1. history of s aehan in dustries, inc. 2. introduction to ... · pressure vessel, cleaning agents...

0T able of Co nte nts

T ec h n o lo g y E x pr es s C S M

1. H istory of S AE H A N In dustries , Inc.

2. Int roduction t o Rever se O s m os is M em bran es

3. SA EH A N 's M em bran e E lem ents and F ilter C art rid ges

4. W at er C hem istry and P re tre atm ent

5. System D esign

6. System O p er at ion

7. Cleanin g in Place , C leaning an d D isin fec tion

8. Replacing E lem ents, As sem b ly and L oading of P re ssure V essel s.

9. Troubles hooting

10. Ap pen dix

CSMR ev e rse O s m o sis M e m b ra ne

1. In t r o d u c t i o n to

S A E H A N In d u s t r ie s .

1-1 . G eog rap h ic a l Loca tio n of the B us i ne ss

U ni ts o f the C om p an y

1-2 . H is to ry of the C om p a ny

1-3 . H is to ry of R eve rse O sm o sis M em b ran e

D eve lo p m en t of the C om p a ny

1-4 . O rgan iz atio n C ha rt o f the M em b ran e

B us in es s U nit

1In troduct ion to SA EH A N Industries

T e c h n o log y Ex p re s s C S M- 1 -

1 -1 . O ff ic e a nd P la n ts

■ Headquater :

Saehan Building 254-8 Kongduk-dong, Mapo-gu, Seoul, Korea.

■ R & D :

c/o Samsung Central Research, 14 Nongsuri, Kihung-up, Yongin-goon,

KyungKi-do, Korea.

■ Kyungsan Plant :

1 Joongsan-dong, Kyungsan City, Kyungbuk, Korea.

Products : Elite(fabrics), Elasto(elastic fabrics) Fabrics for Y shirts,

Codray products, R ev e rse O s m o sis M em b ra n e s.

■ Kumi Plant :

287 Gongdan-dong, Kumi city, Kyungbuk, Korea.

Products : Films for Audio, Video, and Packaging. Polyester fabrics, fibers, and

chips(for the production of bottles)

1. O ff ic e a n d P lan ts

1 In troductio n to SA EHA N In dustries

G lo b a l In it iat iv e 2 1 - 2 -

1 -2 . H is to r y o f S a e ha n Ind u s tr ie s

■ 1972 : Founded as Cheil Synthetics Co.,a subsidiary of Samsung Conglomerate.

■ 1974 : Kumi plant 1 began to produce PET staple fibers.

■ 1977 : Became a public company.

■ 1982 : Began to produce PET filaments.

■ 1985 : Began to produce PET films.

■ 1987 : Exported a PET plant to Barodaraion Co. In India.

■ 1988 : Exported a Weaving plant to Yasonta Co. In India.

■ 1988 : Founded Cheil Ciba-Geigy as a subsidiary.

■ 1989 : Started a joint venture with Yasonta Co. In India.

■ 1990 : Started PP non-woven fabrics and Spundex fabric.

■ 1991 : Built a plant for Baroda Rayon Co. In India.

■ 1994 : Kumi plants passed ISO-9002.

■ 1 99 4 : S u cc e ss fu l R & D w o rk o n r ev e rse o s m o s is (R O )

m e m b ran e p ro d u c t io n a n d s a les o f h o m e R O ele m en ts .

■ 1 99 5 : K u m i P lan t 2 w a s b u ilt fo r th e p ro d u c t io n o f P E T , P E T f ilm , P E T f ib e r,

P S n o n -w o v en , an d inc in e rat o rs.

■ 1 99 6 : B e g an th e p ro d u ct io n an d sa le s o f in d u st rial R O ele m en ts .

■ 1 99 7 : R e ce ive d IR 52 A w a rd (a K o rea n p r es t ig io u s inv e n t io n a w a rd ) an d K T

m a rk (n a t io n al ce rt if ic at io n o f h ig h q u a lit y) fo r n e w R O m e m b ra n es .

■ 1 99 7 : T h e m e m b ra n e p ro d u ct io n fac ility in K y u n g s an w a s c ert if ie d fo r IS O

1 40 0 1 an d 90 0 2.

■ 2 00 0 : W Q A m e m b e r a n d N S F ce rt if ic at io n .

2. H is to ry of S a e h a n In d u s tr ies

1In troduct ion to SA EH A N Industries

T e c h n o log y Ex p re s s C S M- 3 -

1 -3 . H s ito r y o f S a e ha n R O M e m br a ne D e v e lo p m e nt

■ 1990 - Started basic research on RO membranes at the R&D

center of Cheil Synthetic Co.

■ 1993 - Membranes showing high salt rejection and high flux were developed.

■ 1994 - Began the production of the 13 inch wide RO membrane and completed

RO element test lines.

- Manufactured home RO elements(supplied to samsung for the production

of home RO systems), which received a Samsung Technology Award.

- Received approval from Japanese FDA in March and FDA in June

■ 1995 - Began to build the manufacturing facility in Kyungsan to produce

the 40 inch wide RO membrane for industrial brackish water and

seawater applications.

■ 1996 - Began to produce the 40 inch long RO elements for industrial

applications

- Exported the elements to Asian countries including China, Taiwan and

India.

- Special purchase agreements with ASI, Argo and

Ionics were set up in regard to parts for the RO systems including

pressure vessel, cleaning agents and antiscalants.

■ 1997 - Received IR 52 Award(a Korean prestigious invention award) and

KT mark(national certification of high quality).

Increased production capacity of 8040 elements for industrial applications

to 40,000 per year at Kyungsan plant.

- Increased market share for home RO elements to 50% in Korea.

- Received ISO 14001 and 9002 approval.

3. H s i to ry of S a e h a n RO M e m b ra n e D e ve lop m e n t

1 In troductio n to SA EHA N In dustries


1 -4 . Or ga n iza tio n C ha rt o f F il te r C om pa ny D iv is io n o f

S AE H AN Ind u s tr ie s .

4. Org a n iza tion C h a rt of F i lte r C o m p a n y D iv is ion of

S A E H A N In d u stries.

S a le s

Membrane Team

(Domestic Sale)Membrane

Export Team

MicrofiltrationTeam

(Domestic andExport)

Filter C o mpany

TechnicalService and

Development

Membrane

Development

P rod u c tio n

M e m br a ne

R & D

S y s te m

B us in e s s

T e c hn ic a l

S up p o rt

Microfilter

Production

Membrane

Production

Large ScaleDrinking

Water ProductionSystem

(System Set upand Parts)

Large Scale IndustrialRO System

(System Design, Partsand Accessories)


2. In t r o d u c t io n to R e v e r s e

O s m o s is RO M e m b r a n e

2-1. O vervi ew of R O M em b ran e

Ap li ca tio n s

2-2. The ory o f R O M em b ran e

2-3. Type s o f R O M em b ran e

2-4. G en eral P rop e rty of S aeh a n

R O M em b rane s(C S M )

2-5. S ol u te R e jec tio n Prope rty o f

C S M R O M em b ran es

2In troductio n to R everse O sm osis R O M e mbrane

T ec h n o lo g y E x pr es s C S M- 6 -

2 -1 . Ov e r v ie w of R O M e m br ane App l i c a tio n s

R a pi d te ch n ol o g i ca l a dv a nc e m e n t b rou g h t ab o ut m an y be n efi ts to our life a n d also

co n cu r re n tl y c au s ed p o l lu ti on to o u r e n v iron m en ts s uc h as c on tam in ati n g w a te r so u rce s

fro m i n d us try eff l u e nts .

M o reo v er , na tu ral fr es h w a ter s ou rce co u l d not m e et the e v er i nc rea s in g w a te r

de m an d fro m our g row in g p o pu l ati o n a nd i n du s tr ie s . A d di t i on a l ly , u ne v en s e as o na l ra i n

fall m ak e s the w ate r s h or ta g e pr ob l em w o rse .

D is tillat i o n ha s b e en tra d it i o na l ly u s ed to o b tai n p u re w a ter from co n ta m in a ted w ate r

so u rce s . O th er p roc e sse s su c h as i o n e xch an g e an d e l ec tro di a l ys is h av e b e en

em p l oyed for w ate r p u r if i ca ti o n s i n ce 19 5 0. R ec e ntl y , RO m em b ra ne w as a p p l ie d to the

de s al ina ti on of s e a w a ter an d b ra ck ish w a ter .

RO h a s b ee n p ro ve n to be the m o s t ec o no m ical tec h no l og y not o nl y for the

de s al ina ti on of w ate r c o nta i ni n g s al ts , but also for pu r ify i n g w a ter co n tam ina te d with

he a vy m e ta l s, pe s ti c id e s an d o the r c o n ta m in a nts .

A n d a l so RO c a n be u s ed for rec yc lin g w a s te w ate r a n d re c la im ing u s efu l m a te r ial s

fro m w as te s tr ea m s uc h as the rec o ve ry of d ye s from dye i nd u s try e ff l ue n ts . T h e

ap p l icati o ns for RO also in c lu d e foo d a nd be v era g e p roc e ss in g . In a dd i t io n , a dv a nc e s

in th e fi el d s of b io te ch n ol o g y an d p ha rm ac e uti cal s, co u pl e d w i th a dv a nc e s in ne w

m e m br an e d ev e l op m en t, a re m a k ing m e m b ran e s an im p o r ta n t s ep a ra ti o n s tep , w h i ch

off e rs en e rg y sa v in g s w i tho u t l ea d i ng to th e rm a l d eg ra d ati o n of the p ro du c ts .

1. O v e rview of RO M e m b ra n e A p p lica tions

2 In troductio n to R everse O sm osis R O M em brane


2 -2 . T he ory o f R e v e rs e O s m os is M e m bra ne

T h e p he n om e n on of o s m o s is is i llu s tra te d in the F i g ure be l ow .

A se m ip e rm e a bl e m e m bra n e (R O m e m br an e ) is p l ac e d be tw e en two

co m pa r tm en ts . An RO m e m b ra n e is co n s iste d of a s up p o r ti n g l a ye r w i th 50 ㎛ in

th i ckne s s an d a b ar r ie r l a ye r with ab o ut 0 .2㎛ in th ickn e ss . T h e ph e no m e no n of o sm os is

oc cu rs w h en pu re w a te r f lows from a di lute s a l in e s ol u tio n in o ne co m pa r tm e nt th ro ug h

the RO m e m bra n e i nto a h i g he r c o nc e ntra te d sa l in e s o lu ti on in th e o the r c a u s ing a

rise in the h e i gh t of the sa l t s o lu ti on in th e c om p a r tm e n t of the h i g he r co n ce n tr ate d

so l uti o n.

T h e w a ter f l ow will s to p w h en the p re ssur e of the c ol u m n of the s a l t s ol u tio n e q ua l s

to th e d i ff ere n ce in c h em ica l p ote n ti a l b e tw ee n the two a qu e ou s s ol u ti on s . T h e

eq u ilibr iu m po i n t of the w ate r c o l um n h ei g h t in term s of w a te r pre s su re a ga i n s t the

m e m br an e is c al led o s m o ti c pre s su re .

If a force is ap p l ie d to thi s c ol u m n of w a ter , the d i rec ti on of w a te r f l o w th ro ug h the

m e m br an e c an be re v er se d. Th i s ph e n om e no n is c al led rev e rse o s m o s is . T hi s re v ersed

fl o w pro d uc e s a p ure w ate r fro m the s al t s o lu ti o n, s i n ce th e m em b ra ne is n o t

pe rm ea b le to s a l t

2. Th e o ry of R e v e rse O sm o s is M e m b r a ne

RO membrane

Osmosis

Dilute

Solution

H2O H2O

Pressure

Concentrated

Solution

Dilute

Solution

Concentrated

Solution

RO membrane



2 -3 . T y pe s of R O M e m bra ne s

◎ A s y m m e tric M em b ra n e - - - C e llu lo s e A c eta te (C A ) M em b ra n e

H i sto r ica l ly, the a s ym m e tr ic m e m bra n e is form e d by c a s ti n g a thi n fi lm a c eto n e-

b as e d s ol u tio n of c e llu lo s e a ce tate (C A ) po l ym e r , w h i ch w as d e v el o pe d by Lo e b an d

S o ur ira ja n in 1 9 62 a n d the first c o m m e rcia lly v i a b l e RO m em b ra ne .

T h e res u l ti n g CA m e m b ran e h a s an a s ym m e tr ic s tru c tu re w i th a d e ns e s ur fa c e

l a ye r of a b ou t 0.1 - 0.2 ㎛ w h i ch is re s po n s ib le for th e s al t rej e c ti o n p rop e r ty. T h e

rest of the m e m bra n e, w h ich is 1 00 -2 00㎛ th i ck a nd su p p or ts the thi n s u r fac e l a ye r

m e ch a ni cal ly, is s p on g y an d p or ou s , a nd h a s hi g h w a ter p erm e ab i lity. S a l t re j e c ti o n

a nd w ate r fl u x of a CA m e m b ran e c an be c o ntro l le d by va r iati o ns in te m pe ra ture

a nd du ra tio n of the an n ea l in g s tep .

◎ T h in F ilm C o m p o s it e M e m b ran e - - - P o ly am id e (P A ) M e m b ra n e

T h in fi lm co m po s ite (TFC ) p o lyam ide m e m b ran e s are co n s iste d of a p oro u s

s up p o r t l a ye r a n d a thi n fi lm d e ns e l a ye r w hi ch is a c ro s s li n k ed m e m bra n e s k in

a nd is fo rm ed in s i tu on the p or ou s s up p o r t l a ye r , u su a l ly m ad e of p ol ysu l fon e .

T h e thi n fi lm d e ns e l a ye r is a c ro s slin ke d a ro m a ti c po l ya m id e m ad e fro m i n te r fa c ia l

p ol ym er iza ti on r ea c ti o n of a p ol y f un c tio n al am ine su c h as m -p h en yl e n ed i a m in e with

a p o ly f u nc ti on a l a c id c hl o r id e s u ch as tr im e s oy l c hl o r id e . T h is T F C m an u fac tur in g

p ro ce d ure en a bl e s i nd e pe n de n t o pti m izati o n of the d i sti nc t p ro p er ti e s of the s u pp o r t

a nd sa l t re j ec ti ng s k in . T h e T F C m e m b ran e is c ha ra c te r ize d by h i g he r s p ec if ic

w a te r f l ux an d h ig h e r s al t r ej e c ti o n tha n c e llu lo s e a ce tate m e m bra n e s .

◎ C o m p a riso n o f P o ly am id e T F C M e m b r an e s w ith C e llu lo se A ce ta te( C A )M e m b ran e s .

As m en ti on e d ab o ve , th e F T C m e m br an e s ex h i bi t h i g he r w ate r f l u x a nd h i gh e r

s al t rej e c ti o n th an CA m em b ra ne s w h i ch h a d be e n u se d w i d el y u nti l the

c om m er cial i n tro du c tio n of T F C m e m b ran e s in 19 8 1. T F C m e m bra n es a re s ta bl e

o ve r a w i d er pH ra n ge a n d op e ra bl e at l o w e r p re ssure th a n CA m e m bra n es .

D e ta ile d co m p ar iso n s be tw e e n the two typ es of m e m bra n es a re s h ow n in the ta bl e

b el o w .

3. Typ e s of RO M e m br a n es



P a ram e te rs P A M e m b ran e C A M e m b ran e

O p e ra t in g p H ra ng e 2 ~ 12 4~ 6

O p e ra t in g P re ss u re (K g /㎠ ) 15 30

S alt

R eje c t ion

(% )

T D S 99 + 98

S ilic a (S iO 2) 99 + < 95

S a lt R e je ct io n C h an g e

a f te r 3 y ea rs9 9 % →98 .7% 98 % →9 6%

C h lo rine T o le ran c e < 0.1 p p m 1 p pm

M em b ra n e F o u lin g H i g h Lo w

3. Types of RO Membranes



2-4. General Property of Saehan RO Membranes(CSM)

◎ G en e ra l S e pa ra t io n P ro p e rty o f R O M e m b ra n e

1) In o rg an i c so l ute s are re j ec ted by RO m e m b ra n e be tter tha n o rg an i c so l u te s .

O rg a ni c s ol u tes w ith m o l ec u l ar w e ig h t(M W ) l ar ge r th a n 10 0 a re a l so well

rej e c te d by the m e m b ra n e

2) Io n i za bl e s o lu te s are rej e c te d b e tte r th a n no n - io n i za bl e s o l ute s .

3) Io n i za bl e s o lu te s with hi g h er ch a rg es a re re j e c te d b ette r th an l o w e r c h ar ge s .

Fo r ex a m pl e s , al u m in u m i on (A l 3+) is re j ec te d be tter tha n m a gn e s iu m

io n (M g 2+) which is in tu rn re j e c te d b ette r th an s o di u m i o n( N a+).

4) Th e re j e c ti o n of in o rg an i c so l ute s d ep e nd also on th e s ize of the i o n s an d

the s i ze of h yd ra te d i o ns . T h e b i gg e r th e i o ns a n d the h yd ra ted i o n s , the

be tter the y are rej e c tte d

5) Th e b i gg e r the n o n- io ni zab l e so l u tes (the h i g he r th e m ol e c ul a r w ei g h t) , the

be tter the re j e c ti o n.

6) G a se s w i th MW l o w e r th an 1 0 0 c an e a s ily p e rm e a te th ro u gh the m em b ra ne .

Fo r ex a m pl e , the rej e c ti o n of a m m o ni a , c hl o r ine , c a rbo n d i ox ide , o x yg e n an d

hydro g en su l f id e is v er y low.

7) Th e re j e c ti o n of w e ak a c id s is low, w h i ch a l so d ep e n ds on the MW of the

ac id s . T h e re je c tio n of the fol low in g a c id s is de c rea s in g in the ord e r of c i tr ic

ac id , ta r ta r ic a c id a nd a c eti c a c id as the MW of the a c ids de c re a se s .

◎ C h a ra cte ris t ic P ro p e rty o f C S M R O M e m b ra n es .

1) H i g h pe rm ea te fl ux a n d h ig h s a lt rej e c ti o n .

2) C he m ica lly s ta b le in a w i de r an g e of pH (p H 2 - 12)

3) Lo n g m e m b ran e life tim e .

4) R es ista nt to a bi o l og i ca l a tta c k.

5) O p er ab l e in a w i d e ran g e of pr es su re (2 0 - 10 0 0p s i)

6) O p er ab l e at a w i d e ran g e of tem p er atu re (4 - 45 ℃)

7) Ec o n om ica l

4. General Property of Saehan RO Membranes(CSM)



2-5. Solute Rejection Property of CSM RO Membranes

◎ S o lu te R e jec t io n o f B E a n d B N G ra d e o f C S M

N O S o lu te R e jec t io n (% ) M o le c u lar W e ig h t

1 N a F 99 42

2 N aC N 98 49

3 N aC l 99 58

4 S i O 2 99 60

5 N a H C O 3 99 84

6 N a N O 3 97 85

7 M g C l 2 99 95

8 C a C l 2 99 11 1

9 M gS O 4 99 12 0

10 N i S O 4 99 15 5

11 C uS O 4 99 16 0

12 F o rm a l de h yd e 35 30

13 M e tha n ol 25 32

14 E tha n ol 70 46

15 Is o p rop a no l 92 60

16 U rea 70 60

17 La c tic a c id (p H 2 ) 94 90

18 La c tic a c id (p H 5 ) 99 90

19 G l u c os e 98 18 0

20 S u c ro s e 99 34 2

21 C hl o r in a ted P e s tic id e s 99 -

22 B O D 95 -

23 C O D 97 -

5. Solute Rejection Property of CSM RO Membranes

※ T h e s ol u te re j ec ti on b a se d on the fo l lo w i n g test c o nd i t i on s . :

2 00 0 pp m s o l ute , 2 25 p s i, 25 ℃, a nd p H 7

CSMReverse Osmosis Membrane

3. Product Specifics ofCSM Membranes

3-1. CSM Element Nomenclature

3-2. List of Customers Using

CSM Elements

3-3. Brackish Water Elements

3-4. Sea Water Elements

3-5. Nanofiltration(NF) Elements

3-6. Ultrafiltration(UF) Elements

3-7. Tap Water Elements

3-8, Microfilteration(MF) Cartridges

3P roduc t S pecifics of CSM Membranes

Tec h no logy Express CSM- 17 -

1. CSM Element Nomenclature

RE8040-BN

① Types of Element

- RE : Reverse Osmosis Element

- UE : Ultrafiltr ation Element

- NE : Nanofiltration Element

② Diameter of Element

- 80 : 8.0 inch

- 40 : 4.0 inch

- 25 : 2.5 inch

③ Length of Element

- 40 : 40 inch

- 21 : 21 inch

④ Element Applications

- B : for Brackish Water

- T : for Tap Water

- S : for Sea Water

⑤ Related to Flux and Rejection

- N : Normal (Regular)

- E : Extended Membrane Area(High Flux)

- R : High Rejection

- L : Low Pressure Element(High Flux)

3 P roduc t S pec ifics of CSM Membranes

G lobal Ini tiative 21 - 18 -

2. Brackish Water Elements

Parameter UNITBRACKISH WATER MEMBRANES

RE2521-BN RE2540-BN RE4021-BN RE4040-BN RE8040-BN

Performance

Flux

GPD 250 600 800 2000 9000

m3/D 0.9 2.3 3.0 7.6 34.1

Salt rejection % 98.0 98.0 98.0 99.0 99.5

Test

Conditions

pressure psig 225 225 225 225 225

temperature ℃ 25 25 25 25 25

NaCl(MgSO 4) ppm 2000 2000 2000 2000 2000

pH - 6.5-7 6.5-7 6.5-7 6.5-7 6.5-7

recovery % 15 15 15 15 15

Operating

Limits

max. SDI - 5 5 5 5 5

max. turbidity NTU 1 1 1 1 1

max. chlorine concentration ppm 0.1 0.1 0.1 0.1 0.1

max. feed flow rate m3/hr 1.3 1.3 4.0 4.0 15

min. concentration flow m3/hr 0.22 0.22 0.91 0.91 3.6

max. pressure drop/element psig 15 15 15 15 15

max. temperature ℃ 45 45 45 45 45

pH range - 3-10 3-10 3-10 3-10 3-10

max. recovery % 20 20 20 20 20

Size

effective areaft 2 12 27 35 75 345

m2 1.1 2.5 3.3 7.0 31.8

element length inch 21 40 21 40 40

element diameter inch 2.5 2.5 4.0 4.0 8.0

permeate tube inner diameter inch - - - - 1.125

permeate tube outer diameter inch 0.75 0.75 0.75 0.75 -

permeate tube intrusion inch 1.1 1.0 1.1 1.0 0.51

Max.

Permeate

Flow

surface water

GPD 210 470 630 1340 6180

m3/D 0.8 1.8 2.4 5.1 23.5

softened well water

GPD 260 630 790 1710 7890

m3/D 1.0 2.4 3.0 6.5 30.0

RO permeate

GPD 340 760 970 2080 9630

m3/D 1.3 2.9 3.7 7.9 36.6




Parameter UNIT

BRACKISH WATER MEMBRANES

RE4040-BL RE8040-BL RE4040-BE RE8040-BE

Performance

Flux

GPD 2600 12000 2200 11000

m3/D 9.8 45.4 8.3 41.6

Salt rejection % 99.0 99.0 99.0 99.5

Test

Conditions

pressure psig 150 150 225 225

temperature ℃ 25 25 25 25

NaCl(MgSO 4) ppm 2000 2000 2000 2000

pH - 6.5-7 6.5-7 6.5-7 6.5-7

recovery % 15 15 15 15

Operating

Limits

max. SDI - 5 5 5 5

max. turbidity NTU 1 1 1 1

max. chlorine concentration ppm 0.1 0.1 0.1 0.1

max. feed flow rate m3/hr 4.0 15 4.0 15

min. concentration flow m3/hr 0.91 3.6 0.91 3.6

max. pressure drop/element psig 15 15 15 15

max. temperature ℃ 45 45 45 45

pH range - 3-10 3-10 3-10 3-10

max. recovery % 20 20 20 20

Size

effective areaft 2 85 400 85 400

m2 7.9 37.2 7.9 37.2

element length inch 40 40 40 40

element diameter inch 4.0 8.0 4.0 8.0

permeate tube inner diameter inch - 1.125 - 1.125

permeate tube outer diameter inch 0.75 - 0.75 -

permeate tube intrusion inch 1.0 0.51 1.0 0.51

Max.

Permeate

Flow

surface water

GPD 1808 8484 1530 7180

m3/D 6.4 32.1 5.8 27.3

softened well water

GPD 2304 10822 1950 9160

m3/D 6.7 40.9 7.4 34.8

RO permeate

GPD 2413 13216 2370 11200

m3/D 9.1 50.0 9.0 42.4




Parameter UNITTAP WATER MEMBRANES

RE2521-TE RE2540-TE RE4021-TE RE4040-TE RE4040-TL

Performance

FluxGPD 300 700 1050 2000 2200

m3/D 1.1 2.6 4.0 7.6 8.4

Salt rejection % 98.0 98.0 98.0 98.0 98.0

Test

Conditions

pressure psig 225 225 225 225 150

temperature ℃ 25 25 25 25 25

NaCl(MgSO 4) ppm 2000 2000 2000 2000 2000

pH - 6.5-7 6.5-7 6.5-7 6.5-7 6.5-7

recovery % 15 15 15 15 15

Operating

Limits

max. SDI - 5 5 5 5 5



max. feed flow rate m3/hr 1.3 1.3 4.0 4.0 4.0




pH range - 3-10 3-10 3-10 3-10 3-10

max. recovery % 20 20 20 20 20

Size

effective area

ft 2 12 27 35 75 85

m2 1.1 2.5 3.3 7.0 7.9



permeate tube inner diameter inch - - - - -

permeate tube outer

diameterinch 0.75 0.75 0.75 0.75 0.75


Max.

Permeate

Flow

surface water

GPD 210 470 630 1340 1340

m3/D 0.8 1.8 2.4 5.1 5.1

softened well water

GPD 260 630 790 1710 1710

m3/D 1.0 2.4 3.0 6.5 6.5

RO permeate

GPD 340 760 970 2080 2080

m3/D 1.3 2.9 3.7 7.9 7.9



3. Sea Water Elements

Parameter UNITSEA WATER MEMBRANES

RE2521-SN RE2540-SN RE4021-SN RE4040-SN RE8040-SN

PerformanceFlux

GPD 225 500 600 1500 6200

m3/D 0.9 1.9 2.3 5.7 23.4

Salt rejection % 99.2 99.2 99.2 99.2 99.2

Test

Conditions

pressure psig 800 800 800 800 800

temperature ℃ 25 25 25 25 25

NaCl(MgSO 4) ppm 32000 32000 32000 32000 32000

pH - 6.5-7 6.5-7 6.5-7 6.5-7 6.5-7

recovery % 8 8 8 8 8

Operating

Limits

max. SDI - 5 5 5 5 5



max. feed flow rate m3/hr 1.3 1.3 4.0 4.0 15




pH range - 3-10 3-10 3-10 3-10 3-10

max. recovery % 10 10 10 10 10

Size

effective area

ft2 12 27 35 74 310

m2 1.1 2.5 3.3 6.9 29.0



permeate tube inner diameter inch - - - - 1.125

permeate tube outer inch 0.75 0.75 0.75 0.75 -

permeate tube intrusion inch 1.1 1 1.1 1 0.51

Max.

Permeate

Flow

surface water

GPD 180 420 550 1100 4970

m3/D 0.7 1.6 2.1 4.2 18.9

softened well water

GPD 260 570 760 1530 6790

m3/D 1.0 2.2 2.9 5.8 25.8

RO permeateGPD 340 760 970 1950 8660

m3/D 1.3 2.9 3.7 7.4 32.9



4. Nanofiltration(NF) Elements

Parameter UNITNANOFILTRATION MEMBRANES

NE2540 NE4021 NE4040 NE8040-N NE8040-E

PerformanceFlux

GPD 1000 1300 2700 12000 14000

m3/D 3.8 4.9 10.2 45.4 52.9

Salt rejection % 99.5(40) 99.5(40) 99.5(40) 99.5(40) 99.5(40)

Test

Conditions

pressure psig 225 225 225 225 225

temperature ℃ 25 25 25 25 25

NaCl(MgSO 4) ppm 2000 2000 2000 2000 2000

pH - 6.5-7 6.5-7 6.5-7 6.5-7 6.5-7

recovery % 15 15 15 15 15

Operating

Limits

max. SDI - 5 5 5 5 5



max. feed flow rate m3/hr 1.3 4.0 4.0 15 15




pH range - 3-9 3-9 3-9 3-9 3-9

max. recovery % 20 20 20 20 20

Size

effective area

ft2 27 35 75 345 345

m2 2.5 3.3 7.0 31.8 31.8



permeate tube inner diameter inch - - - 1.125 1.125

permeate tube outer

diameterinch 0.75 0.75 0.75 - -


Max.

Permeate

Flow

surface water

GPD 530 680 1500 6890 8038

m3/D 2.0 2.6 5.7 26.2 30.4

softened well water

GPD 760 970 2080 9630 11235

m3/D 2.9 3.7 7.9 36.6 42.5

RO permeate

GPD 950 1200 2600 12000 14000

m3/D 3.6 4.6 10.0 45.7 52.9



5. Ultrafiltration(UF) Elements

Parameter UNITULTRAFILTRATION MEMBRANES

UE4040-PF UE8040-PF

PerfomanceFlux

GPD 3500 14000

m3/D 13.2 52.9

Rej' Molecular weight(MWCO) - 50K~100K 50K~100K

Test

Condition

pressure psig 20 20

temperature ℃ 25 25

Concentration - pure water pure water

pH - 6.5-7 6.5-7

Operating

Limits

max. SDI - 5 5

max. turbidity NTU 1 1

max. chlorine concentration ppm 10 10

max. feed flow rate m3/hr 4.0 16

min. concentration flow m3/hr 0.91 3.6

max. pressure drop/element psig 15 15

max. temperature ℃ 45 45

pH range - 3-10 3-10

max. recovery % 20 20

Size

effective area

ft 2 75 345

m2 7.0 31.8

element length inch 40 40

element diameter inch 4.0 8.0

permeate tube inner diameter inch - 1.125

permeat tube outer diameter inch 0.75 -

permeat tube intrusion inch 1.0 0.51

Max.

