Abstract
The use of low-cost sensors for air quality measurements has become very popular in the last few decades. Due to the detrimental effects of particulate matter (PM) on human health, PM sensors like photometers and optical particle counters (OPCs) are widespread and have been widely investigated. The negative effects of high relative humidity (RH) and fog events in the mass concentration readings of these types of sensors are well documented. In the literature, different solutions to these problems – like correction models based on the Köhler theory or machine learning algorithms – have been applied. In this work, an air pre-conditioning method based on a low-cost thermal dryer for a low-cost OPC is presented. This study was done in two parts. The first part of the study was conducted in the laboratory to test the low-cost dryer under two different scenarios. In one scenario, the drying efficiency of the low-cost dryer was investigated in the presence of fog. In the second scenario, experiments with hygroscopic aerosols were done to determine to which extent the low-cost dryer reverts the growth of hygroscopic particles. In the second part of the study, the PM10 and PM2.5 mass concentrations of an OPC with dryer were compared with the gravimetric measurements and a continuous federal equivalent method (FEM) instrument in the field. The feasibility of using univariate linear regression (ULR) to correct the PM data of an OPC with dryer during field measurement was also evaluated. Finally, comparison measurements between an OPC with dryer, an OPC without dryer, and a FEM instrument during a real fog event are also presented. The laboratory results show that the sensor with the low-cost dryer at its inlet measured an average of 64 % and 59 % less PM2.5 concentration compared with a sensor without the low-cost dryer during the experiments with fog and with hygroscopic particles, respectively. The outcomes of the PM2.5 concentrations of the low-cost sensor with dryer in laboratory conditions reveal, however, an excess of heating compared with the FEM instrument. This excess of heating is also demonstrated in a more in-depth study on the temperature profile inside the dryer. The correction of the PM10 concentrations of the sensor with dryer during field measurements by using ULR showed a reduction of the maximum absolute error (MAE) from 4.3 µg m−3 (raw data) to 2.4 µg m−3 (after correction). The results for PM2.5 make evident an increase in the MAE after correction: from 1.9 µg m−3 in the raw data to 3.2 µg m−3. In light of these results, a low-cost thermal dryer could be a cost-effective add-on that could revert the effect of the hygroscopic growth and the fog in the PM readings. However, special care is needed when designing a low-cost dryer for a PM sensor to produce FEM similar PM readings, as high temperatures may irreversibly change the sampled air by evaporating the most volatile particulate species and thus deliver underestimated PM readings. New versions of a low-cost dryer aiming at FEM measurements should focus on maintaining the RH at the sensor inlet at 50 % and avoid reaching temperatures higher than 40 ∘C in the drying system. Finally, we believe that low-cost dryers have a very promising future for the application of sensors in citizen science, sensor networks for supplemental monitoring, and epidemiological studies.
Users
Please
log in to take part in the discussion (add own reviews or comments).
