A

bopm(otn(©

K

1

ociiksswtbpo

0d

Available online at www.sciencedirect.com

Journal of Hazardous Materials 154 (2008) 839–845

Kinetics of the biodegradation of green tableolive wastewaters by aerobic

and anaerobic treatments

J. Beltran ∗, T. Gonzalez, J. GarciaDepartamento de Ingenieria Quimica y Energetica, Universidad de Extremadura, 06071 Badajoz, Spain

Received 20 October 2005; received in revised form 26 October 2007; accepted 31 October 2007Available online 7 November 2007

bstract

The biodegradation of the organic pollutant matter present in green table olive wastewater (GTOW) is studied in batch reactors by an aerobiciodegradation and by an anaerobic digestion. In the aerobic biodegradation, the evolution of the substrate (in terms of chemical and biochemicalxygen demand), biomass, and total polyphenolic compounds present in the wastewater are followed during the process, and a kinetic study iserformed using Contois’ model, which when applied to the experimental results provides the kinetic parameter of this model, resulting in aodified Contois’ equation (q = 3.3S/(0.31S0X + X), gCOD/gVSS d−1). Other kinetic parameters were determined: the cellular yield coefficient

YX/S = 5.7 × 10−2 gVSS/gCOD) and the kinetic constant of cellular death phase (kd = 0.16 d−1). Similarly, in the anaerobic digestion, the evolutionf the substrate digested and the methane produced are followed, and the kinetic study is conducted using a modified Monod model combined with

he Levenspiel model, due to the presence of inhibition effects. This model leads to the determination of the kinetic parameters: kinetic constant wheno inhibitory substance is present (kM0 = 8.4 × 10−2 h−1), critical substrate concentration of inhibition (TP* = 0.34 g/L) and inhibitory parametern = 2.25).

2007 Elsevier B.V. All rights reserved.

digest

pidatcp

bpc

eywords: Green table olive; Wastewater; Aerobic biodegradation; Anaerobic

. Introduction

The manufacturing of table olives is an industrial processf great economical importance in the Southern Mediterraneanountries, like Spain, Portugal, Italy and Greece [1], with twomportant varieties: black and green types. The processingnvolves several steps: the selected olives are first cleaned andept in a solution with sodium chloride (brine) for better con-ervation. Then, they are later treated with sodium hydroxideolutions (lyes) for debittering of fruit, followed by severalashes with water to remove the excess of alkali. Finally,

he olives are pitted and packed, sometimes in a new cover

rine. In the case of green olives, previously to pack they areut into sodium chloride brine in which lactic fermentationccurs.

∗ Corresponding author. Tel.: +34 924 289385; fax: +34 924 289385.E-mail address: jbelther@unex.es (J. Beltran).

Scmmesa

304-3894/$ – see front matter © 2007 Elsevier B.V. All rights reserved.oi:10.1016/j.jhazmat.2007.10.102

ion; Kinetic study

Although this type of industry works seasonally, the wholerocess requires large volumes of tap water mainly for clean-ng and washing the olives in several stages, as well as forifferent uses in the production plants. The volume of wastew-ters generated in the process is very important. Therefore,he most frequent procedure of elimination being the dis-harge of untreated waters in the environment into evaporationonds [2].

These wastewaters contain moderate to large chemical andiological oxygen demand, together with a significant content ofolyphenolic compounds which in general are toxic to biologi-al processes [3,4] and inhibit the efficiency of these processes.ince only a few data have been reported with regard to the appli-ation of physical [5,6], chemical [7,8] and biological [9,10]ethods for the purification of wastewaters generated in the

anufacturing of green table olives, the biodegradation of these

ffluents by means aerobic and anaerobic processes has beentudied in the present research. In both cases, the objectivesre to report data for the removal of the organic matter and

840 J. Beltran et al. / Journal of Hazardou

Nomenclature

BOD5 biochemical oxygen demand (g O2/L)COD chemical oxygen demand at each reaction time

(g O2/L)COD0 initial substrate concentration (gCOD/L)CODf final substrate concentration (gCOD/L)−dS/dt substrate degradation rate (g/(L d))dX/dt rate of production of biomass (g(X)/(L h))k0 empirical constant defined by Eq. (7) (L d/g(S))k1 empirical constant defined by Eq. (7) (d)kd kinetic constant of cellular death phase (d−1)kM anaerobic rate constant defined by Eq. (18) (h−1)kM0 anaerobic rate constant without inhibitory effects

