Monitoring VOCs in Ambient Air

Increased media coverage of air pollution in recent years has led to heightened public interest in air quality. Since the first air pollution measurements were taken in the 1950s, the breadth of measurements recorded has widened in-line with what is known about the range of substances present in the air we breathe, together with a better understanding of emission sources and the related health impacts.

Motor vehicle emissions have, in recent years, been the predominant source of concern for many of the pollutants that the public are exposed to. These pollutants include Nitrogen Dioxide (NO2) and particulate matter (PM) where exceedances of EU limit values over the last 10 to 15 years in the UK, have drawn the most public attention.

Examples of less well-known pollutants present in ambient air include Non-Methane Volatile Organic Compounds, or NMVOCs, often referred to simply as VOCs. In the past the sources of these VOCs have also been predominated by vehicle emissions. However new technologies have been effective in mitigating vehicle emissions of VOCs such as benzene over the last twenty years. In the present day, concerns over VOCs emissions are now more focussed on emissions from solvents and glues, in addition to a diverse range of other sources beyond those.

Sources, trends and seasonal variation

This article will consider the sources, trends and seasonal variation in ambient VOCs concentrations within the UK, as well as looking at how to re-focus measurement priorities to target a different range of VOCs which are becoming of greater interest for policy.

Natural gas often contributes to the emission of ethane and other VOCs, for instance through industrial distribution pipeline leaks. Figure 1 shows hourly ethane concentrations in the suburban area of Greenwich, London, for each hour of the day, averaged across 2018. The chart
shows that for every hour there is a strong seasonal pattern in ethane concentrations with much higher concentrations in winter.

Several VOCs are emitted from petroleum fuel, particularly butane, pentane and xylene isomers as well as hexane and benzene. In recent years, the introduction of bioethanol to fuel mixtures has added ethanol to this list. Emissions of benzene are predominated by incomplete combustion, which is common in vehicle emissions – where a lower ratio of oxygen means carbon is not completely oxidised, resulting in the production of CO2 emissions that include various VOC substances.

If we look at concentrations of benzene for the same area of London in 2018 (Figure 2), we see again that for every hour of the day, there is a strong seasonal pattern in benzene concentrations with much higher concentrations in winter.

Although the sources of ethane and benzene are largely different, we can see a similar variation in annual concentrations in both Figures 1 and 2 due simply to the meteorological conditions resulting in poorer pollutant dispersion during winter.

Natural emissions sources, such as forest fires ensure that there is a consistent overall background level of global atmospheric VOCs. In addition, there are further seasonal contributions due to biogenic sources of VOCs from trees and plants. For instance, isoprene – known to be emitted from trees, particularly in the summer months when sunshine is more concentrated – is measured by the UK Automatic Hydrocarbon network.

The plot in Figure 3 shows trends for isoprene across London for the whole of 2018 for each hour of the day, showing peaks in summer during daylight hours.

Historically, information on ambient measurements of VOCs and emission inventory estimates have been crucial to policy makers in tracking the impact of their decisions. Particularly when regulating emissions from road vehicles in order to effectively and efficiently mitigate emissions of VOCs in urban areas (Ricardo Energy & Environment, 2016)1. There are a multitude of VOCs emitted from combustion activities, with great variations in substances depending upon the relevant fuels. Of the common VOCs found in vehicle emissions, both benzene and 1,3-butadiene would be expected, but
over time mitigation of these toxic VOCs has been effective in reducing both estimated emissions and measured ambient concentrations.

In Figure 4 we can see how the ambient concentrations of benzene, particularly on Marylebone Road, in London, compare against the emission estimates of benzene with the UK National Atmospheric Emissions Inventory information on emissions from road vehicles (NAEI). The trends from 2002 to 2008 show how replacing benzene with toluene in petroleum and the introduction of vehicle exhaust catalysts and evaporative canisters have been used to reduce emissions of benzene from vehicle exhaust fumes

The more toxic VOCs such as benzene and 1,3-butadiene, regulated under EU directives and UK legislation (Table 1), have not consistently exceeded limit values for several years. This is likely to be one of the reasons why emissions and health impacts of VOCs in the UK have not generally been of high public concern.

