Membrane Biofouling
Director and Prof. Chung-Hak LEE,
Institute of Environment Protection and Safety Water Environment – Membrane Technology Lab. School of Chemical Engineering,
Seoul National University, KOREA
Water Environment-Membrane Technology Lab., Seoul National University ApplicationApplication ofof MembraneMembrane ProcessesProcesses inin WaterWater EnvironmentEnvironment
Fusion Tech Hydraulics Mol. Biol. Sur. Chem. Nano Particle 난배양성 미샘물 CFD 모니터링 Biofilm Homo. Catal.
Space station Airplane
Building Recreation
House Industrial water
Stream water Groundwater
Water Environment-Membrane Technology Lab., Seoul National University MBR : Worldwide Buisiness
~ thousands of installations of MBRs in the World
Vivendi (France), Memcore (USA) Zenon (Canada), Mitsubishi, Kuboda (Japan), etc..
~ 750 MBRs in Korea since last 6 years
Water Environment-Membrane Technology Lab., Seoul National University Commercial MBR plants in Korea *2003(2002)
Membrane Highest Membrane Starting Number of Module type Company Trade mark Material capacity Fitration Mode Manufacturer Year Installation 3 (PORE SIZE, ㎛ ) (M /d)
MRC KEC & PE SMAS & HANT 1997 403(300) 4,000(1,400) (JAPAN) HEC (0.4)
ZENON PVDF SAE-HAN ZENOGEM 2000 10(7) 1,000(300) (CANADA) (0.035)
KMS HOLLOW PE KMS - 2002 150(100) 600(225) (KOREA) FIBER (0.4) MEMBRANE
SKC, PSF E.N.E. KOLON KIMAS Ⅰ,Ⅱ 1998 10(10) - DEAD-END (0.1) (KOREA) FILTRATION
天津膜天社 RAPAH PVDF - 2002 10 - (CHINA) TECH (0.1-0.4)
YUASA Polyolefin ZENIX ENG & JIN NIX-MBR 1999 68(67) 4,000(900) (JAPAN) WOO ENV. (0.4) PLATE PURE MEMBRANE PURE CPVC ENVI-TECH - 2002 20(2) 350(250) ENVI-TECH (0.25 ) (KOREA)
MEMBRATEK PES AQUATECH BIOSUF 1995 50(50) 2,000(2,000) (SOUTH AFRICA) (40,000Da) CROSSFLOW TUBULAR FILTRATION MEMBRANE PSF RUSSIA ZENIX ENG. - 1996 13(13) 200(-) (30,000Da)
TOTAL 734(559)
Water Environment-Membrane Technology Lab., Seoul National University Membranes used for domestic MBR processes
Hollow fiber Plate Tubular
MF MF UF Out-In Out-In In-Out Submerged Crossflow/ Submerged Crossflow
Water Environment-Membrane Technology Lab., Seoul National University Factors Affecting Membrane Performance
Water Environment-Membrane Technology Lab., Seoul National University Biofouling
: Membranes in contact with the broth of activated sludge reactor will be colonized within short time by microorganisms, leading to the formation of a composite layer known as biofilm.
