Particulate and Colloidal Fouling

Flow rates and good chemistry are the critical factors in successfully cleaning membranes fouled by colloidal material. The high pH RoClean formulations contain dispersants that help push particles away from each other and away from the membrane surface, allowing a more effective clean.

THE FO O!"#$ %RE &'()TO(& OR "#*"C%TOR& OF &" T %#* CO O"*% FO+ "#$,
High pressure differential on elements. &urface water feed supply to system. High turbidity in feedwater. Element telescoping. *iscoloration of the membranes.

Fig 1, Element severely fouled with resin.

Fig 0, Element severely fouled with sand.

C E%#"#$ %))RO%CH

ow pH clean with either RoClean )-.- or /.- followed by a high pH clean with RoClean 011.

Consider reverse flow and air sparging during cleaning

procedure.
Heat cleaning solution to ma2imum allowed by the membrane

Figure -, Colloidal fouling of a membrane surface.

manufacturer.

&TE)& TO )RE3E#T"#$ &" T %#* CO O"*% FO+ "#$,

Conduct &*" 4&ilt *ensity "nde25 and turbidity testing to

determine colloidal fouling potential.

Conduct laboratory filtration study to determine fouling

potential, filterability of the water and optimum Ro6uest7 coagulant dosage.

Consider a multi media pilot filter study. +sing a microscope, inspect the spent &*" pads from the

((F feed and effluent to compare filter efficiency.

+se properly designed multi media filtration and consider the

need for air scour or surface wash.

Consider the use of a coagulant to increase the effectiveness

of the multi media filter.

+se clarifiers or other filtration e8uipment for high turbidity

waters.

Organics Fouling (non-biological)

THE FO O!"#$ %RE &'()TO(& OF OR$%#"C& FO+ "#$,
(embrane discoloration.

ow flow on individual membrane test data.

)ossible high pressure differential on individual membrane

test data.

THE RECO((E#*E* C E%#"#$ %))RO%CH "&,

% low pH clean using either RoClean )-.- or /.-, followed

Fig 1, (embrane surface with severe organics fouling.

by a high pH clean using either RoClean )111 or, for severe cases, RoClean )110 or 010. The combination of a low pH clean followed by the high pH solution is e2tremely effective in removing organics. The low pH cleaner helps brea9 the bridge between the organics and the membrane. The high pH solution then lifts the foulant off the membrane surface. This is why there is sometimes a color discharge only when using the high pH cleaner. *on:t be misled though, the low pH clean was a vital step in the cleaning regime.
Fig 0, (embrane fouled with humic acid.

&TE)& TO )RE3E#T"#$ OR$%#"C FO+ "#$ "#C +*E,

Remove the organics from the RO feedwater with a properly

designed multimedia filter with coagulant addition. (etal salts blended with a polymer are usually the most effective coagulant for high organic waters.

Figure -, &pent cleaning solution.

#OTE&,
Figure - is an e2ample of a spent cleaning solution. The foulant was humic acid, dissolved by formulated cleaners.

Iron and Manganese Fouling

THE FO O!"#$ %RE &'()TO(& OF "RO# %#* (%#$%#E&E FO+ "#$,
*iscoloration of membranes. )oor salt re;ection on individual membrane test data.

ow flow on individual membrane test data.

)ossible high pressure differential on individual membrane

Fig 1, "ron fouled membrane.

test data.
High iron or manganese values reported in feedwater. High pressure differential reported on first array.

The recommended cleaning approach is a low pH clean 4RoClean )<.-5 followed by a high pH clean 4RoClean )111 or 0115. Results can be further improved by heating the cleaning solution to the ma2imum allowed by the membrane manufacturer.

Fig 0, "ron fouled membrane.

&TE)& TO )RE3E#T"#$ "RO# %#* (%#$%#E&E FO+ "#$ "#C +*E,

Conduct a complete and accurate water analysis. "f the water supply contains a high amount of ferrous iron, it is
Figure -, "nitially, the foulant was a mystery, but it was later determined to be (anganese.

important to prevent it from o2idi=ing. O2idation can be caused by e2posure to chlorine or aeration of the water 4which occurs in the multimedia filter or other similar e8uipment5.
"ron and manganese can be removed using a greensand filter

regenerated with potassium permanganate. However, if potassium permanganate is overdose or is not properly rinsed from the greensand filter, it will o2idi=e the membrane surface.
"ron can be removed by chlorinating the water then removing

the o2idi=ed iron or manganese with a multimedia filter.

%**"T"O#% #OTE&,
"ron in water can be found as ferrous or ferric. Ferrous iron is dissolved iron that has not precipitated. E2posure to air turns ferrous iron into ferric 4o2idi=ed5 iron, which essentially, has become rust. Ferrous iron can e2ist in relatively high levels and not precipitate so long as it is not o2idi=ed. "t becomes a filtration issue if it becomes o2idi=ed. Ferric iron fouling can be removed using RoClean )<.-.

Sulfate Scale
&'()TO(& OF &+ F%TE &C% E "#C +*E,
(aterial e2truding from the downstream end of the last

membranes in the system.

Foulant will not dissolve when introduced to a dilute

hydrochloric acid solution.

)oor salt re;ection, low flow, and>or high pressure differential
Fig 1, &E( photograph of barium sulfate scale.

on individual test data.

?@ 2 /.@ element weight e2ceeds /. pounds. &ite reports interruption in the antiscalant or acid in;ection. &ite reports scale in the last vessel or piping of the
Fig 0, &E( photograph of barium sulfate scale.

concentrate stream.

THE RECO((E#*E* C E%#"#$ %))RO%CH 3%R"E&,

"f some of the material dissolved in a dilute hydrochloric

solution, begin the clean with a low pH product such as RoClean /.-.
"f none of the material dissolved in a dilute hydrochloric
Figure -, % magnified view of one of the crystals.

solution, then s9ip the low pH clean and use only a high pH solution, RoClean ?11.
"f possible, heat the cleaning solution to the ma2imum

temperature allowed by the membrane manufacturer.

%**"T"O#% #OTE&,
Aarium sulfate scale on a membrane, pressure vessel, or pipe feels li9e fine grit sandpaper. This abrasiveness can cause damage to a membrane surface during system operation. !hen the system is started and stopped, the ve2ar 4feedspacer5 material shifts slightly. This shifting can cause the barium sulfate scale to scratch the membrane surface causing permanent damage and resulting in a loss of re;ection.

