Development of an anaerobic ammonium oxidation (ANAMMOX) process reactor for aquaculture wastewater treatment

As a developed country, Malaysia has experienced exponential social and economic growth in numerous sectors. For decades, the aquaculture sector has been a significant source of animal protein for the people of Malaysia (FAO 2017). The rapidly growing aquaculture industry contributed to 47% of the world's fish consumption, which enhances the waste produced in production systems (Luo et al. 2016). This wastewater was excessive in total nitrogen (TN), total phosphorus, organic material, and suspended particles; when it enters the aquatic environment, it causes pollution by reducing the capacity to process these nutrients and altering the food chain and biogeochemical cycles (Gumelar et al. 2022; Ismail et al. 2022). Since the potential threat of nitrogen compounds to the marine environment, more attention has been paid to the removal of nitrogen from wastewater (Ge et al. 2019).

Conventional nitrogen removal technologies, including biological, chemical, or physical processes, are energy-intensive due to high aeration requirements (Hasan et al. 2021). The use of bio-process for instance trickling filters, vortex sand biofilters, and advanced technologies likely bio-floc technology (BFT), recirculating aquaculture systems (RAS), aquaponics, and others frequently applied for aquaculture ecosystems (Ahmad et al. 2021). However, the disadvantages of these systems include system complexity and higher operation (Rahimi et al. 2020) and maintenance costs. These constraints led to nitrogen removal techniques known as anaerobic ammonium oxidation (ANAMMOX), which do not require aeration, involve lower sludge production, and result in significant energy savings, have been considered practical alternatives for aquaculture areas (Ismail et al. 2022; Gumelar et al. 2024).

Recently, ANAMMOX bacteria have been discovered in aquaculture systems indicating that nitrogen-polluted water and sediment could be treated with ANAMMOX (Shen et al. 2016). The utilization of ANAMMOX technology in the elimination of nitrogen attracted interest and thus became popular in the water industry worldwide due to the various advantages it offers (Wen et al. 2020). Ali & Okabe (2015) noted as of 2015, there were approximately 114 large-scale ANAMMOX treatment plants worldwide. In Western Australia, Environmental Engineers International Pty Ltd (EEI) used this technology as a preference to replace the existing standard nitrification-denitrification wastewater treatment plant (WWTP) (Kurup 2018). The WWTP saved about 20% of energy and achieved a 95% reduction in TN (Environmental Engineers International (EEI) 2020).

The purpose of this research was to explore the performance of nitrogen removal as well as to culture locally isolated ANAMMOX bacteria from sludge from an anaerobic lagoon in Malaysia. The anaerobic lagoon of the integrated shrimp aquaculture park was chosen as the inoculum in the reactor. Research by Van Duc et al. (2018) reported 88%, nitrogen removal efficiency (NRE) and 1.0 kg-Nm³.d of nitrogen loading rate (NLR) for enriched marine sediment from a shrimp aquaculture pond. Gumelar et al. 2024 reported a fast start-up ANAMMOX process using inoculum from shrimp pond solid waste using a filter bioreactor. The anaerobic lagoon is subject to eutrophication due to nutrient enrichment from the ecological shrimp hatcheries (Mavraganis et al. 2020). Therefore, it was selected as a potential ecosystem for discovering and cultivating the ANAMMOX bacteria from tropical regions.

Since ANAMMOX bacteria are a potential candidate for nitrogen removal from nitrogenous wastewater, it was crucial to analyze the efficiency of nitrogen removal during the start-up phase. Furthermore, since this technology supports the Sustainable Development Goal (SDG 6), the utilization of the ANAMMOX technology in Malaysia should be promoted for environmental sustainability. The results of this research could justify the application of the ANAMMOX technology for the removal of nitrogen biologically for aquaculture effluent remediation systems in Malaysia.

The physical and chemical sludge characteristics samples were examined using the standard (APHA 2017), procedure for analysis of water and wastewater. The samples collected from three checkpoints were analyzed for direct in-situ and ex-situ experiments. Samples were taken from the sludge of the anaerobic lagoon at a depth of 0.5 to 1 m below the surface. The in-situ experiment included analysis of pH and salinity meanwhile ex-situ such as total solid (TS) and volatile suspended solid (VSS) were analyzed by gravimetric method (APHA 2017).

In addition, the TN and total phosphate (TP) were examined via the spectrophotometric method and colorimetric method, respectively (APHA 2017). The results of the physicochemical characteristics of sludge and seeding sludge are reported in Table 1.

