Comparing methanol and glycerol as carbon sources for mainstream partial denitrification/anammox in an IFAS process.

This study explored the implementation of mainstream partial denitrification with anammox (PdNA) in the second anoxic zone of a wastewater treatment process in an integrated fixed film activated sludge (IFAS) configuration. A pilot study was conducted to compare the use of methanol and glycerol as external carbon sources for an IFAS PdNA startup, with a goal to optimize nitrogen removal while minimizing carbon usage. The study also investigated the establishment of anammox bacteria on virgin carriers in IFAS reactors without the use of seeding, and it is the first IFAS PdNA startup to use methanol as an external carbon source. The establishment of anammox bacteria was confirmed in both reactors 102 days after startup. Although the glycerol-fed reactor achieved a higher steady-state maximum ammonia removal rate because of anammox bacteria (1.6 ± 0.3 g/m2/day) in comparison with the methanol-fed reactor (1.2 ± 0.2 g/m2/day), both the glycerol- and methanol-fed reactors achieved similar average in situ ammonia removal rates of 0.39 ± 0.2 g/m2/day and 0.40 ± 0.2 g/m2/day, respectively. Additionally, when the upstream ammonia versus NOx (AvN) control system maintained an ideal ratio of 0.40-0.50 g/g, the methanol-fed reactor attained a lower average effluent TIN concentration (3.50 ± 1.2 mg/L) than the glycerol-fed reactor (4.43 ± 1.6 mg/L), which was prone to elevated nitrite concentrations in the effluent. Overall, this research highlights the potential for PdNA in IFAS configurations as an efficient and cost-saving method for wastewater treatment, with methanol as a viable carbon source for the establishment of anammox bacteria. PRACTITIONER POINTS: Methanol is an effective external carbon source for an anammox startup that avoids the need for costly alternative carbon sources. The methanol-fed reactor demonstrated higher TIN removal compared with the glycerol-fed reactor because of less overproduction of nitrite. Anammox bacteria was established in an IFAS reactor without seeding and used internally stored carbon to reduce external carbon addition. Controlling the influent ammonia versus NOx (AvN) ratio between 0.40 and 0.50 g/g allowed for low and stable TIN effluent conditions.

Case study: Bioaugmenting the comammox dominated biomass from B-stage to enhance nitrification in A-stage at Blue Plains AWWTP.

A comprehensive case study was undertaken at the Blue Plains wastewater treatment plant (WWTP) to explore the bioaugmentation technique of introducing nitrifying sludge into the non-nitrifying stage over the course of two operational years. This innovative approach involved the return of waste activated sludge (WAS) from the biological nutrient removal (BNR) system to enhance the nitrification in the high carbon removal rate system. The complete ammonia oxidizer (comammox) Nitrospira Nitrosa was identified as the main nitrifier in the system. Bioaugmentation was shown to be successful as nitrifiers returned from BNR were able to increase the nitrifying activity of the high carbon removal rate system. There was a positive correlation between returned sludge from the BNR stage and the specific total kjeldahl nitrogen (TKN) removal rate in A stage. The bioaugmentation process resulted in a remarkable threefold increase in the specific TKN removal rate within the A stage. Result suggested that recycling of WAS is a simple technique to bio-augment a low SRT system with nitrifiers and add ammonia oxidation to a previously non-nitrifying stage. The results from this case study hold the potential for applicable implications for other WWTPs that have a similar operational scheme to Blue Plains, allowing them to reuse WAS from the B stage, previously considered waste, to enhance nitrification and thus improving overall nitrogen removal performance. PRACTITIONER POINTS: Comammox identifying as main nitrifier in the B stage. Comammox enriched sludge from B stage successfully bio-augmented the East side of A stage up to threefold. Bioaugmentation of comammox in the West side of A stage was potentially inhibited by the gravity thickened overflow. Sludge returned from B stage to A stage can improve nitrification with a very minor retrofits and short startup times.

Impact of instrumentation reliability on mainstream suspended partial denitrification-anammox (PdNA).

