Skip to content

Building and upscaling a 3-D reservoir model of a modern mixed-influence delta system and the associated effects on reservoir heterogeneity.

Massey, Thomas

Engineering Honours Degree 2009

University of Adelaide


Reservoir modelling is commonly used to characterise the heterogeneities of a reservoir, to reduce uncertainty in volume and to provide a framework for reservoir simulation. Upscaling of reservoir models is typically required prior to simulation. This requires that a compromise be made between the minimisation of data quantity and the preservation of important reservoir model attributes such as reservoir geometry, pore volume and fluid flow characteristics. The amount of detail initially input to geological models to define these attributes is vulnerable to purging by the upscaling process, potentially resulting in wasted time and money.

A suite of three 3-D reservoir models of a modern mixed-influence delta system is presented, to test the effects of upscaling on reservoir heterogeneity. These models were based on mapping of sedimentary facies at three hierarchical scales in the Mitchell River delta, Northern Queensland and facies thicknesses inferred from regional studies and fieldwork conducted by the WAVE Research Group. The study area is tide-dominated and composed of a mixture of wave, tide and river influenced morphologies with reservoir facies dominated by beach ridges, distributaries channels and channel bars.

Upscaling of these models was undertaken in a lateral and vertical direction while volumetrics and connectivity were monitored. Lateral upscaling resulted in little change in volume however tended to affect the larger, connected facies. Vertical upscaling resulted in larger changes in volume and was controlled by the thickness and location of facies units within the sequence. This change occurred when facies were upscaled to over half their thickness. Connectivity decreased with increased upscaling ratios and indicated key connectivity geometries and heterogeneity scales of <10 and <50 metres respectively. These limits were mainly controlled by beach ridge and channel bar geometries and their relationship with surrounding facies, suggesting they are a limiting factor in the characterisation of mixed-influence depositional environments. Detailed modelling produced better characterisation of these reservoir attributes however they experienced greater manipulation during the upscaling process. Lesser detailed models are suitable if large grid cell dimensions are required.

It is recommended that upscaling of the Mitchell River delta should avoid vertical upscaling above half of major facies unit thickness to maintain volume and connectivity of the system. This could be achieved with cell dimensions up to 20x20x1 metres.

Australian School of Petroleum



T: +61 8 8313 8000
F: +61 8 8313 8030