This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 201132, “The Future of Plunger Lift Control Using Artificial Intelligence,” by Ferdinand Hingerl and Brian Arnst, SPE, Ambyint, and David Cosby, SPE, Shale Tec, et al., prepared for the 2020 SPE Virtual Artificial Lift Conference and Exhibition - Americas, 10-12 November. The paper has not been peer reviewed.
Dozens of plunger lift control algorithms have been developed to account for different well conditions and optimization protocols. However, challenges exist that prevent optimization at scale. To address these challenges, a plunger lift optimization software was developed. One aspect of this software is enabling set-point optimization at scale. This paper will present the methodology to do so, detailing three separate areas working in unison to offer significant value to plunger lift well operators.
Introduction
Even in vertical wells, plunger lift presents significant challenges to optimization. Despite their mechanical simplicity, plunger lifted wells produce large amounts of data, and the combinations of possible set points to optimize the well are many. Additionally, plunger lift wells can present a variety of different types of anomalies and problems that require a robust understanding of the underlying physics and mathematics.
Such problems then are amplified when applied to horizontal well applications. The underlying physics and mathematics applied throughout the years for vertical wells do not produce accurate results for horizontal wells. Additionally, the anomalies produced in horizontal wells are more complex. Finally, typical production engineers and well optimizers now regularly look after more than 150—and often more than 500—wells, creating additional resource constraints to optimizing a field of plunger lift wells.
The presented plunger lift optimization software was implemented by creating a secure connection between the operator’s supervisory control and data acquisition (SCADA) network and the cloud. As new data are generated by the SCADA network, they are automatically transmitted to the cloud and processed.
Plunger Lift Control Algorithm Overview
These algorithms are the software code that determines when the well opens and when the well closes. Even though the algorithms only control well open/close, the plunger moves through four stages of plunger operation to complete one cycle: plunger fall time, casing pressure build time, plunger rise, and after flow (or production). Optimizing each individual stage is critical to ideal well production.
Plunger fall time is the time required for the plunger to descend from the lubricator to the bottomhole assembly (BHA). Currently, operators use the manufacturer’s anticipated fall time, trial and error, previous knowledge, acoustical plunger tracking, and plunger fall applications to set the appropriate fall time in the controller. A “fudge factor” is often applied to help ensure that the fall timer does not expire before the plunger reaches the BHA. Plunger fall time is affected by many changing variables: plunger condition, tubing condition, liquid height, and gas and liquid density. These variables make it difficult for a fall timer set once to represent accurately the correct time required for the plunger to reach the BHA on every cycle.
The Future of Adaptive Tutoring: Wrangling Complexity across Domains, Applications, and Platforms
