Addressing the potential for improvement of urban building stock: a protocol applied to a Mediterranean Spanish case

Journal Pre-proof Addressing the Potential for Improvement of Urban Building Stock: A Protocol applied to a Mediterranean Spanish Case T. Bl ´ azquez, R. Su´ arez, S. Ferrari, J.J. Sendra PII: S2210-6707(21)00253-5 DOI: https://doi.org/10.1016/j.scs.2021.102967 Reference: SCS 102967 To appear in: Sustainable Cities and Society Received Date: 21 January 2021 Revised Date: 21 April 2021 Accepted Date: 23 April 2021 Please cite this article as: Bl´ azquez T, Su´ arez R, Ferrari S, Sendra JJ, Addressing the Potential for Improvement of Urban Building Stock: A Protocol applied to a Mediterranean Spanish Case, Sustainable Cities and Society (2021), doi: https://doi.org/10.1016/j.scs.2021.102967 This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2020 Published by Elsevier. 1 Addressing the Potential for Improvement of Urban Building Stock: A Protocol applied to a Mediterranean Spanish Case T. Blázquez a, *; R. Suárez b; S. Ferrari a; J. J. Sendra b a Department of Architecture, Built Environment and Construction Engineering (ABC), Politecnico di Milano, Milano (Italy) b Instituto Universitario de Arquitectura y Ciencias de la Construcción, Escuela Técnica Superior de Arquitectura, Universidad de Sevilla, Sevilla (Spain) * Corresponding author. E-mail address: mari[email protected] Full postal address: Dept. ABC - via Ponzio 31 – 20133 – Milano, Italy Highlights • Bottom-up protocol to rate urban stock energy behaviour before building upgrading. • Geographic Information System ease to store and analyse building-related data. • As built and upgraded energy models provide for buildings shortcomings. • The protocol is applied to a Mediterranean climate-dominated Spanish city. • The achieved benefits allow stakeholders in decision-making of buildings upgrading. Journal Pre-proof 2 Abstract Current plans to combat the environmental global impact of the building sector have encouraged research on sustainable actions for increasing the renovation rate of many buildings whose performance is far from meeting current energy standards. This paper presents a protocol for analysing the existing residential stock towards planning building upgrading actions. A Geographic Information System with building-related data is used to assess the housing stock following a bottom-up approach. The urban study sample is established using several criteria at building level. The potential for improvement of buildings is ascertained using DOE 2.2 energy models which incorporate standard passive strategies to reduce the current primary energy demand and the associated CO2 emissions. The protocol is applied to the existing public housing developments built in the southern Spanish city of Córdoba, under Mediterranean conditions, during the urban expansion period which spanned 1951 to 1980. The extrapolation of results to the urban study sample shows a 25 % reduction in CO2 emissions and a 40 % decrease in annual energy demand. Furthermore, the impact of typology and position on the energy behaviour of a given building is acknowledged. The benefits achieved aim to help stakeholders in the implementation of energy retrofitting urban plans. Keywords: public housing developments; Mediterranean housing stock; buildings energy performance; urban analysis; GIS for urban building model; UBEM. 1. Introduction Despite joint efforts from EU countries to reduce CO2 emissions further measures must be implemented if Europe is to reach the objectives set for 2030 [1]. A key issue for this is the Journal Pre-proof 3 renovation pace of existing buildings, as most European building stock is still in need of updating [2]. Establishing an effective strategy aiming to meet the European targets for energy efficiency in the housing sector first requires a thorough assessment of the energy deficit of existing buildings before implementing the appropriate retrofit strategies. In keeping with this, in recent years, many researchers have focused on the growing need to learn more about the status quo of the existing building stock [3], especially the residential sector. However, further in-depth research of the building stock (in situ assessment, monitoring of environmental parameters, user surveys, representable archetypes, prevailing patterns within the existing stock) is needed to establish more reliable margins for improvement [4]. Top-down and bottom-up approaches are often employed in the study of urban stock [5]. The former focuses on generalized data (i.e., national, or regional consumptions) attempting to predict individual aspects at a macro-scale within a population or stock [6], while the latter examines specific cases in order to identify common assumptions which can be applied to the rest of the analysed stock [7]. Lately, there has been a prevalence of bottom-up approaches due to a greater availability of disaggregated data and to the options for estimating the energy potential of individual buildings to support the update of urban areas [8]. There is currently no standard procedure for the assessment of urban building stock and various methods leading to a diversification of results can be found in the reviewed literature. According to M. Brogger [9], Urban Building Energy Models (UBEM) need to offer transparent results to ensure consistence in the energy performance of real cases, so that the energy improvement potential of building stocks can be estimated with precision. Several works have been found to validate bottom-up UBEMs with data on a national scale for energy demand [10], space heating and domestic hot water [11], and energy-use to address different Journal Pre-proof 4 climate-change scenarios [12], resulting in many cases in differences between simulated and measured results within a range of 3 % [13]. However, finding real and robust data is not always that straightforward. Macro-scale energy expenditure or consumption results and the environmental performance of individual buildings are not always publicly available, either for reasons of data protection or because of planning required for long-lasting tasks in order to obtain this kind of information. In this