Open-Source Datasets, Models, and Tools for Renewable Energy Community Design Mohammed Qasem, Sesil Koutra, Stephane Brisset, Arnaud Davigny, Benoit Durillon To cite this version: Mohammed Qasem, Sesil Koutra, Stephane Brisset, Arnaud Davigny, Benoit Durillon. Open-Source Datasets, Models, and Tools for Renewable Energy Community Design. Symposium de Génie Électrique SGE 2025, cnrs, ups, Jul 2025, Toulouse, France. ⟨hal-05506759⟩ HAL Id: hal-05506759 https://hal.science/hal-05506759v1 Submitted on 12 Feb 2026 HAL is a multi-disciplinary open access archive L’archive ouverte pluridisciplinaire HAL, est des- for the deposit and dissemination of scientific re- tinée au dépôt et à la diffusion de documents scien- search documents, whether they are published or not. tifiques de niveau recherche, publiés ou non, émanant The documents may come from teaching and research des établissements d’enseignement et de recherche institutions in France or abroad, or from public or pri- français ou étrangers, des laboratoires publics ou vate research centers. privés. Distributed under a Creative Commons CC BY-NC 4.0 - Attribution - Non-commercial use - International License SYMPOSIUM DE GENIE ELECTRIQUE (SGE 2025), 1 - 3 JUILLET 2025, TOULOUSE, FRANCE Open-Source Datasets, Models, and Tools for Renewable Energy Community Design Mohammed QASEM1,2, Arnaud DAVIGNY2, Benoit DURILLON2, Sesil KOUTRA1, Stephane BRISSET2 1 Faculty of Architecture and Urban Planning, University of Mons, St. Havré 88, 7000 Mons, Belgium. 2 Univ. Lille, Arts et Metiers Institute of Technology, Centrale Lille Institute, Junia, ULR 2697 L2EP, F-59000 Lille, France Abstract -Renewable Energy Communities (RECs) show simulation environments. As illustrated in Fig. 2, the primary promising potential for achieving energy transition goals by fostering components required for modeling REC-based distributed collective ownership of energy systems and empowering citizens to energy systems include datasets, models, and simulation take control of energy processes (generating, managing, consuming, frameworks. These tools enable researchers and professionals to storing, or even distributing). To facilitate the study and simulate and explore different configurations of RECs to test implementation of RECs, a comprehensive analysis of their dynamics their hypotheses. Particularly, these components are typically through modeling is needed. Achieving this requires access to data, classified according to their accessibility and licensing terms: models, and tools with open-source options that present a viable and commercial, partially open-source, and fully open-source accessible alternative. However, literature lacks comprehensive solutions [4], [5]. This paper focuses only on fully open-source reviews of state-of-the-art open-source datasets, models, and tools specifically tailored for RECs modeling. To fill this gap, this paper toolkits for their accessibility, flexibility, and innovation by conducts a systematic review to evaluate the available up-to-date open- enabling community-driven development and inclusive design source materials applicable to REC analysis, including electrical, through collaborative problem-solving [6], [7]. Their thermal, and multi-energy system models, and develop a transparency and adaptability align with research interests in comprehensive list of open-source toolkits that effectively support advancing equitable and decentralized energy systems, such as their simulation and optimization. RECs [8]. Keywords: datasets, energy modeling tools, open-source, renewable energy communities. 1. INTRODUCTION In the EU, the buildings sector accounted for approximately 40% of final energy consumption [1] underscoring its pivotal role in realizing the ambitious objectives of achieving carbon neutrality through the integration of clean energy by 2050 [2]. In this context, RECs contribute significantly to the green energy transition by putting citizens at the core of this shift and actively involving them in the decision-making process and daily energy Fig. 2. The main elements of a simplified REC modeling diagram. system operations (see Fig. 1) [3] . In this paper, the publicly accessible datasets, models, and tools related to energy modeling across various building types are compiled and analyzed. For each resource, a brief overview of its structure, scope, and key features is provided, followed by a comparative analysis in terms of scale, purpose, and relevance to specific energy research domains. This analysis supports researchers in selecting appropriate resources for eco-feedback, demand-side management, and sustainable energy applications [9]. Previous studies, such as [7], explored the available datasets, models, and tools tailored for RECs and provided a brief introduction for each. Additionally, the studies [10], [11] Fig. 1. A general architecture of a REC. provided a comparative review of urban building energy modeling (UBEM) frameworks, focusing on their capabilities to Fig. 1 represents a simple structure of a REC, where the simulate and optimize building energy performance. However, members of this kind of energy systems share the energy despite the rapid progress of digital technologies in recent years, produced locally. these studies were limited by the fact that their most recent updates were in 2022, and they only addressed one or two of the To fully understand the complexities of the RECs and essential modeling components (datasets, models, and tools). In evaluate effective implementation strategies, robust modeling contrast, no study has yet comprehensively addressed all three and analytical frameworks are needed; achieving this requires components together. the integration of relevant datasets, computational models, and Therefore, to fill this gap, this review paper aims to conduct Transparency and innovation: Open science enhances a literature review focusing on evaluating, categorizing, and transparency by enabling repeatable analyses, which are critical listing the pros and cons of the available up-to-date open-source for scientific credibility, especially in energy systems modeling materials that support and enhance effective modeling and that often involves complex datasets and methodologies [19]. design by optimization of RECs. This review will give insights Furthermore, the implementation of open-source tools and data- for researchers on energy systems modeling, providing a sharing methodologies enables rigorous peer review and the complete package of open-source datasets, models, and independent validation of results [20]. As emphasized by [21], frameworks that facilitate their research, with an open and open science plays a critical role in enhancing the quality and reproducible science approach. transparency of scientific research. This review paper is structured as follows: the second section Reproducibility and replicability: Open-source practices explores the role of open-source science in advancing research strengthen reproducibility by promoting transparency through