Efficacy of a high-pressure jet device for excess sludge reduction in a conventional activated sludge process: Pilot-scale demonstration

Accepted Manuscript Efficacy of a high-pressure jet device for excess sludge reduction in a conventional activated sludge process: Pilot-scale demonstration Hiroyuki Yoshino, Toshikazu Suenaga, Tadahiro Fujii, Tomoyuki Hori, Akihiko Terada, Masaaki Hosomi PII: DOI: Reference:

S1385-8947(17)30828-8 http://dx.doi.org/10.1016/j.cej.2017.05.084 CEJ 16982

To appear in:

Chemical Engineering Journal

Received Date: Revised Date: Accepted Date:

8 December 2016 25 April 2017 13 May 2017

Please cite this article as: H. Yoshino, T. Suenaga, T. Fujii, T. Hori, A. Terada, M. Hosomi, Efficacy of a highpressure jet device for excess sludge reduction in a conventional activated sludge process: Pilot-scale demonstration, Chemical Engineering Journal (2017), doi: http://dx.doi.org/10.1016/j.cej.2017.05.084

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Efficacy of a high-pressure jet device for excess sludge reduction in a conventional activated sludge process: Pilot-scale demonstration

Running title: Efficacy of a high-pressure jet device for excess sludge reduction

Hiroyuki Yoshinoa, Toshikazu Suenagaa, Tadahiro Fujiib, Tomoyuki Horic, Akihiko Teradaa*, Masaaki Hosomia

a

Department of Chemical Engineering, Tokyo University of Agriculture and Technology, Naka-cho

2-24-16 Koganei, Tokyo 184-8588, Japan b

DPK, 5780-8, Shinyoshida, Tsuzuki, Yokohama, Kanagawa 223-0056, Japan

c

Environmental Management Research Institute, National Institute of Advanced Industrial Science

and Technology (AIST), Onogawa 16-1, Tsukuba, Ibaraki 305-8569, Japan

*Corresponding author: [email protected] Tel/Fax: +81-42-388-7069/+81-42-388-7731

Abstract Application of a high-pressure jet device (HPJD) is promising for the reduction of excess activated sludge from aeration tanks in wastewater treatment plants. Nevertheless, proving the feasibility of an HPJD installed in the return line of a conventional activated sludge (CAS) process remains a challenge. We investigated differences in two pilot-scale CAS processes with or without an HPJD to demonstrate its influence on waste excess sludge reduction and wastewater treatment during operation for 108 days. The application of the HPJD (designated CAS-HPJD) reduced waste activated sludge by 65% compared with the standard CAS. Despite a slight increase in suspended solids concentration from the CAS-HPJD effluent, the total amount of excess sludge from CAS-HPJD was 23% lower than that of CAS. These two CAS processes exhibited comparable performances in the removal of organic carbon, ammonium and phosphate, indicating that sludge solubilization by the HPJD did not undermine microbial activities in activated sludge. In addition, the HPJD allowed oxygenation of the aerobic tank because of intensive mixing and subsequently produced fine bubbles, resulting in a reduction in the air flow rate of 35%. High-throughput Illumina sequencing of prokaryote 16S rRNA genes and Sanger-sequencing of cloned eukaryote 18S rRNA genes revealed that microbial community compositions in the two systems were distinct and were related to the HPJD treatment.

Keywords Activated sludge system, Sludge reduction, High-pressure jet device, Microbial community composition, Oxygen supply

1

1. Introduction The conventional activated sludge (CAS) process has been widely implemented for municipal wastewater treatment to deter negative impacts by discharge of untreated wastewater to water bodies. Despite being used all over the world, the CAS process has faced challenges with high excess sludge production and disposal. The cost for treatment and disposal of excess sludge accounts for about 40% of the total for the operation of a wastewater treatment plant (WWTP) [1]. An increase in the number of WWTPs eventually results in the continuous increase in excess sludge, which rather entails reduction of excess sludge. For example, the amount of excess sewage sludge is increasing by an average of 5% every year in China [2]. In addition, improving WWTPs, driven by the implementation of legislations and directives, e.g. the Urban Waste Water Treatment Directives in the EU, has elevated the amount of sludge production and it will expectedly increase further [3]. Given this context, the development of excess sludge reduction technologies for wastewater treatment to alleviate the economic burden for WWTP operations is of great importance. As an excess sludge reduction technology, a lysis-cryptic growth method has been developed and implemented in practice. The main concept of this method is decomposition of activated sludge into biodegradable compounds by either a chemical, physicochemical or biological unit installed in the return activated sludge line, e.g., ozonation [4], thermal treatment, bead-mill and anaerobic digestion [5,6]. Ozonation provides high sludge reduction efficiency (30%–100%); however, economic and technical problems, e.g., high capital cost, increase in effluent COD and low sludge settlability, remain unresolved [4]. Thermal treatment achieves high sludge reduction efficiency (70%) but requires a robust countermeasure against reactor corrosion and neutralizations [7]. Anaerobic digestion systems are based on the consecutive biotransformation of sludge into methane [8], entailing longer time to stabilize sludge reduction than other abiotic technologies. A high-pressure jet device (HPJD) is one of the most promising solutions to achieve cost-effective 2

and small-footprint excess sludge reduction. An HPJD consists of a high-pressure pump and pipe with two narrow and broader diameters in the middle and at both edges, respectively, to disrupt bacterial cell walls and membranes (Fig. S1) [9]. The high-pressure pump conveys excess activated sludge at a high lateral velocity and applied pressure (6 MPa) whereas it is also supplied from the top port in conjunction with sucking air. The two entry points of excess activated sludge at one end of the pipe generates fine bubbles resulting in cavitation, high friction of bacterial cells and collision with the steel wall at the other side of the pipe [10]. The unique configuration of an HPJD confers multiple physical injuries to microbial cells by friction, cavitation and collision. Preliminary studies have manifested cost-effective and high-throughput sludge reduction by the HPJD [9] and plausible cell disruption mechanisms for Escherichia coli and Bacillus subtilis [10-12]. Given these findings, an HPJD could be a promising lysis-cryptic growth method and its installation in the return line of a CAS process could potentially accomplish the effective reduction of excess sludge. This study was undertaken to investigate the feasibility of an HPJD installed in a CAS process (defined as CAS-HPJD) as an efficient lysis-cryptic growth method. To this end, two pilot-scale CAS processes with or without the HPJD installed in the return line were constructed and operated by receiving a municipal wastewater after the primary sedimentation tank. The effectiveness of the HPJD against excess sludge reduction, fate of phosphorus, floc size in activated sludge and microbial community structure was evaluated.