(h−1)KC Contois’ constant (g(S)/g(X))KS Monod saturation constant (g(S)/L)n inhibitory parameterq specific decomposition rate of substrate

(g(S)/g(X)/d)qmax maximum specific decomposition rate of sub-

strate (g(S)/(g(X) d))S biodegradable substrate concentration at each

reaction time (gCOD/L)S0 initial biodegradable substrate concentration

(gCOD/L)Sf final biodegradable substrate concentration

(gCOD/L)TP0 initial concentration of polyphenolic compounds

in each experiment (g/L)TP* the critical inhibition concentration of polyphe-

nolic compounds (g/L)VM volume of methane produced at each reaction time

(mL)VMF volume of methane produced at the end of each

experiment (mL)VR anaerobic digestor volume (L)VSS volatile suspended solids (g/L)X biomass concentration at each reaction time

(gVSS/L)X0 initial biomass concentration (gVSS/L)XCOD removal reached in every experiment (%)YM methane production yield coefficient (mL

methane/gCOD degraded)YX/S cellular yield coefficient (gVSS/gCOD)

Greek lettersμ specific growth rate of biomass (d−1)

toasm

2

2

isfMpuawmmfpassCc(

2

d1bas

amrmwi

btd1wtodtbar

μmax maximum specific growth rate of biomass (d−1)

o develop kinetic studies of each process. Kinetic constantsbtained can be used for design and operation of aerobic and

naerobic reactors [11,12]. Various kinetic models have beenuccessfully applied including Monod model, modified Monododel, Haldane model, Contois model, etc.

2

o

s Materials 154 (2008) 839–845

. Materials and experimental procedures

.1. Green table olive wastewaters

The wastewaters were obtained from a green table olivendustry located at Aceuchal (Extremadura Community inouthwestern Spain). The analysis of this wastewater was per-ormed according to the procedures described in the Standard

ethods [13], except the total polyphenolic and aromatic com-ounds. The total polyphenolic compounds were measuredsing Folin-Ciocalteau reagent (a mixture of phosphomolybdicnd phosphotungstic acids) after prior extraction of the sampleith ethyl acetate. The result is a blue polymer in an alkalineedium [14]. The aromatic compounds were determined byeasuring the absorbance at 254 nm [15]. The values obtained

or the main chemical parameters and compositions wereH 4.5; alkalinity = 2.45 g CaCO3/L; volatile acidity = 2.22 gcetic acid/L; total solids = 50.8 g/L; mineral suspendedolids = 0.22 g/L; volatile suspended solids = 0.58 g/L; total dis-olved solids = 50.0 g/L; mineral dissolved solids = 21.4 g/L;OD = 49.6 g O2/L; BOD5 = 21.4 g O2/L; total polyphenolicompounds = 3.06 g caffeic acid/L and aromatic compoundsdilution 1:100) = 1.22 (absorbance).

.2. Aerobic biodegradation

The aerobic degradation experiments of GTOW were con-ucted in a cylindrical mixed batch reactor (internal diameter:0 cm; height: 20 cm), which was submerged in a thermostaticath to maintain the temperature constant at 28 ± 0.5 ◦C. Theirflow was fed to the reacting medium through a bubble gasparger at a constant flow rate of 125 L/h at room conditions.

As GTOW does not contain microorganisms capable oferobic biodegradation, a previous stage was necessary to accli-atize bacterial flora to this substrate. For this purpose, the

eactor was inoculated with an activated sludge taken from aunicipal wastewater treatment plant, and several experimentsere carried out with successive additions of GTOW gradually

ncreasing concentrations of COD [16].Once the acclimatization stage was finished, the GTOW

iodegradation experiments were conducted. Thus, and prioro each run, the GTOW was diluted in order to attain the pre-esignated initial concentration of COD for the experiment, andL of this wastewater was introduced into the reactor whichas also inoculated with the previously acclimatized biomass in

he amount required to obtain the desired initial concentrationf biomass for the experiment. During an experiment (always 7ays), samples were withdrawn at regular intervals to analyzehe substrate (chemical and biochemical oxygen demand) andiomass (measured as volatile suspended solids) concentrationsnd the total content of polyphenolic compounds present in theeacting medium.

