| Abstract |
The bulk of the steam at the Geysers geothermal field is produced from fractures in a. relatively impermeable graywacke massif which has lieen heated by an underlying felsite intrusion. The largest of these fractures are steeply dipping right lateral strike-slip faults which are subparallel to the NW striking Clollayomi and Mercuryville faults which form t,lie NE and SW boundaries of the known reservoir. \:lierc the graywacke source rock outcrops at the surface it is highly sheared and fractured over a. wide range of scale lengths. Boreholes drilled into the reservoir rock encounter distinct ìstea.ni entriesî a.t vliicli t,he well head pressure jumps from a few to more t.lian one hundred psi. This observation that stean.is produced from a relatively small number of major fractures has persuaded some analysts to use t,he Uíarren and Root (1963) dual porosity model for rescrvoir simulation purposes. The largest fractures in this model are arranged in a regular 3-D array which partitions the reservoir into cubic ìmat,rixî blocks. The net storage and transport contribution of all t.lie smaller fractures in the reservoir are lumped into average values for the porosity and perniealiility of these matrix blocks which then feed the large fractures. Recent improvements of this model largely focus on a more accurate representation of the t.ra.nsport from matrix to fractures (e.g. Pruess et al., 1983; Ziminerman et al., 1992), but t.he basic geoinetry is rarely questioned. However, it has long been recognized that steam entries often occur in clusters separated by large intervals of unproductive rock (Thomas et al., 1981). Such clustering of fixtures at all scale lengths is one characterist,ic of self-similar distributions in which the fracture distribution is scale-independent. Recent studies of the geometry of fracture networks both in the laboratory and in tlie field are finding that such patterns are selCsiniilar and can be best described using fractal geometry. Theoretical simulations of fracture developmeii t in |