respect, UBEMs offer the possibility to evolve. This means that the databases generated can be established with the means readily available (i.e., cadastre data, age of construction, morphological building features, size, compactness, thermal envelope constructive definition, energy simulation results) and be updated continuously as energy-related data input becomes feasible and is used to support the results. Moreover, the assessment of existing building stock involves a broad sample of items to be addressed. A painstaking cataloguing process for building details can be time-consuming and add complexity to the procedure with no major benefit in results. However, excessive generalization can lead to an increase in cumulative errors and result in overestimations [14]. Therefore, these sorts of assessments must be based on balanced representative samples or building “archetypes” of the analysed stock, in an attempt to simplify the most prevalent architectural, constructive, and energy-related features at different scales (local, municipal, regional, or national) [7]. There is a strong correlation between influential energy-related parameters and urban building energy approaches in terms of the building age, type, and compactness ratio of outer surface to heated volume [15]. To simplify this task, the European project TABULA [16] is an effective tool to test wide scope urban building stock, guiding the generation of representative archetypes based on a study carried out over 13 European countries, which provided exemplary catalogues of local traditional building typologies [17]. Within this database several works have been found on Journal Pre-proof 5 representative national stocks. I. Ballarini [18] in Italy and E. Daskalaki [19] in Greece, among others, develop analyses of the local urban residential building stock in order to run energy performance assessments at territorial scale. P. Florio [20] establishes an algorithmbased relation between the French Enquête Nationale Logements (ENL) with TABULA defined typologies for the French stock, aiming to label existing buildings in order to certify their energy performance in a simplified manner. T. Loga [21] uses TABULA archetypes to generate simplified building stocks for each European country involved in the database, providing a cross-country benchmarking of building energy features at national and regional scale. D. Groppi [22] generates an UBEM to address thermal energy consumptions and usable solar energy potential in two Italian districts, transposing TABULA methodology to classify the building stock analysed by age and typology and correlating each category to a defined compactness ratio. Urban building energy approaches can also be used to plan global strategies or design energy policies to promote building updating [23] by analysing hindrances in the retrofitting process [24]; making the users aware of their homes’ potential for improvement [25]; detecting critical areas that need to be addressed beforehand [13]; or assessing the impact of socioeconomic features in the regeneration potential of the building stock [26]. Indeed, a common framework that blends and coordinates all the involved agents within a sustainable approach to building stock performance is still needed towards a clear path to address integrated energy planning strategies [27]. Lately, GIS tools have become essential in the urban stock analysis developed as they offer many possibilities linking spatial and temporal data with existing buildings and allowing macro-scale analysis at different urban stages in order to promote more sustainable city energy planning [28]. C. J. García-Ballano et al. [29] estimate energy saving in recently retrofitted near zero Energy Buildings using a parametrized GIS and conclude that the Journal Pre-proof 6 buildings’ potential for improvement closely correlate with the age, type, and thermal envelope composition, resulting in an improvement of around 30-58 % in energy behaviour in individual buildings. V. D’Alonzo et al. [30] substitute archetypes and energy simulation methods for GIS tools to estimate energy balance at building level in a sample of 42,000 buildings in the region of Valle d’Aosta (Italy), and achieving a useful methodology for estimating the heating demand of buildings which can also be applied at different scales. Recent studies focus on different parts of the world, examining different countries and climate zones. This is the case with M. Braulio-Gonzalo et al. [31] where a bottom-up approach to the building stock in a Mediterranean Spanish city is used to analyse buildings’ passive energy behaviour and the impact of urban morphology on indoor conditions. Also, in Spain, F. Martín-Consuegra et al. [32] study the impact of energy losses through the building thermal envelopes in an obsolete neighbourhood on the outskirts of Madrid, where the population is exposed to energy poverty. The methodology can be applied to any other building stock, using building cadastre data to estimate energy-related defined indexes (urban overall heat transmission coefficient, energy losses due to conduction across the envelope, urban energy efficiency indicators) by means of GIS tools. Based on data from the Italian cadastre, P. Caputo et al. [33] define a methodology to estimate energy consumption at urban scale by generating building archetypes representing the building stock in Milan (both residential and commercial). As well as the analysis of retrofit barriers in some Italian municipalities [34], they propose a new energy framework at urban scale and GIS tools to encourage building updating by correlating census data and real data, and expecting the results generated to be implemented continuously as more information becomes available [35]. A recent collaboration by P. Caputo on the use of 3D-GIS to study the retrofitting potential of façades with BIPV applied to the Swiss residential sector can be found in [36]. S. Torabi Moghadam et al. [37] present a work-in-progress database, a bottom-up model for estimating energy Journal Pre-proof 7 consumption for space heating of the residential stock applied to a medium-sized city in Italy, combining GIS tools and Multiple Linear Regression (MLR) analysis. By