Supervisors: Assoc Prof Simon Lang, Prof Cedric Griffiths (CSIRO)
Project support: NCPGG
Scholarship support: The University of Adelaide
The project aims to investigate the possibility of using forward stratigraphic modelling (specifically Sedsim) to build a reservoir flow model for an oil field which exhibits stratigraphically controlled lateral and vertical facies variations. The Sedsim model will form the base for a reservoir simulation model, which in turn will be history matched with production data.
The Stag Field in the Dampier Sub-basin (North West Shelf) has been chosen as a case study for simulating a shallow marine siliciclastic reservoir, after consultation with Apache Energy and Santos. The Sedsim computer package can simulate 3D distribution of uncompacted sediments of four different grain sizes on a bathymetric surface, by modelling forward in time sediment transport, deposition and erosion. The first stage of this modelling study consists of running a depositional simulation over an 80x50 km area on the Enderby Terrace, surrounding the Stag field, for a time interval of 10 million years in the Early Cretaceous. The palaeoreconstruction is based on 2D and 3D seismic and well data from around 20 exploration wells. Once the overall grain size distribution in this regional model is achieved -and thus the main depositional controls identified- a more detailed model will be built for the Stag field (approx. 10x15km), simulating the deposition of the sands included in the M. australis palynozone. The reservoir scale model will predict and quantify facies variations and stratal architecture for the inter-well space, beyond the seismic resolution.
As part of the project, sensitivity studies and variable-resolution simulations have been run for different data sets, and have shown that Sedsim can be used to identify an optimum lateral and vertical scale of the reservoir model, as well as capturing the salient lateral and vertical variations and facies continuity.
Subsequent compaction and diagenesis will be estimated for the computer-generated
sediment bodies. Synthetic lithologies will be then translated into distributions
of rock/flow unit properties, thus generating a reservoir model in a format
suitable for a multi-phase flow reservoir simulator.