.4758%;\"\u003e\n \u003cp\u003eHeating controls\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 29.9257%;\"\u003e\n \u003cp\u003e4,363 \u0026euro;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 31.5985%;\"\u003e\n \u003cp\u003e3,750 \u0026euro;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eBoth attic insulation and heating controls were grouped into four categories in the same fashion as solar PV as described above. This allows for direct comparison of the likelihood solar PV installation is followed by further measures against the counterfactuals of attic insulation or heating controls being followed by retrofit. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;After data cleaning and categorization, statistical analyses were conducted using multiple methods. Initial chi-square tests provided a straightforward interpretation of associations. Subsequently, survival analysis modelling was employed, focusing on time-to-event data to understand the influence of covariates on the timing of events, such as applications for additional retrofits. The Kaplan-Meier estimator (Kaplan and Meier, 1958) was utilized for non-parametric estimation of the time-to-event function, allowing for visual representation of event probabilities over time and group comparisons via log-rank tests. The Cox proportional hazards model (Cox, 1972) further examined the relationship between covariates and the hazard rate, providing hazard ratios to quantify the effects of predictors like dwelling size and deprivation index. Additionally, random survival forests, as proposed by Ishwaran et al. (2008), leveraged an ensemble of decision trees to capture complex interactions and nonlinear relationships in censored time-to-event data.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eDescriptive statistics\u003c/h2\u003e \u003cp\u003eA total of 65,110 solar PV installations have been completed since the beginning of SEAI\u0026rsquo;s grant programme of which 1,361 were later followed by additional retrofit projects, totalling 1,839 additional measures (see Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The cumulative investment in these additional retrofits has reached approximately \u0026euro;13.0\u0026nbsp;million, indicating an average of \u0026euro;200 in further retrofits per solar PV installation.\u003c/p\u003e \u003cp\u003eAs the SEAI PV grant has been present for 6 years, only the last 6 years of attic insulation and heating controls data are reported below for comparability.\u003c/p\u003e \u003cp\u003eA total of 33,935 attic insulation installations were finalized during the same period. Among these, 1,063 projects were followed by subsequent retrofits, totalling 1,196 measures and amounting to a total cost of \u0026euro;11.9\u0026nbsp;million or \u0026euro;350 in additional retrofits per attic insulation installation\u003c/p\u003e \u003cp\u003eLastly, there were 24,225 installations of heating controls completed. In total 1,866 of these heating controls resulted in further retrofitting activities amounting to 2,393 measures, generating a total investment of \u0026euro;18.0\u0026nbsp;million. On average, each heating controls installation has led to an additional \u0026euro;740 in retrofits. For detail breakdown of measures following Heating controls installations see Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, Heating controls column.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBreakdown of number of measures installed after initial retrofit shown for solar PV, attic insulation and heating controls over the last 6 years.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003eInitial retrofit\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eType of measure installed at future date following the initial retrofit\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003esolar PV\u003c/p\u003e \u003cp\u003e(65,110)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAttic insulation\u003c/p\u003e \u003cp\u003e(33,935)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eHeating controls\u003c/p\u003e \u003cp\u003e(24,225)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eExternal wall insulation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e178\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e124\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e123\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInternal wall insulation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e23\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCavity wall insulation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e487\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e176\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e464\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHeat pump\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e320\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e28\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDoors\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e39\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWindows\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOther measures\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSolar PV\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e664\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1004\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAttic insulation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e641\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e719\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHeating controls\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e105\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1839\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1196\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2393\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e\u003c/h2\u003e \u003cp\u003eThe descriptive statistics above show that solar PV was followed by additional retrofit the least when compared with attic insulation and especially with heating controls. However, even though there have been a large number of solar PV, attic insulations and heating controls installed over the last 6 years, the level of subsequent retrofit is very small (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The amount of additional retrofitting happening at the same time is substantially larger, especially in the case of attic insulation (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Nonetheless, there are number of issues when looking at the descriptive data in this way. Firstly, it is unclear if solar PV is more or less likely to be followed by a retrofit at future date than attic insulation or heating controls. Secondly, those descriptives look at all the solar PV, attic insulation and heating controls installs in the last 6 years and for many of those, especially the recent ones, there hasn\u0026rsquo;t been enough time to observe additional retrofit that might happen sometime in the future. This is especially problematic for solar PV which saw large increase in applications in the last 3 years (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). To address those shortcomings a series of chi-square tests was conducted next.