3D facies and reservoir property prediction of deepwater turbidite sands; case study of an offshore Niger delta field

Abstract

The distribution of lithofacies and key reservoir properties such as porosity, net to gross and water saturation is essential to describing reservoir heterogeneity predicted within a clastic deep water reservoir unit in the study area. A systematic analysis of core and gamma-ray wireline logs was used to identify the reservoir unit in this study. The identified reservoir unit was mapped on 3D seismic data, and four lithological and fluid sensitive seismic attributes extracted for the reservoir were fed as inputs into an unsupervised artificial neural network (UANN) analysis to define the depositional pattern of the reservoir unit. Afterwards, a 3D sand facies trend model was built using the facies depositional pattern map and upscaled properties from well data. The generated 3D sand probability model from trend modelling was subsequently used to constrain geostatistical facies modelling of the reservoir unit. The systematic stratal analysis from core, well data and the integration of seismic attributes using UANN reveal the reservoir unit as a system of turbidite lobe sands and turbidite channel levees sands. The 3D Geostatistical facies distribution reveals spatial and vertical facies heterogeneity. It indicates high sand amalgamations and static hydraulic connectivity of turbidite lobe sands in the distal part of the study area. Further geostatistical distribution of reservoir properties conditioned to facies indicated good to excellent reservoir quality within the turbidite lobe sands, moderate to low reservoir quality within turbidite channel sands and poor reservoir quality at the turbidite channel margins and the deep-water shale facies. The 3D facies and reservoir property models built in this study would ensure optimisation of the reservoir unit, enable better estimation of hydrocarbon volumes and serve as input data for flow simulation.