The direct influence of ship traffic on atmospheric PM2.5, PM10 and PAH in Venice

https://doi.org/10.1016/j.jenvman.2011.01.016Get rights and content

Abstract

The direct influence of ship traffic on atmospheric levels of coarse and fine particulate matter (PM2.5, PM10) and fifteen polycyclic aromatic hydrocarbons (PAHs) has been estimated in the urban area of Venice. Data analysis has been performed on results collected at three sites over the summer, when ship traffic is at a maximum. Results indicate that monitoring of the PM daily concentrations is not sufficiently detailed for the evaluation of this contribution, even though it could be useful for specific markers such as PAHs. Therefore a new methodology, based on high temporal resolution measurements coupled with wind direction information and the database of ship passages of the Harbour Authority of Venice has been developed. The sampling sites were monitored with optical detectors (DustTrack® and Mie pDR-1200) operating at a high temporal resolution (20s and 1s respectively) for PM2.5 and PM10. PAH in the particulate and gas phases were recovered from quartz fibre filters and polyurethane foam plugs using pressurised solvent extraction, the extracts were then analysed by gas chromatography- high-resolution mass spectrometry. Our results shows that the direct contribution of ships traffic to PAHs in the gas phase is 10% while the contribution to PM2.5 and to PM10 is from 1% up to 8%.

Introduction

Whilst pollutant emissions from land-based sources are gradually decreasing, those from shipping show a continuous increase (Lu et al., 2006, Colvile et al., 2001). Emissions of particles, gases, heat and water vapour from ships can cause changes in the microstructure of marine stratiform clouds and produce the phenomenon known as “ship tracks” (Hobbs et al., 2000). According to global estimates, shipping emits between 0.9 and 1.7 million tons of particulate matter annually with nearly 70% of the emissions occurring within 400 km of the coast (Moldanova et al., 2009). The ship-exhaust particles are composed of elemental, organic and inorganic carbon, sulphate and ash as well as nitrates (Cooper, 2003). Cooper (2003) reported that fuels with higher polycyclic aromatic hydrocarbons (PAHs) contents are likely to give higher PAH exhaust emissions. Moldanova et al. (2009) found that the higher organic carbon content of the particulate matter (PM) collected in the cooled exhaust is accompanied by a much richer composition of PAH species comparing to the few PAH species observed in PM collected in the hot exhaust. Ship emissions contribute significantly to the levels of atmospheric pollution found over the Mediterranean region in the summer (Dalsoren et al., 2009). They can even modify the radiation budget through sulphate aerosol forcing (Marmer and Langmann, 2005). Furthermore, ship emissions when docked and during port manoeuvring can have a substantial negative effect on local air quality. As a result, in coastal areas there is now increasing concern about the contribution that marine vessels have on local air quality. In cases where the harbours are located near to densely populated urban areas the emissions from ships could be particularly important. Several authors have tackled the problem in different harbour areas using different approaches. The results of dispersion modelling at Danish ports (Saxe and Larsen, 2004) showed an annual contribution of around 0.1–0.2 μg m−3 for the harbour of Copenhagen. The Fagerli and Tarrason (2001) simulation results indicate that ship traffic emissions contribute to 5%–10% of PM10 concentrations in large parts of Great Britain, Portugal and Italy. Isakson et al. (2001) reported that in the harbour of Gotenborg (Sweden) the atmospheric particulate mass load in the particle size range of between 0.01 μm and 0.63 μm was increased by an average factor of 1.3 in the ship plumes when measured at distances that varied between 600 m and 2000 m. Receptor models such as factor analysis and Chemical Mass Balance (CMB) have also been used to characterise the contribution of harbour activities to atmospheric aerosol concentrations (Isakson et al., 2001, Alastuey et al., 2007, Minguillon et al., 2008). An analysis of the contribution of ship traffic on the composition of atmospheric aerosol in the harbour area of Melilla (Spain) has been studied by Viana et al. (2009) using a receptor model. The results showed a direct contribution of ship traffic of around 2% to the PM10 levels. They also found a possible contribution to secondary aerosol levels of around 4%, due to both ship traffic in the harbour and the passage of ships through the Straits of Gibraltar. In the same article (Viana et al., 2009) it was also reported an absence of correlation between daily aerosol concentrations with increasing ship traffic confirming the potential usefulness of high temporal resolution measurements. Simulations have been used by Lucialli et al. (2007) and Gariazzo et al. (2007) to study air pollution in Italian harbours. However, difficulties were found when trying to simulate PM10 concentrations. An intensive measurement campaign based on high temporal resolution instruments have been carried out Lu et al. (2006). However, the problem of ship traffic is still open and further investigations are needed to understand the role of harbour activities on air quality.