with a specific focus on its application to REC modeling. The publicly available code, data, and modeling environments [22]. third section provides a comprehensive review of the state-of- For example, open-source energy modeling tools, such as the-art datasets available for various types of energy loads OSeMOSYS provide researchers with accessible, standardized (including electric vehicles (EVs) and heat pumps), energy frameworks to inspect, validate, and adapt methodologies. This generation, and energy storage systems. The fourth section openness not only builds trust in results but also accelerates explores energy models suitable for simulation and innovation by reducing redundant efforts and encouraging optimization, as well as models designed for profile generation. iterative improvements across the scientific community (see While the fifth focuses on open-source tools relevant to these section 4.2). analyses, such as UBEM frameworks based on EnergyPlus [12] (e.g. CityBES, UrbanOpt, …etc.), PyPSA [13] and oemof [14]. Financial accessibility: Reduction of costs compared to Finally, the discussion and concluding sections of this paper proprietary software, enabling broader participation, especially offer a summary and synthesis of the key findings and provide in resource-limited RECs. Reducing barriers to data sharing insights and suggestions for future studies in this area. enables researchers to build upon each other’s work, accelerates discovery, and facilitates the translation of findings into • Load profiles effective policy measures [23]. • Renewable energy profiles Datasets* 2.2. Challenges • Models for generating profiles (load, renewable energy, and battery profiles) Despite the significant benefits of open science for Models • Models for energy systems modeling and design by optimization accelerating research collaboration, improving reproducibility, and enhancing transparency in energy systems modeling, these • Frameworks for energy systems modeling and design by optimization open-access materials face notable challenges, including: • Urban Building Energy Modeling (UBEM) frameworks Frameworks Technical complexity: A major barrier to the adoption of open-access materials is the technical complexity involved in * As mentioned in the Models section, several models are also used to generate datasets. their implementation and customization processes. Moreover, Fig. 3. Categorization of open-source resources for energy modeling and key technical challenges include inadequate institutional simulation presented in this paper. infrastructure, limited support, diverse data, and scalability issues [24], [25]. Fig. 3 presents a categorization of open-source resources employed in energy modeling research, as outlined in this paper. Sustainability of initiatives: Open-source projects often face challenges in maintenance and updates due to limited 2. ROLE OF OPEN-SOURCE SCIENCE IN ENERGY MODELING resources and financial support. Additionally, concerns arise Open-source solutions play a pivotal role in enabling over recognition and publication opportunities, particularly researchers and engineers to design, implement, and optimize when independently generated datasets fail to yield appropriate sustainable energy systems by ensuring transparency, credit or return on investment for researchers [24]. replicability, encouraging collaboration, and reducing financial Interoperability issues: Challenges arise in integrating barriers to entry, thereby supporting the advancement of RECs open-source solutions with existing proprietary systems and on multiple fronts. [15], [16]. This section critically examines heterogeneous data formats. [26]. the role of open-source collaborative initiatives in advancing REC research, highlighting both the opportunities they offer, Data security and privacy: Protecting sensitive energy data and the challenges associated with their adoption and in open-source environments is crucial for ensuring REC implementation. compliance and maintaining stakeholder trust. While this presents a challenge, it also offers an opportunity, as open access 2.1. Opportunities to code facilitates collaborative debugging and enhances overall Open science is pivotal for accelerating advances in energy- system reliability [26]. Privacy and confidentiality concerns systems modeling, especially in cases where the availability of underscore the importance of ownership structures and data data or modeling tools is economically challenging. Open- governance. The absence of standardization and data source datasets and tools may bring several opportunities in governance strategies poses a challenge [24]. energy modeling that can be summarized as follows: 3. DATA FOR RECS Interoperability and collaboration: Interoperability refers to the ability of diverse systems or technologies to collaborate, Over the past decade, interest in open science has grown communicate, and function cohesively [17]. Open-source tools markedly, as reflected by both the proliferation of open‑source play a critical role in advancing interoperability by prioritizing energy datasets and a corresponding increase in related collaboration and community-driven innovation [18]. academic publications, as is seen in Fig. 4. 12 The criteria for selecting the models in this section are 10 twofold. First, models are considered based on their ability to generate relevant datasets, such as load profiles and renewable Number of publications 8 energy production estimates, essential for simulating energy 6 systems. Second, models are selected for their capacity to simulate energy generation and consumption dynamics, as well 4 as their compatibility and interoperability with other modeling 2 tools and platforms, which is crucial for integrated and scalable energy system analysis in decentralized and community-based 0 settings. Consequently, the selected models are classified into 2012 2014 2016 2018 2020 2022 2024 Year two primary groups: one group of models is employed for generating datasets, while the other includes models utilized for Fig. 4. The number of publications that used public open data for modeling modeling and simulating energy systems. energy per year in the Scopus database. 