Modeling the removal of VOC mixtures in biotrickling filters.

A mathematical model was derived for describing removal of mixed VOC vapors in biotrickling filters (BTFs). The model accounts for potential process rate limitation by the availability of oxygen as well as for potential kinetic interactions among pollutants during their biodegradation. Without using any fitted parameter, the model was found capable of predicting experimentally obtained removal rates of mono-chlorobenzene (m-CB) and ortho-dichlorobenzene (o-DCB) vapors. Experimental results reported here show that m-CB removal is better than that of o-DCB. The two compounds were known to be involved in a kinetic cross-inhibition interaction when degraded in suspended culture. However, model sensitivity studies showed that cross-inhibition does not affect BTF performance due to the low pollutant concentrations involved. For the same reason, the influence of oxygen on BTF performance was found to be minimal under the conditions tested. The model was found to predict experimentally obtained values with less than 10% error in the majority of cases. Computations with an earlier model describing VOC removal in conventional biofilters showed that, for the model mixture used in this study (m-CB/o-DCB), removal rates obtained with BTFs are one to more than two orders of magnitude higher than those obtained with conventional biofilters. This is attributed to the larger active specific biofilm surface area in BTFs, obtained through the creation of favorable growth conditions for the biomass, and better moisture control.