desalination process engineering part ii · 2018. 3. 22. · mark wilf ph. d. phone: +1 858 444...

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology

E-mail: [email protected]

Webpage:www.rotechnology.net

138

Desalination Process Engineering

Part II

Mark Wilf Ph. D. 2018

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



139

5. RO System

5.1. Membrane elements and pressure vessels

The basic blocks of RO membrane unit are membrane elements. The membrane elements are

highly standardized. They have the same outside dimensions and in most cases identical

dimensions of the connecting permeate ports. They are interchangeable.

Regardless of type of elements they look practically the same. The dimensions of commercial

elements are 200 mm in diameter and 1000 mm long as shown on Figure 5.1.1. Membrane area

of such element is 37 m2 – 41 m2.

Depending of application and element type elements of the same membrane area will produce

different quantity of permeate in field conditions as indicated by permeate capacity numbers in

Figure 5.1.1.

1 m,

40 “

200 mm,

8”

37.2 m2 (400 ft2)

Brackish application:

~24.2 m3/d (~6,400 gpd)

Wastewater reclamation:

~18.2 m3/d (~4,800 gpd)

Seawater application:

~12.8 m3/d (~3,400 gpd)

Second pass RO:

~30.3 m3/d (~8,000 gpd)

Figure 5.1.1. 200 mm by 1000 mm spiral wound element and corresponding product capacity

values in various applications.

In addition to 100 mm and 200 mm diameter elements, larger, 400 mm diameter elements are

being introduced recently. The relative size and corresponding permeate capacity is shown on

Figure 5.1.2.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



140

8” and 16” diameter elements

8” element

Membrane area

40m2 (430 ft2)

Nominal flow

45 m3/day (12,000 gpd)

Avg. field flow

24 m3/day (6,500 gpd)

Nominal salt rejection 99.8%

16” element

Membrane area

160 m2 (1,700 ft2)

Nominal flow

180 m3/day (47,000 gpd)

Avg. field flow

95 m3/day (25,000 gpd)

Nominal salt rejection 99.8%

Figure 5.1.2. 200 mm and 400 mm diameter, 1000 long spiral wound membrane elements.

In order to be able to apply feed pressure to feed solution and continently separate and collect

permeate and concentrate streams, the elements have to be enclosed in pressure vessels.

The basic configuration of elements in pressure vessel is shown on Figure 5.1.3.

Figure 5.1.3. Configuration of pressure vessel with membrane elements.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



141

Membrane units in commercial desalination systems are configured as having between 6 – 8

elements per vessel. The number of elements per vessel is determined based on number of

concentrate stages and pressure drop along the membrane unit.

In majority of cases, larger number of elements per vessel will result in lass expensive membrane

unit and smaller footprint.

The distribution of recovery rate in seawater membrane unit having 6, 7 and 8 elements per

vessel is illustrated in Table 5.1.1.

Table 5.1.1. Recovery rates of individual elements in pressure vessel according to number of

elements per vessel.

Element position 6 elements/vessel

Recovery rate, %

7 elements/vessel

Recovery rate, %

8 elements/vessel

Recovery rate, %

1 16.1 15.1 13.2

2 15.3 12.9 12.5

3 13.5 11.0 11.1

4 11.3 8.3 8.9

5 10.2 6.8 7.8

6 8.5 4.9 6.3

7

5.1 4.5

8

2.3

The results listed in Table 5.1.1 were calculated for a membrane unit operated at recovery rate of

50%. It is evident from the above results that larger number of elements per vessel will results in

more gradual reduction of recovery rate along the pressure vessel.

As in each case the total system permeate capacity and the total membrane area in the system are

the same, the average permeate flux in the system will be the same regardless number of

elements per vessel.

The systems with different number of elements per vessel are designed to operate at the same

feed salinity and recovery rate. The feed pressure and concentrate pressure will be very similar in

each case. Therefore, the net driving pressure (NDP) and permeate flux distribution in the

pressure vessel will be the same, as shown on Figure 5.1.4.

Large capacity seawater RO desalination systems in configuration of eight elements per vessel

have been operating successfully. Some of the systems have close to 10 years on line operating

record.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



142

Figure 5.1.4. Flux distribution along the length of pressure vessel for 6,7, &8 elements per vessel

membrane unit configurations..

In the same way as membrane elements, internal dimensions of pressure vessels are highly

standardized and the outside dimensions and configurations are similar.

The dimensions of pressure vessels for 100 mm and 200 mm diameter membrane elements are

summarized in Table 5.1.2.

Table 5.1.2. Example of representative dimensions of commercial pressure vessels for RO

application.

Element type - elements number/PV

Inside diameter, mm (in)

Outside diameter, mm (in)

Length, mm (in)

8040 – 1 204 (8.02) 259 (10.2) 1478 (58.2)

8040 – 2 204 (8.02) 259 (10.2) 2494 (98.2)

2468

1012141618202224262830323436

0 0.2 0.4 0.6 0.8 1 1.2

Ave

rag

e e

lem

en

t fl

ux

, l/

m2

-hr

Fraction of pressure vessel length

Average flux distribution in RO seawater systemMediterranean seawater, 50% recovery, avg flux 14.3 l/m2-hr

8/PV

7/PV

6/PV

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



143

8040 – 6 204 (8.02) 259 (10.2) 6558 (258.2)

8040 – 7 204 (8.02) 259 (10.2) 7574 (298.2)

8040 – 8 204 (8.02) 259 (10.2) 8590 (338.2)

4040 – 1 105 (4.13) 127 (5.0) 1194 (47)

4040 – 2 105 (4.13) 127 (5.0) 2210 (87)

4040 – 4 105 (4.13) 127 (5.0) 4242 (167)

5.2.Membrane unit configuration

In membrane unit pressure vessels are arranged in array of parallel units, connected together

through the corresponding ports to feed, permeate and concentrate manifolds.

5.2.1. Single stage and multistage

RO membrane systems are configured as a single stage unit in seawater applications due to low

recovery rate, usually not exceeding 50%.

In a single stage unit all pressure vessels form a single parallel grid, connecting feed and

concentrate manifolds, as shown on Figure 5.2.1. Accordingly feed water passes only once

through the array of pressure vessels. After the passage through the array of pressure vessels,

feed water becomes concentrate and it is discharged form the system, usually back to the body of

water where the feed water was pumped from (ocean). The permeate is collected from individual

pressure vessels into a single manifold connected to the product water system.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



144

FeedConcentrate

Permeate

Concentrate

Permeate

Figure 5.2.1. Single stage membrane unit configuration (Courtesy CH2M Hill)

The recovery rate of a single stage unit configuration is usually do not exceed value of about

65% value due to limitation of a minimum concentrate flow. For this reason brackish and

nanofiltration systems, that operate at higher recovery rates are configured as two, three or even

four concentrate stages units.

A two stage unit configuration is shown schematically on Figure 5.2.2.

The pressure vessels are arranged in array of 4:2. The 2:1 ratio of pressure vessels between

consecutive stages is related to the objectives of maintaining similar feed cross flow rate in

pressure vessels along the system. Approximately in each stage about 50% of feed water is

converted to permeate. Therefore, only half of the original feed flow rate is available for the next

stage. Accordingly, the number of pressure vessels should be half of the number of pressure

vessel in the proceeding stage.

As shown on Figure 5.2.2 the concentrate from pressure vessels in the first stage is collected in

common concentrate manifold and flows according to hydraulic pressure gradient as a feed to the

second stage. There are no flow or pressure regulating valves between the stages.

Mechanical drawing of RO membrane train is shown on Figure 5.2.3. This drawing corresponds

to pressure vessel array of 36:18. The drawing includes two view of membrane unit.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



145

Figure 5.2.2. Schematic diagram of a two stage membrane unit. Pressure vessel array 4:2.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



146

Figure 5.2.3. Mechanical drawings of RO membrane train. Pressure vessels array 36:18.

The feed side of the membrane unit is shown as View A – A . The concentrate side of the

membrane unit is shown as View B – B. The B – B view illustrates the common approach to the

design of interstage piping connections when no iterstage booster pump is being used. The

concentrate collecting manifolds of the first stage are connected directly to the feed manifold of

the second stage.

5.2.2. Sideport, multiport and center port configuration

The membrane unit shown on Figure 5.2.3. utilizes side port type pressure vessel, shown on

Figure 5.2.4. This type of pressure vessel have only two high pressure inlet – outlet ports: one for

feed and the second for concentrate.

In a never pressure vessels design (called multiport configuration) the number of high pressure

ports is increased to four: two feed ports and two concentrate ports. The two ports (feed or

concentrate) are on opposite sides of pressure vessels – 180 degree apart.

The multiport configuration enables connecting number of pressure vessels together through the

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



147

corresponding ports, therefore reducing number of central distributor piping.

Comparable configurations of membrane units using side port and multiport pressure vessels for

single stage and two stage units, is shown schematically in Figures 5.2.5 – 5.2.8.

The number of pressure vessels that can be connected together is determined by the port size and

limit of pressure drop in the feed/concentrate ports or limit of maximum flow velocity through

the ports

Figure 5.2.4. Configuration of a side – port pressure vessel.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



148

320 640

980

360

2850

2050

Figure 5.2.5. Single stage membrane unit with sideport pressure vessels.

2050 (81")

2180 (85")

Figure 5.2.6. Single stage membrane unit with multiport pressure vessels.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



149

980 (38.5")

360 (14")

2850 (112")

2050 (81")

Figure 5.2.7. Two stage membrane unit with sideport pressure vessels.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



150

Figure 5.2.8. Two stage membrane unit with multiport pressure vessels.

The connection of pressure vessels can be in any direction, both horizontal and vertical. An

example of a vertical connection of multiport pressure vessels is shown on Figure 5.2.9

Figure 5.2.9. Unit with multiport pressure vessel in vertical flow configuration (courtesy PUB,

Singapore)

The limits and corresponding ports diameter is provided by the manufacturers of pressure

vessels. The usual limit is maximum pressure drop of up to 0.2 bar and/or maximum flow

velocity of 3 m/sec.

The general approach to assure good flow distribution is that the pressure drop along the pressure

vessel should be significantly higher than the pressure drop in the ports.

The feed and concentrate connection to the grid of parallel connected array of multiport pressure

vessels could be configured in a number of ways. The proffered connection is the “Z”

configuration, however the “U” configuration (feed and concentrate connected to the same side

of the array) is also used quite frequently. In this configuration the feed enters at one corner of

the array and the concentrate is collected at the opposite corner.

When designing and construction a membrane unit that utilizes multiport pressure vessels it is

important to remember that the pressure vessels should not have unconnected ports.

Unconnected ports will create pockets of stagnant water that eventually could become centers of

development of biological activity.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



151

Center port pressure vessel concept has been developed in Netherlands by Vitens Company. The

center port configuration, shown on Figure 5.2.10. allows entry of feed water through two ports

at the end of pressure vessel and collection of concentrate through one port in the center of

pressure vessel.

Figure 5.2.10. Schematic configuration of membrane unit utilizing center port pressure vessels.

The benefit of center port configuration is lower pressure drop in the membrane unit as

illustrated in Table 5.2.1..

According to the comparison included in Table 5.2.1 the center port configuration can provide

about 15% reduction of feed pressure, which translates directly to saving of pumping energy.

Table 5.2.1.. Comparison of side port and center port configuration

Configuration Side port Center port

Number of stages 2 2

Pressure vessels array 40:20 34:18

Elements/vessel 7 8

Total number of elements 420 416

Recovery rate 85% 85%

Average flux rate 24.5 l/m2/hr 24.5 l/m2/hr

Feed pressure 9.0 bar 7.8 bar

Feed pressure reduction 15%

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



152

The reduction of feed pressure is related to reduction of pressure drop due to reduction of

average feed flow rate by half and the same rate of reduction of the path length.

