Specific time-lapse processing techniques are appropriate where two or more surveys have been acquired and a measurable time-lapse response is expected. The key to the technical success of 4D seismic applications lies in the repeatability and resolution of the seismic data. If the processed seismic datasets are not identical then the 4D signal can become masked by, or confused with, 4D noise. Historically, differences in the acquisition characteristics of baseline and monitor 3D surveys have been a significant source of this 4D noise.
It is not always possible to exactly repeat two seismic datasets, and often the surveys to be compared have been recorded with significantly different parameters. A carefully designed time-lapse seismic processing flow will help compensate for these variations, but a poorly chosen one will exacerbate them.
The primary objective of time-lapse seismic processing is, therefore, to maximize the repeatability of the datasets, while attaining sufficient temporal and spatial resolution for detection of the expected subsurface variations. The subsurface variations are normally changes in acoustic impedance, but may be an AVO effect, a time shift, or any other aspect of the seismic data. Combinations of responses can also occur. The expected time-lapse response must be borne in mind when developing the processing flow, and areas where production effects are likely to have occurred must be monitored to ensure that any genuine changes in the seismic response are preserved.