Dalarna University's logo and link to the university's website

The airfoil fin printed circuit heat exchanger (PCHE) has great potential for deployment in advanced Brayton cycles with supercritical CO<inf>2</inf>-based mixtures as working fluid. However, accurately predicting the complex flow and heat transfer behavior in the PCHEs remains challenging. This study employs two prominent machine learning (ML) methods – Artificial Neural Networks (ANN) and Extreme Gradient Boosting (XGBoost) – to predict the thermal-hydraulic performance (i.e., Nusselt number Nu and Fanning friction factor f) of supercritical CO<inf>2</inf>-based mixtures in airfoil fin channels. Beyond conventional predictive modeling, an innovative application of SHAP (SHapley Additive exPlanations) analysis is introduced to provide novel physical understanding of model outputs. The results demonstrate that both ANN and XGBoost exhibit excellent prediction performance, significantly outperforming conventional correlations, with R2 exceeding 0.99 for Nu and approaching 0.95 for f as well as small root mean square error (RMSE) and weighted mean absolute percentage error (wMAPE). The models also effectively capture the structurally-induced fluctuations of Nu and f along the flow direction. The feature selection based on the Spearman correlation coefficient method yields a more compact feature space without compromising predictive capability. SHAP analysis reveals a consistent and dominant influence of heat flux (q) and Reynolds number (Re) on the predictions for two targets, with q primarily affecting Nu, while Re and the densities ratio (ρ<inf>b</inf>/ρ<inf>w</inf>) are crucial for f. Notably, a previously overlooked positive impact of the buoyancy effect on improving hydraulic performance is identified in this study. These findings demonstrate significant potential of explainable ML models in predicting the complex supercritical thermal-hydraulic performance, promoting reliable design and optimization of novel heat exchangers in supercritical Brayton cycles. © 2025 Elsevier B.V., All rights reserved.

Methane, a highly essential industrial raw material, plays a pivotal role in safeguarding national energy security and advancing sustainable development. Due to the expansion of industrial scale and increased integration in modern methane production, the production data exhibits complex multiscale variability over time, which poses great challenges for accurate methane production prediction. Therefore, a novel production prediction model is proposed by employing an improved Long Short-Term Memory Network (LSTM) combining with the multiscale feature fusion method (MSFF) (MSFF-LSTM). The MSFF decomposes the raw industrial process data into multiple two-dimensional tensors based on periods, which can ravel out the complex temporal fluctuations into multiple intraperiod-and interperiod-variations. Then, the methane prediction model is constructed utilizing multiple LSTM models to extract interactive features at various scales. Finally, using a feature fusion module to fuse the prediction results at different scales can fully aggregate local and global features for complementary prediction. Experimental results demonstrate that, compared with other prediction models, the MSFF-LSTM achieves the state-of-the-art results with the mean absolute error (MAE), the mean square error (MSE), coefficient of determination (R2) and the root mean square error (RMSE) of 0.1056, 0.0300, 0.9199 and 0.1733, respectively, which offers the optimization direction for the anaerobic digestion process of straw for methane production.

Accurate building energy consumption prediction is crucial for building energy management. However, a substantial number of buildings lack sufficient data that hinders the application of data-driven models for energy prediction. Transfer learning emerges as a powerful strategy to address the challenge posed by limited data availability. This research proposes a novel model ensemble method based on Multi-Layer Perception (MLP) structure, which can realise selfadaptive weight, to exploit the advantage of existing transfer learning strategies for building energy sequence-to-sequence (Seq2seq) prediction. The overall and stepwise model performance comparisons between traditional transfer models and traditional ensemble methods in 4 different transfer scenarios are conducted to prove the superior performance of proposed method under all the investigated transfer conditions. The impact of prediction step length on the model performance is also investigated. The results show that the proposed method outperforms traditional transfer models and ensemble methods at different prediction steps in all the investigated transfer conditions. Compared to the best performing transfer model, the proposed method can reduce prediction error by 6.83 %-25.08 %. Compared to the best performing ensemble method, the proposed method can reduce prediction error by 6.32 %-36.54 %. The analysis of the selfadaptive weight reveals that the proposed method is capable of dynamically allocating weights to the two transfer models to enhance the prediction accuracy.

