Time-lapse wave-equation migration velocity analysis

Abstract

Time-lapse analysis of seismic data acquired at different stages of hydrocarbon production or fluid/gas injection has been very successful at capturing detailed reservoir changes (e.g., pressure, saturation, fluid flow). Conventional 4D analysis is performed in the time domain assuming a constant baseline model; however, this procedure becomes difficult when the subsurface is significantly altered by production/injection and large time anomalies and complex 4D coda are recorded. We argue that a more robust 4D analysis procedure in these situations requires iterative wave-equation depth imaging and time-lapse velocity analysis. Wave-equation depth migration requires accurate knowledge of the velocity field usually obtained by one of two ways. First, data-space methods are where recorded data are matched to those calculated through a background velocity model. Differences between the two datasets are used in tomographic backprojections to generate velocity model updates. Alternatively, image-space methods are where discrepancies between migrated images (non-flat gathers) are backprojected to estimate velocity model updates. These types of approaches are termed migration velocity analysis (MVA). This abstract focuses on extending 3D wave-equation MVA (WEMVA) approaches to time-lapse velocity analysis. We discuss the differences between 3D and 4D WEMVA inversion goals, and how we leverage the locality of 4D image perturbations to provide high-resolution velocity model updates. We demonstrate the utility of 4D WEMVA analysis in a synthetic CO2 geosequestration experiment by successfully inverting for a velocity perturbation corresponding to a thin layer (<20m) of injected CO2 in a typical North Sea reservoir.