: Biofouling has restricted the widespread application of MBR, because i) it limits the maximum flux obtainable, ii) it leads to substantial cleaning requirements, iii) it shortens membrane life time
Water Environment-Membrane Technology Lab., Seoul National University Development of Biofilm Community
Advances in Microbial Ecology, 1995
J.R.Lawrence,Water Environment D.R.-Membrane Korber, Technology G.M. Lab., Wolffardt,Seoul D.E National Caldwell University Scanning electron micrograph (SEM) of an early bacterial biofilm on the surface of a cellulose acetate (CA) reverse osmosis (RO) membrane. In this case the RO membrane was used to demineralize pretreated municipal wastewater at Water Factory 21 in Orange County, California. Note that the microcolonies appear to have gained foothold in low depressions (imperfections) in the membrane surface where lower hydrodynamic shear forces might be expected. Localized regions of reduced shear allows more time for bacteria to undergo irreversible attachment. Water Environment-Membrane Technology Lab., Seoul National University Scanning electron micrographs (SEMs) of individual cells and microcolonies growing on the permeate (product water) surfaces of polyester Texlon fibers of cellulose acetate (CA) reverse osmosis (RO) membranes. The membranes were fed with a pretreated municipal wastewater at Water Factory 21 in Orange County, California. Note the copius production of extracellular polymeric substances (EPS) by the attached bacteria, especially those cells associated with the larger microcolonies. Such EPS mediates early cell attachment and physically stabilizes and protects the biofilm. Water Environment-Membrane Technology Lab., Seoul National University MBR for advanced wasterwater treatment Side Stream type
Submerged type Water Environment-Membrane Technology Lab., Seoul National University a) Traditional wastewater treatment (전통적인 생물학적 처리공정)
b) External crossflow and side stream (외부 십자흐름 분리형)
c) Internal submerged (내부 침지형)
d) External submerged (분리 침지형)
Water Environment-Membrane Technology Lab., Seoul National University Models for predicting membrane performance (Flux)
1) Filtration Model
2) Resistance in Series Model
3) Film theory Model
Water Environment-Membrane Technology Lab., Seoul National University Filtration Models
- If the solute is thought of as a ‘cake’ of deposited particles,
- the cake (gel, deposit, etc) resistance is obtained from filtration theory. ∆P J = − (1) (Rm + Rc )⋅η
Rc = α ⋅VCb / Am = α ⋅ms / Am −(2)
Rc : solute resistance (1/m) Rm : membrane resistance (1/m) V : cumulative solvent volume (m3)
Cb : bulk solute concentration ( kg/m3) Am : membrane area (m2) ms : mass of solute (kg) α : specific resistance (m/kg)
Water Environment-Membrane Technology Lab., Seoul National University For unstirred conditions Rc grows and combining equation (1) with (2) gives, 1 dV ∆P J (t) = = −(3) Am dt (Rm +α ⋅VCb / Am )⋅η
from eqn.(3) VV t R ⋅η α ⋅VC ⋅η dt = m ⋅dV + b ⋅dV − (4) ∫0 ∫∫2 00∆P ⋅ Am ∆P ⋅ Am
At cons tant pressure, int egration of eqn.(4) gives, 2 Rm ⋅η α ⋅V Cb ⋅η t = ⋅V + 2 ∆P ⋅ Am 2∆P ⋅ Am
Rm ⋅η α ⋅V Cb ⋅η ∴ t /V = + 2 (well − known filtration equation) ∆P ⋅ Am 2∆P ⋅ Am Water Environment-Membrane Technology Lab., Seoul National University If Rm is negligible,
2 ⋅ ∆P ⋅ A2 V 2 = m ⋅t −(5) α ⋅Cb ⋅η
Which predicts that filtrate accumulates according to t1/2
Water Environment-Membrane Technology Lab., Seoul National University Determination of gel layer thickness
From the filtration model, it is possible to estimate the magnitude of the polarized layer (δs) in stirred ultrafiltration.
α : determined by unstirred experiment.
J,Rm : measurement by experiment. From eqn. (1),(2) ms/Am is obtained. ε : obtained from Cg
The effective thickness of the polarized solute layer (gel layer) ;
mS δ S = − (7) Am (1− ε ) ⋅ ρ S
Water Environment-Membrane Technology Lab., Seoul National University Experimental determination of α
Rm ⋅η α ⋅V Cb ⋅η t /V = + 2 ∆P ⋅ Am 2∆P ⋅ Am
α ⋅C ⋅η slope = b t/V 2 2⋅∆P ⋅ Am
V From a plot of t/v vs V, α is obtained experimentally by the slope.