The European Union (EU) has introduced a range of policies to promote energy efficiency, including setting specific targets for energy-efficient renovations across the EU building stock. This study provides a comprehensive environmental and economic assessment of energy-efficient renovation scenarios in a large-scale multifamily building project that is district-heated, considering both the building and the broader urban energy system. A systematic framework was developed for this assessment and applied to a real case in Sweden, where emission factors from energy production are significantly lower than the EU average: 114 g CO2e/kWh for district heating and 37 g CO2e/kWh for electricity. The project involved the renovation of four similar district-heated multifamily buildings with comparable energy efficiency measures. The primary distinction between the measures lies in the type of HVAC system installed: (1) exhaust ventilation with air pressure control, (2) mechanical ventilation with heat recovery, (3) exhaust ventilation with an exhaust air heat pump, and (4) exhaust ventilation with an exhaust air heat pump combined with photovoltaic (PV) panels. The study's findings show that the building with an exhaust air heat pump which operates intermittently with PV panels achieves the best environmental performance from both perspectives. A key challenge identified for future research is balancing the reduced electricity production from Combined Heat and Power (CHP) plants within the energy system.

Climate change is considered as one of the biggest threats that humankind is facing nowadays, with environmental, social, and economic consequences. The building sector is facing multiple climate change impacts, which is becoming more and more vulnerable. This is especially true considering that about 35% of the buildings in the European Union (EU) are over 50 years old and the replacement rate of new building in Europe is low. Therefore, it is expected that much of the existing building stock will be affected by several climate change impacts in near future. Specifically, from the building point of view, these impacts can range from a slight rise in the average environmental temperature and humidity levels to extreme and severe events (such as strong wind and floods), changing in most of the cases, the building performance, and thermal behavior. Among the adaptation strategies to climate change, the envelope optimization, whichever climate type, is the most effective way to reduce the building energy dependency and increase the indoor thermal comfort. In this regard, the integration of phase change materials (PCM) into the building envelope can produce a sort of extra thermal capacity to the building, enhancing its overall energy efficiency. Specifically, when PCM is used without any control systems, it means that it is passively contributing to the building thermal comfort, stabilizing the indoor temperature and reducing both cooling and heating demands. Considering that the effectiveness of PCM application over the building envelope is mostly associated with the selection of the appropriate melting temperature and thickness, in the context of climate change, it is expected that the optimal PCM melting point and amount found for the present period will not be optimal for future and vice versa. Therefore, the present book chapter presents a numerical investigation on the effectiveness of PCMs wall implementation as a resilient building refurbishment solution. Specifically, the book chapter aims at proofing the PCM’s capability of being an effective building refurbishment strategy, under historical and future climate conditions. The whole study is based on dynamic building simulations carried out by IDA ICE tool on a typical residential single zone house in Stockholm (Sweden) and Rome (Italy) cities. The results of the simulations highlight that PCM can contribute to a reduction of cooling demand and improve the indoor thermal comfort under both historical and future climate in Stockholm. In addition, PCM results in slight effectiveness in reducing heating loads, and the total annual energy saving is between −1.5% and −2.4% for the historical period and −1.9% and −5.7% for the future one. In Rome, the incorporation of a PCM layer in the building envelope slightly reduces the cooling demand and enhances the indoor thermal comfort, where the total annual energy saving equals to −1.6% for the historical period. Conversely, no beneficial effects in term of annual energy saving have been observed for future climate condition in Rome.