2. Materials and Methods 2.1 Overview of pilot-scale CAS process Two pilot-scale CAS processes were constructed and operated for 108 days (from May 2013 to August 2013) at a municipal wastewater treatment plant (Kitatama-Ichigo Water Reclamation Center, Fuchu, Tokyo, Japan). Each CAS process comprised an aeration tank (1 m3), a 3

sedimentation tank (0.8 m3) and a concentrated activated sludge return line connecting both tanks, as shown in Fig. 1. The difference between the two CAS processes stemmed from the presence or absence of an HPJD. The HPJD was installed into the return activated sludge line of the system (CAS-HPJD). The dimensions of the HPJD were 1 m in length, with 13 mm and 28 mm pipe diameters in the middle and at both edges, respectively. Activated sludge was transferred by a high-pressure pump (HPJ-160, Tsurumi, Osaka, Japan) via a nozzle ( 1 mm) set at one edge of the pipe. The energy output of the pump was adjusted to ensure an applied pressure of 3–4 MPa in the HPJD. In the CAS-HPJD, the return activated sludge line was bifurcated and the same amount of concentrated activated sludge (5 L/min for each system) was supplied to both the pump and an injection port of the HPJD as illustrated in Fig. 1. The HPJD was operated in an intermittent manner by turning on (1 min) and off (6 min) the pump to achieve a ratio of return activated sludge of 0.4. Activated sludge in the WWTP was inoculated at the start of the CAS operation. 2.2 Operating conditions of the activated sludge systems Reactor operating conditions and the influent wastewater characteristics are shown in Tables 1 and 2, respectively. The influent wastewater was drawn from the overflow of the primary sedimentation tank in the WWTP. Solid retention time (SRT) was determined to achieve comparable mixed liquor suspended solids (MLSS) concentrations in the CAS and CAS-HPJD. The amount of cumulative excess activated sludge by weight was evaluated by multiplying the volumetric withdrawal rate of waste activated sludge and its MLSS concentration. Air flow rates were controlled at 28–38 L/min and 18–25 L/min in the CAS and CAS-HPJD, respectively, to maintain a dissolved oxygen (DO) concentration of 1.0–1.5 mg/L in the aeration tank. 2.3 Chemical analyses Influent and effluent samples were taken three times per week. All samples were filtered through a 0.45 µm glass filter before measurement of wastewater constituents. Dissolved organic carbon 4

(DOC) and total nitrogen (TN) were measured by a TOC analyzer (TOC5000A, Shimadzu, Kyoto, Japan). Concentrations of nitrate (NO3−), nitrite (NO2−), ammonium (NH4+) and phosphate (PO43−) ions were measured by ion chromatography (ICS-90, Thermo Scientific, Waltham, MA, USA). Total phosphorus (TP) was measured by a potassium peroxodisulfate decomposition method [13]. pH, DO and temperature were measured by a multi-meter (D55, HORIBA, Kyoto, Japan). 2.4 Activated sludge sampling Activated sludge samples were taken three times per week. Sludge settlability was evaluated by the sludge volume index (SVI). MLSS concentration and mixed liquor volatile suspended solids (MLVSS) was measured through a 0.45 µm glass filter according to the standard protocol as described elsewhere [13]. Particle size of activated sludge was measured by a particle size analyzer (Partica LA-950V2, HORIBA, Kyoto, Japan). For this purpose, a semiconductor laser (650 nm) and LED (405 nm) were applied with a measurement size range of 0.01–3000 µm. 2.5 Prokaryotic community structure analysis DNA was extracted from 50 mg of activated sludge with the ISOIL Bead Beating Kit (Nippon Gene, Tokyo, Japan) according to the manufacturer’s instructions. Extracted DNA was diluted to 2 ng/µL for PCR. Primers targeting the V4 hypervariable region of 16S rRNA genes for all bacteria and archaea, i.e., 515F (5’-GTGCCAGCMGCCGCGG-3’) and 806R (5’-CCGTCAATTCMTTTRAGTTT-3’) were used. The reverse primer was labeled with a 12-bp barcode [14]. Reaction conditions were as follows [15]: initial denaturation at 98 C for 20 sec, 28 cycles consisting of 98 C for 10 sec, 53 C for 30 sec, 72 C for 120 sec and final extension at 72 C for 7 min. TaKaRa ExTaq HotStart (TaKaRa, Tokyo, Japan) was used as DNA polymerase. PCR amplicons were purified using the AMPure XP Kit (Beckman Coulter, Brea, CA, USA) and Wizard PCR clean-up system (Promega, Madison, WI, USA). The concentration and purity of extracted DNA was measured on a Qubit fluorometer (Thermo Fisher Scientific, Waltham, MA, 5

USA). A 300-cycle Miseq Reagent kit (Illumina, San Diego, CA, USA) was used for DNA amplicons and initial control (bacteriophage phiX; Illumina, CA), followed by sequencing on a Miseq sequencer (Illumina, San Diego, CA, USA). After PhiX, low-quality (Q <30) and chimeric sequences were removed as reported previously [16]. Retrieved sequences were assigned as operational taxonomic units (OTUs) by 97% similarity cut-off using the software QIIME version 1.7.0 [17]. 2.6 Eukaryotic community structure analysis DNA extraction for eukyaryotic community analysis was performed with the same protocol for the prokaryotic community analysis (section 2.5). The 18S rRNA gene primers targeting all eukaryotes, i.e., EukA (5'-AACCTGGTTGATCCTGCCAGT-3') and EukB (5'-TGATCCTTCTGCAGGTTCACCTAC-3') [18], and DNA polymerase TaKaRa ExTaq HotStart (TaKaRa, Tokyo, Japan) were used for PCR. Reaction conditions were as follows: initial denaturation at 98 C for 20 sec, 28 cycles consisting of 98 C for 10 sec, 53 C for 30 sec, 72 C for 120 sec and final extension at 72 C for 7 min. PCR amplicons were subject to cloning after purification by a Wizard PCR clean-up system (Promega Corporation). The pGEM-T (Easy) Vector System (Promega Corporation) was used for ligation. After the vector transformation of supercompetent E. coli cells, the suspension was spread onto an agar medium including 100 µg/mL ampicillin. White colonies were picked up and transferred to a 96-well plate. DNA was eluted by freezing and thawing. Eluted DNA was subject to PCR by vector primers SP6 and T7. Sixty-seven and 77 clones were obtained from the CAS-HPJD and CAS processes, respectively, and these clones were subjected to Sanger-sequencing with a DNA sequencer (3730xl DNA Analyzer; Applied Biosystems, MA, USA). Retrieved sequences were compared with sequences in GenBank using a BLAST search (http://blast.ncbi.nlm.nih.gov/Blast.cgi). 2.7 Statistical analysis 6

