The drilling process is complex, especially for ultra-deep wells, which face the problems of high temperature, high pressure and poor drilling resistance in their formation. In order to establish an ROP (the Rate of Penetration) prediction model for ultra-deep wells, the characteristics of ultra-deep well drilling operations, such as formation temperature and formation pressure, are fully considered in the process of parameter optimization. Combined with the drilling mechanism and mutual information correlation coefficient, the final input parameters are determined. The powerful nonlinear fitting ability of the artificial intelligence method is very suitable for predicting the ROP. Considering the time sequence of multi-source data, this paper combines the powerful timing information-based mining ability of the LSTM (Long Short-Term Memory Neural Network) with the nonlinear fitting ability of FNN (Fully Connected Neural Network), and establishes an intelligent prediction model of the ROP based on a LSTM–FNN. The results show that the average relative error and $R^2$ of the LSTM–FNN model on the data of well 1 and well 2 are better than the FNN and LSTM models. In addition, the accuracy of the LSTM–FNN model on the data of adjacent wells is reduced by only 5%, which further verifies the good mobility of the model.

Periodic changes in wellbore temperature and pressure caused by the cyclic injecting and producing of gas storage wells affect wellbore integrity. To explore the distribution and influencing factors of wellbore temperature and pressure during gas storage well injection-production processes, based on energy conservation, momentum theorem, and the transient heat transfer mechanism of the wellbore, a temperature and pressure coupling model for gas storage injection-production wellbores was established, and a piecewise iterative method was used to solve the model equations. Compared with the field data, the predicted relative errors of the wellhead temperature and pressure were 2.30% and 2.07%, respectively, indicating that the coupling model has a high predictive accuracy. The influences of the injection-production conditions, tubing diameter, and overall heat transfer coefficient on the wellbore temperature and pressure distributions were analyzed through an example. When the gas injection flow rate increased by 1.5 times, the bottomhole temperature decreased by 37%. Doubling the overall heat transfer coefficient resulted in a 10% rise in the bottomhole temperature. An increase of 0.3 times in the gas injection pressure led to a 31% increase in bottomhole pressure. With a 1.5-fold increase in the gas production flow rate, the wellhead temperature rose by 28%, and the wellhead pressure dropped by 20%. The research in this paper can serve as a guide for the optimization design and safe operation of gas storage wells.

To address the insufficient long-term stability of kill fluids in ultra-deep, ultra-high-temperature wells in the Tarim Oilfield, this study systematically evaluates the adaptability of high-density kill fluids under high-temperature and prolonged static aging conditions, with a focus on identifying dominant settling mechanisms. The correlation between the microstructure and macroscopic properties of kill fluids was elucidated through particle size distribution analysis, scanning electron microscopy (SEM), Fourier-transform infrared spectroscopy (FTIR), and rheological characterization. A quantitative grading criterion for settling stability was established using settlement values and the falling rod method. Key findings demonstrate that low-density kill fluids (1.4–1.6 g/cm³) retained rheological stability after 20 days of aging at 220 °C, fulfilling the ≥20-day operational requirements for ultra-deep well testing. In contrast, high-density systems (1.9 g/cm³) exhibited severe particle aggregation after 15 days under identical conditions, with the yield stress-to-plastic viscosity ratio dropping below 0.10 and suspension capacity deteriorating. The apparent viscosity of ultrafine barite-weighted kill fluid increases with temperature, and its settling value is positively correlated with aging time and temperature. The settling mechanism of ultrafine barite-based kill fluids was attributed to reduced surface charge density caused by the decarboxylation of polyacrylate dispersants, which diminished interparticle electrostatic repulsion. The developed “settlement value vs. falling rod time” correlation model and grading criteria lay a theoretical foundation for optimizing kill fluid formulations and evaluating field performance in ultra-high-temperature wells, offering critical engineering insights to ensure safe deep hydrocarbon testing operations.

Determining the representative elementary volume (REV) of tight sandstone remains challenging because CT resolution and field of view impose competing constraints on pore-structure characterization. In this study, six tight sandstone samples were examined using multi-resolution micro-CT with voxel sizes of 0.556, 3.98, and 10.5 μm. A unified stability-index method was used to evaluate geometric, topological, and hydraulic parameters, including porosity, fractal dimension, anisotropy, pore-size distribution, specific surface area, coordination number, sphericity, tortuosity, and permeability. The results show that both CT resolution and field of view influence REV estimates. High-resolution images preserve more micropores, narrow throats, and irregular boundaries, whereas low-resolution images lose fine-pore information but better capture larger-scale structural heterogeneity. Hydraulic parameters generally require larger REVs. Porosity and coordination number show scale-dependent trends similar to those of the hydraulic parameters, indicating that pore-space abundance and network connectivity jointly influence the representative scale of flow. Pore-size distribution and anisotropy provide more reliable REV estimates at high resolution, while fractal dimension, specific surface area, and sphericity are more sensitive to changes in resolution. Therefore, flow-related REV evaluation in tight sandstone should account for resolution limitations and integrate multiple geometric, topological, and hydraulic parameters rather than relying solely on porosity.