Energy Conversion and Management-X最新文献

{"title":"Machine learning-driven building energy optimization: integrating EnsembleXGBoost sensitivity analysis with meta-heuristic algorithms and carbon economics","authors":"Mahdi Asghari,&nbsp;Melika Jahangiri,&nbsp;Khosro Ashrafi","doi":"10.1016/j.ecmx.2026.101619","DOIUrl":"10.1016/j.ecmx.2026.101619","url":null,"abstract":"<div><div>While recent building optimization studies achieve significant energy reductions, they often overlook two critical gaps: (1) proceeding on fixed, non-optimized building geometries, and (2) using machine learning (ML) solely as computational surrogates rather than strategic optimization guides. This study introduces a novel four-stage framework that establishes formalized procedural linkages, with each stage systematically informing subsequent decisions. The methodology integrates Grey Wolf Optimization (GWO) for geometric baseline establishment, a robust EnsembleXGBoost-SHAP sensitivity analysis (SA), NSGA-II for targeted multi-objective parameter optimization, and ASHRAE standards implementation. The EnsembleXGBoost model demonstrated superior predictive accuracy (R²: 0.973–0.995), a finding statistically confirmed by paired t-tests (p &lt; 0.002) and bootstrap stability analysis (CV &lt; 0.12). Critically, validated SHAP insights—identifying key drivers (Infiltration, External Wall, Lighting, Heating Setpoint) accounting for ∼ 80% of performance variation—were used to formally define and constrain the NSGA-II decision variables, transforming SA from an interpretative endpoint to an optimization input. This integrated approach achieved a 59.4% reduction in total operational energy (rounded to 60.0%) across cumulative stages (GWO: 3.95%, NSGA-II: 49.7%, ASHRAE: 15.9%). The plausibility of this reduction is supported by validating the reference building’s baseline EUI (163.98 kWh/m2/year) against regional literature and the technical feasibility of the optimized parameters (U-value 0.09 W/m2K, Infiltration 0.2 ac/h) against empirical studies. Furthermore, the framework simultaneously improved occupant thermal comfort (average PMV improved from −0.75 to −0.15). A comparative economic scenario analysis revealed stark market-dependent viability: US market conditions supported positive NPV by the 16th year (WCP), while subsidized Iranian markets required carbon pricing mechanisms to achieve viability (36th year). The framework advances building optimization by bridging the disconnect between ML-based sensitivity analysis and multi-objective optimization while demonstrating that preliminary geometric optimization significantly improves parametric optimization baselines.</div></div>","PeriodicalId":37131,"journal":{"name":"Energy Conversion and Management-X","volume":"30 ","pages":"Article 101619"},"PeriodicalIF":7.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146189895","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A hybrid degradation prediction method for PEMFC integrating model-based degradation index extraction and Bayesian-optimized Bi-directional long short-term memory","authors":"Chang Ke ,&nbsp;Kai Han ,&nbsp;Yongzhen Wang ,&nbsp;Rongrong Zhang ,&nbsp;Xuanyu Wang ,&nbsp;Ziqian Yang ,&nbsp;Xiaolong Li","doi":"10.1016/j.ecmx.2026.101593","DOIUrl":"10.1016/j.ecmx.2026.101593","url":null,"abstract":"<div><div>Accurately estimating the state of health of proton exchange membrane fuel cell (PEMFC) and predicting the degradation trend are essential prerequisites for effective health management to enhance durability. This paper proposes a generalized hybrid degradation prediction method for PEMFC that is applicable to diverse operating conditions. Firstly, the internal polarization dynamics are characterized via the distribution of relaxation times method, and a third-order equivalent circuit model is established to quantify the polarization losses. The voltage losses are quantified using a polarization curve model. Degradation characteristic analysis from both approaches consistently reveals that deterioration in mass transfer kinetics and charge transfer kinetics is the primary cause of performance degradation. Subsequently, component-level degradation indexes are extracted based on degradation models, and a novel weighted fusion method is proposed to construct a hybrid degradation index characterizing the overall degradation state of PEMFC. Finally, the Bayesian-optimized Bi-directional long short-term memory (Bi-LSTM) model is employed to predict PEMFC degradation trend under various prediction horizons, enabling accurate estimation of remaining useful life (RUL). The results show that the optimized Bi-LSTM achieves higher RUL estimation accuracy than the baseline Bi-LSTM, and the hybrid method outperforms the AutoML-based method and the cascaded echo state network reported in previous studies. For the first stack, the estimation error remains below 7.78%, with a minimum error of 0.50%. For the second stack, the estimation error does not exceed 12.28% overall and drops below 10% when the prediction horizon is within 300 h, with a minimum error of 2.67%.