This work proposes an intelligent grey-wolf-optimizer-improved Apriori algorithm (GWO-Apriori) to mine the association rules of hidden dangers in hydrogen pipeline transmission stations. The optimal minimum support and minimum confidence are determined by GWO instead of the time-consuming trial approach. Experiments show that the average support and average confidence of association rules using GWO-Apriori increase by 29.8% and 21.3%, respectively, when compared with traditional Apriori. Overall, 59 ineffective association rules out of the total 105 rules are filtered by GWO, which dramatically improves data mining effectiveness. Moreover, 23 illogical association rules are excluded, and 12 new strong association rules ignored by the traditional Apriori are successfully mined. Compared with the inefficient and labor-intensive manual investigation, the intelligent GWO-Apriori algorithm dramatically improves pertinency and efficiency of hidden danger identification in hydrogen pipeline transmission stations.

To address the reason of valve leakage, laboratory experiments and Fluent simulations were used to study the failure mechanism. Chemical composition analysis, metallographic examination, and mechanical property tests have verified that the material of the gate valve complies with the API 6A standard specifications. The outcomes of Fluent finite element numerical simulations have disclosed that the airflow velocity in the gate area is positively correlated with the inlet airflow velocity. It increases exponentially as the degree of gate closure diminishes. As the closure degree decreases, the position where the gas impacts shifts toward the gate. The macroscopic appearance of the gate demonstrates that long-term operation in a semi-open state leads to a significant surge in gas flow velocity, resulting in the erosion of the valve body. This erosion has been identified as the root cause of the leakage incident. Additionally, the electrochemical corrosion induced by CO₂ has expedited the failure process. It is advisable to coat the valve surface with a highly wear-resistant and corrosion-resistant material and to regularly monitor the operational status of the valve during daily production. The findings of this research provide substantial guidance for improving the operational reliability and extending the service life of valves.

The interface between dispersed compound nanoprecipitates and metal substrates can act as effective hydrogen traps, impeding hydrogen diffusion and accumulation, thus mitigating the risk of hydrogen embrittlement and hydrogen-induced coating failure. In this study, we considered the precipitation of vanadium carbide (VC) and vanadium nitride (VN) nanoprecipitates on a body-centered cubic Fe (α-Fe) substrate in the Kurdjumov–Sachs (K–S) orientation relationship. To evaluate the stability and hydrogen trapping ability of the interface, we used the first-principles method to calculate the interfacial binding energy and hydrogen solution energy. The results show that the stability of the interface was related to the type and length of bonding between atoms at the interface. The interface zone and the interface-like Fe zone have the best hydrogen trapping effect. We found that hydrogen adsorption strength depends on both the Voronoi volume and the number of coordinating atoms. A larger Voronoi volume and smaller coordination number are beneficial for hydrogen capture. When a single vacancy exists around the interface region, the harder it is to form a vacancy, and the more unstable the interface becomes. In addition to the C vacancy at the Baker–Nutting relationship interface found in previous studies being a deep hydrogen trap, the Fe and V vacancies at the α-Fe/VC interface and the V and N vacancies at the α-Fe/VN interface in the K–S relationship also show deep hydrogen capture ability.

The corrosion behavior of N80 steel in a simulated coalbed–methane environment was investigated in a high-temperature, high-pressure autoclave. Uniform corrosion and maximum pit depth were quantified by weight-loss measurements and ultra-depth-of-field three-dimensional microscopy, respectively, while SEM, EDS, and XRD were used to characterize the morphology, elemental composition, and crystalline phases of the surface corrosion products. A one-factor-at-a-time design was applied at a constant total pressure of 10 MPa to evaluate the effects of nominal CO₂ partial pressure (0.10–1.00 MPa), nominal O₂ partial pressure (0–0.40 MPa), Cl⁻ concentration (3–187 g/L), and temperature (40–75 °C). Increasing the nominal CO₂ partial pressure raised the uniform corrosion rate from 0.712 to 0.930 mm/a but reduced the maximum pitting corrosion rate from 1.691 to 0.280 mm/a, while FeCO₃-containing surface coverage increased. Increasing the nominal O₂ partial pressure intensified both corrosion modes; at 0.40 MPa, the uniform and maximum pitting corrosion rates reached 1.446 and 2.202 mm/a, respectively, and the corrosion-product layer exhibited extensive cracking and spallation. Increasing the Cl⁻ concentration reduced the uniform corrosion rate from 1.078 to 0.839 mm/a but increased the maximum pitting corrosion rate from 0.474 to 1.807 mm/a, indicating a shift in the principal damage risk from average metal loss to localized penetration. The uniform corrosion rate reached a maximum of 1.516 mm/a at 60 °C, whereas the maximum pitting corrosion rate increased continuously to 2.202 mm/a at 75 °C. XRD identified Fe, FeCO₃, Fe₂O₃, Fe₃O₄, and FeOOH. The persistent Fe substrate reflections, interpreted together with the SEM observations, revealed spatially heterogeneous corrosion-product coverage. These results show that the protective contribution of FeCO₃-containing products depends on their surface coverage and visible integrity rather than on phase presence alone. The findings support stringent oxygen-ingress control, targeted pitting protection in high-salinity environments, and enhanced corrosion surveillance of intermediate- and high-temperature well sections.

Accurate and efficient prediction of thermal radiant of hydrogen jet fire is important to schedule safety design and emergency rescue program for hydrogen pipelines. In response, this paper proposes a novel Optuna-improved back propagation neural network (Optuna-BPNN) to estimate hydrogen jet flame radiation. A linear integral approach incorporating leakage rate and jet flame length is theoretically derived to establish dataset for machine learning. Then the Optuna tool is employed to optimize the initial weights and thresholds of the BP neural network. Input matrix of the Optuna-BPNN model includes pipeline diameter, leakage aperture size and hydrogen pressure. 8 sets of experimental data are employed to verify its correctness. When the abnormal data is excluded, the predicted thermal radiation of hydrogen jet fire agrees quite well with experimental results, with average and maximum deviations being 12.4% and 24.4% respectively. Using the linear integral approach, 32,670 thermal radiation data points are generated to train and test the Optuna-BPNN model. The maximum deviation between predicted and theoretical radiant heat flux for training and testing sets are only 4.5% and 6.2%, respectively. Parallel comparison trials using 6 different machine learning algorithms show that the Optuna-BPNN model gives the best mean absolute error, root mean square error and determination coefficient, which proves the effectiveness and feasibility of the developed Optuna-BPNN model in predicting thermal radiation of hydrogen pipeline jet fires.