A new automated drilling program on a rig sets off a battle of wits. Drillers first focus on learning to use it, and then figure out the underlying logic of the software.
“The people learn from that and it’s not too long before they have ideas that can be put into practice that can improve whatever the operation is,” said Matt Isbell, drilling engineering advisor for Hess. The oil company uses six Nabors rigs, equipped with digitally controlled systems to drill in the Bakken.
That conflict is good for Isbell, but not for everyone. “In terms of the designers of these automated systems, they’re always disappointed when people try to beat them, and do successfully beat them,” said Isbell.
The tension is required because the value of the Nabors’ rig automation, developed in partnership with Hess, depends on finding new ways to improve drilling.
While the word automation is associated with machines replacing humans, payroll reduction is not their goal. “People are still critical. Machines do not learn by themselves. People learn. We are still counting on people to be the leaders of this,” he said.
Drilling productivity improvements are likely required to meet Hess’ aggressive plans to increase production and profits in the Bakken.
In its recent third quarter report, the company reported an 8% drop in drilling and completion costs since the first quarter, to $6.7 million per well.
A big reason for that was a shift to plug-and-perf completions. Changing old habits, such as fracturing using sliding sleeves, is part of a drive to create a leaner drilling operation.
“Through the continued application of lean manufacturing, we expect to achieve further cost reductions as we progress toward our targeted drilling and completion cost of $6 million per well,” said Gregory Hill, chief operating officer and president of worldwide exploration and production for Hess, during its third quarter earnings call.
Drilling productivity has been synonymous with drilling wells faster. A chart in a recent paper showed how Hess has roughly halved the average time required to drill a well since 2012, and narrowed the difference between its fastest and slowest jobs (SPE 195818).
“Shave off longer delivery wells, you generate improvement—there is still a little inconsistency there in flat (nonproductive) time and drilling time. We are hard at work at that,” Isbell said.
Doing what they are doing now even faster is not an option. “If we continue to compress these activities, expecting machines or people to do them faster… the physics does not support that,” Isbell said.
Hess does see an opportunity to use automation to build more and better quality wells. During that recent call, the company said it plans to use its fleet of six rigs to push Bakken production up to 200,000 BOE/d by 2021. Then it will use four rigs to sustain that output in a play where rapid decline rates are a given.
This plan is expected to result “in material, free cash flow generation across a range of prices,” said John Hess, chief executive of Hess during the call with financial analysts.
That bit of financial jargon represents the current Holy Grail for shale producers. It means consistent profits that are high enough to satisfy investors, even when oil and gas prices are low, which is what shareholders are demanding.
Same Every Time
The Hess paper on its drilling improvement program outlined the ongoing process that has combined increasingly automated drilling machines and people looking for more productive ways to do it.
Hess calls it drilling operations automation, because there are digital control systems for key tasks, without a central computer coordinating it all.
Nabors is competing with other drillers, such as Precision Drilling, and service companies that have developed systems. Those include Schlumberger, which has developed a central control system to coordinate subsystems, such as the one adjusting the weight on bit and rotation speed to maximize performance as conditions change.
The value of programmable systems depend on their ability to facilitate productive change. Hess and Nabors do not want to automate the status quo. The goal is to create programmable machine tools that generate good data that fosters ideas for how to do things better. Those innovations that prove useful can be quickly programmed into all the Nabors rigs.
These drilling machines have the advantage of sticking with the program. If there is any question about the source in variation during drilling, Isbell said, “I am happy to say it is people.”
Grading the Curve
The prime measure of success has always been time saved drilling, and that has not changed.
The digital control system used to control sliding—the act of changing the well path by turning the curved mud-motor housing just long enough to make a planned course change—saves time. Longer than needed slides are inefficient because sliding is slower than drilling ahead. More precise control can also limit the time spent getting a well back on course if slides are poorly executed.
Hess reports that automated sliding control saves it about 4.5 hours per lateral, assuming there are about 45 slides, which adds up to significant money in a year of drilling. Harder to measure is the payoff for drilling wells that are closer to the original design, allowing others to effectively complete and produce the well for years to come.
Isbell said that the ability to “more tightly manage and control our operational execution of a well” will make it possible to improve well designs, adding value to those spaces.
In the near term, the goal is better hole quality. Engineers describe that as eliminating flaws, such as tight “dogleg” curves that can make it hard to run casing, or lead to destructive contacts between tubing and sucker rods.
