RESEARCH REPOSITORY

This is the author’s final version of the work, as accepted for publication following peer review but without the publisher’s layout or pagination.

The definitive version is available at:

http://dx.doi.org/10.1016/j.watres.2017.03.045

Khalfbadam, H.M., Ginige, M.P., Sarukkalige, R., Kayaalp, A.S. and Cheng, K.Y. (2017) Sequential solid entrapment and in situ electrolytic

alkaline hydrolysis facilitated reagent-free bioelectrochemical treatment of particulate-rich municipal wastewater. Water Research, 11 . pp. 18-26.

http://researchrepository.murdoch.edu.au/id/eprint/36284/

Copyright © 2017 Elsevier Ltd.

Accepted Manuscript

Sequential solid entrapment and in situ electrolytic alkaline hydrolysis facilitatedreagent-free bioelectrochemical treatment of particulate-rich municipal wastewater

Hassan Mohammadi Khalfbadam, Maneesha P. Ginige, Ranjan Sarukkalige, AhmetS. Kayaalp, Ka Y. Cheng

PII: S0043-1354(17)30229-4

DOI: 10.1016/j.watres.2017.03.045

Reference: WR 12782

To appear in: Water Research

Received Date: 23 February 2017

Revised Date: 20 March 2017

Accepted Date: 21 March 2017

Please cite this article as: Khalfbadam, H.M., Ginige, M.P., Sarukkalige, R., Kayaalp, A.S., Cheng,K.Y., Sequential solid entrapment and in situ electrolytic alkaline hydrolysis facilitated reagent-freebioelectrochemical treatment of particulate-rich municipal wastewater, Water Research (2017), doi:10.1016/j.watres.2017.03.045.

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

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

1

Submit to Water Research 1

2

Sequential solid entrapment and in situ electrolytic alkaline hydrolysis 3

facilitated reagent-free bioelectrochemical treatment of particulate-rich 4

municipal wastewater 5

6

Hassan Mohammadi Khalfbadama,b, Maneesha P. Ginigea, Ranjan Sarukkaligeb, Ahmet S. 7

Kayaalpc, Ka Y Chenga,d* 8

9

a CSIRO Land and Water, Floreat, Western Australia, 6014, Australia. 10

b Department of Civil Engineering, Curtin University, Bentley, Western Australia, 6102, 11

Australia. 12

c Water Corporation of Western Australia, Western Australia, 6007, Australia. 13

d School of Engineering and Information Technology, Murdoch University, Western Australia 14

6150, Australia. 15

16

17

*Corresponding author. Tel: +61 8 9333 6158; Fax: +61 8 933 6499. 18

E-mail address: kayu.cheng@csiro.au (Ka Yu Cheng) 19

20

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

2

Abstract 21

This study proposed and examined a novel process for the treatment of particulate-rich 22

wastewater. A two- stage combined treatment process, consisting of an electrolysis filter and 23

a bioelectrochemical system (BES) configuration was designed and evaluated to remove 24

particulate and soluble organic matter from municipal wastewater. The system was designed 25

such that the electrolysis step was used as a filter, enabling physical removal and in situ 26

alkaline hydrolysis of the entrapped particulate matter. The alkaline effluent enriched with 27

the hydrolysed soluble compounds (soluble chemical oxygen demand, SCOD) was 28

subsequently loaded into the BES for removal via bioanodic oxidation. The coupled system 29

was continuously operated with a primary sedimentation tank effluent (suspended solids (SS) 30

~200 mg/L) for over 160 days, during which SCOD and total COD (TCOD), SS removal and 31

current production were evaluated. With no sign of clogging the process was able to capture 32

near 100% of the SS loaded. A high Coulombic efficiency (CE) of 93% (based on overall 33

TCOD removed) was achieved. The results also suggest that the SCOD- laden alkaline 34

liquor from the electrolysis step compensated for the acidification in the bioanode and a final 35

effluent containing low COD with neutral pH was achieved. Overall, since the system can 36

effectively entrap, in situ hydrolyse and oxidise organic matter without external chemical 37

dosing for pH control, it has desirable features for practical application. 38

39

Key Words: municipal wastewater; solid entrapment, electrode-driven alkaline hydrolysis; 40

bioelectrochemical systems; sequential; reagent-free 41

42

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

3

1. Introduction 43

Waste stabilisation ponds (WSPs) are commonly used to treat municipal wastewater 44

particularly that of remote communities (Mara et al., 1998). However, maintenance of stable 45

effluent quality of WSPs is challenging and additional treatment processes are often required 46

to minimise discharge of suspended solids (SS) (mostly in form of algal biomass) and 47

dissolved organic carbon (DOC) into the environment (Ellis & Mara, 1983; Mara et al., 48

1992). Rock filters are one of the widely used downstream polishing technologies designed to 49

entrap and oxidise both particulate matter and DOC discharged from WSPs. They rely on 50

passive diffusion of oxygen from the atmosphere to facilitate aerobic biological oxidation of 51

the entrapped organic matter. However, excessive biomass growth and particulate matter 52

accumulation can hinder passive aeration and could result in anaerobic conditions in the filter 53

compromising organic matter removal and hydraulic throughput of the treatment unit. Hence, 54

alternative technologies are desirable for efficient treatment of particulate-rich liquor such as 55

those emanating from failed WSPs. 56

Recently, there has been an effort to develop new technologies to substitute rock filters 57

(Khalfbadam et al., 2016b). Khalfbadam et al. (2016b) proposed a filter-type 58

bioelectrochemical system (BES) termed as “BES filter” to retain and oxidise high particulate 59

matter and soluble chemical oxygen demand (SCOD) effluent. The tubular, dual chamber 60

BES prototype is composed of an outer anodic chamber that encircles an inner cathodic 61

chamber and the two chambers are separated with a cation exchange membrane. The outer 62

anodic chamber was designed as a filtration unit, whereby a carbon felt was installed as both 63

a physical filter and a bioanode to entrap and bioelectrochemically oxidise the suspended 64

organic matter from the wastewater. The study successfully demonstrated the ability of the 65

BES filter to oxidise SCOD in wastewater with a high Coulombic efficiency (CE) of >80% 66

(based on the SCOD removed). The study also demonstrated effective removal of organic 67

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

4

particulate matter from municipal wastewater (83% of SS removal). However, if the CE 68

calculation was based on the coulombs derived from the entrapped particulates, the CE would 69

be very low (~1%). Such a low CE was most likely due to inefficient microbial hydrolysis of 70

the entrapped particulate matter. Indeed, the system proposed by Khalfbadam et al. (2016b) 71

suffered from the following limitations: (1) inefficient hydrolysis of entrapped particulate 72

matter; (2) the need for active control of pH both in the cathodic and anodic chambers of 73

