A preliminary study on real-time asbestos recognition in tunnel spoil of green stones

This study focuses on the analysis of serpentinite rocks to detect asbestos minerals, with a particular emphasis at discriminating chrysotile from other minerals and matrices (antigorite, lizardite) in serpentinites. The primary objective of this exploratory study is to lay the groundwork for the development of an innovative system for real-time identification of asbestos on the surface of materials and rocks. In more detail, the study emphasizes the advantages of utilization of rapid and reliable portable instrumentation, aided by machine learning (ML) techniques. Serpentinite rock debris from tunnel excavation activities (i.e. tunnel spoil) were collected at the Cravasco railway tunnelling site (Genoa - Italy) of the TEN-T Rhine-Alps Core Corridor. On-site analysis was conducted using the ASD FieldSpec 4 Standard-Res spectrophotoradiometer (350-2500 nm). A two-class classifier was developed based on acquired reflectance spectra, to discriminate between the “Presence of asbestiform fibers” and the “Absence of asbestiform fibers”. The classification model was calibrated through stereomicroscopic inspection and electron microscopy carried on selected sample fractions.