December 2019 /// Vol 204 No. 12


Drilling Advances

Into the weeds

Jim Redden, Contributing Editor

Determining the true value of any innovation is hardly possible, if you have no idea of precisely what you’re measuring against. That truism is particularly applicable when it comes to assessing the real-world benefits of automated slide drilling, where broad performance metrics like on-bottom ROP are insufficient barometers for quantifying whether a new system is successful, as is, or needs tweaking.

So says Scott Coffey, product manager in the Drilling Technology Automated Performance Group of Nabors Drilling Technologies, pointing to the need for more granular, and scalable, key performance indicators (KPI) than those typically employed in a drilling operation. “In our group, we understood early in our journey to automated directional drilling and automated slides that the KPIs that we were using were insufficient to drive the development of automated directional drilling,” he told the September IADC Drilling Engineering Committee (DEC) quarterly Technology Forum in Houston. “Metrics, like well cycle time for on-bottom ROPs, don’t have the granular information necessary to describe the performance of automation.”

To stress the import of granularity in performance metrics, Coffey drew an analogy to the recently deployed On-Road Integrated Optimization Navigation (ORION) system of United Parcel Service, commonly known as UPS. In developing its route optimization technology and associated driver navigation tool, the package delivery heavyweight aggregates data into minute detail that is continuously updated to break down the vehicle route, right and left turns, the time the vehicle spends idling, and even how many times the driver fastens and unfastens the seat belt. “The value to this approach to UPS is tremendous,” Coffey said. “They say if they can eliminate one mile per driver per day, that’s $50 million in savings over the course of a year. The reason this is possible is that UPS recognized the value of measuring these very small KPIs.”

The micro approach. In relating well cycle time to how long it takes a UPS driver to make deliveries during the course of a day, speeding up the process obviously is the overriding goal, but that metric, in and of itself, fails to explain the steps taken to make the improvements, Coffey said. “The other issue with these typical macro KPIs is that they can be skewed by things that happen on a single well, like a motor failure or unexpected trip, or, in the case of a UPS drive, a flat tire. The UPS driver may have an optimized route, but with a flat tire that macro metric is useless.”

For automated slide drilling, Coffey said KPIs had to be devised that would provide detailed measurement of activities directly affected by automation. “And, how do we use those measurements to drive performance of automated systems and make development decisions about what the product should do in the future? Also, how do we take these measurements off one rig and put them on 100 rigs?”

To that end, first consider the anatomy of an automated slide drilling operation: rotate, come off bottom, set the toolface, execute the slide and transition back to rotary drilling. “On top of these (activities), we put KPIs in place that describe the performance and precision of the automated directional drilling and sliding system,” Coffey said.

While components like rotary time are aggregated into a single data set, off-bottom activities are subdivided into the time spent specifically to work torque out of the drillstring, for instance, and the time required to set the toolface. “What’s important to us is what we call the slide rate, which is a speed metric that is the footage drilled divided by the entire rotating time and is inclusive of the off-bottom functions that are impacted by automation. This gives us a better representation of the true cost of the slide,” he said, explaining that the benefits of sliding twice as fast as a manual operation can be partially offset when too much time is spent lining up the toolface.

“We also have some precision metrics to describe the precision of the toolface. Here, we have a slide score to describe the toolface distribution, which is the percentage of the toolface values with “X” degrees of the target toolface, and the burn footage, which is the initial footage drilled in the wrong direction,” Coffey said.

Speed versus precision. Coffey cited a comparative analysis of 120 wells, 80 of which were landed with automated directional drilling and sliding and 40 drilled manually. Focusing strictly on cycle time, the automated wells were drilled some 6% faster than the manual wells. “That’s good to see, obviously, but that information is not useful by itself. We need to go deeper to see exactly what’s happening.”

The time savings attributed to faster toolface orientation amounted to less than one hour in the targeted interval, which may seem inconsequential at first blush. “The critical point, however, is that because these processes are measured and, because they are automated, we have very fine control over this activity and can make improvements and scale those improvements over multiple rigs.”

Digging deeper, however, the data show the manual wells outperforming the automated wells with respect to less burn footage and comparable or better toolface distribution. “This is a great illustration of the trade-offs that exist between precision and speed. The automated slides operated faster, but we’re trading away some of the precision to get that performance,” he said. “This may not be acceptable, but at least we can decide if we need to make changes within the existing automated framework or if it will require additional product development.”

The Authors ///

Jim Redden is a Houston-based consultant and a journalism graduate of Marshall University, has more than 38 years of experience as a writer, editor and corporate communicator, primarily on the upstream oil and gas industry.

Related Articles ///

  • No related articles found


Comments ///


{{ error }}
{{ comment.comment.Name }} • {{ comment.timeAgo }}
{{ comment.comment.Text }}