Seismic attribute analysis for fracture detection and porosity prediction: A case study from tight volcanic reservoirs, Barmer Basin, India

Abstract

The Raageshwari Deep Gas Field is situated in the southern part of the onshore Barmer Basin in Rajasthan, India. The field contains gas condensate with excellent gas quality reservoired dominantly within the tight volcanic rocks. Optimal development and production of the tight reservoirs requires characterization of faults and natural fractures that are important to fluid flow and production. Reservoir quality of the volcanic complex is governed by the distribution of matrix pore spaces and fractures. Use of seismic attributes for detecting effective reservoir pore spaces such as fractured zones and vesicles is challenging considering the limits of seismic resolution in the tight volcanic reservoirs. Poststack seismic attributes used in this study for fracture detection include geometric attributes such as coherency for detection of seismically resolved faults and reflection curvature attributes for discrimination of subseismic faults. Discrete frequency components extracted from seismic data are also used to analyze geologic discontinuities. Besides geometric attributes, seismic wave attenuation from the basic instantaneous seismic attributes — instantaneous amplitude, phase, and frequency — are also key indicators of fractured zones and vesicles that create effective pore spaces. In fractured intervals within the volcanic reservoirs, a decrease in density and slowdown in acoustic propagation velocity yields a relative drop in P-wave impedance, which are important characteristics to determine fractured zones. Detected fractured zones from seismic attributes were validated with wells, image logs, and production data. In addition to fracture detection, predicting subsurface properties, such as porosity distribution away from wells, has always been a fundamental requisite for appropriate infill well planning. Full-azimuth seismic data and selected seismic attributes were used in multiattribute analysis to predict porosity and were correlated with “blind” wells. This study captures the heterogeneous volcanic rock porosity distribution in the field, improving the understanding of varied production behavior observed within the volcanic rock complex.