Implications of the Urban Heat Island on the selection of optimal retrofitting strategies: A case study in a Mediterranean climate

Depósito de Investigación de la Universidad de Sevilla https://idus.us.es/ This is an Accepted Manuscript of an article published by Elsevier in Urban Climate, Vol. 44, on July 2022, , available at: https://doi.org/10.1016/j.uclim.2022.101234 © 2022 Elsevier. En idUS Licencia Creative Commons CC BY-NC-ND 1 Implications of the Urban Heat Island on the selection of optimal retrofitting strategies: a case study in a Mediterranean climate Abstract The Urban Heat Island (UHI) is a significant phenomenon that currently receives much attention from the scientific community due to its importance on the environment, and it is mostly regarded as a merely negative event. However, there is a clear knowledge gap to investigate its influence when choosing optimal retrofitting interventions in buildings. A replicable methodology is thus needed, also to allow the comparison between different studies, which is currently an intricate task. The aim of this study is to develop such a methodology, using hourly rural and urban data to create temperature baselines and improve building simulations. An evaluation of the UHI impacts on building consumptions is presented, together with the repercussions on the energy savings achieved by retrofit interventions. The results considering case studies in Seville (Spain), in a Mediterranean climate, show that the heating demands are always overestimated when the model driven by rural climate data is used instead of the urban climate model. Conversely, the cooling demands are always underestimated in every considered scenario. In addition, the climate does not considerably influence the decision-making process to choose the optimal retrofitting alternatives, although the actual values of energy demand differ greatly. Keywords: Microclimate, Urban climate, Urban Heat Island, Buildings, Retrofitting. 1 Introduction 1.1 Background The rapid development of urban areas around the world has led to a growing concern about the pressure that humanity is exerting on the environment, which poses a threat to our current way of living. Urbanization has deteriorated indoor and outdoor air quality, modified buildings’ ventilation, increased buildings’ energy consumption and transformed the once green outdoor environment into areas with heat-absorbing materials (asphalt for instance), triggering an 2 increase in urban temperatures. A growing number of publications may be found related to the Urban Heat Island (UHI) phenomenon; for instance, the book (Oke et al., 2017) shows a very comprehensive synthesis of recent research on urban climates. This UHI phenomenon explains why urban temperatures are warmer in comparison to rural areas. The energy use in buildings is greatly influenced by the UHI, and researchers have proven that the total building energy use is frequently underestimated when the UHI effect is considered. However, there is an often overlooked fact in scientific literature: the UHI is mostly regarded as a negative phenomenon that occurs only in summer. Yet, the overall UHI effect may be even beneficial for certain weather conditions and buildings, especially in climate zones where the use of energy for heating is prevalent. These winter energy consumption benefits of the UHI are clearly articulated by several studies such as (Boccalatte et al., 2020; Crawl, 2008; Hong et al., 2021; Meng et al., 2020b; Santamouris et al., 2001; Shen et al., 2021). Nevertheless, there is still a lack of studies on the overall UHI effects considering both summer and winter for different climates, which is why further investigations are needed. 1.2 UHI and building energy consumptions Buildings are greatly affected by the weather conditions of their surroundings, particularly air temperatures. Decreasing their energy consumption is very important, since for instance in the case of the European Union buildings account for 40 % of energy end-use (European Commission, 2021). To evaluate the energy performance of these buildings, simulation tools are frequently utilized to estimate the heating and cooling demands, using hourly weather files. Building regulations also require to perform thermal simulations of the buildings’ behavior to demonstrate compliance. In addition, energy models may be calibrated for studying building retrofits, but if the results want to be compared it is important to link the calibration parameters to the time when the model is defined (Guattari et al., 2018). A proper assessment of a building thus requires a precise consideration of the actual climate. However, most studies simulate buildings in isolation or employ reference weather data. This is why there is an increasing concern about using such reference or rural weather data, since it could lead to making incorrect decisions for example regarding energy efficiency in metropolitan areas (Sun and Augenbroe, 2014). Also, it is important to understand and map building energy consumptions and quantify their impact on the urban environment, and data is not yet available at adequate spatial and temporal scales (Olivo et al., 2017). Due to the complexity of having data of urban temperatures with and without the UHI effect to simulate building consumptions, only some studies quantify the UHI impacts (Li et al., 2019). 3 There are however some tools which may be used as weather generators to modify climate data for urban applications, for example ENVI-met (ENVI-met, 2021), the Urban Weather Generator (Bueno et al., 2013b), and other approaches such as those proposed by (Bueno et al., 2014, 2013a; Kim et al., 2018; McLeod et al., 2013). In that regard, to account for the modifications of urban temperatures other works such as (Santos et al., 2018) propose to calibrate the building models taking into account the microclimate at district level. In (Santos et al., 2017), a novel calibration methodology including the UHI effects is proposed, demonstrating the relevance of considering the UHI impacts on the simulation of urban areas. As will be suggested in the present study, an appropriate methodology should thus be adopted in order to adapt the standard weather files that are used in building energy performance simulation tools. With respect to climate data, building simulations that use reference weather data throw unrealistic results, an issue which has been studied in scientific research. However, it is very important to evaluate how much simulations with rural data differ from simulations with urban data, allowing to compute the UHI impact on building energy consumptions. For these reasons, the present study proposes a new replicable methodology based on empirical models and real climate measurements, with the purpose of modifying adequately the weather files used in building simulations. This is done to fill the existing knowledge gaps and quantify the real consequences of the UHI on the selection of optimal retrofitting strategies, as well as look into the discrepancies due to the use of reference weather files. 