Abstract
In carbonate reservoirs, heterogeneity is usually driven by both depositional and diagenetic patterns, which influence the final static and dynamic reservoir properties. Karstification is one of the most crucial diagenetic phases because of the way in which it constrains the permeability behaviour of the field. Defining static rock-type can therefore be problematic due to the highly imbricate karstic phases and their various distributions and extensions. A further major difficulty is the spatial distribution of rock-types in reservoir models, each linked to different sedimentological and karstic processes and each with their own, but interdependent, geometrical characteristics.
This paper presents a modeling workflow applied to a large carbonate mound in Kazakhstan, affected by hydrothermal karstification. This new approach is based on nested stochastic simulations, geologically-driven and derived from the relationships between karst, depositional facies, and proximity to fault/fracture zones. In addition to the initial depositional model, three diagenetic phases and bitumen deposits are modeled successively in order to restore the complex spatial relationship existing between described alterations. The overall workflow is designed to quantify uncertainties at each step of the nested chain, including structural and filling uncertainties.
The results of this study are quantification of uncertainties and volumetric estimations. They underline the importance of karst distribution and density on reservoir volumes ans especially on static recovery factors. They introduce high permeability drain increasing connectivity between injector and producer wells.
This workflow provides new perspectives for reservoir modeling, including the appropriate effective porosities and permeabilities affected by karst features. It leads to multiples realization scheme allowing uncertainty quantification, either in mature or appraisal field development. The outcome of this modeling workflow, after a stage of up-scaling and dynamic simulation, allows a better assessment the potential impact on field behavior.
