Static Modeling of deep water sand rich reservoir of braided fluvial origin, deposited in syn-rift set up is challenging in order to capture the reservoir complexities like presence of low permeability facies and complex fault geometry. It becomes more difficult with limited well controls and lack of effective seismic attributes. To characterize such complexities, a robust rocktype based modeling approach is considered with effective spatial control on properties. A rocktype based static modeling approach using geostatistical techniques guided by conceptual depositional model is adopted here as a case study. Five rocktypes are classified based on hydraulic units (HU), defined by FZI (Flow Zone Indicator) technique using porosity and permeability measurements from core data. A depositional model of braided fluvial setup is conceptualized based on integrated analysis of core data, well log and overall reservoir architecture. Using this conceptual depositional model, trend analysis for different rocktypes is carried out to derive vertical proportion curve (VPC) and variograms which are then used to populate reservoir properties away from the wells. The static model was simulated for dynamic studies to depict production profile scenarios for assessing the commercial viability of the field. Pressure data from appraisal wells has established three different fluid contacts separating the reservoir into different segments. Surgically interpreted faults are used to define the segment boundaries and build a detail as well as appropriate fault network model. Use of rocktypes based on FZI method, instead of Sand-Shale classification scheme, helped in capturing the reservoir heterogeneities within a sand rich system. A property modeling approach using trend analysis along with concept based variogram models have provided better control on properties away from the well especially when the seismic attributes are not that effective. The model was validated in a recently drilled appraisal well. The reservoir properties at this well location have shown great match with the pre drill predictions from the model.

You can access this article if you purchase or spend a download.