1.3 Quantification of UHI impacts Even though the importance of the UHI effect on the energy consumption of buildings is broadly acknowledged by the scientific community, the quantitative impacts are not sufficiently studied (Li et al., 2019). One of the most important difficulties lies in having a long temporal coverage and a fine spatial resolution, which is why mobile air temperature monitoring is a promising method to better understand temperature distributions at fine spatial resolutions across urban areas (Tsin et al., 2016). Also, as noted before, the scientific works that quantify UHI impacts mostly study its negative effects. According to (Roxon et al., 2020), many research works about the UHI have merely focused on the cooling demands of summer months with hot weather, but fail to consider that the UHI may even be beneficial for cold climates due to the fact that the decreased heating energy demands could exceed the higher cooling energy demands. The influence of the UHI on the heating consumption of buildings at hourly or daily time scales have thus remained understudied (Meng et al., 2020a). Some examples of the underlying reasons 4 why winter UHI impacts are often overlooked are suggested by the statements of the following studies:  The negative impacts of summertime heat islands outweigh the benefits of wintertime warming in most cities because the largest population centers are located in (sub)tropical climates (Masumoto, 2015).  Considering that many developing countries are located in tropical or subtropical regions and that urbanization has advanced rapidly, mitigation of UHIs would probably reduce energy consumption (Hirano and Fujita, 2012).  In Delhi, the increased consumption in summer months far outweighs the reduced consumption in winter months (Kumari et al., 2021). Nevertheless, these assertions were made for specific locations and may not be extrapolated to all climates. In fact, there are some studies that explore the tradeoffs in summer and winter, performed for example in the US or in European cities. For instance, the review paper in (Santamouris, 2014) collected existing studies, showing that the cooling demand increased 23% on average but the heating demands were also reduced by 19 %. Apart from that, in previous studies the methodology used to account for the UHI effects on building energy performances varied, thus hindering extracting solid conclusions and comparing the results of different studies (Santamouris, 2014). Even within regions such as London or Beijing, there were significant differences if different studies were considered. Hence, there is a need to improve the comparability of studies so as to quantify and understand the effect of the UHI on buildings in a better way (Yang et al., 2020). A replicable methodology to do so is thus needed, as proposed in the present work. Another important discussion which has not received sufficient attention is the influence of the UHI on different types of buildings. For instance, (Hirano and Fujita, 2012) evaluated the influence of the UHI on energy consumptions in Tokyo considering residential and commercial buildings, stating that the UHI increased commercial energy consumption but decreased residential energy consumption. The present work also includes an evaluation of the UHI impact on the energy consumption of a residential and a commercial building. In addition, as far as the authors know there is an almost negligible amount of research about the influence of the UHI on the quantification of the energy savings achieved by retrofitting strategies. In such a case an accurate evaluation of the actual energy consumptions would not be needed, but instead an accurate assessment of the relative changes in energy use as a result of interventions. This issue is however discussed in (Sun and Augenbroe, 2014), which poses the 5 question whether obtaining proof of energy saving percentages of retrofit alternatives requires the consideration of the UHI in the simulations. According to the results of the previous paper, when building simulations are utilized to evaluate the energy savings it is not so important to consider the UHI impact (if the percentage of relative savings is the only metric). Conversely, in (Dodoo et al., 2017) the influence of several parameters (including outdoor temperatures) is considered for the quantification of energy savings in several retrofitting alternatives, concluding that the differences of saving percentages were not insignificant. Another example is the study shown in (De Masi et al., 2021), which analyzes a residential case study and the impact of climate change projections in the refurbishment design. According to this study, traditional passive measures (insulation, solar control or low emissive windows, cool roof, shading systems) are not resilient when the energy savings are calculated taking into account the climate change. Last of all, (Zou et al., 2021) studies a classroom in a coolingdominant hot and humid climate whose results show that the different climatic conditions can lead to distinct building performance indices and optimal solutions. The effects of climate change are also very important to evaluate the energy consumption of buildings. According to (Yang et al., 2021), who performed a study of 38 European cities with climate change projections, there will be larger needs for cooling and less for heating in the future, although there are differences between zones and the discomfort hours will increase notably in cooling-dominated zones but not so much in those heating-dominated. Climate change may also have some implications regarding Nearly Zero Energy Buildings (NZEB) for instance, which are believed to increase the cooling demands in summer due to the added thermal insulation. In fact, (Baglivo et al., 2022) showed how NZEB constructed in hot climates with increasingly insulated thermal envelopes will suffer from overheating over the years, leading to disproportionate uses of cooling. As stated by (Ascione et al., 2022), who performed a study of a NZEB in a Mediterranean climate, the effect of climate changes could bring to turn on the cooling systems even in the spring season. Also, the study in (D’Agostino et al., 2022) investigates how climate change may impact building heating and cooling loads, indicating that heating could decrease by 38%-57% while cooling could increase by +99%-380%, depending on the location. In summary, even though the potential benefits of the UHI in winter are frequently overlooked, many researchers have already noted that increasing urban temperatures (whether due to climate change or the UHI) result in energy cost savings across many cities. However, there is a 6 clear knowledge gap in the quantification of the UHI influence for the assessment of energy savings in retrofit interventions, which the current work intends to address. 