Online Journal for E&P Geoscientists

Vesnaver, Aldo ^*1; Menanno, Giovanni M.²; Jervis, Michael ³
(1) Earth Sciences, KFUPM-OGS, Dhahran, Saudi Arabia. (2) Earth Sciences, King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia. (3) Expec ARC, Saudi Aramco, Dhahran, Saudi Arabia.

In global-scale seismology, the uncertainty in estimating hypocenter coordinates of major earthquakes is in the order of a few kilometers. A higher precision is not relevant for planetary studies or civil protection. When dealing with micro-earthquakes at a reservoir scale, the accuracy of the hypocenters location is very important; their pattern may highlight non-sealing faults and preferential fluid pathways, which are critical for reservoir monitoring and production optimization. At both scales, however, we cannot quantify the accuracy of our estimates experimentally, as we can for the depth estimates of formation tops, predicted from seismic surveys and validated by well logs. In this paper, we present an indirect validation using an experiment carried out on the Arabian Peninsula, where the induced micro-seismicity due to high-pressure brine injection was recorded. This analysis allows clarification of the reliability of micro-seismic data and provides clues for further improvements in hypocenter estimation. We modeled the synthetic traveltimes for hypocenter locations by mimicking a set of real ones, and compared these estimates with the known coordinates and time origin. In addition to modeling the real recording geometry, comprising both surface and well receivers, we analyzed possible alternative locations, to minimize relocation errors.

The key principle for estimating the hypocenter coordinates is minimizing the misfit between recorded and modeled traveltimes of P- and S-wave arrivals. In addition to the (xH, yH, zH) hypocenter coordinates, a time origin t0 is estimated, using Wadati’s method. This approach requires coupled P- and S-wave arrivals to be picked at a common receiver, which is not always possible. Where applicable, however, it significantly reduces the cross-talk between errors in the time origin t0 and the hypocenter depth zH. When only P-wave arrivals are available, an estimate is still possible, but in our experience, the uncertainties increase by an order of magnitude.