The Effect of Fluid Management on Membrane Filtration - ACS

27 The Effect of Fluid Management on Membrane Filtration

Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: May 21, 1981 | doi: 10.1021/bk-1981-0153.ch027

M. C. PORTER Nuclepore Corporation, 7035 Commerce Circle, Pleasanton, CA 94566

The advent of the Loeb-Sourirajan asymmetric membrane some twenty years ago gave birth to an industry now exceeding 200 million dollars in annual sales. Reverse osmosis (RO) and ultra filtration (UF) were previously only laboratory curiosities. To day, there are many large membrane plants (up to 16 million gal lons per day) in service for applications as diverse as desalinat ing seawater; concentrating serum proteins, or the recovery of paint and other by-products from waste streams. The break-through made by Loeb and Sourirajan was the for mation of a relatively thin skin (0.1 to 2 μm) on the surface of a more open sponge-like structure (Figure 1) which offers reduced resistance to flow while still maintaining the retention required. However, the structure of these membranes precludes use in a con ventional "flow-through" filtration mode due to the accumulation of retained species on the surface, rather than throughout the matrix of the filter (Figure 2). In the "flow-through" mode, the membrane will "plug" or "foul" much more rapidly than a conven tional depth filter because the particle/solute loading capacity is less. However, if the particulates or solutes accumulated on the surface can be dispersed back into the bulk fluid, these mem branes can be used to great advantage since there is relatively little i f any "internal-fouling" of the membrane structure. There is a high probability that a molecule or particle which penetrates the skin will not be trapped within the filter structure but will pass through into the filtrate. Schematically, the pores may be represented by ever-widening cones with no internal constrictions to restrain molecules or particles. Since the successful exploitation of these membranes has been largely dependent on effective fluid-management techniques, i t seemed appropriate for this symposium to review developments in this field over the last twenty years.

0097-6156/81/0153-0407$10.50/0 © 1981 American Chemical Society

In Synthetic Membranes:; Turbak, A.; ACS Symposium Series; American Chemical Society: Washington, DC, 1981.


408

SYNTHETIC

Figure 1.

MEMBRANES:

DESALINATION

Cross-section of Type A asymmetric UF membrane from a Nuclepore

Figure 2. Accumulation of particulates through the matrix of a depth filter and on the surface of an asymmetric membrane


27.

PORTER

Fluid

Management


C o n t r o l of Concentration Techniques

409

P o l a r i z a t i o n With S t i r r i n g and Cross-Flow

The simplest l a b o r a t o r y apparatus f o r reducing the accumulat i o n of r e t a i n e d species on the membrane (sometimes r e f e r r e d to as concentration p o l a r i z a t i o n ) i s s t i l l the s t i r r e d c e l l (Figure 3). Stable f l u x values may be obtained by a magnetically d r i v e n s t i r r i n g bar suspended above the membrane; otherwise, the f l u x declines rapidly. I t i s i n t e r e s t i n g to note that one of the f i r s t i n d u s t r i a l UF u n i t s o f f e r e d by ABCOR was a one f o o t d i a meter drum with a high-speed r e c i p r o c a t i n g s t i r r e r l o c a t e d between two one-foot diameter membranes to minimize concentration p o l a r i z a t i o n (Figure 4) (1) . C u r r e n t l y , the most common f l u i d management technique used i n i n d u s t r y i s "cross-flow through tubes or channels with the f l u i d v e l o c i t y t a n g e n t i a l to the membrane surface (Figure 5 ) . The hydrodynamic shear f o r c e s at the membrane s u r f a c e tend to r e duce the boundary l a y e r and keep the membrane c l e a n . In reverse osmosis, where the s o l u t e s r e t a i n e d are r e l a t i v e l y low i n molecular weight and have a s i g n i f i c a n t osmotic pressure, concentration p o l a r i z a t i o n can r e s u l t i n osmotic pressures considerably higher than those represented by the bulk stream concentration. Higher pressures are required to overcome the osmotic pressure (Figure 6 ) . 11

m where i s the water f l u x through the membrane, AP i s the t r a n s membrane pressure drop, AIT i s the osmotic pressure d i f f e r e n c e across the membrane, and R i s the membrane r e s i s t a n c e to permeat i o n . Thus, higher c r o s s - ? i ow v e l o c i t i e s adjacent to the membrane w i l l tend to decrease the boundary l a y e r and the p o l a r i z a Q t i o n modulus __s where C i s the concentration of r e t a i n e d species c

b at the surface of the membrane and C^ = bulk stream concentration of the r e t a i n e d s p e c i e s . The net r e s u l t i s to i n c r e a s e the f l u x and reduce the s a l t passage through the membrane-which i s more dependent on C than on C^. For u l t r a f i l t r a t i o n , the macromolecular s o l u t e s and c o l l o i d a l species u s u a l l y have i n s i g n i f i c a n t osmotic p r e s s u r e s . In t h i s case, the concentration at the membrane s u r f a c e (C ) can r i s e to the point of i n c i p i e n t g e l p r e c i p i t a t i o n , forming a dynami c secondary membrane on top of the primary s t r u c t u r e (Figure 7 ) . This secondary membrane can o f f e r the major r e s i s t a n c e to flow. g



410

SYNTHETIC

Figure 3.