1.4 Modification of weather data in building simulations We mentioned before that it is very common to use reference files that represent long-term typical weather conditions, but an accurate building performance evaluation should consider instead the actual weather conditions. Hourly weather data are thus needed for building simulations, such as dry bulb temperature, humidity, solar radiation and wind speed. These data are usually based on observations, and a file may be created with typical individual months or using a representative year period with average conditions. One of the most common alternatives in this regard is using Test Reference Year (TRY) or Typical Meteorological Year (TMY, TMY2 and TMY3) files, developed from climate data collected in weather stations that are often located in rural environments around the world. However, climate data obtained outside urban areas are not influenced by the UHI effect and may be considered accurate for studying buildings in districts with low urban densities, but not for buildings in dense urban environments (Guattari et al., 2018). Different approaches have been used in the literature to account for the UHI impacts on building simulations by modifying the weather data. Having high-spatial-resolution and hourly data through on-site measurements is not feasible, but the files may be generated for instance using spatial interpolations from limited weather stations located in a city (Hwang et al., 2020). Developing a methodology for including the UHI impacts in TMY files is very important for cities, so methods such as “morphing” may be used to reflect the average weather conditions while keeping physically realistic data (Hwang et al., 2020). In this regard, the study in (Chan, 2011) reported a method to modify a typical meteorological weather file through a morphing algorithm, taking into account the UHI effect in the summer season through field measurements. Other examples are for instance the research in (Liu et al., 2017), whose objective was to consider micro-scale data that took into account the UHI effect in Singapore, generating both rural and urban TMY files by modifying the temperatures. In (Sun and Augenbroe, 2014), a model was used to generate various hourly UHI magnitudes for different urban contexts, modifying the TMY rural data used in the simulations. On the other hand, it is also possible to couple urban climate models to building simulations. This is done in (Salvati et al., 2020), describing a chain strategy to model urban boundary conditions suitable for annual simulations using dynamic thermal simulation tools. In this way, the UHI effect may be captured. As stated by (Jin et al., 2021), coupling of the building energy 7 models’ approach in an urban climate model is also worthwhile for a more realistic and accurate representation of near-surface energy fluxes and air temperature in mid-latitude cities. 1.5 Aim of this work The literature review has evidenced that there are some clear knowledge gaps when it comes to assessing the real impact of the UHI on the decision-making process for retrofitting, particularly in Mediterranean climates. As stated in (Díaz-López et al., 2021), for instance twothirds of Spanish cities are currently designing and constructing buildings with obsolescent climate data that do not take into account the current or future climate reality, which is significantly affecting previous calculations on the thermal performance of buildings. Thus, the present study proposes a new methodology for quantifying more accurately the energy savings achieved by retrofitting interventions in urban environments, for long periods and different seasons. This is carried out by developing empirical models obtained through real measurements at the reference weather station and at the urban area under study, proposing the modification of the reference weather files used in building energy performance evaluations. The aim is to develop a replicable methodology that would also allow the currently unfeasible comparison between research studies, including different types of buildings and climates. Both the heating and cooling demands are evaluated for rural and urban climate contexts, assessing the real impact of the UHI due to its effect on air temperatures. Even though the UHI modifies the urban climate in more ways than the ambient air temperature (humidity, wind speeds, precipitation, shading and solar radiation may also be modified), our focus is on the influence of the air temperatures, the rest are out of the scope. Nevertheless, the real conditions of solar radiation and wind are considered for the development of the urban temperature model. The present study therefore investigates the influence of the UHI on the assessment of energy savings in retrofit interventions, which includes passive measures such as improving the thermal envelope of the buildings as well as solar control and night ventilation alternatives for summer periods. The methodology is illustrated in a Mediterranean climate, using the city of Seville in southern Spain as a case study, and two different types of building are considered: a residential and a commercial building, representative of the type of construction in the area. 8 2 Data and methods 2.1 Climate data In this section, the necessary information to modify the reference weather data is discussed. First of all, the methodology in the present study requires having temperature data of:  A rural weather station, so as to be able to evaluate the UHI effect.  The weather station where the reference climate for the region under study was developed. In some cases this might be the same as the rural weather station.  