Accurate fault diagnosis in chillers is essential for maintaining optimal energy efficiency and operational reliability in building Heating, Ventilation, and Air Conditioning (HVAC) systems. However, chillers often operate under diverse conditions, which can cause data-driven models trained on specific operational conditions to suffer significant performance deterioration when deployed in new conditions. Transfer learning offers a promising solution by leveraging knowledge from source domains, but its "black-box" nature raises concerns about model interpretability, hindering practical application. To address this challenge, this study developed an evaluation framework integrating Layer-wise Relevance Propagation (LRP), Integrated Gradients (IG), and SHapley Additive exPlanations (SHAP) to assess feature-level interpretability in transfer learning-based FDD model. We modified the 1D CNN architecture specifically for compatibility with interpretation methods. The model achieves an overall accuracy of 97.01 %, with component-level faults showing higher accuracy than system-level faults. All interpretation methods consistently identify physically meaningful discriminative features, and the robustness of the methods is validated through 10 randomized trials. Meanwhile, data volume critically impacts the clarity of interpretation results—larger datasets yield higher feature importance scores, though even 1 % of training data is capable of identifying discriminative features. To simulate real-world chiller operation, multiple cross-condition transfer learning tasks were designed, covering a wide range of operating scenarios. Results demonstrate that the Domain-Adversarial Neural Network (DANN) and Fine-Tuning (FT) improve target-domain accuracy by 25 % over baseline models while preserving physically meaningful discriminative features from the source domain. © 2025 Elsevier B.V., All rights reserved.

The building provides to the occupants a shelter envelope and comfortable interior climate conditions. Starting from our ancestors, climate directly has great influence on both building design and the corresponding building’s overall energy performance. Nowadays, passive design has been getting popular again for taking advantage of the regional climate to maintain a comfort indoor climate, so as reducing or eliminating the dependence on active systems. Along with the prolonged construction service years, the current appeal of passive design is not only facing historical weather but also the changing future climate. It is undoubted that there is an ever-widening disparity between historical weather patterns and current—not to mention future—climate conditions resulting from anthropogenic changes. Consequently, this chapter focuses on this field and presents a preliminary climate-adaptive design study for urban multifamily buildings at early stage. Special attentions are paid to the indoor thermal comfort and minimum energy use from today to the last part of the twenty-first century. The generated future climate data combined with thermal comfort model assessment has been proposed as a new way of including future climate scenarios in preliminary building design for two representative sites, in Rome, Italy, and Stockholm, Sweden. The existing vulnerability to the expected climate conditions from psychometric analysis indicates that (1) the climate trend in Rome would gradually lead to more failures in the majority of conventional adaptive design measures, as the cooling and dehumidification demands would rise from 5.3% to 23.6%, while the heating and humidification demands would decrease from 27% to 16%, and (2) the climate trend in Stockholm would result in an increased comfort period by exploiting more adaptive design measures, since the heating and humidification demands would be reduced from 67% to 53%. However, the cooling and dehumidification demands would increase slightly from 0% to 1.5%. Accordingly, four main key risks are identified: (1) overheating would become a rising increasing public health threat for buildings in Rome that rely exclusively on natural ventilation; (2) open questions remain for the design team in the area of correct cooling load selection, additional space for the future installation and the effectiveness of current cooling device, etc.; (3) occasional heat waves and gradual rising humidity levels are expected to be a vulnerable topic for conventional lightweight building in Stockholm; and (4) buildings with a heavy heating load would tend to have greater cooling demand, especially those with poor ventilation resources or greater internal gains. In conclusion, it is suggested that envelope optimization, whichever climate type, is one of the most efficient and effective adaptation measures toward future climate conditions. After that, a detailed case study with a new container building is proposed accordingly.

