RESULTS

Depth migration results from Biondi et al. 1996 are shown in Figure 3. These correspond to an in-line slice through the model at Y=2.27 km, which intersects two water injection wells.

The left panels are zero-offset and the right panels have a source-receiver offset of 3000 m. The top row corresponds to t=0, the middle row corresponds to t=3 years and the bottom row is the difference between the two.

At t=0, a gas cap extends from the top of the reservoir to approximately 1.72 km depth. After three years, the water injectors cause the gas cap to be unevenly depleted. This effect is clearly visible in the difference sections.

 

results
Figure 3 Migrated synthetic seismograms from a heterogeneous reservoir model. The left panels are zero-offset and the right panels have a source-receiver offset of 3000 m. The top row corresponds to t=0, the middle row corresponds to t=3 years and the bottom row is the difference between the two.

With the insight gained from the plane layer studies above, meaningful, but qualitative, comparisons between the two offsets can be made. The gain applied to the results from both zero- and far-offset surveys was the same, enabling direct comparisons be made between the two.

The clearest differences between the surveys at different offsets are that the large-offset sections have lower resolution, but larger amplitudes than the zero-offset sections. The decrease in resolution is an inherent effect of the migration operator. The increase in amplitudes is due to the fact that the reservoir is mostly oil/gas saturated, and the previous results show that the reflection coeffiecient for both the oil-sands and the gas-sands increase with offset.

The most noticable difference between the two `difference' seismograms, is that the change in amplitude due to the water injection and disappearance of the gas cap is much larger at large offset than in the near-offset sesctions. This highlights the result from the plane layer-modeling that the large-offset section is most sensitive to the presence of gas.

Other differences between the two offsets are harder to interpret; although an `AVO cube' may help this process. This shows some of the difficulties in drawing conclusions from variation in amplitude with offset. Heterogeneities in the subsurface are imaged with different resolution for different offset, and even with the simple lithologies that we have created, it is difficult to identify the source of anomalies.