Известия Томского политехнического университета: Инжиниринг георесурсов (Jan 2018)
Phase-frequency deconvolution of seismic waves
Abstract
Relevance. At present, deconvolution methods are widely used to process data for a variety of applied purposes, such as radio and sonar detection, navigation, hydroacoustics, geophysics, etc. These methods are of great importance in seismic prospecting when searching for oil and gas fields in thin-layered geological environments, where waves reflect from interfaces and interfere with each other thus making it extremely complicated to identify the boundaries of different geological media. The problem becomes even more complicated due to irregular noise in all seismic records. Therefore, to trace the boundaries in thin-layered media, many algorithms for signal compression are used, which are commonly based on deconvolution methods. However, in traditional deconvolution methods phase spectra are neglected and considered to be either minimum-phase or zero-phase spectra. Meanwhile, it is the phase of seismic signals, or rather the complex law of the change in phase spectra, that carries important information on location of reflecting boundaries. Hence, the resolving power of a signal is primarily determined by the complexity of its phase characteristic. Therefore, the development and investigation of the efficiency of a phase-frequency deconvolution algorithm is of great importance for processing and interpreting seismic materials. The aim of the research is to develop an algorithm for phase-frequency deconvolution on the basis of the method of phase-frequency tracking of seismic waves that was proposed earlier; to analyse the efficiency of this algorithm in models of geological environments; to test the algorithm for processing and interpreting general deep point materials obtained at a number of oil fields in the Tomsk region, Russia. Methods: digital processing of space-time signals and fields, discrete Fourier transform, mathematical modeling and a computational experiment. Results. The authors have developed the deconvolution algorithm based on the method of phase-frequency tracking of seismic signals. The algorithm differs from the previous solutions in its capability to analyse instantaneous phase spectra. The paper introduces the characteristic features of the phase spectra that form the basis of the algorithm. Investigations of the algorithm in models of a wave seismic field, as well as experimental processing and interpretation of actual data were carried out. The results obtained confirm the potential of phase-frequency deconvolution to analyse thin-layered geological media.
