1.1 Background

Stansted Airport is London’s third busiest international airport, handling approximately 25.9 million passengers a year. The airport is situated approximately 40 miles north of London, in North West Essex. It is situated outside the general urbanised area of Greater London, and its surroundings are rural.

Stansted Airport Ltd is required, under the terms of its Section 106 Planning Agreement with the Local Authority (Uttlesford District Council), to carry out monitoring of oxides of nitrogen and particulate matter at an agreed location. Prior to 2006, monitoring was required for three months per year; from 2006 onwards, continuous monitoring throughout the year has been required. Ricardo Energy & Environment was contracted by Stansted Airport Ltd to carry out the required programme of air pollution measurements during 2017, the twelth full year of continuous monitoring.

Provisional data are reported to Stansted Airport Ltd monthly throughout the year. This annual report presents and summarises the fully validated and quality controlled dataset for the entire calendar year. Data in the annual report have been processed according to the rigorous quality assurance and quality control procedures used by Ricardo Energy & Environment. These ensure the data are reliable, accurate and traceable to UK national measurement standards.

This report covers the period 1^st January to 31^st December 2017.

1.2 Aims and Objectives

The aim of this monitoring programme is to monitor concentrations of two important air pollutants around the airport. The results of the monitoring are used to assess whether applicable air quality objectives have been met, and how pollutant concentrations in the area have changed over time. The pollutants monitored were as follows:

• Oxides of nitrogen (nitric oxide NO and nitrogen dioxide NO₂), using automatic techniques at two locations: Stansted 3 (High House) and Stansted 4 (Runway).
• Particulate matter (PM₁₀) at Stansted 3 and Stansted 4.
• Particulate matter (PM_2.5) at Stansted 3 and Stansted 4.

The automatic monitoring was supplemented by indicative monitoring of NO₂ using diffusion tubes at five locations around Stansted Airport, and an additional nine locations in Hatfield Forest.

Monitoring data collected at Stansted are compared in this report with:

• Relevant UK air quality limit values and objectives.
• Corresponding results from a selection of national air pollution monitoring sites.
• Statistics related to airport activity.

In addition, periods of relatively high pollutant concentrations are examined in more detail.

1.3 Air Quality Limit Values and Objectives

This report compares the results of the monitoring survey with air quality limit values and objectives applicable in the UK. These are summarised below.

2.1 Pollutants Measured

2.2 Monitoring sites and Methods

2.2.1 Site Locations

Automatic monitoring was carried out at two sites during 2017. These are referred to as Stansted 3 and Stansted 4 (the numbering of the sites continues the sequence used for previous short-term sites in earlier monitoring studies). The location descriptions of both sites fall into the category “other” as defined by the Defra Technical Guidance on air quality monitoring LAQM.TG(16)⁹, (i.e. “any special source-oriented or location category covering monitoring undertaken in relation to specific emission sources such as power stations, car-parks, airports or tunnels”).

These two automatic sites were supplemented by five sites at which diffusion tubes were used to monitor NO₂ on a monthly basis. These were located at the Stansted 3 automatic site, and four sites to the north, east, south and west of the airport. Further to this, an additional nine diffusion tube sites were located around Hatfield Forest from August 2017.

Table 2 describes the monitoring locations. Figure 1 shows a map of the locations of all monitoring sites used in this study. Automatic monitoring sites are shown by purple dots, diffusive samplers by yellow dots.

Figure 1: Locations of the Automatic and Diffusive air monitoring sites around Stansted Airport

The location of the automatic monitoring site at High House (Stansted 3) was agreed with Stansted Airport, Uttlesford District Council and Ricardo Energy & Environment. It is located just outside the eastern perimeter of the airport. It is considered to be close enough to the airport to detect effects relating to airport emissions. It is also close to vulnerable receptors, being located in a nursery school car park. The A120 main road runs approximately 1.5 km to the south of the site. The monitoring apparatus is housed in a purpose-built enclosure. Figure 2 shows a photograph of the Stansted 3 site.