Calcium Carbonate Scale

&'()TO(& OF C% C"+( C%RAO#%TE &C% E "#C +*E,
&cale e2truding out of the downstream end of the last

membranes in the system.

%ll of the scale dissolves when introduced to a dilute

hydrochloric acid solution.

)oor salt re;ection, low flow, and or, high pressure differential
Fig 1, Element severely fouled with CaCO- scale.

on individual membrane test data.

?@ 2 /.@ Element weight e2ceeds /B pounds. &ite reports interruption in the antiscalant or acid in;ection. &ite reports scale in the last vessel or piping of the

C Essential +sers $uide !ord doc is missing Figure 1. D

concentrate stream.

THE RECO((E#*E* C E%#"#$ %))RO%CHE& %RE,

ow pH clean with RoClean /.-. to maintain a pH of -...

"n severe cases, add hydrochloric acid to the cleaning solution

Fig 0, Element severely fouled with CaCO- scale.

TE&T"#$ FOR C% C"+( C%RAO#%TE &C% E,

Calcium carbonate scale is not always as white as shown in Figures 1 and 0. Aut, a 8uic9 test can be conducted to see if the foulant is comprised solely of calcium carbonate.

"n a glass bea9er, ma9e a 1,1 dilution of HCl and *" water. *rop a small sample of the foulant into the solution. "f the foulant contains calcium carbonate, it will bubble 4see Figure -5. Continue adding acid until the bubbling stops or until the scale disappears. "f the bea9er contains residual material after the bubbling has stopped, then the foulant consists of more than ;ust calcium carbonate.
Figure -, %n acid test is used to determine if a foulant is comprised of CaCO- scale.

Biological Fouling
&'()TO(& OF A"O O$"C% FO+ "#$ "#C +*E,
3isible slime on the feed side of the membrane. &ite reports slime in the cartridge filter housing and piping. &ite reports high pressure differential in the first array. Odor "ndividual membrane test data reports high pressure
Fig 1, &ee H%dditional #otes@.

differentials.
(embranes are telescoped.

THE RECO((E#*E* C E%#"#$ %))RO%CHE& %RE,

%dd *A0. biocide to a low pH cleaning solution 4RoClean

Fig 0, (embrane fouled with biological material.

/.- or )-.-5.
Follow the low pH clean with RoClean )111 4for severe cases,

consider )1105.
Heat the high pH solution 4"f )110 is used, temperature range

is E.F G 1..FF5.

&TE)& TO )RE3E#T A"O O$"C% FO+ "#$ "#C +*E,

)roperly dose sodium bisulfite. Clean and saniti=e the pretreatment e8uipment and piping.

Figure -, (embrane fouled with biological material.

 Consider an intermittent biocide treatment. Evaluate the necessity of a continuous in;ection biocide.

%**"T"O#% #OTE&,
Figure 1 is a membrane sample that was ta9en from a single ?@ 2 /.@ element. The first 0.@ of both the membrane and ve2ar 4feedspacer5 were plugged with biological slime while the last 0.@ were free of the foulant.
Figure /, (Aiological fouling of membrane and vessel interior.

ackpressure Damage (Delamination)

A%CI)RE&&+RE "& C%+&E FOR CO#CER# AEC%+&E "T C%# C%+&E *E %("#%T"O# OF RE3ER&E O&(O&"& (E(AR%#E&.
*elamination is a separation of the membrane from the bac9ing material. This type of damage is caused by bac9pressure from the permeate side of the membrane and results in a loss of re;ection. Aac9pressure can be caused by restrictions in the line such as valves, resin beds, or elevated piping. "t can also be caused by improper relief of product pressure in a product staged 4double pass5 RO system. Figure 1 shows an e2ample of severe bac9pressure damage where the pattern of the feedspacer 4ve2ar5 is clearly visible. The bubbling of the membrane shown in Figure 0 is sometimes subtle and difficult to detect. Figure - shows the membrane envelope, comprised of two sheets of membrane and one sheet of tricot 4permeate carrier5 between. The yellow arrow is the tricot, the red arrow is the bac9ing material and the blue arrows are the membrane that was pulled away from the bac9ing material. *elamination allowed easy separation of the membrane material from the bac9ing material. This separation would not be possible on an undamaged element.
Figure Fig 0 Figures 1

Membrane elescoping - !ig" Pressure Differential

TE E&CO)"#$ OF % (E(AR%#E "& C%+&E* A' EJCE&&"3E
*"FFERE#T"% & AET!EE# THE FEE* )RE&&+RE %#* THE CO#CE#TR%TE )RE&&+RE.

The ma2imum pressure differential for a single /.@ long membrane is 1. psi. !hen this pressure is e2ceed, damage to the membrane and its materials of construction can occur.
Fig 1, Telescoped element.

CO#&E6+E#CE& OF TE E&CO)"#$,
*amage to the fiberglass outer wrapping allows water to flow

on the outside of the element. This will reduce crossflow across the membrane surface and increase the fouling potential.
"f the feed spacer 4ve2ar5 moves within the membrane area, it
Fig 0, 3e2ar protruding from the element.

can scratch the membrane surface and cause permanent damage.

$lue lines can be stressed and fail. Flow patterns will be disrupted, resulting in channeling over

the membrane surface and uneven fouling.

The membrane crease will be stressed. Failure can occur at

Figure -, 3e2ar protruding from the element.

these points near the permeate tube.

)RE3E#T"O#, (O#"TOR )RE&&+RE *"FFERE#T"% & %CRO&& THE

E#T"RE &'&TE( %& !E

%& %CRO&& E%CH %RR%'.

"F THE F"AER$ %&& "& *%(%$E*, RE)%"R !"TH 3"#'

T%)E OR ) %CE THE AR"#E &E% & O# THE O))O&"TE E#* "F THE F"AER$ %&& %#* *%(%$E* O# TH%T E#*.

%T* %RE #OT

C E%# THE &'&TE( AEFORE )RE&&+RE& EJCEE* 1.

)&" )ER (E(AR%#E E E(E#T %#* CORRECT %#' EJCE&&"3E FO+ "#$ )ROA E(&.

Determining Scale Potential for #e$erse Osmosis %pplications

%# %CC+R%TE FEE*!%TER %#% '&"& "& CR"T"C% "#
*ETER("#"#$ THE &C% "#$ )OTE#T"% OF % RE3ER&E O&(O&"& %)) "C%T"O#.

