Composición y fuentes del material particulado atmosférico en la zona cerámica de Castellón. Impacto de la introducción de las Mejores Técnicas Disponibles

Resumen

Pollution due to atmospheric particulate matter has been shown to have a considerable impact on health, climate, ecosystems, buildings and visibility. Legislation on the levels of atmospheric particulate matter has undergone changes in recent years. The implementation of the Directive 1999/30/CE has brought about changes in the control of air quality by regulating PM10, and by establishing stricter limits. Moreover, PM2.5 will also be regulated (Proposal of the Directive Clean Air For Europe of October 2006). As regards emissions of particulate matter, the Directive 1996/61/CE, IPPC, has brought about significant changes to the previous legislation. Prior studies on the ceramic producing area of Castellón showed that levels of particulate matter and some trace elements were relatively high. A detailed study was therefore carried out in the period 2002-2005 to assess the levels and composition of particulate matter in the ceramic zone of Castellón, identify emission sources and to evaluate the impact on air quality of the implementation of corrective measures in some industrial premises. Thus, it was sought to establish goals of air quality and to devise a plan of action to reduce emissions and thereby improve air quality. Data on atmospheric dynamics of the study area was collected given its bearing on the dispersion and transport of pollutants. Thus, atmospheric dynamics is characterised by two situations. In winter, it is dominated by the synoptic circulation adapted to the orography, with the prevalence of NW winds. In summer, there are mesoscale circulations of daily periodicity (sea and land breeze, valley and mountain breeze), so that ESE-SEE winds prevail during the day (sea breeze), while NW winds prevail during the night (nocturnal drainage). For the selection of the monitoring sites to study the air quality of the area, two intensive campaigns were carried out in the summer 2002 and in the winter 2002-2003. These campaigns consisted in sampling and chemical analysis of PM10 at ten sites in the area. Subsequently, four sites were selected as representative of the PM10 levels and composition for use in the continuous study 2002-2005: LAlcora-PM, Vila-real, Onda and Borriana-residencia. Moreover, data of PM10 recorded at two additional sites (Almassora and Borriana-rural) were used. During the period 2002-2005, the levels of PM10 were relatively homogeneous and constant at the selected sites of the study area (33-37 µg/m3 as annual means at urban sites and 25-29 µg/m3 at the subrban site of Onda). The levels at Onda are lower than those at the other sites due to the location of this site in a sub-basin slightly isolated from the Millars basin (which dominates the transport of pollutants from the main industrial area) with the result that it is less influenced by industrial emissions. The levels at Borriana-rural and Almassora (32-38 µg/m3 of PM10) were similar to those at the urban sites, although they are suburban sites. This is probably due to their location, which receives more pollutants from the industrial zone owing to the atmospheric dynamics (with NW breezes mainly during the night). This is demonstrated by the higher leveles of PM10 at Borriana-rural and Almassora during the night than during daytime. The levels of PM10 recorded in the study area were compared with those recorded in other zones of Spain without a high industrial influence, observing an excess of 3 to 6 µg/m3 of PM10. At Onda, a high proportion (60%) of the exceedances of the daily limit value of PM10 were due to natural episodes, whereas at LAlcora-PM, Vila-real and Borriana-residencia, a high number of the exceedances of the daily limit value (60-70%) were due to local pollution episodes and/or adverse meteorological conditions. Borriana-rural and Almassora were in an intermediate situation. The mean levels of PM2.5 during the period 2003-2005 were 19 µg/m3 at Onda and 24 µg/m3 at LAlcora-PM and Vila-real. Nevertheless, there were fewer PM2.5 than PM10 data. The ratio PM2.5/PM10 was 0.7 at all the sites. This fine granulometry was due to the presence of high concentrations of mineral matter in PM2.5. The comparison of the PM2.5 levels recorded at LAlcora-PM and Vila-real with those recorded at urban sites in Spain without a high industrial influence indicates that there was an excess of about 3-4 µg/m3 of PM2.5. The main component of both PM10 and PM2.5 was mineral matter. Mineral matter accounted for 28-51% of the bulk PM10 and for 29-42% of the bulk PM2.5, with the highest contribution at LAlcora. The levels of mineral matter in PM10 in the study area were between 3 and 7 µg/m3 higher than those recorded at similar Spanish sites without a high industrial influence, and in PM2.5 the excess was between 3 and 5 µg/m3. Given the excess of bulk mass of PM10 and PM2.5 and of mineral