Abstract

In this study, the influence of the anaerobic mixed feeding rate on granule stability and reactor performance in a conventional sequencing batch reactor (C-SBR) was investigated while treating various industrial wastewaters. A laboratory-scale SBR fed with malting wastewater rich in phosphorus was operated for approximately 250 days, which was divided into two periods: (I) mixed pulse feed and (II) prolonged mixed feed. Initially, no bio-P activity was observed. However, by lowering the feeding rate biological P-removal was rapidly established and no effect on the aerobic granular sludge (AGS) characteristics was observed. Additionally, to investigate the effect of the mixed feeding rate when treating an industrial effluent with low phosphorus content, i.e. brewery wastewater, a laboratory-scale reactor was operated for approximately 400 days applying different mixed feeding rates. Morphological and molecular analysis indicated that a low substrate concentration promoted the enrichment of anaerobic carbon storing filaments when fed with brewery wastewater. Findings suggest that a prolonged mixed feeding regime can be used as a tool to easily establish bio-P removal in a C-SBR system for the treatment of phosphorus-rich wastewaters. It should however be considered that under P-limiting conditions, enrichment of poly-P storing filaments may occur, possibly due to their higher substrate affinity under anaerobic conditions.

INTRODUCTION

The European brewing sector is a major contributor to the European economy, providing up to 2.3 million direct and indirect jobs. The agricultural malting industry provides products and serves the beer sector. Europe is responsible for one third of the worldwide malt production resulting in increased employment within the overall beer sector (Europe Economics & The Brewers of Europe 2016). Globally, efforts are made towards reducing water usage and carbon footprint. Brewery wastewater typically contains high amounts of easily biodegradable compounds (Driessen & Vereijken 2003), while wastewater produced by malting activities typically contains high amounts of particulate matter (Schwarzenbeck et al. 2004). Successful full-scale applications of the aerobic granular sludge (AGS) technology have been reported for the treatment of municipal wastewater, resulting in a reduction of energy consumption of between 20% and 40% and a footprint reduction of up to 33% (Pronk et al. 2015). Due to biomass with excellent settling characteristics, sludge separation is improved drastically, resulting in systems with higher biomass concentrations. In addition, simultaneous nitrification/denitrification (SND) and biological phosphorus removal contribute to the design of compact AGS bioreactors (STOWA 2013). It is generally considered as an attractive technology for the treatment of various industrial wastewaters. However, research using complex and/or particulate wastewaters such as real municipal wastewater (STOWA 2013) and industrial effluents (Schwarzenbeck et al. 2004; Caluwé et al. 2017; Stes et al. 2018) show that more complex substrates may lead to filamentous outgrowth on the granule surface and the co-existence of flocculant and granular sludge. Stable aerobic granule formation when treating industrial wastewaters with a particulate content still remains challenging. The enrichment of slow growing organisms such as glycogen accumulating organisms (GAOs) and phosphate accumulating organisms (PAOs) appears to be critical to obtain stable aerobic granulation (STOWA 2013). PAOs and GAOs have the ability to anaerobically convert volatile fatty acids (VFAs) into intracellular storage polymers, which are used for microbial growth during aeration. The energy source for anaerobic VFA uptake is different for both groups of organisms; for PAOs this energy originates primarily from the hydrolysis of the intracellularly stored poly-P. For GAOs, the main energy source results from the degradation of intracellular glycogen (Oehmen et al. 2007). Tu & Schuler (2013) found that the in-reactor substrate concentration influenced the PAO-GAO competition, favouring PAO over GAO when the mixed feeding rate was decreased (synthetic wastewater). It is suggested that PAOs are able to apply active transport for anaerobic carbon uptake, giving PAOs a competitive advantage when in-reactor substrate concentrations are low (Tu & Schuler 2013). Accordingly, growth of GAOs is negatively influenced by lowering the feeding rate, pointing out the importance of excess P to allow PAOs to proliferate under these operational conditions. Based on these findings, it should be considered that a high chemical oxygen demand/phosphorus (COD/P) ratio of the brewery wastewater will prevent proliferation of granule forming PAOs due to limited P availability. At the same time a lower feeding rate will prevent GAO growth due to an increased energy demand for anaerobic carbon uptake possibly hampering enrichment of granule forming organisms. The COD/P ratio may therefore have a substantial impact on the overall granule characteristics when the feeding rate is lowered. Short anaerobic pulse feeding strategies (several minutes) are mainly tested in laboratory-scale experiments to promote granulation (Val del Río et al. 2012; Caluwé et al. 2017; Stes et al. 2018). Applying this pulse feeding strategy is technically unfeasible in industrial sequencing batch reactor (SBR) systems due to the enormous required feeding flows. Inevitably, existing or newly built wastewater treatment plants (WWTPs) will maintain longer mixed feeding times. This should be considered as an important operational parameter because of the direct impact on the in-reactor substrate concentration during the anaerobic feeding phase. It is of high relevance to investigate the influence of a prolonged anaerobic mixed feeding rate on the overall aerobic granule stability, reactor performance and the bio-P removal activity in an AGS system. It is assumed that three operational factors will be influencing the GAO/PAO competition, i.e. (1) the COD/P ratio (Oehmen et al. 2007), (2) the pH (Filipe et al. 2001) and (3) the anaerobic mixed feeding rate (Tu & Schuler 2013). As described by Oehmen et al. (2007) it is generally assumed that a high influent COD/P ratio favours growth of GAO over PAO due to the limited availability of phosphate. This study focusses on two main objectives. The influence of the anaerobic mixed feeding rate on the AGS stability, reactor performance and bio-P activity was investigated while treating (1) an industrial wastewater with a high phosphorus content, i.e. malting wastewater, and (2) low phosphorus content, i.e. brewery wastewater. It is hypothesised that when lowering the feeding rate, the bio-P removal activity will increase when COD/P ratios are low and PAOs are able to proliferate. However, deterioration of the granular sludge characteristics may occur when the feeding rate is lowered while treating a wastewater characterised by a high COD/P ratio.

