Umeå University's logo

Demand response (DR) is a key strategy for enhancing energy flexibility, allowing buildings to dynamically adjust electricity demand and mitigate supply–demand mismatches—particularly in the context of rising renewable energy integration. Single-family houses (SFHs) are increasingly recognized as decentralized energy actors in advancing DR, owing to their suitability for integrating on-site generation systems such as photovoltaic (PV) panels. In such houses, an energy management system (EMS) coordinates local generation and consumption through DR optimization methods. Due to the high autonomy of single-family houses, effective DR optimization is critical for facilitating occupant participation, especially as thermal comfort significantly affects engagement. Although research in this domain is expanding, a systematic review focusing on DR optimization for SFHs with on-site generation and thermal comfort integration has yet to be conducted. To fill this gap, this review systematically synthesizes existing DR optimization methods in accordance with the PRISMA guidelines. DR optimization approaches are categorized into five groups: rule-based control, mathematical programming, metaheuristic optimization, model predictive control, and artificial intelligence-based methods. It also classifies thermal comfort integration approaches into four types: comfortable temperature zone (CTZ), comfortable temperature deadband (CTD), PMV–PPD, and adaptive thermal comfort models. A mechanistic framework integrating thermal comfort into DR optimization is developed, and a six-dimensional analysis reveals key methodological trade-offs and emerging trends. Finally, the review highlights key research gaps and outlines future directions, including refined thermal comfort metrics, occupant-centric and behavior-aware optimization frameworks, and uncertainty-aware strategies to ensure robust and scalable DR deployment in single-family houses.

Upcoming energy transition initiatives, such as Sweden’s capacity-based electricity tariff in 2027, aims to incentivize changes in occupant behaviour (OB) influencing how occupants schedule and shift energy use. Urban building energy modelling (UBEM) is widely used for city-level energy assessment. However, most UBEM studies assume uniform occupant behaviour and homogeneous building properties due to modelling limitations in incorporating diversity and heterogeneity, which is computational infeasible at large scale. In practice, OB varies across demographics and seasonal routines, while heterogeneous building properties such as U-values, HVAC, archetypes directly affect energy impacts. Oversimplifying OB diversity and urban building heterogeneity can misrepresent energy demand, peak loads, and policy effectiveness of energy transition initiatives.

This study introduces an enhanced framework, referred as DOB-HUBS, that explicitly incorporates OB diversity (e.g., seasonal, demographic) along with urban heterogeneity into UBEM. This enables ability to evaluate behavioural impacts across heterogeneous building clusters and occupants’ seasonal interactions. The approach employs unsupervised K-Means clustering and proportional stratified sampling to identify representative buildings, an automated bottom-up physics simulation workflow programmatically assigning diverse OB schedules and building properties, and an ensemble machine learning model that efficiently extrapolates predictions to the entire stock. This framework is demonstrated in Sweden, by evaluating seasonal impacts of excessive window opening, and behavioural responses to 2027 tariff policy. Results reveal strong seasonal dependence of OB, with deviations of up to 15% with 3.06% (standard deviation) energy use compared to traditional approach. Behavioural responses to 2027 policy could reduce peak loads by 6–17%, with impacts varying across small building clusters. By evaluating OB diversity and urban heterogeneity, DOB-HUBS enhances UBEM, supporting urban planners and policymakers in designing more effective occupant engagement strategies.

Projecting building-stock climate resilience is essential for urban climate adaptation. The archetype-based approach, widely used for stock modelling, represents buildings with a small set of archetypes and scales the results to the full building population. However, building climate performance is highly sensitive to detailed building attributes, such as HVAC systems and controls, envelope thermal properties, and window/shading details, which greatly differentiate the buildings’ climate resilience. This sensitivity often conflicts with the core premise of classical archetype methods, which assume uniform attributes and rely on homogeneous modelling within same building groups to enable archetype’s scalability. The absence of climate-sensitive attributes may constrain the identification of resilient or vulnerable buildings and constrains the design of targeted, effective adaptation measures. This study aims to enhance the archetype methods by proposing a general data-augmented framework for building-stock climate modelling, enabling vulnerable buildings clustering and effective adaptation measures identification. The proposed approach is designed to complement archetype methods through integrating multi-source building data, augmenting archetype models, and performing data-driven analysis to support climate adaptation. It is applied to the residential building stock of Umeå, Sweden, under two Nordic future climate scenarios: a near-term extreme heat year (2030) and a mid-term gradual warm year (2050). The results indicate that the vulnerable buildings were successfully clustered and effective adaptation measures were identified. Building renovations such as adjusting to mechanical ventilation and behaviour adaptations like active curtain use were found to reduce overheating by 26% and 5%, respectively. Overall, this approach extends classical archetype methods for stock-level climate modelling, enabling targeted identification of at-risk buildings and selection of effective adaptation actions.

