Abstract Micro earthquakes with magnitudes between -6 and -2 have been observed in three successive massive injections of water at the hot dry rock geothermal energy development project's demonstration site at Fenton Hill, NM. The injections were part of a program to increase the heat transfer area of hydraulic fractures and to decrease the flow-through impedance between wells in the energy extraction system under construction. The micro earthquakes were used in mapping the location of the extended hydraulic fractures. The large shear- to compressional-wave-amplitude ratio observed in the signals suggests that the microearthquakes result from shear failure, probably on preexisting planes of weakness that inter-sect or make up the main hydraulic system. Failure resulting from increased pore pressure is a likely cause for the micro earthquakes. Knowledge of the failure criterion for the reservoir rocks permits calculation of the pore pressure increases necessary for failure. At Fenton Hill. it appears that pressures 2 Pa (0.00029 psi) above hydrostatic are necessary. For this hypothesis of microearthquake occurrence to hold, the effective reservoir permeability must be four orders of magnitude above that for the bulk rock, as would be the case if there were permeable joints or fractures in the reservoir before hydraulic stimulation. Introduction A hot dry rock geothermal energy extraction system consists of two wells connected by a flow-through fracture system constructed in relatively impermeable rock. Fracture surfaces serve as heat exchangers. Water is injected into the fracture system through one well, returned heated to the surface through a second well, cooled during power generation, and reinjected. Knowledge of fracture system location and dimension is an important element in reservoir construction and development. During construction, fractures are targeted for intersect drilling and are driven between wells. In later evaluations of reservoir capacity and lifetime, the dimensions are an important constraint in modeling efforts. This paper describes a technique for using the location of very small (local magnitude -6 to -2) micro earthquake-., that accompany hydraulic stimulation experiments in low-permeability basement rocks to locathydraulic fractures. Because these events, which are induced by high pore fluid pressures, occur in a narrow band around the hydraulic fracture, their locations can be used to infer fracture location and orientation. As an unexpected by-product from this study, we developed a technique to estimate effective reservoir permeability through use of these locations and a very simple model of the type of rock failure that causes micro earthquakes. In our experience at Fenton Hill. microseismicity (Table 1) usually is associated with massive hydraulic fracturing (MHF) experiments if the maximum injected water volume exceeds a critical volume equal to twothirds to three-fourths of the maximum amount injected into this fracture during all previous MHF experiments. After this critical volume is exceeded, the rate of seismicity rapidly increases from a very low level. Fig. 1 shows a sample histogram of event frequency vs. time for several sample experiments. Microseismic data associated with two MHF experiments and one flow test carried out in low permeability crystalline rock at the Los Alamos Natl. Laboratory, Fenton Hill, hot dry rock demonstration site were available for analysis. During the two MHF experiments, event locations were distributed in a narrow zone around an almost vertical, north-by-northwest- trending plane surrounding the hydraulic fracture. In contrast, during the flow test, event locations were scattered in a much larger, almost spherical volume. This change in the distribution of event locations is caused by transition from one- to three-dimensional (1- to 3D) fluid flow in the medium surrounding the hydraulic fracture. SPEJ P. 523^