Hybrid Integral Transform Solution of a Nonlinear Core-Flooding Model with Capillary Pressure and End Effects

ATZelektronik · Springer Nature · 2026-09-03 · DOI: 10.1007/s11242-026-02351-z

Аннотация

Oil–water displacement in core-flooding experiments is governed by a nonlinear convection–diffusion equation derived from the extended Buckley–Leverett formulation, including saturation-dependent relative permeability and capillary pressure effects. When realistic constitutive relations are combined with nonlinear capillary end-effect boundary conditions, such as those proposed by Chang and Yortsos, the problem becomes particularly challenging because of the strong coupling among saturation, capillary pressure gradients, and phase fluxes at the inlet and outlet boundaries. In this work, a hybrid numerical-analytical solution based on the generalized integral transform technique (GITT) is developed for one-dimensional immiscible displacement with nonlinear Chang-Yortsos boundary conditions. A problem-adapted filtering procedure incorporates the available explicit integral and boundary information, while the nonlinear boundary flux contributions are retained explicitly in the transformed equations through the boundary terms generated through integration by parts. Because the physical outlet condition changes at breakthrough, two auxiliary Sturm–Liouville bases are employed, and the complete pre-breakthrough saturation field is transferred to the post-breakthrough representation through an orthogonal L 2 projection. The nonlinear transformed coefficients are evaluated using a semi-analytical procedure that combines piecewise-linear spatial approximation with closed-form eigenfunction integrals. The resulting nonlinear ODE systems are numerically integrated with the open-source CVODE/SUNDIALS solver and independently cross-verified using DIVPAG/IMSL. Norm-based convergence analyses against a high-resolution reference solution, besides comparisons with MOL/FVM finite volume scheme and SCORES finite difference results, demonstrate the accurate predictions of saturation profiles, pressure drops, production quantities, and local constitutive fields. At reasonable truncation orders, the GITT algorithm exhibits computational costs comparable to those of the optimized finite-volume implementation, while providing a controlled semi-analytical modal representation of the complete transient solution.

Читать у издателя →

Похожие статьи