Low-level alkalinity (pH 9–10) coupled with ultrasonic or mechanical cutting with different energy input for obtaining carbon sources were tested for sludge pretreatment process before anaerobic sludge digestion. The differences between the primary sludge (PS) and waste activated sludge (WAS)-derived dissolved organic matter (DOM) species were evaluated for their bioavailability and affinity (in the form of amino acids) to the bio-nutrient removal (BNR) biomass. Soluble microbial by-product-like substances as the predominant DOM components in the raw PS and WAS increased by 23 and 22%, respectively, after low-level alkaline treatment (pH 9–10) and ultrasonication. In addition, the protein components were degraded further as free amino acids (FAAs). The sludge-derived aspartate, glutamate, followed by arginine were the most commonly used FAAs by the BNR biomass. The pattern of recovering this special sludge-derived carbon source to enhance P removal and recovery in the BNR process is depicted.
Ultrasonic treatment of sludge has the advantage of efficiently producing non-fermentative carbon sources rich in amino acids.
Soluble microbial by-product-like substances were the main DOM components released by ultrasonication treatment of sludge.
The BNR communities showed high affinity to the ionized amino acids contained in the sonication-induced sludge carbon source.
readily-biodegradable chemical oxygen demand
wastewater treatment plants
waste activated sludge
volatile fatty acids
free amino acids
dissolved organic matter
soluble microbial by-product-like
the alkaline and mechanical cutting treatment of WAS
the alkaline and ultrasonic treatment of PS/WAS
fluorescence regional integration
soluble chemical oxygen demand
degree of disintegration
In the bio-nutrient removal (BNR) process, readily biodegradable organic matter is consumed by denitrifiers and phosphorus-accumulating organisms (PAOs) as a carbon and energy source for biological nitrogen and phosphorus removal. The type and quantity of carbon sources determine the community structure of the biomass and the resulting bio-N and -P removal efficiencies. As a rule of thumb, influent readily biodegradable chemical oxygen demand (rbCOD)/total nitrogen and rbCOD/total phosphorus should have a relation of over 8 and 35, respectively. Thus it is a guarantee of an efficient biological nitrogen and phosphorus removal (Park et al. 2011). However, a deficiency in the influent carbon source is reported globally. Especially in China, the COD in wastewater treatment plants (WWTPs) is generally lower than that in other countries (approximately 200–300 mg L−1 and C/N < 4) due to multiple factors including food processing, the wide application of septic tanks, and high groundwater infiltration rates into leaky sewage systems (Liao et al. 2015). Thus, considerable commercial carbon sources (such as methanol) are added in the pre/post denitrification process, which is resulting in high WWTP operating costs and carbon footprints (Lu et al. 2018).
On the other hand, considerable sludge, including primary sludge (PS) and waste activated sludge (WAS), is produced inevitably in WWTPs. The cost of sludge treatment is as high as 64–198 dollars t−1 dry sludge (Yang et al. 2015). The three highest operational cost factions in WWTPs are aeration power, sludge treatment, and supplementary carbon input (Huang et al. 2018). However, with increasingly stringent worldwide emission standards for N and P, sludge treatment and additional carbon source consumption are becoming the leading costs, especially in China. As much as 80.42 million tons of WAS will be produced in China in 2020 (Chen et al. 2020). Although anaerobic digestion has been applied widely as the main sludge treatment process, many factors including the low organic content (30–65%), unstable methane production, lack of skilled operators, and additional alkaline and heating costs constrain its application (Huang et al. 2018; Hao et al. 2020). Dewatering and landfill treatment are still the most commonly used sludge treatment/disposal processes worldwide, which waste the significant resources contained in the sludge. Therefore, the challenging task of developing technologies for efficient carbon and nutrient recovery from sludge in an energy- and cost-efficient way is the focus of current research (MacDonald et al. 2011).