doing this, the authors highlight the most influential factors on building energy consumption (period of construction, heated volume, type of ground floor, occupation factor, air temperature, type of roof and the installed heating power). The final model thus becomes a powerful Multi-Criteria Spatial Decision Support System visualization tool which is used to define green urban systems. Another example of a mix of GIS tools and MLR analysis is included in A. Mastrucci et al. [38], where a bottom-up statistical approach is used on Rotterdam’s residential stock, estimating energy savings at dwelling level through typical retrofitting measures to complete the energy updating of 300,000 housing units within a more sustainable urban plan. Furthermore, bottom-up approaches can be used to forecast the annual energy consumption in future climate scenarios [39]. Although the specific aims are diverse (i.e., predicting energy-saving potential, analysing hindrances in the retrofitting process, raising awareness among users of their homes’ potential for improvement, etc), urban building energy approaches are generally used by political decision-makers, from local authorities to public and private stakeholders, as guidance to boost energy policies and urban planning for building upgrades. Nevertheless, to date few urban studies have been found on the current state of buildings under climate conditions such as those of the southern Mediterranean arc, with important daily fluctuations throughout the year, and with temperatures expected to increase in the coming decades. In fact, the common lack of thermal insulation and centralized active systems in a high portion of the existing stock make the buildings highly sensitive to weather changes. This, in combination with low-income user-profiles results in poor passive energy performance and frequent situations of thermal discomfort indoors [40]. Journal Pre-proof 8 This paper aims to bring to light the prevailing energy status quo of residential building stock located in the Spanish Mediterranean arc. In doing so, it presents a protocol for a territorial assessment of existing public housing developments created in urban expansion processes in the post-war period between 1951 and 1980. The proposed protocol is validated through its application to the existing residential stock of Córdoba, an inland city in the south of Spain. Differently from other works, the present research raises from a thorough description of the analysed building stock through an exhaustive documentation of Archive files and original projects, backing up results based on an extensive sample of real building features. The statistical screening of architectural and constructive prevailing patterns together with a crosscheck analysis of building census and cadastre data brings about the representative energy models that will serve as a base to assess the stock’s passive energy performance in terms of primary energy consumption and CO2 emissions, and to establish suitable standard strategies for large building stocks. The results above, in combination with GIS, enable passive upgrading actions to be applied at urban as well as district levels. The digitalisation of building-related information in a supporting dataset enables the accessibility to the constructive composition of the existing structures and shows the potential for improvement of extensive housing stock, providing public administrations with a valuable tool towards the design of comprehensive plans for a more efficient constructive sector. The introduction in this paper is followed by a further three sections. Section 2 presents the methods and materials employed and provides a definition of the study sample; an architectural, constructive and energy assessment; the linkage of spatial and alphanumerical data in a GIS database; and the generation of base-case and upgraded energy models. Section 3 discusses the results relating to the current state of the buildings and their potential for improvement with standard passive strategies. Finally, Section 4 presents the conclusions and future lines of study for implementing further research in this field. Journal Pre-proof 15 of multi-family buildings with standard typologies (H-plan, linear) with more than 100 housing units are considered in order to avoid incorporating cases with unique features that would hinder the robustness of global results. The buildings considered within the study sample are sorted based on the major indicator of age of construction. ii) level of information 2 (Loi2) focuses on detailed morphological and constructive information of the selected buildings, obtained from an in-depth examination of the available original project files found in the Municipal Archive of Córdoba [57]. Based on this, buildings are chosen considering accessibility to the necessary information. 2.2.2 Step 2: Structure of the GIS dataset Open-source software QGIS (v.3.2.0) [58] has been used to generate the GIS dataset. Through a mass download of shapefile information layers, a supporting map with urban cadastre cartography (municipal, neighbourhood and building levels) [56] is created. Starting with a mother layer with the registers of all the buildings selected, a set of specific information layers with building-related data is generated following the pre-established criteria for outlining the sample (Fig. 3). Information layers link graphical entities to dBase attribute tables. Graphical entities can represent built-up volumes at different levels of detail, from building level to the outlining of the municipal district, holding different data depending on the Loi (Fig. 3). For instance, a graphical entity representing the municipal district will contain the total number of housing units built in the years of study (Loi1), while a building-shaped graphic entity will hold general data for the building, obtained from online sources (Loi1), such as age of construction, typology or position, or a detailed description of the building’s thermal envelope drawn from the consultation of the original project files and plans (Loi2). Journal Pre-proof 16 Finally, the GIS dataset is made up of four Loi1 information layers (including a total of 1491 