AAPG Annual Convention and Exhibition

Datapages, Inc.Print this page

Application of Numerical Outcrops of Fluvio-Deltaic Systems in Building Geological Reservoir Models


Fluvio-deltaic reservoir models for hydrocarbon production are created based on sparse core and seismic data. The characterisation of the heterogeneity and connectivity properties at a sub-seismic scale can be improved by complementing field data with numerical outcrop data that are generated by physics-based numerical models which are capable of simulating high-resolution stratigraphy. We carried out many simulations using Delft3D, a high-resolution open-source process-based numerical model, simulating hydrodynamics, sediment transport and morphological changes in a deltaic setting (Geleynse et al., 2010, 2011). Numerical simulations show the capability of Delft3D to accurately resolve the morphology and stratigraphy of developing river delta distributary networks under different hydraulic and sedimentary forcing, for example in river-, wave-, or tide-dominated setting. Model output comprises an emerging self-organizing distributary channel network, of which metrics, such as channel aspect ratio and drainage density, were calculated. The model was applied to investigate the impact of water and sediment supply, grain size, and substrate composition on deltaic development, showing how allogenic and autogenic processes affect channel properties, grain-size trends, facies distribution and the connectivity of sand bodies. We propose a workflow for the implementation of Delft3D to optimize stochastic reservoir models, by treating the model output as a ‘numerical outcrop’. A database of numerical outcrop built with different boundary conditions provides training sets and images that can be used to characterize deltaic stratigraphy in addition to existing statistic and geometric data collected in the field at outcrops or from modern sedimentary systems. This can be implemented using a geometric (e.g., channel dimensions) or statistic (e.g., channel density) approach, similar to field data and can therefore be directly used in Petrel or RMS modelling software.