BES; and (3) a huge overpotential for reduction of oxygen at BES cathode. It is imperative to 74

overcome the above limitations should this innovative design be considered as an effluent 75

polishing technology for WSPs. 76

Generally, hydrolysis is often the rate-limiting step in biological treatment processes for 77

waste streams rich in organic solids. Various pre-treatment methods (physical, chemical, 78

biological, and electrochemical) have been examined to improve hydrolysis of organic 79

particulate matter. For instance, Khalfbadam et al. (2016a) effectively harnessed hydrolytic 80

bacteria in anaerobic sludge to increase in situ oxidation of algal biomass in a BES. Chu et al. 81

(2001) found that ultrasound pre-treatment (sonication at 0.33 Watt/mL, 2 h) significantly 82

increased the SCOD content of a waste activated sludge (WAS) (from 42 to 1084 mg 83

SCOD/L). Alternatively, drastically altering the pH of the feedstock by dosing acids or alkali 84

could also enhance hydrolysis of organic particulate matter (Chen et al., 2007; Huang et al., 85

2016; Yi et al., 2013). However, these methods require either substantial retrofit of 86

infrastructure or usage of chemicals, and hence may not be desirable for practical application. 87

Recently, Charles et al. (2013) investigated the use of a dual chamber electrochemical cell to 88

enhance anaerobic digestibility of a WAS. With an applied voltage of 12 V, the acidic and 89

alkaline conditions created by the electrolytic reactions in the anodic and cathodic chambers 90

remarkably increased the SCOD in the WAS by 31% and 34%, respectively. Since this 91

method could effectively use electrodes to hydrolyse particulate matter without dosing 92

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

5

external chemicals, it may be suitable for addressing the abovementioned limitations in the 93

BES filter process of Khalfbadam et al. (2016a). 94

In this study, we proposed that by operating the BES filter as an electrolytic cell, alkaline 95

hydrolysis of the entrapped particulate matter could be driven in situ at the electrode (filter). 96

This process would create an effluent with elevated SCOD and alkalinity, which could be a 97

suitable feed stock for a subsequent bioelectrochemical (anodic) treatment. Figure 1A depicts 98

the concept of the proposed two-stage process. In the first stage, the particulate-rich 99

wastewater is fed to the module, whereby the filter-electrode physically removes (filtering) 100

the particulate organic matter from the wastewater. The filter-electrode loaded with 101

particulate matter is then cathodically-driven to create a localised alkaline condition (≥12), 102

facilitating in situ alkaline hydrolysis of the entrapped particulate matter. The elevated SCOD 103

and alkalinity are expected to facilitate current production and help neutralise the acidity 104

liberated from the anodic oxidation reaction at the subsequent stage. Overall, the sequential 105

process is expected to yield a better quality final effluent with a much higher total COD 106

removal efficiency. 107

The aim of this study was to validate the proposed combination of electrolysis and anodic 108

oxidation for effective removal of particulate and soluble organic matter from municipal 109

particulate-rich wastewater. Specifically, the following hypothesises were tested: 110

(1) The increased cathodic pH created by the electrolysis step (first stage) can trigger 111

alkaline hydrolysis of the particulate matter entrapped at the filter-electrode; 112

(2) The alkaline effluent with elevated SCOD content can be harnessed as a more suitable 113

feedstock for the subsequent anodic carbon oxidation step, negating the need of dosing 114

external chemicals (e.g. NaOH) for pH control; 115

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

6

(3) The acidity generated at the anodic reaction (captured in the counter electrolyte) of the 116

first stage can be used to facilitate the cathodic reaction of the second stage, alleviating 117

the overall reliance on external chemicals for pH control. 118

The coupled process was operated in a continuous mode with different hydraulic loadings of 119

municipal primary sedimentation tank (PST) effluent for over 160 days, during which the 120

performance of the system was quantified for a range of parameters such as total COD 121

(TCOD), soluble COD (SCOD), particulate COD (PCOD) and SS removals; electrical 122

current, Coulombic efficiencies and energy consumption. 123

124

2. Materials and Methods 125

2.1. Configuration of the two-stage reactor process 126

Figure 1 illustrates the experimental process used in this study. The process consisted of two 127

identical, hydraulically connected dual-compartment electrochemical reactors (Figure 1). The 128

first reactor (R1) was operated as an electrolytic cell facilitating both entrapment and in situ 129

alkaline hydrolysis of the entrapped particulate solids; the second reactor (R2) was operated 130

as a typical BES for bioanodic oxidation of soluble organic matter. Specific configurational 131

details of the reactor and the electrode materials have been described in Khalfbadam et al. 132

(2016b). In brief, each reactor had a height of 1.5 m (internal diameter 0.1 m) and consisted 133

of two concentric cylindrical stainless steel mesh columns, one with a diameter of 6 cm and 134

the other one with a diameter of 4 cm. The two columns were served as both electrical current 135

collector and physical support for the anode and the cathode, respectively. To prevent 136

corrosion of the stainless steel columns, the columns were pre-treated with pickling and 137

passivation treatments as per American Society for Testing and Materials (ASTM) standards 138

(ASTM A380 and A967) (International Corrosion Services Ltd. Co., Perth, Australia). The 139

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

7

two compartments were isolated by a cation exchange membrane (Ultrex CMI 7000, 140

Membrane International Inc.), which was firmly mounted onto the outer side of the inner 141

stainless steel mesh column and was sealed with epoxy glue to create a water-tight 142

compartment within the module. Carbon felts (MGM-Carbon Industrial, Ltd. Co., China) 143

were used as both the anodic and cathodic electrode materials. The carbon felts were 144

mounted onto the outer side of the larger and inner side of the smaller stainless steel current 145

collectors, respectively. The carbon felt at the larger stainless steel column (i.e. cathode of R1 146

and anode of R2) had a surface area of 0.33 m2. A short distance (approximately 2 cm) was 147

maintained between the anode and the cathode, creating a void volume where the wastewater 148

influent was introduced at the bottom end of the unit. Since the inner column was completely 149

enclosed, the influent wastewater was compelled to channel through the carbon felt 150

facilitating physical removal of SS from the wastewater (Figure 1C). In this study, it was 151

proposed that the particulate-rich wastewater was first processed in R1 for particulate matter 152

removal and electrolytic alkaline-hydrolysis of the entrapped particulate; the alkaline effluent 153

created by R1 was further processed in R2 for bioelectrochemical treatment before the treated 154

effluent was finally discharged (Figure 1C). 155

156

2.2. General operation of the process 157

To mimic effluent of failed WSPs, primary sedimentation tank (PST) effluent (200 mg-158

