Semi-automated Rock Layer Recognition from Borehole Log Data Using Combined Wavelet and Fourier Transform: A Case Study in the KG basin, India

Abstract

Abstract The accurate identification of rock layer interfaces from wireline logs is one of the crucial issue of geological interpretation as we face difficulties in traditional interpretations due to the presence of various kind of high and low frequencies in the log data. Determination of rock boundaries in shallow marine sediments is very important to understand several issues related to the exploration of conventional and non-conventional hydrocarbons. Presently a combined wavelet and Fourier transform-based approach was demonstrated to discriminate the rock layers from wireline log data in Krishna-Godavari basin, which were acquired during the India National Gas Hydrate Program Expedition 02 (NGHP-Exp.-02). Initially, we selected the suitable mother wavelets and optimum level of decomposition to perform the wavelet transform by computing the error between the original log signal(s) and approximation coefficients (cA’s) of the decomposed signal(s) at several decomposition levels. Subsequently, discrete wavelet transform (DWT) of a particular decomposition level was applied to the log signals to obtain the detail coefficients (cD’s), which contain the high-frequency components of the log signals. Further, these high-frequency components of the log signals were utilized in frequency spectrum-based filtering approaches. This approach was adapted to yield information about the abrupt changes occurring in the log signals across the studied depth interval. The combined application of Fourier and wavelet transforms to shallow marine log data demonstrates that this approach is an efficient tool for delineating rock layer boundaries.