Application of Machine Learning to Augment Wellbore Geometry-Related Stuck Pipe Risk Identification in Real Time

Othman, Eswadi Bin; Gomes, Dário; Bidin, Tengku Ezharuddin Bin Tengku; Hashim, Meor M. Hakeem Meor; Yusoff, M. Hazwan; Arriffin, M. Faris; Ghazali, Rohaizat

doi:10.4043/31695-ms

Cited by 10 publications

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Evaluating Multi-target Regression Framework for Dynamic Condition Prediction in Wellbore

Keshavarz,

Elmgerbi,

Vita

et al. 2024

Arab J Sci Eng

View full text Add to dashboard Cite

In recent years, the focus has shifted towards leveraging physics-based modelling and data-driven analysis to predict drilling incidents and anomalies in real time, with the goal of reducing non-productive periods. However, much of this attention has directed at specific drilling operations like drilling and tripping, leaving other vital processes, such as wellbore conditioning, comparatively overlooked. The primary objective of this study is to employ data-driven techniques for predicting the dynamic state of the wellbore by utilising sensor data, operating parameters, and surface measurements. Accurate predictions are pivotal for automating these processes, promising significant savings in both redundant time and associated costs, ultimately elevating operational efficiency.In this research, the surface drilling parameters such as flowrate, rotation speed, block position, and drill string length are incorporated with the surface measurements such as hookload, pressure, and torque during wellbore conditioning operation to predict further surface sensor measurements. Different parameter settings are evaluated to find the best approach. Six supervised learning algorithms are used to select the best prediction method. The findings reveal that considering all surface parameters and measurements yields the most accurate predictions. Among various single and multi-target regression methods, including deep learning approaches, the Gaussian process and random forest models exhibit the lowest prediction errors.By reliably predicting and understanding wellbore behaviour, this research paves the way for more efficient and autonomous drilling operations in the future, bridging a critical gap in the industry's automation capabilities.

show abstract

Evaluating Multi-target Regression Framework for Dynamic Condition Prediction in Wellbore

Keshavarz,

Elmgerbi,

Vita

et al. 2024

Arab J Sci Eng

View full text Add to dashboard Cite

show abstract

Automated Detection of Rig Events from Real-Time Surface Data Using Spectral Analysis and Machine Learning

Robinson¹,

Revheim²

2023

Day 3 Thu, March 09, 2023

View full text Add to dashboard Cite

The authors present a method for automated, high-fidelity detection of rig events characterized by complex temporal signals, such as downlinking, or wave-induced heave affecting floating rigs. These can adversely impact other systems utilizing relevant data streams, for example downlinking via mud pulse telemetry can interfere with detection of pressure changes that might indicate hole cleaning problems. Identifying these events using classification techniques applied to time-domain data is difficult, hence spectral (frequency domain) techniques, combined with Machine Learning (ML), were applied to solving this problem. Surface measurements from a variety of wells, fields, regions, service companies and operators were used to develop and validate the detection methods. Data was preprocessed using time-frequency analysis, and then input to discriminative classifiers to identify rig events of interest. For downlinking state detection, high recall and precision scores (both >93%) were achieved on independent holdout well data, and thus false positive rates were low. Successful detection was demonstrated on wells separate from the training data, hence the method is expected to generalize to new well operations. The detection method enhances situational awareness, and can actively support other software in improved automated decision-making by providing operational context in real-time, such as suppression of false warnings from monitoring pressure or modelled ECD for detecting signs of poor hole cleaning. These techniques are not limited to downlinking or heave detection, and can be applied more generally to scenarios with complex periodic signals.

show abstract

Stuck Pipe Prediction in Utah FORGE Geothermal Wells

Montes,

Ashok,

van Oort

2023

Day 1 Mon, October 16, 2023

View full text Add to dashboard Cite

Four stuck pipe events occurred during the construction of six geothermal wells in the Utah FORGE field, causing several hours of non-productive time (NPT). All of them could have been avoided if effective preventive measures and tools had been used. This paper provides an analysis of these events with the publicly available Utah FORGE dataset and discusses their predictability using a new hybrid approach that can in future be used to avoid stuck pipe incidences at Utah FORGE. Our analysis is primarily based on physics-based simulations combined with well information, such as trajectory, lithology, and the position of the drillstring (specifically the contact points) with respect to the geometry during the time the events occurred. Based on this detailed analysis, we implemented anomaly detection techniques in the time window before each stuck pipe event to automatically identify root causes. Following this, we formulated guidelines that combine the developed early stuck pipe event detection capability with accepted industry operational practices, to prevent similar stuck events in the future. The Utah FORGE stuck pipe events were found to be primarily related to wellbore geometry (i.e., mechanical stuck pipe) and use of non-optimum operational drilling practices. The incidents, analyzed in time domain after dimensional transformations, show clear early signs of anomalies in the time frame before their occurrence. This means they could have been detected and potentially prevented. The guidelines include recommendations for acting pro-actively on early stuck pipe indicators identified by real-time data analysis, as well as making appropriate additional changes during well planning and execution phases, such as changing the bottom hole assembly to be less sensitive to mechanical sticking. The work shows the high value of robust predictive models for stuck pipe prevention. The work will help reduce geothermal well construction (which is expected to expand significantly in the near future) cost by enabling better stuck pipe NPT avoidance. However, the approach is not specific to geothermal well construction and can be expanded to all drilling applications as well as stuck pipe incidents that are non-geometrical in origin. This paper therefore provides a new real-time data analysis system and effective guidelines which, if implemented, will enable better mitigation of stuck pipe risks on future wells.

show abstract

Application of Machine Learning to Augment Wellbore Geometry-Related Stuck Pipe Risk Identification in Real Time

Cited by 10 publications

References 6 publications

Evaluating Multi-target Regression Framework for Dynamic Condition Prediction in Wellbore

Evaluating Multi-target Regression Framework for Dynamic Condition Prediction in Wellbore

Automated Detection of Rig Events from Real-Time Surface Data Using Spectral Analysis and Machine Learning

Stuck Pipe Prediction in Utah FORGE Geothermal Wells

Contact Info

Product

Resources

About