John Henry vs the steam drill: Will the robots win?

FORD BRETT, CONTRIBUTING EDITOR

John Henry is a nineteenth century African American folk hero, celebrated for his extraordinary strength and skill in using a hammer drill to create holes for explosives during railroad construction.  According to legend, when a steam-powered rock drill was introduced to speed up work, John Henry challenged it to a contest to prove that human labor could outperform the machine.  As the story goes, John Henry used two 10-pound (4.5-kg) hammers, one in each hand, to drill a 14-ft (4.2-m) hole in the rock, while the steam drill managed only 9 ft (2.7 m).  John Henry did beat the steam drill.  Sadly, though, his effort was so intense that John Henry “died with his hammer in his hand.”   

For over 100 years, American elementary schoolteachers have used this, and other similar tall tale stories, as hooks to interest kids learning to read. Like Aesop’s fables, the tales generally have some kind of moral.  What’s the moral here?  People are better than machines – OR - Don’t fight machines because you’ll die?   

Another good question is What does John Henry have to do with a drilling technology column in World Oil?  Answer: The next two columns in this series, which look at efforts to automate drilling, might help address the challenges in meeting growing North American natural gas demand.  It seems we can’t get away from humans vs machines.  More on John Henry or the Machine later in this column. 

Background: North America’s gas challenge. Even if A.I. and data center growth underperforms expectations, since planned LNG facilities already have contracted sales, North American natural gas demand is set to grow by ~30 Bcfd (or ~25%) in five years – the largest and fastest growth the industry has ever faced in both percentage and absolute terms. The entire world faces similar unprecedented growth in natural gas demand (more on that in later editions). This edition will look at the situation in North America. 

The huge growth in North American gas production in the past decade has come from the gas associated with unconventional oil production.  Because unconventional oil production won’t grow at the same rate, future gas supply growth must come largely from drilling projects intentionally targeting gas. North America will need something like 100 to 200 additional gas‑focused rigs to meet that demand. Unconventional gas resources are vast and well-understood. Thanks to drillers, productivity per rig has increased roughly eight-fold since 2010, so wells are also now cheap enough to access that resource.  

So, technical factors are not going to be what limits us, but there are still limits to producing this much gas. The real risks lie in the following three bottlenecks: 

• Midstream infrastructure—pipelines, processing, compression, and storage—may not expand fast enough, leaving producers unable to move gas to market.  
• Permitting and regulatory delays, especially in the northeastern U.S., can block both drilling and pipelines, making political and legal opposition a critical constraint.  
• Rapid increases in rig activity reliably degrade drilling performance to the tune of >$10 million  per rig added to the annual rig count, creating a major economic challenge.   

The last article in this series showed how, by increasing lateral length, drillers can help our midstream and permitting brothers and sisters to address the first two bottlenecks.  Increasing laterals from 5,000 ft to 20,000 ft reduces the number of surface locations and total length of gathering pipelines needed by a factor of ~16x (that’s 1600% not 160%!).  And since it’s cheaper (and less disruptive) to drill and case a foot of horizontal hole than it is to run a foot of pipeline on the surface, drillers help mitigate permitting and midstream issues by putting the gathering line in the shale itself. Longer laterals don’t completely solve the first two bottlenecks, but they really do help. This edition, and the next one too, will look at what drillers can do about the third bottleneck.  

Fig. 1. Increased activity creates a costly drilling performance twist-off.

The performance “twist-off” problem–money and safety. Prior columns detail how drilling performance degrades when market forces cause rig activity to increase. I try to keep columns coherent when standing alone, but they do have a red thread. Those interested in the details can find more here.   

Very briefly, every surge in rig count has been accompanied with a statistically significant drop in drilling efficiency—the effect is driven not by equipment scarcity but by the time and expertise required to rebuild high-performing crews (drilling ‘crew’ in the larger sense: the rig crew, service and operator personnel). This Performance Twist-off represents a billion-dollar challenge at scale.  Figure 1 summarizes this effect since 2008, but the effect goes all the way back to at least 1950.    

Feet per Day, and Drilling Cost aren’t the only measures of drilling performance. What we do makes the world a better place—energy makes lives easier and better.  But if we hurt people in the process, we are not making things better.   

Fig. 2. Increased activity increases U.S. oil and gas extraction fatality RATE.

