Development of a Submerged Membrane Bioreactor simulator: a useful tool for teaching its functioning

HAL Id: hal-01069368

https://hal.archives-ouvertes.fr/hal-01069368

Submitted on 29 Sep 2014

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of

sci-entific research documents, whether they are

pub-lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diffusion de documents

scientifiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

Development of a Submerged Membrane Bioreactor

simulator: a useful tool for teaching its functioning

Yusmel González Hernández, Ulises Javier Jáuregui Haza, Claire Albasi,

Marion Alliet-Gaubert

To cite this version:

Yusmel González Hernández, Ulises Javier Jáuregui Haza, Claire Albasi, Marion Alliet-Gaubert.

De-velopment of a Submerged Membrane Bioreactor simulator: a useful tool for teaching its functioning.

Education for Chemical Engineers, Elsevier, 2014, vol. 9, pp. e32-e41. �10.1016/j.ece.2014.03.001�.

�hal-01069368�

Any correspondence concerning this service should be sent to the repository administrator:

staff-oatao@inp-toulouse.fr

Identification number: DOI : 10.1016/j.ece.2014.03.001

Official URL:

http://dx.doi.org/10.1016/j.ece.2014.03.001

This is an author-deposited version published in:

http://oatao.univ-toulouse.fr/

Eprints ID: 11427

To cite this version:

González Hernández, Yusmel and Jáuregui Haza, Ulises Javier and Albasi,

Claire and Alliet-Gaubert, Marion Development of a Submerged Membrane

Bioreactor simulator: a useful tool for teaching its functioning. (2014)

Education for Chemical Engineers, vol. 9 (n° 2). pp. e32-e41. ISSN 1749-7728

O

pen

A

rchive

T

oulouse

A

rchive

O

uverte (

OATAO

)

OATAO is an open access repository that collects the work of Toulouse researchers and

makes it freely available over the web where possible.

Development

of

a

Submerged

Membrane

Bioreactor

simulator:

a

useful

tool

for

teaching

its

functioning

Yusmel

González

Hernández

_,

_Ulises

_Javier

_Jáuregui

_Haza

_,

Claire

Albasi

,

Marion

Alliet

a_{InstitutoSuperiordeTecnologíasyCienciasAplicadas(InSTEC),Cuba}

b_{UniversitédeToulouse,INPT,UPS,LGC,4,AlléeEmileMonso,BP84234,F-31432ToulouseCedex4,France} c_{CNRS,LaboratoiredeGénieChimique,BP84234,F-31432ToulouseCedex4,France}

a

b

s

t

r

a

c

t

Amongthe technologiesused totreat wastewater,the Submerged MembraneBioreactor (SMBR) hasexcellent prospectsbecauseofthepossibilityitprovidesforwaterreuse.Inthiswork,anSMBRcomputersimulatoris devel-oped.Amathematicalmodelwasimplemented,whichintegratedthebiologicaldegradationprocessusingactivated sludgeswiththephysicalseparationprocessusingmembranes.Thesimulatorfunctioningwasvalidatedwith exper-imentalresultsanditsuseinteachingwasevaluatedthroughthedevelopmentofasimulatedlaboratoryrunning forthreeandahalfhours.Thisgaveaccesstotrendsandordersofmagnitudethatwouldtakemorethanfifteen monthstoobtainwithrealexperiments.Itwassuccessfullyusedandacceptedbythestudents.

Keywords: Improvingclassroomteaching;Interactivelearningenvironments;Simulations;SubmergedMembrane Bioreactor

1.

Introduction

Fresh water is becoming known as the “blue gold” of the 21stcentury.Itisanaturalresourcealreadyinshortsupply and it will become even scarcer with increased urbaniza-tionandpopulation,climatechange,andindustrialpollution, makingithumanity’smostpreciousresourceandoneofthe majorenvironmentalissuesofthiscentury(BuzatuandLavric, 2011).For thisreason, manygovernments todayare devot-ingconsiderableresourcesandeffortstothedevelopmentof newtechnologiesforwastewatertreatmentandthe decon-taminationof contaminated sources.An example of these technologiesistheSubmergedMembraneBioreactor(SMBR).

TheSMBRcanbedefinedasasystemthatcombines bio-logicaldegradationofwastewatereffluentswithmembrane filtration(Ciceket al.,1999).Formanyyears,thesesystems haveshown theireffectivenessinthe treatmentof munic-ipalandindustrial wastewater(Jimenezet al.,2010;Santos

∗

Correspondingauthorat:UniversitédeToulouse,INPT,UPS,LGC,4,AlléeEmileMonso,BP84234,F-31432ToulouseCedex4,France. Tel.:+330534323630;fax:+330534323700.

E-mailaddress:marion.alliet@ensiacet.fr(M.Alliet).

et al.,2011).Inthe lasttwo decades,SMBRtechnologyhas grownexponentiallyduetoitsadvantagesoverconventional wastewater treatmentprocesses, such asreduced environ-mentalimpact,improvedeffluentqualityandbetterprocess control(BuerandCumin,2010;Drews,2010).Themajor poten-tial advantage of this technology is found in the field of waterreuse.ThisisbecausetheSMBRcanuseultrafiltration membranesandthusretainbacteria,somevirusesandmany organicandinorganiccomponentsthatareoftenfoundinthe effluentfromconventionalbiologicaltreatments(Lobosetal., 2007;DeLucaetal.,2013).