The collected sludges were then mixed (seeding sludge) in a 15-L canister and tightly sealed. The sample was stored in a freezer before proceeding to be added for operation in the ANAMMOX reactor. For the start-up of the ANAMMOX process in the reactor, this sludge served as inoculum.

Table 2 describes the synthetic wastewater composition. Trace element I included FeSO₄ 5 g; EDTA 2Na 6.37 g and Trace Element II (per litre of distillate water contained EDTA 2 Na 19,11 g; CoCl.6H₂O 0.24 g; ZnSO₄ 0.241 g; MnCl₂.4H₂O 0.99 g; CuSO₄.5H₂O 0.25 g; H₃BO₄ 0.014 g; NaMnO₄.2H₂O 0.22 g; NiCl₂.6H₂O 0.19 g; NaSeO₄.10H₂O 0.024 g (Van Duc et al. 2018; Zulkarnaini et al. 2024). The use of synthetic wastewater in this research was intended to provide a controlled environment for investigating the fundamental aspects of the ANAMMOX process. When the reactor was first started to operate, 70 mg-N/L (NH₄)₂SO₄ and NaNO₂ were added with a ratio of 1:1 to substrate (per litre of distillate water). The concentration of ammonium and nitrite was increased gradually from 70 to 150 mg-N/L throughout the experimental period.

Prior to addition to the reactor, samples collected at three checkpoints were mixed as seeding sludge. A unit of laboratory-scale fixed film bioreactor with a working volume of 1.8 L and a string wound filter as the carrier will be used for the cultivation of ANAMMOX bacteria (filled about 80% in reactor volume), keeping the room temperature constant. This research was conducted as a laboratory-scale study. To prevent heterotrophic bacteria's growth, an aluminum foil was used as cover from direct sunlight (Zulkarnaini et al. 2021a). Synthetic wastewater that acts as substrate was pumped using a peristaltic pump at 24 h, with a flowrate of 1.25 mL/min.

The reactor's operation was separated into three periods known as period I, period II, and period III with different influent concentrations of ammonium and nitrite. 70 mg-N/L of ammonium and nitrite was supplied in period I 100 mg-N/L in period II, and 150 mg/L in period III. Table 3 shows the experimental condition of fixed film bioreactors.

Table 3

Experimental conditions of fixed film bioreactor

Period .	Day .	Concentration (mg-N/L) .		Hydraulic Retention Time (h) .
		.	.
I	0–45	70	70	24
II	50–120	100	100	24
III	130–200	150	150	24

Period .	Day .	Concentration (mg-N/L) .		Hydraulic Retention Time (h) .
		.	.
I	0–45	70	70	24
II	50–120	100	100	24
III	130–200	150	150	24

View Large

Samples were collected twice a week for analysis of the influent and effluent. Nitrate, nitrite and ammonium were analyzed using a UV-Vis spectrophotometer (Shimadzu 1800, Kyoto, Japan) at a specific wavelength (APHA Method, 2017). The pH and salinity were measured using YSL Multiparameter.

The ANAMMOX process was determined via conversion of ammonium, nitrate and nitrite production in the reactor. The quantitative ratio between reactants and products in a chemical process is known as stoichiometry. The stoichiometry ratio is calculated by the ratio of nitrite/ammonium consumption as well as nitrate/ammonium by using the following equations (Zulkarnaini et al. 2021b):

Descriptions:

• = Concentration of influent ammonium

• = Concentration of effluent ammonium

• = Concentration of influent nitrite

• = Concentration of effluent nitrite

• = Concentration of influent nitrate

• = Concentration of effluent nitrate

• HRT = Hydraulic retention time (24 h).

To observe the ANAMMOX bacteria growth, monitoring the concentration of influent and effluent data and microbial profiling were analyzed. Besides, bacterial growth was monitored through visual color changes in the sludge, stoichiometric ratio evaluations, and molecular methods. The sample of influent and effluent were analyzed twice weekly to measure the concentration of ammonium, nitrate and nitrite. Following the standard APHA method, samples were examined using a UV-VIS spectrophotometer.

Period I represent the start-up process using a substrate with a concentration of ammonium and nitrite (70 mg/L) for 45 days. Since the ACE value has been maintained at 50–57% for 20 days, the concentration of ammonium and nitrite has been elevated. Period II began on day 50 until day 120 along with an increment in concentration of ammonium and nitrite from 70 to 100 mg/L. The sludge inside the reactor was removed on day 125. The ACE and NRE values for 15 days remain in the range of 50%, thus the ammonium and nitrite concentration has been elevated from 100 to 150 mg/L. Period III lasts from day 130 to day 200 whereby the ANAMMOX process conditions have shown stability and optimum level.