Stortini et al. (2009) have applied a multivariate statistical analysis using both PCA and PMF to the metal concentrations measured in PM2.5 samples taken at a site near the harbour area of Venice (Sacca San Biagio – St. 1 described later on). This analysis showed the presence of a factor loaded with V and Co that could include a sizeable contribution from ship emissions. Furthermore the correlation between elemental concentrations and wind direction during the PM10 sampling campaign indicated an increase in V, Cd and Pb when the wind was from the N–NNE directions, when the sampler was down wind of the ship traffic and harbour area. However the data analysis reported by Stortini et al. (2009) could not separate ship emissions from industrial/urban emissions. This may have been partly due to the limited number of chemical species considered, and the relatively long sampling time (24 h) used a site at which there is a clear wind direction pattern due to local meteorological circulation. This results in atmospheric mixing of the emissions from all the different sources masking their contributions.

Research work dealing with the analysis of ship plumes shows that in terms of numerical concentrations submicron particles dominate the size distributions (Healy et al., 2009, Ault et al., 2010). However, an analysis of the mass distribution of particles emitted by the large diesel engines of ships, reported by Moldanova et al. (2009), shows a bimodal mass size distribution with a first maxima at around 0.5 μm and a second maxima around 7 μm. This means that ship emissions could influence PM1, PM2.5 and also PM10 concentrations. This is compatible with the results reported in Mazzei et al. (2008) in the area of Genova where source apportionment showed a contribution of a factor loaded with Ni and V, that was at least partially associate with ship emissions, in PM1, PM2.5 and PM10.

In this work, an alternative methodology for evaluating the contribution of ship traffic to local aerosol concentrations has been developed. It is based on the use of short-time scale aerosol concentration measurements, with simultaneous measurement of the wind direction, aimed at obtaining a better correlation between the aerosol concentration and the passage of ships past the sampling point. Measurements taken in the harbour area of Cherbourg (France) have shown that short term concentration peaks can be a good indicator of the contribution of ship traffic (Roupsard et al., 2009) to local air pollution.

The simultaneous use of high temporal resolution measurements of PM concentrations and wind directions has been proven useful in the apportionment of specific industrial sources (Contini et al., 2010, Snyder et al., 2009). The advantage is that they do not need a complete chemical characterisation of the aerosol and overcome the difficulties of receptor models by including meteorological information to help identify specific sources that may contribute only in small wind direction sectors. The main difference in the method proposed here, is that is that ship emissions are from moving sources. So it is not enough to consider wind direction but information is also required on the effective time of ship passages, so the information can be included in the analysis.

The area of Venice, object of this present study, is heavily affected by anthropogenic activities such as industrial emissions from the Porto Marghera industrial area and traffic pollution from the nearby motorway and ring road of Mestre. An assessment of the emissions inventory from passenger ships by the Environmental Regional Protection Agency (ARPAV), for the Venice area, was estimated at 1000 tons/year for NOx, 1100 tons/year for SO2 and 170 tons/year for PM10 (ARPAV, 2007).

