Integrating production expertise with digital capabilities drives next-generation autonomous well control solutions

Onshore horizontal resource plays have propelled U.S. oil, natural gas, and natural gas liquids output to historic highs. At the same time, however, the shale revolution has ushered in new types of production challenges, ranging from fracture-driven well interactions to upsized operational scales and prototypical steep declines. It is a whole new game, and even after two decades of developing best production practices, the industry continues to learn, adapt and innovate. 

According to various sources, typical well decline curves in tight oil plays are estimated to average approximately 70% to 90% within the first 36 months of commencing production. The decline can resemble a cliff more than a curve in the earliest phases of the well lifecycle. The rate of decline slows in time, aided by artificial lift, but the dramatic changes in daily output—and the speed and magnitude with which they occur—make shale plays a highly dynamic operating environment. Constantly changing bottomhole conditions make it difficult to optimize production on a sustained basis on individual wells, let alone across a pad or field. 

That dynamism is reflected in U.S. Energy Information Administration well count data, which indicate that half of total U.S. oil production since 2013—coinciding with the rise of Permian basin horizontal plays—comes from wells producing between 100 and 3,200 boed. That is an exceptionally wide range, encompassing wells on both the high and low ends of the spectrum. They require quite different operating strategies and artificial lift systems. Whether a well has been online for 10 months or 10 years, innovation is the key that enables operators to unlock well productivity and maximize cash flow generation across all phases of the production lifecycle. 

Managing wells with rapidly changing production profiles is challenging the oil and gas industry like never before. Compounding this challenge are several interrelated business-side drivers that mandate increased operational efficiencies, focused capital discipline, ESG attainment and all types of risk exposure mitigation—from personnel safety to supply chain surety. As a result, operators are trying to do more with less and maximize ROI in every producing asset. 

For artificial lift, this translates into safely and efficiently maximizing system run times and bottom-line measurables with minimal unplanned opex outlays. Digital technology to automate, monitor and control wellsite operations is on the critical path to accomplishing these goals. Given the high-impact capabilities built into today’s digital hardware and software platforms, and the unpredictability of unconventional wells, it’s not a surprise that automation has become standard procedure in horizontal resource plays. Roughly eight in 10 shale wells are now equipped with some form of automated production equipment. 

REAL-WORLD PROOF 

Producers are taking different approaches to automating their assets, but there is no doubt that companies of all sizes view digital solutions as a significant part of the answer to the question of how they can do more–and do it better–with less. But the time for what I call “digital buzzword bingo” is over. Operators no longer want to hear about the promise of artificial intelligence, neural networks, machine learning, cloud computing, deep learning, internet of things, etc. Operators are ready for demonstrable, real-world proof. They want to see a clear upside in investing in digital tools in operations where differentiators are traditionally scored by metrics, such as daily production rate, LOE/barrel (bbl) and run time. The value is not in the buzzwords or the digital tools themselves, but in how they are applied within the context of relevant technical expertise and operational knowledge to raise the production performance standard across the board, Fig. 1. 

Fig. 1. In artificial lift applications, the value of digital tools is determined by how the best available digital capabilities can be leveraged within the context of relevant technical expertise and operational knowledge to improve daily rates, LOE/bbl, run times and other critical measures of production performance.

Let’s expound on that by viewing a specialized knowledge-based digital solution for optimizing the performance of, say, a sports car rather than an oil or gas well. The car’s owner can frequent any number of repair shops. But wouldn’t it wise to visit the one with model-specific software, tools and expertise to correctly diagnose problems and take appropriate mechanical or operational actions to remedy them? It's not a stretch to say a producing well is even more engineered and operationally complex than a sports car (and more valuable). Like the car’s owner, the well operator seeks to protect the investment and net present value by heading off potential issues that may degrade performance or lead to operational breakdowns. 

 Accomplishing that requires framing the best available digital capabilities around the best available expertise specific to the oil and gas production domain. Consequently, ChampionX’s goal is to develop digital offerings differentiated by deep technical proficiencies in engineering, optimizing and troubleshooting all types of artificial lift, as well as in developing production optimization software. We have been involved in rod pumping since the days when sucker rods were carved from wood and have been leading the way in architecting digital automation solutions since before the advent of the dial-up modem. 

We know our way around the patch, and we understand through practical experience the technical challenges operators face, because we’ve been right alongside them every step of the way, helping them overcome those challenges. The mission now is to focus the full power of digital technology through the lens of that subject matter expertise to drive high-impact outcomes for customers. What exactly does that mean from the operator’s perspective? ChampionX’s ultimate goal is to use software logic to extract actionable insights from a large amount of data and make precise expert recommendations within a suitable timeframe to affect operational outcomes. Put succinctly, we want to make your life easier with digital solutions. 