</div></div>","PeriodicalId":37131,"journal":{"name":"Energy Conversion and Management-X","volume":"30 ","pages":"Article 101593"},"PeriodicalIF":7.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146080386","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A quasi-2D multiphase flow proton exchange membrane fuel cell model for efficient distributed cell state prediction","authors":"Florian Altmann ,&nbsp;Dominik Kuzdas ,&nbsp;Dominik Murschenhofer ,&nbsp;Johanna Bartlechner ,&nbsp;Christoph Hametner ,&nbsp;Stefan Jakubek ,&nbsp;Stefan Braun","doi":"10.1016/j.ecmx.2026.101584","DOIUrl":"10.1016/j.ecmx.2026.101584","url":null,"abstract":"<div><div>To enhance the durability and performance of proton exchange membrane fuel cells, it is essential to capture both spatial and temporal variations of internal states during dynamic operation. While existing reduced-order models (0D/1D) lack spatial resolution, 3D models are often too computationally expensive for transient simulations. To bridge this gap, we present a quasi-2D, time-dependent multiphase model capable of predicting distributed cell states with high computational efficiency. The model accounts for key transport phenomena, including convection, multicomponent diffusion, capillary effects, and membrane water dynamics via electro-osmotic drag and diffusion. It also includes nitrogen crossover, finite-rate sorption/desorption at membrane interfaces, and heat generation from electrochemical reactions, proton conduction, and phase change. A linearisation scheme combined with Chebyshev collocation ensures low computational cost and near real-time capability. Validation against high-resolution 3D computational fluid dynamics simulations confirms the model’s accuracy in predicting polarisation curves, gas species distributions, liquid water accumulation, and temperature profiles. Dynamic simulations under load transients further demonstrate its ability to capture key physical processes, underpinning the importance of spatially resolved water transport. By enabling fast and accurate simulations of both steady-state and dynamic fuel cell behaviour, the proposed model supports extensive parametric studies, control system development, and predictive diagnostics. Its computational efficiency makes it a valuable tool for improving fuel cell efficiency, longevity, and system-level control strategies.</div></div>","PeriodicalId":37131,"journal":{"name":"Energy Conversion and Management-X","volume":"30 ","pages":"Article 101584"},"PeriodicalIF":7.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146080387","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Chilled water-based hybrid cooling solution for data centers: A comprehensive survey of technologies, developments, and regenerative energy transitions","authors":"Majid J. Almheiri ,&nbsp;Haris M. Khalid ,&nbsp;Abdulla Ismail ,&nbsp;Asif Gulraiz ,&nbsp;Zafar Said","doi":"10.1016/j.ecmx.2026.101612","DOIUrl":"10.1016/j.ecmx.2026.101612","url":null,"abstract":"<div><div>Data centers face new cooling challenges due to the growth of digital infrastructure as well the rise in computing needs. This makes it urgent to find more sustainable ways to manage heat. To address this challenge, this proposed study surveys the chilled water-based hybrid cooling systems as a practical way to meet higher cooling demands while being environmentally responsible. To achieve this, the proposed research uses several methods: 1) it reviews cooling basics, 2) looks at recent technology, 3) studies control systems, and 4) examines real-world case studies. By reviewing current literature and industry examples, the proposed study highlights key performance and sustainability measures that could show the benefits of this choice towards hybrid cooling. Results show that combining water and air cooling with smart sensors, the Internet of Things (IoT), and artificial intelligence (AI) control can greatly improve energy efficiency. This would also make operations more resilient to climate change. The utilization of advanced chillers (chilled water-based hybrid cooling), heat exchangers, and phase-change materials helps transfer heat more efficiently, while using renewable energy can lower carbon emissions. Though there are challenges, such as saving water and working with older systems, the long-term savings and environmental gains make these hybrid systems more important for sustainable data centers. This proposed paper also offers useful insights for industry professionals who are working to adopt greener cooling solutions while keeping data centers reliable and high-performing.