One of the more toxic VOCs – 1,3-butadiene, is emitted to air by vehicles and the production of synthetic rubber, as well as other industrial activities. It is now rarely detected in high concentration in UK national monitoring networks, as it does not persist in air for long periods of time. Annual average observations of benzene are less than 20% and 30% of the limit values for England, Wales and Scotland; and Northern Ireland, respectively, since 2015 (Figure 5).

Measurement techniques

VOCs are more difficult to detect in ambient air, compared with that of their inorganic cousins, such as NO2. Additionally, VOCs include more than 500 different compounds and can be emitted from many different sources. The technology necessary to separate total VOCs into separate compounds and measure them also varies greatly in expense and accuracy. No perfect instrument exists to cover all aspects of measurement. As a result, compromises often must be made when selecting instrumentation.

The most reliable and economically viable means to measure VOCs in ambient air is by Thermal Desorption (TD), followed by Gas Chromatography (GC). This is usually carried out by using a Flame Ionisation Detector (FID) or Photo-Ionisation Detector (PID). This method of online analysis is common around the globe with variations in design because it provides reasonable time resolution – allowing for analysis to be completed in a relatively short period of time, ranging from 15 minutes to two hours. This method is also relatively inexpensive to purchase and operate compared with alternative instrumentation.

In the UK, a non-automatic sampler is used to measure benzene against the 5 µg.m-3 EU limit value. This limit is also adopted in the UK Air Quality Strategy (including a limit of 3.25 µg.m-3 in Scotland). A passive method was also used to measure 1,3-butadiene against the UK Air Quality Strategy limit value of 2.25 µg.m-3 until 2007. However, this ceased operation following consistent measurements below the lower analytical detection limit.

A TD is able to trap and concentrate VOCs onto a cold trap containing two types of sorbents capable of absorbing the full range of required VOCs from the sample. Following sampling, the TD then superheats the trap in order to remove the trapped VOCs and transfer them to the GC for analysis. The GC separates the VOCs using capillary columns before the FIDs (or a similar detector) produce a chromatographic trace (Figure 6).

Presently, the UK Automatic Hydrocarbon Network measures 29 VOCs from a list of 31 suggested in Annex X of the Directive on Cleaner Air for Europe 2008/50/EC (AQD). This list is a similar, smaller list compared with the list adopted by the United States Protection Agency (USEPA) and contains VOCs made up of Hydrogen and Carbon atoms only. Polar VOCs such as carbonyls and alcohols are omitted.

The VOCs listed in the AQD were chosen by a European Expert Group. At the time the selected list of VOCs were identified as the most important precursors to ozone formation in the presence of sunlight and NO2. However, this relatively small selection of VOCs – routinely measured in Europe and more recently in developing countries – is leaving gaps in knowledge regarding the global concentrations of many other important VOCs, often not accounted for. At the time the list was first created, the state-of-the-art TD-GC method was only able to measure non-polar VOCs due to the nature of the nafion dryer used to remove sample moisture.

Research (Dunmore et al, 2015) shows that the EU Directive target list of ozone precursor VOCs, predominantly emitted from petroleum and natural gas probably only captures about 30% of the key ozone precursor substances. It is now thought that longer chain hydrocarbons emitted from diesel vehicles are more significant in winter than previously thought, alongside oxygenated VOCs in summer (Figure 7).

The rising importance of VOCs from solvents, glues and other products

The relative contributions of different sources to overall emissions of VOCs has changed markedly over time. In the early 1990s, these were predominated by motor vehicles and the distribution and use of fossil fuels. There have been marked declines in observations of VOCs emitted from both of these sources over the last 25 years since a peak in 1990. As VOC emissions from these sources have decreased, emissions of VOCs through other means are consequently coming to the fore.

The emissions of VOCs in the present day are now dominated by production of solvents, glues and other products (Figure 8). The majority of emissions post 1990, predominantly contain polar VOCs such as ethanol, methanol and acetone, which are compounds that are not captured by the systems typically used in current air quality monitoring networks.

