Making Sense of Mudrocks

November 12, 2020

A computer program created at the Bureau of Economic Geology is speeding up rock core analysis in mudrocks.

Called CorePy, the program automatically organizes geochemical and geomechanical data points into graphics that showcase the statistical distribution of important rock attributes in real time. The raw data is collected from the core through high-resolution X-ray fluorescence scanning.

Researchers at the bureau’s Mudrock Systems Research Lab team used the programming language Python to create the program, which applies K-means clustering to integrate data analytics with graphical visualization.

“High-resolution X-ray fluorescence is a powerful geochemical tool used to characterize geological core, but geologists have struggled to fully utilize the large multivariate data sets that it generates,” said Research Associate Toti Larson, who leads research at the lab. “The application allows researchers to integrate these complex data sets and visualize the results directly on core photographs, and to apply new data analysis tools, quickly visualize the results and better characterize the core.”

The team has collected core data sets from samples from the Eagle Ford Shale, Permian Basin, Vaca Muerta Formation, Haynesville, Barnett and other mudrock plays. In addition to Larson, the research team included Priyanka Periwal and Evan Sivil, both research scientist associates, and graduate student Esben Pedersen.