4.1. Models for data generation Fig. 4 shows that the number of publications that use open- source materials for energy systems modeling has notably Besides benchmark datasets that are published publicly, increased since 2017. The results are based on a research query numerous models and software programs can be used to generate using these keywords ("energy modeling" AND "open data") in load profiles, renewable energy profiles, or battery storage Scopus database. profiles. The utilization of publicly accessible datasets is a critical 4.1.1. Load profiles data generating models component of the testing, validation, and benchmarking of The publicly accessible models that can be used for simulating models [27], including RECs models. The input generating load profiles are: demandlib, Load Profile Generator datasets, considered the fuel for energy systems modeling, can (LPG), Artificial Load Profile Generator (ALPG), Office Load be classified according to their acquisition method into two MATLAB App, and CREST Demand Model. principal categories: empirical (real‑world) datasets and Demandlib synthetic (generated) datasets. In this section, the empirical publicly available datasets (load and renewable energy profiles) Demandlib is a Python library used for generating time- from diverse regions around the world are reviewed. These resolved electrical and thermal demand profiles at various profiles serve as benchmarks for advancing energy‑modeling frequencies (e.g., 15-minute, hourly, or daily) from annual methodologies and simulation tools, thereby supporting the energy values. This model is typically used in energy-system transition to greener and more sustainable energy systems [7]. modeling and scenario analysis to provide realistic demand inputs for tools like oemof and PyPSA. It covers multiple end- Load profiles: publicly available load profiles datasets can be use sectors, notably residential (single- and multi-family houses) categorized by various criteria, including building type (e.g., and various commercial and industrial categories. A key residential: GREEND, UK-DALE-2017, SustData; limitation of this tool is its reliance on fixed profile templates, commercial/public: Building Data Genome, BLOND, BERDS), which are often based on standardized German energy norms geographical location (e.g., Asia, Europe, North America), and and historical weather data. This approach may struggle to data resolution (e.g., per second, 10 seconds, 1 minute, 15 account for region-specific or culturally distinct energy minutes, hourly, or daily). These classifications encompass consumption patterns outside of the original design context [30]. diverse usage types such as household or industrial consumption, as well as base and flexible loads. LPG Renewable energy profiles: these datasets can be categorized LPG is a tool designed for simulating residential energy based on the type of energy source, such as solar (solar consumption through detailed behavioural modeling of irradiation) or wind (wind speed). Additionally, many of these household occupants. It generates load curves at various datasets incorporate relevant climatic and meteorological resolutions, ranging from 1 minute and 15 minutes to 1 hour, variables, including temperature and humidity, which are based on predefined German household profiles, and outputs the essential for accurate modeling and forecasting. data in CSV format, ensuring compatibility with other simulation tools. However, its application is limited to The publicly available datasets are shown in Appendix 1. residential settings; it does not support commercial or office 4. MODELS FOR RECS building simulations [31]. Models transform data into actionable insights [28] and serve ALPG as critical tools for understanding, predicting, and optimizing the ALPG is a Python tool designed to generate time-resolved performance of distributed energy systems, including RECs, electrical and thermal load profiles with embedded flexibility allowing for better decision-making and resource allocation. constraints for smart grid simulation and control algorithm These models, based on modeling techniques, can be classified evaluation. Unlike traditional static profiles, ALPG simulates into data-driven models and physical models. Also, they can be realistic, minute-resolution consumption behaviours of classified by the degree of physical interpretability into black- controllable domestic devices, enabling benchmarking and box, white-box, and grey-box models [29]. There are no specific comparative analysis of energy management solutions using models focused on modeling RECs. Still, since RECs are standardized input data. The tool integrates thermal modeling considered one of the innovative energy systems, this section techniques and reflects real-world consumption dynamics will focus on the open-source models that can be used for validated through field test measurements at both household and modeling and optimizing energy systems. neighbourhood levels. Importantly, ALPG's output is tailored for use as input data in downstream optimization or simulation tools, offering a robust foundation for evaluating demand-side energy cost savings using local weather data and NREL’s PV management strategies [32]. performance models. While less customizable than tools like pvlib, PVWatts is ideal for early-stage project assessment due to Office Load MATLAB App its ease of use, default assumptions, and reliable integration with The Office Load MATLAB App is a simulation tool NREL’s extensive solar resource database [37]. designed to model electricity consumption in office buildings by GSEE integrating behavioural and physical approaches. It generates load profiles by considering occupancy patterns, appliance GSEE is a Python library designed for rapid and user- usage, and HVAC system operations, validated against real- friendly simulation of PV power output. It offers streamlined world data from Northern European office buildings. While the modeling of PV systems, supporting both fixed and tracking application offers flexibility in choosing time resolutions, the configurations, and is particularly suited for large-scale or long- default or commonly used time step is 10 minutes. The app is term energy assessments. GSEE integrates modules for useful for analysing demand response potential, optimizing estimating diffuse irradiance (via the BRL model), calculating energy use, and assessing the impact of different building irradiance on inclined planes, and interfacing with climate parameters on electricity consumption. However, this tool does datasets in formats like NetCDF. The BRL model is a reliable, not apply to large-scale simulations, which are computationally logistic regression-based approach that efficiently decomposes intensive [33]. global solar radiation into its diffuse and direct components using minimal input data, and its simple yet adaptable structure CREST Demand Model makes it particularly valuable in data-sparse environments such The CREST demand model is designed to simulate thermal as semiarid regions [38]. GSEE is utilized by platforms such as and electrical demand, with a default time step of 1 minute, Renewables.ninja for simulating solar energy generation across within residential settings, primarily for low-voltage network various geographical and temporal scales [39]. analysis. It employs a bottom-up, activity-based approach Renewables.ninja integrated with stochastic programming techniques to capture occupant behaviour and dwelling diversity. By incorporating Renewables.ninja is a web-based platform that simulates reduced-order thermal-electrical networks, the model effectively hourly power output from wind and solar power systems at any represents thermal dynamics while generating high-resolution, global location. Developed at TU Delft and Imperial College validated outputs. Its design prioritizes the accurate timing of London, the tool aims to make scientific-grade energy and demand to reflect the inherently stochastic nature of household weather data accessible to a broad audience. It features validated energy use [34]. models for both solar and wind power and is widely used in energy research and planning [40]. 