This study successfully revealed the importance of probe reliability and sensitivity with ion sensitive electrode (ISE) probes on achieving high partial denitrification (PdN) efficiency; and decreasing carbon overdosing events that cause the decline of microbial populations and performance of PdNA. In a mainstream integrated hybrid granule-floc system, an average PdN efficiency of 76% was achieved with acetate as the carbon source. Thauera was identified as the dominant PdN species; its presence in the system was analogous to instrumentation reliability and PdN selection and was not a consequence of bioaugmentation. Up to 27-121 mg total inorganic nitrogen/L/d, an equivalent of 18-48% of the overall total inorganic nitrogen removed, was achieved through the PdNA pathway. Candidatus Brocadia was the main anoxic ammonium oxidizing bacteria species that was seeded from sidestream and enriched and retained in the mainstream system with observed growth rates of 0.04-0.13 day-1 . Moreover, there was no direct negative impact of methanol's use for post-polishing on anoxic ammonium oxidizing bacteria activity and growth. PRACTITIONER POINTS: Stress testing with ISE sensors revealed the importance of probe reliability and sensitivity on PdN selection and PdNA performance. Up to 121 mg TIN/L/d was achieved via PdNA in a mainstream suspended hybrid granule-floc partial denitrification-anammox (PdNA) system. Candidatus Brocadia was the dominant AnAOB species with observed growth rates of 0.04-0.13 day-1. There was no direct negative impact of methanol's use for post-polishing on AnAOB activity and growth.

Feasibility of packed-bed trickling filters for partial nitritation/anammox: Effects of carrier material, bottom ventilation openings, hydraulic loading rate and free ammonia.

This study pioneers the feasibility of cost-effective partial nitritation/anammox (PN/A) in packed-bed trickling filters (TFs). Three parallel TFs tested different carrier materials, the presence or absence of bottom ventilation openings, hydraulic loading rates (HLR, 0.4-2.2 m3 m-2 h-1), and free ammonia (FA) levels on synthetic medium. The inexpensive Argex expanded clay was recommended due to the similar nitrogen removal rates as commercially used plastics. Top-only ventilation at an optimum HLR of 1.8 m3 m-2 h-1 could remove approximately 60% of the total nitrogen load (i.e., 300 mg N L-1 d-1, 30 °C) and achieve relatively low NO3--N accumulation (13%). Likely FA levels of around 1.3-3.2 mg N L-1 suppressed nitratation. Most of the total nitrogen removal took place in the upper third of the reactor, where anammox activity was highest. Provided further optimizations, the results demonstrated TFs are suitable for low-energy shortcut nitrogen removal.

Low diversity and microdiversity of comammox bacteria in wastewater systems suggest specific adaptations within the Ca. Nitrospira nitrosa cluster.

Studies have found Ca. Nitrospira nitrosa-like bacteria to be the principal or sole comammox bacteria in nitrogen removal systems for wastewater treatment. In contrast, multiple populations of strict ammonia and nitrite oxidizers co-exist in similar systems. This apparent lack of diversity is surprising and could impact the feasibility of leveraging comammox bacteria for nitrogen removal. We used full-length 16S rRNA gene sequencing and genome-resolved metagenomics to compare the species-level diversity of comammox bacteria with that of strict nitrifiers in full-scale wastewater treatment systems and assess whether this comparison is consistent or diverged at the strain-level. Full-length 16S rRNA gene sequencing indicated that Nitrosomonas-like bacteria exhibited higher species-level diversity in comparison with other nitrifying bacteria, while the strain-level diversity (also called microdiversity) of most Nitrospira-like bacteria were higher than Nitrosomonas-like bacteria with few exceptions (one Nitrospira lineage II population). Comammox bacterial metagenome assembled genomes (MAGs) were associated with Ca. Nitrospira nitrosa. The average amino acid identity between principal comammox bacterial MAGs (93% ± 3) across systems was significantly higher than that of the Nitrosomonas-like ammonia oxidizers (73% ± 8), the Nitrospira_A-like nitrite oxidizer (85% ± 4), and the Nitrospira_D-like nitrite oxidizer (83% ± 1). This demonstrated the low species-level diversity of comammox bacteria compared with strict nitrifiers and further suggests that the same comammox population was detected in all systems. Comammox bacteria (Nitrospira lineage II), Nitrosomonas and, Nitrospira_D (Nitrospira lineage II) MAGs were significantly less microdiverse than the Nitrospira_A (lineage I) MAGs. Interestingly, strain-resolved analysis also indicates that different nitrogen removal systems harbor different comammox bacterial strains within the Ca. Nitrospira nitrosa cluster. These results suggest that comammox bacteria associated with Ca. Nitrospira nitrosa have low species- and strain-level diversity in nitrogen removal systems and may thus harbor specific adaptations to the wastewater ecosystem.