</div></div>","PeriodicalId":37131,"journal":{"name":"Energy Conversion and Management-X","volume":"30 ","pages":"Article 101612"},"PeriodicalIF":7.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146189893","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Design and utilisation of a multi-criteria decision support tool to analyse power-to-heat technologies for advancing industrial electrification","authors":"Arman Ashabi ,&nbsp;Mohamed Mostafa ,&nbsp;Andriy Hryshchenko ,&nbsp;Ken Bruton ,&nbsp;Dominic T.J. O’sullivan","doi":"10.1016/j.ecmx.2026.101631","DOIUrl":"10.1016/j.ecmx.2026.101631","url":null,"abstract":"<div><div>Transitioning industrial processes towards sustainability necessitates robust decision-support solutions. This paper introduces FlexiHeat-DST, a cutting-edge web-based tool dedicated to power-to-heat applications. The platform comprises three core functionalities: technology selection, multi-criteria decision analysis, and scenario exploration, offering an integrated approach to identifying optimal electrification pathways. It incorporates eight techno-economic and environmental parameters within a normalization framework to ensure fair comparisons, while an intuitive dashboard allows stakeholders to readily customize applications, adjust parameters, and dynamically visualize outcomes. A practical validation in the steel industry compared four electric-based alternatives for replacing fossil-fuel-intensive furnaces. The results indicated Induction and Resistance furnaces as top performers, scoring 0.669 and 0.578, followed by Plasma and Electric Arc furnaces, which scored 0.500 and 0.367, respectively. Scenario analyses revealed that while Induction and Resistance furnaces excelled under criteria such as technology maturity and installation simplicity, Electric Arc and Plasma furnaces were superior in scenarios prioritising efficiency and decarbonisation. In addition, sensitivity analyses recognised strong correlations between electrification stages, installation complexity, and overall technology rankings. Also, technologies superior in efficiency and carbon reduction typically incurred higher expenditures, adversely affecting their overall scores due to lower ratings in other critical areas such as Technology Readiness Level and lifespan. Finally, the analysis of energy costs and carbon emissions indicated that, despite significant environmental advantages, electrification remains economically challenging under prevailing electricity pricing structures. This underscores the necessity for strategic policy shifts and comprehensive decarbonisation plans within industrial sectors to achieve economically viable and efficient heat decarbonisation.</div></div>","PeriodicalId":37131,"journal":{"name":"Energy Conversion and Management-X","volume":"30 ","pages":"Article 101631"},"PeriodicalIF":7.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146189894","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dual-porosity model for harmonic pulse testing in fractured geothermal reservoir","authors":"Peter A. Fokker ,&nbsp;Eloisa Salina Borello ,&nbsp;Francesca Verga ,&nbsp;Dario Viberti","doi":"10.1016/j.ecmx.2026.101624","DOIUrl":"10.1016/j.ecmx.2026.101624","url":null,"abstract":"<div><div>Well testing and conventional Pressure Transient Analysis (PTA) are fundamental and well-established methodologies for characterizing well and reservoir parameters. However, the applicability of PTA is limited during production or injection operations, since it requires a shut-in of the tested well, and it is significantly affected by interferences from neighboring wells.</div><div>In previous works, we proposed, implemented, and validated against real data a methodology called Harmonic Pulse Testing (HPT). HPT is complementary to PTA. By specifically deploying the periodicity of rate and pressure signals, it has been designed to be applied during ongoing field operations.