14:55 〜 15:15
[SE-06] Investigation of the influence of anisotropic velocity errors on dipping reflector imaging
Anisotropic pre-stack depth migration (A-PSDM) is commonly used for recent seismic imaging. Still, seismic interpretation is often performed in time domain for avoiding instability associated with seismic velocities. Time-to-depth conversion is then conducted using the “best practice” vertical velocity model.
Our recent case study in the North Sea shows nontrivial positioning inconsistency in time domain when comparing legacy A-PSDM with newly acquired anisotropic pre-stack time migration (A-PSTM). The inconsistency induces a significant impact on geological evaluation and volume estimation of prospects.
To investigate the origin of positioning inconsistency, we conducted finite-difference simulation with a three-layer model that represents a geological configuration in the North Sea. The model is characterized by a horizontally layered overburden above a dipping target reflector at 3 km depth. We analyzed sensitivity of the anisotropic velocity parameters ranging within a practical resolution of moveout-base velocity analysis.
Synthetic tests demonstrate that errors in interval vertical velocity (V0) or δ (corresponding to errors in effective Vnmo for A-PSTM) cause substantial distortion in time-domain dipping reflector positioning. The magnitude of misposition increases with the error in V0 or δ and with dip of the reflector. For example, the misposition reaches 32 milliseconds vertically and 105 m laterally for a 30-degree reflector when the error of interval δ at the 3rd layer is 0.33 (just 5% overestimates in effective Vnmo), but such misposition is negligible for a horizontal reflector. Instead, errors in interval ε do not induce significant misposition.
Our simulation results demonstrate that positioning of the dipping reflector is significantly influenced by errors in V0 or δ and hence by effective Vnmo. Such errors can be easily produced by ambiguity of nonhyperbolic moveout inversion or anisotropic travel-time tomography. It is practically important to optimize an anisotropic velocity model using available well data for both A-PSDM and A-PSTM.
Our recent case study in the North Sea shows nontrivial positioning inconsistency in time domain when comparing legacy A-PSDM with newly acquired anisotropic pre-stack time migration (A-PSTM). The inconsistency induces a significant impact on geological evaluation and volume estimation of prospects.
To investigate the origin of positioning inconsistency, we conducted finite-difference simulation with a three-layer model that represents a geological configuration in the North Sea. The model is characterized by a horizontally layered overburden above a dipping target reflector at 3 km depth. We analyzed sensitivity of the anisotropic velocity parameters ranging within a practical resolution of moveout-base velocity analysis.
Synthetic tests demonstrate that errors in interval vertical velocity (V0) or δ (corresponding to errors in effective Vnmo for A-PSTM) cause substantial distortion in time-domain dipping reflector positioning. The magnitude of misposition increases with the error in V0 or δ and with dip of the reflector. For example, the misposition reaches 32 milliseconds vertically and 105 m laterally for a 30-degree reflector when the error of interval δ at the 3rd layer is 0.33 (just 5% overestimates in effective Vnmo), but such misposition is negligible for a horizontal reflector. Instead, errors in interval ε do not induce significant misposition.
Our simulation results demonstrate that positioning of the dipping reflector is significantly influenced by errors in V0 or δ and hence by effective Vnmo. Such errors can be easily produced by ambiguity of nonhyperbolic moveout inversion or anisotropic travel-time tomography. It is practically important to optimize an anisotropic velocity model using available well data for both A-PSDM and A-PSTM.
抄録パスワード認証
抄録の閲覧にはパスワードが必要です。パスワードを入力して認証してください。