Application of center port configuration increases significantly concentration polarization as the

stage recovery rate is achieved in a path length of 4 elements in place of 6 or 7 in the

conventional configuration.

So far the center port configuration has been applied in number of small nanofiltrtaion unit is

Netherland. However, more recently a 55,000 m3/day nanofiltrtaion plant with center port

pressure vessels has started operation at the town of Jupiter, Florida.

5.2.3. Two pass, partial two pass, split partial

Second pass processing of permeate from the first pass unit is implemented when permeate

salinity, or specific constituent(s), produced by the first pass unit exceed the specified values.

Schematic configuration of a two pass system is shown on Figure 5.2.11.

Two Pass RO System

PG

PG

Pressure vessel,

1st pass

Pressure vessel,

1st pass

Pressure vessel,

1st pass

Pressure vessel,

1st pass

Pressure vessel,

2nd pass

Pressure vessel,

2ND pass

Concentrate

pass one

Permeate

pass two

Feed

Permeate

pass one

Concentrate

pass two

FI

FI

FI

Figure 5.2.11. Schematic configuration of a two pass unit.

As shown in the above diagram, first pass permeate is collected and pumped, using a high

pressure pump, as feed water to the second pass unit. The concentrate from the second pass unit

is usually returned to the suction of the first pass high pressure pump. This configuration reduces

the need for additional feed water and additional pretreatment equipment.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



153

Application of two pass system requires additional RO equipment and also reduces overall

system recovery rate, as illustrated on Figure 5.2.12. The necessary additional equipment

includes equipment of the second pass unit and also equipment to increase permeate production

capacity of the first pass unit that will be lost with the second pass concentrate. The required

increase of the first pass unit permeate capacity is direct function the recovery rate of the second

pass unit.

Frequently, the second pass processing of the full flow from the first pass is not required and

partial two pass processing is implemented. Schematic configuration of a partial two pass system

is show on Figure 5.2.13.

Two Pass RO System

10045

5

R1 = ?

R2 = ?

Rt = ? (w/o recirculation)

Rt = ? (w recircirculation)

50

50

95

100*50/100 = 50%

100*45/50 = 90%

100*45/100 = 45.0%

100*45/95 = 47.4%

Figure 5.2.12. Two pass system with second pass concentrate recirculation

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



154

Combined permeate

1st

Pass RO 2nd

Pass RO

Figure 5.2.13. Schematic configuration of a partial two pass processing

The fraction of first pass that will be processed is calculated based on salinities of the first pass

permeate, salinity of the second pass permeate and the required salinity of the product water

(after blending). Due to possible fluctuations of feed water temperature, salinity and membrane

condition, the permeate salinities from the first and the second pass units may fluctuate as well.

Accordingly, the first and second pass units are sized based on the most extreme conditions of

feed water, as defined in project specifications.

The partial two pass processing is effective way of improving product water quality, requiring

smaller additional equipment and lower operating cost than the full two pass system.

The partial two pass processing can be optimized further, applying modification of this process

called “split partial” two pass processing.

The split partial processing takes advantage of internal permeate salinity distribution along the

pressure vessel, shown schematically on Figure 5.2.14.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



155

Combined salinity in pressure vessel

0

100

200

300

400

500

1 2 3 4 5 6 7 8

Element position

Sali

nit

y p

pm

TD

S

Element position in pressure vessel

1 2 3 4 5 6 7 8

Figure 5.2.14. Permeate salinity distribution along the pressure vessel.

The permeate salinity distribution, shown on Figure 5.2.14, is result of two processes:

- Increase of feed salinity along the pressure vessel, therefore increased salinity gradient.

- Increase of feed osmotic pressure, therefore lower NDP and lower permeate flux along

the pressure vessel.

The split partial design utilizes this phenomenon in a configuration shown in Figure 5.2.15.

In the split partial configuration, high salinity permeate, collected from the concentrate end of the

membrane unit, is processed with the second pass system. The low salinity permeate, collected

from the feed end of the membrane unit is used for blending.

The split partial design results in significan reduction of the second pass processing required.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



156

Low salinity permeate

High salinity

permeate

Combined permeate

1st

Pass RO 2nd

Pass RO

Figure 5.2.15. Split partial two pass configuration.

An example of potential saving possible with split partial configuration, compared to a

conventional partial two pass system is illustrated in Table 5.2.2.

Table 5.2.2. Comparison of conventional partial two pass and split partial two pass configuration

Configuration Conventional partial two pass Split partial two pass

First pass capacity (recovery rate) 11500 m3/day (50%) 10700 m3/day (50%)

Second pass capacity (recovery rate) 6700 m3/day (85%) 3000 m3/day (85%)

TDS (boron), ppm 111 (0.75) 118 (0.77)

First pass array 129 PV ( 8 M) 120 PV (8 M)

Second pass array 18:9 PV (8 M) 8:4 PV (8 M)

Power consumption, kwhr/m3 3.60 3.33 (-8%)

Total number of elements 1248 1056 (-15%)

Another illustration of benefits of split partial configuration compared to conventional partial

two pass design is shown on Figure 5.2.16.

The plots on Figure 5.2.16 provides fraction of second pass processing required for a given

salinity of the blended product. For example, for a combined permeate salinity of 300 ppm, the

conventional partial two pass system would require processing of about 24% of the first pass

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



157

permeate. In case of split partial configuration, only 5% processing of the first pass permeate

would be required.

Partial two pass processing

0

100

200

300

400

500

0 5 10 15 20 25 30 35

Fraction of 1st pass processed, %

Co

mb

ine

d p

erm

ea

te

sa

lin

ity

, p

pm

TD

S

Conventional

Split partial

Figure 5.2.16. Comparison of second pass processing required for conventional partial two pass

and split partial two configurations.

The relative effectiveness of the split partial configuration increases with decrease of the fraction

of the first pass that has to be process again. The closer that two pas system is to the full two pass

processing the less attractive the split partial processing is.

The split partial system operates without any obstruction of valve regulation of the relative flow

from both ends of the membrane unit. The flow is regulated by the pump of the second pass. The

higher the flow rate of the feed pump of the second pass unit the higher fraction of the first pass

unit is taken from the concentrate end.

The split partial unit operates without buffer tank. Utilization of buffer tank would reduce

flexibility of operation and effectiveness of the process.

5.3. Membrane cleaning

Membrane cleaning is conducted to recover membrane performance by removing deposits from

the membrane surfaces of membrane elements. The decision to conduct membrane cleaning is

triggered by reduction of membrane water permeability, increased pressure drop and/or increase

of salt passage.

The decision about timing, frequency and type of cleaning procedure applied is usually based on

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



158

the relevant experience gained during operation of the membrane system and analysis of the

foulant deposit.

The limiting parameters of cleaning procedures, such as maximum temperature of cleaning

solution, maximum and minimum pH, average flow rate and pressure, is specified by membrane

manufacturers.

5.3.1. Configuration of membrane cleaning unit

The cleaning of membrane elements in an RO train is conducted using the cleaning unit. The

configuration of cleaning unit is shown in Figure 5.3.1. The cleaning unit consists of cleaning

tank, heater, recirculation pump, cartridge filter and connecting piping. It is recommended to

have an air gap between cleaning tank and piping that returns cleaning solution from the product

manifold of the membrane unit. This is to prevent contamination of the product side of the

membranes with cleaning solutions.

Larger cleaning units also include separate tank for dissolving and mixing of cleaning solutions.

Depending on the size of the system and frequency of cleaning, some systems may require a

dedicated tank and dosing units for neutralization of cleaning solution, prior to disposal.

Materials of construction of the cleaning unit should be selected to withstand low and high pH

cleaning solutions (pH 2 – 11) at temperatures up to 50 C.

The size of cleaning tank and capacity of cleaning pump is determined by the number of pressure

vessels that will be cleaned at one time. During cleaning operation the flow rate of cleaning

solution per vessel should be close to 7 - 9 m3/hr (~ 30 – 40 gpm). The cleaning tank volume

should hold enough cleaning solution volume to provide at least 3 - 5 min of pump capacity. If,

for example, the RO membrane train is a two stage unit with pressure vessels array of 64:32,

then the maximum number of pressure vessels that will be cleaned at one time will be 64.

Accordingly, the maximum flow rate of the cleaning solution to the membrane unit will be 64 * 8

m3/hr = 512 m3/hr. The operational volume of the cleaning tank, equivalent to 5 min of pump

operation, will be 42 m3. The quantity of chemicals required for cleaning is not based only on

operating volume of cleaning tank but consideration must be given to the total empty volume of

the RO unit, which includes, pressure vessels, manifolds and connecting piping. Estimation of

empty volume of membrane unit for cleaning is illustrated in Example 5.3.1.

In large RO systems, the connecting piping from the cleaning unit is permanently attached to all

trains. Valves or removable piping segments are used to connect/disconnect given train or train

segment to the cleaning unit. The connecting manifold should be configured to enable to transfer

of majority of spent cleaning solution after the cleaning to the neutralization tank.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



159

CF

Cleaning solution

storage tank

Cleaning solution

neutralization tank

Heater

Figure 5.3.1. Configuration of membrane cleaning unit that includes neutralization tank.

In the configuration of cleaning system shown on Figure 5.3.1 the cleaning solution

neutralization tank is shown as of similar size as the cleaning tank. The actual size will depend

on the mode and frequency of conducting cleaning – neutralization operations.

Table 5.3.1. General specification of cleaning equipment.

Equipment type Description

Cleaning solution flow 7 – 9 m3/hr/pressure vessel in the train (or

stage) being cleaned

Cleaning solution inlet pressure 2 – 4 bar

Materials of construction

- Metallic components

- Plastic components

- Elastomers

Stainless steel 316/316L

PVS, FRP and HDPE

Teflon (PTFE) or EPDM

Inlet – outlet flanges class ANSI Class 150

Tank immersion heater power output 1kW/m3/hr of recirculation flow.

Cartridge filters flow sizing 1 m3/hr/25cm cartridge length

Cartridge rating 20 micron

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



160

Cleaning pump configuration Corrosion resistant, weatherproof

Main instrumentation

- Level gauges

- pH analyzer

Magnetic type

Provide discrete high and low pH alarms

Pressure gauges Low pressure gauges

5.3.2. Sequence of operation of cleaning unit.

Cleaning operation sequence includes:

1. Flushing RO train with permeate water.

2. Connecting train or train segment to the cleaning unit.

3. Preparing cleaning solution in the cleaning unit.

4. Recirculating cleaning solution for 1 – 4 hr through the RO train.

6. Draining cleaning solution to the neutralization tank.

7. Flushing cleaning solution with permeate (after high pH cleaning) or with feed water.

8. Repeating steps 2 – 5 with next cleaning formulation or reconnecting cleaned train to high

pressure pump and restoring normal operation.

Membrane cleaning, like any other dispersive process, is more effective when conducted at

elevated temperature.

Cleaning should be conducted at temperature of cleaning solution in the range of 35 – 40 C.

Cleaning solutions can be purchased from specialized suppliers or generic cleaning

formulations could be used. Composition of generic cleaning formulations can be obtained

from all major membrane manufacturers.

One of the generic low pH cleaning formulation, frequently used, is 2% solution of citric

acid. pH of such solution is about 2.5. Citric acid cleaning solution is very effective in

removal of deposits of metal hydroxides and dissolving of carbonate scale. If it has been

established that fouling deposit contains mainly calcium carbonate or metal hydroxides,

temporary operation with feed water acidified to low pH (pH = 4.5 – 5) with mineral acid

(H2SO4 or HCl), may be sufficient to restore membrane performance. Cleaning, through

operation of membrane unit at low feed pH, is only possible if discharge of low pH

concentrate is allowed by local regulation at a given site.

