preozonation of primary-treated municipal wastewater for reuse in biofuel feedstock generation

Preozonation of Primary-Treated Municipal Wastewaterfor Reuse in Biofuel FeedstockGenerationAndro H. Mondala,a Rafael Hernandez,a W. Todd French,a L. Antonio Estevez,b Mark Meckes,c

Marlene Trillo,b and Jacqueline Hallaa Renewable Fuels and Chemicals Laboratory, Dave C. Swalm School of Chemical Engineering, Mississippi State University,Mississippi State, MS 39762; [email protected] (for correspondence)b Department of Chemical Engineering, University of Puerto Rico - Mayaguez, Mayaguez, PR 00681c U.S. Environmental Protection Agency, Office of Research and Development, National Risk Management Research Laboratory,Cincinnati, OH 45268

Published online 9 November 2010 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/ep.10514

The results of a laboratory scale investigation onozone pretreatment of primary-treated municipalwastewater for potential reuse in fermentation proc-esses for the production of biofuels and bio-basedfeedstock chemicals were presented. Semi-batch preo-zonation with 3.0% (w/w) ozone at 1 L min21

resulted into a considerable inactivation of the indig-enous heterotrophic bacteria in the wastewater withless than 0.0002% comprising the ozone-resistantfraction of the microbial population. The disinfectionprocess was modeled using first-order inactivationkinetics with a rate constant of 4.39 3 1023 s21.Chemical oxygen demand (COD) levels were reducedby 30% in 1-h experiments. COD depletion was alsomodeled using a pseudo-first-order kinetics at a rateconstant of 9.50 3 1025 s21. Biological oxygendemand (BOD5) values were reduced by 60% up to20 min of ozonation followed by a plateau and someslight increases attributed to partial oxidation of re-calcitrant materials. Ozone also had no substantialeffect on the concentration of ammonium and phos-phate ions, which are essential for microbial growthand metabolism. Preliminary tests indicated that ole-aginous microorganisms could be cultivated in the

ozonated wastewater, resulting in relatively highercell densities than in raw wastewater and compara-ble results with autoclave-sterilized wastewater. Thisprocess could potentially produce significant quanti-ties of oil for biofuel production from municipalwastewater streams. � 2010 American Institute of Chemi-cal Engineers Environ Prog, 30: 666–674, 2011Keywords: ozone, municipal wastewater, kinetics,

biofuels

INTRODUCTION

Municipal wastewater treatment remains a daunt-ing task worldwide due to the rapid increase in pop-ulation and demand for industrial products resultinginto high wastewater volumes and treatment costs. Inthe United States alone, around 16,583 operationalwastewater facilities were treating a total of 34 billiongallons of municipal wastewaters daily within the last4 yr [1]. With this high volume, the total nationwidecost for wastewater treatment and collection was esti-mated to be $189.2 billion annually [2]. One pro-posed strategy to mitigate the environmental impactsand treatment costs of these high-volume wastewaterstreams and enhance environmental sustainability isthrough wastewater reclamation and reuse for a vari-� 2010 American Institute of Chemical Engineers

666 December 2011 Environmental Progress & Sustainable Energy (Vol.30, No.4) DOI 10.1002/ep

ety of applications. Some examples include industrialcooling and washing water, nonpotable domesticwater, irrigation in urban areas (gardens, parks, golfcourse, etc.), restoration of natural water courses [3],and irrigation for agriculture activities [4]. Anotherpotential application that is under investigation byour group is to reuse municipal wastewaters as culti-vation media for microorganisms involved in the pro-duction of biofuels and/or biobased fuel feedstocks.In addition to producing high-value products, thisprocess could degrade the organic pollutants in thewastewater that make up its biochemical oxygendemand (BOD5) as well as its nutrient (N and P poly-ionic species such as nitrate, nitrite, ammonium, andphosphate) levels. Examples of these microorganismsare bioethanol-producing yeasts, and more recently,oleaginous yeasts, bacteria, and algae capable ofaccumulating lipids that could be used for the pro-duction of biodiesel [5].

