(**a**) (**b**)

**Figure 5.** Extract of the PEDRERA model approach, expected visualization in the urbanZEB tool: (**a**) concept design based on financial appraisal and KPIs prioritization; (**b**) scenario visualization for scenario selection according to priorities from stakeholders. Source: authors.

#### *PEDRERA Model Input*

Table 1 provides a summary of the inputs required by the PEDRERA model to properly run simulations. Most of the input data for the algorithms are already collected from public sources and from case study sources and integrated in the urbanZEB tool (e.g., cadaster, energy performance certificates and census, among other) by Semantic Web processing.

Each input is systemized in the model framework in order to predicts the suitable output (KPIs) for given inputs. The input's features are namely: type of **issue** and type of **parameter**, **indicator**, **unit**, **scope** (the goal where input is adopted), **data source** (imported from database, calculated, simulated), **scale** (building, district, urban area/census unit) and **source dataset**.

Therefore, the PEDRERA model will run data through a special wizard able to filter input data which are already displayed and shared on the urbanZEB web platform. Through Geographical Information System (GIS), Big Data and Extract, Transform and Load (ETL) techniques these sources are integrated in a Data Warehouse that allows the dynamic crossing of the different levels of information. The connectors will be implemented with API calls, data dump digestion or in web scrapping techniques based on the characteristics of the sources. In this way, the processes are configured to cover specific case studies on extended urban areas. The repository is implemented as a PostgreSQL/PostGIS database in order to facilitate displaying the data in map visualizations. Data sources that are not published in a structured format such as CSV, RDF, API REST and XLS, among others, are processed using web scrapping techniques. The database file stores all the information resulting from the characterization phase of the project (with all the parameters and indicators) and the prioritization indices. In addition, to facilitate the reading and visualization of the data, an X-ray file (ArcMap) holds the layers of information established in the project and their set of symbols.


**Table 1.** List of main input implemented and managed in the PEDRERA model. Source: authors.

**Table 1.** *Cont.*



**Table 1.** *Cont.*

The inputs are organized according to both tags and specific parameters, and refer to special feature types adopted in the model engine to calculate the KPIs: building attributes, cost analysis, user types, architectural characterization, energy characterization, renovation measures, economies of scale, climate and environment (Figure 6).

As described above, one of the main information required for the definition of potential scenarios is represented by the architectural characterization feature type of the residential stock. For this purpose and with the aim of simplifying the assessment of building performance in large-scale retrofitting programs, different building archetypes are used (Figure 7 and Table 2).

The description of archetypes, which comes from the methodology used in the national Grupo de Trabajo sobre Rehabilitación (GTR) 2011 report and the LTRS 2020—set out in the EPBD 2010/31/EU—allows the understanding of the residential stock based on the segmentation in groups of buildings that present similar conditions and therefore require similar intervention actions [31]. This characterization is focused on setting the parameters that will have the greatest impact on its energy performance, as well as its related weakness, in order to design the best intervention strategy and assess the economic investment required. This phase is mainly based on processing extraction of cadastral data and is available throughout the urban classifications and by generating more than 300 cross variables related to: location, use and areas, type of residential property, year of construction, number of floors, number of dwellings.

In addition, the collected data include the performance of each building element and envelope—facades, internal wall, roofs, floors and ground floors—as well as the incidence form, the windows performance, rate, size and type. In this way it is possible to carry out the geometric modelling of each building, and to define the features and surfaces of each level on façades and roof, detaching patios and interferences with neighboring or surrounding buildings (e.g., shadows cast) that could be adopted for both the energy demand and the renewable energy sources (RES) production KPIs.

**Figure 6.** Main input implemented and managed in the PEDRERA model framework. The inputs are organized in compliance with the four issues, and each of them is adopted in the algorithms for the evaluation of KPIs according to the scopes. Source: PEDRERA.

**Figure 7.** Aerial view showing the filters applied on residential buildings, and presenting the 12 archetypes based on year of construction and typology. Source: urbanZEB web platform.


**Table 2.** List of main building archetypes implemented and managed in the PEDRERA model. Source: urbanZEB web platform.

An overall overview of the construction systems is also made for each building according to the classification into clusters, thus displaying the current building state and enabling the design of post-intervention building scenarios resulting by the application of each intervention menu provided by the model (i.e., input from renovation measures A, B and C as shown previously) (Figure 4, Table 1) within the feature type **renovation measures**. Simultaneously with the architectural characterization parameters, the **energy characterization** inputs are focused on defining the current energy level and the potential from reducing the demand and the energy consumption through the intervention for single buildings. As explained above, for the determination of this set of input, the reference values of the national GTR Reports are used as defined by the Technical Building Code (CTE) and the Institute for Diversification and Energy Saving (IDAE) [2].

