Automated Hole Cleaning Monitoring: A Modern Holistic Approach for NPT Reduction

Abstract

NPT is thought to account for upwards of 30% of the upstream costs in O&G production, half of which can be attributed to downhole drilling problems, within which hole cleaning is arguably the most prominent offender, accountable for tens of millions of dollars in lost time costs. The objective of this paper is to outline a modern, holistic, automated approach to hole cleaning that closely replicates human experience within a suite of software applications, allowing the early identification, and therefore early mitigation and remediation of this significant drilling issue. The principles of hole cleaning as a drilling process are generally considered to be well understood, and both the accurate diagnosis during the drilling operation and the means to either mitigate or remediate the issue exist. While significant technological advances regarding detection of hole cleaning dysfunctions have been made in various individual approaches, the best certainty in identification has still been achieved where a human aggregates the individual results of these. The paper explores how such a holistic approach can be automated through an innovative digital solution. This paper describes the individual advances made in individual system components, including real-time engineering modelling, surface and downhole direct measurements with intelligent agents, and offset experience. We then look at how these various components can be aggregated by an automated higher-level process to produce a result akin to that of an experienced human at the wellsite. Case studies from initial system deployments demonstrate both the validity and the financial impact of such an approach. Several of the ‘component’ services are described in more detail, including topics within torque and drag and pressure regime modelling are advances on current practice. Additionally, the concept of exploring aggregated multiple point-source findings in an application, just as a human would, is a framework that has significant potential when automating the detection of other drilling dysfunctions. These concepts are expected to greatly reduce drilling risks and achieve predictable drilling performance at scale, in various drilling environments.