This article was downloaded by: [RMIT University]On: 10 November 2014, At: 20:44Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Critical Reviews in EnvironmentalScience and TechnologyPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/best20

Recent Advancements in the Treatmentof Municipal Wastewater ReverseOsmosis Concentrate—An OverviewMuhammad Umara, Felicity Roddicka & Linhua Fana

a School of Civil, Environmental and Chemical Engineering, RMITUniversity, Melbourne, VIC, AustraliaAccepted author version posted online: 25 Feb 2014.Publishedonline: 04 Nov 2014.

To cite this article: Muhammad Umar, Felicity Roddick & Linhua Fan (2015) Recent Advancements inthe Treatment of Municipal Wastewater Reverse Osmosis Concentrate—An Overview, Critical Reviewsin Environmental Science and Technology, 45:3, 193-248, DOI: 10.1080/10643389.2013.852378

To link to this article: http://dx.doi.org/10.1080/10643389.2013.852378

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Critical Reviews in Environmental Science and Technology, 45:193–248, 2015Copyright © Taylor & Francis Group, LLCISSN: 1064-3389 print / 1547-6537 onlineDOI: 10.1080/10643389.2013.852378

Recent Advancements in the Treatment ofMunicipal Wastewater Reverse Osmosis

Concentrate—An Overview

MUHAMMAD UMAR, FELICITY RODDICK, and LINHUA FANSchool of Civil, Environmental and Chemical Engineering, RMIT University,

Melbourne, VIC, Australia

Greater use of reverse osmosis (RO)-based processes in municipalwastewater reclamation presents the water industry with a majorchallenge regarding the sustainable management of the resultantRO concentrate (ROC) generated. This has promoted interest ininvestigating different approaches for the cost-effective treatment ofROC to reduce the risks associated with its disposal or reuse. Thisreview highlights and discusses research on the treatment of ROCgenerated mainly from domestic wastewater.

Methods employed for the treatment of municipal ROC arepredominantly physicochemical processes, although biological pro-cesses with varying degrees of success in terms of the removal oforganic content have also been reported. Relatively little attentionhas been paid to the quantification and removal of micropollutants.Due to the recalcitrant nature of some of the organics in ROC, ad-vanced oxidation processes (AOPs) have been demonstrated to bethe most effective for breaking down various pollutants and im-proving the biodegradability of the ROC. UV/H2O2, electrochemi-cal oxidation, and ozonation are among the most studied AOPsfor the treatment of ROC, and some have investigated posttreat-ment changes such as effect on molecular weight distribution andbiodegradability improvement. Although these treatments appearto be promising, high energy consumption and the formation ofharmful by-products during oxidative treatments are challengesthat are yet to be fully addressed. Most studies have been conducted

Address correspondence to Felicity Roddick, School of Civil, Environmental and Chem-ical Engineering, RMIT University, GPO Box 2476, Melbourne, VIC 3001, Australia. E-mail:felicity.roddick@rmit.edu.au

193

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

194 M. Umar et al.

at laboratory scale and therefore large-scale investigation shouldbe carried out to validate the effectiveness of these technologies.

KEY WORDS: reverse osmosis concentrate, wastewater, mem-brane filtration, advanced oxidation processes, ozonation, coagu-lation, biological treatment, UV/H2O2, UV/TiO2, electrochemicaloxidation

1. INTRODUCTION

Wastewater reuse is now being increasingly emphasized as a strategy forconservation of limited fresh water resources (Shon et al., 2006). Pressure-driven membrane processes including microfiltration (MF), ultrafiltration(UF), nanofiltration (NF), and reverse osmosis (RO) are one practical meansof achieving this objective. Due to their larger pore sizes and so higher flux,MF membranes are most commonly used for the removal of suspended par-ticles, turbidity, and microorganisms, whereas UF removes colloids, viruses(Madaeni et al., 1995; van Voorthuizen et al., 2001), and high molecularweight fractions of organic matter. MF and UF are often called low-pressuremembrane filtration as they operate at up to about 8 bar (Khan et al., 2009).Of the high-pressure membrane processes (up to 100 bar), NF and RO areused for the removal of trace organics such as emerging pollutants and dis-solved solids (including ions); the separation capability of NF membraneslies between that of UF and RO membranes (Stephenson et al., 2000). TheRO membranes are used for applications ranging from desalting brackishwater and seawater (Fritzmann et al., 2007) to the treatment of water andwastewater.

RO membranes were the first to be commercialized at large scale andsince then they have gained significant interest in the water industry (UCLA,2011). RO technology has become a practical and affordable means of pro-ducing high-quality recycled water due to its effective and reliable purifica-tion performance, and is now widely used for the polishing of secondaryeffluent in water reclamation schemes such as aquifer recharge, nonpotable,and indirect potable reuses. RO membranes have been proven to be highlyeffective for rejecting a wide variety of organic compounds in feed water: mi-cropollutants such as endocrine-disrupting compounds (EDCs), pharmaceu-ticals and personal care products (PPCPs) and hormones (Snyder et al., 2007;Al Turki et al., 2010), various ions and salts (Lee et al., 2009b), and biologicalmaterials (bacteria, viruses, oocysts, cell fragments) (Comerton et al., 2005).

Over 100 full-scale water reclamation facilities worldwide using mem-brane technology as tertiary treatment were identified by Foussereau et al.(2003). Furthermore, the study revealed that over 97% of the surveyed plants,which required total dissolved solids (TDSs) and organic content removal,used RO membrane technology. Some of the major domestic wastewaterfacilities using RO-based processes include the Sulabaiya reclamation facility

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 195

in Kuwait (375 megaliters per day (MLD)), Orange County, USA (328 MLD),Changi (232 MLD), and Ulu Pandan in Singapore (197 MLD) (Raffin et al.,2013). Other notable examples include West Basin (50 MLD) in California,Kranji (40 MLD), and Bedok (32 MLD) reclamation plants in Singapore. Cur-rently, there are more than 50 plants with capacity >10 MLD worldwide thatuse RO post-UF/MF as a membrane filtration stage postconventional acti-vated sludge for municipal wastewater reuse applications (Graeme Pearce,2013; personal communication, Director, Membrane Consultancy AssociatesLtd, UK). It is expected that there would be at least 10 times the num-ber of plants with capacity <10 MLD than those with capacity >10 MLD(Graeme Pearce, 2013, personal communication; Director, Membrane Con-sultancy Associates Ltd, UK). Although not the focus of this review, thereare 80–100 industrial wastewater plants with capacity <10 MLD that use ROpost-UF/MF for reuse applications (Pearce, 2013, personal communication;Director, Membrane Consultancy Associates Ltd, UK).

The reclaimed water is used for irrigation, groundwater replenishment,indirect potable use, and industrial purpose. In 2008, global water reuse wasjust under 0.2% of total water abstraction, however with a forecast annualgrowth rate of 14%, it has been predicted to outstrip desalination by 2020(Pearce, 2008).

Due to increasing osmotic pressure and membrane scaling issues duringthe RO process, it is generally impractical to obtain 100% water recovery.Approximately 20–30% of the input stream is consequently left as RO con-centrate (ROC, also termed brine or retentate), which contains almost all theconstituents from the RO feed at elevated concentrations. Traditionally, thedisposal of concentrate has been by direct discharge to surface water, sewerdisposal, evaporation ponds, deep well injection, and zero liquid discharge(Lee et al., 2009a). Each disposal method has its own limitations. Further-more not all these disposal methods are suitable for the concentrate arisingfrom municipal wastewater treatment. Direct discharge and sewer disposalare the widely used disposal options, not only for municipal wastewaterROC (Khan et al., 2009), but also for that generated from desalination plants(Mickley, 2008). However, sewer disposal is mostly only suitable for smallplants discharging into large capacity sewage treatment facilities due to thedetrimental effects of the high TDS concentration of ROC on the biologicaltreatment process, as some inhibition may begin to occur when influent TDSexceeds 3 g/L (Voutchkov, 2005). Furthermore, the disposal of ROC to sewerwill result in the buildup of salt level in the long term, and so it is not a sus-tainable option for RO-based WW recycling schemes. Discharging the ROCto the environment particularly, bays or inland aquatic systems, can poseserious health and environmental risks and may have a deleterious impactdepending on the type and concentration of contaminants in the ROC.

Given the increasing use of membranes in generating reusable water, theresultant concentrate volumes facilities need to deal will also increase whichpresent new challenges, particularly for inland cities (Khan et al., 2009). One

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

196 M. Umar et al.

such example is Canberra, Australia, where construction of an RO-basedadvanced reclamation plant was planned in 2007. Sustainable managementof ROC remains one of the major environmental and economic challengesto the water industry as it is one of the greatest limitations to the imple-mentation of membrane processes. Although conventional disposal methodscontinue to be commonly practiced, the issue of the safe disposal of ROChas been increasingly focused upon in recent years to avoid potential long-term environmental and health impacts and to increase the overall yield ofreclaimed water. For example, in Brisbane, the Bundamba advanced wastew-ater treatment plant which contributes purified recycled water to the WesternCorridor Recycled Water Scheme, the largest recycled water scheme in Aus-tralia and the third-largest advanced water treatment project in the world, isrequired to treat ROC and monitor nutrients and metal concentration priorto its discharge into the Brisbane river (Vargas and Buchanan, 2011).

The need for new technical and regulatory approaches to facilitate theexpansion of membrane-based processes and water reuse applications hasbeen emphasized by the U.S. Bureau of Reclamation (USBR, 2003) and theU.S. WaterReuse Foundation (Jordahl, 2006). In addition to minimizing theenvironmental impacts, the economically profitable reuse applications canhelp to offset the costs of treatment processes (Khan et al., 2009). Anotherbenefit of treating ROC is that the wastewater stream is in much smallervolume (about 1

4 of the feed), making the treatment potentially more cost-effective. As a result, there has been growing interest in developing cost-effective ROC treatment strategies over the past decade.

Several treatment schemes have been investigated for the treatment ofROC including physicochemical and biological processes. Advanced oxida-tion processes (AOPs) have been the most widely investigated treatment forROC. However these are energy intensive, and considering that ROC has lowtransmittance and high HO• scavenging potential, these processes may incursignificant cost if used as a single treatment. However integrating suitable preand posttreatment can significantly improve cost effectiveness of the AOPs.This paper provides a review of the recent studies of different technologiesinvestigated for the treatment of ROC with regard to their performance forthe removal of the organic constituents. The focus is on the ROC obtainedfrom domestic wastewater, and where applicable, industrial wastewater. ROCgenerated from brackish and seawater desalination processes is not includedin this review.

2. ROC CONTAMINANTS AND POTENTIALENVIRONMENTAL IMPLICATIONS

The characteristics of ROC used in various studies are given in Table 1.These vary significantly depending on feed water quality, quality of wa-ter produced, type of pretreatment, and the nature of chemicals such as

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

TA

BLE

1.

Com

par

ison

ofm

unic

ipal

RO

Cch

arac

terist

ics

Conduct

ivity

TO

CCO

DTD

SColo

rA

254

DO

CpH

(μS/

cm)

(mg/

L)(m

g/L)

(mg/

L)(P

t.Co)

(/cm

)(m

g/L)

Ref

eren

ces

7.8

5,20

0—

175

—27

453

Bag

asty

oet

al.(2

013)

8.3

5,96

027

.677

——

0.59

5—

Just

oet

al.(2

013)

7.7,

8.2,

8.3

8,30

0,23

,000

—16

4,18

0,23

0—

∼150

,23

50.

67,1.

2434

,47

,53

Um

aret

al.(2

013)

7.66

±0.

18∼6

,000

—13

6±

8—

55.2

±4

0.8,

0.41

42±

2B

agas

tyo

etal

.(2

012)

8.3,

8.5

2,82

0—

65,67

1,68

555

,88

0.43

∼21

Liu

etal

.(2

011,

2012

)7.

823,

870

37−4

091

±9

——

——

Ghys

elbre

chtet

al.(2

012)

7.8,

8.1

7,30

0,12

,760

—14

7,16

8—

101,

228

±50

—42

−62

±4.

5B

agas

tyo

etal

.(2

011a

)∼8

∼4,9

00−5

,000

—∼1

66−1

73—

172−

181

1.2−

1.3

∼56−

58B

agas

tyo

etal

.(2

011b

)7.

7−7.

83,

850,

4,17

0—

—2,

190

—0.

27,0.

2812

,13

Com

stock

etal

.(2

011)

7.5−

7.7

3,97

0−4,

250

——

——

0.91

−1.3

257

Rad

jenovi

cet

al.(2

011)

7.5

5,00

040

——

——

—Zhan

get

al.(2

011)

6.9

±0.

21,

705

±21

18±

260

±v5

1,12

9±

4014

4±

10—

—Zhou

etal

.(2

011a

)8.

322

,300

25−3

612

0−19

014

,745

——

—Zhou

etal

.(2

011c

)7.

91,

700−

4,80

0—

——

——

—Per

ezet

al.(2

010)

7—

—47

0—

——

95va

nAke

net

al.(2

010)

7.5

±0.

21,

990

±25

924

.5±

5—

1,27

6±

166

——

—Le

eet

al.(2

009a

)7.

151,

972

18.4

,15

.6—

——

0.51

—Le

eet

al.(2

009b

)7

10,0

00—

138

5,56

0—

—40

Wes

terh

off

etal

.(2

009)

8.5−

12.3

——

——

——

—D

ialy

nas

etal

.(2

008)

7.72

±0.

532,

025

±15

131

.1±

3.4

——

109

±1

——

Ng

etal

.(2

008)

7.91

,8.2

1,8.

743,

990,

5,06

0,52

90—

151,

171,

218

——

——

van

Heg

eet

al.(2

002,

2004

)

“—”

indic

ates

“notgi

ven”.

197

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

198 M. Umar et al.

antiscalants used (e.g., polyacrylates, polyacrylic acids, or polyphosphates)(Chelme-Ayala et al., 2009; Khan et al., 2009). Addition of acid (such assulfuric acid) may also be needed for pH correction. Both antiscalants andacids influence the chemical equilibrium of the dissolved constituents (vander Bruggen et al., 2003). In some cases, biocides are used to avoid theformation of biofilm on the membrane surface. Hence ROC differs from thefeed water or secondary effluent not only with regard to the concentrationof contaminants but it can also be different in terms of character of theorganic and inorganic pollutants by virtue of the chemicals used prior tothe RO treatment. Limited knowledge on the stability, residence time, andecotoxicity of antiscalants is available (Ahmed et al., 2000), however, it isexpected that their high concentrations may adversely affect the health ofthe receiving ecosystem.

The successful rejection of estrogens, pharmaceuticals, pesticides, andmany other toxic trace pollutants by RO membranes results in their elevatedconcentrations in ROC. A comparison of the various pharmaceuticals foundin ROC exhibited an average concentration factor of 3- to 4-fold that of themunicipal effluent (Benner et al., 2008) warranting their removal from ROCprior to discharge. Concerns related to emerging contaminants such as natu-ral and synthetic hormones, PPCPs, cosmetics, and dioxins are of particularimportance due to their increasingly widespread use, incomplete removalduring wastewater treatment and the fact that they are not regulated in mostcases (Snyder, 2008). Since 2000, about fourteen million synthetic organiccompounds, and thousands more new compounds, have been producedeach year (Al-Rifai et al., 2007). As a result, there is an increasing concernregarding the presence of trace concentrations (μg/L–ng/L) of these com-pounds in municipal secondary effluent due to their toxicity and potential fordisrupting the endocrine systems of animals, even at very low concentrations(Snyder et al., 2007).

Although these micropollutants are found in low concentrations, theirlong-term exposure can have adverse ecological and health effects. For ex-ample, PPCPs can exert toxicity to aquatic organisms, hormone-disruptingagents can interfere with natural hormones, and antibiotics can induce bac-terial resistance (Huber et al., 2005). High concentrations of salts can alsohave consequences such as formation of a dense plume in the vicinity ofthe diffuser system under calm wind conditions, as reported during fieldinvestigations at the Perth desalination plant (Okely et al., 2007), and ionimbalance-triggered toxicity to aquatic flora and fauna (Khan et al., 2009).An increasing regulatory concern is the long-term implication of organicallybound nutrients (Bagastyo et al., 2011a), that is, the nutrients inherentlybound within the structure of the organic compounds.

ROC is also rich in several inorganic constituents such as heavy metalsand various cations and anions (e.g., Cl−, NO2

−, NO3−, PO4

2−, SO42−, Na+,

NH4+, K+, Mg+, Ca+). Depending on the nutrient removal scheme utilized

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 199

in the secondary treatment, ROC may also contain phosphorus and nitrogen,with some of the nitrogen in the form of ammonia which is toxic to manyaquatic species (Walker et al., 2007). The concentration of phosphate inROC can be as high as 40 mg/L for a feed phosphate concentration of5 mg/L (Kumar et al., 2007). Due to its high salt content, the electricalconductivity of ROC is generally very high, and any pH adjustment willincrease it (Lee et al., 2009a). Municipal wastewater ROC with conductivityas high as 23 mS/cm has been reported (Umar et al., 2013). A significantnumber of viruses, bacteria, and cysts may also be present, depending onpretreatment, for example, the type of membrane used (Mickley, 2001).

High concentration of nutrients (phosphorus and nitrogen) can causeeutrophication in the receiving marine waters (Meerganz von Medeazza,2005). Similarly the excessive discharge of nutrients in freshwater en-vironments can contribute to profligate algal growth and subsequentdeoxygenation with serious implications to susceptible waterways (Davisand Koop, 2006).

Depending on the final disposal or reuse application, the ROC may berequired to be treated to an acceptable level of the target contaminants whichis discussed in Section 3.

3. TREATMENT OF ROC

As noted earlier, the conventional methods for ROC management are con-strained due to increased environmental awareness, reuse requirements,and stringent regulations. Reclaimed water guidelines depend on the enduse and are generally based on pH, suspended solids or turbidity, bio-chemical oxygen demand, biological indicators (total or fecal coliforms),nutrient level, and chlorine residual. For example, water reuse guidelinesand regulations for landscape irrigation in California require total coliformcount <2.2/100 mL, turbidity <2 NTU (24 hr median) and chlorine residual>5 mg/L (US EPA, 2004). Similarly in Victoria, Australia, the water qual-ity criteria for irrigation of crops consumed raw and nonpotable uses mustmeet class A recycled water standard (EPA Victoria, 2003). Class A is thehighest quality recycled water with Escherichia coli count <10 per 100 mL,turbidity <2 NTU, BOD <10 mg/L, 5 mg/L SS, and 1 mg/L Cl2 residual.Given that ROC contains contaminants 3–4 times the concentrations foundin secondary effluent, treatment of ROC to an appropriate level is becomingincreasingly important to enable its reuse and reduce the risk it poses to theenvironment by removing the contaminants of concern prior to its disposal.Various treatment schemes have been proposed, including physicochemicaland biological processes and their combinations. Coagulation/flocculation,activated carbon adsorption, oxidative and biological processes, either alone

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

200 M. Umar et al.

TABLE 2. Performance of coagulation for removal of organic content of ROC

Parameters and removal

Type DOC removal (%) Dosage (mM) pH References

Alum 52, 25 1.5 5 Bagastyo et al. (2011a)Iron 34, 38 1.5 5 Bagastyo et al. (2011a)Iron 58% 8.95 7.7 Comstock et al. (2011)FeCl3 26 1 — Zhou et al. (2011a)Ferric ions 5 — — Westerhoff et al. (2009)Alum 42 2 — Dialynas et al. (2008)Iron 52 0.4 — Dialynas et al. (2008)

“—” indicates “not given”.

or in combination, have been studied for the treatment of ROC. Section 4provides an overview of the various processes used.

