Well based modeling and seismic data analysis were used to investigate the potential of Amplitude Variation with Angle (AVA) for fluid discrimination in a high porosity carbonate reservoir in a producing UAE oil field. Gassmann fluid substitution was used to model well log data, which included compressional and shear sonic logs and density logs to produce synthetic well logs representing the reservoir at 100% fluid saturations of brine, oil and gas at reservoir pressure and temperature conditions. The average VP/VS ratio for brine saturated reservoir (~2.0) was observed to be higher than both the oil (~1.7) and gas (~1.6) saturated reservoir cases.
The modeled brine, oil and gas logs were used to calculate the AVA responses at the top and base of a thick, 25-35 % porosity reservoir layer using the Zoeppritz equations. The responses for all three fluids were found to be a Class-IV type AVO anomaly. Seismic amplitude variation on the synthetic CDP gathers was successful at discriminating brine from hydrocarbon but could not differentiate oil from gas. These encouraging results on synthetic data suggest that with good quality seismic data it can be possible to see a difference in AVA responses between brine and hydrocarbon filled porous reservoirs. An AVA study was performed using available relative amplitude CDP gathers along a 2D seismic line extracted from a 3D seismic volume. The results were able to discriminate between areas saturated with brine from those with hydrocarbon. The real seismic results were in good agreement with synthetic model results. The angle stack analysis was successful at improving the signal-to-noise ratio in lower fold seismic data and reduced the input data size requirement when dealing with larger CMP gathers. The effects of varying key reservoir and seismic properties on AVA response were examined to help understand the potential for misinterpretation.