coupling digestion in a pilot-scale uasb reactor and electrochemical oxidation over bdd anode to...

ADVANCED OXIDATION PROCESSES FOR ENVIRONMENTAL PROTECTION

Coupling digestion in a pilot-scale UASB reactorand electrochemical oxidation over BDD anode to treat dilutedcheese whey

Αthanasia Katsoni & Dionissios Mantzavinos &Evan Diamadopoulos

Received: 27 January 2014 /Accepted: 21 April 2014# Springer-Verlag Berlin Heidelberg 2014

Abstract The efficiency of the anaerobic treatment ofcheese whey (CW) at mesophilic conditions was investi-gated. In addition, the applicability of electrochemicaloxidation as an advanced post-treatment for the completeremoval of chemical oxygen demand (COD) from theanaerobically treated cheese whey was evaluated. Thediluted cheese whey, having a pH of 6.5 and a totalCOD of 6 g/L, was first treated in a 600-L, pilot-scaleup-flow anaerobic sludge blanket (UASB) reactor. TheUASB process, which was operated for 87 days atmesophilic conditions (32±2 °C) at a hydraulic retentiontime (HRT) of 3 days, led to a COD removal efficiencybetween 66 and 97 %, while the particulate matter of thewastewater was effectively removed by entrapment in thesludge blanket of the reactor. When the anaerobic reactoreffluent was post-treated over a boron-doped diamond(BDD) anode at 9 and 18 A and in the presence of NaClas the supporting electrolyte, complete removal of CODwas attained after 3–4 h of reaction. During electrochem-ical experiments, three groups of organochlorinated com-p o u n d s , n am e l y t r i h a l om e t h a n e s ( THM s ) ,haloacetonitriles (HANs), and haloketons (HKs), as wellas 1,2-dichloroethane (DCA) and chloropicrin were iden-tified as by-products of the process; these, alongside freechlorine, are thought to increase the matrix ecotoxicity toArtemia salina.

Keywords Agro-industrial effluent . Anaerobic process .

Co-digestion . Electrochemical oxidation . Integratedtreatment . Toxicity . Transformation by-products

Introduction

According to the Food and Agricultural Organization (FAO),cheese is one of the main agricultural products worldwide.The European Union dominates its production and consump-tion, followed by the USA. The dairy industry is one of themain sources of industrial effluent generation. These effluentshave different characteristics according to the product obtain-ed such as yogurt, cheese, butter, and various types of dessertsby means of different processes, such as pasteurization, coag-ulation, filtration, centrifugation, and chilling (Carvalho et al.2013; Rivas et al. 2010). Three are the main streams of cheesemanufacturing wastewater; cheese whey—CW (resultingfrom cheese production), secondary cheese whey—SCW(resulting from cottage cheese production), and cheese wheywastewater—CWW (wash-waters that also contain differentfractions of CW and/or SCW) (Prazeres et al. 2012).

Cheese whey is the most contaminated waste generated inthe production of cheese and its principal components arelactose, proteins (casein), and mineral salts. This waste typi-cally has a very high organic content (COD up to 100 g/L) andlow alkalinity content (0.5 g/L as CaCO3), and it is highlybiodegradable (99 %).

In cheese wastewater management, three different optionscan be considered, namely (i) valorization leading to therecovery of valuable compounds such as proteins and lactose,(ii) biological and/or physicochemical treatments, and (iii) acombination of the above. So far, the high cost of valorizationtechnologies makes the biological and/or physicochemicaltreatments more attractive for CW management. Processessuch as coagulation–flocculation, ozonation, Fenton

Responsible editor: Philippe Garrigues

Α. Katsoni : E. DiamadopoulosDepartment of Environmental Engineering, Technical Universityof Crete, Polytechneioupolis, 73100 Chania, Greece

D. Mantzavinos (*)Department of Chemical Engineering, University of Patras,Caratheodory 1, University Campus, 26504 Patras, Greecee-mail: [email protected]

Environ Sci Pollut ResDOI 10.1007/s11356-014-2960-2

oxidation, and electrochemical oxidation have been reportedfor CW treatment (Prazeres et al. 2012).

Biological technologies, and especially anaerobic diges-tion, are used for the treatment of agro-industrial wastewaters.Anaerobic digestion is typically preferred to aerobic processesdue to the lower cost, energy requirements, and sludge gener-ation, as well as the possibility of energy recovery in the formof biogas. However, disadvantages related with anaerobicdigestion is the long start-up periods required for the acclima-tization of bacteria and the post-treatment of the anaerobiceffluent in order to reach the water discharge standards (Latifet al. 2011).

Anaerobic high-rate digesters are popularly used in waste-water treatment; of them, the up-flow anaerobic sludge blan-ket (UASB) reactors, are by far, the most robust, high-rateanaerobic reactors for wastewater treatment; their main featureis the availability of granular or flocculent sludge, thus achiev-ing high COD removal efficiencies without the need of asupport material.

