Small-spacing twin-well (SSTW) salt caverns have an extensive application prospect in thin or bedded rock salt formations due to their good performance, while they are rarely used in ultra-deep formations. The target strata depth of Pingdingshan salt mine is over 1700 m, and it is planned to apply an SSTW cavern to construct the underground gas storage (UGS). A 3D geomechanical model considering the viscoelastic plasticity of the rock mass is introduced into Flac^3D to numerically study the influence of internal gas pressure, cavern upper shape and well spacing on the stability of an SSTW salt cavern for Pingdingshan UGS. A set of assessment indices is summarized for the stability of gas storage. The results show that the minimum internal gas pressure is no less than 14 MPa, and the cavern should not be operated under constant low gas pressure for a long time. The cavern with an upper height of 70 m is recommended for Pingdingshan gas storage based on the safety evaluation and maximum volume. The well spacing has a limited influence on the stability of the salt cavern in view of the volume shrinkage and safety factor. Among the values of 10 m, 20 m and 30 m, the well spacing of 20 m is recommended for Pingdingshan gas storage. In addition, when the cavern groups are constructed, the pillar width on the short axis should be larger than that on the long axis due to its greater deformation in this direction. This study provides a design reference for the construction of salt cavern gas storage in ultra-deep formations with the technology of SSTW.

Wellbore instability is one of the most serious issues faced in the drilling process. During drilling operations, the cyclic loads applied on the fractured formation progressively modify the initial parameters (i.e., length and width) of the fractured formation, thus resulting into undesirable wellbore instability. In this paper, using a nonlinear finite element software (ABAQUS) as the numerical simulator, a poro-elasto-plastic model has been established which aimed at analyzing the influence of drill string vibration cyclic loads on the development of the wellbore natural fracture. The numerical results showed that the fracture width as a function of time profiles followed a sinusoidal behavior similar to the drill string vibration cyclic load profiles. For different cyclic load magnitudes with constant number of cyclic loads, the highest percentage increase of the fracture width after integration of cyclic loads was 64.77%. Interestingly, the fracture width increased with the fracture length in the near wellbore region while it globally decreased in the region far away from the wellbore. But for constant cyclic load magnitude with different number of cyclic loads, the biggest percentage increase of the fracture width after integration of cyclic loads was slightly lower representing 63.12% while the oscillating period of the fracture width increased with the number of cyclic loads. The parametric study revealed that the drill string vibration cyclic loads were relatively independent of the fracture length and the bottom hole pressure. However, the fracture width and the loss circulation rates were considerably impacted by the drill string vibration and the highest percentage increase of the loss circulation rate after integration of cyclic loads was 14.3%. This study provides an insight into the coupling of the fracture rock development and the continuous cyclic loads generated by drill string vibrations which is an aspect that has been rarely discussed in the literature.

Summary: Artificial intelligence (AI) is revolutionizing several businesses across the world, and its implementation in drilling engineering has enhanced the performance of oil and gas companies. This paper reviews and analyzes the successful application of AI techniques to predict wellbore instabilities during drilling operations. First, a summary of the implementation of AI for the prediction of loss circulation, pipe stuck, and mud window is highlighted. Then, the recent innovations and challenges of the AI adoption in major drilling companies is presented. Finally, recommendations are provided to improve the integration of AI in the drilling industry. This analysis gives deep insight into the main publications and recent advances of the application of AI in drilling engineering and is expected to contribute to the further development of the drilling industry.