A Network Design Study for Optimum Resolution of Reservoir Parameters Based on a Dense Seismic Array at The Geysers Geothermal Reservoir, CA.

The goal of this paper is to conduct a seismic network design study to appraise the best placement of seismic stations for optimum resolution of estimates of reservoir parameters at The Geysers geothermal reservoir in northern California, USA. The design study employs 100 sensors and analyzes the resulting resolution in elastic parameters such as P- and S-wave velocity estimates. The study is based on a one-year period of previously acquired microearthquake locations within a 5 km by 5 km study area, 3D P- and S-wave velocity models, and surface topography. A one-year distribution of seismicity is important, because the number of earthquakes depends on the volume of injected water, which is a function of seasonal availability. Over 3,000 earthquakes were selected, which are densely distributed and cluster in various locations of the region. Based on the hypocenter locations, the 3D velocity models and the various network designs of seismic stations in the study area, synthetic P- and S-wave travel times were computed using an eikonal solver. The algorithms used to estimate model parameter resolution of our seismic velocity estimates were based on (a) methods for computing singular values of large sparse systems values with an iterative Lanczos scheme to obtain quantitative resolution and (b) by 3D inversion for earthquake hypocenters and velocity structure based on absolute and double difference travel times to obtain qualitative resolution estimates. In our paper we present the effects of various network topologies including regular and irregular shaped network geometries with a varying number of sensors and incorporate the influences of surface topography on the resolution of the reservoir parameters.