Imaging Reservoir Quality: Seismic Signatures of Geologic Processes

摘要

Lithofacies successions from diverse depositional environments show distinctive patterns in various rock-physics planes (velocity-porosity, velocity-density and porosity-clay). Four clear examples of decameter-scale lithofacies sequences are documented in this study: (1) Micocene fluvial deposits show an inverted-V pattern indicative of dispersed fabric, (2) a fining-upward sequence of mud-rich deep deposits shows a linear trend associated with laminated sand-clay mixtures, (3) sand-rich deposits show a pattern resulting from the scarcity of mixed lithofacies, and (4) a coarsening-upward sequence shows evidence of both dispersed and horizontally laminated mixed lithofacies, with predominating dispersed mixtures generated by bioturbation. It was observed that carbonate-cemented sandstones are extremely heterogeneous in the project deep-water study area. Those from the base of incisions are usually associated with lower shaliness, lower porosity and higher P-impedance, while from the top of flooding surfaces exhibit higher shaliness, higher porosity and lower P-impedance. One rock physics model that captures the observed impedance-porosity trend is the 'stiff-sand model'. For this model, the high-porosity end-member is unconsolidated sand whose initial porosity is a function of sorting and shaliness, while the low-porosity end-member is solid mineral. These two end points are joined with a Hashin-Shtrikman equation. A systematic variation of quartz:clay ratio from proximal to distal locations was observed in the study area even within a single facies. The quartz:clay ratio changes from [0.5:0.5] to [1:0] along the direction of flow, based on the trends of P-impedance vs. porosity as predicted by the rock model for uncemented sands. The results are in agreement with spill-and-fill sequence stratigraphic model in mini-basin setting. In addition, porosity at the distal location ({approx}25 % to 35%) is higher than the porosity at the proximal location ({approx}20 % to 23%). This trend is explained by a sequence stratigraphic model which predicts progressive increase in sorting by turbidity current along the flow, as well as, quantified by a rock model that heuristically accounts for sorting. The results can be applied to improve quantitative predication of sediment parameters from seismic impedance, away from well locations.