In many geological situations, there is a need to monitor fluid migration and associated phenomena. Acoustic emissions (AE) that result from fluid pressure diffusion from a borehole can potentially provide useful information about migration paths, local stress states and local strength distributions. However, AE may also result from regional differential stress as well as mechanical stress concentrations associated with the injection borehole. To attempt to differentiate between these possible AE generating situations, we performed several experiments in unjointed granite rock samples containing a borehole. Fluid diffusion into and out of the borehole was achieved, under hydrostatic to high deviatoric far-field stress states, while concurrently recording digitized waveforms of AE at 20 transducer locations and measuring acoustic velocity changes. Over 2000 AE events were located with a precision better than 2 mm, and focal mechanisms were obtained for the majority of them. Most acoustic emission events could be modeled by a shear displacement source mechanism giving rise to a quadrature polarity pattern usually associated with a double couple.

A significant fraction of the events, however, were not so simply modeled and may indicate a combination of extension and shear displacement at the source. Very few pure dilatational or compressional events occurred. Results indicate that diffusing fluids, at pressures less than the local tensite strength, produce useful accompanying AE only in the presence of a large differential stress state surrounding the borehole. It is difficult, however, to distinguish events produced solely as a result of fluid pressure diffusion from those resulting from regionally applied shear stresses. A high-pressure fluid leak at a metal-rock interface was easily detected using AE, but mapping of isobaric pore pressure fronts in the presence of background seismicity was less successful. A model is developed to explain the difficulty. The spatial patterns of the acoustic emission locations appeared to be fractal and might be used to infer information about the magnitude and spatial distributions of local strength in the rock. ¿ American Geophysical Union 1990