Removal of organic matter and nitrogen in a single reactor with extremely high sludge age and cyclic aeration Open Access

Important parameters for designing aerobic reactors with optimal performance include: cell retention time or sludge age (SRT = θ_c), biomass (VSS = X), and substrate/microorganism ratio (S/X). In conventional bioreactors, SRT varies between approximately 8 and 15 days; for high-rate reactors, sludge ages of 2–3 days are used, while in extended aeration reactors, they generally do not exceed 30 days. However, to remove a larger amount of organic matter, longer SRT has been used, independent of the biological treatment type used (Orozco 2014; Libhaber & Orozco 2015). This has led to the development of the concept of Extremely High Sludge Age (EHSA), which involves treating domestic wastewater (DWW) with a very high θ_c, resulting in maximum organic matter removal efficiency. This is because the concentration of biomass increases with θ_c, leading to a progressively lower substrate-to-biomass ratio (S/X). Additionally, as biomass is fully retained, the production of excess sludge is minimal (Paul & Liu 2012; Van Haandel & Van Der Lubbe 2012; Orozco 2014). One of the greatest advantages is that the hydraulic retention time (HRT = t_d) of the aeration tank can be reduced in proportion to the increase in the biomass (Mixed Liquor Volatile Suspended Solids – MLVSS) since X* t_d is a constant (Libhaber & Orozco 2015). Due to its length, the model deduction is detailed in Supplementary Material, highlighting an advantage of the EHSA system in space-restricted settings.

Studies that have used high SRT have not only analyzed the removal of organic matter but also dissolved oxygen (DO) (Fan et al. 2017), nutrient removal (Jena et al. 2016), biomass production behavior, sludge sediment ability (Amanatidou et al. 2015b), and sludge recirculation conditions, among others. With SRT between 150 and 300 days, in pilot scale membrane bioreactors used to treat leachate, removals above 90% were generated in: biochemical oxygen demand (BOD₅), total Kjeldahl nitrogen (TKN) and suspended solids (SS) (Ittisupornrat et al. 2021), ensuring not only the removal levels achieved by conventional systems for organic matter, but also offering the advantage of eliminating other sanitary-relevant pollutants that require high sludge retention times (SRTs). This is similar to what is seen when there is complete retention of solids due to extreme starvation conditions, including the consumption of certain biodegradable organic compounds that are not degraded in systems with conventional SRTs, as well as the slow hydrolysis of cell debris (Amanatidou et al. 2015a), or when operating within an Intermittent Cycle Extended Aeration System – ICEAS (Nguyen et al. 2021).

The methods applied for nutrient removal are very varied, ranging from discontinuous sequential reactors (SBR) which combine aerobic, anoxic, and anaerobic phases, by experimenting with the amount of oxygen provided in the aerobic sections, either in separate tanks or in the same reactor, in order to obtain nitrification–denitrification and dephosphatization (Arumugham et al. 2024). When there is intermittent aeration (IA) or the oxygen content is manipulated, the populations of microorganisms specific to each condition can be stimulated and the nutrients can be removed simultaneously. Thus, IA is present in many of the experiments that aim to remove high quantities of nutrients from wastewater, using diverse treatment strategies, (Yoo et al. 1999; Cheng & Liu 2001; Pérez et al. 2002; Cárdenas et al. 2006; Water Environmental Federation 2006; Lee et al. 2015) among others.

IA has also been used to enhance the removal of organic and inorganic compounds in a way similar to that seen in vertical flow wetlands (Wu et al. 2015, 2016; Hou et al. 2018), to remove chemical compounds (Tanwar et al. 2007), in landfill bioreactors used for solids treatment (Sang et al. 2009), in submerged MBRs (Yang & Yang 2011; Wu & He 2012), demonstrating high efficiency in contaminant removal through thermophilic membrane processes, achieving significant reductions in volatile solids by 90% (Collivignarelli et al. 2025). Furthermore, this technology has enabled a circular economy approach by facilitating the reuse of contaminants such as carbon (C), nitrogen (N), and phosphorus (P) (Collivignarelli et al. 2021). This latter technology is considered highly effective for contaminant removal (Abdelrahman et al. 2025), although it implies high maintenance costs for the membranes (Jiang et al. 2023; Nthunya et al. 2024; Bokhary et al. 2025).