The generic high pH cleaning formulations consist of solutions of NaOH in combination with

EDTA or SDBS (surfactant). The caustic cleaning solutions have pH of 10 – 11 and are

effective in removal deposits of organic matter from membrane surface. It has been found

(14) that EDTA or surfactants are essential components of high pH cleaning solutions and

their presence contributes to improved removal of surface deposits that contain Ca ions

imbedded in the organic fouling layer. In majority of cases, fouling layer is of a mixed nature,

it contains a mixture of inorganic and organic matter. The effective cleaning sequence is to

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



161

apply low pH cleaning followed by application of high pH formulation. Prior to use of

cleaning solution containing surfactant, a small scale test should be conducted to check if

designed concentration of surfactant will not result in excessive foaming.

RO systems treating well water feed very seldom required membrane cleaning. Cleaning

frequency is usually less than one cleaning per 2 – 3 years of operation. RO systems treating

surface water or wastewater feed, require more frequent membrane cleaning. For the purpose

of operating cost estimation, budget for cleaning operation is usually based on two – four

cleaning events per year. If more frequent cleanings are required, then it is an indication of

inadequate pretreatment process. Frequently, only segments of membranes in a train require

cleaning. For example membrane elements in the last stage my require cleaning at higher

frequency then the rest of the system. Connections between cleaning unit and membrane train

should be configured to enable cleaning of either the whole train or single stage.

Example 5.3.1. Calculation of annual cleaning cost.

System permeate capacity: 100,000 m3/day (26.4 mgd),

RO unit configuration: 8 trains, 96 PV per train, 7 elements per vessel, 64:32 array

Train segment size for a single cleaning: 98 pressure vessels.

Annual cleaning frequency: 2

Cleaning procedure: low pH cleaning followed by high pH

Free volume of pressure vessels: 98 X 7 X 0.025m3 = 17 m3 (4,500 gallons)

Volume of manifolds (10% of PV): 0.1 x 17 m3 = 1.7 m3 (450 gallons)

Volume of connecting piping (50% of PV): 0.5 x 17 m3 = 8.5 m3 (2250 gallons)

98 PV X 8 m3/hr X 3min/60 min = 39.2 m3 (10,300 gallons)

Total empty volume of membrane unit for cleaning: (17+1.7+8.5) m3 + 39.2 m3 = 66.4 m3

(17,500 gallons)

Chemicals quantity for annual cleaning operation:

Solution 1 – citric acid

2% Citric acid: 0.02 x 66.4 = 1.328 t/cleaning

8 train X 2 cleanings/year X 1.328= 21.2 t

Solution 2 – NaOH + SDBS

0.2% NaOH: 0.002 x 66.4 = 0.133 t/cleaning

8 train X 2 cleanings/year X 0.133 = 2.1 t

0.2% SDBS: 0.002 x 66.4 = 0.086 t/cleaning

8 train X 2 cleanings/year X 0.133 = 2.1 t

Annual cost of cleaning chemicals

Citric acid: 21.2t X $3500/t = $74,200

SDBS: 2.1t X $3500/t = $7,350

NaOH: 2.1t X $350/t = $735

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



162

Total cost of cleaning chemicals per year: $82,285

Cost per water produced: $0.0024/m3 ($0.009/kgallon)

As indicated by above example cost of generic cleaning chemicals is not significant if

cleaning frequency is limited to two cleanings per year. Additional cost, associated with

cleaning operation, that should be considered is loss of production capacity. The system off

line time required for cleaning is in the range of 1 – 1.5 days, which is corresponds to about

0.4% of availability The major expense related to cleaning, at some locations, could be the

disposal cost of spent cleaning solutions.

In majority of cases, cleaning operation is capable to restore some of the lost permeability

and reduce pressure drop. Very seldom salt rejection is improved. Usually, it remains the

same or can even decline after cleaning. This is because foulant layer plugs imperfections

and damaged areas in membrane barrier and effective cleaning opens them again to salt

passage.

If cleaning attempts do not result in sufficient performance improvement, membrane element

replacement is the only practical solution available for additional performance correction. As

it will be discussed in section X.X1 usually a considerable fraction of elements in the system

have to be replaced to achieve noticeable performance improvement. Number of elements

that require replacement could be reduced, if elements with worst performance can be

identified in the RO system. Discussion of such approach is included in chapter X.X2.

5.4. Membrane flushing unit configuration

Flushing of membrane unit is required every time the unit shuts down for planed stoppage of

system operation, shut down due to equipment failure or power outage.

Without flushing of membrane unit on shut down, there is possibility of scale formation,

especially in brackish RO unit or bacterial grow, especially in seawater RO.

Flushing of membrane unit is conducted with permeate water. If permeate water is not

available, the membrane unit can be flushed with clean feed water. Point of supply of

permeate for flushing should be selected to prevent of flow of chlorinated permeate water to

the membrane unit.

If possible, the flushing should be configured so the membrane units will be flushed also,

when shut down is caused by power outage.

The standard recommendations of membrane manufacturers are to flush membrane unit on

shut down with permeate water and then repeat flushing every 5 days if the membrane unit

remains idle.

Flushing involves replacing 1 – 2 empty volumes of the membrane unit, including manifolds

and interconnecting piping. Empty volume for flushing is estimated as illustrated in Example

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



163

5.3.1.

The general specifications of flushing equipment are the same as listed in Table 5.3.1 for the

cleaning equipment.

5.5. RO membrane unit design criteria in accordance with water quality

The design of RO system starts with the ultimate project objective: consistent production of a

design flow of product water at a required quality. The selection of process parameters is

influenced by type of feed raw and design range of water quality parameters: dissolved species

composition and concentration, fluctuation of water temperature and turbidity. The direct design

parameters of the RO membrane unit, such as an average permeate flux rate and system recovery

rate, are based on type of raw water source and projected feed water quality. The optimization of

values of design process parameters, within the recommended ranges, requires understanding the

relation between system operating conditions and stability of membrane performance. The

average permeate flux rate should be within the range that would not accelerate membrane

fouling. The recovery rate should be below the value that would result in excessive saturation

limit of the sparingly soluble salts in the concentrate.

The design process begins with evaluation of feed water quality. Based on feed water quality

and composition the designer decides on selection of membrane element type, average flux rate

(total membrane area in the system) and maximum recovery rate.

5.5.1. Feed water quality parameters

The primary indicators of fed water quality for RO applications are Silt Density Index (SDI) and

turbidity. Membrane manufacturers usually define values of these two indicators as part of

membrane warranty terms. Secondary indicators, such as concentration of suspended particles,

TOC and concentration of sparingly soluble salts, guide system designer to define process

parameters: average permeate flux and system recovery rate.

Some of the feed water quality indicators, including SDI have been discussed briefly in Chapter

1. SDI is very sensitive indicator of presence of suspended solids in water. However, the SDI

readings could vary significantly with type and size of suspended particles present in the water

and type of material of the filter pad used for SDI determination. The required equipment for

determination of SDI is very simple and procedure easy to conduct at field conditions. However,

both the accuracy and reproducibly of results are not very satisfactory. In spite of the above

deficiencies, SDI is universally adopted as a primary indicator of feed water quality in RO

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



164

systems. Determination of SDI takes about 25 – 30 min.

The secondary feed water quality indicator, used in RO applications is turbidity. Turbidity

results, usually expressed as nefelometric turbidity units (NTU), are determined through

measurements of intensity of light scattered by suspended particles in a water sample. Turbidity

sensor could provide continuous measurement of feed water turbidity. Turbidity results are

sensitive to quantity of colloidal particles, their size and shape. Past attempts to correlate

between SDI and turbidity results, even for the same site, demonstrate very weak relations. These

two water quality indicators correlate to the number and size od suspended particles differently.

However, the general rule is that feed water with SDI in the range of 2 – 3 has a corresponding

turbidity below 0.1 NTU, usually at 0.05 NTU range.

5.5.2. Membrane fouling

During the membrane fouling process, performance of membrane elements changes due to

formation of deposits on membrane surface or inside of feed – brine channel. Fouling affects

membrane elements performance. The symptoms of fouling are decrease or increase of water

permeability, increase or decrease of salt rejection and increase of pressure drop across the RO

unit. Later stages of uncontrolled fouling could result in structural damage of membrane or

membrane element. List of fouling factors encountered in RO applications is included in Table

5.5.1.

Table 5.5.1. Summary of membrane fouling categories and their symptoms

Fouling factor Initial fouling stage

effect

Advanced fouling

stage effect

Potential

membrane

damage

Exposure to

strong oxidants

Some permeability

and salt passage

decline. Initially in

the lead element(s)

Increase of

permeability and

significant increase of

salt passage

Irreversible

damage of

membrane barrier

Colloidal matter Some increase of

pressure drop. Initially

in the lead element(s).

Significant increase of

pressure drop, some

decline of

permeability and

increase of salt

passage

Element

telescoping and

extrusion of brine

spacer. Membrane

barrier damage

Dissolved natural

organic matter

Some permeability

and salt passage

Moderate decline of

permeability, same salt

None

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



165

(NOM) decline. passage decline

Biological matter Some increase of

pressure drop. Some

permeability and salt

passage decline.

Initially in the lead

element(s).


pressure drop. Some


passage decline.

Element

telescoping and

extrusion of brine

spacer. Membrane

barrier damage

Inorganic scale Some increase of

pressure drop. Some

permeability decline.

Initially in the tail

element(s).

Severe increase of

pressure drop. Some

permeability decline

and salt passage

increase.

Severe blockage

of feed channels

Petroleum

products, also oil

and grease

Significant

permeability decline in

the lead element(s).

Small effect on salt

passage

Severe decrease of

permeability. Small

effect on salt passage

None at low

concentration. At

high

concentrations

barrier integrity

damage

Composite

foulants (organics

+ colloids)

Some increase of

pressure drop. Some

water permeability

decline. Initially in

the lead element(s).


pressure drop.

Significant

permeability decline

and some salt passage

increase.

Significant

blockage of feed

channels. Element

telescoping.

Membrane barrier

damage

Sharp edged,

micron size

particles in feed

water

Gradual salt passage

increase

Significant salt

passage increase

Cuts and holes in

membrane barrier

Prolonged

exposure to very

high pH (> 12)

Increase of water


passage


water permeability and

salt passage

Chemical damage

of membrane

barrier material

Prolonged

exposure to very

low pH (< 2)

Initial decrease of

water permeability and

salt passage

Significant decrease of

water permeability

Chemical damage

of membrane

barrier material

Low

concentration of

organic solvents

Gradual decrease of

water permeability,

increase of salt

passage

Significant decrease of

water permeability,

increase of salt

passage

Swelling of

membrane barrier

and eventually

chemical damage

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



166

The specific membrane fouling process that could be of concern in desalination system will

depend on application type and composition of feed water. Summary of more frequent potential

fouling processes, grouped by application type, is provided in Table 5.5.2

Table 5.5.2. Summary of fouling processes for various desalination applications and possible

cause of the fouling processes.

Application type Fouling process Likely culprit

Desalination of seawater Particulate fouling High concentration of

particulate matter in

seawater (algae bloom),

carryover of coagulant from

the pretreatment system

Biofouling High concentration of

biological activity in

seawater, use of continuous

chlorination

Desalination of ground

water

Membrane scaling High concentration of silica,

calcium sulfate, calcium

carbonate, barium sulfate in

the concentrate

Particle induced cuts of

membrane barrier

Presence of mineral

particles in feed water.

Reclamation of industrial

streams

Any of the fouling

processes listed in Table

5.5.1.

Potentially any industrial

waste constituent present in

the feed water

Reclamation of municipal

wastewater

Scaling High concentration of

calcium phosphate, calcium

carbonate in the concentrate

Blocking of feed channels High concentration of

colloidal matter in the feed

water.