Period I began at days 0–45 when the ammonium and nitrite concentrations of influent were 70 mg/L. Figure 2 displays the nitrogen conversion inside the reactor. Starting from day 0 to 30, the concentration of ammonia in effluent showed an average ranging from 64.5 to 40.15 mg/L. Then, ammonium concentration in effluent exhibited a gradual and consistent decline with the lowest concentration beings on day 45 at 30.98 mg/L. Moreover, the average concentration of nitrite in the effluent was 40 mg/L as well as 49.62 mg/L as the highest concentration on day 10 and the lowest on day 30 at 43.68 mg/L. When the concentration of effluent on ammonium and nitrite was reduced, the nitrate concentration in effluent moderately increased from 40.23 mg/L from day 10 until day 30 at 50.49 g/L.

Since ACE remains at 50–57% for weeks, the additional influent concentration of ammonium and nitrite was done to elevate ACE. The initial 70 mg/L of influent concentration was elevated toward 100 mg/L for both ammonium and nitrite in period II. As observed, the effluent concentration of ammonia remained lower compared to the influent with the highest and lowest concentration of 68.47 and 39.75 mg/L for day 50 and day 70, respectively. Nitrite is required during the process of ANAMMOX as an electron acceptor. During period II, from day 50 to day 120, the effluent nitrite concentration decreased from 70.64 to 63.60 mg/L. The effluent of nitrate concentration continues at an average of 40 mg/L with 48.57 mg/L as the dense concentration and 41.77 mg/L as the lowest concentration on day 105 and day 60 correspondingly.

Sludge that was inoculated (since the start of the experiment) inside the reactor was removed on day 125 by maintaining the biomass attached to the string wound filter. Day 125 considered ANAMMOX bacteria had established the population and achieved stable nitrogen removal performance. Inoculum was necessary to establish the ANAMMOX process; however, removing the inoculum can help prevent contamination from unwanted microorganisms that compete with ANAMMOX and also enhance the performance.

During this period III, the influent concentration of ammonia and nitrite elevated to 150 mg/L on day 130. The efficiency of nitrogen removal has shown to be decreased when compared to the value obtained in period I. This was noticeable since there was a drastic decrease in the efficiency of ammonium conversion and removal of nitrogen for day 130 and day 140. Apparently, the occurrence of shock loading caused this, which can still happen even despite progressive increment in the concentration of substrate starting 70 to 100 to 150 mg/L.

After day 140, recovery started to take place, and return nitrogen removal performance increased. This is proven by the fact that the ammonium and nitrite effluent concentration simultaneously, then continuously, decreased until day 160. The efficiency of nitrogen removal was stable until the experiment was complete. The concentration of ammonia (effluent) gradually decreased with the lowest being 48.32 mg/L on the day 200. Meanwhile, the concentration of nitrite for effluent was in between 70 and 150 mg/L throughout period III. Besides, maximum effluent concentration of nitrate was on day 160 (52.19 mg/L) and the minimal was on the day 135 (49.26 mg/L).

Performance nitrogen removal data showed fluctuation state during the period I and II, then increased after sludge removal and period III showed a stable state and optimum level until the day 200. The greater the concentration of ammonium and nitrite in the influent, the least the effluent concentration of ammonium and nitrite. During day 200 of reactor operation, it was observed that the optimum levels of ACE, NRE, were at 70.39, 82.49%, and 0.125 kg-N/m³.d of NRR with 0.30 kg-N/m³.d of NLR, respectively.

Furthermore, the highest and lowest ACE were 70.39 and 7.86% for day 200 and the start-up day respectively. The ACE gradually increased up till day 45 at 57.23% and started to decline on day 50 at 31.53%. Likewise, NRE declined 13.36% on day 50 since influent ammonia and nitrite concentration increased from 70 to 100 mg/L. The nitrogen removal not reaching the optimum level during this period I with the highest amount of ACE, NRE, and NRR of 57.23%, 32.26%, and 0.023 kg-N/m³.d, respectively, along with 0.14 kg-N/m³.d of NLR on day 45.