Similarly, IA has played a key role in activated sludge (AS) tanks with fixed media (Singh et al. 2018; Wu et al. 2022; Lan et al. 2024; Rui et al. 2024), in aerobic granular reactors (Long et al. 2015; Campo et al. 2020; Maia et al. 2022; Miyake et al. 2023; Yu et al. 2023; Cong et al. 2024), and in the anaerobic ammonium oxidation process, which has proven essential for achieving denitrification (Regmi et al. 2014; Deng et al. 2021; Wang et al. 2021), by inhibiting nitrite-oxidising bacteria and thus preventing the deterioration of these systems (Miao et al. 2022), among other applications.

The requirement for a carbon source to remove N in series reactors can be met when working in a single reactor with IA (Han et al. 2018; Orozco & Vélez 2020) and returning sludge up to 200%. In addition, it is possible to work with very low DO (0.5 mg/L) to achieve high removals of organic matter and TN (Gogina & Gulshin 2016), even if the tank is divided with partitions, as is the case of the baffled bioreactor operating under IA conditions, the reduction of COD, TN, and total phosphorus (TP) is considerable (Liu & Wang 2017).

The challenges to achieving denitrification and dephosphatization seen in the SBR led to the formulation of new treatment trains, where the anaerobic stage was modified by a prolonged period of inactivity in the AS, called aerobic/extended-idle (A/EI). Better results were obtained in the removal of TP and TN with greater stability than the anaerobic/oxic systems (A/O). This shows that there is no need for the anaerobic phase for nutrient removal (Chen et al. 2013; Wang et al. 2014). Also, the hybrid SBR and biofilm reactor (SBR-BR) simultaneously remove TN and TP, with different times in each of the defined phases providing good living conditions for P accumulating organisms (PAOs), denitrifying phosphorus accumulating organisms (DNPAOs) and nitrifying and denitrifying bacteria (Yin et al. 2015). Furthermore, in sequencing batch biofilm reactors with the optimal ratio between C, N, and P: C: N:P of 140:70:7, IA enhances the role of DNPAOs in biological dephosphating (Mielcarek et al. 2015). This proves once again how DNPAOs, with IA, can play an important role in nutrient removal (Wan et al. 2017; Azhdarpoor et al. 2018).

There is an urgent need to remove ammonia nitrogen from wastewater, and conventional aerobic processes are highly useful, but they require large amounts of energy and generate high quantities of VSS. In this context, applying IA allows for a reduction in energy consumption and facilitates denitrification within the same reactor. On the other hand, when operating with EHSA, excess VSS is significantly reduced. These two applications offer a practical solution for populations with economic deficiencies, helping to remediate discharges that cause eutrophication in receiving water bodies. Another important factor is the low investment costs for existing conventional AS plants, limited only to the installation of an on/off aeration system and analysis of secondary clarifier characteristics. In the literature, existing advanced technologies for nutrient removal are highly effective but are not easily applicable to most populations in developing countries due to their high investment, operation, and maintenance costs, and the demands placed on operating personnel, which necessitates finding practical solutions to introduce these technologies. Therefore, the aim of this research was to assess the effects of EHSA and IA on the efficiency of COD and TN removal from DWW in a full-scale AS reactor.

We used the full-scale, continuous-flow wastewater treatment plant that treats the wastewater of the ‘Hotel Movich-Las Lomas’, located 2,000 m above sea level, in the municipality of Rionegro, Colombia. The plant was designed to treat a flow rate (Q) of 1.56 L/s, with BOD₅ = 325 mg/L, COD = 500 mg/L, SRT = θ_c ≥ 30 day, and MLVSS = X = 3,703 mg/L. The system is made up of a preliminary treatment that consists of double sieving that then delivers the water to an underground tank, the latter acting as a trap for sedimentable and floating solids, a bioreactor of AS with extended aeration and a kinetic selector, a secondary settler and drying beds. The bioreactor includes a selector at the beginning of the tank, in order to control the growth of filamentous bacteria. The aeration tank has a total hydraulic retention time of 14 h, with a volume of 81.1 m³.

Since 2018, we have monitored COD in influent and effluent flows, as well as in sedimentable solids (SedS), both in the ML and in the recirculation sludge in order to establish the EHSA condition. While not performing purges from the bottom of the secondary settler to the drying beds and with 100% recirculation of the sludge, we analyzed data for the parameters noted below to define the stability of the plant in terms of COD removal, the maximum amount of SedS reached, and the buoyancy of the biomass in the secondary settler. The SedS were measured daily in the Imhoff cone and the COD was quantified weekly at the beginning of the first 2 months and then with a frequency of 15 days in the remaining 7 months. During this stabilization of the bioreactor, the on and off cycle in the aeration system was set to 45′:15′, respectively, as determined by the authors' previous explorations.