4. PHYSICOCHEMICAL TREATMENT OF ROCs

4.1. Coagulation

Coagulation is a relatively simple and commonly applied process for thetreatment of water and wastewater. It consists of four distinct mechanisms:(1) compression of the diffuse layer (van der Waals interaction); (2) adsorp-tion to produce charge neutralization (destabilization), (3) enmeshment in aprecipitate when a high dosage of coagulant is used, leading to sweep coagu-lation; and (4) adsorption to permit interparticle bridging (complex betweenparticle and polymer with synthetic organic coagulant) (Vigneswaran andVisvanathan, 1995). The predominant mechanism mainly consists of chargeneutralization of negatively charged colloids by cationic hydrolysis prod-ucts (Duan and Gregory, 2003). Aluminum sulfate, polyaluminum chloride,and iron salts are among the most commonly used coagulants in water andwastewater treatment.

Table 2 lists the studies reported for the treatment of municipal wastew-ater ROC by coagulation. Dialynas et al. (2008) evaluated the performance ofcoagulation with alum and ferric chloride for the concentrate obtained fromthe RO treatment of n membrane bioreactor effluent using a lab-scale RO pi-lot unit. The optimum dosage using alum was 2 mM (as Al3+) with dissolvedorganic carbon (DOC) removal of 42% (initial DOC, 8.5 mg/L). Coagulationusing FeCl3 gave a DOC reduction of 52% (initial DOC, 12.3 mg/L) at anoptimum dosage much less than that of Al3+, that is, 0.4 mM as Fe3+. Usinga higher concentration of FeCl3 (1 mM Fe3+) than Dialynas et al. (2008),much lower removal of DOC (initial concentration, 18 mg/L) was reportedby Zhou et al. (2011a) for the ROC collected from the second stage of anRO process. The authors established that the DOC removal increased withincreasing dosage of coagulant up to 1 mM to give a DOC removal of 26.4%.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 201

However, the high specific ultraviolet absorbance (SUVA) value (4.4, higherthan for the raw sample) of the coagulated ROC indicated that only a smallfraction of the aromatic content was removed. This low removal efficiencywas attributed to the inability of FeCl3 to remove soluble organic compoundswith low molecular weight as it mainly targets organic compounds with highmolecular weight (>10 kDa) (Shon et al., 2006).

Coagulation of ROC (initial DOC, 13 mg/L) obtained from a municipaldrinking water treatment plant was investigated by Comstock et al. (2011) us-ing ferric sulfate at three metal dosages (1.79, 4.48, 8.95 mM Fe3+). Althoughthe initial DOC concentration of ROC was comparable with that of Dialynaset al. (2008), they used a significantly higher dosage of coagulant (8.95 mMFe3+) for a fairly similar level of DOC reduction (58%) due to the differ-ent source water and hence different characteristics of the organic content.Using two different ROC samples of considerably higher concentrations ofDOC (42, 62 mg/L) than reported in the abovementioned studies, Bagastyoet al. (2011a) reported significantly different DOC reductions for both sam-ples for Al3+, that is, 52%, 25%. However, the reduction of DOC for Fe3+ wascomparable for both samples, that is, 34% and 38%. They found an optimumdosage of 1.5 mM for both coagulants at pH 5, with little improvement inthe removal performance for higher dosages. Consistent with the findingsof Dialynas et al. (2008), Fe3+ performed better than Al3+, particularly forremoving the color and chemical oxygen demand (COD) from one of theROC samples as well as good removal of noncolored DON compounds suchas sugars, proteins, amino acids, and other recalcitrant organic compoundswhich were not removed by Al3+. However similar removal efficiencies werereported for both coagulants for the other ROC sample as it was predomi-nantly comprised of large molecular weight color-causing compounds whichwere effectively removed by Al3+.

Coagulation as pretreatment can increase the overall removal of organiccompounds as found by Zhou et al. (2011a) who reported enhanced reduc-tion of DOC by coupling coagulation (1 mM FeCl3) with different AOPs. Foran ozone dosage of 0.45 mg/L, coagulated ROC showed markedly higherDOC reduction (45%) than using O3 alone. Similarly, a DOC reduction of49% was reported when the coagulated ROC was treated by UVA/H2O2 treat-ment compared with the DOC reduction of <3% by the UVA/H2O2 treatmentalone.

Coagulation as pretreatment has not been studied extensively for thetreatment of ROC, and there is a lack of comparison of different coagulantsand the optimum conditions in terms of coagulant dosage and pH. Takinginto account that ROC may contain a significantly high concentration ofsalts (Zhou et al., 2011a; Umar et al., 2013) and coagulation of high-salinitywater has been proposed to occur differently than for low-salinity water(Duan et al., 2002, 2003; Edzwald and Haarhoff, 2011), the process must beinvestigated for its efficiency in removing the organic content of high-salinity

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

202 M. Umar et al.

ROC where applicable. Coagulation is well known to target large molecularweight compounds, however, a significant proportion of the organic contentconsists of low- to medium-MW compounds (Lee et al., 2009a), and thereforecoagulation is not a viable option on its own and is recommended to be usedas a pretreatment. Iron-based coagulants appear to be more effective thanaluminum-based ones in treating ROC, although the organic matter removalefficiency depends significantly on the characteristics of the ROC.

4.2 Advanced Oxidation Processes

AOPs are promising by virtue of the generation and the reaction of extremelyreactive hydroxyl radicals (HO•). Due to their electrophilic nature, HO• cannonselectively oxidize almost all electron-rich organic molecules, eventuallyconverting them to CO2 and water. Generation of HO• leads to the oxidationof organic compounds by radical addition (Eq. (1)), hydrogen abstraction(Eq. (2)), and electron transfer (Eq. (3)) (Legrini et al., 1993).

R• + O2 → ROO• (1)

RH + HO• → R• + H2O (2)

RX + HO• → RX• + + HO− (3)

where R refers to the reacting organic compound.Most AOPs use combination systems such as two oxidants (O3/H2O2),

catalyst and oxidant (Fe2+/H2O2), oxidant with irradiation (H2O2/UV), ox-idant with photocatalyst (H2O2/TiO2/hv) or oxidants with ultrasound (US)(H2O2/US) (Lopez et al., 2004). AOPs have proven highly efficient for treat-ing complex compounds including recalcitrant organics in mature landfillleachate (Umar et al., 2010), anti-inflammatory and analgesic pharmaceuti-cals and antibiotics (Motwani et al., 2011; Ziylan and Ince, 2011), chlorophe-nols (Al Momani et al., 2004; Saritha et al., 2009), and several EDCs (Wertet al., 2009; Bertanza et al., 2010; Mosteo et al., 2010). Various AOPs havebeen reported recently for the treatment of ROC. Section 4.2.1 provides anoverview of these applications, and Table 3 provides a summary of theseinvestigations.

4.2.1. UV/H2O2 TREATMENT

UVC/H2O2 is one of the most commonly used AOPs for the treatment ofwater and wastewater. Due to the successful application of UVC/H2O2 fortreating recalcitrant compounds in domestic and industrial wastewater, its usein the treatment of ROC has recently been investigated by several researchers(Westerhoff et al., 2009; Bagastyo et al., 2011a; Zhou et al., 2011a; Liu et al.,2011, 2012; Justo et al., 2013; Umar et al., 2013). UV/H2O2 treatment iscapable of removing a wide range of organic compounds of different MW

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 203

TABLE 3. DOC reduction of ROC by various AOPs

DOC removal Reaction H2O2

(%) time (min) concentration Process References

TOC, 9.6 61.7 0.54 mg/mg TOC UVC/H2O2 Justo et al. (2013)TOC, <1% 5.4 1.38 mg/mg TOC O3 Justo et al. (2013)26−38 60 3 mM UVC/H2O2 Umar et al. (2013)40−60 120 3 mM UVC/H2O2 Liu et al. (2012)38, 40 120 11.8 mM UV/H2O2 Bagastyo et al. (2011a)43.5 — — UVA/H2O2 Zhou et al. (2011a)72 360 — UVA/TiO2 Zhou et al. (2011a)95 360 — UVC/TiO2 Zhou et al. (2011a)41.1 — — O3 Zhou et al. (2011a)43.6 — — US/O3 Zhou et al. (2011a)68.1 — — UV/TiO2/O3 Zhou et al. (2011a)∼49−76 60 2−6 mM UVC/ H2O2 Liu et al. (2011)∼56 60 2 mM VUV/ H2O2 Liu et al. (2011)COD, ∼20 40 0.228 mol/mol O3 O3 van Geluwe et al. (2011a)43 — 2.4 mM Fe2+/ H2O2 van Aken et al. (2010)19 — 2.4 mM UV/O3 van Aken et al. (2010)50 — 10 mM Fe3+/ H2O2 Westerhoff et al. (2009)75 — 0.7 mol/mol O3 O3/ H2O2 Westerhoff et al. (2009)40 — 10 mM UV/ H2O2 Westerhoff et al. (2009)95 — — UV/TiO2 Westerhoff et al. (2009)23.4 20 — O3 Lee et al. (2009b)49, 41 — — UV/TiO2 Dialynas et al. (2008)29, 34 — — US Dialynas et al. (2008)

“—” indicates “not given”.

(Dwyer and Lant, 2008). The generation of HO• occurs by photolysis of H2O2

according to the following reaction (Baxendale and Willson, 1957):

H2O2 + hv → 2HO• (4)

Due to its weak acidity, H2O2 can dissociate to H+ and HO2− (Eq. (5)):

H2O2 → H+ + HO−2 (5)

HO2− can be a source of HO• under UV irradiation (Legrini et al., 1993) as

shown in Eq. (6):

HO−2 + hv → HO• (6)

Decomposition of H2O2 through dismutation is another way of HO•

generation (Legrini et al., 1993) as described in Eq. (7):

H2O2 + HO−2 → H2O + O2 + HO• (7)

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

204 M. Umar et al.

At high local HO• concentration, recombination of HO• occurs to form H2O2

(Legrini et al., 1993) as given in Eq. (8):

HO• + HO• → H2O2 (8)

Formation of hydroperoxyl radicals (HO2•) which possess markedly lower

oxidizing activity than HO• takes place in the presence of excess H2O2 viathe following reaction (Eq. (9)):

H2O2 + HO• → HO•2 + H2O (9)

A list of the recent studies on AOPs, including the UV/H2O2 process, for thetreatment of ROC is given in Table 3. Using a lab-scale batch recirculationsystem, Westerhoff et al. (2009) investigated the efficiency of the UVC/H2O2

process for treating ROC with an initial DOC concentration of 40 mg/L. ADOC reduction of 40% was reported using 10 mM H2O2 at pH 4 (UV fluencenot given). Low DOC reduction (2.3 ± 2.8% at pH 6.9) was reported by Zhouet al. (2011a) for UVA/H2O2 treatment of raw ROC for a UV fluence of 277.2kJ/m2 (irradiation intensity 7.7 mW/cm2, time 60 min) and H2O2 dosage of10 mM. This low DOC reduction may have been due to the higher molarabsorption coefficient of H2O2 at 254 nm for UVC than at 360 nm for UVAas highlighted by Liu et al. (2011).

Liu et al. (2011) compared the performance of UVC (254 nm) and VUV(254 + 185 nm) alone and in combination with H2O2 to treat three differentROC samples (DOC ∼ 20–25 mg/L) prepared from municipal secondaryeffluents using a batch lab-scale RO rig. VUV alone gave markedly higherDOC reduction (34%) compared with the UVC alone treatment (21%), whichwas attributed to the formation of HO• in situ via water photolysis by the185 nm component of VUV irradiation. However the time required achievingthis reduction was very high (3 hr, UV fluence 1,392.12 kJ/m2). Therefore,they used various dosages of H2O2 (1–6 mM) combined with UVC irradiationand reported enhanced performance with proportional increase in reactionrate constant with increasing H2O2 dosage up to 2 mM. Scavenging of HO•

occurred for H2O2 dosage >2 mM, which was attributed to the formationof less reactive HO2

• (Eq. (9)) (Buxton et al., 1988). The kinetics for CODreduction by the UVC/H2O2 process were expressed as pseudo-first-orderfor 30 min of reaction. The reduction of DOC was faster at high H2O2

dosages, particularly in the first 1 hr of reaction (UV fluence 464.04 kJ/m2).The reduction of DOC and COD reached a plateau after 1 hr treatmentdue to the depletion of H2O2 and reduction in the concentration of easilybiodegradable organic compounds. The authors noted that the VUV/2 mMH2O2 process (UV fluence 642.24 kJ/m2) gave comparable COD reduction(57%) to UVC/6 mM H2O2 treatment (55%) after 1 hr.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 205

Different from the reduction of COD, the DOC reduction (∼56%) byVUV/2 mM H2O2 was slower and less (76%) than the UVC/6 mM H2O2 treat-ment, but still greater than for the UVC/2 mM H2O2 treatment (49%) (Liuet al., 2011). The oxidation of organics was suggested to follow differentpathways for both the UV-based AOPs. Apart from the continuous genera-tion of HO• via water photolysis, the better performance of the VUV/H2O2

process can be attributed to initiation of the generation of hydroperoxylradicals. Although these radicals possess lower oxidation strength than thehydroxyl radicals, their generation without compromising the generation ofhydroxyl radicals can improve the oxidation efficiency through the additionalbreakdown of aromatic organics (Oppenlander et al., 2005).

Compared with the DOC and COD reductions discussed above, muchgreater reduction in A254 (>90%) and color (>95%) was noted for theUVC/H2O2 and VUV/H2O2 treatments (Liu et al., 2011). These large reduc-tions were attributed to the breakdown of the conjugated and chromophoricstructure of the organic compounds. Similar to the trends for DOC and CODreduction, the rate and extent of reduction in A254 and color were greater forthe VUV/H2O2 system, and the fluorescence excitation emission matrix spec-tra revealed that this was related to the more rapid and greater breakdownof the fluorophores.

A batch study on the treatment of two ROC samples (DOC of 42 and62 mg/L) collected from two different WWTPs was conducted by Bagastyoet al. (2011a) using the UVC/H2O2 process (UV fluence not given). Althoughthe initial DOC concentration was comparable with that of reported by West-erhoff et al. (2009), they used a considerably higher H2O2 dosage of 400 mg/LH2O2 (11.8 mM) for a UV power consumption of 3.1 kW h/m3 for 120 min.Although the samples were of different initial concentrations, they exhibitedfairly comparable COD (55 and 50%) and DOC removal (38 and 40%). In-creasing the H2O2 dosage to 600 mg/L (17.6 mM) gave almost similar CODand DOC reductions. The reduction in DON was 32% and 27% for the twoROC samples using 400 mg/L H2O2. This poor DON removal was possiblydue to the high proportion of low-MW organic compounds which could havebeen neutral or positively charged compounds with low reactivity towardsoxidation by the UVC/H2O2 process (Bagastyo et al., 2011a).

The impact of salinity and initial DOC concentration on the UVC/3 mMH2O2 treatment of ROC was investigated by Liu et al. (2012). Over 2 hrof reaction (UV fluence 928.08 kJ/m2), the reduction of DOC was influ-enced little by increasing the salinity 4-fold (original electrical conductiv-ity of 2,820 μS/cm). Increasing the initial DOC concentration (from 21to 26 and 30 mg/L) revealed similar trends in the reduction of normal-ized COD, DOC, and A254. Although the net reduction was higher for thehigher initial DOC concentration, the residual DOC concentration was alsohigher. The target residual DOC concentration of <10 mg/L was obtainedafter 30 min UVC/3 mM H2O2 (UV fluence 232.02 kJ/m2) with downstream

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

206 M. Umar et al.

biological treatment (as biodegradable dissolved organic carbon (BDOC)test), suggesting the applicability of biological posttreatment for improvedoverall performance and cost effectiveness as discussed in detail in Section8.

The potential of the UVC/H2O2 process for treating three ROC samplesof very different initial DOC concentrations (34, 47 and 53 mg/L) was inves-tigated by Umar et al. (2013). Unlike Liu et al. (2011, 2012), the authors usedROC samples collected from the municipal wastewater reclamation facility ofa WWTP. After 1 hr UV/3 mM H2O2 treatment (UV fluence 464.04 kJ/m2), thereduction of DOC and COD for the three samples was 26–36% and 25–37%,respectively, which was lower than that reported by Liu et al. (2011, 2012).Under similar irradiation conditions of UV fluence and H2O2 dosage, theyreported DOC and COD reductions of 56% and 47%, respectively, and thiscan be attributed to the different initial characteristics (mainly DOC, COD,TDS, color, and alkalinity) of the samples used by Umar et al. (2013). Assimilar H2O2 concentration was used in both studies, the difference in theinitial DOC concentrations led to different H2O2/DOC ratios, that is, 4–5 (Liuet al., 2011) and 2–3 (Umar et al., 2013). Although different DOC reductionswere reported for the different H2O2/DOC ratios, the reductions in color(>80%) and A254 (75–80%) were consistent with Liu et al. (2011, 2012).

pH is one of the important parameters affecting the efficiency of theUV/H2O2 process. Liu et al. (2012) investigated the effect of pH (pH 4, 6, 8.5,and 10) on the reduction of organic compounds by UVC/H2O2 treatment andfound that it decreased with increasing pH. Low organic content reductionat high pH values was attributed to the presence of bicarbonate/carbonate(HCO3

−/CO32−) species which are strong •OH scavengers (Eqs. (10) and

(11)) (Weeks and Rabani, 1966).

HCO−3 + H2O

• → CO−•3 + H2O, k = 1.5 × 107 M−1s−1 (10)

CO2−3 + HO• → CO−•

3 + H2O, k = 4.2 × 108 M−1s−1 (11)

The CO3−• ion has much lower oxidation potential and is highly selective

in reacting with organic compounds compared with HO• (Liao et al., 2001).Carbonate and bicarbonate species exist in approximately the same pro-portions at pH 10, whereas reducing the pH to 8.5 results in the exclusivepresence of HCO3

− (Oppenlander, 2003). As indicated by the reaction con-stants given in Eqs. (10) and (11), since CO3

2− reacts with HO• almost twomagnitudes faster than HCO3

−, a reduction in pH from 10 to 8.5 can im-prove the efficiency of the UVC/H2O2 process. A further reduction in pH to6 leads to approximately 50% of the HCO3

− being transformed to H2CO3

which has very low reactivity with HO•, whereas at pH 4 the main inorganiccarbon component is dissolved CO2, only 0.1% of which reacts with waterto form H2CO3 (Oppenlander, 2003). Thus the scavenging effect caused by

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 207

the HCO3−/CO3

2− species is almost completely eliminated at pH 4 leadingto better oxidation performance. The reduction of DOC increased to 40% atpH 4 compared with <10% at pH 7 under similar conditions of UVC/H2O2

treatment which is in agreement with Westerhoff et al. (2009). Zhou et al.(2011a) reported an increase in DOC reduction from 2.3 at pH 6.9 to 17%when the pH was reduced to 5.3 during UVA/H2O2 treatment.

UVC/H2O2 appears promising for oxidizing recalcitrant organic com-pounds in ROC, however, for wastewater such as ROC with inherent lowtransmittance and high HO• scavenging potential, the process may incurhigh energy requirements limiting its economic feasibility. The oxidation ef-ficiency of UVC/H2O2 can be improved taking into consideration the wave-length of UV irradiation and optimization of UV and H2O2 dosages. A step-wise or continuous addition of H2O2 can be considered to improve removalefficiency as reported by Liu (2011), and reduce overall chemical consump-tion as established in other hard-to-treat wastewaters such as landfill leachate(Primo et al., 2008; Hermosilla et al., 2009). A significant improvement in thedegradation efficiency can be achieved by reducing the pH to an appropriatevalue to avoid scavenging of hydroxyl radicals by HCO3

−/CO32−. However,

then the pH of the treated water has to be raised to neutral for its reuse whichinvolves the addition of chemicals. The use of chemicals such as H2O2 is gen-erally needed for effective oxidation which adds to the cost of the treatmentprocess. Pretreatment for the UVC/H2O2 process can significantly improvetransmittance of ROC, leading to reduced chemical consumption (H2O2) andradiation dose. Although the UVC/H2O2 process has increasing applicationsin treating drinking water and tertiary treated water, to date its applicationto ROC treatment has been limited to lab studies.