Response of mixed community anammox biomass against sulfide, nitrite and recalcitrant carbon in terms of inhibition coefficients and functional gene expressions.

Anaerobic ammonium oxidation (anammox) has evolved as a carbon and energy-efficient nitrogen management bioprocess. However, factors such as inhibitory chemicals still challenge the easy operation of this powerful bioprocess. This research systematically evaluated the inhibition kinetics of sulfide, nitrite, and recalcitrant carbon under a genomic framework. The inhibition at the substrate and genetic levels of sulfide, nitrite and recalcitrant carbon on anammox activity was studied using batch tests. Nitrite inhibition of anammox followed substrate inhibition and was best described by the Aiba model with an inhibition coefficient [Formula: see text] of 324.04 mg N/L. Hydrazine synthase (hzsB) gene (anammox biomarker) expression was increased over time when incubated with nitrite up to 400 mg N/L. However, despite having the highest specific nitrite removal (SNR), the expression of hzsB at 100 and 200 mg N/L of nitrite was more muted than in most other samples with lower SNRs. Sulfide severely inhibited anammox activities. The inhibition was fitted with a Monod-based model with a [Formula: see text] of 4.39 mg S/L. At a sulfide concentration of 5 mg/L, the hzsB expression decreased throughout the experiment from its original value at he beginning. Recalcitrant carbon of filtrate from thermal hydrolysis process pretreated anaerobic digester had a minimal effect on maximum specific anammox activity (MSAA), and thus the value of the inhibition coefficient could not be calculated. At the same time, its hzsB expression profile was similar to that in the control. Resiliency and recovery tests indicated that the inhibition of nitrite (up to 400 mg N/L) and recalcitrant carbon (in 100% filtrate) were reversible. About 32% of MSAA was recovered after repeated exposures to sulfide at 2.5 mg/L, while at 5 mg/L, the inhibition was irreversible. Findings from this study will be helpful for the successful design and implementation of anammox in full-scale applications.

Enhancing bioflocculation in high-rate activated sludge improves effluent quality yet increases sensitivity to surface overflow rate.

High-rate activated sludge (HRAS) relies on good bioflocculation and subsequent solid-liquid separation to maximize the capture of organics. However, full-scale applications often suffer from poor and unpredictable effluent suspended solids (ESS). While the biological aspects of bioflocculation are thoroughly investigated, the effects of fines (settling velocity < 0.6 m3/m2/h), shear and surface overflow rate (SOR) are unclear. This work tackled the impact of fines, shear, and SOR on the ESS in absence of settleable influent solids. This was assessed on a full-scale HRAS step-feed (SF) and pilot-scale HRAS contact-stabilization (CS) configuration using batch settling tests, controlled clarifier experiments, and continuous operation of reactors. Fines contributed up to 25% of the ESS in the full-scale SF configuration. ESS decreased up to 30 mg TSS/L when bioflocculation was enhanced with the CS configuration. The feast-famine regime applied in CS promoted the production of high-quality extracellular polymeric substances (EPS). However, this resulted in a narrow and unfavorable settling velocity distribution, with 50% ± 5% of the sludge mass settling between 0.6 and 1.5 m3/m2/h, thus increasing sensitivity towards SOR changes. A low shear environment (20 s-1) before the clarifier for at least one min was enough to ensure the best possible settling velocity distribution, regardless of prior shear conditions. Overall, this paper provides a more complete view on the drivers of ESS in HRAS systems, creating the foundation for the design of effective HRAS clarifiers. Tangible recommendations are given on how to manage fines and establish the optimal settling velocity of the sludge.

