A flow simulation-driven time-lapse seismic feasibility study is performed for the Amberjack field that leverages existing multi-vintage 4D time-lapse seismic data. The focus is a field consisting of stacked shelf and deepwater reservoir sands situated in the Gulf of Mexico in Mississippi Canyon Block 109 in 1,030 ft of water. The solution leverages seismic interpretation, seismic inversion, earth modeling, and reservoir simulation [including embedded petro-elastic modeling (PEM) capabilities] to enable the reconciliation of data across multiple seismic vintages and forecast the optimal future seismic survey acquisition in a closed-loop. The overarching feasibility solution is integrated and simulation-driven involving multi-vintage seismic inversion, spatially constraining the petrophysical property model by seismic inversion, and performing reservoir simulation with the embedded PEM. The PEM is used to compute P-impedance and Vp/Vs dynamically, which enables tuning to both historical production and multi-vintage seismic data. The process considers a hybrid fine-scale 3D geocellular model in which the only upscaling of petrophysical properties occurs when the P-impedance from seismic inversion is blocked to the 3D geocellular grid. This process minimizes resampling errors and promotes direct tuning of the simulator response with registered seismic that has been blocked to a geocellular earth model grid. The results illustrate a three-part simulation-to-seismic calibration procedure that culminates with a prediction step which leads to a simulation-proposed time-lapse seismic acquisition timeline that is consistent with the calibrated reservoir simulation model. The first calibration tunes the model to historical production profiles. The second calibration reconciles the dynamic P-impedance estimate of the simulated shallow reservoir with that of the seismic inversion blocked to the 3D geocellular grid. The combination of these two steps outline a seismic-driven history matching process whereby the simulation model is not only consistent with production data but also the subsurface geologic and fluid saturation description. Large and short wavelength disparities in the P-impedance calibration existing between the simulator response and the time-lapse seismic data are attributed to resampling errors as a result of seismic inversion-derived P-impedance being blocked to the 3D geocelluar grid, as well as sparse well control in the earth model which leads to the obscuring of some asset-specific characteristics. The results of the third calibration step show how the time-lapse seismic feasibility solution accurately confirms prior seismic surveys undertaken in the asset. Given this confirmation, the solution achieves a suitable prediction of seismic-derived rock property response from the reservoir simulator as well as the optimal future time-lapse seismic acquisition time.