10th International Aerosol Conference September 2 - September 7, 2018 America's Center Convention Complex St. Louis, Missouri, USA
Abstract View
Investigating the Performance of a Low-Cost PM Monitor (Dylos) against Dusttrak DRX for Different Indoor and Outdoor PM Sources
MEHDI AMOUEI TORKMAHALLEH, Obaidullah Mohiuddin, Fatemeh Mohammaddezashibi, Madina Obaidullah, Hamed Sharifi, Chemical and Aerosol Research Team, Nazarbayev University
Abstract Number: 1277 Working Group: Low-Cost and Portable Sensors
Abstract Recently, several low-cost, portable optical monitors including Dylos Air Quality Monitor became available in the market and need to be evaluated against the well-reference monitors for public use. United State Environmental Protection Agency (EPA) air sensor workshop recommended several criteria such as linearity of response, the precision of measurements, limit of detection, concentration resolution, response time, interference equivalents, and relative humidity (RH) and temperature influences for the investigation of new air quality sensors. Since the availability of low-cost monitors, researchers have investigated possible correlations with low-cost PM monitors and respective well referenced high-cost monitors. Dylos and Grimm EDM180 showed a medium correlation (R2 = 0.533). Another study was conducted for the correlation of Portable University of Washington Particle (PUWP) monitors with optical and gravimetric PM2.5 reference monitors. The pairwise correlation between them was around (R2- 0.86-0.89). When the test was conducted on eight different sites for measuring PM2.5, the correlation was found to be (R2 = 0.53).
In the present study two instruments were used to examine the air quality. A Dylos DC1700 (a low-cost equipment), was used to measure number concentrations of the particles. Dylos monitors the particle number concentrations in two sizes, larger than 0.5 µm (hereinafter called small size) and larger than 2.5 µm (hereinafter called larger size). The difference between the measurements by Dylos for the two sizes offers the particle number concentration for particle size from 0.5 to 2.5 µm. A DustTrak DRX Model, TSI, USA was employed to measure PM mass concentrations (PM1, PM2.5, PM4, PM10, Total particles). The two monitors were placed near the indoor or outdoor sources with at least one meter away from each other. Dusttrak was zero checked, before starting the experiments. The instruments were operated simultaneously at the same time. Two approaches were employed to compare the Dylos measurements and the Dusttrak data. The difference between Dylos large and small sizes (fine range) was compared with the Dusttrak PM2.5 measurements. Additionally, Dylos large data (coarse range) were compared with the Dusttrak Total PM2.5. Three PM sources including mixed grilling (ground beef meat) and construction particles (Number of measurements=188) generated at faculty residence of Nazarbayev University, outdoor PM (Number of measurements= 9746) in a rural area, Borovoye, Kazakhstan, and smoking particles at Middle East Technical University Northern Cyprus Campus (Number of measurements= 1624) were examined. The operation length were approximately 3hr, 8 days, and 26hr (4 days each 7 hours approximately), respectively. The data for three more sources including beauty shop, construction dust and grilling are being analyzed to be presented during the conference. For Mixed Grilling and Construction source, the correlation coefficient values for fine and coarse fractions were found to be 0.61 and 0.66, respectively, indicating a moderate correlation between Dylos and Dusttrak for this particular source. This moderate correlation could be due to the presence of significant amount of coarse size particles that can be detected by Dylos. A wide range of correlation coefficient values was obtained at four different smoking tests for coarse particles varying from 0.0061 to 0.569. Nevertheless, combining all smoking data, poor correlation values for fine (R2 = 0.17) and coarse (R2 = 0.20) particles were obtained that could be attributed to the presence of PM1 as the dominant size fraction which cannot be completely detected by Dylos. Poor correlation coefficient values were observed for both fine (4×10-5) and coarse (6.1×10-3) particles during the monitoring of outdoor PM in Borovoye, While we are continuing testing different sources until the IAC conference, our conclusion up to this point is that Dylos may show a poor correlation for sources generating particle with major size of PM1. However, it could show an acceptable correlation for sources producing mainly coarse particles.