| Popis: |
We integrated pre-stack 3D seismic data with petrophysical, geological and production data to build a closed-loop reservoir model. All four data types were honored at the appropriate scale in their respective domains using a petrophysical seismic inversion workflow. Starting from an initial fine-scale geocellular model of the subsurface built using state of the art methods, we forward modeled it into synthetic seismic data through user-defined petro-elastic models, that is, a set of rock physics relationships used to compute compressional and shear velocities, and density. Reflection coefficients were calculated and convolved with the wavelet to obtain the forward modeled seismic response. The modeled seismic data were compared with the measured seismic data and the discrepancy was gradually minimized through a simulated annealing optimization. Additional perturbation of the initial model was performed through comparing modeled production performance with the measured production data. This perturbation/updating process stops when the modeled performance matches the measured performance data. The workflow was applied to a sand reservoir in the Gulf of Mexico. It is shown that the resulting product is a closed-loop reservoir model consistent with log, seismic and production data at their respective scales and domains. The workflow overcomes the many limitations of the linear "throw-over-the fence" approaches widely used in reservoir modeling. A potential expensive deep-water well was analyzed using precise delineation and evaluation of the reservoir and was found to have a high risk of failure so the well program was cancelled. |
