440
© IWA Publishing 2015 Water Science & Technology
|
71.3
|
2015
Continuous flow aerobic granular sludge reactor for dairy wastewater treatment C. Bumbac, I. A. Ionescu, O. Tiron and V. R. Badescu
ABSTRACT The focus of this study was to assess the treatment performance and granule progression over time within a continuous flow reactor. A continuous flow airlift reactor was seeded with aerobic granules from a laboratory scale sequencing batch reactor (SBR) and fed with dairy wastewater. Stereomicroscopic investigations showed that the granules maintained their integrity during the experimental period. Laser diffraction investigation showed proof of new granules formation with 100–500 μm diameter after only 2 weeks of operation. The treatment performances were satisfactory
C. Bumbac (corresponding author) I. A. Ionescu O. Tiron V. R. Badescu National Research and Development Institute for Industrial Ecology – ECOIND, 71–73 Drumul Podu Dambovitei Street, Sector 6, 060652 Bucharest, Romania E-mail: [email protected]
and more or less similar to the ones obtained from the SBR. Thus, removal efficiencies of 81–93% and 85–94% were observed for chemical oxygen demand and biological oxygen demand, respectively. The N-NHþ 4 was nitrified with removal efficiencies of 83–99% while the nitrate produced was simultaneously denitrified – highest nitrate concentration determined in the effluent was 4.2 mg/L. The removal efficiency of total nitrogen was between 52 and 80% depending on influent nitrogen load (39.3–76.2 mg/L). Phosphate removal efficiencies ranged between 65 and above 99% depending on the influent phosphate concentration, which varied between 11.2 and 28.3 mg/L. Key words
| aerobic granular sludge, continuous flow, dairy wastewater treatment
INTRODUCTION The biological treatment of municipal or industrial effluents in wastewater treatment plants (WWTPs) is often accomplished by means of the application of conventional activated sludge processes, a one-century-old technology, which was firstly designed to oxidise organic matter and afterwards adapted for nutrient removal. However, recent intensive research is looking for solutions to improve treatment efficiencies while reducing investment and operational costs (de Kreuk et al. ; Adav et al. ), thus setting the aerobic granular sludge technology as the new standard for domestic and industrial wastewater treatment (Giesen et al. ). Since its discovery (Mishima & Nakamura ), aerobic granular sludge has drawn the attention of researchers worldwide as a promising technology, as it offers several advantages over conventional activated sludge systems in terms of performance and cost efficiency (both investment and operational) by simultaneously removing organic load and nutrients (N and P) (de Bruin et al. ; Gao et al. ) or even its ability to treat wastewater with high organic loads or toxic substrates in only one aerobic reactor (Moy et al. ; Figueroa et al. ). doi: 10.2166/wst.2015.007
The technology was developed mainly in column type sequential biological reactors (SBR) with a large height to diameter ratio (McSwain et al. ; Schwarzenbeck et al. ; Wichern et al. ; Mosquera-Corral et al. ). The use of aerobic granular sludge technology for different wastewater treatment applications is one of the recent research trends. The technology evolved rapidly from laboratory-scale to pilot- and ultimately to full-scale applications (Giesen et al. ; Winkler et al. ) for municipal wastewater treatment. However, the key feature of this technology is the sequencing batch operation with batch influent, effluent and the possibility of washing out poor settling sludge. Thus, the application of aerobic granular sludge technology is suitable especially where investments in a completely new SBR type WWTP are envisaged or an existing SBR WWTP is considered for upgrading. The optimisation of reactor design and conditions for aerobic granules to perform under continuous flow operation could open new perspectives for potential upgrading, with mere investment, of the existing conventional
441
C. Bumbac et al.
|
Continuous flow aerobic granular sludge reactor
WWTPs from large population agglomerations or industry in order to increase the treatment capacity and performances. There are already a few reports that studied the possibility to obtain and/or maintain aerobic granular sludge in continuous flow reactors (Chen et al. ; Juang et al. ; Liu et al. ; Zhou et al. ). However, little work a has been done to cultivate aerobic granules in a continuous flow bioreactor with a simple structure that is realistic for engineering (Zhou et al. ), and even fewer studies have focussed on evaluating the treatment performances of continuous flow aerobic granular sludge reactors (Jemaat et al. ; Zhao et al. ). This study focusses on the treatment performances of an aerobic granular sludge continuous flow reactor treating dairy industry wastewater and granules progression against time.