For the calculation of energy costs and CO2 emissions of materials and intervention belonging to a typological cluster, the Catalan database of construction elements (BEDEC) from the Institut de Tecnologia de la Construcció de Catalunya (ITeC) is available and adopted. Furthermore, according to a PEDs perspective, and the positive energy balance referred to a cluster of buildings (Building Portfolio) at the neighborhood scale, the calculation of the energy performance (EP) follows the overarching framework of the EN ISO 52000-1:2017. The new EN ISO 52000 family of standards to assess the EP of buildings offers a great flexibility in the calculation according to different choice of assessment methods defined [32]. Otherwise, a stochastic model supports a wide volume of simulations from the very beginning of the renovation process by using statistical information [33].

A territorial **socio-economic index** (IST) from Catalonia region provided by the IDESCAT [34] is adopted. This information, based on the Spanish National Institute of Statistics (INE) census, means a synthetic index by small areas (census units) that aggregates in a single value several socio-economic characteristics of the population.

That information is very relevant not only for the evaluation of the economic effort allowable by the users or the revolving funds required to be covered by the Public Sector, but also to predict the impact of the user' profile on electric consumes or load while designing PEDs. The index concentrates information on the employment situation, educational level, immigration and income of all people living in each territorial unit, based on 6 sectoral indicators. The IST is a relative index, with no units of measurement. A reference value for Catalonia region is established equal to 100, thus each unit is valuable in comparison with this average value. Values per decile are also referred: the first decile refers to the areas with the lowest socio-economic level and the tenth decile to that ones with the highest level.

The PEDRERA model provides a list of **renovation measures**—integrated in the renovation strategy menu—that, based on operative, environmental, economic and social input gathered from the selection of buildings, will enable users to evaluate the estimative metric computation, but also to simulate the energy improvement on the building according to established menu of intervention already available within the model, as well as the other benefits on welfare and security. The renovation measures included in the PEDRERA model are functional in the large-scale refurbishment projects for every cluster of residential buildings selected, as previously described (Figure 1). The menu of intervention refers to 3 main targets of intervention, with a specific type of solution that can be implemented or replaced within the residential buildings:


With regards to the **cost analysis**, the inputs are mainly managed within the model instead to be gathered from an external data source. Inputs are finalized to provide specific information on potential **investment KPIs** and several co-benefits to be achieved through the renewal process, and that are profitable and/or feasible for the actors involved in the process. Hence, the building renovation assessment does address and disclose the co-benefits from the renovation measures adopted and the economies of scale, the RES production, the appreciation in housing value, the increase of building lifespan, the impact on the environment and the improved health benefits for both the households and the healthcare system [28]. At the same time, the data and related benefits motivate and empower all stakeholders and target groups to do action. In this way, the city planners can focus on the most beneficial areas in both energy and quality renovation efforts, while the investors can get better access on the information of building stock, which eases making the financial decisions related to large-scale renovation projects benefits and risks. Then, data increase building owner's interest on the performance of their buildings when they can compare their consumption and renovation need, and they can see the potential derived from the renovation and the related businesses of energy improvements (e.g., the estimation if the roof renovation can be combined with PV panel installation and the benefits to establish an energy community) as shown in Table 3.

#### **3. Results**

Preliminary results obtained by the ongoing PEDRERA project refer to the definition of the conceptual framework of the model and in regards to: (a) data sources aggregation according to the four domains described above; (b) input required for the KPIs calculation (Table 1) that can be assumed to assess different "scopes". Aligned with this vision, the model is powered by the integration of the processed input for the calculation of the most relevant KPIs (outputs) algorithms according to each process phase and stakeholder's profile (Table 3).

#### *PEDRERA Model Ouput*

Table 3 summarizes the outputs provided by running a typical simulation with urbanZEB tool powered by the PEDRERA model. It must be noted that the outputs are assigned to five main targets related with the **scopes**: (i) financial appraisal, (ii) renovation strategy, (iii) energy community, (iv) welfare and security, (v) marketing.

In order to complete the panel of KPIs requested from large actions on building stock, the project focuses also on the perspectives of selected target groups that can be considered as key actors for urban development. The main considered stakeholders that would manage scenarios from the data-driven model are mainly:


**Table 3.** List of output (KPIs) of the PEDRERA model according to the "scopes" and stakeholder engagement for each phase of the process. Source: authors.