VALUING THE DATA  

If data are the currency of the oil and gas business, artificial intelligence and machine learning are the gold standards. They are what give data relational value. Not surprisingly then, the process of digital automation and optimization begins by obtaining well data, using a variety of downhole and surface sensors. These digital devices function as the system’s eyes and ears to perceive reservoir, wellbore and equipment conditions, such as pressure, temperature, flowrates, rpm and loads. 

Horizontal well operators are increasingly installing in-well sensors to collect “live” downhole measurements rather than approximations. ChampionX provides ruggedized downhole sensors that continuously transmit data, such as pressure, temperature, and vibration measurements to surface through tubing encapsulated cable. The technology enables automated control of gas lift injection rates and electrical submersible pump (ESP) speeds, for example, to improve system performance and maximize production based on real-time parameters. 

But not every application requires a full suite of downhole sensors. What makes sense from an ROI perspective on a new pad well with an ESP or jet pump may be completely impractical for a well just across the lease road on beam pump or plunger lift. Fortunately, sensing solutions do not have to be complicated or expensive to be highly effective. 

A case in point is the new SmartSpin wireless rod rotator sensor. Quick and easy to deploy by attaching directly to the polished rod, the sensor provides verifiable results to confirm proper rod rotation for prolonged service life. The SmartSpin sensor is compatible with all beam pump field equipment and is agnostic to the brand of rotator, controller and software. Cameras are another type of surface sensor that are helping operators keep an eye fixed on the well‒literally. Options range from infrared imaging cameras for emission detection to basic motion detecting cameras for site security.  

ChampionX recently deployed a new module within its flagship XSPOC production optimization software that creates a gallery for operators to easily review, classify and prioritize photos for further investigation. Particularly in regions where proposed regulations may soon require daily well surveillance, including Canada and California, this is an affordable option for visually monitoring locations without having to dispatch personnel into the field. Cameras can be “on duty” 24/7 to foster improved HS&E and ESG practices by reducing windshield time, mitigating wellsite risk exposures, and enhancing operational awareness. 

Regardless of the sensing device, the acquired data are typically transmitted to a remote terminal unit (RTU) residing on the well site. For artificial lift, RTUs include pump-off controllers for reciprocating rod pumps and variable speed drives (VSDs) for rotating equipment, such as ESPs and progressing cavity pumps (PCPs). The RTU is the brawn of an automated well operation, providing localized control and initiating responses to pre-set parameters (i.e., regulating pump rate, based on motor torque). 

As technology continues to advance, logic is being pushed from the cloud to the edge to reside within the controller’s architecture. Edge computing allows more real-time analysis and decision-making capacity without the latency speed issues associated with iteratively sending data or receiving commands via the cloud. 

ChampionX’s SMARTEN suite of automation and control solutions is designed for producing assets on rod pump, ESP, gas lift, plunger lift, and PCP, but also has application in chemical injection pumping. Its easy-to-use intuitive interface helps field personnel quickly check the health status of the lift system and easily address issues or make adjustments. 

RUNNING THE SHOW  

Perhaps the single, most important piece of the production optimization puzzle is the software that runs the show. The software is the central nervous system–the brains of the operation. It collects raw data acquired by sensors, acts as the memory bank to store and retrieve information, processes and sorts incoming data, and communicates with the controller and other on-site devices. 

It’s obviously not possible to have a team of production engineers analyze data and make operational adjustments around the clock, but the logic in XSPOC may be the next best thing. It analyses all data from the well and supporting facilities, and then sends expert recommendations, such as setpoint changes or parameter updates, to the controller. The platform has been entrusted by more than 200 operators across the globe to host data on 135,000 wells to help make faster and better-informed decisions for maximized returns. 

At the risk of engaging in more buzzword bingo, let’s break down four primary attributes of XSPOC and how each helps optimize production: physics-based diagnostics, AI, IOT/SCADA communications, and open architecture. Physics-based diagnostics is the foundation of XSPOC. The software uses industry-leading multiphase flow correlations and NODAL analysis to analyze well data and makes sense of what’s happening at any point in time from the bottom of the hole to the surface. 

AI assists field personnel in performing their jobs faster and more efficiently. XDIAG analysis on downhole pump cards uses pattern matching to interpret downhole data and enhance the operator’s understanding of a well’s behavior, so they can make more-informed decisions. Multivariate correlations are used with ESPs to simultaneously analyze multiple parameters and identify possible operational anomalies and failure modes in real time. AI-based autonomous control is our newest venture. It allows users to have XSPOC identify and implement optimized set points for rod pump wells on a continuous basis. This same functionality has now expanded to gas lift wells, with ESP next on the list. 

The software can connect with almost any device in the field, expanding automation and optimization well beyond artificial lift controllers to facilities devices and other data gathering sources to develop a fieldwide network inside the software. Open architecture makes it possible to integrate XSPOC with other software, including third-party applications that an operator may already be deploying. This adaptable approach allows organizations to enable two-way data communication or one-way communication, depending on security needs. 