MATERIAL AND METHODS

Reactor set-up

Two fully automated laboratory-scale SBRs were used during this study. The anoxic SBR fed with malting wastewater, SBRM, had a working volume of 15 L (H/D = 3.5) and was operated at room temperature (18–22 °C) for approximately 200 days. The aerobic SBRB, fed with brewery wastewater had a working volume of 13 L (H/D = 1.1) and was operated at room temperature (18–22 °C) for 400 days. Each reactor was provided with a mechanical stirrer (IKA RW20 digital), a dissolved oxygen (DO) (Endress + Hauser, Oxymax W COS51D) and pH (Endress + Hauser, Orbisint CPS11–7AA21) sensor and an aeration system consisting of an aeration pump (Ubbink Air 1000) and an air diffuser at the bottom of the reactor (AngelAqua DY 104-A). A Siemens PLC (LOGO! Logic Module) was used for process control. Sensor data were recorded and visualised by an Ecograph T RSG35 graphic display recorder (Endress + Hauser). Monitoring of the DO set-points was done by the same recorder. A schematic overview of the experimental set-up can be found in supplementary data I (available with the online version of this paper).

Industrial wastewaters and seed sludge

The SBRM was fed with wastewater from a local malting company. The wastewater used for SBRB originated from a local brewery/bottling plant. Sufficient nutrient availability in the brewery wastewater was ensured by manual dosage of nitrogen (urea, 30%) and phosphorus (phosphoric acid, 75%) resulting in a final COD:N:P ratio of approximately 100:2:0.5. To avoid clogging of the feeding tubes, all wastewater was sieved (pore size: 1 mm) to remove grains and large particulate matter. Minimum, maximum and average influent concentrations are summarised in Table 1.

Table 1

Malting and brewery (incl. N and P dosage) wastewater composition

  CODt (mg O2·L−1CODs (mg O2·L−1TN (mgN.L−1TP (mgP.L−1COD/P pH EC (μS/cm) 
Malting 
 Min 1,256 1,079 59.6 19.8 36 5.50 1,668 
 Max 3,661 3,350 118 47.2 165 7.07 3,140 
 Av 2,403 1,964 85.4 33.7 76 6.38 2,289 
 Stdev 491 443 12.6 7.54 30 0.44 357 
Brewery 
 Min 1,768 1,046 23 12.7 90 4.9 802 
 Max 7,934 6,430 166 43.7 433 7.2 3,280 
 Av. 4,473 3,663 77 24.6 200 5.9 1,949 
 Stdev 1,212 1,160 24 7.7 82 0.6 452 
  CODt (mg O2·L−1CODs (mg O2·L−1TN (mgN.L−1TP (mgP.L−1COD/P pH EC (μS/cm) 
Malting 
 Min 1,256 1,079 59.6 19.8 36 5.50 1,668 
 Max 3,661 3,350 118 47.2 165 7.07 3,140 
 Av 2,403 1,964 85.4 33.7 76 6.38 2,289 
 Stdev 491 443 12.6 7.54 30 0.44 357 
Brewery 
 Min 1,768 1,046 23 12.7 90 4.9 802 
 Max 7,934 6,430 166 43.7 433 7.2 3,280 
 Av. 4,473 3,663 77 24.6 200 5.9 1,949 
 Stdev 1,212 1,160 24 7.7 82 0.6 452 

All influent and effluent concentrations were measured using Hach test kits (Mechelen, Belgium); total COD (CODt) and soluble COD (CODs): LCK014, LCK514; TN-N: LCK 338, LCK138; NH4+-N: LCK305, NO2N: LCK342, NO3N: LCK339, TP-P: LCK350 and LCK348, respectively. All samples were filtered using glass microfibre filters (particle retention: 0.6 μm, Macherey-Nagel MN GF-3) before measuring the CODs, PO43−-P, NH4+-N, NO2-N and NO3-N concentrations. Seed sludge for SBRM originated from a parent laboratory-scale SBR which was operated to promote aerobic granulation while treating malting wastewater. An overview of the operational parameters of the parent laboratory-scale SBR can be found in supplementary data II (available with the online version of this paper). SBRB was seeded with flocculant sludge from the existing local WWTP treating the brewery wastewater.

Reactor operation

In this study, aerobic granulation was promoted by applying a metabolic selection pressure to enhance growth of specific groups of slow growing micro-organisms associated with successful AGS formation, i.e. PAOs and GAOs. Therefore, an anaerobic feast/aerobic famine regime was applied for both SBR systems which is known to favour growth of granule forming organisms over floc forming and/or filamentous organisms. The SBRM was operated for approximately 200 days, divided into two periods based on changes in anaerobic mixed feeding rate. The SBRB was operated for approximately 400 days, during which the feeding rate was lowered on day 182. To quantify the difference in feeding rate from a substrate loading point of view, the ratio of the organic loading rate (OLR) to the feeding time per cycle, i.e. OLRfeeding, was calculated using the subsequent formula:  
formula

This parameter is introduced to compare different anaerobic mixed feeding rates since it will result in different in-reactor substrate concentrations. An overview of the adjustments in SBR operation during the experiment is shown in Table 2.

