Analysis of the primary and secondary contributions to PM10 concentrations of the two largest coal-fired power plants of Italy

Energy production by coal-fired power plants have impact on climate and on local air quality. Although coal consumption decreased by over 40% between 1990 and 2014 in Europe, this trend has not been uniform for all member States. Implementing EU directives, the Italian law (D. Lgs. 46/2014), introduced stricter emissions standards, even if the majority of Italian coal-fired power plants are already well-equipped with controls for NOx, SO2 and PM, and have been able to 'opt in' to the new legislation without further modification. This work present a source apportionment (SA) study on PM10 samples collected nearby the two largest Italian coal-fired power plants: Federico II (FEDII) and Torrevaldaliga Nord (TVN). The SA was performed to evaluate the impact of the emissions in the surrounded areas using the multi-model approach based on PMF and CMB receptor models integrated with CALPUFF dispersion model, specifically developed for coal-fired power plants in Contini et al. (2016). The study allowed to estimate the contributions of the two power plants to primary PM10 and to secondary (sulphate) comparing the influence of the two plants in the surroundings areas. Measurement sites and methods The FEDII station is located in south-eastern Italy (Apulia Region) where four measurement sites were selected at different distance from the sources and four measurement campaigns were done (simultaneously at the four sites) between summer 2015 and autumn 2016. Daily PM10 samples (418 in total) were collected using low-volume (2.3 m3/h) dual channel samplers (Hydra and Swam, Fai Instruments) and an Explorer Plus sampler (Zambelli). For each day, samples were collected on quartz (Whatman Q-grade) and Teflon 47 mm filters (Whatman PTFE). The TVN station is located in central Italy (Latium region) near the Tyrrhenian coast. Three sites were selected and six measurement campaigns were performed between 2010 and 2014 collecting 347 daily samples using the same instrumental set-up and the same substrates as in the FEDII area. Measurements were carried out in different seasons in order to have results representing the annual average conditions. Quartz fibre filters were used for the determination of OC and EC using the NIOSH5040 protocol (TOT) with a Sunset Laboratory instrument. Teflon filters were used for the determination of the major water-soluble ions (via HPIC) and of the major metals (via ED-XRF). In total 19 chemical species were characterised and used in SA. The SA was done using EPA-PMF5.0 and EPACMB8.2 models. Some of the input profiles for CMB were obtained experimentally including that of the power plants. Three independent estimates of primary contribution were obtained: one from CMB and two from PMF. Diagnostic Si/Al ratio and CALPUFF dispersion model was used to separate crustal and power plant contribution in the PMF model because the two profiles are collinear being both characterised mainly by Si and Al (Contini et al. (2016). Results SA studies performed for TVN and FEDII power stations identified similar PM10 sources even if the study areas were different: crustal, resuspended dust, coal-fired power plant, road traffic, secondary sulphate, secondary nitrate, marine, biomass burning and harbour/industrial. The relative contributions found in the two areas are shown in Fig. 1, as average of the results found for the different measurement sites in each area. Globally (sum of primary and secondary sulphate), FEDII contribution ranges between 3% and 5% of PM10 at the different sites and TVN contribution ranges between 3% and 4%. The differences in the contributions of the two plants reflect the differences in the local meteorological conditions as well as the structural differences of the plants (i.e. the height of the stacks). FIGURE F