MEMBRANES:

DESALINATION

Conventional stirred-cells

Figure 4. ABCOR drum ultrafilter with high-speed reciprocating stirrer (I)


27.

PORTER

Fluid

411

Management

ULTRAFILTRATE

t t t t t t t t t

MEMBRANE

7} BOUNDARY • ••-V LAYER * *• • ••lw RETENTATE

•

FEEO IN

\ BOUNDARY J LAYER Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: May 21, 1981 | doi: 10.1021/bk-1981-0153.ch027

MEMBRANE

1

4 I I I I I I 4

i

ULTRAFILTRATE Figure 5.

Crossflow fluid management in tubular membrane configuration

LOW SOLUTES J (=J s

w

C

C )

LAMINAR

SUBLAYER

midi

Figure 6.

p

J «K

« C b

w

m

(AP-Air)«

p

A v r o

MEMBRANE »l

Concentration polarization with RO membranes

MEMBRANE

HIGH M. W SOLUTES

—>

J ,C w

p

Cn«Ch

»

m -. .

LAMINAR rAlfC SUBLAYER CAKE (NON N E W T O N I A N )

Figure 7.

AP

(

lkJc

(SLIME)

Concentration polarization with UF membranes


(AP-AT)

412

SYNTHETIC

MEMBRANES:

DESALINATION

In the steady-state c o n d i t i o n , t h i s " g e l - l a y e r " , as i t i s sometimes c a l l e d , w i l l grow i n thickness u n t i l the pressure-acti v a t e d convective transport of s o l u t e with solvent toward the membrane surface (^ C) j u s t equals the concentration g r a d i e n t a c t i v a t e d d i f f u s i v e transport away from the surface D -j— ( F i g ure 8 ) . W

In cases where the l i m i t i n g r e s i s t a n c e to flow i s the " g e l l a y e r " and the concentration a t the membrane s u r f a c e ( C ) i s f i x e d a t a constant g e l concentration (C ) , t h i s simple d i f f e r e n t i a l equation: ^ Downloaded by UNIV OF CALIFORNIA SAN DIEGO on June 2, 2015 | http://pubs.acs.org Publication Date: May 21, 1981 | doi: 10.1021/bk-1981-0153.ch027

g

J C = D i£w s dx

(2)

may be solved to y i e l d the r e s u l t D

j

=

w

^

6

C l

n

C

^ C

= b

k

l

n

(3)

JL

C

fe

where 6 i s the boundary l a y e r t h i c k n e s s , and k i s the mass-transf e r c o e f f i c i e n t . Thus, under g e l - p o l a r i z e d c o n d i t i o n s , the water f l u x (J^) i s i n v a r i a n t with transmembrane pressure drop or perm e a b i l i t y and i s dependent only on the boundary l a y e r thickness (6) and s o l u t e p r o p e r t i e s such as d i f f u s i v i t y (D ) and g e l concent r a t i o n (C ) . The ?nvariance with pressure i s demonstrated i n Figure 9. Above some threshold pressure, f l u x becomes independent of pressure. Any i n c r e a s e i n pressure w i l l cause a t r a n s i e n t i n c r e a s e i n f l u x which r e s u l t s i n more s o l u t e transport to the membrane. Since the g e l - c o n c e n t r a t i o n (C ) cannot increase and the backd i f f u s i v e transport away from fee membrane i s unchanged, an accumulation of s o l u t e at the membrane w i l l thicken the g e l - l a y e r and increase r e s i s t a n c e to flow u n t i l the f l u x i s reduced to i t s former v a l u e . I t i s a l s o obvious from Figure 9 that the backd i f f u s i v e transport c o n t r o l s the f l u x through the membrane. Inc r e a s i n g the s t i r r e r speed improves the mass transport away from the membrane by reducing the boundary l a y e r thickness ( 6 ) . Lower bulk stream concentrations (C^) improve the mass transport by i n c r e a s i n g the concentration g r a d i e n t . Lower concentrations become g e l - p o l a r i z e d at higher threshold pressures. Equation 3 i n d i c a t e s that a semilog p l o t of f l u x against concentration should be a s t r a i g h t - l i n e i n t e r c e p t i n g the h o r i z o n t a l a x i s at the g e l concentration (C ) . When the bulk-stream concentration (C^) equals the g e l concentration (C ) there i s no d r i v i n g f o r c e f o r removal of s o l u t e from the membrlne. The g e l l a y e r increases i n thickness u n t i l the f l u x i s zero. Figure 10 provides experimental confirmation f o r a number of p r o t e i n s o l u t i o n s and c o l l o i d a l suspensions (_2) . The i n t e r c e p t s with the h o r i z o n t a l a x i s are reasonable values f o r the g e l c o n c e n t r a t i o n .