Figure 2 shows how the Performance Twist-off bottleneck affects safety, as well as feet per day and well cost. The sad figure shows that the fatality rate of U.S. oil and gas extraction workers changes as rig count changes. (Note: these fatalities are for all oil and gas extraction activities, not just drilling). When drilling activity goes up the fatality RATE goes up. More activity logically means more people in harm’s way and—unfortunately—more safety incidents. This plot shows that it is worse than that—the upwardly sloping line means that the busier we are, the more dangerous things become. The fatality RATE goes up—not only are there more workers, but they are also harmed more frequently.  

The Performance Twist-off problem is real, a bit too real. And it absolutely doesn’t have to be this way. 

Back to John Henry and the steam drill…One way to beat the Performance Twist-off is automation. Build a machine that will reliably replicate known performance—use a ‘Steam Drill.’  Another way would be to have crews capable of reliably duplicating the performance that humans have already proven possible—hire and train “John Henrys.”  This column will look at what one drilling contractor, Helmrich and Payne (H&P), is doing to automate the cost-constrained land rig market by carefully using the best of both machines AND people.  

What follows is the result of conversations and tours that the folks at H&P were kind enough to provide, describing the system they have built to automate their land-based fleet. Much of recent rig automation activity has been focused—rightly—on deepwater or offshore drilling operations, where saving 15 min. a day can add up to >$40 million over the course of a year.  Here we look at automated land rigs., 

(For background, State of Drilling System Automation is a link to an SPE presentation posted by John De Wardt, chairman of the SPE’s Drilling Automation Technical Section. It is a nice survey of what many leading drilling organizations—BakerHughes, SLB, NOV and others—are doing to improve automation. It’s a good starter for those who’d like to learn more). 

Many thanks to H&P’s Patrick Gustafson, operations manager of HP Rig 918; Rhett Stricker, senior manager, Training & Development; and John Baer, director, Shared Services, for their time, insights and candor in explaining H&P’s automation journey. It’s their work, and that of others at H&P, which is described here. 

Machines AND people: How H&P automates land rigs. H&P’s automation approach is to incrementally and continuously apply proven technology to its operations. H&P is starting by automating two specific things: drill pipe connections/tripping and drilling ahead, and then incrementally rolling out a proven approach to its fleet. But their approach is not just about installing machines. It’s also about making sure the crew can operate and maintain the system. This article focuses on automated tripping and drilling connections (including HWDP). The next in this series will describe what they are doing to automate the process of drilling ahead. 

Fig. 3. One of the three robotic pipe handling arms. Image: Helmerich and Payne, H&P Rig 918.

To move pipe, the H&P rig floor automation system uses NOV’s AtomTRX system.  For reliability, AtomTRX adapts off-the-shelf Yaskawa robots that have a 60-year history of industrial use to the rig floor.  For connections, H&P system uses a Hexgrip automated floor wrench, along with specialized retractable sub-lifters, mud bucket, pipe dope tool, and robot ends.  The system makes up all connections—drill pipe and the BHA—but H&P has only automated tripping. BHA handling is mechanized, but the crew makes it up, as with any other mechanized rig.   

To be sure that what they apply in the field works, H&P has set up a full-scale test facility, with one of their FlexRigs to troubleshoot and test every system before they use it on real operation—they call the test rig H&P 918 (918 is Tulsa’s area code, where H&P’s office and test facility are located). They set up the rig in 2022 and have used it to first mechanize, then automate, the rig for tripping and drilling connections. Figure 3 shows one of the off-the-shelf robotic pipe handling arms with its custom attachment guiding the drill pipe.   

John Henry: Human in the loop. While it is theoretically possible that humans could be taken completely out of the loop while tripping, H&P has elected to keep a human in the loop to ensure that the automatic system is working as anticipated, and that the well is responding as it should. Without human interaction at appropriate points in a connection, the system halts. So, H&P’s automated tripping system is not like a fully self-driving taxi, where you can take a nap going down the road. It is more like one of the newer cruise control systems, where the car does adjust speed and steer to stay in lane but needs to sense that you still have your hands on the wheel or the cruise control shuts off.   

Fig. 4. H&P driller supervising an automated trip. Image: Helmerich and Payne.

I guess it’s conceivable that the trip could eventually be fully automatic.  H&P is thinking about what might be possible but won’t add that capability until they are very sure it will work every time. I’m betting even then, they’d first start with just trips inside cased hole.  