Cheese whey constitutes a difficult substrate to treat anaer-obically (especially in highly loaded reactors) due to its highorganic content and low bicarbonate alkalinity. Indeed, thehigh level of carbohydrates in it promotes the growth of acid-forming bacteria, but it has negative effects on methane-producing bacteria (Hanson 1982). Moreover, whey may con-tain an increased concentration of mineral salts, such as Na+,which might become detrimental to the anaerobic reactorefficient operation (Gelegenis et al. 2007). To cope with thesedifficulties, co-digestion of cheese whey with other wastes,industrial or even domestic wastewater, has been investigated(Gannoun et al. 2008; Minhalma et al. 2007). These studieshave shown that diluted CW is easier to treat as it has beenenhanced with nutrients and buffer capacity.

Treatment of industrial effluents by a combination of sep-aration, biological, and advanced oxidation processes is con-ceptually advantageous (Comninellis et al. 2008). Of thelatter, boron-doped diamond (BDD) electrochemical oxida-tion is an environmentally acceptable remediation technologysince it can achieve increased mineralization rates of theorganic pollutants, as well as due to its robustness, versatility,amenability to automation, and for its little or no need foraddition of chemicals (Anglada et al. 2009). However, theformation of undesirable oxidation by-products, such as chlo-rinated organic compounds, has been reported during theelectrochemical oxidation over BDD and other anodes(Anglada et al. 2011; Gotsi et al. 2005). For this reason,speciation of oxidation by-products, especially of chlorinatedorganics, as well as the identification of the main factors thataffect their formation, needs to be determined.

This study aims at treating CW by sequential anaerobicdigestion in a pilot-scale UASB reactor followed by an emerg-ing advanced oxidation technology, the BDD electrochemicaloxidation. Information regarding the electrochemical

oxidation of cheese whey is scarce in the literature. Anaerobicdigestion efficiency was assessed in terms of COD removaland biogas production, while BDD oxidation efficiency wasevaluated in terms of dissolved organic carbon (DOC), COD,and color removal; moreover, major organochlorinated by-products, as well as toxicity to Artemia salina, were alsodetermined.

Materials and methods

Cheese whey

Cheese whey was taken from a cheese factory located inChania, W. Crete, Greece. The factory operates over a periodof 8 months (between October and June) producing mainlythree different kinds of cheese with the total quantity of wheybeing about 500 t.

In all cases, the original effluent was diluted with primarydomestic sewage taken from the Technical University campuswastewater treatment plant to achieve initial concentrationsequal to or less than about 6 g/L COD and then fed to theUASB reactor. Characteristics of raw cheese whey, domesticsewage, and the diluted cheese whey used to feed the UASBreactor are shown in Table 1. The ratio of cheese whey toprimary domestic sewage ranged between 1/12 and 1/29.

At the start-up of the reactor, inoculum sludge was takenfrom the anaerobic digester of the municipal wastewater treat-ment plant of Chania, W. Crete, Greece. The total mass ofsludge added to the pilot-scale reactor was almost 20 kg; thus,the v/v ratio of sludge/wastewater was 3 %.

UASB experiments

A pilot-scale UASB reactor with internal diameter, totalheight, working volume, and total tank capacity of 40 cm,215 cm, 550 L, and 600 L, respectively, was employed. Allparts of the reactor were made of ANSI 316 stainless steel. Anoutlet weir was provided at the top, which was connected to anoutlet pipe to the effluent collection tank. The reactor wasequipped with five sampling ports. Biogas was collected fromthe headspace on the top of the reactor via a gas collectingsystem. The gas collecting and measuring system consisted ofa gas–solid–liquid separator, a gas collecting pipe, a glasswater trap used as a potential flame trap, and a wet-tip gasmeter. The UASB reactor is shown in Schematic 1.

At the start-up period, the bioreactor, seeded with theinoculum sludge, was filled with diluted CW at 5 g/L COD,and left to stabilize for almost 2 months, in order to allow thebacterial community to acclimatize. After the start-up period,the UASB system was operated in a daily-continuous modethrough pumping of fresh feed into the reactor with a peristal-tic pump, while effluent samples were collected daily. The

Environ Sci Pollut Res

hydraulic retention time (HRT) was 3 days. The reactor wasoperated at mesophilic conditions (32±2 °C) for 87 dayscirculating the hot water through the external reactor jacketwith an external circulator. After the acclimatization of seedculture, the dilution ratio was gradually decreased in order to

increase the organic concentration (from about 0.75 to 6 gCOD/L) and study the effects of different feed strengths on thereactor performance. Part of the sludge generated during an-aerobic treatment was periodically removed from the reactor.Mixed liquor volatile suspended solids (MLVSS) and mixed

Table 1 Characteristics of rawcheese whey, domestic wastewa-ter, and diluted wastewater usedas the feed of UASB reactor

ND not determined

Parameters Raw cheese whey Domestic sewage Diluted cheese whey forUASB reactor feed

pH 5–5.5 7–7.5 6.8–8.2

Alkalinity (as mgCaCO3/L) 450–500 ND 1,000–2,000

COD (g/L) 75–100 0.6–0.65 0.75–6

BOD5 (g/L) 23–28 0.25–0.3 0.25–1.5

DOC (g/L) 8–10 0.05–0.06 0.8–1

NH4-N (mg/L) 40–50 40–45 ND

TN (mg/L) 60–80 60–70 ND

TS (g/L) 67–70 ND ND

VS (g/L) 58–60 ND ND

TSS (g/L) 5–7 0.33 1.36

VSS (g/L) 4–6 ND 1.1

Schematic 1 The UASB reactor used in this work


liquor suspended solids (MLSS) were 5.8 and 7.5 g/L,respectively.

