Using Finite-Difference Modelling To Understand And Optimise Time-Lapse Acquisition And Processing Strategies

Throughout the recent challenges in the marine seismic industry, both time-lapse and ocean-bottom acquisition programs have remained robust. As the industry requirement for these techniques increases, it is important that we fully understand if the survey we acquire will overcome the specific challenges for the given field and provide the data and insight required by the end users. Recent advances in model building, forward-modeling algorithms, and compute architecture, coupled with the depth of knowledge and information available on the areas being considered for time-lapse acquisition programs, allow us to run efficient and accurate forward simulations to answer these design questions and benchmark our time-lapse processing workflows. Here, we present results from a forward-looking finite-difference modelling exercise that uses a synthetic visco-elastic model. As we learn what changes occur within the reservoir and overburden, we can assess the impact of likely acquisition variables on our ability to resolve the true 4D signal. Furthermore, we can look at implications of acquisition geometry variations on seismic processing steps in isolation, which is simply not possible with real data.