Semi-Analytical Method for Accurate Calculation of Well Injectivity During Hot Water Injection for Heavy Oil Recovery

Abstract

Representation of wells in numerical simulation of petroleum reservoirs is a challenging task due to large difference in typical scales of grid blocks (tens to hundreds meters) and wells (~0.1 m), with high pressure and saturation gradients around wells. Although a variety of grid refinement techniques can be used for local simulations, they have limited application in field-scale problems due to huge model dimensions. Thus, auxiliary quasi-stationary local solutions (so-called inflow performance relations) are used to relate well flow rate with well and grid block pressures. These auxiliary solutions are strictly derived for linear cases and generalized to non-linear problems by using grid-block averaged values of fluid and reservoir properties. In the case of hot water injection for heavy oil recovery, this results in significant errors in well injectivity calculations due to large temperature and saturation gradients dynamically influencing viscosity and relative permeability distributions around the well. In this paper we propose a method which combines a semi-analytical solution of the hyperbolic Entov-Zazovsky problem for non-isothermal oil displacement with integration for pressure distribution taking into account nonlinear dependencies of fluid viscosities and relative permeabilities on temperature and saturations. Both constant injection rate and constant well pressure cases are considered. Example calculations demonstrate that the method helps to avoid underestimation of well injectivity in non-isothermal problems caused by grid-block averaging of fluid and reservoir properties in conventional inflow performance relations.