Electrochemical oxidation experiments

Experiments were conducted in a DiaCell (type 100) singlecompartment electrolytic flow-cell manufactured by AdamantTechnologies (Switzerland). Two circular electrodes of 0.1 mdiameter made of BDD on silicon were used as the anode andcathode; each electrode area was 70 cm2 and the distancebetween them 0.01 m. In a typical run, the already anaerobi-cally treated cheese whey was loaded in a vessel and contin-uously recirculated in the cell through a peristaltic pump. In allcases, the working volume was 10 L. A spiral coil immersedin the liquid and connected to tap water supply was used toremove the heat liberated from the reaction. All experimentswere conducted at ambient temperature; nonetheless, temper-ature was found to increase slowly with treatment time (re-spective profiles are shown in Fig. 1), but it never exceeded34 °C at the end of each experiment which lasted for up to 7 h.Experiments were conducted at two current intensity values, 9and 18 A, with NaCl or Na2SO4 as the supporting electrolyte.

Analytical methods

COD was determined by the dichromate method. The appro-priate amount of sample was introduced into commerciallyavailable digestion solution containing potassium dichromate,sulfuric acid, and mercuric sulfate (Hach Europe, Belgium)and the mixture was then incubated for 2 h at 150 °C in a CODreactor (Model 45600-Hach Company, USA). COD concen-tration was measured colorimetrically using a DR/2010 spec-trophotometer (Hach Company, USA). The pH was measuredusing a pH meter (PHM92 LAB).

DOC was measured with a Shimadzu 500ATOC analyzer.Prior to analysis, the samples were filtered through a mem-brane filter with a pore size of 0.45 μm.

The analysis of TS, VS, total suspended solids (TSS),volatile suspended solids (VSS), alkalinity, and the VFA/alkalinity ratio was done according to standard methods(APHA 1998). True color was determined spectrophotomet-rically at 410 nm against reference Pt–Co solutions, asdescribed in detail by Hongve and Akesson (1996).

Chlorinated volatile organics in the aqueous phase caneasily be extracted using pentane as the extraction solvent.In this study, 10 mL of sample was pipetted into 40-mL screwcap vials. Then, 2 mL of pentane was added and the vials wereshaken vigorously for 1 min and allowed to stand for 3 min tofacilitate phase separation. Two milliliters of pentane extractwas analyzed on a Carlo Erba gas chromatograph equippedwith a Ni63 electron capture detector at 300 °C. A DB-5,60 m×0.32 mm I.D., 0.25-mm film thickness column wasemployed (J&W Scientific). The temperature program was

from 35 °C (15 min) to 100 °C (1 min) at 5 °C/min and from100 to 260 °C (2 min) at 15 °C/min. The injector temperaturewas set at 250 °C. Residual chlorine was measured accordingto standard method 4500 Cl-B method I (APHA 1998).

The anaerobic biodegradability of cheese whey was deter-mined using the biochemical methane potential (BMP) testwhere cumulative biogas production is monitored as a func-tion of time; the test was carried out in 250 mL glass bottles at35 °C with inoculum sludge obtained from the UASB reactor.Each bottle was fed with 100 mL of the medium solution(Angelidaki et al. 2009; EPA 1996) containing 10 % v/v of theinoculum sludge and 50 mL of the wastewater at differentconcentrations, 1 and 3 g/L. Bottles containing glucose at50 mg DOC/L were also prepared and used as controls inorder to check the inoculum response toward standard sub-strates. Two more bottles were used as blanks containing themedium solution but no substrate. Each concentration wascarried out in triplicate.

The ecotoxicity of cheese whey was determined againstA. salina, which is a common organism in seawater. A. salinawas preferred over bacteria because the latter are easily ame-nable to chlorine disinfection (Tsolaki et al. 2010).Dehydrated cysts of A. salina were put into a 500-mL bottlefilled with deionized water, which was previously aerated for30 min. The bottle was then placed in the incubator for 24 h at29 °C. Following incubation, A. salina were exposed for 24 hto anaerobically treated CW samples, prior to and after elec-trochemical oxidation, to record frequencies of immobiliza-tion of ten species in 2 mL. Cheese whey samples wereneutralized to pH=7 with 2.75 w/v NaCl and 0.021 w/vNaHCO3. Each sample was run in triplicate.

