Velocity Modeling in Gas Sagging Area

Abstract

Velocity model provides direct linkage between time domain and depth domain. Time to depth conversion of interpreted surfaces, faults and seismic data is an important part in the 3D geological modeling for volumetric assessment. The data used for velocity modeling may include seismic velocity data, well velocities (from checkshot, VSP, and/or synthetic seismograms), well markers, and surfaces. Conventionally geophysicist uses checkshot data, stacking velocity, or checkshot integrated with stacking velocity directly. These methods are widely used for depth conversion in a normal area (i.e. without gas sag or pull-up effect in surface seismic interpretation). In this project, the existence of shallow gas cloud in the area of study is causing sag effect instead of anticlinal structure on the seismic data. The common solution for the gas sag is manually interpreting the anticline at the affected area. However, manual seismic horizons interpretation to correct the sag to anticline is not applicable for this project as seismic inversion work is involved. The manual correction in seismic horizons interpretation will introduce an artifact in the seismic inversion process. Thus the seismic horizons interpretation has to honor the sag effect. The depth conversion from the conventional velocity modeling method above would not able to solve the gas sag effect, where the sag exists instead of anticline in the depth domain. A simple but effective workflow is introduced to overcome the problem where the seismic velocity was corrected first, and then calibrated with the well velocity. Considering a point - ‘X’, which is unaffected by shallow gas beside the gas sag is chosen as reference for time and velocity (TWT difference factor derivation and seismic velocity correction). A TWT difference factor is derived between the point ‘X’ and gas sag. Then the seismic velocity is corrected with the TWT difference factor, thus remove unreliable velocity within the gas sag area. This corrected seismic velocity is then calibrated with well velocity and the latter result is used as an input for velocity modeling. The structural dip of depth converted structure surfaces are QCed with average dip from the OBMI to ensure no anomaly on dip changes. The residual depth of the velocity model is within 10m.