Click on the buttons below to access this repo trhough Binder or Google Colab:
Instructors: Lalitha Venkataramanan (SLB), Chicheng Xu (Aramco), Andy McDonald (LR), Vikas Jain (SLB)
- About the Course
- Using the Notebooks
- About the Instructors
- Volve Dataset
- Kansas Geological Survey (KGS)
This workshop will focus on the applications of Artificial Intelligence (AI) and Machine Learning (ML) to the upstream O&G industry. Consisting of two half-days, the workshop will provide an introduction to machine learning, lay out sample workflows and steps for ML applications and summarize some of the used cases in the industry.
The workshop will cover both supervised and unsupervised learning and highlight applications such as QA/QC, outlier detection, facies mapping and learning complex functional mapping.
Hands-on tutorials with Python codes to analyze a publicly available data set will also be provided.
This repository contains a series of notebooks that will be used during the workshop.
There are multiple options for following along:
The first option is to clone the repository to a local folder and run using your own version of Python & Jupyter. If you do this, please make sure you install the required packages listed in requirements.txt. See here: https://docs.github.com/en/github/creating-cloning-and-archiving-repositories/cloning-a-repository for details on how to clone a repository and here https://note.nkmk.me/en/python-pip-install-requirements/ for details on how to install the required packages
However, if you are restricted from installing Python or are unable to load the required libraries, you can use the binder link https://mybinder.org/v2/gh/SPWLA-ORG/spwla2021_ml_workshop/HEAD or use Google Colab: https://colab.research.google.com/. Google Colab will require some setup and instructions on how to do this can be found here: https://medium.com/analytics-vidhya/how-to-use-google-colab-with-github-via-google-drive-68efb23a42d
Lalitha Venkataramanan is the Reservoir Performance: Data Science Advisor in Schlumberger. She is also a Scientific Advisor and the Associate Editor for NMR-Petrophysics. She is regional distinguished speaker for SPWLA for 2018-19. She is on the board of SIAM and NSERC as well as Business-Industry-Government Math network. Her current interests include machine learning, mathematical modeling and inversion, optimization, probability and stochastic processes. Trained as an Electrical Engineer, she obtained her M.S. and Ph.D. degrees from Yale University in 1998. She has co-authored 40+ peer-reviewed publications and has over 24 granted US patents and 18 pending patent applications.
Dr. Chicheng Xu is currently working in Aramco Houston Research Center as a research petrophysicist with a focus area in Petrophysics Data-Driven Analytics that utilizes advanced computational techniques and artificial intelligence/machine learning for interpretation, classification, and modeling based on multi-scale data integration. He obtained his PhD degree in petroleum engineering from the University of Texas at Austin in 2013 and had previously worked in various technology departments of Schlumberger, BP, and BHP Billiton. He was selected to receive the SPE Gulf Coast Regional Formation Evaluation Award in 2018 and the outstanding Associate Editor service award of SPE REE journal in 2020.
Andy McDonald is a Petrophysicist and has been with Lloyd’s Register for 9 years. He provides petrophysical expertise to software development projects, has a strong interest in Python programming and applications of Machine Learning to the geoscience domain. He holds an MSc in Earth Science and a BSc (Hons) in Geology & Petroleum Geology. Andy has co-authored several technical papers for SPWLA and SPE conferences covering machine learning, heavy oil, and low salinity waterflooding.
Vikas Jain is currently working in Schlumberger’s Analysis and Interpretation business as Principal Petrophysicist and Global Domain Head of Petrophysics and Acoustics, based out of Houston, U.S.A. His focal area is development of custom and intelligent solutions, leveraging machine learning and artificial intelligence, to help solve common upstream challenges. Previously he was managing Interpretation Engineering projects of NMR Answer Products and Next Generation Data-Driven Petrophysics in Schlumberger’s Engineering Center in Houston. He started his career with Schlumberger in 2001 as a field engineer. Since then, he has held positions in operations, sales, marketing, domain and engineering. He has authored more than 35 publications and 15 patent applications/grants.
In 2018, Equinor released the entire contents of the Volve Field to the public domain to foster research and learning. Data includes:
- Well Logs
- Petrophysical interpretaions
- Reports
- Core measurements
- Seismic data
- Models
- And more
The data is licensed under the Equinor Open Data Licence.
The Volve Field is located some 200 km west of Stavanger in the Norwegian Sector of the North Sea. Hydrocarbons were discovered within the Jurassic aged Hugin Formation in 1993. Oil production began in 2008 and lasted for 8 years (twice as long as planned) until 2016, when production ceased. In total 63 MMBO were produced over the field's lifetime and reached a plateau of 56,000 B/D.
Details for the Volve Field and the entire dataset can be found at: https://www.equinor.com/en/what-we-do/norwegian-continental-shelf-platforms/volve.html
The full licence agreement can be found here: https://www.equinor.com/content/dam/statoil/documents/what-we-do/Equinor-HRS-Terms-and-conditions-for-licence-to-data-Volve.pdf
Data released by the Kansas Geological Society is public domain and was covered in the Supervised Learning section by Chicheng. The data can be accessed here: http://www.kgs.ku.edu/PRS/Ozark/well_1_32.html
The well used for the analysis can be found in the data subfolder with the file Supervised_Learning - WELLINGTON_1_32_LAS_SP_ADDED.las
To load LAS files into Python, you can use the LASIO library. A general intoduction to using LAS files in Python can be found at the following links:
https://andymcdonaldgeo.medium.com/loading-and-displaying-well-log-data-b9568efd1d8
https://towardsdatascience.com/loading-multiple-well-log-las-files-using-python-39ac35de99dd