Applying a sludge-derived carbon source in the carbon-source-deficient BNR processes yields multiple benefits: satisfying the carbon source requirement for efficient bio-nutrient removal (Qiu et al. 2019; Wang et al. 2019), increasing sludge dewatering characteristics, and decreasing the total sludge volume for disposal (Ahn et al. 2002). To date, research on winning carbon sources from sludges has mainly focused on WAS. However, the types of carbon sources contained in PS and their bioavailability to the BNR biomass have been mostly neglected. The fermentation of organic matter in wastewater already occurs in the sewage pipeline before entering the WWTPs. Thick biofilms are distributed ubiquitously in the sewer system, which benefit from the dark, nutrient-rich, and warm environment suitable for their growth. During its migration, sewage organic matter is captured and partially degraded by fermentative bacteria, forming small organic molecules or volatile fatty acid (VFA) complexes (Jin et al. 2018). It is speculated that degradation byproducts can be adsorbed in situ into the cell aggregates (by van der Waals or chemical bonding forces), forming PS precursors. This implies that a high-quality carbon source derived from PS can be obtained by low-intensity mechanical disintegration without fermentation of PS.
Mechanical pretreatment processes such as ultrasonic radiation, desizing machines, hydrodynamic cavitation, and grinding (Kampas et al. 2007; Mancuso et al. 2019) are the most frequently reported methods for effectively disintegrating WAS. Among them, ultrasonic radiation has been widely studied and developed (Tian et al. 2018). Considerable soluble chemical oxygen demand (SCOD) release was observed during ultrasonic treatment of WAS. The organic matter was released preferentially during ultrasonic treatment, thus avoiding the introduction of excessive nutrients into the mainstream with the ultrasonic pretreated side streams (Li et al. 2014). Compared to a strong alkali addition (e.g. alkaline conditions pH > 12), an initial critical relatively low pH edge is required to achieve sludge solubilization with low energy input. A low-level alkaline treatment (pH > =9.5) coupled with ultrasonication is an effective sludge disintegration method (Tian et al. 2018). This combined treatment has also been applied on digested sludge (post-treatment), and soluble microbial by-product-like (SMP-like) substances. For this treatment, humic acid-like (HA-like) substances increased by 0.9 and 2.2 times, respectively, from which the fermented sludge carbon source can also be recovered (Tian et al. 2016). However, information is still lacking on the specific carbon and nitrogen components released, which determine the effects of a carbon-supplemented BNR process, and on the change in dissolved organic matter (DOM) components with energy input. Furthermore, the components of this type of carbon source derived directly from non-fermented PS and their recovery potential as the supplementary carbon source for BNR have not been studied.
In this study, two mechanical sludge treatments, ultrasonication and mechanical cutting, coupled with low-level alkaline treatment, were studied firstly and compared. The purpose of this investigation was to: (1) compare the quantities and species of organic matter released during ultrasonic treatment and the alkaline and mechanical treatment of sludge. (2) estimate the impact of energy input on the components of DOM released from sludge; and (3) explore the affinity of PAO-rich biomass to the specific components in sludge carbon sources. The results of our study will provide a new strategy for recovery and reuse of carbon and nutrient resources directly from wastewater via non-fermented sludge, as well as to recover biogas efficiently in the subsequent anaerobic digestion of this pre-treated sludge.
MATERIALS AND METHODS
Sludge samples and treatments
Sludge samples were collected from a municipal WWTP in Songjiang, (Shanghai), China, which uses the Anaerobic/Anoxic/Oxic process for BNR. The characteristics of sludge after thickening are shown in Table 1. According to literature (Tian et al. 2018; Mancuso et al. 2019), a pH range of 9–10 was defined as ‘low-level alkaline condition’. Approximately 0.32 g or 0.61 g NaOH were consumed for every liter PS or WAS to achieve pH 9 and 10, respectively.