The minimal amount of ACE, NRE, NRR, and NLR for period II happened on day 50 with 31.53, 13.36%, 0.013 kg-N/m³.d, with 0.20 kg-N/m³.d of NLR, respectively. Moreover, during period II continuous increment ACE from 31.53 to 44.17% and NRE from 35.32 on day 56 to 53.06% on day 120 as well as 0.054 kg-N/m³.d of NLR including 0.20 kg-N/m³.d of NRR. Considering moderate ACE and NRE that remain at an average of 40–50% for two months, adjustment at the concentration of influent nitrite and ammonium was applied.

The influent's nitrite and ammonium concentrations were constant at 150 mg/L during period III. The reactor undergoes a shock loading phase and causes a decline in nitrogen removal. As a result, ACE and NRE decreased to 13.07 and 10.78%, respectively, and 0.017 kg-N/m³.d of NRR. At day 140, the recovery process began and an increase in ACE and NRE was observed. The highest NRE was 82.49 on day 200 and the lowest was −8.06% on start-up day. The amount of NLR remained at an average 0.14 kg-N/m³.d, 0.20 kg-N/m³.d, and 0.30 kg-N/m³.d for start-up day, day 40, day 100 and day 200, respectively.

Zulkarnaini et al. (2018) mentioned that the type of reactor and the temperature were the main factors in determining the ANAMMOX process' stoichiometry. The ratio of stoichiometric for this study was 1 mol ammonium to 1.00 mol nitrite removal to the 0.21 nitrate production (1:1.00:0.21). In summary, the ANAMOX bacteria were efficient at nitrogen removal in the bioreactor because the ratio was 1:1 between ammonium and nitrite removal. In addition, the ANAMMOX bacteria removed both ammonium and nitrite (1:1) from the wastewater and produced a small amount of nitrate (0.21 ratio) as a by-product.

In this study, the existence of ANAMMOX bacterial species in the sludge of an anaerobic lagoon in an integrated shrimp farming facility in Terengganu was successfully demonstrated. It was found that anaerobic conditions in a solid film bioreactor with a coiled polypropylene filter as the biomass carrier can promote the growth performance of ANAMMOX bacteria under continuous feeding and constant salinity, temperature and pH. It was also proved that ANAMMOX bacteria were competent in nitrogen removal with the optimum level of ACE, NRE, and NRR at 70.39, 82.49%, and NLR of 0.30 kg-N/m³.d with 0.125 kg-N/m³.d after day 200 of the operational period. The ANAMMOX process is highly effective compared to other nitrogen removal systems as it achieves a high NRE of 82.49% and operates with a competitive NLR of 0.30 kg-N/m³·d. Moreover, it was more economical than traditional nitrification-denitrification systems and advanced technologies like MBBR and MBR, as it requires less aeration and has lower operational costs (Wang et al. 2022).

In conclusion, the enrichment of shrimp sediments has demonstrated the presence of ANAMMOX bacteria in our tropical country. Findings from this research offer helpful knowledge for the ANAMMOX process' enhancement and its application on an industrial scale. In addition, it is crucial to manage water quality for the aquaculture ecosystem to ensure shrimp health and quality and provide a favorable environment for aquatic life growth.

There are a few recommendations that can be suggested for further improvement in this research. The findings from this research supply a great platform for the potential of the ANAMMOX technology for treating aquaculture wastewater. First, perform enrichment and cultivation of ANAMMOX bacterium at constant temperature and pH for the reactor's greatest ANAMMOX performance. Excessively low or high pH is detrimental to the ANAMMOX system's ability to effectively remove nitrogen from the wastewater. Moreover, salinity has a significant effect on the removal of nitrogen and could influence the microbial diversity of ANAMMOX species.

To further investigate ANAMMOX cell morphology, it is advisable to subject the enriched ANAMMOX culture to electron microscopy analysis, such as transmission electron microscopy (TEM) along with analysis on scanning electron microscopy (SEM). By identifying the distinct cell compartmentalization of the microorganisms, the morphological analysis of ANAMMOX cell structure can further establish the enriched bacteria as ANAMMOX.

Implementation of an ANAMMOX process was advocated in biological nitrogen removal. Therefore, it is proposed to explore the ANAMMOX process using other reactors such as membrane bioreactors and sequence batch reactors. For ammonia-nitrogen removal, it is expected that further investigation will give a stronger foundation for the anticipated technical implementation of ANAMMOX technology in Malaysia.

This research was supported by the Ministry of Higher Education (MOHE) through the Collaboration Research Grant (CRG 15) (Grant No. 53461).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.