Publication

Non-hazardous industrial waste (NHIW), primarily consisting of manufacturing process residues, has long been overlooked in waste reporting, regulation, and reuse. Limited information about NHIW generation with spatial and sectoral details has impeded the systematization of reuse efforts to move towards a resource-efficient economy. In this article, we develop a methodological framework that makes the best use of fragmented and limited observational data to infer the confidence intervals of NHIW generation by sector, location, and year across the United States. The framework decomposes the quantity of NHIW into two factors: the activity level (economic output) and the waste intensity factor (waste tonnage generated per unit of output). It statistically infers the probability distribution of the waste intensity factor and extrapolates waste tonnages to the entire country. In our demonstrative application of the method, we provide an updated estimate for spent foundry sand and find that its total amount in the United States decreased from 2.2-7.1 million tons in 2004 to 1.4-4.7 million tons in 2014. The spatial distribution, however, was highly uneven, with over 90% of the waste generated in 10% of the counties, indicating great variations in reuse potentials and benefits among regions. Our methodological framework makes a significant departure from existing estimations that usually rely on averaging limited observations or expert judgments biased by subjectivity. Detailing spatial and sectoral distributions and temporal trends in NHIW generation and reuse benefits, our study could inform more systematic strategies on waste and materials management to build a circular economy. (C) 2020 Elsevier Ltd. All rights reserved.