A two-tailed t-test was applied using the software SPSS (IBM Co., New York, USA) to confirm significant differences in acquired data between the CAS-HPJD and CAS processes. The rejection threshold was set at 5% (α = 0.05). 2.8 Accession numbers The sequence data acquired in this study were deposited in the MG-RAST database (http://metagenomics.anl.gov/) as “Microbial community transitions in a conventional activated sludge system installing a high pressure jet device” under the following ID: DRA005138.

3. Results 3.1 Wastewater treatment performance The performances of the CAS-HPJD and CAS processes are shown in Fig. 2. Removal efficiencies of DOC (Panel a), NH4+ (Panel b) and dissolved total nitrogen (DTN) (Panel c) were comparable between CAS and CAS-HPJD processes (p < 0.05). Throughout the operation, the average DOC removal efficiencies were 60.5 ± 8.26% and 54.8 ± 10.7% for the CAS-HPJD and CAS, respectively. The average NH4+ and DTN removal efficiencies were 91.3 ± 13.8% and 27.8 ± 16.2%, respectively, for the CAS-HPJD, and 92.2 ± 10.5% and 25.1 ± 15.8%, respectively, for the CAS. These removal efficiencies were not statistically significantly different (p > 0.05). Dissolved biochemical oxygen demand (BOD) removal efficiency changed little in both the systems (Table S1). On days 38 and 83, the pump in the CAS-HPJD malfunctioned, rendering it incapable of pumping return activated sludge to the aeration tank. On day 62, aeration was temporarily halted to suppress overgrowth of Daphnia sp. in the CAS. Despite these incidents, the reactor performances recovered in a short time. Taken together, the installation of the HPJD did not deteriorate the performance of the CAS in terms of the organic carbon and nitrogen removal. The concentrations of phosphate ions and TP and the composition of phosphorus constituents were 7

evaluated to elucidate the fate of phosphorus in the systems. Average effluent PO43− concentrations in the CAS-HPJD and CAS processes were similar (p < 0.05), at 1.66 ± 0.30 and 1.68 ± 0.34 mg-P/L, respectively (Fig. S2a). Effluent TP concentrations in the two processes were likewise similar (Fig. S2b). The release of phosphorus from activated sludge by HPJD treatment was detected but did not affect phosphorus concentrations in the effluent (for details, see Supplementary Information and Fig. S3). The average DO concentration in the aeration tank was 1.3 ± 0.4 mg/L in the CAS-HPJD, which was slightly but significantly higher than that in the CAS (1.0 ± 0.4 mg/L; p < 0.05; Fig. 2d). Aeration rate in the CAS-HPJD was reduced by 35% to ensure that the DO concentration was comparable with that of the CAS. 3.2 Effect of HPJD on activated sludge reduction Time series of MLSS concentrations and SVI values are shown in Fig. 3. MLSS concentration sharply decreased on days 38, 62 and 83 because of the pump malfunction and overgrowth of Daphnia sp. as described in the previous section. Nevertheless, no significant difference in MLSS concentration was observed between the CAS-HPJD and CAS processes (p > 0.05). Average MLVSS/MLSS ratios of the activated sludge in the CAS-HPJD and CAS were 0.90 ± 0.05 and 0.86 ± 0.04, respectively, and were not statistically different (p > 0.05). SVI in the CAS-HPJD was not significantly deteriorated throughout the operation whereas SVI sharply increased from day 6 to 27 in the CAS (Fig. 3b). Microscopic observation confirmed the presence of filamentous bacteria only in activated sludge of the CAS process (data not shown). The average SVIs, attained from the period when the operations of CAS and CAS-HPJD were stable but the dates (on days 20, 22 and 24) having bulking events in CAS were excluded, were 140 ± 50.1 mL/g and 101 ± 57.9 mL/g in the CAS-HPJD and CAS, respectively, and were not statistically different (p > 0.05). The average suspended solids concentration in the CAS-HPJD increased by 40% in comparison with that of the 8

CAS (Fig. S4). The time-dependent change in cumulative amount of excess activated sludge is shown in Fig. 4. The amount of waste activated sludge was 65% lower from the CAS-HPJD than from the CAS during the whole operation period. Taken the cumulative amount of suspended solids overflowed from the supernatant into consideration as excess activated sludge, percentage of excess sludge reduction by the HPJD throughout the operation was 23%. 3.3 Effect of the HPJD on floc size in activated sludge The time course of median diameter of activated sludge is shown in Fig. 5. Average diameter of activated sludge flocs was approximately 70 ± 8.75 µm on day 7 in both systems. In the CAS, median diameter of activated sludge sharply increased to approximately 170 ± 24.0 µm by day 20 and stayed constant until day 97, finally increasing up to 280 ± 26.0 µm at the end of the operation. The median diameter of activated sludge in the CAS-HPJD slightly decreased to approximately 50 µm on day 48 and was then constant until the end of the operation. 3.4 Transition of Prokaryotic community structure As many as 700–1200 OTUs were identified from 30,000–50,000 sequences of high-throughput sequencing based on 16S rRNA gene amplicons (Table S2). Alpha-diversity indices, i.e., Chao1, Shannon and Simpson reciprocal, of activated sludge in the CAS-HPJD and CAS before and after the operation were estimated (Table S2). The diversity indices were similar at the beginning of the reactor operation and increased only in the CAS at the end of the operation. Microbial community compositions at the phylum and family levels are shown in Fig. 6. At the phylum level, the most dominant phylum was Proteobacteria in both systems. Relative abundances of the phyla Firmicutes and Chloroflexi were lower in the CAS-HPJD (0.5% and 9.0%, respectively) than in the CAS (2.6% and 10.9%, respectively). On the contrary, the relative abundances of the phyla Chlorobi and Acidobacterium were higher in the CAS-HPJD (3.2% and 3.3%, respectively) than in the CAS 9

