The classical Buckley-Leverett theory applies to 1D linear flow of two immiscible phases under the fundamental assumption that the flow rate (or total volumetric flux) is constant as a function of time. One phase is injected into the medium at a constant rate, thereby displacing the other phase. If the displacing phase is instead injected at constant pressure and the outlet pressure is also constant, the problem is still well-defined; however, the classical Buckley-Leverett theory does not apply. This is because the injected phase and the in-situ phase have different properties such as viscosities. If the boundary pressures are kept constant, the flow rate will, therefore, vary over time. The main result of this paper is to show that the solution of the displacement problem can be obtained from the constant-rate solution through an analytical determination of the flow rate as a function of time, given constant-pressure boundaries. The theory developed in this paper also provides an analytical solution for the location of the displacement front at any given time, the time for frontal breakthrough at the outlet end, and the pressure distribution as a function of time inside the medium.
It is demonstrated through computed examples that the constant-flow-rate solution, in general, cannot be used to approximate the corresponding solution for constant-pressure boundaries because the variation in the flow rate is very significant. A standard numerical method has also been applied and compared with the analytical solutions, demonstrating that fine numerical-simulation grids are required for acceptable comparison with the analytical solution.