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBreakdown of number of measures installed together (at the \u0026ldquo;same time\u0026rdquo;) with initial retrofit shown for solar PV, Attic insulation and Heating controls during last 6 years.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003eInitial retrofit\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNumber of Measures installed with initial retrofit (solar PV/attic insulation/heating controls)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003esolar PV\u003c/p\u003e \u003cp\u003e(65,110)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAttic insulation\u003c/p\u003e \u003cp\u003e(33,935)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eHeating controls\u003c/p\u003e \u003cp\u003e(24,225)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eExternal wall insulation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e751\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2,993\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e507\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInternal wall insulation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e209\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1,866\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e356\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCavity wall insulation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1,179\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13,817\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e395\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHeat pump\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1,722\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3,266\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDoors\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e863\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1,747\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e43\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWindows\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e808\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1,712\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e43\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOther measures\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e211\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e693\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003esolar PV\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1,740\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e219\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAttic insulation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2,030\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,246\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHeating controls\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e508\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1,933\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8,281\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e29,767\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3,084\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e\u003ch2\u003eChi-square test results\u0026nbsp;\u003c/h2\u003e\n\u003cp\u003eTo address the issue of unobserved future retrofits, the data were coded based on the number of retrofits occurring within a two-year period following the completion of a solar PV installation, attic insulation, or heating controls (collectively referred to as \u0026quot;Measure\u0026quot;). Data from the most recent two years were excluded from the analysis, as it remains uncertain whether additional retrofits will occur following installation made during that time. A two-year period was chosen to balance having enough time between the completion of the measure and any subsequent retrofits, while minimizing the exclusion of relevant data. This adjustment effectively reduces the available data from six years to four.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe category \u0026quot;retrofit followed by measure\u0026quot; follows the same two-year rule, but to avoid excessive data loss, retrofits from as far back as eight years are considered. The following rules apply:\u0026nbsp;\u003c/p\u003e\n\u003col start=\"1\" type=\"1\"\u003e\n \u003cli\u003eA retrofit is counted as being followed by a Measure only if it occurred within two years after the Measure\u0026apos;s completion and no retrofit occurred in the two years preceding the Measure.\u0026nbsp;\u003c/li\u003e\n\u003c/ol\u003e\n\u003col start=\"2\" type=\"1\"\u003e\n \u003cli\u003eA Measure is counted as being followed by a retrofit only if it occurred within two years after the completion of the retrofit.\u0026nbsp;\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eThree sets of chi-square tests were conducted:\u0026nbsp;\u003c/p\u003e\n\u003col start=\"1\" type=\"1\"\u003e\n \u003cli\u003eThe first set of tests compared all four categories: \u0026ldquo;Measure is followed by retrofit,\u0026rdquo; \u0026ldquo;Retrofit is followed by Measure,\u0026rdquo; \u0026ldquo;Measure and retrofit at the same time,\u0026rdquo; and \u0026ldquo;Measure only.\u0026rdquo; The findings indicate that both solar PV vs attic insulation (X-squared = 16501, df = 3, p-value \u0026lt; 2.2e-16, Cramer\u0026rsquo;s V = 0.38) and solar PV vs heating controls (X-squared = 261, df = 3, p-value \u0026lt; 2.2e-16, Cramer\u0026rsquo;s V = 0.044) were significantly different in four categories (Table 4). However, it appears that the difference is mostly driven by \u0026ldquo;Measure and retrofit at the same time\u0026rdquo; category.