Identifying Shallow Water Flow in Offshore Malaysia Using Multicomponent Data and FWI Approach

Abstract

Abstract This paper presents the identify of the probably shallow water flow (SWF) conditions in offshore Malaysia by estimating and analyzing the compressional velocity (Vp), shear velocity (Vs), compressional attenuation (1/Qp), shear attenuation (1/Qs), and Density inverted from multicomponent seismic data using full waveform inversion (FWI). Vp, Vs, and Density were estimated through a join inversion of compressional PP and converted PS seismic data. The attenuation parameter is estimated from the quality factor Q which is derived using spectral ratio method. The estimated attenuation was used to build velocity model for inversion, based on the dispersion and attenuation theory. Kennett's full reflectivity inversion approach is used to achieve better velocity estimation especially at shallow (top-hole) zone where SWF typically occurs. Vp/Vs and Poisson's Ratio (rigidity) are then derived from inverted Vp, Vs and Density. The SWF is identified through some anomalies in inverted properties. A 2D multi-components seismic data from offshore Malaysia was used for study. The data is affected by gas clouds which result poor amplitude in some part of study area. The inversion result is considered well, showed by very good correlation between synthetic and real seismic data. Depth image of inverted properties showed good velocity estimation at shallow depth (0-500m). The inversion updated the initial velocity model quite well. Low density and anomalously low Vp and Vs were exhibited, indicated an unconsolidated sediments. Increasing Vp/Vs and Poisson's Ratio indicates that the unconsolidated sediments loss of cohesion due to the increase of pressure which is typical for over-pressured SWF sediment. However, it should be noted that, the seismic data used is an exploration (not a hazardous) data, and anomalous are found at relatively shallow depth (not more than 500 m in depth or about 400 ms in TWT). The inverted property's value may be varied for other data type or location.