27.

PORTER

Fluid

413

Management

Cb

JC-D


•

£ -o t •

• • •• •• t • • • • •

Figure 8.

Steady-state gel polarization model

Figure 9. Flux-pressure data for BSA



414

SYNTHETIC

MEMBRANES:

DESALINATION

P r o t e i n s o f t e n form gels at concentrations between 20 and 30 percent. Likewise the " g e l c o n c e n t r a t i o n " f o r c o l l o i d a l p a r t i c l e s should be equivalent to a value f o r close-packed spheres between 60 and 75 percent. Higher t a n g e n t i a l v e l o c e t i e s (or r e c i r c u l a t i o n rates) should decrease the boundary l a y e r thickness (6) and i n c r e a s e the mass t r a n s f e r c o e f f i c i e n t (k) i n Equation 3 r e s u l t i n g i n higher slopes of the f l u x vs concentration curve (Figure 11) without changing the g e l - c o n c e n t r a t i o n . Even with e f f e c t i v e cross-flow fluid-management techniques, membrane f l u x w i l l u s u a l l y decay with time due to tenacious dep o s i t s on the membrane s u r f a c e . I t has been observed that the rate of decay i s minimized with high t a n g e n t i a l v e l o c i t i e s . Further, detergent c l e a n i n g of membrane surfaces i s a c c e l e r a t e d by higher t a n g e n t i a l v e l o c i t i e s (Figure 12) ( 3 ) . I t should be obvious from the above that fluid-management techniques which improve the mass-transfer c o e f f i c i e n t (k) w i t h minimum power consumption are most d e s i r a b l e . However, i n some cases, low-cost membrane c o n f i g u r a t i o n s with i n e f f i c i e n t f l u i d management may be more cost e f f e c t i v e . In any case, i t i s important to understand q u a n t i t a t i v e l y how t a n g e n t i a l v e l o c i t y and membrane/hardware geometry a f f e c t s the mass-transfer c o e f f i c i e n t . The mass and heat t r a n s f e r analogies make p o s s i b l e an e v a l u a t i o n of the mass-transfer c o e f f i c i e n t (k) and provide i n s i g h t i n t o how membrane geometry and f l u i d - f l o w c o n d i t i o n s can be speci f i e d to optimize f l u x ( 4 ) . For laminar flow: k

= i -

6

2 d

TJO-33 rjO'67 0 33 0 33 h *

( 4 )

T

where^U = t a n g e n t i a l f l u i d v e l o c i t y h = e q u i v a l e n t h y d r a u l i c diameter of the tube or channel L = channel or tube l e n g t h For t u r b u l e n t flow: TJO-08 k

=

°-

0 2 d

0.2 h

0

n * v - * 0

1

6 7

7

(5)

where v i s the kinematic v i s c o s i t y Equations 4 and 5 have been used to p r e d i c t f l u x values f o r a v a r i e t y of macromolecular s o l u t i o n s and channel geometries ( 4 ) . The t h e o r e t i c a l values were i n good agreement with the experiment a l v a l u e s . F i g u r e 13 i l l u s t r a t e s the 0.33 power dependence on w a l l shear r a t e per u n i t channel length (U/d, L ) .



27.

PORTER

Figure 12.

Fluid

415

Management

The effect of detergent circulation rate (O, 1.51L/s, 0.57L/s) and time on Flux (3)

1.14L/s, A ,



416

SYNTHETIC

MEMBRANES:

DESALINATION

0.3 0.2 0.1 0.08 0.06 ^

0.04

x => 0.02

g £

o.oi ->© 0.008 '

E 0.006

+

•

•

c: 0.004 h

•

O • o

= 0.002[ 0.001 10

CHANNEL CHANNEL CHANNEL LENGTH (cm) DEPTH (in.) TYPE 0.010 41 RECTANGULAR RECTANGULAR 6.5 0.005 TRIANGULAR 6.5 0.017 TRIANGULAR 6.5 0.035 TRIANGULAR 13 0.005 TRIANGULAR 13 0.017 TRIANGULAR 13 0.035 TUBULAR 40 0.008

1

100 WALL SHEAR RATE CHANNEL LENGTH

Figure 13.

