Application of FWI in Central Luconia, Malaysia

The Fxx field is located in the Central Luconia basin offshore Sarawak, with varying water depths from 80 to 95 meters. Hydrocarbon accumulations are found in clastic field at the shallower level as well as in the deeper carbonate build up. In the central part of the Fxx field, absorption and scattering by near surface channels appears to cause data degradation, highlighted by extraction of seismic amplitude (shown in Figure 1 left). This near surface gas- filled channel system, which leaked from the F field, have caused prominent poor, wipe out data zone. In this paper, we present a case study of the use of FWI and tomographic model updating workflow, with the aim to improve the seismic imaging quality of the wipe out zone. Although such technique is not new to address this issue, we wish to demonstrate that this processing solution has not been accessed in this area

Picking accurate shallow RMO is generally very difficult on conventional seismic because of the limited usable offset range and resulting low effective fold. To overcome this limitation we inverted for shallow velocities using full waveform inversion.

We performed simultaneous FWI inversion of velocity and eta. The FWI result suggested a low velocity anomaly at a location where we see an event with high amplitude in the migration section. This probably corresponds to a shallow gas in the channel body (see highlighted red circles in Figure 2 ). The bright amplitudes (low velocities due to gas) shown on the left map view in Figure 2 corresponds very well with the channel bodies whereby sand deposited on the edges.

After the FWI update, we merged the FWI model with the deeper model and proceeded with travel time tomography updating. Significant attention was given to the construction of a ‘geologically plausible’ velocity field in the deeper section. In the tomography update, the main geological interfaces are used as soft contrast to preserve the velocity break in the model. We have also included the dip information in the grid for the tomography inversion. An anisotropic model with incorporation of epsilon and delta was built in this manner.

The final FWI + tomography velocity model shows a network of shallow low velocity channels associated with gas that matches similar features in the reflection data. The resulting velocity model provides a better match to well logs, and better flattens migrated gathers, compared to the starting model. The final results will be shared, and the business impact will be discussed.