Basin-wise Concentration Prediction for Saline Aquifers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
There is a need for improved solutions to recover valuable elements, such as metals like lithium, from saline aquifers, as existing methods are costly and cumbersome.
Innovation Solution
A computer-implemented method of machine-learning predictive basin-wise models to predict the concentration of elements in saline aquifers, using geochemical variables and ensemble-learning models like XG Boost or Random Forest, which allows for accurate predictions without direct measurement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If direct measurement methods are used to determine element concentration in saline aquifers, then measurement precision is improved, but loss of time and productivity deteriorate due to costly and cumbersome processes
Solution Approach 1:
The patent replaces direct physical measurement methods with a machine learning-based predictive system. The model uses geochemical variables (input features) to predict element concentrations (output), substituting cumbersome direct measurement processes with computational prediction that achieves comparable precision while dramatically improving productivity and reducing costs
2Productivity
If basin-wise machine learning models are used to predict element concentration, then productivity is improved through fast predictions, but measurement precision may worsen compared to direct measurement
Solution Approach 1:
The patent performs preliminary training of basin-specific machine learning models using historical geochemical data and measured element concentrations. This preliminary action creates pre-trained models that can rapidly predict concentrations for new locations without requiring time-consuming direct measurements, thus improving productivity while maintaining precision through the quality of pre-training data
Solution Approach 2:
The patent creates predictive copies of the relationship between geochemical variables and element concentrations by training models on existing measured data. These model copies capture the underlying patterns and can predict concentrations for new locations with high accuracy, enabling fast predictions that maintain measurement precision while improving productivity
3Reliability
If multiple basin-specific models are trained with ensemble learning, then reliability is improved through better predictions, but device complexity worsens due to multiple models and sub-models
Solution Approach 1:
The patent segments the prediction problem by creating separate basin-specific models for different geological basins. Each model is trained on local data and captures basin-specific geochemical patterns. This segmentation improves reliability by accounting for regional variations while organizing complexity into manageable, independent modules that can be selected based on the target basin
Solution Approach 2:
The patent combines multiple prediction sub-models into an ensemble learning framework where predictions from individual models are aggregated. This merging improves reliability through ensemble diversity and error cancellation, while the modular structure allows efficient computation and management of the combined model system
Data Source
AI summary
It is hereby proposed a computer-implemented method of machine-learning a plurality of predictive basin-wise models. Each predictive basin-wise model is configured for predicting a concentration of an element at a given location in a saline aquifer of a respective basin. The machine-learning method comprises, for each basin and with respect to a predetermined set of one or more geochemical variables, providing a dataset, and learning the predictive basin-wise model based on the dataset. The dataset comprises, for respective saline aquifer locations of the basin, training samples. Each training sample includes a measurement of one or more geochemical variables of the predetermined set. Each training sample further includes a respective ground truth value. The ground truth value represents a concentration of the element at the respective saline aquifer location. Such a method forms an improved solution for analysis of a saline aquifer with respect to a given element of interest.