Stochastic Time-lapse Inversion of a CO2 Sequestration Synthetic Seismic Data

Abstract

In the domain of time lapse interpretation the development of tools to create scenarios that can explain the time lapse anomalies directly from the seismic data and, at the same time, evaluate the uncertainty associated with the available input data can help to accelerate the incorporation of this important information to flow simulations. The purpose of this study is to investigate whether the time-lapse seismic inversion can detect changes of the seismic response caused by injection of distinct amounts (10,000 t and 30,000 t) of 80/20% CO2-CH4 mixture into sandstone reservoir in South Australia using synthetic seismic data. The number of realizations as well as the informed uncertainties of input parameters plays a big role in the unveiling of subtle effects of small volumes of injection as this one. The presented findings suggested that the current approach has succeeded in using synthetic data to test the limits of this recent technique.