Nitrous oxide (N2O) is not only a potent greenhouse gas with approximately 300 times global warming potential of carbon dioxide (CO2), but it is also a major sink for stratospheric ozone. Wastewater treatment systems are a recognized source of N2O. During biological wastewater treatment, N2O is mainly generated from biological nitrogen removal (BNR), which involves both nitrification and denitrification. Recently, ammonia-oxidizing bacteria (AOB) are identified as the major contributor to N2O production in wastewater treatment plants. However, the mechanisms of N2O production by AOB are still not fully understood. This thesis aims to experimentally assess and mathematically model the effect of several key operational parameters on N2O production by AOB as well as the contributions of different N2O production pathways to total N2O emission.
In order to achieve the aim, a nitrifying culture comprising primarily AOB and nitrite oxidizing bacteria (NOB) was enriched in a sequencing batch reactor (SBR). The reactor was fed with synthetic water containing 1 g N/L ammonium (NH4+) and performed complete nitrification with 100% conversion of NH4+ to nitrate (NO3-). This thesis firstly investigated the effect of dissolved oxygen (DO) on the N2O production by the enriched nitrifying sludge. On this occasion nitrite accumulation was minimised by augmenting nitrite (NO2-) oxidation through the addition of an enriched NOB sludge into the batch reactor. It was demonstrated that the specific N2O production rate (N2OR) increased from 0 to 1.9 ± 0.09 (n=3) mg N/hr/g VSS with an increase of DO concentration from 0 to 3.0 mg O2/L, whereas the N2O emission factor (the ratio between N2O nitrogen emitted and the ammonium nitrogen converted) decreased from 10.6 ± 1.7% (n=3) to 2.4 ± 0.1% (n=3). The site preference measurements, which is calculated as the difference between central and terminal nitrogen isotopomer signatures in N2O molecule, indicated that both the AOB denitrification and hydroxylamine (NH2OH) oxidation pathways contributed to N2O production, and DO had an important effect on the relative contributions of the two pathways. This finding is supported by analysis of the process data using an N2O model integrating both pathways. As DO increased from 0.2 to 3.0 mg O2/L, the contribution of AOB denitrification decreased from 92% − 95% to 66% − 73%, accompanied by a corresponding increase in the contribution by the NH2OH oxidation pathway.
The combined effect of NO2- and DO on N2O production by AOB was further studied using the enriched nitrifying culture. At each investigated DO level, both the biomass specific N2O production rate and the N2O emission factor increased as NO2- concentration increased from 3 mg N/L to 50 mg N/L. However, at each investigated NO2- level, the maximum biomass specific N2O production rate occurred at DO of 0.85 mg O2/L, while the N2O emission factor decreased as DO increased from 0.35 to 3.5 mg O2/L. The analysis of the process data using the two-pathway N2O model indicated that the contribution of AOB denitrification pathway increased as NO2- concentration increased, but decreased as DO concentration increased, accompanied by a corresponding change in the contribution of NH2OH oxidation pathway to N2O production. The AOB denitrification pathway was predominant in most cases, with the NH2OH oxidation pathway making a comparable contribution only at high DO level (e.g. 3.5 mg O2/L).
The effect of inorganic carbon (IC) on N2O production by AOB was also investigated using the enriched nitrifying culture over a concentration range of 0 – 12 mmol C/L, encompassing typical IC levels in wastewater treatment reactors. Batch experiments were conducted with continuous CO2 stripping or at controlled IC concentrations. The results revealed a linear relationship between N2OR and IC concentration (R2 = 0.97) within the IC range studied, suggesting a substantial effect of IC on N2O production by AOB. Similar results were also obtained with another AOB culture treating anaerobic sludge digestion liquor. The fundamental mechanism responsible for this dependency is unclear; however, in agreement with previous studies, it was observed that the ammonia oxidation rate (AOR) was also influenced by the IC concentration, which could be well described by the Monod kinetics. These resulted in an exponential relationship between N2OR and AOR, as previously observed in experiments where AOR was altered by varying dissolved oxygen and ammonia concentrations. It is therefore possible that IC indirectly affected N2OR by causing a change in AOR. The observation in this study indicates that alkalinity (mostly contributed by IC) could be a significant factor influencing N2O production and should be taken into consideration in estimating and mitigating N2O emissions in wastewater treatment systems.
None of the existing N2O models are able to capture N2O dynamics caused by the variation of IC concentration revealed in this study. A mathematical model that describes the effect of IC on N2O production by AOB is developed and experimentally validated. The IC effect is considered by explicitly including the AOB anabolic process in the model, which is coupled to the catabolic process with the use of the Adenosine triphosphate (ATP) and Adenosine diphosphate (ADP) pools. The calibration and validation of the model were conducted using experimental data obtained with two independent cultures, including a full nitrification culture and a partial nitritation culture. The model satisfactorily describes the N2O data from both systems at varying IC concentrations. This new model further enhances our ability to predict N2O production by AOB in wastewater treatment systems under varying IC conditions.
By conducting side-by-side comparison of the two single-pathway models with experimental data reported in literature, it has been demonstrated that none of the single-pathway models can reproduce all the N2O data, probably due to the fact that the two pathways are affected by operational conditions differently. The two-pathway N2O model help to understand the metabolic mechanism of N2O pathways and enhanced our ability to estimate N2O emission under various conditions, but the previously proposed single-pathway models have more simplified structure and fewer parameters to calibrate, which bring convenience to implication. Therefore, this thesis tests the predictive ability of two single-pathway models based on the AOB denitrification pathway and the NH2OH oxidation pathway, respectively, to describe the N2O data generated by the well- established N2O model that incorporates both pathways and provides theoretical guidance on how to use these two single-pathway models as well as the two-pathway model under various conditions. The modeling results suggested that (1) The model based on the AOB denitrification pathway should be used under the conditions with constant DO concentration and applied either at low DO concentration (< 0.5 mg O2/L) with any non-inhibitory NO2- concentration or at higher DO (≥ 0.5 mg O2/L) with relatively high NO2- but non-inhibitory concentration (≥ 1.0 mg N/L); (2) The model based on the NH2OH oxidation pathway can be applied under the conditions of relatively high DO concentration (≥ 1.5 mg O2/L) with any non-inhibitory NO2- concentration; (3) under other conditions, the two-pathway model should be applied.
The above listed findings experimentally reveal the effect of several key factors on N2O production by AOB and mathematically identified the shift of N2O production pathways under varying operational conditions, which would potentially be beneficial for the design and operation of full- scale wastewater treatment plant with the aim of N2O mitigation. However, advanced techniques like matatranscriptonmics are essential to provide solid evidences for the source identification of N2O production by AOB in further studies.