(0.04% and 0.8%, respectively). The family Sphingomonadaceae in the class Alphaproteobacteria was dominant with relative abundances of 19%–20% in both reactors on day 1. Bacterial community compositions between the two systems were not statistically different on day 1; however, they were distinctly different on day 108. Relative abundance of the most dominant genus was below 10% in the CAS, which is consistent with the diversity data that the evenness of microbial communities, based on the Shannon and Simpson reciprocal index, were higher in the CAS than in the CAS-HPJD (Table S2). In stark contrast, the most dominant genus (in the family Xanthomonadaceae) occupied 35% of the total population in the CAS-HPJD, contributing to the low evenness of microbial communities in this process. 3.5 Difference in Eukaryotic community structures Cloning and Sanger-sequencing analysis based on 18S rRNA genes was performed to investigate the eukaryote population (Fig. 7). We have confirmed that eukaryote communities in CAS and CAS-HPJD were identical because of the same inoculum used for this study (Data not shown). Eukaryote community structure was remarkably distinct between the CAS-HPJD and CAS processes on day 108. Fungi were confirmed as the most predominant (66/77 clones) in the CAS but far less abundant (7/68 clones) in the CAS-HPJD. Instead, Mesomycetozoa (40/68 clones) was dominant in the CAS-HPJD. Metazoa was only found in the CAS. Other clones that are closely related (similarity >97%) to Euglenozoa (7/68 clones) and Amoebozoa (5/68 clones) were detected in the CAS-HPJD. The relative abundance of protozoa was lower in the CAS than in the CAS-HPJD. To consolidate the results from the 18S rRNA gene analysis, microscopic observation was conducted to detect eukaryote members present in activated sludge samples. As a result, small eukaryotes with sizes of <100 µm, such as Bodo sp., were observed in the CAS-HPJD. Relatively large eukaryotes with sizes of >100 µm, e.g., the genus Macrobiotus, were frequently found in the CAS (Fig. S5). The combination of 18S rRNA gene-based cloning/sequencing and microscopic 10

observation identified the distinct eukaryotic niches shaped in the activated sludge depending on the installation of the HPJD.

4. Discussion 4.1 Sludge reduction by HPJD and influence on water quality This work presents the first pilot-scale demonstration that the amount of excess activated sludge was reduced by the installation of a high-pressure jet device (HPJD). During the 108-day operation fed with municipal wastewater after a primary sedimentation tank, the CAS-HPJD reduced the amount of waste activated sludge by 65% in comparison with the CAS alone. Previously, sludge reduction by the HPJD had been performed in batch tests, but these tests were incapable of demonstrating the efficacy of the HPJD connected with the CAS [9]. The reduction percentage of excess sludge and overflowed suspended solids by the HPJD was 23%. Given that a batch test by the HPJD to treat excess activated sludge showed approximately 40% reduction at maximum [9], the value obtained in the pilot-scale study (23%) should have been higher. This might be because of malfunction of the high-pressure pump and clogging of activated sludge in a nozzle that occasionally happened during the operation. The percentage of excess activated sludge reduction was still not as high as that by ozonation, which can achieve 100% excess sludge reduction [19]. Nevertheless, this study likewise demonstrated that the HPJD provided a function of oxygenation in the aeration tank, potentially saving operating costs for aeration. Actually, high oxygen supply is necessary in a lysis-cryptic growth system with ozonation because of an increase in DOC released from intracellular organic compounds [20]. This pilot-scale test exhibited that the HPJD has an oxygenation ability; the detailed discussion was described in section 4.3. Therefore, the oxygen enrichment effect by the HPJD might have facilitated the decomposition of DOC increased by the solubilization of activated sludge. Moreover, energy consumption of the HPJD for bacterial cell 11

disruption in activated sludge is much lower (1.18–3.53 kJ/g-MLSS) than those of other lysis-cryptic sludge reduction technologies (2.16–78.5 kJ/g-MLSS) [9]. Taken together, the pilot-scale demonstration manifested that the HPJD is cost-effective and therefore promising as a novel sludge reduction technology. The continuous operation of the CAS-HPJD and CAS processes showed that the performances of DOC and NH4+ removal were not affected by the HPJD as shown in Fig. 2a and b. This was not the case for a previous study implementing ozonation as a lysis-cryptic method, where dissolved COD removal was exacerbated [4]. On the contrary, it was reported that the increase in DOC concentration after the application of an HPJD to excess activated sludge was approximately 5% of total organic carbon [9], but the increased percentage of DOC could be offset because of the mineralization by activated sludge in an aeration tank. The MLSS in the return activated sludge and the return ratio in this study were approximately 4000 mg/L and 0.4, respectively. Herewith, the intrinsic dissolved organic carbon loading rate (ILRDOC) derived from activated sludge decomposition by the HPJD was calculated as: ILRDOC  X MLSS _ RAS  0.05 

M carbon  0.4 Q inf M cell

(Eq. 1)

where XMLSS_RAS is the MLSS concentration of the return activated sludge (g/m3); Qinf is the inflow rate of the incoming wastewater (m3/day); 0.05 is the ratio of the solubilization of activated sludge after the data by Suenaga et al. [9]; Mcarbon is the molecular weight of carbon content (60 g-C/mol) in a bacterial cell expressed as C5H7NO2 and Mcell is the molecular weight of a cell (113 g-C5H7NO2/mol). This equation provides an ILRDOC of approximately 170 g/day in the CAS-HPJD. Because the DOC concentration in effluent did not go as high in the CAS-HPJD as in the CAS (Fig. 2a), it is likely that DOC eluted by the microbial cell decomposition was mineralized by activated sludge. The specific dissolved organic carbon removal rate (SRRDOC) was defined as: 12

