Abstract
Complex fluid distribution pattern being investigated in a shallow unconsolidated viscous oil bearing multi-layered sandstone reservoir system in North Kuwait using the "Rock Type" classification approach. These rock types are characterized by petrophysical properties that can be linked to genetic stratigraphy, depositional lithofacies, and secondary diagenesis to recognizing Rock Type framework using an integrated geomodel building approach. The concept that "Rock Type" is a product of rock fabrics to form a link between geologic description and petrophysical properties; thus there is a predictable relationship between petrophysics and stratigraphy. To capture such complexity, we use core-based rock-fabric method to determine and map reservoir characterization variations fieldwide. Identifying these complex variations in petrophysical relationships by geological cycle provides a strong quantitative basis for developing an accurate framework for flow modeling and simulation.
In this paper, Rock Type is based on about 150 well core data and later propagated to over 800 well logs. Ten primary lithotypes have been predicted using Self Organizing Maps (SOM) and a Cluster Analysis Model. They are defined based on core analysis data using Mercury Injection Capillary Pressure (MICP) parameters, capillary pressure, water saturation, core porosity and permeability, in order to assess the effectiveness of the combined model. These petrophysically defined distinct lithotypes are further merged into seven "Rock Type" for geomodel construction. This approach has been successfully applied to several other viscous oil reservoirs worldwide. This provides a rational opportunity to anticipate risk and uncertainty associated with reservoir fluid distributions and production profile prediction with enhanced confidence.
This developed geomodel framework will be subsequently used in dynamic modelling and history matching to improve the static description of the reservoir. Consequently, this would significantly contribute in flow modeling and reservoir simulation to short and long term production forecast and development profile predictions.