Adsorption of organics Present of high

concentration of NOM in

the feed water

Biofouling Insufficient concentration of

chloramines in the feed

water

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



167

Oxidation of membrane

barrier

Presence of free chlorine

(insufficient ammonia

concentration) in the feed

water

Fouling can be prevented or mitigated by reduction of concentrate of the foulant in the

pretreatment step, adjustment of feed water pH, addition of scale inhibitor, operation at average

flux rate and recovery that would not result in high fouling rate.

5.5.2.1.Oxidative degradation of membrane performance

The sensitivity of membrane performance to presence of strong oxidants was known at the very

onset of development composite polyamide membrane. The early tests of exposure of polyamide

membrane to free chlorine, indicated limited initial stability followed up by a significant decline

of salt rejection. Feed water to RO membrane unit has to be completely dechlorinated. The better

solution is to totally avoid introduction of free chlorine to RO feed water.

The only exception is formation and maintaining of chloramine in wastewater reclamation

systems. RO membrane operating in municipal wastewater reclamation systems are exposed to

high concentration of organics in the feed water that rapidly adsorb on membrane surface during

the initial stage of operation. This condition enables use of chloramines to prevent membrane

biofouling. With the presence of chloramines in the feed water in the concentration range of 2 – 4

ppm, salt passage increases at a rate of about 100% within 3 – 5 years of operation. This is an

acceptable rate for majority of applications.

Use of chloramines for other RO application, almost always results in very rapid salt passage

increase. The possible reason for slower effect of chloramines on salt rejection change in

wastewater application is presence of organic layer on membrane surface. Most likely, this

organic layer reacts with chloramines, decreasing its concentration at the membrane surface. In

addition, the organic matter present in the feed, plugs pinholes and other defects in membrane

barrier, effectively reducing the salt transport.

Presently, the use of chloramines in RO systems desalting tertiary municipal effluent is a

common operational procedure. The operating conditions are designed to maintain chloramines

level at 2 – 4 ppm in the feed to the membrane unit. If level of ammonia in feed water is too low

to produce sufficient concentration of chloramines, ammonia could be added, usually in the form

of ammonium chloride, adjacent to chlorine (or hypochlorite) injection point.

5.5.2.2.Colloidal fouling

Ground water sources, even these contaminated by agricultural or industrial effluents, usually

have very low concentration of colloidal matter. Water undergoes natural filtration during

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



168

passage through a sand layer in the underground aquifer. Brackish water RO systems very

experience colloidal fouling.

Presence of colloidal fouling in seawater systems will depend on process selection and operation

of seawater pretreatment unit. In seawater systems utilizing conventional pretreatment, colloidal

fouling is possible. In seawater systems utilizing membrane pretreatment (MF or UF) with good

membrane integrity, the colloidal fouling is not very likely.

Colloidal fouling is not a common phenomenon in membrane wastewater reclamation systems.

Membrane filtration of the secondary municipal effluent, used as pretreatment, prior to RO,

practically removes all colloidal matter. The other municipal source, effluent from membrane

bioreactor (MBR), is practically of the same quality as tertiary effluent after MF or UF filtration.

Presence of colloidal fouling, in RO units operating in municipal wastewater reclamation

systems, is usually a result of lack of integrity of membrane barrier in the pretreatment system. In

some cases, colloidal particles could form as a result of precipitation from a saturation solution

of scale forming salts. Another possible source of colloidal particles could be coagulant used in

the membrane filtration system. The coagulant particles could pass through micropores of MF

membrane or form in the RO system as the feed volume is reduced and concentration of soluble

metal ions increases into a saturation range.

Presence of colloidal matter in RO feed could result in very rapid permeability decline. Colloidal

matter, mixed with organics, forms on the membrane surface a very dense layer of low

permeability. Colloidal particles could also become a crystallization centers for sparingly

soluble salts that are at saturation, and induce formation of scale. Use of membrane filtration as a

pretreatment step, prior to RO, would prevent in most case colloidal fouling of RO membranes.

5.5.2.3.Fouling by organic matter

Membranes in both brackish and seawater RO desalination system could experience fouling by

dissolved organics. However, the origin of organic matter is usually different in both cases.

Majority of brackish water sources have very low concentration of dissolved organics. However,

some low salinity well waters, treated by nanofiltrtaion system may have high concentration of

dissolved organic matters, usually in the form of humic substances. Treatment of these water

sources require use of special, low fouling, membrane elements.

Clean seawater contains very low concentration of organic matter, usually below 1 ppm of TOC.

However, seawater in the area of heavy marine traffic or in the areas containing oil fields,

seawater could be contaminated with petroleum products. Exposure of RO membranes to feed

water containing petroleum product will result in rapid decline of water permeability. Petroleum

products in seawater have to be removed completely, below the detection limit, prior to

introduction of feed water to membrane unit.

The municipal and industrial wastewater contains high and variable concentrations of

organics. Some fraction of the organics is in a form of suspended colloidal particles. Majority of

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



169

these will be removed during the pretreatment step of membrane filtration.

The reduction of concentration of dissolved organics by MF or UF filtration is not significant.

Some reduction of TOC, up to about 30% could be achieved if coagulation is applied prior to

membrane filtration. The reduction mechanism is through adsorption of organic compounds on

coagulant particles and subsequent removal by membrane filtration.

In majority of cases, the effect of dissolved organics on membrane performance is reduction of

water permeability, associated with some decrease of salt passage (improvement of salt

rejection).

The extend of water permeability decline and the decline rate depends on composition of organic

mixture in the feed, characteristics of membrane surface and composition and ionic strength of

feed water. In RO reclamation units with well functioning pretreatment, the long term water

permeability decline is expected to be in the range of 30% - 40%. Usually, the decline irate is

high at the startup, reaching 10% - 20% and then levels off with time. Accordingly, the feed

pumping system has to be designed to provide sufficient feed pressure to compensate for the

above decline of permeability over time. Except for permeability decline, fouling of RO

membranes by dissolved organics, usually does not result in other adverse effect on membrane

performance parameters (build up of feed channel pressure drop or increase of salt passage).

In some isolated cases organic scale inhibitor, added to the feed to prevent scaling, could react

with coagulant and foul the membrane. Some dissolved industrial organics may cause membrane

swelling. Membrane manufacturers should be consulted regarding potential compatibility issue

with industrial type feed constituents.

5.5.2.4.Biofouling

Biofouling, which is the phenomena of bacterial film formation in RO systems, could represent a

serious operation problem. Biuofouling manifest itself in increase of feed channel pressure losses

and productivity decline. Systems contaminated with established bacterial grow are very difficult

to clean and restore to the original performance level.

The assumption is that every body of water contains microorganisms in equilibrium with the

local nutrients supply. The water born microorganisms easily attached themselves to the surfaces

in the RO systems and form colonies. The attachment to the surfaces is through excretion of

extra cellular polymeric substance (EPS), composed mainly of polysaccharides. On wetted

surfaces, the microbiological cells and surrounding EPS form biofilm, which can grow at rapid

rate if sufficient nutrients and energy are available. The structures of biofilms are not uniform,

depending on type of microorganisms and environmental conditions: pH, temperature, flow

velocity, age and variety of other parameters.

The biofouling occurrence is limited mostly to RO systems processing surface water (including

seawater) or ground water from shallow wells. Majority of RO systems processing well water

experience little or no biofouling. Number of procedures have been developed to identify

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



170

condition of availability of nutrient and to monitor biofilm grow.

RO treatment of municipal effluents represents unique situation of high biofouling potential of

RO feed water and development of effective way of controlling microorganisms population. In

the past, municipal effluents were processed using cellulose acetate membranes. Therefore, it

was a common practice to chlorinate RO feed water to prevent microorganisms grow. Due to

presence of ammonia and high concentration of organics in the effluent the free chlorine was

converted to chloramines. Today, almost all RO systems treating municipal effluents use

composite polyamide membranes. Based on field experience it has been established that in

wastewater applications salt rejection of composite membranes, in presence of chloramines

concentration of 2 – 4 ppm, is sufficiently stable to provide 3 – 5 years effective membrane life.

The above range of chloramines concentration is also sufficient to effectively mitigate membrane

biofouling in RO unit.

5.5.2.5.Inorganic scale and determination of permeate recovery rate.

The concerns for potential of scale formation in RO systems are related to the very nature of the

RO process: removal of water from the feed stream and increase of concentration of the ions

present in the feed water.

During the RO process concentration of all constituents increases due to reduction of the feed

water volume. This increase of concentration, expressed by concentration factor (CF) is function

of permeate recovery.

CF = 1/(1 – R) (1.1)

There is possibility for some of the dissolved constituents to nucleate and precipitate, if the

concentration product of salt forming ions exceeds its solubility product: Ksp. For a given salt of

composition CmAn in equilibrium of solid phase salt (S) with dissolved ions, the Ksp is defined

as:

CmAn(S) = mC+n + nA-m (1.2)

Ksp = [C+n ]m [A-m]n (1.3)

SI = [C]m [A]n / Ksp (1.4)

Where C stands for cation and A for anion, m and n are valency coefficients. Brackets [ ] indicate

molar concentration of a given ion in solution. SI is the saturation index, indicating excess

concentration of a given salt in comparison to its saturation value.

The Ksp is determined through measurement of ions concentrations in solution at saturation

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



171

conditions (in equilibrium with a solid phase of a given salt). Ksp has a specific value for each

salt and it is function of temperature and ionic strength of the solution.

In RO seawater desalination systems concentration of scaling constituents is very low and the

solubility is significantly increased due to high ionic strength. Therefore, scaling in seawater RO

membrane units is very unlikely.

Brackish water sources could have significant concentration ions that could form scaling

compounds. Usually, the recovery rate of brackish RO systems is limited by the saturation limits

of sparingly soluble salts. The salts of concern are calcium carbonate, calcium sulfate, strontium

sulfate, barium sulfate and silica.

In wastewater reclamation systems, the salt of concern is mainly calcium carbonate, less

frequently calcium phosphate. In some isolated cases, potential for precipitation of barium

sulfate has to be considered.

Calcium carbonate is the most common scaling constituent in natural occurring waters and also

in wastewater. However, it is also the easiest to control either with pH adjustment or use of scale

inhibitor.

In solution, calcium ions are in equilibrium with bicarbonate and carbonate species as shown in

the following equations:

H2CO3 = H+ + HCO3- (5.1)

HCO3- = H+ + CO3

-2 (5.2)

Ca+2 + CO3-2 = CaCO3 (5.3)

At sufficiently high concentrations of Ca+2 and CO3-2, crystallites of CaCO3 could nucleate and

form a scale. The calcium carbonate system is quite complex. Saturation conditions are not just

function of concentrations of Ca, CO3 and HCO3 ions but also influenced by concentration of

hydrogen ion (pH) and other ions that contribute to water alkalinity.

Attempts to define relations for saturation conditions in potable water networks lead to

development of a number of saturation indexes. The calcium carbonate saturation index

developed by Langelier for potable water networks has been adopted by RO industry as an

indicator of saturation conditions in concentrate stream of brackish water RO systems. The

Langelier Saturation Index (LSI) is calculated according to relations:

LSI = pH - pHs (5.4)

Where pH I is the actual pH of the water and pHs is pH that corresponds to saturation

concentrations of ions forming calcium carbonate.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



172

K2 = [H+] X [CO3-2]/[ HCO3

-] (5.5)

Ksp = [Ca+2] X [CO3-2] = [Ca+2] X[HCO3

-] X K2/ Hs+ (5.6)

Hs+ = [Ca+2] X [HCO3

-] X K2/ Ksp (5.7)

- log[Hs+] = pHs = - log[Ca+2] - log[HCO3

-] + log [Ksp/ K2] (5.8)

LSI = pH - pHs = pH - pCa - pAlk +pK (5.9)

Where K2 is second dissociation constant to carbonic acid (H2CO3), Ksp – solubility constant of

calcium carbonate at given pH and temperature. Other parameters represent molar concentrations

of relevant species in the solution.

