Fluid and Lithology Prediction within a Coal Sequence using Seismic Attribute Modelling and Analysis (Gippsland Basin)

Abstract

In the Latrobe Group of the Gippsland Basin (Australia) the seismic response of the reservoir is masked by the presence of coal seams. These coal seams possess a large contrast in their acoustic properties (density and P-wave velocity) to those of the surrounding rocks (sands and shales). This causes strong sidelobe interference and coherent noise that overprints the more subtle sand/shale and porefill responses that we aim to detect. This has prevented reliable delineation of existing fields and the possible discovery of new fields in the area. To study the effect of coal seams on seismic attributes we modelled the seismic response of sand-shale-coal sequences. Synthetic seismograms of a ?coaly sequence? were built from simple models consisting of three to four layers and also from blocked well log data. The synthetic seismograms were computed for these models using convolutional modelling as well as the reflectivity method, which takes into account multiple reflections and mode conversions. Input models created by randomly shuffling the blocked logs were used to analyse the sensitivity and robustness of AVO attributes to variations in rock and fluid properties. Fluid effects were detectable on the far-offset amplitudes for coal sequences with no more than fifteen to twenty percent coal content. For larger amounts of coal, fluid detection becomes ambiguous regardless of the distribution of coal layers in the sequence. A useful attribute for predicting coal content is the near-offset average absolute amplitude. A better attribute for detecting fluid effects can be formed as a linear combination of the far-offset event amplitude and the near-offset average amplitude. These attributes were applied to a 3D prestack depth migrated dataset to characterise the ?coaly sequence? and calibrate it to nearby wells. The combined fluid attribute predicted the porefill in all five wells and indicated possible hydrocarbon extents in the known fields.