Buildings contribute to global energy consumption and greenhouse gas emissions, making energy-efficient interventions important for sustainable development. In practice, the design and evaluation of such interventions commonly rely on correlation-based predictive models, which describe statistical associations but provide limited insight into the causal mechanisms linking environmental changes, occupant perceptions, and behavioral responses. As a result, interventions may produce outcomes that differ from expectations. This study introduces the Occupant-Centric Building Intervention Framework (OCBIF) designed to assess effectiveness of building energy interventions related to OBI. It establishes causal relationships among environments, occupant characteristics and perceptions, and adaptive actions by adopting the Driver–Need–Action–System (DNAS) concept. It employs Immersive Virtual Environments (IVEs) to simulate building contexts to allow observations related to occupant-building interaction (OBI) for final validation of building interventions. The case study is demonstrated using scenarios related to building interventions aiming to reduce heater uses. The causal analysis yields insights into thermal comfort and OBI. The results show that thermal sensation mediates the effect of indoor temperature on heater interaction, while age and system accessibility are additional causal influences on OBI. This causal structure explains why changes in indoor temperature do not translate directly into behavioral responses and why identical thermal interventions can lead to heterogeneous outcomes across occupants. Findings revealed that OCBIF bridged gaps in building intervention research providing actionable insights for stakeholders to design interventions enhancing energy efficiency while considering OBI.

Accelerating climate change is fundamentally transforming indoor thermal environments, intensifying heat exposure and variability that threaten occupant health, productivity, and well-being. In response to changing indoor conditions, occupants adopt adaptive behaviors such as adjusting clothing, opening windows, or operating HVAC systems that in turn reshape indoor environments. Understanding these complex human-environment interactions under future climate scenarios is critical for developing resilient and occupant-centric building strategies. This study proposes a hybrid experimental framework that integrates immersive virtual environments (IVEs) with a climate-controlled physical laboratory to investigate thermal-related occupant behaviors under future climate scenarios. The experimental setup combines visual immersion through virtual scenarios with precise control of thermal stimuli, enabling realistic simulation of future indoor conditions. Behavioral responses, physiological signals (e.g., heart rate, skin temperature), and psychological assessments (e.g., perceived thermal comfort, stress) were systematically collected from participants exposed to varied thermal scenarios. The collected multi-dimensional data provide a basis for modeling occupant behavior patterns and identifying the physiological and psychological factors that drive adaptive responses. The study further outlines the future integration of reinforcement learning-based occupant behavior models with building energy simulations via co-simulation, enabling closed-loop modeling of human-building interactions. This approach contributes to advancing climate-resilient building design, supporting the development of adaptive control strategies grounded in empirical occupant data.

Polyurethane (PU) as a popular polymer is widely recognized for its exceptional thermal insulation properties, making it a critical material for applications requiring effective heat transfer management. Integrating Phase Change Materials (PCMs) into PU (PU-PCMs) has emerged as a highly effective strategy for enhancing building envelope performance, ensuring greater indoor thermal stability, and mitigating temperature fluctuations. This study introduces an Enhanced Multi-Scale Modeling approach to investigate the thermal conductivity and energy efficiency of PU-PCMs, with a particular focus on their comparative performance in subarctic and tropical climates. By combining Molecular Dynamics (MD) simulations with Finite Element Methods (FEM), the model seamlessly bridges molecular-scale interactions with macroscopic thermal behavior. Using a Representative Volume Element (RVE)-FEM framework, microscopic properties are translated into engineering-scale parameters, enabling accurate and efficient predictions of material performance across diverse climatic conditions. The findings demonstrate the capability of PU-PCMs to significantly enhance energy efficiency and indoor thermal comfort. In a case study of a single-family house, PU-PCMs achieved a 7.376% reduction in energy consumption and increased comfort hours under Stockholm's subarctic climate. Moreover, the adaptability of PU-PCMs was validated in both tropical and subarctic environments, consistently stabilizing indoor temperatures, reducing HVAC energy demands, and improving occupant comfort. These results underscore the potential of PU-PCMs as a passive thermal management solution, advancing sustainable building practices. The proposed multi-scale model offers a computationally efficient and precise tool for optimizing material design in energy-sensitive applications, reinforcing the versatility and significance of PU-PCMs across a range of climatic conditions.