A weather station in the urban area. In this study a temperature datalogger (resolution of 0.1 °C and accuracy of ± 0.5 ◦C) was placed at a 3 meters height in a representative area of the city center of Seville. A fan-aspirated radiation shield is included so as to avoid a hot bias in the temperature measurements. The reference climate is usually a Typical Meteorological Year file which provides a composite climate for building energy simulation tools. In the case of Seville, this file is provided by the official Technical Building Code in Spain (CTE). It was composed using several years of measurements at the weather station located at Seville’s airport, which allowed to create this weather file with the typical climate in the area. As stated above, in some cases the rural weather station and the reference weather station could be the same. However, in the particular case of Seville, the study in (Romero Rodríguez et al., 2020b) showed that the temperatures at the airport are influenced by the city due to the prevailing wind directions in the region. For this reason, in our study the rural weather station of “La Rinconada” is used instead for UHI considerations. The location of the 3 sources of information is shown in Figure 1. 15 3 Results and discussion 3.1 Development of the modified rural and urban weather files 3.1.1 Temperature baseline of the city center The baseline of the city center is composed by a winter and a summer model. To obtain the coefficients of the model in summer or winter, first we need hourly data of the weather station that was used to obtain the reference climate. In our case, this means the data at the airport. Also, to train the model we need the real measurements of the city center. To obtain the coefficients of the summer model, hourly data from 15/06/2019 to 29/06/2019 was used: two weeks in summer. In the case of the winter model, the data used to train the model dated from 15/01/2020 to 29/01/2020. The rest of the periods in which measurements were available were used for the validation of the models. It is important to make sure that the training periods contain a sufficient variation of wind speeds and cloud cover, to be as much representative as possible of the actual weather conditions during summer and winter. Otherwise, some of the complexities of the UHI effect would not be captured and the model would not be accurate. As may be seen in Figure 4, in our case the data used for the baselines during the training periods show very similar tendencies in their percentiles compared to those of the whole summer and winter periods. Both low and high values of wind speed were available in the baselines so as to train the models properly, although it is true that more data with low wind conditions would have been desirable in the summer training period. Even so, the accuracy of the estimations of the summer model are very precise. As for the ratio of diffuse radiation, the distributions are very similar in summer and a bit more different in winter, but still, days with no cloud cover and days with a lot of cloud cover were available during the training periods, so the models could be trained in a precise way. 16 Figure 4: Adequacy of the climate data used during the training periods. Once the summer and winter models were obtained, they were used to create the baseline equation of the city center in order to estimate the urban temperature depending on the climate conditions at the reference weather station. Since measured temperature data was available for a whole year, Figure 5 shows the validation for the whole period, producing very satisfactory results with a coefficient of determination (R2) of 0.969 and a Root Mean Square Error (RMSE) of 1.23. These results are very accurate compared to previous literature as shown in (Romero Rodríguez et al., 2020a), particularly for hourly measurements and such a long timeframe. Figure 5: Validation of the city center baseline. 17 The accuracy of the estimations made by the baseline may also be observed in Figure 6, which shows a comparison between the hourly real and estimated temperatures in a period far from that used to train the summer and winter models. Figure 6: Comparison between the estimated and the measured temperatures in September. Model of the city center. 3.1.2 Temperature baseline of the rural weather station The baseline of the rural weather station was obtained in the same way as that of the city center. It was composed by a winter and a summer model, obtained through hourly data in the same periods: 15/06/2019 to 29/06/2019 for the summer model, and 15/01/2020 to 29/01/2020 for the winter model. Figure 7 shows the validation for the whole period, producing also very satisfactory results with a R2 of 0.971 and a RMSE of 1.25. Figure 8 is also included to compare the hourly evolution of the real and the estimated temperatures. Figure 7: Validation of the rural weather station baseline. 18 Figure 8: Comparison between the estimated and the measured temperatures in September. Model of the rural weather station. 3.1.3 Typical meteorological year for the city center and the rural weather station First, we ensured that the baselines of the city center and the rural weather station are accurate enough so as to estimate their temperatures depending on those at the reference weather station of the airport. Then, we may use the reference weather file of the latter and apply the baselines so as to modify the weather. In this way, with the temperatures, wind speed and solar radiation of the reference climate we may infer what the temperatures would have been at the city center or at the rural weather station. Thus, it is possible to obtain their modified typical meteorological year files. Figure 9 shows a comparison between the temperatures of the reference climate and those modified for the city center and the rural weather station using their baselines, for a period of several days during summer. Several conclusions may be obtained from the figure. First, it may be seen that there is a clear UHI effect in the city center during the night, with maximum UHI intensities being reached around sunrise. Also, it may be observed that as previously mentioned the temperatures at the airport (where the reference climate was obtained) are influenced by those at the city. In addition, the temperatures at noon are generally higher at the city center compared to the rural weather station. 