Optimizing power fluid in jet pump oil wells

Abstract

A method for optimizing power fluid to a network of jet pump wells is established. Jet pump performance is modeled by numerically solving a system of equations for specific throat and nozzle geometries. An upper boundary is established of the most efficient geometries for each well which creates a continuous function relating power fluid to oil production. Jet pump oil wells are segregated into networks which share a common power fluid surface pump. These boundaries are added together to create a non-linear objective function. To solve power fluid distribution in a network, a reduced Newton method is applied that incorporates active constraints. System constraints are that the total network power fluid is at or below surface pump capacity and that each power fluid rate is non-negative. Upon successful convergence, the power fluid estimate per well is passed to a discrete algorithm to choose between either a high or low power fluid jet pump. A computer program is developed capable of implementing the optimization method. This program is successfully tested on an eight well network, determining whether an additional well can be supported with existing equipment. The continuous optimization is fast, converging in four iterations to an answer. Any engineer can run this program, providing the benefit of a unified approach to decision making. This is a significant improvement, since no previous methods have been found in literature on how to distribute power fluid across a jet pump network.