matter in both fractions (comparing levels of the study area with those recorded at other urban and suburban sites of Spain), and given the knowledge of the degree of implementation of corrective measures in the industrial premises of the area, it is possible to establish goals of reduction between 3 and 5 µg/m3 in the annual mean of PM10 and of 2-3 µg/m3 in the annual mean of PM2.5 at urban sites. This reduction is mainly based on the reduction of the mineral load. As for the seasonal variation of the PM10 components, the mineral matter showed no clear pattern because of the random distribution of the pollution episodes and the occurrence of natural contributions of mineral material throughout the year. The anthropogenic origin of this component was confirmed by lower levels during the weekends given the reduction in industrial activity and traffic. The levels of OM+EC, mainly emitted by traffic, showed a maximum in winter and a minimum in summer owing to the lower dispersive conditions in winter, which favour the accumulation of pollutants in the surroundings of the emission sources. The sulphate levels were higher in summer than in winter, as expected, given the higher rate of conversion of SO2 in summer (amply documented) and given the prevalent breeze regime in summer in the study area. This breeze regime brings about the transport of emissions of the industrial estate located on the coast to the interior. The remaining major components showed no clear pattern. Nevertheless, levels of V, Ni and Na were higher in summer than in winter because of the diurnal sea breeze in summer that transports V and Ni (emitted by the coastal industrial estate) and the marine Na to the interior. The evaluation of the levels of trace elements in PM10 allowed the identification of As, Pb, Zn, Zr and Tl as tracers of ceramic emissions (including frit manufacture) in the area of study from 2002 to 2005. These elements were identified by comparing the levels recorded in the study area with those registered in other zones without a high industrial influence. Moreover, the elements with annual mean levels lower than 3 ng/m3 and the elements with high annual mean levels due to sporadic episodes (and not due to high levels throughout the year) were not considered for the identification of the tracer elements. The Borriana site underwent a higher influence of industrial sources as indicated by the higher levels of trace elements at this site than those recorded at the other sites in the area. Interpretation of the daily variability of the levels of trace elements together with the review of the literature on chemical profiles of emission sources allowed the identification of the main sources of the most relevant trace elements in the area: - Zr, Zn, Pb, As, Tl, Ba and Cs: manufacture and use of glaze components (including frit manufacture). - Se, Ce, Cd y Cr: manufacture and use of some ceramic pigments. - Sb: generally associated with traffic although it may be emitted by the manufacture and use of some ceramic pigments. - Rb, Li, La, Sc y Pr: emissions of mineral matter (mainly generated in the manufacture of ceramic tiles). - Co: emissions of mineral matter and manufacture and use of some ceramic pigments. - Cu: traffic. - Mn: generally associated with traffic although in the study area it is associated with mineral matter. - V: power plant and petrochemical plant. - Ni: power plant and petrochemical plant although it may be emitted by the manufacture and use of some ceramic pigments. During the study period (2002-2005), a decrease in the levels of Li, Zn, As, Rb, Cs, Ba, La y Pb was recorded at LAlcora, Vila-real and Onda (especially from mid 2004). This decrease can be attributed to the progressive implementation of corrective measures in some plants of the ceramic industry (mainly in the fusion stage for the manufacture of frits) in order to meet the requirements of the IPPC. A principal component analysis followed by a multilinear regression, which allowed the identification and quantification of the sources of PM10, was carried out. Five common factors were identified as sources of PM10 in the study area: mineral, regional background, industrial 1, traffic and sea spray. The mineral factor is characterised by the contribution of Al2O3, Ca, K, Mg, Fe, Ti, Li, Rb, Sr, Y, La, Pr, Nd and Mn and by the high correlation of these elements and components with the levels of PM10. This factor includes sources such as: ceramic industry (mainly the stages emitting mineral matter), transport of powdery materials, clay extraction, African intrusions and soil resuspension. The regional background factor is characterised by the contribution of SO42-, NH4+, V and Ni, and a high correlation of these elements and components with the levels of PM10. This factor includes, in addition to the regional background, the influence of the industrial estate on the coast. The industrial 1 factor is made up of K, Zn, As, Rb, Cs, Tl and Pb and is attributed to the