19 Figure 9: Comparison of temperatures for the typical weather files considered. By considering the modified weather files in our case, we may also compare the average hourly temperature profiles for each month independently. Figure 10 shows for example the months of December and July, allowing us to compare the temperatures at the rural weather station with those at the city. As it may be seen, the UHI Intensity (UHII) is never higher than 2 °C in December, while in July values around 5 °C are reached before sunrise. Figure 10: Average hourly temperature profiles of the city center and the rural weather station in December and July. We may also calculate for example the Heating Degree Hours (HDH) for the rural and urban weather files, or the Cooling Degree Hours at Night (CDHN). The HDH are a measure of how much and for how long the temperatures are below a base heating temperature of 20 °C. On the other hand, the CDHN are calculated similarly but during night hours (from 00 to 08h) for temperatures below 25 °C. The CDHN we are introducing should not be confused with the typical Cooling Degree Hours (CDH) found in the literature, since we intend to discuss the UHI influence on night ventilation strategies, which take advantage of the lower temperatures in summer during the night to reduce the cooling demands. This is why for the CDHN we are considering the temperatures below and not above the cooling threshold. 20 The values of HDH and CDHN for each month are shown in Figure 11 and Figure 12 respectively. It may be seen that in winter, the HDHs are lower at the city center. This suggests that the UHI effect in the city results in lower heating demands, and also that the rural climate considered in reference weather files may overestimate the real heating demands. Also, the CDHN are much lower during summer in the city center, which means that the night cooling potentials will be actually lower since there are less hours when the ventilation systems could use outdoor temperatures to reduce the cooling demands. We will discuss this further when we look at the results of the energy saving alternatives. Figure 11: Heating Degree Hours for each month. Figure 12: Cooling Degree Hours at Night in the summer months. 21 3.2 Building simulations 3.2.1 Energy demands: base case scenarios Once the modified climate files for the rural weather station and the city center have been obtained, it is possible to calculate the energy demands of the case study buildings by using any building energy performance simulation software. In our case, we performed accurate hourly simulations using the computer program Unified LIDER-CALENER software Tool, explained in more detail in Section 2.3.2. The simulation of the residential and the commercial buildings considered in the present work allowed the calculation of their heating and cooling demands in their base case scenarios. The results are shown in Table 2. The retrofitting alternatives will be considered later in the next sections. Building Climate Heating demand [kWh/m2] Cooling demand [kWh/m2] Residential Rural 53.19 10.96 Urban 33.65 (-37%) 25.35 (+131%) Office Rural 31.12 34.30 Urban 17.28 (-44%) 46.43 (+35%) Table 2: Heating and cooling demands of the buildings in the base case scenarios. First of all, it may be observed that the cooling demands are much higher for the urban climate. This result might be expected, due to the strong UHI effect in the area which increases urban temperatures. In the case of the residential building for instance the cooling demand is almost 2.5 times higher for the urban climate (+131% difference). This evidences that relying on the results of rural or reference weather files (frequently obtained in rural surroundings) when performing building simulations would clearly underestimate the cooling demands and energy consumptions. As for the heating demands, it is also interesting to see that they are greatly reduced when the urban climate is considered, both in the residential (-37%) and the office building (-44%). This means that performing simulations with the rural or reference weather files would overestimate the heating demands and energy consumptions. Due to these opposing effects when considering the urban climate (increased cooling demands but reduced heating demands), the total energy consumption could be higher or lower, depending on the relative differences as well as the type of air conditioning system. Before 22 discussing this issue, we will introduce the results of the different retrofitting alternatives in the next section, so as to understand their influence on the energy demands under urban and rural climates. 3.2.2 Energy demands: retrofitting alternatives In this section, the building models were modified to study all the retrofitting alternatives shown in Table 8 and Table 9 of the Appendix of supplementary data. Once every model was available, they were simulated considering both the rural and the urban climates. All in all, a total of 1800 simulations were carried out. 3.2.2.1 Residential building In the case of the residential building, the first matter we will discuss is the effect of the climate on the heating demands. Figure 13 shows the results of this comparison for all the retrofitting alternatives, showing the heating demands (left figure) as well as the percentages of error committed (right figure) when using the urban climate instead of the rural climate. This percentage is calculated as the relative difference of the urban demand with respect to the rural demand. In the case of the demands, a 45-degree line has been drawn. It may be observed that the differences of demand are clearly linear, and that the urban weather produces lower heating demands in every scenario. The percentage of error ranges from -37% to -54%. Also, we can see that the more we reduce the heating demand using retrofitting alternatives, the higher the error committed. We may conclude that performing simulations with the rural or reference climate would always result in an overestimation of the heating demands, regardless of the retrofitting alternative. In addition, the UHI produces a clearly positive effect in winter by reducing the heating demands, which as mentioned in the literature review is a frequently overlooked side benefit. These differences are in this case not negligible. 23 Figure 13: Comparison of the heating demands for the urban and the rural climates (left) and percentage of error committed (right). Residential building. The markers show the 540 retrofitting alternatives. On the other hand, we also need to investigate the effect of the UHI on the cooling demands. Just as for the heating demands, Figure 14 shows a comparison of the results for the rural and the urban climate in every retrofitting scenario. Here, the tendencies are different than those observed for the heating demands. From these results we may conclude that using the rural or reference climate in the simulations would always result in an underestimation of the cooling energy demands compared to the urban climate, regardless of the retrofitting alternative. Also, the lower the cooling demand, the higher the percentage of error. This is the case of the retrofitting alternatives used for reducing the cooling demands in summer: solar control and night ventilation strategies. The cooling demands may even be 12 times higher for the urban climate in such cases. The reason for this lies mostly on the impact of the night ventilation strategies: since the temperatures at night are much higher in the urban climate, increasing night ventilation does not reduce the cooling demands as much as it would in a rural climate with much lower temperatures. Figure 14: Comparison of the cooling demands for the urban and the rural climates (left) and percentage of error committed (right). Residential building. The markers show the 540 retrofitting alternatives. 