Table 2

Different operational conditions considering the SBR cycle

Duration of SBR cycle phase SBRM
 
SBRB
 
Period I Period II Period I Period II 
Day 1–85 Day 86–200 Day 1–181 Day 182–400 
Mixed idle phase (min) 10 10 10 10 
Total anaerobic phase (min) 90 90 120 120 
Aerobic phase (min) 155 205 205 195–205 
Anoxic phase (min) 80 30 
Settling (min) 15 15 15 15–25 
Effluent withdrawal (min) 10 10 10 10 
Total cycle time (min) 360 360 360 360 
Feeding time (min) 10 ± 4 49 ± 19 22 ± 8 41 ± 13 
FT/TTa (%) 4 ± 1 16 ± 3 6 ± 2 13 ± 4 
Duration of SBR cycle phase SBRM
 
SBRB
 
Period I Period II Period I Period II 
Day 1–85 Day 86–200 Day 1–181 Day 182–400 
Mixed idle phase (min) 10 10 10 10 
Total anaerobic phase (min) 90 90 120 120 
Aerobic phase (min) 155 205 205 195–205 
Anoxic phase (min) 80 30 
Settling (min) 15 15 15 15–25 
Effluent withdrawal (min) 10 10 10 10 
Total cycle time (min) 360 360 360 360 
Feeding time (min) 10 ± 4 49 ± 19 22 ± 8 41 ± 13 
FT/TTa (%) 4 ± 1 16 ± 3 6 ± 2 13 ± 4 

aFT/TT: anaerobic mixed feeding time to total cycle time ratio.

For both reactors, a relatively constant F/M ratio of 0.15 kgCOD.(kgMLSS.day)−1 was applied by adjusting the feeding time with respect to the influent COD concentration and the mixed liquor suspended solids (MLSS) concentration. As a result, the ratio of the SBR anaerobic mixed feeding time to the total cycle time (FT/TT) varied during the experiment. For both laboratory-scale experiments, the DO concentration was maintained between 1.0 and 1.5 mgO2.L−1 using an on/off aeration control strategy, and no pH control was applied. For SBRM and SBRB, the sludge retention time was kept constant at 43 and 63 days by the automatic removal of 353 mL and 380 mL, respectively, of the sludge mixture during effluent withdrawal.

Sludge characteristics and sludge staining

All sludge samples used for analyses were taken at the end of the anoxic (SBRM) or aerobic (SBRB) cycle phase. The MLSS concentration was measured by filtering 5 mL of homogeneous sludge mixture over a glass microfibre filter, which was subsequently washed with demineralised water and dried for 24 h at 105 °C. The sludge volume (SV) was determined as described by APHA/AWWA/WEF (1998). The median particle size distribution by volume, DV50, was measured using a Malvern Mastersizer 3000 (Malvern, UK) as described by Stes et al. (2018). Weekly analysis of the sludge was performed to investigate the evolution of the sludge morphology using a CX21FS2 Olympus microscope. Gram (modified Hücker method) and Neisser stainings were performed according to the methods described by Jenkins et al. (2004) in the attempt to identify specific filamentous organisms present in the system.

Microbial community composition by 16S rRNA gene amplicon sequencing

Genomic DNA was purified according to the NaTCA method (McIlroy et al. 2008) from sludge samples taken in triplicate from the reactor at regular time intervals. A sequencing library pool targeting the V1-3 region of the 16S rRNA gene was generated as described by Karst et al. (2016), with minor modifications. Briefly, Phusion High-Fidelity DNA polymerase (Thermo Scientific) and barcoded primers (IDT) were used for library polymerase chain reaction (PCR) with 10–20 ng of DNA as template, and the denaturation step was carried out at 98 °C. The resulting library pool was submitted to a final purification step by gel extraction using NucleoSpin Gel and PCR Clean-up (Macherey Nagel), and diluted to obtain a 4 nM library pool. Amplicon sequencing was carried out on an Illumina MiSeq system at the Centre for Medical Genetics (Edegem, Belgium) with the MiSeq Reagent Kit v3 (Illumina). The obtained paired-end reads were processed with the UPARSE pipeline (Edgar 2013). As a reference database for taxonomy prediction, MiDAS (version 2.1) was used, which is a manually curated SILVA 16S rRNA taxonomy (release 1.23 Ref NR99) that proposes a name for all the abundant phylum- and genus-level taxa present in activated sludge, anaerobic digesters and influent wastewater (McIlroy et al. 2015).

In-situ cycle measurements during SBR operation

Due to the SBR operational strategy and subsequently the presence of granule structures, the enrichment of GAO/PAO like organisms is expected. In-situ cycle measurements during SBR operation are performed to determine the anaerobic carbon uptake and the degree of anaerobic phosphorus release and the aerobic phosphorus uptake rate. This was done to investigate the impact of the feeding rate on the GAO/PAO competition, favoring PAOs when the feeding rate was lowered. To determine the carbon and phosphate profiles, grab samples were taken (1) before and (2) after feed, (3) before aeration, (4) during aeration and (5) at the end of the SBR cycle. Grab samples were filtered immediately followed by the analytical measurements. From these results, the anaerobic carbon uptake [%], anaerobic P release [mg P.g MLSS−1] and aerobic P uptake rates [mg P·(g MLSS.h)−1] were calculated.