\u0026nbsp;\u003c/li\u003e\n\u003c/ol\u003e\n\u003col start=\"2\" type=\"1\"\u003e\n \u003cli\u003eThe second set of tests recoded the data into two categories: \u0026ldquo;Measure is followed by retrofit\u0026rdquo; and \u0026ldquo;Measure is not followed by retrofit,\u0026rdquo; the latter of which includes \u0026ldquo;Retrofit is followed by Measure,\u0026rdquo; \u0026ldquo;Measure and retrofit at the same time,\u0026rdquo; and \u0026ldquo;Measure only\u0026rdquo;. This test directly assesses whether a difference in the probability of additional retrofits exists between solar PV and attic insulation/heating controls. The findings found solar PV vs attic insulation (X-squared = 1.54, df = 1, p-value \u0026lt; 0.21, Cramer\u0026rsquo;s V = 0.006) and solar PV vs heating controls (X-squared = 0.75, df = 1, p-value = 0.39, Cramer\u0026rsquo;s V = 0.004) were not significant.\u0026nbsp;\u003c/li\u003e\n\u003c/ol\u003e\n\u003col start=\"3\" type=\"1\"\u003e\n \u003cli\u003eThe third set of tests excluded the categories \u0026ldquo;Retrofit is followed by Measure\u0026rdquo; and \u0026ldquo;Measure and retrofit at the same time,\u0026rdquo; resulting in a dataset containing only \u0026ldquo;Measure is followed by retrofit\u0026rdquo; and \u0026ldquo;Measure only\u0026rdquo;. It was found that solar PV vs Attic insulation (X-squared = 627.94, df = 1, p-value \u0026lt; 2.2e-16, Cramer\u0026rsquo;s V = 0.091) was significant indicating that Attic insulation was more likely to be followed by retrofit than solar PV when multi-measure applications and \u0026ldquo;Retrofit is followed by Measure\u0026rdquo; of solar PV/Attic insulation are excluded. Solar PV vs Heating controls (X-squared = 0.65, df = 1, p-value = 0.42, Cramer\u0026rsquo;s V = 0.004) found no significant difference indicating that solar PV was no more likely to be followed by retrofit than Heating controls.\u0026nbsp;\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003eTable 4: Probabilities of a measure (solar PV/attic insulation/heating controls) being followed by retrofit in next 2 years displayed by 4 categories. Last 2 years of data are excluded to allow for two-year window to see if additional retrofit materializes.\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"623\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width: 123px;\"\u003e\n \u003cp\u003eInitial Measure\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"4\" style=\"width: 501px;\"\u003e\n \u003cp\u003eProbabilities of:\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003eMeasure only\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 139px;\"\u003e\n \u003cp\u003eMeasure and retrofit at the same time\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eRetrofit is followed by Measure\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003eMeasure is followed by retrofit\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 123px;\"\u003e\n \u003cp\u003eSolar PV\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e87.2%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 139px;\"\u003e\n \u003cp\u003e5.3%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003e4.3%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003e3.2%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 123px;\"\u003e\n \u003cp\u003eAttic insulation\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e28.0%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 139px;\"\u003e\n \u003cp\u003e65.0%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003e4.0%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003e3.0%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 123px;\"\u003e\n \u003cp\u003eHeating controls\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003e87.1%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 139px;\"\u003e\n \u003cp\u003e7.6%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003e2.2%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 142px;\"\u003e\n \u003cp\u003e3.1%\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eThe findings from the chi-square tests indicate that solar PV is no more likely to be followed by retrofit than attic insulation or heating controls. If anything, Attic insulation might be more likely followed by retrofit than solar PV when looking only at first-time standalone Measures. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eHowever, there are number of limitations. The last two years of data were excluded to allow enough time to observe whether additional retrofit materializes or not. This is of particular concern for solar PV as majority of solar PV applications were only completed in last two years (see Figure 1) and hence many datapoints from solar PV were excluded. Also, this two-year window is limiting as it does not consider retrofits that happened more than two years after the initial measure. Moreover, the chi-square test does not allow to control for covariates such as dwelling type, dwelling size, deprivation index, Building energy rating, etc. To address these limitations survival analysis was explored next.\u0026nbsp;\u003c/p\u003e\n\u003ch1\u003eSurvival analysis model results\u0026nbsp;\u003c/h1\u003e\n\u003cp\u003eThis section presents the results of the survival analysis, which utilized a comprehensive dataset that includes 6 years of data for solar PV and 16 years for attic insulation and heating controls. The analysis employed two sets of survival models to assess the likelihood of additional retrofits following initial measures:\u0026nbsp;\u003c/p\u003e\n\u003col start=\"1\" type=\"1\"\u003e\n \u003cli\u003eUsing the full dataset, containing 160k attic insulations, 101k heating controls, and 65k solar PV installations.\u0026nbsp;\u003c/li\u003e\n\u003c/ol\u003e\n\u003col start=\"2\" type=\"1\"\u003e\n \u003cli\u003eUsing a partial dataset from which the \u0026ldquo;Measure and retrofit at the same time\u0026rdquo; and \u0026ldquo;Retrofit is followed by Measure\u0026rdquo; categories were excluded, resulting in 15.7k attic insulations, 71.4k heating controls, and 44.4k solar PV installations. The purpose of this was to investigate whether solar PV installation is likely to be followed by subsequent retrofitting where it is a first-time, standalone measure.\u0026nbsp;\u003c/li\u003e\n\u003c/ol\u003e\n\u003ch2\u003eKaplan-Meier plots \u0026nbsp;\u003c/h2\u003e\n\u003ch3\u003eFull dataset\u0026nbsp;\u003c/h3\u003e\n\u003cp\u003eAttic insulation is least likely to be followed by retrofit compared to solar PV and heating controls (Figure 2). Attic insulation shows a rapid initial uptake with minimal follow-up retrofits after two years, while solar PV and heating controls see a steady increase in follow-up retrofits showing almost a linear trend. However, the level of additional retrofit is limited to about 1% per year for solar PV and heating controls.