SCOD/L, 150 mg-SS/L) of a local municipal wastewater treatment plant was used 159

(Khalfbadam et al., 2016b). The effluent was collected weekly and stored in a fridge (4ºC) to 160

minimise compositional changes. The wastewater was continuously fed to the bottom of R1 161

at a defined flow rate (first stage). The effluent of R1 was withdrawn from a port located near 162

the top of the reactor through a peristaltic pump and introduced to the bottom of R2 (second 163

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

8

stage). The final effluent was discharged from the outer (anodic) chamber of R2 by hydraulic 164

pressure via an effluent port located near the top of the reactor. The wastewater in the outer 165

chambers of both reactors was continuously recirculated (633 mL/min). The inner chambers 166

of R1 and R2 were filled with deionised water (2 L). To ensure supply of dissolved oxygen 167

was not limiting the cathodic reaction (presumably oxygen reduction) at the cathode of R2, 168

the catholyte of R2 was continuously aerated (833 mL air/min). Two peristaltic pumps (Cole- 169

Parmer, Victoria, Australia) operated at a same flow rate (17 mL/min) were used to mix the 170

two counter electrolytes (in the inner chambers) between the two reactors, with one pump 171

transferring the electrolyte from R1 to R2, and the other one returning the electrolyte from R2 172

to R1. This step was to facilitate the use of the acidity created by the anodic reaction of R1 173

for neutralising the alkalinity created by the cathodic reaction of R2. 174

Two adjustable digital power supplies (Array 3645 A; Array Electronics, Australia) were 175

used to separately apply electrical voltage to both reactors. The positive and negative poles of 176

the power supplies were connected to the anode and cathode, respectively (Figure 1B). To 177

enable measurement of electrical current in both reactors, an external resistor (1 ohm) was 178

connected between the cathode of both reactors and their respective negative terminals of the 179

power supply. Electrical current was determined from the voltage measured across the 180

resistor according to Ohm’s Law. A silver-silver chloride (Ag/AgCl) reference electrode 181

(MF-2079, RE-5B, BASi Bioanalytical Systems, Inc., USA) was mounted in the anodic 182

chamber of R2 to enable measurement of the anodic potential of the BES. All electrode 183

potentials (mV) reported in this study refer to values against Ag/AgCl reference electrode (ca. 184

+197 mV vs. standard hydrogen electrode). The process was operated at ambient pressure 185

and temperature (22-25ºC). 186

187

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

9

2.3 Determination of the optimal pH for enhancing particulate matter hydrolysis (serum 188

bottle experiment) 189

A separate experiment was first conducted to determine the susceptibility of the particulate 190

matter in the PST effluent for acid/alkaline hydrolysis, and to define the optimal pH for 191

enhancing the SCOD content in the PST effluent. Freshly collected PST effluent was loaded 192

into four 1-L Imhoff cones and was allowed to settle for one hour. Thereafter, two distinct 193

sediment layers were visualised at the bottom of each cone. The lower layers, which 194

contained mostly inert sand-like particles were carefully discarded through the decant port of 195

each Imhoff cone. The upper layers were carefully collected and pooled together for 196

experimentation. The TCOD and SCOD contents of the composite particulate matter sample 197

were 5361 mg/L and 251.3 mg/L, respectively. Hence, the SCOD/TCOD ratio was 4.7%. 198

To evaluate the impact of acidic and alkaline pH on solubilisation of particulate matter, the 199

composite particulate matter sample (630 mL) was equally divided into seven serum bottles 200

(90 mL each). The initial pH value in bottles 1–6 was adjusted to 2, 5, 7, 8, 10 and 12, 201

respectively, by adding sodium hydroxide (1M NaOH) or hydrochloric acid (1M HCl). The 202

bottle 7, in which the pH was not adjusted, was used as the control (pH = 7.05). The bottles 203

were incubated on an orbital shaker (room temperature, 225 rpm) for 24 h. Samples were 204

taken from each bottle at 0, 3, 7 and 24 h over a 24 h period for SCOD analysis. Percent 205

solubilisation of particulate was calculated for each setting according to: (SCOD measured – 206

251.3 mg/L)/ 5361 mg/L × 100%. 207

208

2.4. Impact of applied voltage on solubilisation of entrapped particulate matter in R1 209

The impact of various applied voltages (0, 3, 6 and 10 V) on the hydrolysis of entrapped 210

particulate matter in the filter type electrolysis unit (R1) was evaluated before connecting 211

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

10

both R1 and R2 as a coupled process. This experiment was to determine the most suitable 212

applied voltage for driving the alkaline hydrolysis of the entrapped particulate matter in R1. 213

Fresh PST effluent was continuously loaded into R1 at a hydraulic retention time (HRT) of 214

32 h (flow rate of 5.02 mL/min) over a period of 14 days. After each voltage and prior to the 215

next setting, the applied voltage was set to zero and the HRT was reduced to 6 h for a period 216

of at least 24 h, to ensure the electrode-filter was recharged with sufficient organic particulate 217

matter build up prior to the next applied voltage setting. Current, working electrolyte pH and 218

effluent SCOD for various settings were monitored and compared. 219

220

2.5. Influence of applied voltage on the performance of R2 221

Prior to this study, R2 was started up and had been operated for approximately two years in 222

microbial fuel cell (MFC)-mode as reported in Khalfbadam et al. (2016b). It was identified 223

that the current production and SCOD removal efficiency were largely limited by the poor 224

cathodic reaction (Khalfbadam et al. 2016b). Hence, in this study R2 was coupled with an 225

external power supply to overcome such limitation. This mode of operation (i.e. as a 226

microbial electrolysis cell (MEC) was evaluated before connecting both R1 and R2 as a 227

coupled process. The effects of various applied voltages (0, 150, 250, 400, 600 and 800 mV) 228

on R2 were evaluated over a period of 60 hours. Current production, electrode potentials, 229

electrolyte pH and anodic SCOD removal rates were recorded. In order to ensure the supply 230

of COD was not limiting, the R2 was continuously fed with fresh municipal PST effluent at a 231

