Optimizing the Geothermal Drilling Process Using Artificial Intelligence Methods

In deep geothermal projects, the drilling process can account for a significant proportion of the overall project cost of up to 70%. To provide a solution to minimize common uncertainties and the associated costs, a data driven AI-based drilling advisory systems is being developed within the OptiDrill project. The system applies machine learning based models to optimize the drilling process for geothermal wells and at the same time increases the economic attractiveness and accessibility of geothermal energy. The advisory system consists of four modules, each addressing a different aspect of the drilling process. The modules focus on the areas of drilling performance prediction and optimization, drilled lithology prediction, drilling problem detection and well completion and stimulation optimization. This presentation provides an overview of the OptiDrill project with a focus on the developments based on AI methods. It introduces and presents software modules focused on drilling process performance prediction and optimization as well as drilled lithology prediction. Both modules utilize artificial neural networks trained on historical drilling data from oil, gas, and geothermal projects to predict target values, such as ROP and drilled formation lithology. These predictions offer valuable insights to drillers, contributing to a more effective and seamless drilling process.