Water Environment-Membrane Technology Lab., Seoul National University Properties of α
α properties are given by the Carman-Kozeny relationship:
180(1−ε) α = 2 3 ρS ⋅dS ⋅ε
α : specific resistance of biofilm
ε : Porosity ρ : density (floc density)
ds : size (floc diameter)
Water Environment-Membrane Technology Lab., Seoul National University Theoretical determination of α
∆P J = − (1) µ ⋅(Rm + RC )
if RC >> Rm , ∆P J = − (2) µ ⋅ RC
Water Environment-Membrane Technology Lab., Seoul National University Theoretical determination of α
from Carman-Kozeny Equation,
∆P K ⋅ µ ⋅(1−ε )2 ⋅ S 2 ⋅v = 3 − (3) δC ε
∆P : pressure drop (Pa)
δC : thickness of cake (m) ν : superficial velocity (m/sec) ε : porosity of filter media (dimensionless) µ : viscosity of fluid (Pa⋅sec) K : Kozeny-Carman constant ( ≅ 5)(dimensionless) S : specific surface (area/volume) of particle (1/m) d : particle diameter (m)
Water Environment-Membrane Technology Lab., Seoul National University π ⋅ d 2 If the particle is spherical, S = = 6 / d − (4) π ⋅ d 3 / 6
6 If the particle is non-spherical, S = Ψ ⋅ d
6/Ψ : shape factor Ψ= 1 for spherical particle. Ψ : sphericity (dimensionless) Surface area of equivalent-volume sphere = True surface area
Water Environment-Membrane Technology Lab., Seoul National University From eqn. (3) ∆P ε 3 v = 2 2 δ C K ⋅ µ ⋅ (1− ε ) ⋅ S
∆P ε 3 v = 2 2 δ C 5µ ⋅ (1− ε ) ⋅ (6 / d)
∆P d 2 ⋅ε 3 v = 2 −(5) δ C 180µ ⋅ (1− ε )
Water Environment-Membrane Technology Lab., Seoul National University 1 dQ J = = v Q : volume of flow (m3/sec) Am dt 2 Am : membrane area (m ) ∆P d 2 ⋅ε 3 J = 2 − (6) δ C 180µ ⋅ (1− ε )
From eq. (2) and (6)
180δ ⋅ (1− ε ) 2 R = C − (7) C d 2 ⋅ε 3
Water Environment-Membrane Technology Lab., Seoul National University RC = α ⋅ mS / Am − (8)
α : specific cake resistance (m/kg)
mS : mass of particle deposited (kg) 3 ρS : particle density excluding water (kg/m )
mS = δ C ⋅ Am ⋅ (1− ε ) ⋅ ρ S − (9) from eqn. (7), (8) & (9)
180 ⋅ (1− ε ) α = 2 3 − (10) ρ S ⋅ d ⋅ε
Water Environment-Membrane Technology Lab., Seoul National University - The effect of pressure on α is frequently expressed by the relationship.
s α0 : constant α = α0 ⋅∆P s : compressibility factor
- For compressible solids, typical values of s = 0.2~0.7 (Fig. 2)
∆Pg : pressure drop across the deposited solute
R ∆P = ∆P g = ∆P − R ⋅η ⋅ J g m Rm + Rg
Water Environment-Membrane Technology Lab., Seoul National University Fig. 2. Specific resistance vs Pressure
Water Environment-Membrane Technology Lab., Seoul National University Carman-Kozeny Equation
Hydraulic permeability(Ph) d 2 ⋅ ε 3 P = p h 180 ⋅−()1 ε 2
dp : particle diameter (m) ε: porosity of the cake layer (dimensionless)
Combined with Resistance-in-series model µε⋅−()1 2 R ∝ c 2 3 d p ⋅ ε μ : viscosity of fluid (Pa·sec)
Water Environment-Membrane Technology Lab., Seoul National University Models for predicting membrane performance (Flux)
1) Filtration Model
2) Resistance in Series Model
3) Film theory Model
Water Environment-Membrane Technology Lab., Seoul National University Resistance Models
- Hagen-Poiseuille and Film Theory Model does not describe the entire pressure-flux behavior, i.e. , pressure-controlled at low pressures, pressure-independent at high pressures. Therefore ultrafiltration performance is also frequently interpreted by Resistance-in-series relationship.
