In this study, the Multilinear Engine (ME-2) receptor model was applied to speciated particulate matter concentration data collected with two different measuring instruments upwind and downwind of a steelworks complex in Port Talbot, South Wales, United Kingdom. Hourly and daily PM samples were collected with Streaker and Partisol samplers, respectively, during a one month sampling campaign between April 18 and May 16, 2012. Daily samples (PM10, PM2.5, PM2.5-10) were analysed for trace metals and water-soluble ions using standard procedures. Hourly samples (PM2.5 and PM2.5-10) were assayed for 22 elements by Particle Induced X-ray Emission (PIXE). PM10 data analysis using ME-2 resolved 6 factors from both datasets identifying different steel processing units including emissions from the blast furnaces (BF), the basic oxygen furnace steelmaking plant (BOS), the coke-making plant, and the sinter plant. Steelworks emissions were the main contributors to PM10 accounting for 45% of the mass when including also secondary aerosol. The blast furnaces were the largest emitter of primary PM10 in the study area, explaining about one-fifth of the mass. Other source contributions to PM10 were from marine aerosol (28%), traffic (16%), and background aerosol (11%). ME-2 analysis was also performed on daily PM2.5 and PM2.5-10 data resolving 7 and 6 factors, respectively. The largest contributions to PM2.5-10 were from marine aerosol (30%) and blast furnace emissions (28%). Secondary components explained one-half of PM2.5 mass. The influence of steelworks sources on ambient particulate matter at Port Talbot was distinguishable for several separate processing sections within the steelworks in all PM fractions.
Keywords: Hourly resolution; ME-2; Receptor modelling; Source apportionment; Steelworks.
Copyright © 2014 Elsevier B.V. All rights reserved.