The biogas volume was measured with a drum-type Ritterwet gas flow meter (Ritter Apparetebau GmbH & Co). Meth-ane and CO2 contents of the biogas were monitored on a gaschromatograph (Shimadzu 14B) equipped with a thermalconductivity detector and two columns, i.e., molecular sieve(MS) 5A and porapak N operated at 40 °C.

Results and discussion

BMP test

As seen in Fig. 2, 80 % of the total biogas production wasachieved within the first 10 days at 3 and 1 g/L initial COD,while productionwas nearly complete (i.e., at least 95% of thetotal volume) after 14 days. Comparing the experimentalmethane yield (Bexp), which was about 20 % of the biogas asmeasured by gas chromatography, to the theoretical value(Btheor=0.35 L CH4/g COD removed), one can assess thedegree of anaerobic biodegradability (ABD) as follows(Raposo et al. 2011):


%ABD ¼ 100Bexp

Btheorð1Þ

This was 30 and 21 % at 1 and 3 g/L, respectively. Therelatively low anaerobic biodegradability at both concentra-tions may be due to different parameters (Angelidaki et al.2009). Espozito et al. (2012) mentioned that substrate toxicityis a likely inhibitor of anaerobic biodegradability for differentorganic substrates. As also seen in Fig. 2, biogas productionincreases as the ratio of COD to VSSinoculum (S/X ratio) in-creases. The concentrations of 1 and 3 g/L COD correspond toan S/X ratio of 0.17 and 0.5, respectively.

Anaerobic digestion experiments

The pH value and alkalinity are important indicators in ananaerobic reactor, especially for the activities of acetogenicbacteria and methanogens. Changes in pH and alkalinity canaffect the stability of the anaerobic reactor. Values of pHbetween 6.8 and 7.4 generally provide optimal conditionsfor the methanogens, whereas values between 6.4 and 7.8are considered necessary to maintain adequate activity(Grady et al. 1999). As for alkalinity (i.e., the capacity toneutralize acids expressed in g CaCO3/L), an anaerobic di-gester requires a bicarbonate alkalinity of 1 to 3 g CaCO3/Lfor stable operation (Wilcox et al. 1995). In this work, pHranged between 6.8 and 8.2 and alkalinity between 1 and 2 gCaCO3/L. The volatile fatty acids (VFA)-to-alkalinity ratio is

Fig. 1 Temperature profileduring the electrochemicaloxidation experiments

Fig. 2 Cumulative biogasproduction during the BMP test ofCW at different substrateconcentrations, 1 and 3 g/L, and asubstrate to inoculum ratio (S/X)of 0.17 and 0.5, respectively


typically used as an index to evaluate anaerobic system sta-bility. When the ratio is less than 0.3–0.4 (equiv. aceticacid/equiv. CaCO3), the process is stable and without the riskof acidification (Rincon et al. 2008). In this work, the VFA/alkalinity ratio was always kept below 0.3.

As can be seen in Fig. 3, COD removal ranged between 66and 97%with an average of 89% during the 87-day period ofoperation and a tendency for a stable removal after day 40 ofoperation.

These results are in good agreement with previous studies;for instance, Ergüder et al. (2001) claimed that the UASBreactor was very efficient in treating diluted CW resulting inCOD removal between 95 and 97 % at HRT between 2.1 and2.5 days. Similar results were obtained by Blonskaja andVaalu (2006) who reported 98 % COD removal at a HRT of2.5 days in a UASB reactor.

Fluctuations in COD removal can be attributed to fluctua-tions in influent COD concentration, which ranged between0.75 and 6 g/L, as can be seen in Fig. 4. Some variations ininfluent COD between the first and the last day of each feed(periodically repeated every 3 days) can be explained by thedegradation of CWoccurring in the feed tank due to the actionof domestic sewage aerobic microorganisms. Another possi-ble explanation for the observed discrepancies in the influentCOD is the amount of whey fat that was stuck to the feed tank,thus changing the organic loading. The feed tank was used forthe dilution of domestic and agro-industrial wastewaters. Upto day 30 of operation (first operational period), the dailybiogas production remained low, with the maximum valuebeing 21 L/day, whereas after day 30 (second operationalperiod) the average biogas production increased up to 75 L/day. Up to day 28, the reactor was fed with influent CODconcentrations up to 3 g/L, while for the remaining of themonitoring period, the average influent COD concentrationwas 4.3 g/L. In general, the biogas production increased whenthe influent COD concentration was more than 3 g/L, whichcorresponds to an organic loading rate (OLR) more than 550 gCOD/(L·day), while it was limited at influent COD concen-trations lower than 3 g/L.

Effluent COD concentrations were in the range 130–250and 130–404 mg/L during the first and the second operationalperiods, respectively. The cumulative biogas production isshown in Fig. 5; it is observed that biogas production beganon day 6 and gradually increased up to day 55, whereas fromdays 56 to 69 less than 2 L/day was produced, quantity whichwas increased again after day 70. These different productionrates are due to the different OLRs in the reactor. OLR, whichis the ratio of inlet COD over the HRT, is an importantparameter significantly affecting microbial ecology and per-formance of UASB systems (Latif et al. 2011).