MATERIALS AND METHODS Reactor set-up and operation For the experiments, a Plexiglas-built rectangular-shaped bioreactor with a working volume of 10 L and a height to width ratio of 1.3 was used (Figure 1). The reactor had a built-in granule pre-settling area consisting of a 4 cm wide lamella positioned at a 45 angle before the reactor overflow wall towards the granules settling and recovery area. This lamella provides a small turbulence free area that allows the granules to return to the reactor mixed liquor. However, smaller granules (with lower density) overflow to the granules settling area – a 2.5 cm wide settler attached to the width of the overflow wall of the reactor with a base communication port with the reactor through which most of W
Water Science & Technology
|
71.3
|
2015
the granules are recovered into the reactor. A second lamella positioned also at 45 angle against the wall, beneath the slot where granules return from the settler, provides shelter from turbulence and favours a circular downwards current that helps the process of granules recovery. The loose, fluffy granules and flocculated sludge is washed out with the effluent towards the secondary settler (volume of 2 L). Thus, poor settling sludge is withdrawn as excess sludge using a peristaltic pump, while the effluent passes through the final overflow to the collecting vessel. The influent was fed continuously using a Verder scientific peristaltic pump (flow 1 L/hour). This ensured a 10-hour hydraulic retention time (HRT) in the aerobic reactor, which was followed by an approximately 5-minute settling time in the granules settler and approximately 2 hours more in the final settler, summing a total time of approximately 12 hours. Dissolved oxygen was supplied continuously by means of air flow (6–8 L/minute) through fine air bubble diffusers that ensured the airlift fluidised bed type movement of biomass within the bioreactor and the necessary dissolved oxygen for the metabolic requirements of the bacteria. W
Seed sludge and feed wastewater The reactor was inoculated with aerobic granules (100– 3,000 μm in diameter) obtained previously by the authors in a laboratory-scale column type aerobic granular sludge SBR treating the same influent – dairy industry wastewater. Thus, 2 L of settled sludge with a sludge volume index of 40 mL/g was added as initial inoculum in the continuous flow reactor. The biomass concentration at start-up was approximately 5 g/L. Influent wastewater was collected from a local milk-processing factory and was characterised by relatively high organic load and nutrients concentrations (see Table 1). Analytical methods During the experimental period (10 weeks), in order to evaluate treatment performances, the following physical–chemical
Table 1
Figure 1
|
Schematic representation of continuous flow aerobic granular sludge reactor.
|
Quality characteristics of influent dairy wastewater
BOD5
Ntotal
NHþ 4
PO34
Parameter
COD (mg/L)
(mg/L)
(mg/L)
(mg/L)
(mg/L)
Variation range
1,296– 2,328
581– 1,213
39.3– 76.2
28–49
11.2–29
Average
1,662
897
57
36
20
442
C. Bumbac et al.
|
Continuous flow aerobic granular sludge reactor
and biological parameters were monitored: pH – using a Consort multi-parameter; dissolved oxygen concentration using WTW Oxi320; organic matter content as chemical oxygen demand (COD) according to standard SR ISO 6060–1996 and as biochemical oxygen demand (BOD5) according to standard SR EN 1899/1-03; 2-02; concentration of main 3 macronutrients: NHþ 4 , NO2 , NO3 and PO4 (mg/L) using ICS 3000, DIONEX ion chromatograph and according to standards: SR EN ISO 14911-03 (for NHþ 4 ) and SR EN ISO 10304/1-09 (for the quantification of NO 2 , NO3 and 3 PO4 ). The size distribution (within 10–2,000 μm) of the granules was determined by the technique of laser diffraction using a Malvern Mastersizer 2000 equipped with the liquid samples module Hydro 2000S. The changes in morphology of the granules were monitored by stereomicroscopy using a Motic trinocular stereomicroscope DMW-143 with built-in camera and dedicated software Motic Images Plus 2.0 for image processing/capture/sample measurement.
RESULTS AND DISCUSSION Throughout the experiment, the continuous flow aerobic granular sludge reactor was operated at an average organic loading rate of 3.84 kg COD/m3/day. The investigations
Figure 2
|
Water Science & Technology
|
71.3
|
2015
focussed on effluent quality and treatment performances assessment and evaluation of continuous flow feeding upon aerobic granules development and integrity.