1000 y/L,

10,000

(sec-cm)"

Flux dependence on wall shear rate in laminar flow


27.

PORTER

Fluid

Management

All


Membrane-Hardware Geometry The c o n s i d e r a t i o n s above have been u t i l i z e d by v a r i o u s manuf a c t u r e r s i n d e s i g n i n g e f f i c i e n t u l t r a f i l t r a t i o n and reverse osmosis equipment. The most s t r a i g h t forward and t r o u b l e - f r e e geometry i s a simple tubular c o n f i g u r a t i o n (see F i g u r e 14) w i t h diameters l a r g e enough to pass a l l extraneous matter. As a p r a c t i c a l matter, oneinch tubes may be e a s i l y cleaned by sponge-rubber b a l l s which are forced through the tubes w i t h h y d r a u l i c p r e s s u r e . The disadvantages are cost and volume per u n i t area. P l a t e and frame systems o f f e r a great d e a l of f l e x i b i l i t y i n o b t a i n i n g smaller channel dimensions. Equations 4 and 5 show that the i n c r e a s e d hydrodynamic shear a s s o c i a t e d with r e l a t i v e l y t h i n channels improves the mass-transfer c o e f f i c i e n t . Membrane r e p l a c e ment costs are low but the l a b o r i n v o l v e d i s h i g h . For the mostp a r t , p l a t e and frame systems have been troublesome i n high-pressure reverse osmosis a p p l i c a t i o n s due to the p r o p e n s i t y to l e a k . The most s u c c e s s f u l p l a t e and frame u n i t from a commercial standpoint i s that manufactured by The Danish Sugar C o r p o r a t i o n L t d . (DDS) (Figure 15). The labor i n t e n s i v e replacement of membranes i n p l a t e and frame systems has been f a c i l i t a t e d i n the "leaf-module" design of Dorr O l i v e r (Figure 16). Here a number of p l a t e s are assembled i n a d i s p o s a b l e c a r t r i d g e where the process stream flows over the p l a t e s and the permeate i s ducted to a common header. Tubular systems can a l s o be converted to thin-channel devices w i t h the use of "volume displacement rods". In one such design (7), manufactured by Amicon and Romicon, a s p l i n e d core has the membrane wrapped around i t , s e a l e d , and braided to form t h i n channels between the core and the membrane (Figure 17). These braided tubes are then potted i n a s h e l l and tube module where the permeate i s c o l l e c t e d on the s h e l l s i d e ( F i g u r e 18). As with any r e s t r i c t e d channel system, s t r a i n e r s or pref i l t e r s must be provided to remove d e b r i s which would otherwise c l o g the channels. Thin-channel tubular modules such as that p i c t u r e d i n F i g u r e 18 provide extremely high f l u x - v a l u e s at h i g h t a n g e n t i a l v e l o c i t i e s . However, to achieve maximum performance, the power requirements f o r pumping are high; t h i s i s p a r t i a l l y due to l a r g e " p a r a s i t e drag" a s s o c i a t e d with the s p l i n e d core. The advent of hollow f i b e r membranes provided a low-cost membrane element with n e g l i g i b l e p a r a s i t e drag. In the case of u l t r a f i l t r a t i o n , optimum f l u i d management d i c t a t e d flowing the process stream down through the lumen of the h o l l o w - f i b e r ( F i g u r e 19); the s k i n i s on the i n s i d e w a l l . T h i s i s not p o s s i b l e f o r reverse osmosis dufe to the high pressure of the process stream; i n t h i s case, the feed stream i s on the o u t s i d e of the hollow f i b e r and the permeate flows through the f i b e r lumen (Figure 20).



418

SYNTHETIC

MEMBRANES:

DESALINATION

EPOXY RE* IN FORCED FIBERGLASS BACKING Figure 14.

ABCOR

1-in. tube (1)


Fluid

Management


PORTER

Figure 15.

DDS plate and frame UF module (5)



420

SYNTHETIC

Figure 16.

MEMBRANES:

DESALINATION

Dorr Oliver leaf-module design (6)

American Institute of Chemical Engineers Symposium Series

Figure 17.