Permeate

Flow

surface waterGPD 2600 12000

m3/D 9.9 45.7

softened well water

GPD 3000 17700

m3/D 11.4 67.2

RO permeate

GPD 3360 19900

m3/D 12.8 75.6



◎ S P IRAL W OUND TYPE M INI RE VE RS E OS M OS IS M ODULE

6. Tapwater RO Elements

FULX REJECTION PRESSURE

(GPD) (%) (psig) Temp. Feed TDS Recovery

(℃) ppm % A B C D E

RE1810-30 30 96 60 25 250 10~20 45±.5 256 25 14 17

RE1810-50 50 96 60 25 250 10~20 45±.5 256 25 14 17

RE1812-25 25 96 60 25 250 10~20 45±.5 298 22 22 17

RE1812-35 35 96 60 25 250 10~20 45±.5 298 22 22 17

RE1812-60 60 96 60 25 250 10~20 45±.5 298 22 22 17

RE1812-80 80 96 60 25 250 10~20 45±.5 298 22 22 17

RE2012-100 100 96 60 25 250 10~20 48.5±.5 298 23 12 17

RE1812-LP 33 93 20 25 100 10~20 45±.5 298 22 22 17

RE2010-LP 33 93 20 25 100 10~20 48.5±.5 298 23 12 17

RE2012-LP 50 93 20 25 100 10~20 48.5±.5 298 23 12 17

RE3010 30 92 15 25 100 10~20 84±.5 238 27 8 17

TEST CONDITION SIZE SPEC(mm)

MODEL NO



■ Operating Limits

- Maximum Operating Pressure 125psig

- Maximum Feed Flow Rate 2gpm(RE3010:6.0gpm)

- Maximum Operating Temperature 45℃

- Feed Water pH Range 3.0~10.0

- Maximum Feed water Density Index 1.0NTU

- Maximum Feed Silt Density Index 5.0

- Free Chlorine Tolerance <0.1ppm

※ Consult Saehan for other operating conditions.

FEED

E

Concentrate

Permeate

D C

B

A

O-RING

6. Tapwater RO Elements

CSMReverse Osmosis Membrane

4. Water Chemistry andPretreatment

4-1. Introduction

4-2. Feed Water Analysis

4-3. Prevention of Scale Formation

4-4. Solubility Product Calculations

4-5. Silica Scale Prevention

4-6. Colloidal Fouling Control

4-7. Biological Fouling Prevention

4-8. Organic Fouling Prevention

4-9. Jelcleer : Positively ChargedMedia Filter

Technology Express CSM

4Water Chemistry and Pretreatment

- 23 -

4-1. Introduction

The efficiency and life of a reverse osmosis(RO) system depends on effective

pretreatment of the feed water. The pretreatment includes any process which can

minimize fouling, scaling and membrane degradation to optimize product flow, salt

rejection, product recovery and operating costs.

Fouling is the entrapment of particulates such as inorganic and organic colloids.

Examples of the inorganic colloids are iron floc, silica, clay and silt. The organic

colloids are mostly consisted of organic polymers and microorganisms.

Scaling is the precipitation and deposition within the RO system of sparingly soluble

salts such as calcium carbonate(CaCO3), calcium sulfate(CaSO4) and barium sulfate

(BaSO4).

A proper pretreatment scheme for the feed water will depend on feed water source,

feed water composition and application. There are three types of feed water. Well

water has a low Silt Density Index(SDI) (typically < 2) and low bacteria count, thus

requiring a simple pretreatment scheme. Surface water, on the other hand, is

characterized by a high SDI and can have a high bacteria count. Pretreatment for

surface water is more elaborate than for well water, requiring the addition of a

coagulant, clarification, and multimedia filtration.

Once the feed water source has been determined, an accurate analysis of the feed

water composition including bacteria count should be done to determine the dosage

size of a coagulant, a scale inhibitor(antiscalant) and a biocide for the pretreatment

1. Introduction

Global Initiative 21

4 Water Chemistry and Pretreatment

- 24 -

4-2. Feed Water Analysis

A complete and accurate water analysis must be provided before an RO system is

designed.

The water analysis report should contain the type and the concentration of all

constituents in the water. The constituents are consisted of dissolved ions, silica,

colloids, and organic(TOC)

■ Typical dissolved anions are as follows :

bicarbonate(HCO3-), carbonate(CO32-), hydroxide(OH-), sulfate(SO4

2-), chloride(Cl -),

fluoride(F-), nitrate(NO 3-), sulfide(S2-) and phosphate(PO4

2-)

■ Typical dissolved cations are shown below :

calcium(Ca2+), magnesium(Mg 2+), sodium(Na +), potassium(K+), iron(Fe2+ or Fe3+),

manganese(Mn2+), aluminum(Al 3+), barium(Ba2+), strontium(Sr 2+), copper(Cu 2+) and zinc

(Zn2+).

Certain combinations of cations and anions form sparingly soluble salts in water and

scaling of a reverse osmosis membrane may occur when the salts are concentrated

within the RO element beyond their solubility limit. Typical sparingly soluble salts and

their solubility product limit are shown in Table 1.

Substance Formula Temp.(℃ ) Solubility Product

Aluminium Hydroxide

Barium Carbonate

Barium Sulfate

Calcium Carbonate

Calcium Fluoride

Calcium Sulfate

Cupric Sulfide

Ferric Hydroxide

Ferrous Hydroxide

Magnesium AmmoniumPhosphate

Magnesium Carbonate

Magnesium Hydroxide

Manganese Hydroxide

Strontium Carbonate

Strontium Sulfate

Zinc Hydroxide

Al(OH)3

BaCO3

BaSO4

CaCO3

CaF2

CaSO4

CuS

Fe(OH)3

Fe(OH)2

MgNH4PO4

MgCO3

Mg(OH)2

Mn(OH)2

SrCO3

SrSO4

Zn(OH)2

20

16

25

25

26

10

18

18

18

25

12

18

18

25

17.4

20

1.9×10- 33

7×10-9

1.08×10- 10

8.7×10-9

3.95×10- 11

6.1×10-5

3.5×10- 45

1.1×10- 36

1.64×10- 14

2.5×10- 13

2.6×10-5

1.2×10- 11

4×10- 14

1.6×10-9

2.81×10-7

1.8×10- 14

Table 1 : Solubility Products of Sparingly Soluble Inorganic Compounds

2. Feed Water Analysis



- 25 -

In an RO system the most common sparingly soluble salts encountered are CaSO

4, CaCO3 and silica. Other salts creating a potential scaling problem are CaF2, BaSO4

and SrSO4, though less prevalent. Other ions causing problems are described below.

Sulfates are present in a relatively large concentrations in most raw waters. Their

concentration can be artificially increased when sulfuric acid is added to water to adjust

pH. In this case, Ba++ and Sr++ must be analyzed accurately at 1㎍/ℓ(ppb) and 1㎎/ℓ

(ppm) level of detection, respectively, since BaSO4 and SrSO4 are much less soluble

in water than CaSO4 and moreover, barium and strontium sulfate scales are extremely

difficult to redissolve.

Alkalin ity consists of negative ions which include bicarbonate,carbonate and

hydroxide. Most of the alkalinity in naturally occurring water sources is in the form of

bicarbonate alkalinity(HCO 3-). Below a pH of 8.3, the bicarbonate alkalinity will be in

equilibrium with a certain concentration of dissolved carbon dioxide. At a pH greater

than 8.3, HCO 3- will be converted to the carbonate form(CO3

2-). With water sources of

pH above 11.3, hydroxide(OH-) will be present.

Water can dissolve carbon dioxide from the air, forming carbonic acid(H2CO3). The

acidic water will tend to dissolve calcium carbonate from the ground as it passes over

or through the calcium carbonate rock. Most naturally occurring water sources are close

to saturation in calcium carbonate which is in equilibrium with calcium bicarbonate,

depending on the pH of the water. Calcium bicarbonate is much more soluble in water

than calcium carbonate. It the water is concentrated in an RO system, calcium

carbonate salt is likely to precipitate in the system. Thus the use of a scale inhibitor

or lowering the pH below 8 by an acid injection is required in most RO systems.

Nitrates are very soluble in water and thus will not precipitate in an RO system,

Nitrates are a health concern since, when ingested by mammals including humans,

they are converted to nitrites which interfere with hemoglobins to exchange oxygen in

blood. This can cause serious problems especially for fetus and children. For this

reason, it is desirable to maintain a nitrate concentration below 40 ㎎/ℓ in drinking

water. Typical nitrate removal by RO is in the range of 90-96%.

Iron and manganese are present in water either in a divalent state, which is

soluble in water, or in a trivalent state, which forms insoluble hydroxides. The soluble

iron(Fe2+) can come from either a well water or the rust of pump, piping and tanks,

especially if acid is injected upstream of the equipment. If the iron or manganese

concentration is greater than 0.05 ㎎/ℓin an RO feed water and they are oxidized by

air or an oxidizing agent to the trivalent state, then the insoluble hydroxides(Fe(OH)3 and

Mn(OH)3) will precipitate in the system, when the water pH is neutral or higher. They

can also catalyze the oxidative effects of residual oxidizing agents, possibly accelerating

the membrane degradation. Thus iron and manganese must be removed at the




- 26 -

pretreatment step(see the pretreatment section).

Aluminum is usually not noticeably present in naturally occurring water sources.

With its valence of 3+ like iron(Fe3+), aluminum will form very insoluble hydroxide(Al

(OH) 3) at the normal operating pH range of 5.3 to 8.5 in an RO system. Because of

the high charge characteristics, alum(Al 2(SO4)3) or sodium aluminate(NaAlO2) is used to

coagulate negatively charged colloids in the pretreatment of surface waters. Care must

be taken not to employ the coagulant excessively to result in carrying the residual

aluminum over to the membranes.

A concentration of aluminum greater than 0.01mg/l in the dialysis water is a health

concern for kidney dialysis patients. In this regard, iron(FeCl 3 or Fe2(SO4)3) may be

preferred as a coagulant.

Copper and Zinc are not appreciably detected in natural water sources.

Sometimes, it is possible to pick up trace amounts from piping materials. Their

hydroxides (Cu(OH) 2 and Zn(OH) 2) will drop out of solution over the operating pH

range of 5.3 to 8.5. Because of the low concentrations of copper and zinc, their

precipitants will foul an RO system only if allowed to precipitate over an extended

period of time without cleaning the system. However, more serious situation may

develop, when an oxidizing agent such as hydrogen peroxide is present together with

copper or zinc, to degrade the membranes rapidly.

Sulfides are present as a dissolved gas, hydrogen sulfide(H 2S). Hydrogen sulfide

gas can be removed by running the water through a degasifier or oxidizing them by

chlorine or air to insoluble elemental sulfur followed by media filtration.

Phosphates have a strongly negative charge(3-) and a tendency to react with

multivalent cations such as Ca2+, Mg2+ ,Fe2+ , Fe3+ to give insoluble salts. Calcium

phosphate has a very limited solubility at neutral pH and an even lower solubility at

higher pH. The use of a scale inhibitor or lowering the pH of the feed water below 7

is a measure to control the phosphate precipitate.

Sil ica is naturally present in most feed waters in the range of 1-100㎎/ℓand exists

mostly in the silicic acid form(Si(OH)4) below a pH of 9. At low pH, Silicic acid can

polymerize to form a colloid(colloidal silica). At high pH above 9, it dissociates into the

silicate anion(SiO32-) and can precipitate as a salt with calcium, magnesium, iron or

aluminum. Silica and silicates are difficult to redissolve. Ammonium bifluoride solutions

are somewhat successful at cleaning silica.However, ammonium bifluoride is

considered a hazadous chemical posing problems for disposal.Silica present in an RO

feed water at a concentration greater than 20㎎/ℓ

may pose a potential for silica scaling.




- 27 -

Colloids(Suspended Solids) Analysis

Sil t density index(SDI), also known as the Fouling Index(FI), is a good guide line

to determine the colloidal fouling potential of RO feed water. The source of colloids in

RO feed waters is varied and often includes bacteria, clay, colloidal silica and iron

corrosion products. Pretreatment chemicals used in a clarifier such as alum, ferric

chloride or cationic polyelectrolytes can also cause colloidal fouling if not removed in

the clarifier or through proper media filtration.

SDI measurement should be done prior to designing an RO pretreatment system and

on a regular basis during RO operation(at least once a day for surface waters). The

test measures the rate of fouling of a 0.45㎛ filter membrane by the following

procedure.

Place the membrane filter(47㎜, 0.45㎛) on its support, bleed water pressure on

carefully, tighten the O-ring seal and fix the support vertically.

Adjust feed pressure to 2.1bar(30psi) and measure initial time t0, necessary to filter 5

00㎖ of sample water(feed pressure to be kept constant by continuous adjustment).

Continue filtering the water for 15 min at 2.1bar(30psi). If the filter is plugged up in 1

5min, use the 10 or 5min measurement. After 15 minutes, measure again time, t1,

necessary to filter 500㎖.

SDI is calculated as follows ;

SDI = 100 ×(1-t0/t1)

T

T is the time before the second flow measurement(5,10 or 15 minutes).

The guideline is to maintain SDI at less than or equal to 5 for RO feed water.

Turbidity is another guideline as an indicator for the rate of RO membrane fouling.

Turbidometers(also called nephelometers) measure the scattering of light caused by

various suspended solids in the water sample. Water samples having turbidity reading

greater than 1 will tend to foul the membranes. These readings are typically given in

nephelometric turbidity units(NTU). Like the SDI test, turbidity is only an indicator of

fouling potential. High turbidity does not necessarily mean that the suspended solid is




- 28 -

going to deposit on the RO membrane.

In fact, there are some membrane foulants such as surfactants and soluble polymers

that do not scatter light to register a turbidity reading. Although turbidity and SDI are

not perfect in predicting the behavior of the colloidal foulants, they are useful for

characterizing an RO feed water. For examples, SDI greater than 5 and NTU greater

than I strongly suggest that some coagulants should be used in the clarifier step of

the feed water pretreatment, followed by media filtration. The feed water with SDI less

than 5 and NTU less than I may just require media filteration or cartridge filters without

the necessity of coagulation of the colloids.

It is also necessary to have some guidelines for controlling the amount of the

coagulant added to the clarifier, since excessive addition of the coagulant should be

avoided. Two possible guidelines are zeta potential measurement and streaming

current detector.

Zeta Potential is a measurement of the overall charge characteristics of the

suspended solids in the water, when the water containing the charged colloids flows in

one direction between the oppositely charged electrodes. Most colloids in natural water

sources have negative charges which helps to repel each other to keep them

suspended in solution. In overall, the colloidal water will show negative zeta potential.

These negative charges can be neutralized by the addition of cationic coagulants such

as aluminum sulfate(alum) and ferric chlorrde. The coagulants are portionwise added

until zeta potential read zero(neutral). The colloids without charges do not repel each

other and are more likely to coagulate into larger particulate groups which can be

easily filtered out by media filters.

Streaming Current Detector uses a mechanical plunger to create a high water

velocity, which causes movement of the ions surrounding negatively charged

suspended colloids. It measures the eletrical current generated by the moving ions. If

the charge of the colloids has been neutralized by the addition of a coagulant, there will

be very little current generated by the streaming current detector.

The last constituents in the feed water to be analyzed are bacteria count and organic

compounds. There are two ways to count bacteria in the water. One method is to

collect bacteria by filtering a measured quantity of water through a membrane filter

followed by culturing the retained organisms and counting the developed colonies under

low power magnification. The second method is to count

the retained microorganisms on the filter plate directly under a fluorescent microscope

after staining the microorganisms with dyes such as acridine orange. Direct count

methods should be preferred, because they are much faster and more accurate than

culture techniques.




- 29 -

Examples of organic compounds in the feed water are oils, surfactants, water soluble

polymers, and humic acid. The organic compounds are collectively analyzed by Total

Organic Carbon(TOC), Biological Oxygen Demand(BOD) and Chemical Oxygen

Demand(COD). Identification of individual organic substance may require more

elaborate analytical tools such as chromatography(HPLC) and GC-Mass spectrometry.

Removal of the organic compounds from the feed water at the pretreatment step

should be considered when TOC exceeds 3 ㎎/ℓ.




- 30 -

4-3. Prevention of Scale Formation

Scaling of an RO membrane may occur when sparingly soluble salts are

concentrated in the RO element beyond their solubility limit. Sparingly soluble salts are

listed below in the order of decreasing scaling problem :

CaCO3 > CaSO4 > Silica > SrCO3 > BaSO4 > SrSO4 > CaF2 > CaSiO3 > MgSiO3 >

MgSiO3 > Ca3(PO4)2 > Fe(OH)2

Calcium sulfate(CaSO4) is more soluble than BaSO4 and SrSO4. However, calcium

ion (Ca2+) is present in natural water sources more abundantly than Ba2+ and Sr2+ and

thus CaSO4 will cause more scaling problem than BaSO4 and SrSO4. On the other

hand, BaSO4 and SrSO4 are difficult to redissolve once precipitated. Hence, scaling of

the two salts should be avoided.

The most frequent scaling problems come from calcium carbonate(CaCO3) because

it precipitates fast, once concentrated beyond its solubility limit and also most natural

waters are almost saturated with respect to CaCO3. CaCO3 scaling including SrCO3 and

BaCO3 can be prevented by acid addition, a scale inhibitor, softening of the feed water,

preventive cleaning and low system recovery.

CaSO4 scaling including BaSO4, SrSO4 and CaF2 is preventable by the same

methods as CaCO3 scaling except the acid addition. In fact, using sulfuric acid to lower

pH for the prevention of CaCO3 scaling would increase the probability of the sulfate

scaling.

Acid Addition

The solubility of CaCO3 depends on the pH as shown in the following eq uation.

CaCO3 + H Ca2+ + HCO 3-

The equilibrium can be shifted to the right side to convert CaCO3 to soluble Ca(HCO

3)2 by adding an acid (lowering pH). The acid used should be of food grade quality.

Sulfuric acid is commonly employed, but hydrochloric acid is preferred in the case of

high scaling potential due to CaSO4, SrSO4 and BaSO4.

In order to avoid calcium carbonate scaling, the pH of the concentrate stream in an

RO system should be lower than the pH of saturation(pHs) where the water of the

concentrate stream is in equilibrium with CaCO3. This relationship is expressed by the

Langelier Saturation Index(LSI) for brackish waters and Stiff & Davis Stability Index(S&

DSI) for sea waters.

3. Prevention of Scale Formation



- 31 -

LSI = pH - pHS (TDS < 10,000㎎/ℓ)

pHs = (9.3 + A + B) - (C + D)

where : A = (Log10 [TDS] - 1) / 10

B = -13.12 ×Log10 ( ℃ + 273) + 34.55

C = Log10 [Ca2+ as CaCO3] - 0.4

D = Log10 [alkalinity as CaCO3]

Values in brackets are moles/ ℓ except TDS which is in ㎎/ℓ.

(Reference : ASTM D3739-88 Calculation and Adjustment of

the Langelier Saturation Index for Reverse Osmosis).

S & DSI = pH - pCa - pAlk - K

Where : pCa is the negative log of the calcium concentration in moles/ ℓ.

pAlk is the negative log of the total alkalinity concentration in moles/ ℓ.

K is a constant which is a function of water temperature and ionic

strength.

(Reference : ASTM D4582-86 Calculation and Adjustment of the Stiff and

Davis Stability Index for Reverse Osmosis)

Concentration factor is used as shown below to calculate the concentration of

the constituents in the concentrate stream from that in the feed water.

Concentration factor =1

1-recovery

Where recovery = permeate flowrate / feed flowrate

In reality, concentration polarization should be taken into account to get more

accurate scaling potential as shown below.

Concentration factor =concentation polarization

1-recovery

Concentration factor = concentration polarization / 1 - recovery




- 32 -

The concentration polarization depends on the turbulence of the bulk stream in the

RO element and varies from 1.13 to 1.2, meaning that the concentration of salts at the

membrane surface is 13% to 20% greater than in the bulk stream.

To control calcium carbonate scaling by acid addition alone, the LSI or S & DSI in

the concentrate stream must be negative. A rule of thumb is to lower the feed water

pH to 6.0. If a high quality scale inhibitor is used, the LSI in the concentrate stream

can be as high as 1.8(refer to the inhibitor manufacturer's literature for reference

points). This will reduce or eliminate the acid consumption, and also could decrease the

potential for corrosion due to the acid.

Scale Inhibitor Addition

Scale inhibitors(antiscalants) slow the precipitation process of sparingly soluble salts

by being absorbed on the forming salt crystals to prevent the attraction of the

supersaturated salt to the crystal surfaces. In this situation the crystals never grow to a

size or concentration sufficient to fall out of suspension. Furthermore, many scale

inhibitors have some dispersive qualities which involves surrounding particles of

suspended salt or organic solids with the anionically charged scale inhibitor. Now the

anionically charged particles will repel each other to prevent the agglomeration of the

particles to larger particles that may precipitate.

Scale inhibitors effective in controlling carbonate scaling, sulfate scaling and calcium

fluoride scaling are listed below.

Sodiumhexametaphostphate(SHMP) is most widely used because it offered good

inhibition at a low cost. However, care must be taken in order to avoid hydrolysis of

SHMP in the dosing feed tank(a fresh solution should be made every 3 days).

Hydrolysis would not only decrease the scale inhibition efficiency, but also create a

calcium phosphate scaling risk.

SHMP should be dosed to give a concentration in the concentrate stream of 20㎎/ℓ.

The dosage into the feed stream can be calculated by the equation :

20 × (1 - recovery).

Organophosphonates are an improvement over SHMP in that they are more

resistant to hydrolysis though more expensive. They offer scale inhibition and dispersion

ability similar to SHMP.




- 33 -

Polyacrylic acids(PAA) are good at both scale inhibition and dispersion. The

usual molcular weight of PAA is 2000 to 5000. PAA with higher molecular weight

distribution in the range of 6000 to 25000 showed the best dispersion ability at the

sacrifice of scale inhibition ability. In general, PAA are more effective than SHMP.

However, precipitation reactions may occur with cationic polyelectrolytes or multivalent

cations such as aluminum or iron to foul the membrane.

Blend Inhibitors are a combination of low and high molecular weight of PAA or a

blend of low molecular weight PAA and organophosphonates for excellent dispersive

and inhibitor performance.

There are many manufacturers of scale inhibitors such as BF Goodrich, Arrowhead,

Betz, Grace Dearborn, Calgon, FMC, Degremont, Nalco, King-Lee and Maxwell

Chemicals, etc. Please, consult with them for the chemical identity of the scale

inhibitor brands and their compatibility with RO membranes. They can usually be

considered compatible with PA membranes up to 50 ppm concentration in the

concentrate.

RO permeate should be used when diluting the scale inhibitors, since calcium

present in the feed water may form a precipitate with the scale inhibitors at high

concentrations. Precautions must also be taken so that there is no microbial growth in

the inhibitor dilution tank. Make certain that no significant amounts of cationic polymers

are present when adding an anionic scale inhibitor.

Softening with a Strong Acid Cation Exchange Resin

The scale forming cations such as Ca2+, Ba2+, Sr2+, Mg2+, and Fe2+ can be removed by

a cation exchange resin which could be a good alternative way to prevent scale

formation for small or medium size brackish water RO systems. The resin has to be

regenerated with NaCl at the saturation point. A drawback of this process is its

relatively high sodium chloride consumption, which might cause an environmental

(disposing the brine) or an economic problem. Alternatively, a recent counter current

regeneration technique using Dowex Monosphere resins can minimize the NaCl

consumption to 110% of the stoichiometrical value.

Softening with a Weak Acid Cation Exchange Resin .

The weak acid cation exchange resin can remove only Ca2+, Ba2+ and Sr2+ linked to

bicarbonate and release H+, thus lowering the pH to a minimum value of 4.2 where the

carboxylic acid groups of the resin are no longer dissociated. Therefore, it is only a

partial softening and ideal for waters with a high bicarbonate content.




- 34 -

The advantages of softening with a weak acid cation exchange resin

are : almost stoichiometrical consumption(105%) of acid for regeneration to

minimize the operating costs and the environmental impact ;

reduction of the TDS value of the water by the removal of bicarbonate salts

to result in the lower permeate TDS.

The disadvantages are :

residual hardness due to incomplete softening ;

variable pH of the treated water from 3.5 to 6.5 depending on the

degree of exaustion of the resin.

At pH < 4.2, the passage of mineral acid may increase the permeate TDS. It is

therefore recommended to use more than one resin column in parallel and to

regenerate them at different times in order to level out the pH. Other possibilities to

avoid extremely low pH values are CO2 removal or pH adjustment by NaOH.

Lime Softening

Lime (Ca(OH) 2) reacts with soluble calcium or magnesium bicarbonate to remove

carbonate hardness as shown in the following equations ;

Ca(HCO 3)2 + Ca(OH) 2 → 2CaCO3↓ + 2H2O

Mg(HCO 3)2 + 2Ca(OH)2 → Mg(OH) 2 + 2CaCO3↓ + 2H2O

The non-carbonate calcium hardness can be further reduced by adding sodium

carbonate (soda ash) :

CaCl2 + Na2CO3 → 2NaCl + CaCO3

The lime-soda ash process can also reduce the silica concentration. When sodium

aluminate and ferric chloride are added as coagulants, the lime-soda ash will precipitate

calcium carbonate and a complex consisted of calcium, aluminum and iron silicate.

Silica can be reduced to 1㎎/ℓ by adding a mixture of lime and porous magnesium

oxide. The lime softening can also reduce barium, strontium and organic substances

such as humic acid significantly. The effluent from this process needs media filtration

and pH adjustment prior to the RO elements. Lime softening should be considered for

brackish water plants larger than 200m3/h(1.2 million gpd).




- 35 -

Preventive Cleaning

In some applications such as small systems, preventive membrane cleaning allows

the system to run without dosage of acids or scale inhibitors or softening. Typically,

those systems operate at low recovery in the range of 25% and the membrane

elements are replaced after 1-2 years. The simplest way of cleaning is a forward flush

at low pressure by opening the concentrate valve. Short cleaning intervals are more

effective than long cleaning times, e.g. 30seconds every 30 minutes. Cleaning can also

be performed with cleaning chemicals, which may be done occasionally between the

short cleaning intervals.

Adjustment of System Recovery, pH and Temperature.

The precipitation of dissolved salts can be avoided by keeping their concentration in

the concentrate stream below the solubility limit, which can be achieved by reducing

the system recovery, raising temperature, and increasing or decreasing pH.

To be more quantitative for the above operation, solubility product(limit) for each

sparingly soluble solute should be calculated under conditions in the concentrate

stream as discussed in the following sections. Silica is usually the only reason for

adjusting the above three operating variables as a scale control method, since these

adjustments have high energy consumption due to low system recovery or other

scaling risks such as CaCO3 scaling at high pH. For small systems, a low recovery

combined with a preventive cleaning program might be a convenient way to control

scaling.




- 36 -

4-4. Solubil ity Product Calculations

To determine the scaling potential, the ion product IPc of a sparingly soluble salt in

the concentrate stream should be compared with the solubility product Ksp of the salt

under conditions in the concentrate stream(Ksp is a function of temperature and ionic

strength). If IPc < Ksp, no scale control is necessary.

The concentration of ion species in the concentrate stream is usually not known

unless measured experimentally, but can easily be estimated from the concentration in

the feed stream by multiplication with the concentration factor CF= 1 / ( 1 - r ). Where

r is recovery ratio(expressed as a decimal).

The ionic strength of the feed water is :

If = ½Σ(m i ×Zi2 )

Where m i = molal concentration of ion i (mol/kg)

Zi = ionic charge of ion i

Where the water analysis is not given in molal (or molar) concentrations, the

conversion is as follows :

m i =1000Ci

MW i

Where Ci = concentration of ion i in ㎎/ℓ

MW i = molecular weight of ion i

The ionic strength Ic of the concentrate stream is obtained from :

Ic = If ×1

(1 - r)

With the ionic strength of the concentrate stream, the solubility product Ksp of the

sparingly soluble salt can be obtained.

To make sure that scaling will not occur, the IPc for CaSO4, BaSO4, SrSO4 and CaF2

should be less than 0.8 Ksp of the corresponding salts, respectively. If IPc > 0.8Ksp,

one of the scale preventing methods discussed in the previous section must be used.

If proper scale inhibitors are used, IPc could be greater than Ksp as shown in the

4. Solubility Product Calculations



- 37 -

following equation.

IPc ≤ 2.0 Ksp for CaSO4 if PAA or organophosphonates are employed.

IPc ≤ 1.5 Ksp for CaSO4 if SHMP is used.

IPc ≤ 50 Ksp for BaSO4

IPc ≤ 10 Ksp for SrSO4

IPc ≤ 100 Ksp for CaF2

Barium sulfate is the most insoluble of all alkaline-earth sulfates. In most natural

waters, barium is present at a level close to precipitation in the concentrate stream.

The critical feed concentration of BaSO4 may be as low as 15㎍/ℓ in sea waters, 5㎍/

ℓin brackish waters or even 2㎍/ℓif sulfuric acid is added to brackish waters.

4. Solubility Product Calculations



- 38 -

4-5. Silica Scale Prevention

In addition to BaSO4 scaling, silica scale is also difficult to redissolve. Thus silica

scaling has to be prevented. The presence of Al3+ and Fe3+ complicates the silica

scaling via formation of insoluble aluminum and iron silicates. Therefore, if a silica

scaling potential exists, aluminum and iron must be removed by 1㎛ cartridge filtration

and preventive acid cleanings.

The calculation of the silica scaling potential requires the following data of the feed

solution : SiO2 concentration, temperature, pH and total alkalinity.

The SiO2 concentration in the concentrate stream is calculated from the SiO2

concentration in the feed solution and the recovery of the RO system :

SiO2c = SiO2f ×1

1 - r

Where : SiO2c = silica concentration in concentrate as SiO2 in ㎎/ℓ

SiO2f = silica concentration in feed as SiO2 in ㎎/ℓ

r = recovery of the RO system expressed as a decimal

Calculate the pH of the concentrate stream from the pH of the feed stream using

the following equation.

pH = Log10 ([alkalinity as CaCO3] / [CO2]) + 6.3

Alkalinity of concentrate =Alkalinity of feed

1 - r

5. Silica Scale Prevention



- 39 -

Obtain the solubility of SiO2 as a function of temperature :

Example are shown in the following table :

T(° C) Solubility of SiO2 (mg/ℓ )

5

10

15

20

25

30

35

85

96

106

118

128

138

148

Temperature of the concentrate is assumed equal to temperature of feed solution.

Obtain the pH correction factor for the concentrate pH.

Since the solubility of silica increases below a pH of about 7.0 and above a pH of

about 7.8, the actual solubility of SiO2 in the concentrate stream can be further affected

by the pH of the concentrate stream and thus is obtained by multiplying the solubility

of SiO2 at a specific temperature by the pH correction factor to give the corrected

solubility(SiO2cor).

For examples, pH correction factor are 1.0 at pH 7.8 and 1.5 at pH 8.5,

respectively. See ASTM D4993-89 for more details. Compare the silica concentration in

the concentrate(SiO2c) of the RO system with the pH corrected silica solubility(SiO2cor). If

SiO2c is greater than SiO2cor, silica scaling can occur and adjustment is required.

The easiest way to prevent the silica scaling is to lower recovery. Reiteration of the

calculations can be used to optimize recovery with respect to silica scaling, once a

reverse osmosis system is operating.

