Seismic Acquisition for 4D Monitoring
In this chapter, we will discuss how to decide and honor the acquisition requirements for good 4D monitoring. Let us assume that we think we know what production-induced seismic changes we are looking for, that we know what we want to measure, and that we have an estimate of the size of changes we are looking for. We saw some ways of doing this in Chapter 3.
Those requirements should be translated into seismic acquisition specifications. In the early days, the simple and direct approach was to make synthetic seismograms of the changes expected, add noise as measured or judged from previous data in the area, and then see whether we could expect to see those changes. If we thought we would see the expected changes, we could shoot another survey over the field. Just to pass economic hurdles, the second survey often had a dual objective, such as enhanced resolution.
Sometimes this worked and sometimes it did not. We found that poststack matching of dissimilar surveys appeared to work only if the shooting direction of the surveys was the same. If we were meticulous and really tried to repeat the previous survey parameters, we were surprised to find that the 4D differences often were better than expected. After reading Chapter 4, on repeatability, we can understand this. If we repeat the surveys as closely as possible, much of the “noise” repeats and the differences are indeed better because the repeating noise is differenced away.