Integrating Static and Dynamic Data for Calibration of Stratigraphic Forward Modeling

Traditional reservoir models use geostatistical techniques but often oversimplify geological features. Stratigraphic Forward Modeling (SFM) offers a physics-based alternative but is challenging to calibrate with field data. We propose a multi-objective calibration workflow integrating static geological and dynamic production data for better calibration.

The methodology balances static geological fidelity with dynamic behavior of the field using a multi-objective optimization framework. Static calibration employs a novel facies comparison function inspired by well-log correlation techniques. Dynamic calibration involves converting SFM outputs into petrophysical properties and refining these transformations using regional-scale pressure and connectivity data.

To overcome computational challenges, we use surrogate modeling via dimensionality reduction and Gaussian Process Regression, reducing the number of required simulations and improving optimization efficiency. The result is a set of optimal solutions, representing multiple best-fit models that vary based on the relative weight given to static versus dynamic data. The workflow was tested on a Brazilian Aptian carbonate reservoir of the Campos Basin.

By integrating static and dynamic data within SFM, this approach enhances reservoir modeling accuracy while maintaining geological integrity. It bridges the gap between geologists and engineers, enabling more reliable predictions and improved decision-making in reservoir development.