RESULTS AND DISCUSSION

Low COD/P malting wastewater

Reactor performance

The feed to mass ratio (F/M) was kept relatively constant by adjusting the feeding volume in function of the influent COD concentration and the MLSS concentration. For SBRM the average F/M ratio was 0.15 ± 0.03 kg COD.(kg MLSS.day)−1. Due to changes in MLSS concentration and the aim to work at a constant F/M ratio, the organic loading rate (OLR) varied strongly, with an average OLR of 0.99 ± 0.43 kg COD.(m3.day)−1. To investigate the influence of the anaerobic feeding rate on the sludge characteristics and metabolic processes associated with the presence of slow-growing organisms, the feeding rate was lowered, resulting in an increase in the average anaerobic mixed feeding time from 10 ± 4 min up to 49 ± 19 min. This adjustment in feeding regime implicates a lower in-reactor substrate concentration during the anaerobic mixed feeding phase. The average OLRfeeding was 0.69 ± 0.18 and 0.34 ± 0.09 kgCOD(m3.h)−1 during period I and II, respectively.

The COD, N and P removal efficiencies were calculated for both operational periods, showing stable carbon and nitrogen removal efficiencies throughout the experiment. Nitrogen removal was obtained through the two-step nitrification/denitrification process, showing high removal efficiencies during the complete experiment. Effluent total nitrogen (TN) concentrations were consistently below 15 mg N.L−1 (the Belgian discharge limit) with NO3 as the main effluent nitrogen compound with a maximum concentration of 5.1 mg NO3-N.L−1. Introduction of the prolonged feeding strategy did not have any impact on the N removal process. However, the adjustment of the feeding pattern had a major impact on the P removal efficiency due to the rapid increase of bio-P activity during period II. As can be seen in Figure 1, high effluent phosphate concentrations (>20 mg PO43–-P.L−1) were measured during period I while during period II the effluent phosphate concentrations were consistently below 2 mg PO43–-P.L−1. The average P removal efficiency was 26 ± 13% during period I, which is remarkably low compared to 89 ± 8% during period II. For SBRM, it was assumed that phosphorus removal may not occur during operation, especially during period I (short feeding time) due to the competitive advantage for anaerobic carbon uptake by GAOs. The results considering the absence of bio-P activity during period I confirm the findings by Filipe et al. (2001) that GAOs show a competitive advantage for anaerobic carbon uptake when pH values are below 7.25. Additionally, the strong increase of bio-P removal when the anaerobic mixed feeding was prolonged is in line with results described by Tu & Schuler (2013). They found that lowering the feeding rate in a laboratory-scale SBR fed with synthetic wastewater led to a shift in the GAO/PAO competition, favouring PAOs over GAOs. This was explained by the fact that PAOs contain an additional energy source for anaerobic carbon uptake, i.e. internally stored poly-P. Our results seem to confirm the statement made by Liu et al. (1997) that the PAO/GAO competition is an energy based competition for anaerobic carbon uptake which tends to favour poly-P containing PAOs over GAOs when in-reactor substrate concentration are low. To gain more insight in the shift in metabolic processes occurring within the system, carbon and phosphate concentrations were profiled by the use of in-situ cycle measurements.

Figure 1

Evolution of the COD, nitrogen and phosphorus removal efficiencies and effluent phosphorus concentrations for SBRM (dotted line: introduction of prolonged feeding phase).

Figure 1

Evolution of the COD, nitrogen and phosphorus removal efficiencies and effluent phosphorus concentrations for SBRM (dotted line: introduction of prolonged feeding phase).

In-situ cycle measurements

In-situ cycle measurements were performed showing minor anaerobic phosphate release during period I while anaerobic COD uptake from 74 up to 94% was observed. Together with the sludge settling characteristics and morphology (see further), the results indicate the presence of GAO like organisms in the system during period I. During period II, anaerobic carbon uptake varied between 74% and 89% while a drastic increase of the anaerobic P-release and aerobic P-uptake rates was observed, as can be seen in Table 3.

Table 3

Evolution of the anaerobic P-release and aerobic P-uptake rate for SBRM

  Day Anaerobic P-release [mg P.(g MLSS)−1Aerobic P-uptake rate [mg P.(g MLSS.h)−1
Period I 1.19 0.36 
20 0.61 0.41 
34 1.27 0.60 
Period II 96 6.12 3.16 
159 5.99 9.42 
196 5.71 5.70 
  Day Anaerobic P-release [mg P.(g MLSS)−1Aerobic P-uptake rate [mg P.(g MLSS.h)−1
Period I 1.19 0.36 
20 0.61 0.41 
34 1.27 0.60 
Period II 96 6.12 3.16 
159 5.99 9.42 
196 5.71 5.70 

Together with the increased phosphorus removal efficiencies up to 98%, it can be concluded that lowering the feeding rate has induced the enrichment of PAOs over GAOs in the system. Our results confirm the findings of Tu & Schuler (2013) that applying a low feeding rate promotes bio-P activity and seems to be critical. This is, to our knowledge, the first time that the influence of the feeding time on the bio-P activity is investigated while treating an industrial wastewater in a C-SBR.