24 3.2.2.2 Office building In the case of the office building, very similar tendencies may be observed. Figure 15 and Figure 16 show the results. The heating demands are again always overestimated by the rural climate in every scenario, with errors committed between -44 % and -60 %. With respect to the cooling demands, they are always underestimated by the rural climate, and errors up to 619 % are reached when the cooling demands are very low. The reason is the same as for the residential building: night ventilation allows to take advantage of the low temperatures during the night to reduce the cooling demands almost completely in the rural climate, while the urban climate does not achieve this due to the strong UHI effect at night. Figure 15: Comparison of the heating demands for the urban and the rural climates (left) and percentage of error committed (right). Office building. The markers show the 360 retrofitting alternatives. Figure 16: Comparison of the cooling demands for the urban and the rural climates (left) and percentage of error committed (right). Office building. The markers show the 360 retrofitting alternatives. 31 4 Conclusions and future work The present work has demonstrated the possibilities offered by the development of urban and rural temperature baselines for evaluating the real impact of the UHI on buildings, frequently regarded as a merely negative phenomenon in scientific literature. A replicable methodology has been established which would allow the comparison of results between different studies, which is currently unfeasible due to the varying nature of the methods that are used. In addition, the UHI effects on the energy consumption of buildings have been discussed, as well as the UHI influence on the results of heating and cooling saving estimations of retrofitting alternatives. The methodology was illustrated for a Mediterranean climate considering the case study of Seville in southern Spain, where an empirical model with an R2 of 0.969 and an RMSE of 1.23 was obtained for the temperature baseline of the city center. In winter months such as December, the UHII reached maximum values of 2 °C, while in July values around 5 °C were found before sunrise. In this case, it was possible to establish that the heating demands in a residential and an office building were always overestimated by the rural climate when compared to the urban climate. On the other hand, the cooling demands were always underestimated. After simulating hundreds of retrofitting alternatives, it was also possible to conclude that while the results of heating demand savings (in percentage) were similar for both climates, significant differences could be observed for the cooling alternatives. Also, we could see that the more the energy demands were reduced using retrofitting alternatives, the higher the errors committed. From the results of the retrofitting scenarios it may also be concluded that the climate did not greatly influence the retrofitting decision-making process to choose the optimal alternatives, although differences were found in solar control strategies and the actual values of energy demand differed enormously. As for the PECs, they were higher for the urban climate in almost every scenario of the residential building and in every scenario of the office building. This means that the annual UHI effect is negative in most cases but not always, and therefore it should be carefully evaluated for different climates, types of buildings and construction. With regards to future work, one of the shortcomings of the present study was the consideration of only two different buildings: a residential building and an office building. Therefore, a very interesting study to consider for the future would be the simulation of many different buildings, changing geometries, building orientations, building use, and so on. Also, different climates could be considered around the world, which would allow to reach more solid conclusions regarding the overall positive or negative UHI effect depending on the relative importance of 32 the heating and cooling demands. Furthermore, it would be possible to study in more detail the feasibility of night ventilation strategies depending on the climate and the UHI intensities. In addition, we should also mention that the study shown in (Romero Rodríguez et al., 2020a) presented a methodology to obtain urban temperature baselines through mobile transects. With that method, if mobile transects were carried out in a certain region it would be possible to obtain a temperature baseline for any particular location, providing a more sophisticated way of correcting for local conditions. Once the models were available and by following the approach of the present study to modify the weather files, any building of the region could be simulated with an extremely accurate consideration of its real weather conditions. Such a study would therefore highlight the differences in building energy consumptions due to the varying UHI intensities within any region, which may differ substantially and are not yet fully understood in scientific literature. 5 Acknowledgments This study was funded by the European Commission under the project LIFE WATERCOOL - Water efficient systemic concept for the climate change adaptation in urban areas (LIFE18 CCA/ES/001122). The project was also co-financed by the European Regional Development Funds (ERDF). Appendix: Supplementary data Heating reduction alternatives Cooling reduction alternatives Level of improvement Walls [W/m2K] Roof [W/m2K] Floor [W/m2K] Windows [W/m2K] Linear transmittance of Thermal Bridges Infiltrations + ventilation Solar control (g-value of windows in summer) Night ventilation [1/h] 1 1.38 0.63 1.90 5.6 Base case n50=5.1 0.77 4 2 0.56 0.44 0.56 3.1 50% reduction n50=3 0.39 8 3 0.42 0.33 0.44 2.6 75% reduction n50=3 (with demand control ventilation) 0.1 10 4 0.27 0.21 0.32 2.1 5 0.28 0.215 0.32 2.12 Table 8: Catalogue of demand reduction alternatives for the residential building. 