Building Energy Performance Modelling and Simulation

The building sector is currently facing two conflicting challenges—the urgent need to reduce energy consumption and greenhouse gas emissions, and the continuous growth of expectations regarding thermal comfort, indoor air quality, and health [1]. Addressing such a dual challenge requires robust analytical tools. Building Energy Modelling (BEM) and Building Performance Simulation (BPS) have therefore become indispensable tools to support energy-efficient design, optimize building operations, and guide retrofit strategies [2,3,4].

The recent literature highlights five major directions for BEM application [5]:

• Performance-driven design of new and retrofitted buildings;
• Operational optimization of HVAC and energy systems;
• Integration with real-time data and the development of digital twins;
• Urban Building Energy Modelling (UBEM) to support city-level energy planning;
• Building-to-grid interaction for demand response and resilience.

Simulation models in this field can be broadly divided into physical (forward) models—based on energy and mass balance equations—and data-driven (inverse) models, which rely on statistical or machine learning techniques. While the former offer strong physical interpretability, the latter provide efficiency and adaptability, particularly when combined in hybrid approaches [6].

The application of simulation methods has a long history [6]. Early overviews highlighted the growing potential of building simulation for design and operation [7], while more recent developments have demonstrated its use in co-simulation frameworks [8] and as a methodological basis for retrofitting strategies in existing buildings [9]. A wide range of software tools, such as EnergyPlus, ESP-r, and TRNSYS, now support both research and industry practice, each with their own strengths and limitations [10].

Contemporary research highlights the role of simulation in enhancing performance-driven design [5], as well as in operational optimization of building systems through advanced control strategies [8]. Furthermore, simulation has extended beyond the building scale towards UBEM, which provides valuable insights for city-scale planning and integration with renewable energy systems [11,12,13].

Optimization remains a key application of BPS, with multi-objective algorithms such as NSGA-II enabling simultaneous improvement of energy efficiency and thermal comfort [14]. Deterministic and probabilistic approaches to optimization have been applied for their ability to handle complex, non-smooth performance functions [15]. Case studies further demonstrate the value of optimization in tailoring retrofit and comfort strategies to local contexts [16].

Finally, current advances increasingly integrate data-driven methods into traditional simulation workflows. Hybrid frameworks combining physical modelling with machine learning enable predictive maintenance, advanced control, and digital twin applications [5]. The rapid expansion of sensor networks and IoT technologies is accelerating these developments, pointing towards a future where BPS will be tightly integrated with real-time building management and urban energy systems [17,18].

BPS methods based on macroscale models are complemented by computational fluid dynamics (CFD) methods for a detailed analysis of the distribution of air parameters in the room, which allows, for example, for the assessment of local human thermal sensations or the distribution of pollutants in the zone.

Against this background, this Special Issue of Energies, “Building Energy Performance Modelling and Simulation”, brings together a diverse set of research contributions. The published papers demonstrate the breadth of applications of simulation and modelling techniques—from building retrofits and new design concepts, through data-driven optimization, to advanced computational fluid dynamics (CFD) analysis. Together, they form a comprehensive picture of current advances in building energy performance research.

• Retrofitting and Climate-Resilient Design

The issue features three contributions that focus on the retrofit of existing housing stock. Ferdyn-Grygierek and Grygierek present a comparative analysis of passive and active retrofit solutions in ageing multifamily housing in Central Europe. Their co-simulation approach, combining EnergyPlus and CONTAM, reveals the high potential of hybrid passive measures such as solar protective glazing and reflective roofs to reduce summer overheating while maintaining winter energy efficiency. The work highlights the importance of tailoring retrofit strategies to future climate conditions.

A complementary perspective is offered by Menconi et al., who analyze retrofit strategies for traditional listed dwellings in the UK. Their study underlines the tension between heritage preservation and energy efficiency, proposing systemic approaches to passive retrofitting that respect the thermo-hygrometric balance of historic constructions. This contribution demonstrates that responsible retrofit can simultaneously safeguard cultural value and contribute to climate targets.

