Shale formations are the main source of borehole stability problems during drilling operations. Suboptimal predictions of borehole failure may partly be caused by neglecting the anisotropic nature of shales: Conventional wellbore stability analysis is based on borehole stresses computed from isotropic linear elasticity (Kirsch solution) with the assumption of no induced pore pressure. This is very convenient for a practical implementation but does not always work for shales. Here, anisotropic wellbore stability analysis was performed targeting an offshore gas field to investigate in particular the impact of elastic anisotropy on borehole failure predictions. Stress concentration around a circular borehole in anisotropic shale was calculated by the Amadei solutions, and induced pore pressure was obtained from the Skempton parameters based on anisotropic poroelasticity. Borehole failure regions and modes were then predicted using the effective stresses and those are apparently consistent with observations. A comparison with the conventional approach suggests the importance of accounting for elastic anisotropy: Predicted failure regions, modes, and also the associated mud weight limits can be completely different. This observation may have significant implications for other fields since shale often show strong elastic anisotropy.

When underground piping for a water supply or gas supply is buried, some obstacles such as unknown pipes or stones can be encountered. To avoid such situations, an examination of the underground condition from the ground surface is needed. For such an examination, the ultra-shallow reflection method is used. We use a giant-magnetostriction vibrator as the seismic source. In this study, first, the P-wave directivity of the giant-magnetostriction vibrator is clarified. Second, the experiment of detecting a buried concrete block is conducted. In this experiment, the cross-correlation analysis is applied. However, the arrival time of the wave reflected from the concrete block is not confirmed. On the basis of the P-wave directivity of the giantmagnetostriction vibrator, the causes are assumed. To obtain good results, the magnified cross-correlation analysis is proposed. By this analysis, the depth at which the concrete block is buried can be estimated with high accuracy.

Velocity-model building with a good understanding of anisotropy is one of the important elements in a prestack depth migration (PSDM) project. During a recently conducted PSDM project, we observed that (1) the application of negative delta/epsilon is necessary to achieve reasonable depthing and gather flatness, and (2) azimuthal variation in gather flatness exists in the interval where negative delta/epsilon are necessary. A geologic explanation of these interesting observations was necessary to justify the velocity-model-building result. Investigation of stress magnitude/orientation and theoretical anisotropy of unconsolidated sandstone under nonhydrostatic stress revealed that these observations can be explained by stress-induced anisotropy. This concept may be applicable to other fields, and the existence of azimuthal anisotropy and value of anisotropy parameters can be roughly estimated by the method described.

Core velocity measurements are an essential part of any 4D seismic feasibility study. During recently conducted core velocity measurements, we found some interesting results regarding velocity anisotropy and hysteresis. These findings include: (1) the stress sensitivity of velocity varies depending on the propagation direction, (2) velocities measured during loading have a significantly larger stress sensitivity than those measured during unloading, and (3) horizontal effective stress has a noticeable impact on velocity anisotropy. We conducted rock physics analysis and 1D seismic forward modeling, incorporating velocity anisotropy, and found that the estimated 4D seismic signal is largely affected by velocity anisotropy and hysteresis. These findings suggest the importance of considering the velocity measurement direction and the nature of the stress change to obtain a realistic 4D seismic signal. Neglecting these considerations may lead to a significantly underestimated or overestimated modeled seismic response.