Therefore, the effluent ofan SMBR may be suitable for directreuseor watersupply forareverseosmosisprocess. ThatisoneofthereasonswhyresearchintheSMBRfieldis increasingcontinuouslyatpresent,duethecommercialand scientificinterestthatithasaroused(Stephensonetal.,2000; Van Nieuwenhuijzen et al., 2008). Nevertheless, the effec-tiveapplicationofmembranebioreactors(MBRs)islimitedby

Nomenclature

A membranearea(m2₎

C sludgeconcentration(kg/m3₎

Cd coefficientofdragandliftingforces

dp particlesize(m)

Fl liftingforce

Fa suctionforce

G apparentshearintensityofthefluidturbulence (s−1₎

g gravitationalconstant(m/s2₎

Gi apparentshearintensityofthefluidturbulence

ontheithsectionofthemembranesurface(s−1₎

J overallflux(m3_/m2_s)

Ji localfiltrationfluxthroughtheithmembrane

section(m3_/(m2_day))

Mtd massofsludgeinthedynamicsludgefilmcake

adheringtothemembranesurface(kg/m2₎

Mtf(i) mass of sludge in the stable sludge cake

attachedtotheithmembranesection(kg/m2₎

Mtd(i) massofsludgeinthedynamicsludgefilmcake

intheithmembranesection(kg/m2₎

n totalnumberofsectionsinthemembrane sur-facearea

qa aerationintensity(Lm−2s−1)

QBG coarsebubbleflow(L/s)

Rm intrinsicresistanceofthemembrane(m−1)

Rp porefoulingresistance(m−1)

rp specificporefoulingresistanceintermsof

fil-tratevolume(m−2₎

RT overallfiltrationresistance(m−1)

rtd specificfiltrationresistanceofdynamicsludge

film(m/kg)

Rtd resistanceofdynamicsludgefilm(m−1)

Rtf resistanceofstablesludgecakelayer(m−1)

rtf specificfiltrationresistanceofsludgecakelayer

(m/kg)

RTS(i) filtrationresistancefortheithmembrane

sec-tion(m−1₎

Si asectionofthemembranesurfacearea

SI concentrationofsolubleundegradableorganics

(gCOD/m3₎

SMBR SubmergedMembraneBioreactor SMP solublemicrobialproducts

SO2 concentrationofdissolvedoxygen(g/m3)

SRT sludgeretentiontime(days)

SS concentration of easily biodegradable

sub-strates(gCOD/m3₎

SSMP concentration of soluble microbial products

(gCOD/m3₎

t time(s)

taBG timeofcoarsebubbleaeration(min)

tf filtrationtime(min)

TMP trans-membranepressure(Pa)

tpaBG timewithoutcoarsebubbleaeration(min)

tpf relaxationtime(min)

tSTOP timetosimulate(days)

V bioreactorvolume(m3₎

Vf water productionwithin afiltrationperiodof

anoperationcycle(m3_/m2₎

XH concentrationofordinaryheterotrophic

orga-nisms(gCOD/m3₎

XI concentration of particulate undegradable

organics(gCOD/m3₎

XS concentration of slowly biodegradable

sub-strates(gCOD/m3₎

XTSS concentration of total suspended solids

(gTSS/m3₎

˛ stickinessofthebiomassparticles

ˇ erosionratecoefficientofthedynamicsludge film

1t timestep(s)

compressioncoefficientforthedynamicsludge film(kgm−3_s−1₎

εa fractionofthemembranesurfacearea(or

dis-tance ratio to the bottom of the membrane module)wheretheshearintensityisincreasing ε fractionofthemembranesurfacearea(or dis-tance ratio to the bottom of the membrane module)

f filtrationtimeinanoperationcycle(min)

s densityofsludgesuspension(kg/m3)

reduction index of cake compression coeffi-cient

s viscosityofsludgesuspension(Pas)

membranefoulingandtheassociatedcostandenergyburdens (MennitiandMorgenroth,2010).Atthesametime, experimen-tationinthesetypesofinstallationsisveryexpensiveandtime consuming.

Ontheotherhand,itisnecessarytotakealltheelements mentionedaboveintoaccountinthetrainingofengineersand ofthestaffthatwilloperatetheSMBR.Itisessentialtodevelop toolsthatcanhelpinthelearningprocess,bothatuniversities andatoperatortrainingcentres.Thedevelopmentof simula-torsisanecessitysincetheyconstituteaplatformtoenhance virtuallaboratories(Corteret al., 2011). Virtuallaboratories can provideadynamic Problem-BasedLearningexperience wherestudentsengageinanauthentic,industriallysituated task.Theysimulatewhatexpertengineersdoinpractice,and areverydifferentincharacterfromthephysicallaboratoryat university(Koretskyetal.,2011).Anotheradvantageofa simu-latorisitsvalueinthetrainingprocessfromtheresearchpoint ofview:tohelptosolveproblemsthatareasyetunsolved. Sim-ulatorsarealsoanimportantsupportforthestudyofprocess optimization.

Theuseofsimulatedexperimentscanconsiderablyreduce thecostofalaboratorycourse,increasethenumberof exper-imentsinthelearningprocessandenableexperimentstobe carriedoutthatwouldotherwiseinvolveworkingwith dan-gerousmaterialsand/orindangerousconditions(Skorzinski etal.,2009).Forallthesereasons,themathematicalmodelling ofanSMBRandthedevelopmentofasimulatorofthis pro-cessprovidesanalternative thatcansolvemany problems. Theobjectiveofthisworkistodevelopacomputersimulator ofanSMBRandtoshowitspotentialinteachinghowsuch processeswork.

2.

Materials

and

methods

A computer simulator consists of three main parts: the mathematical model, the numerical solution method and the graphical interface. The integrated model proposed by

Fig.1–ConceptualschemeoftheintegratedmodelproposedbyZarragoitiaetal.(2008).

Zarragoitiaetal.(2008)wasusedtobuildtheSMBRsimulator. ThemodelwasimplementedusingthePascalprogramming language.Thedifferentialequationsthatappearinthemodel were solvedusing the fourth-order RungeKutta numerical method.