.3. Anaerobic digestion

The anaerobic digestion experiments of GTOW were carriedut in a magnetically stirred spherical batch anaerobic diges-

rdous

twbrm31taa

ofmwiwtmd

diGaaiCmd

3

3

owbaoc

sapee

cte

X

wc

TE

E

AAAAAAAAA

TE

T

001234567

J. Beltran et al. / Journal of Haza

ion reactors (internal diameter: 17 cm). They were providedith a cover containing inlets for loading feedstocks and bub-ling the inert gas (required for unloading), and outlets foremoving effluents and venting biogas. The reactors were sub-erged in a thermostatic bath at a constant temperature of

5 ± 0.5 ◦C. The methane produced was measured by usingL Boyle–Mariotte reservoirs. Previously, the biogas passed

hrough a sodium hydroxide solution to retain carbon dioxide,nd the volume of methane produced was determined from themount of water displaced by the gas.

Similarly to the aerobic biodegradation experiments, a previ-us acclimatization stage was necessary to adapt the biomassrom an anaerobic digester of a municipal wastewater treat-ent plant to this substrate [17]. Thus, anaerobic digestersere inoculated with this biomass and successive increment-

ng volume of wastewater were loaded into the reactors,hich provided increasing initial substrate concentrations in

he digesters, from 0.5 to 3 gCOD/L. In every experiment, theethane volume produced and the initial and final COD was

etermined.The anaerobic digestion experiments were run by loading

ifferent volumes of the original GTOW leading to increasenitial concentration of substrate to be degraded. The load ofTOW in every experiment was introduced into the digester

fter separating the same volume of liquid from the digesterfter settling 10 h in order to avoid biomass losses. Each exper-

ment was continued until the methane production and theOD removal ceased. At regular intervals throughout an experi-ent, the substrate concentration and the methane released were

etermined.

w

ac

able 1xperimental and kinetic results in the aerobic biodegradation process

xpt. COD0 (g/L) TP0 × 102 (g/L) S0 (g/L) X0 (g/L) XCOD

-1 41.1 2.53 30.4 0.17 63-2 41.3 2.55 27.7 0.70 63-3 41.1 2.53 28.2 1.60 60-4 42.1 2.60 30.0 2.27 67-5 9.57 0.59 5.6 0.52 61-6 12.6 0.78 7.3 0.42 67-7 17.3 1.09 9.6 0.49 53-8 31.8 1.96 17.2 0.44 49-9 41.6 2.57 27.3 0.51 55

able 2volution of several parameters in experiment A-1 taken as an example in the aerobi

ime (d) COD (g/L) X (g/L) BOD5 (g/L) TP (g/L)

41.1 0.17 16.7 2.90.62 38.8 0.47 2.24

31.7 0.62 2.0127.0 0.78 1.8825.8 1.00 1.3819.5 1.12 11.3 0.3516.9 1.09 0.2715.8 0.86 0.1415.2 0.76 7.4 0.08

Materials 154 (2008) 839–845 841

. Results and discussion

.1. Aerobic biodegradation

GTOW was degraded by aerobic microorganisms, in a groupf experiments where the initial substrate concentration, COD0,as varied between 9.57 and 41.6 gCOD/L, and the initialiomass concentration, X0, ranged from 0.17 to 2.27 gVSS/L,ccording to the values described in Table 1. In the evaluationf initial biomass concentration was taken into account the VSSontent present in the original wastewater.

During an experiment, the concentration evolution of the sub-trate (chemical and biochemical oxygen demand), biomass,nd total polyphenolic compounds were followed. As an exam-le, Table 2 shows the values obtained for these variables inxperiment A-1, with similar trends observed in the remainingxperiments.

