Wellbore Health Characterization for Drilling Optimization

Publikationen: Thesis / Studienabschlussarbeiten und HabilitationsschriftenMasterarbeit

Standard

Wellbore Health Characterization for Drilling Optimization. / Les, Borna.
2024.

Publikationen: Thesis / Studienabschlussarbeiten und HabilitationsschriftenMasterarbeit

Harvard

Les, B 2024, 'Wellbore Health Characterization for Drilling Optimization', Dipl.-Ing., Montanuniversität Leoben (000).

APA

Les, B. (2024). Wellbore Health Characterization for Drilling Optimization. [Masterarbeit, Montanuniversität Leoben (000)].

Bibtex - Download

@mastersthesis{08c5d006c5a844a89c7ee6f02094313b,
title = "Wellbore Health Characterization for Drilling Optimization",
abstract = "Given the scale of oil and gas projects and the complexity of systems involved, it is evident that ramifications of poor wellbore execution can manifest throughout the whole well lifecycle. Minimizing these ramifications calls for proactive management of wellbore quality already during well construction. However, the term `wellbore quality¿ remains loosely defined in this context, although commonly associated with adherence to a certain time, conditions, and cost constraints. While numerous studies suggest a positive correlation between increased wellbore quality and drilling performance, proving this beyond a reasonable doubt remains challenging due to the complexity of the drilling environment. Nonetheless, improved downhole conditions are generally agreed to enhance the likelihood of meeting well objectives, through either improved operational sequences or drilling performance, establishing a presumed link between wellbore quality and downhole conditions. Many efforts have been undertaken so far to characterize or forecast downhole problems through data-driven and/or physics-based models, there is, to the author¿s knowledge, no established method for quantifying wellbore acceptability. To tackle this issue, this thesis introduces a data-driven framework designed to diagnose and quantify signs of compromised wellbore health, aiming to provide means of quantifying the degree of wellbore acceptability in real time. The study introduces a high-dimensional concept to monitor the drilling process¿s health by creating a healthy reference baseline, combining deep learning (DL) and readily available data-derived metrics. Much like medical diagnostics, drilling anomalies are interpreted as symptoms, and hypothesized to correlate with a deterioration of wellbore health, with their intensity and frequency seen as indicators of wellbore health deterioration. The measure of drilling conditions acceptability is proposed in the form of the wellbore health value metric referred to as the `Symptom severity¿ index. The proposed index aims to refine wellbore quality management during drilling by deepening insight into downhole conditions, thereby facilitating timely interventions and mitigating unwarranted remedial actions. Ultimately, the index is intended to optimize drilling process sequences, thus helping to conserve resources and minimize emissions associated with the well construction process.",
keywords = "Drilling Health Characterization, Drilling Data Analytics, Wellbore Quality, Advisory System, Machine Learning, Deep Learning, Long Short-Term Memory Networks, DBSCAN, Bohrlochgesundheit, Bohrdatenanalytik, Bohrlochqualit{\"a}t, Beratungssystem, Machine Learning, Deep Learning, Long Short-Term Memory Networks, DBSCAN",
author = "Borna Les",
note = "no embargo",
year = "2024",
language = "English",
school = "Montanuniversitaet Leoben (000)",

}

RIS (suitable for import to EndNote) - Download

TY - THES

T1 - Wellbore Health Characterization for Drilling Optimization

AU - Les, Borna

N1 - no embargo

PY - 2024

Y1 - 2024

N2 - Given the scale of oil and gas projects and the complexity of systems involved, it is evident that ramifications of poor wellbore execution can manifest throughout the whole well lifecycle. Minimizing these ramifications calls for proactive management of wellbore quality already during well construction. However, the term `wellbore quality¿ remains loosely defined in this context, although commonly associated with adherence to a certain time, conditions, and cost constraints. While numerous studies suggest a positive correlation between increased wellbore quality and drilling performance, proving this beyond a reasonable doubt remains challenging due to the complexity of the drilling environment. Nonetheless, improved downhole conditions are generally agreed to enhance the likelihood of meeting well objectives, through either improved operational sequences or drilling performance, establishing a presumed link between wellbore quality and downhole conditions. Many efforts have been undertaken so far to characterize or forecast downhole problems through data-driven and/or physics-based models, there is, to the author¿s knowledge, no established method for quantifying wellbore acceptability. To tackle this issue, this thesis introduces a data-driven framework designed to diagnose and quantify signs of compromised wellbore health, aiming to provide means of quantifying the degree of wellbore acceptability in real time. The study introduces a high-dimensional concept to monitor the drilling process¿s health by creating a healthy reference baseline, combining deep learning (DL) and readily available data-derived metrics. Much like medical diagnostics, drilling anomalies are interpreted as symptoms, and hypothesized to correlate with a deterioration of wellbore health, with their intensity and frequency seen as indicators of wellbore health deterioration. The measure of drilling conditions acceptability is proposed in the form of the wellbore health value metric referred to as the `Symptom severity¿ index. The proposed index aims to refine wellbore quality management during drilling by deepening insight into downhole conditions, thereby facilitating timely interventions and mitigating unwarranted remedial actions. Ultimately, the index is intended to optimize drilling process sequences, thus helping to conserve resources and minimize emissions associated with the well construction process.

AB - Given the scale of oil and gas projects and the complexity of systems involved, it is evident that ramifications of poor wellbore execution can manifest throughout the whole well lifecycle. Minimizing these ramifications calls for proactive management of wellbore quality already during well construction. However, the term `wellbore quality¿ remains loosely defined in this context, although commonly associated with adherence to a certain time, conditions, and cost constraints. While numerous studies suggest a positive correlation between increased wellbore quality and drilling performance, proving this beyond a reasonable doubt remains challenging due to the complexity of the drilling environment. Nonetheless, improved downhole conditions are generally agreed to enhance the likelihood of meeting well objectives, through either improved operational sequences or drilling performance, establishing a presumed link between wellbore quality and downhole conditions. Many efforts have been undertaken so far to characterize or forecast downhole problems through data-driven and/or physics-based models, there is, to the author¿s knowledge, no established method for quantifying wellbore acceptability. To tackle this issue, this thesis introduces a data-driven framework designed to diagnose and quantify signs of compromised wellbore health, aiming to provide means of quantifying the degree of wellbore acceptability in real time. The study introduces a high-dimensional concept to monitor the drilling process¿s health by creating a healthy reference baseline, combining deep learning (DL) and readily available data-derived metrics. Much like medical diagnostics, drilling anomalies are interpreted as symptoms, and hypothesized to correlate with a deterioration of wellbore health, with their intensity and frequency seen as indicators of wellbore health deterioration. The measure of drilling conditions acceptability is proposed in the form of the wellbore health value metric referred to as the `Symptom severity¿ index. The proposed index aims to refine wellbore quality management during drilling by deepening insight into downhole conditions, thereby facilitating timely interventions and mitigating unwarranted remedial actions. Ultimately, the index is intended to optimize drilling process sequences, thus helping to conserve resources and minimize emissions associated with the well construction process.

KW - Drilling Health Characterization

KW - Drilling Data Analytics

KW - Wellbore Quality

KW - Advisory System

KW - Machine Learning

KW - Deep Learning

KW - Long Short-Term Memory Networks

KW - DBSCAN

KW - Bohrlochgesundheit

KW - Bohrdatenanalytik

KW - Bohrlochqualität

KW - Beratungssystem

KW - Machine Learning

KW - Deep Learning

KW - Long Short-Term Memory Networks

KW - DBSCAN

M3 - Master's Thesis

ER -