manufacture and use of glaze components, mainly the manufacture of ceramic frits. The traffic factor is characterised by the contribution of OC+EC and NO3-. Finally, the sea spray is characterised by the contribution of Na. Moreover, a second mineral factor was identified at Borriana, associated with the soil emissions, and another industrial factor (industrial 2) was identified at LAlcora, Vila-real and Borriana, characterised by the contribution of Zr, attributed to some sources of this element, including the manufacture of frits. The contribution of the mineral factor was similar at the four sites, varying between 9 and 11 µg/m3, and was relatively constant over the study period given that the significant changes in the channelled emissions of the ceramic industry (one of the main sources of mineral matter) occurred prior to this study (2002). The contribution of the industrial 1 factor was estimated at 4-5 µg/m3 at Vila-real, L'Alcora and Borriana, and at 2 µg/m3 at Onda. This contribution was reduced significantly over the study period, attaining 0.3 µg/m3 at Onda and 1.2-1.4 µg/m3 at L'Alcora and Vila-real in 2005. Nevertheless, the contribution of this source at Borriana in 2005 was still relatively important (4 µg/m3). Moreover, the source contribution to the levels of each element and compound analysed in PM10 was calculated. More than 68%, 58% and 61% of As, Pb and Zn, respectively, was emitted by the source industrial 1, whereas the contribution to Tl levels was not so elevated, which indicated that this element was apportioned to a lower degree by other sources. The contribution to levels of Zr showed that this element was emitted by different sources, the majority of which were related to the ceramic industry. The emissions of PM10 generated by the ceramic plants were calculated from 2000 to 2006, including tile and frit manufacture. The emission of PM10 from the storage and handling of raw materials for the tile body (fugitive emissions) did not vary significantly over the study period given that the implementation of corrective measures of high efficiency was limited. The channelled emissions of particulate matter from the manufacture of tiles dropped from 2001 to 2002, remaining relatively constant thereafter. The emissions generated in the manufacture of frits underwent a reduction of 90% in 2006 with respect to the emissions in 2000; this decrease was more marked throughout 2004. In July 2006, the amount of PM10 emitted by the ceramic plants was 5250 tonnes/year, of which 34% were fugitive emissions, 3% were generated by the manufacture of frits, and the remaining 63% were channelled emissions of the manufacture of tiles, with a high contribution (60%) of the emissions from the spray drying stage. If the Best Available Techniques were implemented in all the ceramic plants, the potential reduction would be 3200 tonnes/year of PM10, mainly based on the reduction of the fugitive emissions (52%) and the emissions from the spray drying stage (36%). The correlation between the evolution of the generated emissions and the levels of some pollutants in ambient air was studied. It was observed that the evolution of levels of As, Pb, Zn and Cs in ambient air was parallel to the evolution of the emission of PM10 from the manufacture of frits with a marked decrease over the period 2002 to 2006. This confirmed the view that the main source of these trace elements in ambient air was the manufacture of frits. On the other hand, the levels of mineral matter in ambient air as well as the emissions of PM10 of mineral composition (emitted during storage and handling of raw materials of the tile body, milling, spray drying, pressing and drying) remained relatively constant over the study period. Based on these results, goal levels for some components were identified, and these should be interpreted as the desirable mean annual levels in the study area. Thus, goal levels of PM10 and PM2.5 are 30 µg/m3 and 20 µg/m3, respectively. As regards trace elements, goal levels are: 2 ng/m3 of As, 1 ng/m3 of Cd, 5 ng/m3 of Ni, 80 ng/m3 of Pb, 100 ng/m3 of Zn, 15 ng/m3 of Zr, and 1 ng/m3 of Tl, as annual means. In the light of our findings, a number of strategies to reduce levels of PM10 and some trace elements are suggested. Thus, some measures to reduce mineral matter emissions from tile manufacture (for the reduction of PM10 levels) and to cut down emissions from frit, glaze and pigment manufacture (for the reduction of trace elements levels) are recommended. Some of these measures are as follows: a) storage and handling of raw materials for the tile body in closed premises; b) paving and washing of roads in the proximity of the factories; c) maintenance of lorries to avoid load loss and washing of truck wheels at the exit of industrial premises; d) implementation of bag filters in spray dryers and fusion kilns for frit manufacture; and f) reduction in the use of raw materials containing heavy metals as impurities or as main components.