HRT of 6 hours, corresponding to a flow rate of 26.8 mL/min. To eliminate the impact of pH 232

changes, both the anolyte and catholyte were maintained at pH 7 by feedback-dosing NaOH 233

(4M) and HCl (1M), respectively. Similar to the previous study, the counter electrolyte of R2 234

was aerated to ensure that the cathodic reaction was not limited by dissolved oxygen. 235

236

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

11

2.6. Evaluation of the effect of hydraulic retention time (HRT) and applied voltage on 237

the performance of the sequential process 238

After separately testing the R1 for alkaline hydrolysis of the entrapped particulate matter and 239

and R2 for bioanodic oxidation of SCOD, the performance of the coupled process was 240

assessed. As described above, the two reactor modules were hydraulically connected (Figure 241

1). Applied voltages (setting 1: 6 and 0.5 V for R1 and R2, respectively (6V/0.5V); setting 2: 242

10 and 0.6 V for R1 and R2, respectively (10V/0.6V)) and HRT (6, 12, 18, 24 and 32 h) were 243

selected as the process variables. The coupled process was evaluated in continuous mode 244

over a period of 150 days. Freshly collected PST effluent was loaded into R1at different flow 245

rates (26.8, 13.4, 8.9, 6.7 and 5 mL/min) to achieve the respective HRTs of 6, 12, 18, 24, 32 h 246

in R1. The final effluent was discharged from R2 at the same flow rate and hence both R1 247

and R2 were operated at identical HRT. Each set of applied voltage and HRT was maintained 248

for 15 days to ensure steady-state operation. Process parameters such as TCOD removal, 249

SCOD removal, working electrolyte pH, current production and CE were quantified for each 250

stage and were compared amongst various settings. 251

252

2.7. Effect of spiking algal biomass to the influent of R1 on current production in R2 253

The effectiveness of using the coupled process to entrap and hydrolyse algal biomass for 254

current production was evaluated in a separated experiment. The concentrated algal biomass 255

was prepared from a synthetic algal medium as described in Khalfbadam et al. (2016a). The 256

algal biomass containing medium was loaded into Imhoff cones, and the biomass was 257

allowed to settle for 3 hours before collection. The collected sediments were centrifuged 258

(4000 × g for 30 minutes) and the supernatants were discarded. The pelleted algal solids were 259

resuspended in deionised water, centrifuged again and the supernatant was discarded. This 260

step was to remove SCOD. The pelleted algae was finally resuspended in a medium (300 261

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

12

mL). The TCOD and SCOD concentrations of this algae-containing medium were 8981 mg/L 262

and 51 mg/L, respectively. Prior to the spiking test, the coupled process was steadily operated 263

with municipal PST effluent for at least one week (HRT 32 h). When the current of R2 was 264

stable, the pre-concentrated algae-containing medium was injected to the influent line of R1. 265

The same procedure was repeated when the current of R2 had returned to the baseline level 266

for >1 HRT. Coulombic efficiencies (CE) were determined using the current induced by the 267

spikes and the TCOD content of the spikes. 268

269

2.8. Analyses and calculations 270

The voltage (i.e. the potential differences between the anode and the cathode) (V) was 271

measured using a digital multimeter placed across the external resistor and the anodic and 272

cathodic potentials were measured against respective reference electrodes (Ag/AgCl). pH was 273

monitored using a portable pH meter (TPS, Australia). 274

The current (I, mA) of R1 was determined from the cell voltage according to Ohm’s Law, 275

=

, where (mV) is the measured cell voltage across an external resister (Ω). Power 276

(P, µW) was calculated according to equation = × . It was used to calculate electrical 277

energy consumption of R1 and R2 over a specified period of time (h). 278

Soluble COD measurements were carried out after filtering the samples through 0.22 µm 279

syringe filters to remove any suspended solids. Total COD measurements were carried out on 280

unfiltered samples. SCOD and TCOD analyses were performed using HACH reagents (cat 281

no. TNT 821; method 8000, LR) and a spectrophotometer (GENESYS 20, Thermo 282

Scientific). CE (i.e., the percentage of electrons recovered as anodic current from the TCOD 283

removed) of the anodic reaction was calculated as detailed in literature (Logan et al., 2006). 284

Particulate matter was quantified by measuring SS. SS was measured as described in standard 285

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

13

methods for examination of water and wastewater (APHA, 1992). All analytical 286

measurements were carried out in duplicate. 287

288

3. Results and discussion 289

290

3.1. Increasing pH to 12 enabled efficient hydrolysis of PST effluent particulate 291

A comparative study was conducted to examine the effect of pH (2 to 12) on the hydrolysis 292

of PST particulate matter. Figure 2 shows how SCOD concentration and percent 293

solubilisation of particulate matter changed as a function of pH. Clearly, both acidification 294

(pHs 2 and 5) and basification (pHs 10 and 12) could increase SCOD concentration and 295

solubilisation of the PST particulate matter. However, compared with other pHs, pH 12 296

enabled the highest solubilisation of the particulate matter (13%) (Figure 2B). Also, at pH 12 297

the SCOD concentration was significantly increased by 4-fold within a short time (from 298

235.2 to 948 mg/L in 3 hours) (Figure 2A). The slight decreases in SCOD recorded near 299

circum-neutral conditions (pH 7, 8 and control) could be a result of microbial degradation of 300

soluble organic matter. Overall, the result confirmed that pH 12 was the most effective 301

condition to facilitate hydrolysis of the PST particulate matter. 302

303

3.2. Cathodic-driven alkaline hydrolysis of entrapped particulate matter in R1 304

With the previous experiment confirming the effect of alkaline pre-treatment for improving 305

hydrolysis of the PST particulate matter, it was decided to couple the R1 with an external 306

power supply to approach the desired alkaline condition. To determine how much applied 307

voltage was needed, the changes in SCOD, pH of working electrode and current production 308

in R1 with different applied voltages (0, 3, 6 and 10 V) were quantified (Figure 3). 309

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

14

Although applying 3 V increased the current of R1 from 0 mA to 98 mA, no remarkable 310

change in both working electrolyte pH and SCOD were recorded compared to the control (no 311

applied voltage) (Figure 3A, C). In contrast, doubling the applied voltage to 6 V remarkably 312

increased the pH from 6.94 to 11.0 within 5.5 h (Figure 3B). This also coincided with an 313

increase in current (from 0 to 1154 mA) and SCOD (from 196 to 271 mg/L). Further 314

increasing the applied voltage to 10 V resulted in a more rapid basification of the catholyte 315

