Machine learning opportunities for geothermal drilling operations: An overview

Abstract

Geothermal energy has been providing low carbon and reliable renewable resource for electricity generation since the 1950’s; however, the total geothermal installed capacity comprises only 0.5% of total renewables-based capacity as of 2021, growing only ~3.5 annually. High upfront costs and resource risks associated with geothermal development proved a challenge for future investments and wide-scale adoption.
As heat is harnessed from the depths of the earth, drilling wells towards a viable resource is critical to the success of the project. Drilling has only ~83% success rate in the operation phase and accounts for 35-40% of the total capital expenditure of the project. Improving drilling performance by de-risking drilling operations will be a game-changer in pushing interest towards geothermal development.
With the recent advent of artificial intelligence (AI) technology, there is a renewed interest in looking at Machine Learning for optimising different drilling applications. Machine learning is a subfield of AI that automates the modelling of complex data sets for problem specific tasks. Combining these models with a problem specific decision model leads to AI. This paper summarises the challenges of geothermal drilling operations and gives an overview of machine learning applications in drilling operations.