The maximum biogas production of 75 L was reached inday 46 corresponding to 15 L of CH4 (20 % of the biogas) andalmost 60 L of CO2, as measured by gas chromatography.

However, this was less than the theoretical value of 0.35 LCH4/g COD removed; specifically, CH4 production in day 46should have been 192 L instead of 15 L, which is the mea-sured value at 3.4 g/L influent and 0.4 g/L effluentconcentration.

The presence of suspended solids in the feedstock consti-tutes the most important problem in the application of UASBtechnology. The accumulation of suspended solids in thereactor leads to an increase in the sludge bed height and adecrease in the fluidized zone. As a consequence, there is agradual decrease in the sludge activity, and therefore, slowhydrolysis of the suspended solids occurs, methanogenic ac-tivity decreases, and reduced volumes of biogas are produced(Kalogo and Verstaete 1999). In some instances, a good CODremoval efficiency can be observed although the conversionefficiency to methane may be poor. Such a removal may bedue to the physical entrapment of solids in the sludge bed (VanHaandel and Lettinga 1994).

Electrochemical oxidation

Following anaerobic treatment of diluted cheese whey atabout 6 g/L influent COD and 94 % removal, the resultingeffluent at 182–290 mg/L COD was added with salt to en-hance conductivity (i.e., at about 6.5 dS/m) and oxidized at18 A to evaluate its electrochemical degradability. Represen-tative results are shown in Fig. 6. COD removal was 97 and89 % after 2 h at 18 Awith 3.3 and 1.7 g/L sodium chloride,respectively, but it decreased to 66 %, when NaCl was re-placed by 10 g/L Na2SO4. Interestingly, doubling Na2SO4

concentration appears to have a detrimental effect on conver-sion, and this has also been observed in previous studiesconcerning the electrochemical oxidation of micro-pollutants(Frontistis et al. 2011) and landfill leachates (Turro et al.2011). The addition of NaCl to the effluent caused an increasein COD removal due to the role of indirect oxidation inducedby chloride radicals; this effect has been demonstrated inseveral previous studies (Frontistis et al. 2011; Turro et al.2011). An extra experiment was performed without addingsalt; the inherent conductivity of the effluent was 2.1 dS/m,and this gave rise to 6.5 A of current. As can be seen in Fig. 6,the 4-h COD removal was 32 %, substantially lower thanthose values achieved in the presence of NaCl or Na2SO4.

Two mechanisms are thought to be responsible for organicmatter electrochemical degradation, namely (i) direct anodicoxidation where the pollutants are adsorbed on the anodesurface and destroyed by the anodic electron transfer reactionand (ii) indirect oxidation in the liquid bulk which is mediatedby the oxidants that are formed electrochemically; such oxi-dants include chlorine, hypochlorite, hydroxyl radicals,ozone, and hydrogen peroxide (Israilides et al. 1997). In thepresence of NaCl, chlorohydroxyl radicals are also formed onthe anode surface and then oxidize the organic matter.


Reactions between water and radicals near the anode can yieldmolecular oxygen, free chlorine, and hydrogen peroxide. Freechlorine and oxygen can further react on the anode yieldingsecondary oxidants such as chlorine dioxide and ozone, re-spectively (Gotsi et al. 2005). In the case of alkaline condi-tions, hypochlorite, chloride ions, and hydroxyl radicals are allimportant (Gotsi et al. 2005).

The electrochemical treatment was also very effective indecolorizing the wastewater. The feed to the electrochemicalreactor had a color of 1,500–2,000 TCU, which decreased to50–100 TCU after 4 h of oxidation with NaCl and 200 TCUwith Na2SO4. The efficient electrochemical decolorization ofindustrial effluents including olive mill wastewaters (OMW)(Kotta et al. 2007), textile dyehouses (Tsantaki et al. 2012),and landfill leachates (Turro et al. 2011) has been reported inthe literature.

A second set of electrochemical oxidation experiments wasthen performed in an attempt to characterize the treated efflu-ent in terms of COD and DOC removal patterns, the formation

of organochlorinated compounds, and their ecotoxicity. Theseexperiments were performed at two chloride concentrations(1.7 and 3.3 g/L) and two values of current intensity (9 and18 A), while the initial COD concentration was between 139and 412 mg/L. As seen in Fig. 7, electrochemical oxidationresulted in nearly complete COD removal after 3–4 h irre-spective of the salinity level, initial COD, and current intensi-ty. Conversely, mineralization was not complete since DOConly partially decreased even after prolonged (7 h) treatment;for example, 50 % DOC removal was achieved for eitherchloride concentration at 18 A, whereas at 9 A the removalwas 12 and 31 % at 1.7 and 3.3 g/L salinity, respectively.Characteristics of anaerobically treated CW after 7 h of elec-trochemical post-treatment at 1.7 g/L NaCl and 18 A are givenin Table 2.

The concept of combining anaerobic digestion with elec-trochemical post-oxidation over dimensionally stable anodesto treat OMW (i.e., a heavily polluted agro-industrial effluent)has recently been demonstrated by Goncalves et al. (2012).

