Elsevier

Atmospheric Environment

Volume 42, Issue 2, January 2008, Pages 323-336
Atmospheric Environment

Modelling the impact of elevated primary NO2 and HONO emissions on regional scale oxidant formation in the UK

https://doi.org/10.1016/j.atmosenv.2007.09.021Get rights and content

Abstract

Recent increases in the fractional contribution of NO2 to NOx emissions from the road transport sector in Europe are well documented in the literature. A photochemical trajectory model has been used to simulate the impact of increasing the NO2 fraction on the chemical evolution of air masses arriving at the TORCH field campaign site in the southern UK during late July and August 2003, a period which included a widespread photochemical pollution episode associated with a heat-wave. The impact of partial emissions of NOx in the form of nitrous acid (HONO) has also been considered. The model incorporates emissions of NOx, CO, SO2, methane and a detailed speciation of non-methane volatile organic compounds (VOC), coupled with a comprehensive description of the chemistry of secondary pollutant formation. An increase in the fractional contribution of NO2 to NOx emissions from 0% to 30% (v/v) results in a 2.49 ppb increase in the simulated campaign mean mixing ratio of oxidant (defined as the sum of O3 and NO2). This is almost exclusively in the form of O3, and represents a ca. 7% increase in the simulated campaign mean O3 mixing ratio. Consideration of 156 events, at 6-hourly resolution throughout the campaign period, indicates that oxidant increments during the heat-wave period are generally simulated to be greater than those for the remainder of the campaign, with a maximum increment of ca. 12 ppb. The increases in oxidant mixing ratios are shown to derive from both the direct effect of increased NO2 input, and indirectly from the enhanced regional-scale chemical processing that this promotes. An illustrative increase in the fractional contribution of HONO to NOx emissions from 0% to 5% (v/v) results in increases in the simulated campaign mean mixing ratios of O3 and oxidant of 1.51 and 1.15 ppb, respectively, with the smaller increment for oxidant reflecting a decrease in the NO2 mixing ratio resulting from a notably enhanced NOx oxidation rate. The oxidant increments during the heat-wave period are once again simulated to be greater than those for the remainder of the campaign, with a maximum increment of ca. 11 ppb, which results exclusively from the enhanced regional-scale chemical processing that the HONO emissions promote. The impact of increased fractional NO2 and HONO emissions on the rate of oxidation of NOx to nitrate is also illustrated and discussed.

Introduction

Nitrogen oxides (NOx=NO+NO2) are released into the atmosphere from a variety of sources. In highly populated regions, the combustion of fossil fuels (e.g., from road transport) represents the dominant source of NOx, with emission mainly in the form of nitric oxide, NO. However, considerable attention has recently been focussed on the contribution to the emissions which is in the form of nitrogen dioxide, NO2 (commonly referred to as “primary NO2”) and the impact this has on air quality in the urban environment (e.g., Latham et al., 2001; Jenkin, 2004a; Carslaw, 2005; Carslaw and Beevers, 2005; Carslaw et al., 2007; AQEG, 2007). For example, the UK Air Quality Expert Group (AQEG) has recently collated and reported observational evidence for a general increase in the fractional contribution of NO2 to NOx emissions over the period 2002–2005, based on monitoring data from ca. 80 automatic network sites in the UK (AQEG, 2007). The observations suggest that site-to-site variations exist in the magnitude of the change, but are indicative of an average increase of about 5% (v/v) in the percentage contribution of NO2 to NOx emissions over the period, such that contributions approaching 30% are now apparent at a number of locations. The observed increase in primary NO2 was attributed to two main factors: (i) an increasing penetration of EURO III light duty diesel vehicles fitted with oxidation catalysts into the vehicle fleet. The results of emissions test measurements were presented by AQEG (2007), which demonstrate that these vehicles emit between 20% and 70% of NOx in the form of NO2; and (ii) the fitting of catalytically regenerating particle traps to bus fleets. This technology uses a purposely elevated NO2 emissions fraction to facilitate combustion of soot at low temperatures (<250 °C), thereby allow the particle trap to be regenerated at accessible temperatures. On the basis of projected vehicle fleet composition, AQEG (2007) also estimated that NO2 would account for ca. 18% of NOx emissions on average in the UK by 2010, and ca. 23% in central London. Observations from other European countries (AQEG, 2007; Hueglin et al., 2006; Palmgren et al., 2007) are indicative of comparable increases in the NO2 emissions fraction over widespread areas of Europe.