4.2.2. UV/TIO2 PHOTOLYSIS

UV-driven photodegradation of organic pollutants, particularly using TiO2

as a photocatalyst, has received significant attention due to its ability tobreak down organic pollutants and even achieve complete mineralizationthrough the generation of strongly reactive-free radical oxidants (Ahmed,2003; Miranda-Garcıa et al., 2011). The photocatalytic and hydrophilic prop-erties of TiO2 make it almost an ideal catalyst due to its high reactivity, lesstoxicity, chemical stability, and low cost (Fujishima et al., 2000).

Heterogeneous photocatalysis using UV/TiO2 is one of the most com-mon photocatalytic processes and is based on the absorption of photonswith energy higher than 3.2 eV (wavelengths lower than ∼390 nm) initiatingexcitation related to a charge separation event (band gap) (Boroski et al.,2009). Generation of high-energy excited states of electron and hole pairs oc-curs when wide band gap semiconductors are irradiated with light of energyhigher than their band gap energy (Miranda-Garcıa et al., 2011). It results inthe promotion of an electron in the conductive band (eCB

−) and formationof a positive hole in the valence band (hVB

+) as shown in Eq. (12) (Boroski

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

208 M. Umar et al.

et al., 2009). The hVB+ oxidizes H2O molecules and OH− ions resulting in

the generation of HO• (Eqs. (13) and (14)) (Baird, 1997).

TiO2 + hv → e−CB + h+

VB (12)

H2O(ads) + vh+VB → HO• + H+ (13)

OH−(ads) + h+

VB → HO• (14)

Additionally, eCB− can react with O2 forming the anionic radical superoxide

(Eq. (15)), which can subsequently lead to generation of hydrogen peroxideand HO• (Baird, 1997).

O2 + e−CB → O•−

2 (15)

A list of studies conducted on the treatment of ROC using the UV/TiO2

process alone or in combination with other AOPs is given in Table 3.A suspension of TiO2 was used with irradiation by UVA in a batch-type

laboratory scale photoreactor for the treatment of ROC by Dialynas et al.(2008). For a reaction period of 60 min, only a small difference in the reduc-tion of DOC was reported after doubling the dosage of TiO2, that is, 41% and49% for 0.5 g/L and 1 g/L TiO2, respectively. This was probably due to the in-creasing opacity caused by the suspended catalyst (Gogate and Pandit, 2004).Initial drop in DOC concentration was rapid which corresponded to the ad-sorption of DOC onto TiO2 and oxidation of easily degradable organic mat-ter, however, the reduction slowed after 10 min indicating the recalcitrant na-ture of the remaining organic compounds and possible saturation of the TiO2.

Westerhoff et al. (2009) reported a maximum DOC reduction of up to95% with the UVC/TiO2 process followed by biological treatment (Table 2).The authors reported that the degradation efficiency was nearly independentof the TiO2 dosage between 1 and 5 g/L, which can be attributed to in-creasing opacity due to the catalyst limiting the reaction rate. Comparison ofUV/TiO2 treatment with other processes (coagulation, Fenton, and O3/H2O2)for treating ROC suggested that UV/TiO2 is the most effective both in termsof DOC reduction and energy efficiency (Westerhoff et al., 2009). Greaterreduction of absorbance at 254 nm than DOC was noted implying that mostof the organic compounds were degraded in the following sequence: bulkorganics → aldehydes → carboxylic acids → carbon dioxide. The UV/TiO2

process reached a plateau at a DOC reduction of 87% for an energy inputof 9 kW h/m3, indicating the conversion of organic compounds to interme-diates refractory to degradation by the UV/TiO2 treatment. The majority ofthe partially oxidized organic compounds were identified as oxalate, acetate,and propionate which are not readily degraded by the UV/TiO2 process. Thereaction of HO• with these compounds was considered as the rate-limitingstep. As these compounds can be degraded biologically, the authors thenemployed biologically acclimated sand reactors and reported a final water

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 209

quality of around 2 mg/L DOC (initial concentration 40 mg/L). This sug-gests that the integration of UV/TiO2 with biological treatment can resultin enhanced removal performance and economic benefits. The removal oftrace organics was also effective, and all the pharmaceutical compoundsmonitored were reduced to below the detection limit of 2 ng/L.

Similar to UV/H2O2 treatment, the UV/TiO2 process performed better atpH 5 than at pH 7 (Westerhoff et al., 2009). However pH below 5 gave noadditional benefit in terms of oxidation efficiency. The higher DOC reductionat pH 5 was attributed to the absence of carbonate species which is an activeHO• scavenger as noted in Section 4.2.1 (Eqs. (10) and (11)). As the pHwas increased to 7, the carbonate/bicarbonate species reduced the steady-state HO• concentration and thus reduced the oxidation efficiency of theprocess. Furthermore, the performance at pH 5 was probably enhanced bythe greater affinity of the organic compounds for TiO2 at that pH (pHIEP ∼6.25 for P25) due to the availability of a larger number of positively chargedsurface sites. pH is the most important parameter affecting the zeta potentialof TiO2 and hence its affinity for the organic matter. Under acidic conditions,the positively charged surface of the TiO2 leads to greater affinity towardsorganic matter than under alkaline conditions.

Zhou et al. (2011a) studied various AOPs including UVA/TiO2 andUVC/TiO2 for the treatment of ROC. The TiO2 dosage used was 1 g/L andlight intensity was 7.1 and 9.1 mW/cm2 for UVA and UVC lamp, respectively.They showed that UVC/TiO2 was more effective for reducing the DOC ofROC, and the improvement was more pronounced for samples pretreated bycoagulation with FeCl3 compared with the raw ROC samples. The treatmentof coagulated samples showed a DOC reduction of 95% (UVC/TiO2) and72% (UVA/TiO2) after 6 hr. A reaction time of >6 hr led to almost com-plete breakdown of the aromatic content to small molecular weight organiccompounds. The authors also investigated UVA/TiO2 treatment with O3 andreported that the presence of O3 gave an additional 38% DOC reduction(cumulative DOC removal of 52%), and when coagulated ROC was treatedby the UVA/TiO2/O3 process, the overall DOC reduction was 68% after 1 hrcompared with 42% under similar conditions without O3. The improved DOCreduction for UVA/TiO2/O3 was attributed to the continuous generation ofHO• through electron trapping by O3 on the TiO2 surface in addition to theindividual ozonation and photocatalysis steps (Zhou et al., 2011a) accordingto Eqs. (16) and (17):

e−CB + O3 → •O−

3 (16)

O−3 + H+ → HO• + O2 (17)

The use of the UV/TiO2 system has been proven efficient for the degradationof organic compounds at lab scale. Coagulation and biological treatment aspre and posttreatment, respectively, can significantly improve the overall

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

210 M. Umar et al.

degradation efficiency and cost effectiveness. However, when the UV/TiO2

system is used with suspended catalyst particles as in the case of Westerhoffet al. (2009) and Zhou et al. (2011a), concerns may arise regarding theirseparation and recycling which can be an inconvenient, time-consuming,and expensive process (Ray and Beenackers, 1998). Moreover, the solutionwill be less penetrable by UV radiation (i.e., have low UV transmissivity)in the presence of the suspended catalyst particles (Ray and Beenackers,1998). Therefore the dosage of catalyst needs to be carefully selected, andinvestigation of the effect of pH and other variables is needed to achievemaximum process efficiency. Moreover the large-scale application of theUV/TiO2 system is limited by several factors such as the wide band gap(3.2 eV), the development of more efficient novel catalysts, and systems forretrieval and reuse of the catalyst are needed to improve the prospects ofindustrial application.

4.2.3. OZONATION

Due to its high oxidation potential, ozone is extremely effective for severalapplications in water and wastewater treatment such as color removal, odor,and taste control, disinfection and removal of resultant by-products, degrada-tion of organic compounds and biodegradability improvement of recalcitrantwastewaters (van Geluwe et al., 2011a; Zhou et al., 2011a). Organic contam-inants are oxidized through direct reaction with molecular ozone or throughindirect reactions with free radicals (HO•) produced by the decomposition ofozone (Broseus et al., 2009). Direct electrophilic attack by molecular ozone(ozonolysis) is highly selective and takes place under acidic or neutral condi-tions, or in the presence of radical scavengers that inhibit the chain reactionresponsible for ozone decomposition. Ozone has been shown to preferen-tially remove molecules with low oxidation state (low C/O ratio) and highdegree of unsaturation (low H/C ratio) (These and Reemtsma, 2005), whichis due to the markedly higher rate of reaction between ozone and aromaticrings (106–109 M−1 S−1) than saturated reaction products (10−5 and 101 M−1

S−1) (van Geluwe et al., 2011b). However, due to the presence of severaltypes of radical scavengers in wastewater such as carbonate and bicarbon-ate, oxidation by molecular ozone is the dominant mechanism, particularlyat lower ozone dosages (Nakada et al., 2007). The reaction of molecularozone with organic compounds results in the formation of carboxylic acidsas end products that are not further oxidized by molecular ozone (Agustinaet al., 2005). The compounds susceptible to ozonolysis possess C–C doublebonds, specific functional groups (e.g., OH, CH3, OCH3), and anions (thoseof N, P, O, S) (Alvares et al., 2001).

Under alkaline conditions, or in the presence of solutes that promotethe radical-type chain reaction and HO• formation, the indirect reaction pre-dominates due to the extremely rapid and nonselective nature of HO• (rateconstant of 109 M−1 s−1) (Staehelin and Hoigne, 1985). The rate of HO•

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 211

generation depends on the water matrix, particularly its pH, alkalinity, type,and content of organic matter (von Gunten, 2003). Ozone decompositiontakes place in chain reactions which include initiation (Eqs. (18) and (19)),propagation (Eqs. (20)–(24)), and termination steps as given below:

4.2.3.1. Initiation. The reaction between OH− ions and O3 results inthe formation of hydroperoxide ion (HO2

−) and is considered as the firstreaction of the mechanism (Beltran, 2004) (Eq. (18)).

O3 + OH− → HO−2 + O2 (18)

The hydroperoxide ion then reacts with O3 to give:

HO−2 + O3 → HO•

2 + O−•3 (19)

4.2.3.2. Propagation. Further decomposition of ozone takes placedue to its reaction with O2

•− that leads to the formation of ozonide ion(O3

•−) (Eq. (20)):

O3 + O•−2 → O•−

3 + O2 (20)

The ozonide ion ultimately protonates to form HO3• (Eq. (21)), which is then

converted to HO• (Eq. (22)). The HO• can react with O3 to form HO4• (Eq.

(23)) followed by the decomposition of HO4• to HO2

• (Eq. (24)).

O•−3 + H+ → HO•

3 (21)

HO•3 → HO• + O2 (22)

O3 + HO• → HO•4 (23)

HO•4 → HO•

2 + O2 (24)

4.2.3.3. Termination. This step involves any recombination of HO•,HO2

•, and O2.Ozonation of ROC has been reported either alone or in combination

with other AOPs (Table 3). Ozonation has been used as a pretreatment forthe natural organic matter content of NF concentrate to improve perme-ate flux (van Geluwe et al., 2011a). Decomposition of organic matter wasstudied in terms of optical density (275 nm) and was compared with thedecomposition of COD. It was noted that almost 50% reduction in opti-cal density occurred in the first 10 min followed by slow oxidation due tothe transformation of unsaturated bonds into oxygenated reaction productssuch as aldehydes, ketones, and carboxylic acids. The reduction of CODwas considerably lower, with a decrease of 19–25% after 30 min, which re-mained almost unchanged (van Geluwe et al., 2011a). The slow reduction inCOD can be attributed to inefficient reaction of ozone with the by-products(von Gunten, 2003). A higher drop in the optical density compared with

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

212 M. Umar et al.

the overall COD removal was due to the selective oxidation by ozone ofthe unsaturated bonds under the reaction conditions (pH 7.7–8.4). An O3

residual appeared after 28 min of reaction and reached a stable aqueousconcentration (pseudo Henry plateau) after 115 min and 90% reduction inthe optical density. The removal of unsaturated bonds resulted in significantdecrease in the concentration of hydrophobic XAD-7 acids corresponding toa decrease of 86% hydrophobic COD. However it represented only 10% ofthe total COD, the overall reduction of which was 22%.

Ozone as an individual process has been reported for the degradationof organic content of ROC by Zhang et al. (2011) and Lee et al. (2009b)for a comparable initial total organic carbon (TOC) and COD concentra-tions of ∼18 and 60–65 mg/L, respectively. Zhou et al. (2011a) investigatedthe degradation of organic compounds of ROC using O3 alone (1 L/min;17.6 ± 8.3 mg/hr) and reported a DOC and COD reduction of 22% and14%, whereas the corresponding color reduction was 90%. Lee et al. (2009b)investigated the reduction of organic content in batch and continuous labexperiments. The reduction of TOC in batch experiments was 25% after20 min ozonation (10 mg O3/L) corresponding to 1.65 mg O3 consumedper mg TOC removed. The improvement in TOC reduction was <2% whenthe ozone dosage was increased from 6 to 10 mg/L, suggesting that theremaining organic compounds were recalcitrant to ozone. As direct ozoneattack is the prevailing mechanism at neutral pH, and given that the authorsperformed ozonation at pH 7.1, the organics were believed to be convertedto carboxylic acids which are recalcitrant to molecular ozone (Lee et al.,2009b). The reduction of A254 under similar conditions was 75% whereascomplete decolorization was achieved for ozone dosage of as low as 3 mgO3/L. They determined COD reduction in addition to the reduction of CODin continuous lab experiments with respective reductions of 26% and 65%.The consumption of O3 was 0.36 ± 0.08 mg for each mg TOC reduced. Ithas to be noted that although a fairly comparable organic reduction (TOCand DOC) was reported, the reduction of COD was markedly different inthese studies in spite of the comparable initial COD concentration.

Lee et al. (2009b) employed ozone to enhance the biodegradability ofthe organic content of ROC (TOC 15.6 mg/L) before biological activated car-bon (BAC) treatment in a batch study. They subjected ozonated ROC (ozonedosage 6.0 mg/L, contact time 20 min) to the BAC system. Ozonation ledto an increase in biodegradability due to the effective breakdown of unsat-urated hydrocarbons and aromatic compounds which was indicated by thelarge reduction (75%) of A254 that led to an increase in biodegradability from<2 to 5.3 mg/L. The BAC treatment gave at least twice the removal of O3

alone and the O3–BAC process achieved 69.8% and 88.7% removal of CODand TOC, respectively. The A254 was three times lower for O3-BAC than BACalone. The ratio of TOC/A254 can be considered as a tool to assess the ap-plicability of biological posttreatment (Tambo and Kamei, 1978). The initial

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 213

TOC/A254 of 35 increased to 107 after ozonation for 20 min confirming the ef-fectiveness of ozone in breaking down the large compounds and enhancingthe biodegradability of the treated ROC. An almost similar TOC/A254 ratio of105.5 was reported by Wang et al. (2007) during the ozonation of secondaryeffluent (TOC of 8–10 mg/L) after a relatively short time of 10 min. There-fore, prolonged contact time for ROC may be needed to achieve effectivedegradation of its organic content. As there was a significant increase in theratio of TOC/A254, the ozonated ROC was suitable for subsequent biologicaltreatment.

Ozone in combination with UV irradiation generated from a mediumpressure UV lamp has been investigated for the degradation of the organiccontent of ROC (van Aken et al., 2010). They performed a batch study onthe degradation of a very high organic content ROC (DOC of 95 mg/L) atozone flow rate of 4 g/hr and reported a fairly low DOC reduction of 19%,but the process improved the biodegradability of the treated concentrate asindicated by a 13% increase in the BOD/COD ratio. The O3/UV treatmentremoved 70% of the adsorbable organic halogens (AOX).

Combination of ozone with other AOPs has been investigated by Zhouet al. (2011a). For raw concentrate, combination of other AOPs with O3

(UVA/O3, US/O3, UVA/H2O2/O3, and US/H2O2/O3, UVA/TiO2/O3) did notshow any significant improvement in DOC reduction than exhibited by O3

alone, suggesting that the selective oxidation by molecular O3 proved moreefficient than the nonselective HO• oxidation which is consistent with thefindings of Nakada et al. (2007). A greater reduction in aromatic content(SUVA) of raw ROC was observed with O3-based AOPs compared with otherAOPs (UVA/H2O2, UVA/TiO2, US/H2O2, US/UVA/TiO2). Although the O3-based AOPs exhibited better DOC reduction than non-O3-based AOPs, theorganic content in treated ROC was not susceptible to further mineralizationemphasizing the role of nonselective HO• oxidation for improving the overalltreatment performance. Ozone in combination with H2O2 (0.7 mol H2O2 permol of O3 dosage) was investigated by Westerhoff et al. (2009) who reporteda DOC reduction of 75%. However the dosage of O3 used to achieve thisremoval was very high (1,000 mg/L). The effect of simultaneous addition ofH2O2 (0.228 mol H2O2 per mol O3 after 10 min of ozonation) was studiedby van Geluwe et al. (2011a) in terms of decrease of optical density ofNF concentrate. They reported a marginally enhanced reduction of 80%compared with 73% when ozone was used alone after 40 min.

Thus it can be concluded that ozonation is promising for the oxidationof the organic content of ROC, particularly as pretreatment, due to its abil-ity to convert large organic compounds into smaller organic compounds.The formation of oxygenated saturated bonds, particularly carboxylic acids,hinders the extensive degradation of organic content by ozone. Completemineralization is generally impossible with ozone alone, and an appropri-ate combination of ozone with another AOP and a biological process as

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

214 M. Umar et al.

posttreatment can improve the overall performance. Research on the oxi-dation of individual organic species and their rate constants is needed tounderstand the degradation mechanism and to address the posttreatmentissues related to ozone toxicity as discussed in Section 10.

4.2.4. ELECTROCHEMICAL OXIDATION

Electrochemical treatment appears to be a promising approach to treat ROCdue to indirect bulk oxidation through the generation of hypochlorite (elec-trogeneration of active chlorines) (Eqs. (25) and (26)), and simultaneousoxidation of total ammonium nitrogen (TAN) and recalcitrant organic com-pounds (Perez et al., 2010; Zhou et al., 2011c).

2Cl− → Cl2 + 2e− (25)

Cl2 + H2O → HClO + H+ + Cl− (26)

Electrochemical oxidation can be easily controlled, is robust and simple,offers in situ generation of oxidants and does not involve the use of chemi-cals. The high salinity of ROC means that it has high electrical conductivitywhich leads to reduced ohmic resistances and improved cost-effectivenessby reducing the energy demand (Bagastyo et al., 2011b).

Several electrode types (BDD, Ti/SnO2-Sb, Ti/PbO2, Ti/Pt-IrO2, Ti/IrO2-Ta2O5, Ti/RuO2-IrO2) have been investigated for the treatment of ROC,however, boron-doped diamond (BDD) electrodes have been proven tobe promising due to their potential for generating oxygen evolution interme-diates, mainly HO• (Panizza et al., 2008). Anodic degradation of pollutants isprincipally due to the generation of electrogenerated HO• (Chen et al., 2003;Zhou et al., 2011b) as expressed by Eq. (27):

BDD + 2H2O → BDD(HO•) + 2H+ + 2e− (27)

The effectiveness of different electrochemical processes for the treatmentof ROC is given in Table 4. Using BDD electrodes, van Hege et al. (2002)reported >80% decolorization and 50% reduction of A254 for a charge (Q)input of 2 Ah/dm3 using ROC produced from a two-stage RO pilot installationprocessing mixed domestic and textile wastewater. The ROC had a highCOD (151–218 mg/L) and low biodegradability (average BOD28/COD of0.3). The authors reported that ClO− was the predominant active chlorinecomponent because: (1) the pH remained above 7 during the treatment; (2)the concentration of ClO− increased with electrolysis time, and (3) activechlorine was consumed during oxidation of organic compounds and wasnot detectable in the presence of substantial DOC and TAN levels. Theyconcluded that DOC and TAN were predominantly degraded through indirectoxidation by hypochlorite in the bulk solution.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 215

TABLE 4. Performance of electrochemical oxidation for treatment of ROC

COD DOC Contact Currentreduction reduction Energy time density Electrode(%) (%) (kW h/g COD) (min) (A/m2) type References

100 48, 59 — — — BDD Bagastyo et al. (2012a)60−74 41−51 — — — BDD/MM Bagastyo et al. (2012b)— ∼31, 9 0.35, 0.70 (DOC) 2−23.5 hr, 75 250 Ti/RuO2/IrO2 Radjenovic et al. (2011)100 — 0.158 90 250 BDD Zhou et al. (2011c)100 — 0.048 120 250 Ti/RuO2/IrO2 Zhou et al. (2011c)41 — 0.05 120 250 Ti/RuO2/Ta2O5 Zhou et al. (2011c)— 100 — 300 200 BDD Perez et al. (2010)— 30, 36 — 30 — Si/BDD Dialynas et al. (2008)56.2 — — — 200 BDD van Hege et al. (2004)11.7 — — — 200 RuO2 van Hege et al. (2004)35, 51 — 0.188, 0.083 — 167 BDD van Hege et al. (2002)

“—” indicates “not given”.

