Seismic Inversion: Technical Details
Process 1
Frequency Compensated Inversion
Process 2
Coloured Inversion
Process 3
Frequency Domain Constrained Sparse Spike Inversion
Seismic data is band-limited, reducing resolution and quality. LFC Inversion includes both seismic data and well data, where well data serves to add the low frequency information below the usable seismic frequency range and to constrain the inversion.
Wavelets derived and the most appropriate wavelet is chosen. Accurate wavelet estimation is critical to the success of any seismic inversion. The inferred shape of the seismic wavelet may strongly influence the seismic inversion results and, thus, subsequent assessments of the reservoir quality.
To extend the frequency bandwidth available, low-frequency data is derived from log data. The autocepstral frequency bounds were derived and used to guide the low frequency compensated inversion, the resulting acoustic impedance is computed for an appropriate time range. A scaling factor which relates the well and seismic reflectivity is applied during the low frequency compensated inversion process.
Coloured Inversion is a method developed by Lancaster and Whitcombe (2000) that can rapidly invert to relative elastic attributes by deriving an inversion operator that shapes the seismic spectrum to the spectrum of well log data, e.g. impedance. There’s no need for wavelet extraction or a low frequency model and the output is not biased by the trend in the well data. This technique is highly recommended in areas where there is not enough well control to generate a reliable absolute inversion.
(FD –CSSI) Sparse Spike is a method of obtaining an estimated broadband reflectivity series from band limited seismic data which seeks to minimise reflectivity spikes to arrive at a physically acceptable solution whilst satisfying a number of horizon constraints including wavelet trade-off and low frequency information. Mathematically the technique involves the construction and solution of a large sparse matrix and uses the L1-norm –this is significant for mathematical and geological reasons. This is one of the most advanced and computationally expensive techniques in quantitative interpretation. This is applied to a subset of the pilot area.