Laboratory Evaluation and Calibration of Three Low-Cost Particle Sensors for Particulate Matter Measurement

TLDR

Portable low‑cost particle sensors are attractive for measuring particulate matter, yet their data quality and calibration protocols remain poorly documented. The study evaluates three light‑scattering sensors (Shinyei PPD42NS, Samyoung DSM501A, Sharp GP2Y1010AU0F) using EPA‑adapted calibration methods. Calibration was performed against a SidePak, a scanning mobility particle sizer, and an AirAssure PM2.5 monitor, assessing linearity, precision, detection limits, composition and size dependence, and humidity/temperature effects. All three sensors exhibited strong linearity (R² > 0.89) over 0–1000 µg/m³, but precision varied (15–90 µg/m³), outputs differed up to tenfold with particle composition and size, and humidity also affected responses.

Abstract

Particle sensors offer significant advantages of compact size and low cost, and have recently drawn great attention for usage as portable monitors measuring particulate matter mass concentrations. However, most sensor systems have not been thoroughly evaluated with standardized calibration protocols, and their data quality is not well documented. In this work, three low-cost particle sensors based on light scattering (Shinyei PPD42NS, Samyoung DSM501A, and Sharp GP2Y1010AU0F) were evaluated by calibration methods adapted from the US EPA 2013 Air Sensor Workshop recommendations. With a SidePak (TSI Inc., St. Paul, MN, USA), a scanning mobility particle sizer (TSI Inc. ), and an AirAssure™ PM2.5 Indoor Air Quality Monitor (TSI Inc.), which itself relies on a GP2Y1010AU0F sensor as reference instruments, six performance aspects were examined: linearity of response, precision of measurement, limit of detection, dependence on particle composition, dependence on particle size, and relative humidity and temperature influences. This work found that: (a) all three sensors demonstrated high linearity against SidePak measured concentrations, with R2 values higher than 0.8914 in the particle concentration range of 0–1000 μg/m3, and the linearity depended on the studied range of particle concentrations; (b) the standard deviations of the sensors varied from 15 to 90 μg/m3 for a concentration range of 0–1000 μg/m3; (c) the outputs of all three sensors depended highly on particle composition and size, resulting in as high as 10 times difference in the sensor outputs; and (d) humidity affected the sensor response. This article provides further recommendations for applications of the three tested sensors.Copyright 2015 American Association for Aerosol Research