A New Workflow for Better 4D Seismic Interpretation

4D seismic images are often replete with artifacts of the time-lapse methodology, unrelated to actual changes in properties within the reservoir. These artifacts introduce uncertainty into the interpretation of flow fronts and pressurization and can greatly diminish the value of the 4D data. In many cases the artifacts originate from seismic signal and noise which cannot be, but, ideally should be exactly repeated from one survey to the next. Practical limitations of seismic acquisition and processing technology inherently limit data repeatability. Current 4D workflows rely heavily upon seismic repeatability. Current workflows utilize subtraction of data (or images) from repeated surveys to localize changes in properties to the time interval between surveys. These subtraction-based methodologies therefore assume that repeated signal and noise will subtract to zero provided subsurface properties have not changed over time. One can relax repeatability requirements by direct inversion for the changes in properties between two surveys. This new 4D workflow avoids subtraction altogether and can greatly reduce artifacts and uncertainty in the resulting images. Improvement is particularly evident for time-lapse relative to the pre-production conditions of the reservoir. These pre-production properties have frequently been recorded through a legacy seismic survey with particularly poor repeatability.