</div><div>In this work, we present a new analytical solution for HPT in naturally fractured reservoirs. The proposed solution is also applied to geothermal systems, as it is coupled with a radial composite model capable of approximating the thermal front. The model has been validated against well-established analytical and numerical models under different scenarios. The calculation steps for converting the numerical dual-porosity model into storativity ratio and inter-porosity flow coefficient are also provided.</div><div>The results of a validation exercise demonstrate that our model is robust against potential interference from other wells and allows the detection of the thermal front. The methodology can therefore be successfully applied during ongoing operations in naturally fractured geothermal reservoirs.</div></div>","PeriodicalId":37131,"journal":{"name":"Energy Conversion and Management-X","volume":"30 ","pages":"Article 101624"},"PeriodicalIF":7.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146080232","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A fuzzy two-phase model for renewable energy system optimization under uncertainty: from operational to strategic scenario planning","authors":"Hamed Fazlollahtabar","doi":"10.1016/j.ecmx.2026.101600","DOIUrl":"10.1016/j.ecmx.2026.101600","url":null,"abstract":"<div><div>Integrating volatile renewable energy into power grids requires robust optimization methods that avoid restrictive probabilistic assumptions. This study proposes a novel two-phase fuzzy model bridging operational planning and strategic market analysis. Phase 1 uses a Mamdani-type fuzzy inference system to generate realistic scenarios for solar/wind availability and demand. Phase 2 formulates a fuzzy linear program to minimize costs under uncertainty, the results of which inform a fuzzy Nash equilibrium model to analyze market participant strategies. Validated on a 10-node model with real California ISO data, our approach reduces system costs by 18.7% and maintains 99.2% reliability, outperforming stochastic and robust benchmarks. The model provides a practical decision-support tool for system operators and policymakers in renewable-dominated energy markets, demonstrating that fuzzy logic effectively captures real-world imprecision without complex data requirements.</div></div>","PeriodicalId":37131,"journal":{"name":"Energy Conversion and Management-X","volume":"30 ","pages":"Article 101600"},"PeriodicalIF":7.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146080234","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Security domain modeling and suppression of commutation failure for LCC-HVDC considering AC-DC coupling","authors":"Shoudong Xu ,&nbsp;Jinxin Ouyang ,&nbsp;Mingyu Pang ,&nbsp;Taiyu Xiao ,&nbsp;Chao Xiao","doi":"10.1016/j.ecmx.2026.101608","DOIUrl":"10.1016/j.ecmx.2026.101608","url":null,"abstract":"<div><div>In line-commuted converter based high voltage direct current (LCC-HVDC), commutation failure (CF) is a main source of danger. The CF is usually suppressed by regulating the single electrical quantity. However, for the active and reactive power of inverter station, the combination of DC current and voltage, AC bus voltage, firing angle (FA) and extinction angle (EA) is the determining influence. The regulation of any electrical quantity leads to variations in power, thereby influencing the AC bus voltage and subsequently inducing variations in the DC current, DC voltage, and EA. The neglect of AC-DC coupling affects the suppression of CF and limits power transmission. The formulation of the active and reactive power of inverter station has been deduced. Considering the constraints imposed by DC current and firing advance angle (FAA), the power feasible range is established. The feasible range of AC bus voltage and power is built considering AC-DC coupling. The commutation security domain of inverter station to avoid CF is modeled. To maximize the active power, formulations of DC current and FAA are proposed. Taking into account the power transmission, an improved suppression method of subsequent CF is proposed. The CIGRE standardized model is utilized to verify the method.</div></div>","PeriodicalId":37131,"journal":{"name":"Energy Conversion and Management-X","volume":"30 ","pages":"Article 101608"},"PeriodicalIF":7.