Feasibility was examined for estimating in-situ stresses by using the so-called drilling-induced tensile fracture (DTF) of a borehole. DTF is macroscopically a longitudinal crack consisting of many small parallel cracks which are oblique to the borehole axis, and is characterized by the circumferential position (θmN) along the borehole wall and the inclination (γm) of the small cracks with respect to the borehole axis. First, we studied the stress field induced by the stress concentration on the borehole wall due to a far stress field and examined the effect of factors such as the far stress field, the tensile strength of a surrounding rock and the Poisson's ratio, on the characteristics (θmN, γm) of DTF. Then we discussed the feasibility of estimating in-situ stress by using a set of data of θmN and γm observed along a deviated borehole. Finally, we attempted to apply the basic concept of estimating the stress field by using θmN and γm to well TG-2 which was drilled by NEDO (New Energy and Industrial Technology Development Organization) in the Yunomori geothermal field. The application was fairly successful.
In-situ heat extraction tests were carried out at Toyoha mine. An artificial fracture was made by hydraulic fracturing and used as a heat exchange plane. An initial direction of the artificial fracture was controlled by disk-shaped slot which had been created by abrasive water jet. After hydraulic fracturing test, circulation tests were conducted twice. Two circulation tests named 93b and 94a consist of one injection borehole and four, seven production boreholes, respectively. Although seven production horeholes had been drilled, water was produced mainly from only one production borehole. In experiment 93b and 94a, water recovery were 44 and 35%, respectively. After circulation tests, the rock mass around boreholes was blasted and removed to observe the fractures. From the observation, the artificial fracture initiated from the tip of the disk-shaped slot was estimated to be connected to two natural fractures. On the basis of the test results and observation result, a fracture model including artificial fracture and natural fractures in a rock mass was constructed. The constructed fracture model was simplified to satisfy the connectivity between boreholes and fractures and a numerical simulation using FEHM (Finite Element Heat and Mass Transfer Code) was carried out. After trial and error, permeabilities of fractures were decided to match the observed temperature drawdown during the circulation test 93b. A close agreement between observed and calculated values was obtained. Using the parameters which gave a best fit to the circulation test 93b, numerical simulation for the circulation test 94a was also conducted. There was an apparent difference between observed and calculated temperature drawdown. The reason for this discrepancy might be that the code could not simulate convectional phenomena within the borehole. There was another possibility that the permeability of a fracture had been increased after experiments 93b. By using a higher fracture permeability, a temperature drawdown for the circulation test 94a was re-calculated and a good agreement between calculated and measured values was obtained. This re-calculation supports the assumption that the fracture permeability had increased after experi
Dominant fluid flow paths in geothermal reservoirs are deformable due to tectonic stresses and fluid pressure. A numerical method for analysing tracer response is proposed by taking account of deformation of flow path, fluid flow and advective diffusion of tracer concentration. The dominant fluid flow path is supposed as a planar horizontal plane and two wells (the injection and production wells) are connected to the planar flow path. The fluid flow is governed by two equations. One is the equation of continuity which includes two unknowns, i.e. fluid pressure and asperity height. The other is the equation concerning with the asperity height derived from the Hertzian theory of contact. These coupled equations are simultaneously solved by means of a finite difference method. Using the results of fluid flow analysis in the dominant fluid flow path, tracer concentration analysis is made to include the effects of diffusion and adsorption to rock. Total Variation Diminishing (TVD) scheme is applied to solve numerically the equation of tracer concentration. As numerical examples the effects of fluid pressure at the injection well, injection process and diffusion coefficient on the tracer response are discussed.