Finally,the simulatorgraphicalinterfacewasdeveloped. Designingeducationalsoftwareinterfacesisacomplextask, given its strong domain dependency and multidisciplinary nature.Itrequirestheteachers’knowledgeandpedagogical beliefstobeincorporated intothe interface,posinga chal-lengetobothteachersanddesignersastheyhavetoactas partnersfromtheearliestphasesoftheprocess,sharingtheir knowledge (Perry and Schnaid, 2012). In the present case, thesimulatorgraphical interfacewasbuiltusing the facili-tiesprovidedbyDelphi2009forobject-orientedprogramming, inordertoachieveafriendlygraphicalinterfaceenablingthe assignationandmanipulationofdifferentoperating param-eters,aswellasobservationofhowthevariablesofinterest behaveovertime.

2.1. Mathematicalmodelanditsimplementation

ThemathematicalmodelproposedbyZarragoitiaetal.(2008)

integrates the biological degradation process by activated sludges with the physical separation process using mem-branes.

Inordertofacilitatetheevaluationofthemodel,the selec-tionofequationsandbiologicalprocessesconsideredduring modellingwaslinkedtothecharacteristicsofthe experimen-tal reactor and its operatingconditions. However, thefinal structureofthemodeloffersthepossibilityofaddingother process rates and stoichiometries. The conceptualscheme ofthemodeldevelopedisshowninFig.1.Itgivesthemain

relationsoperatingduringsimulationandalsothe informa-tionflowestablishedamongthedifferentpartsofthemodel duringcalculation.Themodelisdividedintothreesections, the first considers the biological behaviour (stoichiometry and kinetics), the second is related to membrane fouling evolutionandthebehaviourofallfiltrationresistances,and thelastconsistsofasetofperiodicequationsthatrepresent theprocess associatedwithcoarsebubbleaeration,feeding anddiscontinuousfiltration.

Tosimulatetheactivatedsludgeprocess,amodifiedmodel wasestablishedconsideringtheformation-degradation kinet-icsofsolublemicrobialproductsproposedinthemodification ofASMldevelopedbyLuetal.(2001),butadaptingthese equa-tionstoastrictlyaerobicSMBR.Thebiologicalmodelconsists ofasystemofdifferentialequationsobtainedfromthe Peter-sonmatrix(Zarragoitiaetal.,2008).

Ontheotherhand,inthephysicalseparationmodel,the mainprocessisthemassattachedtothemembranesurface. ThisprocessisdescribedbyEq.(1):

dMtd dt = 24CJ2 24J+CddpG −ˇ(1−˛)GM 2 td Vft+Mtd (1)

ThefirsttermofEq.(1)representsthenetmassdepositedon the membranesurfaceduetothe equilibriumbetweenthe suctionandliftingforces(Fig.2),whilethesecondterm rep-resentsthemassremovedbytheshearforcescausedbythe coarsebubbleaeration(LiandWang,2006).

Shearforcesarenotuniformlydistributedoverthe mem-brane surface.When modellingthedepositionofsludgeon themembranesurface,inthecasewherethemembrane mod-uleresemblestoacylindricalobject,thevariationofGvalues

Fig.2–Modellingthemembranefoulingprocess.

inthevicinityofthemembranemodulecanbeestimatedas follows:

- atthebottomofthemembranemodule,Gvalueisassumed thesmallestpossiblevalue,whichisone-tenthofits maxi-mumvalue,

- fordistanceratiostothebottomhigherthanεa,Gvalueis

assumedtotakeitsmaximumvalue,

- betweenthetwo,sinusoidalgrowthisassumed(Fig.2). Therefore, Eqs. (2) and (3) are used to calculate the Gi

values. These equations must be modified for the case of moduleswith differentgeometryor differentaeration sys-tems,e.g.withairnozzleslocatedatmultipleheightsofthe membranemodule,whicharenotfrequentduethestructural complexitiesandchangesthatthehydrodynamicsofthe sys-temintroduces(Zarragoitiaetal.,2008).

G(ε)=





















q

q















Forthemodellingofmembranefouling,themembrane sur-face was divided into sections of equal area (Si). It was

consideredthateachareaSiwastraversedbyaflowofequal

magnitude,whichwascalculatedbythefollowingequation:

Ji=

J

n (4)

Eq. (4) was adaptedso as to calculate the massof sludge depositedintheithsectionofthemembranesurface.Liand Wang,(2006)foundthat128sectionsofequalarea(Si),

guar-anteedagoodapproximationtosystembehaviour.

Also,theoverallflux(J)inEq.(1)isreplacedbythelocal filtrationfluxthroughtheithmembranesection(Ji)andthe

apparentshearintensityofthefluidturbulence(G)isreplaced bytheactualshearintensityontheithsectionofthe mem-branesurface(Gi).Hence,withthesetransformationsEq.(1)

canbeexpressedas: dM_td(i) dt = 24CJ2 i 24Ji+CddpGi − ˇ(1−˛)GiM2_td(i) Vf(i)t+Mtd(i) (5)

Thistypeofinstallationworksinperiodiccyclesoffiltration andcleaningbycoarsebubbleaerationso,whenthesystemis filtering,themassattachedtothemembranesurfaceis calcu-latedusingEq.(5),butwhenthesystemisnotfiltering,theonly processthatcanoccurisremovalofthesludgefromthe mem-branesurface.InthiscaseEq.(6)isusedtocalculatethemass attachedtothemembranesurface,whichisamodificationof Eq.(5).

dM_td(i) dt =−

ˇ(1−˛)GiM2_td(i)

Vf(i)f+Mtd(i) (6)

Foreachnewtimestep:Mtf(i)(t+1t)=Mtf(i)(t)+Mtd(i)(t).

