Intelligent Control Strategy Based on the Expert System for Zonal Water Injection in All Reservoir Intervals

During waterflooding development, zonal water injection is facing with production conflicts such as complex injection-production relationship and dynamic changes in the flooded field, and results in serious interlayer interference during interval adjustment. In this paper, based on real-time monitoring and automatic control in the whole process of zonal water injection with preset cable, an intelligent control strategy for zonal water injection into all well sections is proposed, which is based on the expert system. In other words, using expert knowledge and control experience, predictive values for adjusting the opening of the plugs in various intervals are established on a BP neural network prediction model; as a result, the accuracy of injection distribution into every interval can be controlled within 10%. The optimal openings of the plugs can improve the automatic adjustment efficiency in all well sections, reducing the driving load of the high-pressure and dynamically sealing mechanism of the electric control system of the downhole water distributor, making it stable and durable. Field experiments have been carried out on 6 wells, and the results show that, compared with traditional automatic interval adjustment method, this control strategy can increase waterflooding efficiency by more than 50% on average. The more layers are injected and controlled, the more effective the result will be.