Granule characteristics

Evolution of the granule settling characteristics, DV50 and sludge morphology were investigated to determine the influence of the feeding regime on the overall granulation state and stability. During the first period of the experiment, the settling characteristics showed SVI values below 30 mL.g−1. In addition, the average DV50 was 209 ± 3 μm suggesting good selection for granule forming organisms. Meanwhile, filamentous outgrowth at the granule surface gradually decreased.

Granule settleability only slightly decreased during period II, resulting in final SVI10 and SVI30 values of 78 mL.g−1 and 47 mL.g−1, respectively on day 200 of the experiment. These are still considered as values representing good settling granular sludge. Lowering the anaerobic feeding rate did not influence the overall granule morphology as shown in Figure 2. However, it can be observed that smaller granular structures developed during period II. The final DV50 was only 153 ± 1 μm on day 189. This phenomenon can be explained by the fact that lower in-reactor substrate concentrations may lead to substrate diffusion limitations and subsequently the development of smaller but still compact granule structures. It can be concluded that lowering the feeding rate from 10 ± 4 min to 49 ± 19 min had no major impact on the granule settling or morphological characteristics. However, the feeding rate had a major impact on the reactor performance considering the increased bio-P activity during period II of the experiment. These results point out that applying a low anaerobic mixed feeding rate may be an easily applicable operational strategy to promote biological phosphorus removal in an AGS system for the treatment of malting wastewater in a compact C-SBR. Considering all results, a low in-reactor substrate concentration is thought to select for (smaller) PAO-like granules over GAO, which is assumed to be due to the presence of an additional energy source, i.e. poly-P, for anaerobic carbon uptake. This, however, raises some questions concerning the applicability of a low feeding rate for the treatment of wastewaters without excess phosphorus, like brewery wastewater.

Figure 2

Overview of the sludge morphology for SBRM during period I and period II.

Figure 2

Overview of the sludge morphology for SBRM during period I and period II.

High COD/P brewery wastewater

Reactor performances

For this experiment, a laboratory-scale SBRB treating brewery wastewater was operated for approximately 400 days, divided into two periods based on the anaerobic mixed feeding rate. The average F/M ratio and OLR were 0.20 ± 0.09 kg COD.(kg MLSS.day)−1 and 0.93 ± 0.42 kg COD.(m3.day)−1, respectively. On day 182, the mixed feeding rate was decreased resulting in an increase in the average feeding time from 22 ± 8 min during period I up to 41 ± 13 min during period II. Consequently, the average OLRfeeding decreased from 0.78 ± 0.18 to 0.25 ± 0.07 kg COD.(m3.h)−1. Stable COD, N and P removal efficiencies of 98 ± 1%, 99 ± 1 and 96 ± 4%, respectively, were obtained throughout the experiment. When comparing period I and II, the overall removal efficiencies were stable and the differences between both periods were found to be negligible.

In-situ cycle measurements

Throughout the experiment, in-situ cycle measurements were performed to evaluate the influence of the feeding rate on the bio-P activity; that is, the anaerobic P release and the aerobic P uptake rate. When treating brewery wastewater, the in-reactor pH is expected to be above 7.25 promoting minor bio-P activity as described by Stes et al. (2018). Since the pH is known to play a key role in the competition between GAO/PAO (Filipe et al. 2001), this parameter was taken into account during the study. The average pH was 8.1 ± 0.1 during a typical SBR cycle (day 366), which is generally known to favour PAOs over GAOs. As can be seen in Table 4, an increase in the anaerobic P-release was observed during period I. These results suggest that even though the COD/P ratio of the influent was high, minor enrichment of some granule forming PAO-like organisms may have occurred during period I. These findings are in line with the results reported by Stes et al. (2018) where stable aerobic granules showed minor bio-P activity when the pH > 7.2 while treating brewery wastewater. Due to the alkaline conditions, it is believed an additional energy source is required to overcome the pH gradient across the cell membrane in order to take up carbon anaerobically. Since PAOs can provide this additional energy requirement through poly-P degradation, they tend to have a competitive advantage over GAOs (Filipe et al. 2001). It is clear that the in-reactor pH can not be neglected in an anaerobic-aerobic operated SBR system.

Table 4

Evolution of the anaerobic P-release and the aerobic P-uptake rate for SBRB

  Day Influent COD/P [%] Anaerobic P-release [mg P.g MLSS−1Aerobic P-uptake rate [mg P·(g MLSS·h)−1
Period I 117 0.29 0.20 
15 189 0.47 0.36 
29 109 0.36 0.17 
34 138 0.15 0.08 
43 138 0.20 0.13 
57 135 0.38 0.15 
119 204 1.48 – 
142 92 1.68 0.37 
162 195 2.79 1.26 
182 126 3.15 2.54 
Period II 211 270 1.34 1.98 
247 226 3.39 – 
283 370 1.49 1.98 
289 199 2.14 1.80 
379 222 1.50 1.67 
393 177 0.88 1.87 
400 190 1.88 2.48 
  Day Influent COD/P [%] Anaerobic P-release [mg P.g MLSS−1Aerobic P-uptake rate [mg P·(g MLSS·h)−1
Period I 117 0.29 0.20 
15 189 0.47 0.36 
29 109 0.36 0.17 
34 138 0.15 0.08 
43 138 0.20 0.13 
57 135 0.38 0.15 
119 204 1.48 – 
142 92 1.68 0.37 
162 195 2.79 1.26 
182 126 3.15 2.54 
Period II 211 270 1.34 1.98 
247 226 3.39 – 
283 370 1.49 1.98 
289 199 2.14 1.80 
379 222 1.50 1.67 
393 177 0.88 1.87 
400 190 1.88 2.48 