Using the same ROC sample, van Hege et al. (2004) trialed four anodes(PbO2, SnO2, RuO2, and BDD) to reduce the concentration of COD andTAN. Their preliminary studies indicated that PbO2 and SnO2 anodes werenot suitable for this particular ROC due to the scaling caused by Ca(OH)2 andMg(OH)2 on the anode surface, and corrosion of the PbO2 anodes due to adrastic rise in the pH in the one-compartment electrolytic cell. The excessiveprecipitation made the electrode surface unsuitable for further oxidation andresulted in a steady increase of cell potential. The RuO2 and BDD electrodeswere effective but increase in current density did not influence the removal ofCOD and TAN as they were the same for various current densities (100, 200,and 300 A/m2), which was consistent with their previous study. The increasein chlorine concentration indicated that the degradation was mainly throughindirect bulk oxidation by hypochlorite. However the authors mentionedthat the HO• produced on the BDD surface also played an important role inthe degradation of organic content. Therefore BDD achieved better removalof COD and TAN as indicated by higher rate of removals, that is, 74.5and 17.9 mg/A h (Q, 1.6 Ah/dm3) compared with RuO2 with correspondingremoval rates of 20 and 13.5 mg/A h (Q, 1.9 A h/dm3). Although the absolutevalues were not given for the abovementioned charge inputs, the COD andTAN reductions for Q of 1 A h/dm3 were 56% and 48% for BDD and 12% and42% for RuO2 anodes, respectively. Both the electrodes were highly efficientin the removal of color, and >90% removal was reported for a charge inputof 3 A h/dm3. However the energy consumption was quite high due to thelow conductivity (average 4.8 mS/cm) of the ROC.

A batch study was carried out by Dialynas et al. (2008) to investigate thepotential of BDD electrodes for treating municipal wastewater ROC undercurrent intensities of 3.6 and 17.8 A. A sharp decrease in DOC (20%) occurredin the first 3 min after which the rate of removal decreased significantly. Theremoval of DOC (initially 10 mg/L) was 30% and 36% at 3.6 and 17.8 A,

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

216 M. Umar et al.

respectively, after 30 min. Therefore approximately 5 times greater currentintensity gave only an additional 6% DOC removal, suggesting that sufficientimprovement in oxidation may take a much longer time, making the processuneconomic. However it is difficult to draw any definitive conclusion basedon this study as the authors studied only two current intensities and did notconsider the effect of other parameters, notably pH.

The application of BDD electrodes was investigated by Perez et al.(2010) in a batch study to treat ROC produced from a tertiary wastewatertreatment plant. They reported an increase in the reduction of COD and TANwith increasing current density, which was complete at a current density of200 A/m2 after 5 hr. Almost similar results were reported for treating landfillleachate for COD and ammonium with BDD electrodes (Cabeza et al., 2007).

Electrodialysis was used to treat raw and biologically (willow field)treated ROC generated from a domestic WWTP with a view to improvingthe water recovery from 75% to 95% by reducing the conductivity of thetreated ROC for groundwater infiltration (Zhang et al., 2011; Ghyselbrechtet al., 2012). They conducted pilot-scale experiments in batch and feed-and-bleed mode. Decarbonation of the ROC (pH was reduced to 5.5–6.5) wascarried out to reduce the scaling potential. A 75% reduction in conductivitywas reported in batch mode at a current density of 50 A/m2 and a flow rateof 300 L/h. The feed-and-bleed mode was investigated to assess the systemstability, and a flow rate of 75 L/hr and 45 L/hr was reported to give con-ductivity reduction of 75% at a current density of 30 and 40 A/m2 for rawand biologically pretreated ROC samples, respectively. The electrodialysis-treated effluent was deemed suitable for biological treatment in a WWTP.However, the reduction of TOC was only 10%, and the application of ozona-tion was suggested to improve the biodegradability of the effluent prior tothe biological treatment.

To avoid precipitation by insoluble species (Ca, K, Mg) at high pH,Bagastyo et al. (2011b) used a divided two-compartment electrochemical sys-tem in batch-scale laboratory experiments. They used 0.1 M HCl as catholyteto maintain low pH (2–3) and prevent scaling in the cathodic compartment.Among the five types of mixed metal oxide electrodes used, the Ti/Pt-IrO2

anode was the most efficient in terms of instantaneous current efficiency(ICECOD) of 50% for a Q of 0.55 A h/dm3, followed by Ti/SnO2-Sb (36%)and Ti/PbO2 (27%), whereas the lowest ICECOD of 18% was reported forTi/IrO2-Ta2O5 and Ti/RuO2-IrO2. The greater efficiency of the Ti/Pt-IrO2 an-ode was related to the enhanced generation of free available chlorine (FAC)(319 ± 22 mg/L) and greater chlorine-mediated electrolysis. The correspond-ing value of FAC for Ti/SnO2-Sb anode was 110 ± 7 mg/L whereas a lowconcentration of <25 mg/L was produced using the Ti/RuO2-IrO2, Ti/IrO2-Ta2O5, and Ti/PbO2 anodes. The removal of DON was fairly similar for theTi/Pt-IrO2 and Ti/SnO2-Sb anodes (57 ± 16% and 58 ± 3%, respectively),and complete removal of color was achieved for these anodes at a specific

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 217

electrical charge of 0.55 A h/dm3. The removal of ammonia was higher thanCOD (16%) for both Ti/Pt-IrO2 and Ti/SnO2-Sb anodes, verifying the dom-inance of the indirect oxidation mechanism (Bagastyo et al., 2011b). Theirfindings are consistent with those of Deng and Englehardt (2007) who re-ported that the removal of ammonia was high when indirect oxidation wasthe dominant mechanism. Due to the low anode surface area (anode surfacearea per active volume, 0.21/cm) used, the degradation of organic content bydirect oxidation was fairly low with 16% DOC removal obtained for Ti/Pt-IrO2 and Ti/SnO2-Sb anodes, and 9 ± 2% for the other three electrodes(Ti/PbO2, Ti/IrO2-Ta2O5, and Ti/RuO2-IrO2). Compared with the DOC re-moval, a high reduction in the SUVA was reported, that is, 40% and 43% forTi/Pt-IrO2 and Ti/SnO2-Sb anodes, respectively, representing the breakdownof the double bonds and the aromatic fraction of the organic matter that actsas a precursor to the formation of oxidation by-products.

Mixed metal oxide anodes such as RuO2/IrO2-coated titanium have beenwidely used in wastewater treatment (Chen et al., 2007) due to their highstability and cost effectiveness, particularly when compared with the BDDelectrodes. The use of RuO2/IrO2-coated titanium was investigated in batchand continuous modes for the treatment of ROC obtained from a mixtureof secondary treated effluents from four wastewater treatment plants (Rad-jenovic et al., 2011). The authors performed continuous experiments witha step-wise increase in the current density (over the range 1–250 A/m2)whereas the batch mode was studied at a current density of 250 A/m2. Thecontinuous mode was operated for 75 min at each current density, and thesamples were collected after 2, 4, 7, and 23.5 hr in batch mode. The effi-ciency of each mode was evaluated in terms of the decrease in DOC, SUVA,and trace organic contaminants (see Section 7). They also investigated theinfluence of various parameters such as operational mode, current density,and applied electrical charge. For a similar current density (250 A/m2) andcharge (0.77 A h/dm3), the reduction in DOC was 30.7 ± 1.8% and 8.9 ± 1.4%for batch and continuous modes, respectively, and the reduction in SUVAwas 28.7 ± 1.8% in both modes. The partial reduction of DOC and SUVA inboth batch and continuous systems indicated the accumulation of oxidationintermediates which were generated by the oxidative cleavage and openingof the aromatic moieties particularly, during the continuous mode. In batchmode, the addition of auxochrome substituents (−Cl, NH2Cl, −Br, −OH) tothe aromatic rings resulted in an initial increase of the SUVA value due tobetter reduction of the DOC compared with the A254 absorbance (Radjenovicet al., 2011).

In a batch study, Zhou et al. (2011c) reported that the BDD anodeshowed the highest oxygen and chlorine evolution potential, that is, 1.98 ±0.02 V and 1.28 ± 0.02 V, respectively. They compared the overall effec-tiveness and energy utilization of three different electrodes. BDD and twodimensionally stable anodes (DSAs) (Ti/IrO2-Ta2O5 and Ti/IrO2-RuO2) were

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

218 M. Umar et al.

used to treat high salinity ROC obtained from a steel plant. The COD removalwas linear for all three electrodes, and the efficiency of the anodes was inthe order of BDD > Ti/IrO2-RuO2 > Ti/IrO2-Ta2O5 for the same chargeinput. The low removal performance of Ti/IrO2-Ta2O5 can be attributed toless generation of FAC as mentioned by Bagastyo et al. (2011b). The BDDand Ti/IrO2-RuO2 anodes gave complete removal of COD after 1.5 and 2hr, respectively, at a current density of 250 A/m2 (Zhou et al., 2011c). ForBDD, both direct oxidation due to the high evolution potential of oxygenand indirect oxidation by chlorate led to the removal of organic compounds.However the removal of COD for Ti/IrO2-RuO2 was attributed mainly toindirect oxidation by active chlorine due to its low chlorine evolution poten-tial. The Ti/IrO2-Ta2O5 anode was the least effective for COD removal, anda decline in removal was reported for current density higher than 250 A/m2.Oxidation of organic content occurred through electrogeneration of activechlorine at low current density (<250 A/m2) for the Ti/IrO2-Ta2O5 anodeand at high potential it proceeded via side reaction of oxygen evolution thatled to a decline in the removal of COD (Zhou et al., 2011c). They reportedsimilar COD removal efficiencies for the same charge passed, regardless ofthe applied current density, which is consistent with the earlier findings ofvan Hege et al. (2002) as discussed earlier in this section.

The effect of chloride, nitrate, and sulfate ions using Ti/Pt-IrO2, Ti/SnO2-Sb, and Si/BDD anodes was investigated for the electrochemical oxidationof ROC (Bagastyo et al., 2013). The chloride concentration of the ROC wasreduced 10-fold by electrodialysis to enable electrooxidation with nitrateand sulfate ions as dominant ion mediators. The Coulombic efficiency ofCOD reduction (CECOD) was 15.9% for complete COD removal at a Q of3.3 Ah/dm3 using Si/BDD anode during chlorine-mediated (0.05 M NaCl,Cl− = 1,318 mg/L) electrooxidation of the electrodialyzed ROC (ROCED). TheTi/SnO2-Sb gave COD reduction of 93% (CECOD = 8.9%), whereas Ti/Pt-IrO2

(CECOD = 8.3%) gave the lowest COD reduction (87%) for Q = 5.6 A h/dm3.Under similar conditions of specific charge (5.6 A h/dm3), the reduction ofDOC was lower than that of COD reductions with Si/BDD achieving a max-imum DOC reduction of 40% followed by 31% and 28% for Ti/SnO2-Sb andTi/Pt-IrO2, respectively. Similarly, the Si/BDD anode performed better thanTi/SnO2-Sb and Ti/Pt-IrO2 anodes in the presence of NO3

− or SO42− dur-

ing electrooxidation of ROCED with low chloride concentration (142 mg/L).Although the reduction in COD for Si/BDD anode was lower, that is, 60%and 74%, compared with the high chloride concentration mentioned above,the mineralization rate was higher, that is, 41% and 51% in the presence ofNO3

− or SO42−, respectively. The efficiency in terms of DOC and COD re-

ductions was in the order Si/BDD>Ti/SnO2-Sb > Ti/Pt-IrO2, the mixed metalanodes performing poorly under conditions of low compared with high chlo-ride concentration which was attributed to their dependence on HClO/ClO−

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 219

species for effective oxidation. The reduction in color was generally similarat the end of treatment for all studied conditions and anodes.

pH is one of the important factors that determines the type of oxidationdue to the formation of different oxidizing species at various pH values. Itis well known that free chlorine is predominant in the pH range 1–3, HClOat pH 5, and ClO− at pH > 8 (Gordon and Tachiyashiki, 1991). The effectof circumneutral (6–7) and acidic pH (1–2) on electrochemical oxidationof ROC using BDD anodes was studied by Bagastyo et al. (2012). Theyreported complete removal of COD at a specific electrical charge of 5.2 and6.6 A h/dm3, but comparatively low DOC removal of 48% and 59% at acidicand circumneutral pH, respectively. They attributed oxidation of the organiccontent of the ROC to enhanced participation of Cl2/HClO species underlow pH conditions whereas HO•-led oxidation appeared to be the dominantmechanism at circumneutral pH, which is in agreement with Zhou et al.(2011c) who studied the effect of pH using various electrodes includingBDD anodes. The BDD was more effective at the high original pH value(8.3) whereas the Ti/IrO2-RuO2 electrode was effective over a wide rangeof pH values due to its strong ability to form active chlorine. Contrary tothe situation for BDD and Ti/IrO2-RuO2 anodes, the removal of COD on theTi/IrO2-Ta2O5 electrode more likely occurred by anodic oxidation insteadof indirect oxidation as the lower pH value was found to be favorable dueto the generation of HO• on the anode. In general, the removal by BDDelectrodes was due to both direct and indirect oxidation reactions whereasTi/IrO2-RuO2 electrodes mainly functioned via indirect chlorine oxidationand Ti/IrO2-Ta2O5 electrodes via direct anodic oxidation.

The BDD electrode has been proven more efficient due to the produc-tion of active chlorine in addition to chlorate formation during electrolysis.The superior chlorine generation of the BDD electrode is related to its highselectivity towards the chlorine evolution reaction (Ferro et al., 2000). Al-though the generation of active chlorine species enables oxidation of theorganic component of the ROC, it also leads to the formation of significantconcentrations of chlorinated organic by-products such as trihalomethanesand haloacetic acids (HAAs). Furthermore, the BDD electrode is generallyconsidered expensive, and therefore several other electrodes have been in-vestigated. Limited information on the effect of operational parameters isavailable. Optimization of process variables such as current density, treat-ment time, cell design, and pH is critical to improve process efficiency,energy efficiency and to reduce the formation of hazardous halogenated by-products as discussed in detail in Section 10. The removal of micropollutantsusing electrochemical oxidation has been investigated, but the mechanismsof their removal under different operational parameters and at various con-centrations remain unclear and require further investigation. The formationof hazardous by-products and fouling of the electrodes are major drawbacks

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

220 M. Umar et al.

of electrochemical treatment which need to be addressed to improve theoperational performance and reduce the cost of the process.

4.2.5. CAPACITIVE DEIONIZATION

Capacitive deionization (CDI) is a low-pressure electrochemical process ca-pable of removing dissolved ions from ROC with high water recovery (Leeet al., 2009a). It is based on applying an electrical field to generate a potentialdifference between two electrodes which attract and subsequently removeions from the solution. A bench-scale study was performed for the treatmentof ROC by Ng et al. (2008) to increase the water recovery of a wastewa-ter reclamation process for NEWater (Singapore) to >90%. They used BACas a pretreatment to the CDI process and obtained removals of 78%, 91%,and 92% of TOC, TN, and electrical conductivity, respectively, and a waterrecovery of 90%. The removal efficiencies of anions and cations were >90%.

The treatment of BAC pretreated ROC was investigated by Tao et al.(2011) in a pilot-scale study. The concentration of TOC decreased from 31.1to <0.1 mg/L, and there was a high removal of ions (>90%) and TDS (>99%).The quality of the CDI-treated water followed by RO treatment was betterthan the NEWater with a recovery of up to 93%. The authors reported theeffect of various combinations of pretreatments (BAC + MF, BAC + MF +O3, BAC + MF + O3 + SBS) on the fouling of the CDI cell. The addition ofsodium bisulfite (SBS) increased the operation time from 112 to 292 hr beforecleaning of the CDI cell was needed. After 5 days of continuous operation,the authors trialed various cleaning solutions to restore the pressure, howeverit was not restored to the initial value as most of the organic content was notremoved by either acid or alkali cleaning.

The CDI process has been proven efficient for the removal of TOC, TDS,and ions from ROC pretreated by BAC + UF (Lee et al., 2009a). The removalof TDS and ions (i.e., Na+, K+, Mg2+, Ca2+, Cl−, NO3

−, SO42−, and PO4

3−)was high, that is, >87% and >88%, respectively, and these findings are inagreement with Ng et al. (2008). The removal of TOC and PO4

3− was 50–63%and 52–81%, respectively. A high organic content can significantly reduce theoperational time but its reduction by pretreatment can improve the removalperformance without substantial pressure buildup and/or drop in permeateflow rate. In addition to the low initial DOC after BAC + UF pretreatment,the adjustment of pH to 6.3–6.5 improved the CDI operation time by at leasttwice. The reduction in CDI fouling at low-feed pH is in accordance with thefindings of Kim et al. (2002). In another of their studies, Lee et al. (2009b)investigated the ozone–BAC process as pretreatment to the CDI treatment ofthe ROC samples used in the previous study. The process was able to attain>80% removal of inorganic anionic and cationic species. Compared with theBAC-UF-CDI process reported earlier, the integrated ozone-BAC-CDI processproduced better quality water with TOC removal of 96%. Given that the initialwater quality was fairly similar, the difference in the performance was likely

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 221

due to the increased biodegradability of the organic content of the ROC afterozonation and subsequent removal by BAC treatment.

The CDI process is an emerging technology and it appears to be promis-ing for treating ROC. Selection of an appropriate pretreatment method isimportant with regard to fouling of the CDI cell and final water quality. Thefouling can be reduced by pretreatment and reducing pH (Lee et al., 2009a),periodically switching the potential of the electrodes (Kerwick et al., 2005),and by applying a pulsed field to the electrodes (Perez-Roa et al., 2006).Organic fouling and calcium phosphate scaling can be reduced by usingozone and SBS (Tao et al., 2011). Because CDI is a low-pressure process,it generally has a lower energy requirement than its counterparts. For ex-ample, it requires three times less energy than the electrodialysis reversalprocess (AWWA, 1999). Other critical operational parameters which need tobe considered are the quality of feed, pH, and water recovery rate. The devel-opment of new electrode materials and improved process control strategiescan make CDI a competitive technology for the treatment of ROC.

4.2.6. FENTON OXIDATION

Although the Fenton process alone (Lak et al., 2012) or in combinationwith other AOPs (Hu et al., 2011) or biological processes (Badawy et al.,2009) has been well studied for the treatment of domestic and industrialwastewater, its application to the treatment of ROC is limited. The massratio of Fe2+ and H2O2 is an important parameter in the Fenton process interms of the cost and effectiveness of the treatment (Deng and Englehardt,2006), and it is well established that Fenton oxidation is most effective at lowpH. Few studies have been conducted to investigate the Fenton process forremoving the organic content and improving the biodegradability of ROC.Westerhoff et al. (2009) reported a DOC reduction of 50% using 10 mMeach of Fe2+ and H2O2 at pH 3. A comparison of O3/UV treatment with theFenton process (H2O2, 2.4 mM) was carried out by van Aken et al. (2010).They reported an average COD reduction of 43%, which was better thanO3/UV treatment (19%), whereas the reduction in AOX was fairly similar forboth the treatments (66–70%). However, the Fenton process was almost fourtimes less efficient for improving the biodegradability of the treated ROC.