For the CAS-HPJD:

SRRDOC 

For the CAS:

SRRDOC 

(Cinf  Ceff )  Qin  ILRDOC X MLSS  V (Cinf  Ceff )  Qin

(Eq. 2) (Eq. 3)

X MLSS V

where Cinf and Ceff are DOC concentrations in influent and effluent, respectively; XMLSS is the MLSS concentration in the aeration tank; Qin is an influent flow rate (m3/day); and V is the aeration tank volume (m3). Assuming that DOC eluted by the HPJD treatment was decomposed, SRRDOC was 0.207 g-C /g-MLSS/day in the CAS-HPJD at an HRT of 6 h and MLSS of 950 mg/L. The obtained SRRDOC was higher than that without an HPJD (0.030 g-C /g-MLSS/day) by a factor of 7. The increase in the activity of DOC removal was plausibly caused by an increase in surface area triggered by sludge disaggregation as observed in Fig. 5. Reportedly, floc size is a driver of activity in activated sludge [21], supporting our result of the higher SRRDOC in the CAS-HPJD than in the CAS in this study. The maximum applied pressure for the HPJD treatment was 6 MPa, which is much lower than the counterpart for other lysis-cryptic methods, e.g., high pressure homogenizer (HPH) (70 MPa) [22]. Because of the low applied pressure in this study, deterioration of treated water quality because of the potential release of recalcitrant cell components was deterred. Phosphate was not removed in both the CAS-HPJD and CAS processes; however, there was no deterioration in phosphate removal in the CAS-HPJD when compared with the CAS. Discharge of phosphate has been reported as an important challenge for sludge solubilization [19, 23]. Although phosphate was not significantly leaked to the effluent in the CAS-HPJD in comparison with that in the CAS, phosphate elution was observed in batch tests as shown in Figs. S2 and S3. The discrepancy in phosphate release between the batch and continuous experiments is possibly because of the re-accumulation of phosphate into microbial cells of the activated sludge. Nonetheless, this is speculative, hence clarification of this discrepancy warrants future investigation. 4.2 Effect of HPJD on bacterial and eukaryote community structures 13

The microbial community analysis in this study manifested that the HPJD treatment selectively exerted a physical impact on numerous bacterial species present in activated sludge. This is particularly supported by the distinct bacterial community structure (Fig. 6) and diversity (Table S2) between the CAS-HPJD and CAS processes. It should be noted that microbial community structure in CAS was also changed after 108 day-operation likely due to temperature [24] and concentration and composition of influent wastewater [25]. Distinct differences in community structure were also observed for a lysis-cryptic system with ozonation [24] or alkaline treatment [26]. Although it remains challenging to ensure reproducibility of microbial community compositions in identically operated CASs [25], implementation of the lysis-cryptic system HPJD in the CAS would be a governing factor to determine microbial community composition. An interesting trait of the HPJD is selective bacterial disruption based on bacterial cell robustness. Xie et al. [11] noted that the susceptibilities of Gram-positive and Gram-negative bacterial cells to HPJD treatment were different. This result indicates that the community structure in activated sludge was shaped by means of the HPJD. On the contrary, it is less likely for ozonation to selectively structure microbial communities because the intensity of cell disruption by ozonation is much higher than that by the HPJD, resulting in the complete breakdown of bacterial cells regardless of their robustness. The time-course and high-resolution analysis on the transition of microbial communities is currently underway. Disaggregation of activated sludge flocs in conjunction with distinct bacterial community structure (Fig. 6) would likely affect eukaryote community structures (Fig. 7). Community structure analysis based on 18S rRNA genes revealed the presence of Fungi and Mesomycetozoa, which were not able to be detected by microscopic analysis. According to the cloning and sequencing results in Fig. 7, the disruption effect by the HPJD triggered the appearance of Fungi, resulting in stark differences in its relative abundances (10% and 85% in the CAS-HPJD and CAS processes, respectively). 14

Fungi have been frequently detected in activated sludge systems by 18S rRNA gene analysis [27, 28], corroborating the authenticity of our analysis. However, metabolic functions of Fungi have not been thoroughly clarified in activated sludge. Understanding their ecophysiology is required to determine the roles of Fungi in the CAS-HPJD. The microscopic observation revealed that the dominant eukaryotes detected in the CAS-HPJD were smaller in size than those in the CAS. It was demonstrated that the sizes of protozoan as predators were positively correlated with those of preys [29]. Analogously, our result suggests that the size of activated sludge flocs was correlated with that of the eukaryotes present. Given this, the emergence of Protozoa and Metazoa is likely governed by the degree of floc disaggregation by the HPJD (Fig. 5). Indeed, smaller eukaryotes, e.g., Bodo sp. and Monas sp., were often found in the system with the HPJD. Their potential as predators is unknown; therefore, the interaction of these eukaryotes and prokaryotes in activated sludge needs to be elucidated. In general, predation of Protozoa and Metazoa largely contributes to excess sludge reduction [30]. Hence, an increase in their abundance is of importance to lower the SS concentration in effluent [31]. A future challenge is to enrich these microorganisms to contribute to synergistic sludge reduction by the lysis-cryptic-based treatment with the HPJD and the associated eukaryote predation of disaggregated flocs. 4.3 Possibility of oxygenation and the challenge of installing an HPJD in a CAS Interestingly, the implementation of an HPJD provided a function of oxygen enrichment in the pilot-scale CAS test, resulting in a 35% reduction in the air flow rate in the CAS-HPJD when compared with the CAS (Table 1 and Fig. 2d). In addition to the primary advantage of the HPJD in terms of reduction of excess activated sludge, this ‘secondary’ advantage is of significance because aeration is responsible for the dominant operating cost in activated sludge systems as previously reported [32, 33]. The decreases in the excess sludge reduction and aeration energy are key issues in the development of a low-cost activated sludge system. Although the air flow rate was lower in the 15