(from 7.0 to 12 within 4.5 h), and more profound increases in both current (from 0 to 1294 316

mA) and SCOD concentration (from 205 to 335 mg/L) (Figure 3). 317

Since applying 6 and 10 V enabled the SCOD concentration in R1 to approach plateau within 318

≤6 h (Figure 3A), it would be sufficient to operate R1 with the applied voltage for a much 319

shorter time frame. Overall, the results confirmed that the alkaline condition (pH>11) could 320

facilitate hydrolysis of the entrapped particulate organics. Such conditions could be created 321

by driving the electrode-filter of R1 as a cathode with an external voltage supply of >6V. 322

323

3.3. Augmenting R2 with an applied voltage to enhance SCOD removal rate 324

It was previously shown that COD removal by R2 was limited by the poor cathodic oxygen 325

reduction (Khalfbadam et al. 2016b). One way to overcome this constrain was to switch the 326

operation into MEC mode, whereby external voltage is applied to enable faster reaction 327

kinetics at the bioanode (Cheng et al., 2012). Hence, before connected with R1, R2 was 328

examined with different applied voltages (0, 150, 250, 400, 600 and 800 mV) and the 329

corresponding effects on anodic potential and COD removal rates were quantified (Figure 330

S1). The results indicated that addition of voltages (up to 400 mV) effectively increased the 331

anodic potential (from -570 mV to -430 mV), which favoured a higher biogenic electrical 332

current (from 100 to 310 mA) (Figure S1A and B). Such improvement also coincided with a 333

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

15

notable rise (>3 fold) in SCOD removal rate. With an applied voltage of 400 mV, the SCOD 334

removal rate increased from 0.08 to 0.25 kg/m3.d (HRT 6 h) (Figure S1D). 335

This experiment confirmed that the activity of the established anodic biofilm in R2 was 336

largely hindered by the cathodic reaction, and that by augmenting with electrical voltages, 337

SCOD removal could be markedly improved. Nonetheless, it must be mentioned that in this 338

experiment the process was operated with active pH corrections (i.e. dosing of NaOH and 339

HCl). For practical application, this requirement should ideally be omitted. 340

341

3.4. Characterisation of the sequential process and influence of hydraulic retention time 342

and applied voltage 343

The above experiments showed that R1 and R2 could effectively facilitate in situ alkaline 344

hydrolysis of the entrapped particulate matter and bioanodic oxidation of SCOD, 345

respectively. Hence, a subsequent experiment was conducted with the two modules 346

connected hydraulically as a sequential process (see Figure 1). The process was operated in 347

continuous mode for a prolonged period (150 days) with various HRTs (6, 12, 18, 24 and 348

32 h) and applied voltages (6V/0.5V; 10V/0.6V). Key process variables were quantified 349

(Figure 4). 350

351

3.4.1 The sequential process could enable an excellent SS removal, and R1 (stage 1) 352

was responsible for the majority (>60%) of removal 353

The results again confirmed that the “electrolytic electrode-filter” of R1 could effectively 354

remove SS (Figure 4D). For both applied voltage settings tested, increasing the HRT of R1 355

(from 6 to 32 h) also improved SS removal (from 62% to 96%) in R1 (data not shown). 356

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

16

Extending the HRTs beyond 6 h may have offered an extra SS retention (e.g. settling) for the 357

improved SS removal recorded (Figure 4D, at days 15 and 90). As expected, further SS 358

removal was attained by R2, giving a near complete overall SS removal (>99%) (Figure 4D). 359

Similar trends for SS removal efficiencies were recorded with the higher applied voltage 360

setting (10V/0.6V, days 75 to 150). These results suggested that the sequential process could 361

enable an excellent SS removal, and R1 (stage 1) was responsible for the majority (>60%) of 362

removal. The continuous loading of wastewater did not result in any sign of clogging in R1 363

during the period of operation for over 150 days , suggesting that the entrapped particulate 364

matter was solubilised in situ via alkaline hydrolysis. This implies the prevalence of a self- 365

cleansing mechanism in R1. 366

367

3.4.2 Current, pH and COD removal 368

In general, increasing the HRT did not notably affect the current and SCOD augmentation in 369

R1, as only slight decreases in current and slight increases in SCOD were recorded (Figure 370

4A and E). The result also corroborates with the previous conclusion (in Section 3.2) that the 371

two voltages tested (6 and 10V) were suitable to allow efficient cathode-driven alkaline 372

hydrolysis, and that a short HRT of 6 h was adequate for R1 to attain good efficacies. Clearly, 373

at all tested settings, R1 could create the desirable alkaline conditions (pH 11.9-13.8) 374

facilitating in situ hydrolysis of the entrapped SS and as such, resulted in a higher SCOD in 375

the R1 effluent (Figure 4C, E). We also noted that both the R1 current and SCOD 376

concentrations in R1 effluent were generally higher (20%) with a higher applied voltage (10 377

vs. 6V) (Figure 4A, E). This further highlights the functional feature of R1 as an electrolytic-378

filter for particulate organics removal and hydrolysis. 379

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

17

In contrast, the current of the subsequent stage (R2) decreased with increasing HRT, 380

indicating that the bioanode activity of R2 was likely limited by the SCOD from R1 (Figure 381

4A). Yet, in terms of overall COD removal (both SCOD and TCOD), operating the process at 382

higher HRT appeared to be essential, as only by extending the HRT to 32 h the overall 383

removal of SCOD and TCOD could then reach >80% (Figure 4E, F). When R1 was applied 384

with 6 V and 10 V, increasing the HRTs (from 6 to 32 h) decreased the current of R2 from 385

390 mA to 235 mA (1.18 to 0.712 A/m2) and from 430 mA to 258 mA (1.30 to 0.781 A/m2), 386

respectively (Figure 4A). A decrease in current trends coincided with an increase in CE from 387

38.8 % to 93.3 % (with applied voltage of 6 V/0.5V) and from 44.6 to 84.8 % (with applied 388

voltage of 10 V/0.6 V) (Figure 4B). For both voltages, the highest CEs were achieved with 389

the longest HRT (32 h), suggesting that operating the process at higher HRTs was also 390

favourable to obtain higher CEs. 391

On the other hand, it is well known that anodic biofilms of BESs are highly sensitive to pH 392

changes and often have a functional pH window near circumneutral range (Cheng et al. 2010; 393

Rozendal et al., 2008). In our study, such a stable pH environment was maintained by a 394

continuous neutralisation between the alkalinity created by the cathodic reaction in R1 and 395

the acidity created by the anodic reaction in R2. As shown in Figure 4C, the pH of the 396

wastewater stream (i.e. working electrolye) was decreased from 11.3 - 13.3 (in R1) to 6.87 - 397