Fig. 3 COD removal in UASBexperiments. Straight line showsaverage value over an 87-dayperiod of operation

Fig. 4 Daily biogas production(bars) and variation in CODinfluent

(symbols) in the pilot-scale UASBreactor


They reported that an anaerobically digested OMW at1,100 mg/L COD could be completely treated (in terms ofCOD, phenols, and color removal) by electrochemical oxida-tion over a RuO2 electrode; this would require 45 h at a currentdensity of 6.9 mA/cm2.

The temporal profiles of COD and DOC shown in Fig. 7imply that the electrochemical oxidation of CWoccurs mainlythrough partial rather than total oxidation reactions. Thismeans that organic compounds are predominantly convertedto more oxidized species (i.e., consistent with high CODreductions) without further oxidation to carbon dioxide (i.e.,consistent with relatively low to moderate DOC reductions).To compare the rates of COD and DOC decrease, one canassume pseudo-first-order kinetics (this is typical for the ad-vanced electrochemical oxidation of low concentration efflu-ents (Mavros et al. 2008)) and compute the respective kineticconstants from the data of Fig. 7, i.e.:

−dc

dt¼ kappc⇔ln

coc¼ kappt ð2Þ

where c is the concentration of COD or DOC and kapp is anapparent kinetic constant. The average values are 1.44±0.44and 5.5±3.7 10−2 min−1 for COD and DOC, respectively,showing a discrepancy of up to two orders of magnitude.

To quantify the relative contribution of partial and totaloxidation reactions, the index μ can be defined as follows(Jochimsen and Jekel 1997):

μ ¼ CODpartial

COD0−COD;where CODpartial ¼ COD0

DOC

DOC0−COD

ð3Þ

The index μ, which corresponds to the degree of CODremoved via partial oxidation only, approaches zero whentotal oxidation predominates and one when partial oxidationreactions mainly occur. Figure 8 shows the efficiency ofpartial oxidation μ with time for the data of Fig. 7. Partialoxidation seems to be dominant since μ is, in most cases,higher than 0.5 (i.e., more than 50 % of the overall CODremoval is due to partial oxidation reactions). Interestingly,

Fig. 5 Cumulative biogasproduction in the pilot-scaleUASB reactor

Fig. 6 Normalized COD profileduring the electrochemicaloxidation of anaerobically treatedCW as a function of supportingelectrolyte concentration and type


these findings do not support the commonly accepted behav-ior of BDD anodes as promoters of the mineralization oforganics. It is well-documented (Kapałka et al. 2009) thatBDD acts as a generator of quasi-free hydroxyl radicals,which mediate oxidation reactions taking place in a layer justa few nanometers away from the electrode surface. In thissense, radicals are formed rapidly and are very reactive to-wards mineralization due to their weak interaction with BDD.

The effect of current intensity and salinity on the formationof chlorinated organic compounds was also studied. Figure 9shows temporal concentration profiles for trihalomethanes(THMs), haloacetonitriles (HANs), haloketons (HKs), and1,2-dichloroethane (DCA). The latter alongside chloroformwere the main compounds formed in terms of concentration(i.e., an order of magnitude greater than HANs and HKs). It isalso notable that the concentrations of the above by-productsexceed the maximum values (i.e., 80, 80, and 5 μg/L forTHMs, HANs, and DCA, respectively) set by USEPA fordrinking water. In addition to the aforementioned by-products, chloropicrin was also detected at low concentrations(i.e., <20 μg/L). In a recent work of our group (Anglada et al.2011), the same set of by-products was checked for during theBDD electrochemical oxidation of landfill leachates

inherently containing 2.6 g/L chloride. With the exception ofchloropicrin, all the other compounds were identified withchloroform being the dominant one (i.e., about half of thetotal concentration).

Overall, the concentration profiles of these compoundsexpectedly depend on both salinity and current intensity sincethese two factors are likely to influence degradation kineticsand, consequently, by-products distribution. Prolonged elec-trochemical oxidation generally decreased their concentrationand this is in line with the COD and DOC patterns shown inFig. 7, while increased current intensities favored by-productformation. Moreover, previous studies dealing with the fate ofby-products generated during water chlorination reported thatelectrochemical oxidation and/or reduction processes are ca-pable of eliminating this type of pollutants (Scialdone et al.2010; Fiori et al. 2005).

