Abstract

In this study, the gas-phase hydrodynamics within a 6.5 L biotrickling filter were quantified using pulse-injection residence time distributions (RTDs) recorded with a cost-effective metal oxide (MOx) sensor. This was conducted at a constant gas flow rate of 8 L min-1 across trickling liquid velocities (TLV) ranging from 0 to 10 m h-1, using three different packings: polyurethane foam, Tri-Packs Jaeger spheres, and wood bark. A hybrid section-based model was developed and evaluated against two closed-closed axial dispersion variants (ADcc+CSTR and ADcc+PFR). Model discrimination using AIC/BIC and R2 indicated that the optimal model description is contingent upon the packing material and operating regime: the hybrid model most accurately represented foam and most wood cases, whereas ADcc+CSTR was sufficient for spheres under irrigation, and ADcc+PFR consistently underperformed. Employing the optimal model in each scenario, the Peclet number increased with TLV for PUF and spheres (with diminishing returns at the highest TLV), whereas wood exhibited a low Peclet number and non-monotonic behavior. By explicitly partitioning the mean residence time between the buffer zones and the bed, the hybrid model effectively separates the boundary hold-up from the core dispersion, yielding a more precise representation of the packed bed Peclet number. Overall, MOx-based RTD measurements, in conjunction with model comparison, offer a cost-effective in situ diagnostic tool for differentiating boundary mixing from bed transport, thereby informing packing selection and operating parameters in gas-phase biofiltration.