In the last decade, the oil and gas industry has witnessed the emergence of rotary steerable systems (RSS) that led to certain achievements that were not possible with conventional mud motors. Rotary steerable systems enable faster drilling, smoother wellbores, and extended-reach drilling (ERD).

Still, a drilling assembly built around an RSS tool is sensitive to factors such as bit type, operating parameters, type of drilling fluid, lithology, and borehole diameter. Therefore, it is not always optimal for a human operator (typically a directional driller) to select and issue the right commands to the drilling tool at the right time. This situation becomes even more challenging when those decisions have to be made in a matter of minutes.

Steering the well accurately is paramount to geo-steering, optimal well placement in the reservoir, and anti-collision.

This paper discusses a new system that monitors all real-time data that is available, “learns” the steering behavior of the drilling assembly, and uses the acquired information to create more accurate projections for the directional driller. The system recommends the optimal command to direct the drilling tool according to plan.

This tool is currently being used by a number of field locations as an advisor. The next version of the system, being field tested, autonomously issues downlink commands directly to the RSS tool, making it an automatic trajectory controller. The concept is very similar to the autopilots used in commercial airplanes today.

The anticipated value from this technology is better service quality and higher performance that can be delivered predictably and consistently.

Preliminary results look very encouraging. During the evaluation period, the system has accumulated approximately 5,000 running hours, corresponding to approximately 72 runs. These tests are in various hole sizes and well profiles, using different RSS tools.