A COD reduction of ∼23% at pH 4 was reported by Huang et al. (2011)during the treatment of ROC generated from a metal plating wastewater recy-cle system. Increasing the H2O2 dosage from 2.6 to 5 mM (Fe2+/H2O2 ratioof 0.8) increased the COD reduction to ∼52%, but no further increase inCOD reduction occurred for H2O2 dosage higher than 5 mM. Fenton oxida-tion followed by hydroxide precipitation (10% w/w NaOH) and coagulationwith the remaining Fe2+/Fe3+ gave Cu and Ni removal of 87.3% and 85.7%,respectively, at a precipitation pH of 8. Coupling the Fenton process with abiological-activated filter reduced the concentration of COD and target heavymetals to meet the local discharge standards.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

222 M. Umar et al.

One of the major drawbacks of the Fenton oxidation is the generationof a considerable amount of sludge which has to be further treated and thusincreases the cost of the treatment. More insights into possible options formanaging the sludge such as its recycling and/or reuse are needed to makethe process economic and competitive with other AOPs.

5. BIOLOGICAL PROCESSES

Most of the organic content present in ROC is bio-refractory (Zhou et al.,2011a), and biological processes are generally considered ineffective for itstreatment. The application of biological treatment, mainly using BAC eitheras a pre or posttreatment, has been reported in literature. The removal oforganic compounds occurs due to adsorption to the activated carbon andbiological degradation through extracellular enzymes excreted by the biofilmthat forms on the activated carbon (Yapsaklia and Cecenb, 2010).

The potential of willow fields for reducing the concentration of nutrients(nitrogen and phosphorus) and organic content (TOC and COD) of ROC wasinvestigated by Ghyselbrecht et al. (2012). Of the 10 willow species, theyselected eight based on a preliminary salt tolerance test. Using a willow testfield of 28.33 m2 and 500 L/hr of ROC, they reported a ∼20% reduction ofTOC and DOC whereas the reduction of total nitrogen and total phosphoruswas 32%. Considering the low reductions reported in this study, in additionto the relatively low conductivity (3–6 mS/cm) of the ROC compared withother studies, the application of such treatment is highly case-specific, andthe performance may vary depending on seasonal variations such as rainfalland the quality of ROC, particularly in terms of conductivity.

As noted in Section 4.2.5, pretreatment of ROC in BAC columns hasbeen investigated for the removal of organic content by Ng et al. (2008) andLee et al. (2009a). For an empty bed contact time (EBCT) of 40 min, thereduction of TOC and COD was 25.0 ± 5.9% and 39.6 ± 8.3%, respectively(Ng et al., 2008). They also reported 34% color reduction and 10% totalnitrogen removal, which was possibly due to the mineralization of organicnitrogen to ammonium–nitrogen under anaerobic conditions. Almost 90%of the organic content was removed by BAC when treating 770 L of ROC,however the normalized TOC of the BAC effluent increased sharply by aboutfive times during the treatment of this volume due to the exhaustion ofthe granular-activated adsorption (GAC) adsorption capacity. Therefore, thebiological activity of the BAC was not the predominant mechanism of organiccontent removal at the initial stage of the treatment. However, biologicalremoval was considered to be contributing significantly in the later stage,after the treatment of 770 L of ROC. The TOC plateaued after treating 2200 LROC, indicating a similar rate of biodegradation of the organic content andof adsorption by the activated carbon. The reduction in the concentration of

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 223

inorganic ions by the BAC column was negligible. For an EBCT of 60 min,Lee et al. (2009a) reported TOC and COD removals of 23.5 ± 6% and 23.7 ±5.5%, respectively. The BAC effluent was subsequently treated by submergedUF membranes which gave an additional 16% TOC removal correspondingto an overall efficiency of 39.9 ± 9% for the combined BAC-UF process. BAChas also been used as a posttreatment for ozonated ROC and in combinationwith ozone prior to CDI treatment as discussed in Sections 4.2.3 and 4.2.5,respectively.

Having already undergone secondary treatment, generally ROC is con-sidered to mainly comprise of organics refractory to biological treatment,therefore biological treatment is thought to be more effective as a post-treatment to AOPs. However where applicable, ROC can be returned to thebiological treatment stage although it may impose certain limitations. For ex-ample, the high concentration of organic compounds and salts may reducethe efficacy of biological treatment by inhibiting microbial growth and canincrease the energy requirements during the subsequent RO stage. The highconcentrations can also reduce the lifespan of the membrane. One way ofdealing this could be to dilute the ROC to an appropriate level with sec-ondary effluent prior to returning to the biotreatment stage. However, thismay lead to gradual increase in the concentration of salts. Although AOP-treated ROC can be returned to the secondary biotreatment stage, the issueof salt level may restrict this option.

6. OTHER METHODS OF ROC TREATMENT

Activated carbon adsorption is one of the most commonly used methodsfor the treatment of water and wastewater containing a wide variety of or-ganic and inorganic compounds. A couple of studies using activated carbonadsorption have been conducted for the treatment of ROC (Dialynas et al.,2008; Zhou et al., 2011a). Dialynas et al. (2008) used GAC and reported aninitial rapid drop in DOC followed by moderate removal due to slow diffu-sion of the organic matter in the GAC pores. Equilibrium was achieved after4 days with a high removal of DOC (91.3% at GAC dosage of 5 g/L). How-ever the treatment time reported (5 days) was too long for its application inpractice. Moreover, the use of GAC only transfers the pollutants to the adsor-bent, and further treatment is required to treat these adsorbed compounds.Zhou et al. (2011a) conducted equilibrium adsorption experiments with GACand powdered activated carbon (PAC). They illustrated that the Freundlichisotherm well described the equilibrium adsorption of DOC for both GACand PAC with removals of 88% and 95%, respectively, using a dosage of5 g/L. However it was noted that the remaining organic matter (hydrophilicorganic compounds of large molecular weight) was not removed, even atdosages higher than 5 g/L.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

224 M. Umar et al.

TABLE 5. ROC treatment by adsorption

Parameters and removal

Material DOC removal (%) Dosage (g/L) References

MIEX 24, 43 10, 15 mL/L Bagastyo et al. (2011a)MIEX 43 10 mL/L Comstock et al. (2011)GAC 88 5 Zhou et al. (2011a)PAC 95 5 Zhou et al. (2011a)GAC 91.3 5 Dialynas et al. (2008)

Adsorption by MIEX resin has been reported by Bagastyo et al. (2011a)for two ROC samples of different initial organic concentrations. For a similarcolor removal efficiency of ∼80% after 20 min contact time, a high optimumdosage (15 mL/L) of resin was established for one of the ROC samples (ROC2) with high initial DOC (62 ± 5 mg/L) compared with 10 mL/L for ROC 1with lower initial DOC concentration (42 ± 4 mg/L) (Table 5). The removalof COD under optimum conditions was around 28% (ROC 1) and 35% (ROC2), whereas the removal of DON was high (47%) for ROC 1 with lower initialconcentration (5.3 ± 0.6 mg/L) than ROC 2 (7.8 ± 0.3 mg/L) for which theremoval was around 26%.

Comstock et al. (2011) used a MIEX concentration of 20 mL/L and re-ported a DOC reduction of 51% whereas the reduction of A254 was compar-atively higher (results not shown), and the MIEX resin exhibited preferentialremoval of A254-absorbing organic matter relative to the DOC.

The treatment techniques discussed above have the potential to be usedfor the degradation/removal of the organic content of ROC, but it is difficultto draw a general conclusion as little work has been done on them. Moreresearch is therefore needed to establish the effectiveness of these processestaking into account the requirement of regenerating the adsorbents to reducecost and further treatment of the adsorbates, which may make the treatmenttrain more complex.

7. REMOVAL OF MICROPOLLUTANTS FROM ROC

Some studies have investigated the removal of micropollutants, and as seenin Table 6, most of the studied compounds are those that are generally ex-pected to be in high concentration in municipal wastewaters. Most of the mi-cropollutants mentioned in Table 6 were removed except where mentionedotherwise. Benner et al. (2008) investigated the degradation of pharmaceu-ticals during ozonation of ROC generated from the effluent of a WWTP.The reaction rates for ozone and HO• with four beta blockers (acebutolol,atenolol, metoprolol, and propranolol) were determined. At pH 7, acebutolol,atenolol, and metoprolol reacted with ozone with an apparent second-order

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 225

TABLE 6. Concentration of micropollutants in ROC

Compound Benner Westerhoff Perez Ben Abdelmelek Justo(μg/L) et al. (2008) et al. (2009) et al. (2010) et al. (2011) et al. (2013)

Atenolol 2.9 — 1.452, 2.779 2.634 1.028Atrazine — 0.006 — — —Caffeine — 0.015 0.0339, 0.05 0.708 —Carbamazepine 3.4 ± 0.2 1.281 — 0.134 1.038DEET — 0.873 — 0.766 —Diclofenac 1.5 ± 0.1 — — — 0.605Gemfibrozil — — 5.92, 9.87 6.98 —Ibuprofen 1.33 ± 0.07 — 7.50, 0.0212 — —Iomeprol 3.9 ± 0.5 — — 0.386 —Naproxen 0.98 ± 0.06 0.015 4.161, 9.223 1.416 1.080Metoprolol 0.88 ± 0.03 — — 0.47 —

“—” indicates “not given”.

rate constant of about 2 × 103 M−1 s−1, whereas for propranolol it was105 M−1 s−1. Huber et al. (2003) reported similar rates of reaction betweenozone and micropollutants with ozone-reactive moieties. Markedly high re-action rates were reported between HO• and selected compounds, that is,0.5–1 × 1010 M−1 s−1 (Benner et al., 2008). The reaction rates between HO•

and 24 micropollutants were reported to be between 108 and 1010 M−1 s−1

(Jin et al., 2012). An ozone dosage of 5 mg/L resulted in quantitative re-moval of propranolol in 0.8 s due to the high second-order rate constantfor its reaction with ozone. It was further suggested that the naphthalenemoiety of propranolol resulted in about two orders of magnitude increase inreactivity. An ozone dosage of 10 mg/L oxidized 70% of metoprolol in 1.2 s,implying that oxidation of metoprolol required high ozone dosage due to itslow reaction rate constant. Although bromate was formed, the concentration(35 μg/L) was lower than the discharge guideline of 3 mg/L for the aqueousecosystem (Hutchinson et al., 1997).

Westerhoff et al. (2009) investigated the removal of 16 pharmaceuticalsusing a batch mode Photo-Cat R© system which consists of a reactor, UV lamp(wavelength not given), and ceramic ultrafilters which are used to separateTiO2 from water. The ROC was diluted 3-fold with nanopure water prior totreatment due to the large volume (20 L) required to operate the system.The authors reported that all the compounds were below the detection limit(2 ng/L) after treatment.

Using BDD electrodes, Perez et al. (2010) investigated the removal of 10emerging pollutants (PPCPs and EDCs) at concentrations found in ROC. Theyreported markedly varying initial concentrations of the studied compounds(Table 6), which was attributed to the different sampling periods. Perez et al.(2010) reported that the removal efficiency of all the emerging contaminantswas in the range 90–98% after 60 min of electrooxidation with the exceptionof ibuprofen for which the removal was only 55% after 2 hr. The removal of

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

226 M. Umar et al.

the emerging contaminants was not affected by current density between 20and 100 A/m2 and this was attributed to their low concentration. Furthermore,the kinetics of oxidation were not influenced by the initial concentration ofthe micropollutants. This was expected considering their low concentrationand hence likely low mass transfer resistance offered by the diffusion ofpollutants from the bulk solution to the anode surface, which controls theoverall kinetics.

In their study of the removal of spiked trace organic contaminants(28 pharmaceuticals and pesticides) from ROC in batch and continuousmode, Radjenovic et al. (2011) reported that the removal of persistent pol-lutants showed marked variations. Twenty compounds completely disap-peared when a current density of ≥150 A/m2 (Q ≥ 461.5 A h/m3) and250 A/m2 was applied in continuous and batch mode, respectively. The com-pounds with high affinity towards FAC showed markedly different removalsfor the same applied charge. For example, diclofenac, acetaminophen, sul-fadiazine, norfloxacin, ranitidine, and lincomycin showed ≥88% removalwhereas trimethoprim, gemfibrozil, caffeine, diuron, and metoprolol exhib-ited <50% removal. Considering that the initial concentrations of these com-pounds were fairly similar, the low removals of some of the compoundswere more likely due to their low oxidation by FAC rather than low masstransfer coefficients. However, carbamazepine which has lower affinity to-wards chlorine was found to degrade more quickly, whereas enrofloxacin,which is known to react very slowly with FAC, disappeared completely.

The incomplete oxidation of some compounds by FAC and completedisappearance of others with low affinity towards chlorine suggest that otheroxidants in the bulk solution and/or surface reactions play a vital role (Rad-jenovic et al., 2011). The compounds recalcitrant to electrochemical oxida-tion were characterized by either the absence of nucleophilic substituents(ibuprofen, phenytoin, metolachlor, and DEET), or by the presence of elec-trophilic groups on the aromatic ring (atrazine, triclopyr, iopromide, and2,4-D). The authors investigated the effect of prolonged exposure to oxi-dation and, except for triclopyr, complete disappearance of all the studiedpesticides and pharmaceuticals was reported in batch mode for a suppliedcharge of 1450 A h/m3, implying higher charge requirements for the oxida-tion of recalcitrant contaminants.

The efficacy of HO• generated during the degradation of PPCPs in ROCwas investigated through selective generation of HO• under a N2O atmo-sphere by employing γ -irradiation (Ben Abdelmelek et al., 2011). They ana-lyzed ROC for 27 PPCPs of which nine compounds were not found, and theother 18 were present in concentrations of 0.1–7.9 μg/L. Those identifiedat higher concentrations were atenolol (2.6 μg/L), erythromycin (8 μg/L),gemfibrozil (7 μg/L), and naproxen (1.4 μg/L). These compounds havebeen reported to be inefficiently removed during biological treatment (West-erhoff et al., 2005; Kim et al., 2007). Ben Abdelmelek et al. (2011) determined

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 227

the absolute reaction rate constants of four pharmaceuticals (trimethoprim,naproxen, nalidixic acid, and venlafaxine) with HO• by competition kinet-ics. The reaction rates were in the range of 3–8.9 × 109 M−1 S−1, which arecomparable to those determined for other commonly reported compoundssuch as atenolol (7.05 × 109 M−1 S−1), caffeine (6.4 × 109 M−1 S−1), carba-mazepine (8.80 × 109 M−1 S−1), DEET (4.95 × 109 M−1 S−1), metoprolol (8.39× 109 M−1 S−1), and ofloxacin (7.66 × 109 M−1 S−1) (Huber et al., 2003; Songet al., 2008, 2009; Santoke et al., 2009; Wols and Hofman-Caris, 2012). A cor-relation between fluorescence excitation-emission matrix spectra and PPCPoxidation was obtained. The three peaks identified as UV humic-like peak(peak A), visible humic-like peak (peak C) and protein-like peak (peak T)decreased in intensity with increased dosage of the γ -radiation. The protein-like peak correlated well with the removal of PPCPs. For example, whenthe removal of PPCPs reached 80–100%, a corresponding decrease of 80%was noted for the protein-like peak compared with 40–50% reduction of theUV and visible humic-like peaks. Consequently the authors suggested mon-itoring the protein-like fluorescence of ROC as a rapid and an inexpensivemethod for the quantitative estimation of degradation of PPCPs, particularlyunder treatment plant conditions.

More recently, the removal of 11 pharmaceuticals from ROC generatedfrom a municipal WWTP using ozone and the UVC/H2O2 process was in-vestigated (Justo et al., 2013). Ozonation was carried out at various ozonedosages, that is, 0.14–2.78 mg O3/mg TOC. Atenolol and carbamazepinewere the most resistant to ozonation and showed lower degradation (<50%)under low O3 dosage conditions (≤0.41 mg O3/mg TOC). However, all thepharmaceuticals investigated were reported to disappear before the initialozone demand (IOD; initial ozone dosage needed to obtain a measurabledissolved ozone concentration) (3.27 mg O3/mg TOC) was met. Sulfamethox-azole, sulfamethazine, and naproxen showed good removals (>70%) even atlow concentration of ozone (0.14 mg O3/mg TOC), which was attributed tothe presence of electron-rich functional groups that are highly reactive withmolecular ozone (Reungoat et al., 2012). However, some other pharmaceu-ticals containing electron-rich functional groups (diclofenac, trimethoprim,and carbamazepine) were not effectively removed at low ozone dosage,contrary to the results reported by Reungoat et al. (2012) for WWTP ef-fluent. Justo et al. (2013) attributed their conflicting results to the differentcharacteristics of their ROC and so reactivity of the organic matter.

The compounds that needed higher ozone dosage for effective degra-dation (atenolol, carbamazepine, and diclofenac) disappeared more rapidlyduring UVC/H2O2 treatment at low H2O2 concentration (0.07 mg H2O2/mgTOC) (Justo et al., 2013). Trimethoprim and paroxetine exhibited the low-est degradation (<50%) during the UVC/H2O2 treatment at similar H2O2

concentration. However, the degradation of paroxetine increased with in-creasing H2O2 concentration, and all the pharmaceuticals disappeared at

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

228 M. Umar et al.

≤0.54 mg H2O2/mg TOC, except trimethoprim which was not completelydegraded even when the concentration was increased to 0.72 mg H2O2/mgTOC. This trend was unexpected as all the pharmaceuticals under investi-gation have comparable reaction rates (as mentioned above), and as such,similar removals have been reported in another study for water collectedfrom a tertiary wastewater facility (Rosario-Ortiz et al., 2010).

It is evident from the examples discussed above that not much attentionhas been paid to the characterization and removal of micropollutants in ROC.Only a few studies have investigated the removal of micropollutants ROCand determined their reaction rate constants with HO•. None of the stud-ies have reported the levels of several other important trace contaminantssuch as bisphenol A and nonylphenol. Considering that these pollutantsare at concentrations 3- to 4-fold higher than in the feed water, and thattheir release into the environment may impose toxicity at any level of thebiological hierarchy (Klavarioti et al., 2009), they pose greater risk to thereceiving ecosystem and therefore must be a focus in future studies. Thecriterion for the selection of target compounds appears to be their presencein high concentrations, rather than their potential harmful effects, which ispartly driven by the difficulty of analyzing the trace organics in low con-centrations. Another important observation is the conflicting results (ROCcf. other water/wastewater) obtained for some micropollutants, and moreimportantly, no experimental investigation has been undertaken to elucidatethis behavior.

8. AOPs AND BIODEGRADABILITY IMPROVEMENT

It is well established that chemical oxidation processes, including AOPs,produce BDOC or assimilable organic carbon (Volk and Lechevallier, 2002).Complete mineralization of the organic compounds of ROC is generally notpossible or cost effective, and several studies have investigated the use ofAOPs with a view to breaking down large molecular weight compoundsinto simpler forms that are more biodegradable and easily removed by sub-sequent biological treatment. For example, the cost of UVC/H2O2 treatmentis generally considered to be high but it can be reduced by minimizing theuse of H2O2 and irradiation time to achieve partial oxidation and henceenhanced biodegradability which can facilitate the implementation of down-stream biological treatment.

Westerhoff et al. (2009) investigated the increase of biodegradability (asBDOC) during UVC/TiO2 treatment and measured the production of fiveorganic acids. The raw sample had only 2 mg/L (5%) biodegradable DOCwhich increased to 12 mg/L (30%) after an energy consumption of 2 kWh/m3, representing a 6-fold increase in biodegradability. Further treatmentdecreased the BDOC concentration which was due to the mineralization

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 229

of the simple organic acids (acetate, propionate, formate, pyruvate, andoxalate) formed during the UV-mediated process. Compared with UVC/TiO2

alone treatment, the authors reported a 35% reduction of energy for a similarresidual DOC concentration of 10 mg/L by incorporating downstream BDOCassay.