However, as with all municipal wastewater reutili-zation projects, certain pretreatment schemes must beimplemented, particularly in reducing the pathogenicrisk and microbial toxicity of these wastewaterstreams [3]. For reusing primary-treated (i.e. screen-ing, grit removal, sedimentation) wastewater in fer-mentation processes involving specific microbialstrains, the main pretreatment objectives are to elimi-nate the indigenous microbial population in thewastewater that may compete with the pure strainsfor substrates in the media, improve wastewater qual-ity and biodegradability, and reduce toxicity. Amongthe various water and wastewater treatment methodsavailable, ozone has been selected for this studybecause of its well-documented strong disinfectingpower against a wide variety of bacteria, viruses, andother pathogenic organisms [6–9]. Compared withchlorination, ozone pretreatment requires shortercontact times, produces less harmful residuals due toits rapid decomposition, and could be generatedonsite, resulting into fewer shipping and handlingproblems. When compared with ultraviolet (UV) radi-ation, ozonation results into less regrowth of inacti-vated microorganisms after the treatment process [8].Moreover, ozone has also been shown to reduce theBOD5, chemical oxygen demand (COD), total organiccarbon (TOC), total solids, turbidity, and improve thecolor of domestic wastewater because of its strongoxidizing power, hence improving the quality of thewastewater stream [6]. In most cases, ozone has alsobeen shown to improve the biodegradability ofdomestic wastewaters by oxidizing recalcitrant materi-als in the wastewater into more biodegradable matterthus decreasing the COD while the BOD5 remainedconstant or increased slightly [6, 10–12]. Dissolved ox-ygen levels in wastewaters have also been shown toincrease to saturation levels following ozonation sinceozone molecules are relatively unstable and decom-pose to oxygen quickly in aqueous media [13]. Thisrobustness and versatility of ozone in reacting withvirtually every component of the wastewater can beattributed to two different routes through whichozone reacts with organic and inorganic compoundsin the wastewater: by direct molecular attack, or by

the action of hydroxyl radicals which are producedby the decomposition of ozone molecules in aqueoussolutions. The latter mechanism has been shown tobe more unselective with rate constants ranging from108 to 1010 M21 s21 [14]. Although capital and treat-ment costs and power requirements for ozonation arerelatively higher than those of chlorination and UV,the benefits appear to outweigh the extra costs,which will eventually decrease with continuedresearch to improve current technologies [6].

Given the perceived benefits of ozone, the objec-tive of the current study was to determine the feasi-bility of applying ozone as a pretreatment step forprimary-treated municipal wastewater for potentialreutilization as fermentation media for oil-producingoleaginous microorganisms. Semibatch ozonationexperiments were conducted to determine the effectof ozone on heterotrophic bacteria that are indige-nous to the municipal wastewater. Furthermore, ozo-nation effects on BOD5, COD, and nutrient ion (N-NH4

1, N-NO32, N-NO2

2, P-PO432) concentrations were

investigated as these parameters represent the sub-strates available in wastewater for consumption bymicroorganisms for the production of biofuels. Ozonedisinfection and COD oxidation kinetic models weretested to obtain kinetic parameters necessary for pro-cess scale-up and design. Afterwards, test cultivationruns using an oleaginous microorganism consortiumdeveloped at the Renewable Fuels and ChemicalsLaboratory-Mississippi State University on ozonatedwastewater media were conducted to determine theeffectiveness of the ozonation process for the pro-posed reutilization scheme.

MATERIALS AND METHODS

Wastewater Sampling and CharacterizationThe primary-treated wastewaters tested in this

study were obtained from a municipal wastewatertreatment plant in Tuscaloosa, AL. Grab samples werecollected from each of three operational primaryclarifiers in the plant and stored in polyethylene bot-tles. Care was taken to ensure that representativesamples were obtained and used in the ozonationexperiments. Samples were taken at designated areasaround the primary clarifiers at the same day of theweek and time of the day throughout the duration ofthe study. The wastewater samples were then trans-ported under ice to the laboratory and used in theozonation experiments within 2 h of arrival to mini-mize changes in the microbial and physicochemicalcharacteristics of the wastewater.

The raw and ozonated wastewater samples werecharacterized for BOD5 and COD according toAWWA/APHA Standard Methods 5210B and 5220D,respectively [15]. Heterotrophic bacteria were enum-erated by a membrane filtration method according toAWWA/APHA Standard Method 9215D. Serial dilu-tions were prepared using sterile 0.85% (w/v) NaClsolution. Triplicate 10 mL dilutions of the primaryeffluent samples were then filtered through sterile 47-mm diameter nitrocellulose membrane filters with a0.45-lm mean pore size (Fisherbrand, Pittsburgh, PA)

Environmental Progress & Sustainable Energy (Vol.30, No.4) DOI 10.1002/ep December 2011 667

using a vacuum filtration assembly (Nalgene Inc.,Rochester, NY). The used filters were placed asepti-cally in plates containing commercially availableDIFCO Plate Count Agar (Fisher Scientific, Pittsburgh,PA) containing (in gL21): pancreatic digest of casein5.0, yeast extract 2.5, dextrose 1.0, and agar 15. Theplates were then incubated at 22 6 28C for 5 days.