7.33 (in R2). In other words, it was a synergy between the two stages (R1 and R2) that made 398

the final process effluent neutral. The results suggest that without external chemical dosing 399

for pH control, the detrimental problem of acidification of anodic oxidation process was not 400

encountered. 401

After each HRT and prior to change to the next HRT, both counter electrolytes of R1 and R2 402

were discarded and renewed with fresh DI water. Such renewal did not retard the rapid 403

acidification of counter electrolyte of R1, as a sharp decrease in pH (pH ~2) was recorded 404

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

18

within less than 2 hours (Figure 4C). Exchanging both counter electrolytes in a close loop 405

enabled the neutralisation of the counter electrolyte of R2. However, the rate of acidification 406

in counter electrolyte of R1 was much faster than alkalinisation of counter electrolyte of R2. 407

As a result, the concentration of proton in both counter electrolytes was acidic (pH 2-3). In a 408

separate study, Khalfbadam et al. (2016b) demonstrated that the performance of a BES 409

process can be improved when the cathodic pH was maintained in acidic conditions by 410

adding HCl. However, in this study creation of a net acidity from R1 without adding external 411

chemicals enabled the overall more acidic counter electrolyte that favoured the cathodic 412

reaction of R2. This feature is another advantage of the proposed process. 413

414

3.5. The sequential process could effectively hydrolyse algal cells for more efficient 415

bioanodic oxidation 416

To test if the sequential process could readily treat fresh intact algal biomass for current 417

generation, a spiking experiment was carried out (Figure 5). The experiment was started by 418

first operating the process with municipal wastewater in a continuous mode with HRT of 32 419

h. After approximately 7 days, the current of R2 became stable at around 260 mA (0.788 420

A/m2) (Figure 5A). At this point, an algal biomass aliquot was added to the influent stream of 421

R1, and within 3.5 h the current of R2 increased and continued to rise reaching a maximum 422

of 420 mA (1.27 A/m2) after 16 h. Subsequently, the current gradually returned to the initial 423

background level. Similar trends was recorded when the experiment was repeated with the 424

same amount of algal injection (Figure 5A). The two current peaks also coincided with slight 425

decreases in the working electrolyte pH of R2 (from 7.6 to 6.4), which were probably caused 426

by additional protons released from the faster anodic reaction (Figure 5B). The two algae-427

induced current peaks corresponded to an average CE of 41%. 428

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

19

Overall, the results suggested that the sequential process could effectively hydrolyse fresh 429

algal cells for bioanodic oxidation (removal). Future studies are necessary to further develop 430

the process for algae-laden streams treatment. 431

432

3.6. Energetic consideration for the sequential treatment process 433

In this study, electrical energy was consumed to facilitate alkaline hydrolysis (stage 1) and 434

bioanodic oxidation (stage 2). Figure 6 and Table S1 summarise the treatment and energetic 435

performance of the sequential process. In general, the energy consumptions in both stages 436

were dependent on both HRT and applied voltage. Also, as expected R1 consumed 437

significantly (two orders of magnitude) more energy than R2 (Figure 6A, B). For stage 1 438

electrolytic hydrolysis (R1), increasing the HRT from 6 to 32 h increased the energy 439

consumption from 92.1 to 274.7 kWh/kg SCODincreased and 67.5 to 303.7 kWh/kg 440

SCODincreased with applied voltage of 6 and 10 V, respectively (Figure 6). Interestingly, these 441

values are lower than that reported for alternative particulate matter hydrolysis pre-442

treatments. For instance, an approximately two-fold higher energy consumption of 633 443

kWh/kg SCODincreased was reported by Chu et al. (2001), who used ultrasound as a pre-444

treatment step to enhance SCOD in WAS (from 42 mg/L to 1084 mg/L). Hence, operating 445

the electrode-filter to achieve in situ alkaline hydrolysis of particulate is realistic in terms of 446

energy requirement. 447

Unlike in R1, increasing the HRT of R2 (6 to 32 h) only slightly increased the energy 448

consumption, from 1.48 to 1.66 kWh/kg SCODremoved (12%) and 1.58 to 1.74 kWh/kg 449

SCODremoved (10%), respectively with applied voltage of 0.5 and 0.6V (Figure 6A, B). This 450

suggests that changing the HRT did not impact much on the energetic performance of R2 451

bioanode in terms of SCOD removal. The energetic performance of R2 (stage 2) is 452

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

20

comparable with other wastewater treatment MEC processes. For example, an energy input of 453

2.59 kWh/kg SCODremoved was recorded by Ivanov et al. (2013) in a MEC process to enable 454

SCOD removal of 74% from a municipal wastewater. In fact, the energy consumption by R2 455

is within the range of energy demand in conventional activated sludge processes (0.7-2 456

kWh/kg SCODremoved) (Pant et al., 2011; Tchobanoglous & Burton, 1991). This suggests that 457

augmenting R2 with applied voltage could be a viable option to remove SCOD hydrolysed 458

from the entrapped particulate (R1). Nonetheless, further studies are desirable to examine 459

operation of R2 as a MFC to recovery electrical energy or as a MEC to recover hydrogen, as 460

this may improve the overall energetic performance of the sequential process. 461

In terms of TCOD removal, increasing HRT and applied voltage generally increased the 462

overall energy demand of the sequential process (Figure 6C). Of all tested settings, the most 463

suitable setting was HRT 24 hours with applied voltages of 6V (R1):0.5V (R2), which 464

enabled a high TCOD removal (80%) with relatively low energy consumption, 44 kWh/kg 465

TCODremoved (Figure 6C). Since the sequential process developed here is the first of its kind 466

to enable simultaneous removal, hydrolysis and oxidation of particulate organics from 467

wastewater, no comparable processes are available for direct comparison. 468

469

3.8. Implication and perspective 470

This study clearly demonstrated the effectiveness of the proposed sequential process (Figure 471

1). The findings also illustrated several key advantages of this approach: (1) the use of 472

electrode as a physical filter ensured a more reliable SS removal (here close to 100%) and 473

offered a pre-concentration mechanism to localise the SS at one place (i.e. electrode-filter in 474

R1); (2) the use of applied voltage to drive cathodic production of alkaline in the vicinity of 475

the electrode-filter enabled efficient hydrolysis of the organics; (3) the cathode-driven 476