Few years after introduction by Langelier the saturation index was modified to account for ionic

strength in a form of correction factor proportional to water salinity.

pHs = (9.3 + A + B) - (C + D) (5.10)

where: A = (Log10 [TDS] - 1) / 10 (5.11)

B = -13.12 Log10 (°C + 273) + 34.55 (5.12)

C = Log10 [Ca+2 as CaCO3] - 0.4 (5.13)

D = Log10 [alkalinity as CaCO3 ] (5.14)

The parameter A is related to ionic strength of the solution. The value of A increases with

increase of salinity. Parameter B reflects the changes of calcium carbonate solubility and changes

of equilibrium of carbonic acid dissociation with temperature. The value of B decreases with

temperature increase. In practical applications the LSI is either calculated using computer

programs or monograms based on pH and composition of concentrate stream. Water solution has

potential for CaCO3 scaling at LSI > 0 and it is assumed that saturation prediction using LSI is

reliable up to salinity of about 5000 ppm TDS.

For very high water salinity and seawater applications the LSI was modified to account for

increased ionic strength by Stiff and Davis. The Stiff and Davis saturation index (SDSI)

introduces empirical constant K in calculations pHs to account for high ionic strength of seawater

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



173

concentrates. Stiff and Davis determined value of K experimentally in the high range of ionic

strength, that covers salinities encountered in RO seawater applications.

SDSI = pH – (9.3 + K – pCa – pAlk) (5.15)

Where pH has the same meaning as in LSI equation and K is a constant found in monograms.

For high salinity solutions (seawater concentrates) the SDSI value is about 1 – 1.3 units lower

then calculated according to the LSI relations. In low salinity applications the values of LSI and

SDSI are similar.

Example 5.5.1. Calculation of Langelier saturation index

Brackish water system is design to operate at recovery ratio of 80%.

Feed water feed has TDS = 1700 ppm, ionic strength 0.04 and the following concentrations of

the relevant ions:

Ca = 300 ppm, HCO3 = 250, pH = 7.3, temp = 25 C.

After acidification and pH adjustment to 6.5, HCO3 = 166 ppm, CO2 = 78 ppm

Concentration factor for 80% recovery rate, CF = 1/(1-R) = 5

Approximate concentrations in the concentrate:

TDS = 8500 ppm, ionic strength 0.18, Ca = 1500 ppm, HCO3 = 830 ppm, CO2 = 78 ppm

Calculation of concentrate pH.

pH = pK1+ log ([HCO3]/[CO2]

K1 is the equilibrium constant of the carbonic acid dissociation reaction. For water temperature

of 25 C and dilute solutions K1 value is 4.2X10-7, pK1 = 6.37.

Then the

Calculation of pHs and LSI

Following equations 15.13 – 15.16:

pHs (feed) = 9.3 +0.22 + 2.08 – 2.17 – 2.13 = 7.48

LSI (feed) = 6.5 – 7.48 = - 0.98

pHs (concentrate) = 9.3 +0.29 + 2.08 – 2.87 – 2.83 = 5.97

LSI (concentrate) = 7.26 – 5.97 = 1.29

The LSI provides qualitative indication about saturation condition of solution that contains

calcium and carbonate ions. The LSI does not enable prediction about quantity of CaCO3 that

could potentially precipitate from solution. The saturation index that provides indication about

quantity of CaCO3 that could precipitate is the calcium carbonate precipitation potential (CCPP).

The calculations of CCPP are more laborious than calculations of LSI. The calculations of CCPP

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



174

are usually conducted using specialized computer programs.

The saturation relation of calcium sulfate is simpler to calculate than the one for calcium

carbonate. For RO calculations it is assumed that its solubility depends only on concentrations of

calcium and sulfate ions, temperature and ionic strength. The effect of ionic strength on

solubility of calcium sulfate is quite significant as shown in the following example of

calculations of saturation indexes (SI) of CaSO4 for water salinities corresponding to brackish

and seawater RO systems:

Example 5.5.2. Calculation of calcium sulfate saturations

For CaSO4; Ksp = (temp/25)0.152 X1.8X10-3 X IS 0.75

Brackish water system

R = 75%, TDS of concentrate = 4000 ppm, IS = 0.07

Ca = 900 ppm, SO4 = 2400 (concentrations in concentrate)

Ksp = 2.5X10-4

SI = ([900/40000] [2400/96000])/2.5X10-4 = 2.25

Seawater system

R = 50% , TDS of concentrate = 80,000 ppm, IS = 1.60

Ca = 900 ppm, SO4 = 6000 ppm (concentrations in concentrate)

Ksp = 2.6X10-3

SI = ([900/40000] [6000/96000])/2.6X10-3 = 0.54

In the example above the product of calcium and sulfate ions concentrations in solution is much

higher in concentrate stream of the seawater system as compared to the brackish system.

However, due to the differences of ionic strength, the calcium sulfate is above saturation level in

the brackish system and below saturation in the seawater system.

In wastewater reclamation systems calcium sulfate practically never reaches a saturation

potential due to limits of calcium and sulfate ions concentrations in potable water supply. In

some isolated cases barium sulfate could be a recovery rate limiting constituents. Solubility of

barium sulfate is five orders of magnitude lower than solubility of calcium sulfate.

Another constituent that sometimes can present problem in RO applications is silica. Silica

concentration in wastewater effluents will depend on its concentration in potable water, which is

low at majority of locations. Silica can be present both in colloidal and reactive (soluble) form.

In the past, the safe limit of silica concentration in concentrate was considered as being about

140 – 170 ppm (as SiO2). In the last decade a new scale inhibitors were introduced that are

effective in maintaining much higher concentration of silica in solution. Some suppliers of these

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



175

specialty inhibitors claim safe limits for silica concentration as high as 300 ppm. When treating

brackish sources with significant silica concentration an extreme caution should be exercised

with maintaining the designed recovery rate as silica scale is very difficult to remove from

membrane elements.

Beside calcium carbonate, the major scaling compound of concern in RO wastewater

reclamation systems is calcium phosphate. Calcium phosphate presence is almost uniquely

associated with household or industrial effluents at various concentrations. In majority of RO

system treating municipal effluents concentration of phosphates is low and does not result in

membrane scaling.

The chemistry of calcium – phosphate compounds is quite complex and precipitants of calcium

and phosphates can be formed as number of chemical formulations of different solubility. The

prevailing assumption is that most likely scale forming compounds are: calcium phosphate

dihydrate (DCPD-Brushite) and Tricalcium phosphate (TCP). At locations where phosphates

concentration is high, formation of scale during RO operation could be prevented by using scale

inhibitors or acidification. The effectiveness of scale inhibitor in prevention of formation of

calcium phosphate scale is not well documented. Results of one experimental work of evaluation

of commercial scale inhibitors indicate very low effect on rate of scale formation. Reduction of

feed pH is very effective in prevention of phosphate scaling. The solubility limits for DCPD

were developed experimentally and follows the equation that provides maximum value for Ca

and P product in solution according to concentration of hydrogen ion (pH):

Ca X P = 103.85 + 1011.55 X [H+] (5.17)

For TCP an empirical relation for calculation of saturation index was developed by BETZ

Company:

65.0

)log(2)log(log755.11 4 tPOCapHIndex

(5.18)

Where

pH - pH of the concentrate

Ca - Calcium concentration in the concentrate, ppm CaCO3

PO4 - Phosphate concentration in the concentrate, ppm

t - Water temperature, C

Positive value of the phosphate saturation index indicates that scaling could occur. Scaling can

be controlled by use of scale inhibitor or acidification of feed water.

Table 5.5.3. provides values of concentrate pH required for maintaining phosphate saturation

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



176

index below the threshold value of 1 at the recovery rate range of 75% - 85% for the

concentration of calcium and phosphate in the feed water and concentrate as listed below.

Table 5.5.3. Controlling phosphate scaling through pH adjustment

Parameter Feed water Concentrate

@ R = 75%

Concentrate

@ R = 80%

Concentrate

@ R = 85%

pH 8.1 6.8 6.5 6.1

Ca, ppm 300 1200 1500 2000

P, ppm 3 12 15 20

PO4, ppm 9.2 36.7 45.9 61.2

Phosphate Saturation

Index (Eq. 15.19)

- 0.5 0.74 0.74 0.72

However, acidification of RO feed, required to prevent formation of calcium phosphate scale,

could be very expensive if significant quantity of acid would be required due to high alkalinity

concentration.

Example 5.5.3. Calculation of quantity of acid required to prevent calcium phosphate scaling.

For concentration of bicarbonate ion [HCO3-] in feed water is 300 ppm as CaCO3, calculate

quantity of sulfuric acid required:

a) To reduce feed pH from 8.1 to 7.

b) To bring pH of concentrate from feed pH of 8.1 to the pH values listed in Table 3.

Feed water temperature 25 C.

Values of K1, K2 and KW are: 4.45X10-7, 4.42X10-11 and 9.17X10-15

Calculate concentration of [CO2], [CO3=], [H+] and [OH-] in feed according to the following

equations:

3

1

2 HCOK

HCO (5.19)

3

2

3

2HCO

H

KCO (5.20)

pHH 10 (5.21)

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



177

)(10 pKwpHOH (5.22)

According to the above relations, the concentrations (mol/l) are:

[H+] = 7.94X10-9

[OH-] = 1.15X10-6

[HCO3-] = 6.00X10-3

[CO2] = 1.07X10-4

[CO3=] = 3.34X10-5

The concentration of total alkalinity (AT) and total carbonate species (CT) are calculated

according to equations 5.23 and 5.24. The assumption is made that contribution of phosphate

species to alkalinity can be neglected.

AT = [OH-] + [HCO3-] + 2[CO3

-2] – [H+] (5.23)

CT = [HCO3-] + [CO3

-2] + [CO2] (5.24)

Accordingly

AT = 6.07X10-3

CT = 6.14X10-3

Acidification to adjust pH has no effect on CT. Therefore, new concentration of CO2 can be

calculated according the equation 5.25.

2

211

2

][][1

][

H

KK

H

K

CCO T (5.25)

[CO2] = 1.13X10-3 mol/l (49.7 ppm)

Concentrations of other components of the alkalinity are calculated according to equations 5.19 –

5.22

The total feed alkalinity and total inorganic carbon at pH = 7 are listed below:

AT1 = 5.02X10-3

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



178

CT1 = 6.14X10-3

Quantity of sulfuric acid required to acidify feed water from pH of 8.1 to pH of 7.0 is

(AT – AT1)X eqv H2SO4 =(6.07X10-3 - 5.02X10-3) X49000 = 51 mg/l (as 100%)

The next step is to estimate what should be the feed pH required for a required pH of the

concentrate.

An approximate value can be derived by calculation of concentrate pH for a designed recovery

rate.

The calculations of alkalinity species and pH of the concentrate are based on the assumption that

CO2 is not rejected by the membrane, i.e. concentration of CO2 in feed and concentrate are the

same, and the other dissolved species are rejected completely.

Accordingly concentration of carbonate species in concentrate, CTC will be:

CTC = (CT1 – [CO2])/(1-R) + [CO2] (5.26)

For a concentration of carbonate species at feed water pH of 7.0 and recovery rate R = 80%

CTC = (6.14X10-3 - 1.13X10-3)/(1-0.80) + 1.13X10-4 = 3.03X10-2 mol/l

Substituting the value of CTC and [CO2] in equation 5.25, concentration of [H+] and pHC of the

concentrate can be calculated.

pHC = 7.8

Accordingly, for concentrate pH, listed in Table 5.5.3. the required feed pH should be about 0.8

units lower for the determined pH of the concentrate..

Following the calculations of alkalinity and quantity of sulfuric acid for feed acidification, shown

in the first part of this example, the quantity of sulfuric for required pH of the concentrate,

starting with feed pH of 8.1, is given in Table 5.5.4.