33 Heating reduction alternatives Cooling reduction alternatives Level of improvement Walls [W/m2K] Roof [W/m2K] Floor [W/m2K] Windows [W/m2K] Linear transmittance of Thermal Bridges Infiltrations + ventilation Solar control (g-value of windows in summer) Night ventilation [1/h] 1 0.67 0.90 1.96 3.37 Base case n50=7.1 0.69 0 2 0.6 0.4 0.6 2.3 80 % reduction n50=2.5 0.35 4 3 0.32 0.25 0.5 1.9 85 % reduction n50=2.5 (with demand control ventilation) 10 4 0.27 0.22 0.48 1.8 n50=2.5 (with heat recovery) 5 0.18 0.15 0.19 1.4 Table 9: Catalogue of demand reduction alternatives for the office building. Level of improvement Is in the optimal zone? (1 yes, 0 no) Thermal envelope Linear transmittance of Thermal Bridges Infiltrations + ventilation Solar control Night ventilation Rural Urban 1 3 4 1 3 1 0 2 1 4 1 3 1 0 2 2 4 1 3 1 0 2 3 4 1 3 1 0 3 2 4 1 3 1 0 3 3 4 1 3 1 0 1 1 4 2 3 0 1 2 1 4 2 3 0 1 2 2 4 2 2 0 1 2 3 4 2 2 0 1 3 2 4 2 3 0 1 3 3 4 2 3 0 1 1 2 4 2 3 1 1 1 3 4 2 3 1 1 2 2 4 2 3 1 1 2 3 4 2 3 1 1 Table 10: Level of improvement of the 10 optimal retrofitting alternatives. Office building. 34 6 References Aparicio Ruiz, P., Guadix Martín, J., Salmerón Lissén, J.M., Sánchez de la Flor, F.J., 2014. An integrated optimisation method for residential building design: A case study in Spain. Energy Build. 80, 158–168. https://doi.org/10.1016/J.ENBUILD.2014.05.020 Ascione, F., De Masi, R.F., Gigante, A., Vanoli, G.P., 2022. Resilience to the climate change of nearly zero energy-building designed according to the EPBD Recast: monitoring, calibrated energy models and perspective simulations of a Mediterranean nZEB living lab. Energy Build. 262, 112004. https://doi.org/10.1016/j.enbuild.2022.112004 Baglivo, C., Congedo, P.M., Murrone, G., Lezzi, D., 2022. Long-term predictive energy analysis of a high-performance building in a mediterranean climate under climate change. Energy 238, 121641. https://doi.org/10.1016/j.energy.2021.121641 Boccalatte, A., Fossa, M., Gaillard, L., Menezo, C., 2020. Microclimate and urban morphology effects on building energy demand in different European cities. Energy Build. 224, 110129. https://doi.org/10.1016/j.enbuild.2020.110129 Brandão de Vasconcelos, A., Pinheiro, M.D., Manso, A., Cabaço, A., 2016. EPBD cost-optimal methodology: Application to the thermal rehabilitation of the building envelope of a Portuguese residential reference building. Energy Build. 111, 12–25. https://doi.org/10.1016/j.enbuild.2015.11.006 Bueno, B., Hidalgo, J., Pigeon, G., Norford, L., Masson, V., 2013a. Calculation of air temperatures above the urban canopy layer from measurements at a rural operational weather station. J. Appl. Meteorol. Climatol. 52, 472–483. https://doi.org/10.1175/JAMC-D-12-083.1 Bueno, B., Norford, L., Hidalgo, J., Pigeon, G., 2013b. The urban weather generator. J. Build. Perform. Simul. 6, 269–281. https://doi.org/10.1080/19401493.2012.718797 Bueno, B., Roth, M., Norford, L., Li, R., 2014. Computationally efficient prediction of canopy level urban air temperature at the neighbourhood scale. Urban Clim. 9, 35–53. https://doi.org/10.1016/j.uclim.2014.05.005 Chan, A.L.S., 2011. Developing a modified typical meteorological year weather file for Hong Kong taking into account the urban heat island effect. Build. Environ. 46, 2434–2441. https://doi.org/10.1016/j.buildenv.2011.04.038 Crawl, D.B., 2008. Estimating the impacts of climate change and urbanization on building performance. J. Build. Perform. Simul. 1, 91–115. 35 https://doi.org/10.1080/19401490802182079 D’Agostino, D., Parker, D., Epifani, I., Crawley, D., Lawrie, L., 2022. How will future climate impact the design and performance of nearly zero energy buildings (NZEBs)? Energy 240, 122479. https://doi.org/10.1016/j.energy.2021.122479 De Masi, R.F., Gigante, A., Ruggiero, S., Vanoli, G.P., 2021. Impact of weather data and climate change projections in the refurbishment design of residential buildings in cooling dominated climate. Appl. Energy 303, 117584. https://doi.org/10.1016/j.apenergy.2021.117584 Díaz-López, C., Verichev, K., Holgado-Terriza, J.A., Zamorano, M., 2021. Evolution of climate zones for building in Spain in the face of climate change. Sustain. Cities Soc. 74. https://doi.org/10.1016/j.scs.2021.103223 Dodoo, A., Yao Ayikoe Tettey, U., Gustavsson, L., 2017. On input parameters, methods and assumptions for energy balance and retrofit analyses for residential buildings. Energy Build. 137, 76–89. https://doi.org/10.1016/j.enbuild.2016.12.033 ENVI-met, 2021. ENVI-met:Decode urban nature with Microclimate simulations. URL: https://www.envi-met.com/. EU, 2010. Directive 2010/31/EU of the European Parliament and of the Council of 19 May 2010 on the energy performance of buildings (recast). Off. J. Eur. Union 13–35. https://doi.org/doi:10.3000/17252555.L_2010.153.eng European Commission, 2021. 2020 climate & energy package [WWW Document]. URL https://ec.europa.eu/clima/policies/strategies/2020_en Ferrara, M., Fabrizio, E., Virgone, J., Filippi, M., 2014. A simulation-based optimization method for cost-optimal analysis of nearly Zero Energy Buildings. Energy Build. 84, 442–457. https://doi.org/10.1016/j.enbuild.2014.08.031 Guattari, C., Evangelisti, L., Balaras, C.A., 2018. On the assessment of urban heat island phenomenon and its effects on building energy performance: A case study of Rome (Italy). Energy Build. 158, 605–615. https://doi.org/10.1016/j.enbuild.2017.10.050 Hamdy, M., Hasan, A., Siren, K., 2013. A multi-stage optimization method for cost-optimal and nearly-zero-energy building solutions in line with the EPBD-recast 2010. Energy Build. 56, 189–203. https://doi.org/10.1016/j.enbuild.2012.08.023 36 Herrando, M., Cambra, D., Navarro, M., de la Cruz, L., Millán, G., Zabalza, I., 2016. Energy Performance Certification of Faculty Buildings in Spain: The gap between estimated and real energy consumption. Energy Convers. Manag. 125, 141–153. https://doi.org/10.1016/J.ENCONMAN.2016.04.037 Hirano, Y., Fujita, T., 2012. Evaluation of the impact of the urban heat island on residential and commercial energy consumption in Tokyo. Energy 37, 371–383. https://doi.org/10.1016/j.energy.2011.11.018 Hong, T., Xu, Y., Sun, K., Zhang, W., Luo, X., Hooper, B., 2021. Urban microclimate and its impact on building performance: A case study of San Francisco. Urban Clim. 38. https://doi.org/10.1016/j.uclim.2021.100871 Hwang, R.L., Lin, T.P., Lin, F.Y., 2020. Evaluation and mapping of building overheating risk and air conditioning use due to the urban heat island effect. J. Build. Eng. 32, 101726. https://doi.org/10.1016/j.jobe.2020.101726 Jin, L., Schubert, S., Fenner, D., Meier, F., Schneider, C., 2021. Integration of a Building Energy Model in an Urban Climate Model and its Application. Boundary-Layer Meteorol. 