In the age of digitalization and automatization of drilling operations, it is time to move the detailed drilling operating procedure from its classical text format, only intended for human interpretation, to a structured representation that can be utilized efficiently by computer systems. The immediate benefit of this transformation is that progress tracking during an operation can be fully automated. To verify that automatic progress tracking can be performed with a digital version of the detailed operating procedure, simulated operations have been fed to the digital procedure and activity recognition software. The matching quality is thereafter assessed for each step of the planned drilling and non-drilling operating procedure. The detailed digital operating procedure itself is automatically generated from general descriptions contained in the operation program like tubular strings and fluids, and operational limits. The testing procedure utilized five operations generated on the fly by a simulator: running in hole and tagging bottom, setting liner hanger and releasing running tool, pull out of hole, spacing out a completion seal assembly in a liner top and setting and testing a production packer. The simulated operations were inspired by actual operations recorded in a real-time data server. The simulator runs remotely and is accessible through different application programming interfaces. The simulated data stream is very realistic compared to recorded data but offers the possibility to make variations on the simulated sequence and therefore test the reliability of the operation recognition algorithm. From a technical point of view, there are no differences where the real-time data flow comes from and what is the source of the data, whether it comes from a real-time data server, drilling simulator or any other source of the real-time data. All the simulated operation steps were successfully recognized by the system for the test case scenarios and proper notifications were triggered when required which proves the effectiveness of both the implemented system and the simulator. The new system is a unique approach to automatically digitalize operating procedures and activities and track them. The method also automatically warns when execution deviates from the plan and allows the instant deployment of best practices.

The exact definition of all types of activities in well construction, from spud to completion, is an area of great challenges for an automation system to function successfully in. In an operation plan, these activities can be categorized into three subgroups: standard and repetitive sub-activities, customized sub-activities, and manual sub-activities. A digitalized detailed operation procedure (DOP) provides the appropriate context by defining the machine-readable version of these activities. Digitalizing a DOP is achieved by splitting it into individual parts, categorizing them into activities, checks, risks, lessons learned, best practices, and operational parameters and constraints. This paper focuses on manual sub-activities (e.g. close/open valves or nippling up) in which start and end time are acknowledged manually; pre-defined sub-activities (e.g. drilling and tripping) which are defined by repetitive tasks (e.g. drilling, lower string, lifting string and connections); and custom sub-activities (e.g. recording slack off and pick up weight or pressure tests) which can be customized for a tool, company, or operation with limited tasks (e.g. increase flowrate or reduce rotation). Adopting this approach guarantees:An unbroken consistency between the planned, executed and reported phases of an operation as the digitalized DOP is directly connected to the automation system and the detailed log of executed activities are recorded.Enabling automation of longer sequences as the rig state transition between activities is managed seamlessly. This also applies when it is interrupted by manual operations like making up casing hangers.Quality check and improvement of the automatically performed activities by using real-time models. This is not limited to the normal repetitive sub-activities such as drilling or tripping. They can be used for every related custom sub-activity, e.g., hydraulic modelling after the opening of a circulation sub or torque and drag calculations after releasing a liner running tool. Having a software library consisting of categorized, distinguishable manual and automatic activities lets the user describe with precision the operational tasks in a standardized digital format. With each automatic activity being readable and recognizable by the rig automation system, the system can execute longer and broader sequences of activities securely and cost-effectively, where the sequence is well defined.