4.1.2. Renewable energy data generating models This group of models aims to generate renewable energy feedinlib (oemof) production profiles: Photovoltaic Geographical Information feedinlib is a model designed within the oemof tool that System (PVGIS), pvlib, PVWatts, Renewables.ninja, Global models feed-in time series from renewable energy sources, Solar Energy Estimator (GSEE), OEMOF’s feedinlib, and atlite. specifically wind and solar PV systems to simulate location- PVGIS specific power generation from renewables. feedinlib is designed to integrate seamlessly into energy system models built PVGIS is a web-based tool that is developed for providing with oemof, providing reproducible and transparent inputs for solar irradiation and PV performance data worldwide (except simulation and optimization. It supports different PV and wind polar regions). It includes modules for estimating grid- turbine models and allows for batch processing of multiple connected and off-grid (battery‑backed) PV yields, generating locations or assets [41]. solar radiation time series (hourly/daily/monthly), and compiling typical meteorological year climate datasets [35]. atlite pvlib atlite is a Python library developed for the conversion of meteorological data into energy-relevant time series, including pvlib is a Python and MATLAB toolbox designed for outputs for PV systems, wind power generation, solar thermal simulating the performance of PV energy systems. It provides a energy, heating demand, and hydropower potential. Optimized comprehensive collection of over 100 empirical and physics- for large-scale energy system simulations, it offers efficient based models from peer-reviewed literature, covering solar memory and CPU usage. The library supports integration with position algorithms, irradiance models, thermal models, and PV high-resolution climate datasets such as ERA5 and COSMO- electrical models. With contributions from a growing global user REA6 and is compatible with energy modeling frameworks like base, pvlib supports high-level workflows for complete PyPSA. Built upon the x-array data structure, atlite provides "weather-to-power" simulations and includes tools for fetching extensive spatial and temporal flexibility, making it a widely and standardizing weather data and has become a reliable adopted tool in power system modeling and energy research standard for PV performance analysis [36]. [42]. PVWatts 4.1.3. Battery profile generating models PVWatts Calculator is a web-based tool designed to estimate A battery profile is defined as a set of performance the electricity production and energy value of grid-connected PV characteristics (e.g., state-of-charge, cycle depth, number of sign systems. It provides a simple interface for users, ranging from changes, length of resting periods, energy between sign changes) homeowners to project developers, to input key system of battery storage under various operating conditions [43]. These parameters such as location, system size, module tilt, orientation, profiles are essential for comprehending battery behavior, and system losses. Based on this input, PVWatts calculates optimizing performance, and ensuring safety in various monthly and annual AC energy output and the corresponding applications, such as EVs and grid storage systems [43]. The models that are tailored for battery profile generation, and Balmorel mentioned in [7], are: QUantum Electronic Structure (QuESt), Open-Source Energy Storage Model (OSESMO), and Balmorel is an energy system optimization model written in EnergyBoost. GAMS modeling language and built in a generic, extensible modular structure. This model is designed to simulate and QuESt 2.0 optimize the generation, transmission, storage, and consumption of electricity and heat. This is achieved by assuming the QuESt 2.0 is a Python-based platform developed for presence of perfectly competitive markets. The objective is to advanced energy storage analytics, aiming to simplify these maximize social welfare, while considering a wide range of analyses and democratize access to these tools. Evolving from technical, economic, and regulatory constraints [49]. its original version, QuESt 2.0 offers a centralized suite of tools (e.g. App Hub, Workspace, and QuESt GPT), that enable NEMO streamlined workflows, AI-driven insights, and scalable data processing. Its key innovations include a unified interface, NEMO is a high-performance model developed in Julia and integration of generative AI for complex analytics, and modular used for modeling and optimizing electrical energy systems. extensibility. Unlike traditional tools focused on isolated tasks, This tool can be used in stand-alone mode or with the Low QuESt 2.0 supports holistic and collaborative energy storage Emissions Analysis Platform (LEAP) as a front-end. NEMO evaluation, aligned with U.S. Department of Energy objectives facilitates the incorporation of energy storage capacity into the by democratizing access to advanced energy storage analytics long-term simulation of power system expansion. This is tools [44]. particularly relevant when assessing the significance of balancing intermittent renewable energy, especially in scenarios OSESMO with a high penetration of renewable energy sources [50] [51]. OSESMO is an energy storage dispatch optimization tool Other models have been built for a specific function to solve originally developed in MATLAB and now being translated into a challenge in the energy systems. Such as the URBS model, Python for broader accessibility and future web deployment. which is designed for linear programming tasks related to energy OSESMO evaluates the dispatch (charge/discharge) of energy dispatch and expansion planning that enables users to flexibly storage to evaluate the GHG emissions of an energy storage define multiple regions, associated energy conversion and project. It supports analysis of behind-the-meter and solar-plus- storage processes, and interregional transmission lines. Its storage systems with a focus on time-of-use arbitrage, demand modular structure allows for the integration of emerging charge management, solar self-consumption, and GHG technologies, such as Power-to-X, making it suitable for diverse emissions reduction [45]. applications [52]. The IDEAS model, a Modelica-based model, allows simultaneous transient simulation of thermal and EnergyBoost electrical systems at both building and feeder levels. The EnergyBoost is a Python-based control software for home AMIRIS model is designed for modeling electricity markets and battery systems, designed to run on a Raspberry Pi. It integrates focuses on the business-oriented decisions of actors, who