|Parameter .||Unit .||Value .||.|
|Primary sludge .||Waste activated sludge .|
|Total solid||mg L−1||23,020–24,561||23,265–24,102|
|Volatile solid||mg L−1||14,160–14,753||14,329–15,023|
|Total COD||mg L−1||13,253–17,253||13,269–16,598|
|Parameter .||Unit .||Value .||.|
|Primary sludge .||Waste activated sludge .|
|Total solid||mg L−1||23,020–24,561||23,265–24,102|
|Volatile solid||mg L−1||14,160–14,753||14,329–15,023|
|Total COD||mg L−1||13,253–17,253||13,269–16,598|
Schematics of the experimental setups are shown in Figure S1. For the low-level alkaline and ultrasonic treatment of PS (A&UPS) and WAS (A&UWAS), an ultrasonicator (S-450D, BRANSON, Shanghai, China) with a maximum power of 450 W was equipped with a titanium probe (3/4 inch, 17 mm) and a transducer. It operated at a fixed frequency of 20 kHz at an 30% of maximum amplitude. In each sonication experiment, 200 mL of sludge were placed in a 250 mL reactor wrapped with a thermal isolator, with the probe located 1.5 cm under the sludge level. The average actual input power (46 W) of the ultrasonicator was recorded with a power monitor (HY-001, HYELEC, China). PS and WAS samples were processed for 0–10 or 0–20 min, and the corresponding specific energy (SE) input (see below) ranges were 0–5,750 and 0–11,500 kJ kgTS−1, respectively.
Batch tests were carried out to determine the biomass affinity for different types of FAAs in the sludge carbon source. Here, we only considered the anaerobic carbon source consumption. Both the extracellular and intracellular carbon sources can serve as the carbon source for denitrification. However, only the intracellular carbon source (mainly in form of polyhydroxyalkanoates, PHAs) can be used for P uptake by the PAOs in the aerobic phase. Therefore, if a certain fraction of the carbon source can be consumed in the anaerobic phase by the PAOs, it can be applied also for endogenous denitrification in the anoxic phase. In this study, we mainly focus on this fraction of the carbon source (tested for bioavailability and compositions), which may cause an underestimation of the carbon source amount usable for nitrogen removal via extracellular carbon source degradation pathways.
Two samples of low-level alkaline and ultrasonic treatment (2,875 and 8,625 kJ kgTS−1 for PS and WAS, respectively), after adjusting the pH to 7 with 2 M HCl solution, were used for the batch test. The supernatants of treated sludge samples were centrifuged (4 °C, 10,000 rpm min−1, 15 min) and filtered (15 μm paper filter). The filtrates were used as the raw solution of the reclaimed sludge carbon source in the batch tests to investigate the components that were readily absorbed and used by the biomass.
The batch test biomass was taken from an alternating anaerobic aerobic biofilter, which had been operating for 5 years. The biofilm was washed with 0.85% NaCl solution, which maintains the osmotic pressure for the microorganisms and removes the biofilm-adsorbed organic matter. Three equal weights (0.03 ± 0.001 dry weight gram) of biomass were incubated in serum bottles with phosphate media to deplete biomass-sequestered PHAs. The composition of the liquid phosphate media was as follows: NaH2PO4·2H2O (70 mg L−1), MgSO4·7H2O (126 mg L−1), KCl (72 mg L−1), NaHCO3 (100 mg L−1), CaCl2 (20 mg L−1), FeSO4 (2.4 mg L−1). The harvested biomass was incubated in serum bottles. The sludge carbon source in the form of filtrate (150 mL) was added, before a N2 purge (20 min) was applied to achieve anaerobic conditions. The supernatant after filtration and centrifugation (4 °C, 10,000 rpm min−1, 15 min) was used to detect changes in FAA concentration after the microbial consumption for 12 h at 25 °C (the operating temperature of the biomass parent reactor). Results from the triplicate serum bottle experiments are shown as mean ± standard deviation (SD).
Three-dimensional excitation-emission matrix (EEM) fluorescence spectroscopy and fluorescence regional integration (FRI) analysis
Fluorescence measurements were conducted with a spectrofluorometer (Fluoro Max-3, QM/TM, USA) at an ambient temperature of 24 °C. All samples were filtered with Whatman 0.7 μm glass microfiber filters before EEM analysis. To obtain fluorescence EEMs, excitation wavelengths were incrementally increased from 220 to 420 nm at 5-nm steps and for emission wavelengths from 230 to 650 nm at 1-nm steps. The EEM spectra are plotted as an elliptical shape of contour. The X-axis represents the emission spectra, while the Y-axis represents the excitation spectra. The fluorescence intensity data were normalized. The spectra of all samples were acquired under the same operating conditions to make the data comparable (Chen et al. 2003). The FRI method was applied to quantitatively compare the relative differences among the characteristic DOM species in the different samples (Chen et al. 2003).