The conversion chart in Table 1 outlines the ions that must be 9nown in order to complete an antiscalant pro;ection. % complete analysis should also include the pH of the water at the time of sampling. For ions such as %luminum, Aarium, "ron, and &trontium, it is important to ensure that the testing methods can detect ion levels in the range of ...1 ppm. The %vista %dvisorK computer program allows the user to enter siteL specific feedwater and system data to,
Recommend an antiscalant and dosage for a specific
Fig 0, %vista %dvisorK dialog. Fig 1, Conversion chart.

application.
*etermine the scaling potential of a feedwater. Calculate chemical in;ection rates. *etermine ma2imum system recovery based on scaling

potential. "t can also be a very useful system troubleshooting tool.

%ntiscalant Dosing Calculations

Errors in chemical dosing are common, but also very preventable. They typically occur when chemical feed pumps are not calibrated correctly or when dosage calculation errors are made. The purpose of the information below is to provide a 8uic9 and easy dosage calculation.

&TE)& TO *ETER("#E *O&%$E )%R%(ETER&,

1. "dentify the optimum ppm dosage of the appropriate 3itec7 antiscalant. This can be determined using the %vista %dvisorK *osing Report 4E2ample shown in Figure 15. 0. *etermine the desirable dilution in the chemical day tan9 4see Figure 05. 1.M is the ma2imum recommended dilution. -. "dentify the feedflow rate to the RO system. /. Calculate the dosage using the values determined in &teps 1L - and the calculation below.

Fig 1, %vista %dvisorK *osing Report.

Table 1, 6uic9 Reference *ilution Chart.

*O&%$E CO#F"R(%T"O#
+sing a drawdown assembly 4as shown in Figure -5, verify that the proper ml>minute dosage is being delivered to the system. "f there is no drawdown tube, calibrate the pump by placing the pump suction line into a graduated cylinder filled with the solution to be pumped. (easure the drop in solution over a oneLminute period to verify the ml per minute in;ection rate. %d;ust the pump accordingly until it is correct. The %vista %dvisor computer program is available to recommend and dose 3itec antiscalants based on site specific feedwaters. &ystem flow in gallons per minute J ppm *osage of %ntiscalant J M *ilution Constant N 4H(@ value in Table 15 ml>minute in;ection rate
Fig -, *rawdown assembly diagram.

Multi Media Filter Operation

(embrane manufactures typically specify that RO feedwater should have a &*" 4&ilt *ensity "nde25 of O B.. and a turbidity value O..0. #T+. ow &*" and turbidity values are believed to mean that the potential for colloidal fouling is reduced. !ith proper application, design, and operation, a multi media filter 4((F5 can achieve these goals.
Fig 1, Turbidity vs. Time.

T+RA"*"T'
!hen a ((F is put onLline following a bac9wash, the turbidity value can sometimes be similar to the feedwater turbidity. Over a brief period of time, the turbidity should drop and stabili=e to an acceptable level. %s the system operates, the turbidity rises. !hen the turbidity increases by 1.M, a bac9wash should be initiated. "f the ((F is not bac9washed, the turbidity in the effluent may eventually e2ceed the feedwater turbidity.
Fig 0, *eltaL) vs. Time.

)RE&&+RE *RO) 4*E T%G)5

"t is important to monitor the deltaL) of the ((F. "ncreases in deltaL) signal the filter is wor9ing and removing particles. The ((F should be bac9washed at or before the deltaL) reaches ten 41.5.
Fig -, Turbidity vs. *osage.

*E T%L) %#* T+RA"*"T' %RE "#*E)E#*E#T

"t is important to understand that deltaL) and turbidity brea9through can be independent of each other. The ((F should be bac9washed based on whichever occurs first.

T+RA"*"T' 3&. *O&%$E

!hen dosing a coagulant, high turbidity does not always indicate an inade8uate dosage. "t is important to understand that both underdosing and overdosing can cause turbidity to rise. %dditionally, significant overdosing can cause the effluent turbidity to e2ceed the feedwater #T+. The optimal dosing range at the bottom of the curve usually has a spread of 1 to 0 ppm.

Fig /, )hoto shows the interior of a failed ((F bed. The garnet, anthracite and sand are intermi2ed and the surface of the bed is cratered.

"()ORT%#T O)ER%T"#$ T")&

Fre8uently test ((F feed and effluent turbidity. +se a drawdown tube to verify coagulant dosages. +se reliable pressure gauges to measure ((F deltaL). "nitiate ((F bac9wash based on turbidity, deltaL) or time.

Silt Densit& Inde' (SDI)

%lthough the values do not directly correlate to the fouling potential of a specific water, the &ilt *ensity "nde2, or &*", test is considered to be an industry standard for measuring the colloidal fouling potential of spiral wound membranes. For &*" test results to be accurate, the feed line to the 9it must be connected to the raw water line representative of the feedwater to the RO system. Ensure that all of the air is purged from the apparatus and that the feed pressure is ad;usted to -. psig.
Fig 1, &*" pad of the Feedwater.

&*" TE&T )ROCE*+RE,

The initial time re8uired to fill a B.. ml graduated cylinder is measured and recorded as t.. % measure of the time re8uired to collect B.. ml volumes is noted again at B, 1. and 1B minutes after the initial start. These times are recorded as tB, t1. and t1B respectively.
Fig 0, &*" pad of the downstream ((F effluent, no coagulant addition.

C% C+ %T"O# OF &" T *E#&"T' "#*EJ 4&*"5,

The &*" value is then calculated using the following e8uation, &*" N 4 1 L t. > t1B 5 1.. T

t( N "nitial time in seconds re8uired to collect a B.. ml sample. t)* N Time in seconds re8uired to collect a B.. ml sample after fifteen min. N Total test time in minutes.
Fig -, &*" pad of the downstream ((F effluent, with coagulant addition.

RECO((E#*E* &*" 3% +E,

The ma;or membrane manufacturers typically recommend maintaining an &*" value of -.. to B.. for feedwater to a reverse osmosis system.

E3% +%T"#$ &)E#T F" TER&,

Figures 1 G - are photos of actual &*" pads ta9en at a single customer site. The benefit of in;ecting coagulant ahead of the multimedia filter is apparent by comparing Figure 0 and Figure -. The

Fig /, Typical &*" apparatus.

