Even in mature oil and gas provinces, unexpected subsurface complexity may challenge budgeted seismic reservoir characterization workflows to become adapted to a higher degree of customization during data preconditioning. In the process of providing a trend cube of sandstone porosity and automatic fault extraction to populate the property distribution and structural framework of a static model over the mature Emlichheim oil field, northwest Germany, many unforeseen data quality issues are encountered that necessitate rigorous well log and seismic data conditioning prior to analysis and interpretation. Specifically, insufficient noise suppression, ambiguous wireline log responses, missing curve log data, noncompliant amplitude-versus-angle gathers, and inadequate compensation of velocity anisotropy need to be addressed. These topics pose serious challenges to automatic fault extraction, seismic attribute analysis, machine learning, artificial neural network technology, the selected inversion method, Bayesian lithology prediction, and fuzzy math to transform elastic impedances into reservoir porosity. Application of multiple inversion methods generates the individual components of new earth models (sand geobodies, alternative elasticity-to-porosity transforms, etc.) that are used for advanced porosity modeling. This new information allows to update the existing static models of a mature oil field.
The Integrated Seismic Reservoir Characterization (ISRC), Study in Amboy Field of Niger Delta Oil Field – Nigeria