Lee et al. (2009b) reported an increase in the biodegradability of ROCafter ozonation. The authors noted that the ratio of BOD5 to TOC increasedby 1.8- to 3.5-fold using 3–10 mg O3/L and contact time of 10–20 min. Theenhancement in biodegradability was attributed mainly to the breakdown oforganic compounds with molecular weight in the range of 10–100 kDa andgeneration of intermediates of less than 1 kDa. Similarly Zhou et al. (2011a)noted an increase in the biodegradability of raw ROC by 4- to 7-fold whentreated with O3-based AOPs. An even greater improvement (7- to 20-fold)occurred when the ROC was pretreated by coagulation prior to AOPs. Ozona-tion (1 L/min, 24 mg/min O3) as posttreatment of electrodialysis-treated ROCwas also reported to improve the biodegradability (BOD5/A254), that is, from42.3 to ∼109 and ∼116 using 576 mg/L ozone at pH 7.8 and 6.5, respectively(Zhou et al., 2011a).

Using a ROC generated from a municipal secondary effluent, Liu et al.(2011) determined the biodegradability after various treatments by using theBDOC method. The original biodegradability of the ROC increased from 11to 22% and 26% after 3 hr of UVC (UV fluence 1,392.12 kJ/m2) and VUV (UVfluence 1,926.72 kJ/m2) treatment, respectively. The addition of 2 mM H2O2

increased both the biodegradability and the overall DOC reduction, howeverfurther increase in H2O2 dosage to 4 mM led to a decrease in biodegradabilitywithout improving the DOC reduction. Fairly similar results were obtainedusing VUV/2 mM H2O2 treatment although the removal of DOC was a littlehigher than for the UVC/2 mM H2O2 process. In another study, Liu et al.(2012) evaluated the biodegradability of ROC using UVC/3 mM H2O2 treat-ment after various irradiation times. The biodegradability almost doubled(to 23%) after 10 min of treatment and reached a maximum value of 35%after 30 min of irradiation and then gradually decreased to 30% at 120 min.The increase in biodegradability after 30 min was correlated to the corre-sponding enhanced breakdown of the conjugated and fluorophoric com-pounds which were the major source for the production of biodegradableproducts.

Umar et al. (2013) reported a reduction in the BDOC after 10 min ofUVC/3 mM H2O2 treatment of a high salinity municipal wastewater ROC,which was followed by a gradual increase with increasing irradiation time.The initial reduction in BDOC was attributed to the mineralization of someof the biodegradable fraction of the DOC and little breakdown of the largemolecular weight organic compounds at that time. However the BDOC in-creased such that an almost 2-fold improvement (from 13% to 25%) in thebiodegradability was reported after 60 min of the treatment.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

230 M. Umar et al.

The foregoing studies demonstrate the effectiveness of various AOPs forenhancing the biodegradability of ROC. Most of the studies discussed abovehave used the BDOC assay to determine the biodegradability. Although theBDOC test is useful in determining the biodegradability, it may underesti-mate the proportion which can be removed when a biological treatmentprocess is utilized, as found by Buchanan et al. (2008) who used BAC totreat the effluent from a UV-based AOP drinking water treatment process.Investigation of the application of downstream biological treatment at pilotscale is therefore needed to determine its potential application at industrialscale.

9. EFFECT OF TREATMENT ON DIFFERENT MOLECULARSIZE COMPOUNDS

An understanding of the molecular size distribution of organic compounds inwater and wastewater is important to determine the potential for the removalof different pollutants by different treatment methods (Shon et al., 2006).Some studies have investigated the molecular weight distribution of ROCbefore and after treatment. The MW distribution of DOC in untreated ROCwas in the ranges of <1 kDa (47%) and 1–10 kDa (37%) (Zhou et al., 2011a).These results are consistent with the findings of Bagastyo et al. (2011a) whoreported that almost half of the organics contributing to the COD had MW<1 kDa. They also found that the dissolved organic nitrogen (DON) in twodifferent ROC samples was predominantly composed of organic compoundswith MW <1 kDa (Bagastyo et al., 2011a). According to Lee et al. (2009a),more than 80% of the organics of ROC consisted of compounds with MW<10 kDa. These results are in close agreement with the MW distribution ofDOC in secondary effluents (Li et al., 2006; Jarusutthirak and Amy, 2007).

AOPs have been found to effectively treat compounds over a broadrange of MW, however the removal of low MW compounds is generally low.The large MW compounds react faster as they tend to be more aromatic innature and may have a larger number of reaction sites available to react withHO• (Frimmel et al., 2000). Furthermore, the large MW compounds reactfaster as they have higher molar absorptivities than low MW compounds(Thomson et al., 2004). For example, an initial drop in DON was observedto be due to the degradation of high MW compounds, and the remaining or-ganics with MW <3 kDa showed low removal (10%) suggesting the inabilityof UV/H2O2 for removing recalcitrant low MW compounds in ROC (Bagastyoet al., 2011a). Similarly, Liu (2011) found that the removal of large MW com-pounds was high after UVC/3 mM H2O2 treatment in the first 15 min ofreaction but later shifted to low MW compounds (≤0.5 kDa). The reductionin the compounds of MW ∼ 1 kDa was 3% in the first 15 min but increasedto 85% after 120 min of irradiation, implying the need for long irradiation

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 231

time for effective breakdown of this fraction. Similar results were reported forUVC/TiO2 treatment with little decrease in compounds of MW <1 kDa (Zhouet al., 2011a). The authors reported a significant change in the MW distri-bution for the ROC treated with either UVC/TiO2 or coagulation-UVC/TiO2.Coagulation (FeCl3) followed by UVC/TiO2 led to increased reduction of thefraction with MW <1 kDa from 47% to 77%, demonstrating the contributionof coagulation.

The difference in the distribution of various size fractions of ROC be-fore and after different O3 dosages and contact time was reported by Leeet al. (2009b). They observed that the organic compounds with large MWwere converted to those with MW <1 kDa. It was most effective for thebreakdown of the compounds with MW 10–100 kDa as shown by their re-duction (36–72%) after 10 min, and increasing the contact time from 10 to20 min improved the breakdown of this fraction by another 20%. Howeverthe organic compounds with MW >100 kDa were the least removed. Thiscan be attributed to the surface trimming effect of ozone that led to breakageof some of the bonds on the surface but not complete disintegration of themolecules, thus keeping the bulk organic content intact (Jansen, 2005). Theorganic compounds with MW <1 kDa increased from 37% to 65% due tothe breakdown of the large MW fractions. Similar findings were reported byZhou et al. (2011a) who showed that an increase in the low MW (<1 kDa)fractions from 47% to 60% occurred after ozonation of raw ROC, with greaterbreakdown of fractions in the range of 10–100 kDa and least breakdown forthose with MW >100 kDa.

Coagulation by alum is known to be efficient for removing compoundslarger than 10 kDa (Shon et al., 2006). Bagastyo et al. (2011a) investigatedthe coagulation of ROC taken from two full-scale MF/RO plants. Alum suc-cessfully removed large MW chromophoric compounds, and most of theresidual compounds were less than 3 kDa, whereas FeCl3 targeted large- tomedium-size compounds such as humics, and the remaining organics weremainly in the lower MW range (Bagastyo et al., 2011a). Both COD and DOCwere removed in all size ranges, with the removal of large MW compoundsbeing favored. The highest residual fraction after alum and FeCl3 coagula-tion comprised molecules of 0.5–1 kDa. Zhou et al. (2011a) also reportedthe greater effectiveness of coagulation for removing hydrophobic organiccompounds of high MW (100 kDa).

MIEX adsorption was effective for removing a wide size range of com-pounds (Bagastyo et al., 2011a). Most of the COD and DOC reduction wasdue to the removal of the organic compounds with MW 5–10 kDa (50–75%)and ≤1–5 kDa (20–40%). The removal of color (∼50%) and DON (up to70%) occurred over all size ranges. Thus MIEX was effective for removingcolored and other organic fractions (>1 kDa) over a broad range of MW,which is consistent with the findings of others for drinking water treatment(Drikas et al., 2003; Boyer and Singer, 2006). The 5–10 kDa MW compounds

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

232 M. Umar et al.

were less targeted by the MIEX resin at dosages lower than the optimum(10–15 mL/L), therefore establishing the optimum dosage of resin is crit-ical for enhancing the process efficiency. However, MIEX targets anionicmolecules and so cannot remove all the organic compounds. Charge densityand SUVA have been suggested as useful parameters for predicting the ef-fectiveness of MIEX treatment for the removal of organic matter (Boyer et al.,2008).

10. FORMATION OF DBPs AND TOXICITY OF WATERAFTER TREATMENT

The formation of DBPs is one of the major concerns associated with theapplication of AOPs after chlorination. The generation of active chlorine canlead to the formation of DBPs for some AOPs such as electrochemical oxi-dation. The formation of total trihalomethanes (THMs), for example, CHCl3+ CHBrCl2 + CHBr2Cl + CHBr3, during electrochemical oxidation of ROCusing two circular electrodes (BDD on silicon anode and stainless steel cath-ode) was investigated by Perez et al. (2010) who observed increasing THMswith increasing current density. At a current density value of 200 A/m2, theconcentration of THMs formed was ∼200 μg/L, which is well above the limitin European (100 μg/L) and USEPA (80 μg/L) regulations for drinking water.The need for optimization of the process to minimize the operational timeand the current density to increase the process efficiency and reduce theformation of THMs was emphasized.

Bagastyo et al. (2011b) reported total HAAs formation even at low charge(0.17 A h/dm3). The formation of HAAs for the different electrode typesdecreased in the order Ti/SnO2-Sb > Ti/Pt-IrO2 > Ti/RuO2-IrO2 ≈ Ti/PbO2 >

Ti/IrO2-Ta2O5. The formation of HAAs increased with increasing electrolysistime and was between 0.6 and 2.7 mg/L after 22 hr for these electrodes.The high formation of HAAs for the Ti/Pt-IrO2 and Ti/SnO2-Sb anodes wasattributed to the enhanced formation of FAC at the surface of the anodesand in the bulk solution by HO• and other oxidants (Lei et al., 2007). Theformation of THMs increased with increasing specific electrical charge andat Q = 0.55 A h/dm3, the concentration of THMs varied between 82 and312 μg/L for the different anodes; the formation of THMs decreased in theorder Ti/Pt-IrO2 > Ti/RuO2-IrO2 > Ti/SnO2-Sb ≈ Ti/PbO2 > Ti/IrO2-Ta2O5.So the anodes that were most efficient led to high formation of oxidationby-products.

Bagastyo et al. (2012) studied the formation of total THMs and HAAs attwo pH values: acidic (pH 1–2) and circumneutral (pH 6–7) at specific elec-trical charge of 5.2 and 10.9 A h/dm3 using BDD anodes. The concentrationof total THMs and HAAs was reported to increase >3- and >18-fold for a Qvalue of 5.2 A h/dm3 at acidic pH, respectively. An even higher formation

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 233

of total THMs and HAAs was observed at circumneutral pH, that is, 13- and>24-fold, respectively, leading to corresponding concentrations of 13 and29 μM from initial values of 0.9 and 1.2 μM. Further increase in specificcharge led to degradation of these by-products at both pH values. They alsoinvestigated the influence of electrodialysis as a pretreatment of the ROC forreducing the formation of THMs and HAAs using Si/BDD, Ti/SnO2-Sb, andTi/Pt-IrO2 anodes (Bagastyo et al., 2013). A large decrease in the formationof total THMs and HAAs was reported using the Si/BDD anode when the Cl−

concentration was lowered from 1,320 mg/L (37.2 mM) to 142 mg/L (4 mM),that is, from 9.1 to 0.2 μM and 62.4 to 1.1 μM, respectively. A similar trendwas observed for the mixed metal oxide anodes.

Liu et al. (2012) determined the THM formation potential (THMFP) oftreated and untreated ROC and reported an increase in the THMFP from 1.22to 1.51 mg/L after 30 min of UVC/3 mM H2O2 treatment, however it reducedto the original value (1.22 mg/L) after 75 min. The application of biologicaltreatment to the treated ROC further reduced the THMFP to 0.86 mg/L.Therefore biological treatment can not only enable enhanced mineralization,but also reduce the potential for the formation of THMs during subsequentdisinfection of the treated water using chlorine.

As described in Sections 8 and 9, partial oxidation is generally usefulin terms of breaking down large MW compounds to those with low MW.Incomplete mineralization may result in the formation of intermediates thatmay be more toxic than the parent compounds (Rizzo, 2011). However, Liuet al. (2012) determined the toxicity of raw and UVC/3 mM H2O2 treatedROC and found that both raw and treated ROC showed no toxicity after50 and 75 min of exposure as measured by the Microtox test. Hence theconcentration of toxic by-products formed was considered to be insignificant.

The effect on toxicity (as measured by the Microtox test) of ROC treatedusing various AOPs with and without coagulation pretreatment was investi-gated by Zhou et al. (2011a) and Justo et al. (2013). Zhou et al. reported 60%inhibition of Vibrio fischeri using raw ROC whereas some reduction of toxic-ity was noted after coagulation and most of the AOPs without pretreatment.UVA/TiO2/O3 treatment led to the lowest ecotoxicity (25% inhibition) com-pared with other AOPs such as US and UVA/US/TiO2. The ROC pretreated bycoagulation followed by some AOPs (UVA/TiO2, UVC/TiO2, US/UVA/TiO2)exhibited substantially lower ecotoxicity, whereas O3-based AOPs exhibitedhigher toxicity with over 40% inhibition. Justo et al. (2013) reported con-trasting results with no inhibition of V. fischeri for raw and ozonated andUVC/H2O2 treated ROC. Furthermore, varying findings can be found in liter-ature in terms of the ecotoxicity of ozonation-treated ROC, with some studiesclaiming acute toxicity to aquatic invertebrates and fish (Petala et al., 2006;Stalter et al., 2010a, 2010b) and higher plants (Monarca et al., 2000; Stal-ter et al., 2010a), whereas others reported a reduction in the toxicity aftertreatment with ozone (Cao et al., 2009; Reungoat et al., 2010; Misık et al.,

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

234 M. Umar et al.

2011). There is a lack of information on the ozone dosage (g O3) consumedper g of organic content degradation (Hollender et al., 2009; Schaar et al.,2010; Stalter et al., 2010a), which is crucial for comparing different studies toreach a reliable conclusion in terms of DBPs formation and ecotoxicity aftertreatment. It is difficult to make a direct comparison of the different studiesdue to different characteristics of ROC, the types of test organism used andhence different toxicity response, and level of ozone dosage and treatmenttime.

Electrochemical oxidation of ROC with a RuO2/IrO2-coated titaniumanode was evaluated for its baseline toxicity to V. fischeri (the Microtoxtest organism) by Radjenovic et al. (2011). The acute toxicity results werereported in toxic equivalent (TEQ) values. The authors reported increasedtoxicity with increasing current density in both batch and continuous modes.The participation of chlorine and bromine in indirect oxidation was themain factor leading to the transformation of organic compounds to theirhalogenated derivatives. The TEQ increased from 4.3 ± 0.1 mg/L to 151 ±4.9 mg/L at 250 A/m2 in batch mode whereas it reached 209.1 ± 14.4 mg/Lin continuous mode at the same current density.

Limited studies have been conducted on the ecotoxicity of ROC, andmore work is required using samples with different characteristics, and morecomprehensive toxicity evaluation such as cytotoxicity, genotoxicity, and es-trogenicity is needed to study the impact of various treatments. It appearsthat the formation of oxidation by-products can be reduced by choosing anappropriate pretreatment and AOP, and optimizing the conditions. Basedon the existing studies, AOPs such as UVC/H2O2, ozonation, and electro-chemical oxidation can increase the toxicity of the ROC, and therefore theformation of toxic by-products (particularly halogenated DBPs) has to beconsidered before discharge or reuse of the treated water.

11. COMPARATIVE ENERGY CONSUMPTION FOR ROCTREATMENT PROCESSES

A comparison of energy requirements reported for various treatments is givenin Table 7. Dialynas et al. (2008) performed a comparative analysis of energyconsumption for photocatalysis, electrochemical oxidation (BDD electrodes),and sonolysis of ROC. They concluded that photocatalysis was the mostefficient with lowest energy consumption (2.6 kW h/g DOC) followed byelectrochemical oxidation (4.6 kW h/g DOC), whereas sonolysis was themost expensive (220 kW h/g DOC) based on the first 30 min of reactiontime.

The estimated power consumption for the treatment of 100 m3/d of ROCwas 0.85 kW h/m3 for a pilot-scale CDI process with BAC as pretreatmentstep (Lee et al., 2009a). The CDI process is an attractive alternative for the

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 235

TABLE 7. Comparison of energy requirements for various treatments

Process/Material Energy (kW h/g)a References

UVC/H2O2 12.5 (DOC) Liu et al. (2012)Ti/IrO2-RuO2 0.350, 0.704 (DOC) Radjenovic et al. (2011)BDD 0.158 (COD) Zhou et al. (2011c)Ti/IrO2-RuO2 0.048 (COD) Zhou et al. (2011c)BDD 0.059 (COD) Perez et al. (2010)UVC/TiO2 0.24 (DOC) Westerhoff et al. (2009)Sonolysis 220 (DOC) Dialynas et al. (2008)BDD 4.6 (DOC) Dialynas et al. (2008)UV/TiO2 2.6 (DOC) Dialynas et al. (2008)BDD 0.078 (COD) van Hege et al. (2002, 2004)

aEnergy consumed per gram of COD or DOC, as per brackets.

treatment of ROC due to the significant reduction in the cost of CDI modules(CTDM module manufacturing costs which were US$75,000 per module in1997 decreased to US $1000 per module in 2004) (Welgemoed and Schutte2005).

Westerhoff et al. (2009) calculated the energy requirement for a rep-resentative ROC sample for a final target DOC concentration of 10 mg/L;the energy required for UV/TiO2 treatment was 9.6 kW h/m3 (0.24 kW h/gDOC), which was estimated to cost US $0.95/m3. They emphasized that theenergy requirements could be reduced by up to 30% (0.15 kW h/g DOC)to achieve similar DOC reduction if a low UV dosage was applied followedby biological treatment with a sand filter. According to electrical energy perreaction order (EE/O) analysis, the authors concluded that the UV/TiO2 pro-cess had equal or lower EE/O than O3/H2O2 treatment; however, the authorsdid not take into account the requirements for TiO2 handling and associatedenergy requirements.

A comparison of the energy consumption for BDD and DSA (Ti/IrO2-RuO2, Ti/IrO2-Ta2O5) anodes showed that the BDD anode had the highestenergy requirement (0.158 kW h/g COD) while Ti/IrO2-RuO2 had the low-est (0.048 kW h/g COD) at a current density of 83.3 A/m2 (Zhou et al.,2011c). Further increase in the current to 50 mA substantially increased theenergy requirements for BDD (0.203 kW h/g COD) and Ti/IrO2-Ta2O5 an-odes (0.130 kW h/g COD). Energy consumed per g of DOC reduction usingTi/IrO2-RuO2 in batch and continuous mode was calculated by Radjenovicet al. (2011) to be 0.350 and 0.704 kW h, respectively, at a current density of250 A/m2. Although the batch system had lower energy requirements, thelow throughput could result in a cost higher than that estimated by the au-thors. Perez et al. (2010) also calculated the energy consumption of electro-chemical treatment using BDD electrodes and obtained 0.059 kW h/g COD(initial COD 125 mg/L) at applied current density of 50 A/m2. In anotherstudy, the specific energy consumption per g COD (initial COD 218 mg/L)and TAN removal for ROC was 0.078 and 0.428 kW h, respectively, using

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

236 M. Umar et al.

BDD electrode at applied current density of 200 A/m2 (van Hege et al., 2002,2004). Although the initial COD values and current densities were different,the energy consumption per g COD was comparable in the studies con-ducted by van Hege et al. (2002, 2004) and Perez et al. (2010). However vanHege et al. (2002, 2004) reported much higher energy consumption (0.18kW h/g COD) for the second sample with lower initial COD (171 mg/L)concentration for a charge input of 1.6 A h/dm3. Although the authors didnot provide an explanation for why the energy requirement was higher forthe second sample, the lower chloride concentration of this sample, that is,595 cf. 804 mg/L for first the ROC sample, may have negatively impactedthe oxidation as the authors considered that electrogenerated hypochloriteswere the main oxidants when using the BDD electrode. Similarly the energyrequirement for the removal of TAN was greater for the second ROC sample(0.782 kW h/g) than for the first sample (0.428 kW h/g).