Lime plus soda ash softening can be used in the pretreatment system to decease

the SiO2 concentration in the feed stream. The maximum allowable recovery against

silica scaling can be increased significantly by increasing the water temperature using a

heat exchanger.

A dispersant such as high molecular weight polyacryate scale inhibitor is helpful in




- 40 -

silica scale control by slowing agglomeration of scale particulate.




- 41 -

4-6. Collo idal Fouling Control

The removal of suspended and colloidal particles can be done by media filtration,

crossflow microfiltration and cartridge microfiltration only for the raw waters with an SDI

slightly above 5. For raw waters containing high concentrations of colloidal matter

showing SDI well above 5, the coagulation and flocculation process is necessary before

media filtration.

Ferric sulfate or ferric chloride is usually used as a coagulant to destabilize the

negative surface charge of the colloids to result in coagulation and further to entrap

them into the freshly formed ferric hydroxide microflocs. Aluminum coagulants are also

effective, but not recommended because of possible fouling problems with residual

aluminum. Cationic polymers may be used as both coagulants and flocculants.

To further agglomerate the hydroxide microflocs for better filterability, flocculants can

be used in combination with coagulants. Flocculants are soluble high molecular weight

polymers such as polyacrylamides which may contain cationic, anionic, or neutral active

groups.

The hydroxide flocs are allowed to grow and to settle in specifically designed reaction

chambers or clarifiers. The hydroxide sludge is removed, and the supernatant water is

further treated by media filtration.

Care must be taken not to allow coagulants and flocculants or the hydroxide flocs to

escape from media filters to reach the RO membranes. Furthermore, reaction of the

residual coagulants and flocculants with a scale inhibitor added after the media filter

can cause a precipitate to form, which is very hard to be cleaned. Several RO plants

have been heavily fouled by a gel formed from cationic polyelectrolytes and scale

inhibitors.

Media filtration

Media filters use a filtration bed consisting of one or more layers of media granules

which are sand and anthracite. The grain size for fine sand filter is in the range of 0.3

5 to 0.5㎜ or 8×12 mesh to 60 mesh and for anthracite filter 0.6 to 0.9 ㎜.

A multimedia filter is designed to make better use of the bed depth in the removal of

a greater volume of suspended solids. This is achieved by loading larger(irregular

shaped) media granules of lower density such as anthracite over smaller media

granules of higher density such as sand. The larger granules at the top remove the

larger suspended particles, leaving the smaller particles to be filtered by the finer

media, thus to result in more efficient filtration and longer runs between cleaning.

6. Colloidal Fouling Control



- 42 -

The design depth of the filter media is normally about 0.8m(31 inches) minimum with

a 50% freeboard for the media expansion during backwashing. The multimedia filters

are usually filled with 0.5m(20 inches) of sand covered with 0.3m(12 inches) of

anthracite.

The multimedia filters can be operated by either gravity or a pressure. A higher

pressure drop can be applied for higher filter beds or smaller filter grains or higher

filtration velocities. The design filtration flow rates are usually 10-20 m/h, and the

backwash rates are in the range of 40-50 m/h.

For feedwaters with a high fouling potential, flow rates of less than 10m/h and / or

second pass media filtration is preferred.

The available pressure is usually about 5m of head for gravity filter and 2 bar(30 psi)

to 4 bar(60 psi) for pressure filters. Periodically, the filter is backwashed and rinsed to

remove the deposited matter, when the differential pressure increase between the inlet

and outlet of the filter is 0.3 to 0.6 bar(4 to 9 psi) for the pressure filter and about 1.4

m for the gravity filter. Backwash time is normally about 10min.

Frequent shut-downs and start-ups should be avoided, because a velocity shock will

release previously deposited particulate matter.

Oxidation Filtration

Some well waters, usually brackish waters, contain divalent iron (Fe2+), manganese

(Mn2+) and sometimes sulfide in the absence of oxygen. If such a water is exposed to

air or is chlorinated, Fe2+, Mn 2+ and sulfide are oxydized to Fe3+, Mn 3+ and elemental

sulfur, respectively, which form insoluble colloidal hydroxides and elemental sulfur as

shown in the following equations :

4Fe(HCO 3) 2 + O2 + 2H2O → 4Fe(OH)3 + 8CO2

4Mn(HCO 3)2 + O2 + 2H2O → 4Mn(OH) 3 + 8CO2

2H2S + O2 → 2S + 2H2O

Iron fouling occurs more frequently than manganese fouling, since iron is present in

the raw waters more abundantly than manganese and the oxidation of iron occurs at a

much lower pH. Thus iron fouling is still possible even if the SDI is below 5 and the

level of iron in the RO feed water is below 0.1㎎/ℓ.

An RO system can treat the well water containing iron(Fe2+) in a closed system

without an exposure to air or to any oxidizing agent, e.g. chlorine. A low pH is

favorable to retard Fe2+ oxidation. At pH<6 and oxygen<0.5㎎/ℓ, the maximum

permissible Fe2+ concentration is 4㎎/ℓ.




- 43 -

Oxidation and filtration can be done in one step by using a filter media greensand

coated with MnO2 to oxidize Fe2+. Greensand is a green(when dry) mineral glauconite

and can be regenerated with KMnO4 when its oxidizing capability is exhausted. After the

regeneration, the residual KMnO4 has to be thoroughly rinsed out in order to avoid an

oxidation damage of the membranes. This technique is used when Fe2+ is present in

the raw water in less than 2㎎/㎖.

An alternative to the greensand is a product called Birm(a registered trademark of

Clack Corporation), which is a light silica coated with manganese dioxide. It catalyzes

the oxidation of iron and manganese by dissolved oxygen. It does not lose the catalytic

activity and thus does not need to be regenerated using KMnO4. However, Birm requires

a certain concentration of dissolved oxygen in the feedwater at an alkaline pH. The

alkaline condition may increase a possibility of CaCO3 precipitation. Birm is less

expensive than greensand.

Crossflow Microfiltration and Ultrafiltration

Crossflow microfiltration can effectively remove most suspended matter, depending

on the pore size of the membranes. Normally microfiltration MF membranes with a

pore size in the range of 1 to 10㎛ are used, depending on the feed water quality.

When the silica concentration in the concentrate stream exceeds the calculated

solubility, MF membranes with 1㎛ pore size are recommended to minimise the

interaction with other colloids such as iron and aluminum colloids.

Now the fouling problem due to colloid deposit is transferred from the RO membrane

to the MF membrane. In fact, the fouling of MF membranes is more severe and more

often than RO membranes because of a high specific permeate flow. However, it is

easier to clean a fouled MF membrane than an RO membrane, using periodic

backflush cleanings which have been proven very effective for cleaning MF

membranes, not for RO membranes. Chlorine can be added to the wash water in

order to prevent biological fouling. Continuous microfiltration membranes(CMF)

equipped with an automated backflush washing mechanism are available in the market

(US Filter). It is usually economical to employ multimedia filters prior to MF

membranes since the cheaper media filters will take a majority of suspended matter to

reduce the costly cleaning frequency of MF membranes and thus to extend the

membrane life time.

Cheaper disposable cartridge filters may be an alternative to the MF membranes and

replaced before the pressure drop has increased to the permitted limit, but latest after 3

months. Replacing cartridge filters more often than every 1 to 3 months usually

indicates a problem with the pretreatment. If it is desirable to remove soluble organic




- 44 -

or inorganic polymers, then ultrafiltration(UF) membranes could be used.




- 45 -

4-7. Biological Foul ing Prevention

All raw waters contain microorganisms such as bacteria, algae, fungi, viruses and

higher organisms. Microorganisms can be regarded as colloidal matter and removed by

the pretreatment. However, it is very difficult to remove all the microorganisms and a

few of them may escape the pretreatment to reproduce and form a biofilm downstream.

MF membranes could stop most of the microorganisms, but still some will escape

through the pores or defects of the membranes to reach the RO membranes. The

symptoms of the biologically fouled RO system are an increase of the differential

pressure and a membrane flux decline.

The potential for biological fouling is higher with surface water than well water. The

concentration of bacteria in water is directly related to the biological fouling potential.

The Total Bacteria Count (TBC) is a method to determine the total number of

viable microorganisms in a water sample according to ASTM F60 by filtering a

measured quantity of water through a membrane filter. Subsequently, the organisms

retained on the filter surface are cultured on the proper nutrient medium for several

days to develop colonies, which are then observed and counted at low power

magnification.

This culture technique can be applied to monitor the microbial activity from the intake

through the subsequent treatment steps up to the concentrate stream and the permeat

e.

Direct Bacteria Count techniques employ filtration of the water sample and

counting the retained microorganisms on the filter plate directly under a fluorescent

microscope after they are stained with acridine orange or INT stain. INT stain can tell

the difference between living cells and dead cells.

The germicidal efficiency of free residual chlorine is directly related to the

concentration of undissociated HOCl which is 100 times more effective than the

hypochlorite ion OCl -.

The fraction of undissociated HOCl increases with decreasing pH and temparture.

At pH = 7.5(25℃), 50% HOCl

At pH = 6.5(25℃), 90% HOCl

At pH = 7.5( 5℃), 62% HOCl

In high salinity waters, less HOCl is present (30% at pH 7.5, 25℃, 40,000㎎/ℓTDS)

Chlorine can react with ammonia to give various chloramines in a series of stepwise

4-7 Biological Fouling Prevention



- 46 -

reactions :

HOCl +NH 3 → NH 2Cl + H2O

HOCl +NH 2Cl → NHCl 2 + H2O

HOCl +NHCl2 → NCl 3 + H2O

Chloramines also have a germicidal effect, albeit lower than that of chlorine. One

advantage with the chloramines is that they do not oxidize the RO membranes.

However, There will be always some residual unreacted HOCl which can still oxidize

the membranes and thus care must still be taken when chloramines are used as a

disinfectant.

To determine the optimum chlorine dosage, best point of injection, pH and contact

time to prevent biofouling, the ASTM D1291, Standard Practice for Determining Chlorine

Requirement of water should be applied to a representative water sample.

Dechlorination

The residual chlorine has to be dechlorinated before it reaches the RO membrane.

CSM RO membrane has some chlorine tolerence, but eventual degradation may occur

after 200-1000 hours of exposure to 1 ㎎/ℓ of free chlorine, depending on the pH,

temperature and residual transition metals such as iron in the feed water. Under

alkaline pH conditions, chlorine degrades the membrane faster than at neutral or acidic

pH. At an acidic pH, chlorine becomes more effective as a disinfectant.

An activated carbon(AC) bed is very effective in dechlorination of the residual

chlorine in RO feed water according to the following reaction :

C + 2HOCl → CO2 + 2HCl

Sodium Metabisulfite(SMBS) or sodium Bisulfite(SBS) is most commonly used

for removal of free chlorine and as a biostatic, as shown in the following reaction.

Na2S2O5 +H 2O → 2NaHSO 3

NaHSO 3 + HOCl → HCl + NaHSO 4

Stoichiometrically, 1.34㎎ of sodium metabisulfite will react with 1.0㎎ of free chlorine.

In practice 3.0mg of sodium metabisulfite is normally used to remove 1.0㎎ of chlorine

to make sure that all the chlorine is reduced.

The injection point of the SMBS solution is preferably upstream of the cartridge

filters in order to keep the residual chlorine up to the filters to prevent microbial growth

in the filters.

The SMBS solution should be filtered through a separate cartridge and injected




- 47 -

through static mixers for good mixing into the RO feed line. The absence of chlorine

should be monitored using an oxidation - reduction potential(ORP) electrode down

stream of the mixing line.

Disinfection by Ultraviolet(UV) Irradiation

UV at 254㎚ is known to have a germicidal effect and has been used especially for

small-scale plants. No chemicals are added and the equipment needs little attention

other than periodic cleanings or replacement of the mercury lamps. However, UV

treatment is limited to relatively clean waters, because colloids and organic matter

interfere with the penetration of UV into depth of the turbid water.

Biological Activity Control by Sodium Bisulfite

Sodium bisulfite(SBS) concentrations in the range of up to 50㎎/ℓ in the feed

stream of sea water RO plants have proven effective to control biological fouling.

Colloidal fouling has also been reduced by the method. SBS is also helpful in

controlling calcium carbonate scaling by supplying protonium(H+) ions as shown below.

2NaHSO 3 + CaCO3 → Na2SO3 + Ca2+ + HCO 3- + HSO 3

-




- 48 -

4-8. Organic Fouling Prevention

Adsorption of organic substances on the membrane surface causes flux loss.

Sometimes the adsorption is irreversible when the organic substances are hydrophobic

or positively charged polymers with high molecluar weight. Such organic substances

including organics present as an emulsion must be removed in the pretreatment.

Organics occuring in natural waters are usually humic substances in concentrations

between 0.5 and 20㎎/ℓ TOC. Pretreatment should be considered when TOC exceeds

3㎎/ℓ. Humic substances can be removed by a coagulation process with hydroxide

flocs or by ultrafiltration or by adsorption on activated carbon.

Oils (hydrocarbon or silicon-based) and greases contaminating the RO feed water at

levels above 0.1㎎/ℓ should be removed by coagulation or activated carbon. Once the

membranes are fouled by oils and greases, they can be cleaned off with alkaline

cleaning agents if the flux has not declined by more than 15%.

Trihalomethane(THM)

THMs(e.g. chloroform) are produced from a chemical reaction between free chlorine

and trace organics. THMs are considered potentially carcinogenic, so are a concern

when found in drinking water. And also trace THMs can be a major concern in the

semiconductor industry, if present in a feed water.

About 90% of THMs can be rejected by RO membranes. Activated carbon can also

adsorb THMs well. A complete removal of THMs is possible by using both activated

carbon and RO membranes.

4-8 Organic Fouling Prevention



- 49 -

4-9. Jelcleer : Positively Charged Media Filter

It has been always desirable not to use coagulants and flocculants in the

pretreatment, since it is difficult to use an exact amount of coagulants and thus there

is always a possibility that either uncoagulated colloids or excess (residual) coagulants

could escape the media filter to foul the RO membranes. Recently a new type of

media filter has been introduced commercially. The new media filter called Jelcleer(Bet

z) has positive charges on the media surface and thus does not need a coagulant.

(Saehan is the exclusive distributor of Jelcleer in Korea).

Preparation of Jelcleer

Jelcleer is ground glass beads coated first with polyacrylate esters, which are bound

chemically and physically to the glass bead surface. The second coating of the beads

with polyamines containing some quaternary ammonium groups is carried out in situ in

the media filter housing. A sketch of the structure of the Jelcleer with double coatings

is shown in the figure below.

<JELCLEER MEDIUM>

GLASS

PAA

AMINE

The spherical beads with a diameter in the range of 0.26 to 0.33㎜ is opaque with

blue tint and has a polydispersity(size distribution) of 1.7. The beads have a density of 9

7lb/ft3 with specific gravity of 2.3 to 2.6.

Theory of Jelcleer Filtration

As shown in the cartoon below, negatively charged colloids are electrostatically

attracted to the positively charged surface of the beads. When the population of the

colloids attached to the bead surface grows, the hydrophobic interation between

adjacent colloids near or on the bead surface increases to agglomerate the colloids.

When the size of the agglomerated colloids becomes large enough to experience the

shear force of the downward flow greater than the electrostatic attraction force between

the agglomerated colloids and the beads, then they drop off the bead surfaces and

flow downward to accumulate in the bottom of the Jelcleer filter.

4-9 Jelcleer : Positively Charged Media Filter



- 50 -

Jelcleer

Filtration

Mechanism

+

+

+

+

--

--

--

--

-

--

-

-

-

----

--

-

-

-

--

- -

-

-

-

-

-

-

- - - - - --

-

-

-

-

-

-

-

-

--

-

--

--

----

-

--

-

-

-

--

-----

- - --

-

-

-

--

COLLOIDS

FLOCCULATEDCOLLOID

JELC LEER

Now the surfaces of the beads in the upper part of the filter are available again to

attract the colloids and the same process repeats until the agglomerated(coagulated)

colloids fill up to the top of the filter. Then simple backwashing will remove all the

colloids out of the filter housing and the Jelcleer media is ready again to do the same

job repeatedly.

Operating Limits

pH range : 4 - 9

Temperature : 2 - 50℃

Minimum bed height : 30 inches

Maximum chlorine tolerance : 0.1ppm

Backwash flow rate : 12gpm / ft2 at 25℃

Normal operating flow : 2 - 7 gpm / ft2

Backwash period : regularly e.g. weekly depending on the raw water quality,

or whenever differential pressure increases to the upper limit.




- 51 -

Advantages of Jelcleer filtration over Multimedia Fil tration.

Advantages of Jelcleer filters over multimedia filters

are : No need of coagulant addition in the pretreatment ;

Ability to remove various size of colloids in the range of 0.05 to 15㎛ more

efficiently than multimedia filters which do not remove well

colloids of 1 to 2㎛ size ;

Production of low turbidity water (≤0.2 NTU) regardless of feed

water quality while the water quality(NTU) produced by

multimedia filters depend on the feed water quality ;

The water quality produced by Jelcleer is close to that by

microfiltration which usually gives water with very low

turbidity(<0.2NTU) regardless of the feed water quality.

Jelcleer can be accommodated by the multimedia housing and

thus no special housing construction is required;

Backwashing frequency is longer than the multimedia filters

Applications of JelCleer

Jelcleer is very efficient in absorbing colloids from clay and silica. The raw water

with the turbidity as high as 10 NTU and the color less than 10 PU( Platinum Unit )

can be transformed by Jelcleer to water of 0.2NTU with SDI in the range of 1 to 5.

If the feed water is precleaned by coagulation, Jelcleer can produce high quality feed

water for the RO membranes, which can reduce the cleaning frequency of the RO

membranes to cut the operating cost and prolong the membrane life time.

However, organic substances in the raw water may foul the surface of Jelcleer. The

fouled Jelceer should be cleaned and coated with the polyamines again. If the

frequency of cleaning and recoating process is too short to be economical, then the

organic substances must be removed by other methods prior to the Jelcleer treatment.

And also the raw waters contaminated heavily with algae and bacteria can not be

converted by Jelcleer alone to water with 0.2NTU and SDI < 5.



5. S y s t e m D e s ig n

5-1 Introd uc tio n

5-2 S ystem D es i gn G u id e lin e s

5-3 B atch vs . C ontin u ou s Proce ss

5-4 S in g le - P ress u re Vess e l

S ystem

5-5 S in g le - Array S ystem

5-6 M u lti - Array S ystem

5-7 D ou b le Pass System

5-8 N um b er of E lem e n ts and

P res su re V ess e l s for S ystem

5-9 Imp o rtan t P aram e ters for

S ystem D es ig n

5-10 Testin g o f S ystem D es ig n s for

U nu su a l App l ica tion s

5-11 S ystem C om p o n en ts

T ec h n o lo g y E xp re s s C S M

5System De sig n

- 53 -

5 -1 . In trod u ct io n

A fte r the pr etre a tm e n t of the ra w w a ter , the tre ate d fe ed w ate r e n ters an RO w a ter

de s al ina ti on s ys te m . T he g o a l of an e ff i cie nt RO sys te m for a c e r tai n re q ui re d

pe rm ea te flo w is to m i n i m ize fe e d pre s su re a nd m e m br an e c os ts (n um b e r of e l em e nts )

w hi le s al t re j e c ti o n an d re c ov e ry sh o ul d be m a x im ize d . T h e op ti m u m d es ign is

in fl ue n ce d by the re l a ti ve i m p o r tan c e of the s e as p ec ts (e .g. rec o ve ry vs. m e m b ran e

co s ts ) re l a te d to o p era ti ng p a ra m e te rs. T h e d es ired s a lt re je c tio n is u su a l ly a ch i e va b le

but the re co v ery d e te rm in e d by a p e rm e a te fl o w is a ff e c te d by m a n y fa c tors .

An RO sys te m d es ign s ta r ts with pr io r it i zin g the re l a ti o n sh i p b etw e e n the d e s ire d

pe rm ea te flo w an d o p era ti ng pa ra m e te rs fo l low e d by op ti m izing the p ar am e ters w i th i n

ph ys ica l l i m its le v ie d by bo th RO m em b ra ne e l e m e n ts an d the fee d w a te r p os se s sing a

po ten ti al of sc a lin g an d fo ul in g. F o r ex a m p l e s , the re c ov e ry of bra c kish w a te r s ys tem s

is l i m ite d by the s ol u bi lity of s pa r ing l y so l u bl e s a lts a n d co l lo i d al fo u l in g p ote n ti a l of the

feed w a ter up to 8 8 % (a ch i e va b l e on l y by m u lt i -a r ra y s ys te m ). Th e re c ov e ry by a

s i n gl e b ra ck ish w a ter e le m e nt is l i m ite d to 15 % by the s c al ing an d fou l ing p o ten ti al of

the fe ed w ate r w ith S D I of 3 to 5.

On the oth er h an d , in se a w a te r d e sa l in a tio n , the li m it of 30 to 4 5 % re co v ery

(ac h i ev a bl e o n ly by m u l t i pl e e l em e n ts in s e r ie s ) is m ai n l y im p o se d by the o s m o tic

pre s su re of the c o n ce n tr ate s trea m , b e ca u se a typ ica l se a w a te r e l em e n t is du ra b le

on l y up to 69 bar (1 00 0 ps i). T he r ec o ve ry by a s i n gl e s e a w a te r el e m en t is l i m ited to 1

0% of th e s ea w a ter fee d w i th S D I le s s th a n 5.

W h i le in s e a w a ter s ys te m s the p e rm e a te fl u x is re l ati ve l y l ow e ve n at m ax im u m

al low ed p re ssu re, the p e rm e a te fl u x c ou l d be very h ig h in b ra ck ish w a te r sys te m s

w i th o ut re ac h in g the l im it of 41 b a r (5 9 5 p s i) for b ra ck ish w a te r el e m en ts . A l th o ug h it is

te m p ti n g to i nc re as e the pe rm ea te fl ux in o rd er to m i ni m ize the c o s ts for m e m b ra n e

el e m en ts , the fl u x ha s to be l im ite d in o rd er to av o i d fo u l in g a nd s c al in g. F r om

ex p er ie nc e , the fl u x l i m it in s ys te m d e s ig n de p en d s on the fo ul ing te n de n cy of the fee d

w ate r .

1. In tro d u c tion

G lo b a l In it iat iv e 2 1

5 System D esign

- 54 -

5 -2 . S y s te m D e s ig n G ui d e lin e s

As m e n ti o n ed a b ov e , the m a i n fac to rs in fl ue n c in g s ea w ate r s ys te m de s ig n s are the

os m o ti c pre s su re a nd th e p hys ica l d u ra bi lity of se a w a te r e l em e nts , w h i le th o s e

aff e c ti n g b rac k ish w ate r sys te m d es ign s a re s ca l in g a nd fo u l in g p ote nti a l of the feed

w ate r . H en c e, se a w a ter s ys tem s co u l d be bu i lt rel a ti ve ly e a s ily by de s ig n in g the

pe rm ea te flo w ra te w i th i n the li m it of the two fa c to rs w i th a m i ni m al atte nti o n to sc a l in g

an d fou l ing p o ten ti al of s e aw ate r . T h e u su a l s ea w ate r re c ov e ry of 30 to 4 0 % c an be

ob tai n e d from s i n gl e -a r ra y s ys te m s.

C om p a rati ve l y, bra c kish w a te r s ys tem s re q ui re m or e el a b ora te d es ign s m ai n l y du e

to v a r io u s s ca l in g a nd fo u l in g p ote nti a l s of th e fe ed w ate rs an d a dd i t i on a l ly d ue to m u l ti -

ar ra y s ys tem s w h en m o re th an 5 0 % re co v e ry is d e s ired . A sys te m d e s ig ne d w i th h ig h

pe rm ea te flu x ra te s is li kel y to e xp e r ie n ce h i g he r fo u l in g ra te s a n d m o re fre q ue n t

ch e m ica l c l ea n i ng . T h e s ilt d e ns ity i n de x (S D I) v al u e of the p re trea te d fee d w a ter

co r re l ate s well with the a m ou n t of fou l ing m a ter ia l p res e nt.

E x p er ie nc e fro m the co r re l ati o n of the S D I va l u e with the m e m bra n e fou l in g tre n d

ca n set the l i m its on pe rm ea te fl ux a n d el e m en t re c o ve ry for d iff e re nt typ e s of w a te rs ,

which is the s tar ti n g p oi n t of the sys te m d e s ign gu i de l in e s s ho w n in T a b le 1 be l ow .

2. S y ste m D e s ign Gu ide line s


5System De sig n

- 55 -

T a b le 1. S ys te m D e s ig n G ui d e lin es

Fe ed S o u rc e

W e ll

W a te r /

S o ft e ned

W a te r

S o ft e ned

S u rfa c e

W a te r

S ur fa ce

W a te rSe a W a te r

F e ed S D I < 3 3 - 5 3 - 5 < 5

M a x, % R e co ve ry p e r

E lem e nt19 17 15 10

M a x , P e r m e ate

Flo w R a te p e r

E lem e nt,

g p d(m 3/d )

2 .5"

D iam e te r7 1 0 (2 .7) 5 0 0(1 .9) 5 0 0 (1 .9) 5 0 0 (1 .9)

4 "

D iam e te r2 1 0 0(8 .0 ) 1 8 70 (7 .1) 1 7 4 0(6 .6 ) 1 5 00(5 .6 )

8 "

D iam e te r7 4 00(2 8) 6 6 0 0(2 5) 5 8 00(2 2) 5 8 00 (2 2 )

M a x. F e ed

Flo w R a te p e r

E lem e nt,

g p m (m 3 /d )

2 .5"

D iam e te r5 .7(1 .3) 5 .7 (1 .3) 5 .7(1 .3) 5 .7(1 .3 )

4 "

D iam e te r1 8 (4 .1) 1 8 (4 .1 ) 1 8 (4 .1) 1 8 (4 .1)

8 "

D iam e te r6 2 (1 4.1) 6 0 (1 3.7) 5 5 (1 2.6) 6 0 (1 3.7)

M in .

C on c entra te

Flo w R a te p e r

E lem e nt,

g pm (m 3 /h )

2 .5"

D iam e te r1(0.22) 1(0.22) 1(0.22) 1(0.22)

4 "

D iam e te r4(0.91) 4(0.91) 4(0.91) 4(0.91)

8"

D iam e te r16 (3 .6) 1 6 (3 .6 ) 16 (3 .6) 1 6( 3.6 )

T h e g ui d e lin es in T a b l e 1 are ba s ed on a c o nti n uo u s pr oc e ss w i th a well d es ig n ed

an d o pe ra ted p re tre atm e nt s ys te m . E xcee d i ng th e l i m its will re s ul t in m o re fre qu e n t

c l e an i n gs th a n ab o ut four t i m es a yea r , a n d a re du c e d m e m b ran e l i fe.

2. Fe e d W a te r A n a lys is


5 System D esign

- 56 -

5 -3 . B a tc h v s . C on tin u o u s P roc e s s

T h e m a jo r ity of RO s ys te m s a re d e s ign e d for c on ti nu o us o p e rati o n with co n s ta n t

pe rm ea te flo w an d c o ns tan t s ys te m re co v er y as s ho w n in F i g u re 1. V ar ia ti o n s in fee d

w ate r te m pe ra tur e an d fo ul ing e ff e c ts a re c o m p e ns a ted by a dj u s ti n g the feed pr es su re .

▶

F e ed P e rm e a te

C on c e ntra te

F i g ure 1 : C o nti n uo u s RO P ro ce s s

In c e r tai n a p p l ica ti on s , w h e n rel a ti ve l y s m al l v o lu m e s (b atc he s ) of s pe c ia l fe ed

w ate rs o ccur d isco n ti n u ou s ly , e.g . w a s tew ate r or in d us tr ial pro c es s s ol u tio n s , the b a tc h

op e rati o n m od e is p refe r re d . T he feed w a ter is co l le c ted in a ta n k an d tre ate d

su b se q ue n tl y . T h e pe rm ea te is re m o v e d an d the co n ce n trate is re c yc le d b ac k to the

ta n k . At the e n d of the b a tch p ro c es s, a sm a l l vo l um e of c o nc e ntra te re m a i n s in the

feed tan k . After thi s h as b e en d ra i ne d , the m e m b ran e s are typ i ca l ly c l ea n ed b e fore th e

ta n k is filled a g ai n w i th a n e w b a tch . F i g ure 2 sh o w s the ba tch o p e rati o n m od e .

▶

F e e d/

C on c en tra te P e rm e a te

F i g ure 2 : B atc h RO P ro c es s

A m o d i f icati o n of the b a tch m o d e is the s em i-ba tc h m o d e. T he fe e d tan k is re filled

with feed w a ter al re ad y d ur in g op e ra ti o n . T h e ba tch is te rm in a te d w i th the fee d ta nk full

of c o nc e n tra te . T h is a l lo w s a s m a lle r ta nk to be u se d .

B a tch sys te m s a re u s ua l ly d e s ig ne d w i th a c on s tan t fe ed p re s su re a nd a d e c lini n g

pe rm ea te flo w w h i le the fee d b ec o m es m o re c on c en tra ted . T h e g ui d el in es g i ve n in

3. B a tch vs. C o n tinu o u s P ro c e s s


5System De sig n

- 57 -

Ta b l e 1 s ho u ld be a p p lie d to ba tc h sys te m s as w e l l. T he pe rm ea te flo w l im its h o w e ve r ,

are c o ns e rv ati ve a n d m ay be e xcee d ed , if an a p p r op r iate c l ea n i ng fre q ue n cy is ta ke n

in to a ccou n t.

S o m e a dv a nta ge s of the ba tc h pro c e ss o ve r the c o nti n uo u s pro c e ss

a re : S ys te m re c ov e ry ca n be m ax im ize d ba tc h by b a tc h ;

C l ea n i ng is e a s ily i m p le m e nte d .

T h e d i sa dv a nta g es

a re : No co n s ta n t p erm e ate q u al ity ;

La rg e r p um p re q ui re d ;

H i gh e r to tal run n i ng c o s ts .

3. B a tch vs. C o n tinu o us P ro c e ss


5 System D esign

- 58 -

5 -4 . S in g le -P re s s ure V e s s e l S y s te m

A S in g l e-P r es su re V e s se l S ys tem co n s ists of a p re ssur e ve s se l w i th up to se v en

m e m br an e e l em e nts , w h i ch a re c o nn e c te d in s er ie s . Th e c o nc e ntra te of the first

el e m en t b ec o m es th e fe ed to the se c o nd , a nd so o n . T h e pro d uc t tu b es of all

el e m en ts are co u pl e d a nd c o nn e c te d to the ve s se l p erm e ate p o r t. T he pe rm ea te p or t

m a y be lo c ate d on the fe e d an d or on the co n ce n trate e n d of the v e ssel .

S i n gl e -P re s su re V es sel S ys te m s are ch o se n w h e n on l y on e or few m e m br an e

el e m en ts are ne e de d for the sp e c if i ed pe rm ea te flow.