Analysis of FAAs
According to literature (Xia et al. 2017), before FAAs analysis the samples and 10% sulfosalicylic acid were mixed at a volume ratio of 1:4 to precipitate the protein and centrifuged for 30 min (18,000 rpm min−1, 4 °C). The supernatants were filtered through 0.45 and 0.22 μm membranes. Aliquots (20 μL) of the supernatants were injected into an automated amino acid analyzer (L-8900, Hitachi, Tokyo, Japan). The FAA content in the solution was determined by the external standard method. The detection wavelength of proline was 440 nm, and the detection wavelength for other FAAs was 570 nm.
The sludge parameters such as total solids, volatile solids, NH4+-N and soluble orthophosphate (SOP) were measured according to standard methods (APHA 2017). pH was measured using a pH instrument (Cond 3310 SET2, GUYU, China). The analyses of SCOD, SOP, and NH4+-N were carried out with the filtered (0.45 μm) supernatant of the sample after centrifugation (4 °C, 10,000 rpm, 15 min). All the measurements were carried out in triplicate. The standard deviations are shown in the form of error bars.
Statistical analysis of the data was conducted using IBM SPSS Statistics (19.0) software (http://www-01.ibm.com/software/analytics/spss/).
RESULTS AND DISCUSSION
Release of DOM by the different treatment processes
As shown in Figure 1(a) and 1(b), the release of SCOD occurred rapidly, which indicates the successful disintegration of sludge flocs and the release of DOM by the low-level alkaline and ultrasonic treatment. The maximum DDCOD achieved was 22.1 ± 0.6% in the PS-treated supernatant at the equivalent SE of 2,875 kJ kgTS−1, and 29.9 ± 0.5% in the WAS-treated supernatant at the equivalent SE of 11,500 kJ kgTS−1. However, the SCOD release slowed down with SE increased from 8,625 to 11,500 kJ kgTS−1 (Figure 1(b)). In addition, the release of NH4+-N and SOP from the sludge flocs or cells also indicated the disintegration of sludge (Figure S2), which was attributed to the synergistic effects of both the chemical and mechanical processes. The most damage to the cell walls and cell membranes of the biomass occurred at pH ≥ 10 (Xiao et al. 2015). Cells are more fragile in an alkaline environment and are more susceptible to damage caused by cavitation. The ultrasound-induced temperature increments also contributed to the dissolution of SCOD from sludge.
Degradation of low molecular weight organic matter and re-flocculation of the particles may contribute to the decline of SCOD and DDCOD (Figure 1(a)). On the one hand, ultrasonic-induced hydroxyl radicals can accelerate the hydrolysis of complex organic matter (Song et al. 2006). This effect would be enhanced under alkaline conditions (Tian et al. 2015). Therefore, it is not difficult to degrade organic matters of low molecular weight and oxidize the reducing inorganic substances by sonication. Furthermore, the accumulation of the released polymers (e.g. protein) would facilitate re-flocculation of the released organic matter. The more the sonication input was, the more flocculation and re-adsorption occurred (Gonze et al. 2003). Changes in the ammonium concentration of the supernatants of A&UPS (in Figure S2 (a)) seem to support this explanation; for example, the release of ammonium from the disintegrated organic remaining constant when the sonication input was higher than 2,875 kJ kgTS−1.
The release of DOM after A&MWAS is shown in Figure 1(c) and 1(d). With the increase in SE, the SCOD and DDCOD continuously increased, reaching 1,336 ± 48 mg L−1 and 9.6 ± 0.4% at 17,173 kJ kgTS−1. However, if the SE was increased from 8,586 to 11,448 kJ kgTS−1, a peak in the enhancement rate of SCOD was achieved. Therefore, considering energy consumption; for example, SE input of 11,448 kJ kgTS−1, the performance of the treatments using different pH conditions were observed. If the initial treatment pH was increased to 10 (low-level alkaline), the SCOD released by sludge increased to 3,126 ± 34 mg L−1 with a corresponding DDCOD value of 22.8 ± 0.3%. High-speed mechanical cutting destroyed the floc structure of the sludge and the turbulent flow intensified the mass transfer and the hydro-mechanical shear forces, causing disintegration of the cell walls and organic matter. In addition, with alkali treatment the released organic matter was more susceptible to degradation by the mechanical cutting forces.