CAS-HPJD than in the CAS, the DO concentration in the CAS-HPJD was still higher than that in the CAS (p < 0.05; Fig. 2d). The high DO concentration obtained in the CAS-HPJD was caused by the return activated sludge with fine bubbles generated by air aspiration during the HPJD operation. This oxygen enrichment function was put forward in our previous studies [9, 11] and this advantage was proven in the pilot-scale CAS in this study. Assuming that the aeration cost is 50% of the total operating cost of a WWTP, the cost of a WWTP could be reduced by 18% [33]. To the best of our knowledge, this is the first report to demonstrate that the pilot-scale HPJD system contributed to an increase in the DO concentration in the aeration tank. Although a relatively long SRT was applied in the CAS-HPJD (13 days), the SVI did not deteriorate in comparison with that in the CAS (Fig. 3b). This is probably because of the breakage of filamentous bacteria that has been reported as the causative agent for bulking [34]. In fact, the high-throughput sequencing of 16S rRNA gene amplicons revealed that the relative abundances of the phylum Chloroflexi accounted for 8.9% and 10.9% in the CAS-HPJD and CAS, respectively. Instead, the major downside of the HPJD installation in the CAS was that it led to higher SS concentration in the effluent than in the CAS alone (Fig. S2). The high SS concentration in the effluent of the CAS-HPJD was putatively derived from the dispersion of flocs in the activated sludge, maintaining a small floc size (Fig. 5). The small flocs partially overflowed from the CAS-HPJD system; solving this problem requires optimization of the HPJD-operating conditions and retrofitting of the sedimentation tank. This challenge has also been found in a previous study employing high-pressure homogenization (HPH) [22]. Currently, a continuous experiment using a retrofitted sedimentation tank is underway, to prevent excessive SS concentrations in the effluent.

5. Conclusion The installation of an HPJD into a CAS and its operation for 108 days clearly revealed that the 16

CAS-HPJD process reduced the excess activated sludge from the return line by 65% and the total SS from the system by 23% in comparison with those of the CAS. The HPJD did not undermine bacterial metabolic activities in the aeration tank, resulting in comparable efficiencies of DOC and NH4+ removal between the CAS-HPJD and CAS alone. In addition to the achievement of cost-effective sludge reduction, the HPJD conferred intensive mixing and resultant fine bubbles, leading to the reduction of cumulative air flow by 35%. The intensive mixing induced by the HPJD disaggregated sludge flocs, and the relatively long SRT likely shaped distinct prokaryote and eukaryote community structures at the end of the operation. Taken together, our comparative pilot-scale study of the CAS-HPJD and CAS demonstrated the effectiveness of the HPJD in reducing excess activated sludge and providing oxygen enrichment in the aeration tank as the nexus of a small-footprint and cost-effective treatment system.

Acknowledgements This work was supported by the Adaptable and Seamless Technology Transfer Program through the target-driven R&D of Japan Science and Technology Agency (AS2311489E), a Grant-in-Aid for Scientific Research (A) (26249076) and a Grant-in-Aid for JSPS Fellows (201508427) from the Ministry of Education, Culture, Sports, Science and Technology of Japan. The authors also acknowledged Prof. Ryuichi Sudo, Prof. Osamu Nishimura and Dr. Nobuo Tanaka at Tohoku University for guidance of eukaryotic detection by microscopy, Prof. Kengo Kubota at Tohoku University for discussion on the primers for eukaryotes, Dr. Tomo Aoyagi at National Institute of Advanced Industrial Science and Technology for technical guidance for high-throughput sequencing, Prof. Hidehiro Kamiya for technical guidance for particle size analysis, and Mr. Hatsuhiro Matsuda and Ms. Noriko Moriya at Bureau of Sewerage in Tokyo Metropolitan 17

Government for assistance in operating the pilot-scale reactor systems.

Conflicts of Interest The authors declare that there are no conflicts of interest relating to this manuscript.

References [1] A.H. Hassani, A. Nejaei, A. Torabian, Excess sludge minimization in conventional activated sludge pilot plant by three chemical matters, Int. J. Environ. Res. 5 (2011) 981-988. [2] B. Tang, L. Yu, S. Huang, J Luo, Y Zhuo, Energy efficiency of pre-treating excess sewage sledge with microwave irradiation, Biores. Technol. 101 (2010) 5092-5097. [3] A. Kelessidis, A.S. Stasinakis, Comparative study of the methods used for treatment and final disposal of sewage sludge in European countries, Waste Manage 32 (2012) 1186-1195. [4] Y. Wei, R.T. Van Houten, A.R. Borger, D.H. Eikelboom, Y. Fan, Minimization of excess sludge production for biological wastewater treatment, Water Res. 37 (2003) 4453-4467. [5] J.T. Novak, D.H. Chon, B.A. Curtis, M. Doyle, Biological solids reduction using the Cannibal process, Water Environ. Res. 79 (2007) 2380-2386. [6] A. Khursheed, A.A. Kazmi, Retrospective of ecological approaches to excess sludge reduction, Water Res. 45 (2011) 4287-4310. [7] V.K. Tyagi, S.-L. Lo, Application of physico-chemical pretreatment methods to enhance the sludge disintegration and subsequent anaerobic digestion: an up to date review, Reviews in Environmental Science and Bio/Technology 10 (2011) 215-242. [8] C. Park, D.H. Chon, High-Rate Anaerobic side-stream reactor (ASSR) processes to minimize 18