&*" test in Figure - was ta9en on the ((F effluent of an onLsite pilot filter that was drawing feedwater from the same source as Figure 0.

urbidit&
Turbidity is an important water 8uality indicator for almost any treatment application. Turbidity represents the presence of dispersed, suspended solidsLparticles not in true solution and often includes silt, clay, algae and other microorganisms, organic matter and minute particles. &uspended solids obstruct the transmittance of light through a water sample and impart a 8ualitative characteristic, 9nown as turbidity, to water. Turbidity is not a direct measure of suspended particles in water. "nstead, it is a measure of the scattering effect such particles have on light. The way in which suspended particles scatter light is very comple2. )article si=e, shape, and nature all effect scattering. "n addition, colored solutions adsorb light and can affect turbidity. 3ery small particles 4..0 microns5 scatter light e8ually in the forward and bac9ward direction. arger particles 41 micron5 scatter light primarily in the forward direction. Forward scattering is intensified as the concentration of suspended solids increases. Turbidity is determined electronically by the apparatus shown in Figure 1. ight from a tungsten filament 4usually5 passes through the sample to be measured. ight scattered at an angle of E. degrees is measured. #inety degrees is chosen because it is very sensitive to light scatter. *ividing scattered light intensity by that transmitted in a forward direction compensates for the affect of sample color. %n instrument of this type is 9nown as a ratio turbidimeter.
Fig 0, Turbidity vs )article Count. Fig 1, Hach Turbidimeter.

roubles"ooting +uide
Product Flo, *ecreasing "ncreasing in #o antiscalant &econd &cale Fouling %ntiscalant stage underfed Salt #e-ection Delta Pressure Direct Cause Indirect Cause Correcti$e %ction Carry out mass balance, confirm that antiscalant is present in the concentrate stream and ensure shutdown flush operating to determine source of problem Clean the system Clean the system "mprove the preLtreatment performance

Colloidal Fouling

)oor pretreatment

*ecreasing

&top train and clean before the *) increases beyond )oor B0 psi 4-.P bar5 )retreatment "mprove pretreatment as Compaction colloidal fouling on the front &ystem operation end membrane normally outside limits accompanies first element compaction "ncreasing in First stage +se of nonL approved products Carryover from pretreatment processes

"ncompatible Chemicals

Review chemical compatibly Clean the system

Aiofouling

"neffective or absent biocide program

Carry out a bacterial survey *isinfect>clean the entire system and piping %dd or increase biocide in;ection and continue to monitor until the system proves to be under control "nvestigate source water and improve pretreatment Clean the system Compacted elements are irreparably damaged. Replace effected membranes and determine

#o Change

Organic #o Change Fouling #o Change

Oil and>or )olyecltroclye residual in RO feedwater

"ncreasing

Compaction High pressure +neven flu2

how to reduce pressure and reLbalance flow rates O2idation *amage Free chlorine or use of an incompatible cleaner %ttempt to use Resi=e to restore performance temporarily Replace affected elements

(embrane Failure "ncreasing *ecreasing #o Change

)roduct bac9pressure or vacuum , Remove source of abrasion, bac9pressure>vacuum installation "mprove pretreatment damage to permeate tube "f possible, probe elements to determine which membrane4s5 in the vessel are effected Replace oLrings, shim vessel *etermine how to avoid QrammingQ the membranes during system startLup and shut down

(ovement or OLRing lea9 improper installation

.ater Mat"
*ETER("#"#$ !%TER E6+"3% E#T&,

One /0S0 +allon .ater N One Cubic Foot .ater N 102 ft0 of ,ater N ) grain "ardness N

-.<?B iters -<?B milliliters 4m 5 ?.-/ pounds <./? +& $allons 1.. )&" 1<.1 ppm

C% C+ %T"#$ )%RT& )ER (" "O# TO $R%"#& )ER +& $% O#,

%. To convert ppm of hardness to grains per +& $allon, divide by 1<.1, ppm hardness Calculation3 1<.1 N grains>+& $allon

0B. ppm hardness 4'ample3 1<.1 N 1/.P1 grains per gallon

A. To convert grains per +.&. $allon to parts per million of hardness, multiply by 1<.1, Calculation3 4'ample3 4$rains hardness5 R 41<.15 N ppm hardness 41/.P1 grains5 R 41<.15 N 0/E.?- ppm hardness

C. &oftener Capacity, Estimate 0B,... grains removal capacity per cubic foot of resin, 4Cubic feet of resin5 40B,...5 Calculation3 4Feed hardness in grains5 4'ample3 4-B ft-5 40B,... grains>ft-5 N &oftener capacity in gallons

C% C+ %T"#$ RE3ER&E O&(O&"& &'&TE( RECO3ER',

Calculation3 4'ample3 4)ermeate $)(>Feed $)(5 41..5 N M Recovery 4<B $)(>1.. $)(5 41..5 N <BM Recovery

C% C+ %T"#$ RE3ER&E O&(O&"& &% T RESECT"O#,

Calculation3 4'ample3 41 L 4permeate T*&>feed T*&55 2 1.. N M re;ection 41L 40.>1...55 1.. N E?M

C% C+ %T"#$ THE CO#CE#TR%T"O# F%CTOR,

%. Recovery Aased Concentration Factor, Calculation3 4'ample3 1 > 41Lrecovery5 N concentration factor 1 > 41L..<B5 N /

A. Conductivity Aased Concentration Factor, 4Concentrate Conductivity G )ermeate Conductivity5 4Feed Conductivity G )ermeate Conductivity5

Calculation3

concentration factor

4/... us G - u&5 4'ample3 41... us G - u&5 N /

CO#3ERT"#$ TE()ER%T+RE
%. *egree Fahrenheit to *egree Celsius, Calculation3 4FF G -05 4..BBP5 N FC

4'ample3

4<<FF G -05 4.BBP5 N 0B..0FC

A. *egree Celsius to *egree Fahrenheit, Calculation3 4'ample3 FC > 4..BBP T -05 R 1.. N FF 0BFC > 4.BBP T -05 R 1.. N <P.?FF

C% C+ %T"#$ CHE("C% "#SECT"O# R%TE& "# (" " "TER& )ER ("#+TE,
Calculation3 Feed $)( R Chem ))( R 4?.-/5 N $)(>minute R -<?B N ml>mn 4chemical lb per gallon5 R 1,...,... Feed $)( N 0.. Chemical dosage ))( N 0.. Chemical U per gallon N 1../? 40.. R 0 R 4?.-/5 R -<?B5 1../? R 1,...,...

