| In recent years,China’s air pollution prevention and control situation are serious,with fine particulate matter(PM2.5)and ozone(O3)as the primary pollutants of air quality exceedances occurring frequently,endangering public health.Emissions of pollutants,mainly volatile organic compounds(VOCs)and nitrogen oxides(NOx),are important triggers for PM2.5 and O3 generation.Mapping the total emissions of pollutants and formulating targeted emission reduction policies are the first priority to solve air quality problems.However,in studies about emission inventories,there is a large gap between the total amount of pollutant emissions counted by various parties,and there is a lack of systematic analysis of control technologies.It is urgent to sort out the process of source inventory preparation,analyze the key factors that affect the accuracy of the inventory,and lay the foundation for pollutant emissions calculated accurately and emission reduction strategies formulated reasonably.In this paper,with 2018 as the base year,four typical cities with industry representativeness within Zhejiang Province were selected.Data were collected from multiple sources,and the emission factor method and material balance algorithm were used to construct a multi-pollutant emission inventory of industrial sources containing 109 types of emission processes.In 2018,the total emissions of VOCs in the four cities were 33.2,57.3,15.6,and 20.2 thousand tons,and the amount of NOx emissions were 16.9,2.7,14.0,and 0.029 thousand tons,and PM2.5 emissions were 26.1,21.7,24.1,and 0.66 thousand tons,respectively.The emission results provide a database and basis for further development of emission reduction targets.This article innovatively proposes to measure industry pollution by VOCs emissions per unit of industrial output value.Through the calculation,several industries with relatively serious pollution per unit of output value are derived,which providing a reference to promote industry remediation for the government.The key problems in the process of source inventory preparation were analyzed and studied,and the key factors that affect the accuracy of the inventory were:the accuracy of emission factors,irregularities in the data processing process,lack of verification of end-of-pipe control efficiency,etc.In response to these problems,ideas are proposed to solve them:implementing localized actual measurement of emission factors,standardizing the content of emission source surveys,standardizing the data processing process with the help of computational methods such as database combined with machine learning,building a management platform for end-of-pipe control facilities.For the construction process of source inventory,the uncertainty of the constructed emission inventories of the four cities is analyzed by using the Monte Carlo method and applying Cristal Ball software.The relative standard deviation of the total amount of VOCs emissions of the four cities was ± 21.4%,and the uncertainty of the 95%confidence interval was[-36.3%,+47.0%].The uncertainty of the solvent use sources is mainly due to the significantly higher emission factor values than the other sources,while the uncertainty of process sources mainly originated from the discounting process in data processing.For the data processing process with huge manual demand in source inventory construction,the application of computer science methods on text intelligent classification was explored.A product name correspondence database was constructed,which can achieve 62.7%intelligent classification of product names with an accuracy of 92.8%;for product names that cannot be identified by the database,an innovative method of calculating text similarity using word vectors was used to complete intelligent classification with an accuracy of 51.7%.With the combination of the two methods,the accuracy of intelligent classification can reach 78.1%.Finally,this paper selects the two industries with the highest VOCs and NOx emissions among the four cities:the printing and dyeing industry and the thermoelectric industry,and a pollution control technology evaluation system that includes four major evaluation indicators:technical indexes,environmental indexes,economic indexes,and management indexes was established to make a comprehensive assessment of the pollution control technologies currently used in the two industries.Under the weighting conditions set in this paper,for the treatment of low concentration VOCs waste gas in the printing and dyeing industry,adsorption+regenerative catalytic oxidation(RCO)technology has the highest comprehensive score value,with good environmental benefits and industry applicability;in the cogeneration industry,selective catalytic reduction(SCR)technology has the highest comprehensive score value,with superior denitrification performance,but relatively high costs.If increasing the weighting value of economic indicators,selective catalytic reduction/selective non-catalytic reduction combined technology(SCR/SNCR combined technology)is more recommended. |
PDF Full Text Request