The air quality of Venice is influenced not only by industrial and port activities, but also by the local climatology, that is determined by its geographical position within the Po Valley. The local air circulation and the consequent dispersion of pollutants are influenced by sea breezes: during the night, winds come prevalently from the NE, and pollutants are transported towards the sea, while in the morning, when winds come prevalently from the SE, pollutants are transported towards the hinterland.

The aim of the work presented here was to assess the direct contribution of ships on the coarse and fine particulate matter and PAH pollution whilst in the Venice harbour area and to compare the results with the emission inventories produced at a local level. The work is based on the measurement of primary emissions at three measurements sites located within a 2 km radius from the harbour area and the nearby shipping canal (Fig. 1). In this canal there is a constant traffic of ferries, cruise ships and smaller vessels. As these three sites are located near the emission sources, it is hoped that the hourly concentration results will reflect the direct contribution of ship emissions to primary particle levels. Ship traffic possibly contributes to the levels of secondary inorganic aerosols (especially sulphate particulate matter, PM) due to sulphur emissions in the exhaust plumes. However the fast chemical transformation/reactions that can occur in the plumes involve only very fine particles (Song et al., 2003, Chen et al., 2005) so they are likely to have only a small contribution to PM2.5 and PM10 levels. The transformation and accumulation rate of sulphur dioxide and sulphate into particles with a dimensional range of around 1 μm occurs over tens of hours and contributes about 1–3%/h to the background atmospheric levels (Rodhe, 1978). This means that the secondary aerosol contribution from ship emissions will be distributed over a large area. This makes it difficult to extract the contribution of ship traffic to secondary aerosol from the data, as aerosol from distant sources will be present. So samples collected at our experimental stations will be well mixed with aerosol from the regional background. Alastuey et al. (2007) carried out a factor analysis of their data from the chemical characterisation of aerosol collected in the harbour area of Tarragona (Spain). Their results showed a factor loaded with Ni and V that was likely associated to ship emissions but was not correlated with sulphate or SO2 concentrations, they concluded that this may be because the ship contribution was masked by contributions from other sources. Minguillon et al. (2008) carried out aerosol measurements of the fine and quasi-ultrafine fraction in the Los Angeles Long Beach harbour area, a Chemical Mass Balance (CMB) analysis showed a ship emission profile loaded with V and Ni that was not correlate with sulphur. These secondary particles are therefore not considered relevant to how ships in harbours pollute their nearby surroundings (Saxe and Larsen, 2004).

Section snippets

PM2.5, PM10

In this work PM2.5, and PM10 measurements have been taken at three sites located along the Giudecca channel, the deep water passage for ships moving to and from the Venice cruise ship terminal (Fig. 1). On arrival, ships navigate through the Giudecca channel, in some cases they navigate under their own power and others (according to the size of the ship) are towed. Therefore an arriving ship passes St. 2 first, then they pass near St. 1 and finally they dock in the harbour area near St. 3.

Polycyclic aromatic hydrocarbons (PAHs)

In this work, samples of 15 PAH (Acenaphthene (Ace), Acenaphtylene (Acy), Fluoranthene (Fl), Phenantrene (Phe), Anthracene (An), Fluorene (Flu), Pyrene (Py), Benzo(a)anthracene (B(a)An), Chrysene (Cr), Benzo(b)fluoranthene (B(b)F), Benzo(k)Fluorathene (B(k)F), Benzo(a)pyrene (B(a)Py), Dibenz(a,h)anthracene (DB(a,h)An), Benzo(g,h,i)perylene (B(ghi)Pe) and Indeno(1,2,3 c-d)pyrene (IPy)) were obtained in operationally defined particulate and gas atmospheric phases at St.1 between March and October

Discussion of results

A summary of the PM2.5 and PM10 concentrations measured at the three sites is reported in Table 1, the summer period (in which the ship traffic is more intense) is shown separately. Daily PM concentrations at the different sites show a similar trend and are well correlated with those of the regional monitoring network. However, it is difficult to extract the contribution of ships from the atmospheric aerosol long-term or daily averages. This because the expected contribution is usually small,

