Characterization ambient seismic noise in an oil field
Seismic noise; surface waves; seismic interferometry; noise sources; cross-correlation
We use a 5-h long experiment with 182 vertical 2 Hz sensors deployed on the surface to characterize noise before and during a hydraulic fracking monitoring in the Potiguar Basin, NE Brazil. We observe that the seismic noise is mainly from electromagnetic inductions and from machinery vibration near the wellhead, and within 2 km from the array center from pupjacks, pipelines, roads, and industrial facilities. We also report a resonance composed of body wave coming from the treated area which is only present when the injection takes place. We interpret this resonance resulting from fluid filled fractures in the subsurface maintaining waves that reverberate in the fractures. Different strategies were employed to cross-correlate and stack the data to Ambient Noise Seismic Interferometry: classical geometrical normalized cross-correlation (CCGN), phase cross-correlation (PCC), linear stacking and phase weighted stacking (PWS). PCC and PWS are based on the instantaneous phase coherence of analytic signals. Because of the unsuitable distribution of the noise source and geometry of acquisition, spurious arrivals arise in the correlograms. We proposed a simple methodology to attenuate these unwanted effects which consists of applying Linear Moveout (LMO) correction, stack the data in the shot domains and f-k filtering. The dispersion curves after this processing are improved and the results from the phase velocity analysis are consistent with those data from the literature.