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

21

hydrolysis step created an alkaline SCOD enriched stream, facilitated a more efficient TCOD 477

removal. The alkaline stream could be effectively neutralised in the subsequent BES process 478

without reliance on external chemical dosing for pH control, enabled a discharge of a pH 479

neutralised final effluent. The need of dosing chemicals for sustaining process operation has 480

been a well-known stumbling block for BES technology (Rozendal et al., 2008). A complete 481

negation of this requirement represents an important step towards practical application. 482

This study also demonstrated the feasibility of using the sequential process to polish 483

particulate matter- laden effluent such as algal biomass emanating from WSPs. Although the 484

complexity of the system and the demand of energy supply may not seem comparable to the 485

simplicity of a rock filter, one unique feature of the proposed process is the inclusion of the 486

electrode-driven hydrolysis step to render the filtration unit (R1) as a self-cleansing filtration 487

device. Similar to a rock filter, the hydrolysis step (R1) filtered out most of the particulate 488

matter from the effluent. However, instead of relying on natural decomposition of the 489

entrapped particulate, the use of electrolytic-cathode continuously enable the solubilisation of 490

the particulate matter at an acceptable rate, preventing clogging and not necessitating any 491

back wash of the filter. This feature is desirable for maintaining longevity of the treatment 492

unit. 493

Although this study has considerably contributed to the advancement of BES technology, 494

further studies will be required to optimise the performance of the individual and combined 495

stages of the process. For instance, in this study, for each stage one single reactor with equal 496

HRT was used. However, the results suggested that the optimal operational regime of the 497

coupled process may be achieved by coupling the two stages with varying HRTs. For 498

instance, the optimal HRTs for R1 and R2 were found to be 6 h and 24 h, respectively. Such 499

HRTs may be provided by manipulating the reactors volume or combination of different 500

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

22

numbers of modules for the two stages. To this end, the modular configuration of the 501

electrochemical-filter could be a promising feature for large scale implementation. Further 502

research is certainly required to develop the technology. 503

504

4. Conclusions 505

This study reports for the first time a combination of an electrolysis filter and a BES process 506

to achieve particulate and soluble organic matter removal from municipal wastewater. Based 507

on the results, the following points are concluded: 508

509

• The first step electrolysis filter could remove near 100% of SS from wastewater 510

influent. Electrolysis enhanced in situ hydrolysis of entrapped particulate matter in the 511

filter. 512

• The use of alkaline cathodic effluent after the electrolysis step as anodic influent of 513

BES process eliminated the need of active pH control. 514

• Electrochemically assisted alkaline hydrolysis enhanced the subsequent 515

bioelectrochemical oxidation process, resulted in high overall TCOD removal of 87% 516

and a CE of 93%. 517

• The electrolysis stage generated an acidic solution (pH~2) in the counter electrolyte 518

which was harnessed by the subsequent proton- requiring cathodic reaction. 519

• Operating the electrode-filter to achieve in situ alkaline hydrolysis of particulate is 520

realistic in terms of energy requirement. 521

5. Acknowledgments 522

This project was funded by the Water Corporation of Western Australia and CSIRO Land and 523

Water. The Australian Commonwealth Government is acknowledged for the International 524

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

23

Postgraduate Research and Publication Scholarship provided to Hassan Mohammadi 525

Khalfbadam. Dr Kaveh Sookhak Lari and Dr Bradly Patterson (CSIRO Land and Water) are 526

thanked for their valuable comments. 527

6. References 528

529

APHA. 1992. Standard methods for the examination of water and wastewater. wastewater. 18th 530

Edition. American Public Health Association, American Water Works Association and the Water 531

Environment Federation, Washington, DC. 532

Charles, W., Ng, B., Cord-Ruwisch, R., Cheng, L., Ho, G., Kayaalp, A. 2013. Enhancement of waste 533

activated sludge anaerobic digestion by a novel chemical free acid/alkaline pretreatment using 534

electrolysis. Water Science and Technology, 67(12), 2827-2831. 535

Chen, Y.G., Jiang, S., Yuan, H.Y., Zhou, Q., Gu, G.W. 2007. Hydrolysis and acidification of waste 536

activated sludge at different pHs. Water Research, 41(3), 683-689. 537

Cheng, K.Y., Ho, G., Cord-Ruwisch, R. 2010. Anodophilic biofilm catalyzes cathodic oxygen 538

reduction. Environmental Science & Technology, 44(1), 518-525. 539

Cheng, K.Y., Ho, G., Cord-Ruwisch, R. 2012. Energy-efficient treatment of organic wastewater 540

streams using a rotatable bioelectrochemical contactor (RBEC). Bioresource Technology, 541

126, 431-436. 542

Chu, C.P., Chang, B.V., Liao, G.S., Jean, D.S., Lee, D.J. 2001. Observations on changes in 543

ultrasonically treated waste-activated sludge. Water Research, 35(4), 1038-1046. 544

Ellis, K., Mara, D.D. 1983. Stabilization ponds: design and operation. Critical Reviews in 545

Environmental Science and Technology, 13(2), 69-102. 546

Huang, C., Liu, C.C., Sun, X.Y., Li, J.S., Shen, J.Y., Han, W.Q., Liu, X.D., Wang, L.J. 2016. 547

Hydrolysis and acidification of waste activated sludge enhanced by zero valent iron-acid 548

pretreatment: effect of pH. Desalination and Water Treatment, 57(26), 12099-12107. 549

Ivanov, I., Ren, L.J., Siegert, M., Logan, B.E. 2013. A quantitative method to evaluate microbial 550

electrolysis cell effectiveness for energy recovery and wastewater treatment. International 551

Journal of Hydrogen Energy, 38(30), 13135-13142. 552

Khalfbadam, H.M., Cheng, K.Y., Sarukkalige, R., Kaksonen, A.H., Kayaalp, A.S., Ginige, M.P. 553

2016a. A bio-anodic filter facilitated entrapment, decomposition and in situ oxidation of algal 554

biomass in wastewater effluent. Bioresource Technology, 216, 8. 555

Khalfbadam, H.M., Ginige, M.P., Sarukkalige, R., Kayaalp, A.S., Cheng, K.Y. 2016b. 556

Bioelectrochemical system as an oxidising filter for soluble and particulate organic matter 557

removal from municipal wastewater. Chemical Engineering Journal, 296, 225-233. 558

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

24

Logan, B.E., Hamelers, B., Rozendal, R.A., Schrorder, U., Keller, J., Freguia, S., Aelterman, P., 559

Verstraete, W., Rabaey, K. 2006. Microbial fuel cells: Methodology and technology. 560

