Chronofacies discontinuities: Precise approach for reservoir layering -- An example from carbonate reservoir, Offshore Abu Dhabi, UAE

Mohamed N. Bushara, Saleh M. Baslaib, Zakum Development Company (ZADCO), Abu Dhabi, UAE Abstract Chronofacies analysis is performed on five carbonate cores to identify facies discontinuities and sedimentological surfaces. These surfaces presumably bound discrete lithofacies strata that are genetically correlative, petrophysically distinct, and depict flow-controlling parameters. They may vary from minor temporal breaks such as flooding surfaces, depositional hiatus and erosion to major depositional shifts and unconformities. Bounding surfaces are important not only for their packaging nature of connected rock fabric but also they are commonly flow barriers and fracture prone zones. On cores they are found as soft burrowed surfaces with thin argillaceous laminae, grit surfaces representing local transgressive lag, depositional contacts, or erosional scour where matrix-filled bioclasts are seen partially sheared. The sedimentary discontinuities in each well were first calibrated to log response, correlated across other wells and subsequently projected onto core Kv, Kh, and profiles for further refinements. Surfaces that do not correlate beyond one well and not show vertical or lateral reservoir property contrasts were eliminated. The resultant zones were then scrutinized with dynamic data (RFT and TDT) profiles to ensure some level of pressure contrast between these zones. The result is an integrated layering model that is pseudo flow units and best characterizes rock heterogeneities. Appreciable results are attainable if description model is taken fieldwide and continuously been conditioned with new field data. Introduction Reservoir layering and facies zonation of carbonate rocks in B field (Fig. 1) offshore Abu Dhabi, pose an increasingly greater challenge as modeling techniques evolve and the field enters stimulation phase of recovery. The accuracy of layering in predicting reservoir performance and delineating upswept oil and water encroachment pockets, however, relies on good geologic description and characterization of heterogeneities that control fluid displacement. Building a perfect model is often hampered by scarce geologic data and complexity associated with qualitative data transformation to numbers usable by reservoir engineering. History matching practices, whether they are done on saturation, production or pressure decline curves, often adjust geologic models to attain acceptable margins of error. This is commonly carried out without regard to incorporation of detail reservoir description elements that are probably the cause for the error. Instead, reservoir engineers tend to manipulate rock properties in simulater grid blocks to match those used to initize the model to begin with. This results in a circular data reproduction which does not validate any geologic description, and leads to failure in predicting reservoir performance. Therefore, it is utterly critical to begin with geologically sound deterministic framework that can be populated with fine scale heterogeneities. Early attempts at layering Zulu-A zone was carried out by Total CFP (Campagnie Francaise Des Petoles). Based on the systematic alternation of porous non-porous stylolized zones. A key curve to this layering model was the CNL derived porosity. As the result six layers were defined L1-2, L3, L4, L3, L6 and L7, along with four dense zones (D1-2, D3, D4 and D6). Layers L3, L5 and L7 were shown to be the most porous intervals. Subsequent studies did not introduce significant changes to the basic layering model. Nonetheless finer subdivisions of three layers were reported on the basis of core data crossplots of and Kh. These fine layers are found to correspond to shallowing upward cycles capped by Bacinella. P. 219^