1) Gel-Polarized Model.
For an ideal membrane and solute, Hagen-Poiseuille equation can be rewritten as, ∆P J = Rm : Intrinsic Membrane Resistance determined Rm ⋅η using pure water as feed.
Water Environment-Membrane Technology Lab., Seoul National University Resistance Models
With a real feed, the membrane resistance by itself may be only a small part of the total resistance and there may be a series of additional resistances.
∆P ∆P J = = (Rm′ + RP )⋅η (Rm + R f + Rg + RBL )⋅η
Rf : Fouling layer resistance (specific membrane-solute interactions, either by surface deposition or pore fouling)
RP : Resistance attributable to the polarized solute. Rg : Resistance due to gel-polarized layer RBL : Resistance of the viscous, but non-gelled boundary layer.
Water Environment-Membrane Technology Lab., Seoul National University RP = Rg+RBL
RP = ø∆PT (a function of applied pressure) - i.e. any increase in ∆PT simply increases the thickness of the gel layer, and increase Rg.
∆P ∴ J = T (Rm′ +φ∆PT )⋅η
Water Environment-Membrane Technology Lab., Seoul National University Remark ;
1) At High pressure, RP >> R´m J will become independent of ∆PT and approach the limiting value 1/ø
2) At low pressure, when polarization is less (Cw< Cg)
Rg=0, ∴J is pressure-dependent. (i.e. ‘pre=gel’ condition) ∆P J = T (Rm′ + RBL )⋅η
3) This model suffer from having to obtain the constants experimentally.
Water Environment-Membrane Technology Lab., Seoul National University - Unit of Rm; ∆P J = Rm ⋅η ∆P N / m2 ∴ R = T = =1/ m m J ⋅η m3 N ⋅ ⋅sec m2 ⋅sec m2
Rm is useful for modeling purpose and for evaluating the effectiveness of the cleaning procedures.
Water Environment-Membrane Technology Lab., Seoul National University Resistance in series Models
∆P J = (Rm + R f + Rc )⋅η
Rm : Intrinsic Membrane Resistance
Rf : Fouling layer resistance (specific membrane-solute interactions, either by surface deposition or pore fouling) Å Bulk Compositions
Rc : Cake layer resistance Å Biofilm
.
Water Environment-Membrane Technology Lab., Seoul National University Models for predicting membrane performance (Flux)
1) Filtration Model
2) Resistance in Series Model
3) Film theory Model
Water Environment-Membrane Technology Lab., Seoul National University The Film-Theory Model
CW (Cg)
Cb
Water Environment-Membrane Technology Lab., Seoul National University The Film-Theory Model
i) Longitudinal mass transport within the boundary layer is assumed negligible (mass transfer within the film is one dimensional) ii) In the steady state, the solute flux is constant throughout the film and equal to the solute flux through the membrane.
• A material balance for the solute in a different element gives the equation.
J S = CP ⋅ JV = CJV − D(dC / dx)
JS : solute flux JV : solvent flux Cb : bulk solute concentration δ : thickness of the boundary layer
CP : permeate solute concentration Cw : solute concentration at the membrane surface D : solute diffusion coefficient
Water Environment-Membrane Technology Lab., Seoul National University Boundary Condition ; C = Cb at x=0 CW at x= δ
dC D ⋅ = J (C − C ) dx V P Cw dC δ C D ⋅ = J ⋅dx → D ⋅ln(C − C ) w =J ⋅δ ∫ C − C ∫ V P C V Cb P 0 b D C − C C − C w P w P ∴ JV = ln = kS ⋅ln δ Cb − CP Cb − CP
Cw JV = kS ⋅ln (if CP = 0) Cb
D/δ = kS ks : mass-transfer coefficient
Water Environment-Membrane Technology Lab., Seoul National University - If Cw → Cg
C g J lim = k S ⋅ ln (Gel polarizati on Model ) Cb
Cg : Gel Concentration Jlim : limiting flux
• Observed retention ; S=1-Cp/Cb , True retention ; R=1-Cp/Cw The concentration polarization ; M=Cw/Cb = 1-S+S exp(JV/ks) So, M can be calculated from the measurement of the retention and the permeate flux, when ks is known.