Introducing bioflocculation boundaries in process control to enhance effluent quality of high-rate contact-stabilization systems.

High-rate activated sludge (HRAS) systems suffer from high variability of effluent quality, clarifier performance, and carbon capture. This study proposed a novel control approach using bioflocculation boundaries for wasting control strategy to enhance effluent quality and stability while still meeting carbon capture goals. The bioflocculation boundaries were developed based on the oxygen uptake rate (OUR) ratio between contactor and stabilizer (feast/famine) in a high-rate contact stabilization (CS) system and this OUR ratio was used to manipulate the wasting setpoint. Increased oxidation of carbon or decreased wasting was applied when OUR ratio was <0.52 or >0.95 to overcome bioflocculation limitation and maintain effluent quality. When no bioflocculation limitations (OUR ratio within 0.52-0.95) were detected, carbon capture was maximized. The proposed control concept was shown for a fully automated OUR-based control system as well as for a simplified version based on direct waste flow control. For both cases, significant improvements in effluent suspended solids level and stability (<50-mg TSS/L), solids capture over the clarifier (>90%), and COD capture (median of 32%) were achieved. This study shows how one can overcome the process instability of current HRAS systems and provide a path to achieve more reliable outcomes. PRACTITIONER POINTS: Online bioflocculation boundaries (upper and lower limit) were defined by the OUR ratio between contactor and stabilizer (feast/famine). To maintain effluent quality, carbon oxidation was minimized when bioflocculation was not limited (0.52-0.95 OUR ratio) and increased otherwise. A fully automated control concept was piloted, also a more simplified semiautomated option was proposed. Wasting control strategies with bioflocculation boundaries improved effluent quality while meeting carbon capture goals. Bioflocculation boundaries are easily applied to current wasting control schemes applied to HRAS systems (i.e., MLSS, SRT, and OUR controls).

Nitrogen removal capacity and carbon demand requirements of partial denitrification/anammox MBBR and IFAS processes.

A pilot study was conducted to investigate the carbon demand requirements and nitrogen removal capabilities of two mainstream partial denitrification/anammox (PdNA) processes: a two-zone, moving bed biofilm reactor (MBBR) process and an integrated fixed-film activated sludge (IFAS) process. The first MBBR zone conducted PdNA, while the second operated as an anammox zone. Operation of the IFAS process was conducted in two phases. The first phase of the operation involved minor external carbon addition, while the second phase of the operation involved controlled external carbon addition. The MBBR process produced an average effluent TIN concentration and chemical oxygen demand (COD)/TIN ratio of 2.81 ± 1.21 mg/L and 2.42 ± 0.77 g/g. The average effluent TIN concentrations and COD/TIN ratios for the IFAS process were 4.07 ± 1.66 mg/L and 1.08 ± 0.38 g/g during phase 1 and 3.30 ± 0.96 mg/L and 2.18 ± 0.99 g/g during phase 2. Despite having relatively low and unstable partial denitrification (PdN) efficiencies, both mainstream PdNA processes exhibited low effluent TIN concentrations and carbon requirements compared to nitrification/denitrification. Successful operation of the PdNA IFAS process indicates that mainstream PdNA can be implemented with minimal capital costs in a water resource recovery facility's second anoxic zone. PRACTITIONER POINTS: Low effluent TIN concentrations can be maintained in mainstream PdNA MBBR and IFAS processes with low external carbon demand. MBBR and IFAS PdNA processes do not require consistent or high PdN efficiencies to maintain low effluent TIN concentrations. IFAS and MBBR PdNA processes exhibit similar TIN and NH3 removal efficiencies. PdNA can be implemented in a second anoxic zone, using IFAS technology for anammox retention, with minimal capital costs.

Startup strategies for mainstream anammox polishing in moving bed biofilm reactors.

