Repeatability and Intermediate Precisions of the CPMA-Electrometer System

Abstract

The centrifugal particle mass analyzer (CPMA)-electrometer system has been proposed as an alternative method of calibration for devices that measure black carbon. This system is proposed to replace the current calibration standard of NIOSH 5040. The repeatability and intermediate precision of the CPMA-electrometer was determined through a series of experiments that examined an individual components of the system to determine the overall expected level of variation of the system. The key measurement components of this measurement system include the CPMA, a Faraday cup, an electrometer and a flow controller. These devices work together to classify a charged aerosol for a specific mass to charge ratio, capture the charge, and measure the current flow. Two devices were compared to a known standard, the electrometer and the flow controller. Though these tests it was found the flow controller had an accuracy of less than 1%, a repeatability of less than 0.4% and an M=2 intermediate precision of 0.3% while the electrometer has a bias error of 2%. The Faraday cups were testing using the two Faraday cup-electrometer systems in parallel. Measurements were collected on both devices simultaneously where the repeatability was found to be less than 0.8% for the majority of tests, and an M=2intermediate precision of 2.2%. The CPMA was tested using a tandem CPMA-CPMA experiment where one device classified the particles to a known set point and the second device scanned across a range of values to determine the peak value. Data was normalized by the initial set point on the first device. Through this it was found the CPMA had an average repeatability of 0.8% while the M=2 intermediate precision was 1.3%. Using these intermediate precisions an estimate of the overall systems intermediate precision was calculated as 2.6%. Tests were also conducted on the CPMA-electrometer system where comparison measurements were collected on a challenge instrument. Between measurements that were collected across a number of days it was found a repeat calibration tests were likely to have less than 5% variation, a great improvement over NIOSH 5040 and a comparable result to the calculated intermediate precision.