While the driller does monitor and interact with the tripping and drilling connection process, it is entirely hands-free. No one is on the rig floor, nor in the derrick. In fact, the system is designed, so that if anyone is on the rig floor or the derrick, the process halts. Figure 4 shows that the driller is still involved in the process, even though all the physical activity of moving and making up the pipe is automated and moves completely by itself.   

Robots need people, too. My wife’s new car has cruise control that will steer to stay in the lane, speed up and slow down, and stop if she doesn’t interact with the steering wheel every so often.  And it took no training to use. She got in, and it worked.   

Because “Hands free” and automated tripping/connections are more involved than turning on cruise control, and because robots need to be maintained, H&P is not just installing equipment and wishing crews good luck. They have formal crew training to ensure that they can operate and maintain all the systems, so that they can be sure everything will work reliably and as advertised from the very first time it’s installed on a rig.   

Fig. 5. Rig crew training at purpose-built, duplicate, mechanized rig floor training facility. Image: Helmerich and Payne.

To do this, to train crews for automation, H&P uses its full-sized automated test rig (H&P 918), and, for better crew training, has a duplicate rig floor at their Tulsa facility, where they conduct a three-day training session for a full train, full crew (driller and floor hands) on their days off (they do get paid) on how the system works and how to maintain it. See Fig. 5.  

H&P also has a one-week drillers’ automatic tripping training program, where they can practice interacting with the automated system on the full-sized rig to build muscle memory. 

Benefits of automated tripping (so far). Several benefits can be realized through the automation of tripping, as follows: 

Reliably safe: H&P, like other contractors, takes safety very seriously and has an admirable safety record.  But it’s very hard to beat ZERO, and with hands-free automated trips, there are ZERO people in harm’s way.  With automated trips and connections, there is no one on the rig floor or in the derrick. In fact, the system shuts down if someone is on the floor.  Automated tripping and drilling connections guarantee no safety incidents. The H&P approach has no ‘cobots’… robots do everything automatically, without humans on the floor. 

Reliable performance: H&P’s goal was for automation to allow rigs to reliably achieve as good or better performance than the P50 trip times for manual rigs. With the connection and tripping process now in control, targets can be set and reliably achieved. H&P has achieved targets from the first time that the system was used in a real operation, probably because the system was designed, tested and debugged on H&P 918. Since crews are fully trained on the system’s operation and maintenance, and since the industrial robots they use have a 60-year history, there hasn’t been any significant loss of performance from equipment failure. 

Better rig maintenance (!): One unanticipated downside of the industry’s recent improvement in drilling performance is that crews have less and less time to perform routine maintenance. Back in the old days, when a crew might make two connections in 12 hrs of drilling, and when rig moves took seven days, there was plenty of time for a crew to do routine maintenance. When connections come every 30 min., and rig moves take 2 hrs, it’s hard (impossible?) to keep maintenance current. Automated tripping provides space, so maintenance can happen when it should. 

H&P has no plans for automation to reduce crew size below a normal five-man crew. When spud-to-spud time is on the order of two weeks, there are plenty of ways for the rig crew to add value, ensuring equipment is properly maintained.   

Current plans to scale. To roll out automation to their existing fleet, H&P has a “leap frog” process that automates a rig on a move from one pad to the next. The rig crews will have already been trained on days off, so that the rig and crew will be automated and ready to go. So far, the process has let the automated rigs meet performance targets. 

Mechanization. To date, H&P has mechanized and trained about >30% of its fleet for hands-free connections.  They plan to continue mechanizing rigs, and because they fully train the crew to operate and maintain the system before it’s used in an operation, they have not experienced teething pains when they mechanize a rig. 

Trip and connection automation: To date, besides the fully automated the rig they have set up in their Tulsa yard, H&P has one that been working – and meeting targets - in the Permian.  An additional Permian rig will be automated in June 2026, with three more planned by the first quarter of 2027.  

One final man vs machine story. I had chosen ‘John Henry vs the steam drill: Will the robots win?’ as this edition’s title. After completion, I fed the body of this text into Copilot and asked what it thought the title should be. Here is what CoPilot said: 

“My Top 3 Recommendations, if I had to narrow it down for impact and readership”: 

1. Don’t Be John Henry: Why drilling needs automation to survive the next boom 
2. John Henry, the steam drill, and the future of drilling 
3. Beating the steam drill this time. 

Question for you: Did the Robot beat me? 

Until next time, I hope to start a conversation with any of you on how we can all help Drilling Advance.  If you have any comments, ideas or corrections, please email me at ford.brett@petroskills.com and I promise I’ll respond.