10th International Aerosol Conference September 2 - September 7, 2018 America's Center Convention Complex St. Louis, Missouri, USA
Abstract View
Possibilities and Limitations of Low-cost PM Sensors
CHRISTOF ASBACH, Michael Spreitzer, Michael Bässler, Thorsten Schultze, Jörg Lindermann, Heinz Kaminski, Bryan Hellack, Ana Maria Todea, IUTA, Duisburg, Germany
Abstract Number: 739 Working Group: Low-Cost and Portable Sensors
Abstract The development of low-cost particulate matter (PM) sensors has gained rapid pace in the recent years. These sensors offer numerous new possibilities to both researchers as well as the general public through citizen science activities. As an example, the OK Lab initiative uses Nova Fitness SDS 011 low-cost sensors to measure the spatio-temporal distribution of PM10 and PM2.5. They started in Stuttgart, one of Germany’s most polluted cities, but meanwhile the network has extended all over Germany (www.luftdaten.info). The costs for such sensors range from a few Euros/Dollars for simple, photometric sensors to a few hundred Euros/Dollars for size resolving optical spectrometers. Besides their potential use for measuring ambient PM10 or PM2.5 concentrations or indoor or workplace air pollution, they can also be applied in a variety of laboratory based studies in the field of aerosol science. However, in all these applications, several limitations apply that need to be known in order to judge whether or not the use of a low-cost sensor or a sensor network is capable of delivering the desired results. For example, the particle shape and refractive index have to be known or assumed for their calibration. In case of photometric sensors, assumptions regarding the particle size distribution are additionally needed for calibration. If the actual aerosol properties differ from the ones used during calibration, the measurement results will be biased. Temperature and humidity were shown to affect the accuracy of both, photometric sensors and spectrometers.
In order to elucidate their possibilities and limitations, we studied in detail the response of three different sensor types, namely the Sharp Dust Sensor (photometric, approximately 7.50 €/9 US$), Nova Fitness SDS011 (photometric, approximately 15 €/18 US$) and Alphasense OPC-N2 (size resolving spectrometer, approximately 400 €/480 US$). Between three and ten specimens of each sensor type were tested. All sensors have undergone a thorough lab study using a variety of test aerosols, representing a wide range of typical size distributions and refractive indices. A limited number of the sensors were aged with diesel soot in an accelerated aging test, mimicking one year operation at the EU annual PM10 limit value of 40 µg/m³. The Sharp and SDS 011 were additionally tested in an ambient air quality measurement station in parallel with reference instrumentation.
It was shown that the low cost sensors can in principle deliver reproducible results, but the accuracy, especially of the photometric sensors, strongly depends on the correctness of the assumptions made during the calibration. Furthermore, measurement errors because of changing ambient conditions and aging of the sensors must not be ignored. While these sensors open a wide range of ambient, indoor and laboratory applications of these sensors, care must be taken if they are applied in areas with varying particle properties and/or changing temperature and humidity levels.
The measurements carried out in the laboratory and in the field will be presented and the results discussed in view of the possibilities and limitations of the low-cost sensors.
Acknowledgement The research presented here has received partial funding from the German Federal Institute of Occupational Safety and Health (BAuA) and the German Federal Environmental Protection Agency (UBA). The support is gratefully acknowledged.