From Well-Driven Seismic Data Processing to Seismic Reservoir Characterisation by 3D Seismic Imaging of Ordovician Karst Caves in West China

Abstract

Summary Accurate idenfication of oil bearing caves within Ordovician Karst carbonate formation utilizing conventional low-fold 3-D seismic data has been a rigorous challenge. We executed a well-driven seismic data processing workflow which includes prestack time migration and reservoir characterization by prestack simultaneous AVO (amplitude variation with offset) inversion for estimation of physical rock properties and postmigration multi-attribute analysis. Delineation of oil bearing caves is achieved by the interpretation of the inverted attributes, postmigration attributes, and attenuation factor analysis with angle stacks. As a result of this work, two highly productive wells were drilled which and successfully reached the target of oil bearing caves. Background and challenges The production is mainly from the epikarst (the uppermost weathered zone) and the caves whose scales vary from tens of centimeters to over ten meters. Drilling experience demonstrates coexistence of oil bearing, oil and water bearing, water bearing, and shale-filled caves. High drilling success rate is associated with epikarst targets while high production rate is associated with oil bearing cave targets (Fangjian Xue, 2010) in the area. High-production rate is the primary objective -for our study area. The 3-D survey includes three blocks overlapping conventional 3-D data sets during the last fifteen years with different parameters. Normal fold in the two blocks is 24 and 72 in the third block. Based on the spectral analysis of the raw data, recoverable signal band at the target level is about 8–40 Hz in one block and 6–45 Hz in the other two blocks. The legacy seismic processing datasets show poor image with very low resolution and fair signal to noise (S/N) ratio at the target interval. Additional available data include several vertical seismic profiles (VSP), check shots and complete log measurements from dozens of wells, five dipole sonic measurements for the Ordovician and the overlain partly clastics carboniferous sections, and various well test data. Various processing challenges (Wang et al., 2011) associated with poor acquisition and complex subsurface geology (Feng et al., 2011) affect the accuracy of the seismic image. Nevertheless, there are no significant lateral velocity variations above the target level that would require prestack depth migration. Main challenges in reprocessing the 3-D seismic data sets are (1) ensuring consistency of seismic amplitude, frequency and phase in merging the three survey data sets, (2) enhancing S/N ratio, (3) improving seismic image and relative amplitude preservation, and 4) improving seismic resolution. Additional challenges are AVO inversion and identification of fluid properties in the caves.