Figure 2: Photo of the Stansted 3 air monitoring site

Stansted 4 is located at the north-eastern end of the main runway, within the airport perimeter. It is intended to monitor any effects on air quality related to airport emissions. The location of Stansted 4 is included in Figure 1, and a photograph is provided in Figure 3.

Figure 3: Photo of the Stansted 4 air monitoring site

3.1 Quality assurance and Quality control

In line with current operational procedures within the Defra Automatic Urban and Rural Network (AURN) ¹⁰, full intercalibration audits of the Stansted air quality monitoring sites took place at six-monthly intervals. Full details of these UKAS-accredited calibrations, together with data validation and ratification procedures, are given in Appendix 1 of this report. In addition to instrument and calibration standard checking, the air intake sampling systems were cleaned and all other aspects of site infrastructure were checked.

Following the instrument and calibration gas checking, and the subsequent scaling and ratification of the data, the overall accuracy and precision figures for the pollutants monitored at Stansted are summarised in Figure 4.

When using diffusion tubes for indicative NO₂ monitoring, the LAQM Technical Guidance LAQM.TG(16)¹¹ states that correction should be made for any systematic bias (i.e. over-read or under-read compared to the automatic chemiluminescent technique, which is the reference method for NO₂). Throughout this study, diffusion tubes have been exposed alongside the automatic NO_x analyser at Stansted 3. These co-located measurements were used for bias adjustment of the annual mean diffusion tube data from the other sites.

The diffusion tube methodologies used for this monitoring programme provide data that are accurate to ? 25 % for NO₂. The limits of detection vary from month to month, but typically equate to 0.4 μg m^-3 for NO₂. Diffusion tube results that are below 10 times the limit of detection have a higher level of uncertainty associated with them. All were above this threshold.

4.3 Diffusion Data

Table 8 shows the NO₂ diffusion tube results for 2017. Tubes were exposed in triplicate at all sites. The results shown are the means of those triplicate measurements. The full dataset is shown in Appendix 4. The analyst provided diffusion tube data to two decimal places. These have been rounded to one decimal place in the table below, but are quoted as integer values in this report, in accordance with the reported uncertainty of the method.

Two results were rejected as they were suspected to be spurious and three tubes were missing on collection. Details of these results are shown in Table 10. All results considered to be “outliers”; results much lower than those of the other two co-exposed tubes, subsequently resulting in rejection.

Across the five Stansted sites, annual mean NO₂ concentrations measured with diffusion tubes ranged from 17.3 to 33.2 μg m^-3. At Stansted 3, where diffusion tube results could be compared directly with data from automatic monitoring, the (rounded) annual mean concentration was 22.8 μg m^-3. This was in line with the annual mean of 23 μg m^-3 obtained using the reference technique (the chemiluminescence analyser).

Diffusion tubes are affected by several artefacts, which can cause them to under-read or over-read with respect to the reference technique. It has therefore become common practice to calculate and apply a “bias adjustment factor” to annual mean NO₂ concentrations measured by diffusion tubes, using co-located diffusion tube and automatic analyser measurements. This bias adjustment factor is calculated as the ratio of the automatic analyser result to the diffusion tube result. This factor can then be used to correct the annual means measured at the other monitoring locations. The bias adjustment factor was calculated using unrounded values from all months. On this basis, the bias adjustment factor was calculated to be 0.96.

The annual mean values from the other four Stansted diffusion tube sites as well as the nine Hatfield Forest sites were all corrected using the same bias adjustment factor.

In accordance with LAQM.TG(16)¹³, 2017 data for Hatfield Forest diffusion tube sites have been annualised because all sites have less than nine months worth of data. Across the nine Hatfield Forest sites, annual mean NO₂ concentrations measured with diffusion tubes ranged from 14.7 to 18.3 μg m^-3.

Please note:

• Only the annual mean concentration (not individual monthly values) should be adjusted in this way. This is because diffusion tube bias can vary considerably from month to month due to meteorological and other factors.
• Even after application of a bias adjustment factor, diffusion tube measurements remain indicative only.

