Modeling denitrification byproducts during bioelectrochemical denitrification mediated by hydrogen oxidizing bacteria

Denitrifying bioelectrochemical systems (d-BES) represent a promising approach for nitrate (NO 3 – ) remediation from contaminated groundwater. In these systems, hydrogen (H 2 ), produced through water electrolysis at cathodic electrodes, serves as the electron donor for autotrophic denitrification. However, process efficiency is often limited by low temperatures and the formation of undesirable intermediates such as nitrite (NO 2 – ), nitrous oxide (N 2 O), or ammonium (NH 4 + ) by competing dissimilatory nitrate reduction to ammonium (DNRA) pathway. This study aims to develop and validate a mechanistic biofilm model that predicts nitrogen transformations and by-product formation during hydrogenotrophic denitrification at different temperatures. A one-dimensional biofilm model was formulated, explicitly coupling electrochemical H 2 generation with four-step hydrogenotrophic denitrification and DNRA processes. The model was calibrated and validated using batch data from d-BES operated at 20, 10, and 4°C. Global sensitivity analysis identified key parameters controlling NO 3 – removal and by-product formation, which differed with temperature. Cathodic H 2 efficiency and boundary-layer thickness were the most influential parameters. The model accurately reproduced experimental dynamics of NO 3 – and N 2 O across all temperatures, obtaining Theil’s Inequality Coefficients (TIC) below 0.3, while for NH 4 + a better fit (TIC = 0.12) was obtained at higher temperature (20°C). The model additionally revealed strong spatial segregation of microbial guilds, with denitrifiers dominating the outer biofilm layers, while DNRA bacteria prevailed near the electrode surface, where H 2 is produced. The proposed framework provides a predictive basis for optimizing d-BES operation and developing digital twins for NO 3 – removal under cold, real-field groundwater conditions.