Ho Chi Minh City University of Transport

This study proposes a method to recover liquefied natural gas (LNG) cold energy from the fuel gas supply system (FGSS) of a two-stroke ME-GI dual-fuel (DF) marine engine to enhance energy utilization efficiency. LNG cold energy was employed to reduce the scavenging air temperature (SAT) through a CaCl2-based secondary refrigerant loop integrated into the engine cooling system. Thermodynamic analysis showed that approximately 12.3% of the required scavenging air cooling heat flux can be recovered at full load. Transient crank-angle-resolved CFD simulations, validated against experimental data (maximum deviation < 8%), were conducted to evaluate combustion and emission impacts under varying SAT conditions. Reducing SAT from 37 degrees C to 17 degrees C in DF mode increased indicated mean effective pressure (IMEP) by approximately 3.8%, reduced specific gas consumption by 3.7%, and significantly decreased NO emissions by up to 36.5% and soot emissions by 47.6%, while CO2 emissions decreased by 1.8%. Considering both performance enhancement and emission reduction, operating the engine in DF mode with SAT controlled at approximately 17 degrees C is recommended. The proposed system demonstrates a practical pathway for improving thermal efficiency and reducing greenhouse gas (GHG) emissions in LNG-fueled marine propulsion systems.

This study investigated the eccentric inclined load on the foundation in the three-dimensional field to deal with rectangular and square footings, which were not addressed in the two-dimensional field. With the conditions, the cohesive-frictional slope was incorporated with the finite element limit. At the top of that, utilizing machine learning to predict the ultimate bearing capacity of the foundations without using the mechanical theories. The model conducted in this paper is the Rhinopithecus Swarm Optimization Algorithm. The input parameters were explored in this paper, including slope angle, embedment depth ratio, shape ratio, setback ratio, cohesion ratio, internal friction angle, inclined load angle, and eccentric load ratio. The results indicated that the trend of the load-bearing capacity under the effect of the input data is plain. It had a trend suitable for the mechanical theories. Moreover, the failure mechanism was also very suitable; there were no scenarios exhibiting unusual behavior. In addition, the machine learning outcomes reached high accuracy with the train and test data corresponding to R2 = 99.99

This study investigates the spatio-temporal variation of stable isotopes (delta & sup1;O-8, delta & sup2;H), deuterium excess (d-excess) in rainwater and surface water in the Mekong Delta, Vietnam to identify hydrological processes and freshwater-seawater interaction. The delta & sup1;O-8 and delta & sup2;H of the rainwater samples from five sites experienced seasonal change, with an average delta & sup1;O-8 and delta & sup2;H of -5.35 parts per thousand and -35.74 parts per thousand, respectively. D-excess ranged from -13.25 parts per thousand to 26.09 parts per thousand with an average of 7.07 parts per thousand, suggesting mixing of moisture sources and post-depositional evaporation. Surface waters were generally enriched (-4.23 parts per thousand for delta & sup1;O-8; -30.16 parts per thousand for delta & sup2;H), with inland river waters the most depleted (-6.17 parts per thousand for delta & sup1;O-8) and coastal waters the most enriched (-1.60 parts per thousand to -1.87 parts per thousand for delta & sup1;O-8), due to effects of tides and evaporation. Isotopic concentrations became more positive closer to the coast, and significant and positive relationships were observed between chloride and delta & sup2;H (r = 0.74) and delta & sup1;O-8 (r = 0.60; p < 0.001). Rainwater presented the highest d-excess values, indicating a greater evaporative influence on surface waters. These observations suggest that the isotope-salinity relation may support to understand the influences of the mixing of water sources, seasonal recharge, and saltwater intrusion. The results also indicate that stable isotopes could be recommended as reliable hydrological indicators in some deltaic regions subjected to clearly distinct climatic and anthropogenic influences for analyzing hydrological processes and providing insights into water management.

Dark fermentation has garnered significant interest due to its dual benefits, including wastewater treatment and bio-H2 production. However, due to its complexity and nonlinearity, the process is difficult to model through conventional approaches. Data-based machine learning (ML) methods can handle this challenge but suffer from a lack of generalization and a black-box nature. This study employs Bayesian optimization and Shapley Additive exPlanations (SHAP) to overcome these challenges. Bayesian optimization was used for training hyperparameters, fine-tuning, and obtaining improved results from Artificial Neural Network (ANN) and Extreme Gradient Boosting (XGBoost) in forecasting bio-H2 production from wastewater through dark fermentation. Model development incorporated key operational factors such as biomass concentration, Ni, Fe, chemical oxygen demand, pH, hydraulic retention time, acetate, ethanol, butyrate, and the acetate-to-butyrate ratio. Among the tested models, the Bayesian-optimized XGBoost model achieves a low Mean Squared Error (MSE) of 0.009 and 0.001 for training and testing phases, respectively, along with a strong Coefficient of Determination (R2) of 0.9875 for training and 0.9812 for testing, demonstrating a strong correlation between predicted and observed values, indicating that the created model is reliable in forecasting bio-H2 production. Furthermore, SHAP analysis was employed to determine the relative influence of input parameters on bio-H2 yield, identifying Ni, butyrate, and ethanol as the top three most influential factors. These findings offer valuable insights for enhancing the efficiency of bio-H2 production through dark fermentation, contributing to the advancement of sustainable biofuel technologies.

The growing concerns about greenhouse gas emissions and air pollution from maritime transport have led to increasing interest in researching cleaner and more sustainable fuel options. Thus, this work presented the feasibility, limitations, and potential benefits of using methanol as a sustainable alternative fuel for marine engines. This work examined various methanol production and application aspects, such as production processes, needs, infrastructure and availability, engine performance and emission characteristics, and cost. In the first stage, this work highlighted the importance of methanol in ocean shipping to achieve decarbonization goals and assessed the infrastructural availability for supplying methanol to ships. In the next stage, the methanol production process's input sources and critical characteristics were evaluated entirely. Also, the methanol properties and applications in marine engines under various strategies were comprehensively analyzed. In the third stage, the methanol cost for different production approaches and applications was scrutinized. The problems and possibilities for bunkering and storage facilities when using methanol for maritime engines were also thoroughly analyzed. Finally, the challenges and solutions for the methanol application for marine engines were critically presented. Overall, the present work provided a comprehensive assessment of the potential role of methanol in the maritime sector, aiming to establish sustainable maritime practices. More importantly, this work intends to inform policymakers, academics, and industry stakeholders about the prospects and challenges of using methanol as an alternative fuel for marine engines with a view to the decarbonization strategy and Sustainable Development Goals of the maritime sector.