4.4 Comparison with air quality objective

Details of the UK air quality standards and objectives specified by Defra are provided in Table 1.

The AQS objective for hourly mean NO₂ concentration is 200 μg m^-3 which may be exceeded up to 18 times per calendar year. At both Stansted 3 and Stansted 4 there were no recorded hourly mean NO₂ concentrations in excess of the hourly mean AQS objective of 200 μg m^-3, the sites therefore met the AQS objective for this pollutant.

The annual mean NO₂ concentrations measured at Stansted 3 and Stansted 4 during 2017 were 23 μg m^-3 and 12 μg m^-3 respectively. Both automatic sites were therefore within the annual mean AQS objective for NO₂ of 40 μg m^-3 for protection of human health and the objective of 30 μg m^-3 for protection of vegetation and ecosystems.

The bias-adjusted annual mean NO₂ concentrations measured at the five diffusion tube sites were all well within the AQS objective of 40 μg m^-3.

PM₁₀ was measured at both Stansted 3 and Stansted 4. At Stansted 3, the number of days when the 24-hour mean was in excess of 50 μg m^-3 was four. At Stansted 4, the number of days when the 24-hour mean was in excess of 50 μg m^-3 was two. These are both well within the maximum permitted number of exceedances (35), therefore these sites met the AQS objective for 24 hour mean PM₁₀.

PM_2.5 was measured at both Stansted 3 and Stansted 4. At Stansted 3 and Stansted 4, the annual mean was 9.2 μg m^-3 and 8.7 μg m^-3 respectively. These are both well within the annual mean objective of 25 μg m^-3, therefore these sites met the AQS objective for annual means for PM_2.5.

4.6 Periods of Elevated Pollution

As well as the AQS Objectives, a Daily Air Quality Index (DAQI) is used in the UK to communicate information about current and forecast air quality to the public. The Index is based on a scale of 1-10, divided into four bands (Low, Moderate, High and Very High): this provides a simple indication of pollution levels, similar to the pollen index. Low air pollution is between 1 and 3, Moderate is between 4 and 6, High is between 7 and 9, and Very High is 10 on the scale. This is intended to allow sensitive people to take any necessary action.

The concentration ranges associated with each band within the index are presented in Appendix 2.

PM₁₀ concentrations at Stansted 3 and Stansted 4 went into the Moderate band on four and two occasions respectively, throughout 2017.

NO₂ concentrations at both sites remained within the Low band throughout 2017.

The historic Air Quality Index data presented at the Department of Environment, Food & Rural Affairs (Defra) UK-AIR website¹⁴ shows air quality index bands that go from 4 (Moderate) to 10 (Very High) for most of the UK regions around the 22^nd January, 13^th February and 26^th September. These pollution episodes are consistent with the period of elevated PM and NO_x concentrations measured in the monitoring stations at Stansted, and explanations for this pollution events follow below:

DAQI for 22^nd January 2017

• There was widespread moderate/high pollution recorded across Southern England and Wales between Sunday 22^nd January and Tuesday 24^th January 2017. This was due to a high pressure system resulting in calm, settled conditions bringing about high PM₁₀ and NO₂. Calm conditions would have led to poor dispersion of local traffic emissions. Furthermore, elevated PM concentrations occured due to extensive PM_2.5 levels brought about by wood burning, the highest of any PM_2.5 episode since the DAQI was set up in 2012.

DAQI for 13^th February 2017

• Moderate PM₁₀ levels were recorded at Stansted 3 on Monday 13^th February 2017. This can be seen in the wider DAQI above with other areas of Eastern England showing an index of high. Air masses originated from Northern Europe including areas such as Poland and Germany, this demonstrates a contribution of PM from further afield.

DAQI for 26^th September 2017

• Moderate pollution was recorded across parts of Southern England on 26^th September 2017. This was mainly attributed to polluted air masses adding to existing UK emissions. These air masses brought with them agricultural, industrial and traffic related emissions, causing elevated levels of PM as seen on this date and the day after at both Stansted sites.

