The 1st AAPG/EAGE PNG Geosciences Conference, PNG’s Oil and Gas Industry:
Maturing Through Exploration and Production

Datapages, Inc.Print this page

Numerical Sand Box as a Tool for Hydrocarbon Exploration: Applications to the Hides Anticline and the Western Papuan Fold and Thrust belt


Analysing the geometry and the geological evolution of subsurface structures is fundamental to the characterisation of reservoirs and the search for hydrocarbons, particularly in fold and thrust provinces. Field data bring fundamental insights but are often insufficient to understand the geological system. Imaging techniques, such as seismic, may be limited in quality and/or have insufficient penetration to fully understand the large-scale structures. Thus it has been necessary to utilize geometrical modelling, backed up by occasional analog experiments using sand, clay and silicone to simulate the formation of large geological structures. These have proven to be very valuable but their setup is time consuming and reproducibility is often an issue. It is also difficult to run a large number of simulations to fully explore the parameter spaces and quantify the uncertainty. Darnault & Hill (AAPG in press) report the results of 17 analog modelling of the Hides anticline by IFPEN run over an 18 month period. We present UWGeodynamics, a finite element software which facilitates numerical-mechanical and thermo-mechanical modeling in 2D and 3D at all scales. Whilst this can be for the expert user on a supercomputer, we discuss a simple setup for the desk geologist that simulates sandbox modelling. The advantage of the UWGeodynamics sandbox modelling is that over 300 2D models can be run overnight on a super computer (or in the cloud) or 10+ models overnight on a laptop as opposed to one analog model/month. This allows us to fully investigate the range of all inputs/variables overnight and produce a movie for each experiment. These can be compared both visually and statistically to determine not only the best outcome but also the range of reasonable outcomes. A major benefit is that we can determine which parameters are important or sensitive in the model and which have little effect. To replicate sandbox modelling of the Hides anticlines (after Darnault & Hill in press) we ran 1050 2D models in 2 days and reproduced their simple ductile, brittle and complex ductile rheologies. We were able to produce the same outcomes with different mixes of parameters, including strain rate, the angle of the fault ramp and subtle variations in material strength and coefficient of friction. This allows a much improved understanding of the structural evolution. Our aim now is to test many other possibilities to improve the fit to the observed structural geometries from the field. We then aim to expand the models to 3D (requiring a supercomputer or cloud computing) to test variations along strike. The same modeling can be applied to other structures along strike in PNG, such as Muruk and Kutubu, and to compressional or extensional structures elsewhere in the world.