Effects of pH on Urine Nitrification: From Microbial Selection to Process Performance

Abstract

Separation of urine at the source allows for the efficient recovery of nitrogen, phosphorus and other nutrients and reduces the load of these nutrients on wastewater treatment plants. Before urine can be used as a fertilizer, a treatment for nitrogen stabilization is required. Nitrification is a well-suited process for urine stabilization and is also the targeted approach in bioregenerative life support systems for Space such as in the MELiSSA project. However, due to the limited alkalinity and the high residual ammonium concentration, urine nitrification without alkalinity addition is very pH sensitive, which can compromise stable urine nitrification. Usually, the pH is controlled with the influent. Nevertheless, process failures can occur, namely nitrite accumulation, complete cessation of nitrification, and the growth of acid-tolerant ammonia-oxidizing bacteria (AOB), which can further lower the pH during periods of limited urine and lead to the emission of harmful nitrogen oxide gases. A robust and reliable process is essential for decentralized treatment and especially for space applications. In addition, the carbon footprint must be minimized to ensure an environmentally sustainable process. The main objective of this thesis was to elucidate the various aspects of pH as the main control parameter for urine nitrification to provide a robust and environmentally sustainable process. Four pH-related aspects were investigated during this PhD: (i) the effect of operational pH on nitrite accumulation and microbial community composition, (ii) the kinetic and genomic characteristics of acid-tolerant AOB selected at low pH, (iii) the enhancement of microbial diversity and process robustness by short-term pH fluctuations, and (iv) the N2O emissions during stable urine nitrification and the operating conditions, especially in terms of pH, leading to a higher carbon footprint. To this end, several methodological approaches were used. Data from pilot-scale urine nitrification reactors were analyzed to determine the pH range for stable nitrification. Experiments with lab-scale reactors revealed the long-term effects of different pH set-points. The kinetic parameters for AOB and nitrite-oxidizing bacteria (NOB) were determined with respirometric activity tests. Microbial species were identified and quantified with 16S rRNA gene-based amplicon sequencing and metagenome sequencing. Further insights into the nitrification process were gained with dynamic modeling of the biological and chemical processes. Finally, continuous off-gas measurements allowed to determine the N2O emissions and, as a next step, calculate the carbon footprint of urine nitrification and fertilizer production. While biological ammonia oxidation of urine was possible over a pH range from 2.5 to 8.5, stable urine nitrification was mostly observed between 5.8 and 6.7. One of four distinct AOB species became dominant depending on the pH. During stable urine nitrification, AOB of the Nitrosomonas europaea lineage were found. The NOB probably belonged to the Nitrobacter genus but could not be unambiguously assigned. Higher pH set-points resulted in higher concentrations of free ammonia (NH3) due to the acid-base equilibrium and generally higher ammonia oxidation rates, making nitrite accumulation more likely. At pH 7 and 8.5, nitrite accumulated, resulting in partial nitritation and AOB, closely related to Nitrosomonas halophila and Nitrosomonas stercoris, respectively, became dominant. The NH3 turning point, i.e., the concentration at which a further increase in concentration leads to a decrease in AOB activity, was 12 mg N L 1, which was only reached at pH values above 7. NOB were more sensitive to NH3 and nitrous acid (HNO2) inhibition than the AOB. The total nitrite-nitrogen (TNN = NO2- N + HNO2-N) turning point for NOB was pH dependent, e.g. 10 mg N L 1 at a pH of 5.8 and 30 mg N L 1 at a pH of 6.7. Continuous nitrite measurement is recommended for stable nitrification to ensure that this TNN turning point is not exceeded to avoid positive feedback and nitrite accumulation. Acid-tolerant AOB proved to be a problem, as they consistently grew during periods without influent. To better understand acid-tolerant AOB, a novel AOB strain was enriched at pH 5, for which the name “Candidatus (Ca.) Nitrosacidococcus urinae” was proposed. “Ca. Nitrosacidococcus urinae” could oxidize ammonia despite pH values as low as 2.5, high HNO2 concentration of up to 15 mg-N L 1 and low NH3 concentration of 0.04 mg N L 1. However, ammonia oxidation under acidic conditions and high nitrogen levels resulted in nitrogen losses of about 10% due to chemical nitrite oxidation and was highly sensitive to process disturbances. Short periods of less than 12 hours without oxygen or influent resulted in a complete cessation of ammonia oxidation with a recovery time of up to two months. Nitrosacidococcus members were present only in small numbers in urine nitrification reactors operated at pH values above 5.8. It was shown that the activity of “Ca. Nitrosacidococcus urinae” decreased strongly at a pH of 7, which correlated with the limited availability of dissolved iron at higher pH and is possibly related to the absence of a siderophore system in “Ca. Nitrosacidococcus urinae”. Even within the limits for stable urine nitrification, nitrite accumulation can occur. To further increase the robustness of urine nitrification, two different strategies were tested: operating the two-position pH controller (influent on/off) with a narrow pH control band, which is commonly used for urine nitrification, at 6.20 and 6.25 (∆pH=0.05) and operating it with a wide pH control band at 6.0/6.5 (∆pH=0.5). The rationale was that the fluctuations would result in greater microbial diversity due to niche partitioning and, thus, a more stable process. While the diversity of the entire microbiota was similar in both reactors, the diversity of nitrifiers was higher in the wide-pH reactor. Nevertheless, the wide-pH reactor was slightly more affected by process disturbances resulting in nitrite accumulation. In addition, the N2O emissions from the wide-pH reactor were twice as high as those from the narrow-pH reactor, most likely due to nitrite fluctuations. Based on these results, a narrow control band is recommended for pH control in urine nitrification. From an overall performance point of view, the carbon footprint needs to be considered. During stable urine nitrification, 0.4% to 1.2% of the total nitrogen load was emitted as N2O with an average N2O emission factor (EFN2O) of 0.7%, which is in the same range as nitrification of municipal wastewater. Additional N2O was produced during anoxic storage between nitrification and granular activated carbon (GAC) filtration for micropollutant removal with an estimated EFN2O of 0.8%, resulting in an EFN2O of 1.5% for the ammonium nitrate fertilizer production from urine. N2O emissions during nitrification can be reduced by 60% by avoiding phases with no or low oxygen and by keeping the nitrite concentrations below 5 mg N L 1. Eliminating the storage tank between nitrification and GAC filtration can prevent N2O formation during intermediate storage. Overall, the N2O accounted for 45% of the operational carbon footprint of 14 kg-CO2,eq kg-N-1 for the urine fertilizer production, including nitrification, GAC filtration and distillation. Using electricity from renewable sources and applying the proposed N2O mitigation strategies would possibly reduce the carbon footprint by 85%. The studies presented in this thesis clearly showed that the bottleneck of urine nitrification is nitrite oxidation by NOB. In contrast, AOB are very versatile and different AOB species become dominant depending on the pH set-point. Future research projects should identify the most important NOB species and improve strategies to avoid or mitigate nitrite accumulation. In conclusion, the pH is a decisive process control parameter for urine nitrification by influencing the selection and kinetics of nitrifiers. Narrow pH control within the range of 5.8 to 6.7 and continuous nitrite measurement to ensure TNN concentrations below 5 mg-N L 1 enable efficient and stable production of an ammonium nitrate fertilizer from urine with a low carbon footprint for terrestrial and space applications.