Application of Machine Learning in Different Stages of Oil Reservoir Development

Geological and oilfield big data is exponentially expanding. The traditional methods used to identify reservoirs and predict production cannot use historical information and new data effectively. The processes of well logging interpretation and pipeline non-destructive examination (NDE) are time consuming and subjective. Numerical flow simulation models do provide a relatively reliable and appropriate approach to conduct a reservoir analysis, but they are laborious and time consuming. In today’s big data environments, it is increasingly necessary to develop an effective and dependable technique to maximize the benefits of a growing data explosion and extract useful information within all the oilfield data. A machine learning method incorporates various algorithms that provide powerful functions in an oilfield. Massive static and dynamic data is put into training models to identify valuable features and learn nonlinear relationships between different variables and output targets. Advanced models using the benefits of machine learning (ML) will help operators to implement classification and/or prediction tasks. This study compares various ML methods applied to different stages from oil and gas exploration to transportation in oilfields: reservoir identification, prediction of production in new and old wells and non-destructive examination (NDE) of pipelines. These ML methods are proven useful and fast to resolve reservoir classification and production prediction challenges. This work provides a set of systematic ML methods and their respective pertinent predicting parameters providing useful experiences and references for industry and future relative research.

Name thesis:degree_name

Doctor of Philosophy (PhD)

Discipline thesis:degree_discipline

Engineering – Chemical & Petroleum

Grantor dc:publisher.institution

Schulich School of Engineering

Year dc:date.issued

2023

Author dc:creator

• Wei, Liu

Advisor dc:contributor.advisor

• Chen, Zhangxing

Committee members dc:contributor.committeemember

• Roman, Shor
• Qingye, Lu
• Haiping, Huang
• Yuntian, Chen

• Reservoir identification
• Production prediction
• Pipeline safety
• Machine learning

Statement dc:rights

• University of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission.

Language dc:language.iso

en

DOI

https://dx.doi.org/10.11575/PRISM/41555

OAI identifier oai:identifier

oai:ucalgary.scholaris.ca:1880/116713