Estimation of Acoustic Impedance of the Tight Sandstones Using Seismic Inversion Methods: A Case Study from Whicher-Range Gas Field in the Perth Basin, Australia

Acoustic impedance can be considered as a measure of rock strength to the propagation of waves, which is a product of the density and compressive wave velocity of the rocks. Geophysicists are trying to estimate the acoustic impedance of rock layers from reflective seismic data, called seismic inversion. The difference of acoustic impedance in the common layer of layers reflects the seismic waves. Acoustic impedance is the property of rock itself, and the harder the rocks are, the higher the acoustic impedance is. Seismic inversion can be used as a reliable method for (1) estimating reservoir properties quantitatively and (2) obtaining quantified values of acoustic impedance from seismic data, well-logging data, and interpreted horizons. In general, acoustic impedance has a strong relationship with petrophysical properties such as fracture, saturation and porosity. The purpose of this study was to estimate the initial model of acoustic impedance using different seismic inversion methods to determine reservoir characteristics and compare their results in the Whicher-Range field of Perth Basin, Western Australia. For this purpose, two wells and a seismic section from the field were used. All steps were performed using the Hampson-Russell software. First, the depth-time relationship between well logs and seismic data was corrected. Then, the compatibility of the synthetic seismogram resulting from the seismic wavelet convolution with the series of reflections produced at the well location was investigated with seismic data. Various inversion algorithms including Model-based, Band-limited, Sparse Spike (Linear Programming) and Sparse Spike (Maximum Likelihood) were inspected. The correlation coefficient in all methods shows an acceptable value. Using wavelet extraction and investigation of different inversion algorithms, the model-based method, as the best method with high correlation coefficient and less error among other algorithms, was used for initial modeling. In the interval of 2050 to 2250 milliseconds of the studied section, acoustic impedance decreases significantly, indicating porosity in the reservoir area.