Petrophysics, Reservoir, Completions, Software
Automated Well Log Editor SPWLA Paper
Automated Well Log Editor SPWLA Paper

Automated Well Log Editor SPWLA Paper

I am proud to announce that PetroRes and LR-Senergy have successfully published and presented a paper this year at the annual SPWLA symposium. Thanks to Andy McDonald and Tegwyn Perkins for co-authoring the paper with me.

With the recent ‘machine learning and data science’ renaissance it was an appropriate time to finally publish on the Automated Well Log Editor that was developed almost 10 years ago now (and widely used since). Nothing flatters me more than to know this piece of software has shown up at many operating companies over the years and is now included as part of Interactive Petrophysics. It is a time saving utility that does an excellent job at conditioning well logs–a process that petrophysicsts usually try to avoid. With this utility, we make it fun.

The basis of the paper is an algorithm that utilizes combinatorics and multiple-linear regression in order to predict and automatically repair (viz. splice) well log curves. In the paper, we discuss the theory behind the algorithm, some applied best practices, pitfalls, nuances, and taking things a bit further than what is currently implemented in IP. One of the most important improvements is the use of training data leveraged from multiple wells to make predictions. Concepts such as automatically generating badhole flags are also discussed.

The paper is titled “Novel Methodology For Automation Of Bad Well Log Data Identification And Repair” presented at the 62nd annual SPWLA logging symposium, May 19th, 2021. I look forward to further developments improving this technology and making it available to a wider audience.