4'ample3

N 1.0. milliliters per minute of neat chemical

1.0. milliliters per minute N milliliters per minute of a diluted solution )ercent dilution 5ote3 #ote, E2press percentages as decimals.

C% C+ %T"#$ CHE("C% *" +T"O#& 4%)) "C%A E TO 3ER' *" +TE &O +T"O#&R5,
Calculation3 ?.-/ R 3d R Cd > 1.. R *n N 3n 3n N 3olume of neat chemical re8uired, gallons 3d N *ilution volume, gallons Cd N )ercent dilution by weight *n N #eat solution density, lb>gallon 3d N 0.. gallons Cd N BM *n N 1../? lb>gallon ?.-/ R 0.. R B > 1.. R 1../? NN <.EP gallons 5ote3 RTo calculate more concentrated solutions, use %vista:s Chemdose program.

."ere3

4'ample3

.ater Mat"
6uestion3 The water temperature is ?.FF. !hat is the temperature in FCV %ns,er3 0P.P?FCV

6uestion3 The water temperature is 1BFC. !hat is the temperature in FF %ns,er3 BEFF

'ou:ve been advised to in;ect - ppm of 3itec -... into a system feed flow of 6uestion3 1.. gallons per minute. The 3itec -... weighs 1../? pounds per gallon. !hat ml>min dose rate is re8uiredV %ns,er3 ..E.0 m >min

6uestion3 % storage tan9 contains 1. feet of water. !hat is the head pressure at the bottomV %ns,er3 /.-/ )&"

6uestion3 Convert 0<B ))( of hardness into grains per +& gallons. %ns,er3 1P..?

6uestion3 Calculate the volume of neat chemical re8uired to ma9e B.. gallons of a -M solution of chemical. *ensity of neat chemical is E.P pounds per gallon. %ns,er3 1- gallons

Scaling and Antiscalants

Scaling means the deposition of particles on a membrane, causing it to plug. Without some means of scale inhibition, reverse osmosis (RO) membranes and flow passages within membrane elements will scale due to precipitation of sparingly soluble gas, such as calcium carbonate, calcium sulfate, barium sulfate and strontium sulfate. Most natural waters contain relatively high concentrations of calcium, sulfate and bicarbonate ions. n membrane desalination operations at high recovery ratios, the solubility limits of gypsum and calcite e!ceed saturation levels leading to crystalli"ation on membrane surfaces. #he surface bloc$age of the scale results in permeate flu! decline, reducing the efficiency of the process and increasing of operation costs. #he effects of scale on the permeation rate of RO systems is illustrated in the following figure. %ollowing an induction period, plant flow decrease rapidly. #he length of this period varies with the type of scale and the degree of super saturation of the sparingly soluble salt.

&s it is evident from the graph, the induction period for calcium carbonate is much shorter than that for sulfate scales, such as calcium sulfate. t is economically preferable to prevent scaling formation, even if there are effective cleaners for scale. Scale often plugs RO element feed passages, ma$ing cleaning difficult and very time consuming. #here is also the ris$ that scaling will damage membrane surface. #here are three methods of scale control commonly employed'

acidification ion e!change softening antiscalant addiction.

Acidification' acid addiction destroys carbonate ions, removing one of the reactants necessary for calcium carbonate precipitation. #his is very effective in preventing the precipitation of calcium carbonate, but ineffective in preventing other types of scale. &dditional disadvantages include the corrosivity of the acid, the cost of tan$s and monitoring e(uipment and the fact that acid lowers the p) of the RO permeate. Ion exchange softening' this method utili"es the sodium which is e!changed for magnesium and calcium ions that are concentrated in the RO feed water, following the chemical e(uations' *a+, , +-a. /0 +-a, , *a.+ Mg+, , +-a. /0 +-a, , Mg.+ (-a. represents the sodium e!change resin). When all the sodium ions have been replaces by calcium and magnesium, the resin must be regenerated with a brine solution. on e!change softening eliminates the need for continuous feed of either acid or antiscalant.

Antiscalants' they are surface active materials that interfere with precipitation reactions in three primary ways'

#hreshold inhibition' it is the ability of an antisclant to $eep supersaturated solutions of springly soluble salts. *rystal modification' it is the property of an antiscalants to distort crystal shapes, resulting in soft non adherent scale. &s a crystal begin to form at the submicroscopic level, negative groups located on the antiscalant molecule attac$ the positive charges on scale nuclei interrupting the electronic balance necessary to propagate the crystal growth. When treated with crystal modifiers, scale crystals appear distorted, generally more oval in shape, and less compact. 1ispersion' dispersancy is the ability of some antiscalants to adsorb on crystals or colloidal particles and impart a high anionic charge, which tends to $eep the crystals separated. #he high anionic charge also separates particles from fi!ed anionic charges present on the membrane surface.

1ispersancy #hreshold Mechanism 1uring the past two decades new generations of antiscalants have emerged commercially, in which the active ingredients are mostly proprietary mi!tures of various molecular weight polycarbo!ylates and polyacrylates. *alculation procedures e!ist for predicting the li$elihood of scale formation. 2se of these predictors depends upon an up3to3date water analysis and a $nowledge of system design parameters. #he ions contained in the feed water concentrate though the RO system, the point of ma!imum scale potential is the concentrate stream. &ntiscalant type and dosage is therefore based upon the mineral analysis at this point. t is important to find the optimi"ation of antiscalant treatment with respect to type and dosage, identifying the proper antiscalant to use and the dosage3induction type relationship for the e!tended level of super saturation. 4enntech can help you in the selection of the best antiscalant for your particular application. Economical analysis &cid addition is not very cost effective because of the cost of acid, tan$s and monitoring e(uipment. 2nless removed by degasification, e!cess of carbon dio!ide contained in the

permeate of acid3fed systems increases the cost of ion e!change regeneration. &ntiscalants are relatively cheap products and have no additional costs. When compared to either acid or antiscalant addition, the main disadvantage to softening is cost factoring in e(uipment costs. #hrough a present worth analysis there is no level of hardness in which softening competes economically with antiscalants addition. #he following table gives a cost comparison between softening and antiscalant treatment options for different levels of hardness, on a basis of a RO system designed to produce 56 gpm (75 m89h) of permeate at 56: recovery.