Changes in DOM species within different treatments
In order to differentiate the detailed changes of DOM components in the supernatants during two low-level alkaline and mechanical treatments, we applied three-dimensional EEM fluorescence spectroscopy and FRI analysis in this study.
DOM species during alkaline and sonication processes
The EEM spectra in Figure 2 indicate that the predominant types of fluorescent substances in the raw sludge samples can be differentiated. For example, Peak A, located in region IV, represents SMP-like substances. Peak B, located in region V, represents HA-like substances, and Peak C, located in region I, represents tyrosine-like or aromatic protein substances. No obvious fluorescence peak in region I was found for raw WAS, and the presence of peak C occurred at 575 kJ kgTS−1.
The FRI method was used to further elucidate changes in DOM fractions with different energy consumption. As shown in Figure 3 for both PS and WAS sludge, SMP-like and the HA-like substances (represented by regions IV and V) were the predominant DOM species and accounted for 68 and 78% of the total DOM species in PS and WAS samples, respectively. Generally, HA-like (region V) and fulvic acid-like (region III) substances are considered to be refractory to biodegradation (Guo et al. 2014). The aromatic proteins (regions I and II) and SMP-like substances (region IV) are referred to as accessible and easily biodegradable DOMs, respectively (Liu et al. 2011; Guo et al. 2014). The total fraction of biodegradable DOM (regions I, II, and IV) increased by 23 and 21% in A&UPS and A&UWAS, respectively, which indicates that low-level alkaline treatment and ultrasonication can enhance the biodegradability of sludge.
There was no significant change in the fraction of aromatic protein components (region I plus II) with more SE input during sonication. Since proteins and polysaccharides are the main components of PS and WAS, ultrasonication can induce the rupture of cell walls and the release of proteins. Furthermore, the released proteins can be degraded into low-molecular components (e.g. FAAs) by low-level alkaline and ultrasonic treatment. However, the amount and species of FAAs that can be harvested have not been investigated before.
DOM species in the A&MWAS process
Figure S3 shows the EEM spectra of the samples obtained in the treatment of A&MWAS and the corresponding FRI analysis data. Similar to the results for A&UWAS, there were three peaks in A&MWAS with the highest peak located in region IV, which is representing SMP-like substances. A noticeable decline in HS-like substances (region V) in the supernatant was observed after the A&MWAS treatment. Based on the FRI analysis, the total fraction of region V and region III decreased from 44% to 36% after the combined treatment. This indicates that the biodegradability of sludge was improved further. However, the proportion of the organic matter refractory to biodegradation was 36%, which was higher than that for A&UWAS with 16%. In addition, the increase in the aromatic protein fraction implies that the protein components of sludge may be degraded by all types of treatment. However, a further conversion of protein components needs additional investigation.
Differentiation of FAAs from protein component degradation
The changes in various FAA species in the supernatants of the A&UPS and A&UWAS assays are shown in Figure 4. The total concentration of FAAs in the raw sludge supernatants was lower than 10 μmol L−1 (5.34 ± 0.01 and 0.71 ± 0.08 μmol L−1 for PS and WAS, respectively). Compared with the total amount of FAAs in the raw sludge, the total amount of FAAs increased in the supernatants of A&UPS and A&UWAS by 737- and 9557-fold for PS (3,935.66 ± 31.37 μmol L−1) and WAS (6,785.23 ± 8.40 μmol L−1), respectively. For example, the concentration of aspartate in the supernatant of PS increased 640 times (from 0.4 ± 0.002 μmol L−1 to 262 ± 5 μmol L−1). Alanine among the different types of released FAAs was highest in the supernatants of A&UPS and A&UWAS. For example, the concentrations of alanine in the supernatants of the pre-and post-sonicated PS were 0.9 ± 0.002 μmol L−1 and 667 ± 3 μmol L−1, respectively. In addition, the low-level alkaline and ultrasonic treatment resulted in the emergence of new FAA species in the supernatants. For example the concentration of leucine increased to 380 ± 6 μmol L−1 (PS) or 795 ± 1 μmol L−1 (WAS), respectively. In our study, the ultrasonic density was 0.31 W mL−1, which means that the contribution of hydroxyl radicals to an oxidation might be within 19% (Wang et al. 2005). In addition, hydroxyl radical can oxidize peptide bonds to release FAAs (Song et al. 2006). For PS, the concentrations of organic matter and protein decrease along the sewer delivery length. Complex VFAs and FAAs can be produced by specific microbial communities (Jin et al. 2018) in the sewer system. Therefore, low-weight molecular organic matter can be adsorbed physically by the primary particles of PS, which is resulting in a large amount of FAA production from a low energy input during the A&UPS.