As observed in Fig. 1 for the series varying initial substrateoncentration, this variable decreases continuously with reactionimes, as could be expected. The total removal reached in everyxperiment is defined as

COD = COD0 − CODf

COD0100 (1)

here COD0 and CODf represent the initial and final substrateoncentration. Table 1 shows the values obtained for XCOD,

hich ranged between 49 and 67%.On the other hand, the total polyphenolic compounds, TP;

lso decreases continuously with time (see Table 2), and theonversions obtained, XTP (defined in a similar form as for the

(%) XTP (%) XBOD5 (%) KC/qmax (gVSS d/gCOD) r2

97 56 3.0 0.99297 69 2.4 0.99092 66 2.6 0.99598 69 2.3 0.995

100 29 0.53 0.987100 40 1.1 0.997100 34 1.5 0.992

99 59 1.8 0.98197 63 2.8 0.955

c biodegradation process

S (g/L) −dS/dt (g/L d) q (gCOD/gVSS d) μ (d−1)

30.4 8.0 46.5 2.3828.1 7.1 15.1 0.7521.0 6.6 10.6 0.5116.4 5.3 6.8 0.2815.1 4.1 4.0 0.12

8.9 3.0 2.7 0.036.3 2.0 1.9 −0.065.2 1.2 1.4 −0.184.6 0.5 0.6 −0.32

842 J. Beltran et al. / Journal of Hazardou

Fp

CAf

(atpos

b[

wtit

−

Fp

we

teCtmcft

q

wt

mtl

aapt(

q

tbdn

ig. 1. Evolution of COD with bioreaction time in the aerobic biodegradationrocess. Experimental conditions are given in Table 1.

OD, Eq. (1)), for every experiment are also shown in Table 1.s can be seen, a high removal in the range 92–100% is reached

or this parameter in all cases.With respect to the biomass, as it is seen in experiment A-1

Table 2) and in Fig. 2 for some of experiments, its evolutiongrees well with the typical growth-cycle phases for batch cul-ivations [18]. As Fig. 2 shows for some of experiments, theopulation of microorganisms increases from the first momentf the culture, indicating that the lag phase is not present in thisystem due to use a biomass acclimatized previously [19].

In the exponential growth phase, the rate of production ofiomass (X) is well described by a first order kinetic equation20]

dX

dt= μX (2)

here μ is the specific growth rate of biomass. Simultaneouslyo the production of the cells, the substrate (S) is degraded, andts rate is also proportional to the mass of cells present, according

o the expression

dS

dt= qX (3)

ig. 2. Evolution of biomass with bioreaction time in the aerobic biodegradationrocess. Experimental conditions are given in Table 1.

ttwdc

ceutautgatctac

s Materials 154 (2008) 839–845

here q is the specific substrate degradation rate, single param-ter which characterizes the degradation process.

The literature [21] provides several expressions which relatehe specific rates (μ and q) to the substrate concentration, forxample Monod, Contois, Haldane models, etc. Among them,ontois’ equation [22] usually gives excellent fits to experimen-

al results. This model considers some effects of inhibition byedium constituents, such as substrate or products [18]. In the

ase of the specific degradation rate, this model proposes theollowing equation for q as a function of the substrate concen-ration

= qmaxS

KCX + S(4)

here qmax represents the maximum rate of substrate degrada-ion and KC is the Contois’ constant.

In order to obtain the specific kinetic parameters for thisodel, qmax and KC or a ratio between them, which consti-

utes the objective of the present kinetic study, Eq. (4) can beinearized in the form [23]

1

q= 1

qmax+ KC

qmax

X

S(5)

According to Eq. (5), a plot of 1/q versus X/S must lead tostraight line for every experiment conducted, whose intercept

nd slope will be 1/qmax and KC/qmax, respectively. For thisurpose, the specific rate q must be previously evaluated for eachime of bioreaction, by transforming its definition expressionEq. (3))

= − dS

X dt(6)

For this evaluation, the term dS/dt is calculated by fittinghe experimental data (S, t) to a third polynomic expressiony least-square regression analysis using Excel worksheet, andividing by the biomass concentration X at any time. It must beoted that S represents the biodegradable substrate concentra-ion, which is determined by substracting from the COD at anyime the non-biodegradable substrate concentration. This valueas determined by adjusting experimental data to an exponentialecay curve, using Origin 8.0 software. The asymptotic valueorresponds to non-biodegradable COD [24].

According to this procedure, Table 2 shows the values cal-ulated for −dS/dt and q in the experiment A-1, selected as anxample. Once the specific rate, q, are known, Eq. (5) can besed as described before. After least-square regression analysis,he slopes and intercepts are determined for all the experiments,nd they are summarized in Table 1. In general, the small val-es of 1/qmax (including some negative values, generated byhe regression analysis, but obviously without meaning) sug-est high values for qmax that cannot be calculated and reportedccurately from the experimental results [25]. On the other hand,he slope, KC/qmax, clearly fall into two groups. In the first one,

orresponding to experiments A-1-A-4 and A-9 (where the ini-ial biomass X0 was varied and COD0 remained nearly constantround 41 gCOD/L), the values of KC/qmax are very close, indi-ating that the biomass has no effect on this biokinetic constant.