In addition, results show that anaerobic P release as well as the aerobic P-uptake rates are somewhat higher during period II compared to period I. These results illustrate that when treating brewery wastewater with a high COD/P ratio, the anaerobic mixed feeding rate has no significant influence on the bio-P activity. These findings are less clear compared to those in SBRM, which showed a strong increase in the bio-P activity when the feeding rate was decreased. It is suggested that during this experiment, next to the pH and the feeding rate, the high COD/P ratio had a major contribution to the competition between GAO and PAO-like organisms, possibly preventing PAOs proliferating.

Sludge characteristics

The evolution in granule settleability, granular size and morphology was investigated to determine the influence of the feeding regime on granule characteristics and stability when treating brewery wastewater. When comparing the sludge settling characteristics, results show a significant difference between the two operational periods. Figure 3 shows the evolution of the DV50 and (D)SVI values during the 400 days of operation.

Figure 3

Evolution of DV50 and (D)SVI10,30 for SBRB. (dotted line: introduction of prolonged feeding time).

Figure 3

Evolution of DV50 and (D)SVI10,30 for SBRB. (dotted line: introduction of prolonged feeding time).

When comparing the (D)SVI values, it is clear that selection of well-settling sludge was established during period I, resulting in a decrease in (D)SVI values below 50 mL.g−1. The prolonged mixed feeding strategy was introduced on day 184. Hereafter, the (D)SVI10 values increased strongly up to 322 mL.g−1 on day 197. Subsequently, the settling characteristics drastically deteriorated, often resulting in very high SVI30 values above 200 mL.g−1 during period II. It is clear that increasing the FT/TT ratio had a major negative impact on sludge settleability when treating brewery wastewater. Additionally, it became clear that during period II the settling characteristics deteriorated rapidly when the difference between the influent CODt and CODs concentration, i.e. ΔCOD, increased (data not shown). This is in line with findings reported by de Kreuk et al. (2010) that the presence of particulate matter in the influent has a negative influence on the overall AGS characteristics. However, the increase of influent ΔCOD had a greater impact on the settling characteristics during period II compared to period I. This suggests that lowering the feeding rate and thus the in-reactor substrate concentration had a negative impact on the overall stability of the sludge settleability. In Stes et al. (2018) successful aerobic granulation in an SBR treating brewery wastewater was reported showing very low SVI30 values (<50 mL.g−1). They applied a non-mixed (static) pulse feeding strategy to obtain a maximum substrate concentration during the subsequent anaerobic mixed feast phase. The brewery wastewater used by Stes et al. (2018) was also characterised by strong fluctuations of the particulate matter content but showed no significant effect on the sludge settling characteristics. It is clear that a high substrate gradient during the anaerobic feast phase had a positive impact on the sludge settling characteristics and the stability of the system towards variations in influent composition. Evolution of the sludge morphology was investigated by microscopic analysis and is shown in Figure 4.

Figure 4

Evolution of the sludge morphology during periods I and II from SBRB (including an example of a Neisser stained sludge sample).

Figure 4

Evolution of the sludge morphology during periods I and II from SBRB (including an example of a Neisser stained sludge sample).

