In-situ Effectiveness of Portable and Central Air Cleaners
RAFSAN NAHIAN, Jeffrey Siegel, University of Toronto
Abstract Number: 216
Working Group: Indoor Air Purification Technologies, Best Practices, and their Health Impacts
Abstract
Portable and central air cleaners are commonly tested with standardized laboratory protocols. However, the laboratory-tested performance values may not accurately reflect real-world performance since the effectiveness of air cleaners can be influenced by the variation of indoor and outdoor sources, room volume and mixing, and background loss rates of pollutants. This study evaluates the in-situ effectiveness of air cleaners to capture the actual performance and uses low-cost sensors. Each experiment was conducted between one to two weeks period alternating between placebo and air cleaner active operation and the effectiveness was calculated by comparing the PM2.5 concentrations during adjoining placebo and air cleaner operating conditions. The results indicate that the effectiveness of air cleaners is not static but varies considerably within the same environment due to the variation of PM sources and background loss rates. For example, the median PM2.5 effectiveness for three portable air cleaners varied between 0% to 94% in residential, classroom, and office spaces. The effectiveness for a central system in a residence with electric and conventional media filters ranged from 0% to 50%. Within this variation, the portable air cleaner with a higher clean air delivery rate (CADR) and central air cleaner with a high-efficiency filter consistently demonstrated higher effectiveness across all environments although the degree of improvement was not proportional to the increase of CADR or filter efficiency. While larger spaces generally showed less effectiveness, this was not always the case as the effectiveness was influenced by the ventilation and sinks and sources of pollutants in the specific indoor environment. These findings emphasize the importance of understanding how air cleaners perform in various real-world conditions for selection and deployment.