Treatment performances Organic load, expressed as COD was removed with efficiencies, which varied in the range of 81–93% (Figure 2(a)) while the BOD removal efficiencies ranged between 85 and 94% (Figure 2(a)). The remaining COD concentrations varied from 250 to 101 mg O2/L. This was correlated to the concentration of total suspended solids (TSS), which varied between 55 and 186 mg/L (Figure 2(b)), mostly due to the negative impact of flocculated sludge washout with the effluent. BOD residual concentration in the effluent ranged between 24.7 and 69 mg/L. Eighty-three to above 99% of the influent N‐NHþ 4 was nitrified while the resulting nitrate was simultaneously denitrified – the highest nitrate concentration accumulated in the effluent as a result of nitrification during the experiment was 4.2 mg/L (Figure 2(d)). Thus, the aerobic granules proved again their capabilities to adapt in the same aerobic reactor to the different process conditions required for aerobic and anaerobic processes. Total nitrogen removal efficiencies
Graphic representation of COD and BOD removal efficiencies (a); COD, BOD5 and TSS effluent concentrations (b); total nitrogen and phosphate removal efficiencies (c); and remaining nutrient concentrations in the effluent (d).
443
C. Bumbac et al.
|
Continuous flow aerobic granular sludge reactor
varied between 52 and 80% (Figure 2(c)) – which was correlated with the variation of influent nitrogen load (Table 1). Phosphate removal efficiency varied during the experiment between 65 and above 99% – a variation that correlated with the influent phosphate load, which ranged between 11.2 and 28.3 mg/L. The treatment performances were satisfactory and more or less similar to the ones obtained by Ionescu et al. () in a sequentially operated bioreactor operating at the same HRT, and also similar to the performances obtained by the conventional biological WWTP of the dairy products factory operating at an HRT of 24 hours. The overall quality of the effluent fits the Romanian limits imposed by the Governmental Decision 352/2005 for discharging into municipal sewerage system and exceeds the limits imposed for discharging in surface waters. Granule integrity/stability The experiment focussed also on the evaluation of aerobic sludge granule evolution in time. During the experiment, sludge granules maintained their integrity and settling capacity. However, the measurements performed using laser diffraction techniques indicated a slight change in the
Figure 3
|
Water Science & Technology
|
71.3
|
2015
pattern of granules size distribution, more exactly a new set of granules with diameters between 100 and 500 μm started to appear after only 2 weeks of operation (compare Figure 3(b) and (a)). The granule size distribution patterns after 5 and 10 weeks showed a slow decrease of the volume of larger granules and a slight increase in the volume of small granules and even flocculated sludge (
© IWA Publishing 2015 Water Science & Technology
|
71.3
|
2015
Continuous flow aerobic granular sludge reactor for dairy wastewater treatment C. Bumbac, I. A. Ionescu, O. Tiron and V. R. Badescu
ABSTRACT The focus of this study was to assess the treatment performance and granule progression over time within a continuous flow reactor. A continuous flow airlift reactor was seeded with aerobic granules from a laboratory scale sequencing batch reactor (SBR) and fed with dairy wastewater. Stereomicroscopic investigations showed that the granules maintained their integrity during the experimental period. Laser diffraction investigation showed proof of new granules formation with 100–500 μm diameter after only 2 weeks of operation. The treatment performances were satisfactory
C. Bumbac (corresponding author) I. A. Ionescu O. Tiron V. R. Badescu National Research and Development Institute for Industrial Ecology – ECOIND, 71–73 Drumul Podu Dambovitei Street, Sector 6, 060652 Bucharest, Romania E-mail: [email protected]
and more or less similar to the ones obtained from the SBR. Thus, removal efficiencies of 81–93% and 85–94% were observed for chemical oxygen demand and biological oxygen demand, respectively. The N-NHþ 4 was nitrified with removal efficiencies of 83–99% while the nitrate produced was simultaneously denitrified – highest nitrate concentration determined in the effluent was 4.2 mg/L. The removal efficiency of total nitrogen was between 52 and 80% depending on influent nitrogen load (39.3–76.2 mg/L). Phosphate removal efficiencies ranged between 65 and above 99% depending on the influent phosphate concentration, which varied between 11.2 and 28.3 mg/L. Key words
| aerobic granular sludge, continuous flow, dairy wastewater treatment
INTRODUCTION The biological treatment of municipal or industrial effluents in wastewater treatment plants (WWTPs) is often accomplished by means of the application of conventional activated sludge processes, a one-century-old technology, which was firstly designed to oxidise organic matter and afterwards adapted for nutrient removal. However, recent intensive research is looking for solutions to improve treatment efficiencies while reducing investment and operational costs (de Kreuk et al. ; Adav et al. ), thus setting the aerobic granular sludge technology as the new standard for domestic and industrial wastewater treatment (Giesen et al. ). Since its discovery (Mishima & Nakamura ), aerobic granular sludge has drawn the attention of researchers worldwide as a promising technology, as it offers several advantages over conventional activated sludge systems in terms of performance and cost efficiency (both investment and operational) by simultaneously removing organic load and nutrients (N and P) (de Bruin et al. ; Gao et al. ) or even its ability to treat wastewater with high organic loads or toxic substrates in only one aerobic reactor (Moy et al. ; Figueroa et al. ). doi: 10.2166/wst.2015.007
