Air Pollutant Emission Inventory of Waste-to-Energy Plants in China and Prediction by the Artificial Neural Network Approach

Abstract

The waste-to-energy (WTE) plant has been deployed in 205 cities in China. However, it always faces public resistance to be built because of the great concerns on flue gas pollutants (FGPs). There are limited studies on the socioeconomic heterogeneity analysis and prediction models of WTE capacity/ FGP emission inventories (EIs) based on big data. In this study, the incinerator level emission factors (EFs) in 2020 of PM, SO2, NO x , CO, HCl, dioxins, Hg, Cd + Tl, and Sb + As+ Pb + Cr + Co + Cu + Mn + Ni were calculated based on 322,926 monitoring values of all the 481 WTE plants (1140 processing lines) operating in China, with uncertainties in the range of ±34.70%. The EFs were significantly 45–96% lower than the national standard (GB18485-2014) and had negative relationships with local socioeconomic elements, while WTE capacity and FGP EIs had significantly positive correlations. Gross domestic product, area of built district, and municipal solid waste generation were the main driving forces of WTE capacity. The WTE capacity increased by 150% from 2015 to 2020, while the total emission of PM, SO2, CO, dioxins, Hg, and Sb + As + Pb + Cr + Co + Cu + Mn + Ni decreased by 42.46–88.24%. The artificial neural network models were established to predict WTE capacity and FGP EIs in the city level, with the mean square errors ranging from 0.003 to 0.19 within the model validation limits. This study provides data and model support for the formulation of appropriate WTE plans and a pollutant emission control scheme in different economic regions.

Growth of the WTE industry in China and environmental performance

“…Themelis 4 HCl, PM, dioxin, and heavy metals (Hg, Cd+TI, Sb+As+Pb+Cr+Co+Cu+Mn+Ni). Furthermore, previous studies examined the driving factors of MSW generation in China 12 , the relationships between incineration amounts and factors such as economic size and population density at the provincial level 13 and the correlation between socioeconomic factors and city-level emission factors 11 . However, factors governing the emission trends of multiple air pollutants from the WTE industry after the introduction of the 2014 emission standard remain unclear.…”

“…In contrast to NOx, SO2, HCl, PM, and CO, which are continuously monitored, the heavy metals (HM) and dioxins emissions are not required to be monitored continuously; and few plants report them to the provincial "supervisory emission monitoring system (SEMS)" 11 . In 2020, more than 400,000 data on nine pollutants were collected based on CEMS (continuous emissions monitoring system) and SEMS nationwide.…”

“…To meet these emission limitations, incineration plants have installed more advanced flue gas purification equipment. For example, the denitrification devices combine selective non-catalytic reduction unit (SNCR) + selective catalytic reduction (SCR) devices 20 . SCR has higher removal efficiency and investment cost than the widely used SNCR system.…”

“…where EFLF,p,x,CO2 represents the estimated CO2 emission factors of different MSW content x from landfill, kg/dry ton MSW, to be 1100, 829, 2708 of the CO2 EFs for paper, wood, and food residue, respectively; GWPj represents the global warming potential of NOx (8), CO (3), CH4 (21), and nonmethane volatile organic compound (NMVOC) (11).…”

Fault monitoring method of domestic waste incineration slag sorting device based on back propagation neural network

“…The experiment analyzed data from more than 400 factories in China and found that environmental footprint had a significant negative correlation with social economy. The constructed model could be used to effectively monitor urban exhaust gas emissions, with monitoring errors ranging from 0.003 to 0.09 [ 9 ].…”