Rock-Typing of the Lower Cretaceous Formation in Onshore Abu Dhabi: An Innovative Workflow Integrating Pore Types for the Prediction of Field-Scale Dynamic Properties

Abstract Rock-types are the key building blocks of reservoir models and crucial for field development. As such, they must capture the key geological and reservoir quality variability. This study presents an innovative workflow through which geological and reservoir quality understanding, gained from detailed core, thin-section observations and laboratory-based petrophysical data, are integrated with open hole log data and KrPc properties to define reservoir rock-types in the lower cretaceous Formation. The strong link with the geology allows prediction of more realistic spatial distribution of dynamic properties. Geological rock-types were defined as the first step, based on their pore system types. The focus on pore system type, rather than the traditional facies type, allows the integration of sedimentological as well as diagenetic factors in the rock-typing process. Facies groups sharing the same pore network were then linked to their dynamic properties via the analysis and the classification of an extensive set of MICP data, and these combined rock-types subsequently linked to the log signatures using machine learning techniques calibrated on core samples. Finally, water saturation functions are established in rock-type specific saturation height functions. The focus on the pore system types is a crucial element of innovation, which is especially important in carbonate rocks, where diagenesis can significantly modify the original pore system, such that same facies can have very different reservoir behavior. Based on this approach, rocks dominated by macropores, mixed macro and micropores or by sole micropores were identified and grouped together. Furthermore, a specific group referring to intercrystalline microporosity within dolomitized fabric was created to allow better link with logs. The better reservoir properties in lower cretaceous are found in pervasively dolomitized rock; however, our results show that we can distinguish, within the dolomite pore types, three sub-classes of dolomites dominating the variable dynamic behavior of the reservoir. These are due to the combination between the degrees of cementation and connected macro/micro-pores. The water saturation observed in logs is very precisely reflected by the results of Brooks-Corey saturation modeling based on these rock-types. The log-based prediction deriving from the application of this innovative workflow is successful to a conformance of 75%, which makes the log-based detection a sound input for geographical interpretation of the rock-type occurrence across the field. The results prove that the definition of Reservoir Rock-Types based on the pore network, rather than conventional facies, helps to improve reservoir properties’ prediction in carbonate reservoirs, thanks to the integration of petrophysical properties with the geological understanding of their origin (i.e. overprinting sedimentary and diagenetic events). This novel method constitutes a big step forward for rock-typing in carbonates, compared with conventional rock-typing methods, with which reservoir property prediction was almost impossible.