Gas hydrate physical properties imaging by multi-attribute analysis — Blake Ridge BSR case history

Abstract

We analysed the bottom simulating reflector (BSR) of the Blake Ridge, by using the ‘multi-attribute’ analysis. This technique allows prediction of petrophysical/geological parameters along seismic lines (velocity, porosity, density, resistivity etc.) starting from log data. Seismic attributes can be thought of as all parameters derived from seismics, by assuming that a seismic trace is the real component of a complex one. The first attributes that can be calculated are the instantaneous amplitude, the instantaneous phase and the instantaneous frequency. All the other attributes that will be considered are derived from the previous ones. The algorithm that allows the correlation between logs and seismics is the generalised multiple linear regression. This equation is solved by a least squares approach, called ‘training phase’, which effectively consists of training the seismics to predict the reservoir parameter of interest at the tie locations. The resulting function is then applied to the seismic profile, generating a target log predicted section. We obtained five sections (VSP and P-wave velocity, density, porosity and resistivity), each of them representing the distribution of the corresponding property along the profile; they permitted better characterisation of the nature of the BSR present in the Blake Ridge area. The obtained-velocity section was finally translated in terms of gas hydrate and free gas distribution by using theoretical approach.