This study evaluated startup strategies for mainstream polishing anammox moving bed biofilm reactors (MBBRs) without anammox bacterial (AMX) biomass inoculation. Two types of startups were tested: anammox only (no external carbon addition) and partial denitrification/anammox (PdNA) with glycerol addition. Reactors were started with either virgin carriers or carriers with a preliminary biofilm from a mainstream aerobic integrated fixed-film activated sludge (IFAS) process. Three pilot-scale startups were completed under the following conditions: anammox-only with preliminary biofilm carriers, PdNA with preliminary biofilm carriers, and PdNA with virgin carriers. AMX presence was confirmed via quantitative polymerase chain reaction (qPCR) after 57, 57, and 77 days, respectively. Prior to AMX detection, average influent concentrations of ammonia and nitrite ranged from 1.7-2.7 mg/L and 0.98-1.8 mg/L, respectively. This study demonstrated that AMX can be grown on carriers without AMX seeding under mainstream conditions (temperature 17-29°C, low ammonia and nitrite), regardless of whether nitrite came from upstream or partial denitrification within the reactor. This study also showed that using preliminary biofilm carriers can decrease startup time by approximately 1 month. These results address critical questions for moving mainstream anammox processes to full-scale implementation, and suggest that PdNA MBBRs are feasible and sustainable for full-scale ammonia, nitrate, and nitrite polishing to meet stringent total nitrogen requirements. PRACTITIONER POINTS: This research will help utilities develop methods for starting up mainstream anammox MBBRs without the barrier of anammox biomass seeding. Preliminary biofilm carriers accelerated startup time in a PdNA MBBR, but a virgin carrier reactor started up in a similar timeframe, contrary to expectations. Also, contrary to expectations, high concentrations of ammonia and nitrite are not necessary for startup of an anammox or PdNA MBBR.

Mainstream partial denitrification-anammox in sand and expanded clay deep-bed polishing filters under practical loading rates and backwashing conditions.

This study focused on evaluating the feasibility of expanded clay and sand as media types for mainstream partial denitrification-anammox (PdNA) in deep-bed single-media polishing filters under nitrogen and solids loading rates as well as backwash conditions similar to conventional denitrification filters. The surface roughness and iron content of the expanded clay were hypothesized to allow for enhanced anammox retention, nitrogen removal rates, and runtimes. However, under the tested loading rates and backwash conditions, no clear benefit of expanded clay was observed compared with conventional sand. This study showed the feasibility of PdNA in filters with both sand and expanded clay with PdN efficiencies of 76% and 77%, PdNA rates of 840 and 843 g N/m3 /d and TIN removal rates of 960 and 964 g N/m3 /d, respectively. Glycerol demands were 1.5-1.6 g COD added per g TIN removed , thus indicating potential carbon savings up to 75% compared with conventional denitrification. Overall, this study showed for the first time PdNA filters performing at nitrogen removal rates double that of previous PdNA studies under realistic conditions while providing insights into the media choice and backwashing conditions. Future research on expanded clay backwash conditions is needed to provide its full potential in PdNA filters. PRACTITIONER POINTS: Hydraulic and TSS loading rates similar to conventional denitrification can be applied in PdNA filters. Conventional sand can be used when retrofitting conventional denitrification filters into PdNA filters. Carbon savings up to 75% can be achieved with glycerol when retrofitting conventional filters into PdNA filters.

Media selection for anammox-based polishing filters: Balancing anammox enrichment and retention with filtration function.

Retrofitting conventional denitrification filters into partial denitrification-anammox (PdNA)- or anammox (AnAOB)-based filters will reduce the needs for external carbon addition. The success of AnAOB-based filters depends on anammox growth and retention within such filters. Studies have overlooked the importance of media selection and its impact on AnAOB capacity, head loss progression dynamics, and shear conditions applied onto the AnAOB biofilm. The objective of this study was to evaluate viable media types (10 types) that can enhance AnAOB rates for efficient nitrogen removal in filters. Given the higher backwash requirement and lower AnAOB capacity of the conventionally used sand, expanded clay (3-5 mm) was recommended for AnAOB-based filters in this study. Owing to its surface characteristics, expanded clay had higher AnAOB activity (304- vs. 104-g NH4 + -N/m2 /day) and higher AnAOB retention (43% more) than sand. Increasing the iron content of expanded clay to 37% resulted in an increase in zeta potential, which led to 56% more anammox capacity compared to expanded clay with 7% iron content. This work provides insight into the importance of media types in the growth and retention of AnAOB in filters, and this knowledge could be used as basis in the development of PdNA filters. PRACTITIONER POINTS: Expanded clay showed the lowest head loss buildup and most likely will result in longer runtime for full-scale PdNA applications The highest AnAOB rates were achieved in expanded clay types and sand compared with smaller media typically used in biofiltration Expanded clay resulted in better AnAOB retention under shear, whereas sand could not withstand shear and required more frequent backwashing Expanded clay iron coating enhanced AnAOB enrichment and retention, most likely due to increased surface roughness and/or positive charge.