The rise of solvent and other product use as a percentage of total emissions in the UK (Figure 9) is significant.

The VOC emissions from these sources are different to those emitted from petrol vehicles. Possibly one of the most interesting and perhaps, surprising source of VOCs to emerge is that of domestic solvent use – that is, products we use in our home on a daily basis. As you next walk down the aisle in your local supermarket, just take a minute to consider that you are looking at hundreds of aerosol products containing VOCs.

Emissions from such products not only pollute the inside of our homes, but can also contribute significant emissions to ambient air by leaking outside through poorly ventilated buildings or open windows.

The data presented by the NAEI shows how the predominant source has changed since 1990, but more needs to be done to reduce VOC emission from solvents.

Future targets and mitigation

Through the National Emissions Ceilings Directive (NECD) 2016/2284/ EU, the UK is obligated to achieve targets set for 2020 and 2030. These targets include NMVOC targets. This directive sets a target of 32% and 39% reductions in VOC emissions compared with 2005 concentrations for 2020 and 2030 respectively.

Presently, the NAEI forecasts that the UK will achieve the target for 2020, however the NAEI data suggests that the UK is not likely to achieve the target for 2030.

Figure 11 shows both 2020 and 2030 emissions ceilings for the UK and the projections for 2020, 2025 and 2030 against 2016 emissions respectively. The target currently forecasts that the UK is likely to exceed by approximately 83 kilotonnes (Clean Air Strategy, 2018), with a large proportion shown to be from solvent production.

The current information provided by emissions inventories weighs heavily in favour of including polar species to the suite of VOCs routinely measured by national monitoring networks, over and above the ozone precursor list. Capturing concentrations of polar VOCs such as ethanol, methanol and certain carbonyl species is likely to give data users a much clearer picture of the most abundant VOCs currently present in the air we breathe. Such insight would help policy makers to make better informed decisions about how they can begin to implement regulations to help further reduce pollution emissions and improve upon existing measures.

Improving measurement capabilities for VOCs

The range and composition of the VOCs that should ideally be included in the future of the Automatic Monitoring Network in the UK presents a new challenge for both instrument manufacturers and operators. Including polar species requires the need for more advanced engineering solutions to replace conventional commercial analysers, or the use of more than one instrument. Some bespoke research machines have been developed by universities to overcome these problems, but these have yet to be adopted into production by manufacturers. It is likely that the need for capturing such species will drive further innovation in the technology used for more widespread operation of online gas chromatography systems in future.

Indeed, it will be the only route to providing a comprehensive and accurate picture of VOC concentrations in ambient air – as we enter a new era, where policy makers must find a way to understand and address UK solvent emissions in order to meet the 2030 NECD ceiling.

Citations
Abreu, P., Carslaw, D., Davies, T., Dernie, J., Kent, A., Stacey, B., Telling, S and Wakeling, D, UK Hydrocarbons Network Annual Report for 2015. 2016. Available online at: https://uk-air.defra.gov.uk/assets/ documents/reports/cat13/1611011540_2015_HC_Net_Report_Issue_1.pdf Defra, Clean Air Strategy 2018. Available online at: https://consult.defra.gov.uk/environmental-quality/cleanair-strategy-consultation/user_uploads/clean-air-strategy-2018-consultation.pdf Accessed 27/07/2019 Dunmore, R.E., Hopkins, J.R., Lidster, R.T., Lee, J.D., Evans, M.J., Rickard, A.R., Lewis, A.C., and Hamilton, J.F. Diesel-related hydrocarbons dominate reactive carbon in megacities. Atmospheric Chemistry & Physics, 15, 9983-9996, 2015. UK-AIR Air Information Resource, Defra. Available online at: https://uk-air.defra.gov.uk/. Accessed 27/07/2019 UK National Atmospheric Emissions Inventory, BEIS. Available online at: http://naei.beis.gov.uk/. Accessed 27/07/2019