Model predicted NO production from membrane-aerated biofilm reactor is greatly affected by biofilm property settings.

Even though modeling has been frequently used to understand the autotrophic deammonification-based membrane-aerated biofilm reactor (MABR), the relationships between system-specific biofilm property settings and model predicted NO production have yet to be clarified. To this end, this study investigated the impacts of 4 key biofilm property settings (i.e., biofilm thickness/compactness, boundary layer thickness, diffusivity of soluble components in the biofilm structure, and biofilm discretization) on one-dimensional modeling of the MABR, with the focus on its NO production. The results showed that biofilm thickness/compactness (200-1000 μm), diffusivity of soluble components in the biofilm structure (reduction factor of diffusivity: 0.2-0.9), and biofilm discretization (12-28 grid points) significantly influenced the simulated NO production, while boundary layer thickness (0-300 μm) only played a marginal role. In the studied ranges of biofilm property settings, distinct upper and lower bounds of NO production factor (i.e., the percentage ratio of NO formed to NH removed, 5.5% versus 2.3%) could be predicted. In addition to the microbial community structure, the NO production pathway contribution differentiation was also subject to changes in biofilm property settings. Therefore, biofilm properties need to be quantified experimentally or set properly to model NO production from the MABR correctly. As a good practice for one-dimensional modeling of NO production from biofilm-based reactors, especially the MABR performing autotrophic deammonification, the essential information about those influential biofilm property settings identified in this study should be disclosed and clearly documented, thus ensuring both the reproducibility of modeling results and the reliable applications of NO models.