The concentrations of nitrite, nitrate, ammonium,and phosphate were determined using two ion chro-matography methods for the analysis of inorganic cat-ions and anions in water and wastewater describedby Dionex Corporation [16, 17]. The equipment usedwas an ICS-3000 ion chromatograph (Dionex Corp.,Sunnyvale, CA) equipped with a 250 3 4-mm IonPacAS11-HC anion exchange analytical column and 50 34 mm AG11 guard column for nitrite, nitrate, andphosphate analysis; a 250 3 4 mm IonPac CS16 cat-ion exchange analytical column and 50 3 4 mmCG16 guard column for ammonium analysis, and twosuppressed conductivity detectors. Before analysis,the samples were pretreated by passing them throughC18 Sep-Pak cartridges (Waters Corp., Milford, MA)that were preconditioned with 5 mL of methanol and5 mL of Type II ASTM water (18.2 MX-cm resistivity).The samples were then filtered through 0.45-lm sy-ringe filters (Millipore, Billerica, MA). The instrumentdetection limits were 0.01 mg L21 for nitrite and ni-trate, 0.02 mg L21 for phosphate, and 0.01 mg L21 forammonium. Table 1 shows the average values andstandard deviations of the wastewater characteristicsof primary clarifier effluent grab samples obtained atdifferent times for the performance of the ozonationexperiments. Chemical analyses indicated no signifi-cant variations in wastewater characteristics amonggrab samples obtained at the same time.

Ozone Generation and MonitoringOzone was produced from compressed dry air by

corona discharge using an Ozonology LC-1234 ozonegenerator (Ozonology, Inc., Northbrook, IL). Theozone concentrations in the inlet and outlet gasstreams were monitored using a UV spectrophotome-ter ozone detector (PCI Ozone and Control Systems,West Caldwell, NJ). The residual ozone in the waste-water samples was measured using commerciallyavailable Ozone CHEMets Test Kits (CHEMetrics Inc.,Calverton, VA).

Ozonation ExperimentsEqual amounts of primary-treated wastewater grab

samples were mixed together to create a homoge-nous mixture, which will now be referred to as thetest wastewater. One liter of the test wastewater wasthen transferred into a 1-L glass reactor (Chemglass,Vineland, NJ) equipped with a gas diffuser and mixer.Gas containing 3.0% (w/w) ozone was bubbledthrough the test wastewater continuously at 1 6 0.05L min21 for 60 min at room temperature with an agi-tation rate of 100 rpm Samples were withdrawn fromthe reactor at different times during the ozonationexperiment and analyzed for residual ozone concen-tration, BOD5, COD, heterotrophic bacteria popula-tion, and N-NH4

1, N-NO32, N-NO2

2, and P-PO432 con-

centrations. Prior to analysis, the residual ozone inthe samples was neutralized with 0.1% (w/v) Na2S2O3

solution. The exhaust gas from the reactor waspassed through a glass column packed with Carulite�200 manganese dioxide/copper oxide catalyst (CarusCorp., Peru, IL) to destroy any remaining ozonebefore being released into the atmosphere.

Microbial Cultivation Experiments on OzonatedWastewater

The ozonated test wastewaters were then tested aspotential growth media for a consortium of oleagi-nous microorganisms (OM) developed at MSU-RFCL.The consortium contains 12 yeast strains and 1 bacte-ria that have well-documented lipid-accumulatingabilities at certain media conditions. Eight hundredmilliliters of ozonated wastewater contained in 2-Lshake flasks with (a) no glucose supplement andamended with (b) 1 g L21 or (c)10 g L21 glucosewere inoculated with 30 mL of the OM culture andincubated at 278C with shaking at 110 rpm Each treat-ment was done in triplicates. Samples of the cultureswere obtained at daily intervals and were measuredfor optical density at a wavelength of 600 nm using aGenesys 20 Spectrophotometer (Thermo Scientific,Waltham, MA). The level of microbial growth in theozonated wastewater cultures was then correlatedwith the optical density measurements. For compari-son, cultivation runs involving raw and autoclaved(1218C, 240 kPa, 15 min) primary effluent wastewaterwere also conducted using the same conditions.

RESULTS AND DISCUSSION

Disinfection KineticsFigure 1 shows the average inactivation trend of

the indigenous heterotrophic bacteria in the munici-pal wastewater primary effluent using ozone. Thedata were presented as typical semi-log plots of thefraction of surviving microorganisms, N/N0 as a func-tion of the transferred ozone dose (TOD); where N0

and N are the number of viable microorganisms ini-tially and at contact time t, respectively. The TODvalues are obtained by calculating the differencebetween the ozone dose supplied and the amount ofozone leaving the reactor at different contact timeintervals and represent the amount of ozone utilized

Table 1. Characteristics of the primary-treatedmunicipal wastewater used in this study.