Some Ozone episodes were recorded over the summer period:

• Widespread moderate ozone was recorded on Saturday 17^th and Sunday 18^th June because of high temperatures (reaching 30°C). This increased heat and sunshine increases the transfer of nitrogen oxides and other ozone precursors into ozone.
• Similarly, on the 20^th to 22^nd June more hot weather contributed to elevated ozone levels in much of England.

4.7 Time Variation plot

4.7.3 Diurnal Variation

Bottom left graphs in Figure 13, Figure 14 and Figure 15 show diurnal variation in pollutant concentrations, as measured at Stansted 3 and Stansted 4.

Both sites showed typical urban area daily patterns for NO₂. Pronounced peaks can be seen for these pollutants during the mornings, corresponding to rush hour traffic at around 06:00. Concentrations tend to decrease during the middle of the day, with a broader evening rush-hour peak in NO₂ building up from early afternoon. At both sites the afternoon peak in NO₂ was of a higher order of magnitude compared to the morning peak and then stayed at elevated levels for much of the night. This is to be expected as concentrations of oxidising agents in the atmosphere (e.g. ozone) tend to increase in the afternoon, leading to enhanced oxidation of NO to NO₂.

PM_2.5 concentrations at both sites, shown in Figure 14 exhibit similar diurnal trends. Concentrations decrease throughout the morning until around 13:00, and then proceed to increase back to a similar concentration of that in the morning. This can be attributed to two factors, the first being emissions of primary particulate matter. The second relates to the emissions of sulphur dioxide and NO_x that can react with other chemicals in the atmosphere to form secondary sulphate and nitrate particles, resulting in elevated levels of PM₁₀ and PM_2.5.

Average PM₁₀ concentrations, show in Figure 15 exhibited different diurnal trends for Stansted 3 and Stansted 4. At Stansted 3 concentrations increase during the morning rush hour period and stay at these elevated levels throughout the day with peaks that coincide with increased local emissions from the car park next to the site. When considered that PM₁₀ has a higher settling velocity than PM_2.5, it helps to explain the temporal variation in concentrations. Whereas at Stansted 4 concentrations decrease until early afternoon followed by a quicker increase until it peaks at around 23:00.

NO₂

Figure 13: Temporal variation in NO₂ concentrations

PM_2.5

Figure 14: Temporal variation in PM₂₅ concentrations

PM₁₀

Figure 15: Temporal variation in PM₁₀ concentrations

4.8 Source investigation

In order to investigate the possible sources of air pollution that were monitored at Stansted airport, meteorological data were used to add a directional component to the air pollutant concentrations. Wind speed and direction data was gathered using data gathered from the National Oceanic and Atmospheric Administration (NOAA) meteorological database. The QA/QC procedures for checking of these data are not known.

Figure 16 shows the wind speed and direction data. The lengths of the “spokes” against the concentric circles indicate the percentage of time during the year that the wind was measured from each direction. The prevailing wind direction was 200 ? to 230 ?, which shows that the prevailing wind direction was clearly from the south west. Each “spoke” is divided into coloured sections representing wind speed intervals of 2 m s^-1 as shown by the scale bar in the plot, followed by a final interval of 8.4 m s^-1. The mean wind speed was 4.13 m s^-1. The maximum measured wind speed was 14.4 m s^-1. Some of the highest wind speeds occurred during February and November 2017.

Figure 16: Wind rose showing wind speed and direction from the on-field anemometer at Stansted Airport in 2017.

4.11 Relationship with airport activity

Figure 26 shows the monthly total aircraft movements and monthly passenger numbers at Stansted, from Jan 2005 onwards against monthly mean concentrations of NO₂ at Stansted 3 and Stansted 4. These are plotted on a normalised scale to illustrate the trends between these three variables.

As highlighted in previous reports in this series, there is a clear seasonal pattern in air traffic movements at Stansted: numbers are higher in the summer and lower in the winter. By contrast, concentrations of NO₂ tend to show the opposite pattern, being highest in January and November.