are supervised learning models to forecast the next-day solar represented as agents, in the energy system [53]. Finally, the generation and household demand, alongside physical models Dispa-SET model is used for solving energy balancing and simulating inverter output and battery state-of-charge (SoC). flexibility problems and modeling the power system on different EnergyBoost formulates an optimal control problem to schedule scales (e.g., district, region, country) [54]. battery charging/discharging over a finite horizon, using both model-based and model-free control methods, and aims at 5. FRAMEWORKS FOR MODELING RECS optimizing residential energy use [46]. Frameworks are essential tools that integrate data and models within software platforms to support the simulation, 4.2. Models for energy systems modeling analysis, and optimization of building energy performance. In In this section, the models used for modeling energy systems this paper, based on their purpose and primary modeling (electrical, thermal, and multi-energy systems) as well as for approaches, these frameworks are classified into two main optimization are briefly summarized and explained. The models groups: energy system optimization frameworks and UBEM explored here are: Open-Source Energy Modeling System frameworks. (OSeMOSYS), Balmorel, and Next Energy Modeling System for Optimization (NEMO). 5.1. Energy system modeling frameworks These frameworks are primarily designed to model and OSeMOSYS optimize energy flows and interactions within multi-sectoral OSeMOSYS is a completely developed optimization model energy systems and are typically used to simulate energy that is utilized for long-run integrated assessment and energy systems (electrical, thermal and multi-energy systems) at planning. This tool is designed to develop energy systems various spatial and temporal scales. (electrical, thermal and transport) models for different scales oemof (districts, cities, states, countries, continents, or even the globe). The low learning curve and the availability of OSeMOSYS for oemof is a modeling framework used to construct free extend the availability of energy modeling to a wide group comprehensive energy system models and incorporate various of researchers and users [47]. Based on this model, the modeling approaches. It is particularly well suited to flexibly GENeSYS-MOD model, an optimization model that uses the model complex cross-sectoral systems, linking the heat, power, CPLEX solver, is developed [19]. Just like OSeMOSYS, and mobility sectors. The framework includes various packages GENeSYS-MOD consists of multiple blocks of functionality, such as oemof.solph for linear optimization of energy systems, which work as separate entities that can be changed or extended oemof.outputlib and oemof-visio for results processing, feedinlib [48]. and windpowerlib for renewable generation profiles, and demandlib for generating load profiles [55]. PyPSA residential consumption patterns, respectively, yet their domain- specific design underscores a persistent need for more PyPSA is a framework designed to simulate and optimize generalized frameworks that can seamlessly accommodate energy power systems over multiple periods. It encompasses diverse building typologies and end-use sectors. conventional generators with unit commitment, variable renewable generation, storage units, integration with other Finaly, in regard to the frameworks, physics-based models energy sectors, and mixed alternating and direct current deliver high-fidelity simulations of building energy dynamics at networks. PyPSA is designed to facilitate seamless extensibility the expense of scalability and computational speed. EnergyPlus- and ensure optimal scaling, both within the context of extensive based tools (e.g., UrbanOpt, AutoBEM, ComStock, CityBES, networks and in the management of extensive time series [13]. and UMI) leverage detailed thermodynamic calculations to capture multi-zone interactions and temporal variations, making 5.2. UBEM frameworks them ideal for district-level energy planning, but challenging for UBEM frameworks are primarily developed to simulate large-scale deployment. Conversely, reduced-order RC model– energy performance in buildings at the urban scale, with a focus based platforms (e.g., CitySim, SimStadt, OpenIDEAS, CEA, on building stock analysis and district-level energy planning. and TEASER) employ simplified data-driven methods to These frameworks, summarized in [4], [5], can be classified achieve rapid simulations while preserving key thermal based on the modeling methodology to which they are built upon behaviours, albeit with reduced spatial and temporal resolution. reduced-order resistance-capacitance (RC) and heat-balanced A balanced integration of these methodologies could enhance physics model-based frameworks. the development of adaptable, efficient frameworks capable of Heat-balanced physics model-based frameworks: These addressing diverse urban contexts and scales. tools are based on models that rely on fundamental 7. CONCLUSION thermodynamic principles to simulate the detailed energy behavior of buildings. These frameworks simulate multi- To conclude, this paper underscores the critical role of open- thermal-zone dynamics and allow for detailed temporal and source datasets and tools in supporting the modeling and spatial resolution of energy flows, and are typically used for optimizing the energy systems, including RECs. Through a high-resolution, dynamic simulations. Frameworks in this literature review, an updated and comprehensive inventory of category include: EnergyPlus-based tools such as UrbanOpt, open-source datasets, models, and frameworks has been AutoBEM, ComStock, CityBES, and UMI. developed. These resources address key aspects of energy systems modeling, including energy generation, consumption, Reduced-order RC model-based frameworks: These storage, and distribution. The primary findings of this research platforms simplify physics to improve computational efficiency include: while preserving key thermal behaviors. These models often use resistance-capacitance (RC) networks or data-driven approaches • A thorough evaluation of existing open-source to estimate energy demand. Tools in this category include: materials, highlighting their strengths and limitations. CitySim, SimStadt, OpenIDEAS, CEA, and TEASER. • A structured categorization of these tools based on their relevance and application to REC modeling. A detailed comparison of the characteristics of these frameworks is provided in Appendix 3, highlighting their core This review provides a robust foundation for advancing the modeling approaches, input, output and temporal resolution. adoption and optimization of RECs, contributing significantly to This comparative overview facilitates understanding of the the achievement of European climate and energy goals. trade-offs between model fidelity and computational efficiency. Finally, this work aims to lower barriers to entry for REC 6. DISCUSSION development by