pH 7.0 6 0.1COD (mg L21) 396 6 94BOD5 (mg L21) 133 6 48N-NH4

1 (mg L21) 22 6 3N-NO3

2 (mg L21) N.D.N-NO2

2 (mg L21) 4.2 6 0.9P-PO4

32 (mg L21) 8.9 6 1.1Heterotrophic bacteria (3107 CFU mL21) 3.3 6 2.0

ND, not detected.


for disinfection and oxidation. The plot shows thatthe general inactivation trend follows a multiphasiccurve consisting of an initial shoulder followed by ashort exponential drop and a tailing-off patterntowards the end of the run. The initial shoulderwhich represents a low initial disinfection rate thatsubsequently increased may be attributed to severalfactors such as delays in ozone diffusion and/or mul-tiple oxidation targets such as the pre-existing oxidiz-able material in the wastewater. On the other hand,the tail may be caused by aggregation of microbialcells, cellular debris, or suspended solids that mayproduce a shielding effect from the effects of ozone,or microbial subpopulations with varying resistancesto ozone inactivation [18]. The latter is expected sincethe heterotrophic bacterial population in the waste-water is thought to consist of different bacterialstrains. This same pattern has been found in previousstudies involving the disinfection of heterotrophicbacteria from water where a major portion of the het-erotrophic bacteria was rapidly inactivated but therealways remains an ozone-resistant subpopulation,hence the tailing-off pattern [18]. In another study,the existence of the tail was explained as a result ofother competing reactions in the wastewater matrixwith ozone which are favored over microbial inacti-vation at that extent of the ozonation process [19].Despite this, the remaining ozone-resistant microbialpopulation represents on average less than 0.0002%of the initial heterotrophic bacteria counts.

Several previous studies using ozone as a water/wastewater disinfectant have noted the difficulty andcomplexity of developing kinetic models that accu-rately describe inactivation kinetics using ozone.Models based on reaction mechanisms are particu-larly susceptible to error due to the complexity ofmulticomponent matrices like wastewater, unstableresidual ozone due to ozone autodecomposition and/or competing reactions with other organic and inor-ganic compounds in the treatment matrix, and rapidinactivation rates [14, 20–22]. For this reasons, the use

of empirical models based on kinetic rate laws aremore widely used. In this study, two exploratorymodels were used based on the following direct reac-tion of ozone with microbial cells:

O3 þ N�! products (1)

where N represents the number of viable heterotro-phic bacterial cells in colony-forming units (CFU) permL of wastewater. The first model expresses the rateof microbial inactivation as

� dN

dt¼ kNNCO3

(2)

where kN is the disinfection rate constant and CO3is

the dissolved ozone concentration. Since the averageinitial pH of the wastewater samples were close toneutral (see Table 1) and was not significantlyaffected by ozonation (pH range 6.78–6.97), it wasassumed that hydroxide ion concentrations were lowso that ozone oxidation caused by the free radicalmechanism was negligible [23]. It was also assumedthat O3 is present in large excess with respect to theinitial number of viable microorganisms; thereforeEq. 1 is simplified into a pseudo-first-order rate equa-tion or more commonly known as Chick’s ratelaw [18]

� dN

dt¼ k�

NN (3)

where k*N is the pseudo-first-order inactivation rateconstant equal to kNCO3

. Integration of Eq. 3 leads tothe following equation:

lnN

N0¼ �k�N t (4)

The values of the parameters were estimated usingthe SOLVER optimization package in MicrosoftEXCEL�. The goodness-of-fit of the proposed modelwith the experimental data were measured by calcu-lating the correlation coefficient:

R2 ¼ 1�P

yi � yp� �2

Pyi � �yi

8:

9;

2 (5)

where yi represents an observed value and yp repre-sents the model predictions.

To determine the apparent order of ozone inacti-vation, the simple power rate law was proposed as asecond model:

� dN

dt¼ kNN

m (6)

The experimental data were then fit into a nondimen-sionalized integral of Eq. 6 given as

Figure 1. Survival ratios of heterotrophic bacteria inprimary-treated wastewater (in semi-log scale) as afunction of transferred ozone dose.


N

N0¼ 1

ð1þ KNm�10 tÞ1=ðm�1Þ (7)

where K ¼ m�1ð Þk�NNm�10

¼ m� 1ð ÞkN . The resulting modelfit with the pooled experimental data is shown inFigure 2 and the calculated disinfection kinetic param-eters are summarized in Table 2. As shown in Table 2,the nonlinear regression fit (R2) 5 0.77 of the pooledexperimental data with the pseudo-first-order relation-ship described by Eq. 4 resulted into a calculated reac-tion rate constant k*N 5 4.39 3 1023 s21. When thepower law model was used, the calculated apparentreaction order is close to the first order (m 5 1.09)while the apparent reaction rate constant obtained isk*N 5 7.37 3 1023 s21, which represents the slope ofthe curve at t 5 0. The power model also resultedinto a better correlation coefficient (R2 5 0.84) thanthe pseudo-first order approach.

