Document Type : Research Paper
Authors
1
Department of Chemical Technologies, Iranian Research Organization for Science and Technology (IROST), P.O. Box 33535111, Tehran, Iran
2
Department of Petroleum Engineering, Amirkabir University of Technology (Tehran Polytechnic), Tehran, Iran
3
School of Chemical Engineering, College of Engineering, University of Tehran, Iran
4
Department of Polymer Engineering & Color Technology, Amirkabir University of Technology, Tehran, Iran
Abstract
Understanding and projecting lithium pricing fluctuations will become increasingly important as the world shifts toward cleaner energy sources and electric mobility. Accurate prediction of lithium prices using accessible and reliable methods is essential for sustainable lithium extraction practices. This study employs machine learning (ML) techniques to forecast the costs of future lithium production, with an emphasis on the relationship between the supply of lithium, the use of EVs, and the renewable energy. To find the most precise approach, five ML models are investigated: K-Nearest Neighbors (KNN), Support Vector Classifier (SVC), Linear Support Vector Regression (SVR), Linear Regression, and Multi-Layer Perceptron (MLP) Regression. KNN emerges as the top performer among them, exhibiting robust prediction capabilities, and it is used as the applied ML model in this work. Eight likely scenarios that could affect lithium prices in the upcoming years are examined. These scenarios include advances in lithium extraction technology, alterations to EV usage patterns, and rising demands for the production of renewable energy. Results reveal that the ML model forecasts a notable upward trend in lithium prices. After 2022 to 2029, as the production of EVs and renewable energies increases, lithium prices are projected to rise substantially from USD 20,973/ton to USD 37,745/ton. From 2029 to 2037, with growing demand for batteries, the price is predicted to rise from USD 37,745/ton to USD 40,747/ton. Despite potential short-term fluctuations due to external factors, the model indicates that the long-term trajectory remains upward in lithium prices by 2040.
Keywords
Main Subjects
)