Full-scale transition from denitrification to partial denitrification-anammox (PdNA) in deep-bed filters: Operational strategies for and benefits of PdNA implementation.

This study shows for the first time more than 2 years of operation of a mainstream anammox application at full-scale under temperate climate. This implementation of partial denitrification-anammox (PdNA) in deep bed filters at the HRSD York River treatment plant was demonstrated to achieve the benefits of shortcut nitrogen removal without nitrite oxidizing bacteria (NOB) out-selection. The transition from denitrification to PdNA filters required bleeding ammonium to the filters using an optimized ammonium versus NOx (AvN) control in the upstream aeration tanks and maintaining a nitrate residual in the filter effluent through feedforward/feedback control. The latter actions led to savings of 85% in methanol, 100% in alkalinity, and 35% in capacity enhancement. Up to 6 mg NH4 + -N/L with an average of 2.2 ± 0.98 mg NH4 + -N/L was removed through the anammox pathway, which accounted for about 15% of the overall plant nitrogen removal. Anammox enrichment was confirmed by activity testing and molecular analysis. The large excess of AnAOB capacity present in the filters (5-10 times more than normal operation) resulted in stable and reliable operation through winter conditions and showed potential for further intensification. PRACTITIONER POINTS: For the first time, long-term mainstream anammox was established full-scale through PdNA implementation in deep-bed filters. PdNA implementation required upstream aeration control optimization to provide a blend of ammonium and nitrate to the filters. Efficient anammox enrichment and retention resulted in reliable PdNA performance under different seasonal and influent conditions. PdNA implementation resulted in significant methanol and alkalinity savings and upstream capacity enhancement as ammonia removal depended less on aerobic nitrification. In the event of NOB out-selection and presence of nitrite, carbon savings in PdNA polishing filters can be enhanced via partial nitritation-anammox.

Mainstream short-cut N removal modelling: current status and perspectives.

This work gives an overview of the state-of-the-art in modelling of short-cut processes for nitrogen removal in mainstream wastewater treatment and presents future perspectives for directing research efforts in line with the needs of practice. The modelling status for deammonification (i.e., anammox-based) and nitrite-shunt processes is presented with its challenges and limitations. The importance of mathematical models for considering N2O emissions in the design and operation of short-cut nitrogen removal processes is considered as well. Modelling goals and potential benefits are presented and the needs for new and more advanced approaches are identified. Overall, this contribution presents how existing and future mathematical models can accelerate successful full-scale mainstream short-cut nitrogen removal applications.

Partial denitrification-anammox (PdNA) application in mainstream IFAS configuration using raw fermentate as carbon source.

