document preview

Abstract
As an important parameter of multiphase flow in porous media, relative permeability curve is often calculated by JBN method which ignores the capillary pressure. However, owing to the fine throat and the complex construct in extra low permeability reservoir, the capillary pressure is very significant. In addition, researches have shown that the flow in low permeability media doesn’t follow the Darcy’s law and there is threshold pressure gradient occurring in the percolation. Therefore, the traditional method is no longer available for the extra low permeability reservoirs and a new method is needed urgently to resolve this problem.

In this paper, the relative permeability of extra-low permeability core was calculated by three methods: JBN method, fractal model and modified JBN method. The fractal model was set up based on the capillary pressure curve and the fractal geometry theory. The modified JBN method taken into the effects of capillary pressure and threshold pressure gradient, and a semi-log fit function other than polynomial function was used to improve the accuracy.

The results show that, comparing with JBN method, the fractal model and the modified JBN method can predict the relative permeability more accurately. And the modified JBN method is the best. The water cut calculated by the new method matched the production data better than that calculated by JBN method.

The new method overcomes the limitations of previous researches. It can be used to calculate relative permeability for extra low permeability reservoirs more accurately and help to develop extra low permeability reservoirs more reasonably.

Introduction
Relative permeability is an important parameter of multiphase flow in porous media, which is the required parameter for the plan of oilfield development, dynamic prediction and numerical simulation. The relative permeability curve can predict oil production rate, ultimate recovery and water cut. At present, the measure methods of relative permeability are usually the followings: steady method; unsteady method; capillary pressure method; centrifugal method and production data method (Firoozabadi, A., Aziz, K. 1988, Li Kewen et al. 1989, E.Ebeltoft et al. 1998, C. Wouter Botermans et al. 2001, Janos Toth et al. 2002). The fore two methods are usually used in laboratory researches.

In steady method, the fluid was injected as a fixed ratio until the equilibrium state of saturation and pressure was established. The key focus in this experiment is to remove or reduce the saturation gradient caused by the capillary pressure effect. There are several methods: Penn. State method, single core method, immobilizing fluid method, Hassler method, Hafford method and disperse injection method. Steady method is the most direct method and is the most easy to explain method. But it is difficult to get the relative permeability curve with continuous saturation distribution and it is only to get one point one time. Therefore, it is very time-consuming. In addition, the development effect predicted by the data gotten from steady method may cause very much error because the saturation gradient is changing continuously during oilfield development.

Unsteady method is that the core was saturated by one fluid and then it is displaced by another fluid. Welge, Efros, Johnson, Jones develop and improve this method, and the JBN method was created, which is extend to three-phase relative permeability calculation.