What does the subsurface look like at locations we have not drilled? How can we predict rock properties in 8 km depth? How can we decide where to drill a well with economic success?

In this course, we will dive into techniques to explore what we do not know about the subsurface – and learn how to take decisions under uncertainty. Forecasting future behavior of geothermal installations, ground water flow, oil and gas reservoirs are integral parts of running a successful business.

Together we will integrate geology with statistics, image analysis, pattern reproduction as well as classical spatial estimation techniques. You will leave this class with an overall understanding on the meaning and use cases of statistical modeling techniques. No equations - promise! Exercises will stay available to you beyond the end of the class, as well as the core software tools. Besides open source software, we will also work with a major proprietary reservoir modeling software package.

After this class, you will have acquired an understanding of the fundamentals of methods like variogram, Kriging, stochastic simulations, geologic object-based simulation, and reproduction of features from training images. Let’s understand how to digitize the subsurface and run data mining deep down in the earth.

Instructor: Dr. Andre Jung (Shell / Baker Hughes)