Surprisingly, although the SE of the A&MWAS treatment is 1/3 higher than the SE of A&UWAS, a lower amount of FAAs was released after A&MWAS at the same pH, as shown in Figure S4. The total concentration of FAAs (1,990 ± 12 μmol L−1) in A&MWAS is only 29% of A&UWAS. Therefore, the disintegration effect of the alkaline and mechanical cutting treatments on sludge is mainly resulting from the hydraulic shear forces and centrifugal force owing to the high-speed rotation. However, except for the hydraulic shear forces, the collapse of cavitation bubbles and the oxidation of hydroxyl radicals arising from sonication also cause the degradation of proteins. More importantly, NaOH catalyzes the production of hydroxyl radicals, which further enhance the hydrolysis of proteins to amino acids under ultrasonic and alkaline treatment (Tian et al. 2015).
Research on extracting and utilizing amino acids from sludge to produce value-added products has been reported (Liu et al. 2009; Su et al. 2014). However, to the best of our knowledge, the use of an amino acid-rich sludge supernatant as a carbon source is another novel environmentally friendly and value-added pathway. Liu et al. utilized hot HCl solution (pH 0.5, 121 °C, 5 h) to digest WAS to obtain a protein solution, and then further hydrolyzed the solution to obtain amino acids under hot and acidic conditions (120 °C, 10 h) (Liu et al. 2009). Similarly, Su et al. extracted amino acids from WAS under hot acidic conditions (100 °C, 14 h) in their corrosion inhibition experiments (Su et al. 2014). Although a large amount of FAAs can be obtained from the low-level alkaline and ultrasonic disintegration of sludge, the potential function of these sludge-derived FAAs has not been reported.
Bioavailability of sludge-derived protein components (FAAs)
Different types of sludge-derived carbon sources were added to the BNR biomass in batch tests under anaerobic conditions to test their bioavailability. The biomass was enriched with conventional PAOs and denitrifying bacteria, as shown in Figure S5. The uptake of the FAAs is shown in Figure 5.
For both sludge carbon sources derived from PS and WAS the FAAs – aspartate, serine, arginine, glutamate, and histidine – were the best utilized ones. The uptake seems to be related to the chemical properties of the FAAs. Acidic and alkaline FAAs (aspartate, glutamate, followed by arginine) were the most used by the BNR biomass. Serine is an amino acid with a high uptake belonging to polar FAAs. After the low-level alkaline and ultrasonic treatment hydrophobic FAAs accounted for a high fraction of the sludge-derived sonication products (55 ± 0.5% for PS and 61 ± 0.003% for WAS). However, the BNR communities used them to a limited extent: only 18 ± 0.6% or 24 ± 0.3% uptake occurred from PS or WAS, respectively.