improving data availability and modeling tools. The future work can address two interconnected challenges in Despite the critical role of publicly accessible data in RECs modeling: First, we propose developing open datasets, for informing research and policy, the vast majority of these datasets different geographical contexts, tailored to RECs, integrating are produced by institutions in Europe, US, and UK, resulting in synthetic data generation to overcome gaps in real-world data. a pronounced regional bias that complicates their applicability Second, we aim to adapt or build open-source energy models in diverse geographic contexts. based on the already developed models (e.g., PyPSA, oemof) to Regarding the models used for generating load profile explicitly address REC-specific dynamics like peer-to-peer datasets, since they are calibrated on historical data from a single trading, grid constraints, and policy frameworks. By bridging geographical region, their applicability, and thus their accuracy, these gaps, this work seeks to empower RECs with actionable is inherently constrained to contexts and climate regimes similar tools for optimizing decentralized systems and informing to those for which they were originally developed. Models, such policies that accelerate equitable energy transitions. as CREST demand model, adopt a bottom-up, stochastic, 8. ACKNOWLEDGEMENTS activity-based methodology that enhances the representativeness of thermal–electrical demand timing; The authors express their gratitude to the Hauts-de-France however, this approach introduces considerable computational Region and the University of Mons for funding the project. complexity and data input requirements. Advanced load profile generators, such as ALPG, further extend these capabilities by 9. REFERENCES embedding device-level flexibility constraints and leveraging [1] D. D’Agostino, P. M. Congedo, P. M. Albanese, A. Rubino, and C. field-validated behavioral simulations to produce minute- Baglivo, “Impact of climate change on the energy performance of building envelopes and implications on energy regulations across Europe,” Energy, resolution profiles. However, the accuracy of these profiles vol. 288, p. 129886, Feb. 2024, doi: 10.1016/J.ENERGY.2023.129886. remains contingent on the availability and quality of underlying [2] Rep. Plan, “Communication from the commission to the european measurement datasets. Similarly, tools such as the Office Load parliament, the european council, the council, the european economic and MATLAB App and the LoadProfileGenerator (LPG) employ social committee and the committee of the regions,” European behavioral and physical modeling to simulate office and Commission: Brussels, Belgium, 2018. [3] A. Caramizaru, A. Uihlein, and others, Energy communities: an overview and a checklist for data managers,” J Agric Food Res, vol. 16, p. 101157, of energy and social innovation, vol. 30083. Publications Office of the Jun. 2024, doi: 10.1016/J.JAFR.2024.101157. European Union Luxembourg, 2020. [25] J. Wang et al., “Data sharing in energy systems,” Advances in Applied [4] D. Kong, A. Cheshmehzangi, Z. Zhang, S. P. Ardakani, and T. Gu, “Urban Energy, vol. 10, p. 100132, Jun. 2023, doi: building energy modeling (UBEM): a systematic review of challenges and 10.1016/J.ADAPEN.2023.100132. opportunities,” Energy Effic, vol. 16, no. 6, Aug. 2023, doi: [26] S. Butler et al., “Considerations and challenges for the adoption of open 10.1007/s12053-023-10147-z. source components in software-intensive businesses,” Journal of Systems [5] M. Ferrando, F. Causone, T. Hong, and Y. Chen, “Urban building energy and Software, vol. 186, p. 111152, Apr. 2022, doi: modeling (UBEM) tools: A state-of-the-art review of bottom-up physics- 10.1016/J.JSS.2021.111152. based approaches,” Sustain Cities Soc, vol. 62, p. 102408, Nov. 2020, doi: [27] K. Blagec, J. Kraiger, W. Frühwirt, and M. Samwald, “Benchmark 10.1016/J.SCS.2020.102408. datasets driving artificial intelligence development fail to capture the [6] J. Manchanda, L. Boettcher, M. Westphalen, and J. Jasser, “The Open needs of medical professionals,” J Biomed Inform, vol. 137, p. 104274, Source Advantage in Large Language Models (LLMs),” Dec. 2024, Jan. 2023, doi: 10.1016/J.JBI.2022.104274. Accessed: May 12, 2025. [Online]. Available: [28] I. H. Sarker, “Data Science and Analytics: An Overview from Data- https://arxiv.org/pdf/2412.12004v1 Driven Smart Computing, Decision-Making and Applications [7] H. Kazmi, Í. Munné-Collado, F. Mehmood, T. A. Syed, and J. Driesen, Perspective,” SN Comput Sci, vol. 2, no. 5, p. 377, Sep. 2021, doi: “Towards data-driven energy communities: A review of open-source 10.1007/S42979-021-00765-8. datasets, models and tools,” Sep. 01, 2021, Elsevier Ltd. doi: [29] A. Petrucci, F. K. Ayevide, A. Buonomano, and A. Athienitis, 10.1016/j.rser.2021.111290. “Development of energy aggregators for virtual communities: The energy [8] N. Gorman, I. MacGill, and A. Bruce, “How to support the adoption of efficiency-flexibility nexus for demand response,” Renew Energy, vol. open-source energy system modelling software? Insights from interviews 215, p. 118975, Oct. 2023, doi: 10.1016/J.RENENE.2023.118975. with users and developers,” Energy Res Soc Sci, vol. 111, p. 103479, May [30] U. Krien et al., “demandlib (oemof) - creating heat and power demand 2024, doi: 10.1016/J.ERSS.2024.103479. profiles from annual values - v0.1.6 ,” Jan. 2019, Zenodo. doi: [9] L. Pereira, F. Quintal, R. Gonçalves, and N. J. Nunes, “SustData: A Public 10.5281/zenodo.2553505. Dataset for ICT4S Electric Energy Research,” ICT for Sustainability [31] N. Pflugradt, P. Stenzel, L. Kotzur, and D. Stolten, “Loadprofilegenerator: 2014, ICT4S 2014, pp. 359–368, Aug. 2014, doi: 10.2991/ICT4S- An agent-based behavior simulation for generating residential load 14.2014.44. profiles,” J Open Source Softw, vol. 7, no. 71, p. 3574, 2022. [10] D. Kong, A. Cheshmehzangi, Z. Zhang, S. P. Ardakani, and T. Gu, “Urban [32] G. Hoogsteen, A. Molderink, J. L. Hurink, and G. J. M. Smit, “Generation building energy modeling (UBEM): a systematic review of challenges and of flexible domestic load profiles to evaluate demand side management opportunities,” Energy Effic, vol. 16, no. 6, p. 69, 2023, doi: approaches,” in 2016 IEEE International Energy Conference 10.1007/s12053-023-10147-z. (ENERGYCON), United States: IEEE, Apr. 2016, p. 1279. doi: [11] M. Ferrando, F. Causone, T. Hong, and Y. Chen, “Urban building energy 10.1109/ENERGYCON.2016.7513873. modeling (UBEM) tools: A state-of-the-art review of bottom-up physics- [33] D. A. Brodén, K. Paridari, and L. Nordström, “Matlab applications to based approaches,” Sustain Cities Soc, vol. 62, p. 102408, Nov. 2020, doi: generate synthetic electricity load profiles of office buildings and 10.1016/J.SCS.2020.102408. detached houses,” in 2017 IEEE Innovative Smart Grid Technologies- [12] “EnergyPlus.” Accessed: Jan. 10, 2025. [Online]. Available: Asia (ISGT-Asia), 2017, pp. 1–6. https://energyplus.net/ [34] E. McKenna and M. Thomson, “High-resolution stochastic integrated [13] T. Brown, J. Hörsch, and D. Schlachtberger, “PyPSA: Python for Power thermal–electrical domestic demand model,” Appl Energy, vol. 165, pp. System Analysis”. 