We collected data on the removal of organic matter and nutrients at the treatment plant from March 2019 to January 2022. It is important to note that when the system was shut down (off cycle), there was neither mixing in the bioreactor nor sludge recirculation. Under this cycle, six composite samples proportional to the flow were collected over 24 h, taking samples in the influent, effluent, and ML. In addition, data were simultaneously recorded on the behavior in the bioreactor for the variables: DO, pH, and redox potential (ORP), in its on- and off-phases of the aeration system cycle. Data were collected over approximately 2 h of operation during the daytime. A quantitative analysis was made of the information obtained, using mean values, standard deviations, and coefficients of variation.

Following the Standard Methods for the Examination of Water and Wastewater (ed. 22), we quantified total and soluble chemical oxygen demand (COD_Tot and COD_Sol; method S.M 5220 D), TKN (method S.M 4500-Norg B and 4500-NH₃ B, C), total nitrogen (TN; by calculation), nitrites (⁠⁠; method S.M 4110 B), nitrates (⁠⁠; S.M 4110 B), TP (method S.M 4500-P B, E), orthophosphates (⁠⁠; method S.M 4110 B), total suspended solids (TSS; method S.M 2540 D), and volatile suspended solids (VSS; method S.M 2540 D, E) in the plant influent and effluent. We also measured pH, ORP, DO, and temperature with the HQ40d HACH multiparameter instrument. In the AS bioreactor, we quantified TSS, VSS, SedS, pH, DO, ORP, and temperature. Influent, effluent, and recirculation flows were quantified volumetrically.

We observed stability in organic matter, in the COD and SedS data collected over 9 months. COD had an average value of 525 mg/L for the influent and 40 mg/L for the effluent, while the SedS for the bioreactor was 580 mL/(L·h) and 860 mL/(L·h) in the recirculation (settler bottom). Between the 9th and 10th months, sludge was observed floating in greater quantities than usual in the secondary settler, which implied evacuation of the floating sludge in the clarifier toward the drying beds. In other words, at around 270 days biomass loss was seen in the floating sludge of the settler. With this as a reference, observation continued during the years 2019 and 2022, in which sludge flotation reoccurred at a frequency of every 9 or 10 months, and with few solids discharged into the drying beds. It is important to highlight that purges were never performed from the bottom of the secondary settler.

The EHSA condition was generated by not purging sludge from the bottom of the settler, recirculating the biomass, and skimming the floating sludge from the secondary settler to the reactor. Likewise, the average stability in the SedS for the reactor was between 450 and 650 mL/(L·h) and for the bottom of the secondary settler between 700 and 960 mL/(L·h), thus obtaining efficiencies greater than or equal to 90% in the reduction of COD.

Table 1 presents the parameters monitored in the influent and effluent of the plant through six composite sampling events, with 24 samples collected per event at intervals of 50–70 min, in order to alternate sampling between the off and on phases. Table 2 calculates the removal for total and soluble COD, TN, TP, and TSS, as well as their standard deviations and coefficients of variation. In general, when reviewing Table 1, it is observed that the plant complies with the discharge parameters for DWW in COD and TSS (Resolution 631 of 2015 in Colombia), with COD less than 180 mg/L and TSS less than 90 mg/L. The discharge also complies with BOD₅ given the COD values seen in the effluent. Although the influent COD varied considerably in the last four samples, since the average value was calculated at 525 mg/L, it can be seen that the removal of TN was not greatly affected by this change, because NT increases its removal up to 90% (see Table 2). The N in the effluent came out in the form of in the last four samples but in low quantities. This indicates that, due to an increase in the removal of TN, the nitrification and denitrification processes continued in the AS reactor, despite variability in the COD values. The values in COD_Sol at ≤40 mg/L and the low quantity of TSS in the effluent (between 10.3 and 19.0 mg/L), represent a low passage of solids into the effluent. This favors the EHSA condition and indicates the adequate functioning of the settler.

