Abstract
Dry-Out is one of the most critical phenomena in foam flooding, especially in Surfactant-Alternating-Gas. Commercial reservoir simulators are equipped with different Dry-Out models which have two main issues; first, even for the water saturation less than the limiting value, the current models are not able to dry out the foam thoroughly, second, the data from steady-state foam flooding experiments are required to estimate the model parameters. This study aims to address these issues in two separate parts. Firstly, it proposes a modified Dry-Out function to tackle the inadequacies of models in the vicinity of limiting water saturation without any discontinuity. It employs the error function to reveal the foam coalescence and cause the singularity in the current models to disappear. This model is verified with foam flooding experimental data. Secondly, the Genetic Algorithm and Nonlinear Least Square are used to develop the new approaches to estimate the Dry-Out parameters using the steady and unsteady state foam flooding experimental data, respectively. The set of steady and unsteady state foam flooding is conducted to validate the proposed approach for the model's parameters estimation. Eventually, foam flooding is numerically simulated to verify the modified model and parameters estimation approaches. The modified model indicates an acceptable fit with observed data compared with commercial simulators. The novel methodology for obtaining fitting parameters from unsteady state flooding yields precise parameters, thereby obviating the need for time-intensive steady state flooding experiments.