the production of excess sludge, Water Environ. Res. 87 (2015) 2090-2097. [9] T. Suenaga, M. Nishimura, H. Yoshino, H. Kato, M. Nonokuchi, T. Fujii, H. Satoh, A. Terada, M. Hosomi, High-pressure jet device for activated sludge reduction; Feasibility of sludge solubilization, Biochem. Eng. J. 100 (2015) 1-8. [10] L. Xie, Q. Bao, A. Terada, M. Hosomi, Single-cell analysis of the disruption of bacteria with a high-pressure jet device: An application of atomic force microscopy, Chem. Eng. J. 306 (2016) 1099-1108. [11] L. Xie, Q. Bao, T. Suenaga, H. Yoshino, A. Terada, M. Hosomi, Identification of a predominant effect on bacterial cell disruption and released organic matters by a high-pressure jet device, Biochem. Eng. J. 101 (2015) 220-227. [12] L. Xie, A. Terada, M. Hosomi, Disentangling the multiple effects of a novel high pressure jet device upon bacterial cell disruption, Chem. Eng. J. (in press) doi.org/10.1016/j.cej.2017.04.067 [13] A.D. Eaton, L.S. Clesceri, A.E. Greenberg, M.A.H. Franson, American Public Health Association., American Water Works Association., Water Environment Federation., Standard methods for the examination of water and wastewater, 19th ed., American Public Health Association, Washington, DC, 1995. [14] J.G. Caporaso, C.L. Lauber, W.A. Walters, D. Berg-Lyons, J. Huntley, N. Fierer, S.M. Owens, J. Betley, L. Fraser, M. Bauer, N. Gormley, J.A. Gilbert, G. Smith, R. Knight, Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms, Isme. J. 6 (2012) 1621-1624. [15] T. Aoyagi, S. Hanada, H. Itoh, Y. Sato, A. Ogata, M.W. Friedrich, Y. Kikuchi, T. Hori, Ultra-high-sensitivity stable-isotope probing of rRNA by high-throughput sequencing of isopycnic centrifugation gradients, Environ. Microbiol. Rep. 7 (2015) 282-287. [16] H. Itoh, M. Aita, A. Nagayama, X.Y. Meng, Y. Kamagata, R. Navarro, T. Hori, S. Ohgiya, Y. 19

Kikuchi, Evidence of environmental and vertical transmission of Burkholderia symbionts in the oriental chinch bug, Cavelerius saccharivorus (Heteroptera: Blissidae), Appl. Environ. Microbiol. 80 (2014) 5974-5983. [17] J.G. Caporaso, J. Kuczynski, J. Stombaugh, K. Bittinger, F.D. Bushman, E.K. Costello, N. Fierer, A.G. Pena, J.K. Goodrich, J.I. Gordon, G.A. Huttley, S.T. Kelley, D. Knights, J.E. Koenig, R.E. Ley, C.A. Lozupone, D. McDonald, B.D. Muegge, M. Pirrung, J. Reeder, J.R. Sevinsky, P.J. Tumbaugh, W.A. Walters, J. Widmann, T. Yatsunenko, J. Zaneveld, R. Knight, QIIME allows analysis of high-throughput community sequencing data, Nat Methods 7 (2010) 335-336. [18] B. Diez, C. Pedros-Alio, T.L. Marsh, R. Massana, Application of denaturing gradient gel electrophoresis (DGGE) to study the diversity of marine picoeukaryotic assemblages and comparison of DGGE with other molecular techniques, Appl. Environ. Microbiol. 67 (2001) 2942-2951. [19] S.T. Yan, L.B. Chu, X.H. Xing, A.F. Yu, X.L. Sun, B. Jurcik, Analysis of the mechanism of sludge ozonation by a combination of biological and chemical approaches, Water Res. 43 (2009) 195-203. [20] L. Chu, S. Yan, X.H. Xing, X. Sun, B. Jurcik, Progress and perspectives of sludge ozonation as a powerful pretreatment method for minimization of excess sludge production, Water Res. 43 (2009) 1811-1822. [21] N. Cicek, J.P. Franco, M.T. Suidan, V. Urbain, J. Manem, Characterization and comparison of a membrane bioreactor and a conventional activated-sludge system in the treatment of wastewater containing high-molecular-weight compounds, Water Environ. Res. 71 (1999) 64-70. [22] W. Lan, Y. Li, Q. Bi, Y. Hu, Reduction of excess sludge production in sequencing batch reactor (SBR) by lysis-cryptic growth using homogenization disruption, Biores. Technol. 134 (2013) 43-50. 20

[23] P. Sui, F. Nishimura, H. Nagare, T. Hidaka, Y. Nakagawa, H. Tsuno, Behavior of inorganic elements during sludge ozonation and their effects on sludge solubilization, Water Res. 45 (2011) 2029-2037. [24] S.T. Yan, H. Zheng, A. Li, X. Zhang, X.H. Xing, L.B. Chu, G. Ding, X.L. Sun, B. Jurcik, Systematic analysis of biochemical performance and the microbial community of an activated sludge process using ozone-treated sludge for sludge reduction, Biores. Technol. 100 (2009) 5002-5009. [25] S. Isazadeh, P.O. Ozcer, D. Frigon, Microbial community structure of wastewater treatment subjected to high mortality rate due to ozonation of return activated sludge, J. Appl. Microbiol. 117 (2014) 587-596. [26] P. Yan, J. Wang, Y.P. Chen, F.Y. Ji, Y. Shen, F. Fang, Y.L. Lin, J.S. Guo, H. Zhang, W.J. Ouyang, Investigation of microbial community structure in an advanced activated sludge side-stream reactor process with alkaline treatment, Int. Biodeter. Biodegr. 104 (2015) 356-362. [27] K. Matsunaga, K. Kubota, H. Harada, Molecular diversity of eukaryotes in municipal wastewater treatment processes as revealed by 18S rRNA gene analysis, Microbes Environ. 29 (2014) 401-407. [28] T.N. Evans, R.J. Seviour, Estimating biodiversity of fungi in activated sludge communities using culture-independent methods, Microb. Ecol. 63 (2012) 773-786. [29] M.W. Hahn, M.G. Hofle, Grazing of protozoa and its effect on populations of aquatic bacteria, FEMS Microbiol. Ecol. 35 (2001) 113-121. [30] X. Hao, Q. Wang, Y. Cao, M.C. van Loosdrecht, Evaluating sludge minimization caused by predation and viral infection based on the extended activated sludge model No. 2d, Water Res. 45 (2011) 5130-5140. [31] P. Holubar, T. Grudke, A. Moser, B. Strenn, R. Braun, Effects of bacterivorous ciliated 21

protozoans on degradation efficiency of a petrochemical activated sludge process, Water Res. 34 (2000) 2051-2060. [32] L. Rieger, R.M. Jones, P.L. Dold, C.B. Bott, Ammonia-based feedforward and feedback aeration control in activated sludge processes, Water Environ. Res. 86 (2014) 63-73. [33] L. Amand, G. Olsson, B. Carlsson, Aeration control - a review, Water Sci. Technol. 67 (2013) 2374-2398. [34] L.B.M. Speirs, S.J. McIlroy, S. Petrovski, R.J. Seviour, The activated sludge bulking filament Eikelboom morphotype 0914 is a member of the Chloroflexi, Environ. Microbiol. Rep. 3 (2011) 159-165.

