Abstract
Abstract
Well construction is a lengthy, complex and an expensive process. A drilling rig can spend several months to complete the operation in a single well. When analyzing the type of operations performed by rigs, we can see that almost 30% of the time is consumed while making new hole (i.e., drilling time), while the majority of the rig time represents what is called the flat time. Any optimization of flat time can significantly reduce the well construction cost.
Flat time is a group of operation types that involve no drilling operations, such as BHA making, drill pipe tripping, hole conditioning, casing running, cementing, etc. For instance, the performance of two specific types of these operations is only determined by rig crew and rig capabilities, and is not influenced by hole condition or formation related factors. These operations are the drill pipe tripping in cased hole (DPTCH) and the drill pipe connection operations (CONN). In typical well construction operations, the DPTCH and CONN operations represent 30% and 20% of the flat time, respectively.
In this paper, we discuss the potential benefits of developing and deploying an automated solution to monitor and alert the performance of these two operations in real time. The paper will explain how the solution can employ advanced machine learning (ML) algorithms to process rig operation sensor data and realize rig states. Then, we discuss the impact on the operation when applying this solution to optimize the DPTCH and CONN operations.
A developed solution of real-time monitoring and alerting can contribute to a significant time and cost savings. It also can improve early warning signs and detection of operation troubles. Also, it can be the foundation of many real-time operation optimization solutions.