Table 5.5.4. Quantity of sulfuric acid required to maintain concentrate pH below the threshold

value of saturation index of calcium phosphate.

Recovery rate, % 75 80 85

Required concentrate pH 7.0 6.7 6.3

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



179

Required feed pH 6.2 5.9 5.5

Dosing rate of H2SO4, ppm (100%) 173 219 260

Contribution to product water cost, $/m3 at

sulfuric acid cost of $500/t

0.087 0.109 0.130

The results listed in Table 5.5.4 indicate that cost of acidification, to prevent calcium phosphate

scaling, could be quite significant and should be consider in the early stages of evaluation of

project feasibility.

The solubility of sparingly soluble salts is affected by ionic strength.

The Ksp values (at 25 C and for low ionic strength solutions) or concentration limit of common

salts that could form scale in RO systems are listed in Table 5.5.5

.Table 5.5.5. Ksp or concentrations limit of scale forming compounds common to RO

Compound Formula Ksp or concentration limit, (ppm)

Calcium sulfate CaSO4 2.5X10-5

Barium Sulfate BaSO4 2.0X10-10

Tricalcium Phosphate Ca3(PO4)2 2.8X10-30

Calcium Phosphate Dihydrate CaHPO4 2H2O 2.2X10-7

Calcium Carbonate CaCO3 LSI < 0, S&DSI < 0

Reactive Silica H4SiO4 (120 – 160 ppm)

The potential for formation of calcium sulfate scale and blocking of feed channels in the spiral

wound element is demonstrated in the following example:

Example 5.5.3. Calculation of calcium sulfate scaling potential in the RO concentrate.

For the following concentration of Ca and SO4 in the concentrate calculate scaling potential of

calcium sulfate:

[Ca] = 1000 ppm = 0.025 mol

[SO4] = 2400 ppm = 0.025 mol

Ksp = 2.25x10-4

Saturation [Ca] = [SO4] = (2.25x10-4)^1/2 = 0.015 mol

Excess [Ca] = [SO4] = 0.010 mol

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



180

MW CaSO4 = 40,000+96,000 = 136,000mg/l

Excess [CaSO4] = 0.010 mol = 0.010X136,000 = 1,360 mg/l

The RO system operates at 80% recovery rate. At average flux of 26l/m2-hr, 7 elements per

vessel the concentrate flow in the last element is: 26X36.8X7X(1-0.80)/0.80 = 1674 l/hr

Potential CaSO4 deposit = 1674X1.36 = 2276 g/hr = 2.28 kg/hr

Assuming specific density of CaSO4 = 4g/cm3, volume of excess CaSO4 that could precipitate is

RO element is (2.28 X1000)/4 = 570 cm3/hr.

Free volume of feed channels in SW element is:

100 cm X100 cm X0.075 cm X 20 = 15000 cm3 (about 50%)

The above calculations shows that at the saturation conditions scale could lead in a short time to

complete blockage of tail elements.

Potential for scale formation is extremely important issue in brackish and wastewater

reclamation applications. At some locations, it may determine the maximum recovery rate for the

RO systems. Due to variability of water compositions and limited level of understanding of the

relevant salt solutions systems at saturation in RO conditions, it is quite difficult to make

accurate predictions about scaling. RO industry adopted limits for individual salts based

literature data and some field experience (Table 5.5.6. – Conservative limits). Due to lack of

accurate analytical models, developed for RO applications, these limits include significant safety

margins.

Manufacturers and suppliers of scale inhibitors are continuously introducing new scale inhibitors

that enable operation at higher levels of concentrations then those initially proposed by

membrane manufacturers (Table 5.5.6. – Possible limits). So far, the experience with the

subsequent products introduced over the years was quite positive. Seldom could any problems of

scale formation be related to malfunction of scale inhibitor if applied according to manufacturer

specifications. Scale inhibitors prevent scale formation by retardation of the nucleation process

of scale forming crystals. The mechanism of prevention of crystal grow is either through

threshold effect, crystal structure distortion, dispersion or sequestration.

The dosing rate of scale inhibitor is determined by supplier of chemicals based on feed water

composition and recovery rate. Feed water analysis should include information on concentration

of iron. High concentration of iron containing compounds in concentrate may reduce

effectiveness of some scale inhibitors. Required concentration of scale inhibitor seldom exceeds

10 ppm of active ingredient in the concentrate stream.

Table 5.5.6. Practical limits of saturation values in RO applications

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



181

Compound Formula Ksp Conservative

limits

Possible limits

(*)

Calcium

sulfate

CaSO4 2.5X10-4 2.4 X Ksp 4 X Ksp

Barium Sulfate BaSO4 2.0X10-10 60 X Ksp 100 X Ksp

Reactive Silica H4SiO4 Not defined 160 ppm 250 ppm

Calcium

Carbonate

CaCO3 LSI < 0

S&DSI < 0

LSI < 1.8

S&DSI < 0.5

LSI < 2.5

S&DSI < 1.5

Calcium

Phosphate

Ca3PO4 CPSI < 1.0 < 1.0

(*) Saturation limits recommended by suppliers of scale inhibitors

In brackish and wastewater reclamation RO units the designed recovery rate is determined based

on composition of feed water and saturation limits of sparingly soluble salts, as listed in Table

5.5.6.

For brackish applications, the common approach is to run computer projection program of one of

major manufacturers of RO membrane elements, which will flag the saturation limits.

The next step is to contact number of suppliers of scale inhibitors to get their opinion about

possible maximum recovery rate with a given feed water composition. The ultimate decision is

made by the system designer based on his confidence level, but the common practice is to follow

recommendations of supplier of scale inhibitors.

In seawater applications the recovery rate is limited by the osmotic pressure of the concentrate,

the feed pressure and required permeate salinity. The common practice is to design seawater RO

units for feed pressure not to exceed 82 bar (1200 psi). The preferred feed pressure limit is 70 bar

(1000 psi). Up to 82 bar a standard pressure vessels made of polymeric materials are available.

Above 82 bar feed pressure, special pressure vessels made of stainless steel tubing are required,

which in majority of cases would result in significant increase of system cost.

The common range of permeate recovery rate in RO application is listed in Table 5.5.7.

Table 5.5.7. Common ranges of permeate recovery rate in RO applications.

Membrane type Feed water source Common range of permeate

recovery rate

Nanofiltration membranes Low salinity wells 85% - 90%

Brackish membranes Brackish wells 75% - 85%

Brackish membranes Tertiary effluent 80% - 85%

Brackish membranes Second pass (RO permeate feed) 85% - 90%

Seawater membranes Atlantic – Pacific Ocean 50% - 55%

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



182

Seawater membranes Mediterranean 45% - 50%

Seawater membranes Arabian Gulf 40% - 45%

Effect of recovery rate on system operating parameters is summarized in Table 5.5.8.

Table 5.5.8. Process parameters affected by permeate recovery rate

Permeate salinity Increased recovery rate results in higher salinity of

permeate

Feed pressure Higher recovery rate results in higher feed pressure

System size Higher recovery rate results in smaller size of flow

related equipment (feed water supply, pumps and

piping). Has no effect on number of membrane

elements.

Membrane fouling rate Higher recovery rate may result in higher rate of

fouling including scaling

Frequency of membrane cleaning Higher recovery rate may result in higher frequency

of membrane cleaning due to higher rate of fouling

5.6. Average permeate flux

Similarly to the permeate recovery rate, permeate flux is very important operational

parameters that determines stability of membrane unit operation and the economics of the

desalination process.

The RO membrane units are designed based average permeate flux, which is the unit

permeate capacity divided by total membrane area installed in the system.

The selected permeate flux range should result in stable operation of the membrane unit. The

range of average permeate flux is part of membrane manufacturer design recommendations.

The average permeate flux range is selected by the system designer based on feed water

source which is related to the expected feed water quality.

The common ranges of average permeate flux according to water source is listed in Table

5.6.1.

Table 5.6.1. Common ranges of average permeate flux rate in RO applications.

Membrane type Feed water source Common range of average

permeate flux, l/m2/hr

Nanofiltration

membranes

Low salinity wells or surface water

with membrane pretreatment

26 – 29

Nanofiltration

membranes

Surface water – conventional

pretreatment

20 - 26

Brackish

membranes

Brackish wells 26 - 29

Brackish

membranes

Tertiary effluent 17 - 20

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



183

Brackish

membranes

Second pass (RO permeate feed) 30 – 37

Seawater

membranes

Open intake – conventional

pretreatment

12 - 15

Seawater

membranes

Open intake – membrane

pretreatment

15 - 19

Seawater

membranes

Beach wells 15 - 19

The value of permeate flux affects performance of the membranes and economics of the

desalination process.

Process parameters affected by the value of permeate flux are listed in table 5.6.2.

Table 5.6.2. Process parameters affected by permeate flux

Permeate salinity High flux rate results in lower salinity of

permeate

Feed pressure High flux rate results in higher feed pressure

System size High flux rate results in smaller size of

membrane unit due to lover number of

membrane elements and pressure vessels

Membrane fouling rate High flux rate results in higher rate of

fouling

Frequency of membrane cleaning High flux rate results in higher frequency of

membrane cleaning due to higher rate of

fouling

5.7.Membrane unit design procedure

The objective of membrane unit design is to design a system that will produce required

quantity and quality of product water in the whole range of feed water quality and

temperature as listed in project specifications.

The designed have to consider the most extreme conditions of raw water quality, salinity and

temperature in defining membrane unit configuration and the operating parameters.

The basic design parameters are recovery rate, average permeate flux, membrane type and

array, train sizes, number of trains and configuration, selection of major equipment (pumps,

instrumentations, etc..) and unit layout.

5.7.1.1.Permeate capacity and permeate quality limits

The membrane unit is sized according to specification of system capacity, taking into

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



184

consideration plant load factor and possible schedule of product water flow delivery

fluctuation. The load factor is based on planed system down time (membrane cleaning,

equipment maintenance, etc..) and additional time set up for emergencies (safety margin).

Usually down time imposed by the client and power outages are excluded from the above

considerations.

Well operated system could have load factor as high as 95%, however, majority of the

desalination systems have load factor in the range of 90% - 95% of the available

operating time.

Permeate quality limits are defined by specification of the final product water quality. In

defining the limits of permeate quality, considerations must be given to the following

conditions:

- Fluctuation of raw water quality

- Fluctuation of raw water temperature

- Aging of membrane elements and expected increase of salt passage

- Increase of product water salinity due to addition of chemicals during the permeate

stabilization process.

- Safety margin applied to mitigate project risk factor.

Permeate quality is calculated utilizing computer programs provided by membrane

manufacturers. The programs calculate required feed pressure and permeate quality.

System designer has to clarify with membrane manufacturer if the values calculated by

the computer program will be accepted as the membrane manufacturer warranty limits or

additional safety margin will be applied.

5.7.1.2.Selection of average permeate flux, recovery rate and array

The average permeate flux and recovery rate are determined according to the application

type, source of raw water and type of pretreatment applied. The array of pressure vessels

is function of number of stages and system size.

The representative ranges of average flux rate, recovery rate, number of stages and

number of elements per vessel for typical RO applications are listed in Table 5.7.1.

The average flux rate is determined based on water source and pretreatment type, which

stipulates expected feed water quality.

The recovery rate of nanofiltrtation, brackish and wastewater applications is determined

based scaling potential of feed water. In seawater applications the recovery rate is in the

range of 40 – 55%, increasing with decreased salinity of seawater source.

The number of stages is function of recovery rate and number of elements per vessel.

Some nanofiltrtaion membrane units that operate at recovery rate of 90% could have up

to 4 stages. With number of stages of 3 to 4, the recommended number of elements per

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



185

pressure vessel is 6. This is in order to maintain acceptable pressure drop along the

membrane unit. However, membrane unit includes interstage booster pump, than the

number of elements per vessel can be increased to 7.