items and 28828 pieces of information) and eleven Loi2 information layers (including 7091 items and 126236 pieces of information). Fig. 3. Transposition of the vertical criteria adopted for outlining the study samples in the supporting GIS map. 2.2.3 Step 3: Characterization of the study sample 2.2.3.1 Architectural and constructive characterization To address the architectural nature of the study sample, buildings are sorted into categorized information layers associated to the corresponding Loi and the scale approached (Table 1): i) At Loi1, the analysis covers groups of buildings to single building level. Groups of buildings are characterized through indicators that are equally applicable to all the buildings Journal Pre-proof 17 covered, such as age of construction, total number of housing units and total built-up surface. Buildings are classified by prevailing characteristics, such as typology, orientation, position, and the number of storeys. ii) At Loi2, the study covers from building level to housing unit scale. Buildings are characterized by their compactness (ratio between the enclosed volume by the thermal envelope of the whole building and the addition of the thermal exchanging surfaces of the enclosing envelope). Housing units are defined by their usable surface, number of bedrooms (as an indicator of how many people live in them), and floor-to-ceiling height. Table 1. Scope and categorized information layers associated to architectural indicators. Scope Title Categories GIS of architectural characterization of housing buildings Loi1 Number of housing units <200 200-400 400-600 600-800 800-1000 >1000 Construction decade 50 60 70 80 Typology H-plan building linear building residential tower others Prevailing orientation N-S E-W Northeast-Southwest Northwest-Southeast Building position freestanding inner-positioned end-positioned Number of storeys 2-4 4-6 6-8 >8 Loi2 Dwelling usable surface <50 50-65 65-80 >80 Number of bedrooms 3 4 Type plan floor-to-ceiling height 2.20-2.35 2.35-2.50 2.50-2.65 2.65-2.80 V/S ratio <1.00 1.00-1.50 1.50-2.00 2.00-2.50 >2.50 To detect typical constructive systems employed in the study sample, buildings are clustered following categorized information layers that address the composition of several thermal envelope elements (Table 2). These include the roof composition and trafficability, outer wall composition, ratio of openings per opaque surface in each façade, window joinery and glazing and the composition of internal partitions in contact with non-habitable spaces. Journal Pre-proof 18 Table 2. Scope and categorized information layers associated to constructive systems. Scope Title Categories GIS of constructive definition of housing buildings Loi2 Type of roof pitched Andalusian Catalan double roof non defined Roof trafficability crossable non crossable Wall solution double layer single layer Ratio of openings per opaque surface <20 % >60 % 20-40 % 40-60 % Windows joinery metallic frame non defined Windows glazing non defined single glazing 3/3 single glazing 6/6 Wall solution adjacent to non-habitable spaces 1 foot ½ foot double-brick layer double-air brick various non defined non existent 2.2.3.2 Energy characterization In line with the Spanish Technical Building Code [43] and the Royal Decree of Energy Certification of Existing Buildings [59], an acknowledged ad hoc Spanish software named CE3X (v.2.3) with DOE 2.2 calculation engine [60] is used to assess the buildings’ current energy performance and foreseeable energy improvement through standard passive strategies. The software carries out a calculation procedure based on comparison scenarios which refer to a forecast of the energy performance of the existing buildings. This is supported with a wide dataset of results from the energy simulation of representative buildings considering the possible combinations in terms of age of construction and typology (number of storeys, constructive system, the ratio of glazing surface and the main façade orientation), the different Spanish climate zones and the four main orientations. When the user generates a building energy model, the software provides an energy indicator which includes a letter from A to G in a decreasing scale of efficiency (A being the most efficient), and a corresponding value in kgCO2/m2, regarding the primary energy consumption and the subsequent CO2 emissions, as a result of the “comparison scenarios” [61]. For each building, the calculation offers a general Journal Pre-proof 19 indicator, together with five other itemized ones for heating and cooling demands (in kWh/m2) and heating, cooling, and domestic hot water emissions (in kgCO2/m2). After the simulation of all the buildings considered, results are transferred to the corresponding information layers to sort buildings by energy rating both at a prior-to retrofit stage and under an upgraded scenario (Table 3). Table 3. Scope and categorized information layers associated to the buildings’ energy rating. Scope Title Categories GIS of energy assessment of housing buildings Loi2 Energy rating D E F Heating demand <15 kWh/m2 15-30 kWh/m2 30-45 kWh/m2 45-60 kWh/m2 >60 kWh/m2 Cooling demand <20 kWh/m2 20-30 kWh/m2 >30 kWh/m2 GIS of scenarios for energy improvement of housing buildings Loi2 Upgraded energy rating D E Upgraded heating demand <15 kWh/m2 15-30 kWh/m2 30-45 kWh/m2 >45 kWh/m2 Upgraded cooling demand <15 kWh/m2 15-20 kWh/m2 >20 kWh/m2 To close the cyclical approach and return to the study sample scale, the reductions achieved in energy demand and CO2 emissions obtained with the representative energy models are extrapolated to the rest of the buildings included within the study sample but not accounted for within the Loi2. Following a similar procedure to that identified in related works [62], buildings are clustered by year of construction and number of housing units per building. Then, averaged results for energy demand and emissions in each building cluster are converted through a coefficient-ratio between the gross and net residential surface of the buildings. Eventually, the converted results are correlated with the overall number of buildings within the study sample to broaden the impact of the enhancements. Journal Pre-proof 