Effects on BOD5 and CODThe trends of COD, BOD5, and BOD5/COD of the

test wastewater as a function of ozone contact timeare shown in Figure 3 as a ratio of the values of thesaid parameters at time t and the initial values. Theaverage COD reduction data from three representa-tive runs show an almost linear reduction trend inthe total COD of the wastewater. The maximum CODreduction level reached was approximately 30%. Onthe other hand, the BOD5 reduction rates were fastestwithin the first 20 min of ozonation and reached amaximum reduction of approximately 60%. Beyondthis point, the BOD5 levels remained fairly constantand in some instances increased slightly. This obser-vation was consistent with numerous previous waste-water ozonation studies [6, 10, 11, 13, 24] in whichthe authors attributed the slight increases in theBOD5 of the wastewater to the partial oxidation of re-calcitrant materials into a biodegradable fraction. Thisphenomenon has been taken advantage of in otherstudies, particularly those in which the goals of ozo-

nation were to improve wastewater biodegradabilityand removal of biorefractory organics such as aro-matics and unsaturated compounds to assist in thesubsequent biological oxidation using the activatedsludge process [25]. Furthermore, the BOD5/COD ra-tio, which is generally used as an indication of thebiodegradability level of wastewater streams was alsoshown to follow a trend similar to that of BOD5. Arelatively high BOD5/COD indicates a more biode-gradable wastewater stream and vice versa. Althoughthe BOD5/COD decreased significantly during thefirst 20 min of ozonation, it appeared to level off untilthe end of ozonation run.

The total COD concentration has usually beenadopted as a parameter to indicate the overall concen-tration of ozone-reacting compounds since it includesboth organic and inorganic materials that constitutewastewaters, including microbial cells [11, 26]. LowCOD levels usually indicate the presence of low mo-lecular weight carboxylic acids such as acetic, oxalic,and maleic acid, which are considered nontoxic andmore susceptible to biological degradation [14]. Hence,the kinetics of COD oxidation was also consideredusing a pseudo-first-order kinetic model similar to Eq.3, assuming that ozone is present in excess

� dCOD

dt¼ k�CODCOD (8)

Integrating Eq. 8 leads to the following pseudo-first-order relationship

Figure 2. Ozone disinfection kinetic model testing.

Table 2. Parameter estimates for the proposed ozonedisinfection kinetic models.

Model kN* (s21) m R2

Pseudo-first order, Eq. 4 4.39 3 1023 N/A 0.77Power law, Eq. 7 7.37 3 1023 1.09 0.84

Figure 3. Variation of BOD5, COD, and BOD5/CODratio with ozone contact time.


lnCOD

COD0¼ �k�CODt (9)

where k*COD is the apparent COD oxidation rate con-stant. Figure 4 shows that the linear regression fit of theexperimental COD reduction data resulted in a satisfac-tory correlation (R2 5 0.70) and a calculated rate con-stant of 9.5 3 1025 s21, which is much smaller than thedisinfection rate constants. Hence, it can be inferredthat microbial inactivation kinetics was favored overCOD reduction. It may also be considered an artifact ofthe transformation of recalcitrant COD to BOD, whichis also encompassed by the total COD of the waste-water. This may have resulted into an apparent reduc-tion of the rate of total COD oxidation.

Effects on Nutrient Ion ConcentrationThe concentrations of ammonium and phosphate

remained fairly constant throughout the ozonationexperiment with minor fluctuations while nitrite levelsincreased slightly as shown in Figure 5. Theoretically,both nitrite and ammonia (which exists in equilibriumwith ammonium) are oxidizable by ozone to producenitrate in the wastewater [27]. Nitrate was not detectedinitially in the primary effluent but after 20 min of ozo-nation, it was detected by the IC and increased up to aconcentration of 5 mg L21. These observations wereconsistent with previous studies that showed ozonationdid not significantly reduce the concentration of nitro-gen species in the wastewater since it appeared thatozone preferentially oxidizes carbonaceous materialsrather than nitrogenous ones [6]. Only very high ozonedoses have been shown to significantly oxidize nitrogencompounds into nitrate [28]. It is also possible that theoxidation of microbial cells and other organic sus-pended solids in the wastewater by ozone could havereleased these ions in the wastewater matrix to offsetthose that have been oxidized.

