Stuck pipe prediction in deviated wellbores: a numerical and statistical analysis.

Due to the significant non-productive times and recovery costs associated with stuck pipe events in oil and gas drilling operations, there is value in being able to predict an impending stuck pipe event. To achieve this, the use of numerical cuttings transport (hole cleaning) models and statistical analysis of real-time drilling data is proposed by this research. Current cuttings transport models are based on unhindered, free settling in the wellbore and do not adequately account for the effect of vortices created as the drill string rotates about its axis. This thesis addresses both shortcomings, and presents improved cutting transport models that consider hindered centrifugal settling of drilled cuttings, effect of Taylor vortices and Van der Waals forces. The implication is that the resulting cuttings settling velocity used to estimate critical transport velocities and flow rates are more representative. The transport ratio, a measure of the hole cleaning efficiency, is consequently more realistically predicted. Although several proprietary automated stuck pipe prediction tools exist in the industry, this research found that they broadly fall into five main groups. It is also apparent that current capabilities do not simultaneously and continuously combine real-time data, offset wells data and well design analytical models in a single approach. On that basis, this thesis presents an integrated stuck pipe prediction concept that utilizes all three data streams, called the "ROW" approach. The concept presented in this thesis was then coded into a tool called the stuck pipe index (SPI). The SPI tool risk assessment is determined in real-time and is referenced by a traffic light alert system (green – amber – red), to warn the user of an impending potential stuck pipe situation. The numerical models developed in this research estimate critical velocities to within 10 – 15% and show strong agreement with published empirical data. Combined with the cuttings transport numerical models developed in this research and other publicly available well design models (such as hydraulics, and torque and drag), the SPI tool has been tested with several case histories and proven to detect stuck pipe events with warning alerts significantly ahead of the event. The tool has equally been deployed in real-time with >90% success rate and without spurious alerts recorded. The results thus confirm that the developed numerical models and the "ROW" approach are robust, and offer an improvement to current industry capabilities in terms of accuracy and sensitivity to changing downhole wellbore conditions.

Name dc:type.qualificationname

PhD

Level dc:type.qualificationlevel

Doctoral

Grantor dc:publisher.institution

Robert Gordon University

Year dc:date.issued

2022

Author dc:creator

• Egbe, Peter Ido

Advisor dc:contributor.advisor

• I. Iyalla and B. Oyeneyin

• Failure prediction
• Stuck pipe prediction
• Wellbores
• Critical transport velocities
• Wellbore cuttings transport

Language dc:language

en

Identifier

oai:rgu-repository.worktribe.com:2071873
https://doi.org/10.48526/rgu-wt-2071873

Author Identifier

0000-0002-4128-7648

OAI identifier oai:identifier

oai:rgu-repository.worktribe.com:2071873