Sizing up rig automation

Automation has long been the epicenter of rig innovation. Safety and efficiency are the traditional drivers, but operational limits, such as those presented by extended-reach drilling (see the NOV/Doyon Rig 26 article on page 37), are an ever-growing part of the mix.

With greater demands on all fronts, rig automation has become an increasingly complex multi-disciplinary process. Development of successive generations of automation depends on the industry’s ability to attract, educate and learn from the next generation of drillers, and the driller’s ability to blend a spectrum of advanced skills with the hard realities of drilling a well.

DRILLBOTICS COMPETITION

SPE’s Drillbotics® competition is right in the middle of that equation. Now in its sixth year, the international contest between university teams is the product of SPE’s Drilling Systems Automation Technical Section (DSATS), which promotes adoption of automation techniques. Joining the effort this year are other SPE technical sections: Wellbore Positioning and Drilling Uncertainty Prediction.

Fig. 1. Drillbotics miniature rigs built by 2018 teams (Left to Right) from Texas A&M University, Clausthal University of Technology (TUC), and The University of Oklahoma. Image: Drillbotics.

The competition’s tantalizing objective is to create a miniature rig and downhole sensors to hands-off drill a quality, on-target wellbore, Figs. 1 and 2. The challenge has captured enthusiastic and growing support from universities and industry, as well as an impressive cadre of Drillbotics team graduates.

Fig. 2. Bit exits sandstone sample in the directional competition. Image: Drillbotics.

Drillbotics started in 2015 with four teams. In the 2020 competition, now underway, there are undergraduate, master and doctoral students from a variety of disciplines participating, Fig. 3.Drillbotics estimates that 300 students at each university are exposed to the competition. The current competition concludes with a final presentation and rig demonstration by the teams in May and June 2020.

Fig. 3. University rigs lined up for the drilling competition at a Halliburton facility in Houston. Image: Drillbotics.

The 2015 competition emphasized vertical drilling and autonomous drillstring mechanics. In 2019, directional drilling was added, and 2020 introduced modeling of a virtual rig and well. The 2020 competition requires teams to hit multiple targets at varying vertical depths and X/Y coordinates, using closed loop control, based on downhole data.

Fig. 4. NTNU circuit board highlights skillset diversity. Note logo, upper right.

By design, the complex undertaking requires a collective effort from diverse teams, with mechanical and electrical engineers building the rig, control engineers creating a real-time control system, and petroleum engineers handling drilling dysfunctions and mitigation techniques, Fig. 4.

The drilling competition is held in North America and Europe. Regional teams set up their rigs and begin drilling at the same time. Starting from a vertical 1-in. pilot hole, they kick off from vertical at any point below 4 in. from the surface. Drilling stops at 3 hr, or when the last team’s bit exits the rock sample.

Of course, the job isn’t finished until the paperwork is done. The teams submit monthly reports accounting for labor hours and costs—there’s a $10,000 limit on rig and BHA spending—and at the end of the competition, they must summarize their final design, and explain any changes from the initial design. The winner presents a paper at the next SPE/IADC Drilling Conference.

COMPETITION INNOVATIONS

In addition to education and experience, the teams produce an impressive spread of innovative tools and control processes. A few novel results include:

3D printed drilling motor. A 3D printed positive displacement motor (PDM) was developed by the 2019 Norwegian University of Science and Technology (NTNU), Trondheim, team to prove its PDM concept and advance the design. The effort included investigation of pitch and number of stages, and stage lengths.

Custom bit design. A drill bit developed by the 2018 NTNU team sparked commercialization interest from a drill bit technology company. The non-aggressive PDC design provided a high-RPM, low-torque response in sandstone applications.

Percussion bit design. In 2018, Clausthal University of Technology (TUC), Germany, took a unique approach to drilling, using a percussion bit. Building on their design from the prior competition, they improved ROP with a 1.5x increase in impact energy.