Another parametertobetakeninto accountisthe pore foulingresistance.Porefoulingoccursasaresultofthe adhe-sionofsomesludgeparticleshavingadiameterlessthanor oftheorderofthediameteroftheporesontheinteriorwalls (Mengetal.,2009).Then,thefiltrationresistancefortheith

membranesectionisdeterminedbythefollowingexpression: R_TS(i)=R_m(i)+R_p(i)+R_td(i)+Rtf(i) (7)

where R_m(i)=const. (8) R_p(i)=rp m

X

Initiation

Input data

t =0

Determination of the sludge

concentration (C)

Determination of the mass

to the membrane (M

)

Determination of the overall

resistance (R

)

t < t

t=t+ t

Output parameters

End

Yes

No

Determination of the

trans

-membrane pressure (TMP

)

Fig.3–Generalalgorithmfortheimplementationofthe mathematicalmodel.

Oncethevalueofthefiltrationresistancefortheith mem-branesectionhasbeendetermined,thevalueoftheoverall resistancecanbecalculatedbymeansoftheexpression(12).

1 RT = n

X

Finally,thevalueofTMPcanbedeterminedusingDarcy’slaw

(13):

TMP=sJRT (13)

Thegeneralschemeofthealgorithmforimplementingthe mathematicalmodelisshowninFig.3.

Forthecalculation,thesludgeconcentrationisdetermined bysolvingthe systemofdifferential equationsforthe bio-logical system. Then, the mass of sludge attached to the membranesurfaceanditsresistancetofiltrationare deter-mined. Later, the overall resistance value is evaluated in theexpressionofDarcy’slawsothattheTMPvaluecanbe obtained.Theoutputparametersofthesimulatorareshown ateach1stimeinterval,whichisthestepusedbythe fourth-orderRungeKuttanumericalmethodduringthecalculation.

Nowadays,manyunitprocessmodelsareavailableinthe fieldofwastewatertreatment.Allofthesemodelsusetheir own notation, which causes problems for documentation, implementation and connectionofdifferent models (using differentsetsofstatevariables)(Corominasetal.,2010).For thisreason,inthiswork,theuniversalnotationproposedby

Corominasetal.(2010)todescribemathematicalmodelsis used.

2.2. Evaluationofthesimulator

Todemonstratethesimulatorcapacitytodescribethe perfor-manceofarealmembranebioreactor,theexperimentaldata reportedbyZarragoitiaetal.(2008)wereused.Tables1–3show theworkingconditionsinwhichtheexperimentswere con-ducted,the characteristics ofthe wastewaterusedand the characteristicsoftheactivatedsludgeatthebeginningofthe experiments,respectively.

2.3. Practicalapplicationofthesimulatorinchemical engineeringeducation

The simulator was used in chemical engineering educa-tion in December 2012and inDecembre 2013atthe Ecole Nationale Supérieure des Ingénieurs en ArtsChimiques Et Technologiques(ENSIACET),France.Theuseofthissoftware wasincorporatedintothesyllabusofthe“Watertreatment” courseinthefifthyearofhighereducationcorrespondingto theChemicalEngineeringCareersunit.Theobjectivesofthe practicalactivityare thatthestudentsbecomefamiliarized withawater treatmentprocess,that theyunderstand how theSMBRprocessworksandthattheyanalyzetheinfluence ofoperatingparametersonthefunctioningoftheprocess.For thispurposepracticalworklastingthreeandahalfhoursina simulatedlaboratorywasdeveloped.Itwasstructuredinthree steps:

1. Analysisofthemathematicalmodelimplementedinthe simulatorinordertobetterunderstandtheresultsobtained withthesimulator.

2. Resolutionofapracticalproblemusingthesimulator. 3. Writingofthereportbythestudents,withtheresultsand

discussionofallthesimulatorpredictions.

2.3.1. Practicalactivity

Apracticalexercisewasdevelopedusingexperimentaldata from a real plant (Zarragoitia et al., 2008). Theexercise is describedbelow.

AnMBRpilotplantwithasubmergedmembrane configu-rationislocatedinthetownofBrax,France.Thisplantisfed withavolumetricflowof0.09m3_day−1_{ofrealwastewater,the}

characteristicsofwhichareshowninTable1.Specifically,a polysulfonehollow-fibremembranemodulesuppliedby Poly-men (pore size=0.2mm, surface area=0.3m2_{) is immersed}

directlyinanaerobictankwithavolumeof10L.Thesludge retentiontimeisaround30days.Thebiologicalreactoris aer-atedandstirredbyafinebubbleairdiffuserlocatedatthe bottomofthereactor.Asecond airblower,locateddirectly atthebottomofthemembranemodule,producesanairflow rateof6Lm−2_s−1_{togeneratecoarsebubblesthatcausestrong}

turbulencesoastocleanthesurfaceofthemembraneand thuslimitmembranefouling.Membranefiltrationiscarried outinconventionalsequentialcyclemode:9minwith filtra-tionand1minwithoutfiltration(relaxationtime).Duringthe relaxationtime,thecoarsebubbleaerationison,sothe mem-branecleaningiscarriedoutinacyclemode:1minofcoarse bubbleaerationand9minwithoutsuchaeration.

Themaintenanceofthesystemisperformedwhen trans-membranepressure reachesitscriticalvalue,whichinthis caseisfixedat60kPa.ThemeantemperatureinsideofMBR

Table1–WorkingconditionsinwhichtheexperimentswereperformedintheSBRM.

Temperature(◦_{C) q}

a(Lm−2s−1) tf(min) tpf(min) taBG(min) tpaBG(min) SRT(days) J(m3/(m2day))

7.3 11 10 4 2 8 40 0.25

Source:Zarragoitiaetal.(2008).

Table2–Characteristicsofthewastewatertobetreated.

XSST(mg/L) XS(mg/L) XI(mg/L) XH(mg/L) SS(mg/L) SI(mg/L) SPMS(mg/L) SO2(mg/L)

50 70 40 10 220 10 60 0.2

Source:Zarragoitiaetal.(2008).