The EE/O for UVC/H2O2 treatment of ROC was estimated to be 7.6and 12.5 kW h/g DOC at 10 and 120 min reaction time, respectively (Liuet al., 2012). A substantial decrease in EE/O was noted for the UVC/3 mMH2O2 process followed by biological treatment (as BDOC), and the energyrequired for 30 min of reaction (initial DOC 25.7 mg/L) was calculated to beapproximately 2.4 kW h/g DOC for a DOC reduction of 63%. However, toachieve an additional 20% DOC removal (to 83%), the energy consumptionfor the extended reaction period of 120 min was 5.3 kW h/g DOC, repre-senting a significant increase in the overall energy requirement. Given thatboth Westerhoff et al. (2009) and Liu et al. (2012) calculated the cost for asimilar final DOC residual of 10 mg/L, the UVC/TiO2 process appears to bemore energy and thus cost-effective, but the difference in initial DOC levelsused in these studies (40 and 25.7 mg/L, respectively) and the difference inthe nature of organic content along with the TiO2 handling issue, have to beconsidered when making direct cost comparisons.

It is difficult to make a direct energy consumption comparison for thedifferent treatment processes due to different initial concentrations and typesof organics present, and different treatment conditions used. A general costcomparison shows that photocatalysis, CDI, and UV-based AOPs are compet-itive processes for the treatment of ROC. Ti/IrO2-RuO2 anodes appear to bethe most economical material while BDD electrodes are still considered to beexpensive, and to establish which electrode works best is also a major issue.The cost of treatment is a function of target level, characteristics, and initialconcentration of pollutants which vary significantly. It is therefore difficult togive a particular number, however, in general terms an EE/O value ≤10 kWh/m3 is regarded as an economically acceptable power requirement (Parsons,2004). The energy requirements of AOPs, as mentioned earlier in this review,can be reduced by incorporating appropriate pre and posttreatment. Furtherresearch on the optimization of electrochemical processes, including opti-mization of cell design, can substantially reduce the overall treatment cost.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 237

12. CONCLUSIONS

Increasing application of RO-based processes in water treatment and wastew-ater reclamation require sustainable management of the resultant concentratestreams due to the toxic and recalcitrant nature of some of the organics in theROC. Considering the complexity and hazard potential of ROC constituents,AOPs are perhaps the most promising treatments due to their effective re-moval and detoxification of the organic compounds. High energy consump-tion is one of the barriers to the use of oxidative treatments such as UV/H2O2.The oxidation efficiency of UV-based AOPs can be further improved by se-lecting an appropriate wavelength of UV irradiation, and optimization of UVand H2O2 doses. The UVA/TiO2 system has been proven efficient in thedegradation of organic compounds at lab scale, and some critical researchgaps related to optimizing the band gap energy of the catalyst, the separationof the TiO2 from the treated water and its recycling need to be addressedfor potential large-scale applications.

Ozonation is a promising process for the oxidation of the organic con-tent of ROC and improving its biodegradability. Electrochemical oxidation isone of the most commonly investigated treatments for ROC but little informa-tion on the optimization of current density, treatment time, cell design, andpH is available. However the formation of hazardous by-products, particu-larly halogenated organic compounds, is one of the main concerns associ-ated with the oxidative treatments (especially electrochemical oxidation), andthere is conflicting information available on the toxicity of the treated water,particularly for ozonation. Coagulation and biological treatment as pre andposttreatment, respectively, can significantly improve the reduction in DOClevels and ecotoxicity. Therefore, AOPs would be a reasonable approach toimprove the biodegradability of ROC making downstream biological treat-ment feasible since they generate biodegradable organic compounds suchas organic acids. Adopting this approach would also make energy intensiveAOPs more economically feasible by reducing the required energy inputs.

The long-term environmental and health implications of ROC manage-ment have yet to be fully understood, and the presence of emerging pol-lutants such as EDCs and PPCPs adds to the complexity of the concentratematrix. Further research is required on the quantification of emerging con-taminants in ROC which entails improvements in existing technologies forbetter and easier analysis at low concentrations, that is, at concentrationscomparable to their levels found in the environment. More comprehensiveevaluation of toxicity of the treated ROC should be carried out to understandthe environmental risks. Research on the detailed kinetics and modeling ofAOPs-mediated degradation of the organic compounds, including EDCs andother micropollutants and their interactions with the other components ofROC such as halide ions, is required to understand the degradation mecha-nisms, toxicity potential, and DBPs formation. It is envisaged that effective

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

238 M. Umar et al.

and economically viable ROC treatment can significantly improve the sus-tainability of the RO-based wastewater treatment processes.

REFERENCES

Agustina, T. E., Ang, H. M., and Vareek, V. K. (2005). A review of synergistic effect ofphotocatalysis and ozonation on wastewater treatment. Journal of Photochem-istry and Photobiology C: Photochemistry Reviews 6, 264–273.

Ahmed, M., Shayya, W. H., Hoey, D., Mahendran, A., Morris, R., and Al-Handaly,J. (2000). Use of evaporation ponds for brine disposal in desalination plants.Desalination 130, 155.

Ahmed, S. (2003). Probing the extent of [O2] depletion and [Cl−] production si-multaneously at an illuminated TiO2 surface employing simple electrochem-ical methods. Journal of Photochemistry and Photobiology A: Chemistry; 154,229–234.

Al Momani, F., Sand, C., and Esplugas, S. (2004). A comparative study of the ad-vanced oxidation of 2,4-dichlorophenol. Journal of Hazardous Materials 107,123–129.

Al-Rifai, J. H., Gabelish, C. L., and Schafer, A. I. (2007). Occurrence of pharmaceuti-cally active and non-steroidal estrogenic compounds in three different wastew-ater recycling schemes in Australia. Chemosphere 69, 803–815.

Al Turki, A. A., Tadkaew, N., Mcdonald, J. A., Khan, S. J., Price, W. E., and Nghiem,L. D. (2010). Combining MBR and NF/RO membrane filtration for the removal oftrace organics in indirect potable water reuse applications. Journal of MembraneScience 365, 206–215.

Alvares, A. B. C., Diaper, C., and Parsons, S. A. (2001). Partial oxidation by ozone toremove recalcitrance from wastewaters—A review. Environmental Technology22, 409–427.

AWWA. (1999). American Water Works Association (AWWA) manual, reverse osmo-sis and nanofiltration, Denver, CO, USA: American Water Works Association.

Badawy, M. I., Wahaab, R. A., and El-Kalliny, A. S. (2009). Fenton-biologicaltreatment processes for the removal of some pharmaceuticals from industrialwastewater. Journal of Hazardous Materials 167, 567–574.

Bagastyo, A. Y., Batstone, D. J., Kristiana, I., Gernjak, W., Joll, C., and Radjenovic,J. (2012). Electrochemical oxidation of reverse osmosis concentrate on boron-doped diamond anodes at circumneutral and acidic pH. Water Research 46,6104–6112.

Bagastyo, A. Y., Batstone, D. J., Rabaey, K., and Radjenovic, J. (2013). Electro-chemical oxidation of electrodialysed reverse osmosis concentrate on Ti/Pt-IrO2, Ti/SnO2-Sb and boron-doped diamond electrodes. Water Research 47,242–250.

Bagastyo, A. Y., Keller, J., Poussade, Y., and Batstone, D. J. (2011a). Characterisa-tion and removal of recalcitrants in reverse osmosis concentrates from waterreclamation plants. Water Research 45, 2415–2427.

Bagastyo, A. Y., Radjenovic, J., Mu, Y., Rozendal, R. A., Batstone, D. J., andRabaey, K. (2011b). Electrochemical oxidation of reverse osmosis concentrate

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 239

on mixed metal oxide (MMO) titanium coated electrodes. Water Research 45,4951–4959.

Baird, N.C. (1997). Free radical reactions in aqueous solutions: examples fromadvanced oxidation processes for wastewater and from the chemistry in airbornewater droplets. Journal of Chemical Education 74, 817–819.

Baxendale, J. H., and Willson, J. A. (1957). Photolysis of hydrogen peroxide at highlight intensities. Transactions of the Faraday Society 53, 344–356.

Beltran, F. J. (2004). Ozone reaction kinetics for water and wastewater treatmentsystems. Boca Raton, FL: Lewis Publications.

Ben Abdelmelek, S. B., Greaves, J., Ishida, K. P., Cooper, W. J., and Song, W.(2011). Removal of pharmaceutical and personal care products from reverseosmosis retentate using advanced oxidation processes. Environmental Scienceand Technology 45, 3665–3671.

Benner, J., Salhib, E., Ternesa, T., and von Gunten, U. (2008). Ozonation of reverseosmosis concentrate: Kinetics and efficiency of beta blocker oxidation. WaterResearch 42, 3003–3012.

Bertanza, G., Pedrazzani, R., Zambarda, V., Dal Grande, M., Icarelli, F., and Baldas-sarre, L. (2010). Removal of endocrine disrupting compounds from wastewatertreatment plant effluents by means of advanced oxidation. Water Science andTechnology 61, 1663–1671.

Boroski, M., Rodrigues, A. C., Garacia, J. C., Sampaio, L. C., Nozaki, J., and Hioka, N.(2009). Combined electrocoagulation and TiO2 photoassisted treatment appliedto wastewater effluents from pharmaceutical and cosmetic industries. Journalof Hazardous Materials 162, 448–454.

Boyer, T. H., and Singer, P. C. (2006). A pilot-scale evaluation of magnetic ion ex-change treatment for removal of natural organic material and inorganic anions.Water Research 40, 2865–2876.

Boyer, T. H., Singer, P., and Aiken, G. R. (2008). Removal of dissolved organic matterby anion exchange: Effect of dissolved organic matter properties. EnvironmentalScience and Technology 42, 7431–7437.

Broseus, R., Vincent, S., Aboulfadl, K., Daneshvar, A., Sauve, S., Barbeau, B.,and Prevost, M. (2009). Ozone oxidation of pharmaceuticals, endocrine dis-ruptors and pesticides during drinking water treatment. Water Research 43,4707–4717.

Buchanan, W., Roddick, F., and Porter, N. (2008). Removal of VUV pre-treatednatural organic matter by biologically activated carbon columns. Water Research42, 3335–3342.

Buxton, G. V., Greenstock, C. L., Helman, W. P., and Ross, A. B. (1988). Criticalreview of rate constants for reactions of hydrated electrons, hydrogen atomsand hydroxyl radicals (•OH/•O−) in aqueous solution. Journal of Physical andChemical Reference Data 17, 513–886.

Cabeza, A., Urtiaga, A., Rivero, M.-J., and Ortiz, I. (2007). Ammonium removalfrom landfill leachate by anodic oxidation. Journal of Hazardous Materials 144,715–719.

Cao, N., Yang, M., Zhang, Y., Hu, J., Ike, M., Hirotsuji, J., Matsui, H., Inoue, D., andSei, K. (2009). Evaluation of wastewater reclamation technologies based on invitro and in vivo bioassays. Science of the Total Environment 407, 1588–1597.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

240 M. Umar et al.

Chelme-Ayala, P., Smith, D. W., and El-Din, M. G. (2009). Membrane concentratemanagement options: A comprehensive critical review. Canadian Journal ofCivil Engineering 36, 1107–1119.

Chen, J., Shi, H., and Lu, J. (2007). Electrochemical treatment of ammonia in wastew-ater by RuO2–IrO2–TiO2/Ti electrodes. Journal of Applied Electrochemistry 37,1137–1144.

Chen, X., Chen, G., and Yue, P. L. (2003). Anodic oxidation of dyes at novel Ti/B-diamond electrodes. Chemical Engineering Science 58, 995–1001.

Comerton, A. M., Andrews, R. C., and Bagley, D. M. (2005). Evaluation of an MBR-ROsystem to produce high quality reuse water: Microbial control, DBP formationand nitrate. Water Research 39, 3982–3990.

Comstock, S. E. H., Boyer, T. H., and Graf, K. C. (2011). Treatment of nanofiltra-tion and reverse osmosis concentrates: Comparison of precipitative softening,coagulation, and ion exchange. Water Research 45, 4855–4865.

Davis, J. R., and Koop, K. (2006). Eutrophication in Australian rivers, reservoirs andestuaries—A southern hemisphere perspective on the science and its implica-tions. Hydrobiologia 559, 23–76.

Deng, Y., and Englehardt, J. D. (2006). Treatment of landfill leachate by the Fentonprocess. Water Research 40, 3683–3694.

Deng, Y., and Englehardt, J. D. (2007). Electrochemical oxidation for landfill leachatetreatment. Waste Management 27, 380–388.

Dialynas, E., Mantzavinos, D., and Diamadopoulos, E. (2008). Advanced treatmentof the reverse osmosis concentrate produced during reclamation of municipalwastewater. Water Research 42, 4603–4608.

Drikas, M., Chow, C. W. K., and Cook, D. (2003). The impact of recalcitrant organiccharacter on disinfection stability, trihalomethane formation and bacterial re-growth: An evaluation of magnetic ion exchange resin (MIEX R©) and alum coag-ulation. Journal of Water Supply: Research and Technology-AQUA 52, 475–487.

Duan, J., Graham, N. J. D., and Wilson, F. (2003). Coagulation of humic acid by ferricchloride in saline (marine) water conditions. Water Science and Technology 47,41–48.

Duan, J., and Gregory, J. (2003). Coagulation by hydrolyzing metal salts. Advancesin Colloid and Interface Science 100–102, 475–502.

Duan, J., Wang, J., Graham, N., and Wilson, F. (2002). Coagulation of humic acid byaluminium sulphate in saline water conditions. Desalination 150, 1–14.

Dwyer, J., and Lant, P. (2008). Biodegradability of DOC and DON for UV/H2O2

pre-treated melanoidin based wastewater. Biochemical Engineering Journal 42,47–54.

Edzwald, J. K., and Haarhoff, J. (2011). Seawater pre-treatment for reverse osmosis:Chemistry, contaminants, and coagulation. Water Research 45, 5428–5440.

EPA. (2003). Use of reclaimed water: Guidelines for environmental management.Southbank: EPA Victoria.

Ferro, S., Battisti, A. D., Duo, I., Comninellis, C., Haenni, W., and Perret, A. (2000).Chlorine evolution at highly boron-doped diamond electrodes. Journal of theElectrochemical Society 147, 2614–2619.

Foussereau, X., Paranjape, S., and Reardon, R. (2003). The current status of the useof membranes for wastewater treatment. Alexandria, VA: WEFTEC Proceedings,January, 1, 2003.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 241

Frimmel, F. H., Hesse, S., and Kleiser, G. (2000). Characterisation and control indrinking water. In S. E. Barrett, S. W. Krasner, and G. L. Amy (Eds.), Naturalorganic matter and disinfection byproducts. ACS symposium series 761 (pp.84–95). Washington, DC: American Chemical Society.

Fritzmann, C., Lowenberg, J., Wintgens, T., and Melin, T. (2007). State-of-the-art ofreverse osmosis desalination. Desalination 216, 1–76.

Fujishima, A., Rao, T. N., and Tryk, D. A. (2000). Titanium dioxide photocatalysis.Journal of Photochemistry and Photobiology C: Photochemistry Reviews 1, 1–21.

Ghyselbrecht, K., Van Houtte, E., Pinoy, L., and Verbauwhede, J. (2012). Treatment ofRO concentrate by means of a combination of a willow field and electrodialysis.Resources Conservation and Recycling 65, 116–123.

Gogate, P. R., and Pandit, A. B. (2004). A review of imperative technologies forwastewater treatment I: Oxidation technologies at ambient conditions. Advancesin Environmental Research 8, 501–551.

Gordon, G., and Tachiyashiki, S. (1991). Kinetics and mechanism of formation ofchlorate ion from the hypochlorous acid/chlorite ion reaction at pH 6–10. En-vironmental Science and Technology 25, 468–474.

Hermosilla, D., Cortijo, M., and Huang, C. P. (2009). Optimizing the treatment oflandfill leachate by conventional Fenton and photo-Fenton processes. Scienceof the Total Environment 407, 3473–3481.

Hollender, J., Zimmermann, S. G., Koepke, S., Krauss, M., Mcardell, C. S., Ort,C., Singer, H., von Gunten, U., and Siegrist, H. (2009). Elimination of organicmicropollutants in a municipal wastewater treatment plant upgraded with afull-scale post-ozonation followed by sand filtration. Environmental Scienceand Technology 43, 7862–7869.

Hu, X., Wang, X., Ban, Y., and Ren, B. (2011). A comparative study of UV–Fenton,UV–H2O2 and Fenton reaction treatment of landfill leachate. EnvironmentalTechnology 32, 945–951.

Huang, R. M., He, J. Y., Zhao, J., Luo, Q., and Huang, C. M. (2011).Fenton-biological treatment of reverse osmosis membrane concentrate froma metal plating wastewater recycle system. Environmental Technology 32,515–522.

Huber, M. M., Canonica, S., Park, G.-Y., and von Gunten, U. (2003). Oxidation ofpharmaceuticals during ozonation and advanced oxidation processes. Environ-mental Science and Technology 37, 1016–1024.

Huber, M. M., Gobel, A., Joss, A., Hermann, N., Loffler, D., McArdell, C. S., Ried,A., Siegrist, H., Ternes, T. A., and von Gunten, U. (2005). Oxidation of phar-maceuticals during ozonation of municipal wastewater effluents: A pilot study.Environmental Science and Technology 39, 4290–4299.

Hutchinson, T. H., Hutchings, H. J., and Moore, K. W. (1997). A review of the effectsof bromate on aquatic organisms and toxicity of bromate to oyster (Crassostreagigas) embryos. Ecotoxicol. Environ. Saf. 38 (3), 238–243.

Jansen, R. H. S. (2005). Ozonation of humic substances in a membrane contac-tor mass transfer, product characterization and biodegradability (PhD thesis).University of Twente, The Netherlands.

Jarusutthirak, C., and Amy, G. (2007). Understanding soluble microbial products(SMP) as a component of effluent organic matter (EfOM). Water Research 41,2787–2793.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

242 M. Umar et al.

Jin, X., Peldszus, S., and Huck, P. M. (2012). Reaction kinetics of selected microp-ollutants in ozonation and advanced oxidation processes. Water Research 46,6519–6530.

Jordahl, J. (2006). Beneficial and nontraditional uses of concentrate. Alexandria, VA:Water Reuse Foundation.

Justo, O., Gonzalez, J., Acena, S., Perez, D., Barcelo, C., Sans, S., and Esplugas, S.(2013). Pharmaceuticals and organic pollution mitigation in reclamation osmo-sis brines by UV/H2O2 and ozone. Journal of Hazardous Materials 263, 268–274

Kerwick, M. I., Reddy, S. M., Chamberlain, A. H. L., and Holt, D. M. (2005). Electro-chemical disinfection, an environmentally acceptable method of drinking waterdisinfection? Electrochimica Acta 50, 5270–5277.

Khan, S. J., Murchland, D., Rhodes, M., and Waite, T. D. (2009). Management of con-centrated waste streams from high-pressure membrane water treatment systems.Critical Reviews in Environmental Science and Technology 39, 367–415.

