Acoustic impedance inversion using stacking velocities: Hussar Example

Heather Jeanne Eileen Lloyd, Gary F. Margrave

In band-limited acoustic impedance inversion low frequencies must be provided from a non-seismic source. Impedance logs, estimated at wells from standard sonic and density logs, are commonly used however other sources such as impedance sections created from stacking velocities can be used. Stacking velocities must first be conditioned by interpolating on a grid that is the same size as the seismic data. For this purpose the PCHIP (Piecewise Cubic Hermite Interpolating Polynomial) algorithm was used. The next step is to convert the stacking velocities to interval velocities. Densities can be estimated using Gardner’s equation and the interval velocities and from there an impedance section can be calculated by multiplying the densities and interval velocities. Information from this impedance section in the 0-2 Hz band was combined with the seismic data from Hussar using the BLIMP (Band-Limited IMPedance) algorithm. The inversion has high lateral variation so a second inversion was calculated using the mean of the impedance section. The mean inversion had a percent error of 11% between 0.2 and 1.05 seconds where as the regular inversion had a percent error of 12% in the same interval. Good results were obtained using the stacking velocities from standard processing but these had sparse picks. Better results are likely to be obtained with finer stacking velocity picks and constraining the conversion of stacking velocities to interval velocities.