20 2.2.4 Step 4: Open-source dataset as a tool for assessing housing stocks To promote access to the information stored in the GIS dataset, an easy-to-handle tool with a user-friendly interface is generated with an open-source plug-in, qgis2web, for the software used [63]. The tool provides qualitative or visual features of the buildings through several interactive web maps to illustrate the spread of building parameters throughout the sample analysed. Furthermore, thanks to the detailed building description in the GIS dataset generated, the tool provides quantitative or statistical outcomes through simple calculations about the number of cases sharing similar features. The tool is based on use of the web map GIS of identification and definition of housing buildings (see supplementary material) to support the GIS dataset. This web map presents building-related data sorted by information layers with different scopes: i) within Loi1, urban nucleus, neighbourhood, groups of buildings and buildings; ii) within Loi2, groups of buildings, buildings, structure, and thermal envelope constructive systems (roof, floor in contact with outdoor air, floor in contact with the soil, vertical partitions in contact with nonhabitable spaces, horizontal partitions under/on top of non-habitable spaces, façades, and openings). Graphical entities within each information layer will appear providing their host layer is switched on, and the information of each building will appear in a pop-up window whenever a graphical entity is selected. Another four web maps stem from the supporting one: i) GIS of architectural characterization of housing buildings; ii) GIS of constructive definition of housing buildings; iii) GIS of energy assessment of housing buildings; and iv) GIS of scenarios for energy improvement of housing buildings. In these maps, buildings are sorted into categorized information layers according to building-related indicators. The five web maps generated can be launched in any internet browser and can be consulted in the supplementary material of this article. Journal Pre-proof 21 2.3 Representative energy models As mentioned above, the energy characterization of the buildings is carried out through the energy simulation of both the prior-to-retrofit state and the predicted scenarios for improvement. Considering the wide range of buildings analysed, in order to account for each possible combination energy models will be developed according to i) building typology, distinguishing between H-plan, linear, residential tower and others; ii) building position with respect to other adjacent or nearby buildings, classed as end-positioned (for those standing at the ends of a row of buildings), inner-positioned (for those standing in between of a row of buildings) or freestanding; and iii) the building orientation, according to the Appendix A of Building Code Basic Document on Energy Saving [43]. In all, a total of 212 different buildings representing 781 buildings are defined. To shorten the time spent on the evaluation of this high number of cases, energy models will be generated at building level, just taking into consideration the volume covered by residential use for the purposes of energy simulation. Accesses to the building and vertical core (staircases or lifts), as well as ground-floor commercial premises, garages, and other nonresidential spaces, are considered adjacent zones. Input data will involve parameters recorded such as location, climate zone and thermal envelope definition, together with standard use and operation profiles, centralized active thermal conditioning (if applicable) and calculated domestic hot water supply [43]. To foresee the buildings’ potential for improvement, the reduction in energy demand and corresponding primary energy consumption and CO2 emissions are considered, comparing asbuilt energy models depicting prior-to-retrofit state with upgraded ones, which incorporate standard passive strategies for energy improvement. In line with the desired urban-scale Journal Pre-proof 22 applicability, the protocol entails an initial analysis at building level and a later transposition at urban level, to measure the global impact over the study sample. 2.3.1 Base-case energy models The definition of as-built energy models implies: • Current location and outdoor conditions, obtained from a downloadable weather file [64] with reference climate data for zone B4 [43], also corresponding to Csa [44]. • Normalized profiles for residential activity (internal loads) according to [43] depending on the use, density of internal heat sources (low, medium, or high) and period of use (8, 12, 16 and 24 hours). • The constructive and energy characterization of the buildings’ thermal envelope. For each individual element, the following are included: surface of thermal envelope, U-Value, mass, orientation, depth, perimeter, use of non-habitable spaces, number of similar items, G-Value, and solar protections. • Thermal bridges, which are considered through linear U-Values, such as the intersection between roof-façade, slab-façade, inner pillar-façade and corner pillar-façade, exterior floor-façade, soil floor-façade, opening edges-façade, and shutters box-façade. • The existing thermal conditioning systems. According to the original files [57], the analysed stock generally lacks centralized active systems (HVAC). In line with [43], energy models are implemented with natural gas-based heating production and electricitybased cooling production. • The ventilation rates for private residential use are set at a default value according to a standardized profile of operation [43]. Journal Pre-proof 23 • The domestic hot water supply relies on an LPG boiler without accumulator. For private residential use, total daily water consumption is set according to [43]. 