Read more' http'99www.lenntech.com9antiscalants.htm;i!""+<#ua&=7W

Pre$enting Scale Formation in #e$erse Osmosis S&stems

Posted on Sune 1?, 0.10 by admin

By Wes Byrne

Reverse osmosis 4RO5 systems behave similar to a boiler or a cooling tower in that they cause the dissolved salts in their ma9eup water to become more concentrated. %s pure water permeates the RO membrane, the salts are left behind in a concentrated stream. "f the solubility of any particular salt is e2ceeded, the potential e2ists that the salt will form scale directly on the membrane surface or possibly within the flow channels through the membrane element. This scale formation will occur first in the tail end of the RO system, within the last elements through which the water flows before e2iting the system in its concentrate stream. The scale will cause the RO system to re8uire increased pressure to achieve the same permeate flow rate, and the permeate conductivity may increase in the tail end of the system. Restoring performance will re8uire cleaning with an acidic solution, although membrane cleaning is only going to be effective with certain types of scale. "t is unli9ely to be fully effective with silica scale, or with most of the sulfate salts. "f cleaning cannot restore

performance, membrane replacement becomes the only option for restoring the lost RO performance. For most water sources, preventing scale formation is not a necessarily difficult challenge. "t can usually be accomplished by applying one or more of the following three options to the RO feed stream, 1. %cid in;ection 0. !ater softening -. "n;ection of a scale inhibitor The specific re8uirements for the water source can be predicted using +.&. !ater &ervices: &cale "nhibitor )ro;ection )rogram 4&"))5, which ta9es into account the particular ma9eup of the water source, the water temperature and pH, and the salt concentration factor relative to the desired permeate recovery of the RO system. The right method will often depend on the priorities of the application. %cid in;ection is effective at preventing calcium carbonate scale, but it is only marginally effective at preventing sulfate or silica scale. % ma;or disadvantage of acid in;ection for many applications is it results in the formation of carbon dio2ide that will readily permeate the RO membrane. "t may then pose a removal burden for downstream ion e2change or deaeration e8uipment. ime softening will reduce the water hardness, al9alinity, and dissolved silica to a point that an RO unit might be able to operate at a low permeate recovery with reduced potential for scale formation. However, increased permeate recovery that ta9es full advantage of the RO feed pressure and minimi=es water waste will re8uire that the water hardness be reduced to a much greater e2tent. This then would re8uire =eolite softening of the RO feed water, which uses a strongLacid cation e2change resin that is regenerated using sodium chloride. %n advantage of this approach is it enables caustic 4sodium hydro2ide5 to be in;ected upstream of the RO system to raise its feed water pH, which will result in increased removal of al9alinity, dissolved silica, and boron. Operation at elevated pH will also reduce the potential for the formation of silica scale. The most cost effective method for preventing scale formation 4where applicable5 is in;ecting a scale inhibitor. +sing similar chemistry to the antiscalants used in boilers and cooling towers, the formation of scale crystals can be slowed sufficiently to allow those crystals to e2it the RO system in its concentrate stream. "n this manner, saturation points may be safely e2ceeded for calcium carbonate and sulfate salts as long as the RO system is operating. !hen the RO unit shuts down, it is critical that the system be flushed of

supersaturated salts using permeate water, or by allowing low pressure feed water to displace the water held up within the system. &ome of the scale inhibition products used in the early years of the RO industry consisted of a single polymer that tended to lose its solubility when in;ected into certain water sources, such as ones that contained iron. This would result in heavy fouling of RO prefilter cartridges, as well as of the membrane elements. )roducts such as +& !ater:s RO B.- and B./ are blend products that contain both polymers and phosphonates can be utili=ed in the RO systems. They wor9 synergistically to improve their solubilities, while providing superior prevention against scale formation. "n fact, they will also assist in 9eeping iron in suspension to reduce its potential for fouling the RO system. The appropriate dosage can be determined using the &")) program. !ith the correct application of the chosen scale prevention method, most RO systems should never have to e2perience scale formation.

Membrane autopsy results from the Genesys Membrane laboratories in Madrid indicate that 35% of the foulants identified on reverse osmosis (RO membrane surfaces durin! autopsy are or!anic in nature containin! si!nificant levels of biomass"
Aiofouling in RO systems can be defined as the growth of biomass on a membrane surface which is sufficient to cause operational problems. The effects of biofouling will show as an increase in differential pressure 4d)5, with a conse8uent reduction in flu2, and increase in pumping costs. &evere biofouling may result in membrane degradation and physical damage. *uring operation these symptoms 4d)5 are usually observed in the lead elements of the system, closest to the source of contamination, however telescoping and compaction may occur in the rear elements *uring offLline periods biofouling can occur throughout the entire plant if shutdown procedures are not followed correctly. Conse8uently on start up problems of reduced flu2, increased d) and salt passage may occur across the entire system. %t the RO design stage analysis of feed water, &*" tests and ionic water analyses indicate the colloidal fouling and scaling potential of the water. ess thought is given to the microbiological content of the water and the conse8uent biofouling potential of the operational system. This is usually only given consideration when membrane performance is affected.

Every RO system has a different biofouling potential. The feed water source, preLtreatment 4both chemical and physical5, maintenance, cleaning regime and system downtime all contribute to the biofouling potential of a system. &ite specific financial and operational factors must be considered when selecting a suitable method of control. "t is therefore difficult to recommend a QsingleQ best treatment regime for all biofouling issues. Ta9ing this into account the ob;ective of this application guide is to give an overview of the most common methods of control and also the available methods for determining the presence of biofouling within a membrane system. Identification3 The definitive method for identifying the presence and nature of biofoulants on a membrane is via autopsy. Care must be ta9en to select the correct element for autopsy with biofouling usually occurring in the first elements of a system. "n order to enable representative samples to be ta9en the membrane must be pac9aged correctly 4free of &A& and>or biocides5 and despatched within 0/ hours of removal from the plant. &E(LE*%J 4&canning Electron (icroscopy L *ispersive JLray %nalysis5 and microbiological identification and counting techni8ues are used to e2amine the membrane surface for the presence of microorganisms. Contamination source3 &imilar procedures applied to preLtreatment cartridge filters or ../B Wm &*" papers can be used to evaluate the source of the contamination. This is a suitable method for determining the presence and type of fouling 4by colloidal mater and some metal o2ides5 in the feed water, but does not give an indication of the degree of fouling affecting the membrane elements. %lso bacteria counts cannot be 1..M reliable as in some cases there may not be bacteria present in the feed water, but there are nutrients which will assist the bacteria growing on the membranes. Aiofouling Control L Chemical preLtreatment OnL ine Chlorination X Halogens 4#aClOTCl0 gasTClO0, chloramines, Aromine, etc5 +se of halogens is limited due to the potential o2idation of polyamide membranes. The effectiveness of chlorine depends mainly on feed water pH and e2posure time 4usually 0.L-. minutes of contact time is re8uired5. The o2idation reaction is catalysed by the presence of trivalent cations 4Fe-T>%l-T5 common to RO feed waters. Typically on surface waters chlorine is added continuously to the inta9e to give a free residual chlorine concentration of ..BL1..mg>l. "n order to protect polyamide membranes from o2idation deLchlorination is re8uired upstream of the membranes using sodium