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePartial dataset\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWhen considering only first-time standalone measures, attic insulation and heating controls are both more likely to be followed by retrofit compared to solar PV (Figure 3). There is a dramatic increase in the likelihood of attic insulation being followed by additional retrofits, resulting from the exclusion of a large number (about 80%) of attic insulation installations for this partial dataset. Attic insulation is primarily part of more extensive retrofits including multiple measures. The level of additional retrofit increased slightly but is still limited to about 1.5% per year for solar PV and heating controls.\u0026nbsp;\u003c/p\u003e\n\u003ch2\u003eCox Model\u0026nbsp;\u003c/h2\u003e\n\u003ch3\u003eFull dataset\u0026nbsp;\u003c/h3\u003e\n\u003cp\u003eThe findings from the Cox model analysis indicate that all covariates are statistically significant due to the large size of the dataset (Table 5). It shows that solar PV is more likely to be followed by retrofit than attic insulation but less likely than heating controls, which aligns with the Kaplan-Meier plot. Moreover, a better BER after the initial measure increases the likelihood of applying for additional retrofits. Additionally, apartments and mid-terraced dwellings are less likely to seek additional retrofits than detached and semi-detached houses. The analysis also suggests that more affluent neighbourhoods and larger homes are more inclined to apply for subsequent retrofits, although the effect sizes are small as indicated by hazard ratios (exp(coef)).\u0026nbsp;\u003c/p\u003e\n\u003ch3\u003ePartial dataset\u0026nbsp;\u003c/h3\u003e\n\u003cp\u003eWhen considering only first-time standalone measures, attic insulation and heating controls both have a higher probability of additional retrofit when compared to solar PV (Table 6), controlling for covariates. Dwelling size and the deprivation index have a smaller influence when compared to full dataset, but other covariates follow the same pattern.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 5: Output from Cox model (All Data). Total number of datapoints \u0026ndash; 330,680 with 13,726 of events (initial Measure being followed by a retrofit).\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eCovariate\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003ecoefficient\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003eexp(coef)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003ese(coef)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003ez\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003ePr(\u0026gt;|z|)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eAttic insulation\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.8848\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.4127\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0402\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e-21.983\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eHeating Controls\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e1.0638\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e2.8975\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0356\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e29.872\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eDeprivation index\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e0.0236\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e1.0239\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0029\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e7.974\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e1.54e-15\u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eDwelling size\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e0.1526\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e1.1648\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0276\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e5.516\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e3.46e-08\u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eSemi-detached\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e0.1082\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e1.1143\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0209\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e5.156\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e2.52e-13\u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eMid-terraced\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.2678\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.7650\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0372\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e-7.192\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e6.40e-13\u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eApartment\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.5259\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.5909\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0710\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e-7.408\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e1.29e-13\u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eUnknown dwelling\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-1.8973\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.1499\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0650\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e-29.160\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eBER A\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e1.0214\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e2.7772\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0462\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e22.076\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eBER B\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e0.6277\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e1.8734\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0214\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e29.314\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eBER