An increase in the fraction of NOx emitted as NO2 not only influences levels of NO2 in the local environment, but can also potentially have an impact on local levels of ozone (O3) and on oxidation processes occurring on regional scales. As described in detail elsewhere (e.g., Leighton, 1961), the local-scale impact results from the well-established chemical coupling of NOx and O3, which typically occurs on the timescale of minutes, by the following reactions:NO+O3NO2+O2NO2+hν(+O2)NO+O3

Because of this strong chemical coupling, the term “oxidant” is sometimes used as a collective term for NO2 and O3 (e.g., Kley et al., 1994; Clapp and Jenkin, 2001; Jenkin, 2004b). This reaction cycle thus partitions NOx between its component forms of NO and NO2, and oxidant between its component forms of O3 and NO2, but conserves both NOx and oxidant. As a result, oxidant derived from background O3 is partitioned between the forms of NO2 and O3, with a progressively greater proportion in the form of NO2 as NOx increases as a result of received emissions. Similarly, oxidant derived from emitted NO2 is subsequently partitioned between the forms of NO2 and O3, with a progressively greater proportion in the form of O3 as NOx decreases with dilution. The emission of an elevated fraction of NOx in the form of NO2 therefore potentially has a direct local-scale impact on ambient concentrations of both NO2 and O3.

The photochemical processing of emitted NOx and volatile organic compounds (VOC) on local-to-regional scales leads to the conversion of NO to NO2 and therefore the formation of oxidant (e.g., Jenkin and Clemitshaw, 2000), as illustrated in Fig. 1. The photochemical processing is driven by reactions of free radicals, which can be generated from the photolysis of O3, and from its reactions with emitted alkenes. Consequently, the emission of an elevated fraction of NOx in the form of NO2, and its resultant direct effect on local O3 concentrations, also has a secondary impact on oxidant formation on local-to-regional scales by stimulating radical formation. The additional oxidant generated is also partitioned between the forms of NO2 and O3, with the relative contributions depending on the level of NOx at the given location.

In addition to the recent interest in the magnitude and trends in primary NO2 emissions, historical observations made at urban roadside locations, in tunnels and in laboratory facilities are consistent with a small fraction (0.5–1%) of oxidised nitrogen being emitted in the form of nitrous acid, HONO (e.g., Pitts et al., 1984; Kirchstetter et al., 1996; Martinez-Villa, 2001; Kurtenbach et al., 2001; Gutzwiller et al., 2002). Although this is a small fraction, it is potentially significant from a chemistry point of view, because HONO photolyses efficiently to generate free radicals, thereby promoting VOC oxidation and additional NO-to-NO2 conversion:HONO+hνOH+NO

Given that HONO is known to be generated by surface reactions of NO2 with water vapour and with semi-volatile exhaust organics (e.g., Gutzwiller et al., 2002; Ramazan et al., 2006; and references therein), it is possible that the recent increase in the primary NO2 emissions fraction has also been accompanied by an increase in the fractional emission of HONO, as a result of surface reactions in the tailpipe. Indeed, due to the well-documented interferences of oxidised nitrogen species such as HONO for measurements of NO2 made with chemiluminescent analysers (e.g., Winer et al., 1974; Grosjean and Harrison, 1985; Fehsenfeld et al., 1987), the EU reference method applied in the majority of studies, it is possible that a currently unknown fraction of the reported NO2 increases may actually be in the form of HONO.

In the present paper, an investigation of the impact of emissions of NO2 and HONO on regional scale oxidant formation is described. A photochemical trajectory model (PTM) containing speciated emissions of 124 non-methane VOC, and a comprehensive description of the chemistry of VOC degradation, has been used to simulate the chemical evolution of boundary layer air masses arriving at a field campaign site in the southern UK during late July and August 2003, a period which included a widespread photochemical pollution episode. The sensitivity of the simulated oxidant mixing ratios throughout the campaign to changes in the fraction of oxidised nitrogen emitted as NO2 and HONO is examined, and the impact on the rate of oxidation of NOx to nitrate is also illustrated and discussed.