Table3–Characteristicsoftheactivatedofsludgeatthebeginningoftheexperiments.

XSST(mg/L) XS(mg/L) XI(mg/L) XH(mg/L) SS(mg/L) SI(mg/L) SPMS(mg/L) SO2(mg/L)

5500 10 10 550 130 10 90 10

Source:Zarragoitiaetal.(2008).

is20◦_{Cduringtheentirecampaign.Tables2–4showthedata}

necessaryfortheMBRoperation:thesupplycharacteristics, the characteristics ofthe mixture insidethe MBR and the propertiesofthe activatedsludge and cakeformed onthe membranesurface,respectively.

2.3.1.1. Activities. Usingthesoftware

1. SimulatetheoperationoftheMBRover8days.Describethe behaviourovertimeofthefollowingparameters:

• Trans-membranepressure. • Totalsuspendedsolids. • Chemicaloxygendemand. • Filteredwatervolume. • Solublemicrobialproducts. • Overallresistancetofiltration. • Dissolvedoxygen.

2. Determinethetimenecessaryforthesystemtoreachthe criticaltrans-membranepressureworkingintheoperating conditionsdescribedabove.

3. Whatneedstobedoneifthesystemreachesthecritical trans-membranepressure?

4. In order to understand the influence of the following variablesonthetrans-membranepressure,performa sen-sitivityanalysisintheindicatedranges:

• Filtrationflux(±5%ofestablishedvalue).

• Specificfiltrationresistanceofthesludgecakelayer(±5% ofestablishedvalue).

Discusstheresultsofthesensitivityanalysis.

Table4–Propertiesoftheactivatedsludgeandcake

formedonthemembranesurface.

Parameters Value

Erosionratecoefficientofthedynamicsludgefilm 3.5×10−4

Stickinessofthebiomassparticles 0.7

Particlesize(m) 1×10−4

Specificfiltrationresistanceofthedynamicsludge film(m/kg)

1×1015

Specificfiltrationresistanceofthesludgecakelayer (m/kg)

1×1015

Compressioncoefficientforthedynamicsludgefilm (kgm−3_s−1₎

2.5×10−5

Coefficientofthedragandliftingforce 0.4 Reductionindexofthecakecompressioncoefficient 0.1

5. For the purposeof determiningthe best operating con-ditions,plot thetimenecessary forthe systemtoreach thecriticaltrans-membranepressureversusthefollowing variables(maintainingtherestoftheoperatingconditions constant):

• Filtration frequency (Filtration time/Relaxation time=9:1,7:3,5:5,3:7and1:9)

• Aerationintensity(3,6,9,12and15Lm−2_s−1₎

Discusstheresultsobtained.

2.3.2. Anonymousquestionnaire

Attheendofthepracticalactivity,thestudentswereaskedto fillinananonymousquestionnairegivingtheiropinionabout thesimulatedlaboratory(Table5).

3.

Results

and

discussion

3.1. Descriptionandoperationofthesimulator

For teaching use, the SBRM computer simulatordeveloped shouldbeuserfriendlyandprovideaneasilyaccessible intro-ductiontothesubject.Sinceotherusesareadvancedtraining andresearch,manyparametersshouldbeeasilymodifiable.

The simulatorshows ageneral standardscheme ofthe SMBR,whichallowsthemainstructuralcomponentsofthe systemtobeapprehended,sothattheusercangainabetter understandingoftheinstallationperformanceandthusa bet-terunderstandingoftheprocessesthatareinvolvedinthese typesofinstallations(Fig.4).

Thesimulatorallowstheusertostudyoftheinfluenceof the35modelinputvariables(thosepresentedinTables1–4, bioreactor volume, membranesurfaceand time tobe sim-ulated) on 16 output parameters, which can be displayed graphicallyornumerically.Theseoutputparametersare:

1. Trans-membranepressure.

2. Resistanceofthestablesludgecakelayer. 3. Resistanceofthedynamicsludgefilm. 4. Porefoulingresistance.

5. Overallresistance. 6. Chemicaloxygendemand. 7. Dissolvedoxygen.

Table5–Questionspresentedtothestudents.

No Questions Stronglyagree Agree Disagree Stronglydisagree Idon’tknow

1 Thislaboratoryisinteresting     

2 Thislaboratoryhelpedmetorepresent therealfunctioningofamembrane bioreactor

    

3 Withthisexercise,Ihaveabetter understandingoftheinfluenceofthe membranebioreactorfunctioning parametersontheprocessperformance

    

4 Iwasabletounderstandthefunctioning ofthesimulatorbymyself

    

5 Theparticipationoftheteacherhelped meunderstandthefunctioningofthe simulatorbetter

    

6 Theproblempresentedtomewasclearly defined

    

7 Thisactivityenabledmetounderstand whyacomputersimulatorsisanuseful toolinchemicalengineering

    

8 Thislaboratoryiswellsituatedinmy educationprogramme

    

9 Thislaboratoryisrelevanttomy programme

    

10 Theactivityhelpedmetoconsolidatethe conceptsexposedintheclassroom

    

9. Solubleundegradableorganics. 10. Solublemicrobialproducts. 11. Particulateundegradableorganics. 12. Slowlybiodegradablesubstrates. 13. Ordinaryheterotrophicorganisms. 14. Totalsuspendedsolids.

15. Massattachedtothemembrane. 16. Filteredwatervolume.

The simulator enables the user to select, before calcu-lating,the time-scale overwhich resultswillbepresented.

Thistime-scalecanbeseconds,minutes,hoursordays.The precisionofcalculationdoesnotchangewiththetimescale selectedbecauseitisalwaysperformedinseconds.