bisulphite solution 4$enesys RE*5 or activated carbon filtration. "t is recognised that chlorine o2idi=es and brea9s down natural organic matter 4#O(5 present in the feed water to more easily biodegradable byLproducts providing a nutrient source to microLorganisms. "n addition as no chlorine is present on the membrane surface biofilm growth can occur re8uiring more fre8uent saniti=ation Off line chlorination 4system saniti=ation5 This process is designed to limit the amount of bacterial ingress into the membrane system from contaminated preLtreatment systems. "n order to saniti=e the preLtreatment and distribution system chlorine is applied at regular intervals determined by the degree of fouling. Feed water is sent to drain before reaching the membranes and OR)>Redo2 is used to measure chlorine levels before returning to normal service. This method is aimed at saniti=ing the preLtreatment system reducing bacterial contamination into the RO and as such has no direct effect on e2isting biofilm growth on the membrane surface. This process should therefore be used in con;unction with approved nono2idising biocides such as $enesol -.. !hile o2idising biocides are unable to control bacteria on the membrane surface $enesol -. has the advantage that it can reach the membrane surface 9illing the bacteria with no detrimental effect on the polyamide structure. The dose and fre8uency of application is determined by the degree of bacterial ingress to the system and the conse8uent effects on operating parametersY flu2, d) and salt re;ection. Hydrogen )ero2ide X )eracetic %cid and formaldehyde Hydrogen pero2ide or a combinationof hydrogen pero2ide>peracetic acid and also formaldehyde are sometimes used as a method of disinfecting RO membranes. However there is a danger of o2idation of polyamide membranes which can be further catalysed by the presence of iron and manganese which limits the practical application of this method. 5on-o'idising biocides *ue to the potential damage to polyamide RO membranes caused by o2idising biocides it is often preferable to use nonLo2idising biocides to control biofouling. These biocides are applied through normal dosing pump or introduced via C") tan9 and do not re8uire additional QneutralisingQ treatment or carbon filtration $enesol -., fast acting biocide, OffLline &aniti=ation, dose at -..mg>l for periods up to P. minutes. OnLline application, in nonLpotable applications $enesol -. dose at /..mg>l for -. minutes.

$enesol -0, membrane preservative, Off line &aniti=ation, dose at 1,...L1,B..mg>l 4..1L..1BM5 for a period of PL? hours followed by an al9aline clean. OnLline *osage, dose at <..L1...mg>l 4...<L..1M5 for ?L10 hours once every BL01 days depending on the severity of the fouling. "n cases of severe fouling or for seasonal operation at higher ambient temperatures it can also dosed at BLB.ppm continuously. $enefloc %AF is a flocculant with biocidal and algaecidal properties. % dosage of 0L1.mg>l can reduce the silt density "nde2, replace iron and aluminium coagulants and inhibit algae and bacteria growth in prefilters, cartridges and membranes. )reventative treatment L membrane cleaning, "n reality biofouling in an RO system is successfully controlled using a combination of preventative biocide treatment in con;unction with a well designed and effective membrane cleaning programme. The e2tremes of pH combined with surfactants used in cleaning will 9ill the ma;ority of bacterial cells, However inefficient cleans may not remove the remaining nutrient rich biomass from the system giving any remaining viable cells the opportunity to rapidly multiply into viable populations. % well designed and applied cleaning regime will remove the biomass from the system preventing rapid reLgrowth. +sing different chemicals will help to avoid resistance to a particular biocide developing within the microbial population. Efficient removal of biofilm is achieved using al9aline cleaning chemicals and depends on temperature, pH, contact time and good cleaning practice. "n addition combining al9aline detergents and biocide cleaning steps will improve santitisation. OffLline preservation, %ll membrane environments contain a viable population of microorganisms which will grow and degrade the membranes if the system is not treated correctly on shut down. "t is recommended by membrane manufacturers that the system must firstly be flushed with permeate to remove highly concentrated water and avoid mineral deposition. %ny plant that will be stopped for over 0/ hours should be treated to limit potential microbiological fouling. 5on-o'idising biocide3

$enesol -0 for membrane preservation and storage, 0/L-P hours -..mg>l N ...-M -PL1P? hours B..mg>l N ...BM 1L/ wee9s ?..mg>l N ...?M 1LP months 1...mg>l N ..1M ZP months drain and refill every P months &odium Aisulfite 4&A&5 L $enesys RE*, +sed as a method for inhibiting growth of aerobic bacteria at a concentration of 1M. Care must be ta9en to avoid the ingress of air into the system if using this method. "t must be noted that &A& is not a biocide it wor9s by removing o2ygen from the water, so aerobic bacteria will not survive. "t has no effect on anaerobic bacteria and the solution must be refreshed regularly when it becomes e2hausted. Conclusions %ll a8uatic environments have a bacterial population and if not controlled in an RO plant biofouling may cause significant operational problems with a conse8uent impact on costs and performance. "ncreased cleaning fre8uency, increased pumping costs, reduced permeate 8uality and decreased membrane life ma9e good bacterial control essential.%ll systems are different and it is therefore difficult to assess the fouling potential. % combination of biocide application with thorough cleaning practice is an efficient way to control microbial activity and ensure efficient plant operation. %lways follow membrane manufacturers[ guidelines with respect to application of o2idising biocides. Correct preservation of offLline systems is vital to ensure that membranes do not degrade during prolonged shut down and that the system is not repopulated on restart.

