Iterative Geostatistical Time Lapse (4D) Seismic Inversion

Baraa Osman Abdo Osman

Key information

Authors:

Baraa Osman Abdo Osman (Baraa Osman)

Supervisors:

João Miguel Figueiredo Narciso (João Narciso); Leonardo Azevedo Guerra Raposo Pereira (Leonardo Azevedo Guerra Raposo Pereira)

Published in

12/19/2023

Abstract

Time-lapse (4D) seismic analysis is crucial for the containment and conformance assessment of CCS projects. However, interpreting CCS data for 4D signals like the Sleipner dataset faces challenges from noise, velocity effects, and thin-bed tuning. Advances in data processing and quantitative analysis methods offer promise for more accurate interpretations. Seismic reservoir characterization relies on inverted elastic subsurface models to predict the distribution of porosity, permeability, and fluid saturation. Probabilistic seismic inversion methods provide high-resolution models with uncertainty assessment. Our proposed iterative geostatistical time-lapse inversion methodology predicts the difference in acoustic impedance \ac{Ip} using available seismic data. It considers all available vintages simultaneously and uses geostatistical simulations for perturbation, driving convergence and better exploration of the model parameter space. We apply this methodology to monitor the growth of the CO2 plume from 1994 to 2006 using the 2007 processing of the Sleipner's seismic data.