Water Environment-Membrane Technology Lab., Seoul National University Remark;
1) This model will be valid only in the pressure-independent region. (there is no pressure term) (Fig. 4.13)
2) Flux will be controlled by the rate at which solute is transferred back from the membrane surface into the bulk fluid.
3) Flux can only be improved by enhancing ks as much as possible, such as by reducing the thickness of the boundary layer (δ)
4) Cg is fixed by physicochemical properties of the feed.
Lm : permeability coefficient
Water Environment-Membrane Technology Lab., Seoul National University Water Environment-Membrane Technology Lab., Seoul National University Remark;
5) Jlim should be independent of membrane properties. It is determined by Cg, Cb and ks (hydrodynamic conditions) ( But Fane showed that ‘gel-polarized’ behavior with identical solutions
and hydrodynamics produced different Jlim values when membranes of differing permeability, Lm )
6) Feed solutions of various macrosolutes with concentration did not give zero flux.
7) If Cg is a ‘gel’ concentration, it should depend only on the nature of the solute, but it appears to vary with system-hydrodynamics. 8) This model appears to have physical limitations although it still remains the most convenient model from a practical point of view.
Lm : permeability coefficient Water Environment-Membrane Technology Lab., Seoul National University Flux Paradox
Film theory model – for macromolecule solutions
Water Environment-Membrane Technology Lab., Seoul National University Film theory model – for colloidal suspensions
Water Environment-Membrane Technology Lab., Seoul National University Flux Paradox
1) For macromolecular solutions, the agreement between theoretical and experimental ultrafiltration rates is within 15~30%.
2) For colloidal suspensions, experimental flux values are often one to two orders of magnitude higher than those indicated by the Lévéque and Dittus-Boelter relationships. But the reason is not clear.
3) In colloidal suspensions, the diffusion coefficient calculated from the ultrafiltrate flux using the Lévéque and Dittus-Boelter equations is generally from one to three orders of magnitude higher than the theoretical Stokes-Einstein diffusivity.
4) Minor adjustments in molecular parameters such as diffusivity (D), kinematic
viscosity (ν), or gel concentration (Cg) are incapable of resolving order of magnitude discrepancies.
Water Environment-Membrane Technology Lab., Seoul National University 5) Back-diffusive transport of colloidal particles away from the membrane surface into the bulk stream (D· ∂C/∂x) is substantially augmented over that predicted by the Lévéque and Dittus-Boelter relationships.
6) For colloidal suspensions, mass transfer from the membrane into the bulk stream is driven by some force other than the “concentration gradient”.
M.C. Porter’s opinion (1972) : Tubular Pinch Effect is responsible for this augmented mass transfer.
Water Environment-Membrane Technology Lab., Seoul National University Tubular Pinch Effect
• Segré and Silberberg : the first to publish the observations of the tubular pinch effect. (As the particles were flowing through a tube, the particles migrated away both from the tube wall and the tube axis, reaching equilibrium at an eccentric radial position.)
Water Environment-Membrane Technology Lab., Seoul National University Forces acting on a suspended particle
Charge repulsion Shear induced diffusion Inertial lift Diffusion
Axial drag Drag torque Axial velocity profile Permeation Sedimentation drag van der Waals attraction Semipermeable membrane
J J J
Fig.Ⅳ-1. Forces and torques acting on a charged, spherical particle suspended in a viscous fluid undergoing laminar flow in the proximity of a flat porous surface.[modified from Wiesner(1992)]
Water Environment-Membrane Technology Lab., Seoul National University Water Environment-Membrane Technology Lab., Seoul National University 10 -2
Back-transport velocity(m/s) 10 -3
10 -4
10 -5
10 -6
10 -7
10 -8 10 -3 10 -2 10 -1 10 0 10 1 10 2 Particle diameter (µm)
Fig. 7.7. Back-transport velocity based on different migration
Watermechanisms. Environment-Membrane Technology Lab., Seoul National University 10 -3
1000 Total 10 -4 sec ) / /hr )
100 2
m Interaction induced ( t r 10 -5 o sp Shear induced ort ( L/m Diffusion 10
10 -6 Back tran Lateral migration Back transp 1
10 -7 .01 .1 1 10 100
Particle size ( ㎛ )
Fig. Ⅳ-4. Comparison of different models explaining a critical flux over a range of particle size.