F e e d w a te r en te rs th e s ys tem th ro u gh th e s hu t-o ff v al ve an d flows thro u gh the

ca r tr idg e fi lte r to the hi g h p re ssure p u m p.

F ro m the h ig h p re ssur e pu m p, the fe ed w ate r f lows to the fee d i n le t c o nn e c ti o n of

the v es sel . T h e p ro du c t s tre a m s ho u l d le a v e the v e ssel at no m o re tha n 0.3 ba r ( 5

PS I ) o v er atm o sp h er ic p res sure .

In o the r w o rds , in a n y e v en t, the ba c k pre s su re of the pe rm ea te m us t not be gre a ter

th a n the fe e d pre s su re in o rd er not to d a m a g e the m e m br an e s .

T h e c on c en tra te l ea v es th e c on c en tra te ou tl et co n ne c ti o n at e ssen ti al ly the fe ed

pre s su re .

P re s su re d ro p will u s ua l ly a m ou n t to 0.3 to 2 b a r (5 ~ 30 P S I) from feed i nl e t to

co n ce n tr ate o u tl e t, d ep e nd i n g on th e n um b e r of m em b ra ne e l e m e n ts , the fe ed fl o w

ve l oc ity a n d the te m pe ra ture . T h e c on c en tra te flo w is co n trol led by the co n ce n tr ate fl o w

co n tr ol v a l ue .

T h e s ys tem rec o ve ry is c on tro lle d by th i s v a l ve a nd m u s t n ev e r e xce ed th e d es ign

va l ue .

4. S ing le-P re s su r e V e s s e l S y s te m


5System De sig n

- 59 -

5 -5 . S in g le -A rra y S y s te m

In a s i n gl e -a r ray s ys te m , two or m or e m o d ul e s a re a r ra ng e d in pa ra l le l . F e ed

pro d uc t a n d co n ce n trate a re c o nn e c te d to m an i fol d s . O th e r a sp e c ts of the s ys te m a re

the s am e as in a s i ng l e -pre s su re v es sel s ys te m . S i ng l e- ar ra y s ys tem s are typ i ca l ly

us e d w h ere th e s ys tem rec o ve ry is l es s tha n 5 0 % . e.g . in s ea w ate r d e sa l in a tio n .

An e x am p l e of a s i n g l e-a r ra y sys te m is o utl ine d in F i g ure 3 : E ac h of the thre e

pre s su re v es se l s ho u se s s i x C S M R e l e m e n ts R E 8 04 0 -S N .

R E 804 0-S N

R E 804 0-S N

R E 804 0-S N

▶

C o n ce n trate V a l ve →

H i gh P re s su re

P um p

F e ed

C a r tr idg eF i lte r

P erm e ate

C on c en tra te

F i g ur e 3 : S i ng l e -A r ra y S ys te m

5. S ing le-A r ra y S y s te m


5 System Design

- 60 -

5-6. Multi-Array system

Systems with more than one array or stage are used for higher system recoveries

without exceeding the single element recovery limits. Usually two arrays will suffice for

recoveries up to 75%, and three must be used for higher recoveries 87.5%. These

numbers are based on the assumption that standard pressure vessels with six

elements are used.

Generally speaking, the higher the system recovery, the more membrane elements

have to be connected in series. In order to compensate for the permeate that is

removed and to maintain a uniform feed flow to each array, the number of pressure

vessels per array decreases in the direction of feed flow.

A typical two-array system using a staging ratio of 2:1 is shown in Figure 4. The

staging ratio is defined as the ratio of pressure vessels in two adjacent arrays,

upstream vessels: downstream vessels.

R E 804 0-B N

R E 804 0-B N

R E 804 0-B N

ConcentrateValve

▶High Pressure

Pump

Feed

CartridgeFilter

Permeate

Concentrate

R E 804 0-B N

R E 80 40 -BN

R E 8 04 0-B N

Figure 4 : Two-Array System

6. Multi-Array system


5System Design

- 61 -

5-7. Double Pass System

A double pass system is the combination of two RO systems where the permeate of

the first system (first pass) becomes the feed for the second system (second pass).

Both RO system may be of the single-array or multi-array type, either with continuous

flow or with concentrate recirculation. With this concept, a product conductivity of < 1㎲

/㎝ can be achieved without any contamination from organic matter and pyrogens

including bacteria. Thus applications of double pass systems can be found in the

production of pure water for pharmaceutical, medical and semiconductor industries.

Figure 5 shows a schematic flow diagram of a double pass system.

▶

CartridgeFilter

Concentrate

Concentrate Valve

▶

Feed

High PressurePump RO Ⅰ RO Ⅱ

ProductHigh Pressure

Pump

Figure 5 : Double Pass System

The concentrate of RO ll is recycled back to the feed of RO l, because its quality is

usually better than the system feed water.

The concentrate of RO is of high quality (RO permeate). RO II can be designed for

a higher recovery than RO I and with fewer membrane elements.

Instead of having a separate high pressure pump for the second pass, the whole

system can also be operated with one single high pressure pump, provided the

maximum permissible feed pressure of the membrane element is not exceeded ( 41

bar(600psi) for brackish water elements).

The second pass is then operated with the permeate backpressure from RO I. Care

must be exercised that the permeate backpressure at no time exceeds the feed

pressure by more than 0.3 bar(5psi)

7. Double Pass System


5 System Design

- 62 -

5-8. Number of Elements and Pressure Vessels for System Design

As mentioned in the introduction, RO systems are usually designed for a specified

amount of water production per day ( total permeate flow QT). Then a number of

membrane elements (NE) required to produce QT is estimated from dividing QT by

average permeate flow per element (QA)

NE =QT

QA

In most standard applications, the average permeate flow per element(QA) is about

75% of the maximum permeate flow per element(QM) which is shown in Table 1 in the

section of system design guide lines.

QA = 0.75 QM

Thus equation (1) can be converted to equation 3.

NE =Qt

0.75QM

And also the number of pressure vessels, NV is obtained from deviding NE by PE

which is the number of elements per pressure vessel.

Nv =NE

PE

Standard vessels contain six elements. Nv is rounded to the next highest whole

number.

With QT , NE , NV and the analysis of the feed water source, a system can be

selected. The selected system must then be verified using Saehan CSM Pro 2000

computer simulation program. This program calculates the feed pressure and the

permeate quality of the system including the operating data of all individual elements.

Furthermore, the system design can easily be optimized by changing the number and

type of elements and their arrangement (e.g. single-array or multi-array). However the

optimization should fall within the physical limit of the elements and the empirical limit

of the recovery rate according to the guidelines of the fouling potential of the feed

waters.

8. Number of Elements and Pressure Vessels for System Design

(1)

(3)

(4)

(2)


5System Design

- 63 -

For examples, Two-array systems with 6-element vessels effectively employ twelve

spiral wound elements in series and are generally capable of operating at an overall

recovery rate of 60 to 75%. For such systems the average individual recovery rate per

element will vary from 7 to 12%. To operate a two-array system at an overall recovery

much higher than 75% will cause an individual element to exceed the maximum

recovery limits(e.g. 15% for a feed water with SDI in the range of 3 to 5). Then, a third

array will have to be employed to place eighteen elements in series for a recovery rate

lower than 15% per element.

If two-array systems are operated at too low a recovery(e.g. <60%), the feed flow

rates to the first-array vessels can be too high causing excessive feed/concentrate-side

pressure drops and potentially damaging the elements. As a result, systems with lower

than 60% recovery will typically utilize single-array configurations.

8. Number of Elements and Pressure Vessels for System Design


5 System Design

- 64 -

5-9. Important Parameters for System Design

The performance of an RO system is usually defined by its feed pressure, its

permeate flow and its salt passage(salt rejection). Two simple equations show the

relationship among the parameters.

Q = A ×S ×( Pf-Δπ-ΔPf c

)2

Q = permeate flow

A = membrane permeability confficient

S = membrane surface area

Pf = feed pressure

ΔPfc = concentrate side pressure drop

Δπ = osmotic pressure of the feed water

Equation 1 indicates that the permeate flow Q is directly proportional to the surface

area S times a net permeation driving force.

(Pf - Δπ-ΔPf c

)2

On the other hand, the salt passage is by diffusion and thus the salt flux N is

proportional to the difference in salt concentration between both sides of the

membrane as shown in equation 2.

N = B × (Cfc - CP)

B = salt diffusion coefficient

Cf c = feed - concentrate average concentration

CP = permeate concentration

The equations 1 and 2 can be used to calculate the performance of a single RO

element and also of a multi RO element system if average values for

(Pf-Δπ-ΔPf c / 2) and (Cf c-Cp), temperature, and number of elements are known.

When a large number of elements are combined in a system with a complex series-

9. Important Parameters for System Design

(1)

(2)


5System Design

- 65 -

parallel-series configuration and only inlet operating variables are known, system

performance calculation becomes more complex and tedious. Feed pressures and salt

concentration for each element is the feed to the following element. However, it is

possible to do the complex element to element calculation by a computer simulation

program.

The first step is to guess the applied pressure Pf needed to produce the water flow

rate (Q) required. Equation 1 can be converted to equation 3.

Pf =Q

+ Δπ+ΔPf c

A×S 2

To solve equation 3 for the feed pressure, the total area of the membrane from the

total number of elements is summed to give S. The concentrate osmotic pressure(Δ

π) is calculated from the average concentration of the feed and the concentrate which

is obtained from the nominal recovery and salt rejection of the element. The

differential pressure ΔPf c may be estimated from the average of ΔPfc calculated from

flow to the first element in series and the last element in the array.

The membrane permeability coefficient (A) is provided by the manufacturer or

obtained from the slope of the curve of the permeate flow (Q) versus the operating

pressure (Pf - Δπ).

Now using the guessed applied pressure(Pf), the computer calculates feed flow rates,

permeate flow rates and salt concentrations on an element by element basis through

the system. The total calculated permeate flow (Q) is lower than the target flow, then

the pressure is raised proportionally and if Q is higher than the target flow, then the

pressure is lowered proportionally.

The analysis of the concentrate exiting the last array is then checked for solubility

and pH limits for various scaling salts. If the solubility limits are exceeded, Q or the

pressure may be decreased, but not so low to drop below the minimum concentrate

flow rate. If it does, then more staging (number of arrays) is required.

Another limiting factor is that in any event, the maximum feed flow rate and permeate

flow rate of the first element in the first array must not be exceeded. If they do, then a

physical damage to the element and fouling would occur. Controlling the feed pressure

or adding more elements may be necessary to avoid such problems.

To be accurate and closer to the real situation, equation 1 could have several

additional variables as shown in equation 4.


(3)


5 System Design

- 66 -

Q = A(π)×S×TCF×FF×[ Pf-Δπ-ΔPf c

-Pp]2

π = πf×Cfc

× pfCf

π = average feed - concentrate osmotic pressure

πf = osmotic pressure of feed

C fc= concentration of the concentrate

C f = concentration of the feed

pf = concentration polarization

Pf = Exp [0.7×Y]

Y = recovery of the element

TCF = Exp [ 2900×{1/298 - 1/(273+T)}]

TCF = temperature correction factor

FF = Fouling Factor ≒ 0.8

Δπ = π - πp

πp = permeate side osmotic pressure

Pp = permeate pressure

A varies from 0.5 to 1.25 depending on the average osmotic pressure(π) and

specific to the element itself. For brackish water with TDS up to 2500 ppm. A is about

1.3.

Cp = B×Cf c × pf × TCF ×S

Q

Cp = permeate concentration

Saehan's CSM PRO 2000 computer program utilizes equations 4 and 9. Using

iterative trial and error solution, it projects the performance of given systems, and

optimizes the design of the system.

(4)

(5)

(6)

(7)

(8)

(9)



5System Design

- 67 -

5-10. Testing of System Designs for Unusual Applications

For the desalination of well characterized waters with known SDI and composition,

the RO performance can be projected with reliable accuracy by the computer program.

However, testing is recommended to support the proper system design in special

situations such as :

Unknown feed water quality ;

Waste waters ;

Special permeate quality requirements ;

Very high system recoveries ;

Very large plants ;

First , a screening test is to select the right membrane and to obtain an idea about

flux and rejection properties of this membrane in the special situations. Usually a

small piece of flat sheet membrane is used for the screening test.

Second, using a 2540 size element, scale up data such as permeate flow and salt

rejection as a function of feed pressure and system recovery are obtained. In

subsequent batch mode test, leading the permeate into a separate containment and

returning the concentrate to the feed tank, permeate flow and salt rejection are

monitored until the permeate flow has declined to a very low value such as 0.09

gpm. From the batch tests, an indication of membrane stability and fouling effects can

be revealed.

Third, a pilot test is run in the field in a continuous operation mode. The pilot plant

has at least one 8040 size element, preferably an arrangement of elements similar to

the arrangement in a large scale system. The permeate flow of the pilot plant should

be at least 1% of the large scale plant flow. The test should be run for more than 30

days. The objective is to confirm the system design and to optimize operating

parameters as well as to minimise the risk involved in large scale plants.

10. Testing of System Designs for Unusual Applications


5 System Design

- 68 -

5-11. System Components

High Pressure Pump

The high pressure pump must provide smooth and continuous flow to the RO

membrane elements and also be sized to provide the necessary flowrate at the desired

pressure. Its energy consumption is one of the major expenses of RO system

operation. There are two types of high pressure pumps, centrifugal pumps and piston

pumps. Centrifugal pumps, which operate by spinning the fluid with the pump impellers

are more energy-efficient than piston pumps, which use various numbers of mechanical

plungers to create pressure. The centrifugal pumps provide more smooth and

continuous flow than the piston pump, and is controlled by a throttling valve on the

discharge line. The piston pump cannot be throttled, so pressure is controlled by a

back pressure valve installed in a by pass line from the pump discharge to the pump

suction. the pressure from piston pumps tends to pulsate, creating surges, which could

damage RO membrane elements and therefore should be controlled using a damper

(accumulator).

Pressure Vessel

Pressure vessel (membrane element housing) are designed for specific pressure

applications. Most pressure vessels are overdesigned for safety reasons to withstand a

pressure at 1.5 times the rated operating pressure. The vessel materials are usually

FRP (fiber glass reinforced plastic) and sometimes stainless steel (316L) for special

applications such as very high pressure applications(>800psi).

Pressure vessels are available with different diameters, lengths, and pressure ratings.

The smaller vessels with diameters in the range of 1.5 to 2.5 inch are usually supplied

by Payne. All other vessels with diameter in the range of 2.5 to 8 inch and the

pressure rating up to 1000psi are manufactured by Advanced Structures Inc.

Alarms and Shutdowns

There is always the possibility of a malfunction in the RO system or with the

pretreatment. Instruments can be used to monitor the quality of the RO feed water as

well as the performance of the RO system. The instruments such as flow meters,

pressure gauges, pH meters and conductivity meters can be connected with an audible

alarm that will sound if some parameter is not within design specifications. Furthermor

e, if the parameter is significantly out of the specifications to damage the high pressure

pumps and the RO elements or to produce an unacceptable permeate water, then the

11. System Components


5System Design

- 69 -

alarm should warrant automatic shutdown of the RO system. Suggested alarms and

shutdowns are shown below :

Low inlet pressure damaging the high pressure pump;

Too high feed pressure damaging the RO elements;

High feed temperature damaging the RO elements;

High permeate pressure damaging the RO elements;

Too high concentration of colloidal matter or sparingly soluble

salts in the feed damaging the RO elements;

Oxidizing agents in the feed damaging the RO elements;

Low concentrate flow fouling the RO membranes;

Oil in the feed fouling the RO membranes.

Instruments, Valves, and Equipment

As mentioned above, instruments and valves are necessary not only to detect a

malfunction in the RO system, but also to ensure proper routine operation of the

system. The necessary instruments, valves, and other equipment are listed below in

more details.

Pressure gauges to measure the pressure drop across the cartridge filter, the

pressure on the pump inlet line and discharge line, the feed pressure to the membrane

element(s), the pressure drop between feed and concentrate of each array, and

eventually the pressure in the permeate line. Liquid-filled gauges should contain

membrane compatible fluids such as water or glycerine in place of oils or other water

immiscible liquids.

Flow meters to measure feed, concentrate and total permeate flow rate, also

permeate flow rate of each array. pH meter in the feed line after acidification to control

carbonate scaling potential. Conductivity meters in the feed line, in the brine line, and

in the permeate line to determine permeate quality and salt rejection. Sample ports on

the feed, concentrate and permeate line (total permeate and permeate of each array) to

be able to evaluate system performance. A sample port on each pressure vessel

permeate outlet is recommended to facilitate troubleshooting.

Feed inlet valve to shut down the plant for maintenance and preservation. Valve on

the pump discharge line or pump bypass line to control feed pressure during operation

and feed pressure increase rate during start-up. Check valve on pump discharge line.

Check valve and atmospheric drain valve on permeate line to prevent the permeate



5 System Design

- 70 -

pressure from exceeding the feed pressure. Flow control valve on the concentrate

line to set the rec overy. (Caution:back-pressure valve must not be used). Valve in

the permeate line to provide permeate drain during cleaning and start-up. Valves in the

feed and concentrate line (and between arrays) to connect a cleaning circuit.

A small draw-back tank is necessary in the permeate line to provide enough volume

for osmosis backflow when a seawater system shuts down. Without the tank, air could

be sucked into the membrane elements to dry the membrane (flux loss) and to

contaminate the permeate side by airborn bacteria and fungii.

A shut down flush system flushes the feed-concentrate line with pretreated feed

water or with permeate water after shut own, especially when scale inhibitors are used,

and also in the case of a seawater system.

Materials of System Construction

The materials of system construction including the RO elements, pumps, pressure

vessels, pipes, valves and instruments should be compatible with the pressures,

vibrations, and temperatures during the RO system operation. The materials must also

be resistant to the potential corrosion attacks caused by the high chloride content of the

feed water and the concentrate stream, and the chemicals used for membrane cleanin

g.

Non-metalic materials such as plastics and fiberglass are widely used not only for

economic reasons (e.g. pressure vessels and pipes), but also for preventing corrosion

and chemical attacks, usually in the low pressure (<10 bar) applications.

However, it is usually necessary to use metals for the high-pressure (10-70 bar/ 200-

1,000PSI) parts such as pumps, piping and valves. Carbon and low alloy steels do not

have sufficient corrosion resistance, and their corrosion products can foul the

membranes.

Stainless steel type AISI 316 L with <0.03% C is recommended for the pipe system

for RO plants with concentrate stream TDS below 7000 ppm. For TDS higher than 700

0ppm, stainless steel type 904L is preferred for pipes and bends for welding and

stainless steel type 254 SMO should be used for flange connections, valves, and

pumps where crevices occur.



6. S y s t e m O p e r a t i o n

6-1 . In trod u ctio n

6-2 . In itia l S tart-up

1) C hec k l is t befo re Start-up

2) S tart-up S eq u e nc e

6-3 . R ec o rd K ee p i n g for m ain tena n ce

1) P retrea tm en t S ystem

2) R O System

3) M ain ten an c e Log

6-4 . D ata N orm al iza tio n

1) S al t R ejec tion

2) P ress u re D ifferen ce

3) N orm a liz ed Perm e ate Flo w R ate

6-5 . C ontro l o f M icroorgan is m G row th

in R O S ystem


6Sy stem O peratio n

- 72 -


S u c ce ss ful lo n g term p e r for m a nc e of the RO s ys te m d e pe n ds on p ro pe r op e rati o n

an d m ai n ten a n ce of the sys te m .

T h i s in c lu d es w el l p l an n ed pre tre atm e nt an d i n it i a l p la n t s tar t-u p . M e m b ran e fo ul ing

an d s ca l in g , w h i ch c a n red u c e pe rm ea te fl ow ra te an d s al t r ej e c ti o n , are ca u se d

pr im a r ily by a po o r p re trea tm en t a n d also an i m p ro pe r o p e rati o n. M e ch a ni cal a n d

ch e m ica l d a m a g es of th e RO s ys tem i nc lu d i ng th e m em b ra ne s m a y a l so a r ise fro m

im p ro pe r op e rati o ns . R ec o rd ke e p in g a nd da ta n orm a liza ti on is re q u ire d to k no w the

ac tua l p l a nt p e r for m a nc e a n d to lo c ate th e s ou rce of any pro b l em . D ata n o rm a l izati o n

with refe re nc e to the in i t i al s ta r t-u p sys te m p er fo rm an c e is us e ful to s h o w a n y

pe r fo rm a n ce c h an g es d u r in g the fol low i ng o p era ti on s .

C om p l ete re c ord s a re re q ui re d in ca s e of a sys te m p e r form a nc e w a r ra nty c l ai m

in c lu d in g e l em e n ts .



6 System O p erat ion

- 73 -

6 -2 . In i t ia l S ta r t-up

1 ) C h ec k lis t b efo re S tart -u p

B e for e s ta r t i ng up the RO s ys tem , it is i m p or ta n t to m ak e s ure th a t the w h ol e

pre tre atm e nt pro c es s is w o rking ac co rd i ng to th e s pe c if i ca ti on s . If the ch e m ica l

ch a rac te r is ti c s of th e ra w w ate r is c h an g ed , th e n a full a na l ys is of the w a ter e nte r in g

the RO u n it sh o ul d be d on e so th at p ro p er m ea s ure m e nts c a n be m a d e to put the

va r iab l es un d er co n trol .

F a c to rs a ff ec ti ng the fe ed w ate r q u al ity a n d thu s the sys te m d e s ig n are

as fo llo w s : F l o w , S D I, T u rb i di ty , T em p e ratu re , p H , T D S , r es idu a l ch l o r in e ,

an d b ac te r ia c o un ts .

A n d a l so the fo l low i n g m e ch a ni cal i n s pe c tio n s of the RO s ys te m are re co m m e n de d

for the i n i ti a l s tar t-u p .

● O p era ti on a l c on d i ti o n s of m e d i a fi lter s a n d ca r tr idg e fi lte rs.

● F e e d l in e is p ur ge d a nd fl u sh e d b efo re p res sure v e sse ls a re c on n ec te d.

● C he m ical ad d it i o n l i ne s a nd v a l ve s .

● P ro p er m i xing of c h em ica l s in the fee d s trea m .

● S a fety s hu t off of th e RO s ys tem w h en the c he m ical do s ag e p um p s are

s h ut do w n .

● C om p l ete c h lo r ine r em o va l p r ior to the m e m br an e s

● In s tru m e nta ti on for p ro pe r op e ra ti o n a nd m o n ito r ing of the p retre a tm e n t a nd

RO s ys te m .

● In s ta l la t i on an d c al ibr ati o n of su c h i ns tru m e n tati o n

● In s ta l la t i on of p re s su re re l ie f p ro te c tio n .

● P i p i ng a n d s ec u r in g p res sure v e sse ls for o p era ti on an d c le a n in g m o de .

● L u br ica ti on an d p rop e r ro ta tio n of p um p s .

● V a l ve s for p erm e ate l i n e, feed flow, an d re j ec t f l o w c o ntro l a re in o pe n p o s it io n .

● In i t ia l fe ed fl o w is l i m ite d to l es s tha n 5 0% of op e rati n g fee d flow.

2 ) St art -u p S e q u en c e

● B e fo re s ta r t i ng th e i n it i a l o pe ra ti on s e qu e nc e , the pre tre atm e nt se c tio n s ho u l d be

th o rou g hl y r i n se d to fl us h out de b r is a n d oth e r c on ta m in an ts w i th ou t a l lo w i n g the

fe e d to en ter the el e m en ts .

● M a ke s u re tha t all va l ve s e tt in g s are co r re c t. T he fe e d pr es su re c o ntro l a nd

2. Initial S ta rt-u p



- 74 -

c o nc e ntra te c on tro l v al ves s h ou l d be ful ly o p en .

● U se th e fee d w a te r at a l o w fl o w ra te to e xp e l the air o u t of the e l e m e n ts an d

p re s su re ve s se l s at a ga u ge p re s su re of 30 to 60 psi for m o re th an 30 m i n ute s .

All p er m e ate a n d c on c en tra te f lows s h ou l d be di re c te d to an ap p ro ve d w a s te

c o l le c ti o n d rai n d u r in g fl us h i ng . At th i s po i n t, all p i pe c o nn e c ti o ns an d v al ves a re

c h ec ke d for l e ak s .

● A fte r the sys te m h a s be e n fl us h ed , c l os e the fee d p res sure c o ntro l v a l ve , but

m a k e su re th at the c o nc e ntra te c on tro l va l ve is o p en .

● O p en th e fee d p re ssur e co n trol va l ve l i t t le by l i t tl e so th at fe e d pr es su re d o es n o t

e x ce ed 4.0 k g /c m 2 (60 psi) an d th en s ta r t the hi g h p res su re p u m p .

● In c re a se op e ni n g of the fe e d pre s su re v al ve s l o w l y to el e v ate the fee d p re ssure

a n d fee d fl ow ra te to the e l e m e n ts un ti l the d e s ign co n ce n tr ate fl o w is re a ch e d.

T h e n s l o w l y c lo s e the c o nc e ntra te c on tro l v al ve un ti l the ra ti o of pe rm ea te fl ow to

c o nc e ntra te fl ow a pp ro a ch e s the de s ig ne d re c ov e ry rati o .

● R ep e at op e ni n g of the fe e d pre s su re c on tro l v al ve an d c l os ing of the c on c en tra te

c o ntro l v a lve u nti l the de s ig n p erm e ate a n d co n ce n trate f lows are o bta i ne d , w h i le

c h ec kin g the s ys te m p res su re to e n su re th at it do e s not e x ce e d the u p p er de s ig n

l i m it.

● A fte r a d ju s tin g the two v al ves , c a lcul a te the s ys te m re co v er y a n d c om p are it to

th e s ys tem d e s ig n va l u e.

● C he c k c h em ica l a d di t i on s of a c id , s ca l e i nh i b ito r , an d s o di u m m eta b isu lf i te .

M e as u re pH , c on d uc ti vity, c a lc iu m h a rdn e ss , a nd a l ka l in i ty le v e ls to c al cul a te the

L a ng e l ie r S a tu rati o n Ind e x (L SI) or the S ti ff & D a v is S tab i lity Ind e x (S & D S I) for a

p o ss ib i lity of sc a le fo rm ati o n.

● T a k e the first rea d i ng of all op e ra ti n g p ara m e te rs a fte r al low in g the sys te m to ru n

for o n e h ou r . R e a d the pe rm ea te co n d uc ti vity fro m e ac h p re ssure ve s se l a nd

i d e nti fy a ny v es se l w ith a n y m a l fun c tio n .

● A fte r 24 to 48 ho u rs of op e rati o n, rec o rd all p l a nt pe r fo rm a n ce d a ta su c h as feed

p re s su re, di ff e re nti a l p res su re , te m p e ra tu re , f l o w s , re co v ery ra ti o an d c o nd u c ti vity

re a d in g . And also an a l yz e co n s ti tu en ts of fee d w a te r , co n ce n tr ate , a nd p e rm e a te

w ate r s a m p l e s . C o m p a re s ys te m p e r form a nc e to d es ign v a l ue s . U s e the i n i t i al

s ys te m p er fo rm a n c e in fo rm a ti on as a re fe ren c e for e va l ua ti ng fu tu re sys te m

p e r form a nc e . M e a s ure s ys te m p er fo rm an c e reg u l ar ly du r in g the first w ee k of

o p er ati o n.

2. Initial S ta rt-u p



- 75 -

6 -3 . R e c ord K ee p i n g fo r M a in te na nc e

All re l ev a nt da ta of the RO sys te m is s tron g l y re c o m m e nd e d to be re c or de d p ro pe r ly.

Th e y are not o nl y n ec e s sa ry for fo l lo w i n g the pe r fo rm a n ce of the s ys te m , but also

va l ua b l e too l s for tr ou b le s h o oti n g an d a l so n ee d ed in th e c as e s of w a r ra nty c la i m s.

T h e RO s ys tem pe r fo rm a n ce d e pe n d s for the m o s t pa r t on the pr op e r o pe ra ti on of

the p re tr ea tm en t a n d th u s the o p e rati n g ch a ra c te r is t i cs of the p re trea tm en t e q ui p m en t

for the fo l lo w i n g ite m s sh o ul d be re co rd e d an d m a in ta in e d .

1 ) Pr etr ea tm en t S y ste m

● D i sch a rg e pre s su re of a ny well or b oo s ter p um p s an d p re ssur e dro p of all fi lters

i n c lu di n g s an d fi lte r , m u l ti m ed i a filte r , a nd a c tivate d c ar bo n fi lte r . T h e d ata of the

p re s su re dr op m a y in d i ca te w h en a b a ckw a sh for the fi lte rs is ne e d ed .

● T o ta l re s idu a l c hl o r ine c o nc e ntra ti on in th e RO fe ed .

● R eg e ne ra tio n p e r io d of w a ter s oftn er if u se d for the rem o v al of h a rd ne s s.

● In l et an d o utl e t p re ssure of m i cro fi lte rs an d c ar tr id ge fi lters . An i n c re as e in the

d i ff e ren ti al pre s su re b etw e e n i nl e t a n d ou tl et pre s su re m a y i n d icate th e ti m e for

c l e an i ng an d b ac kw as h i ng of the m ic rofi lte rs or a re p l ac e m e n t of c ar tr id ge fi lters .

● S i lt D e n s ity In d ex ( SD I) a nd tu rb i di ty of the RO feed s tre a m . M e a su re S D I an d

th e tur bi d i ty b e fo re a n d afte r all the fi lters .

● C on s um p tio n of a c id a n d any oth e r c h em ica ls s u ch as c o ag u la n ts an d s ca l e

i n h ib i tor s. S c a le i n h ib i t i on c a n also be ac co m p lish e d by pH co n tr ol (us u al ly pH

6 -7 ) , d e pe n di n g on the a m ou n t of h ard n es s in the fe e dw ate r .

2 ) R O O p e ra t in g D a ta

T h e fo llo w i ng d a ta m us t be re co rd e d fr eq u en tl y, p re fer r ab l y on c e per sh i ft.

● C om p l ete w a te r a n al ys is of th e fe ed , p er m e ate ,a nd c o nc e ntra te w a ter a nd th e ra w

w ate r b e fore p re tre atm e nt o n ce at s ta r t-up an d e ve ry w e e k th e re afte r .

● pH of the fe ed , p e rm e a te an d c on c en tra te w a te r . T e m p era tu re of the fe e d w a te r .

● L a ng e l ie r S a tu rati o n Ind e x (L SI) of the c o nc e ntra te w a ter fro m the l a s t a r ray (for

c o nc e ntra te w a ter < 1 0 ,00 0 p pm TD S ) .

● S ti ff a n d D a v is S tab i lity In d ex (S & D S I) of the c o nc e ntra te w a ter fro m the l a s t

a r ra y (for c on c en tra te w ate r > 1 0,0 00 pp m T D S ) .

