Interference between simultaneously shooting seismic vessels is a problem in areas with high exploration activity If seismic interference exceeds the noise limits set to assure data quality during acquisition, the interfering vessels usually commence time sharing It is a widely held opinion within the industry that limits based on random noise levels may cause unnecessary downtime when uncritically applied to seismic interference (SI) Revised noise limits for SI have been proposed These have taken into account both the dip of the noise and the amplitude level These limits have been based upon the assumption that the noise is to be removed by dip-filtering on shot records This will limit their application It will be demonstrated how to decompose the interference noise into elements through which the level of contamination of the stack can be judged On this basis a revised noise limit for SI that is independent of the filtering methods is proposed.
The noise limits were tested by using ‘synthetic’ data, where pre-recorded seismic interference noise given the desired characteristics were superimposed on a ‘noise-free’ testline.
Seismic interference noise appears on the shot records as a coherent wave train with characteristic rate of repetition and duration The interference signal also has a stable RMS level and moveout which only slowly varies along the profile Interference noise is due to energy travelling as shallow refractions and mode propagation in the water column (Calvert et al 1984) SI noise may carry over long distances In the extreme case, vessels at more than 100 km distance have been subject to timesharing operations.
The effect of the interference noise on the data quality depends on the following four parameters
Relative moveout of the incoming noise from trace to trace over the shot record This is decided by the interfering vessel's relative position and movements
The recurrent frequency of the SI on the recordings This reflects the shot interval of the interfering vessel
The RMS noise level and duration These depend on the vessel's separation, size of airguns, water depth, seabed geology, seabed topography and sea state
Fig. 1 Seismic Interference contaminated shot-records (Available in full paper)
The effect on the data of these parameters can be studied separately. Their effect can to some extent be quantified, as will be detailed in subsequent setting
One shot records, SI appears as coherent signals with a constant moveout (Fig 1)On the CP-gather, the SI is mapped as apparent randomly spaced high amplitude wave trains (Fig 2) The extent to which SI will be suppressed in the stack is therefore largely independent of its initial moveout. One the stacked section the SI will reappear as stripes with dips depending on the initial moveout plus the stacking velocity function (Fig 3)
Fig. 2 NMO-corrected CDPs with interference noise (Available in full paper)
Fig. 3 Stacked section of the testline contaminated with SI, seen as dipping event on the lower part of section 25-mlcrobar SI noise level (Available in full paper)