Ozone Consumption and Residual FormationReactions between ozone and oxidizable matter

in the wastewater matrix are typically fast in addi-

tion to ozone autodecomposition so that in simplebatch ozonation processes, the ozone supplied isquickly consumed and any residual ozone is usuallyundetected [22]. To produce a measurable residualozone concentration in the wastewater matrix,which is usually required in most ozone oxidationand disinfection kinetic models, semi-batch experi-ments are performed instead. In this study, a com-pressed, dry-air stream containing 3.0% (w/w) ozonewas supplied continuously at approximately 1.0 Lmin21 into single batches of the test wastewater.This corresponds to an ozone feeding rate of 36 mgL21�min or a total ozone dose of 2160 mg L21 after60 min contact time. After taking the differencebetween the ozone dose supplied and the amountof ozone leaving the reactor at different contacttime intervals, the transferred ozone doses (TOD)or the amounts of ozone consumed by the waste-water matrix were obtained. On average, theTODs make up approximately 42% of the totalozone supplied. It should also be noted that com-pared with previous studies dealing with ozonation

Figure 4. COD oxidation kinetic model testing.Figure 5. Effect of ozone contact time on the nutriention composition of primary-treated wastewater.

Figure 6. Residual ozone formation and total ozonedose as a function of contact time.


of less polluted secondary effluents, the TOD lev-els used in this study were relatively higher dueto higher ozone demands associated with higherCOD levels and suspended solids concentration inprimary-treated wastewater.

Residual ozone levels (Figure 6) were initiallylow in the first 20 min of ozonation after which itincreased substantially at the 20 to 30 min ozona-tion period. From 30 min onwards, a constantozone residual of approximately 2.8 mg L21 wasobtained. A constant ozone residual is usuallythought of as an indication of complete oxidationof oxidizable materials in the wastewater matrix byozone. On the other hand, McCarthy and Smith [6]instead define the ozone residual as the 98% air oroxygen that makes up the bulk of the gas mixturethat was supplied into the reactor. According tothem, the continuous supply of ozone supersatu-rates the wastewater with oxygen, which originatedfrom the feed gas and from the spontaneous decayof ozone molecules to oxygen.

To verify whether or not the observed ozone re-sidual represents the saturation level, the saturationconcentration of ozone at the conditions of the ozo-

nation experiments (208C and pH 7) was calculatedusing Henry’s law:

C�O3

¼ PO3

H(10)

where C *O3is the saturation concentration of ozone,

PO3is the partial pressure of ozone, and H is the

Henry’s coefficient, which for ozone was estimated tobe 10.3 3 106 Pa dm3 mol21 at 208C and pH 7 byKuosa et al. [29]. The calculated ozone saturationconcentration (8.5 mg L21) however, is relativelylarger than the observed ozone residual (2.8 mg L21).It is possible that a kinetic steady state was achievedat 30 min wherein the rate of ozone addition is thesame as the rate of ozone consumption or decay.

Interestingly, the TOD continuously increased withozone contact time (Figure 6) despite the ozone resid-ual plateau and apparent reduction of the rates of mi-crobial inactivation and BOD5/COD degradation asshown in their respective plots. This implies thatozone is still being consumed by the wastewater ma-trix. As mentioned earlier, the pH of the wastewaterwas virtually unaffected by ozone and remained closeto neutral therefore the rate of ozone autodecomposi-

Figure 7. Test cultivation runs using MSU-RFCL oleaginous microorganism consortium on ozonated, raw, andautoclaved primary-treated wastewater. (a) No glucose supplement, (b) 1 g L21, and (c) 10 q L21 glucose sup-plement.


tion via the hydroxyl radical is relatively slow with apseudo first-order rate constant k < 1024 s21 [27].Although not explicitly quantified, the majority of theTOD might have been utilized for the degradation ofrefractory compounds comprising the COD albeit at aslower rate, since approximately 75% of initial levelsstill remained even after 60 min of ozone contact. Theozone supplied could have also reacted with inorganiccompounds in the wastewater that were not quantifiedas well such as H2S/HS

- and SO322, with rate constants

in the range of 108–1010 M21�s21 at pH 6–8 [27].

