Quantitative Modelling of Trace Elements in Hard Coal.

Area of Science:

• Environmental Science
• Analytical Chemistry
• Geochemistry

Background:

• Coal remains a dominant global fossil fuel, necessitating clean technologies and analytical tools for sustainable development.
• Environmental regulations highlight the need to monitor hazardous trace elements in coal.
• Effective coal quality assessment is crucial for managing environmental impacts.

Purpose of the Study:

• To apply quantitative Partial Least Squares (PLS) methods for modeling trace element concentrations in hard coal.
• To develop predictive models based on coal's physical and chemical parameters.
• To identify and manage potentially hazardous trace elements emitted from coal processing.

Main Methods:

• Utilized quantitative Partial Least Squares (PLS) regression for trace element concentration modeling.
• Applied robust PLS methods to handle outliers in a dataset of 132 coal samples from Poland.
• Analyzed 24 physical and chemical parameters to predict concentrations of 13 elements (As, Ba, Cd, Co, Cr, Cu, Mn, Ni, Pb, Rb, Sr, V, Zn).

Main Results:

• Developed robust PLS models with good fit and prediction abilities for trace element concentrations.
• Achieved root mean square error below 10% for most constructed models.
• Prediction errors (RMSECV) exceeded 10% for only three of the developed models, indicating high accuracy.

Conclusions:

• Demonstrated a unique application of chemometric methods for trace element analysis in coal.
• The developed PLS models provide valuable tools for assessing coal quality and environmental risk.
• Contributed to advancing analytical methodologies for the coal industry and environmental monitoring.