Thin-channel tube with splined core (1)



PORTER

Fluid

Management

421

American Institute of Chemical Engineers Symposium Series

Figure 18.

Figure 19.

Thin-channel tubular module (1)

Nuclepore Type A UF hollow-fiber showing the direction of flow


422

SYNTHETIC

MEMBRANES:

DESALINATION


U l t r a f i l t r a t i o n h o l l o w - f i b e r modules are u s u a l l y made with a s h e l l and tube c o n f i g u r a t i o n . The f i b e r s are potted at both ends of the module with the f i b e r lumen open f o r r e c i r c u l a t i o n of the process stream (Figure 21). N a t u r a l l y , s t r a i n e r s or p r e f i l t e r s must be u t i l i z e d to e l i m i n a t e plugging of the f i b e r s . At Nuclepore, i t has been shown that l a r g e r diameter hollow f i b e r s , 1.5 to 3mm i n i . d . , are much l e s s prone to f o u l i n g . F o r t u n a t e l y , a l l UF hollow f i b e r systems can be back-washed and are amenable to a number of c l e a n i n g techniques. The b a s i c membrane/hardware geometries-tubes, p l a t e and frame, and hollow f i b e r s can of course a l l be operated w i t h f l u i d management techniques which are r e l a t i v e l y e f f i c i e n t or n o n - e f f i c i e n t . The remainder of t h i s paper w i l l adress i t s e l f to novel fluid-management techniques which u t i l i z e the b a s i c membrane/hardware geometries but which seek to augment the mass-transfer coeff i c i e n t even f u r t h e r . Pulsed-Flow

F l u i d Management

Kennedy et a l (8) have achieved permeation i n c r e a s e s of more than 70 percent while concentrating sucrose s o l u t i o n s with reverse osmosis by superimposing a harmonic pulse generator (Figure 22) on the normal flow down a 1.31 cm i . d . tubular membrane. T y p i c a l r e s u l t s are shown i n Figure 23. T h i s study suggests that pulsed r e verse osmosis may o f f e r a more economical means f o r reducing membrane area. Kennedy c a l c u l a t e d that to d u p l i c a t e the 80 percent permeation i n c r e a s e of Figure 23 by the conventional means would r e q u i r e a s i x - f o l d increase i n v e l o c i t y . At Nuclepore, we have found that a h o l l o w - f i b e r module ( F i g ure 24) can be operated i n a dead-ended, through-flow mode on tap water with a 30 minute c l e a n i n g pulse every 24 hours to r e s t o r e flux (Figure 25). The e x i t v a l v e at the end of the u n i t p i c tured i n Figure 24 i s simply opened up to f l u s h out the i n s i d e of the hollow f i b e r s . With DI water, the c l e a n i n g pulse was not r e quired u n t i l a f t e r one-week of o p e r a t i o n . T h i s f l u i d management technique e l i m i n a t e s the need f o r pumps or power; the r e s u l t i s a small compact u n i t u s e f u l i n removing b a c t e r i a or pyrogens. Turbulence Promoters Considerable i n t e r e s t has been generated i n turbulence promoters f o r both RO and UF. Equations 4 and 5 show considerable improvements i n the mass-transfer c o e f f i c i e n t when operating UF i n turbulent flow. Of course the penalty i n pressure drop i n c u r red i n a t u r b u l e n t flow system i s much higher than i n laminar flow. Another way to i n c r e a s e the mass-transfer i s by i n t r o d u c i n g t u r bulence promoters i n laminar flow. T h i s procedure i s p r a c t i c e d e x t e n s i v e l y i n enhanced heat-exchanger design and i s now e x p l o i t e d i n membrane hardware design.


27.

PORTER

Fluid

Management


HOLLOW FIBER PRINCIPLE

Figure 20.

RO hollow-fiber showing direction of flow

Figure 21.

Nuclepore hollow-fiber module end potting


423

S Y N T H E T I C

424

M E M B R A N E S :

D E S A L I N A T I O N

Permeate Membrane

k


aiming

section

|—I'I—

^

Jacket

/

Pressure gauge

s

Piston

v3 Scotch

-r5i—— C o n c e n t r a t e

Pressure control valve

yoke

By-pass Moyno pump

Feed tank

Piston pump

Temperature .control ( 25°C)

Chemical Engineering Science Figure 22.

Schematic of pulsed-flow RO equipment (8)

10

J

2 0 Frequency,

L

3 0

J_

4 0

50

60

70

cycles/min Chemical Engineering Science

Figure 23.

RO permeation rate as a function of pulsing frequency ($)



27.

PORTER

Fluid

425

Management

Figure 24.

