Abstract

Coloured Seismic Inversion (SCI) is a simple, fast and cost effective way of inverting seismic reflection data from an interface attribute to a layer-based attribute. The process creates a volume that is rotated 90 degrees with respect to the reflection data and whose amplitude spectrum matches that of well log impedance in a gross sense. The result of the process is a relative impedance seismic volume, which has in effect been shaped to well data in terms of amplitude and frequency. In addition no wavelet estimation is needed, and although an interpretive low frequency model is not used, the technique provides a robust inversion that honours the impedance trend of available well data. This process can be undertaken by non-specialist, seismic interpreters, is very fast to perform and the result benchmarks well against unconstrained sparse spike inversion.

Due to the ease of use of the application and robust result the SCI method is simple to implement as part of a standard interpretation workflow and has been used by ConocoPhillips Indonesia to invert carbonate and clastic reservoirs across the region. Two case studies are presented in this paper.

SCI was applied to full stack and offset cube seismic reflection data over the Belida Field, West Natuna as part of a major reservoir characterisation study of this mature oil field. Due to the fast track nature of the project this approach was preferred to a more quantitative seismic inversion method. The resultant full stack coloured impedance volume has been loaded directly into the geological model and used as a guide to facies distributions away from well control.

The coloured impedance volume was also used in a calibration of well data to multi attribute seismic data (including Pre-Stack Inversion attributes) to derive Vshale and had one of the highest correlation coefficients of any of the attributes to Vshale.

An example from carbonate reservoirs, offshore East Java is also presented. This project is again a fast track interpretation and the SCI has provided a volume that has significantly reduced interpretation time as reflectors are more laterally continuous than the reflectivity volume and can thus be auto-picked to a greater extent. In addition, as the volume is layer based, changes in relative impedance can be correlated directly to changes in facies.