● F e e d flo w pr es su re a fter hi g h p re ssure p u m p.

● F e e d, pe rm ea te a nd c o nc e ntra te fl ow p res sure of e a ch a r ra y. P res sure d ro p per

3. R e c o rd K e e p ing for M a inte n a n c e



- 76 -

c a r tr id g e an d p e r a r ra y.

● P e rm ea te a nd c o nc e ntra te f lows of ea c h ar ray . C a l cu l ate re c ov e ry rati o to e n su re

th a t it d oe s not go b e yo nd the d e s ign li m it.

● C on d uc ti vity /T D S of the fee d , p erm e ate a n d co n ce n trate s tre am s for e ac h a r ra y.

T h e T D S of the RO c o nc e ntra te c an be u s ed a l on g w i th the fee d w a te r T D S to

c a l cu la te an a ve ra ge m e m br an e c on c en tra tio n w h i ch c a n be us e d to ca l cu l ate an

a v er ag e m em b ra ne s a l t re j ec ti on as s ho w n in th e fo llo w i ng e q ua ti o ns .

p e rm e a te fl o w × p er m e ate T D S + c o nc e ntra te fl ow × c o nc e ntra te T D S

= fe ed fl o w × fee d T D S

A v er ag e s al t r ej e c ti o n( % ) = 〔1-

pe rm ea te T D S

〕×100( fe ed T D S + c o nc e ntra te T D S ) / 2

● C al ibra ti o n of all g au g es a n d m ete rs b as e d on m a n ufa c tu re r 's rec o m m e nd a tio n s

at l e as t o n ce ev e ry th re e m on th s . Im p or ta n t g au g es in c lu d e pH m e te rs , f lo w

m e ters , p re ssure ga u ge s , a nd c o nd u c ti vity m e te rs. It is r ec o m m e nd e d tha t the

w ate r pH is ve r ifi e d w e e k ly an d on a w e e k ly sc h ed u l e, th e pH p ro b e sh o ul d be

p l a ce d i n to b u ff er s ol u ti on s w i th p ar ti cul a r pH va l ue s , c al ibr ati n g the p ro be to

th o se v a l ue s . If the v a l ue s a re d r ift i n g ev e ry ti m e the pH m ete r is c h ec ke d , the

re fe ren c e p rob e n e ed s to be rep l a ce d or h a ve i ts K C l so l uti o n re pl e n ish ed .

3 ) M a in ten a n c e L o g

● R ec o rd re gu l a r m a i nte na n ce .

● R ec o rd m ec h an i ca l fa i lu re s an d re p la c em e n ts or ad d i ti o n s of RO d ev ice s an d

p re trea tm en t e q ui p m en t s u ch as c a r tr id g e filter s.

● R ec o rd a ny c ha n ge of m e m bra n e e le m e nt lo c ati o ns with el e m en t s er ia l n um b e rs

● R ec o rd c al ibra ti on of a l l g au g es a n d m ete rs.

● R ec o rd all c l e an i ng s of RO m e m bra n es i n c lu d in g d a te , d u rati o n of c l e a ni n g .

C le a ni n g a ge n ts an d c o nc e ntra ti on , s o lu ti on p H , te m p era tu re, f lo w rate a n d

p re s su re.

3. R e c or d K e e p ing for M a inte n a n ce



- 77 -

6 -4 . D a ta N or m a li za tio n

T h e p er fo rm an c e of an RO sys te m is i n fl u e nc e d by c h an g es in the fee d w a te r T D S ,

feed pr es su re , te m p e ra tu re a n d re co v ery ra ti o. D ata n o rm a l iza ti o n is a p roc e s s to

co n ve r t the re al pe r fo rm a n ce of the RO s ys tem in to a form which c an be c om p a red to

a gi ven re fe ren c e p er fo rm a n c e which m ay be the d e s ig ne d p e r form a nc e or the

m e a su re d i ni t i a l p er fo rm an c e.

A di ff e re nc e b etw ee n the no rm al ize d da ta a nd the i n it i a l or d e s ig ne d p er fo rm an c e

m a y i nd i ca te the re a re s om e p ro b le m s in the s ys tem as s h ow n be l o w .

● M e m b ra n e fo u l in g a nd / or s ca l in g

● M e m b ra n e ch e m ica l d a m a g e - p o or s al t r ej e c ti o n d ue to a c h em ica l c ha n ge in

th e m em b ra ne s tru c tu re by ex ce s sive e x po s ure to c h l or in e or e x tre m e p H .

● M e ch a ni cal fa i lu re - a b rok e n O -r in g or e l em e nt gl u e l i ne .

● H yd rau l ic p l u gg i n g - th e p res e nc e of fo ul a n ts ( la rg e s ize c ol loi d s ) or s ca l e s tic ke d

to the fl ow c ha n ne l s p ac in g b e tw ee n the m em b ran e l e av e s of sp i ra l - w ou n d

e l e m e n ts .

T h e p ro bl e m s c o ul d be i d en ti fi e d e ar ly w h en the n o rm a l ized d a ta are rec o rd ed d a ily.

Th re e re p res e nti ve va r iab l es su c h as s a l t re je c ti on , n or m a l ized d i ff e ren ti al pre s su re ,

an d n orm a lize d p e rm e a te fl o w ra te are ca l cu l ate d fro m the RO op e ra ti n g d ata . T h us , th e

eff e c ts of th e a bo v e four p ro b le m s ca n be d ire c tly m o ni to red by the th re e v ar ia bl e s as

shown b el o w .

● S a l t re j ec ti on

● N orm a lize d d i ff ere n tia l p re s su re

● N orm a lize d p e rm e a te fl o w ra te

3. R e c o rd K e e p ing for M a inte n a n c e



- 78 -

1 ) Sa lt R eje c t io n

S a l t re j ec ti on is th e m os t w id e l y k n o w n m eth o d of m o n i to r ing th e p er fo rm an c e of an

RO sys te m , th ou g h any pr ob l em in the s ys tem co u l d ha v e be e n n oti ce d a nd co r re c te d

so o ne r by m on i tor in g o th e r p a ram e ter s s u ch as n o rm a l ized d i ff e ren ti al pr es su re a n d

pe rm ea te flo w r ate , b efo re the pro b l em aff e c ts th e s al t r ej e c ti o n .

T h e re are two d iff e re n t m eth od s c o m m o nl y u se d to c al cul a te sa l t re j ec ti on ,

de p en d in g on the T D S v al u e at the m e m b ran e s ur fa c e. O n e m eth o d sh o w n b el o w

uti lize s the RO fee d w a te r T D S for the m e m bra n e s ur fa c e T D S .

S a lt re je c tio n (% ) =

(F e ed T D S - P e rm ea te T D S )

×1 00

F e e d T D S

T h i s m e th od will g ive a l o w er s al t r ej e c ti o n tha n the ac tu al in d i vidu a l e le m e nt s a l t

rej e c ti o n . T he ex ten t of the v ar ia ti o n will de p en d on the r ec o ve ry of the RO sys te m .

T h e o the r m e tho d u s es a m a the m a ti cal av e rag e of the fee d a nd c o nc e ntra te T D S to

ap p ro x im a te the av e rag e T D S w i thi n the RO s ys tem , w hi ch m a y be c l o s er to the re a l

TD S at the m e m b ran e s ur fa c e. T hi s m eth o d will a l so n o rm a l ize for c ha n ge s in s al t

rej e c ti o n d ue to c h an g es in th e RO p e rm e a te r ec o ve ry. T hu s an a ve ra ge fe e d T D S

pro v id e s a m ore a c cu ra te way to c a lcu la te s al t r ej e c ti o n :

A ve ra ge fe e d T D S =

fe ed T D S + c o n ce n tr ate T D S

2

S a lt re je c tio n (% ) =

av e ra ge fe e d T D S - pe rm ea te T D S

×1 00

a ve ra g e fee d T D S

If th e c on c en tra te T D S h a s not b e en m e a su re d, it ca n be e s ti m a te d us ing th e

pe rm ea te re co v ery of the sys te m w h e re re c ov e ry is a r ati o of the pe rm ea te fl ow r ate to

the fe ed fl o w ra te :

C o n ce n trate T D S = fe e d T D S ×1

1 - re c ov e ry ra ti o

T h e ra te of rej e c tio n v ar ie s for e ac h of the pa r ti cu l ar s al ts in the feed w a ter an d th us

a va r iati o n in the i o n c om p os iti o n of an RO fee d w a te r will re su l t in a c ha n ge in th e

ov e ral l p e rce nt rej e c ti o n of the T D S . H e n ce , it is s u gg e s te d to re c ord an i n di v id u al i o n

rej e c ti o n in o rde r to h a v e a ba s is for fu tu re p e r form a nc e c om p a r ison , w he n a s ys tem

4. D a ta N o rm a liz a tion



- 79 -

s ta r ts with a n ew m em b ra ne .

T h e d ata on the i nd i vidu a l i on rej e c ti o n s is also he l pf ul in d i a gn o s in g so m e s ys tem

m a l fun c tio n . F o r e x am p l e, a rej e c ti o n c al cul a ted u s in g a d i va l en t i o n su c h as ca l cium

ca n tel l a d i ff e ren c e be tw e e n a m ec h an i ca l l e ak in the s ys te m a n d m em b ra ne

de ter io ra ti o n . M e c ha n i ca l d am a ge s in m e m b ra n es , g l ue l i n es an d O -r in g s will re s u lt in

a s i m ilar d ec re as e in re j ec ti on for b o th m o n ov a l en t a n d di val e n t i on s , w h i le , in the

ca s e of m e m b ran e d ete r iora ti on , th e re j ec ti on d e c line will be m ore s e ve re for

m o n ov a l en t i o ns .

2 ) D if fe ren t ia l P re s su re ( Δ P )

D iff e re n ti a l p re ssure is th e d i ff ere n ce b e tw e en th e fee d p re ssur e an d the co n ce n trate

or br in e p res su re e x it i n g the e n d of the e l em e n ts . It is a m e a su re of the pre s su re d ro p

as the fee d w a te r p as ses th ro ug h the flo w ch a nn e l s of all the e l e m e n ts in the s ys tem .

At co n s ta n t f l ow r ate , an i nc re as e in the d i ff e ren ti al pr es su re i n di cate s tha t l a rg e

co l lo i da l p a r ti cles or p h ys ica l de b r is s u ch as p u m p s h av ing s , i n org a ni c s ca l e s an d

bi o film p ar ti cu l a te s a re b l oc kin g the fl o w c h an n el s . T h e te l e scop i n g of sp i ral w ou n d

el e m en ts ca n a l so c a us e an i nc re as e in the d i ff e ren ti al pr es su re , w h i ch is a fu nc ti on of

the p er m e ate a n d c on c en tra te fl o w ra tes . S i n ce th e se rate s m ay v a ry d a i ly d ue to

va r iati o n in w a te r te m p era tu re or s om e o the r ch a ng i n g pa ra m ete rs, the a c tu a l

di ff e ren ti a l pr es su re s h ou l d be no m al ize d a ccor di n g to the fo l lo w i n g eq u a ti o n to

co m pa re w i th the i ni t i a l di ff e re nti a l p re ssure .

N o rm a l ized

D i ff er en ti al

P re s su re

= a c tu a l ΔP ×(2 ×i ni t i a l c on c en tra te fl o w + i ni t i a l pe rm ea te flow) 1.5

(2 ×c on c en tra te fl o w + p erm e ate flow) 1.5

A pe rcen t c h an g e(e .g . 10 % ) in n o rm a l ize d di ff e re nti a l p res su re c o ul d s u gg e s t w h en to

c l e an an RO s ys te m .

3 ) N o rm aliz ed P erm e a te F lo w ra te

N orm a lize d p er m e ate fl o w ra te is the m o s t im p o r tan t m o ni to r in g p a ram e ter for an RO

sys te m . N o rm a l iz in g for the e ff e c ts of pre s su re , tem p e ratu re , a nd s o l ute c o nc e ntra ti on

on p e rm e a te fl o w ra te will e n a bl e the re su l t in g fl o w v a lu e to re fl ec t c ha n g es d u e to

ch a rate r ist i cs of the m e m bra n e, the m e m bra n e s ur fa c e, a n d the i n te rgi ty of the

m e m br an e e l em e nts or ve s se ls . T h us the n or m a l ized p e rm e a te fl o w ra te c an be u s ed to

m o n i to r the fo l lo w i n g pro b l em s :

4. D a ta N o rm a liz a tion


6System Operation

- 80 -

① The extent of fouling and scale formation on the membrane surface, causing adecrease in the permeate flowrate.

② Membrane compaction, causing a decrease in the flowrate.

③ The integrity of the membrane system such as mechanical leaks in thesystem, causing an increase in the flowrate

④ The extent of membrane deterioration, causing an increase in the flowrate.

The normalized permeate flowrate can be obtained by the following equation :

QN = QO ×

Pi -ΔPi

- Ppi - πi

×2 TCFi

Po -ΔPo

- Ppo - πo

TCFo

2

QN = normalized permeate flowrate.

Qo = measured (actual) permeate flowrate.

Pi= initial operating pressure, Po = actual operating pressure

ΔPi = initial differential pressure, ΔPo = actual differential pressure

Ppi = initial permeate pressure, Ppo = actual permeate pressure

πi= initial osmotic pressure, πo = actual osmotic pressure

TCFi = initial temperature correction factor

TCFo = actual temperature correction factor

4. Data Normalization


6 System Operation

- 81 -

6-5. Control of Microorganism Growth in RO System

Biofouling is one of the most common and severe problems in the operation of RO

systems. Thus it is very important to control the microorganism growth using properly

designed and operated pretreatment including an equipment for efficient chlorination and

dechlorination (before RO membrane) process.

It is a good idea to build the systems in such a way to reduce spots of stagnant

flow such as blind long pieces of piping where microorganisms can easily be settled to

grow.

Most of all, checking the microorganism load in the system and sanitizing the system

periodically is the usual and best way to keep the microbiological activity under control.

The frequency of sampling of microorganisms and analysis depends on the risk of

biofouling. A daily check of the feed water after dechlorination and a weekly check of all

points as shown below is recommended for surface water plants.

① Intake before chlorination.

② After a clarifier or similar sedimentation process.

③ After filtration units (sand, multimedia, activated carbon or other).

④ After dechlorination (usually after cartridge filtration or just before entering

RO units in the case that sodium bisulfite is used for the dechlor ination).

⑤ Concentrate water.

⑥ Permeate water.

The following system checking is very helpf ul in controlling microbial activity. Open

basins or tanks should be disinfected properly at the open source and the part of the

system down-stream from it should also be sanitized frequently. The air breathing

(ventilation) systems of sealed tanks should be equipped with HEPA filters.

The backwash of the media filters should be done with water free of microorganisms

or sufficiently chlorinated water. In general, the water used for disinfection, flushing and

cleaning solutions should be of good quality and free of microorganisms (e.g. permeate

water).

The whole pretreatment system including all piping, tanks, manifolds, the retention

tanks and filters should be disinfected before each start-up following shutdown times

and periodically even when the RO plant is being operated continuously.

5. Control of Microorganism Growth in RO System


7. C le a n in g a n d

D i s in fe c t io n


7-2 . Tim i n g for C le an in g

7-3 . C le an in g Tank and O ther

E qu ip m en ts

7-4 . C le an in g P roce d u re

7-5 . C le an in g C hem ic a l s

7-6 . D is in fec tion


7C leanin g an d D isin fec tio n

- 84 -


F o u lin g of RO m em b ra ne is m o re or l e ss a n o rm a l p h en o m e n on of m o s t RO

sys te m s s i n ce th e p retre a tm e n t of the feed w a ter pr io r to the RO m em b ra ne is

ba s ic ia l ly d es ign e d to red u ce fo u l in g s ub s ta n ce s as m uc h as p os sib l e an d te ch n ica lly

co u ld n o t re m ov e all of th e m . F or tu n ate l y, w ith c o r re c t c l e an i n g freq u en c y, m o s t

fo u l an ts c an be re m ov e d from the m e m b ra n e. T he c l ea n i ng fre q ue n cy c o ul d be

m i n i m ize d as l on g as the p re trea tm en t is w el l m ai n ta in e d w i tho u t u ps e t c on d i ti o n s

su c h as an u n co n tr ol led c h an g e in fee d w a ter c om p o s it io n a n d un c on tro l le d b io l o gi ca l

co n ta m in a tio n . S o m eti m e s m i sta ke s in the sys te m o p era ti on s u ch as to o h ig h re c ov e ry

an d fai lu re of ch e m ica l d o s in g sys te m s co u l d en d up with fou l in g the m em b ra ne .

T h e fo ul ing of m e m bra n e s ur fa c es re s ul ts in l o w e r p er m e ate fl o w ra te a nd /o r l o w e r

sa l t re j ec ti on . In c re a se d p re ssure dro p b etw ee n the feed a nd c o nc e ntra te s id e c a n also

oc cu r fro m th e fou l ing .

C le a ni n g the fo u l ed m e m br an e s ca n be a ccom p l ishe d by s u i tab l e c l e a ni n g a ge n ts at

al ka l ine (u p to pH 12 ) a n d a c id ic (p H 2 ) c o nd i t i on s b ec a us e S a e ha n 's C S M RO

m e m br an e s are s ta bl e at th e pH c o nd i t io n s a nd at an e l ev a te d te m pe ra ture ( 45℃).

M a ny fo ul a nts , p a r ti cul a r ly c l a y- typ e s o i ls , ca n c o m p a c t with tim e as the fo u la n t l aye r

in c rea s es in d e pth . As the fou l an ts c om p a c t, it will b ec o m e m o re d i ff i cu l t to rem o v e

th e m d ur in g c l e a ni n g. T hu s the tim e of c le a ni n g m u s t n o t be d el a ye d too l o n g.

1. Introduction


7 Cle anin g an d D isin fect ion

- 85 -

7 -2 . T im in g fo r C le a n in g

E l e m en ts s ho u ld be c l ea n e d im m ed i a te l y w h en o n e of the fo l lo w i n g s ym pto m s is

de tec te d :

Lo s s of 10 to 1 5% in n o rm a l ize d pe rm ea te fl ow ra te.

D ec re as e of 0 .5% in s a l t re je c tio n .

T h e d i ff ere n tia l p re ssu re ( fee d p re ssure - c on c en tra te pr es su re ) ΔP i nc rea s es by 15

% from the re fer en c e co n di t i o ns ( in i t i al p e r for m a nc e e s ta b l ishe d d ur in g the first 24 to

48 h o urs of o p era ti on ) .

It s h ou l d be no te d tha t it is i m p or ta n t to no rm al ize fl o w a n d sa l t c on te nt of the

pe rm ea te ac cord i n g to the n o m a l izati o n p roc e du re d e sc rib e d in Se c ti on 6.

2. Timing for Cleaning



- 86 -

7 -3 . C le a n in g T a nk a nd Othe r E qu ip m e nts

T h e m i xing ta n k for c l e a ni n g a ge n ts sh o ul d be m ad e of p ol yp rop y l en e or F R P w h i ch

is r es is tan t to pH in th e ra n ge of 1 to 12. T he c l e an i n g ag e nts work b e tte r at an

el e va te d tem p era tu re. (e .g . 3 5 -40℃). T he c l ea n i ng te m pe ra tur e sh o ul d not be be l o w 15

℃ at w h i ch the c l e a ni n g ra te is very slow. C oo l ing m a y also be re q ui re d to a vo i d

ov e rhe a tin g . So h ea ti ng an d c oo l in g e q ui p m en ts m ay be n ec e ssary to c o ntro l the

te m p era tu re of the c l ea n i ng s o l uti o n.

T h e s i ze of the ta n k n ee d s to be l a rg e en o ug h to c on ta i n the v o lu m e of the c l e an i n g

so l uti o n a pp ro x im ate l y eq u al to th e v ol u m e of the p re ssu re ve s se l p l us th e v ol u m e of

the fe ed a n d re tu rn p i p es . If it is d i ff i cu lt to ca l cu l ate th e e xa c t v ol u m e of the p i pe s , it

ca n be as su m e d to be a b ou t 2 0 % of the v e sse l vo l u m e . A p p ro pr ia te pu m p , v al ve s ,

fl o w m e te rs, a n d pr es su re g a ug e s ho u l d be i n s ta l le d to c o ntro l the flow.

3. Cleaning Tank and Other Equipments



- 87 -

7 -4 . C le a n in g P r oc e du re

1. F i ll the c l ea n i ng ta n k with RO p erm e ate w a te r . T h e vo l u m e of c l e an i ng so l uti o n

s h ou l d be su ff i cie nt to fill all the pre s su re v es sel s a nd p i p e li n e s . Add the

c a l cu la te d am o u nt of th e c l ea n in g c h em ica l s to the tan k . U s e a m i xe r or

re c ircul a te the s o l uti o n with the tra n s fe r p u m p to e n s ure th a t all c h em ica l s a re

d i ssov e d an d w e l l-m ixe d be fo re c i rcu l a tin g the so l u ti o n to the el e m en ts .

2. D r ai n m o s t of the w a te r fro m the RO sys te m to p re ve n t the d i luti o n of the

c l e an i ng so l uti o n by w ate r w ith i n the RO sys te m .

3. H e a t the so l u ti o n to the te m p era tu re re co m m e n de d by th e m a nu fac tu rer to

i m p rov e c l ea n i ng e ff e c ti ve n e ss .

3. P u m p the p re h ea ted c l e an i n g so l u ti o n to the v e s se l at co n di t i o ns of l o w fl o w rate

(a b o ut h a l f of tha t s h ow n in Ta b l e 1 ) a n d l ow p res sure to d i sp l ac e the pr oc e ss

w ate r re m a in i n g in the v e ssel . D um p the d ispl a c ed w ate r u n til th e p res e nc e of

th e c le a n in g s ol u ti on is e v id e nt in th e RO c o nc e ntra te s ys te m or in th e re tur n

p i p e, i n d i ca te d by the pH an d te m p e ra tu re of the c l e a ni n g s ol u tio n . A d j us t f l o w

ra te a nd pre s su re a ccor di n g to the T a b le 1. O p en th e RO c o nc e ntra te thro tt l in g

v a l ve c om p l ete l y to m i n i m ize o p era ti ng p re s su re d ur in g c l e a ni n g. U s e on l y en o ug h

p re s su re to re c ircu l ate the c l e a ni n g s ol u tio n w i th ou t p e rm e a te co m in g o u t.

T a b le 1 : F e ed F l o w R ate of C l e a ni n g S o lu ti on p e r P re s su re V e ssel du r ing

R e c ircu l ati o n

E le m e n t D iam e te r( in )

M a x im u m F ee d F lo w R a te

( g pm ) (m 3/h )

2.5 5 1.1

4 10 2.3

8 40 9

5. R e c yc le the c o nc e ntra te to the c l e an i ng s o l uti o n tan k u nti l the de s ire d te m pe ra ture

is m a i nta i ne d th rou g ho u t the sys te m . O b s e rve a n y in c rea s e in the tu rb id i ty to

j u d ge e ff i cien c y of the c l ea n i ng s o lu ti o n, e s pe c ia l ly in the ca s e of an al kal ine

c l e an i ng so l uti o n or a d ete rg en t s o l uti o n. If the c l ea n i ng s o l uti o n be c om e s tu rbi d

or co l or ed , d ra in the s o lu ti on a n d re s ta r t w ith a fre s hl y p rep a re d c l e a ni n g s ol u tio

n. C h ec k the pH du r ing a c id c l e an i n g. T h e a c id is c on s u m e d w h e n it d isso l ve s

4. Cleaning Procedure



- 88 -

i n o rga n i c p re c ip i ta te s . If the pH in c rea s es m o re th an 0 .5 pH u n it, ad d m or e ac id.

6. T u rn the pu m p off a nd a l lo w the e l e m e n ts to so a k . S om e tim e s a so a k p er io d of

a b ou t 1 h o ur is s uff i cie nt. F o r di ff i cu l t fo ul ing an e x te n de d s o ak p e r io d is

b e ne fi cial ; s o a k the e l em e n ts for 10 -1 5 ho u rs. To m a i n ta i n a h i gh te m pe ra tur e

d u r in g an e x te nd e d s oa k p er io d, use a s l ow re c ircu l a ti o n ra te (ab o u t 10 % of th at

s h ow n in T ab l e 1).

7. C i rcul a te the c l ea n i ng s o l uti o n at the ra te s sh o w n in T a bl e 1 for 3 0 -60 m i n u te s .

T h e h i gh fl o w ra te fl us h es out the fou l an ts l o os e ne d fro m the m e m br an e s ur fa c es

by the c l e an i n g. If the e l em e nts a re h e av ily fou l e d, a fl o w rate w hi ch is 50 %

h i g he r th a n sh o w n in T a bl e 1 m a y ai d c l ea n i ng . At h i g he r fl o w r ate s e xces sive

p re s su re dr op m a y be a p ro b le m . T h e m ax im u m re c om m en d ed d ro ps are 1 .4 b ar

(2 0 P S I) p e r e l em e n t or 4.1 bar (60 P S I) per m u l t i-e l e m e n t v es se l . T h e d ire c ti o n of

f l o w d u r in g c l ea n i ng m u s t be th e s am e as d ur in g n orm a l op e ra ti o n to a vo i d

te l es co p i ng of th e e le m e nts .

8. D r ai n the us e d c le a ni n g s ol u ti on o u t of the sys te m . A na l yz e a s a m p l e of the us e d

s o l uti o n to de ter m ine th e typ e s a nd th e a m ou n t of su b s ta n ce s ( fo ul ing m a te r ia l s)

re m o ve d fro m the m e m br an e e l em e nts . T h e re su l ts co u l d tel l the de g re e of

c l e an i ng an d the ca u se s of fou l ing .

9. RO p e rm e a te or g oo d q ua l ity w a te r ( f i lte re d, S D I < 3), fre e of b ac te r ia a n d c hl o r in

e, c o nd u c ti vity < 1 0,0 0 0 s /c m is us e d for f l us h in g out the re s id u al c l e a ni n g

s o l uti o n. Th e m i n im u m flu s h out te m p era tu re is 20 ℃ to p re v en t p re c ip ita ti on .

1 0 .Th e RO p l a nt is s ta r ted up a ga i n re s um ing n o rm a l o p era ti ng co n di t i on s . H o w e v er ,

th e p erm e ate m u s t be d ra i ne d u n ti l c o nd u c ti vity an d pH re tu rns to n o rm a l . And

a l so th e p erm e ate s i de dra i n in g is n ec e ssa ry w he n a no th er c le a ni n g c yc le w i th

a n oth e r c l ea n i ng c h em ica l is to fol low . D u r ing the r i n se out s te p , the op e rati n g

p a ra m e te rs s h ou l d be n ote d to j ud g e the c l ea n i ng e ff i cie nc y an d to d ec ide if

a n oth e r c l ea n i ng is re q ui re d. If the s ys te m ha s to be s hu t-d ow n afte r c l ea n i ng for

l o n ge r th a n 24 h ou rs , the e l e m e n ts sh o ul d be s tore d in a p re se rv a ti o n s ol u ti on

s u ch as 1% s od i u m b isul f i te a nd 0 .5 % form a ld e hyd e.F o r m u l ti -a r ra y s ys te m s,

c l e an i ng sh o ul d be c ar r ie d out s e pa ra tel y for e a ch a r ra y. Th i s ca n be

a c co m pl ish ed ei th er by u s in g o ne c l e an i ng pu m p an d o p era ti ng o n e a r ra y at a

ti m e, or us ing s e pa ra te c l e a ni n g p um p for e a ch a r ra y.

4. Cleaning Procedure



- 89 -

7 -5 . C le a n in g C he m ic a ls

C ho o s in g r i gh t c l ea n i ng c h em ica l s is i m p or ta n t s in c e h arsh a nd fre q ue n t c le a ni n g will

sh o r ten the m em b ra ne l i fe , a nd s o m e ti m e s a w r on g c ho i ce of c l ea n i ng c h em ica l s ca n

w or se n the fo u lin g s i tu ati o n. T he c l e an i ng will be m or e eff e c ti ve if it is tai lor ed to the

sp e c if ic fo ul in g pro b l em . T h e refo re , the typ e of fou l a nts s h ou l d be de ter m ine d p r io r to

c l e an i n g, th e re are he l p f ul ways to d ete rm in e the typ e of fou l an ts as sh o w n b el o w :

A na l yz e the pl a n t pe r fo rm a n ce da ta ;

A na l yz e the fe e d w a ter to fin d p o te n tia l fo u lin g su b s ta n ce s ;

C h ec k the r es u lts of p re v io us c l e an i ng s w h i ch m a y i nd i ca te sp e c if i c fo u l in g

su b s ta n ce s ;

A na l yz e the fo u l an ts c ol lec te d with a m em b ra ne fi lte r us e d for S D I m e as u rem e n t;

A na l yz e the de p os its on the c ar tr id ge fi lte r ;

Ins p ec t the i nn e r s u r fac e of the fe e d l in e tu bi n g a nd th e fe ed e n d sc ro ll of th e

RO e le m e nt. If it is r ed d ish- bro w n , fo ul in g by i ro n is po s sibl e . B i o lo g i ca l fo ul ing

or an org a n ic m a te r ia l de p os it is o fte n s l im y or ge l ati n o us .

T a b le 2 li s ts su i tab l e c le a n in g c he m ica ls d e pe n di n g on the typ e of fo u l an ts . T he a c id

c l e an e rs are to r ed i ssol ve in o rg an i c de p os its i n c lu d in g i ro n , w h i le the al kal ine c l e an e rs

are to re m o ve o rg a ni c fou l in g i n c lu di n g b io l o gi cal m a tter . S ul fu r ic a c id s h ou l d not be

us e d for c l e a ni n g b ec a us e of the risk of ca l cium su l fate s c al ing .

F o r th e p re pa ra tio n of the c l e a ni n g s ol u tio n s , RO p erm e ate is p re fe r re d , but

pre fi lte re d raw w a te r m a y be u se d . T h e ra w w a te r c o ul d h av e s o m e b u ff er in g a bi lity, so

m o re a c id or h yd ro x id e m ay be n ee d ed to r ea c h the d e s ired pH l ev e l, w h ich is a bo u t 2

for ac id c l e an i n g an d a b ou t 12 for a lka lin e c l e a ni n g at 30 ℃, r es p ec ti ve l y. At 35 ℃, the

pH l i m it is in the ra n ge of 2 to 1 1 , a nd at 50 ℃ the o pe ra bl e pH ran g e is 3 to 10.

5. Cleaning Chemicals



- 90 -

T a b le 2. C l ea n in g C h e m ical s for C S M M e m b ra ne

Foulant Cleaning Chemical Comments

Inorganic salts

(CaCO3,CaSO4,BaSO4)

0.2% Hydrochloric Acid.

0.5% Phosphoric Acid.

2.0% Citric Acid.

Best

O.K.

O.K.

Metal Oxides

(Iron)

0.5% Phosphoric Acid.