D\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.2300\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.7944\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0303\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e-7.571\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e3.71e-13\u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eBER EFG\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.2038\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.8156\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0498\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e-4.086\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e4.39e-13\u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eBER Unknown\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e0.2087\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e1.2321\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0313\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e6.657\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e2.79e-13\u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eTable 6: Output from Cox model (Partial Data). Total number of datapoints \u0026ndash; 131,484 with 13,726 of events (initial Measure being followed by a retrofit).\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eCovariate\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003ecoefficient\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003eexp(coef)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003ese(coef)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003ez\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003ePr(\u0026gt;|z|)\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eAttic insulation\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e2.1863\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e8.9026\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0403\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e54.202\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eHeating Controls\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e1.2125\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e3.3619\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0372\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e32.588\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eDeprivation index\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e0.0059\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e1.0059\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0029\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e1.991\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0464\u003csup\u003e*\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eDwelling size\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e0.0774\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e1.0805\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0282\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e2.746\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0060\u003csup\u003e**\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eSemi-detached\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.1748\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.8396\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0212\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e-8.229\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eMid-terraced\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.6613\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.5161\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0374\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e-17.646\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eApartment\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-1.1039\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.3315\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0711\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e-15.506\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eUnknown dwelling\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-2.1529\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.1161\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0651\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e-33.061\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eBER A\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e1.3854\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e3.9964\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0477\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e28.993\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eBER B\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e0.9165\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e2.5007\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0216\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e42.337\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eBER D\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.4190\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.6576\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0304\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e-13.757\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eBER EFG\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e-0.7156\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.4888\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0502\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e-14.244\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 132px;\"\u003e\n \u003cp\u003eBER Unknown\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94px;\"\u003e\n \u003cp\u003e0.2110\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e1.2350\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e0.0323\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e6.521\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003e\u0026lt; 2e-16 \u003csup\u003e***\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003ch2\u003eRandom Forest Model\u0026nbsp;\u003c/h2\u003e\n\u003ch3\u003eFull dataset\u0026nbsp;\u003c/h3\u003e\n\u003cp\u003eThe random forest model does not generate regression outputs or dummy variables for categorical covariates, but it provides variable importance metrics (Table 7) that align closely with Cox model findings. A comparison of models suggest that the Kaplan-Meier and random forest models yield similar results, while the Cox model consistently shows a somewhat lower probability of additional retrofit.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 7: Variable importance as produced by random forest model (all dataset).