Table 1

Monitoring physical and chemical parameters

45′:15′ Cycle (on: off) .
Parameters .	8/03/2019 .	22/03/2019 .	12/04/2021 .	26/05/2021 .	1/06/2021 .	4/06/2021 .
	Influent
CODTot (mg/L)	506 ± 32	618 ± 40	444 ± 28	371 ± 24	284 ± 18	331 ± 21
CODSol (mg/L)	–	–	–	–	–	–
TN (mg N/L)	46.2 ± 2	57.7 ± 2	39.9 ± 2	34.3 ± 1.7	16.9 ± 0.8	35.6 ± 1.8
TKN (mg/L-N)	46.2 ± 2	57.7 ± 2	39.9 ± 2	34.3 ± 1.7	16.9 ± 0.8	35.6 ± 1.8
(mg/L-N)	<1	<1	<5	<1.13	<1.13	<1.13
(mg/L-N)	<0.05	<0.05	<0.40	<0.12	<0.12	<0.12
TP (mg/L)	3.5 ± 0.07	4.7 ± 0.09	2.8 ± 0.05	4.3 ± 0.08	2.1 ± 0.04	3.9 ± 0.08
(mg/L-P)	3.0 ± 0.1	2.5 ± 0.1	2.5 ± 0.1	<0.65	<0.65	1 ± 0.06
TSS (mg/L)	53 ± 2	107 ± 4	110 ± 8	61 ± 4	53 ± 3	61 ± 4
VSS (mg/L)	–	–	–	–	–	–
pH	6.9	7	6.8	6.8	6.7	6.7
Average pH	6.8
T (°C)	22.3	22.5	21.3	19.3	21	21.3
Average T (°C)	21.3
	Effluent
CODTot (mg/L)	73 ± 5	67 ± 4	37 ± 2	32 ± 2	40 ± 3	40 ± 3
CODSol (mg/L)	< 25	40 ± 3	<25	25 ± 2	<25	18 ± 1
TN (mg N/L)	18.9	31	12.2 ± 0.6	8 ± 0.38	1.9 ± 0.1	3.6 ± 0.2
TKN (mg/L-N)	18.5	31	<5	<5	<5	<5
(mg/L-N)	< 1	< 1	12.2 ± 0.6	8 ± 0.38	1.9 ± 0.1	3.6 ± 0.2
(mg/L-N)	0.41	< 0.05	<0.40	<0.12	<0.12	<0.12
TP (mg/L)	0.7	3.3 ± 0.07	3.2 ± 0.06	3.7	1.6 ± 0.03	2.3 ± 0.04
(mg/L-P)	< 0.48	0.7	2.5 ± 0.1	1.6	1.0 ± 0.06	0.9 ± 0.05
TSS (mg/L)	19 ± 1	19 ± 1	14 ± 1	12	15 ± 1	10.3 ± 0.7
VSS (mg/L)	17	17	13.9	12	15 ± 1	<10
pH	7.1	7.2	6.9	6.8	7.3	7.5
Average pH	7.1
T (°C)	22	21.4	21.3	21	21.5	21.6
Average T (°C)	21.5