Also, the simulator allows the numerical results to be exported toa txt file. Thisoption may permitthe user to process theseresultsusingother computationaltools. Sim-ilarly,thegraphicalresultscanbesavedinbmpimageformat. Anotheradvantageisthateachcalculationcanbesavedina fileusingbmsformat,anextensioncreatedforthissoftware. Finallythesimulatoropensthepossibilityofasensitivity anal-ysiswithrespecttovariousparameters.

Fig.5–Comparisonbetweentheexperimentaldata(points)andsimulationresults(line).Theleft-handdiagramshowsthe workingconditionsoftheexperiments.

3.2. Comparisonofsimulatorperformancewith experimentaldata

Tosubstantiateandjustifytheuseofthecomputersimulator tostudyanSMBR,itisessentialtoknowthelevelof approxi-mationtowhichthemathematicalmodelcanreproduceSMBR operation.Forthisreason,thesimulationresultswere com-paredwithexperimentaldata.

Theparameterchosenwasthetrans-membranepressure becauseofitsimportanceintheoperationoftheSMBR(Meng etal.,2009;Fenuetal.,2010).Fig.5showstheexperimentaland calculatedvaluesofTMP.Ameanrelativeerrorofestimation of15%wasobtained.

This result can be considered acceptable for predicting the behaviour ofa systemof suchcomplexity because, as explainedabove,thesimulatorwasbuiltbyimplementinga modelthatcombinesbiologicaldegradationwiththefiltration process.Thebiologicalsystemmodelling introducesahigh percentageoferrorastheinputvaluesofbiologicalvariables correspondtothemeanvaluesmeasuredduringthe experi-ment(Zarragoitiaetal.,2008).

Even with15% oferror, thesimulator already givesthe trendsintheevolutionofphysicalquantitiesandtheorderof magnitudeoftheirvalues,whichistheinformationsought.

3.3. Practicalapplicationofthesimulatorinchemical engineeringeducation

Beforebeginningthepracticalactivity,theteachergavethe studentsabriefoverviewofthemodelimplementedinthe simulatorasa“blackbox”.Oncethestudentshadreceived theorientationinformationandthenecessarymaterials,they werereadytostarttheproposedexercise.Theinstructorwas athandduringtheentirepracticalactivity,andwasavailable toclarifyanypointsthestudentshaddoubtsabout.Finally,the studentspreparedareportwiththeexercisesolution.When theyhad finishedtheworkinthesimulatedlaboratory,the studentsansweredtheanonymousquestionnaire,whichwas drawnupintheaimofknowingthestudents’opinionsabout thesimulatedlaboratory.

Theevaluationsofhowwelltheobjectivesofthispractical activitywereattainedweremade:

- duringtheactivityitself,bythediscussionbetweenthe stu-dentsandtheteachingstaff,

- fromthereportsgivenbythestudents, - bythestudents’answerstothequestionnaire.

3.3.1. Considerationsaboutthepracticalactivity

Thepracticalactivitywascarriedoutsuccessfullybythe stu-dents. They showed their abilities in the use of computer programsand,ingeneral,theymanagedthesimulatorwith success. Nevertheless, there were some students who had difficultysolvingthistaskbecausetheydidnotunderstand the functioning of anSMBR correctly and others who had problems with the simulator language. However, with the instructor’shelp,theyfinishedtheproposedexercisecorrectly. Thestudents’correctuseofthedifferentsimulatortoolsand their understanding of the SMBR operation was evaluated fromtheiranalysisoftheresultstheyreported.

Thereportswere correctedand gradedaccordingtothe Frenchnorm,whichgivespointsoutof20,withthe follow-ingappreciation:10=pass,12=quitegood,14=good,16=very good,18=excellent,and20=congratulations.Theaveragewas 13.6/20withaminimumof12/20andamaximumof16/20, whichisarathergood result.Parts1–4were achievedvery well,withonlyminormistakes. Part5was completedina morevariable way,mainlyduetoalackoftime(andtothe Frenchwayofteaching,whichdiscriminatesusingtime).

3.3.2. Students’opinions

Thestudents’responsestothequestionnairearepresentedin

Fig.6.Agradingscaleobtainedbyusingnumericalequivalents forthe opinions:“Strongly agree”=20,“Agree”=13.33, “Dis-agree”=6.67,“Strongly disagree”=0 (in orderto correspond totheFrenchgradingsystem, whichisoutof20)hasbeen added.Foreachofthestatementsproposedinthe question-naire,a“grade”isindicated,whichwasobtainedbyaveraging theanswers.Toanalyzetheseresponses,thequestionswith themostnumerousanswers“Disagree”were consideredas wellastheoneswithlessgoodgrades.Thestudents’ evalu-ationswereverypositive.Thissimulatedlaboratoryaroused greatinterestinmorethan95%ofthestudents.

As noticed by the teaching staff during the practical activity, although some students had some difficulties in understandingthefunctioningofthesimulatorbythemselves (Q4),theparticipationoftheteacherhelpedtheminthistask (Q5).

Slightlymorethan10%ofthestudentsdidnotagreethat thislaboratorywasrelevanttotheirprogrammeandthe low-est evaluationconcerned thesituation ofthis laboratoryin theeducationprogramme(Q8).Adiscussionwiththestudents showedthatanadditionalexperimentalactivitymayhelpto improvethisimpression.Theteachingstaffisthinkingabout aconvenientandnottooexpensivewaytoincludeit(visittoa

Fig.6–Resultsoftheanonymousquestionnaire(seeTable5). watertreatmentplant,visittoaresearchexperimentaldevice,

shortexperimentalpracticalactivity,etc.)

3.4. Advantagesofsimulatorinteaching

Laboratory exercises, field observations and field trips are a fundamental part of many earth science and environ-mental science courses. Field observations and field trips cansufferfrom constraintsofdistance,time, expense,and the scale,safety,or complexity ofreal-worldenvironments (Ramasundaram etal., 2005). Corteret al.(2011) statethat themajorityofcomparativestudieshaveconcludedthat sim-ulation is a good substitute forhands-on labsin teaching courseconceptsandtheirapplicationbut someresearchers haveproposedthatsimulationmightbemosteffectivewhen itisintegratedasacomplementarypartofacourseinvolving hands-onlaboratoryactivity.