Table 5.7.1. Range of RO design parameters according to application and feed water

source

Application Feed water

source

Average flux

range, l/m2/hr

Common

range of

permeate

recovery rate

Number of

stages

Number of

elements per

vessel

Nanofiltration Low salinity

wells

26 – 29 85% - 90% 3 – 4 6 – 7

Brackish Brackish

wells

26 – 29 75% - 85% 2 - 3 6 – 7

Wastewater

reclamation

Secondary

effluent with

membrane

filtration

17 – 20 80% - 85% 2 – 3 6 - 7

Second pass

RO

Second pass

(RO permeate

feed)

30 – 37 85% - 90% 2 – 3 6 - 8

Seawater Surface

intake

12 – 15 40% - 55% 1 6 - 8

Seawater Beach wells

or intake with

membrane

filtration

15 – 19 40% - 50% 1 6 – 8

Membrane units in brackish systems that operate at recovery rate up to 85% are

configured as 2 – 3 stage units. At recovery rate higher than 85%, usually a 4th stage is

added, frequently with a dedicated booster pump.

For configuration of 2 – 3 stages the number of elements per vessel could be 6 – 7,

usually 7 element configuration for a 2 stage unit.

Wastewater reclamation membrane units are configured in similar way as in brackish

applications.

Membrane units operating as a second pass RO, treat very clean feed water (RO

permeate). Therefore, the common limitations of average flux rate and recovery are not

applicable here. In most cases the second pass membrane units are configured as a two

stage units with 7 – 8 elements per vessel.

Presently, majority of seawater RO membrane units are configured as a single stage units

with number of elements per vessel 7 – 8.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



186

5.7.1.3.Selection of membrane type

Considerations for selection of membrane elements include type of application, parameters of

feed water and required performance in respect of permeate quality and feed pressure.

Membrane are categorized according to applications they are effective:

Nanofiltration for color removal

Nanofiltration for sulfate reduction

Nanofiltration for hardness reduction

Low pressure brackish RO

Low fouling brackish

High rejection brackish RO

Low pressure seawater RO

High rejection seawater RO

Most of the membrane manufacturers offer membrane elements in all above categories with very

similar performance. The membrane elements from different manufacturers have the same

dimensions and are interchangeable in pressure vessels. Therefore, the EPC or end user have

significant flexibility with selection of membrane supplier among major membrane

manufacturers.

Table 5.7.2 includes examples of correlation between membrane elements and applications.

Except for very specific nanofiltrtaion applications, in other application there is number of

membrane elements type that can be selected. The listing provided in Table 5.7.2. is only

representative. Additional commercial membrane elements are available from major

manufacturers.

Table 5.7.2. Examples of representative membrane elements models according to applications

Application Membrane selection Representative membrane

Reduction of color and

organics

Loose nanofiltration (tight

UF)

HYRACoRe

Reduction of sulfates Loose nanofiltration SR90-400

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



187

Reduction of hardness, sulfates, iron and organics

Regular nanofiltration, Low fouling nanofiltration

NF90, NF-270, ESNA-LF, SU620F

Desalting low salinity

brackish water

Low pressure brackish RO ESPA4+, ESAP2+,

BW30XLE-440, TMG20-430

Wastewater reclamation Low pressure brackish RO, Low fouling RO

ESPA2+, LFC3, BW30-400FR, TMG20-430

Desalting high salinity

brackish, nitrate reduction

High rejection brackish RO BW30-400, ESPA2, CPA3,

Desalting low salinity, low temperature seawater

Low pressure seawater RO SW30XLE-400, SWC5

Desalting high salinity, high

temperature seawater

High rejection seawater RO SW30HR-380, SWC4+

Table 5.7.3. Representative offering of nanofiltration membrane elements

Element model Hydracore ESNA-LF SU620F NF-90 NF-270

Membrane area, m2

(ft2)

37.1

(400)

37.1 (400) 37.1 (400) 37.1 (400) 37.1

(400)

Permeate flow, m3/d

(gpd)

31.0

(8,200)

29.5 (7,800) 21.9

(5,800)

37.9 (10,000) 47.3

(12,500)

Salt rejection, 50.0 80.0 55.0 97.0 97.0

Test flux rate, l/m2-

hr (gfd)

34.8

(20.5)

33.2 (19.5) 24.7 (14.5) 42.5 (25.0) 55.9

(32.9)

Permeability, l/m2-

hr-bar (gfd/psi)

7.7

(0.31)

7.2 (0.29) 8.7 (0.35) 11.9 (0.48) 15.7

(0.63)

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



188

Relative salt

transport: salt

passage*flux rate

17.4

(10.2)

6.6 (3.9) 11.1 (6.5) 1.3 (0.8) 1.7 (1.0)

Table 5.7.4. Representative offering of brackish membrane elements (*)

Element

model

ESPA2+ ESPA4+ TMG20-430 BW30-

XLE440

BW30 LE-

440

Membrane

area, m2

(ft2)

40.0 (430) 40.0 (430) 40.0 (430) 40.9 (440) 40.9 (440)

Permeate

flow, m3/d

(gpd)

41.6 (11,000) 49.2

(13,000)

41.6 (11,000) 48.1

(12,700)

48.1

(12,700)

Salt

rejection,

99.60 99.60 99.50 99.0 99.30

Test flux

rate, l/m2-hr

(gfd)

43.5 (25.6) 51.3 (30.2) 43.5 (25.6) 49.1 (28.9) 49.1 (28.9)

Permeability,

l/m2-hr-bar

(gfd/psi)

5.0 (0.20) 8.2 (0.33) 6.2 (0.25) 7.7 (0.31) 6.0 (0.24)

Relative salt

transport:

salt

passage*flux

rate

0.261

(0.153)

0.308

(0.181)

0.218

(0.128)

0.491

(0.289)

0.344

(0.202)

(*) Brackish membrane elements are also used for wastewater reclamation applications.

Except for nanofiltration elements offering, where the range of element performance is relatively

wide, in the brackish and seawater groups of element offering the performances of elements

belonging to the same group are very similar.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



189

The element nominal performances are defined through limited number of parameters. These are:

membrane area, permeate flow and salt rejection.

The nominal permeate flow and nominal salt rejection values depend not only on membrane

property but also on nominal test conditions. The nominal test conditions applied in testing of

seawater elements are similar. However, the testing conditions for brackish elements vary

significantly in respect of feed salinity and feed pressure.

For this reason, comparison of membrane elements in a given category cannot be based nominal

performance alone. Good indicator for comparison of various membrane elements belonging to

the same category (i.e. brackish, seawater ..) are parameters of water permeability (also called

specific flux) and relative salt transport.

Table 5.7.5. Representative offering of seawater membrane elements

The water permeability (WP) is calculated by dividing element nominal permeate flux by the net

driving pressure (NDP). The units of WP are l/m2/hr/bar.

Element

model

SWC4+ SWC5 TM820-400 SW30HR-LE SW30HR-

XLE

Membrane

area, m2 (ft2)

37.1 (400) 37.1 (400) 37.1 (400) 37.1 (400) 37.1 (400)

Permeate

flow, m3/d

(gpd)

24.6 (6,500) 34.1 (9,000) 24.6 (6,500) 26.5 (7,000) 34.1(9,000)

Salt rejection, 99.80 99.80 99.75 99.75 99.70

Test flux rate,

l/m2-hr (gfd)

27.6 (16.3) 38.2 (22.5) 27.6 (16.3) 31.3 (18.4) 38.2 (22.5)

Permeability,

l/m2-hr-bar

(gfd/psi)

1.0 (0.04) 1.5 (0.06) 1.0 (0.04) 1.2 (0.05) 1.5 (0.06)

Relative salt

transport: salt

passage*flux

rate

0.055 (0.032) 0.076 (0.045) 0.069 (0.041) 0.078 (0.046) 0.114 (0.067)

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



190

Nominal permeate flux (NFLUX) is the nominal permeate flow (NPF) divided by the membrane

area of the element (MA).

NFLUX = NPF/MA (5.27)

NDP = Pf – Pos – 0.5* DP (5.28)

WP = NFLUX/NDP (5.29)

Where Pf is nominal test feed pressure

Pos is osmotic pressure of average feed solution during the test

DP is pressure drop along the element during test, usually 0.2 bar.

The relative salt transport (RST) is calculated as a product of nominal flux and salt passage (SP).

RST = NFLUX * SP

Example 5.1. Calculation of water permeability and relative salt tratnsport

Nominal element performance

Permeate flow 48 m3/day

Salt rejection 99.5%

Membrane area 40 m2

Nominal test conditions

Feed salinity 2,000 ppm NaCl

Feed pressure 15.5 bar

Feed temperature 25 C

Recovery rate 15%

NFLUX = 48 * 1000/(24 * 40) = 50 l/m2/hr

AFS = 2000 * (1 + 1/(1 – 0.15))/2 = 2,176 ppm NaCl

Pos = 0.00076*2176 = 1.65 bar

NDP = 15.5 – 1.65 – 0.1 = 13.75 bar

WP = 50/13.75 = 3.63 l/m2/hr/bar

RST = 50 * (100 – 99.5) = 25

Values of WP and RST are indicative of membrane performance at field conditions.

RO system utilizing membrane elements with higher value of water permeability will require

lower feed pressure.

RO system utilizing membrane elements with higher value of relative salt transport will produce

permeate of higher salinity.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



191

The values of WP and RST are related. Higher value of WP is usually associated with higher

value of RST.

5.7.1.4.Membrane train size and configuration

Size of RO membrane train is determined based on system permeate capacity, desired flexibility

of variability of output capacity and convenience of system maintenance (membrane cleaning).

Very small systems up to few thousands m3/day capacity are usually designed as a single train

units. The contingency capacity is realized by maintaining sufficient supply of spare equipment,

either installed or stored in the plant warehouse.

Larger systems are built as a multi train configurations. Depending on size the number of

membrane trains can vary from 2 to 10 – 20. In principle, larger number of trains results in

increasing system cost. The additional cost results from pumping equipment, valves,

instrumentation and control equipment associated with each train.

In addition to consideration of flexibility of water production, the train size is also determined by

the logistic of membrane cleaning. Membrane train with large number of pressure vessels will

require large size cleaning systems to conduct membrane cleaning. This is one of the reason that

the train size is limited to about 200 pressure vessels in a very large RO systems.

In large capacity brackish and nanofiltration systems two hydraulic configurations of membrane

trains are possible, as shown on Figure 5.7.1.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



192

Stag 2

Stag 3

Stage 1

Stage 3

Stage 3

Stage 1

Stage 2

Stage 3

Stage 1

Stage 2

Stage 3

Stage 1

Stage 2

Stage 3

Stage 1

Stage 2

Stage 1

Stage 2

Stage 1

Stage 2

Satge 3

Stage 3

Stage 3

Stag 1

Figure 5.7.1. Alternative configuration of membrane trains in large capacity RO plants.

Accordingly, the subsequent desalination stages (1 – 3) can be aggregated together and form

integrated membrane trains (left configuration). Alternatively, the last stages can be grouped

together forming a separate train(s) that process the combine concentrate from trains consisting

of stages 1 and 2. The second alternative reduces operational flexibility of water production but

provides some saving of the system capital cost.

Regarding the type of pressure vessels used, it is current tendency to utilize more multiport

pressure vessels over traditional side port configuration in order to reduce cost of RO trains as it

is discussed in section 5.2.2.

5.7.2. Utilizing computer programs in membrane unit design

The calculations of projected performance of RO membrane unit is conducted utilizing computer

programs developed by membrane manufacturers. The information on membrane products and

computer programs for projection of membrane unit performance are available, free of charge,

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



193

form all commercial RO membrane manufacturers through their web pages. Various performance

projection programs are quite similar in functionality, design of user interface, input values

required and output format.