The final samples of the experiments shown in Fig. 7 weretested for their toxicity to A. salina; working at 3.3 g/Lchloride concentration resulted in 100 % inhibition even aftera threefold dilution of the sample. A lower chloride concen-tration (1.7 g/L) also led to complete inhibition after a twofoldsample dilution, which decreased to 40–60% inhibition after athreefold sample dilution. Although the ecotoxicity data can-not be directly related to the profiles shown in Fig. 9 due to thecomplex nature of the water matrix, it is likely that ecotoxicityis, to some degree, due to the organochlorinated transforma-tion by-products; this argument is substantiated by the fact thatthe anaerobically treated sample exhibited no ecotoxicity priorto electrochemical oxidation. This said, A. salina death mayalso be due, to a large extent, to the presence of free chlorine;this was measured at 851 and 1,028 mg/L after 7 h of

Fig. 7 Normalized COD (solidlines) and DOC (dashed lines)profiles during theelectrochemical oxidation ofanaerobically treated CWas afunction of NaCl concentrationand current intensity

Table 2 Characteristicsof anaerobically treatedCWafter electrochemicalpost-treatment (1.7 g/LNaCl, 18 A, 7 h)

BDL below detectionlimit

pH 7.6

Conductivity (dS/m) 6.5

DOC (mg/L) 28.4

COD (g/L) BDL

Color (TCU) 50


Fig. 8 Temporal variation of theefficiency of partial oxidationindex (μ) for the data of Fig. 7

Fig. 9 Temporal profiles of a trihalomethanes, b haloacetonitriles, c haloketons, d 1,2-dichloroethane during electrochemical oxidation at the conditionsof Fig. 7


treatment at 9 and 18 A, respectively, in the case of 3.3 g/Lchloride, while the corresponding values were 393 and355 mg/L, at 1.7 g/L chloride.

Conclusions

The conclusions drawn from the present study are summa-rized as follows:

& Anaerobic digestion of diluted cheese whey in a pilot-scale UASB reactor was capable of achieving an averageCOD removal of 89 % under various organic loadings.Biogas production generally increased with increasinginfluent COD concentration although its CH4 contentwas lower than the theoretical one.

& Anaerobically pretreated effluents still needed somepolishing to meet receiving water discharge standards. Inthis work, this was accomplished by means of electro-chemical post-oxidation over BDD electrodes with NaClas the supporting electrolyte; complete COD removal anddecolorization were attained in up to 4 h of operation.Conversely, mineralization was not complete since DOConly partially decreased even after prolonged (7 h)treatment.

& Although the use of chloride ions usually enhances elec-trochemical degradation due to the formation and oxida-tive action of secondary species, the main drawback is stillthe side formation of undesired organochlorinated by-products. Several such compounds were analytically de-termined in this work; their presence, alongside the for-mation of residual chlorine, is thought to be responsiblefor the increased post-treatment ecotoxicity.

Acknowledgments This research has been co-financed by the Europe-an Union (European Social Fund—ESF) and Greek national fundsthrough the Operational Program “Education and Lifelong Learning” ofthe National Strategic Reference Framework (NSRF)—ResearchFunding Program: Heracleitus II. Investing in knowledge society throughthe European Social Fund.

References

Angelidaki I, Alves M, Bolzonella D, Borzacconi L, Campos JL, GuwyAJ, Kalyuzhnyi S, Jenicek P, van Lier JB (2009) Defining thebiomethane potential (BMP) of solid organic wastes and energycrops: a proposed protocol for batch assays. Water Sci Technol 59:927–934

Anglada A, Urtiaga A, Ortiz I (2009) Contributions of electrochemicaloxidation to waste-water treatment: fundamentals and review ofapplications. J Chem Technol Biotechnol 84:1747–1755

Anglada A, Urtiaga A, Ortiz I, Mantzavinos I, Diamadopoulos E (2011)Boron-doped diamond anodic treatment of landfill leachate:Evaluation of operating variables and formation of oxidation by-products. Water Res 45:828–838

APHA (1998) Standard methods for the examination of water and waste-water, 20th edn. American Public Health Association, Washington

Blonskaja V, Vaalu T (2006) Investigation of different schemes foranaerobic treatment of food industry wastes in Estonia. Proc EstAcad Sci: Chem 55:14–28

Carvalho F, Prazeres RA, Rivas J (2013) Cheese whey wastewater:characterization and treatment. Sci Total Environ 445–446:385–396

Comninellis C, Kapalka A, Malato S, Parsons SA, Poulios I,Mantzavinos D (2008) Advanced oxidation processes for watertreatment: advances and trends for R&D. J Chem TechnolBiotechnol 83:769–776

EPA (1996) Fate, transport and transformation test guidelines; OPPTS835.3400 Anaerobic Biodegradability of Organic Chemicals

Ergüder TH, Tezel U, Güven E, Demirer GN (2001) Anaerobic biotrans-formation and methane generation potential of cheese whey in batchand UASB reactors. Waste Manag 21:643–650

Espozito G, Frunzo L, Liotta F, Panico A, Pirozzi F (2012) Bio-methanepotential tests to measure the biogas production from the digestionand co-digestion of complex organic substrates. Open Environ Eng J5:1–8

Fiori G, Rondinini S, Sello G, Vertova A, Cirja M, Conti L (2005)Electroreduction of volatile organic halides on activated silver cath-odes. J Appl Electrochem 35:363–368

Frontistis Z, Brebou C, Venieri D, Mantzavinos D, Katsaounis A (2011)BDD anodic oxidation as tertiary wastewater treatment for theremoval of emerging micro-pollutants, pathogens and organic mat-ter. J Chem Technol Biotechnol 86:1233–1236

