Transport and frictional properties of core samples from Taiwan Chelungpu-fault Drilling Project and its association with the heat generation due to frictional heating

Taiwan Chelungpu-fault Drilling Project (TCDP) was started from 2002 to investigate the faulting mechanism of the 1999 Chi-Chi earthquake. TCDP was succeeded in penetrating the Chelungpu fault and recovered core samples from two holes, Hole A and Hole B. In Hole B, three fault zones, FZB1136 (1134-1137 m), FZB1194 (1194-1197 m), and FZB1243 (1242-1244 m), were recognized in the core samples (Hirono et al., 2006). Micro- textual observation and rock magnetic analyses of fault zones implied the evidence of heat generation, though the temperature did not reach the melting point. Borehole temperature measurement in Hole A observed the very low temperature anomaly around the fault zone (Kano et al., 2006). These results suggest the low degree of the frictional heating due to very low friction during the slip. The possible low friction might be explained by the slip within clay rich fault gouge with low shear strength. The other possible mechanisms are dynamic weakening behaviors of the fault zone, such as thermal pressurization and elast-hydrodynamic lubrication. To demonstrate the assumptions, the transport properties and the strength of the fault rocks are measured using core samples. Core samples of three fault zones in Hole B (FZB1136, FZB1194, and FZB1243) are selected for our laboratory experiments. Permeability and specific storage for fault rocks were measured under high confining pressure up to 100 MPa. Nitrogen gas was used as a pore fluid, and gas permeability was transformed to water permeability from gas permeability dependence on pore pressure of Klinkenberg equation. In FZB1136, permeability for fault breccia showed around 10-16 m2 at 1km depth which is similar value to that for host rock of siltstone and fracture rocks. In FZB1194, permeability of black fault gouge was about 10^{-15} m2, is larger than surrounding rocks. Frictional tests were also conducted using fault gouge samples with less than 100 μm of grain size. Tests are performed under the condition of 24 MPa of confining pressure, 10 MPa of pore fluid pressure using distilled water, and 50°C. Gouge samples (0.500g) were set between Arminius dummy samples to reproduce the slip plane, and gouges were sheared at the maximum displacement of 1.1 mm under 0.1 μm/sec of slip velocity. Difference in the shear strength are identified between black gouge, gray gouge and gouge within fault breccia zones, though all frictional coefficients are so large that we can not explain the low temperature anomaly of the fault zone. Therefore dramatic slip weakening might have occurred during the slip event. Mathematical approaches to evaluate the temperature rise history combined with our experimental data are necessary for further studies.