Environmental Science & Technology, 40(17), 5181-5192. 561

Mara, D.D., Cogman, C.A., Simkins, P., Schembri, M.C.A. 1998. Performance of the Burwarton 562

Estate waste stabilization ponds. Journal of the Chartered Institution of Water and 563

Environmental Management, 12(4), 260-264. 564

Mara, D.D., Mills, S.W., Pearson, H.W., Alabaster, G.P. 1992. Waste Stabilization Ponds - a Viable 565

Alternative for Small Community Treatment Systems. Journal of the Institution of Water and 566

Environmental Management, 6(1), 72-78. 567

Pant, D., Singh, A., Van Bogaert, G., Gallego, Y.A., Diels, L., Vanbroekhoven, K. 2011. An 568

introduction to the life cycle assessment (LCA) of bioelectrochemical systems (BES) for 569

sustainable energy and product generation: Relevance and key aspects. Renewable & 570

Sustainable Energy Reviews, 15(2), 1305-1313. 571

Rozendal, R.A., Hamelers, H.V.M., Rabaey, K., Keller, J., Buisman, C.J.N. 2008. Towards practical 572

implementation of bioelectrochemical wastewater treatment. Trends in Biotechnology, 26(8), 573

450-459. 574

Tchobanoglous, G., Burton, F.L. 1991. Wastewater engineering. Management, 7, 1-4. 575

Yi, X., Luo, K., Yang, Q., Li, X.M., Deng, W.G., Cheng, H.B., Wang, Z.L., Zeng, G.M. 2013. 576

Enhanced Hydrolysis and Acidification of Waste Activated Sludge by Biosurfactant 577

Rhamnolipid. Applied Biochemistry and Biotechnology, 171(6), 1416-1428. 578

579

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 1. (A, B) Schematic diagram of the two- stage sequential solid entrapment and in situ electrolytic alkaline hydrolysis particulate-rich wastewater treatment process; (C) cross and top sections of the two stages (red arrows show wastewater flow path).

4 cm (cathode)

6 cm (anode)

4 cm (anode)

6 cm (cathode)

stainless steel mesh (current collector) Ion exchange

membrane

PST effluent

Recirculation

1.5 m

(B)

R1- stage 1

Final effluent

Recirculation

+-

Power supply+ -

Power supply

R2- stage 2

Cathode Anode Anode Cathode

(A)

Abiotic cathode

Abiotic anode Abiotic cathode

Bio- anode

Alkaline SCOD enriched stream

Acidic water Treated effluent

Neutral low COD stream

Less acidic water

Neutral influentwith Primary

effluent

Filter type Electrolysis cell

Filter typeBES cell

(C)

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 2. Changes in (A) SCOD concentration over time and (B) solubilisation of organic

particulate matter at different pHs.

0

200

400

600

800

1000

1200

0 3 7 24

pH-2 pH-5 pH-7 pH-control pH-8 pH-10 pH-12

SCO

D (

mg/

l)

Time (h)

A

Sol

ubili

satio

n of

pa

rtic

ulat

e m

atte

r (%

)

B

-2

0

2

4

6

8

10

12

14

pH-2 pH-5 pH-7 pH-control pH-8 pH-10 pH-12

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 3. Effect of applied voltage on (A) SCOD content (B) working electrolyte pH and (C)

current production of the filter type electrolysis unit (R1).

0

100

200

300

10 V6 V3 V0 V

0

2

4

6

8

10

12

14

10 V6 V3 V0 V

0

200

400

600

800

1000

1200

1400

0 5 10 15 20 25 30

10 V6 V3 V0 V

SC

OD

(m

g/L

)pH

Cur

rent

(m

A)

Time (h)

A

B

C

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 4. Effect of HRT and applied voltage on (A) current production of R1 and R2; (B)

working electrode potential of R1 and R2 and coulombic efficiency of R2; (C) working and

counter electrolytes pH of R1 and R2; (D, E, F) suspended solids (SS), soluble chemical

oxygen demand (SCOD) and total chemical oxygen demand (TCOD) concentrations of R1

and R2, and overall removal efficiency of the whole process. Vertical dotted arrows in C

indicate renewal of the counter electrolyte of R1 and R2 with fresh DI water.

Ele

ctro

lyte

pH

02468

101214 WE- R1 WE- R2 CE- R1 CE- R2

0

10

20

30

40

0

500

1000

1500Current- R1 Current- R2 HRT

Cur

rent

(mA

)

0

20

40

60

80

100

-2000

-1500

-1000

-500

0

WE- R1

WE- R2

Coulombic efficiency

Ele

ctro

de p

oten

tial

(mV

vs.

Ag/

AgC

l)

Time (d)

A

B

C

6V/ 0.5V 10V/ 0.6VApplied voltage

Medium renewal

0

20

40

60

80

100

0

30

60

90

120

150

Influent-R1Effluent-R1Effluent-R2Overall removal

SS

(m

g /L

)

-20020406080100

0

100

200

300

SCO

D (

mg

/L)

0

20

40

60

80

100

0100200300400500600

0 30 60 90 120 150

TC

OD

(m

g /L

)

Cou

lom

bic

effi

cien

cy (%

)H

RT

(h)

SS o

vera

ll

rem

oval

(%

)S

CO

D o

vera

ll

rem

oval

(%

)T

CO

D o

vera

ll

rem

oval

(%

)

D

E

F

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 5. Current generation in R2 induced by the addition of algal biomass at R1. The

vertical arrows indicate addition of the algal biomass. (HRT= 32 h, applied voltages= 10 V

for R1 and 0.6 V for R2.)

0

100

200

300

400

500Algae to R1

Time (d)

Cur

rent

of

R2

(mA

)

Algae to R1

02468

101214

0 2 4 6 8 10 12 14

WE- R1WE- R2

Wor

king

elec

trol

yte

pH

A

B

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 6. Energetic consideration of the sequential process operated with different HRT and

applied voltages. A and B: specific energy consumption in each stage; C: overall energy

consumption and TCOD removal. Dotted hexagons in C identify the most suitable HRT of 24

h for achieving TCOD removal (80% removal at 44 kWh/kg TCODremoved).

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Highlights (Maximum 85 character per bullet point, including spaces) :

• Novel combination of an electrolysis filter and a BES process for sewage treatment.

• In-situ electrolytic alkaline hydrolysis of entrapped particulates in the filter.

• With no signs of clogging, the combined process removed 100% of SS from influent.

• The combined process did not require external chemicals for pH control.