Abstract
In the petroleum industry, software systems designed to monitor real-time drilling parameters are constantly evolving, with the aim of providing improved capabilities in the early detection of dysfunctions during the construction of wells. When optimally applied and embedded within a drilling team, such software systems have been proven to be of great benefit to the operation by helping to avoid major drilling problems. The trust in and the implementation of the information provided by such software are crucial factors of success, and herein lay the biggest challenge.
This paper investigates the application of a cutting-edge, early drilling problem detection system for the drilling of a complex, shallow horizontal exploration well. It determines if the use of the system was successful and looks at improvements that can be made to the system for future operations. The study was performed by a combination of reviews of the software output, interviews with key personnel involved, experience's from other cases, and theoretical knowledge.
The case well was drilled offshore Norway in the Barents Sea, with a rig contracted solely for this well. The operator employed a permanent drilling team in the local office in Stavanger, with key personnel located on the rig. The company providing the early drilling problem detection system had an engineer at the rig site as part of the drilling team. The system installation was coordinated by the local office while being supported in setup and organizational matters by the well engineering global support team at the operator's head office in Vienna. The implemented system simulated the physics in the well bore to provide early identification of drilling problems, and produced dynamic look-ahead simulations to assist drilling decisions. Key factors such as communication patterns, when and which functionalities of the system were used, how the results were fed back into the operational decision making, the organizational set up, and the physical placement of the various personnel and the technology used will be investigated. Additionally, the paper will look at the learnings and potentials for future implementations of such a system.