45′:15′ Cycle (on: off) .
Parameters .	8/03/2019 .	22/03/2019 .	12/04/2021 .	26/05/2021 .	1/06/2021 .	4/06/2021 .
	Influent
CODTot (mg/L)	506 ± 32	618 ± 40	444 ± 28	371 ± 24	284 ± 18	331 ± 21
CODSol (mg/L)	–	–	–	–	–	–
TN (mg N/L)	46.2 ± 2	57.7 ± 2	39.9 ± 2	34.3 ± 1.7	16.9 ± 0.8	35.6 ± 1.8
TKN (mg/L-N)	46.2 ± 2	57.7 ± 2	39.9 ± 2	34.3 ± 1.7	16.9 ± 0.8	35.6 ± 1.8
(mg/L-N)	<1	<1	<5	<1.13	<1.13	<1.13
(mg/L-N)	<0.05	<0.05	<0.40	<0.12	<0.12	<0.12
TP (mg/L)	3.5 ± 0.07	4.7 ± 0.09	2.8 ± 0.05	4.3 ± 0.08	2.1 ± 0.04	3.9 ± 0.08
(mg/L-P)	3.0 ± 0.1	2.5 ± 0.1	2.5 ± 0.1	<0.65	<0.65	1 ± 0.06
TSS (mg/L)	53 ± 2	107 ± 4	110 ± 8	61 ± 4	53 ± 3	61 ± 4
VSS (mg/L)	–	–	–	–	–	–
pH	6.9	7	6.8	6.8	6.7	6.7
Average pH	6.8
T (°C)	22.3	22.5	21.3	19.3	21	21.3
Average T (°C)	21.3
	Effluent
CODTot (mg/L)	73 ± 5	67 ± 4	37 ± 2	32 ± 2	40 ± 3	40 ± 3
CODSol (mg/L)	< 25	40 ± 3	<25	25 ± 2	<25	18 ± 1
TN (mg N/L)	18.9	31	12.2 ± 0.6	8 ± 0.38	1.9 ± 0.1	3.6 ± 0.2
TKN (mg/L-N)	18.5	31	<5	<5	<5	<5
(mg/L-N)	< 1	< 1	12.2 ± 0.6	8 ± 0.38	1.9 ± 0.1	3.6 ± 0.2
(mg/L-N)	0.41	< 0.05	<0.40	<0.12	<0.12	<0.12
TP (mg/L)	0.7	3.3 ± 0.07	3.2 ± 0.06	3.7	1.6 ± 0.03	2.3 ± 0.04
(mg/L-P)	< 0.48	0.7	2.5 ± 0.1	1.6	1.0 ± 0.06	0.9 ± 0.05
TSS (mg/L)	19 ± 1	19 ± 1	14 ± 1	12	15 ± 1	10.3 ± 0.7
VSS (mg/L)	17	17	13.9	12	15 ± 1	<10
pH	7.1	7.2	6.9	6.8	7.3	7.5
Average pH	7.1
T (°C)	22	21.4	21.3	21	21.5	21.6
Average T (°C)	21.5

View Large

Unlike the TN, the TP was affected in its removal as it dropped considerably from 80.8% in the first sample to 0% in the third sample (see Table 2). Phosphorus removal is dependent on the anoxic stage and the amount of carbon present, and when observing Figure 4, the DO for the third sampling was the highest reported, slightly above 2 mg/L, agreeing with what was observed by several authors wherein increased DO in the reactor was related with the rainy season or with low organic load in the system due to dilution, both impacting the removal of phosphorus (Ittisupornrat et al. 2021). The suspension of activities in the hotel as a result of the COVID-19 pandemic, where the influent flow was reduced to an average value of 0.72 L/s (see Table 1), may have affected this phenomenon.

The average pH in the influent was 6.8 and 7.1 in the effluent, thus very stable and complying with the Colombian discharge standards (pH between 6 and 9). The removal percentages in COD_Tot varied between 85.6 and 91.7%, which are considered high although values up to 95% had been reported in the system stabilization process. Regarding nitrogen, the removal ranged between 46.3 and 90.0%, values considered adequate for nitrogen removal in this AS reactor, including during post-COVID-19 pandemic conditions which generated variability in flow rate and COD due to the decrease in numbers of guests at the hotel. During the pandemic, the system operated for over a year without receiving wastewater or aeration, with retained biomass under anoxic and even anaerobic conditions, triggering autolysis and cell death. This led to phosphorus solubilization, low P removal, and the release of organic compounds and ammoniacal nitrogen, acting as an internal load within the reactor. Additionally, the presence of ELEA conditions (favoring nitrifier development) and AI (enhancing anoxic denitrification) resulted in high nitrogen removal post-pandemic. By June 2021, P removal resumed, and N removal stabilized, possibly due to microbial community restructuring toward a more specialized composition, dominated by nitrifying and denitrifying bacteria.

Figures 2 and 3 show how the bioreactor can remove organic matter and nitrify with low DO contents, with values between 0.1 and 0.5 mg/L measured in most of the samples. In the sample from 12 April 2021 (Figure 4), the DO peaked at 2.24 mg/L, when the hotel was just beginning to resume activities after the isolation mandated by the national government due to the COVID-19 pandemic, and when there was a lower number of guests and thus the lowest flow rate (see Table 1). We saw that the ORP in the reactor can reach values greater than 360 mV and can drop below 100 mV. It should be noted that, with such a short shutdown time of 15′, the ORP can drop between 50 and 100 mV units, permitting a high amount of nitrogen removal in this anoxic stage. Regarding pH, there is stability in the bioreactor, with values generally between 6.8 and 7.0, which helps positively in this treatment.