Assessment of Ozonated Primary Effluents asCultivation Media

Figure 7 shows the trend in the growth of theoleaginous microorganism consortium in ozonatedprimary-treated wastewater compared with raw andautoclaved wastewater with and without glucose sup-plement. In Figure 7a, it can be seen that there is atwo-fold increase in the optical density of the cultureusing ozonated wastewater whereas in autoclavedwastewater, no significant change was observed. Inraw wastewater, a four-fold reduction in the opticaldensity was observed. However, the optical densitylevels obtained in this setup were too low to make adefinitive comparison in the cell growth. This mightbe due to the low BOD5 levels in the wastewater bothbefore and after ozonation. Hence, glucose was usedto supplement the residual BOD5 for use as an addi-tional substrate for microbial growth. As shown in Fig-ures 7b (1 g L21 glucose added) and 7c (10 g L21 glu-cose added), the oleaginous consortium exhibited bet-ter growth in ozonated as opposed to raw wastewater,resulting in six- and four-fold increases, respectively.In raw wastewater, an initial increase in OD wasobserved in 24 h followed by a decrease up to 72 h,after which the OD appeared to increase again. Thiscould indicate competitive growth kinetics betweenthe oleaginous microorganisms and the indigenousmicrobes in the wastewater. Based on these observa-tions, preozonation of the wastewater resulted in theelimination of competing microorganisms for sub-strates, giving way for the oleaginous consortium toproliferate more freely in the wastewater. Another pos-sible factor is the presence of more readily biodegrad-able matter as a result of ozone oxidation of recalci-trance in the wastewater. Moreover, it can also beseen in that the growth of oleaginous microorganismsin ozonated wastewater was comparable with that inautoclave-sterilized wastewater containing 1 g L21 glu-cose supplement and slightly higher with 10 g L21 glu-cose. This could be attributed to the fact that ozonedecomposition in the wastewater resulted into anincreased dissolved oxygen levels which would befavorable for the aerobic growth of the oleaginousmicroorganisms whereas heat sterilization by means ofthe autoclave could have released most of the dis-solved oxygen in the wastewater.

CONCLUSIONS

This study has demonstrated that ozone is a robustpretreatment process for primary-treated municipal

wastewater before its application in fermentativeprocesses for biofuel feedstock generation. After 60min of semibatch ozonation with 3.0% (w/w) ozoneat 1 L min21, less than 0.0002% of the initial hetero-trophic bacteria population remained in a primary-treated wastewater effluent. COD levels were alsoreduced by 30% while BOD5 levels remained con-stant and increased slightly following a 60% reductionwithin the first 30 min of ozone contact due to theconversion of recalcitrant compounds to biodegrad-able materials that could be readily utilized by micro-organisms as substrates. Ammonium and phosphateconcentrations were also unaffected by ozonation;hence, this could reduce the necessity to supplementthe wastewater with nutrient salts to induce microbialgrowth. Cultivation experiments have demonstratedthat oleaginous microorganisms could be grown inozonated wastewater to produce cell density levelscomparable to growth in autoclaved-sterilized waste-water, and better than in raw wastewater. Therefore,ozone pretreatment was effective in eliminatingmicrobial competitors for nutrients and substrateswithout adversely affecting the biodegradability andnutrient content of the wastewater to give way forthe cultivation of specialized microorganisms capableof producing large quantities of oil. This processcould then potentially generate millions of gallons ofrenewable and sustainable feedstock for biofuel pro-duction using an established infrastructure.

ACKNOWLEDGMENTS

This article was developed under CooperativeAgreement No. CR-83361801 awarded by the U.S.Environmental Protection Agency. EPA made com-ments and suggestions on the document intended toimprove the scientific analysis and technical accuracyof the document. However, the views expressed inthis document are those of Mississippi State Univer-sity and EPA does not endorse any products or com-mercial services mentioned in this publication.

LITERATURE CITED

1. U.S. Environmental Protection Agency. (2008).Clean watersheds needs survey 2004 report toCongress Appendix C. Available at: http://www.epa.gov/cwns/2004rtc/appendixa.pdf.Last accessed August 3, 2010.

2. U.S. Environmental Protection Agency. (2008).Clean watersheds needs survey 2004 report to Con-gress executive summary. Available at: http://www.epa.gov/cwns/2004rtc/exesumm.pdf. Lastaccessed August 3, 2010.

3. Cao, N., Yang, M., Zhang, Y., Jianying, H., Ike,M., Hirotsuji, J., Matsui, H., Inoue, D., & Sei, K.(2009). Evaluation of wastewater reclamationtechnologies based on in vitro and in vivo bioas-says, Science of the Total Environment, 407,1588–1597.

4. de Koning, J., Bixio, D., Karabelas, A., Saglot, M.,& Schaefer, A. (2008). Characterisation and assess-ment of water treatment technologies for reuse,Desalination, 218, 92–104.


5. Li, Q., Du, W., & Liu, D. (2008). Perspectives ofmicrobial oils for biodiesel production, AppliedMicrobiology and Biotechnology, 80, 749–756.

6. McCarthy, J.J., & Smith, C.H. (1974). A review ofozone and its application to domestic wastewatertreatment, Journal AWWA, 66, 718–725.

7. Beltran, F.J., Garcia-Araya, J.F., & Alvarez, P.M.(2001). Domestic wastewater ozonation: A kineticmodel approach, Ozone Science & Engineering,23, 219–228.

8. U.S. Environmental Protection Agency. (1999).Wastewater technology fact sheet ozone disinfec-tion. Available at: http://h2o.enr.state.nc.us/tacu/documents/ozonedis.pdf. Last accessedAugust 3, 2010.

