Classification and predictive leaching risk assessment of construction and demolition waste using multivariate statistical and machine learning analyses

Abstract

Managing construction and demolition waste (CDW) poses serious concerns regarding landfilling and recycling because of the potential release of hazardous elements after leaching. Ceramic materials such as bricks, tiles, and porcelain account for more than 70% of CDW. Fourteen samples of different CDW products from Ferrara (Northeast Italy) were subjected to geochemical analyses, including leaching tests, in accordance with UNI EN 12457–2. The interaction between ceramics and concrete was examined, highlighting the influence of mixed environments on the leaching behavior. Results were compared with an extensive database of more than 150 samples collected from the literature on different CDW types worldwide. Multivariate statistical analysis and machine learning were used to classify the CDW compositions based on the bulk chemical data. Various metrics—contaminant factors (C_f and C_d) and hazardous quotients (HQ and HQ_m)—were introduced to quantify the key environmental hazards of leachates. The results of this study underscore the potential of the proposed approaches in automating CDW classification and predicting C_f and HQ using only the starting bulk chemical composition. The findings enhance CDW management practices and support sustainability efforts in the construction industry.