445–461, Mar. 2016, doi: 10.1016/J.APENERGY.2015.12.089. [14] “A modular open source framework to model energy supply systems.” [35] T. Huld, R. Müller, and A. Gambardella, “A new solar radiation database Accessed: Jan. 10, 2025. [Online]. Available: https://oemof.org/ for estimating PV performance in Europe and Africa,” Solar Energy, vol. [15] L. Bosek, J. Pickard, S. Chuang, D. Batts, and J. Musgrave, “Transparent 86, no. 6, pp. 1803–1815, Jun. 2012, doi: Collaborative Sharing: Adopting Open-Source Methodology for 10.1016/J.SOLENER.2012.03.006. Improving Business Processes and Public Trust,” Journal of Technology [36] K. S. Anderson, C. W. Hansen, W. F. Holmgren, A. R. Jensen, M. A. Studies, doi: 10.21061/jts.418. Mikofski, and A. Driesse, “pvlib python: 2023 project update,” J Open [16] M. G. Bertram, J. Sundin, D. G. Roche, A. Sánchez-Tójar, E. S. J. Thoré, Source Softw, vol. 8, no. 92, p. 5994, Dec. 2023, doi: and T. Brodin, “Open science,” Current Biology, vol. 33, no. 15, pp. 10.21105/joss.05994. R792–R797, Aug. 2023, doi: 10.1016/J.CUB.2023.05.036. [37] A. P. Dobos, “PVWatts Version 5 Manual,” May 2014. doi: [17] C. Dictionary, “Cambridge advanced learner’s dictionary & thesaurus,” 10.2172/1158421. Retrieved from Cambridge website: https://dictionary. cambridge. [38] M. Guermoui, J. Boland, and A. Rabehi, “On the use of BRL model for org/dictionary/english, 2017. daily and hourly solar radiation components assessment in a semiarid [18] J. Klimt, N. Eiling, F. Wege, J. Baude, and A. Monti, “The Role of Open- climate,” Eur Phys J Plus, vol. 135, no. 2, Feb. 2020, doi: Source Software in the Energy Sector,” Energies 2023, Vol. 16, Page 10.1140/epjp/s13360-019-00085-0. 5855, vol. 16, no. 16, p. 5855, Aug. 2023, doi: 10.3390/EN16165855. [39] S. Pfenninger and I. Staffell, “Long-term patterns of European PV output [19] T. Prike, “Open Science, Replicability, and Transparency in Modelling,” using 30 years of validated hourly reanalysis and satellite data,” Energy, in Towards Bayesian Model-Based Demography: Agency, Complexity vol. 114, pp. 1251–1265, Nov. 2016, doi: and Uncertainty in Migration Studies, Cham: Springer International 10.1016/J.ENERGY.2016.08.060. Publishing, 2022, pp. 175–183. doi: 10.1007/978-3-030-83039-7_10. [40] S. Pfenninger and I. Staffell, “Long-term patterns of European PV output [20] J. F. DeCarolis, K. Hunter, and S. Sreepathi, “The case for repeatable using 30 years of validated hourly reanalysis and satellite data,” Energy, analysis with energy economy optimization models,” Energy Econ, vol. vol. 114, pp. 1251–1265, 2016. 34, no. 6, pp. 1845–1853, Nov. 2012, doi: [41] U. Krien, G. Plessmann, S. Günther, B. Schachler, S. Bosch, and C. 10.1016/J.ENECO.2012.07.004. Kaldemeyer, “feedinlib (oemof) - creating feed-in time series - v0.0.12 ,” [21] S. Pfenninger, J. DeCarolis, L. Hirth, S. Quoilin, and I. Staffell, “The Jan. 2019, Zenodo. doi: 10.5281/zenodo.2554102. importance of open data and software: Is energy research lagging [42] F. Hofmann, J. Hampp, F. Neumann, T. Brown, and J. Hörsch, “atlite: A behind?,” Energy Policy, vol. 101, pp. 211–215, Feb. 2017, doi: Lightweight Python Package for Calculating Renewable Power Potentials 10.1016/J.ENPOL.2016.11.046. and Time Series,” J Open Source Softw, vol. 6, no. 62, p. 3294, Jun. 2021, [22] L. A. Barba, “Defining the role of open source software in research doi: 10.21105/JOSS.03294. reproducibility,” 2022, Accessed: Apr. 15, 2025. [Online]. Available: [43] D. Kucevic et al., “Standard battery energy storage system profiles: https://opensource.org Analysis of various applications for stationary energy storage systems [23] S. Wang et al., “Open science 2.0: revolutionizing spatiotemporal data using a holistic simulation framework,” J Energy Storage, vol. 28, p. sharing and collaboration”, doi: 10.1007/s43762-025-00165-1. 101077, Apr. 2020, doi: 10.1016/J.EST.2019.101077. [24] A. I. Ugochukwu and P. W. B. Phillips, “Open data ownership and [44] R. J. Concepcion, “QuESt: an Energy Storage Application Suite.,” May sharing: Challenges and opportunities for application of FAIR principles 2019. [Online]. Available: https://www.osti.gov/biblio/1639201 [45] R. C. Mann and Others, “OSESMO: Open Source Energy Storage Data 2020 7:1, vol. 7, no. 1, pp. 1–13, Oct. 2020, doi: 10.1038/s41597- Model,” 2018. [Online]. Available: 020-00712-x. https://github.com/RyanCMann/OSESMO [65] M. Pipattanasomporn et al., “CU-BEMS, smart building electricity [46] “Welcome to EnergyBoost’s documentation! — EnergyBoost 1.0.0 consumption and indoor environmental sensor datasets,” Scientific Data documentation.” Accessed: May 08, 2025. [Online]. Available: 2020 7:1, vol. 7, no. 1, pp. 1–14, Jul. 2020, doi: 10.1038/s41597-020- https://sustainable-computing.github.io/EnergyBoost/index.html 00582-3. [47] M. Howells et al., “OSeMOSYS: The Open Source Energy Modeling [66] S. K. Lien, H. T. Walnum, and Å. L. Sørensen, “COFACTOR Drammen System: An introduction to its ethos, structure and development,” Energy dataset - 4 years of hourly energy use data from 45 public buildings in Policy, vol. 39, no. 10, pp. 5850–5870, Oct. 2011, doi: Drammen, Norway,” Scientific Data 2025 12:1, vol. 12, no. 1, pp. 1–10, 10.1016/J.ENPOL.2011.06.033. Mar. 2025, doi: 10.1038/s41597-025-04708-3. [48] K. Löffler, K. Hainsch, T. Burandt, P.-Y. Oei, C. Kemfert, and C. von [67] R. Trivedi, M. Bahloul, A. Saif, S. Patra, and S. Khadem, “Comprehensive Hirschhausen, “Designing a Model for the Global Energy System- Dataset on Electrical Load Profiles for Energy Community in Ireland,” GENeSYS-MOD: An Application of the Open-Source Energy Modeling Scientific Data 2024 11:1, vol. 11, no. 1, pp. 1–12, Jun. 2024, doi: System (OSeMOSYS)”, doi: 10.3390/en10101468. 10.1038/s41597-024-03454-2. [49] F. Wiese et al., “Balmorel open source energy system model,” Energy [68] O. Parson et al., “Dataport and NILMTK: A building data set designed Strategy Reviews, vol. 20, pp. 26–34, Apr. 2018, doi: for non-intrusive load monitoring,” 2015 IEEE Global Conference on 10.1016/J.ESR.2018.01.003. Signal and Information Processing, GlobalSIP 2015, pp. 210–214, Feb. [50] K. Handayani, I. Overland, B. Suryadi, and R. Vakulchuk, “Integrating 2016, doi: 10.1109/GLOBALSIP.2015.7418187. 100% renewable energy into electricity systems: A net-zero analysis for [69] C. Beckel, W. Kleiminger, R. Cicchetti, T. Staake, and S. Santini, “The Cambodia, Laos, and Myanmar,” Energy Reports, vol. 10, pp. 4849– ECO Data Set and the Performance of Non-Intrusive Load Monitoring 4869, Nov. 2023, doi: 10.1016/J.EGYR.2023.11.005. Algorithms”, doi: 10.1145/2674061.2674064. [51] “NEMO: the Next Energy Modeling system for Optimization | SEI.” [70] “EMBED : A Dataset for Energy Monitoring through Building Electricity Accessed: Apr. 04, 2025. [Online]. Available: Disaggregation.” Accessed: Apr. 07, 2025. [Online]. Available: https://www.sei.org/tools/nemo-the-next-energy-modeling-system-for- http://embed-dataset.org/ optimization/ [71] C. Shin, E. Lee, J. Han, J. Yim, W. Rhee, and H. Lee, “The ENERTALK [52] M. Stüber and L. Odersky, “Uncertainty modeling with the open source dataset, 15 Hz electricity consumption data from 22 houses in Korea,” framework urbs,” Energy Strategy Reviews, vol. 29, p. 100486, 2020, doi: Scientific Data 2019 6:1, vol. 6, no. 1, pp. 1–13, Oct. 2019, doi: https://doi.org/10.1016/j.esr.2020.100486. 