Nuclepore BST-1 hollow-fiber module

60

100

200

300

400

500

600

THROUGH PUT (GALS)

Figure 25.

Tap water flux decay and regeneration with flush on BST-1


426

SYNTHETIC MEMBRANES:

DESALINATION

Thomas and Watson (9) of the Oak Ridge N a t i o n a l Laboratory have shown that detached s p i r a l wire turbulence promoters, p o s i tioned away from an RO membrane s u r f a c e by small wire runners, markedly increased the r e j e c t i o n of s a l t s and the permeation r a t e through the membrane. "Detached promoters" of t h i s type may be designed to minimize stagnant regions; i n a d d i t i o n , they are r e l a t i v e l y easy to i n s t a l l . Figure 26 presents t h e i r r e s u l t s f o r the r e j e c t i o n of 0.01 M MgC&2 400 p s i on a dynamic membrane as a f u n c t i o n of | J ^ R e ^ * ^ where U i s the l i n e a r v e l o c i t y down the tube, v i s the permeation r a t e , and N i s the Reynolds number. I t w i l l be noted that the greatest e f f e c t of the turbulence promoter was observed at the lowest v e l o c i t i e s , where the r e j e c t i o n increased from 25 to 72 percent. At the highest t a n g e n t i a l v e l o c i t i e s , the improvement was much l e s s , from 90 to 93 percent. In a d d i t i o n , Thomas and Watson observed an i n c r e a s e i n permeation r a t e v a r y i n g from 10 to 50 percent. Thus, with turbulence promoters, the same r e j e c t i o n and f l u x , as i n an unpromoted system, may be obtained at a considerable reduction.


a t

25

P r o b s t e i n et a l (10) i n v e s t i g a t e d the use of detached s t r i p type turbulence promoters i n the u l t r a f i l t r a t i o n of bovine serum albumin i n laminar flow. His apparatus i s shown i n Figure 27; the detached s t r i p type promoters tested were c i r c u l a r c y l i n d e r s with a diameter (D) approximately one-half (0.46) of the channel height and were across the center of the channel c r o s s - s e c t i o n , transverse to the flow. T y p i c a l f l u x data with two interpromoter spacings (AL) are shown i n Figure 28 as a f u n c t i o n of the cross-flow r a t e . The f l u x increased by a f a c t o r of 3 f o r the best case. Though P r o b s t e i n d i d not p l o t h i s data i n t h i s way, i t i s i n t e r e s t i n g to note that the empty channel f l u x has a p r e d i c t a b l e 0.33 power dependence on tangential velocity. With the turbulence promoters, the slope s h i f t s c l o s e r to the 0.7-0.8 power dependence normally observed i n turbulent flow. U n f o r t u n a t e l y , data are not a v a i l a b l e i n Probs t e i n ' s paper on the increased pressure drop a s s o c i a t e d with the turbulence promoters, but i t would appear that the f l u x to power r a t i o i s g r e a t l y improved with turbulence promoters. In current p r a c t i c e , turbulence promoters most o f t e n take the form of a net or screen m a t e r i a l which also serves as a feed chann e l spacer between two membranes. For example, the f a m i l i a r s p i r a l wound modules (Figures 29) used e x t e n s i v e l y i n reverse osmosis and to a l e s s e r extent i n u l t r a f i l t r a t i o n use a p l a s t i c screen m a t e r i a l as the feed channel spacer. This i s a l s o used i n some p l a t e and frame systems (Figure 30). S p h e r i c a l turbulence promoters have a l s o been used i n tubular systems (Figure 31). In the author's experience, these s p h e r i c a l promoters are not as e f f e c t i v e as detached s p i r a l wire promoters (9). P a r t i c l e s c o l l e c t i n the dead stagnant areas between the spheres and f o u l the system. The spheres serve more e f f e c t i v e l y as volume displacement spheres than as turbulence promoters.


PORTER

Fluid

427

Management to

/

/

f 25

/

t


50,

/

A IT

8

AV

90 a?

-i

r

n

{/

O N O TUP B U L E N C E P R O M O T E R • D E T A C H E D T U R B U L E N C E PROMC) T E H

1

1

f

1

Industrial and Engineering Chemistry Process Design & Development

Figure 26.

Effect of detached spiral turbulence promoters on rejection of 0.01 MgCl at 400 psi (9) 2

Desalination

Figure 27.

Probstein's apparatus for investigating the effect of turbulence promoters in laminar-flow VF (10)



428

SYNTHETIC

B S A

2 %

1 CROSS

MEMBRANES:

DESALINATION

DATA OF PROBSTTIN E T A I .

i — 4 FLOW

i RATE

i

U

U

(L/MIH) Desalination

Figure 25.