Introduction to Reverse Osmosis

A technique used in processes requiring high-quality, purified water, such in semiconductor processing or biochemical applications, is reverse osmosis. It can be used to treat boiler feedwater, industrial wastewater, or process water. Reverse Osmosis is a water purification technique that reduces the quantity of dissolved solids in solution ( ucera, !"#. It was first developed in the $%!&'s by the () government to provide fresh drin*ing water for the +avy, and since then, advances have made it much more feasible for obtaining purified water from wastewaters produced in many industrial applications. RO uses waterline pressure to push raw wastewater against a special semipermeable membrane. It is essentially a molecular squee,ing process which causes -.& molecules to separate from the contaminants. /he separated water molecules then pass thru to the inside of the membrane on to a holding reservoir. /he contaminants are washed from the membrane and disposed of. Recently, RO has been used in treating boiler feedwater, in addition to industrial and process wastewaters. 0oilers are found throughout the chemical processing industry and the primary method to treat boiler wastewater is an ion-e1change based deminerali,ation. -owever, RO has been demonstrated to be more cost effective than this deminerali,ation process ( ucera, !"#.

Example of RO system

Problems With Reverse Osmosis:

It is necessary to establish feedwater quality guidelines to optimi,e system performance and prevent the three main problems associated with RO2 scaling, fouling, and degradation of ROmembranes ( ucera !!# /hese problems tend to decrease system productivity because they reduce wastewater purity. )caling occurs on RO membranes when the concentration of scale-forming species e1ceeds saturation, producing additional solids within the RO feedwater. )calants include such chemical species as calcium carbonate, calcium sulfate, barium sulfate, strontium sulfate, and reactive silica ( ucera !!#. )ince these species have very low solubilities, they are difficult to remove from RO membranes. )caling decreases the effectiveness of the membranes in reducing the solids and causes more frequent cleanings. A scale on a membrane provides nucleation sites that increase the rate of formation of additional scale ( ucera !!#.

Methods to minimize scaling

In order to minimi,e scaling, pretreatment methods involving chemical or ion e1change techniques are used. Ion e1change methods remove scale-forming species from the RO feedwater, while chemical techniques change the characteristics of the RO

feedwater so that crystal formation is not favored. An e1ample of a chemical technique to prevent fouling is lime softening, which involves chemical processes that reduce the hardness of the wastewater, essentially preventing material from precipitating out. ( ucera !3# 4ime, soda, ash, and +aO- are used to convert soluble calcium and magnesium to insoluble calcium carbonate and magnesium hydro1ide. 5agnesium hydro1ide tends to absorb silica, another scalant. /hese solids are then collected as sludge from the bottom of the 6softener6. Another softening procedure involves ,eolite in an ion e1change process. A strong acid cation resin in the sodium is used to remove scale-forming cations, suchas calcium, magnesium, barium, and iron. ( ucera, !3# /hese cations are e1changed with the sodium to yield 6soft water6, that is, water of low hardness. Another pretreatment technique to prevent scaling is acidification, which specifically reduces the crystalli,ation of calcium carbonate. )ulfuric acid is most commonly used in this process, but can often increase the formation of sulfate scales. /herefore, where sulfuric acid cannot be used, hydrochloric acid is substituted. ( ucera, !7#. Often used with acidification, or by itself, are antiscalants. Antiscalants are chemicals added to wastewater to minimi,e scale carbonate or sulfate based scale (!7#. /hey consist of acrylates and phosphonates which inhibit the precipitation of carbonate or sulfanates.

Methods to prevent fouling

/he second problem with reverse osmosis is with the fouling of membranes. 8ouling occurs when suspended solids, microbes and organic material deposit on the surface of the membrane. )oluble heavy metals, such as iron, can be o1idi,ed within membrane modules and foul the membranes. Another problem is from colloidal sulfur, which when o1idi,ed from -.) can foul RO membranes. ( ucera !!#. 9olloidal sulcar tends to be very stic*y and therefore

can attach easily to the surface of RO membranes -ydrogen sulfide would be found most commonly in well-water. /he primary methods used to combat fouling are mechanical processes that physically remove the suspended solids or chemical treatments the deactivate the foulant. 9oagulation is one technique that neutrali,es the negative surface of the suspended solids, allowing the particles to cometogether. ( ucera !7# /hese large particles are then easy to remove from the water using filtration. /he most common coagulants used are cationic polymers, inorganic salts, and aluminum and iron salts. Inorganic solvents tend to form large particles, while catonionic polymers require much less product for coagulation. )imilar to coagulation is the clarification method, which destabli,es suspended particles through charge neutrali,ation (!:#. /hese particles conglomerate and are removed using sedimentation or filtration techniques. One particular type of filtration uses manganese greensand as a filter to remove soluble iron and manganese from the water source. /his is generally done by o1idi,ing iron and manganese and physically removing the precipitates in the manganese greensand bed. 9hlorination is the primary technique to minimi,e microbiological foulants, as it is very effective against a wide variety of microbes and can be easily deactivated using sodium metabisulfite. (!%#. After chlorination, activated carbon filters can be used to remove chlorine and reduce organics. -owever, activated carbon tends to foster microbial growth by providing nutrients for microbes, so it is not a very effective filtration technique. 8inally, to treat -.) containing feedwater, which can form colloidal sulfur, a combination the above techniques is used. 8irst, the water is o1idi,ed to precipitate the sulfur, which is then coagulated and filtered. Any colloidal sulfur that may have formed is converted to thiosulfates with the addition of sulfite. 8inally, chlorination is done to convert the thiosulfates to sulfates.(3&#

Methods to Minimize Membrane

egradation

/he final problem with reverse osmosis is membrane degradation. It occurs when the membranes are e1posed to conditions that destroy the polymers used to create the membranes. )ome membranes are susceptible to hydrolysis at high and low p-, while others are degraded by e1posure to o1idi,ers such as chlorine. ( ocera !3# /o prevent membrane degradation by acidic or al*aline waters, a corrective amount of acid of base should be added to the feedwater to ma*e the p- appro1imately neutral. /o prevent o1idation reactions, dechlorination is used. (!7# 0ac* to the 5icrobiology and 5embrane /reatment of ;astewater -omepage
References
6 ucera, <ane6. Properly Apply Reverse Osmosis 9hemical =ngineering >rogress. 8ebruary $%%7. >gs !"-3$.