The very low concentration of DO and high efficiency in removing organic matter and nitrogen are in agreement with what was reported in several studies from 2016 to 2021, among which three (Gogina & Gulshin 2016; Singh et al. 2018; Ittisupornrat et al. 2021) stand out. These confirmed that regardless of the treatment train used, whether Batch type in the same tank or in separate tanks, high concentrations of DO are not necessary.

The AS bioreactor had an average VSS of 2,661 mg/L, with a maximum of 4,070 mg/L and a minimum of 2,040 mg/L. This average value was below that reported in the design (3,703 mg/L), which means greater recirculation to increase the amount of biomass could be used or that there is a greater capacity to accommodate a higher microbial population. This could be because the average recirculation flow rate was 68% due to failures in the pumping system in most of the samples. The VSS/TSS ratio is 98% and the SedS were not kept in the range 450–650 mL/(L·h), with an average of 392 mL/(L·h), a maximum of 650 mL/(L·h), and a minimum of 250 mL/(L·h). This minimum value occurred because of conditions caused by the COVID-19 pandemic, which led to a drastic decrease in wastewater production. The pH was very stable with an average of 7.

We used the following plant design parameters and the data collected in the investigation cycle, with an average reactor temperature of 20.3 °C, for the equations used to develop the proposed EHSA model, which is presented in Appendix A.

K_m is a 0.025 affinity or saturation coefficient, mg/L COD; K_e is a 0.05 endogenous respiration coefficient, 1/day; K_o is a 5 maximum substrate removal rate, mg COD/mg VSS-day; Y is a 0.5 coefficient of yield, mg VSS/mg COD; K_c is a 0.02 Contois saturation constant, mg COD/mg VSS; S_o is a COD_Inf influent substrate for each cycle, mg/L; X is aMLVSS biomass concentration in the reactor, mg/L MLVSS; V is a 81.1 Reactor volume, m³; X_e is the VSS_Ef biomass concentration in the secondary settler effluent, mg/L; Q is an average flow rate to be treated, L/s or m³/s; Q_w is an excess sludge flow rate from the reactor, L/s or m³/s; t_d is the hydraulic retention time, h or d.

It should be noted that the value of (X*t_d) will depend on the value of the substrate S_o. In other words, for a certain value of S_o, S depends solely on t_d, which means that the lower the t_d, the higher the X. In Table 3, when the t_d is lowered to 10 h, the biomass, X, increases. Therefore, if a considerably high biomass X is available, the t_d can be significantly reduced, which can have advantageous consequences in the design of wastewater treatment systems by minimizing the volume of the reactors.

We observed an average θ_c of 180 days, with this being highly affected by the MLVSS and very sensitive to both the variation of the flow rate to be treated and the VSS_Ef. On the other hand, despite differing from the biomass actually measured in the reactor, itself an important determinant of the t_d present, a good nitrogen removal efficiency was seen with a θ_c between 84 and 382 days, and with an insignificant sludge production in the wastewater treatment plant of the Movich Las Lomas hotel. This takes into account the capacity and design of the secondary settler because it has sufficient conditions to receive the maximum biomass that can be generated in the bioreactor. This situation was corroborated in the period 2018–2021 with the analysis of the VSS_Ef, which is equivalent in load to 0.4% of the biomass generated in the bioreactor. This is consistent with the theoretical aspect developed by Orozco (2014) regarding EHSA and other studies where the influence of θ_c has been considered, such as Ittisupornrat et al. (2021), which achieved good results in the removal of organic matter and nutrients with sludge ages between 150 and 300 days.

This trend means that the biomass produced is equal to the biomass consumed in the endogenous phase, which can make the EHSA condition applicable to AS reactors, requiring only that appropriately designed secondary settlers be used.

It is important to remember that biological phosphorus removal is based on purging sludge that contains appreciable phosphorus concentrations. In this case study, we observed high variability in TP removal. Upon analysis of the mechanism of this process, since sludge is not purged from the reactor in a considerable way and since there is a low DO in the bioreactor, the variation that occurs in its removal in this studied cycle is understandable because at some moments it is assimilated or absorbed in the biomass and at other times it is being released or solubilized in the liquid, precisely due to the high age condition of the sludge. However, it is worth highlighting some justifications that have been given in the literature, regarding the possibility of removing P in a treatment train without an anaerobic stage. In the removal of biological phosphorus, there are two processes: the release of P and the intracellular absorption of P. The release of phosphorus occurs in the anaerobic phase together with the absorption of an external carbon source and the formation of polyhydroxyalkanoates (PHA) or polyhydroxybutyrates (PHB) via the consumption of volatile fatty acids (VFA). This process requires energy that is provided by the fragmentation of poly-P; that is, by the release of phosphorus. Then in the aerobic or anoxic phase, the PHB or PHA stored internally is oxidized and is used for growth, phosphate absorption, and formation and maintenance of glycogen. The PAO has this function, but there can be strong competition with the Glycogen Accumulating Organisms (GAOs), noting that the latter do not release or absorb phosphorus while consuming considerable VFA, and thus lower the efficiency in P removal in A/O systems (Chen et al. 2013). This highlights the advantage of not having an anaerobic phase in treatments aimed at removing phosphorus.