It can be concluded that during period I dense granular structures developed and only a limited amount of filamentous organisms were present. Preservation of the large granules within the system was successful throughout period I. The morphology during period I is in strong contrast with period II where less dense, medium large granule structures, small flocs and filamentous organisms dominated the overall sludge morphology. These observations indicate that the increase of the (D)SVI values during period II are caused by the increased presence of smaller flocs and filamentous structures. In the attempt to identify the filamentous organisms and to confirm the presence of poly-P activity, i.e. intracellular poly-P granules, Neisser staining's were performed. As previously discussed, the in-situ cycle measurements showed bio-P activity within the system which was relatively high during period II considering the high COD/P ratio. Sludge samples were taken at the end of the aerobic phase to ensure a maximum poly-P content within the cells. Remarkably, no accumulation of poly-P was observed within the granular sludge structures while it was clearly present within the filamentous organisms (see Figure 4). Only a limited number of filamentous organisms are known to have the capacity to store poly-P granules. In addition, Gram-staining responses were positive for the filamentous structures (data not shown). Considering the SBRB operational strategy (anaerobic feast to promote carbon conversion into intracellular polymers), and based on the staining results and the filament morphological characteristics, only two specific filamentous organisms are suggested to be present in SBRB; that is, Thiothrix spp. or M. parvicella (Jenkins et al. 2004). Both species are known to be able to accumulate carbon intracellularly as poly-hydroxyalkanoates (PHA) and have the capacity to store phosphorus intracellularly (Rossetti et al. 2005; Rubio-Rincón et al. 2017). Molecular analysis by 16S-rRNA gene amplicon sequencing was performed in order to gain insight into the overall biomass composition of the seed sludge (sample 1) and additionally identify the filamentous organisms at the end of period II (sample 2). In supplementary data IIIa-b a detailed overview of the microbial composition at different taxonomical levels can be found (supplementary data III is available with the online version of this paper). For sample 1 and 2, up to 97% and 93%, respectively, of the resulting sequences could be classified at phylum level and up to 62% and 68%, respectively, at genus level. For sample 1, Planctomycetes and Bacteroidetes were the most abundant phyla representing 30% and 25% of all bacteria, respectively. For sample 2, Proteobacteria and Bacteroidetes were the most abundant phyla, reaching up to 45% and 39% of the bacteria, indicating a shift in microbial composition during this study. The read abundance for C. Accumulibacter (PAO), known as an anaerobic carbon storing organism associated with granule formation, slightly increased from 0.00 ± 0.04% in sample 1 up to 0.14 ± 0.05% in sample 2. For Defluviicoccus (GAO) the read abundance in the seed sludge was 0.71 ± 0.07% and 0.76 ± 0.06% at the end of period II. No enrichment of other known PAOs or GAOs was observed during this study (see supplementary data IIIc for the resulting read abundances for all known GAOs and PAOs). The suggestion, based on microscopic observations that M. parvicella may be present in the system was countered by the 16S rRNA amplicon sequencing analysis, indicating the complete absence in both sludge samples. However, the presence and strong enrichment of the Thiothrix genus (Proteobacteria) was confirmed by 16S rRNA amplicon sequencing analysis with an average read abundance of 0.0 ± 0.0% in the seed sludge and 7.3 ± 2.3% at the end of period II. The enrichment of Thiothrix in an anaerobic-aerobic granular sludge system was also observed by Stes et al. (2018) treating brewery wastewater in an anaerobically fed SBR. Since the presence of sulphur compounds was not taken into account during this study, measurements of the influent sulphur concentrations are absent. However, the anaerobic feeding strategy and the enrichment of Rhodobacter (Proteobacteria), known as a sulphate-reducing organism, indicate the presence of sulphur compounds in the brewery wastewater. The read abundance for Rhodobacter increased from 0.28 ± 0.03% in sample 1 up to 1.95 ± 0.28% in sample 2. By applying a prolonged anaerobic phase, sulphate reduction was found to be promoted and subsequently to induce growth of filamentous sulphur bacteria in an anaerobic/aerobic SBR system (Baetens et al. 2001). It is likely that enrichment of Thiothrix spp. was promoted by the prolonged anaerobic SBR phase, which was applied to promote anaerobic carbon uptake. Rubio-Rincón et al. (2017) showed that, in the presence of sulphide, Thiothrix caldifontis has the capacity to store carbon anaerobically as intracellular polymers and contribute to bio-P removal using stored poly-S as an additional energy source. These findings may explain our observations showing enrichment of poly-P storing filamentous organisms, i.e. Thiothrix, when applying an anaerobic/aerobic SBR operational strategy for the treatment of brewery wastewater. However, minor filamentous outgrowth was observed during period I of the experiment, suggesting growth of granule-forming organisms is promoted over filamentous organisms during period I. It is suggested that, like M. parvicella, other filamentous carbon storing organisms show a higher substrate affinity compared to granule-forming organisms when in-reactor concentrations are low during anaerobic mixed feeding. It is expected that sulphur compounds were present in the brewery wastewater. This complements the explanation of why more filamentous outgrowth, that is, Thiothrix, was observed during period II compared to period I due to an increased energy demand for anaerobic carbon uptake when in-reactor substrate concentrations decline.

SUMMARY AND CONCLUSION

The aim of this study was to investigate the influence of the anaerobic mixed feeding rate on the aerobic granule stability, reactor performance and bio-P activity while treating an industrial wastewater with a relatively low (malting) and high (brewery) COD/P ratio. In both cases, selection of carbon storing organisms was promoted by applying a feast/famine regime through anaerobic/aerobic/anoxic SBR cycle operation. It can be concluded that for SBRM a decrease of the OLRfeeding resulted in a strong increase in bio-P activity and is therefore assumed to promote growth of PAO over GAO. The effect on the granule settleability was only minor, and dense granular structures were preserved. This is to our knowledge the first time that the positive effect of a prolonged feeding time on the bio-P removal activity in an AGS system has been investigated while treating an industrial wastewater in a C-SBR. When treating brewery wastewater characterised by a high COD/P ratio, successful granule formation was achieved by applying an anaerobic mixed pulse feeding strategy. Sludge morphology was dominated by dense granule structures showing good and stable settling characteristics suggesting successful selection for granule-forming organisms. A decrease of the anaerobic mixed feeding rate resulted in deterioration of the granular sludge combined with filamentous outgrowth. In-situ cycle measurements showed no increase in bio-P activity, while Neisser staining showed intracellular poly-P granules within filamentous organisms. This shift in sludge morphology towards enrichment of filamentous organisms had a negative impact on the settleability of the biomass. It is suggested that the high influent COD/P ratio prevented proliferation of PAO-like organisms within the system, promoting enrichment of high affinity filaments over granule forming groups. Through this study, the importance of the anaerobic mixed feeding rate in a C-SBR system is confirmed. When considering application of the AGS technology in conventional mixed SBR systems, the anaerobic mixed feeding rate should be taken into account. In this study, a new operational parameter, OLRfeeding, was defined to allow quantification of the feeding rate from a substrate loading point of view.

ACKNOWLEDGEMENTS

This work was supported by the University of Antwerp, Pantarein Water BVBA and the Flanders Innovation and Entrepreneurship Agency (grant number 150723).

