15 Feb 2023
Affordable, reliable and sustainable Geothermal Energy has significant global potential and could play an important role in the Energy Transition.
Understanding the structure of porous rocks is essential for harnessing this type of energy. For this reason, the objective of the 2022 SPE Europe Geohackathon was to design an automated method for identifying the areas occupied by detrital quartz grains, their diagenetic overgrowth, other minerals and pore space on Scanning Electron Microscope (SEM) images of sandstone rocks.
The competition ran from the 25th of October to the 25th of November 2022 and was preceded by sixteen ‘bootcamps’ covering a number of different skills related to Data Science and using Python and MATLAB coding.
An RPS team (Adam Turner, Adolfo Perez and Ali Bahrami, entered under the team name ‘Triple A’) defined an unsupervised image-detection algorithm, written in Python, that first aligned and segmented the SEM images to then colour the defined segments according to the legend proposed by the competition. The training dataset for the algorithm included more than 4,000 of these images.
Figure 1 shows an example of the SEM images included in the dataset and the results obtained from the algorithm. Yellow corresponds to the original quartz grains, red to the overgrowth, green to other minerals and black to porosity.
The judges assessed the algorithm with a testing dataset unavailable to the participants. They scored the algorithm generated images based on the number of correctly labelled pixels divided by the total number of pixels. The RPS team achieved a final score of 60.19%.
This result earnt the team third place in the competition overall, an excellent achievement given the team of reservoir engineers had very limited or no previous experience in image processing, that all work had to be done around extensive existing project commitments and that they faced stiff opposition, with teams from over 20 countries registering for the event.
The RPS team demonstrated great determination in solving projects that matter and showing that they are open to new opportunities that this complex world of Energy Transition brings.
The algorithm is publicly available on ( https://geusgitlab.geus.dk/geohackathon/triplea ).
Adolfo Perez
Adam Turner
Your contact information:
All fields are mandatory *
RPS is committed to protecting and respecting your privacy. We will only use your personal information to administer your account and to provide the products and services you have requested. We would also like to contact you about our products and services, as well as other content that may be of interest to you.