**Table 3.** *Cont.*

The stakeholders will be different beneficiaries of the simulated KPIs according to each phase of the renovation process as reported in Table 3. Due to the analysis of the current state and the model's accuracy, stakeholders are allowed to influence the model's computation phase by selecting and modifying the desired settings for both the calculation and data, by providing detailed input. Moreover, according to a wide PEDRERA approach beyond a reliable business tool, the model is meant to offer additional services, empowered by its next implementation into the urbanZEB tool web platform. For this reason, it is necessary to first establish each participant stakeholder and their characteristics clearly in order to effectively define the model results.

With regard to the **financial appraisal**, the design of the model algorithms has been concluded and tested on real large-scale renovation processes (i.e., the "ACR Pirineus" intervention on 32 buildings in Santa Coloma-Barcelona) [35]. In addition to its validation, the sensitivity analysis of the financial appraisal model was conducted (Figure 8) on 200 scenarios in order to determine how target variables (KPIs) are affected based on changes in other input variables. This simulation analysis refers also to show the outcome of a decision given a certain range of variables of the inputs (e.g., the execution budget (PEM), the number of entities involved in the process, the percentage of grants and of defaulters, the operational cost from the private partner or the financial costs in a shifting PPP model) in order to evaluate, for example, the revolving funds impact on the Public Administration or the payment rates covered by each user type, among others. Since various private and public financial mechanisms for energy renovations in buildings are currently available, the financial appraisal can range from well-established and traditional

mechanisms such as grants, subsidies and loans to emerging new models and other oriented PPP models. The test beds demonstrated that, although public bodies are typically involved in large-scale retrofitting projects, the majority of them are only partially engaged, often playing a role in the subsidy plan or, occasionally, by allowing the legal framework to adapt to local conditions. At the same time, several end-user typologies need to be identified in order to better assess different scenarios and apply the most suitable solutions suited on their profile. For this scope five types of users have been stated according to the number of fees and the duration of the loan (between 5 and 10 years) (Table 4). Due the economic vulnerability of some low-income users to cover the cost of the renovation effort, for this reason a specific grant program has been identified for eligible homeowners (vulnerable persons) who are unable to pay their fees. It consists of the registration of a charge in the Property Register equivalent to the overall fee to be paid. This charge must be paid to the City Council in that case of the transfer of ownership (sale, inheritance or other). Thus, the financial capacity of the Public Sector determines the acceptable limit of the revolving funds.


**Table 4.** List of user types (UT) adopted in the financial appraisal (Figure 8). Source: authors.

Figure 8 presents the analysis considering two main KPIs related to the end-user perspective (average monthly payments) and the impact of revolving funds size on the Public Sector to cover vulnerable users (UT4). The graphics shows how the revolving funds will surpass 250 K€ when 3 different scenarios occur: (a) grant is lower than 15% of the total cost; (b) investment is high and the number of entities involved is considerable (more than 250 dwellings); (c) the proportion of UT4 increases from 10% to 20%. The amount of 250 K€ represents the suitable value adopted in the testbed and it is adjustable according to the effort that each Public Sector can assume to support limited number of parallel operations of the renovation process. The baseline scenario shown in Figure 8 represents the end-user and the Public Sector perspective when the conditions of Table 5 occur, referring to the Santa Coloma test bed.

The conclusions of the sensitivity analysis for the economic model demonstrate the model is robust enough to allow for different breakdowns between user types, variations in operational costs, variations in financial costs (i.e., interest rates), investment per dwelling and number of entities involved. In those cases, robustness refers to whether final monthly end-user payments remain lower than 100€ and savings offer incentive to undergo a largescale retrofitting operation. Moreover, large operations with a high number of entities (i.e., 500) or more vulnerable users that may require access to municipality grants do increase both financial need and municipal resources in terms of operational cost and size of revolving fund. In such cases, the size of the operation can be a limiting factor for the Public Sector. Furthermore, the debt financing in the form of loans represents a more sustainable means of up-scaling energy efficiency investments as loans can provide liquidity and direct access to capital, as well as support the cashflow during the process period. Loans can be more relevant for energy efficiency measures attached to high upfront costs, especially in deep renovation projects which comprise a package of multiple intervention measures. Despite this, the market interest rate (TAE) deviations are included in the model, and the result of the financial appraisal reveals how the fluctuance of interests has a strong impact on the financial cost during the loan period, and how the business model will be consequently affected by the financial cost increase.