2.3.2 Upgraded energy models To improve the estimated energy performance of the buildings analysed, a set of standard passive strategies has been incorporated into the representative energy models. After a sensitivity analysis in previous studies [41], an all-inclusive proposal (AIP) has been defined integrating the following hypothesis: AIP = HTI + HW + HV1 + HV2 (0.4 ACH) where: • HTI implies the addition of thermal insulation. Façades show improvement with a 50-mm CO2 PUR injection into the existing air-chamber, while roofs are implemented depending on whether they are flat (with 60-mm filter slab) or sloping (with 80-mm filter slab). This strategy led to the U-values of walls and roofs complying with current regulations [43], with a decrease of 70 % and 80 % W/m2K respectively and U-value deviations below limit by 4 % in walls and 12 % in roofs (Table 4). • HW means implementing window glazing depending on orientation. Existing single glazing is replaced with insulated glazing units (IG units) with a 16-mm air chamber and variable glazing thickness (4-mm external and 6-mm internal) to improve acoustic behaviour, together with Low-Emissive coatings for north-facing openings and Solar Control finish for south-facing ones. Like the previous strategy, it helps to meet U-values [43] and provides an overall reduction of 75.5 % and 73.7 % W/m2K, respectively (Table 4). Moreover, the Solar Heat Gain Coefficient (SHGC) values go from 0.83 to 0.62 and 0.52 in the respective orientations. Journal Pre-proof 24 Table 4. Averaged improvement in implemented thermal envelope solutions with AIP (HTI and HW) throughout the analysed stock and deviation according to the limit U-value for a B4 climate zone [43]. Position Averaged U-Value Upgraded U-Value Limit U-Value Average Improvement (%) Average Deviation (%) Opaque elements Vertical 1.80 0.54 0.56 70.0 3.6 Horizontal (roof) 1.89 0.39 0.44 79.4 11.4 Transparent elements North-oriented (low-emissive) 5.7 1.4 2.3 75.5 39.1 South-oriented (solar control) 5.7 1.5 2.3 73.7 34.8 • HV1 is for increasing window airtightness by replacing general metallic joinery with PVCmade ones, reducing air-leakage with Type 4 folding frames (3 m3/h/m2 under 100 Pa) [65]. • HV2 sets up a controlled renewed air-rate through a hygro-adjustable ventilation system that relies on occupation and shifts to natural ventilation during the night in summer [43]. For the purposes of calculation, air rates have been estimated at 0.4 ACH. 3. Results and discussion The following section offers an overview of the achieved tool and the results from its application to the housing buildings built in Córdoba between 1951 and 1980. 3.1 Identification and definition of the study sample buildings The application of the methodology to the housing stock built between 1951 and 1980 in the city of Córdoba has enabled the identification of 232 groups of buildings, consisting of 1699 buildings and 30079 housing units dating from the period studied. According to the Journal Pre-proof 31 however, a trend to enlighten structural solutions by incorporating non-ventilated airchambers in roofs and opaque façade solutions in the period studied, although these were not efficient enough. Generally, façades are made mostly of two-layer brick solutions with plaster finishes: an outer double-brick wall, a non-ventilated vertical air chamber, and an inner single-brick partition, with an averaged U-value among the analysed buildings of 1.8 W/m2K (Table 4). These have been detected in 70.12 % of the total executed façade surface and are found in 71.70 % of the buildings studied. The use of single-layer brickwork solutions during the 1950s is also notable. Roofs are largely composed of horizontal slabs with non-ventilated solutions, which are found in over 48 % of the surface of the study sample buildings, while pitched roofs with a fibrocement finish layer are found in around 40 % of the roof surfaces analysed. The lack of any kind of insulation results in U-values of up to 1-2.5 W/m2K of the roofs identified (Table 4), commonly resulting in dense and massive solutions. In most cases roofs could not be walked on. But for more than 50 % of the study sample, for which no description of the openings was found, windows are generally double-glazed (with two 3-mm panes of glass) or single-glazed (one 6-mm piece of glass) and aluminium frames. No inner air-chambers or thermal breaks were identified. Solar protection in the form of blinds was preferred for use on bedroom windows but only identified in 50% of the buildings analysed. In addition to building shape, building orientation has been addressed for its importance in a southern Mediterranean climate. Four main categories – North-South (N-S), East-West (EW), Northeast-Southwest (NE-SW) and Northwest-Southeast (NW-SE) – were established for ease of analysis despite the fact that no regular pattern is observed among the buildings Journal Pre-proof 32 analysed. H-plan buildings present a similar percentage in all directions, while linear buildings are mostly in N-S (34.48 %) and NW-SE (40.00 %) orientations. Residential towers with square floor plans are also found in the four main orientations (N-S-E-W) (See supplementary material). 3.3 Energy characterization of the study sample buildings The following section acknowledges the energy status of the buildings studied estimating the amount of CO2 emissions and their annual energy demand through energy simulation, as well as the potential improvement of their energy performance with upgraded energy models. 3.3.1 Base-case energy models: Energy grading, CO2 emissions and annual energy demand The application of the protocol to the housing sample has enabled the estimation of energy indicators and CO2 emissions for the existing buildings. According to the energy rating scale of residential buildings, cases range from D to F, although most of them are rated E (96.6 %) (Table 6). A general trend towards a better performance can be seen, with E-rated buildings’ initial averages ranging between 36.0 kgCO2/m2 and 32.9 kgCO2/m2 when examining the final years of the period under study (Table 6). Journal Pre-proof 33 Table 6. Averaged energy rating and percentage of similar housing buildings according to their decade of construction. Averaged energy grade (kgCO2/m2) Percentage of buildings (%) Decade of construction D E F D E F 1951-1960 - 36.0 45.0 - 11.1 2.9 Fraction built 1951-1960 0.0 79.4 20.6 1961-1970 18.9 34.4 - 0.5 51.9 - Fraction built 1961-1970 1.0 99.0 0.0 1971-1980 - 32.9 - - 33.6 - Fraction built 1971-1980 0.0 100.0 0.0 Total 18.9 34.1 45.0 0.5 96.6 2.9 Equally, the annual energy needs