The technology was developed mainly in column type sequential biological reactors (SBR) with a large height to diameter ratio (McSwain et al. ; Schwarzenbeck et al. ; Wichern et al. ; Mosquera-Corral et al. ). The use of aerobic granular sludge technology for different wastewater treatment applications is one of the recent research trends. The technology evolved rapidly from laboratory-scale to pilot- and ultimately to full-scale applications (Giesen et al. ; Winkler et al. ) for municipal wastewater treatment. However, the key feature of this technology is the sequencing batch operation with batch influent, effluent and the possibility of washing out poor settling sludge. Thus, the application of aerobic granular sludge technology is suitable especially where investments in a completely new SBR type WWTP are envisaged or an existing SBR WWTP is considered for upgrading. The optimisation of reactor design and conditions for aerobic granules to perform under continuous flow operation could open new perspectives for potential upgrading, with mere investment, of the existing conventional
441
C. Bumbac et al.
|
Continuous flow aerobic granular sludge reactor
WWTPs from large population agglomerations or industry in order to increase the treatment capacity and performances. There are already a few reports that studied the possibility to obtain and/or maintain aerobic granular sludge in continuous flow reactors (Chen et al. ; Juang et al. ; Liu et al. ; Zhou et al. ). However, little work a has been done to cultivate aerobic granules in a continuous flow bioreactor with a simple structure that is realistic for engineering (Zhou et al. ), and even fewer studies have focussed on evaluating the treatment performances of continuous flow aerobic granular sludge reactors (Jemaat et al. ; Zhao et al. ). This study focusses on the treatment performances of an aerobic granular sludge continuous flow reactor treating dairy industry wastewater and granules progression against time.
MATERIALS AND METHODS Reactor set-up and operation For the experiments, a Plexiglas-built rectangular-shaped bioreactor with a working volume of 10 L and a height to width ratio of 1.3 was used (Figure 1). The reactor had a built-in granule pre-settling area consisting of a 4 cm wide lamella positioned at a 45 angle before the reactor overflow wall towards the granules settling and recovery area. This lamella provides a small turbulence free area that allows the granules to return to the reactor mixed liquor. However, smaller granules (with lower density) overflow to the granules settling area – a 2.5 cm wide settler attached to the width of the overflow wall of the reactor with a base communication port with the reactor through which most of W
Water Science & Technology
|
71.3
|
2015
the granules are recovered into the reactor. A second lamella positioned also at 45 angle against the wall, beneath the slot where granules return from the settler, provides shelter from turbulence and favours a circular downwards current that helps the process of granules recovery. The loose, fluffy granules and flocculated sludge is washed out with the effluent towards the secondary settler (volume of 2 L). Thus, poor settling sludge is withdrawn as excess sludge using a peristaltic pump, while the effluent passes through the final overflow to the collecting vessel. The influent was fed continuously using a Verder scientific peristaltic pump (flow 1 L/hour). This ensured a 10-hour hydraulic retention time (HRT) in the aerobic reactor, which was followed by an approximately 5-minute settling time in the granules settler and approximately 2 hours more in the final settler, summing a total time of approximately 12 hours. Dissolved oxygen was supplied continuously by means of air flow (6–8 L/minute) through fine air bubble diffusers that ensured the airlift fluidised bed type movement of biomass within the bioreactor and the necessary dissolved oxygen for the metabolic requirements of the bacteria. W
Seed sludge and feed wastewater The reactor was inoculated with aerobic granules (100– 3,000 μm in diameter) obtained previously by the authors in a laboratory-scale column type aerobic granular sludge SBR treating the same influent – dairy industry wastewater. Thus, 2 L of settled sludge with a sludge volume index of 40 mL/g was added as initial inoculum in the continuous flow reactor. The biomass concentration at start-up was approximately 5 g/L. Influent wastewater was collected from a local milk-processing factory and was characterised by relatively high organic load and nutrients concentrations (see Table 1). Analytical methods During the experimental period (10 weeks), in order to evaluate treatment performances, the following physical–chemical
Table 1
Figure 1
|
Schematic representation of continuous flow aerobic granular sludge reactor.