This research examined the feasibility of raw fermentate for mainstream partial denitrification-anammox (PdNA) in a pre-anoxic integrated fixed-film activated sludge (IFAS) process. Fermentate quality sampled from a full-scale facility was highly dynamic, with 360-940 mg VFA-COD/L and VFA/soluble COD ratios ranging from 24% to 48%. This study showed that PdNA selection could be achieved even when using low quality fermentate. Nitrate residual was identified as the main factor driving the PdN efficiency, while management of nitrate conversion rates was required to maximize overall PdNA rates. AnAOB limitation was never observed in the IFAS system. Overall, this study showed PdN efficiencies up to 38% and PdNA rates up to 1.2 ± 0.7 g TIN/m2 /d with further potential for improvements. As a result of both PdNA and full denitrification, this concept showed the potential to save 48-89% methanol and decrease the carbon footprint of water resource recovery facilities (WRRF) by 9-15%. PRACTITIONER POINTS: Application of PdNA with variable quality fermentate is feasible when the nitrate residual concentration is increased to enhance PdN selection. To maximize nitrogen removed through PdNA, nitrate conversion rates need enhancement through optimization of upstream aeration and PdN control setpoints. The IFAS PdNA process was never anammox limited; success depended on the degree of PdN achieved to make nitrite available. Application of PdNA with fermentate can yield 48-89% savings in methanol or other carbon compared with conventional nitrification and denitrification. Integrating PdNA upstream from polishing aeration and anoxic zones guarantees that stringent limits can be met (<5 mg N/L).

Unravelling adaptation of nitrite-oxidizing bacteria in mainstream PN/A process: Mechanisms and counter-strategies.

Stable suppression of nitrite-oxidizing bacteria (NOB) is still a major challenge for the implementation of partial nitritation and anammox (PN/A) in mainstream treatment. Despite numerous suppression strategies demonstrated, it is increasingly recognized that NOB could develop resistance to these strategies, threatening the long-term stability of the mainstream PN/A process. This study aims to understand adaption mechanisms and develop counter-strategies to overcome the adaptation. To this end, three previously-demonstrated suppression strategies, including NOB inactivation via side stream sludge treatment with free ammonia (FA), the use of low dissolved oxygen (DO), and the use of anammox to scavenge nitrite, were stepwise applied, over a period of 800 days, to a laboratory-scale reactor treating effluent from a high-rate activated sludge (HRAS) plant. FA sludge treatment alone sustained nitrite accumulation for about two months, after which NOB adaptation occurred causing PN to fail. The FA adaptation was induced by a shift in the NOB community from Nitrospira to Ca. Nitrotoga. The latter was found to have higher resistance to FA and also a higher maximum specific growth rate. Low DO at 0.2-0.4 mg O2 L-1 was then applied, in conjunction with FA treatment, which successfully eliminated Ca. Nitrotoga and re-established PN. However, new and unidentified NOB with a higher apparent oxygen affinity emerged in three months, again leading to PN failure. Lastly, as the third strategy for NOB suppression, anammox was introduced as an in-situ nitrite-scavenger. The combo-strategy delivered reliable NOB suppression with no further adaptation in the remaining experimental period (eight months). The resulted one-stage PN/A reactor achieved a nitrogen removal efficiency of 84.2 ± 5.37%. A control reactor, operated in parallel under the same conditions but without FA treatment, only achieved 10.4 ± 4.6% nitrogen removal, with anammox completely outcompeted by NOB in the last phase.

Stoichiometric and kinetic characterization of an acid-tolerant ammonia oxidizer 'Candidatus Nitrosoglobus'.

Recently, acidic (i.e. pH<5) nitrification in activated-sludge is attracting attention because it enables stable nitritation (NH4+ â†’ NO2-), and enhances sludge reduction and stabilization. However, the key acid-tolerant ammonia oxidizers involved are poorly understood. In this study, we performed stoichiometric and kinetic characterization of a new acid-tolerant ammonia-oxidizing bacterium (AOB) belonging to gamma-proteobacterium, Candidatus Nitrosoglobus. Ca. Nitrosoglobus was cultivated in activated-sludge in a laboratory membrane bioreactor over 200 days, with a relative abundance of 55.1 ± 0.5% (indicated by 16S rRNA gene amplicon sequencing) at the time of the characterization experiments. Among all known nitrifiers, Ca. Nitrosoglobus bears the highest resistance to nitrite, low pH, and free nitrous acid (FNA). These traits define Ca. Nitrosoglobus as an adversity-strategist that tends to prosper in acidic activated-sludge, where the low pH (< 5.0) and high levels of FNA (at parts per million levels) sustained and inhibited all other nitrifiers. In contrast, in the conventional pH-neutral activated-sludge process, Ca. Nitrosoglobus is less competitive with canonical AOB (e.g. Nitrosomonas) due to the relatively slow specific growth rate and low affinities to both oxygen and total ammonia. These results advance our understanding of acid-tolerant ammonia oxidizers, and support further development of the acidic activated-sludge process in which Ca. Nitrosoglobus can play a critical role.

