Processing and imaging time-lapse datasets requires great care and attention to detail which is why our experience really makes a difference.

The aim of 4D processing is to attenuate the 4D noise caused by changes in acquisition parameters or environmental conditions, and to emphasise the 4D signature of the reservoir caused by changes in fluid, pressure and stress.

Some processes with particular application to 4D projects include:

Statics
Static differences between vintages occur in both land and marine 4D surveys. They are caused by changes in the near-surface, weathering layer, water column or water surface conditions. CGGVeritas has developed a range of strategies and algorithms for both land and marine 4D surveys including application of deterministic GPS-based tidal and waveheight corrections and statistical 4D corrections for water column statics.

Regularization
The regularization of 4D datasets uses bin centring and interpolation of missing data to create either regularly sampled datasets on a common grid, or map monitor survey data to a base survey grid. CGGVeritas has developed REVIVE, a Fourier regularization technique using up to 5 dimensions (x, y, offset, azimuth and time) which provides dramatic increases in the accuracy of the interpolation which benefits 4D processing.

4D Binning
4D binning is an essential step in time-lapse processing which selects the most compatible data for 4D processing. Traces are selected from all of the vintages making up the 4D dataset which are best matched in terms of source, receiver and midpoint position, offset and azimuth. The binning criteria can be extended to include statistical measurements of 4D data quality such as cross-correlation, predictability and NRMS.

NRMS repeatibility attribute map of 4D difference before 4D binning.	After 4D binning. The red stripes (indicating low repeatability due to differences in the acquisition) have been minimized.

De-striping
Systematic amplitude variations related to the acquisition geometry create a pattern or acquisition footprint. The footprint within each vintage is emphasised in the 4D difference. CGGVeritas utilises a 4D de-striping technique where data from all vintages are analysed for systematic variations with respect to acquisition attributes, such as sail line, and scalars are calculated to compensate for them.

Matching
Matching is used to minimize residual 4D differences caused by acquisition and environmental variations which cannot be addressed by deterministic corrections. A range of sophisticated multi-vintage matching techniques have been developed by CGGVeritas to suit a wide range of 4D datasets including the merging and 4D matching of streamer and OBC datasets. Other applications include the accurate estimation of time-shifts for geomechanical inversion.

Land and OBC
Land and OBC 4D surveys have additional processing needs and advantages. These can include multi-component processing and merging of streamer and OBC datasets in obstructed fields. With many of these surveys providing wide-azimuth datasets CGGVeritas can also bring to bear its unparalleled experience of wide-azimuth processing, including:

• Tau-px-py noise attenuation and deconvolution.
• 3D high-resolution Radon de-multiple.
• REVIVE 5D wide-azimuth interpolation.
• AGORA elastic modelling and subtraction of the direct arrival.