Putting our air quality sensors to the test

Scientists use a varied array of sensors to detect and measure UK air quality, looking at a wide range of pollutants. They need to have confidence in the data gathered with these sensors to make valid assessments of the hazards posed by air pollution. Metrology, the science of measurement, is an essential element in the process.

As part of the Clean Air Programme, a project led by the National Physical Laboratory has reviewed the latest air quality monitoring technologies in use today to ensure these sensors provide reliable, repeatable data.

Are low-cost sensors up to the job?

Researchers need a large array of relatively low-cost sensors to get a detailed picture of air pollution in a particular area. For nitrogen dioxide (NO₂), a major air pollutant produced from burning fossil fuels, diffusion tubes can be used to measure its atmospheric concentration. These small tubes allow air to diffuse through them, and any NO₂ reacts with a chemical sorbent supported on a mesh inside the closed end of the tube. Once the tube has been open for a set period, typically four weeks, its contents can be analysed and the concentration of NO₂ in the air calculated.

Although they are easy to use, these sensors can encounter problems, due to wind or other environmental factors. This can affect diffusion of gases into them, adversely impacting their measurements. To investigate this issue, the team carried out field tests on Marylebone Road in central London. They found that placing the tubes in protective shelters or altering their design to make them more wind-proof significantly improved the reliability of the readings. The length of time they are used can also be refined, given better knowledge of how long they last until they are saturated. These findings have been used in the UK Urban NO₂ Network, operated by Defra to monitor air quality at around 300 sites nationwide.

Improving methods and specifications

The team also looked at assessing the reliability of existing air quality data. Focussing on the Breathe London community sensing network, they developed a new methodology for auditing quality assurance and control. This methodology highlighted a strong adherence to proper data processing and management across the network. It can also be used to ensure the quality of future air quality network projects. In addition, the team are working on network-level statistical tools that enable insights such as individual sensor uncertainty assessments and analyses of long-term drifts in readings.

Standardisation is also a key part of developing a reliable and useable network of air quality sensors. To foster this standardisation, the team has developed a new European Technical Specification for the performance evaluation of air quality sensors (CEN TC264 WG42). This specification document presents an internationally agreed basis for performance requirements and test protocols for gaseous air quality sensors, broken down into defined classes of measurement applications. A standard procedure like this allows sensor developers and manufacturers to work to common standards, as well as defining how new sensors should be tested to demonstrate improvements over older models. Users of air quality data can also benefit from understanding the permitted tolerances and uncertainty associated with the particular class of measurement equipment they are using. As a follow up, Defra commissioned NPL to develop British Standard PAS 4023, which is a UK guide to using low-cost sensors for air quality measurements.

This project has already had a direct effect on improving air quality measurement in the UK. SPF Clean Air project lead Tom Gardiner of the National Physical Laboratory said, “This work has enhanced the air quality community’s understanding of monitoring uncertainties, and how to better account for them, while laying the groundwork for this monitoring to be an integrated part of the Clean Air framework. The work provides both suppliers and users of commercial products with a basis for assessing their readiness for deployment, ultimately helping improve our air quality monitoring capabilities and our understanding of the hazards posed by air pollution.”