Robust Nitritation Sustained by Acid-Tolerant Ammonia-Oxidizing Bacteria.

Oxidation of ammonium to nitrite rather than nitrate, i.e., nitritation, is critical for autotrophic nitrogen removal. This study demonstrates a robust nitritation process in treating low-strength wastewater, obtained from a mixture of real mainstream sewage with sidestream anaerobic digestion liquor. This is achieved through cultivating acid-tolerant ammonia-oxidizing bacteria (AOB) in a laboratory nitrifying bioreactor at pH 4.5-5.0. It was shown that nitrite accumulation with a high NO2-/(NO2- + NO3-) ratio of 95 ± 5% was stably maintained for more than 300 days, and the obtained volumetric NH4+ removal rate (i.e., 188 ± 14 mg N L-1 d-1) was practically useful. 16S rRNA gene sequencing analyses indicated the dominance of new AOB, "Candidatus Nitrosoglobus," in the nitrifying guild (i.e., 1.90 ± 0.08% in the total community), with the disappearance of typical activated sludge nitrifying microorganisms, including Nitrosomonas, Nitrospira, and Nitrobacter. This is the first identification of Ca. Nitrosoglobus as key ammonia oxidizers in a wastewater treatment system. It was found that Ca. Nitrosoglobus can tolerate low pH (<5.0), and free nitrous acid (FNA) at levels that inhibit AOB and nitrite-oxidizing bacteria (NOB) commonly found in wastewater treatment processes. The in situ inhibition of NOB leads to accumulation of nitrite (NO2-), which along with protons (H+) also produced in ammonium oxidation generates and sustains FNA at 3.0 ± 1.4 mg HNO2-N L-1. As such, robust PN was achieved under acidic conditions, with a complete absence of NOB. Compared to previous nitritation systems, this acidic nitritation process is featured by a higher nitric oxide (NO) but a lower nitrous oxide (N2O) emission level, with the emission factors estimated at 1.57 ± 0.08 and 0.57 ± 0.03%, respectively, of influent ammonium nitrogen load.

The inhibitory impact of ammonia on thermally hydrolyzed sludge fed anaerobic digestion.

This study evaluated the impact of ammonia on mesophilic anaerobic digestion (AD) with thermal hydrolysis pretreatment (THP) treating a mixture of primary sludge and waste activated sludge and operated under constant organic loading rate of 9 kg COD/m3 /d. Free ammonia concentrations in the digesters were varied between 37 and 966 mg NH3 -N/L, while maintaining all other operational conditions constant. A decrease in volatile solids reduction from 54 ± 5% (at <554 mg NH3 -N/L) to 35 ± 6% at the maximum free ammonia concentration of 966 mg NH3 -N/L was observed at steady-state conditions. No impact of free ammonia on final dewaterability was detected. Free ammonia thus mostly limited methanogenesis. A free ammonia Monod inhibition constant of 847 ± 222 mg NH3 -N/L for methanogens was estimated based on the digester steady-state methane rates dynamics. This study showed that current THP AD digesters (typically 110-260 mg NH3 -N/L) operate under 12%-18% ammonia inhibition for methanogenesis. Operation under SRT of 15 days, about 2 times more than needed to retain methanogens, can compensate for lower methanogens rates and avoid performance impacts. The later shows a good potential to operate under higher free and total ammonia concentration without jeopardizing performance. PRACTITIONER POINTS: Only from a free ammonia concentration above 554 mg NH3 -N/L, decreased volatile solids reduction and biogas yield were observed. A volatile solids reduction of 35 ± 6% at maximum free ammonia concentration of 966 mg NH3 -N/L was still achieved. A Monod inhibition constant for methanogens of 847 ± 222 mg NH3 -N/L was estimated. It was estimated that current THP AD systems (110-260 mg NH3 -N/L) operate under 12%-18% NH3 inhibition for methanogenesis.