The Dunga field Aptian reservoir is very thin bedded and heterogenous at a fine (cm) scale below log resolution and presents distinct challenges for reservoir modelling. This paper describes methods for upscaling detailed thin bed reservoir description and facies mapping from core and log data to build a 3D geological facies model.
The Dunga Field is penetrated by approximately 190 wells, of which 9 are cored. Field wide correlation of the main reservoir units is sub-seismic and below biostratigraphic resolution. A novel method using step outs (cluster correlations) from the key cored wells to the closely spaced vertical well penetrations has provided a more detailed correlation and mapping of reservoir units. The method has provided consistent and field wide definition of depositional trends and architecture of deltaic mouthbars which form the main reservoir sandbodies.
Facies modelling of the heterogeneous reservoir was built by integrating and upscaling detailed core data, well data and depositional trends. Four facies, identified from log and core evaluation, are modelled using stochastic method (SIS), where facies fraction was controlled by up-scaled cells. The variogram directions of facies model are guided by sand body depositional trends. Depth appropriate facies probability maps are used for modelling clean sand and heterolithics facies to guide facies distribution.
The detailed study carried out to build the 3D geological facies model is helping to achieve better production history matching and is being used for future well planning.