J. Beltran et al. / Journal of Hazardous

FC

IaoT

eeoo(pii

q

afais

Y

aw

μ

gtiobmtt

−

wde

μ

iisdtaaecddf1f

3

doidcaa

TE

E

NNNNN

ig. 3. Kinetic study in the aerobic biodegradation process. Influence of initialOD on KC/qmax (r2 = 0.936).

n the second group, experiments A-5 to A-9 (with X0 constantnd the substrate concentration S0 being modified), an incrementf KC/qmax is observed when S0 increases, as indicated in Fig. 3.herefore, a linear relationship can be proposed [16,26]

KC

qmax= k0S0 + k1 (7)

According to Eq. (7), a plot of KC/qmax against S0 inxperiments A-5 to A-9 leads to the determination of bothmpirical constants. After regression analysis, the valuesf k0 = 9.1 × 10−2 ± 0.01 L d/gCOD and k1 = 0.32 ± 0.23 d arebtained. With the results obtained, the initial Contois’ modelEq. (4)) can be modified. Thus, the intercepts close to zero in theresent system, suggest that the term 1/qmax can be eliminatedn Eq. (5) because KCX � S. Therefore, Eq. (4) is transformednto

= qmaxS

KCX(8)

Other kinetic parameters useful in the design of bioreactorsre related with the biomass evolution during the whole cycleor batch cultivation [21], like the cellular yield coefficient, YX/S,nd the kinetic constant of cellular death phase, kd. The first ones defined as the ratio mass of biomass produced per mass ofubstrate consumed, and can be expressed by the equation

X/S = −dX

dS(9)

nd taking into account the definition Eqs. (2) and (3), it can beritten

= YX/Sq (10)

f

vm

able 3xperimental and kinetic results for the anaerobic digestion process

xpt. COD0 (g/L) TP0 × 102 (g/L) XCOD (%) V

-1 0.6 3.7 92-2 1.0 6.2 81-3 1.75 10.8 94-4 2.5 15.4 81 1-5 3.0 18.5 89 1

Materials 154 (2008) 839–845 843

However, this expression only holds for the exponentialrowth phase. For the whole growth cycle of microorganisms,he death phase must be also taken into account, when the declinen the cells number takes place. As Bailey and Ollis pointedut [18], relatively few studies have been made on this phase,ecause many industrial batch microbiological processes are ter-inated before the death phase begins. Usually the death rate of

he microorganism’s population during this period is assumedo follow a first order kinetics [20]

dX

dt= kdX (11)

here kd is the previously mentioned kinetic constant for theecrease rate of biomass in the death phase. Therefore, the globalxpression for the specific growth rate μ must be

= YX/Sq − kd (12)

According to this equation, a plot of μ values versus q valuesn the experiments must lead to a straight line whose slope andntercept will be YX/S and kd, respectively. For this purpose, thepecific growth rate of biomass μ was evaluated. That is, the termX/dt was determined by fitting the experimental data (X, t) to ahird polynomic expression by means of a regression analysis;nd later, this term was divided by the biomass concentration X atny time [27]. Table 2 shows the values obtained for both param-ters (q and μ) in the experiment A-1 taken as an example: itan be clearly seen the decreasing values of μ with reaction timeuring the exponential growth phase, and their negative valuesuring the death phase. After least-square regression analysisor all the experimental values, it is obtained a value of 5.70−2 ± 0.4 × 10−2 g VSS/gCOD for YX/S and 0.16 ± 0.02 d−1

or kd.

.2. Anaerobic digestion

The anaerobic digestion process of GTOW has been con-ucted by a group of experiments where the initial concentrationf substrate was varied between 0.6 and 3.0 gCOD/L as is shownn Table 3. The final conversion in these experiments is alsoepicted in Table 3, calculated using Eq. (1). Again, the substrateoncentration decreased continuously with the degradation times can be expected. The values obtained for the final conversionre between 81 and 94%, indicating that most of the substrate

ed to the digester is biodegraded anaerobically.