9. Paraskeva, P., & Graham, N.J.D. (2002). Ozona-tion of municipal wastewater effluents, WaterEnvironment Research, 74, 569–581.

10. Beltran, F.J., Garcia-Araya, J.F., & Alvarez, P.M.(1997). Impact of chemical oxidation on biologi-cal treatment of a primary municipal wastewater.1. Effects on COD and biodegradability, OzoneScience & Engineering, 19, 495–512.

11. Alvares, A.B.C., Diaper, C., & Parsons, S.A.(2001). Partial oxidation by ozone to removerecalcitrance from wastewaters - a review, Envi-ronmental Technology, 22, 409–427.

12. Hagman, M., Tykesson, E., Hjorth, B., & Jansen,J.L.C. (2007). Oxygen uptake rate measurementsfor evaluation of ozonation of municipal waste-water, Environmental Technology, 28, 177–183.

13. Paraskeva, P., Lambert, S.D., & Graham, N.J.D.(1998). Influence of ozonation conditions on thetreatability of secondary effluents, Ozone Science& Engineering, 20, 133–150.

14. Rivas, J., Gimeno, O., & Beltran, F.J. (2009).Wastewater recycling: Application of ozone basedtreatments to secondary effluents, Chemosphere,74, 854–859.

15. APHA, AWWA, & WEF (1998). Standard methodsfor the examination of water and wastewater(20th Edition), Washington, D.C.: American Pub-lic Health Association.

16. DIONEX Corporation. (2001). Application note 135:Determination of inorganic anions in wastewater byion chromatography. Available at: http://www.dionex.com/en-us/webdocs/4085_AN135_V13.pdf.Last accessed August 17, 2009.

17. DIONEX Corporation. (2001). Application note141: Determination of inorganic cations andammonium in environmental waters by ion chor-matography using the IonPac CS16 column.Available at: http://www.dionex.com/en-us/

webdocs/4211_AN141_V15.pdf. Last accessedAugust 17, 2009.

18. Gyurek, L.L., & Finch, G.R. (1998). Modeling watertreatmment chemical disinfection kinetics, Journalof Environmental Engineering, 124, 783–793.

19. Finch, G.R., & Smith, D.W. (1989). Ozone dose-response of Escherichia coli in activated sludgeeffluent, Water Research, 23, 1017–1025.

20. Zhou, H., & Smith, D.W. (1994). Kinetics ofozone disinfection in completely mixed system,Journal of Environmental Engineering, 120, 841–858.

21. Zhou, H., & Smith, D.W. (1995). Evaluation of pa-rameter estimation methods for ozone disinfec-tion kinetics, Water Research, 29, 679–686.

22. Hunt, N.K., & Marinas, B.J. (1997). Kinetics ofEscherichia coli inactivation with ozone, WaterResearch, 31, 1355–1362.

23. Ko, C.H., Hsieh, P.H., Chang, M.W., Chern, J.M.,Chiang, S.M., & Tzeng, C.J. (2009). Kinetics ofpulp mill effluent treatment by ozone-based proc-esses, Journal of Hazardous Materials, 168, 875–881.

24. Nebel, C. (1973). Ozone disinfection of industrial-municipal secondary effluent, Journal WPCF, 45,2493–2507.

25. Beltran, F.J., Garcia-Araya, J.F., & Alvarez, P.M.(1999). Integradtion of continuous biological andchemical (ozone) treatment of domestic waste-water. 2. Ozonation followed by biological oxida-tion, Journal of Chemical Technology & Biotech-nology, 74, 884–890.

26. Preist, S., Munter, R., & Siirde, E. (1988). Kineticdescription of industrial wastewater ozonationprocesses, Ozone Science & Engineering, 10,953–957.

27. Balbon, G., Bellamy, W., Bourbigot, M., Daniel,F., Dore, M., Erb, F., Gordon, G., Langlais, B.,Laplanche, A., Legube, B., Martin, G., Massche-lein, W., Pacey, G., Reckhow, D., & Ventresque,C. 1991. Fundamental aspects. In B. Langlais, D.Reckhow, & D. Brink (Eds.), Ozone in watertreatment: Application and engineering (pp. 11–132), Chelsea, MI: Lewis Publishers.

28. Nebel, C., Unangst, P.C., & Gottschling, R.D.(1973). An evaluation of various mixing devicesfor dispersing ozone in water, Water and SewageWorks, 120, 6–12.

29. Kuosa, M., Laari, A., & Kallas, J. (2004). Determi-nation of the Henry’s coefficient and mass trans-fer for ozone in a bubble column at different pHvalues of water, Ozone Science & Engineering,26, 277–286.


preozonation of primary-treated municipal wastewater for reuse in biofuel feedstock generation

Documents