Dynamic Pumping Unit Control Using Variable Frequency Drives

Through the motion of a rod pump well, stress oscillations typically appear in the rod string at the beginning of both the upstroke and downstroke phases. This dynamic phenomenon has several adverse consequences on the well. The load in the road string is drastically increased, thus reducing its service life; the plunger velocity is higher thus increasing erosion and wear on the pumps and the stress on the gearbox and the pumping unit also increases general wear and tear.

It is well known that reducing a unit Stroke-per-Minute (SPM) will reduce the severity of the stress oscillations, at the cost of production. Simple, once-per-stroke, intra-stroke speed changes are used today on long-stroke rotaflex units to reduce equipment wear at the top and bottom of the stroke, and more rarely in wells with gas issues to minimize the pounding buckling effect on the rod string during the downstroke motion of a pumping unit. 

In this paper, we present how multiple dynamic intra-stroke motor speed adjustments can reduce stress or increase production. We also show how the motor speed can be automatically computed to obtain a system that dynamically adapts to any well.
Examples of reducing the stress in the system while maintaining or increasing production are shown on multiple wells in Oklahoma and in the Permian basin. We show how a theoretical control model was developed, and the results of its implementation through AI models running a Variable-Frequency-Drive (VFD) via a machine-to-machine connection.

This paper shows a real-world example of how AI can be used to build flexible well control models which bring drastic positive outcomes. The result is a system that can be used on any rod-pumped VFD-powered well and will deliver optimal production at minimal wear.