Since prediction of shallow gas is essentially a prediction of pore fill, we have chosed to focus on seismic parameters which are significantly affected by a change in pore fill. These parameters are: Seismic velocity, attenuation, amplitude versus off-set and seismic response. To study these parameters, different techniques are required, and in this paper we apply the techniques on real data for shallow gas prediction. If all the techniques described could be applied at the same site, the result should be a reliable prediction. The attenuation of seismic energy is a promising parameter, but it requires that refracted waves are recorded. The amplitude versus offset is expected to be a powerful prediction technique in the future, but at the present time more development is needed. This technique also has the potential of estimating the saturation under favourable conditions. Single trace inversion is also a promising technique, and with further development and linked to velocity modelling it will become a powerful tool in shallow gas prediction. Direct use of seismic velocities in shallow gas prediction requires refracted arrivals. These data are of special importance where no borehole data are available. All techniques will benefit from borehole data, and they are absolutely necessary for a quantitative estimation of relevant parameters.
The prediction of shallow gas is an important task which will make drilling operations safer and can also contribute to a reduced rig time. The most important thing is to be able to predict this shallow gas with reliability so that blow-out can be avoided in any drilling operation.
The strategy for shallow gas prediction from seismic data will change from place to place depending on the water depth, the depth of the gas pocket below the sea floor, the velocity gradient from the sea floor to the gas pocket and the length of the seismic streamer.
The most significant seismic parameters which can be applied in the prediction of shallow gas are the seismic velocity, the seismic response and the absorption of seismic energy. In this paper we will demonstrate how these parameters can be analysed and used in connection with real data. The procedures are simple, and since they are based on the most fundamental parameters in seismics, they happen to work in practice
The seismic velocity is amongst other factors dependent on the pore fill. Figure 1 shows the seismic velocities as a function of porosity for a shallow sand which is water and gas saturated respectively. As seen from the figure, the P-wave velocity is most strongly affected by the pore fill. This is mainly due to the great difference between the incompressibility for water and gas. Figure 2 shows how the incompressibility or bulk modulus changes with gas saturation in the pore fluid. The difference in S-wave velocity for a water and gas saturated sand (see Figure 1) 1s due to the difference in bulk density.
It may be difficult, or impossible, to estimate the seismic velocity from reflection data with sufficient accuracy to be used as an indicator for shallow gas. Shallow gas is frequently associated with high amplitude anomalies in the seismic section. The problem in reflection seismology is to find a relevant reflector below this anomaly with sufficient response to obtain a good estimate of the interval velocity in the anomalous zone.
For shallow depths, refracted arrivals are superior to reflected arrivals ~n detailed velocity analysis. Berge et al.(1983,1985) have demonstrated the practical use of refracted arrivals in marine seismic data. In this paper we will demonstrate some additional proce