Predicting the distribution coefficient of cesium in solid phase groups using machine learning.

Area of Science:

• Environmental Science and Engineering
• Nuclear Waste Management
• Geochemistry

Background:

• Radioactive contaminant migration, particularly ¹³⁷Cesium, presents significant challenges for nuclear waste storage.
• The distribution coefficient (K_d) is crucial for evaluating contaminant mobility but is sensitive to environmental factors.
• Accurate K_d prediction is essential for robust safety and risk assessments in radioactive waste disposal.

Purpose of the Study:

• To develop and evaluate machine learning models for predicting ¹³⁷Cs distribution coefficients (K_d) in various environmental media.
• To identify key environmental and geochemical variables influencing ¹³⁷Cs sorption and mobility.
• To provide tools for improved environmental risk assessment and safety analysis in nuclear waste management.

Main Methods:

• Utilized the Japan Atomic Energy Agency Sorption Database (JAEA-SDB) containing 14 input variables.
• Developed and trained three machine learning models: Random Forest (RF), Artificial Neural Network (ANN), and Convolutional Neural Network (CNN).
• Preprocessed data using normalization and log-transformation; evaluated model performance using R² and RMSE.

Main Results:

• RF, ANN, and CNN models achieved high prediction accuracy with R² values exceeding 0.97, 0.86, and 0.88, respectively.
• Variable importance analysis identified environmental media, initial radionuclide concentration, solid-phase properties, and solution conditions as significant predictors.
• Models demonstrated robust performance across diverse environmental conditions, indicating their applicability for K_d prediction.

Conclusions:

• Machine learning models, particularly RF, offer accurate prediction of ¹³⁷Cs K_d values, crucial for assessing contaminant mobility.
• The developed models can enhance safety analyses and long-term risk assessments for nuclear waste disposal.
• Accurate K_d prediction aids in preventing potential hazards and contamination from radioactive materials in the environment.