Section snippets

The photochemical trajectory model

The PTM used in the present study is based on that which has been widely applied to the simulation of photochemical O3 formation in north-west Europe (Derwent et al., 1996, Derwent et al., 1998; Jenkin et al., 2002a). It has been used recently to simulate O3 and organic aerosol formation, and the detailed speciation of organic compounds, in relation to observations made during the tropospheric organic chemistry experiment (TORCH) campaign at Writtle, Essex, UK (51.74°N; 0.42°E), a rural site

Base case simulation

Fig. 2 shows a comparison of the base case simulated O3 mixing ratios at the arrival point with those observed for the entire campaign period. The model is able to recreate the general variations of O3 throughout the 5-week campaign, which reflect the variety of prevailing meteorological conditions. As discussed in detail previously (Utembe et al., 2005; Johnson et al., 2006a; Lee et al., 2006), and identified in the figure, the campaign can be broadly subdivided into five periods: (i) The

Conclusions

The results of the simulations presented above indicate that an increase in the fractional contribution of NO2 to NOx emissions to 30% (a level which could result from the projected penetration of existing technology into the vehicle fleet) would be accompanied by an increase in oxidant mixing ratios on local to regional scales, resulting from both the direct effect of increased NO2 input, and indirectly from the enhanced regional-scale chemical processing that this promotes. For the rural

Acknowledgements

The work described in this paper was funded by the Department for Environment, Food and Rural Affairs (Defra) under contracts EPG 1/3/200 and AQ03508. M.E.J. also gratefully acknowledges support from the UK Natural Environment Research Council (NERC), via provision of Senior Research Fellowship grants NER/K/S/2000/00870 and NE/D008794/1.

References (32)

Cited by (35)

  • Investigation of vehicle cold start primary NO<inf>2</inf> emissions inferred from ambient monitoring data in the UK and their implications for urban air quality

    2019, Atmospheric Environment
    Citation Excerpt :

    The influence of those interferences might lead to overestimation of NO2 in urban areas (Dunlea et al., 2007; Harrison et al., 2012). In addition, direct emissions of HONO, thought to comprise a small, but still significant and uncertain, component of vehicle exhaust (Crilley et al., 2016; Jenkin et al., 2008; Kurtenbach et al., 2001), is also a potential interferent in NO2 measurements. However, in our analysis (see section 2.3 below) all the above potential interferences would be expected, to a first approximation, to co-vary with NO2 and hence have limited effect upon the inferred primary NO2/NOx ratio.

  • Investigation of the impact of short-timescale NO<inf>x</inf> variability on annual mean oxidant partitioning at UK sites

    2014, Atmospheric Environment
    Citation Excerpt :

    [OX]L is believed to be mainly derived from primary emissions of NO2, and is therefore often represented by the term f-NO2. [ NOx], where f-NO2 is the fraction of NOx emitted as NO2 (e.g. AQEG, 2007, 2009; Jenkin et al., 2008; Grice et al., 2009; Simmons and Seakins, 2012). [ OX]B provides a quantification of the background [O3] which would exist at the given location in the notional absence of NOx, i.e. when [OX]L = 0, and the local-scale chemical coupling described by reactions (1) and (2) above cannot occur.

  • Chemical composition of PM10 in Switzerland: An analysis for 2008/2009 and changes since 1998/1999

    2012, Atmospheric Environment
    Citation Excerpt :

    Another hypothesis is, that the increasing average ozone concentrations in Switzerland and other European countries might enhance the formation of nitric acid during the night (see e.g. Seinfeld and Pandis, 1998 for associated reactions) and thus the formation of nitrate. The increasing fraction of NO2 in NOx from road traffic exhaust emissions (Grice et al., 2009) might also contribute to increasing average ozone as indicated by Jenkin et al. (2008) and therefore favours the above mentioned mechanism. However, an atmospheric chemical modelling study that includes all relevant processes is required for clarification of the observed trend of particulate nitrate.

View all citing articles on Scopus
1

Current address: School of Chemistry, University of Bristol, Bristol BS8 ITS, UK.

View full text