22

Figure captions Fig. 1 Schematic diagram of (a) the pilot-scale conventional activated sludge (CAS) system and (b) the system with an HPJD (CAS-HPJD)

Fig. 2 Influent and effluent water qualities of the conventional activated sludge (CAS)-HPJD and CAS: (a) dissolved organic carbon (DOC) removal efficiency; (b) Ammonium removal efficiency; (c) dissolved total nitrogen (DTN) removal efficiency; and (d) pH and dissolved oxygen (DO) concentration

Fig. 3 Time variation of (a) mixed liquor suspended solids (MLSS) concentration and (b) sludge volume index (SVI) in the conventional activated sludge (CAS)-HPJD and CAS

Fig. 4 Cumulative amount of excess activated sludge from the conventional activated sludge (CAS)-HPJD and CAS

Fig. 5 Time variation of median diameter of activated sludge flocs in the conventional activated sludge (CAS)-HPJD and CAS

Fig. 6 Bacterial community compositions at the phylum and family levels in the conventional activated sludge (CAS)-HPJD and CAS

Fig. 7 Eukaryotic community compositions in (a) the conventional activated sludge (CAS) and (b) the CAS-HPJD on day 108 (n represents the number of clones)

23

Excess sludge

(a)CAS Influent

Sedimentation tank

Aeration P tank

Aeration

1 m3

Effluent

0.8 m3

P

Return sludge

(b) CAS-HPJD Influent

P

Excess sludge Sedimentation tank

Aeration P tank

Aeration

1 m3

0.8 m3

P

P HPJD

Return sludge

High pressure pump

Fig. 1

24

P

Effluent

100

(a)

90

90

80

80

Ammonia removal efficiency [%]

70 60 50 40 30

20

(b)

70 60 50 40 30

20

CAS-HPJD 10

CAS-HPJD 10

CAS

(a) 0

0 0

20

40

60

80

100

120

0

20

40

Time [day] 100

60

80

100

120

Time [day] 10

(c) (c)

3.5

(d)(d)

CAS-HPJD 9

pH

80

8

DO

70

7

60

6

90

CAS

pH [-]

DTN removal [%]

CAS

Trouble

(b)

50

CAS-HPJD CAS CAS-HPJD CAS

3

2.5

2

5

40

4

30

3

20

2

10

1

1.5

1

0.5

0

0 0

20

40

60 80 Time [day]

100

0 0

120

20

40

60

Time [day]

Fig. 2 (e)

25

80

100

120

DO concentration [mg/L]

DOC removal efficiency [%]

100

1600

600

CAS-HPJD

(a)

(b)

CAS-HPJD

CAS

1400

CAS 500

1200

SVI [mL/g]

MLSS [mg/L]

400

1000 800

300

600 200

400

Pump trouble

100

200

Stop aeration 0

0 0

20

40

60

80

100

120

0

20

40

60

Time [day]

Time [day]

Fig. 3

26

80

100

120

10000 9000

Cumulative excesssludge [g]

8000 7000 6000 5000 4000 3000 2000 CAS-HPJD 1000 CAS 0 0

20

40

60

Time [day]

Fig. 4

27

80

100

120

350 CAS-HPJD

Median diameter [μm]

300

CAS

250 200 150 100 50

0 0

50

100 Time [day] Fig. 5

28

150

(a) Other

90%

TM7

80%

Verrucomicrobia

70%

Chlorobi

60%

Firmicutes

50%

Acidobacteria

40%

Bacteroidetes

30%

Chloroflexi

20%

Planctomycetes

10%

Actinobacteria

0%

Proteobacteria

Relative abundance [-]

100%

CAS 1day

CAS-HPJD CAS 1day 108day

CAS-HPJD 108day

(b) 100% 90%

Relative abundance [-]

80% 70%

Other

Microthrixaceae

Pirellulaceae

Bradyrhizobiaceae

Gordoniaceae

Hyphomonadaceae

Gemmataceae

Methylocystaceae

Isosphaeraceae

Caulobacteraceae

Planctomycetaceae

Mycobacteriaceae

Chitinophagaceae

Micrococcaceae

Rhodospirillaceae

Caldilineaceae

Comamonadaceae

Hyphomicrobiaceae

Rhodobacteraceae

Xanthomonadaceae

60% 50% 40% 30% 20% 10% 0% CAS 1day

CAS-HPJD 1day

CAS 108day

CAS-HPJD 108day

Fig. 6

29

Sphingomonadaceae

(a)

(b) Mesomycetozoa

Euglenozoa Fungus

Amoebozoa Platyhelminthes Opisthokonta Unclutured Other

n=77

n=68 Fig. 7

Table 1 Operating conditions (n = 41) in the conventional

activated sludge (CAS) and CAS-HPJD (SRT; sludge retention time, HRT; hydraulic retention time, MLSS; mixed liquor suspended solid) SRT [day] HRT [hour] MLSS [mg/L] Air flow rate [L/min]

CAS-HPJD 13 6.38 947.8 ± 260.7 18–25

3

Influent flow rate [m /day] Discharge rate of waste activated sludge [m3/day]

CAS 6

28–38 3.76 ± 0.49

0.036 ± 0.017

30

0.091 ± 0.009

Table 2 Average properties of influent wastewater constituents (n = 41) (DOC; dissolved organic carbon, TN; total nitrogen, TP; total phosphorous) DOC [mg/L] 13.7 ± 4.7 TN [mg-N/L] 21.3 ± 6.3 NH4+ [mg-N/L]

15.8 ± 3.9

TP [mg-P/L] PO43− [mg-P/L]

2.92 ± 0.81 1.68 ± 0.47

Temperature [°C]

25.9 ± 2.1

31

Highlights 

A high-pressure jet device (HPJD) was installed in an activated sludge system.



The system with an HPJD reduced wasted activated sludge by 65%.



The installation of an HPJD did not deteriorate C and N removal performances.



Microbial community compositions were distinct by introducing an HPJD.



An HPJD assisted oxygenation in an aerobic tank.

32