But this numbers-driven industry has not agreed on a way to quantify it. “Quality metrics are an interesting one for me because it really is what you define it to be,” Isbell said.
Hess is measuring quality based on the degree the wellbore diverges from the straight lines and curves of the well plan, which is known as tortuosity.
The line between acceptable well-path variations when drilling fast through unpredictably changing rock, and excessive tortuosity, is based on input from those completing and producing Hess wells, Isbell said.
Those specifications also include inputs from drillers because “the unfortunate reality is that these objectives are often at odds,” the paper said. For example, many short slides will reduce tortuosity, but their use means it will take more time to drill the well.
The value of a better well is harder to estimate than the cost of a few hours more of drilling time, but “the production side is where you make money. The drilling side is a sunk cost,” Isbell said.
Still, the goal is “delivering higher quality wells at a lower cost” he said.
To get beyond drilling spending decisions based on a single measure, Hess decisions makers judge the value of automation using a formula based on improved well delivery, quality, and safety, which define the cost.
Some benefits are easy to measure. For example, controlling drilling settings to avoid excessive stress that damages a bottomhole assembly (BHA) saves on time lost for unexpected, costly repairs.
“As you have better control of the drilling process we will see an improvement in the life of the BHA,” Isbell said, adding, “That’s where most current activities are focused.”
Other benefits are hard to figure. Drillers have struggled to answer the question: What is the dollar value of drilling limits that prevent an accident from happening? It is also difficult to know how the undulations of a wellbore look inside the casing, which tends to smooth the path. It is also hard to locate the low spots most likely to accumulate water and slow production.
Hess management has remained committed to the long-term drilling automation program during a tough period in the shale business when many companies have slashed technology development budgets.
“If we decided to do something else, management would be against it because this is the path we are on,” Isbell said.
Customer Driven
Automation can facilitate changes, if all the people involved “think of this differently,” Isbell said.
“If we can manage a system this way, it impacts the design of everything from your MWD (measurement while drilling) to the way we train our people,” he said.
The people driving drilling improvement are inside and outside Hess. Isbell says customer input from well users is critical. This is a different use of the word customer in drilling. It normally describes the oil company hiring a drilling contractor.
Anyone who has talked to completion and production engineers about hole problems soon sees they are demanding customers. Given the importance of improved well performance, they have influence. “One thing that we learned early on is that business-led projects tend to gain traction faster,” Isbell said.
The customers also include those in charge at drilling contractor and service companies. They are in control on the drilling sites, which is the hub for innovation.
There are many voices to consider, and they sometimes offer conflicting advice. “Automating drilling requires many highly detailed discussions about the drilling and operations practices. That is kind of where we are today,” Isbell said.
While Hess is seeking input from more people, those on the rig still have influence.
“The rig manager and the company man are really the ones that are in charge of judging how well these systems work,” Isbell said, adding that, “It has to make a driller’s life easier or they will turn it off.”
Working Relationships
Before automated systems were added to rigs, Hess started changing where, and how, the people guiding directional drilling worked.
In 2016, Hess moved directional drillers to a remote operations center where two advisors would monitor three wells each, with the assistance of advisory software providing drillers turn-by-turn directions.
“Directional drillers are optimizing how they use their time. Going from one rig to six made them a lot busier,” Isbell said. Hess had learned that three wells per person was the maximum number per person, and found that they learned faster when tracking multiple wells.
Digital controls can outperform humans at tirelessly calculating the ideal settings for a slide, which include adjusting multiple rig components to ensure the weight on bit and rotations per minute are at the right level while sliding.
There are still situations in these unpredictable formations that demand human input. Directional drillers are still required while drilling the lateral because the path through most productive rock has to take an unexpected turn.
For that reason, directional drillers are located near geosteering advisors who interpret logging data to determine if the drill is penetrating the most productive rock.
If the logging data show that a fault has shifted the productive zone down 15 ft, those experts must deal with their conflicting priorities. The geosteering advisor’s goal is maximizing the time in the pay zone while the driller needs to avoid an abrupt turn that would add tortuosity.
Over time, data from digitally controlled equipment will identify new problems to solve.
“As you get the ability to more tightly manage and control your operational execution of a well, then you have the ability to improve the well design, which accumulates more value. And you can also tie that better information into your completions and production,” Isbell said.