\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"378\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 199px;\"\u003e\n \u003cp\u003eVariable\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 179px;\"\u003e\n \u003cp\u003eImportance\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 199px;\"\u003e\n \u003cp\u003eType of measure\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 179px;\"\u003e\n \u003cp\u003e0.0935\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 199px;\"\u003e\n \u003cp\u003eBER following measure\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 179px;\"\u003e\n \u003cp\u003e0.0207\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 199px;\"\u003e\n \u003cp\u003eDwelling type\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 179px;\"\u003e\n \u003cp\u003e0.0196\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 199px;\"\u003e\n \u003cp\u003eDwelling size\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 179px;\"\u003e\n \u003cp\u003e0.0152\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 199px;\"\u003e\n \u003cp\u003eDeprivation index\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 179px;\"\u003e\n \u003cp\u003e0.0039\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003ch3\u003ePartial dataset\u0026nbsp;\u003c/h3\u003e\n\u003cp\u003eThe pattern of variable importance metrics (Table 8) for the partial dataset is comparable to full dataset findings. A comparison of models show that all three models are predicting very similarly for the first five years since following installation of the initial measure, diverging later.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 8: Variable importance as produced by random forest model (Partial data).\u0026nbsp;\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"378\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 189px;\"\u003e\n \u003cp\u003eVariable\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 189px;\"\u003e\n \u003cp\u003eImportance\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 189px;\"\u003e\n \u003cp\u003eType of measure\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 189px;\"\u003e\n \u003cp\u003e0.0679\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 189px;\"\u003e\n \u003cp\u003eBER following measure\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 189px;\"\u003e\n \u003cp\u003e0.0549\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 189px;\"\u003e\n \u003cp\u003eDwelling type\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 189px;\"\u003e\n \u003cp\u003e0.0347\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 189px;\"\u003e\n \u003cp\u003eDwelling size\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 189px;\"\u003e\n \u003cp\u003e0.0085\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 189px;\"\u003e\n \u003cp\u003ePobal deprivation index\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 189px;\"\u003e\n \u003cp\u003e0.0045\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eThe survival analysis shows agreement across all three models suggesting that most important covariate affecting an increased probability of future retrofit is a dwelling\u0026rsquo;s BER. Dwellings with a better BER are more likely to retrofit again compared to dwellings with a poorer BER. The impacts of the deprivation index and size of dwelling are least important. The probability of additional retrofit in a given year following the initial measure for solar PV only is between 1% (Figure 4) based on full dataset analysis and 1.5% (Figure 5) as given by the partial dataset analysis. \u0026nbsp;\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis research explores the potential for solar photovoltaic (PV) installation to act as a catalyst for subsequent home retrofitting. The findings provide valuable insights into how the adoption of renewable energy technologies may influence broader residential energy retrofitting behaviours, particularly in Ireland, where the urgency around climate action necessitates transformative changes in home energy consumption patterns (DECC, 2024).\u003c/p\u003e \u003cp\u003eThe data indicate that subsequent retrofitting activities are minimal compared to the extensive installation of solar PV systems. Of the 65,000 solar PV installations, only 1,839 have applied for further retrofits. Findings from survival analysis reveal that only 1-1.5% of solar PV installations seek additional retrofits each year. Similar trends are observed in the cases of attic insulation and heating controls, where additional retrofits are also relatively uncommon albeit to lesser degree. Furthermore, the results of the survival analysis demonstrate almost linear hazard functions (indicating the probabilities of additional retrofits) over time, suggesting that these retrofits occur at nearly constant rates.\u003c/p\u003e \u003cp\u003eThe results from both chi-square tests and survival analysis indicate that solar PV does not inherently catalyse additional retrofitting to the same extent as other energy efficiency measures, such as attic insulation and heating controls. The tests reveal that heating controls are more likely to be followed by additional retrofits than solar PV, while attic insulation is more frequently installed as part of a package of measures. Reported increases in environmental motivations (Kastner and Stern, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) or shifts in awareness, behaviour, and attitudes (Bergman et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2009\u003c/span\u003e) following solar PV installation did not translate into additional retrofits. A meta-analysis by Kormos and Gifford (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) may provide an explanation, as it found significant discrepancies between self-reported pro-environmental behaviour and actual behaviour.\u003c/p\u003e \u003cp\u003eImportantly, solar PV installations could be counterproductive, diverting resources from more effective measures such as insulation or heating retrofits. However, it remains unclear whether solar PV competes directly with these other measures. It is uncertain if homeowners who choose to install solar PV ever consider other, more effective retrofits. For instance, solar PV might only be considered against other types of expenditures (e.g., vacations), making it a more desirable option than forgoing improvements altogether. Further research is needed to explore this possibility.