Estimation of two-phase petroleum reservoir properties by regularization

Abstract

An algorithm is developed, based on the theory of regularization and on spline approximation, to estimate the absolute permeability in two-phase petroleum reservoirs from noisy well pressure data. The regularization feature of the algorithm converts the ill-posed estimation problem to a well-posed one. The algorithm, which employs the partial conjugate gradient method of Nazareth as its core minimization technique, automatically chooses the regularization parameter based on the non-regularized estimation. It is shown that regularized estimation is more stable and insensitive to the choice of initial guess as compared to non-regularized conventional estimation.

Additional details