Fjffcc* o/ turbulence promoters on flux in UF of BSA (10)

Figure 29.

Spiral-wound module construction (1)



27.

PORTER

Figure 30.

Fluid

429

Management

Plate-and-frame cassette system using a screen spacer between membranes (11)

Figure

31.

Spherical turbulence promoters (1)


430


Static

SYNTHETIC MEMBRANES:

DESALINATION

Mixers

A disadvantage of many of the i n s e r t s used f o r turbulence promotion i s the high pressure-drop and/or stagnant regions ass o c i a t e d with t h e i r o b s t r u c t i v e nature. Middleman et a l (13) (14) have s t u d i e d the use of s t a t i c mixers - a twisted tape i n s e r t (Figure 32) w i t h a l t e r n a t i n g r i g h t and l e f t - h a n d p i t c h which m i n i mizes the two drawbacks noted above. S t a t i c mixers set up a h e l i c a l f l o w - p a t t e r n , which e s t a b l i s h e s secondary flows (perpendicular to the main d i r e c t i o n of f l o w ) . F u r t h e r , the p e r i o d i c a l t e r n a t i o n of the flow generates v o r t i c e s which f u r t h e r enhance mass-transfer i n the neighborhood of the tube w a l l . The i n s e r t i o n of a s t a t i c mixer i n a 0.53 i n c h RO tubular membrane, f o r feed concentrations of 0.08 to 0.35m NaC£, showed an i n c r e a s e i n f l u x of about 25 percent a t N =20 and of 266 percent at N =1500. The T l u x improvement was even more dramatic i n the case of u l t r a f i l t r a t i o n . A one-inch tube was used to u l t r a f i l t e r a 1 percent polymer l a t e x emulsion. F i g u r e 33 shows the improvement i n f l u x at v a r i o u s Reynolds numbers with the s t a t i c mixer and F i g u r e 34 shows the r a t i o of f l u x e s (promoted/empty tube) as a f u n c t i o n of Reynolds number. The r e s u l t s are c o n s i s t e n t with the expected behavior of the s t a t i c mixer. At very low Reynolds numbers, the s w i r l i n g flow generated by the mixer i s i n s u f f i c i e n t to a l t e r convection at the membrane surface to any a p p r e c i a b l e degree. At very high Reynolds numbers, the l e v e l of turbulence i n the unpromoted tube i s so high that the a d d i t i o n of a s w i r l i n g component of flow does not b r i n g about a major r e d u c t i o n i n the g e l l a y e r r e s i s t a n c e . Hence, there i s an intermediate Reynolds number where the f l u x r a t i o i s maximized. Figure 34 shows a maximum f l u x r a t i o of 4.9 at a Reynolds number of 1 0 . 6

4

Secondary Flow One of the b e n e f i t s of s t a t i c mixers l i k e those mentioned above are the secondary flow patterns set up by the s w i r l i n g h e l i c a l flow. Secondary flow, i . e . flow perpendicular to the main d i r e c t i o n of flow a l s o occurs whenever f l u i d passes through a curved tube or channel. The phenomenon i s caused by c e n t r i f u g a l f o r c e s , and can be r e a d i l y understood by r e f e r r i n g to F i g u r e 35, the c r o s s s e c t i o n of a h e l i c a l l y c o i l e d tube. Near the tube's center the a x i a l v e l o c i t y i s g r e a t e s t , causing c e n t r i f u g a l f o r c e s to act most s t r o n g l y . F l u i d i s thrown outward and replaced by r e c i r c u l a t i n g f l u i d which flows inward along the w a l l s . In both laminar and turbulent flow, two strong symmetrical patterns are normally e s t ablished .



27. PORTER Fluid Management


431

SYNTHETIC

M E M B R A N E S *.

DESALINATION


432

Figure 35.

Secondary flow patterns in the cross-section of a helically coiled tube


27.