**Figure 8.** Extract of the PEDRERA sensitive analysis on the financial appraisal KPIs: the end-user perspective (UT2 monthly rate fee) and the Public Sector perspective (City revolving funds size) to cover vulnerable type users.


**Table 5.** List of input from Santa Coloma test bed, adopted as a baseline in the financial appraisal (Figure 8). Source: authors.

101


**Table 5.** *Cont.*

#### **4. Discussions and Conclusions**

In this paper, we presented a description of the approach, purpose, methods and first results obtained by the ongoing research project PEDRERA focused on the design of a PEDs oriented renovation model. The input and KPIs considered for the sensitive analysis shown in the paper (Tables 1 and 3, and Figure 8) mean the main aspects covered by the PEDRERA model reach each "scope", according to a wider PED vision where both energy efficiency and production strategies are considered together with the operative, social, economic and financial aspects.

The PEDRERA model is currently being developed with Python, a well-known programming language, the same as the urbanZEB tool. Once the programming phase will be completed then the software will be fully implemented in the web platform, thus delivering the multiple stakeholders' engagement in large-scale renovation actions, and will support the prediction of the sustainability and positive outcomes of distinct renovation scenarios.

Other areas of development will go in the direction of the Renovation Wave, with different solutions and services that can help to face challenges in terms of supporting renovation program and seeking for innovative financial frameworks. With this perspective, the PEDRERA model implementation will provide a very comprehensive service able to support and promote renovation actions by multidimensional and dynamic scenarios analysis, as well as the prediction of the potential impacts and benefits from feasible measures both at building and district level. In our future work, we plan to further develop a web platform tool service from the PEDRERA model, in order to boost and design large-scale renovation actions as well as engage the different stakeholders in the renovation process. Further steps will be to adopt the platform prototype in ongoing EU-wide efficient retrofitting projects at district level. Specific case studies will be selected to ensure the platform performance is tested under different conditions including climate aspects, boundary conditions, uses, building typologies, intervention levels, conservation conditions and other aspects.

Coherently to the deployment of the Clean Energy Transition and the Driving Urban Transition (relevant to the Renovation Wave), the PEDRERA model already establishes appropriate mechanisms to analyze the interdisciplinary aspects addressed by the energy communities (ECs). This can only be done by leveraging a range of very advanced analysis including urban modelling and interoperability of data as well as information from a spread digitalization of cities. According to this pathway, semantic frameworks will help to spatialize, organize and normalize information, as also making possible the graphic representation, the insertion of algorithmic-logical models as well as the realization of complex questions. Nevertheless, there is still a lack of tools/services oriented to the relevant key stakeholders of the building renovation process. Thus, one of the main contributions of the project regards leading and delivering innovation of energy saving and renewables-related services but also improving the consciousness on renovation projects opportunities from the very beginning in order to force and support large-scale actions.

Given to today's pressing scenarios of energy transition and climate change, and to the economic worldwide circumstances worsened by the COVID-19 pandemic, the energy poverty is one of the main issues that will profoundly characterize urban environments in the next years. The lack of capital is clearly one of the most pressing issues: although many large-scale private funds are eager to find and finance bankable projects, nevertheless the fragmented nature of the renovation market and actors (at least until solutions to deliver high volumes of renovation are available) hinders their interest and ability to fund building renovation at a large scale. In order to take informed decisions in their respective realms, these stakeholders need to have access to information which suits their knowledge and capacities. Furthermore, the possibilities of using city model maps are inherently unlimited and can be addressed to current and future city issues that may arise.

**Author Contributions:** Conceptualization, P.C., J.P., J.S.; methodology, P.C., J.P., J.S., J.A.A. and A.B.F.; formal analysis, P.C., J.P., J.S., J.A.A. and A.B.F.; investigation, P.C., J.P., J.S., J.A.A. and A.B.F.; writing—original draft preparation, P.C., J.P., J.S., J.A.A. and A.B.F.; writing—review and editing, P.C.; visualization, P.C. and A.B.F.; supervision, P.C., J.P., J.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** The research leading to these results has received funding from the European Union's Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement No 712949 (TECNIOspring PLUS) and from the Agency for Business Competitiveness of the Government of Catalonia. TECNIOspring PLUS. Investigator: Paolo Civiero, Project: PEDRERA. Positive Energy Districts renovation model. IREC (Barcelona—ES).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