Therefore, the uptake of the FAAs was classified according to their acidity, basicity, and hydrophilicity, as indicated by the background colors of the columns in Figure 5. The carbon number and the side chain structure of FAAs jointly determine their utilization efficiency by the BNR communities. Thus, further investigations are required. Sakamoto et al. (2014) used Ralstonia eutropha to test the bioavailability of twenty natural amino acids as a carbon source for the synthesis of PHAs. They found that leucine was the best one. Investigations on single amino acids with different characteristics as alternative carbon sources for the BNR process have attracted much attention (Li et al. 2018; Rey-Martinez et al. 2019). Another attractive FAA is glycine, which shows moderate absorption. Studies have shown that Candidatus Accumulibacter can use many metabolic pathways including glycine cleavage and to produce acetyl coenzyme A (used to synthesize PHAs). This causes exogenous absorption of glycine. The use of glycine alone as a carbon source causes the highest P release relative to other FAAs from full-scale biomass (Oyserman et al. 2016). However, whether this special type of mixed carbon source affects biological communities and its effect on biological phosphorus removal need further investigation.
Research implications for future BNR process development
A schematic process diagram of deriving carbon sources from PS and (WAS) and their potential application for BNR enhancement and P recovery in the main stream of a WWTP is depicted in Figure 6. In this configuration, the FAAs-enriched sludge carbon source (e.g. derived from A&UPS) can benefit from the accumulation of some putative (D)PAOs (Nguyen et al. 2015; Liu et al. 2019). Since the FAAs can be taken up directly from substrates in solution and accumulated as alternative intracellular carbon sources, they will fuel phosphorus uptake in some putative PAO species (e.g. Tetrasphaera) (Liu et al. 2019) and can even improve the stress resistance of putative PAOs (e.g. Microlunatus) (Feehily & Karatzas 2013). Although knowledge gaps exist regarding unconventional PAOs (e.g. Tetrasphaera) in the BNR processes (Liu et al. 2019), once their metabolism is determined, the carbon source-containing FAAs, especially the sludge-derived carbon source discussed in this study, may become a viable supplement for the process.
In addition, the development and application of main stream anammox is an emerging trend in the BNR system. The combined low-level alkaline and ultrasonic treatment may supplement sludge-derived carbon or nitrogen sources for partial-denitrification or in the mainstream anammox process, thus facilitating the removal of N in a cost-effective way. Dilute influent ammonium and insufficient nitrite are the bottleneck factors for maintaining a stable anammox population in the mainstream anammox process (Du et al. 2020). Compared with the partial nitrification-anammox process, the partial denitrification-anammox process has advantages in being capable of providing sufficient and stable ammonium and nitrite supplements at ambient or low temperatures (Du et al. 2017). However, additional carbon sources are still needed in this process. Since limited amounts of ammonium are released simultaneously with the carbon sources (Figure S2), they may provide sufficient carbon sources for partial denitrification and ammonium for the pending mainstream partial denitrification-anammox process. In addition, recent progress based on the co-existence of anammox and denitrification has been reported (Wang et al. 2016; Ji et al. 2020). In such a mixed anammox-based system, heterotrophic denitrifiers and anammox bacteria benefit from their synergistic growth relationship: heterotrophic denitrifiers utilize organic carbon sources to produce nitrite and alleviate the inhibitory effect of organic matter, while the anammox bacteria provide nitrate for heterotrophic denitrifiers.
The application of sludge fermentate or seeding the fermentation sludge in the reactor to provide easily biodegradable SCOD has frequently been reported in recent years (Wang et al. 2016; Ji et al. 2020). However, the release of a large fraction of easily biodegradable SCOD (e.g. SMP-like) and the alternative FAAs as carbon source derived from sludge with a low energy input and a short processing time were also achieved in this study. Therefore, the sludge-derived carbon source preparation process presented here can be applied to the traditional BNR process or potentially in the pending varied mainstream anammox process in the future.
Combined low-level alkaline and mechanical processes can induce DOM release from sludge and the SMP-like substances (predominant DOM) increase up to 23%. However, high release of FAAs was only detected in the supernatant of A&UPS and A&UWAS. The acidic or alkaline FAAs were the best bioavailable FAAs to the BNR communities. The concept of producing a sludge-derived carbon source on site for further application in the BNR process can be a strategy for solving the problem of an insufficient carbon source encountered in some WWTPs. Also it may be a possibility for retrofitting the conventional BNR process into an EBPR combined with a P recovery process.
The authors thank the financial supporting program of National Natural Science Foundation of China (21777024) and National Key Research and Development project (2019YFC0408503).
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.