As the global demand for renewable energy and environmental protection continues to grow, anaerobic digestion of food waste as an effective way of resource recycling and energy production has attracted widespread attention. And forecasting methane generation with precision throughout the anaerobic digestion (AD) process is crucial for optimizing the process and improving energy recovery efficiency. Therefore, this paper proposed a new time series prediction model based on the iTransformer method to accurately predict the biogas production during the AD of food waste. The iTransformer uses the attention mechanism to capture the inter-variable relationships, and sequentially processes the historical observations features layer by layer along the time dimension through the feedforward network to capture the complex dynamic characteristics of production process data and build a predictive model. Finally, the proposed method is used to forecast the methane yield and carbon dioxide emissions during the AD of food waste. Compared with the gate recurrent unit (GRU), the autoregressive integrated moving average (ARIMA), the long short-term memory network (LSTM) and Transformer methodologies, the proposed iTransformer method based time series prediction method performs well in the productivity prediction of Garment Employees (PPGM) dataset and the AD dataset, where the mean square error (MSE), coefficient of determination (R2), and accuracy are 0.0231, 0.9036, and 95.9118% on the PPGM dataset, and the MSE, R2, the root mean square error (RMSE) and accuracy are 3946.9602, 0.9949, 7.1596, and 98.5517% on the AD dataset, respectively. Moreover, the impact of different operational parameters on the AD process can be optimized through the prediction results to increase biogas production and reduce carbon emissions.

Previous research has shown a lack of studies with comparisons between primary (virgin) and secondary (re-used) building materials, and their embodied emissions. The creation of different scenarios comparing the environmental impact of virgin vs. re-used materials is also motivated by the scarcity of raw materials in the world and the emergency of mitigating greenhouse gas (GHG) emissions from buildings. The aim of this study was to investigate scenarios, including new vs. re-used building products, applying the LCA method for a wooden single-family building. The findings showed a 23% reduction potential for total released (positive) CO₂e when comparing the Reference scenario with Scenario I, using re-used wooden-based materials. Further, Scenario II, using all re-used building materials except for installations, showed a 59% CO₂e reduction potential compared to the Reference scenario. Finally, Scenario III, which assumes all re-used building products, showed a 92% decreased global warming potential (GWP) impact compared to the Reference scenario. However, when including biogenic carbon and benefits (A5 and D module), the Reference scenario, based on newly produced wooden building materials, has the largest negative GHG emissions. It can be concluded that the re-use of building products leads to significant carbon savings compared to using new building products.

To effectively reduce building energy consumption, a novel full fresh air system with a heat source tower (HST) and a borehole heat exchanger (BHE) was proposed for space cooling and dehumidification in this paper. The cooling system only adopts geothermal energy to produce dry and cold fresh air for space cooling and dehumidification through the BHE and HST, which has the advantage of non-condensate water compared to BHE systems integrated with a fan coil or chilled beam. Based on the established mathematical model of the cooling system, this paper analyzed the system characteristics, feasibility, operation strategy, energy performance, and cost-effectiveness of the proposed model in detail. The results show that the mathematical model has less than 10% error in estimating the system performance compared to the practical HST-BHE experimental set up. Under the specific boundary conditions, the cooling and dehumidification capacity of this system increases with the decrease in the air temperature, air moisture content, and inlet water temperature of the HST. The optimal cooling capacity and the system COP can be achieved when the air-water flow ratio is at 4:3. A case study was conducted in a residential building in Shenyang with an area of about 1800 m2. It was found that this system can fully meet the cooling and dehumidification demand in such a residential building. The operation strategy of the cooling system can be optimized by adjusting the air-water flow ratio from 4:3 to 3:2 during the early cooling season (7 June-1 July) and end cooling season (3 August-1 September). As a result, the average COP of the cooling system during the whole cooling season can be improved from 6.1 to 8.7. Compared with the air source heat pump (ASHP) and the ground source heat pump (GSHP) for space cooling, the proposed cooling system can achieve an energy saving rate of 123% and 26%, respectively. Considering that the BHE of the GSHP can be part of the proposed HST-BHE cooling system, the integration of the HST and GHSP for space cooling (and heating) is strongly recommended in actual applications.