Quantifying Time-lapse Seismic Signal Detection for the Otway Project Using Prestack Migration

Abstract

During Stage 2 of the Otway CCS Australian project it is planned to inject a small, up to 15,000 tonnes of gas, into a saline aquifer located at depth of 1500 m. In CO2 sequestration, the ability to detect CO2 plumes is one of the main purposes of using time-lapse seismic imaging. The detectability of CO2 in seismic time-lapse surveys relies on two main factors: a sufficiently strong signal and sufficiently small noise. Therefore, to model time-lapse seismic records, we need to model not only the seismic response of the geology and the plume but also the time-lapse noise. Because plume detection is determined by the S/N, the ability to model realistic time-lapse noise is crucial in any feasibility study. In this work, we propose a more realistic approach by adding band-limited random noise to the pre-stack data (shot gathers) to match the S/N of field data. Using these noisy gathers we then compare the detectability of CO2 plume by using pre- and post-stack Kirchhoff migrations.