The advancement of artificial intelligence (AI) in material design and engineering has led to significant improvements in predictive modeling of material properties. However, the lack of interpretability in machine learning (ML)-based material informatics presents a major barrier to its practical adoption. This study proposes a novel quantitative computational framework that integrates ML models with explainable artificial intelligence (XAI) techniques to enhance both predictive accuracy and interpretability in material property prediction. The framework systematically incorporates a structured pipeline, including data processing, feature selection, model training, performance evaluation, explainability analysis, and real-world deployment. It is validated through a representative case study on the prediction of high-performance concrete (HPC) compressive strength, utilizing a comparative analysis of ML models such as Random Forest, XGBoost, Support Vector Regression (SVR), and Deep Neural Networks (DNNs). The results demonstrate that XGBoost achieves the highest predictive performance ((Formula presented.)), while SHAP (Shapley Additive Explanations) and LIME (Local Interpretable Model-Agnostic Explanations) provide detailed insights into feature importance and material interactions. Additionally, the deployment of the trained model as a cloud-based Flask-Gunicorn API enables real-time inference, ensuring its scalability and accessibility for industrial and research applications. The proposed framework addresses key limitations of existing ML approaches by integrating advanced explainability techniques, systematically handling nonlinear feature interactions, and providing a scalable deployment strategy. This study contributes to the development of interpretable and deployable AI-driven material informatics, bridging the gap between data-driven predictions and fundamental material science principles.

Graphene-based polymer nanocomposites show great potential for thermal management, but accurately predicting their thermal conductivity remains challenging due to multiscale structural complexity and parameter uncertainty. We propose an innovative approach integrating interpretable stochastic machine learning with multiscale analysis to predict the macroscopic thermal conductivity of graphene-based polymer nanocomposites. Our bottom-up framework addresses uncertainties in meso- and macro-scale input parameters. Using Representative Volume Elements (RVEs) and Finite Element Modeling (FEM), we compute effective thermal conductivity through homogenization. Predictive modeling is powered by the XGBoost regression tree-based algorithm. To elucidate the influence of input parameters on predictions, we employ SHapley Additive exPlanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME), providing insights into feature interactions and interpretability. Sensitivity analyses further quantify the impact of design parameters on material properties. This integrated method enhances prediction accuracy, reduces computational costs, and bridges data-driven and physical modeling, offering a scalable solution for designing advanced composite materials for thermal management applications.

The utilization of immersive virtual environments (IVEs) has emerged as a pivotal tool in enhancing observation of occupant-building interaction (OBI) in non-existing and pre-operational buildings (e.g., buildings under-designed, renovated, and retrofitted). The data derived from IVEs are critical in developing Building Predictive Models (BPMs) that prioritize occupant comfort and optimize building performance. Nevertheless, a persistent challenge is the collection of sufficiently large sample sizes from IVEs, often resulting in data sets inadequate for creating accurate and dependable BPMs. To address the gap, the generation of synthetic data is one promising solution. Mega-trend diffusion (MTD) is particularly adept at managing the nuances of small, mixed-type, and imbalanced data sets aligning with the natures of the IVE data sets. This study explores MTD-based algorithms such as baseline MTD, baseline MTD with class probability function, and k-Nearest Neighbors MTD (kNNMTD), all of which are adept at addressing the inherent data challenges. Various small data sets associated with OBI in IVEs were used to test these algorithms. The fidelity of the synthetic data sets is assessed using the Pairwise Correlation Difference (PCD) and accuracy of Artificial Neural Networks (ANNs) trained on the synthetic data sets with several modeling structures. A variety of findings indicated strength and limitations of the algorithms, where some areas need further investigation. At this stage, the evaluation based on this study found that the kNNMTD produced synthetic data sets that were closest to the experimental data set (i.e., the smallest PCD), contributing to the most accurate ANN models.

As part of Sweden's commitment to carbon neutrality, various municipalities have established energy efficiency targets. Achieving these targets requires decision-making knowledge at the local level, particularly concerning the energy retrofitting of existing building stocks. It includes understanding of the thermal properties (U-values) of building stock, their impact on energy use, and the potential energy retrofits. Our research focuses on assessing the thermal properties performance and their impact on energy use of residential building stocks in Umeå, Sweden. We employ explainable artificial intelligence (XAI) integrated with machine learning regression framework to elucidate how different building thermal features influence the building's energy use and also the correlations among these features. The findings highlight the significant impact of building floor area on energy use, followed by location, age, etc. Among thermal properties, the exterior walls have high impact and attic floor has the lowest impact on the energy use of Umeå's residential building stock. Ultimately, this study provides municipality-level decision-making insights for planning energy retrofitting initiatives for Umeå building stock.