Several analyses have been published in the specialized literature on the use and performance of distinct prestack seismic attributes as direct hydrocarbon indicators (DHI). The main idea is to use combinations of those attributes, in particular the P and S impedances (I P and I S ), to minimize the dry rock frame contribution, that cause the increase of the pore fluid discrimination capacity.The Batzle proposal (2001) refers to the calculation of K-G (where K and G represent the bulk modulus and the shear modulus, respectively). Several publications of Goodway et al. since 1997 have systematically presented the advantages of the parameter lambda*rho (Lamé parameter lambda λ multiplied by the density ρ as an excellent fluid indicator).More recently, a tutorial by Russell et al. (2003) presents the attribute I P 2 -CI S 2 with C being a function of local (V P /V S ) 2 where V P and V S are the P-and S-velocities for dry rock condition. As discussed by Russell et al., the Goodway (1997) and Hedlin (2000) proposals may be considered as particular cases of the I P 2 -CI S 2 attribute. In this study we examine the performance of several DHI attributes in the various scales involved in exploration geophysics, from rock samples to seismic.A total of 592 rock samples from Tertiary (188) and Cretaceous (404) siliciclastic reservoirs of Brazilian offshore fields were analyzed. The 3D seismic data set used in this study illustrates examples of siliciclastic oil-bearing Tertiary and Cretaceous intervals from offshore Brazil.The Cretaceous reservoirs are adjacent to igneous rocks with very high impedance properties, creating a greater challenge in the fluid identification process. Because of these high contrast zones, the use of fluid indicator attributes derived from conventional AVO weighted stacking methods is not appropriate. In order to prevent the pitfalls due to wavelet effects, the seismic attribute analysis was mainly based on the attributes generated from the inverted AVO data, particularly from the impedances (I P and I S ) derived from a simultaneous elastic inversion process. A 2D section containing three wells that crosses both the Tertiary and Cretaceous reservoirs was chosen to illustrate the attributes analysis on the seismic scale ( Figure 1).
Rock samples, logs, and seismic data analysis.From the Gassmann equation and laboratory measurements of V P , V S , and ρ (dry, under real subsurface effective pressure), the elastic behavior of the samples was calculated according to two scenarios: 100% brine and 80% oil with 20% brine.Tables 1a and 1b summarize the results for each attribute analyzed, including I P -I S , λρ, Poisson's ratio, V P /V S , K-G, I P , fluid factor, and I P 2 -CI S 2 calculated for five distinct C values. We have also separated the attribute evaluations for two distinct sample groups, Cretaceous and Tertiary sandstones.These tables show the mean and standard deviation of the attribute value related to each scenario analyzed and also a fluid indicator coefficient which diagnoses the sen...