The anaerobic digestion process can be also followed by theolume of methane formed during each experiment, which waseasured at regular times (see Fig. 4 for experiment N-4). There

Mf (mL) YM (mL/gCOD) kM × 10−2 (h−1) r2

325 295 6.7 0.997430 265 5.5 0.892890 270 3.1 0.951115 276 2.6 0.980303 245 1.4 0.915

844 J. Beltran et al. / Journal of Hazardous Materials 154 (2008) 839–845

Foi

wtatyi

Y

w

chiarHie

ptt

μ

wc

μ

e

−

w

−

Fs

wtl(

fii

V

l

eemeI1cti

mkp

k

wstt(

ig. 4. Evolution of the methane production with bioreaction time in the anaer-bic digestion. Experiment N-4 (r2 = 0.976). Experimental conditions are givenn Table 3.

as a high production of methane at initial times of bioreac-ion, while this formation decreased at higher times. Table 3lso shows the values of the volume of methane accumulated athe end of each experiment, VMf, and the methane productionield coefficient, YM (mL of methane/gCOD degraded), whichs defined in the form [28]

M = − dVM

dCOD= VMf

COD0XCODVR(13)

here VR is the volume of digester (2 L).As can be seen in Table 3, VMf increases when initial substrate

oncentration also increases, as could be expected. On the otherand, the methane production yield coefficient, YM, calculatedn each experiment is between 245 and 295 mL/g, with an aver-ge value of 270 mL/g. This value is higher than other valueseported by several authors in similar processes: 150 mL/g byamdi et al. [29] for olive oil mill wastewater, 187 mL/g by Ben-

tez et al. [30] for winery wastewater and 250 mL/g by Benitezt al. [31] for black table olive wastewater.

In any anaerobic process, the rate of production of biomass isroportional to the mass of cells present according to Eq. (1). Inhis case, Monod’s equation is the most used model to determinehe specific growth rate [32]

= μmaxS

KS + S(14)

here KS is the Monod saturation constant. For low substrateoncentrations, KS � S, Eq. (14) can be reduced to

= μmaxS

KS(15)

By introducing Eqs. (10) and (6) into Eq. (15), the followingquation is obtained

YX/S

dS

dt= μmax

KSXS (16)

hich can be transformed into

dS

S= μmax

YX/SKSX dt = kMdt (17)

ta

e

ig. 5. Kinetic study in the anaerobic digestion. Determination of kM rate con-tant. Experiment N-4 (r2 = 0.98). Experimental conditions are given in Table 3.

here the cellular concentration, X, has been consideredo be constant during the process, due to the well knownow cellular yield coefficient, YX/S, for anaerobic bacteria0.02–0.06 gVSS/gCOD) [20].

Now, from Eq. (13) for the methane production yield coef-cient, expressions for dS and S are obtained, which are

ntroduced into Eq. (17) leading to

dVM

VMf − VM= kM dt (18)

Therefore, its integration with the assumption that at t = 0,M = 0, the following expression is obtained [33,34]

n

(VMf

VMf − VM

)= kMt (19)

According to this equation, Fig. 5 shows the plot obtained forxperiment N-4 taken as example. A good arrangement of thexperimental points around a straight line confirms the agree-ent with the proposed model. A regression analysis to the

xperimental data results in kM values are provided in Table 3.t is observed that these constants decrease from 6.7 × 10−2 to.4 × 10−2 h−1 with the increase in the initial substrate con-entration. This can be attributed to an inhibition effect dueo the substrate present (i.e., polyphenolic compounds) or tontermediate products formed [35].

For the evaluation of this inhibition process, the Levenspielodel is used [23,30]. In this model the anaerobic rate constant

M is related to the concentration of polyphenolic compoundsresent in the reacting medium by means the expression

M = kM0

(1 − TP0

TP∗

)n

(20)

here kM0 is the anaerobic rate constant when no inhibitoryubstance is present; n is an inhibitory parameter related tohe inhibitory power of the substance; TP0 represents the ini-ial concentration of polyphenolic compounds in the experimentshown in Table 3); and TP* is the critical inhibition concentra-

ion of polyphenolic compounds, value when the cells cease theirctivity because of the inhibition process [36].