for primary energy decrease during the study years (Fig. 7). In fact, heating energy demands initially reached around 80 kWh/m2, decreasing over time but remaining above 35 kWh/m2 in all buildings, thus increasing the annual global energy demand (Fig. 7). The results for cooling energy demand are within the 10-30 kWh/m2 range, although linear buildings dating from the 1970s occasionally display lower values (Fig. 7). Fig. 7. Annual mean primary energy demand (kWh/m2) during the study period according to building typology. Journal Pre-proof 34 3.3.2 Upgraded energy models: Improved energy performance and reduction of CO2 emissions Buildings in which the AIP has been implemented have been rated from D to E according to the energy rating scale for residential buildings. Table 7 compares CO2 emissions and average energy rated for the evaluated buildings prior to and after the incorporation of the AIP, sorted by decade of construction and presented along with the percentage of buildings with similar energy performance. A general improvement can be observed as none of the buildings is F-rated after the AIP. 97.1 % of the buildings within the E-range display an average value of 26.2 kgCO2/m2, 2.9 % of the buildings are rated D, with an average value of 18.06 kgCO2/m2, which overall constitutes a 30 % reduction in CO2 emissions (Table 7). Table 7. Average energy rating of buildings and percentage of buildings with similar rating, before and after AIP implementation. Averaged energy rating (kgCO2/m2) % of buildings with similar rating Before AIP After AIP Before AIP After AIP Decade of construction D E F D E F D E F D E F 1951-1960 - 36.0 45.0 - 27.1 - - 11.1 2.9 0.0 14.0 - 1961-1970 18.9 34.4 - 17.6 25.9 - 0.5 51.9 - 2.1 50.3 - 1971-1980 - 32.9 - 19.3 26.4 - - 33.6 - 0.8 32.8 - Total 18.9 34.1 45.0 18.1 26.2 - 0.5 96.6 2.9 2.9 97.1 - 3.3.3 Impact of the AIP on the energy performance of the study sample Figure 8 shows the achieved deviations in annual CO2 emissions and primary energy demand for heating and cooling after the implementation of the AIP in the buildings assessed and the transposition of results to the study sample. Journal Pre-proof 35 An overall improvement of results can be observed throughout the study sample, highlighting the major potential for a better performance of these buildings. A mean decrement of 23.3 % in CO2 emissions is achieved for current environmental behaviour, from a previous mean of 35.7 kgCO2/m2 to an upgraded one of 27.4 kgCO2/m2 (Fig. 8). Following the transposition of results to the study sample, a major impact of the AIP on the heating demand reduction of the analysed buildings is observed. Considering a previous average value of 69.3 kWh/m2 for annual heating, the AIP offers a reduction of over 33.4 % with an average heating demand of 46.2 kWh/m2 (Fig. 8), although there is still room for improvement. As regards annual cooling, the AIP results in a full amelioration of cooling needs over 41.6 %, ranging from a prior average value of 29.3 kWh/m2 to an average of 17.0 kWh/m2 annual cooling after the retrofit (Fig. 8). Globally, annual energy demand has decreased by 35.8 % on average, from an initial mean value close to 98.6 kWh/m2 to a lower mean value of 63.2 kWh/m2. Fig. 8. Annual mean CO2 emissions, primary energy demand and deviation (%) before and after AIP implementation. Sorting the study sample by years of construction can highlight a greater need for upgrading the older stock. Greater deviations can be observed in the buildings dating from 1951-60, with reductions of 21-34.2 % for CO2 emissions in most cases, 29.8-53.9 % in heating demand, and 37.4-48.3 % in cooling demand (Fig 9). Towards the end of the period studied variability Journal Pre-proof 36 in these ranges is just 1-3 %, with reductions of 21.0-22.2 % in CO2 emissions, and 27.7-30.3 % and 39.3-40.7 % in heating and cooling demands, respectively (Fig. 9). In all, major reductions of 24.6 % in CO2 emissions, 36.5 % in heating demand, and 43.5 % in cooling demand can be achieved, resulting in some cases with reductions of almost 35 % in CO2 emissions and 55 % in primary energy demand (Fig. 9). Fig. 9. Annual deviation (%) for CO2 emissions (a; left) energy demand for heating (b; middle) and cooling (c; right) after AIP implementation. As part of the analysis of the impact of the AIP on these buildings the annual mean deviations in energy demand have been addressed through morphological parameters such as typology, position, and orientation, which impact the building’s energy performance (Figs. 10 and 11). In this sense, significant deviations are observed in buildings with a higher exposed surface, such as linear buildings (with average decreases in heating and cooling of 38.5 % and 42.0 %, respectively) (Figs. 10a and 11a) as well as freestanding buildings (with average decreases in heating and cooling of 39.1 % and 44.8 %, respectively) (Figs. 10b and 11b), other than more compact buildings such as residential towers (with average decreases in heating and cooling of 26.3 % and 39.7 %, respectively) (Figs. 10a and 11a) or inner-positioned buildings (with average decreases in heating and cooling of 33.5 % and 40.8 %, respectively) (Figs. 10b and 11b). Journal Pre-proof 37 Clear differences can be observed between the range of deviations obtained in freestanding buildings and those located in end or interior positions. While the former present deviations of around 30.0-47.2 % for heating and 39.4-48.6 % cooling demand, inner-positioned buildings range from 29.1-36.5 % for heating and 38.9-41.0 % for cooling, while those at the end are 27.7-33.0 % for heating and 39.4-41.5 % for cooling (Figs. 10b and 11b). Results do not shed much light on the impact of the buildings’ orientation on their energy behaviour. Nevertheless, the orientations with the most solar exposure present noticeable deviations thanks to the strategies implemented (Figs. 10c and 11c). Heating deviations around 28.1-42.3 % for Northeast-Southwest orientations show that these buildings may benefit from higher solar gains in winter (Fig. 10c), while other buildings with different orientations present 8-9 % ranges, around 28.4-36.5 % for East-West ones, 27.6-36.5 % in the North-South ones and 28.4-37.2 % in the Northwest-Southeast ones. Similarly, both Southeast and Southwest orientations increase the AIP benefit in the buildings, with average deviations of 41.8-42.3 % of cooling demands, and reaching maximum reductions of almost 50 % of the original cooling loads (Figs. 10c and 11c), picturing that solar control strategies also contribute to a better energy performance under a Mediterranean climate. Fig. 10. Deviation in annual heating demand (%) sorted by typology (a; left) position (b; middle) and orientation (c; right) after AIP implementation. Journal Pre-proof 38 Fig. 11. Deviation of annual cooling demand (%) sorted by typology (a; left) position (b; middle) and orientation (c; right) after AIP implementation. 