PORTER

Fluid

Management

433

I t i s w e l l known (15, 16, 17) that c o i l e d tube heat exchange r s possess s u p e r i o r heat t r a n s f e r c h a r a c t e r i s t i c s because of secondary flow e f f e c t s . At small curvatures, the Dean number, d e f i n N = N -V ed as De Re ~ vhere a and R are tube and c o i l r a d i i respect i v e l y , governs tne transport processes i n c o i l e d tubesjhigher Dean numbers causing higher t r a n s p o r t r a t e s . Dravid et a l (17) has shown experimentally that the heat t r a n s f e r c o e f f i e c i e n t i n a c o i l e d tube v a r i e s as i n the f u l l y developed r e g i o n of the boundary l a y e r . By analogy, the mass-transfer c o e f f i c i e n t f o r laminar flow i n s p i r a l or c o i l e d channels should vary as N r a t h e r than as N as p r e d i c t e d i n equation 4 (see reference 4 ) . UF data (human albumin-laminar flow) taken by the author i n a s p i r a l t h i n channel u n i t (Figure 36) confirm the 0.5 power dependence on the Reynolds number (Figure 37) r a t h e r than the usual 0.33 power dependence p r e d i c t e d by equation 4 and observed experimentally i n l i n e a r thin-channel u n i t s (Figure 38). S r i n i v a s a n and T i e n (18) have made an a n a l y t i c a l study on the mass-transfer c h a r a c t e r i s t i c s of reverse osmosis i n curved tubular membranes. The i n c r e a s e i n mass-transfer due to secondary flow r e s u l t e d i n a s u b s t a n t i a l r e d u c t i o n i n the w a l l c o n c e n t r a t i o n (the p o l a r i z a t i o n modulus) f o r N^ =100 and a/R=0.01 (see F i g u r e 39). Further, the production c a p a c i t y (permeation rate) was markedly increased (see Figure 40). In the reverse osmosis f i e l d , only one commerical design has e x p l o i t e d the e f f e c t of secondary flow and i t i s no longer a v a i l able (Figure 41). a


7

0

0

,

3

,

5

3

e

Particulate

Scouring

Davies (20) used powdered a c t i v a t e d carbon i n c o n j u n c t i o n with u l t r a f i l t r a t i o n of a c t i v a t e d sludge to adsorb s o l u b l e organic c o n s t i t u e n t s which might otherwise pass through the membrane u n t i l the biomass can metabolize them. The r e d u c t i o n i n e f f l u e n t COD i s shown i n Figure 42. An a d d i t i o n a l b e n e f i t of the a c t i v a t e d carbon was the scouring of the membrane s u r f a c e to remove membrane f o u l a n t s . Not only was the exponential decay i n f l u x a r r e s t e d , the f l u x was r e s t o r e d to i t s i n i t a l value as shown i n Figure 43. B i x l e r and Rappe (21) a l s o obtained UF data showing that i n creased f l u x values could be obtained by i n t r o d u c i n g s o l i d p a r t i c u l a t e m a t e r i a l s i n t o the process stream (Figure 44). Bixler argued that these p a r t i c l e s augmented the back d i f f u s i o n from the membrane s u r f a c e by mixing of higher concentration s o l u t e s near the membrane with lower concentration s o l u t e s more remote. He a l s o recognized the scouring a c t i o n of the beads i n c l e a n i n g the membrane. T h i s author has p r e v i o u s l y suggested that the mechanism of f l u x enhancement might be due to the " t u b u l a r pinch e f f e c t " ( 4 ) . The lower d e n s i t y (0.94 g/cc) MMA beads show l e s s tendency to


434

SYNTHETIC

MEMBRANES:


ULTRAFILTRATE

DESALINATION

MEMBRANE

" S K I N " SIDE O F ANISOTROPIC MEMBRANE

CONTACTS

DIAFLO

PROCESS

S T R E A M IN C H A N N E L

FLOW C H A N N E L SPACER

T O RECYCLE RESERVOIR Industrial and Engineering Product Research & Development

Figure 36.

Spiral thin-channel unit (4)



PORTER

Fluid

435

Management

RECIRCULATION RATE I

I I I

(CC/MIN)

I Industrial and Engineering Product Research & Development

Figure 37.

UF flux as a function of recirculation rate for UF of HSA in a spiral thin-channel unit (4)

RECIRCULATION RATE (GPM)

-l

2

3

1

1—j

I

I I

I

4 5fi7 8 10

-l

20

1 i 1 ii i 30 40 50 Industrial and Engineering Product Research & Development

Figure 38.

UF flux as a function of recirculation rate for UF of HSA in a linear thin-channel unit (4)


436

SYNTHETIC

MEMBRANES:

DESALINATION

4.0 B, B

2

X

* 7.4074

- 0.25 - 0.01

( N R ^ J J . 1000

3.0


STRAIGHT

TUBE

2.0 CURVED

4.0

O

TUBE

12.0

8.0

6.0

Desalination

Figure 39.

Wall concentration (dimensionless) vs. axial coordinate for straight and curved tubes fl8j

o^o

1 B, X

- 7.4074 -o.oi

CURVEO TUB!

The Effect of Fluid Management on Membrane Filtration - ACS

Recommend Documents