Applying a non-linear regression analysis to thexperimental results, the following results are obtained:

rdous

k0Eaa±

4

•

•

•

•

•

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

of effluents from the green olive debittering process, Appl. Microbiol.

J. Beltran et al. / Journal of Haza

M0 = 8.4 × 10−2 ± 0.25 h−1, n = 2.25 ± 3.47 and TP* = 0.34 ±.34 g/L (r2 = 0.966). In order to validate the robustness of theq. (20), the theoretical values of kM have been calculatednd compared with the experimental ones. There is a generalgreement between both values, with discrepancies within8%, which confirm the proposed model.

. Conclusions

From the results, it is possible to conclude that:

The aerobic biodegradation of green table olive wastewatersachieves a significative reduction of the COD between 50and 70%, and an important removal of the total polypheno-lic compounds around 97%, while the biomass follows thetypical growth cycle phases for batch cultivations.The application of Contois’ model to the experimental resultsleads to the evaluation of the biokinetic rate constant.In the anaerobic digestion, an important removal of CODbetween 81 and 94% is obtained. As well a mean value of themethane production yield coefficient of 270 mL/g is deduced.The use of a modified Monod’s model provides the valuesof the global kinetic constants which decrease with increas-ing initial substrate concentration in the range 1.4 × 10−2

to 6.7 × 10−2 h−1, indicating that some inhibition effects bysubstances present take place.For the evaluation of this inhibition process, the Levenspielmodel is used and its parameters are calculated.

eferences

[1] A. Garrido, A. Fernandez, R.M. Adam, Table Olives. Production and Pro-cessing, Chapman & Hall, London, 1997.

[2] G.C. Kopsidas, Wastewater from the table olive industry, Water Res. 28(1994) 201–205.

[3] R. Borja, A. Martin, A. Garrido, Anaerobic digestion of black-olivewastewater, Bioresour. Technol. 45 (1993) 27–32.

[4] R. Borja, C.J. Banks, A. Garrido, Kinetics of black-olive wastewater treat-ment by the activated-sludge system, Process Biochem. 29 (1994) 587–593.

[5] M. Brenes, P. Garcia, A. Garrido, Regeneration of Spanish style green tableolive brines by ultrafiltration, J. Food Sci. 53 (1988) 1733–1736.

[6] A. Garrido, P. Garcia, M. Brenes, The recycling of table olive brine usingultrafiltration and activated carbon adsorption, J. Food Eng. 17 (1992)291–305.

[7] F.J. Benitez, J.L. Acero, T. Gonzalez, J. Garcia, Application of ozone andadvanced oxidation processes to the treatment of lye-wastewaters from thetable olives industry, Ozone Sci. Eng. 24 (2002) 105–116.

[8] F.J. Benitez, J.L. Acero, A.I. Leal, Purification of storage brines from thepreservation of table olives, J. Hazard. Mater. B 96 (2003) 155–169.

[9] G.G. Aggelis, H.N. Gavala, G. Lyberatos, Combined and separate aerobicand anaerobic biotreatment of green olive debittering wastewater, J. Agric.Eng. Res. 80 (2001) 283–292.

10] A. de Castro, M. Brenes, Fermentation of washing waters of Spanish-stylegreen olive processing, Process Biochem. 36 (2001) 797–802.

11] M.H. Zwietering, I. Jongenburger, F.M. Rimbouts, K. Van Riet, Model-

ing of the bacterial growth curve, Appl. Environ. Microbiol. 56 (1990)1857–1881.

12] M. Pelillo, B. Rincon, F. Raposo, A. Martin, R. Borja, Mathematical mod-elling of the aerobic degradation of two-phase olive mill effluents in a batchreactor, Biochem. Eng. J. 30 (2006) 308–315.

[

Materials 154 (2008) 839–845 845

13] APHA, Standard Methods for the Examination of Water and Wastewater,16th ed., American Public Health Association, Washington, 1985.

14] J.D. Box, Investigations of the Folin-Ciocalteau phenol reagent for thedetermination of polyphenolic substances in natural waters, Water Res. 17(1983) 511–525.

15] J. Beltran, J. Torregrosa, T. Gonzalez, J.R. Dominguez, Analisis Quimicode Aguas Residuales, Universidad de Extremadura, Badajoz, 1994.