Gannoun H, Khelifi E, Bouallagui H, Touhami Y, Hamdi M (2008)Ecological clarification of cheese whey prior to anaerobic digestionin upflow anaerobic filter. Bioresour Technol 99:6105–6111

Gelegenis J, Georgakakis D, Angelidaki I, Mavris V (2007) Optimizationof biogas production by co-digesting whey with diluted poultrymanure. Renew Energy 32:2147–2160

Goncalves MR, Marques IP, Correia JP (2012) Electrochemical mineral-ization of anaerobically digested olive mill wastewater. Water Res46:4217–4225

Gotsi M, Kalogerakis N, Psillakis E, Samaras P, Mantzavinos D (2005)Electrochemical oxidation of olive oil mill wastewaters. Water Res39:4177–4187

Grady CP, Daigger GT, Lim HC (1999) Biological wastewater treatment,2nd edn. Marcel Dekker, New York

Hanson G (1982) End product inhibition in methane fermentation.Process Biochem 17:45–49

Hongve D, AkessonG (1996) Spectrophotometric determination of watercolour in Hazen units. Water Res 30:2771–2775

Israilides C, Vlyssides A, Mourafeti V, Karvouni G (1997) Olive oilwastewater treatment with the use of an electrolysis system.Bioresour Technol 61:163–170

Jochimsen JC, Jekel MR (1997) Partial oxidation effects during thecombined oxidative and biological treatment of separated streamsof tannery wastewater. Water Sci Technol 35:337–345

Kalogo Y, Verstaete W (1999) Development of anaerobic sludge bed(ASB) reactor technologies for domestic wastewater treatment: mo-tives and perspectives. World J Microb Biotechnol 15:523–534

Kapałka A, Fóti G, Comninellis C (2009) The importance of electrodematerial in environmental electrochemistry: Formation and reactiv-ity of free hydroxyl radicals on boron-doped diamond electrodes.Electrochim Acta 54:2018–2023

Kotta E, Kalogerakis N,Mantzavinos D (2007) The effect of solids on theelectrochemical treatment of olive mill effluents. J Chem TechnolBiotechnol 82:504–511

Latif AM, Ghufran R,WahidAZ, AhmadA (2011) Integrated applicationof upflow anaerobic sludge blanket reactor for the treatment ofwastewaters. Water Res 45:4683–4699

Mavros M, Xekoukoulotakis NP, Mantzavinos D, Diamadopoulos E(2008) Complete treatment of olive pomace leachate by coagulation,


activated carbon adsorption and electrochemical oxidation. WaterRes 42:2883–2888

Minhalma M, Magueijo V, Queiroz DP, Pinho MN (2007) Optimizationof “Serpa” cheese whey nanofiltration for effluent minimization andby-products recovery. J Environ Manag 82:200–206

Prazeres RA, Carvalho F, Rivas J (2012) Cheese whey management: areview. J Environ Manag 110:48–68

Raposo F, Fernandez-Cegri V, De la Rubia MA et al (2011) Biochemicalmethane potential (BMP) of solid organic substrates: evaluation ofanaerobic biodegradability using data from an internationalinterlaboratory study. J Chem Technol Biotechnol 86:1088–1098

Rincon B, Borja R, Gonzalez JM, Portillo MC, Saiz-Jimenez C (2008)Influence of organic loading rate and hydraulic retention time on theperformance, stability and microbial communities of one-stage an-aerobic digestion of two-phase olive mill solid residue. BiochemEng J 40:253–261

Rivas J, Prazeres RA, Carvalho F, Beltrán F (2010) Treatment of cheesewhey wastewater: combined coagulation–flocculation and aerobicbiodegradation. J Agric Food Chem 58:7871–7877

Scialdone O, Galia A, Currer i L, Randazzo S (2010)Electrochemical abatement of chloroethanes in water: reduc-tion, oxidation and combined processes. Electrochim Acta 55:701–708

Tsantaki E, Velegraki T, Katsaounis A, Mantzavinos D (2012) Anodicoxidation of textile dyehouse effluents on boron-doped diamondelectrode. J Hazard Mater 91–96:207–208

Tsolaki E, Pitta P, Diamadopoulos E (2010) Electrochemical disinfectionof simulated ballast water using Artemia salina as indicator. ChemEng J 156:305–312

Turro E, Giannis A, Cossu R, Gidarakos E,Mantzavinos D, Katsaounis A(2011) Electrochemical oxidation of stabilized landfill leachate onDSA electrodes. J Hazard Mater 190:460–465

Van Haandel AC, Lettinga G (1994) Anaerobic sewage treatment.A practical guide for regions with a hot climate. Wiley, NewYork

Wilcox SJ, Hawkes FR, Guwy AJ (1995) A neural network based onbicarbonate monitoring, to control anaerobic digestion. Water Res29:1465–1470


coupling digestion in a pilot-scale uasb reactor and electrochemical oxidation over bdd anode to...

Documents