American Association for Aerosol Research - Abstract Submission

AAAR 35th Annual Conference
October 17 - October 21, 2016
Oregon Convention Center
Portland, Oregon, USA

Abstract View


Ambient and Laboratory Evaluation of a Low-Cost PM sensor

KERRY KELLY, Chris Widmer, Jonathan Whitaker, Anthony Butterfield, University of Utah

     Abstract Number: 75
     Working Group: Instrumentation and Methods

Abstract
Low-cost, light-scattering-based particulate matter (PM) sensors are becoming more widely available and are being deployed in ambient and indoor environments because they can provide improved spatial and temporal resolution PM measurements. In fact, some community groups are posting PM concentrations based on these measurements. Researchers have begun to evaluate some of these sensors under laboratory and environmental conditions. In this study, the PlantPower PMS 1003/3003 was evaluated in a controlled wind-tunnel environment with alumina oxide particles and in the ambient environment during a winter season that was punctuated by several cold-pool events with corresponding high ambient PM levels. Under the ambient conditions tested the PMS sensor correlated well with 24-hour federal reference measurements (RSQ > 0.88) and hourly federal equivalent methods (RSQ > 0.90); these correlations are generally higher than correlations reported by other ambient studies of low-cost sensors. In the wind tunnel, the PMS also correlated well with mass-adjusted DustTrack measurements (RSQ > 0.87). Although the correlations between the PMS sensor and the federal reference/equivalent methods and research-grade instruments was high, the PMS sensor overestimated PM2.5 concentrations under the winter-time ambient conditions tested and underestimated the PM2.5 concentrations in the wind-tunnel environment. This study provides a suggested model for the PMS sensor response, emphasizes the importance of particle composition and size on sensor response, and highlights some of the differences between laboratory and field performance of these low-cost sensors.