Abstract

Solid oxide electrolysis is a promising technology for green hydrogen production. However, its wide thermal operating range, spanning endothermic, thermoneutral, and exothermic regimes, poses significant challenges for thermal management, particularly during load-following operations. Insufficient control of temperature and thermal gradients can result in substantial thermal stresses, leading to reduced durability and, in severe cases, cracking or failure of cells and stacks.
This work investigates the thermal performance of counter-flow and co-flow solid oxide electrolysis cells. A dynamic multi-physics model is developed to simulate coupled electrochemical, thermal, and fluid-flow behaviors. A model predictive control strategy, regulating both cathode and anode flows, is implemented for active thermal control of solid oxide electrolysis cells. The performance of the model predictive control strategy is compared to that of a traditional proportional–integral controller. The results demonstrate that the proposed control strategy outperforms the traditional proportional–integral controller. It effectively maintains the cell temperature at the target value of 800 °C while keeping the maximum thermal gradient across the cell at a moderate level, limited to 6 °C/cm for both co-flow and counter-flow configurations.

Abstract

Renewable and Citizen energy communities offer a collective pathway to decarbonize the building energy sector through shared energy generation, investment, and energy efficiency. Social science studies focus on RECs and CECs are mostly case specific. Current energy system models mostly do not capture individual building heterogeneity. EnerPlanET, an open-source, web browser-based, and highly spatially explicit model, aims at addressing this gap by integrating building energy system with regional renewable and energy infrastructure planning. This paper presents an open-source single building parameterized Python-based model, called Building Energy Model (BuEM). BuEM is intended to become one of EnerPlanET’s modules. It is based on the lumped capacitance 5R1C model that follows the NEN-ISO 52016-1 standard. BuEM calculates thermal energy demands considering detailed 3D building geometry, weather, and a European, commonly used, buildings-related TABULA database. The model can calculate the hourly heating and cooling loads of building for a year in less than 8 seconds using a standard laptop. The model input choices can be made by users, developers, or modelers, where atypical characteristics of every building within a community can be captured, which is highly helpful in regional spatial planning.

Abstract

District heating networks require fast and accurate forecasting tools, but classical numerical models are too slow for real-time use, and standard machine learning approaches cannot adapt to changing topologies. In addition, purely data-driven methods fail to take into account physical phenomena, such as mass conservation. Therefore, we propose a Physics-Informed Graph Neural Network (PIGNN) that incorporates the continuity equation into the loss function to enhance accuracy and convergence. Trained on ca. 32,000 samples from 620 hypothetical topologies, the PIGNN achieves a Pearson correlation coefficient of 0.91 (vs. 0.76 for a baseline GNN), converges faster, and provides a computational speedup of 4–5 orders of magnitude over classical methods. This makes it a promising tool for design exploration and operational planning in district heating networks.

Abstract

In this work, density functional theory (DFT) calculations were carried out to investigate the effect of surface terminations (–O and –F) on the hydrogen evolution reaction (HER) performance of single-atom Pt anchored Ti₃C₂Tₓ MXenes. The optimized structures reveal that surface functionalization induces subtle distortions of the lattice and modifies the electronic properties of the substrate. Gibbs free energy calculations show that bare Ti₃C₂ binds hydrogen too strongly (ΔGH* = –1.16888 eV), while Pt anchoring significantly weakens the binding (–0.89759 eV). Surface terminations further regulate the adsorption strength, with saPt-Ti₃C₂O₂ (–0.82934 eV) and saPt-Ti₃C₂F₂ (0.678677 eV) progressively approaching thermoneutrality. Projected density of states (PDOS) analysis demonstrates that O and F terminations downshift the Pt d-band center and reduce the density of states near the Fermi level, correlating with weaker Pt–H interactions. Charge density difference (CDD) plots further confirm that electron withdrawal by O and F enhances Pt electron deficiency, tuning the interfacial charge transfer and Pt–H bonding characteristics.

Abstract

Achieving carbon neutrality in Hong Kong by mid- century essentially lies in integrating innovative technologies, including hydrogen fuel cell vehicles (HFCVs). This study explores public awareness, perceptions, and acceptance of HFCVs in Hong Kong, using electric vehicles (EVs) as a benchmark. The barriers and opportunities for their widespread adoption are identified. The results show significant differences in public awareness: only 26.0% of the public know about HFCVs, less than half of that of EVs. Despite minimal HFCV driving experience (1.5%), 18% express purchase intent, highlighting latent market potential. However, risk perception and infrastructure gaps remain significant challenges. Forty percent of respondents expressed concerns about the safety issue on the hydrogen. 40% of respondents expressed positive feedback on the local hydrogen refueling station construction. Infrastructure shortages (nearly 50%) are considered as the biggest barrier for wide adoption of HFCVs in Hong Kong. These findings underscore sociotechnical barriers, including limited public understanding of HFCVs’ advantages in heavy-duty applications. To bridge this gap, policymakers in Hong Kong are urged to expand infrastructure for hydrogen energy and implementing targeted public engagement strategies to improve public acceptance on HFCVs. This research offers actionable insights to transform Hong Kong’s latent HFCV receptivity into tangible progress toward decarbonization.