1.0% Sodium Hydrosulfite.

Good

Good

Inorganic Colloids

(silt)

0.1% Sodium Hydroxide (NaOH), 30℃

0.025 Sodium Dodecylsulfate/0.1% NaOH, 30℃

Good

Good

Biofilms

0.1% Sodium Hyudroxide,30℃.

1.0% Sodium EthyleneDiamine Tetra

Acetic Acid (Na4EDTA) and 0.1% NaOH, 30℃

Best

Best When biofilm

contains inorganic scaling

Organics0.025%Sodium Dodecylsulfate/0.1% NaOH, 30℃.

0.1% Sodium Triphosphate/1% Na4 EDTA

Good

Good

Silica

0.1% Sodium Hydroxide, 30℃.

1.0% Sodium EthyleneDiamine Tetra

Acetic Acid (Na4 EDTA) and 0.1% NaOH, 30℃

O.K.

O.K.

T a b le 2 sh o w s the w o rking fo rm ul a for c l ea n i ng s o l uti o ns , but br an d n am e c l ea n i ng

ch e m ica l s are fre q u en tl y u s ed in th e fi el d ra th er tha n s el f-m ad e fo rm u l a ti o n s . M o s t of

the b ra nd n a m e ch e m ica l s are co m pa ti bl e w ith C S M m em b ra ne s in s ho r t te rm tes t.

Th e l o ng te rm c om p a ti b i lity test i n c lud i n g c l e a ni n g e ff i ca c y test sh o ul d be c a r r ie d o ut.

In th e m e an ti m e , th e y c a n be us e d as l o n g as the m e m b ra n e pe r fo rm a n ce is c a refu l ly

m o n i to re d to d ete c t any l on g te rm eff e c ts at an e a r ly s tag e . In an y e ve n t, m ak e s ure

th a t the bra n d na m e c he m ica ls do not co n ta i n c ati o ni c a nd n o n io n i c s u r fac ta nts , a nd

the pH of th e c le a n in g s ol u ti on fro m the c he m ical s d oe s not ex ce e d the li m its at the

sp e c if ie d te m pe ra ture .

5. Cleaning Chemicals


7 Cleaning and Disinfection

- 91 -

7 -6 . Dis infec tio n

If the RO system is suspected to be infected by bacteria or mold, e.g. slimy deposit

or rotten smell, a disinfection should be performed after the cleaning . The procedure

is the same as for cleaning, except the high flow pumping step.

Commonly used disinfectants are formaldehyde, hydrogen peroxide, peracetic acid,

and chlorine. Quaternary ammonium disinfectants, iodine, and phenolic compounds

should not be used because they cause flux losses.

The effective concentration of formaldehyde is in the range of 0.5 to 3%. Care

should be taken in handling this chemical since it is considered a carcinogen.

A 400ppm peracetic acid solution (also containing 2,000 ppm of hydrogen peroxide)

can be used to disinfect the RO system. The biocidal efficacy of peracetic acid is

much higher than hydrogen peroxide. Care must be exercised not to exceed the 0.2%

concentration as a sum of both compounds. Only periodic use is recommended since

continuous exposure at this concentration may damage the membrane. When the

peracetic acid is used, pH adjustment is usually not required.

However, when hydrogen peroxide is employed alone upto 0.2% concentration, the

pH of the solution is preferably adjusted to be 3. This will ensure optimal biocidal

effect and minumum damage to the membrane. If an alkaline cleaning has preceded

disinfection, an acid rinsing is recommended for both sides of the membrane.

Additionally, hydrogen peroxide can attack the membrane more aggresively at

temperature above 25℃ and in the presence of transition metals such as iron and

manganese.

CSM membranes can withstand short term exposure to free chlorine (hypochlorite).

However, eventual degradation may occur after 200-1000 hours of exposure to one

ppm chlorine, depending on feed water characteristics, e.g. pH and the presence of

heavy metals. Thus chlorine is not recommended for disinfecting the membrane, but

can be used in the pretreatment prior to the RO elements.

Disinfection using chloramine, chloramine-T, and N-chloroiso-cyanurate is not

recommended, since their eff ectiveness as disinfectants at low concentration (<3㎎/ℓ)

is limited and the compounds can also slowly damage the membranes.

6. Disinfection

CSMR ev er se O sm os is M em b ran e

8. P r o c e d u r e s fo rR e p la c i n g E le m e n t s

8-1 O pen in g P ress ure V es se l

8-2 . E le m e n t R em o v a l and Loa d i n g

8-3 . C lo s in g Vess e l


8Pr ocedures fo r R eplacin g Elements

- 93 -

8 -1 . Ope n in g P re s s ure V e s s e l

S E T P 1 R E L IE V E P R E S S U R E

1. S h u t off all s o ur ce s of pre s su re a nd rel iev e p re ssur e fr om th e v es sel , fo l lo w i n g

th e s ys tem m a n ufa c tu re r 's rec o m m e nd a tio n s .

S T E P 2 D IS C ON N E C T P E R M E AT E P O R T

1. D i sco nn e c t an d re m ov e p e rm e a te pi p i ng fro m the pe rm ea te p or t of the v e ssel .

S T E P 3 E X AM I N E E N D C LO S U R E

1. E x a m in e en d c l os u re of ve s se l for c o r ros ion . If a n y is e v id e nt, pro c ee d as fol low s

:

A. L o os e n an y d ep o s its with a s m a l l wire b rus h a nd /o r a m e di u m g ra de p i e ce of

S c o tch b r ite .

B. F l u sh aw a y l oo s e ne d d ep o s its with c le a n w a ter .

S E T P 4 R E M O V E IN TE R LO C K

1. U s e a 1 /4 " h ex k e y w re nc h to re m ov e the thre e s ec u r in g s c re w s fro m the ye l lo w

p l a s ti c se c ur in g r i n g. P l a c e on e of the screws a s ide for use on re as se m bl y .

T h re ad the o the r two sc re w s i nto th e thr ea d ed j a c king h o l es in the se c u r in g r i n g

u n til th e y c o nta c t the b e ar in g p la te . If e xces sive co r ro s io n bu i ld-u p p re ve n ts

th re ad i n g,

A. A p p l y pe n etra ti ng fl u i d (su c h as W D -4 0 or L P S -1 ) to i nte r fa c in g are a s of

s e cu r ing r i n g .

(S ec u r in g r i n g m ay h av e b e co m e bo n d ed to l o ck in g r i n g set a n d/o r be a r in g p l ate .)

B. W i th a s c re w d r ive r h a nd l e or s i m ila r to o l , tap th e s ec u r in g r i n g to re le a se the

b o nd .

C. A ga i n a tte m pt to rem o ve r i ng by turn i n g the j a ck in g s c re w s an a d di t i on a l 1/4

tu rn.

2. G r ip the two (2) sc re w s fi rm ly a n d pu l l the se c ur in g r i ng to w a rd s you. O n c e r i ng

is fre e , c are fu lly re m o v e it from the b e ar in g pl a te.

1. O p e n ing P re s su re V e s s e l

RR


8 Pr ocedures for R e placin g Elements

- 94 -

P re ssur e V es se l of a d va n ce d S tru c tu re s Inc . for 8 0 40 E l e m en ts



- 95 -

3. If r i n g c an n ot be fre e d by th e m eth o d, th e n use the s c re w s as j ac kin g sc re w s by

tu rn i ng 1/4 tur n on e a ch screw.

4. O nc e r i n g h as s ta r ted to m o ve , re p ea t p ro c ed u re in s te p 2. a bo v e.

5. U se a c u sh i o ne d m a lle t or h am m er in c on j un c ti on w ith a w oo d b l oc k to tap

b e ar in g p la te fac e . T h i s s h ou l d free l oc kin g r i ng s e gm e n ts .

6. R e fe r to F i g ure . 1 for l o c king r i n g s eg m e nt id e nti f i ca ti on .

7. R o ta te lo c king r in g set so th a t s eg m e nt A (k ey s eg m e nt) is at the 12 o 'c lo c k

p o s it i on . S e g m e n t A c a n no w be rem o ve d . If s e gm e nts will not r ota te, rep e at 1

A, 1B, a n d 5.

8. R e p e at s te p 7 ab o v e for se g m e n ts B & C .

9. A fte r re m ov a l of s eg m en ts , re m o v e all d e br is (c or r os ion p ro du c ts , di rt, e tc .) fr om

th e v es se l e n ds .

BEVELED

ENDS

BEVELED

ENDS

SQUARE

ENDSA

B

C

F i g ure 1.

S T E P 5 R E M O V E H E AD

S T E P 5 A R E M O V A L B Y H A N D

1. T hre a d a s ho r t l e n gth (1 2 i n .) of 1 in. I.D . pi p e i n to the p er m e ate p o r t a nd pu l l

th e h ea d s trai g h t o ut. A sh a rp for ce fu l tu g m ay be re qu i re d to s ta r t the he a d

a s se m bl y m ov ing .

2. If the h e ad s e al re m a i n s in the v es sel b o re , it s ho u l d be re m o ve d at th is ti m e.

1. O p e ning P re ss u re V e s se l



- 96 -

S T E P 5 B R E M O V A L U S IN G H E A D TO OL

1. In se r t th e to ol in to the sh e ll w ith th re ad e d ro ds in l i n e with be a r in g p l ate h o l es .

2. T h re ad th e ro d s in to the be a r in g p l ate h o le s a n d tu rn u n til th e k no b s bo ttom

o ut.

3. G ra sp to o l with bo th h an d s a nd p u ll s tra i gh t out to re m o v e the h e ad . If th e

h ea d will not re l e as e fro m the s h el l,

A. T h re ad a 1" ID p i pe a p p rox im a tel y 2 feet lo n g i nto th e p erm e ate p o r t.

B. C a re ful ly rock the h e ad as se m bl y b ac k a nd fo r th to re l ea s e the s e al .

C. O n ce th e h ea d s ea l h a s be e n b rok e n, co m p le te re m ov a l as in s truc te d in

the O p e ni n g V e ssel se c ti o n .

4. To re m o ve th e too l fro m the he a d , re in s tal l the he a d pa r t-w a y i nto th e s he l l so

th a t the to o l is co m p res se d , th en u n sc rew the rod s .

1. O pe n ing P re ss u re V e s se l



- 97 -

8 -2 . E le m e nt R e m ov a l a nd Loa d in g

Do not p ro ce e d the s te ps for r em o v in g el e m en ts u nti l all p re s su re h as b e en r el iev e d

fro m th e v es se l a n d bo th h ea d s h av e b ee n re m ov e d from the v e ssel .

S T E P 1 R E M O V E E LE M E N T IN TE R FA C E H A R D W A R E

1. R e m o ve th ru s t r i ng fro m d o w ns tre am (co n c en tra te ) en d .

2. R e m o ve a d ap te rs fr om e l e m e n ts at ea c h en d .

S T E P 2 E LE M E N T R E M O V AL

1. R e m o ve e l e m e n ts from ve s se l. C l ea n off a n y e x ce s s l u b r ican t fro m v es sel i n s id e

d i a m e te r b efo re re m ov ing e l e m e n ts . E le m e nts m u s t be re m o ve d in d ire c ti o n of

fe e d flow.

2. F l u sh o u t the ve s se l w i th c l ea n w a ter to re m o v e all d u s t an d d eb r is.

3. E x a m in e m e m b ran e e l em e n t su r fa ce s for a n y i m p er fe c tio n w h i ch c o ul d s c ratc h the

v e ssel bo re . P a y pa r ti cu l ar a tte n ti o n to e dg e s of an ti -te l e scop e d ev ice (A T D /b r ine

s e al ca r r ier ) . If a ny d e fe c ts are fo u nd which ca n no t e a s ily be c or rec te d, co n ta c t

th e e le m e nt m a n ufa c tu re r for c o r re c ti ve ac ti on .

4. U s ing an a p p rox im a te 50 % m i xtu re of g lyc er in e in w a te r , l u b r icate the i ns ide of th e

v e ssel . T hi s m ay b es t be a c co m p l ish ed u s in g a s ui ta bl y s i ze d sw ab s o a ke d in the

m i xtu re . T h is p ro ce d ur e will e a se m e m bra n e e le m e nt l o a di n g a nd r ed u ce c h an c e

of s c ra tc hi n g the ve s se l b ore .

5. L o ad th e first e l em e nt i nto the up s trea m a n d of the v e ssel . L e a ve a few in c he s of

th e e le m e nt pro j e c ti n g from the v es sel to fa c ilita te i nte rco n ne c tio n to the ne x t

e l e m e n t.

6. A p p l y a li g h t f i lm of a n on -p e tro l ium b a se d l u br ican t, s uc h as P ar ke r S u p er O -L u b

e, to the i nte rco n ne c tor O - r in g. (T he am o un t of O - lu b e s ho u ld be j u s t en o ug h to

g i ve a l u s tr e to the O -r ing . E x ce s s O - lub e m us t be re m ov e d to pr ev e nt po s sibi lity

of e l em e n t c on tam ina ti o n) .

7. A s se m bl e th e i nte rcon n ec tor to the l o a de d e l em e nt.

8. L i ne up the n ex t e l em e n t to be l o a de d a nd as se m bl e it to th e i nte rcon n ec tor

a l rea d y as se m bl e d on first e l em e nt.

2. E lem e n t R e m o v a l a n d L o a d ing



- 98 -

CAUTION

M ai n tai n e le m e nt a l ig n m e n t ca re ful ly d u r in g

a sse m b l y pro c es s . Do not al low e l em e nt w e ig h t

to be su p p or te d by in te rco n ne c to r .

M isa lig nm e nt ca n re su l t in da m a ge to

i n te rcon n ec to rs or pe rm ea te tu b es or to

e l em e nt ou te r s ur fa c e.

9. P u s h bo th e le m e nts i n to the v e ssel un ti l a few i nc h es a re p ro j ec ti ng fro m th e

v e ssel . R ep e a t l oa d in g p ro ce s s un ti l all e l em e nts a re i n s ta l le d .

10. W he n the fin a l e l em e nt is in s tal led , p u sh th e e l em e nt s ta ck fo rw a rd u nti l the face

of the fi rs t(do w n s trea m ) e l em e nt is at di m en s io n D as s h ow n in fi g u re 2.

T a k e ca re to a v oi d p u sh i ng e l e m e n ts too far as it c a n be d i ff i cu l t to p us h the s ta ck

in a rev e rse d i rec ti on .

D Downstream element Position

Fi g u re 2

V E S S E L T Y P E D IM E N S IO N " D"

(see figure 2)

E8U/SP 8.40 in.

E8L/SP 8.65 in.

E8B/SP 9.15 in

E8B/SP 9.90 in.

2. E lem e n t R e m o va l a n d L o a d ing


8Procedures for Replacing Elements

- 99 -

Alternate To Measurement Method

Insert a clean thrust ring into downstream end of vessel.

Insert head assembly, without quad seal or adapter, into downstream end of vessel.

Place the two square ended sections of locking ring into locking ring groove (with

squared ends together, stepped side outwards.)

Load elements as described in 5. through 9.

Install upstream adapter per Step 4. and head assembly per section on "Closing

Vessel."

STEP 4 INSTALL ELEMENT INTERFACE HARDWARE

1. Assemble adapter to element permeate tube at each end of vessel.

2. Element Removal and Loading


8 Procedures for Replacing Elements

- 100 -

8-3. Closing Vessel

Do not proceed closing vessel steps until elements and adapters have been installed

in vessel.

Do not proceed closing vessel steps until head has been checked for correct

component assembly, and vessel has been shimmed to prevent movement of the

elements if required

STEP 1 INSPECT SHELL INSIDE SURFACE

1. Inspect the vessel inside surface for any corrosion deposits or other foreign matte

r. If any are found, clean the surface as follows :

A. Using a medium or finer grade of ScotchbirteR and a mild soap solution, clean

each end of the vessel liner surface up to 8" in from each end of vessel.

B. Rinse away all loosened deposits from the shell inside surface using clean

fresh water.

2. Inspect the vessel inside surface for scratches or other damage which could

cause leaks. Vessels that leak must be replaced.

3. Inspect feed and concentrate port seals and attachments for internal and external

damage or deterioration. Contact Advanced Structures, Inc. for guidance, if

damage to the vessel's internal surface or feed / concentrate port, seals or

attachments are discovered during inspection.

STEP 2 SHELL AND HEAD SEAL LUBRICATION

1. Work O-ring lubricant glycerin into the shell from half way up the bevel to

approximately 1/2" in from the bevel. (See Figure 3)

2. Ensure the entire head seal is covered with a thin layer of O-ring lubricant, with

no dirt or dust contamination. Use only glycerin for lubrication.

STEP 3 INSTALL HEAD

STEP 3A INSTALLATION BY HAND

1. Hold the head assembly square to axis of the shell and slide it straight in until a

slight resistance is felt. Do not rotate the head assembly after insertion into the

vessel as this may cause the head seal to become detached.

Figure 3

3. Closing Vessel



- 101 -

Figure 3.

2. Using both hands, firmly push the head in as far as it will go. (A sharp, forceful

thrust may be necessary to enter the head seal into the vessel bore). When the

head is correctly positioned, approximately 1/2" of the locking ring groove will be

exposed.

STEP 3B INSTALLATION USING TOOL

1. Hold the head assembly square to axis of the shell and slide it straight in until a

slight resistance is felt.

2. Slide the head tool into the shell just behind the head. Do not engage threaded

rods.

3. Give a sharp, forceful thrust on the head tool to enter the head into the vessel

bore. Then push into the shell as far as it will go. When the head is correctly

positioned, approximately 1/2" of the locking ring groove will be exposed.

4. Remove the tool by pulling straight out. Do not rotate.(The tool can be obtained

from Advanced Structures, Inc.)

STEP 4 INSTALL INTERLOCK

1. Refer to Figure 1 for correct segment identification.

2. With the head assembly inserted into the shell, install segment B into the bottom

of the shell groove, with the stepped edge facing outwards.

3. Slide segment B counterclockwise making room to install segment C into the

bottom of the shell groove.

4. Slide segments B & C in the shell groove until the square ends meet at the 3 o'

clock position. Hold these in position while installing segment A (the key segment

3. Closing Vessel

1/2"

Bevel


8 Procedures for Replacing Elements

- 102 -

in the 9 o'clock position).

5. Rotate the installed locking ring set counterclockwise until the square ends of

segments B&C are in the 12 o'clock position. (This will prevent the segments

from falling out.) Locking ring segments must be installed with stepped edge

facing outwards.

6. Install the yellow securing ring with its ends flush. Align the three mounting holes

in the ring with their corresponding holes in the bearing plate. Insert the three

securing screws and turn them in about two turns.

7. Press the securing ring in until it seats securely on the bearing plate. Before

inserting securing screws. It is advisable to lightly coat the screw threads with

anti-seize compound, to ease later disassembly.

8. Tighten all three mounting screws until snug. Over-tightening may cause

disassembly problems!

9. Visually inspect locking ring set to ensure it is correctly positioned between shell

and bearing plate.

10. Verify that securing ring is fully seated and held in place by securing screws.

WARNINGINTERLOCKING COMPONENTS MUST BE

CORRECTLY INSTALLED. INCORRECT

ASSEMBLY OR INSTALLATION CAN

RESULT IN EXPLOSIVE HEAD FAILURE.

STEP 5 RECONNECT PERMEATE PIPING

1. Reconnect manifold piping to the vessel permeate port. Using Teflon tape on all

threaded connections will help ensure a leak-free assembly.

Do not tighten a component into thermoplastic permeate port more than one

turn past hand tight.

STEP 6 PRE-PRESSURIZATION CHECKS

It is vitally important that the following checks be carried out before any attempt is

made to pressurize the vessel.

3. Closing Vessel



- 103 -

HEAD ASSEMBLY

Verify the following at each end of the vessel.

1. Head assembly is in good condition, with no evidence of damage or corrosion.

2. Locking ring set is properly in place and yellow securing ring is snugly held in

place by the securing screws.

MEMBRANE ELEMENTS

Verify that…

1. Elements are installed in the vessel.

2. Element adapters are installed at each end of vessel.

3. Thrust ring installed at downstream end of vessel.

PIPING CONNECTIONS

1. Check all piping connections to ensure that they will provide a leak-free seal.

STEP 7 PRESSURIZATION

1. After following the above pre-pressurization checks, pressurize vessel in

accordance with the element specifications.

2. Vessels should be filled solwly to assist trapped air to escape.

2. Vessels should be pressurized slowly to avoide damage to membrane elements

and vessel components.

3. Closing Vessel


9. T r o u b le s h o o t in g


9-2 . Ins trum en t C a lib rat ion s

9-3 . Loca ting H ig h S alt Pass ag e

(Lo w S alt R ejec tio n )

9-4 . E lem e n t Ana lys is

9-5 . C le an in g Test for H ea v i ly

Fou l ed E lem e n ts

9-6 . Ana lyt ica l M etho d s for H ea v i ly

Fou l ed E lem e n ts

9-7 . C au s es of E le m en t Fai lu res and

C orrec tive M eas ures


9Tr oubleshootin g

- 105 -


Any RO s ys tem m a l fun c ti o n m a ni fe s ts i ts el f in a l o ss of s a lt rej e c ti o n , a l o ss of

pe rm ea te flow, an d an i nc re as e in d iff e re n ti a l p re ssur e, r es p ec ti ve l y or co l le c tivel y .

If o n e of the th re e pa ra m ete rs or c om b i ne d o n es d e v ia te s s l ow ly fro m the no rm al ize d

va l ue , it m a y i nd i ca te a no rm al fou l in g a nd sc a lin g w h ich c an be re m ov e d by p ro p er

c l e an i n g.

A fast or an i m m e d ia te p er fo rm an c e de c lin e i nd i ca tes a de fec t or m i sop e rati o n of the

sys te m . It is e ssen ti al in th i s ca s e tha t the pr op e r c o r rec ti ve m e as u re is tak e n as ea r ly

as p o ss ib l e b ec a us e a ny d e la y d ec rea s es the c h an c e of res to r in g the sys te m

pe r fo rm a n ce a n d also it m ay c re ate o the r pro b l em s.

A pre re q ui s ite for e ar ly d ete c ti o n of p ote nti a l p rob l e m s is c o ns is ten t re c or d ke e pi n g

an d p er fo rm an c e no rm al iza ti on , i n c lu di n g p rop e r c a lib rati o n of all i n s tru m e nts . It m a y

not be p os sib le w i th o ut ac cu ra te re ad i n gs to d e te c t a p ro bl e m e ar ly.

A fte r the pr ob l em ha s b ee n d e te c ted , the ne x t s tep is to l o ca l ize the pro b l em a n d to

id e nti fy the ca u se s of the pro b l em . T h i s ca n be do n e u s in g the d a ta of th e re co rd

ke e pi n g l og s h ee t or so m e a dd i t io n al on - lin e m e as u rem e n ts .

If th e d ata a re not su ff i cie nt to d e term in e the c au s es , o n e or m ore m e m br an e

el e m en ts m us t be tak e n out of th e s ys tem an d a na l yz e d by e i the r n o nd e s tru c tive or

de s tr uc ti ve m e th o ds .



9 Tr oubleshootin g

- 106 -

9 -2 . Ins tr um e nt C a lib r a tio n s

In s tru m e nt ca l ib ra tio n s are the first th i n g to ch e ck in a ny tro u bl e sh o oti n g o pe ra tio n

s i n ce w ro ng i n s tru m e nta tio n c a n m i ss a rea l d e c re a se in s a lt rej e c tio n or c a us e a

false a l a rm .

O n-lin e T D S M e te r s

A c cu ra c y of o n - lin e T D S m e te rs c an be v e r if i e d by m e as u r in g the fee d a nd pe rm ea te

w ate r T D S w i th a s ep a ra te ha n dh e ld m e te r .

T h e p er ce nt sa l t re j ec ti on c a l cu l ate d by the h an d he l d m ete r is c o m p a re d to the

va l ue s o b ta i n ed fro m the on - lin e m e te r fo l low i n g the i ns trm e n tati o n. Th e two v a l ve s

sh o ul d be re l ati ve l y c l o s e. If a di screp a nc y a r ises , re c al ib rate the o n- lin e m e ter fol low in g

the i n s tru c ti o n s of the m a nu fa c tu re r . B e fo re re c a lib rati n g, i ns p ec t the pro b e for the

po s sibl e a ccu m u l a ti o n of for ei g n m ate r ial th a t m ay i n ter fe re w i th the rea d i ng . A l so ,

m a k e su re th a t the p ro b es a re m o un te d pro p er ly a ccord i n g the m a n ufa c tu re r 's

sp e c if ica ti o n s . Im p ro pe r ly m o u nte d p rob e s m ay tra p air bu b b le s or h a ve in a de q ua te

w ate r c ircul a ti on th ro ug h the pro b e, w h ich m ay g i ve fa l se re a di n gs .

F lo w m e te r s

F l o w m e ter c al ibr ati o n is also im p o r tan t in an RO s ys tem , s i nc e the c or re c t

m e a su re m en t of the pe rm ea te a nd c o n ce n tr ate fl ow ra te s is c r it i ca l to th e s uc ce s sfu l

op e rati o n of the s ys te m . F or i ns ta nc e , the c o nc e ntra te fl ow rate l o w e r th a n the s ys tem

sp e c if ica ti o n m ay c re ate a s i tua ti o n for fo u l in g a nd s c al in g, a n d th e n orm a lize d

pe rm ea te flo w r ate is a c r it i cal p a ra m e te r to m o ni to r a ny tre n d in fo u l in g or m e m b ra n e

de ter io ra ti o n .

O n e m e th od of c a lib ra ti o n is to di re c t the w ate r fro m the flo w m e ter i nto a v e sse l an d

to m e a su re the tim e th at it ta k es to fill the k n ow n vo l um e . T h i s s h o ul d be re p ea te d

se v era l t i m e s for a c cu ra cy . If the va r iati o n is s l igh t, the va l u es ( th e v ol u m e d iv id e d by

the ti m e ) ca n be av e ra ge d . T he oth er m eth o d is to ca l ib ra te a flo w m e ter to a k no w n

fl o w m e te r which h as b e en c a l ib ra ted p re v io us ly.

P re s s ure S e ns ors

B e s id e s the fe ed pre s su re a nd th e p erm e ate p re ssu re, the m e as u rem e n t of the

di ff e ren ti a l pr es su re is n ec e s sa ry to m o n ito r th e b ui ld up of de p os its on the

m e m br an e s ur fa c e, or in the e l e m e n t fl ow c ha n ne l s. In a ccur ac y in the p re ssu re

rea d i ng s will ca u se to m i ss the tim in g for c l ea n i ng th e s ys tem .

2. In stru m e n t C a libra tion s


9Tr oubleshootin g

- 107 -

T h e a ccur ac y of m e c h an i ca l p re ssure g a u ge s s ho u l d be p e r io d i ca l ly v er if i ed u s in g a

ca l ib ra ted p re s su re ga u g e. E l e c tr on i c pre s su re s en s or s h a ve th e p ote n ti a l for gre a ter

ac cu ra cy . H o w e v er , th e y are su b je c t to s en s or drift a n d da m a ge re s ul t i n g from

v i b ra ti o n of the h i g h pr es su re p u m p s . To re du c e the eff e c ts of v i b rati o n, the s e ns o r

ca n be m ou n te d re m ote l y an d c on n ec te d to the h i gh pre s su re p ip i n g with a l en g th of

s ta i nl e ss s tee l or hi g h- pre s su re n yl o n tub i ng .

A ca l ib ra ti on p re s su re ga u g e m o u nte d w i th a q ui ck-co n ne c t f i t ti n g is u se ful for

ca l ib ra tin g p re ssur e se n so rs .

pH M e te r s a nd T e m pe r a tu re

T h e pH m e ters s h ou l d be reg u l ar ly c al ibra te d us ing b u ff er s ol u tio n s of a k no w n pH .

Sm a l l va r iati o n s in feed w a ter tem p er atu re do not s i g ni f i ca ntl y a ff ec t the pe rcen t s a lt

rej e c ti o n .

T h e te m p e ra tu re re a di n g s are s ti ll i m p or ta n t b ec a us e th ey a re u se d to d ete rm in e the

no rm al ize d pe rm ea te fl ow rate . T h e refo re the ac cu ra c y of the fee d w a te r te m p e ra tu re

rea d i ng s s ho u l d th e re fo re be re g ul a r ly v e r if i e d with an a c cu ra te th e rm o m e te r .

2. In s tr u m e n t C a libra tions


9 Tr oubleshootin g

- 108 -

9 -3 . Loc a t in g H ig h S a lt P a s s a ge (Lo w S a lt R e jec t io n )

A l os s in sa l t re j ec ti on m a y be un i form th ro ug h ou t the s ys te m or it co u ld be l i m ite d

to th e fro nt or to the ta i l en d of the sys te m . It c ou l d be a ge n e ral sys te m fa ilu re or it

co u ld be l i m ite d to on e or few i n d iv id u al ve s se l s. T h e l oc a tio n of the hi g h s al t p a s sa ge

ca n be i so l ate d by fo l lo w i n g the th re e s te p s :

● C h e ck the i nd i vid ua l v e ssel pe rm ea te T D S v a l ue s .

● P ro b e the s u sp e c te d v es se l .

● In di v id u al ly test e ac h e l em e nt in the v es sel .

A w e ll-d es ig n ed sys te m c on ta in s a s a m p l e p or t l oc a te d in the pe rm ea te s trea m fro m

ea c h ve s se l . C a re m us t be ta ke n d ur in g s am p l in g to av o i d m ix in g of the pe rm ea te

sa m pl e w ith p e rm e a te fr om o th er ve s se l s. All p e rm e a te s a m pl e s are the n te s te d for

th e i r co n c en tra ti o n of d isso l ve d s ol ids w ith a T D S m e te r . N oti ce th a t from on e a r ra y

to th e n ex t th e a ve ra ge pe rm ea te T D S u s ua l ly i n c re a se s , b ec a us e the se c on d a r ra y is

fed w i th the co n ce n trate fro m the first a r ra y. To d e te rm in e the sa l t p as sag e of all

ve s se ls fro m th ei r p e rm e a te T D S , the T D S of the fee d s trea m to e a ch a r ra y m u s t a l so

be m e as u re d. Th e s a lt pa s sa ge is th e ra ti o of the p e rm e a te T D S to the feed T D S .

If o n e p res su re v e ssel sh o w s a s i gn i f ica ntl y hi g h er pe rm ea te T D S th a n the oth er

ve s se ls of the sa m e a r ra y, th en th i s ve s se l s ho u l d be p ro b ed . P ro b i ng i n vo l ve s the

in s er ti o n of a p la s ti c tu b e (ap p ro x . 1/4 " for 8" m od u le ) in to the full l en g th of the

pe rm ea te tub e (s e e F ig u re 1).