AUTONOMOUS CONTROL 

Putting all the components together—sensors to collect data, software to analyze and render recommendations, based on expert logic, and the controller to act—defines the state of the art in automation and optimization as it exists today. As noted, however, the next evolutionary step is moving logic from the cloud to the edge and closing the control loop to empower the software to autonomously execute recommended actions in real time. 

This new autonomous operating paradigm compresses data collection-to-action cycle times and minimizes the required human element, along with the risk of human error. Instead of relying on a human supervisor to review software recommendations and authorize changes, what if the software could automatically and instantly make changes within pre-established parameters based on expert logic? That is the journey ChampionX is on right now. 

Intelligent autonomous control adds value by making it possible for production and engineering teams to focus on the most critical and highest-reward assets and nuanced operational challenges, while letting the software take care of the rest. Production teams have limited time and resources. It’s inevitable that they spend most of their efforts on the wells that matter most, giving less attention to low-priority/less-dynamic wells. With autonomous control, XSPOC can consistently apply engineering logic across all assets in a company’s portfolio. No matter the well, its location, or its priority in the hierarchy of assets, the software optimizes its operation on a continuous basis. It’s always working. 

It should be emphasized that autonomous well control is not another digital promise or some vision of a distant future. ChampionX is turning its potential into proof points in rod lift well applications. The software is continuously optimizing setpoints on both POC and VSD applications to maximize production rates without sacrificing equipment performance, Fig. 2. The aim is to generate incremental production gains while also reducing excessive cycling and extending run time to maximize equipment service life, pump efficiency and power consumption. These are the buzzwords we are excited about. 

Fig. 2. ChampionX is turning the potential of autonomous well control technology into real-world proof points in rod pumping applications. Autonomous control software continuously optimizes set-points to maximize production rates without sacrificing equipment performance. The goal is to generate incremental production gains, reduce excessive cycling and extend run time to maximize equipment service life, pump efficiency and power consumption.

The report card is encouraging so far. A U.S. land-based unconventional operator piloted autonomous control on more than 100 rod pump wells. After 60 days, the pilot recorded an average incremental oil production gain of 2 bpd per well. Projected over a full year, the results equate to recovering an additional 500 bbl in each well, correlating to approximately $5 million in incremental annual revenue. In addition, we have introduced autonomous algorithms to proportionally control chemical injection rates in wells on artificial lift. In this case, the technology leverages all the data streaming into the software (production rate, temperature, pressure) to optimize the entire producing asset, including the lift system, by automatically optimizing chemical rates. 

The next milepost is extending autonomous control to ESP wells, beginning with frequency setpoint optimization, Fig. 3. If the logic can be perfected to optimize speeds during critical events, such as startup and restart, the technology can continuously identify the best frequency range to maximize production without damaging or degrading ESP equipment. As autonomous capabilities are further developed and refined using multiple variables to control individual wells, it will gradually expand to multiple-well, multiple-variable control scenarios, such as optimizing gas lift injection in all wells on a pad using one centralized compressor. 

Fig. 3. The next milepost in the well control journey is extending autonomous control to ESP wells, beginning with frequency setpoint optimization to maximize production without degrading ESP equipment. As autonomous capabilities are further developed and refined, it will gradually expand to multiple-well, multiple-variable control scenarios, including optimizing gas lift injection in all pad wells using a single centralized compressor.

NEW POSSIBILITIES  

Our industry has been defined by overcoming seemingly insurmountable engineering and operational challenges, sometimes by necessity and often in collaboration. Looking at the longer-term horizon, it’s impossible to predict all the ways in which autonomous control could eventually be deployed in production operations. However, it’s a safe bet that applications will emerge that are hard to even conceive of today. An entire new realm of possibilities is opened by migrating software logic to the edge, in the thick of the action on the wellsite. 

Meantime, it bears repeating that the development of autonomous well operations is a long journey that has only just begun. Similar to self-driving cars, some oil and gas companies may be incentivized enough by the high-upside benefits to embrace autonomous control from the jump. Others may prefer to take a slower route to adoption. They may want to keep a hand on the wheel, but use select autonomous features–like a driver using adaptive cruise control or lane monitoring–to automate pump speeds or maintain parameters within preset “fairways.” 

But every journey begins with a first step, and our goal is to meet operators where they’re at to embark on the process together. In the end, collaboration and cooperation between service provider and operator will prove fundamental to navigating the learning curve and accelerating the adoption curve for autonomous well control. It’s time to get started. 

About the Authors

Paul E. Mahoney

ChampionX

Paul E. Mahoney is president of production and automation technologies for ChampionX. In this role, he leads the company’s global artificial lift business, which encompasses all forms of artificial lift, automation and digital, and emissions management. Before joining Apergy (a ChampionX legacy company) in 2012, he was president of Dover Artificial Lift. He also served as chief operations officer for Dover and as senior vice president of global sales and operations for Dover Artificial Lift. He began his career at Emerson Electric, where he served in a variety of sales and management roles. Mr. Mahoney holds a BS degree in electrical engineering and physics from the University of Buffalo and an MBA from Seattle University.