Optimal waterflooding management using an embedded predictive analytical model

Abstract

In the petroleum industry, there is an ever-increasing interest in oil recovery processes with high hydrocarbon extraction rates. One of the most common oil recovery processes is waterflooding, which involves the injection of water into a reservoir. This process is often challenging, as there is uncertainty in the reservoir’s properties. In this paper, we propose an optimal waterflooding management methodology for setting the producer and injector wells conditions to maximize the net present value (NPV). Our methodology integrates a predictive analytical model, which models the reservoir performance and forecasts the production rates based on the producer and injector well operating conditions. We applied the methodology in an illustrative example with rock and fluid properties representative from a Colombian oil field. We compared the proposed methodology against two benchmark scenarios in which the bottom-hole pressure of the producer wells is kept low and constant over time, as it is a common operational practice in the oil fields. With the proposed methodology, the profit increases by 27.02%, the oil recovery factor increases by 12%, and the water cut reduces by 4%, over a ten-year planning horizon.