W h i le the RO s ys te m is op e ra ti n g at n orm a l o pe ra tin g c on d i t io n s , w a te r is di ve r te d

fro m th e p erm e ate s tre am of th e v es se l in q u es ti on . A few m in u tes s h ou l d be al lo w ed

to r in s e out th e tu bi n g a nd a l lo w the RO s ys te m to e qu l ilib ra te. T he T D S of

th e pe rm ea te s am p l e from the tu bi n g c an th e n be m e a su re d with a h an d -he l d m e te r

an d the da ta be rec o rd ed . T h i s m e a su re m en t s h ou l d re fl e c t the T D S of the pe rm ea te

be i ng p ro d uc e d by the C S M e l em e n t at tha t l o ca ti on .

E le m e nt N o. # 1 # 2 # 3 # 4 # 5 # 6

TDS

Meter

FeedConcentrate

Permeate

Marks

F i g ur e 1. P rob i n g In d i vidu a l E l em e n ts .

3. L o ca ting H igh Salt P a s s a g e (L o w Salt R e jec tion )


9Tr oubleshootin g

- 109 -

9 -4 . E le m e nt A na ly s is

If h i g h sa l t p as sag e is fou n d o nl y in on e or s e ve ra l e l em e n ts in o n e or s ev e ra l

pre s su re v es se l s, th e n it is m os t l i ke l y th a t the el e m en t(s) co u l d ha v e m ec h an i ca l

da m ag e s su c h as pu n c tu re s on the m e m b ra n e su r fa ce , g l ue li n e fa ilu re, a c r ac ke d

ce n te r fo l d of the m e m b ra n e an d d am a g ed O -r ing s i n c lu di n g b r in e s ea l s. D am a g ed O-

r in g s a nd b r ine s e a ls c a n be v e r if i e d e as ily by v i su a l i ns p ec ti on of th e fa ile d el e m en ts .

D am a ge d m e m b ra n es a n d gl u e l i ne fa i lu re c a n be v i su a l ized o n l y by the a u top s y of

the e l em e nts . A l te rna ti ve l y, th os e p h ys ical da m ag e s c an be v er if i ed by a dye test al o n g

with the test for sa l t re j ec ti on a n d fl ux u s in g a s m al l test li n e c on ta in i n g m e th yl e ne

bl u e or rh od a m in e B. If the dye is d ete c te d v i su al ly or sp e c tro s co p ical ly in the

pe rm ea te, th i s pr ov e s the re is a c on s id e ra bl e d a m a g e in the m e m bra n e or g l ue l i n e.

Th e n the el e m en t c an be a u top s ie d to fin d the c au s es of the d am a ge s .

W h e n the re is a g e ne ra l s ys tem fa i lu re , a fro n t en d e l em e n t or a ta il e n d e le m e nt

sh o ul d be tak e n out of th e v es se l s for ex a m in a ti o n , d ep e nd i ng on w he re the pro b l em is

lo c ate d .

W h e n the p ro b le m c a nn o t be l oc a te d , an el e m en t fro m b o th en d s of the s ys tem

sh o ul d be tak e n. T yp i ca l fro n t en d p ro bl e m s a re d u e to fou l in g a nd typ i ca l tai l-en d

pro b l em s are co m in g fro m s ca l in g . V e ssel s /el e m en ts with the s e pr ob l em s u su a l ly s ho w

lo w pe rm ea te fl ow ra te an d s om e ti m e s a h i gh s a lt pa s sa g e fr om s e ve re fo u lin g an d

sc a lin g.

If th e m em b ra ne s a re d am a g ed by c he m ical s s uc h as c hl o r in e a n d co n ce n trate d aci

d, a hi g h sa l t p a ssag e a l on g w i th a h i gh e r th an no rm al p e rm e a te flo w ra te w ou l d o ccur

us u al ly in all the e l e m e n ts of the first a r ra y. If the a c cide n tal h i g h do s ag e of the

ch e m ica l s in to the sys te m is not co r re c te d i m m e d i ate l y, the m em b ra ne s of the se c on d

ar ra y w o u ld a l so be d a m a g ed .

4. E lem e n t A n a lys is


9 Tr oubleshootin g

- 110 -

9 -5 . C le a n in g T e s t fo r H e a v il y Fou le d E le m e nts

T h e re g ul a r c l ea n i ng p ro c ed u re, which is pa r t of th e s ys tem op e rati o n, us u al ly

res to res a d e c line in l e s s th a n 15 % of the pe rm ea te fl ow rate of the sys te m b a ck to the

no rm al v a l ue . H o w e v er , w h en th e d ec lin e in the fl o w rate is g re ate r th a n 15 % d u e to

m i ssing the c le a ni n g ti m e or an a c cide n ta l p re tre atm e nt u p se t, it is u s ua l ly d i ff icu lt to

rec o ve r the l o s t fl o w rate fu l ly by the n or m a l c l ea n in g . In thi s c as e , an e le m e nt fro m

the fro nt en d or ta i l en d , d ep e nd i n g on th e l oc a ti on of the pr ob l em , sh o ul d be tak e n out

for c l e an i n g tes ts u s ing m o re p ro pe r ch e m ica l s or m ore ag g re ss ive ch e m ical s . W h en

the c l ea n i ng test ha s p ro ve n e ff ec ti ve , the trea tm en t c a n be a p p l ie d to the w ho l e RO

sys te m .

H ow e v e r , c le a ni n g m a y n o t be su c ce ss ful w h en th e m em b ra ne is d a m ag e d, or w h en

the p er m e ate fl o w of th e e le m e nt is b e l ow 5 0% of s p ec if icati o n d ue to h e av y fou l in g

an d s ca l in g . T h en th e e l em e nt is au top s ie d to e xa m in e the m e m b ra n e su r fa ce , the

gl u e l in e , a nd the fo ul ing d e p os its by the m e tho d s s ho w n in the fol low in g s e c ti o n.

5. C lea n ing Test for H e a v i ly Fo u led E lem en ts


9Tr oubleshootin g

- 111 -

9 -6 . Ana ly tic a l M e tho d s fo r H e av il y Fou le d E le m e nts .

V is ua l Ins pec t io n

To a c e r tai n d e gr ee by a v isua l i n s pe c ti o n of the au top s ie d e l em e nt, the

ap p e ara n ce of th e fou l a nt will p ro v id e c l u e s to its n a ture a n d to the d i ff i cu l ty in i ts

rem o va l . L a rge fi lter m ed i a p ar ti cle s l ike a c ti va te d ca rb on will be e v ide n t to the s i g ht.

A bi o l og i ca l fo ul a n t will h a v e a di ff e re nt ap p e ara n ce th a n an in o rg an i c sc a le , a n d

sm e l l d i ff e ren tl y. O the r m e c h an i ca l p ro bl e m s w ith th e e le m e nt su c h as th e b rok e n g lu e

li n e a nd , the d am a ge d fe ed a n d p rod u c t w a te r ch a nn e l m ate r ial s c an be s po tted

v i su a l ly .

D is s o lu tio n in Ac id

If th e d ep o s its on the m e m bra n e ap p e ar to be c rys ta l lin e an d d i ssol ve in an a c id i c

so l uti o n (H C l ) of pH 3 to 4 with s om e g as ev o lu ti on (ca rb on di o x id e) , the n it is l i ke l y

th a t the de p os its a re c o ns iste d of ca rb on a tes s u ch as C a C O 3.

S u l fate s or s i lica will o nl y d isso l ve w i th d i ff icu lty in v ery l o w p H (e .g . p H 1) . If the sc a le

is s o l ub l e in 0 .1 M h yd ro fl u o r ic a c id (H F ) s ol u tio n , it is p os sib l y s i lica .

D y e T e s t

A dye s uc h as m eth yl e ne b l u e or rh od a m in e B c a n be a d de d in a c o nc e ntra ti on of 0.

00 1 to 0 .0 0 5% to the fe ed w ate r of th e test e l e m e n t on an i n d i vid u a l el e m en t test

s ta nd . T h e dye will pe rm ea te the m e m b ran e th ro ug h a rea s of d eg ra d ati o n an d

m e c ha n i ca l l ea k s in th e e l em e nt. Th e dye in th e p erm e ate w a te r c a n be d e tec ted

v i su a l ly or m e as u re d us ing a s p ec tro ph o tom e ter .

A fte r the dye tes t, the el e m en t c a n be au top s ie d to v i su al ly in s pe c t the sp e c if i c

lo c ati o n of the dye pa s sa ge . D a m a ge d a re as m a y pick up m o re dye tha n u n aff e c te d

are a s . C h em ica l a tta c k by c hl o r in e or h i g h do s ag e a c id i n du c ed h ydro l ys is of the

m e m br an e will te n d to re su l t in u ni fo rm ab s or pti o n of the d ye .

O ptic a l M ic ros c op y

W h e n the v i su a l i ns p e c ti o n of the a u to p s ie d el e m en t c a n not re v e al e n ou g h

in fo rm a ti on a b ou t the n atu re of fou l a nts , a h i gh p o w e r l i g ht m i c ro s co pe ca n te ll if a

fo u l an t is b i ol o gi cal or i no rg a ni c s ca l e. It ca n a l so p ro v id e in fo rm a ti on a b ou t th e

c rys ta l lin e s tru c tu re of a s c al e fo rm a ti o n. U s in g po l a r ized l i g ht, the m i cro scop e c a n te l l

6. Analytical Methods for Heavily Fouled Elements.


9 Troubleshooting

- 112 -

the d i ff ere n ce be tw e en c a l cium s u l fa te a nd ca l cium ca rb on a te sc a l es , s i nc e c al c iu m

su l fate c rys ta l ha s m o re tha n o n e refra c tive i nd e x to give a u ni q ue ap p e ara n ce .

S o m eti m es it c a n be rec o g ni ze d u n de r th e m i cro scop e th at a la yer of sc a l e s i tt i ng on

the m e m b ra n e su r fa ce is c o ve re d by a n oth e r l a ye r of o rg an i c an d b i ol o g ical fou l an ts .

In th i s ca s e, it m a y be a m os t e ff ec ti ve c l e an i ng to r em o ve th e o rg an i c an d b io l o gi cal

fo u l an ts first w i th an al kal ine c l e an i n g so l u ti o n , p r io r to atte m pti n g to rem o v e the s c al e

with an a c id i c so l uti o n.

Fourier Transform Infrared Spectroscopy (FTIR)

FTIR can provide additional information about foulants on the membrane surface.

When there is a thick layer of foulants on the membrane, it is preferred to scrape off

a sample of foulants from the membrane surface. The sample is dried and FTIR of

the dried sample is run. If a suitable size of sample cannot be collected from the

membrane surface, an FTIR analysis may be performed directly on the fouled

membrane using a technique called attenuated total reflection (ATR). ATR will give

FTIR spectrum corresponding to the foulants after it subtracts FTIR spectrum of a

fresh membrane from that of the fouled membrane. Sometimes, the substraction

process does not work well to result in a mixture of peaks corresponding to both the

foulants and the membrane.

Organic and inorganic compounds have their own specific FTIR peaks which can in

turn be used to identify the compounds in a mixture. From the peak intensity,

semiquantitative analysis is also possible. If the FTIR spectrum shows peaks due to Si-

O-Si, CO 3, and SO4, then it indicates there are silica, calcium carbonate, and sulfates,

respectively in the foulants. If the spectrum exhibits peaks corresponding to C-H, -CO

-, C-C, and C-N, then it indicates the foulant is organic or biological. Peaks due to C-

H, -CO-, C-C, and phenol groups strongly suggest that the foulant is consisted of

humic acid.

Scanning Electron Microscopy (SEM)

SEM can distinguish much smaller objects than the optical microscope. So SEM

gives clear photographs of particles as small as 0.1㎛ to identify small crystalline and

amorphous inorganic scaling matter, and also the cell structure of microorganisms.

SEM could be very helpful in analyzing the foulants in more detail.

Energy Dispersive X-ray (EDX)

During the process of SEM, X-ray radiation is emitted from the sample due to the



9Troubleshooting

- 113 -

electron bombardment. The X-ray is low energy and characteristic of the elements in

the sample. Thus EDX can identify elements in the sample and even offer a semi-

quantitative analysis of the sample. It can detect very small amount of inorganic

elements in the sample and also identify carbon, nitrogen and oxygen, though less

sensitively. It works the best for analyzing an inorganic scale, but also is useful for

organic sample analysis.

The method can also furnish evidence of halogen damage to the membrane due to

chlorine oxidation by detecting the presence of chlorine attached chemically to the

polyamide.



9 Troubleshooting

- 114 -

9-7. Causes of Element Failures and Corrective Measures

9-7-1. High Salt Passage and High Permeate Flow

Membrane Oxidation

A combination of a high salt passage (low salt rejection) and a high permeate flow

is a typical symptom of the damaged membrane oxidized by oxidizing chemicals

including chlorine, bromine, and ozone. Other oxidizing chemicals such as peracetic

acid, hydrogen peroxide, and N-chloro compounds are less aggressive, but still can

damage the membranes when they are present in excessive amount or coexist with

transition metals. In the case of chlorine and bromine, a neutral to alkaline pH favors

the attack to the membrane. At the early stage of the oxidation, the front end elements

are usually more affected than the rest.

The oxidation damage can be made visible by a dye test on the element or on

membrane coupons after autopsy of the element. All damaged elements must be

replaced.

Leak

A leak of feed or concentrate to the permeate through a mechanical damage of the

element or of the permeate tubing can cause high salt passage and high permeate

flow. The contribution of the leak to the permeate flow may depend on the magnitude

of the damage usually caused by high pressure and vibration. The types of the

damages include leaking O-rings, cracked tubes, telescoping, punctured membranes,

and centerfold cracking.

9-7-2. High Salt Passage and Normal Permeate Flow

Leaking O-ring

Leaking O-rings can be detected by the probing technique (Section 9-3). Inspect O-

rings of couplers, adaptors, and end plugs for correct installation and aging condition.

Replace old and damaged O-rings. O-rings may leak after exposure to certain

chemicals, or to mechanical stress, e.g. element movement caused by water hammer.

Sometimes, they have been improperly installed or moved out of their proper location

during element loading.

7. Causes of Element Failures and Corrective Measures


9Troubleshooting

- 115 -

Telescoping

Telescoping is caused by excessive pressure drop from feed to concentrate. Eight

inch elements are more critical because of their greater feed side area. Make sure that

a thrust ring is used with eight inch elements to support the elements' outer diameter

s. Elements with smaller diameter are supported by their permeate tubing. Severe

telescoping can rupture the glue line or the membrane itself. Telescoping damage can

be identified by probing(section 9-3). The operating conditions leading to excessive

pressure drop are detailed in the section of High Differential Pressure. For an exampl

e, when the pressure pump is started before a drained system has time to fill, the front

end elements will be exposed to higher than normal water velocities. This can hammer

the elements to telescoping which can be prevented by opening the throttling valve

slowly.

Membrane Surface Abrasion

The front-end elements are typically most affected by crystalline or sharp-edged

metallic suspended solids in the feed water. Check the incoming water for such

particles. Microscopic inspection of the front-end elements are typically most affected by

crystalline or sharp-edged metallic suspended solids in the feed water. Check the

incoming water for such particles. Microscopic inspection of the membrane surface will

also reveal the damage. No corrective action is possible. The pretreatment must be

changed to cope with this problem. Ensure that no particles are released from the high

pressure pump.

Permeate Backpressure

When the permeate pressure exceeds the feed/concentrate pressure by more than 0.3

bar (5 PSI) at any time, the membrane may tear. The damage can be identified by

probing. Upon autopsy of the damaged element, the outer membrane typically shows

creases parallel to the permeate tube, usually close to the outer glue line. The rupture

of the membrane occurs mostly in the edges between the feed-sided glue line, the

outer glue line, and the concentrate-sided glue line.

Centerfold Cracking

The regular process for making a spiral wound element requires folding a leaf of

membrane sheet in the center (centerfold). The creased (folded) membrane can break

at the centerfold under certain conditions. Then the salt passage increases with or



9 Troubleshooting

- 116 -

without an increase in the permeate flow. Centerfold cracking may be caused by :

Hydraulic shock during start-up, e.g. by air in the system ;

Too fast pressure increase ;

Increased shear stress ;

Abrasion by scaling and fouling;

Permeate backpressure.

Centerfold cracking typically occurs only after one year or more of improper operatio

n, and only at systems with a high start and stop frequency.

9-7-3. High Salt Passage and Low Permeate Flow

High salt passage combined with low permeate flow is the most commonly occurring

system failure, usually induced by colloidal fouling, metal oxide fouling and scaling.

Colloidal Fouling

Colloidal fouling occurs predominantly in the first array. The problem can be more

easily located when permeate flow meters have been installed in each array separately.

SDI should be checked more frequently to identify the pretreatment upset.

Inspect SDI filters and cartridge filters for deposits. Clean the elements according to

the cleaning procedure, and correct the pretreatment process accordingly.

Metal Oxide Fouling

Metal oxide fouling also occurs predominantly in the first array. Check feed water for

levels of iron and aluminum. Check the materials of construction upstream of the

membranes. Improper construction materials may under go corrosion to shed iron in

the feed water. Inspect SDI filters and cartridge filters for deposits. Clean the

membranes with an acidic cleaning solution. Correct the pretreatment and / or material

selection.

Scaling

Scaling will involve deposits starting on the last array, and then gradually moving to

the upstream arrays. Analyze the concentrate for levels of calcium, barium, strontium,

sulfate, fluoride, silicate, pH and LSI (S & DSI for sea water). Try to calculate the mass

balance for those salts, analyzing also feed water and permeate. Scaling occurs slowly

because of the low concentrations involved except CaCO3.



9Troubleshooting

- 117 -

The crystalline structure of the deposits can be observed under the microscope. The

type of scaling is identified by a chemical analysis or X-ray analysis. Cleaning with acid

and/or an alkaline EDTA solution with subsequent analysis of the spent solution may

also help to identify the type of scalant. In the case of carbonate scaling, adjust the

pH of the pretreatment. For the other salts, either use an appropriate scale inhibitor or

other suitable pretreatment techniques, or lower the recovery. Make sure that the

design recovery is not exceeded.

9-7-4. Low Permeate Flow and Normal Salt Passage

Biofouling

Biofouling of the membrane occurs predominantly at the front end of the system and

affects permeate flow, feed flow, feed pressure, differential, and salt passage in the

way as shown below :

Permeate flow decreases when operated at constant feed pressure and recovery.

Feed flow decreases when operated at constant feed pressure and

recovery.

Feed pressure has to be increased if the permeate flow is maintained at constant

recovery. Increasing the feed pressure will invoke a worse situation, since it

increases the fouling, making it more diff icult to clean later.

Differential pressure increases sharply when the bacterial fouling is massive or

when it is combined with silt fouling.

Since pressure drop across the pressure vessels is a sensitive indicator of foulin

g, installing pressure monitoring devices is strongly recommended for each array.

Salt passage is normal at the beginning, but may increase when fouling becomes

massive.

High counts of microorganisms in water samples from the feed, concentrate, or

permeate stream indicate the beginning or the presence of biofouling.

Corrective measures require disinfection of the whole system including

pretreatment as described in section 7, and optimization of the pretreatment

system to cope with the microorganism growth in the raw water.

An incomplete cleaning and disinfection will result in rapid regrowth of the

microorganisms.

Aged Preservation Solution



9 Troubleshooting

- 118 -

Elements of RO systems preserved in a bisulfite solution can also become

biologically fouled, if the preservation solution is too old, too warm, or oxidized by

oxygen. An alkaline cleaning usually helps to restore the permeate flow.

Incomplete Wetting

Elements that have been allowed to dry out, usually give a very low permeate flow

with a normal salt passage. The lost permeate flow may be recovered by soaking the

elements in a 50:50 mixture of alcohol and water for one or two hours followed by

soaking in water.

9-7-5. Low Permeate Flow and Low Salt Passage

Compaction

Membrane compaction usually results in low permeate flow and low permeate salt

passage (increased salt rejection). CSM membrane does not undergo compaction at

normal operation, but significant compaction may occur at high feed pressure (see

section 5 of System Design), high water temperature (> 45℃) and water hammer.

The water hammer can occur when the high pressure pump starts with air in the

system and full opening of the throttle valve.

The compaction can induce intrusions of the membrane into the permeate channel

spacer fabric, which are visible. Thus, the permeate flow is not only restricted by the

compaction of the polyamide or the polysulfone layer, but also by the reduced cross-

section of the permeate spacer that is available for permeate flow.

Organic Fouling

Organic matter in the feed water can deposit on the membrane surface to cause flux

loss usually in the first array. The deposited organic layer could act as an additional

barrier for dissolved solutes,or plug pinholes of the membrane, to increase salt

rejection.

Organics with hydrophobic characters or cationic groups can produce such an effect.

Examples are hydrocarbons, cationic polyelectrolytes, cationic surfactants, nonionic

surfactants, and humic acids.

Analyse the incoming water for oil and organic matter, and check the SDI filter and

the cartr idge filter for organic deposits. Conduct SDI and TOC measurements on a

more frequent basis. Improve the pretreatment accordingly.



9Troubleshooting

- 119 -

An oil fouling can be removed with an alkaline cleaning agent, for example NaOH

(pH 12) or Henkel P3-ultrasil 10. Cationic polyelectrolytes may be cleaned off at an

acidic pH, if it is not a precipitation product with other compounds, e.g., antiscalants.

Cleaning with alcohol has also proven effective in removing adsorbed organic films.

9-7-6 High Differential Pressure

High differential pressure, also called pressure drop from feed to concentrate,

generates a high force pushing the feed side of the element in flow direction. This

force impacts on the permeate tubes and the fiberglass shells of the elements in the

same vessel. The stress on the last element in the vessel is the highest since it has

to bear the sum of the forces from the pressure drops of all prior elements.

The upper limit of the diff erential pressure per multi-element vessel is 4.1 bar ( 60

PSI), per single element 1.4 bar (20 PSI). When these limits are exceeded, even for a

very short time, the elements might be mechanically damaged to result in telescoping

and/or breaking the fiberglass shell. This type of damage may not disturb the

membrane performance temporarily, but eventually cause flux loss or high salt passag

e.

An increase in differential pressure at constant flow rates usually arises from the

accumulation of debris, foulants and scale within the element flow channels (feed

spacer). It usually decreases the permeate flow. An excessive increase in differential

pressure can occur from operating mistakes such as exceeding the recommended feed

flow (section 5, System Design), and building up the feed pressure too fast during

start-up (water hammer).

Water hammer, a hydraulic shock to the membrane element, can also happen when

the system is started up before all air has been flushed out. This could be the case at

initial start-up or at restart-ups, after the system has been allowed to drain. Ensure

that the pressure vessels are not under vacuum when the plant is shut down (e.g. by

installation of a vacuum breaker). In starting up a partially empty RO system, the

pump may behave as if it had little or no backpressure. It will suck water at great

velocities, thus hammering the elements. Also the high pressure pump can be

damaged by cavitation.

The feed-to-concentrate diff erential pressure is a measure of the resistance to the

hydraulic flow of water through the system. It is very dependent on the flowrates

through the element flow channels, and on the water temperature. It is therefore

suggested that the permeate and concentrate flowrates be maintained as constant as

possible in order to notice and monitor any element plugging that is causing an



9 Troubleshooting

- 120 -

increase in differential pressure.

The knowledge of the extent and the location of the differential pressure increase

provides a valuable tool to identify the cause(s) of a problem. Therefore it is useful to

monitor the differential pressure across each array as well as the overall feed-to-

concentrate differential pressure. Some of the common causes and prevention of high

differential pressure are discussed below.

Failure of Cartridge Fil ters

When cartridge filters are loosely installed in the housing or connected without using

interconnectors, they can shed debris and particles to block the flow channels in the

front end elements. Sometimes cartridge filters deteriorates while in operation due to

hydraulic shock or the presence of incompatible materials.

Media Filter Breakthrough

Fines from multimedia,carbon,weak acid cation exchange resin,or diatomaceous earth

filters may set loose and enter into the RO feedwater. Sometimes some of the

coagulated colloids can pass through the channeling of the filters when the filters are

not regularly backwashed to result in caking and channeling of the filters. The

channeling could occur in very old filters. Cartridge filters should catch most of the

larger particles. Smaller particles can pass through a five micron nominally rated

cartridge filter to plug the lead elements.

Chemical cleaning is diff icult. It can be tried to rinse out the deposits with detergent

s. Separate single-element cleaning is recommended to avoid the transport of removed

particles into other elements. Diatomaceous earth filters should be taken off-line when

such problems are encountered. Soft carbons made from coal should be replaced by

coconut-shell-based carbons with a hardness rating of 95 or better. New media should

be suff iciently backwashed to remove fines before the bed is put into service.

Pump Impeller Deterioration

Most of the multistage centrifugal pumps employ at least one plastic impeller. When

a pump problem such as misalignment of the pump shaft develops, the impellers have

been known to deteriorate and throw off small plastic shavings. The shavings can enter

and physically plug the lead-end RO elements.

Many high pressure pumps are equipped with an optional discharge screen. This

screen will catch most of the shavings. Pump discharge screens should be checked

regularly for shavings or other debris. The screen may be cleaned or replaced. As part

of a routine maintenance schedule, monitoring the discharge pressure of the pumps



9Troubleshooting

- 121 -

prior to any control valves is suggested. If not enough pressure, it may be deterioratin

g.

Scaling

Scaling can cause the tail-end differential pressure to increase. Make sure that scale

control is properly taken into account (see Section 4-3), and clean the membranes with

the appropriate chemicals (see Section 7-5). Ensure that the designed system recovery

will not be exceeded.

Brine Seal Damage

Brine seals can be damaged or turned over during installation or due to hydraulic

impacts. A certain amount of feedwater will flow through the chasm in the damaged

seals to bypass around the element, resulting in less flow and velocity through the

element. This will cause to exceed the limit for maximum element recovery to increase

the potential for fouling and scaling.

As a fouled element in the multi-element pressure vessels becomes more plugged,

there is a greater chance for the downstream elements to become fouled due to

insufficient concentration flow rates within that vessel.

The brine seal damage causing an increase in differential pressure could happen

randomly in any pressure vessel. Early detection for the increase in diff erential

pressure is important for an easy correction of system malfunctions.

Biological Fouling

Biological fouling is typically associated with marked increase of the differential

pressure at the lead end of RO system. Biofilms are gelatinous and quite thick, thus

creating a high flow resistance.

Corrective measures have been described in Section 9-7-4. It is important to

frequently clean out the microbial growth and disinfect the system. It is also suggested

that bacteria samples are taken and analyzed on a regular basis from the feed,

permeate, and concentrate streams.

Precipitated Antiscalants

When polymeric organic anti-scalants come into contact with multivalent cations like

aluminium, or with residual cationic polymeric flocculants which can heavily foul the



9 Troubleshooting

- 122 -

lead elements.

Repeated applications of an alkaline EDTA solution may clean the fouled elements

with some difficulties.



1 0 . A p p e n d ix

10-1. P erfo rma n ce Warran ty

10-2. C hec k l is t P r io r In it ia l System

O pe ratio n

10-3. R eve rse O sm os is O peration Log

10-4. C on ve rs io n of C on ce n tration

U ni ts o f Ion i c S pec ie s

10-5. S tan da rd S ea Water C om p o s ition

10-6. S o lu b il i ty P rod u c ts o f S pa ring l y

So lu b le S alts

10-7. C on du c tiv ity of Acid S ol u tio n s

10-8. C on du c tiv ity of Alka li S o lu tio n s

10-9. C on du c tiv ity of S alt So lu tion s

10-10 . C on d u ctiv i ty o f Ion s

10-11 . S pe ific C ond u c tan c e o f S od iu m

C hlo rid e

10-12 . Ion i zation o f C arbon D io x id e

S o lu tion s as Fun ctio n o f the pH

10-13 . O sm o tic P ress u re of S od iu m

C hlo rid e

10-14 . O sm o tic P ress u re of S o lu tio n s

10-15 . C S M P erfo rma n ce

10-16 . Tem p e ra tu re C orrec tio n Fac to r


10A p pendix

- 124 -

1 0- 1 . Perf orma nce W arranty

Saehan Industries, Inc. guarantees the performance of its reverse osmosis elements

only when the elements are used in the system operated under the following

conditions.

1. Conditions for Handling CSM Elements in general.

a. Keep elements moist at all times after initial wetting

b. Permeate water obtained from first hour of operation should be discarded.

c. CSM elements should be immersed in a protective solution during storage,

shipping or system shutdowns to prevent biological growth and freeze damage.

The standard storage solution contains 18-20 weight percent propylene glycol

and 1 weight percent sodium metabisulfite (food grade). For short term storage

of one week, a 1 weight percent sodium metabisulfite solution is adequate for

inhibiting biological growth.

d. Elements must be in use for at least six hours before formaldehyde is used as

a biocide. The elements exposed to formaldehyde before being in use for six

hours may lose flux.

e. The customer is fully responsible for the effects of incompatible chemicals on

elements. Their use will void the element limited warranty.

2. Conditions for System Design and Equipment

a. The system array, recovery and instrumentation and the design parameters and

components of the system in which the element(s) are employed shall be

consistent with sound engineering practice. Saehan reserves the right to review

system design.

b. Recovery ratio shall be consistent with concentration of sparingly soluble salts.

c. There shall be no membrane scaling caused by failure of the chemical dosing

system (e.g., Ca, Ba, or Sr salts)

d. Adequate provisions against microbiological contamination shall be incorporated

into the system design, as well as into all operating and maintenance

procedures.

1. Performance Warranty


10 Ap pendix

- 125 -

3. Conditions for Feed Water

a. Feedwater SDI (15 min., 30 psi) shall be less than 5.0.

b. Feedwater shall contain no colloidal sulphur.

c. There shall be no membrane fouling by colloidal or precipitated solids.

d. The brine-soluble silica shall be less than 150 ㎎/ℓat 25℃.

4. Conditions for Operating Parameters

a. The element(s) shall not be exposed to pressure greater than 1000 psi for

seawater elements, 600 psi for brackish water elements, and 300 psi for

tapwater elements.

b. Backpressure (where permeate static pressure exceeds reject static pressure)

shall not exceed 5 psi at any time.

c. The element(s) shall be operationally protected against hydraulic shock loading

(water hammer).

d. Feedwater temperature shall be less than 45℃.

5. Conditions for Chemicals on Elements

a. Antiscalants or sequestrant used shall be standard or approved equivalent.

b. During continuous operation the pH shall be no less than 2.0 nor greater than 1

1.0. If pH adjustment is required, use H2SO4 or approved equivalent.

c. There shall be no membrane damage caused by chemical compounds (e.g.,

surfactants, solvents, soluble oils, free oils, lipids, and high molecular weight

natural polymers.)

d. Feedwater shall contain no ozone, permanganate, chlorine or other strong

oxidizing agents.

e. Neither nonionic nor cationic surfactants, nor any other chemical not approved

by Saehan, should be used for membrane cleaning or shall come in contact

with elements.