While it is true that the P removal process has been carried out more in A/O systems, recent research has also studied anoxic-aerobic treatment trains for this same purpose. When analyzing the metabolism present there, it is observed that in the aerobic phase, almost all of the carbon source is consumed, with the energy required for carbon absorption and PHA formation supplied by the tricarboxylic acid cycle (Wang et al. 2014). In the anoxic period only a small amount of energy is used for cell maintenance (the polyphosphates formed do not release as much ⁠) and energy is mainly obtained by the hydrolysis of intracellular stored poly-P sufficient to satisfy this energy requirement during the anoxic period. Therefore, the need for PAO metabolism in anoxic-oxic systems can be satisfied and at the same time, as there is no VFA production, the problem of the existence of GAOs and their competition with PAO is eliminated leading to more stability in P removal, and addressing what has been the greatest difficulty in A/O systems (Chen et al. 2013; also corroborated by Jena et al. 2016 and Liu & Wang 2017, among others).

In this phosphorus removal process, DNPAO may also be present, which is capable of absorbing P and denitrifying in aerobic/anoxic processes. This requires that there is a sufficient carbon source so that PAO and DNPAO do not compete for it, and so that most of the COD is converted into intracellular PHB, permitting denitrification and phosphate absorption to occur under aerobic and anoxic conditions (Yin et al. 2015). This justifies the importance of carrying out other studies with the goal of improving phosphorus removal with 45′:15′ cycles in AS reactors as the only reactor. Additionally, the biomass in these reactors should be characterized to clarify whether PAO and DNPAO are present; research is already being developed by the authors of this article.

With IA in a 45′:15′ cycle and EHSA in a single AS reactor and in a continuous flow, it is possible to eliminate nitrogen from DWW, with the guarantee of good COD removal.

The 45′:15′ cycle shows that denitrification can be achieved with only 15′ of anoxia. This cycle results in an equivalent of 25% energy savings per day (6 h per day in total). In other words, this ensures that there is sufficient nitrification, which provides the electron acceptors necessary for denitrification to occur in the anoxic phase.

It is important to highlight the significant usefulness of aerobic processes for the removal of ammoniacal nitrogen from wastewater. However, these processes require high energy consumption and generate large volumes of sludge. The application of IA addresses the issue of TN removal, while Excess Sludge Elimination through Aeration (EHSA) significantly reduces the discharge of excess sludge. This makes the approach particularly suitable for developing communities, as it only requires the installation of a simple on/off control system and the analysis of the secondary clarifier to verify its operational capacity. This strategy facilitates progress toward tertiary treatment in DWW plants in a straightforward manner, without the financial burden associated with most current advanced technologies.

Other advantages of this application are: no additional carbon sources are needed for denitrification, as required by systems in separate tanks and problems of competition between microorganisms can be avoided.

Since the ORP is so important in this type of reaction, it is reasonable to look for a correlation between ORP and the removal of nitrogen and phosphorus, especially as there was a limitation in this experiment in relation to the correspondence over the time intervals used, likely caused by composite nutrient sampling in 24 h while measurements of the redox potential were monitored every minute during an approximately 2 h interval.

Regarding phosphorus, although the variability (coefficient of variation = 89%) was not acceptable with removal values from the six samples ranging from 0 to 80.8%, these data suggest that it is worth carrying out more studies in this system to further understand and improve its removal, including the importance of variation in organic load, N and P in the influent (Mielcarek et al. 2015), amount of DO, microbial populations and recirculation (Gogina & Gulshin 2016).

We would like to thank the Ministry of Science, Technology and Innovation-MINCIENCIAS of the national government of Colombia for the financial support provided to this research, under call for proposals #727 of 2015.

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

The authors declare there is no conflict.