| Popis: |
The studied lower Cretaceous carbonate ramp deposits are heterogeneous with pervasive diagenetic processes leading to complex pore network and rock texture. Before this study, field development was based on log based water saturation modeling. This saturation modeling was dependent on the hydraulic flow units and porosity classes, which was meant to be, but was not explicitly representative of the defined geologic facies. The assignment of the relative permeability data is also very challenging in the absence of proper Rock Type model. Often in carbonate reservoirs, there is no direct or linear relationship between Reservoir Rock Types (RRT), sedimentological facies assemblages and water saturation distribution across the field. Hence, accurate integration of the sedimentological, diagenetic, depositional environment information and petrophysical properties is essential in building a robust RRT model. This RRT model can then explain the rock-fluid interaction through a reliable saturation height model. In the first part, the paper illustrates the workflow which involves integrating lithofacies, depositional packages, degree of cementation, pore-type from sedimentology and relating resultant pore-typing to core porosity–permeability data. This workflow resulted in a strong reservoir rock typing scheme, which was key in building a robust saturation height model. Precisely, this was achieved by assigning "most-of lithofacies and diagenetic indicators" to each rock type defined. In the second part, a Variable Saturation Height Function (VSHF) was developed using mercury injection capillary pressure (MICP) data. The function was made variable with depth by bringing one or both of the reservoir parameters (Phi and K) into the equation. Most importantly, VSHF explicitly scans the lower and upper Sw boundary of a particular rock type and helps in removing the skewness in the Sw difference histogram between model and log Sw. One of the important steps in the workflow was to normalize the MICP data to log derived Sw values, provided the confidence on the Sw calculation from logs is high. After stress and closure correction, the normalization of the reservoir Pc data was achieved through an independent correction factor. Both of the workflows (rock typing and saturation height modeling) were built based on data from 30 cored wells. The workflows were tested on 15 cored wells and more than 600 non-cored wells. Rock type maps were found to be more correlatable with reservoir quality maps than lithofacies maps alone. This is a result of the diagenetic processes undergone by the rock during and after deposition modifying the original depositional controlled pore architecture. With this approach the water saturation distribution was more consistent with logs and core derived Sw data. The workflows shown in this paper are reliable and can be extended to other carbonate fields’reservoir characterization. |
