Volatile Organic Compounds (VOCs), which may cause short and long term adverse health effects, constitute a significant class of indoor gaseous pollutants. To reduce the VOC concentration of indoor environment, an in-depth understanding of the performance of the air cleaning technologies that address VOCs is important. Although many studies have been conducted to evaluate the performance of air cleaning devices, most of them were performed at elevated concentration level (~ppmv). The performance of the sorbent media at realistic VOC concentration (ppbv) is still not clear due to lack of experiment data and theoretical study. The test conducted under the low concentration is not cost-effective and difficult to perform. Currently, no mathematical model can be effectively used for sorbent media performance simulation at ppb level.

The main objective of the presented research is to investigate the adsorption mechanism, address the existing problems through a model-based testing and evaluation method, develop and validate reliable methodologies to predict the long-term performance of filter sorbent media, when exposed to the realistic indoor concentration of VOCs.

A series of long-term tests were conducted at six concentration levels from 100 ppm to 66 ppb, and different test conditions (particle size, flow rate and sorbent bed length) to reveal the different behavior of the filter media at different concentration levels. A new mechanistic model named convective & diffusion mass transfer model with variable partition coefficient (C&DMT-VP) was proposed to simulate the performance of adsorption-based air cleaning devices under the typical indoor VOC concentrations. The applications of this model were demonstrated, including the determination of the surface diffusion coefficient, prediction of sorbent media filter performance via C&DMT-VP at typical indoor concentration level based on the model parameters determined from different test methods, including the ASHRAE standard test 145.1, ground pellet test and thin layer with ground pellet test. Finally, the proposed methods were compared and validated with the experimental data.

It was found that 1) the partition coefficient varied with the concentration in the form of (Kma = aCb (or log(Kma) = A + B log(C); 2) The C&DMT-VP model incorporating the K(C) relationship significantly improved the representation of the performance at the low concentration as well as being able to represent the high concentration performance as in previous model; 3) The three accelerated methods were able to provide the data needed to determine the Kma( C) function for a given adsorption media's performance at low concentrations typically found indoors.