Data analytics. Innovative algorithms created by the 2017 Texas A&M University (College Station) team used real-time mechanical specific energy (MSE) to characterize rock properties and optimize WOB and RPM drilling parameters. The system identified drilling set points with the assistance of a machine-learning model that predicted drilling parameters. A patent was submitted for the control system technology that the students designed.

INDUSTRY OBJECTIVES

The growing use of automation makes the Drillbotics competition important to the industry’s ability to develop the most suitable technology, says SPE’s new Technical Director, Drilling, David Reid, a founding board member of DSATS and the chief marketing officer for NOV.

 “The competition is a great way for students to learn more about the day-to-day challenges that we face in our industry and how to overcome them prior to joining the workforce,” says Reid. He describes Drillbotics as a two-way street that provides students with automation experience that the industry needs, along with novel approaches and designs. “In the end, both the students and the industry have grown together, as we have learned more about automation and the benefits it brings to our industry.”

Shashi Talya, current Drillbotics committee chair, and Halliburton’s senior technology leader for drilling motors, says, “Students and universities gain from a unique learning opportunity to appreciate and address real-world challenges. The drilling industry benefits from attracting a diverse multidisciplinary talent pool that will advance digital technologies in the industry.”

That multi-disciplinary engagement is core to the Drillbotics concept, says Fred Florence, Drillbotics founder, co-chair and president of Rig Operations. “A key part of the Drillbotics concept was to create graduates, not only in drilling and petroleum engineering disciplines, but other disciplines as well, who could learn to work together and know enough about automation to choose when to use it and when to avoid it.”

Aaron Logan, Drillbotics co-chair and founder of Evolution Engineering, brings a perspective as a sponsor, committee member and judge. “Not only is this a great experience for the students, it is eye-opening to many from the industry as well,” he says.

The creative solutions built by the teams are “approached with an open mind, unhindered by years of experience in the industry, and free from the constraints of industry thinking,” observes Logan. “Although the solutions are not always transferable, this blue ocean approach to design and innovation is what will spark step-change approaches to common problems we all face.”

The competition lets students apply knowledge they have learned as theory, and understand the practical limitations of theory in the real world, he says. “This accelerates the student’s applicability to employability, since real-world, hands-on experience of this caliber is hard to come by and replicate. Many industry veterans will never have the privilege of a ground-up robotic system design and planning, covering facets of each engineering discipline, and actually executing on its erecting and operation,” says Logan.

UNIVERSITY PERSPECTIVE

The SPE competition promotes industry engagement with universities and a level of collaboration that broadens student education and their familiarity with the various disciplines that go into drilling automation, and contributes to career development.

“Industry gets new engineers with practical, multi-discipline experience; students get to meet prospective employers well ahead of graduation,” says Florence.

The process introduces students and universities to new industry issues and concepts. “Both the universities and industry are beginning to see these mini-rigs as a research tool, not a science project.”

“Modeling the drilling process and validating the model with data from the physical rig is a sound practice that can be expanded later to validate the models with data from full-scale drilling rigs,” explains Florence. “This will speed development of new tools and processes at a much lower overall cost to the industry.”

Roman Shor was on the University of Texas 2015 and 2016 teams while he was a PhD student. He then introduced the competition to the University of Calgary, where he is now assistant professor, Department of Chemical and Petroleum Engineering, and the Drillbotics team faculty adviser

“Drillbotics is a unique opportunity for a multi-disciplinary team of students to tackle a challenging problem from the ground up and be exposed to many of the same pressures they will encounter in their future careers,” he says. “Working within a fixed timeline and within a budget, a student must manage design constraints—both competition-imposed as well as those imposed early in the design process—and become familiar with the concept of lead-times for components. Finally, while integrating the various components of the system, they will learn that not everything works as seamlessly as it has in courses, or even as they intended.”

While the control paradigms or mechanization techniques may not be directly transferable to a drilling rig in the field, the industry benefits greatly from seeing new ideas and concepts demonstrated on scaled systems, he explains. “These learnings may range from ideas as simple as state machines for supervisory control to novel ideas for adjustable-bend downhole motors.”