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146080321","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Large-scale climate-neutral district heating and cooling: Integration of local microgrids for thermal distribution","authors":"Mohammad Sameti ,&nbsp;Tao Fan ,&nbsp;Anna Volkova ,&nbsp;Zili Li","doi":"10.1016/j.ecmx.2026.101552","DOIUrl":"10.1016/j.ecmx.2026.101552","url":null,"abstract":"<div><div>Conventional small-scale district heating (DH) systems that serve a limited number of buildings or neighbourhoods often exhibit higher specific capital costs and lower long-term efficiency compared to large-scale 4th or 5th generation DH networks. This is mainly because conventional systems typically operate at higher temperature levels and with less flexibility to integrate multiple renewable and waste heat sources, leading to greater distribution losses and reduced system synergy. In this study, underground metro space is utilized to integrate several small DHC systems into a single large-scale network, and a thermo-economic model is proposed. The metro helps lower the cost of distributing heat and makes it easier to integrate different heat sources. The cost reduction achieved in large-scale DHC systems is attributed to the use of existing conduits, reduced thermal losses, and enhanced heat exchange among the interconnected small DHC units. Additionally, this large-scale DHC network can easily accommodate future growth of consumers along it without extending the Metro and piping infrastructure and with upgrading the pumping capacity. A case study of the Dublin MetroLink, incorporating current 16 and future 26 small DHs, is analyzed to demonstrate the effectiveness of the proposed model. The primary heat sources considered are data centers and underground water from the Dublin Port Tunnel which also functions as the primary heat-transport medium. Each smaller DH along the underground route would utilize its own large-scale heat pump to extract heat from the supply line in the underground space and inject their extra/unused heat back to cover the peak demand in another smaller DHs. He case study showed 17% reduction in annualized cost over its lifetime.</div></div>","PeriodicalId":37131,"journal":{"name":"Energy Conversion and Management-X","volume":"30 ","pages":"Article 101552"},"PeriodicalIF":7.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146080322","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing dynamic economic emission dispatch through renewable energy prediction and multi-objective optimisation","authors":"Arunachalam Sundaram","doi":"10.1016/j.ecmx.2026.101898","DOIUrl":"10.1016/j.ecmx.2026.101898","url":null,"abstract":"<div><div>This study proposes an integrated framework for renewable-energy-assisted Dynamic Economic Emission Dispatch to support the transition toward net-zero power system operation. The framework uniquely combines multi-output renewable energy forecasting with advanced multi-objective optimisation to enhance dispatch efficiency and reliability. A unified feedforward Artificial Neural Network is developed and trained offline using 9 years (2009–2018) of NASA MERRA-2 meteorological data for the Texas region (31.0°N, −100.0°W) to simultaneously predict hourly solar irradiance and wind speed. The predicted values are converted into renewable power and directly incorporated into the DEED formulation. In addition, a secondary ANN-based model is employed to estimate transmission losses, eliminating the need for repeated load-flow calculations and overcoming the limitations of conventional B-loss coefficients. The resulting bi-objective optimisation problem, which minimises fuel cost and emissions, is solved using the Multi-Objective Grey Wolf Optimiser, enabling the generation of a diverse Pareto front in a single run. A fuzzy membership approach is then applied to identify the best compromise solution. The proposed framework is validated on a modified IEEE 30-bus system under multiple scenarios, including solar-only, wind-only, and hybrid renewable integration. Results indicate that hybrid integration achieves a maximum daily fuel cost saving of $540.24 and an emission reduction of 0.19 tons compared to the thermal-only case. These findings demonstrate that the proposed ANN–MOGWO framework provides a scalable and computationally efficient solution for dispatch planning in renewable-rich power systems, thereby reducing operational costs and enhancing environmental performance.</div></div>","PeriodicalId":37131,"journal":{"name":"Energy Conversion and Management-X","volume":"30 ","pages":"Article 101898"},"PeriodicalIF":7.6,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147802427","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}