178, 249– 281. https://doi.org/10.1007/s10546-020-00569-y Judkoff, R., Neymark, J., 1995. International Energy Agency building energy simulation test (BESTEST) and diagnostic method. Natl. Renew. Energy Lab. Size: 296 pages. https://doi.org/10.2172/90674 Kim, H., Gu, D., Kim, H.Y., 2018. Effects of Urban Heat Island mitigation in various climate zones in the United States. Sustain. Cities Soc. 41, 841–852. https://doi.org/10.1016/j.scs.2018.06.021 Kumari, P., Garg, V., Kumar, R., Kumar, K., 2021. Impact of urban heat island formation on energy consumption in Delhi. Urban Clim. 36, 100763. https://doi.org/10.1016/j.uclim.2020.100763 Li, X., Zhou, Y., Yu, S., Jia, G., Li, H., Li, W., 2019. Urban heat island impacts on building energy consumption: A review of approaches and findings. Energy 174, 407–419. https://doi.org/10.1016/j.energy.2019.02.183 Liu, Y., Stouffs, R., Tablada, A., Wong, N.H., Zhang, J., 2017. Comparing micro-scale weather data to building energy consumption in Singapore. Energy Build. 152, 776–791. https://doi.org/10.1016/j.enbuild.2016.11.019 37 Masumoto, K., 2015. Urban heat islands. Environ. Indic. 67–75. https://doi.org/10.1007/978-94- 017-9499-2_5 McLeod, R.S., Hopfe, C.J., Kwan, A., 2013. An investigation into future performance and overheating risks in Passivhaus dwellings. Build. Environ. 70, 189–209. https://doi.org/10.1016/j.buildenv.2013.08.024 Meng, F., Guo, J., Ren, G., Zhang, L., Zhang, R., 2020a. Impact of urban heat island on the variation of heating loads in residential and office buildings in Tianjin. Energy Build. 226, 110357. https://doi.org/10.1016/j.enbuild.2020.110357 Meng, F., Ren, G., Guo, J., Zhang, L., Zhang, R., 2020b. Impact of urban heat island effect on the heating and cooling loads of residential buildings in Tianjin City, China. Prog. Geogr. 39, 1296–1307. https://doi.org/10.18306/dlkxjz.2020.08.005 Ministry of Transport Mobility and Urban Agenda, 2019. Unified LIDER-CALENER software Tool (HULC) URL: https://www.codigotecnico.org/Programas/HerramientaUnificadaLIDERCALENER.html. Monzón, M., López-Mesa, B., 2017. Simplified model to determine the energy demand of existing buildings. Case study of social housing in Zaragoza, Spain. Energy Build. 149, 483– 493. https://doi.org/10.1016/J.ENBUILD.2017.04.039 Oke, T.R., Mills, G., Christen, A., Voogt, J.A., 2017. Urban Climates. Cambridge University Press. Olivo, Y., Hamidi, A., Ramamurthy, P., 2017. Spatiotemporal variability in building energy use in New York City. Energy 141, 1393–1401. https://doi.org/10.1016/j.energy.2017.11.066 Romero Rodríguez, L., Sánchez Ramos, J., Guerrero Delgado, M., Molina Félix, J.L., Álvarez Domínguez, S., 2018. Mitigating energy poverty: Potential contributions of combining PV and building thermal mass storage in low-income households. Energy Convers. Manag. 173, 65–80. https://doi.org/10.1016/j.enconman.2018.07.058 Romero Rodríguez, L., Sánchez Ramos, J., Molina Félix, J.L., Álvarez Domínguez, S., 2020a. Urbanscale air temperature estimation: development of an empirical model based on mobile transects. Sustain. Cities Soc. 63. https://doi.org/10.1016/j.scs.2020.102471 Romero Rodríguez, L., Sánchez Ramos, J., Sánchez de la Flor, F.J., Álvarez Domínguez, S., 2020b. Analyzing the Urban Heat Island: comprehensive methodology for data gathering and optimal design of mobile transects. Sustain. Cities Soc. 102027. https://doi.org/10.1016/j.scs.2020.102027 38 Rosselló-Batle, B., Ribas, C., Moià-Pol, A., Martínez-Moll, V., 2015. An assessment of the relationship between embodied and thermal energy demands in dwellings in a Mediterranean climate. Energy Build. 109, 230–244. https://doi.org/10.1016/J.ENBUILD.2015.10.007 Roxon, J., Ulm, F.J., Pellenq, R.J.M., 2020. Urban heat island impact on state residential energy cost and CO2 emissions in the United States. Urban Clim. 31, 100546. https://doi.org/10.1016/j.uclim.2019.100546 Sağlam, N.G., Yılmaz, A.Z., 2015. Progress towards EPBD Recast Targets in Turkey: Application of Cost Optimality Calculations to a Residential Building. Energy Procedia 78, 973–978. https://doi.org/10.1016/j.egypro.2015.11.036 Salvati, A., Palme, M., Chiesa, G., Kolokotroni, M., 2020. Built form, urban climate and building energy modelling: case-studies in Rome and Antofagasta. J. Build. Perform. Simul. 13, 209– 225. https://doi.org/10.1080/19401493.2019.1707876 Santamouris, M., 2014. On the energy impact of urban heat island and global warming on buildings. Energy Build. 82, 100–113. https://doi.org/10.1016/j.enbuild.2014.07.022 Santamouris, M., Papanikolaou, N., Livada, I., Koronakis, I., Georgakis, C., Argiriou, A., Assimakopoulos, D.N., 2001. On the impact of urban climate on the energy consuption of building. Sol. Energy 70, 201–216. https://doi.org/10.1016/S0038-092X(00)00095-5 Santamouris, M., Sfakianaki, A., Pavlou, K., 2010. On the efficiency of night ventilation techniques applied to residential buildings. Energy Build. 42, 1309–1313. https://doi.org/10.1016/j.enbuild.2010.02.024 Santos, L.G.R., Afshari, A., Norford, L.K., Mao, J., 2018. Evaluating approaches for district-wide energy model calibration considering the Urban Heat Island effect. Appl. Energy 215, 31– 40. https://doi.org/10.1016/j.apenergy.2018.01.089 Santos, L.G.R., Masri, D., Afshari, A., 2017. Inverse estimation of the urban heat island using district-scale building energy calibration. Energy Procedia 143, 264–270. https://doi.org/10.1016/J.EGYPRO.2017.12.682 Shen, P., Liu, J., Wang, M., 2021. Fast generation of microclimate weather data for building simulation under heat island using map capturing and clustering technique. Sustain. Cities Soc. 71, 102954. https://doi.org/10.1016/j.scs.2021.102954 Sun, Y., Augenbroe, G., 2014. Urban heat island effect on energy application studies of office 39 buildings. Energy Build. 77, 171–179. https://doi.org/10.1016/j.enbuild.2014.03.055 Tsin, P.K., Knudby, A., Krayenhoff, E.S., Ho, H.C., Brauer, M., Henderson, S.B., 2016. Microscale mobile monitoring of urban air temperature. Urban Clim. 18, 58–72. https://doi.org/10.1016/j.uclim.2016.10.001 Yang, X., Peng, L.L.H., Jiang, Z., Chen, Y., Yao, L., He, Y., Xu, T., 2020. Impact of urban heat island on energy demand in buildings: Local climate zones in Nanjing. Appl. Energy 260, 114279. https://doi.org/10.1016/j.apenergy.2019.114279 Yang, Y., Javanroodi, K., Nik, V.M., 2021. Climate change and energy performance of European residential building stocks – A comprehensive impact assessment using climate big data from the coordinated regional climate downscaling experiment. Appl. Energy 298, 117246. https://doi.org/10.1016/j.apenergy.2021.117246 Zou, Y., Lou, S., Xia, D., Lun, I.Y.F., Yin, J., 2021. Multi-objective building design optimization considering the effects of long-term climate change. J. Build. Eng. 44, 102904. https://doi.org/10.1016/j.jobe.2021.102904