PAHs

In this work it was not possible to obtain high temporal resolution measurements to estimate the direct influence of ship traffic on organic compounds. So the daily concentrations, coupled with wind direction information, has only been used for PAHs, which are toxic and specific combustion marker compounds. Daily concentrations of ∑PAHs (the sum of 15 PAHs congeners), measured at St.1, in the gas phase range from 0.02 to 7.04 ngm−3 with a mean of 3.20 ngm−3. The concentrations of single PAH

Conclusions

The aim of this work was to investigate the direct influence of ship traffic on atmospheric concentration levels of PM2.5, PM10 and PAH in the area close to Venice. Results indicate that daily concentration averages are not suitable for the analysis of ship emission contributions because they do not appear to be correlated with the actual ship traffic, so a more specific methodology, based on high temporal resolution measurements, has been developed that is suitable for moving sources. The

Acknowledgements

This work has been carried out with the financial support of the Venice Port Authority. The authors wish to thank Mr. I. Ongaro, P. Silviero and L. Zagolin of the University Ca’ Foscari of Venice (Environmental Sciences Department) for their help in setting up and in keeping operative the experimental sampling sites, and specially Prof. P. Cescon, Prof. F. Prodi, Ing. G. Santachiara, Dr. D. Cesari and Ing. C. Elefante for their valuable suggestions and continual insight during the preparation

References (45)

  • N.R. Khalili et al.

    PAH source fingerprints for coke ovens, diesel and gasoline engine, highway tunnels, and wood combustion emissions

    Atmospheric Environment

    (1995)
  • S. Kingham et al.

    Winter comparison of TEOM, MiniVol and DustTrak PM10 monitors in a woodsmoke environment

    Atmospheric Environment

    (2006)
  • G. Lu et al.

    Identification and characterisation of inland ship plumes over Vancouver, BC

    Atmospheric Environment

    (2006)
  • P. Lucialli et al.

    Harbour of Ravenna: the contribution of harbour traffic to air quality

    Atmospheric Environment

    (2007)
  • E. Marmer et al.

    Impact of ship emissions on the Mediterranean summertime pollution and climate: a regional model study

    Atmospheric Environment

    (2005)
  • F. Mazzei et al.

    Characterisation of particulate matter sources in an urban environment

    Science of the Total Environment

    (2008)
  • P.H. McMurry

    A review of atmospheric aerosol measurements

    Atmospheric Environment

    (2000)
  • M.C. Minguillon et al.

    Seasonal and spatial variations of sources of fine and quasi-ultrafine particulate matter in neighborhoods near the Los Angeles-Long Beach harbor

    Atmospheric Environment

    (2008)
  • J. Moldanova et al.

    Characterisation of particulate matter and gaseous emissions from a large ship diesel engine

    Atmospheric Environment

    (2009)
  • F. Prodi et al.

    Aerosol fine fraction in the Venice lagoon: particle composition and sources

    Atmospheric Research

    (2009)
  • K. Ravindra et al.

    Atmospheric polycyclic aromatic hydrocarbons: source attribution, emission factors and regulation

    Atmospheric Environment

    (2008)
  • H. Saxe et al.

    Air pollution from ships in three Danish ports

    Atmospheric Environment

    (2004)
  • Cited by (118)

    • Impact of ship emissions at a high receptor point in the port of Naples

      2022, Atmospheric Environment
      Citation Excerpt :

      Other parameters apart from wind speed and direction as temperature, humidity and atmospheric stability can influence the dispersion of pollutants but they have not been considered in this analysis. Several authors have estimated the average intrinsic primary contribution of in-port shipping on the average atmospheric concentration in a particular wind sector based on differences between measured concentrations of when ships were in the harbour and when they were not (Contini et al., 2011; Merico et al., 2016). This methodology can be applied when a limited number of ships are present in the harbour area.

    View all citing articles on Scopus
    View full text