Anisotropic AVO: Implications for reservoir characterization

Abstract

Routinely applied methods in seismic reservoir characterization such as forward modeling, wavelet extraction, amplitude-variation-with-offset (AVO) analysis, AVO inversion, and interpretation of seismic data usually assume that the earth can be modeled by a stack of isotropic layers. This assumption may cause significant problems where there are nonnegligible differences in the anisotropic parameters between the various lithologies that cause vertical profiles of VP/VS and the anisotropy parameters to be dissimilar. In this case, a significant mismatch between the seismic data and the isotropic synthetic seismogram AVO response will occur, making far-angle stack interpretation difficult. In some cases, the mismatch might be misinterpreted as a data quality issue. In an offshore Western Australia field, three lithofacies (volcanic rock, sandstone, and shale) need to be correctly identified for detailed reservoir characterization. Here, the AVO response of the actual seismic data is significantly affected by velocity anisotropy. Originally, it was thought that the far-angle stack could be used to detect volcanic rock in the field; however, after accounting for the velocity anisotropy effect, it was found that the far-angle stack enables us to identify sandstone. A proper understanding of the anisotropy effect allows the interpreter to use seismic data more effectively, which leads to a more robust estimation of the distribution of lithofacies in the target area.