16] F.J. Benitez, J. Beltran, J. Torregrosa, J.L. Acero, V. Cercas, Aerobic degra-dation of olive mill wastewaters, Appl. Microbiol. Biotechnol. 47 (1997)185–188.

17] F.J. Benitez, J. Beltran, F.J. Real, J.L. Acero, Purification kinetics of winerywastes by ozonation, anaerobic digestion and ozonation plus anaerobicdigestion, J. Environ. Sci. Health A 34 (1999) 2023–2041.

18] J.E. Bailey, D.F. Ollis, Biochemical Engineering Fundamentals, McGraw-Hill, New York, 1985.

19] F.J. Benitez, J. Beltran, J. Torregrosa, J.L. Acero, Treatment of olive millwastewaters by ozonation, aerobic degradation and the combination of bothtreatments, J. Chem. Technol. Biotechnol. 74 (1999) 639–646.

20] R.S. Ramalho, Introduction to Wastewater Treatment Processes, AcademicPress, London, 1983.

21] S. Aiba, A.E. Humphrey, N.F. Millis, Biochemical Engineering, AcademicPress, New York, 1973.

22] D.E. Contois, Kinetics of bacterial growth: relationship between populationdensity and specific growth rate of continuous cultures, J. Gen. Microbiol.51 (1959) 808–814.

23] O. Levenspiel, The Chemical Reactor Omnibook, OSU Book Stores, Cor-vallis, 1979.

24] J. Beltran, J. Torregrosa, J.R. Dominguez, J. Garcia, Aerobic biologicaltreatment of black table olive washing wastewaters. Effect of an ozonationstage, Process Biochem. 35 (2000) 1183–1190.

25] J. Beltran, J. Torregrosa, J.R. Dominguez, J. Garcia, Ozonation ofblack-table-olive industrial wastewaters. Effect of an aerobic biologicalpre-treatment, J. Chem. Technol. Biotechnol. 75 (2000) 561–568.

26] F.J. Benitez, J. Beltran, F.J. Real, J.L. Acero, Enhancement of the ozonationof wine distillery wastewaters by an aerobic pretreatment, Bioprocess Eng.21 (1999) 459–464.

27] J. Beltran, J. Torregrosa, J.R. Dominguez, J. Garcia, Treatment of black-olive wastewaters by ozonation and aerobic biological degradation, WaterRes. 34 (2000) 154–161.

28] F.J. Benitez, J. Beltran, J. Torregrosa, J.L. Acero, Improvement of theanaerobic biodegradation of olive mill wastewaters by prior ozonationpretreatment, Bioprocess Eng. 17 (1997) 169–176.

29] M. Hamdi, C. Festino, C. Aubart, Anaerobic digestion of olive mill wastew-aters in fully mixed reactors and in fixed film reactors, Process Biochem.27 (1992) 37–42.

30] F.J. Benitez, J. Beltran, F.J. Real, T. Gonzalez, Aerobic and anaerobic purifi-cation of wine distillery wastewaters in batch reactors, Chem. Eng. Technol.22 (1999) 165–172.

31] F.J. Benitez, J.L. Acero, T. Gonzalez, J. Garcia, Organic matter removalfrom wastewaters of black olive industry by chemical and biological pro-cedures, Process Biochem. 37 (2001) 257–265.

32] J. Monod, The growth of bacterial cultures, Ann. Rev. Microbiol. 3 (1949)371–376.

33] R. Borja, A. Martin, L.F. Gomez, A. Ramos-Cormenzana, Anaerobic diges-tion of olive mill wastewater pretreated with Azotobacter chroococcum,Resour. Conserv. Recycling 9 (1993) 201–211.

34] R. Borja, A. Martin, M. Luque, M.M. Duran, Kinetic study of anaero-bic digestion of wine distillery wastewater, Process Biochem. 28 (1993)83–90.

35] G.G. Aggelis, C. Ehaliotis, F. Nerud, I. Stoychev, G. Lyberatos, G.I. Zer-vakis, Evaluation of white-rot fungi for detoxification and decolorization

Biotechnol. 59 (2002) 353–360.36] R. Borja, A. Martin, R. Maestro, J. Alba, J.M. Fiestas, Enhancement of the

anaerobic digestion of olive mill wastewater by the removal of phenolicinhibitors, Process Biochem. 27 (1992) 231–237.