In comparison with textbooks and lectures, a learning environment with a computer simulation has the advan-tagesthatstudentscansystematicallyexplorehypothetical situations,interactwithasimplifiedversionofaprocessor system,changethetime-scaleofevents,and practicetasks andsolveproblemsinarealisticenvironmentwithoutstress (vanBerkumanddeJong,1991).

El-Naas (2011) developed a course following the active learning approach, wherestudents are heavily involved in classactivitiesandtheycandirectlyassesstheeffectofinput variablesonthedesignparameters,allowingthemtocarryout “WhatIf”orparametersensitivityanalysis.Forthisheused ExcelandEz-Solve indesigningandanalysing desalination processes.Simulationshaveearnedaplaceintheclassroom asrobustadditions tothe teachers’repertoire,eitherasan additiontothetraditionalteachingmethodsavailableorasa replacementofpartsofthecurriculum(Ruttenetal.,2012).As shownhere,theSMBRcomputersimulatordevelopedisable topredictthebehaviourofvariousoutputparametersinthe short-and long-term.Itallowsuserstostudytheinfluence ofthebioreactorinitialworkingconditionsonthebehaviour ofthemainparametersthatdescribetheplantperformance.

ItalsoallowstheoptimumSMBRoperatingconditionstobe determinedsoastoincreasethemembranelifespan.Atthe sametime,itfacilitatesstudiesrelatedtothesearchforthe optimaldesignparametersoftheplantandtheinfluenceof sludgeproperties,whichwillleadtobetterefficiencyinthe processofwastewatertreatment.Italsofacilitates sensitiv-itystudiesonthemostimportantparametersinthesystem, whichisafundamentalaspecttobeconsideredinfuture mod-ellingworksothatthemajorcontributorstotheestimation errorofthemodelcanbedetermined.Lastbutnotleast,it givesaccesstointermediateparametersthatcannotbe reg-ularlymeasured inrealconditions, suchastheresistances ofthestablesludgecakelayerandthedynamicsludgefilm, ortheconcentrationofsolublemicrobialproducts,butwhich contributetoabetterunderstandingofthewaytheprocess functions (membranefouling,COD,etc.). Itisimportantto notethatthesimulatorwasbuiltinamannerthatmakesit aninvaluabletoolforteachinghowanSMBRworks,sinceit allowstheusertointeractwiththeoperatingconditionsofthe bioreactorandobservetheinfluenceoftheseparameterson thebehaviourofthemaincontrolvariablesofthesystemover time.

Computer modelling has become a helpful tool in the analysisoftheperformanceandeffectivenessofwastewater treatmentsystems(Korniluketal.,2008).Nowadays,industrial firms havebecomevery interested because computer sim-ulationssignificantlylowercostscomparedtoexperimental studies (Kraft et al., 2005).The use ofthe SMBR simulator allows considerable savingof resources and time since,in general,performingexperimentsinrealtimeonanSMBRis veryexpensiveandtimeconsuming.Forthepractical activ-ity, wewere abletoplace24 studentsinfrontoftheirown 12“devices”.Thisallowedthemtovirtuallyimplement sev-eraloperatingconditionsandtoanalyzetheconsequencesof thiscontrol.Theuserfriendlinessofthesimulatorhelpedin this,inatimemuchshorterthanintherealworldandata costcompatible withthe university’sresources.Theroleof theteacherswastosupplythestudentswithallthe practi-calproceduresthatarenotincludedinthesimulator,suchas

chemicalcleaningormembranereplacement,andtokeepthe studentsawareoftherealcontextofinsituexperiments com-paredtoinsilicoones.Inthecaseofthetreatmentplantunder study,theplantneedsaperiodofalmost20daysforthesludge mixturetobecomestabilizedand,eachtimethecriticalvalue oftrans-membranepressureisreached,thesystemmustbe stoppedtocarryoutexpensivechemicalcleaningor replace-mentofthemembrane,oftenbeforetheendofitslifespan (Zarragoitiaetal.,2008;Kimetal.,2011).However,thisstudy canbeperformedinthesimulatorinashorttimeandwith considerablesavingofresources.

4.

Conclusions

An SMBR computer simulatorwas built with recent mod-ellingknowledgeandafriendlyinterface.Theresultsgiven bythesimulatorareaccurateenoughtoprovidethetrends andordersofmagnitudesofphysicalquantitiesneededfor theteaching applicationofthe simulatedMBR.The practi-caluseofthesimulatorwasevaluatedwiththedevelopment ofsimulatedlaboratoryworklastingthreeandahalfhours, whichgaveresultsthatwouldtakemorethanfifteenmonths ofreal-worldexperiments.Itwassuccessfully applied,and achievedthemostdifficultobjectivesofenablingthestudents toanalyzetheinfluenceofoperatingparametersontheSMBR functioningandbeinglargelyacceptedbythestudents.While thishasnotbeentested, itappears clearthatthe dynamic modelusedwouldpermitthetrainingofprofessionals.

References

Buer,T.,Cumin,J.,2010.MBRmoduledesignandoperation. Desalination250,1073–1077.

Buzatu,P.,Lavric,V.,2011.Optimaloperatingstrategiesofa SubmergedMembraneBioreactorforwastewatertreatment. ChemicalandBiochemicalEngineeringQuarterly25(1), 89–103.

Cicek,N.,Franco,J.P.,Suidan,M.T.,Urbain,V.,Manem,J.,1999.

Characterizationandcomparisonofamembranebioreactor andaconventionalactivated-sludgesysteminthetreatment ofwastewatercontaininghighmolecularweightcompounds. WaterEnvironmentResearch71,64–70.