Abstract

A promising method for geological COâ‚‚ storage is through the formation of COâ‚‚ hydrates in sediment pores. However, the uniform distribution of hydrates is essential to ensure storage stability and representative experimental properties. In this study, we investigated pore water redistribution during COâ‚‚ hydrate formation, with a comparative analysis of ice crystallization to elucidate underlying phase transition mechanisms. Experimental results revealed that under temperature gradients, phase change induces local water migration, resulting in heterogeneous water distribution. Key factors include the type of phase transition, temperature gradient, and salt concentration. Increased salinity reduced heterogeneity in water distribution, and at 0.5 mol/L NaCl, migration was largely inhibited, maintaining initial uniformity. A migration model incorporating these parameters effectively captured the behavior. A physical model incorporating these parameters effectively predicted the migration behavior. These findings highlight that similar heterogeneity in COâ‚‚ hydrate distribution may occur in field-scale storage sites, potentially affecting the sealing integrity and long-term stability of the reservoir.

Abstract

Electrification across sectors such as transport, heating, and industry has accelerated in recent years, with battery electric construction machinery emerging as a promising alternative in the construction industry. However, the high charging demand of such equipment can stress the grid through transformer overload and voltage deviations, while local renewable generation introduces challenges of overproduction and intermittency. This paper assesses the grid impact of electrifying construction equipment in a new district in a Swedish city, considering local solar and wind generation, and evaluates renewable-based smart charging strategies aimed at maximizing load matching and self-consumption. Results show that the integration of electric construction machinery and renewables increases voltage fluctuations, with renewables causing more severe overvoltage than machinery-induced undervoltage. Smart charging improves voltage stability by reducing the frequency of renewable-induced overvoltage events. The findings provide practical insights for construction site and system operators, showing that electrification of construction machinery, when paired with renewable-based smart charging, can support grid stability and sustainable operation.

Abstract

This paper presents a comparative study in plasma-assisted ammonia combustion (PAAC) characteristics ignited by different device configurations with same discharge parameters. Tested discharge devices include the nanosecond pin-to-pin discharge, the nanosecond surface dielectric barrier discharge with positive and negative electrode arrangements, which were denoted as nPTPD, nSDBD(+), and nSDBD(–), respectively. The applied discharge devices were connected to a pulse generator and separately mounted into a constant volume chamber test rig. The supplied energy to discharge devices, high frame video morphology, and pressure-derived combustion performance indicators were obtained and analysed to conduct the PAAC performance comparison. The results show nPTPD performs less supplied energy but faster and stronger PAAC than nSDBD. nSDBD(+) conducts higher supplied energy, faster but weaker PAAC than nSDBD(–).

Abstract

The burgeoning global population has heightened concerns regarding energy management and freshwater production. Traditional liquid air energy storage (LAES) systems, while innovative, struggle with low operational efficiency and substantial energy demands. Similarly, conventional desalination methods are impeded by inefficiencies and high energy consumption. Addressing these deficiencies is imperative to more sustainable energy and water resource management. In prior work, our team proposed two advanced polygeneration liquid air energy storage systems coupled with LNG cold energy, solar energy, and hydrate based desalination (LNG-HBD-SALAES). This study further conducts a comprehensive life cycle cost analysis to investigate their potential profitability in typical application scenarios near coastal LNG receiving terminals. In addition, it examines the sensitivity of the overall economic performance to various parameters. The results indicate that the initial capital cost of the LNG-HBD-SALAES project is 103.10 million USD. When the plant is situated in Guangdong Province, the annual operational and production cost is 5.45 million USD, with the total life cycle cost potentially as low as 199.52 million USD. The LNG-HBD-SALAES system demonstrates optimal economic benefits when located in Guangdong Province, where the levelized cost of energy is 0.117 USD/kWh, and the net present value is 196.54 million USD. Additionally, the detailed sensitivity analysis indicates that the purchase cost of equipment and revenue from the LNG regasification process significantly impact the overall economic performance. These two parameters require focused attention in project investment. This thorough evaluation provides dependable groundwork for the initiation and execution of energy storage projects.

Abstract

Recent advancements in Large Language Models (LLMs) have reshaped human-AI interaction across various domains, yet their application in Home Energy and Comfort Management (HECM) remains in its infancy. By extracting insights from recent studies integrating GPT-based agents, semantic sensors, and LLM-driven virtual datasets, we identify both emerging opportunities and unresolved challenges. The proposed analytical framework illustrates that LLMs can act as (1) semantic interfaces translating natural language into executable energy control commands; (2) reasoning engines enabling interpretable multi-objective optimization; (3) synthetic data generators supporting privacy-aware model training; and (4) adaptive mediators fostering human-centered energy behavior understanding. However, concerns remain regarding hallucination, security, local deployment, and model validation. Future research should integrate LLM reasoning with edge intelligence, federated learning, and human-in-the-loop feedback to achieve trustworthy, personalized, and sustainable HECM systems.