REFERENCES

REFERENCES
APHA/AWWA/WEF
1998
Standard Methods for the Examination of Water and Wastewater
,
20th edn
.
American Health Association/American Water Works Association/Water Environment Federation
,
Washington, DC
,
USA
.
Baetens
D.
,
Weemaes
M.
,
Hosten
L.
,
De Vos
P.
&
Vanrolleghem
P. A.
2001
Enhanced Biological Phosphorus Removal: Competition and Symbiosis Between SRBs and PAOs on Lactate/Acetate Feed
.
http://hdl.handle.net/1854/LU-149048 (accessed 8 October 2018).
Caluwé
M.
,
Dobbeleers
T.
,
D'aes
J.
,
Miele
S.
,
Akkermans
V.
,
Daens
D.
,
Geuens
L.
,
Kiekens
F.
,
Blust
R.
&
Dries
J.
2017
Formation of aerobic granular sludge during the treatment of petrochemical wastewater
.
Bioresource Technology
238
,
559
567
.
de Kreuk
M. K.
,
Kishida
N.
,
Tsuneda
S.
&
Van Loosdrecht
M. C. M.
2010
Behavior of polymeric substrates in an aerobic granular sludge system
.
Water Research
44
,
5929
5938
.
Driessen
W.
&
Vereijken
T
, .
2003
Recent Developments in Biological Treatment of Brewery Effluent
.
The Institute and Guild of Brewing Convention
,
Livingstone
,
Zambia
Europe Economics & The Brewers of Europe
2016
The Contribution Made by Beer to the European Economy EU Report – January 2016
.
Filipe
C. D. M.
,
Daigger
G. T.
&
Grady
C. P. L.
Jr
2001
pH as a key factor in the competition between glycogen-accumulating organisms and phosphorus accumulating organisms
.
Water Environment Research
73
(
2
),
223
232
.
Jenkins
D.
,
Richard
M. G.
&
Daigger
G. T.
2004
Manual on the Causes and Control of Activated Sludge Bulking, Foaming, and Other Solids Separation Problems
,
3rd edn
.
CRC Press LLC & IWA Publishing
,
London
,
UK
.
Karst
S. M.
,
Albertsen
M.
,
Kirkegaard
R. H.
,
Dueholm
M. S.
,
Nielsen
P. H.
2016
Molecular methods
. In:
Experimental Methods in Wastewater Treatment
(
Van Loosdrecht
M.
,
Lopez-Vasquez
C. M.
,
Nielsen
P. H.
&
Brdjanovic
D.
, eds).
IWA Publishing
,
London
,
UK
, pp.
285
319
.
McIlroy
S.
,
Porter
K.
,
Seviour
R. J.
&
Tillett
D.
2008
Simple and safe method for simultaneous isolation of microbial RNA and DNA from problematic populations
.
Appl. Environ. Microbiol.
74
,
6806
6807
.
McIlroy
S. J.
,
Saunders
A. M.
,
Albertsen
M.
,
Nierychlo
M.
,
Mcilroy
B.
&
Hansen
A. A.
2015
MiDAS: the field guide to the microbes of activated sludge
.
Database
2015
,
1
8
.
Oehmen
A.
,
Lemos
P. C.
,
Carvalho
G.
,
Yuan
Z.
,
Keller
J.
,
Blackall
L. L.
&
Reis
M. A. M.
2007
Advances in enhanced biological phosphorus removal: from micro to macro scale
.
Water Research
41
(
11
),
2271
2300
.
Pronk
M.
,
de Kreuk
M. K.
,
de Bruin
B.
,
Kamminga
P.
,
Kleerebezem
R.
&
van Loosdrecht
M. C. M.
2015
Full scale performance of the aerobic granular sludge process for sewage treatment
.
Water Research
84
,
207
217
.
Rubio-Rincón
F. J.
,
Welles
L.
,
Lopez-Vazquez
C. M.
,
Nierychlo
M.
,
Abbas
B.
,
Geleijnse
M.
,
Nielsen
P. H.
,
van Loosdrecht
M. C. M.
&
Brdjanovic
D.
2017
Long-term effects of sulphide on the enhanced biological removal of phosphorus: the symbiotic role of Thiothrix caldifontis
.
Water Research
116
,
53
64
.
Schwarzenbeck
N.
,
Erley
R.
,
Mc Swain
B. S.
,
Wilderer
P. A.
&
Irvine
R. L.
2004
Treatment of malting wastewater in a granular sludge sequencing batch reactor (SBR)
.
Acta Hydrochimica et Hydrobiologica
32
(
1
),
16
24
.
Stes
H.
,
Aerts
S.
,
Caluwé
M.
,
Dobbeleers
T.
,
Wuyts
S.
,
Kiekens
F.
,
D'aes
J.
,
De Langhe
P.
&
Dries
J.
2018
Formation of aerobic granular sludge and the influence of the pH on sludge characteristics in a SBR fed with brewery/bottling plant wastewater
.
Water Science and Technology
77
(
9
),
2253
2264
.
STOWA
2013
Nereda ® Praktijkonderzoeken 2010–2012
.
STOWA
, p.
29
.
Val del Río
A.
,
Figueroa
M.
,
Arrojo
B.
,
Mosquera-Corral
A.
,
Campos
J. L.
&
García-Torriello
G.
2012
Aerobic granular SBR systems applied to the treatment of industrial effluents
.
Journal of Environmental Management
95
,
S88
S92
.

Supplementary data