“However, industry’s key win is a crop of new talent that understands the complexity of system design,” says Shor. At the same time, institutions benefit from having a challenging hands-on project for their students to work on, plus the availability of a scaled drilling rig for future and ongoing research work.

Vimlesh Bavadiya was on the Oklahoma University Drillbotics teams in 2015 and 2016. He received his MS degree in data science and analytics in 2019, and an MS degree in petroleum engineering in 2017.

He describes the competition as an intensely holistic experience. ”It involves brainstorming and working with a diverse team; research and simulations; designing the rig and fabrication; programming; performing experiments to optimize the automated drilling process—all to be completed with limited time and resources.”

Interdisciplinary teams foster creative, innovative solutions and a better understanding of subject, such as drillstring vibrations and bit design, says Bavadiya, whose PE master thesis topic was, “Experimental Investigation of the Effects of Rotational Speed and Weight on Bit on Drillstring Vibrations, Torque and Rate of Penetration.”

Mohammad Aljubran was also on the OU team during 2015. He is currently a Stanford MS student. “Drillbotics promotes interdisciplinary collaboration, free of the traditional drilling presumptions,” says Aljubran. “Participants with no work experience on drilling rigs facilitate an unbiased exploration of novel engineering techniques, and exploitation of state-of-the-art designs.”

“I found that Drillbotics was not only meaningful to large-scale drilling automation but also to lab-based drilling R&D efforts that required small-scale rig instrumentation.”

INDUSTRY PERSPECTIVE

Volunteers and sponsors from many rig automation firms support the Drillbotics competition. Sponsor companies and organizations include Baker Hughes, Evolution Engineering, Halliburton, Wintershall, Oxy, ASME, H&P Technologies, Cathedral and INTELLICESS.

John Macpherson serves on the Drillbotics committee and is chief consulting scientist for Baker Hughes. “The Drillbotics competition is innovative on several levels, but it is in terms of cross-disciplinary cooperation that it really shines,” he says.

The competition motivates students to move outside their chosen specialty (be it petroleum, mechanical, electrical or software) and cooperate in solving a difficult problem—the automation of the drilling process, he explains. “Those teams that miss a critical discipline, especially software or firmware, tend to have an uphill battle, while those teams with a diverse group of members tend to shine.”

Innovative exploration of the challenge can yield unanticipated solutions. “The current trend of teams printing their drill bits is one that was not anticipated,” says MacPherson. “It seems to have started with one team deciding that the supplied drill bit was not suited to the task, and investigating printing a ‘fit-for-purpose’ drill bit.” It was a good decision. The team went on to win the competition.

Fig. 5. Drillstring twist-off illustrates demands on system integrity.

The evolution of control systems is another innovative direction in the competition, observes MacPherson. “The earliest control systems seemed to be modeled around ‘classic’ drilling engineering concepts, such as MSE. More recent concepts seem to be better-tailored for the task in hand, recognizing that hitting a boundary with such a flimsy drilling system can catastrophically damage the system. So, focus has shifted to lower-level fast-response controllers,” Fig. 5.

Perhaps one unanticipated characteristic of all teams is the degree of enthusiasm and commitment that they bring to the contest, he says. ”They are extremely proud of what they have achieved—even if they do not win—and leave with a certainty that they will be back next year with a better drilling system. The level of inter-university inspection and analysis that is generated by having the competition in one locality is also impressive,” says MacPherson. “The teams learn from one another.”

Jayesh Jain is a Drillbotics committee member and Baker Hughes senior research engineer in computational sciences, and digital and machine learning leader. He says, “The unprecedented problems that the oil and gas industry currently faces require extraordinary solutions. To stay afloat in a low-price era, safety and efficiency of operations are seen as major sources of cost reduction.”

