AAAR 36th Annual Conference October 16 - October 20, 2017 Raleigh Convention Center Raleigh, North Carolina, USA
Abstract View
Comparison of Four Consumer-Grade Air Quality Monitors
RUIKANG HE, Sanjeevi Thirumurugesan, Daniel Bachman, Dominick J. Carluccio, Rudolph Jaeger, Clinton J. Andrews, Gediminas Mainelis, Rutgers, The State University of New Jersey
Abstract Number: 371 Working Group: Instrumentation and Methods
Abstract Consumer-grade air quality monitors have been on the US market since 2012 and can report a variety of metrics including humidity, temperature, atmospheric pressure, VOCs, carbon dioxide, and particulate matter levels. Many of them stream the data in real-time to smartphone apps or provide data to be downloaded via web. 3 Foobot, 3 AirVisual Node, 1 Particle Matter-Dust Sensor (Libelium Tec.) and 1 Fidas® Frog were involved in this study. The reference instruments were DustTrak DRX for particulate matter and IAQ Monitor (both TSI Inc.) for CO2, humidity, and temperature. Three measurements were done in different environments and with different particles: Polystyrene latex (PSL) beads with diameter of 0.72um and 2.00um; Arizona road dust; Nano silver particles; and dust from cooking event. Temperature, humidity and CO2 were not controlled. All devices were run side-by-side in each environment. The results have shown that for particulate matter measurements, the Pearson correlation coefficients between Libelium, Airvisual, Fidas and DRX were always higher than 0.85, while this value for Foobot dropped to 0.72. Since most correlations were statistically significant (P-value<0.01), it is reasonable to say that Foobot, AirVisual, Fidas and Libelium are linearly correlated with DRX. For temperature and humidity measurements, the Pearson correlation coefficients between Foobot, Airvisual and IAQ monitor were higher than 0.92 (P-value<0.01), and the R-squares were higher than 0.84 when doing linear regressions with IAQ monitor. For CO2 measurements, the Pearson correlation coefficient between Airvisual and IAQ monitor were higher than 0.99 (P-value<0.01), while Foobot had correlation coefficients lower than 0.56 or even not correlated with IAQ monitor. The data show that consumer-grade devices are reliable in some cases; however, testing in different conditions is needed before using them in field studies.