Understanding the intricate processes governing Lithium occurence within geothermal fluids is akin to decoding the Earth’s geological narrative. Lithium, a highly sought-after element, exhibits diverse behaviours influenced by geological and hydrochemical conditions within reservoirs. By discerning correlations and interdependencies among major cations, anions, and fluid characteristics, project partner INLECOM INNOVATION aims to unravel the complex web of factors controlling Lithium concentrations. These insights not only optimise geothermal resource exploration but also contribute to a deeper understanding of Lithium’s broader geological context, crucial given its pivotal role in energy storage technologies.

INLECOM INNOVATION’s task represents a pioneering effort to uncover the intricate relationships between elemental features and Lithium concentrations through the application of advanced Machine Learning methodologies. At the heart of this endeavour lies the development of the CRM AI Tool—a sophisticated platform designed to provide accurate estimations of Lithium concentrations while ensuring interpretability in geothermal data analysis.

INLECOM INNOVATION’s journey commenced with the refinement of the REFLECT Dataset into the upgraded CRM-AI Database, laying a robust foundation for subsequent analyses enabling the identification of relationships between key elements such as Magnesium, Potassium, Calcium, Sodium, Chlorine, temperature, electrical conductivity and Lithium extracted from well samples. Through meticulous data pre-processing and the application of cutting-edge Machine Learning models such as Tree-Based regression, INLECOM INNOVATION enhanced data reliability and accuracy in Lithium concentration estimations.


How does the CRM AI Tool work?

The CRM AI Tool represents a fusion of geological insights and Machine Learning expertise, offering a multifaceted approach to Lithium concentration estimation in geothermal waters. Here’s a glimpse into its workings:

  1. Data Input:Users can input elemental concentrations from geothermal fluid samples into the CRM AI Tool’s interface.
  2. Machine Learning Algorithms:Leveraging sophisticated Machine Learning algorithms, such as Decision Trees, the tool analyses the input data to estimate Lithium concentrations accurately.
  3. Real-time Exploration:With backward navigation through Decision Tree nodes, users can explore the AI model decision process
  4. Explainable AI (XAI):The tool incorporates Explainable AI techniques, including Shapley values, to provide transparency into the decision-making process of Lithium concentration estimation. Users can understand the impact of individual features on Lithium estimations, enhancing interpretability.
  5. Visualisations:The tool generates visualisations such as scatterplot maps, showcasing geographical variations in Lithium concentrations. These visual aids offer users a deeper understanding of spatial patterns and correlations.
  6. Sensitivity Analysis:Additionally, the CRM AI Tool offers Sensitivity Analysis, empowering users with sliders to adjust the ranges of every input element, electrical conductivity, and outflow temperature. This feature allows users to explore how variations in these parameters affect Lithium concentration estimations, providing valuable insights into the sensitivity of the model.

In conclusion, INLECOM INNOVATION’s CRM AI Tool stands as a testament to the power of collaboration between geoscientists and data scientists, paving the way for enhanced exploration and utilisation of geothermal resources.


For more information, check out the infographic below!