LP and ML Sparse Spike Inversion for Reservoir Characterization - A Case Study from Blackfoot Area, Alberta, Canada

Abstract

The present study focuses on Sparse Spike inversion techniques to estimate P-impedance and density profile, the important parameters for characterizing the reservoir. Two types of Sparse Spike inversion method popularly used in the seismic industry are utilised in the present study - Linear Programming sparse spike inversion (LPSSI) and Maximum Likelihood sparse spike inversion (MLSSI). The inversion techniques are applied to process the 3D-3C Blackfoot post-stack seismic data for better characterization of the reservoir. The inversion results show low inverted P-impedances and density around 1058-1068ms. The low values are inferred to be due to the presence of sandstone channel. The high correlation coefficient variation of 0.955 to 0.992 for LPSSI and 0.922 to 0.974 reiterate that both method works well and are able to resolve the channel but the estimated results are better from LPSSI than from MLSSI method. The synthetic relative error showed variation from 0.125 to 0.298 and 0.225 to 0.388 for LPSSI and MLSSI respectively. These numbers indicate that the relative errors from the implementation of LPSSI method are low in comparison to the MLSSI algorithm. The analysis of inversion results suggests that LPSSI method provide better reservoir characterization than MLSSI for Blackfoot seismic data.