A comparison study of bottom‐up and top‐down methods for analyzing the physical composition of municipal solid waste

Abstract

Municipal solid waste (MSW) management is a crucial issue in socioeconomic metabolism and requires multicategory and high‐resolution data, and data on the physical composition of municipal solid waste (PCMSW) are fundamental in MSW research. Extensive financial resources have been invested in the research on field investigations of PCMSW; however, it is time‐consuming and sometimes not truly representative of the studied case. In this work, two bottom‐up and two top‐down approaches were applied for analyzing the PCMSW, namely, field investigation (FI), BP neural network (BPNN), material flow analysis (MFA), and inversion algorithm based on electricity generation of waste incinerator (IAEI). Wuhan City, China, was chosen as the studied case for analyzing and comparing the PCMSW results. The PCMSW values obtained by applying those four methods showed acceptable differences, and the standard deviations of organic fraction, ash and stone, paper, plastic and rubber, textile, wood, metal, glass, and others were 3.94%, 2.77%, 6.57%, 2.22%, 2.49%, 1.36%, 0.53%, 1.19%, and 0.28%, respectively. Use of the MFA, BPNN, and IAEI methods could reduce time and labor spent on manual sampling, sorting, and weighing of MSW, compared to the FI method. BPNN algorithm advances in providing PCMSW data in history and by trajectory throughout the year, whereas IAEI contributes PCMSW data with much higher temporal resolution. Data quality and applicability of different methods were discussed, considering the availability of time, labor, and reference data.

Tracking the material cycle of Italian bricks with the aid of building information modeling

“…Nonetheless, many are carefully collected and summarized, or are estimated by knowledgeable specialists. In our opinion, our results have accuracies of about ±5% from the Raw Materials stage to the Material Stock stage, and ±25% from Waste Management onward because of the well‐known inaccuracy of waste data (Powell & Chertow, 2019; Powell et al., 2016; Zhou et al., 2021). Uncertainties of this magnitude are concerning, especially regarding waste flows, but appear not to significantly compromise the findings of the Italian clay bricks cycle that we present (cf.…”

Data innovation in industrial ecology

“…Validation of such ML-based modeling emerges as an important issue. Zhou et al (2021) thus compared the ability of a neural network to estimate municipal solid waste composition, as an alternative to more time-intensive bottom-up field investigation, or more generic top-down MFA and inversion algorithm techniques.…”

“…Many publications focus on cities and their urban metabolism, leveraging the increased data richness of these dense environments (Arbabi et al, 2021;Ballatore et al, 2021;Clark et al, 2021;Hastie et al, 2020;Xie et al, 2021). Characterizing material stocks, especially infrastructures, is also increasingly facilitated by novel data acquisition and analysis techniques (Arbabi et al, 2021;Ebrahimi et al, 2021;Sprecher et al, 2021;Vilaysouk et al, 2021;Zhou et al, 2021). In particular, waste management and recirculation strategies are propelling a very diverse set of novel data-driven research (Ballatore et al, 2021;Davis & Aid, 2021;Kerdlap et al, 2021;Zhou et al, 2021).…”

“…Characterizing material stocks, especially infrastructures, is also increasingly facilitated by novel data acquisition and analysis techniques (Arbabi et al, 2021;Ebrahimi et al, 2021;Sprecher et al, 2021;Vilaysouk et al, 2021;Zhou et al, 2021). In particular, waste management and recirculation strategies are propelling a very diverse set of novel data-driven research (Ballatore et al, 2021;Davis & Aid, 2021;Kerdlap et al, 2021;Zhou et al, 2021).…”