Abstract:
Indoor environments contain some of the largest sources of human exposure to fine particulate matter (PM2.5). Emissions of PM2.5 can vary greatly from source to source, among different types of environments, and even among the same type of environment, such as residences. Research- and regulatory-grade instrumentation are cost prohibitive for monitoring such variability. Low-cost monitors may offer a solution, though they are prone to some well-documented biases and limitations. We assessed the accuracy of three consumer-grade (Air Quality Egg, PurpleAir, and Speck) and one research-grade (WashU) low-cost PM2.5 monitors during a controlled field study in a test house. We used experimental data from a variety of indoor meals and sources (background, window opening, stir-fry, breakfast, toast, and chili cooking) to evaluate the monitors’ time response and accuracy. Overall, the Air Quality Egg had the best performance during collocation periods, WashU had the best time response, and Speck had the best accuracy when compared to the reference monitors, both by percent accuracy and integrated concentration. All Plantower-based monitors were prone to overestimating concentrations, especially during cooking events, but they also better tracked the peaks as they occurred. Additionally, there were differences between different monitors that contained Plantowers, suggesting that casing, or microprocessors may also affect particulate matter concentration reporting in these devices. Thus, the Plantower-based monitors proved to be more precise, where the Speck was determined to be more accurate. These findings may help consumers and researchers determine how to assess the performance of low-cost air quality monitors.
Implication Statement This work compared three consumer-grade and one lab-developed fine particulate matter monitors to research-grade instrumentation in a “pseudo-field” study. Through controlled perturbations within a test-house, we were able to determine how these low-cost monitors performed in comparison to reference instruments and point out the biases these monitors have. Low-cost monitors have exploded in popularity over the last decade, but consumers are not aware of their biases. We show that the most popular model of sensor (Plantowers) often overestimates cooking particulate matter emissions in comparison to research-grade instruments. This may lead to a misunderstanding of exposure risk for consumers who are not aware of these limitations. These results are also necessary to understand how to use these monitors from a research standpoint, and the limitations to accurate readings one might expect.