It is important to note that emissions from the airport are an important contributor to local concentrations of NO₂. However, this simplistic analysis illustrates how seasonal variation in ambient pollutant concentrations is influenced more by general factors (e.g. meteorological conditions).

4.12 Comparison with other UK sites

Figure 27 provides a comparison between annual mean pollutant NO₂ levels at the Stansted sites and corresponding measurements made at six other monitoring stations (2001 to 2017). Five of these are other AURN monitoring sites in the south and east of England and the sixth is in the vicinity of a major airport. These sites are listed below.

• Canterbury - an urban background site approximately 1.5 kilometres from the centre of Canterbury.
• Thurrock - an urban background site in the town of Thurrock, Essex, approximately 35 metres from the kerb of a busy road.
• Cambridge Roadside - roadside site in the city of Cambridge, where vehicle emissions are the major pollution source.
• Southend-on-Sea - an urban background site situated in an urban public park in a residential area.
• London Harlington - a background monitoring station approximately 1 km north east of the perimeter of Heathrow airport.
• LHR2 - a long-term airside monitoring station at Heathrow, 180 metres north of runway 27R and north east of the central terminal area. This site is not part of the AURN, but data are made available to the public through the Heathrow Airwatch website.

In recent years, annual mean concentrations of NO₂ at the Stansted sites have resembled urban background concentrations measured at similar sites. For example, the concentrations seen at Stansted 3 and Stansted 4 are comparable with those at Southend on-Sea and Thurrock, although they are slightly higher than those reported from Canterbury.

Both Stansted sites have consistently reported lower concentrations than those recorded at London Harlington, Heathrow LHR2, and Cambridge Roadside. Annual mean NO₂ concentrations at Stansted 3 have exhibited a slight general decrease from 2004-2014, followed by a slight increase in 2015 and a stronger decrase in 2016. In 2017 the annual mean concentration at Stansted 3 increased slightly in relation to 2016, and similarly at Stansted 4 in comparison to 2016.

Cambridge Roadside, located at the kerb of a busy road in the nearby city of Cambridge, is included as an example of a site showing constant high annual mean NO₂ concentrations. The site (like many other urban roadside sites in the UK) has consistently recorded annual mean NO₂ concentrations in excess of 29 μg m^-3, substantially higher than those observed at either of the Stansted sites.

The data collected in 2017 shows that there is generally similar concentrations of NO₂ in most of the analysed monitoring sites when compared to 2016 (? 2 μg m^-3). Only Cambridge Roadside seems to escape from this tendency, having registered a decrease of NO₂ concentrations in 2017.

Figure 28 shows annual mean PM₁₀ concentrations at Stansted 3, Stansted 4 and other local sites. Stansted 3 data is “as measured” without VCM correction for data until the end of 2016.

Concentrations of PM₁₀ at Stansted 3 have decreased in comparison to 2016, a similar trend exhibited at London North Kensington. Since 2004, Stansted 3 and LHR2 have shown a very similar pattern from one year to the next. Over the past four years this has been slightly different, with Heathrow having registered a decrease on PM₁₀ annual mean concentration, and Stansted showing a steady increase until 2016. 2017 data however, shows a strong decrease in annual PM₁₀ concentrations at Stansted 3 that reach that of LHR2 (15 μg m^-3). PM₁₀ annual mean concentration at Stansted 3 continue to present similar values since 2007. Concentrations of PM₁₀ at Stansted 4 in its first year of operation show a strong decreased comapred to that of Stansted 3 and LHR2 (12 μg m^-3).

Figure 29: Time series of annual mean PM_2.5 concentrations at nearby sites, 2001

Figure 29 shows annual mean PM_2.5 concentrations for 2017 at nearby sites. Both Stansted 3 and Stansted 4 measured 9 μg m^-3, along with LHR2. Other sites at Leamington Spa and London North Kensington measured considerably higher. As previously explained, PM_2.5 is a widely dispersed pollutant, this can therefore offer a possible explanation to the similar measurements seen at both Stansted sites and LHR2.