Ciuman et al. provide insights into newly designed single-family houses within the framework of the “4E Idea”—energy-saving, ecological, ergonomic, and economic. Their simulations in IDA ICE highlight the role of integrated design combining efficient HVAC systems, renewable energy sources, and cost optimization. The results not only serve as a proof-of-concept but also provide benchmarks for future sustainable residential designs.

• Data-Driven Modelling and Machine Learning

Two papers explore the integration of artificial intelligence and digital methods into building energy assessment. Nassif et al. apply machine learning techniques—ranging from linear regression to neural networks and ensemble methods—to predict heating coil performance in HVAC systems. Their results point to bagging and neural networks as highly promising approaches, capable of improving system efficiency and reducing operational uncertainties.

Tsikas et al. advance this line of research by proposing a BIM-based machine learning application for a parametric assessment of residential building energy performance. By training statistical and AI models on a dataset of 337 BIM-derived instances, they demonstrate that artificial neural networks can serve as accurate surrogate models for complex energy simulations. Importantly, they also deliver a user-friendly interface tool, making such predictive modelling accessible for practical design applications.

• Advanced Thermal and Airflow Simulation

The Special Issue also contains valuable contributions addressing the detailed physical modelling of building components and airflows. Urzędowski et al. use CFD and statistical analysis to investigate how material parameters, such as surface emissivity and roughness, affect the thermal resistance of ETICS wall systems. Their results show the potential of reflective coatings and microstructural optimization to significantly enhance wall performance while reducing thickness, with implications for prefabricated elements.

Hurnik et al. conduct a comprehensive validation of different turbulence models in predicting sidewall jet airflow in rooms. Their study confirms the reliability of the standard k-ε and EVTM models while highlighting the limitations of other approaches. This research provides practical guidance for selecting CFD models in HVAC design, particularly in the early stages where accurate airflow prediction is crucial for comfort assessment.

• Urban and Occupancy-Related Factors

Beyond the scale of individual components and systems, the issue also includes research on urban and human-related aspects of energy performance. Sadłowska-Sałęga and Wąs evaluate the impact of shading from neighbouring buildings in the historic centre of Kraków. Their simulations show that shading can increase heating demand while substantially reducing cooling loads, emphasizing that urban morphology must be accounted for in retrofit strategies and HVAC system sizing.

Nam and Kim explore energy performance improvements in Korean senior centres. Their combination of simulation and economic feasibility analysis reveals that although multiple measures improve efficiency, boiler replacement is the only intervention with a positive long-term economic return without government subsidies. This study is a reminder of the importance of coupling technical performance with financial viability in real-world decision-making.

Norouziasl et al. investigate occupancy factors in a US small office buildings using agent-based modelling combined with EnergyPlus simulations. They identify occupant density as the most influential parameter across climate zones. The study highlights the stochastic nature of occupancy and demonstrates the value of dynamic schedules for more realistic energy modelling.

• Conclusions and Perspectives

Taken together, the contributions in this Special Issue reflect both the maturity and the evolving frontiers of building energy performance simulation. On one hand, the works confirm the central role of simulation in assessing retrofit strategies, system performance, and comfort conditions. On the other, they showcase emerging opportunities—from AI-based surrogate modelling to hybrid physical-data approaches—that promise to accelerate and enhance decision-making in both design and operation.

Looking forward, several trends are likely to shape the field:

• the integration of physics-based and machine learning models into hybrid frameworks;
• increased emphasis on resilience to climate change, particularly in retrofitting existing and historic stock;
• the extension of simulation from individual buildings to urban contexts, incorporating morphology and microclimate;
• and the systematic inclusion of user behaviour and economic feasibility in performance assessments.

We hope that the articles presented here will serve as inspiration and a valuable reference for researchers, practitioners, and policymakers engaged in creating energy-efficient, resilient, and comfortable built environments.