10.1038/s41597-019-0212-5. [53] C. Schimeczek et al., “AMIRIS: Agent-based Market model for the [72] A. Monacchi, D. Egarter, W. Elmenreich, S. D’alessandro, and A. M. Investigation of Renewable and Integrated energy Systems,” J Open Tonello, “GREEND: An Energy Consumption Dataset of Households in Source Softw, vol. 8, no. 84, p. 5041, Apr. 2023, doi: Italy and Austria”, Accessed: Apr. 07, 2025. [Online]. Available: 10.21105/JOSS.05041. http://www.pecanstreet.org/projects/consortium/ [54] “GitHub - energy-modelling-toolkit/Dispa-SET: The Dispa-SET unit- [73] M. Pullinger et al., “The IDEAL household energy dataset, electricity, commitment and optimal dispatch model, developed at the JRC.” gas, contextual sensor data and survey data for 255 UK homes,” Sci Data, Accessed: Apr. 03, 2025. [Online]. Available: https://github.com/energy- vol. 8, no. 1, p. 146, Dec. 2021, doi: 10.1038/S41597-021-00921-Y. modelling-toolkit/Dispa-SET [74] H. Rashid, P. Singh, and A. Singh, “I-BLEND, a campus-scale [55] S. Hilpert, C. Kaldemeyer, U. Krien, S. Günther, C. Wingenbach, and G. commercial and residential buildings electrical energy dataset,” Scientific Plessmann, “The Open Energy Modelling Framework (oemof) - A new Data 2019 6:1, vol. 6, no. 1, pp. 1–12, Feb. 2019, doi: approach to facilitate open science in energy system modelling,” Energy 10.1038/sdata.2019.15. Strategy Reviews, vol. 22, pp. 16–25, Nov. 2018, doi: [75] N. Batra, M. Gulati, A. Singh, and M. B. Srivastava, “It’s Different: 10.1016/J.ESR.2018.07.001. Insights into home energy consumption in India”, Accessed: Apr. 07, [56] C. Gisler, A. Ridi, D. Zujferey, O. A. Khaled, and J. Hennebert, 2025. [Online]. Available: www.goo.gl/01edPS “Appliance consumption signature database and recognition test [76] G. Hebrail and A. Berard, “Individual Household Electric Power protocols,” 2013 8th International Workshop on Systems, Signal Consumption,” 2006. Processing and Their Applications, WoSSPA 2013, pp. 336–341, 2013, [77] J. Z. Kolter and M. J. Johnson, “REDD: A Public Data Set for Energy doi: 10.1109/WOSSPA.2013.6602387. Disaggregation Research”, Accessed: Apr. 07, 2025. [Online]. Available: [57] A. Ridi, C. Gisler, and J. Hennebert, “ACS-F2 - A new database of http://www.enmetric.com appliance consumption signatures,” 6th International Conference on Soft Computing and Pattern Recognition, SoCPaR 2014, pp. 145–150, Jan. [78] Um. UMassTraceRepository, “Smart data set for sustainability,” 2024, 2014, doi: 10.1109/SOCPAR.2014.7007996. IEEE Dataport. doi: 10.21227/g91z-hz32. [79] A. Reinhardt et al., “On the Accuracy of Appliance Identification Based [58] S. Makonin, F. Popowich, L. Bartram, B. Gill, and I. V. Bajić, “AMPds: on Distributed Load Metering Data,” 2012, Accessed: Apr. 07, 2025. A public dataset for load disaggregation and eco-feedback research,” 2013 [Online]. Available: http://www.tracebase.org. IEEE Electrical Power and Energy Conference, EPEC 2013, 2013, doi: 10.1109/EPEC.2013.6802949. [80] J. Kelly and W. Knottenbelt, “The UK-DALE dataset, domestic appliance-level electricity demand and whole-house demand from five [59] S. Makonin, B. Ellert, I. V. Bajić, and F. Popowich, “Electricity, water, UK homes,” Scientific Data 2015 2:1, vol. 2, no. 1, pp. 1–14, Mar. 2015, and natural gas consumption of a residential house in Canada from 2012 doi: 10.1038/sdata.2015.7. to 2014,” Scientific Data 2016 3:1, vol. 3, no. 1, pp. 1–12, Jun. 2016, doi: 10.1038/sdata.2016.37. [81] L. Pereira, F. Quintal, R. Gonçalves, and N. J. Nunes, “SustData: A Public [60] C. Miller et al., “The ASHRAE Great Energy Predictor III competition: Dataset for ICT4S Electric Energy Research,” ICT for Sustainability Overview and results,” Sci Technol Built Environ, vol. 26, no. 10, pp. 2014, ICT4S 2014, pp. 359–368, Aug. 2014, doi: 10.2991/ICT4S- 1427–1447, Aug. 2020, doi: 10.1080/23744731.2020.1795514. 14.2014.44. [82] S. Barker, A. Mishra, D. Irwin, E. Cecchet, P. Shenoy, and J. Albrecht, [61] T. Kriechbaumer and H.-A. Jacobsen, “Data Descriptor: BLOND, a “Smart*: An Open Data Set and Tools for Enabling Research in building-level office environment dataset of typical electrical appliances,” Sustainable Homes,” 2012. 2018, doi: 10.1038/sdata.2018.48. [83] A. Nadeem and N. Arshad, “PRECON: Pakistan residential electricity [62] F. Jazizadeh, M. Afzalan, B. Becerik-Gerber, and L. Soibelman, “A consumption dataset,” e-Energy 2019 - Proceedings of the 10th ACM Dataset for Energy Monitoring through Building Electricity International Conference on Future Energy Systems, pp. 52–57, Jun. Disaggregation,” ACM Reference format, 2018, doi: 2019, doi: 10.1145/3307772.3328317. 10.1145/3208903.3208939. [63] C. Miller and F. Meggers, “The Building Data Genome Project: An open, [84] S. Hasan, A. Blinov, A. Chub, and D. Vinnikov, “Solar PV Generation public data set from non-residential building electrical meters,” Energy and Consumption Dataset of an Estonian Residential Dwelling,” Scientific Procedia, vol. 122, pp. 439–444, Sep. 2017, doi: Data 2025 12:1, vol. 12, no. 1, pp. 1–8, Mar. 2025, doi: 10.1038/s41597- 10.1016/J.EGYPRO.2017.07.400. 025-04747-w. [85] D. Murray, L. Stankovic, and V. Stankovic, “An electrical load [64] C. Miller et al., “The Building Data Genome Project 2, energy meter data measurements dataset of United Kingdom households from a two-year from the ASHRAE Great Energy Predictor III competition,” Scientific longitudinal study,” Scientific Data 2017 4:1, vol. 4, no. 1, pp. 1–12, Jan. 2017, doi: 10.1038/sdata.2016.122. [86] S. Athanasoulias et al., “The Plegma dataset: Domestic appliance-level and aggregate electricity demand with metadata from Greece,” Scientific Data 2024 11:1, vol. 11, no. 1, pp. 1–17, Apr. 2024, doi: 10.1038/s41597- 024-03208-0. [87] J. Chavat, S. Nesmachnow, J. Graneri, and G. Alvez, “ECD-UY, detailed household electricity consumption dataset of Uruguay,” Scientific Data 2022 9:1, vol. 9, no. 1, pp. 1–16, Jan. 2022, doi: 10.1038/s41597-022- 01122-x. [88] C. Quesada, L. Astigarraga, C. Merveille, and C. E. Borges, “An electricity smart meter dataset of Spanish households: insights into consumption patterns,” Scientific Data 2024 11:1, vol. 11, no. 1, pp. 1– 17, Jan. 2024, doi: 10.1038/s41597-023-02846-0. [89] M. Patin, S. Bégot, F. Gustin, and V. Lepiller, “Comprehensive bottom- up methodology for generating high-resolution yearly building load profiles: A case study in temperate oceanic climate,” Energy Build, vol. 320, p. 114584, Oct. 2024, doi: 10.1016/J.ENBUILD.2024.114584.