oa Geology Meets Petrophysics: from Example of a Process-Based Rock Type Methodology for a Khuff Reservoir, North Oman

Carbonate reservoirs are inherently heterogeneous as result of being deposited in laterally variable<br>settings with subsequent overprint of complex and substantial diagenetic processes. The Upper Khuff<br>carbonate reservoir, deposited on a large carbonate/evaporitic ramp (Late Permian - Early Triassic), is<br>a prime example for such reservoir quality variability dictated by its intrinsic carbonate nature.<br>A Rock Type (RT) “process-based” geological-petrophysical approach has been applied to a Upper Khuff<br>reservoir in North Oman in order to improve the reservoir characterization and the understanding of its<br>heterogeneity, and ultimately to derive representative static and dynamic models. This approach is<br>based on the integration of sedimentological/petrographic core observations and the analysis of<br>petrophysical properties in wells.<br>A core-based Rock Type scheme was defined based on Porosity/Permeability from CCA, pore type<br>identification and sedimentological observation on thin sections. Each RT was linked to the original<br>depositional facies and the subsequent diagenetic processes (cementation, dissolution, dolomitization,<br>anhydrite precipitation). Pore throat distribution and capillary pressure curves were also used to<br>calibrate the classification scheme. RTs with enhanced properties correspond to two classes:<br>dolomitized oolitic-skeletal Grainstones and Thrombolites (with intercrystalline, interparticle/mouldic<br>and vuggy porosity) and oolitic/peloidal/skeletal Grainstones and Packstones (with mouldic,<br>interparticle and vuggy porosity).<br>A Neural Network approach was used to implement this RT scheme to all wells, and its results were<br>calibrated to cores. An electro-facies log was obtained by combining Gamma Ray, Porosity and BVSxo<br>logs. This results in a ‘Depositional Facies’ log, comprising: Bioconstructed facies, Grainstones (well<br>and poorly sorted), Packstones to Mudstones and Argillaceous Mudstone. The ‘Lithology’ log was<br>generated from the Gamma Ray, Neutron/Density and PE logs to distinguish amongst Limestone,<br>Dolostone, Anhydrite and Argillaceous Mudstone. The final RT log was generated by merging<br>‘Depositional Facies’ and ‘Lithology’ logs.<br>‘Depositional Facies’ and ‘Lithology’ logs were used to build 3D Facies properties (Depositional Facies<br>and Lithology). The latter were directly merged to generate RT models. Porosity and Permeability were<br>modelled conditioned to RTs which resulted in more realistic distributions.