Corominas,L.,Rieger,L.,Takács,I.,Ekama,G.,Hauduc,H., Vanrolleghem,P.A.,Oehmen,A.,Gernaey,K.V.,van Loosdrecht,M.C.M.,Comeau,Y.,2010.Newframeworkfor standardizednotationinwastewatertreatmentmodelling. WaterScienceandTechnology61(4),841–857.

Corter,J.E.,Esche,S.K.,Chassapis,C.,Ma,J.,Nickerson,J.V.,2011.

Processandlearningoutcomesfromremotelyoperated, simulated,andhands-onstudentlaboratories.Computers& Education57,2054–2067.

DeLuca,G.,Sacchetti,R.,Leoni,E.,Zanetti,F.,2013.Removalof indicatorbacteriophagesfrommunicipalwastewaterbya full-scalemembranebioreactorandaconventionalactivated sludgeprocess:implicationstowaterreuse.Bioresource Technology129,526–531.

Drews,A.,2010.Membranefoulinginmembranebioreactors– characterisation,contradictions,causeandcures.Journalof MembraneScience363,1–28.

El-Naas,M.H.,2011.Teachingwaterdesalinationthroughactive learning.EducationforChemicalEngineers6,e97–e102.

Fenu,A.,Guglielme,G.,Jimenez,J.,Spérandio,M.,Saroj,D., Lesjean,B.,Brepols,C.,Thoeye,C.,Nopens,I.,2010.Activated

sludgemodel(ASM)basedmodellingofmembranebioreactor (MBR)processes:acriticalreviewwithspecialregardtoMBR specificities.WaterResearch44,4272–4294.

Jimenez,J.,Grelier,P.,Meinhold,J.,Tazi-Pain,A.,2010.Biological modellingofMBRandimpactofprimarysedimentation. Desalination250,562–567.

Kim,M.J.,Sankararao,B.,Yoo,C.K.,2011.DeterminationofMBR foulingandchemicalcleaningintervalusingstatistical methodsappliedondynamicindexdata.Journalof MembraneScience375,345–353.

Koretsky,M.D.,Kelly,C.,Gummer,E.,2011.Fundamental ResearchinEngineeringEducation.StudentLearningin IndustriallySituatedVirtualLaboratories.Chemical EngineeringEducation(CEE)45(3),219–228.

Korniluk,M.,Montusiewicz,A.,Piotrowicz,A.,Lagód,G.,2008.

Simulationofwastewatertreatmentsystemswithmembrane separation.ProceedingsofECOpole2(1),41–45.

Kraft,M.,Mosbach,S.,Wagner,W.,2005.Usingawebmoduleto teachstochasticmodeling.ChemicalEngineeringEducation 39(3),244–248.

Li,X.-Y.,Wang,X.-M.,2006.Modellingofmembranefoulingina submergedmembranebioreactor.JournalofMembrane Science278,151–161.

Lobos,J.,Wisniewski,C.,Heran,M.,Grasmick,A.,2007.Membrane bioreactorperformances:effluentqualityofcontinuousand sequencingsystemsforwaterreuse.Desalination204, 39–45.

Lu,S.G.,Imai,T.,Ukita,M.,Sekine,M.,Higuchi,T.,Fukagawa,M., 2001.Amodelformembranebioreactorprocessbasedonthe conceptofformationanddegradationofsolublemicrobial products.WaterResearch35,2038–2048.

Meng,F.,Chae,S.R.,Drews,A.,Kraume,M.,Shin,H.S.,Yang,F., 2009.Recentadvancesinmembranebioreactors(MBRs): membranefoulingandmembranematerial.WaterResearch 43,1489–1512.

Menniti,A.,Morgenroth,E.,2010.MechanismsofSMPproduction inmembranebioreactors:choosinganappropriate

mathematicalmodelstructure.WaterResearch44,5240–5251.

Perry,G.T.,Schnaid,F.,2012.Acasestudyonthedesignof learninginterfaces.Computers&Education59,722–731.

Ramasundaram,V.,Grunwald,S.,Mangeot,A.,Comerford,N.B., Bliss,C.M.,2005.Developmentofanenvironmentalvirtual fieldlaboratory.Computers&Education45,21–34.

Rutten,N.,vanJoolingen,W.R.,vanderVeen,J.T.,2012.The learningeffectsofcomputersimulationsinscience education.Computers&Education58,136–153.

Santos,A.,Ma,W.,Judd,S.J.,2011.Membranebioreactors:two decadesofresearchandimplementation.Desalination273, 148–154.

Skorzinski,E.,Shacham,M.,Brauner,N.,2009.Asimulation programformodellingpollutantdispersionforeducational applications.In:Je ˙zowski,J.,Thullie,J.(Eds.),19thEuropean SymposiumonComputerAidedProcessEngineering.Elsevier, Oxford,pp.1233–1238.

Stephenson,T.,Judd,S.,Jefferson,B.,Brindle,K.,2000.Membrane BioreactorsforWastewaterTreatment.IWAPublishing, London.

vanBerkum,J.J.A.,deJong,T.,1991.Instructionalenvironments forsimulations.Education&Computing6,305–358.

VanNieuwenhuijzen,A.F.,Evenblij,H.,Uijterlinde,C.A., Schulting,F.L.,2008.Reviewonthestateofscienceon membranebioreactorsformunicipalwastewatertreatment. WaterScienceandTechnology57,979–986.

Zarragoitia,A.G.,Schetrite,S.,Alliet,M.,Jáuregui-Haza,U.J., Albasi,C.,2008.Modellingofsubmergedmembrane bioreactor:conceptualstudyaboutlinkbetweenactivated sludgebiokinetics,aerationandfoulingprocess.Journalof MembraneScience325,612–624.