Ozone removal air cleaning devices : test method and performance assessment

Applying ozone removal devices in ventilation systems is an effective way to reduce building occupant exposure to ozone. However, the test method on ozone removal devices has not been fully validated. Little is known about the performance of commercially-available ozone removal devices under realistic usage conditions. Additionally, the device performance in building ventilation systems can be affected by a wide range of environmental and operating conditions. This dissertation aims to 1) assess the test method of ozone removal devices and introduce modifications, 2) evaluate and compare a list of devices and identify important device characteristics, 3) quantify the impact of usage conditions on device performance, and 4) develop a mechanistic model for device performance prediction. A prolonged test protocol “Ozone Stress Test” and several metrics for evaluating device performance were proposed, and results of this new test protocol were compared to the existing ASHRAE Standard 145.2 method. Results show that activated carbon-based devices had a high degradation rate at the beginning of exposure to ozone or exposure to significantly increased ozone, so a short test (i.e., 1-h initial efficiency test and 4-h capacity test in Standard 145.2) overestimated device performance. A total of 14 ozone removal devices that utilize various technologies were tested by the ozone stress test. Results show a wide range of single-pass removal efficiency from 3% to 93% at 70 ppb, and all devices degraded at a slow rate at 500 ppb. The loading and source of carbon had a significant impact on the efficiency of activated carbon filters. The impact of temperature, relative humidity, ozone concentration, approach air speed, and intermittent exposure on the ozone removal performance of activated carbon filters and one metal oxide catalyst filter were systematically evaluated. Within the realistic range, temperature, relative humidity, and ozone concentration had very small effect on device performance, but an increase in air speed decreased efficiency significantly. Activated carbon filters regenerated after a break, while metal oxide did not. The dynamic and quasi-steady-state device performance was successfully explained by a plug flow reactor model.