N₂ injection into coal seams can effectively enhance the gas flow capacity in the late stage of pumping, thereby improving the recovery rate and recovery efficiency of low coalbed methane (CBM). To reveal the thermodynamic flow coupling relationship between the reservoir and the gas phase and its transportation mechanism in the process of thermal N₂ injection, a mathematical coupling model of N₂ injection that considers the deformation of the coal seam, fluid transportation, and temperature change was established. The influence of the seepage heat transfer effect of the coal seam under the effect of N₂ injection on the CH₄ extraction rate was investigated using this model. Results indicate that the action mechanism of N₂ injection in coal seams includes increasing seepage, promoting flow, and displacing gases. A higher initial coal seam temperature results in a smaller thermal expansion and deformation of the coal skeleton during thermal N₂ injection and less pronounced coal permeability increase. A larger initial coal seam permeability results in more favorable N₂ diffusion, which strengthens the flow-promoting effect on CH₄. The effect of gas injection pressure on CH₄ recovery is greater than that of the gas injection temperature: High pressure promotes N₂ seepage, carries CH₄ flow, and increases the CH₄ diffusion effect, whereas higher temperature promotes the desorption of adsorbed gas in the coal seam and improves the recovery rate.

The premixed abrasive jet possesses a strong strike ability and is widely used in oil and gas exploitation, machining, rust removal, and other fields. The superstrong, forceful impact of the premixed abrasive jet is mainly provided by high-speed abrasive groups. Hence, the abrasive velocity is the basis of this research, by applying the distribution law of abrasive impact force. In this paper, the particle velocity of the premixed abrasive jet is analyzed theoretically, and the corresponding particle velocity model is established. The real-time contrast interpolation method is employed to solve the problem of the variable drag coefficient. Factors such as the nozzle structure, average abrasive diameter, abrasive density, and jet flow are utilized to determine the abrasive velocity of the nozzle outlet. The numerical solution for the abrasive velocity is obtained by dividing the high-pressure pipe and nozzle into several sections, along the axis. Finally, the calculated particle velocity is compared with the particle image velocity measurement (PIV), to verify the correctness of the established model. These results demonstrate that the model calculation is in effective agreement with the experimental results. The deviation between the theoretical value and the experimental mean is 0.18 m/s. The standard deviation of the experimental results is 3.81-4.22 m/s, while the average error is less than 4%.

To solve the serious filtering problems of fracturing fluid during active hydraulic fracturing of coalbed methane, which blocks gas migration channels with residual fracturing fluid and results in a rapid decline in coalbed methane production, several preliminary experiment of thermal nitrogen injection was carried out. To explain the process and mechanism of thermal nitrogen injection at the nanoscale, an adsorption and diffusion model was established to study the adsorption and diffusion of methane, water and fracturing fluid during thermal nitrogen injection. The results showed that the self-diffusion coefficient of methane was increased by 23%, 45% and 32%, the self-diffusion coefficient of water molecule was increased by 108%, 185% and 138%, the self-diffusion coefficient of fracturing fluid molecule was increased by 169%, 275% and 181%, and the total number of atoms in the slit hole was decreased by 15%, 25% and 37%, respectively, when hot nitrogen was injected at 2, 4 and 6 Mpa, respectively. This shows that the numbers of methane, water and fracturing fluid molecules in the slit hole were significantly decreased by the thermal and displacement effects of hot nitrogen injection. The self-diffusion coefficient was significantly improved, and the effects of plug removal was remarkable. GRAPHICAL ABSTRACT

Nano-clean fracturing fluids have broad application potential in coalbed methane reservoir fracturing owing to their high stability, good temperature resistance, low filtration loss, and strong frictional resistance reduction. However, the sand-carrying regularity of nano-clean fracturing fluids in coalbed methane reservoirs is unclear, especially for complex fractures with variable directions. This study established a sand transport model that considers proppant collision, wall friction blocking, fracture fluid filtration loss, and the fracture branching angle to study the sand-carrying law of nano-clean fracturing fluids and its influencing factors in complex fractures. The degrees of influence on equilibrium height and placement rate from high to low were the proppant particle size, proppant density, fracturing fluid properties, sand ratio, and pumping discharge volume, and the correlation degrees obtained by grey correlation analysis are 0.862, 0.861, 0.855, 0.854, and 0.832, respectively. As the complexity of the fractures deepens and the resistance increases, the flow rate of the fracturing fluid is reduced, making it difficult for the proppant to enter the branching joints. The sand-carrying performance of a nano-clean fracturing fluid is better than that of a common clear-water fracturing fluid. The fluid-structure coupling model of a nano-clean fracturing fluid can accurately characterize the sand-carrying law of nano-clean fracturing fluids, providing a research basis for optimizing high efficiency sand-carrying fracturing fluid parameters in coalbed methane reservoir fracturing construction.

Accurate prediction of retaining-pile displacement is important for deformation control during staged deep-excavation construction, where monitoring series often exhibit pronounced nonstationarity and multiscale temporal variation. To account for the distinct temporal characteristics of the trend and fluctuation components, this study proposes a CEEMDAN–SVR–PSO-LSTM hybrid framework, termed CSPL. Complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) is first used to decompose the monitored displacement series into a slowly varying residual and oscillatory intrinsic mode functions (IMFs). Support vector regression (SVR) is employed to predict the trend component, whereas particle swarm optimization (PSO)-optimized long short-term memory (LSTM) is used to predict the fluctuation components. The component-wise predictions are then reconstructed to obtain the final displacement prediction. The proposed model is validated using monitoring data from two Zhengzhou Metro projects. For Case 1, the model achieves average R², RMSE, and MAPE values of approximately 0.94, 0.33 mm, and 2.8%, respectively, across different monitoring depths, showing better overall predictive performance than BP, EMD-LSTM, and VMD-GRU. For Case 2, external validation using 90 depth-wise monitoring points over six construction stages yields millimeter-level errors, supporting stable predictive performance under different geological and support-system conditions. The results further indicate that the late construction stages and the upper pile segments deserve particular attention in deformation control. The proposed model provides a data-driven tool for construction-stage displacement prediction and deformation control in underground geotechnical engineering.