Kim, D. H., Moon, S. H., and Cho, J. W. (2002). Investigation of the adsorptionand transport of natural organic matter (NOM) in ion-exchange membranes.Desalination 151, 11–20.

Kim, S. D., Cho, J., Kim, I. S., Vanderford, B. J., and Snyder, S. A. (2007). Occurrenceand removal of pharmaceuticals and endocrine disruptors in South Koreansurface, drinking, and waste waters. Water Research 41, 1013–1021.

Klavarioti, M., Mantzavinos, D., and Kassinos, D. (2009). Removal of residual pharma-ceuticals from aqueous systems by advanced oxidation processes. EnvironmentInternational 35, 402–417.

Kumar, M., Badruzzman, M., Adham, S., and Oppenheimer, J. (2007). Beneficialphosphate recovery from reverse osmosis concentrate of an integrated mem-brane system using polymeric ligand exchanger (PLE). Water Research 41,2211–2219.

Lak, M. G., Sabour, M. R., Amiri, A., and Rabbani, O. (2012). Application of quadraticregression model for Fenton treatment of municipal landfill leachate. WasteManagement 32, 1895–1902.

Lee, L. Y., Ng, H. Y., Ong, S. L., Tao, G., Kekre, K., Viswanath, B., Lay, W., andSeah, H. (2009a). Integrated pretreatment with capacitive deionization for re-verse osmosis reject recovery from water reclamation plant. Water Research 43,4769–4777.

Lee, L. Y., Ng, H. Y., Ong, S. L., Hu, J. Y., Tao, G., Kekre, K., Viswanath, B., Lay,W., and Seah, H. (2009b). Ozone-biological activated carbon as a pretreatmentprocess for reverse osmosis brine treatment and recovery. Water Research 43,3948–3955.

Legrini, O., Oliveros, E., and Braun, A. M. (1993). Photochemical processes for watertreatment. Chemical Reviews 93, 671–698.

Lei, Y., Shen, Z., Huang, R., and Wang, W. (2007). Treatment of landfill leachateby combined aged-refuse bioreactor and electrooxidation. Water Research 41,2417–2426.

Li, L., Zhu, W., Zhang, P., Zhang, Z., Wu, H., and Han, W. (2006). Comparisonof AC/O3–BAC and O3–BAC processes for removing organic pollutants in sec-ondary effluent. Chemosphere 62, 1514–1522.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 243

Liao, C. H., Kang, S. F., and Wu, F. A. (2001). Hydroxyl radical scavenging roleof chloride and bicarbonate ions in the H2O2/UV process. Chemosphere 44,1193–1200.

Liu, K. (2011). Treatment of reverse osmosis concentrate produced from municipaleffluent by UV/H2O2 processes (M Eng. thesis). School of Civil, Environmentaland Chemical Engineering, RMIT University, Melbourne, Australia.

Liu, K., Roddick, F. A., and Fan, L. (2011). Potential of UV/H2O2 oxidation of mu-nicipal reverse osmosis concentrates. Water Science and Technology 63, 2605–2611.

Liu, K., Roddick, F. A., and Fan, L. (2012). Impact of salinity and pH on the UVC/H2O2

treatment of reverse osmosis concentrate produced from municipal wastewaterreclamation. Water Research 46, 3229–3239.

Lopez, A., Pagano, M., Volpe, A., and Di Pinto, A. C. (2004). Fenton’s pre-treatmentof mature landfill leachate. Chemosphere 54, 1005–1010.

Madaeni, S. S., Fane, A. G., and Grohmann, G. S. (1995). Virus removal from waterand wastewater using membranes. Journal of Membrane Science 102, 65–75.

Meerganz von Medeazza, G. L. (2005). “Direct” and socially-induced environmentalimpacts of desalination. Desalination 185, 57–70.

Mickley, M. (2001). Membrane concentrate disposal: Practices and regulation, desali-nation and water purification research and development program. Washington,DC: U. S. Department of the Interior Bureau of Reclamation.

Mickley, M. (2008). Review of concentrate management options. Boulder, CO: Mick-ley and Associates.

Miranda-Garcıa, N., Suarez, S., Sanchez, B., Coronado, J. M., Malato, S., and Mal-donado, M. I. (2011). Photocatalytic degradation of emerging contaminants inmunicipal wastewater treatment plant effluents using immobilized TiO2 in asolar pilot plant. Applied Catalysis B: Environmental 103, 294–301.

Misık, M., Knasmueller, S., Ferk, F., Cichna-Markl, M., Grummt, T., Schaar, H., andKreuzinger, N. (2011). Impact of ozonation on the genotoxic activity of tertiarytreated municipal wastewater. Water Research 45, 3681–3691.

Monarca, S., Feretti, D., Collivignarelli, C., Guzzella, L., Zerbini, I., Bertanza, G., andPedrazzani, R. (2000). The influence of different disinfectants on mutagenicityand toxicity of urban wastewater. Water Research 34, 4261–4269.

Mosteo, R., Miguel, N., Maria, P. O., and Ovelloeiro, J. L. (2010). Effect of advancedoxidation processes on nonylphenol removal with respect to chlorination indrinking water treatment. Water Science and Technology 10, 51–57.

Motwani, P., Vyas, R. K., Maheshwari, M., and Vyas, S. (2011). Removal of sul-famethoxazole from wastewater by adsorption and photolysis. Nature Environ-ment and Pollution Technology 10, 51–58.

Nakada, N., Shinohara, H., Murata, A., Kiri, K., Managaki, S., Sato, N., and Takada,H. (2007). Removal of selected pharmaceuticals and personal care prod-ucts (PPCPs) and endocrine-disrupting chemicals (EDCs) during sand filtra-tion and ozonation at a municipal sewage treatment plant. Water Research 41,4373–4382.

Ng, H. Y., Lee, L. Y., Ong, S. L., Tao, G., Viswanath, B., Kekre, K., Lay, W., andSeah, H. (2008). Treatment of RO brine–towards sustainable water reclamationpractice. Water Science and Technology 58, 931–936.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

244 M. Umar et al.

Okely, P., Antenucci, J. P., Imberger, J., and Martin, C. L. (2007). Field investigationsinto the impacts of the Perth seawater desalination plant discharge on Cock-burn sound. Nedlands, WA: Centre for Water Research, University of WesternAustralia.

Oppenlander, T. (2003). Photochemical purification of water and air. Weinheim:Wiley-VCH.

Oppenlander, T., Walddorfer, C., Burgbacher, J., Kiermeier, M., Lachner, K., andWeinschrott, H. (2005). Improved vacuum-UV (VUV)-initiated photomineraliza-tion of organic compounds in water with a xenon excimer flow-through pho-toreactor (lamp, 172 nm) containing an axially centered ceramic oxygenator.Chemosphere 60, 302–309.

Panizza, M., Kapalka, A., and Comninellis, C. (2008). Oxidation of organic pollutantson BDD anodes using modulated current electrolysis. Electrochimica Acta 53,2289–2295.

Parsons, S. (2004). Advanced oxidation process for water and wastewater treatment.London: IWA Publishing.

Pearce, J. K. (2008). UF/MF pre-treatment to RO in seawater and wastewater reuseapplications: A comparison of energy costs. Desalination 222, 66–73.

Perez-Roa, R. E., Tompkins, D. T., Paulose, M., Grmies, C. A., Anderson, M. A.,and Noguera, D. R. (2006). Effects of localised, low-voltage pulsed electricfields on the development and inhibition of Pseudomonas aeruginosa biofilms.Biofouling 22, 383–390.

Perez, G., Fernandez-Alba, A. R., Urtiaga, A. M., and Ortiz, I. (2010). Electro-oxidationof reverse osmosis concentrates generated in tertiary water treatment. WaterResearch 44, 2763–2772.

Petala, M., Samaras, P., Zouboulis, A., Kungolos, A., and Sakellaropoulos, G. (2006).Ecotoxicological properties of wastewater treated using tertiary methods. Envi-ronmental Toxicology 21, 417–424.

Primo, O., Rivero, M. J., and Ortoz, I. (2008). Photo-Fenton process as an efficientalternative to the treatment of landfill leachates. Journal of Hazardous Materials153, 834–842.

Radjenovic, J., Bagastyo, A., Rozendal, R. A., Mu, Y., Keller, J., and Rabaey, K.(2011). Electrochemical oxidation of trace organic contaminants in reverse os-mosis concentrate using RuO2/IrO2-coated titanium anodes. Water Research 45,1579–1586.

Raffin, M., Germain, E., and Judd, S. (2013). Wastewater polishing using membranetechnology: A review of existing installations. Environmental Technology 34,617–627.

Ray, A. K., and Beenackers, A. A. C. M. (1998). Development of a new photocatalyticreactor for water purification. Catalysis Today 40, 73–83.

Reungoat, J., Escher, B. I., Macova, M., Argaud, F. X., Gernjak, W., and Keller,J. (2012). Ozonation and biological activated carbon filtration of wastewatertreatment plant effluents. Water Research 46, 863–872.

Reungoat, J., Macova, M., Escher, B. I., Carswell, S., Mueller, J. F., and Keller, J.(2010). Removal of micropollutants and reduction of biological activity in a fullscale reclamation plant using ozonation and activated carbon filtration. WaterResearch 44, 625–637.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 245

Rizzo, L. (2011). Bioassays as a tool for evaluating advanced oxidation processes inwater and wastewater treatment. Water Research 45, 4311–4340.

Rosario-Ortiz, F. L., Wert, E. C., and Snyder, S. A. (2010). Evaluation of UV/H2O2

treatment for the oxidation of pharmaceuticals in wastewater. Water Research44, 1440–1448.

Santoke, H., Song, W., Cooper, W. J., Greaves, J., and Miller, G. E. (2009). Free-radical-induced oxidative and reductive degradation of Fluoroquinolone phar-maceuticals: Kinetic studies and degradation mechanism. Journal of PhysicalChemistry A 113, 7846–7851.

Saritha, P., Raj, D. S. S., Aparna, C., Laxmi, P. N. V., Himabindu, V., and Anjaneyulu,Y. (2009). Degradative oxidation of 2,4,6 trichlorophenol using advanced ox-idation processes—A comparative study. Water, Air, and Soil Pollution 200,169–179.

Schaar, H., Clara, M., Gans, O., and Kreuzinger, N. (2010). Micropollutant removalduring biological wastewater treatment and a subsequent ozonation step. Envi-ronmental Pollution 158, 1399–1404.

Shon, H. K., Vigneswaran, S., and Snyder, S. A. (2006). Effluent Organic Matter(EfOM) in wastewater: Constituents, effects, and treatment. Critical Reviews inEnvironmental Science and Technology 36, 327–374.

Snyder, S. A. (2008). Occurrence, treatment, and toxicological relevance of EDCsand pharmaceuticals in water. Ozone: Science and Engineering 30, 65–69.

Snyder, S. A., Adham, S., Redding, A. M., Cannon, F. S., Decarolis, J., Oppenheimer,J., Wert, E. C., and Yoon, Y. (2007). Role of membranes and activated carbonin the removal of endocrine disruptors and pharmaceuticals. Desalination 202,156–181.

Song, W., Cooper, W. J., Mezyk, S. P., Greaves, J., and Peake, B. M. (2008). Freeradical destruction of β-blockers in aqueous solution. Environmental Scienceand Technology 42, 1256–1261.

Song, W., Cooper, W. J., Peake, B. M., Mezyk, S. P., Nickelsen, M. G., and O’Shea,K. E. (2009). Free-radical-induced oxidative and reductive degradation of N ,N -diethyl-m-toluamide (DEET): Kinetic studies and degradation pathway. WaterResearch 43, 635–642.

Staehelin, J., and Hoigne, J. (1985). Decomposition of ozone in water in the presenceof organic solutes acting as promoters and inhibitors of radical chain reactions.Environmental Science and Technology 19, 1206–1213.

Stalter, D., Magdeburg, A., and Oehlmann, J. (2010a). Comparative toxicity assess-ment of ozone and activated carbon treated sewage effluents using an in vivotest battery. Water Research 44, 2610–2620.

Stalter, D., Magdeburg, A., Weil, M., Knacker, T., and Oehlmann, J. (2010b).Toxication or detoxication? In vivo toxicity assessment of ozonation as ad-vanced wastewater treatment with the rainbow trout. Water Research 44, 439–448.

Stephenson, T., Judd, S., Jefferson, B., and Brindle, K. (2000). Membrane bioreactorsfor wastewater treatment. London, UK: IWA Publishing.

Tambo, N., and Kamei, T. (1978). Treatability evaluation of general organic matter.Matrix conception and its application for a regional water and waste watersystem. Water Research 12, 931–950.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

246 M. Umar et al.

Tao, G., Viswanath, B., Kekre, K., Lee, L. Y., Ng, H. Y., Ong, S. L., and Seah, H.(2011). RO brine treatment and recovery by biological activated carbon andcapacitive deionization process. Water Science Technology 64, 77–82.

These, A., and Reemtsma, T. (2005). Structure-dependent reactivity of low molecularweight fulvic acid molecules during ozonation. Environmental Science andTechnology 39, 8382–8387.

Thomson, J., Parkinson, A., and Roddick, F. A. (2004). Depolymerization of chro-mophoric natural organic matter. Environmental Science and Technology 38,3360–3369.

UCLA. (2011). First demonstration of reverse osmosis [Online]. Retrieved July17, 2011, from http://www.engineer.ucla.edu/explore/history/major-research-highlights/first-demonstration-of-reverse-osmosis

Umar, M., Aziz, H. A., and Yusoff, M. S. (2010). Trends in the use of Fenton,electro-Fenton and photo-Fenton for the treatment of landfill leachate. WasteManagement 30, 2113–2121.

Umar, M., Roddick, F. A., and Fan, L. (2013). Assessing the potential of a UV-basedAOP for treating high-salinity municipal wastewater reverse osmosis concen-trate. Water Science and Technology 68, 1994–1999.

USBR. (2003). Desalination and water purification technology roadmap—A reportof the executive committee. Desalination and Water Purification Research andDevelopment Program Report #95.

U.S. EPA. (2004). Guidelines for water reuse, EPA/625/R-04/108, September 2004.Washington, DC: U.S. Environmental Protection Agency and U.S. Agency forInternational Development.

van Aken, P., van Eyck, K., Luyten, J., Degreve, J., and Liers, S. (2010). Advancedtreatment of the reverse osmosis concentrate by the Fenton and O3/UV oxi-dation processes. In J. Drahos (Ed.), International congress of chemical andprocess engineering (CHISA). 19 Prague, Czech Republic: CHISA.

van der Bruggen, B., Lejon, L., and Vandecasteele, C. (2003). Reuse, treatment, anddischarge of the concentrate of pressure driven membrane processes. Environ-mental Science and Technology 37, 3733–3738.

van Geluwe, S., Braeken, L., and van der Bruggen, B. (2011b). Ozone oxidation forthe alleviation of membrane fouling by natural organic matter: A review. WaterResearch 45, 3551–3570.

van Geluwe, S., Vinckier, C., Braeken, L., and van der Bruggen, B. (2011a). Ozoneoxidation of nanofiltration concentrates alleviates membrane fouling in drinkingwater industry. Journal of Membrane Science 378, 128–137.

van Hege, K., Verhaege, M., and Verstraete, W. (2002). Indirect electrochemicaloxidation of reverse osmosis membranes at boron-doped diamond electrodes.Electrochemistry Communications 4, 296–300.

van Hege, K., Verhaege, M., and Verstraete, W. (2004). Electro-oxidative abatementof low-salinity reverse osmosis membrane concentrates. Water Research 38,1550–1558.

van Voorthuizen, E. M., Ashbolt, N. J., and Schafer, A. I. (2001). Role of hydrophobicand electrostatic interactions for initial enteric virus retention by MF membranes.Journal of Membrane Science 194, 69–79.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

Advancements in the Treatment of Municipal Wastewater Reverse Osmosis Concentrate 247

Vargas, C. and Buchanan, A. (2011). Monitoring ecotoxicity and nutrients loadin the reverse osmosis concentrate from Bundamba Advanced Water Treat-ment Plant, Queensland Australia. Water Practice and Technology 6, doi:10.2166/wpt.2011.006

Vigneswaran, S., and Visvanathan, C. (1995). Water treatment processes: Simple op-tions. Boca Raton, FL: CRC Press.

Volk, C. J., and Lechevallier, M. W. (2002). Effects of conventional treatment onAOC and BDOC levels. Journal of American Water Works Association 94,112–123.

von Gunten, U. (2003). Ozonation of drinking water: Part I. Oxidation kinetics andproduct formation. Water Research 37, 1443–1467.

Voutchkov, N. (2005). Alternatives for ocean discharge of seawater desalinationplant concentrate. 20th annual Water Reuse symposium: Water reuse and de-salination: Mile high opportunities, Water Reuse Association, September 18–21,Denver, Colorado.

Walker, T., Roux, A., and Owens, E. (2007). Western Corridor Recycled WaterProject—The largest recycled water scheme in the southern hemisphere. InS. J. Khan, R. M. Stuetz, and J. M. Anderson (Eds.), Water reuse and recycling.Sydney: UNSW Publishing.

Wang, X., Wang, L., Liu, Y., and Duan, W. (2007). Ozonation pretreatment forultrafiltration of the secondary effluent. Journal of Membrane Science 287, 187–191.

Weeks, J. L., and Rabani, J. (1966). The pulse radiolysis of deaerated carbonatesolutions. I. Transient optical spectrum and mechanism. II. pK for OH radicals.Journal of Physical Chemistry 70, 2100–2104.

Welgemoed, T. J., and Schutte, C. F. (2005). Capacitive deionization technology : analternative desalination solution. Desalination 183, 237–340.

Wert, E. C., Rosario-Ortiz, F. L., and Snyder, S. A. (2009). Effect of ozone exposureon the oxidation of trace organic contaminants in wastewater. Water Research43, 1005–1014.

Westerhoff, P., Moon, H., Minakata, D., and Crittenden, J. (2009). Oxidation oforganics in retentates from reverse osmosis wastewater reuse facilities. WaterResearch 43, 3992–3998.

Westerhoff, P., Yoon, Y., Snyder, S. A., and Wert, E. C. (2005). Fate of endocrine-disruptor, pharmaceutical, and personal care product chemicals during simu-lated drinking water treatment processes. Environmental Science and Technol-ogy 39, 6649–6663.

Wols, B. A., and Hofman-Caris, C. H. M. (2012). Review of photochemical reac-tion constants of organic micropollutants required for UV advanced oxidationprocesses in water. Water Research 46, 2815–2827.

Yapsaklia, K., and Cecenb, F. (2010). Effect of type of granular activated carbon onDOC biodegradation in biological activated carbon filters. Process Biochemistry45, 355–362.

Zhang, Y., Ghyselbrecht, K., Meesschaert, B., Pinoy, L., and van der Bruggen, B.(2011) Electrodialysis on RO concentrate to improve water recovery in wastew-ater reclamation. Journal of Membrane Science 378, 101–110.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014

248 M. Umar et al.

Zhou, M., Liu, L., Jioa, Y., Wang, Q., and Tan, Q. (2011c). Treatment of high-salinityreverse osmosis concentrate by electrochemical oxidation on BDD and DSAelectrodes. Desalination 201, 201–206.

Zhou, M. H., Sarkka, H., and Sillanpaa, M. (2011b). A comparative experimentalstudy on methyl orange degradation by electrochemical oxidation on BDD andMMO electrodes. Seperation and Purification Technology 78, 290–297.

Zhou, T., Lim, T. T., Chin, S. S., and Fane, A. G. (2011a). Treatment of organ-ics in reverse osmosis concentrate from a municipal wastewater reclamationplant: Feasibility test of advanced oxidation processes with/without pretreat-ment. Chemical Engineering Journal 166, 932–939.

Ziylan, A., and Ince, N. H. (2011). The occurrence and fate of anti-inflammatory andanalgesic pharmaceuticals in sewage and fresh water: Treatability by conven-tional and non-conventional processes. Journal of Hazardous Materials 187,24–36.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

20:

44 1

0 N

ovem

ber

2014