6. Conditions for Cleaning Elements

a. Cleaning shall be initiated at 10% to 15% normalized product flow decline.

b. The element(s) shall not be exposed during cleaning, or in shutdown periods, to

a pH less than 2 or greater than 12.

c. Cleaning chemicals used shall be standard or approved equivalent.

1. Performance Wrranty


10A p pendix

- 126 -

1 0- 2 . Checklist P rio r to In it ia l Sys tem O p eratio n

1. All piping and equipment is compatible with designed pressure.

2. All piping and equipment is compatible with designed pH range (cleaning).

3. All piping and equipment is protected against galvanic corrosion.

4. Corrosion resistant materials of construction are used for all equipment including

piping and wetted parts of pumps.

5. Pressure vessels are properly piped both for operation and cleaning mode.

6. Pressure vessels are secured to the rack or frame.

7. Check the head assemblies of pressure vessels are properly installed and free of

corrosion.

8. Planned instrumentation allows proper operation and monitoring of the

pretreatment and RO system.

9. Instrumentation is calibrated.

10. Pressure relief protection is installed and correctly set..

11. Provisions exist for preventing the product pressure from exceeding the feed/

brine pressure more than 0.3 bar (5PSI) at any time.

12. Interlocks, time delay relays and alarms are properly set.

13. Provisions exist for sampling permeate from individual modules.

14. Provisions exist for sampling feed, permeate and reject streams from each

array and the total plant permeate stream.

15. Membranes are protected from temperature extremes(freezing, direct sunlight,

heater exhaust.

16. Pumps are ready for operation (lubricated, proper rotation).

17. Fittings are tight.

18. Permeate line is open.

19. Permeate flow is directed to drain.

20. Reject flow control valve is in open position.

21. Feed flow valve is throttled and/or pump bypass valve is partly open to limit

feed flow to less than 50% of operating feed flow.

22. Media filters are backwashed and rinsed.

2. Checklist Prior to Initial System Operation


10 Ap pendix

- 127 -

23. New/clean cartridge filter is installed directly upstream of the high pressure

pump.

24. Feedline, including RO feed manifold, is purged and flushed, before pressure

vessels are connected.

25. Chemical addition points are properly located.

26. Check valves are properly installed in chemical addition lines.

27. Provisions exist for proper mixing of chemicals in the feed stream.

28. Provisions exist for preventing the RO system from operating when the dosage

pumps are shut down.

29. Provisions exist for preventing the dosage pumps from operating when the RO

system is shut down.

30. If chlorine is used, provisions exist to ensure complete chlorine removal prior

to the membranes.

2. Checklist Prior to Initial System Operation


10A p pendix

- 128 -

1 0- 3 . R eve rs e Os mo sis O peratio n Log

Description Design / / / / /

TimeDate

Operating hours

Feed

Free Chlorine(mg/ℓ)

SDI

Turbidity(NTU)

Temperature(℃)

pH

Raw Water

Feed

Concentrate

Permeate

Pressure

(psig, bar, or

㎏ /㎠)

Cartridge In

Cartridge Out

Stage 1 Feed

Stage 2 Feed

Concentrate

Permeate

Differential Pressure

(psig, bar, or ㎏/㎠ )

Cartridge

Stage 1

Stage 2

Flow

(gpm,

or M3/hr)

Feed

Permeate

Concentrate

Recovery(%)

Conductivity

(㎲/㎝)

or TDS

(mg/l)

Feed

Permeate

Concentrate

Salt Rejection(%)

Remarks

3. Reverse Osmosis Operation Log


10 Appendix

- 129 -

1 0- 4 . Co nvers io n o f C oncentratio n Units o f Io n ic Sp ecies

Compound Fomulalonic

Weight

Equiv.

Weight

Conversion to

g CaCO3/ℓ eq/ℓ

Neutral

A m m onia NH 3 1 7. 0 17.0 2. 94 0 . 058 8

C a rbon dioxide CO 2 4 4. 0 44.0 1. 14 0 . 022 7

Silica S iO 2 6 0. 0 60.0 0. 83 0 . 016 7

Positive lons

A lum inu m Al + + + 2 7. 0 9.0 5. 56 0. 11 1

A m m onium NH 4+ 1 8. 0 18.0 2. 78 0 . 055 6

B a rium Ba + + 13 7. 4 68.7 0. 73 0 . 014 6

C a lcium Ca + + 4 0. 1 20.0 2. 50 0 . 050 0

C o pp er Cu + + 6 3. 6 31.8 1. 57 0 . 031 4

F e rro us lron Fe + + 5 5. 8 27.9 1. 79 0 . 035 8

F e rr iclron Fe + + + 5 5. 8 18.6 2. 69 0 . 053 8

H y drog en H + 1.0 1.0 50.0 1 . 000 0

M a gn es ium Mg + + 2 4. 3 12.2 4. 10 0 . 082 0

M a ng ane s e Mn + + 5 4. 9 27.5 1. 82 0 . 036 4

P o ta s sium K + 3 9. 1 39.1 1. 28 0 . 025 6

S o diu m Na + 2 3. 0 23.0 2. 18 0 . 043 5

Negative lons

B icarbo na te H C O 3- 6 1. 0 61.0 0. 82 0 . 016 4

C a rbon at e CO 3- - 6 0. 0 30.0 1. 67 0 . 333 3

C h lo r id e Cl - 3 5. 5 35.5 1. 41 0 . 028 2

F luor ide F - 1 9. 0 19.0 2. 63 0 . 052 6

H y droxide OH - 1 7. 0 17.0 2. 94 0 . 058 8

lo ide l - 12 6. 9 12 6. 9 0. 39 0 . 007 9

N i tra t e NO 3- 6 2. 0 62.0 0. 81 0 . 016 1

P h os p ha te ( tr i -bas ic) PO 4- - - 9 5. 0 31.7 1. 58 0 . 031 5

P h os p ha te (d i -b as ic) H PO 4- - 9 6. 0 48.0 1. 04 0 . 020 8

P h os p ha te (m on o-ba s ic) H 2 PO 4- 9 7. 0 97.0 0. 52 0 . 010 3

S u lfa t e SO 4- - 9 6. 1 48.0 1. 04 0 . 020 8

B isulfa t e H S O 4- 9 7. 1 97.1 0. 52 0 . 010 3

Sulfite SO 3- - 8 0. 1 40.0 1. 25 0 . 025 0

Bisulfite H S O 3- 8 1. 1 81.1 0. 62 0 . 012 3

S u lfide S - - 3 2. 1 16.0 3. 13 0 . 062 5

Silica S iO 2 6 0. 0 60.0 0. 83 0 . 016 7

4. C o nv e rs ion of C o n c e n tr a tion U n its of Io n ic S p e c ies


10A p pendix

- 130 -

1 0- 5 . Standard S ea W ate r C ompositio n

In g re d ie n tsC o n ce n tra t io n

(㎎ / l ite r)

Ion

Barium 0.05

Bicarbonate 152

Bromide 65

Calcium 410

Chloride 19700

Floride 1.4

Iron < 0.02

Magnesium 1310

Manganese <0.01

Nitrate < 0.7

Potassium 390

Silica 0.04~8

Sodium 10900

Strontium 13

Sulfate 2740

TDS 35000 ㎎/liter

pH 8.1

5. S ta n d a rd S e a W a te r C o m p o s i tion


10 Ap pendix

- 131 -

1 0- 6 . So lub ility P roducts o f Sparing ly So lub le Salts

(at ze ro io n ic s treng th)

Material Chemical FormulaTemp'

(℃)Solubility Product

Aluminum Hydroxide Al(OH)3 20 1.9 x 10- 33

Barium Carbonate BaCO3 16 7 x 10-9

Barium Sulfate BaSO4 25 1.08 x 10- 10

Calcium Carbonate CaCO3 25 8.7 x 10- 9

Calcium Floride CaF2 26 3.95 x 10- 11

Calcium Sulfate CaSO4 10 6.1 x 10- 5

Cupric Sulfide CuS 18 8.5 x 10- 45

Ferric Hydroxide Fe(OH)3 18 1.1 x 10- 36

Ferrous Hydroxide Fe(OH)2 18 1.64 x 10- 14

Lead Carbonate PbCO3 18 3.3 x 10- 14

Lead Fluride PbF2 18 3.2 x 10- 5

Lead Sulfate PbSO4 18 1.06 x 10-8

Magnesium

AmmoniumPhosphateMgNH4PO4 25 2.5 x 10

- 13

Magnesium Carbonate MgCO3 12 2.6 x 10- 5

Magnesium Hydroxide Mg(OH)2 18 1.2 x 10- 11

Manganese Hydroxide Mn(OH)2 18 4 x 10- 14

Nickel Sulfide NiS 18 1.4 x 10- 24

Strontium Carbonate SrCO3 25 1.6 x 10- 9

Strontium Sulfate SrSO4 17.4 2.81 x 10-7

Zinc Hydroxide Zn(OH)2 20 1.8 x 10- 14

6. S o lub ility P ro d u cts of S p a ring ly S o lub le S a lts


10A p pendix

- 132 -

1 0 -7 . C on d u c tiv it y o f Ac id s o lu tio n s

1 E - 3 0 . 0 1 0 . 1 1 10 1 0 0

1 0 0

1 5 0

2 0 0

2 5 0

3 0 0

3 5 0

4 0 0

4 5 0

Conduct iv i ty o f So lu t ions, Ac ids

H C l

H N O 3

H 2S O 4

H3P O

4

Co

nd

uct

ivit

y (µ

S/c

m)

C o n c e n t r a t i o n i n m e q / l

7. C o n d u c tivi ty of A c id s o lution s


10 Ap pendix

- 133 -

1 0 -8 . C on d u c tiv it y o f Alk a li S o lu tio n

1 E - 3 0 . 0 1 0 . 1 1 10 1 0 0

0

50

1 0 0

1 5 0

2 0 0

2 5 0

3 0 0

Conduct iv i ty o f Solu t ions, A lka l ies

N a O H

K O H

N H 4O H

Co

nd

uct

ivit

y (µ

S/c

m)


8. Conductivity of Alkali Solution


10A p pendix

- 134 -

1 0 -9 . C on d u c tiv it y o f S a lt S o lu tio n

1 E - 3 0 . 0 1 0 . 1 1 1 0 1 0 0

8 0

9 0

1 0 0

1 1 0

1 2 0

1 3 0

1 4 0

1 5 0

C o n d u c t i v i t y o f S o l u t i o n s , S a l t s

K C l

N a C l

N a2S O

4

N a2C O

3

N a H C O 3

Co

nd

uc

tiv

ity

(µS

/cm

)


0 2 4 6 8 1 0

0

4

8

1 2

1 6

2 0

C o n d u c t i v i t y o f I o n i c S o l u t i o n ( 2 5oC )

H2S O

4

N a O H

H C l

K O H

N H3

N a C l

C O2

Co

nd

uc

tiv

ity

, µS

/cm

C o n c e n t r a t i o n , g / m3( m g / l )

9. Conductivity of Salt Solutions


10 Ap pendix

- 135 -

1 0 -1 0 . C on d u c tiv ity o f Ion s

lon 20℃ (68℉ ) 25℃ (77℉ ) 100℃ (212℉ )

H+ 328 350 646

Na+ 45 50.1 155

K+ 67 73.5 200

NH +4 67 73.5 200

Mg++ 47 53.1 170

Ca++ 53.7 59.5 191

OH- 179 197 446

Cl - 69.0 76.3 207

HCO 3- 36.5 44.5 -

NO -3 65.2 71.4 178

H2PO-4 30.1 36.0 -

CO3- - 63.0 72.0 -

HPO 4- - - 53.4 -

SO4- - 71.8 79.8 234

PO4- - - - 69.0 -

10. Conductivity of Inos


10A p pendix

- 136 -

1 0 -1 1 . S pe c if ic C on d u c ta nc e o f S od iu m C hlo rid e

0 10 20 30 40 50 60 70 80 90 1 0 0

0

10

20

30

40

50

60

70

Specific Conductance of Sodium Chloride

Sp

eci

fic

Co

nd

uct

an

ce (

X1

00

0 m

ho

s/cm

)

Concentrat ion of Sodium chlor ide (X1000 mg/l)

11. Specific Conductance of Sodium Chloride


10 Ap pendix

- 137 -

1 0 -1 2 . Ion i za tio n o f C a rbo n D io x id e S o lu tio n a s Fun c tio n o f the pH

4 5 6 7 8 9 1 0 1 1 1 2

0

2 0

4 0

6 0

8 0

1 0 0

Ionization of CO2 solutions as Function of th pH at 25oC

C O3

-2C O2 H C O

3

-

Mo

l F

ract

ion

X 1

00

pH Value

12. Ionization of Carbon Dioxide Solution as Function of the pH


10A p pendix

- 138 -

1 0 -1 3 . O s m otic P re s s ure o f S od iu m C hlo r id e

0 2 0 0 0 4 0 0 0 6 0 0 0 8 0 0 0 1 0 0 0 0 1 2 0 0 0 1 4 0 0 0

0

2

4

6

8

1 0

1 2

Os

mo

tic

Pre

ss

ure

(k

g/c

m2)

m g / l N a C l

13. Osmotic Pressure of Sodium Chloride13. Osmotic Pressure of Sodium Chloride


10 Ap pendix

- 139 -

1 0 -1 4 . O s m otic P re s s ure o f S o lu t io n

0 5 1 0 1 5 2 0 2 5 3 0

0

1

2

3

4

Lithium Chloride

Sod ium Chlor ide

Seawater

Ethyla lcohol

Magnesium sulfate

Zinc sulfate

Fructose

Sucrose

Increasing

molecular weight

Osm

oti

c P

ress

ure

(MP

a)

C o n c e n t r a t i o n i n w a t e r ( % b y w e i g h t )

14. Osmotic Pressure of Solution


10A p pendix

- 140 -

1 0 -1 5 . C S M P e rfo r m a nc e

1) CSM Performance for Temperature Variation

0 10 20 30 40 50

0

20

25

30

35

40

B L g r a d e

B N , B E g r a d e

GF

D(g

al/

ft2 ·d

ay

)

T e m p e r a t u r e ( o C )

0 1 0 2 0 3 0 4 0 5 0

98.2

98.4

98.6

98.8

99.0

99.2

99.4

99.6

99.8

BL grade

B N , B E g r a d e

Re

jec

tio

n(%

)

T empe ra tu re (oC )

<Test condition>

15. CSM Performance

ParameterFeed Concentration

(ppm)

Pressure

(㎏/㎤)

Temperature

(℃)

Recovery

(%)Remark

BN,BE grade 2,000 15 Variable 15Fix drain at 25℃, 15% recovery

BL grade 2,000 10 Variable 15


10 Appendix

- 141 -

2) CSM Performance for Pressure Variation

2 4 6 8 10 12 14 16 18 20

0

10

20

30

40

50

60

70

B L g r a d e

B N , B E g r a d e

GF

D(g

al/

ft2·

da

y)

P r e s s u r e ( K g / c m2)

0 2 4 6 8 10 12 14 16 18

96.5

97.0

97.5

98.0

98.5

99.0

99.5

100.0

B L g r a d e

B N , B E g r a d e

Re

jec

tio

n(%

)

P r e s s u r e ( K g / c m2)

<Test condition>


(ppm)

Pressure

(㎏/㎤)

Temperature

(℃)

Recovery

(%)Remark

BN,BE grade 2,000 Variable 25 15Fix drain at 15㎏/㎤, 15% recovery

BL grade 2,000 Variable 25 15

15. CSM Performance


10Appendix

- 142 -

3) CSM Performance for Feed Concentration

0 2000 4000 6000 8000 10000

0

5

10

15

20

25

30

35

40

45

B L g r a d eB N , B E g r a d e

GF

D(g

al/

ft2·

da

y)

C o n c e n t r a t i o n ( p p m )

0 2000 4000 6000 8000 10000

9 5

9 6

9 7

9 8

9 9

100

B L g r a d e

B N , B E g r a d e

Re

jec

tio

n(%

)

C o n c e n t r a t i o n ( p p m )

<Test condition>


(ppm)

Pressure

(㎏ /㎤)

Temperature

(℃)

Recovery

(%)Remark

BN,BE grade Variable 15 25 15Fix drain at 2000ppm, 15% recovery

BL grade Variable 10 25 15

15. CSM Performance


10 Appendix

- 143 -

4) CSM Performance for pH

0 2 4 6 8 1 0 1 2 1 4

0

1 0

2 0

3 0

4 0

B N , B E g r a d e

B L g r a d e

GF

D(g

al/

ft2·

da

y)

p H

0 2 4 6 8 1 0 1 2

9 0

9 2

9 4

9 6

9 8

100

B L g r a d e

B N , B E g r a d e

Re

jec

tio

n(%

)

p H

<Test condition>


(ppm)

Pressure

(㎏/㎤)

Temperature

(℃)

Recovery

(%)Remark

BN,BE grade 2,000 15 25 15

BL grade 2,000 10 25 15

15. CSM Performance


10Appendix

- 144 -

5) CSM Performance for Recovery

0 1 0 2 0 3 0 4 0 5 0 6 0

0

5

1 0

1 5

2 0

2 5

3 0

3 5

4 0

B N , B E g r a d e

B L g r a d e

GF

D(g

al/

ft2·

da

y)

R e c o v e r y ( % )

0 1 0 2 0 3 0 4 0 5 0 6 0

95.0

95.5

96.0

96.5

97.0

97.5

98.0

98.5

99.0

99.5

100.0

B N , B E g r a d e

B L g r a d e

Re

jec

tio

n(%

)

R e c o v e r y ( % )

<Test condition>


(ppm)

Pressure

(㎏ /㎤)

Temperature

(℃)

Recovery

(%)Remark

BN,BE grade 2,000 15 25 Variable

BL grade 2,000 10 25 Variable

15. CSM Performance


10 Appendix

- 145 -

10-16. Temperature Correction Factor(TCF)

1) TCF of BN, BE, TN grade

T e m p 0 0 . 1 0 . 2 0 . 3 0 . 4 0 . 5 0 . 6 0 . 7 0 . 8 0 . 9

5 2 . 1 3 4 2 . 1 2 5 2 . 1 1 7 2 . 1 0 8 2 . 1 0 0 2 . 0 9 1 2 . 0 8 3 2 . 0 7 4 2 . 0 6 6 2 . 0 5 8

6 2 . 0 4 9 2 . 0 4 1 2 . 0 3 3 2 . 0 2 5 2 . 0 1 7 2 . 0 0 9 2 . 0 0 1 1 . 9 9 3 1 . 9 8 5 1 . 9 7 7

7 1 . 9 6 9 1 . 9 6 1 1 . 9 5 3 1 . 9 4 5 1 . 9 3 7 1 . 9 3 0 1 . 9 2 2 1 . 9 1 4 1 . 9 0 7 1 . 8 9 9

8 1 . 8 9 2 1 . 8 8 4 1 . 8 7 7 1 . 8 6 9 1 . 8 6 2 1 . 8 5 5 1 . 8 4 7 1 . 8 4 0 1 . 8 3 3 1 . 8 2 5

9 1 . 8 1 8 1 . 8 1 1 1 . 8 0 4 1 . 7 9 7 1 . 7 9 0 1 . 7 8 3 1 . 7 7 6 1 . 7 6 9 1 . 7 6 2 1 . 7 5 5

1 0 1 . 7 4 8 1 . 7 4 1 1 . 7 3 4 1 . 7 2 8 1 . 7 2 1 1 . 7 1 4 1 . 7 0 7 1 . 7 0 1 1 . 6 9 4 1 . 6 8 8

1 1 1 . 6 8 1 1 . 6 7 5 1 . 6 6 8 1 . 6 6 2 1 . 6 5 5 1 . 6 4 9 1 . 6 4 2 1 . 6 3 6 1 . 6 3 0 1 . 6 2 3

1 2 1 . 6 1 7 1 . 6 1 1 1 . 6 0 5 1 . 5 9 8 1 . 5 9 2 1 . 5 8 6 1 . 5 8 0 1 . 5 7 4 1 . 5 6 8 1 . 5 6 2

1 3 1 . 5 5 6 1 . 5 5 0 1 . 5 4 4 1 . 5 3 8 1 . 5 3 2 1 . 5 2 6 1 . 5 2 1 1 . 5 1 5 1 . 5 0 9 1 . 5 0 3

1 4 1 . 4 9 8 1 . 4 9 2 1 . 4 8 6 1 . 4 8 1 1 . 4 7 5 1 . 4 6 9 1 . 4 6 4 1 . 4 5 8 1 . 4 5 3 1 . 4 4 7

1 5 1 . 4 4 2 1 . 4 3 6 1 . 4 3 1 1 . 4 2 5 1 . 4 2 0 1 . 4 1 5 1 . 4 0 9 1 . 4 0 4 1 . 3 9 9 1 . 3 9 4

1 6 1 . 3 8 8 1 . 3 8 3 1 . 3 7 8 1 . 3 7 3 1 . 3 6 8 1 . 3 6 3 1 . 3 5 7 1 . 3 5 2 1 . 3 4 7 1 . 3 4 2

1 7 1 . 3 3 7 1 . 3 3 2 1 . 3 2 7 1 . 3 2 2 1 . 3 1 8 1 . 3 1 3 1 . 3 0 8 1 . 3 0 3 1 . 2 9 8 1 . 2 9 3

1 8 1 . 2 8 8 1 . 2 8 4 1 . 2 7 9 1 . 2 7 4 1 . 2 7 0 1 . 2 6 5 1 . 2 6 0 1 . 2 5 6 1 . 2 5 1 1 . 2 4 6

1 8 1 . 2 8 8 1 . 2 8 4 1 . 2 7 9 1 . 2 7 4 1 . 2 7 0 1 . 2 6 5 1 . 2 6 0 1 . 2 5 6 1 . 2 5 1 1 . 2 4 6

2 0 1 . 1 9 7 1 . 1 9 3 1 . 1 8 8 1 . 1 8 4 1 . 1 8 0 1 . 1 7 5 1 . 1 7 1 1 . 1 6 7 1 . 1 6 3 1 . 1 5 8

2 1 1 . 1 5 4 1 . 1 5 0 1 . 1 4 6 1 . 1 4 2 1 . 1 3 8 1 . 1 3 3 1 . 1 2 9 1 . 1 2 5 1 . 1 2 1 1 . 1 1 7

2 2 1 . 1 1 3 1 . 1 0 9 1 . 1 0 5 1 . 1 0 1 1 . 0 9 7 1 . 0 9 3 1 . 0 8 9 1 . 0 8 5 1 . 0 8 2 1 . 0 7 8

2 3 1 . 0 7 4 1 . 0 7 0 1 . 0 6 6 1 . 0 6 2 1 . 0 5 9 1 . 0 5 5 1 . 0 5 1 1 . 0 4 7 1 . 0 4 4 1 . 0 4 0

2 4 1 . 0 3 6 1 . 0 3 2 1 . 0 2 9 1 . 0 2 5 1 . 0 2 1 1 . 0 1 8 1 . 0 1 4 1 . 0 1 1 1 . 0 0 7 1 . 0 0 4

2 5 1 . 0 0 0 0 . 9 9 6 0 . 9 9 3 0 . 9 8 9 0 . 9 8 6 0 . 9 8 3 0 . 9 7 9 0 . 9 7 6 0 . 9 7 2 0 . 9 6 9

2 6 0 . 9 7 0 0 . 9 7 0 0 . 9 7 0 0 . 9 7 0 0 . 9 7 0 0 . 9 7 0 0 . 9 7 0 0 . 9 7 0 0 . 9 7 0 0 . 9 7 0

2 7 0 . 9 4 0 0 . 9 4 0 0 . 9 4 0 0 . 9 4 0 0 . 9 4 0 0 . 9 4 0 0 . 9 4 0 0 . 9 4 0 0 . 9 4 0 0 . 9 4 0

2 8 0 . 9 1 2 0 . 9 1 2 0 . 9 1 2 0 . 9 1 2 0 . 9 1 2 0 . 9 1 2 0 . 9 1 2 0 . 9 1 2 0 . 9 1 2 0 . 9 1 2

2 9 0 . 8 8 5 0 . 8 8 5 0 . 8 8 5 0 . 8 8 5 0 . 8 8 5 0 . 8 8 5 0 . 8 8 5 0 . 8 8 5 0 . 8 8 5 0 . 8 8 5

3 0 0 . 8 5 9 0 . 8 5 9 0 . 8 5 9 0 . 8 5 9 0 . 8 5 9 0 . 8 5 9 0 . 8 5 9 0 . 8 5 9 0 . 8 5 9 0 . 8 5 9

3 1 0 . 8 3 3 0 . 8 3 3 0 . 8 3 3 0 . 8 3 3 0 . 8 3 3 0 . 8 3 3 0 . 8 3 3 0 . 8 3 3 0 . 8 3 3 0 . 8 3 3

3 2 0 . 8 0 9 0 . 8 0 9 0 . 8 0 9 0 . 8 0 9 0 . 8 0 9 0 . 8 0 9 0 . 8 0 9 0 . 8 0 9 0 . 8 0 9 0 . 8 0 9

3 3 0 . 7 8 6 0 . 7 8 6 0 . 7 8 6 0 . 7 8 6 0 . 7 8 6 0 . 7 8 6 0 . 7 8 6 0 . 7 8 6 0 . 7 8 6 0 . 7 8 6

3 4 0 . 7 6 3 0 . 7 6 3 0 . 7 6 3 0 . 7 6 3 0 . 7 6 3 0 . 7 6 3 0 . 7 6 3 0 . 7 6 3 0 . 7 6 3 0 . 7 6 3

3 5 0 . 7 4 1 0 . 7 4 1 0 . 7 4 1 0 . 7 4 1 0 . 7 4 1 0 . 7 4 1 0 . 7 4 1 0 . 7 4 1 0 . 7 4 1 0 . 7 4 1

3 6 0 . 7 2 0 0 . 7 2 0 0 . 7 2 0 0 . 7 2 0 0 . 7 2 0 0 . 7 2 0 0 . 7 2 0 0 . 7 2 0 0 . 7 2 0 0 . 7 2 0

3 7 0 . 7 0 0 0 . 7 0 0 0 . 7 0 0 0 . 7 0 0 0 . 7 0 0 0 . 7 0 0 0 . 7 0 0 0 . 7 0 0 0 . 7 0 0 0 . 7 0 0

3 8 0 . 6 8 0 0 . 6 8 0 0 . 6 8 0 0 . 6 8 0 0 . 6 8 0 0 . 6 8 0 0 . 6 8 0 0 . 6 8 0 0 . 6 8 0 0 . 6 8 0

3 9 0 . 6 6 1 0 . 6 6 1 0 . 6 6 1 0 . 6 6 1 0 . 6 6 1 0 . 6 6 1 0 . 6 6 1 0 . 6 6 1 0 . 6 6 1 0 . 6 6 1

4 0 0 . 6 4 3 0 . 6 4 3 0 . 6 4 3 0 . 6 4 3 0 . 6 4 3 0 . 6 4 3 0 . 6 4 3 0 . 6 4 3 0 . 6 4 3 0 . 6 4 3

16. Temperature Correction Factor(TCF)


10Appendix

- 146 -

2) TCF of BL grade

Temp 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

5 2.088 2.080 2.072 2.064 2.056 2.048 2.039 2.031 2.024 2.016

6 2.008 2.000 1.992 1.984 1.977 1.969 1.961 1.954 1.946 1.938

7 1.931 1.923 1.916 1.908 1.901 1.894 1.886 1.879 1.872 1.865

8 1.857 1.850 1.843 1.836 1.829 1.822 1.815 1.808 1.801 1.794

9 1.787 1.780 1.774 1.767 1.760 1.753 1.747 1.740 1.733 1.727

10 1.720 1.714 1.707 1.701 1.694 1.688 1.681 1.675 1.669 1.662

11 1.656 1.650 1.644 1.638 1.631 1.625 1.619 1.613 1.607 1.601

12 1.595 1.589 1.583 1.577 1.571 1.565 1.560 1.554 1.548 1.542

13 1.536 1.531 1.525 1.519 1.514 1.508 1.502 1.497 1.491 1.486

14 1.480 1.475 1.469 1.464 1.459 1.453 1.448 1.443 1.437 1.432

15 1.427 1.421 1.416 1.411 1.406 1.401 1.396 1.391 1.385 1.380

16 1.375 1.370 1.365 1.360 1.355 1.351 1.346 1.341 1.336 1.331

17 1.326 1.321 1.317 1.312 1.307 1.302 1.298 1.293 1.288 1.284

18 1.279 1.275 1.270 1.265 1.261 1.256 1.252 1.247 1.243 1.238

18 1.279 1.275 1.270 1.265 1.261 1.256 1.252 1.247 1.243 1.238

20 1.191 1.187 1.182 1.178 1.174 1.170 1.166 1.162 1.158 1.153

21 1.149 1.145 1.141 1.137 1.133 1.129 1.125 1.121 1.118 1.114

22 1.110 1.106 1.102 1.098 1.094 1.090 1.087 1.083 1.079 1.075

23 1.072 1.068 1.064 1.060 1.057 1.053 1.049 1.046 1.042 1.039

24 1.035 1.031 1.028 1.024 1.021 1.017 1.014 1.010 1.007 1.003

25 1.000 0.997 0.993 0.990 0.986 0.983 0.980 0.976 0.973 0.970

26 0.971 0.971 0.971 0.971 0.971 0.971 0.971 0.971 0.971 0.971

27 0.942 0.942 0.942 0.942 0.942 0.942 0.942 0.942 0.942 0.942

28 0.915 0.915 0.915 0.915 0.915 0.915 0.915 0.915 0.915 0.915

29 0.888 0.888 0.888 0.888 0.888 0.888 0.888 0.888 0.888 0.888

30 0.863 0.863 0.863 0.863 0.863 0.863 0.863 0.863 0.863 0.863

31 0.838 0.838 0.838 0.838 0.838 0.838 0.838 0.838 0.838 0.838

32 0.815 0.815 0.815 0.815 0.815 0.815 0.815 0.815 0.815 0.815

33 0.792 0.792 0.792 0.792 0.792 0.792 0.792 0.792 0.792 0.792

34 0.770 0.770 0.770 0.770 0.770 0.770 0.770 0.770 0.770 0.770

35 0.748 0.748 0.748 0.748 0.748 0.748 0.748 0.748 0.748 0.748

36 0.728 0.728 0.728 0.728 0.728 0.728 0.728 0.728 0.728 0.728

37 0.708 0.708 0.708 0.708 0.708 0.708 0.708 0.708 0.708 0.708

38 0.689 0.689 0.689 0.689 0.689 0.689 0.689 0.689 0.689 0.689

39 0.670 0.670 0.670 0.670 0.670 0.670 0.670 0.670 0.670 0.670

40 0.652 0.652 0.652 0.652 0.652 0.652 0.652 0.652 0.652 0.652

16. Temperature Correction Factor(TCF)

1. history of s aehan in dustries, inc. 2. introduction to ... · pressure vessel, cleaning agents...

Documents