Drillbotics helps drive these objectives by instilling innovative thinking in the minds of the next-generation professionals, Jain explains. Creating innovative drilling rigs requires the teams to demonstrate effective utilization of advanced design, simulation, automation/controls, and machine learning techniques.

“Remarkably, the teams were able to demonstrate such feats as autonomous drilling of directional wells and automated whipstock placement, which industry, itself is still attempting to fully realize in field operations,” says Jain.

COMPETITIVE PROCESS

The competition includes creativity, design and drilling phases. It starts with determining rig design; power; the type and accuracy of sensors; data collection; processing and display; where the downhole sensors (a recent addition to contest requirements) can be located, how will it be packaged; and what telemetry will be needed.

Autonomous drilling requires the use of both surface and downhole sensors in conjunction with drilling models and automated controls. The downhole sensors may include vibration sensors and/or directional measurements that are important in both vertical and directional wells. Sensor selection, packaging and data interpretation is a complex undertaking. Students must develop control algorithms that will analyze surface and downhole sensor and make the next set of decisions to steer the bit.

In the design phase, teams must support their engineering choices and corrections, as well as cost versus functionality. Judges consider how much is learned, in addition to machine design. Construction safety is another consideration—including how the heavy rock sample is handled. Challenges range from electrical signal isolation and wiring access, to the rig’s stability and how easily it is moved.

For the virtual rig competition introduced with the 2020 competition, judges will look at model performance compared to the actual rig, as well as assumptions and testing used to validate the model. The build phase for the virtual rig involves creating block diagrams with well-defined inputs and outputs, determining how model uncertainties are considered in the control design, and justifying design decisions to the judges.

The last phase of the competition judges the autonomous drilling process, how well the drilling algorithms functioned, and their performance responding to rock variables and staying on the planned directional well path.

THE BASICS

There is no easy path to TD. Competition teams learn this, as they tackle the many tasks of making hole in a 12-in. W × 24-in. L × 24-in. H (30 × 60 × 60 cm) sandstone block.

Fig. 6. Drilling mediums for earlier vertical wells presented complex (top) and “evil” (bottom) challenges. Horizontal drilling, now using homogenous sandstone, may face similar problems in future competitions. Image: Drillbotics

The nearly homogenous sandstone block is a marked change from the past, and perhaps a short-lived condition. When the competition changed from vertical to directional drilling, the introduction of new issues led to using the homogenous rock. But in prior years, when the test formation was built in the garage of one of the Drillbotics committee members, the decidedly heterogeneous medium included multiple materials hidden in a wooden box, Fig. 6. Students had to automatically react to changes in the formation, whether it was a change in rock strength or formation dip.

“It also included some pretty evil materials, such as the rubber mat that no one had expected,” says Florence. “Once they get a handle on this, we will give them some new problems to deal with, such as directionally drilling through multiple formation layers.”

Fig. 7. These team-supplied bits were used by the West Virginia University 2016 team to gather drilling data to tune their models. Image: Drillbotics.

To drill the sandstone, DSATS provides a 1.5-in. diameter and 2-in. long PDC bit. However, teams are permitted and have taken the opportunity to design their own bits, Fig. 7.

A key aspect of the completion is an understanding of how the BHA, bit and drillstring function together, and how the downhole system measures, samples and transmits the drilling data. This includes design or purchase of components, including drill pipe (aluminum alloy or stainless steel) tool joints, bit subs, drill collars, stabilizers and sensors.

Drilling automation combines data, control and dynamic modeling, so that the control algorithm can respond to differences between expected and actual performance. The teams determine if standard or custom data collection and handling are used, and they get special consideration for novel ways of visually presenting the data. Data must provide bit depth, elapsed time, ROP, MSE, verticality/inclination, vibration and any other variable used to describe rig performance.

CONCLUSION

The gauntlet has been thrown for 2020. Drillbotics teams are busy developing and applying miniature rig technology to the harsh realities of drilling a directional hole in a sandstone block. Their multidiscipline experience will help guide the path of rig innovation for many years to come.