|
Quality characteristics of influent dairy wastewater
BOD5
Ntotal
NHþ 4
PO34
Parameter
COD (mg/L)
(mg/L)
(mg/L)
(mg/L)
(mg/L)
Variation range
1,296– 2,328
581– 1,213
39.3– 76.2
28–49
11.2–29
Average
1,662
897
57
36
20
442
C. Bumbac et al.
|
Continuous flow aerobic granular sludge reactor
and biological parameters were monitored: pH – using a Consort multi-parameter; dissolved oxygen concentration using WTW Oxi320; organic matter content as chemical oxygen demand (COD) according to standard SR ISO 6060–1996 and as biochemical oxygen demand (BOD5) according to standard SR EN 1899/1-03; 2-02; concentration of main 3 macronutrients: NHþ 4 , NO2 , NO3 and PO4 (mg/L) using ICS 3000, DIONEX ion chromatograph and according to standards: SR EN ISO 14911-03 (for NHþ 4 ) and SR EN ISO 10304/1-09 (for the quantification of NO 2 , NO3 and 3 PO4 ). The size distribution (within 10–2,000 μm) of the granules was determined by the technique of laser diffraction using a Malvern Mastersizer 2000 equipped with the liquid samples module Hydro 2000S. The changes in morphology of the granules were monitored by stereomicroscopy using a Motic trinocular stereomicroscope DMW-143 with built-in camera and dedicated software Motic Images Plus 2.0 for image processing/capture/sample measurement.
RESULTS AND DISCUSSION Throughout the experiment, the continuous flow aerobic granular sludge reactor was operated at an average organic loading rate of 3.84 kg COD/m3/day. The investigations
Figure 2
|
Water Science & Technology
|
71.3
|
2015
focussed on effluent quality and treatment performances assessment and evaluation of continuous flow feeding upon aerobic granules development and integrity.
Treatment performances Organic load, expressed as COD was removed with efficiencies, which varied in the range of 81–93% (Figure 2(a)) while the BOD removal efficiencies ranged between 85 and 94% (Figure 2(a)). The remaining COD concentrations varied from 250 to 101 mg O2/L. This was correlated to the concentration of total suspended solids (TSS), which varied between 55 and 186 mg/L (Figure 2(b)), mostly due to the negative impact of flocculated sludge washout with the effluent. BOD residual concentration in the effluent ranged between 24.7 and 69 mg/L. Eighty-three to above 99% of the influent N‐NHþ 4 was nitrified while the resulting nitrate was simultaneously denitrified – the highest nitrate concentration accumulated in the effluent as a result of nitrification during the experiment was 4.2 mg/L (Figure 2(d)). Thus, the aerobic granules proved again their capabilities to adapt in the same aerobic reactor to the different process conditions required for aerobic and anaerobic processes. Total nitrogen removal efficiencies
Graphic representation of COD and BOD removal efficiencies (a); COD, BOD5 and TSS effluent concentrations (b); total nitrogen and phosphate removal efficiencies (c); and remaining nutrient concentrations in the effluent (d).
443
C. Bumbac et al.
|
Continuous flow aerobic granular sludge reactor
varied between 52 and 80% (Figure 2(c)) – which was correlated with the variation of influent nitrogen load (Table 1). Phosphate removal efficiency varied during the experiment between 65 and above 99% – a variation that correlated with the influent phosphate load, which ranged between 11.2 and 28.3 mg/L. The treatment performances were satisfactory and more or less similar to the ones obtained by Ionescu et al. () in a sequentially operated bioreactor operating at the same HRT, and also similar to the performances obtained by the conventional biological WWTP of the dairy products factory operating at an HRT of 24 hours. The overall quality of the effluent fits the Romanian limits imposed by the Governmental Decision 352/2005 for discharging into municipal sewerage system and exceeds the limits imposed for discharging in surface waters. Granule integrity/stability The experiment focussed also on the evaluation of aerobic sludge granule evolution in time. During the experiment, sludge granules maintained their integrity and settling capacity. However, the measurements performed using laser diffraction techniques indicated a slight change in the
Figure 3
|
Water Science & Technology
|
71.3
|
2015
pattern of granules size distribution, more exactly a new set of granules with diameters between 100 and 500 μm started to appear after only 2 weeks of operation (compare Figure 3(b) and (a)). The granule size distribution patterns after 5 and 10 weeks showed a slow decrease of the volume of larger granules and a slight increase in the volume of small granules and even flocculated sludge (
