Building a three-dimensional geological model and property upscaling of a dryland fluvial-lacustrine terminal splay succession of the Neales Cliffs, Lake Eyre, South Australia.
Miller, Tobias J.
Engineering Honours Degree 2008
University of Adelaide
Lake Eyre is the largest salt lake in Australia, and fourth largest terminal lake in the world. The lake lies in the heart of the Australian continent, approximately 950km north of Adelaide, and 900km southeast of Alice Springs. With an area of 11400002 km, the Lake Eyre basin is one of the largest internal drainage areas in the world, and covers one sixth of the continental landmass of Australia.
The Neales Cliffs study area is situated in the lower Neales River on the western side of Lake Eyre. The Neales River forms the northern distributary of the Neales Fan and terminates with the Neales Terminal Splay Complex in the northwestern corner of Lake Eyre.
The Neales Cliff outcrop provides an excellent exposure of a fluvial-terminal splay-lacustrine sedimentary succession. There, the Tertiary Etadunna Formation is erosively overlain by the Quaternary succession of the Neales Formation, which is subdivided into five major units. Three of these units; Unit 1, 2 and 3, are of particular interest in the succession. Units 1 and 3 are predominantly composed of thick, coarse grained sands with good porosity and permeability. Conversely, Unit 2 is a thin zone composed of fluvial muds, with poor porosity and permeability, thus acting as an ideal sealing facies over Unit 1.
With this in mind, these Units were of primary interest when performing the upscaling procdures.
Ephemeral sandy fluvial-lacustrine deltas and terminal splays associated with dryland depositional environments are important reservoirs in many basins around the world.
Research on modern depositional analogues, such as the Neales Cliffs, provides an ideal natural laboratory for further insight into these Terminal Splay complexes.
The Schlumberger Petrel 2007 software package was used to construct the geological models from the basic field data. This software enables the creation of a full scale geological model even from such basic inputs as well logs and facies data. Petrel uses both stochastic and deterministic methods to extrapolate data from a well location throughout the entire model. Primarily, stochastic algortithms were used in this thesis as they best represent “real life” distributions.
The Petrel 2007 software is a very large and powerful program, and with little or no prior knowledge of the software, manipulating it can be difficult. Therefore a very detailed methodology was constructed in this thesis which would allow another researcher to accurately follow it in order to expand the geological model, or create a new, larger Splay model.
In order to construct the model, the basic field well log data was upscaled to a number of progressively coarser models. These models were then analysed to determine which level of upscaling would be acceptable with regard to the preservation of important facies within each unit.
Three increasingly coarse models were created at three different areal grid resolutions; 60, 30 and 15 Layer models at 1x1m, 3x3m and 15x15m.
Results showed that upscaling the original logs to a 60 and 30 Layer model would generally preserve the important facies within each zone. However, the 15 Layer model generally removed the important fine scale facies and therefore would be unsuitable for use in a full scale model.
Important recommendations for future studies are to use the workflow presented in this thesis to construct a full scale model of the Neales Terminal Splay Complex in order for a complete reservoir simulation study to be carried out. In order to do this, ‘pseudo-wells’ would need to be constructed on the Splay itself.
This would allow a further and more detailed insight into the structure and composition of a Splay system, and permit volume calculations and fluid flow simulations. This could enhance the exploitation of current analogous reservoirs around the world.