Schlumberger

Time-Lapse Seismic Processing

Optimal data quality for reservoir monitoring


A suite of algorithms and workflows supports our processing philosophy.

  • Flexible attribute decomposition algorithms enable us to identify and compensate for survey variability that is consistent over a particular acquisition attribute. This is frequently used, for example, to simultaneously decompose horizon time picks from multiple surveys into sail-line-consistent timing corrections relative to a background trend.
  • Normalized root mean square (NRMS) scanning has capabilities similar to attribute decomposition, but it operates poststack and is therefore less susceptible to noise. An example of its use is to identify errors in source-to-detector distances.
  • Spatial repositioning uses spatial crosscorrelation to estimate spatial positioning differences between surveys. The results can be averaged over a survey, sail-line direction, or sail-line number and applied as corrections.
  • Constrained crossequalization expresses the amplitude and phase difference between two datasets in terms of a constant amplitude scaler, time shift, and phase rotation. The crossequalization process can be applied in a survey-averaged manner, trace by trace (with appropriate smoothing), or varying with a particular acquisition attribute.
  • Frequency-variant matching designs and applies frequency-variant matching operators and has the same design and application options as constrained crossequalization. Additionally, the bandwidth over which the matching occurs can be controlled so that it only addresses the frequency range in which the datasets are poorly matched.
  • Nonrigid matching computes and applies sample-by-sample time- and space-variant time shifts between a pair of time-lapse 3D cubes. Displacements in the inline and crossline directions can also be corrected.
  • Time-lapse binning allows selection of the best repeated trace pairs from two or more surveys. Any geometric attribute may be used, together with data-based attributes such as NRMS difference and predictability. The process can search in adjacent common midpoint (CMP) or offset bins for the best matched traces.
  • Many processes that rely on optimization procedures can be parameterized so that data from one survey can be used to define the starting point of the optimization of another survey, improving consistency of operation. This is true, for example, of the matching pursuit Fourier interpolation (MPFI) regularization and interpolation procedure that allows the MPFI results on one survey to guide the interpolation of another.

Global QC 4D Attributes

PrevNextZoom1 of 1