4. Conclusions The proposed protocol in the present research aims to contribute to the body of knowledge by bringing to light the energy status quo of existing residential stock in Mediterranean areas. In the light of results, the collection of real buildings-related data at several detailed levels (Loi1 and Loi2) provides for a supporting source of information towards subsequent assessments of the environmental performance applicable to extensive residential stock. Within a foreseeable common integrative framework for sustainable projects at urban scale, accounting for real buildings features will enable more accurate assessments in the context of the building’s location. The use of GIS technology has become advantageous in the analysis of territorial sets of buildings, incorporating the development of a useful tool into the approach of overarching strategies for energy improvement. This paper tests the suitability of a GIS tool, designed through a protocol for assessing the energy status quo of existing housing stock and its potential for upgrading. The ongoing nature of the tool foresees future lines of implementation to ensure more precise results as more information becomes available. The case study examined, the urban stock of the city of Córdoba, has served as an example to demonstrate the potential effectiveness and scope of this method, which could also be wellJournal Pre-proof 39 suited for assessing other building stocks. The protocol has facilitated the detection of common architectural and constructive patterns throughout the different buildings selected, later used to define an all-inclusive proposal (AIP) to implement the analysed constructions. However, focusing on a specific section of the housing stock results in an unequal size for the variables analysed throughout the study sample (i.e., typologies, orientations). Whilst the accuracy of results is increased in some cases such as H-plan buildings, it is necessary to increase the number of other types of buildings (linear buildings and residential towers) in order to ascertain the extent to which integral passive strategies may improve energy performance in different types of housing stock. Among the study sample, the widespread lack of thermal insulation causes 96.6 % of the buildings to score within the E-range, with average emissions of 34 kgCO2/m2 and annual heating and cooling demands of around 60-70 and 20-30 kWh/m2, respectively. However, among buildings from the second half of the study period a trend can be observed which shows improvement, coinciding with the implementation of lighter constructive solutions thanks to the incorporation of non-ventilated air chambers. The transposition of an all-inclusive proposal (AIP) to the study sample analysed at urban scale leads to an overall benefit of 25 % less of CO2 emissions into the atmosphere, thanks to a reduction of one level in energy ratings among 3 % of buildings from the 1950s (from F to E), as well as in 3% of buildings from the rest of the sample (from E to D). Furthermore, decreases of 36.5 % and 43.5 % in the annual heating and cooling demands respectively are obtained, embodying a major decrease of 37.7 % in primary energy demand at territorial level. The AIP defined implies a general upgrading of the buildings’ thermal envelope, constant and controlled air renewal, night thermal dissipation and solar control, contributing to a greater extent in those cases with greater exposed surface areas. Less compact typologies, such as Journal Pre-proof 40 linear buildings, and freestanding buildings seem to benefit to a greater extent from the improvement brought about by the AIP (39.4-40.7 % average energy demand reduction). Moreover, the AIP provides a more controlled solar incidence inside the buildings, overall, in the most penalized orientations, East-West and Northeast-Southwest, with a reduction of up to 51.6 % and 48.1 % in annual energy demands, respectively). The resultant energy performance indicators of the analysed buildings serve as a reference when similar contexts are addressed. Actually, the application of the proposed protocol to an existing building stock supports its replicability at different scales (e.g., district, urban, territorial) and its adjustable nature to other contexts contributes to broaden the accessibility to information on the existing residential stock in a common European frame. Further lines of research could consider, within the digital transformation of the constructive process, the integration of the generated GIS dataset with a BIM environment, to boost the implementation of the generated tool and assist in the sustainable approach to the project’s management. Moreover, cost-effective assessments of the proposed strategies would provide for an estimation of their investment returns. Indeed, it is expected that the protocol achieved will be incorporated into decision-making processes, encouraging regional public administrations and private stakeholders to boost upgrading plans of the city stock in order to meet the goals set for existing buildings within the current Horizon Europe framework. CRediT authorship contribution statement Teresa Blázquez: Conceptualization, Methodology, Software, Writing, - original draft. Rafael Suárez: Conceptualization, Methodology, Writing - review & editing. Simone Ferrari: Conceptualization, Methodology, Writing - review & editing. Juan José Sendra: Conceptualization, Methodology, Writing – review & editing. Journal Pre-proof