Water Environment-Membrane Technology Lab., Seoul National University Remark :
1) Crossflow microfilters are occasionally operated in the turbulent regime, whereas all of the models described are restricted to laminar flow.
2) Brownian and shear-induced diffusion may be considered simultaneously by adding the diffusion coefficients, although recent simulations have shown that the diffusion coefficients are not strictly additive.
3) The models described are based on idealized suspensions of equi-sized spheres, which do not irreversibly stick to the membrane or cake surfaces but rather are free to diffuse or lift away. Further experiments and models are needed to study Brownian, shear induced diffusion, and inertial lift in real suspensions of non-spherical, deformable particles having both narrow and broad size distributions.
Water Environment-Membrane Technology Lab., Seoul National University 4) Considerable experimental and theoretical research remains to complete our understanding of crossflow microfiltration.
For example, issue of direct membrane fouling by the attachment of particles and precipitates to the membrane pores and surface have not been adequately addressed.
Water Environment-Membrane Technology Lab., Seoul National University 14 Initial size distribution
V
o 12
l
u
m
e
t 10 r
i
c
f
r
e
q 8
u
e
n
c
y 6
∞ )
( 1200 r
%
h
/ 2
) 4 800 m
/
L
(
400
2 x
u l
0 F 0 110 Particle size ( ㎛ )
Fig. Ⅳ-5. Depositing particle size distribution for each flux condition. Size distribution when flux is infinite means the initial particle size distribution -20 measured experimentally (T = 298 K, Ψ0 = 50 ㎷, A = 3.4 x 10 J).
Water Environment-Membrane Technology Lab., Seoul National University Membrane Fouling: Identification and Prevention
MembrMembraanene ffooululiinngg
Physiochemical Microbiological Hydrodynamic Approach Approach Approach
9Quorum sensing 9Process Design 9Back transport velocity mechanism 9Chemical additives 9Critical flux 9Biofouling mechanism 9Membrane surface 9Cell physiology modification 9Micro-organism 9Hybrid system population dynamics
Water Environment-Membrane Technology Lab., Seoul National University Membrane Fouling: Identification and Prevention
Physicochemical Approach:
Examples for submerged MBR systems: • reduce flux (J) • increase membrane aeration • employ physical or chemical cleaning – backflushing (HF only) – relaxation (ceasing permeation whilst continuing aeration) – in-situ clean (chemically enhanced backwash) – ex-situ clean (soak) • all have cost implications
Water Environment-Membrane Technology Lab., Seoul National University Membrane Fouling: Identification and Prevention
Microbiological Approach
• Change of microbial characteristics :
- Floc morphology & size, - Physiological state, - EPSs(extracellular polymeric substances) content
Water Environment-Membrane Technology Lab., Seoul National University Overview of factors leading to membrane biofouling
Environmental Factors in MBR Biofilm Biofilm composition and properties(R c ) • Substrates structure • DO • Air flow rate & Microorganisms Size bubble size ? EPS Shape • pH ? Membrane • Temperature Organic debris Density • Growth mode Inorganic particles Porosity (attached or Ions suspended) BiofoulingJ • Growth phase Stickiness (log or ( ) endogenous) • Cyclic format in ? SBR ? Bulk phase composition ( R f ) • etc
Water Environment-Membrane Technology Lab., Seoul National University Research on MBR in 21C
Quorum sensing mechanism
Biofilm formation mechanism
Cell Morphoplogy & Physiology
Microorganism population dynamics
InnovativeInnovative MBRMBR processprocess
Water Environment-Membrane Technology Lab., Seoul National University