\u003c/p\u003e \u003cp\u003eAdditionally, the research reveals that not only does the probability of an additional retrofit following a solar PV installation remain low, but significant differences are observed when controlling for variables such as dwelling type, size, and energy performance rating (BER). Detached and semi-detached dwellings with better BERs in more affluent neighbourhoods are correlated with higher additional retrofit probabilities, suggesting that energy efficiency investments may be more effectively pursued within specific dwelling contexts.\u003c/p\u003e \u003cp\u003eOne limitation of the study is the lack of information regarding retrofitting measures undertaken outside of the SEAI programs. This potentially underestimates the overall impact of solar PV installations on home energy upgrades by failing to account for homeowners who may have pursued concurrent or subsequent retrofits without governmental grants. It is also important to note that a decision for additional retrofitting cannot be directly attributed to initial PV installation, as it is unknown when that decision was made and what factors influenced it. For example, financial constraints may lead homeowners to stagger their energy efficiency investments rather than pursue multiple upgrades simultaneously.\u003c/p\u003e \u003cp\u003eThis paper highlights the dual nature of energy investments, emphasizing the need for a more comprehensive retrofit strategy that encompasses not only the adoption of renewable technology but also the necessity of improving the energy performance of dwellings through integrated energy efficiency measures. Without addressing the interplay between renewable energy adoption and home energy efficiency upgrades, the potential gains from solar technology may not be fully realized in reducing overall energy consumption and greenhouse gas emissions.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eWhile this study provides significant data and analysis regarding the potential role of solar PV in stimulating retrofit behaviours, it also underscores the complexity of transforming residential energy practices. The findings indicate that simply installing solar panels is insufficient for a substantial shift in energy efficiency uptake. For policymakers, these insights emphasize the importance of fostering an environment that encourages holistic approaches to energy retrofitting. This involves integrating financial incentives, public education, and targeted outreach to guide homeowners through the lifecycle of energy efficiency, from initial installations to meaningful retrofitting across the home. Such a multi-faceted approach will be essential for achieving Ireland's ambitious targets for renewable energy adoption and emission reductions in the coming years, ultimately contributing to a more sustainable future.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eM.B. did all\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eI would like to thank Hanna Julianne, programme manager of behavioural economics unit at Sustainable Energy Authority of Ireland (SEAI) for her comments and support in writing this paper. I would also like to thank SEAI for providing me with access to the raw data to perform this research.This research was conducted as part of my employment at Sustainable Energy Authority of Ireland, where I received funding and support for this work.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe data can be provided on request\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAyompe, L. (2011) \u003cem\u003ePerformance and Policy Evaluation of Solar Energy Technologies for Domestic Application in Ireland.\u003c/em\u003e Doctoral Thesis. Technological University Dublin. https://doi.org/10.21427/D7SW4T\u003c/li\u003e\n\u003cli\u003eBarbose, G. L., Darghouth, N. R., Hoen, B., \u0026amp; Wiser, R. H. (2018). \u003cem\u003eIncome Trends of Residential PV Adopters: An analysis of household-level income estimates\u003c/em\u003e. https://doi.org/10.2172/1433126\u003c/li\u003e\n\u003cli\u003eBeppler, R. C., Matisoff, D. C., \u0026amp; Oliver, M. E. (2021). Electricity consumption changes following solar adoption: Testing for a solar rebound. \u003cem\u003eEconomic Inquiry\u003c/em\u003e, \u003cem\u003e61\u003c/em\u003e(1), 58\u0026ndash;81. https://doi.org/10.1111/ecin.13031\u003c/li\u003e\n\u003cli\u003eBergman, N., Hawkes, A., Brett, D. J. L., Baker, P., Barton, J., Blanchard, R., Brandon, N. P., Infield, D., Jardine, C., Kelly, N., Leach, M., Matian, M., Peacock, A. D., Staffell, I., Sudtharalingam, S., \u0026amp; Woodman, B. (2009). UK microgeneration. Part I: policy and behavioural aspects. \u003cem\u003eProceedings of the Institution of Civil Engineers - Energy\u003c/em\u003e, \u003cem\u003e162\u003c/em\u003e(1), 23\u0026ndash;36. https://doi.org/10.1680/ener.2009.162.1.23\u003c/li\u003e\n\u003cli\u003eDepartment of the Environment, Climate and Communications DECC. (2024). Climate action plan 2024. Government of Ireland. Retrieved March 10, 2025, from https://assets.gov.ie/296414/7a06bae1-4c1c-4cdc-ac36-978e3119362e.pdf\u003c/li\u003e\n\u003cli\u003eCox, D. R. (1972). Regression models and Life-Tables. \u003cem\u003eJournal of the Royal Statistical Society Series B (Statistical Methodology)\u003c/em\u003e, \u003cem\u003e34\u003c/em\u003e(2), 187\u0026ndash;202. https://doi.org/10.1111/j.2517-6161.1972.tb00899.x\u003c/li\u003e\n\u003cli\u003eDepartment of Energy and Climate Change (DECC) (2010a). Feed-in Tariffs Government\u0026rsquo;s Response to the Summer 2009. Consultation. Tech. Rep. Building Britain\u0026rsquo;s Future; Act on CO2. Retrieved March 10, 2025, from https://www.fitariffs.co.uk/library/regulation/100201FinalDesign.pdf\u003c/li\u003e\n\u003cli\u003eDepartment of Public Expenditure and Reform. (2021). National development plan 2021-2030. Irish Government. Retrieved March 3, 2025, from https://assets.gov.ie/200358/a36dd274-736c-4d04-8879-b158e8b95029.pdf\u003c/li\u003e\n\u003cli\u003eDobbyn, J., \u0026amp; Thomas, G. (2005). Seeing the light: The impact of micro-generation on our use of energy. \u003cem\u003eSustainable Consumption Roundtable\u003c/em\u003e. London.