Monitoring and modeling of nitrogen conversions in membrane-aerated biofilm reactors: Effects of intermittent aeration

Nitrogen can be removed from sewage by a variety of physicochemical and biological processes. Due to the high removal efficiency and relatively low costs, biological processes have been widely adopted for treating nitrogen-rich wastewaters. Among the biological technologies, biofilm processes show great advantages as compared to suspended growth processes, allowing for biomass accumulation and retention without the need of external solid separa-tion devices.

The decoupling of solids retention from hydraulic retention is especially useful for slow-growing microorganisms, such as nitrifying bacte-ria, e.g. ammonium-oxidizing bacteria (AOB) and nitrite-oxidizing bacteria (NOB), and anaerobic ammonium-oxidizing bacteria (AnAOB), which are involved in ammonium (NH4+) removal process.

Stability of engineered biological processes requires an appropriate balance between activities of the main microbial groups involved in the system. How-ever, finding proper operational conditions is especially challenging in bio-films. On the one hand, the existence of strong spatial chemical gradients within biofilms increases the difficulty to prescribe environmental conditions that favor any desired biological process.

On the other hand, the presence of multiple simultaneous chemical gradients complicates the performance opti-mization. Mathematical modeling offers a way to describe and analyze multi-ple processes that occur simultaneously in time and space in biofilm systems. This PhD project investigated NH4+ removal process in membrane-aerated biofilm reactors (MABRs), focusing on aeration control, especially the appli-cation of intermittent aeration.

Compared to conventional biofilms which are characterized by co-diffusion, MABRs display counter-diffusion fluxes of substrates: oxygen is supplied through the membrane, whilst NH4+ is provid-ed from the bulk liquid phase. The counter substrate supply not only offers flexible aeration control, but also supports the development of a unique mi-crobial community and spatial structure inside the biofilm.

In this study, lab-scale MABRs were operated under two types of aeration control: continuous versus intermittent aeration. Long-term reactor performance was monitored. Based on bulk measurements of NH4+, nitrite (NO2-) and nitrate (NO3-), mi-crobial activities of individual functional guilds were evaluated.

I found that NOB suppression occurred under intermittent aeration, but not under contin-uous aeration. Relative aeration duration and aeration intermittency were two effective operational factors in regulating MABR performance under inter-mittent aeration. Besides daily bulk monitoring, in situ microprofiles of dis-solved oxygen (DO), pH and nitrous oxide (N2O) were performed.

The sig-nificant temporal fluctuations in local biofilm pH (not DO) during aeration control suggested that pH-related effects drive the changing microbial activi-ties under intermittent aeration, as compared to continuous aeration. Total N2O emissions were dramatically reduced at the onset of intermittent aera-tion, due to the development of an anoxic N2O reduction zone by hetero-trophic bacteria (HB).

To further investigate the causal link between NOB suppression and aeration regime change, a 1-dimensional (1-D) multispecies nitrifying biofilm model was developed in Aquasim software, incorporating a pH calculation. Kinetic parameters to be estimated were chosen based on a local sensitivity analysis, and were estimated from in situ microprofiles.

With the calibrated model, I identified that the periodically varying free ammonia inhibition, which was associated with transient pH variations, was the likely key factor causing NOB suppression in intermittently-aerated nitrifying MABRs. To further investigate the mechanisms of N2O mitigation under aeration con-trol, the 1-D biofilm model was extended to a partial nitritation-anammox (PNA) biofilm model, including description of all relevant biological N2O production pathways.

Sensitive kinetic parameters were estimated with long-term bulk performance data. With the calibrated model, roles of HB and AnAOB were discussed and evaluated in mitigating N2O emissions in auto-trophic nitrogen removal MABRs. Moreover, I developed a 1-D biofilm mod-el in Matlab software describing the counter-diffusion PNA process, aiming at an improved model calibration/evaluation for the highly variable N2O emissions.

Overall, a combination of experimental and modeling efforts were imple-mented to study nitrogen conversions in MABRs. The results showed that intermittent aeration was an efficient strategy to regulate microbial activities in counter-diffusion biofilms, achieving an energy-efficient NH4+ removal process with low N2O emissions.