北海スノーレ油田を対象とした4D震探による油層経年変化可視化の試み

抄録

This paper introduces a workflow to compute the fluid saturation in a reservoir using 4D (time lapse) seismic data. In this approach, the decomposition of the impedance into velocities and density is the key to derive the quantitative reservoir properties using the currently available rock-physics applications. However, the decomposition is a challenging task due to the severe influence of noise on the analysis especially when the seismic data has a maximum-offset angle less than 30°. The proposed method of impedance decomposition simultaneously solves three elastic impedance data derived from the seismic inversion of angle stacks, where the most important reason that makes the decomposition difficult is the effect of noise. Theoretical investigation revealed that the offset-consistent component of noise mostly affects the determination of density but the noise effect is not severe to determine velocities. On the other hand, the effect of the random component of noise appears equally in the determination of velocities and density. Based on this result, the author developed a method of decomposition incorporating rock-physics bounds as the constraints for the analysis. Afterwards, a sequence of rock-physics theories is applied to transform the resultant properties into the bulk modulus and density of the in-situ fluid. Eventually, solving three simultaneous equations for the bulk modulus, density and the total saturation of three possible fluids; we can compute the saturation of water, oil and gas. As a case study, the workflow successfully visualized the change in fluid saturation based on an actual 4D seismic dataset of the Snorre oil field in the North Sea.