\u003c/li\u003e\n\u003cli\u003eFrances, Z., \u0026amp; Stevenson, F. (2019). A relational approach to understanding inhabitants\u0026rsquo; engagement with Photovoltaic (PV) technology in homes. \u003cem\u003eArchitectural Science Review\u003c/em\u003e, 1\u0026ndash;13. https://doi.org/10.1080/00038628.2019.1682962\u003c/li\u003e\n\u003cli\u003eGraziano, M., \u0026amp; Gillingham, K. (2014). Spatial patterns of solar photovoltaic system adoption: The influence of neighbors and the built environment. \u003cem\u003eJournal of Economic Geography\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(4), 815\u0026ndash;839. https://doi.org/10.1093/jeg/lbu036\u003c/li\u003e\n\u003cli\u003eHaas, R., Ornetzeder, M., Hametner, K., Wroblewski, A., \u0026amp; H\u0026uuml;bner, M. (1999). SOCIO-ECONOMIC ASPECTS OF THE AUSTRIAN 200 kWp-PHOTOVOLTAIC-ROOFTOP PROGRAMME. \u003cem\u003eSolar Energy\u003c/em\u003e, \u003cem\u003e66\u003c/em\u003e(3), 183\u0026ndash;191. https://doi.org/10.1016/s0038-092x(99)00019-5\u003c/li\u003e\n\u003cli\u003eIshwaran, H., Kogalur, U. B., Blackstone, E. H., \u0026amp; Lauer, M. S. (2008). Random survival forests. \u003cem\u003eThe Annals of Applied Statistics\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e(3). https://doi.org/10.1214/08-aoas169\u003c/li\u003e\n\u003cli\u003eKaplan, E. L., \u0026amp; Meier, P. (1958). Nonparametric Estimation from Incomplete Observations. \u003cem\u003eJournal of the American Statistical Association\u003c/em\u003e, \u003cem\u003e53\u003c/em\u003e(282), 457\u0026ndash;481. https://doi.org/10.1080/01621459.1958.10501452\u003c/li\u003e\n\u003cli\u003eKastner, I., \u0026amp; Stern, P. C. (2015). Examining the decision-making processes behind household energy investments: A review. \u003cem\u003eEnergy Research \u0026amp; Social Science\u003c/em\u003e, 10, 72\u0026ndash;89. https://doi.org/10.1016/j.erss.2015.07.008\u003c/li\u003e\n\u003cli\u003eKormos, C., \u0026amp; Gifford, R. (2014). The validity of self-report measures of proenvironmental behavior: A meta-analytic review. Journal of Environmental Psychology, 40, 359\u0026ndash;371. https://doi.org/10.1016/j.jenvp.2014.09.003\u003c/li\u003e\n\u003cli\u003eLukanov, B. R., \u0026amp; Krieger, E. M. (2019). Distributed solar and environmental justice: Exploring the demographic and socio-economic trends of residential PV adoption in California. \u003cem\u003eEnergy Policy\u003c/em\u003e, \u003cem\u003e134\u003c/em\u003e, 110935. https://doi.org/10.1016/j.enpol.2019.110935\u003c/li\u003e\n\u003cli\u003ePalm, J., Eidenskog, M., \u0026amp; Luthander, R. (2017). Sufficiency, change, and flexibility: Critically examining the energy consumption profiles of solar PV prosumers in Sweden. \u003cem\u003eEnergy Research \u0026amp; Social Science\u003c/em\u003e, \u003cem\u003e39\u003c/em\u003e, 12\u0026ndash;18. https://doi.org/10.1016/j.erss.2017.10.006\u003c/li\u003e\n\u003cli\u003eRyan, L., Wheatley, J., \u0026amp; Saba, N. (2023). \u003cem\u003eA review of policies for the rollout of rooftop solar PV in Ireland\u003c/em\u003e. Climate Change Advisory Council. Retrieved March 3, 2025, from https://www.climatecouncil.ie/councilpublications/councilworkingpaperseries/WORKIN~2.PDF\u003c/li\u003e\n\u003cli\u003eSEAI. (2024). Energy in Ireland 2024 Report. Sustainable Energy Authority of Ireland. Retrieved March 7, 2025, from https://www.seai.ie/sites/default/files/publications/energy-in-ireland-2024.pdf\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Footnotes","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003e Note that not all SEAI programs record completion date. In these cases, grant payment date was used instead.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003e Note that homeowners could have installed energy efficient measures outside of SEAI programs, but these data are not available.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Photovoltaics, residential, energy retrofit, energy efficiency","lastPublishedDoi":"10.21203/rs.3.rs-6330919/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6330919/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study investigates the impact of solar photovoltaic (solar PV) system installations on subsequent energy retrofitting behaviour among residential dwellings in Ireland, particularly within the context of the nation's commitment to sustainability and renewable energy as outlined in the National Development Plan 2021\u0026ndash;2030. Despite over 65,000 solar PV installations facilitated by the Sustainable Energy Authority of Ireland (SEAI), findings suggest that the installation of solar PV does not significantly encourage additional retrofitting actions when compared to alternative measures, such as attic insulation and heating controls. Using statistical methodologies, including chi-square tests and survival analysis, the study reveals that homeowners who adopt solar PV systems exhibit a minimal propensity to pursue further retrofits, with an average additional investment of only \u0026euro;200 per installation. While the results indicate a nuanced relationship influenced by dwelling type and energy building rating (BER), the hypothesis that solar PV serves as a \"gateway\" to broader energy efficiency improvements is unsupported. The findings emphasize the need for integrated energy strategies that encompass both renewable technology adoption and comprehensive energy performance upgrades to achieve Ireland's ambitious climate targets. This research underscores the importance of fostering holistic approaches to retrofitting that encourage homeowners to invest not just in renewable systems such as solar PV but also in measures that enhance overall energy efficiency such as dwelling insulation.\u003c/p\u003e","manuscriptTitle":"Analyzing solar installations: a catalyst or barrier to subsequent residential retrofits","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-16 16:22:20","doi":"10.21203/rs.3.rs-6330919/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"82df60ab-462d-485b-9bd2-4b8218dc30ae","owner":[],"postedDate":"May 16th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-01-12T16:18:23+00:00","versionOfRecord":{"articleIdentity":"rs-6330919","link":"https://doi.org/10.1007/s12053-025-10402-5","journal":{"identity":"energy-efficiency","isVorOnly":false,"title":"Energy Efficiency"},"publishedOn":"2026-01-07 15:58:57","publishedOnDateReadable":"January 7th, 2026"},"versionCreatedAt":"2025-05-16 16:22:20","video":"","vorDoi":"10.1007/s12053-025-10402-5","vorDoiUrl":"https://doi.org/10.1007/s12053-025-10402-5","workflowStages":[]},"version":"v1","identity":"rs-6330919","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6330919","identity":"rs-6330919","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}