The computer program provides projected values of feed pressure and composition of permeate

and concentrate streams based on system design parameters and according to nominal

performances of selected membrane element.

The input to computer program includes the following information:

- Feed water analysis including water pH and temperature

- Type of water source

- Required permeate flow

- Designed recovery rate

- Selection element type

- Membrane array (number of stages, number of pressure vessels per stage, number of

elements per vessel)

- Fouling factor or membrane age

- Feed water pH

- Feed water temperature

Figure 5.7.2. Computer projections program – water analysis data entry screen

Composition of feed water is entered through water analysis screen (Figure 5.7.2.). The program

converts concentration units, if necessary, and enables adjustment of positive – negative ions

balance.

Following the entry to the water analysis screen, the next step is to entry process parameters,

select membrane model type and membrane array in the membrane unit design screen (Figure

5.7.3).

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



194

Figure 5.7.3. Computer projection program – process parameters and membrane array entry

screen

After entering the required permeate flow and recovery rate, the membrane element type is

selected from the look up window (figure 7.7.4) and the array is being adjusted till the average

permeate flux value is within the range recommended for the type of feed water being processed.

Partial results of calculations are displayed on the screen (Figure 5.7.5) and complete results are

provided through a printout.

In addition to the calculated process parameters, both the screen display of results and the

printout include warnings if some of the process parameters are outside limits recommended by

the membrane manufacturer.

If the warnings are displayed the design of the membrane has to be modified. In case if warnings

pertain to excessive saturation limits than the usual approach is to reduce feed water pH (if

calcium carbonate scaling is of concern) or reduce recovery rate if other constituents (CaSO4,

SiO2, ..) has high potential for precipitation from the concentrate stream.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



195

Figure 5.7.4. Computer projection program – membrane elements look up table

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



196

Figure 5.7.5. Computer projections program – screen display of calculation results

The performance projection printout (Figure 5.7.6) provides detail information on the projected

performances that include pressures, chemical composition and hydraulic parameters of the feed,

permeate and concentrate stream. The printout also includes information on the dosing rate of

acid or caustic designed to be used for pH adjustment of the feed stream.

Some of the computer programs provide option to calculate energy requirement of the high

pressure feed pumps and energy recovery devices.

Addition calculations option included in some program provides capability to calculate chemical

dosage in the post treatment process required to stabilize permeate to prevent corrosion of

product water distribution piping system.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



197

BASIC DESIGN

RO program licensed to: Calculation created by: Project name: Low salinity Permeate flow: 5000.00 m3/d HP Pump flow: 245.1 m3/hr Raw water flow: 5882.4 m3/d Feed pressure: 10.6 bar Permeate recovery: 85.0 % Feedwater Temperature: 25.0 C(77F) Feed water pH: 7.0 Element age: 3.0 years Chem dose, ppm (100%): 0.0 H2SO4 Flux decline % per year: 7.0 Fouling factor: 0.80 Salt passage increase, %/yr: 10.0 Average flux rate: 24.8 lm2hr Feed type: Well Water

Stage Perm. Flow/Vessel Flux Beta Conc.&Throt. Element Elem. Array

Flow Feed Conc Pressures Type No. m3/hr m3/hr m3/hr l/m2-hr bar bar

1-1 180.0 12.3 3.3 32.2 1.25 8.9 0.0 LEWABRANE101S

140 20x7

1-2 28.3 6.5 3.7 10.1 1.03 7.7 0.0 LEWABRANE101S

70 10x7

Raw water Feed water Permeate Concentrate Ion mg/l meq/l mg/l meq/l mg/l meq/l mg/l meq/l

Ca 96.0 4.8 96.0 4.8 0.803 0.0 635.4 31.7 Mg 11.7 1.0 11.7 1.0 0.098 0.0 77.4 6.4 Na 450.0 19.6 450.0 19.6 19.217 0.8 2891.1 125.7 K 6.5 0.2 6.5 0.2 0.442 0.0 40.8 1.0 NH4 0.0 0.0 0.0 0.0 0.000 0.0 0.0 0.0 Ba 0.000 0.0 0.000 0.0 0.000 0.0 0.000 0.0 Sr 0.000 0.0 0.000 0.0 0.000 0.0 0.000 0.0 CO3 0.1 0.0 0.1 0.0 0.000 0.0 0.6 0.0 HCO3 200.0 3.3 209.9 3.4 4.482 0.1 1374.2 22.5 SO4 158.4 3.3 158.4 3.3 1.021 0.0 1050.2 21.9 Cl 669.6 18.9 669.6 18.9 28.325 0.8 4303.5 121.4 F 0.2 0.0 0.2 0.0 0.023 0.0 1.2 0.1 NO3 0.0 0.0 0.0 0.0 0.000 0.0 0.0 0.0 B 0.00 0.00 0.000 0.00 SiO2 24.3 24.3 0.72 157.9 CO2 37.17 36.42 36.42 36.42 TDS 1616.8 1626.7 55.1 10532.4 pH 6.9 7.0 5.3 7.8

Raw water Feed water Concentrate CaSO4 / Ksp * 100: 3% 3% 33% SrSO4 / Ksp * 100: 0% 0% 0% BaSO4 / Ksp * 100: 0% 0% 0% SiO2 saturation: 17% 17% 113% Langelier Saturation Index -0.44 -0.32 2.07 Stiff & Davis Saturation Index -0.43 -0.31 1.61 Ionic strength 0.03 0.03 0.20 Osmotic pressure 1.1 bar 1.1 bar 7.1 bar

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



198

BASIC DESIGN




1-1 180.0 12.3 3.3 32.2 1.25 8.9 0.0 LEWABRANE101S

140 20x7

1-2 28.3 6.5 3.7 10.1 1.03 7.7 0.0 LEWABRANE101S

70 10x7

Stg Elem Feed Pres Perm Perm Beta Perm Conc Concentrate saturation levels

no. pres drop flow Flux sal osm CaSO4 SrSO4 BaSO4 SiO2 Lang. bar bar m3/hr lm2hr TDS pres

1-1 1 10.6 0.5 1.6 40.4 1.14 11.7 1.3 4 0 0 20 -0.2 1-1 2 10.1 0.4 1.5 37.8 1.10 13.2 1.5 4 0 0 23 0.0 1-1 3 9.8 0.3 1.4 35.2 1.16 14.9 1.7 5 0 0 27 0.2 1-1 4 9.5 0.2 1.3 32.6 1.18 16.9 2.1 7 0 0 33 0.4 1-1 5 9.3 0.2 1.2 29.8 1.20 19.3 2.5 9 0 0 40 0.7 1-1 6 9.1 0.1 1.1 26.6 1.23 22.5 3.2 12 0 0 51 0.9 1-1 7 9.0 0.1 0.9 22.7 1.25 26.8 4.1 16 0 0 65 1.2

1-2 1 8.7 0.2 0.7 18.0 1.11 29.6 4.6 18 0 0 72 1.4 1-2 2 8.5 0.2 0.6 14.7 1.10 32.9 5.1 21 0 0 80 1.5 1-2 3 8.3 0.1 0.5 11.9 1.09 36.6 5.6 24 0 0 88 1.6 1-2 4 8.2 0.1 0.4 9.3 1.08 40.7 6.1 26 0 0 96 1.7 1-2 5 8.0 0.1 0.3 7.0 1.06 45.2 6.5 28 0 0 102 1.8 1-2 6 7.9 0.1 0.2 5.1 1.05 50.1 6.8 30 0 0 107 1.9 1-2 7 7.8 0.1 0.1 3.6 1.03 55.3 7.1 32 0 0 111 1.9

Stag

e NDP

bar 1-1 7.2 1-2 2.8

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



199

BASIC DESIGN


********************************************************************* **** THE FOLLOWING PARAMETERS EXCEED RECOMMENDED DESIGN LIMITS: ***

*********************************************************************

Pass 1-1: Conc. polarization factor (beta) too high (1.25). Concentrate saturation of SiO2 too high (113%)

Concentrate Langelier Saturation Index too high (2.07)

The following are recommended general guidelines for designing a reverse osmosis system using Hydranautics membrane elements. Please consult

Hydranautics for specific recommendations for operation beyond the specified guidelines.

Feed and Concentrate flow rate limits

Element diameter Maximum feed flow rate Minimum concentrate

rate 8.0 inches 75 gpm (283.9 lpm) 12 gpm (45.4 lpm)

8.0 inches(Full Fit) 75 gpm (283.9 lpm) 30 gpm (113.6 lpm)

Concentrate polarization factor (beta) should not exceed 1.2 for standard elements

Saturation limits for sparingly soluble salts in concentrate

Soluble salt Saturation BaSO4 6000% CaSO4 230% SrSO4 800% SiO2 100%

Langelier Saturation Index for concentrate should not exceed 1.8

The above saturation limits only apply when using effective scale inhibitor. Without scale inhibitor, concentrate saturation should not exceed 100%.

Figure 5.7.6. Printout of calculation results

5.7.2.1.System performance safety margins

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



200

The results of computer calculations are basis for RO system design and terms of membrane

performance warranty. It is customary for membrane manufacturers to provide system

performance warranty with some safety margins, applied to the results calculated by their

programs. The usual range is 10 – 20% below calculated permeate salinity and 5 – 10% above

calculated feed pressure.

The designer should conduct system performance calculations at two extreme conditions:

- Highest feed water salinity and the lowest feed water temperature to determine the

maximum feed pressure that will be required to produce the designed permeate output.

- Highest feed water salinity and the highest feed water temperature to determine the

maximum permeate salinity

Additional safety margin should be applied by the designer of RO system to account for increase

of salinity as a result of post treatment.

In brackish water systems the salinity increase due to post treatment is usually small. In sweater

treatment the increase of permeate salinity in post treatment can be significant, in the range of

100 – 150 ppm

5.7.3. Configuration of RO membrane unit for high feed salinity operation – 100,000

m3/day product water capacity.

Table 5.7.3.1. Basic process parameters of a 100,000 m3/day SWRO system

Parameter Value

Product capacity, m3/day 100,000

Feed Salinity, ppm TDS 45,000

Feed water temperature, C 20 – 36

Required permeate salinity, ppm TDS < 400

Required permeate boron concentration, ppm < 0.75

The RO membrane system will be configured as eight trains, partial two pass in a “split partial”

configuration.

The first pass permeate trains will produce 104,000 m3/day. It will operate at recovery rate of

45%. The second pass RO units will process about 23% of the first pass permeate. The second

pass trains will produce 20,000 m3/day permeate and operate at recovery rate of 85%. Feed pH

to the second pass RO will be increased to pH up to 10.5 to archive required level of boron in the

blended product water.

The system will be composed of eight first pass trains. The number of second pass trains usually

is smaller than the number of first pass trains. In systems where only part of first pass permeate

is processed by the second pass RO, The number of second pass trains will be half or less of the

number of trains in the first pass.

The schematic configuration of the RO trains arrangement, in the spilt – partial configuration, is

shown on Figure 5.7.3. For the simplicity of presentation the trains are shown as one first pass

train with the corresponding second pass train. In actuality, there will be two or more first pass

trains feeding one second pass train.

Details regarding feed water composition, membrane trains configuration, selection of membrane

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



201

elements and projected system performances are provided in the following computer projections.

46

31

Pass 29

Split partial configuration

2

58

7

Pass 1

Tag 1 2 3 4 5 6 7 8

Flow, m3/hr 18.3 1203.7 1203.7 122.5 122.5 662.0 532.3

TDS, PPM 45034 2889 45034 44416 420 420 82870 110

Pressure,

bar2.0 2.0 2.0 76.2 1.0 12.6 1.0 1.0

9

428.1

135

2.0

1185.4

Tag 1 2 3 4 5 6 7 8

Flow, m3/hr 18.3 1203.7 1203.7 122.5 122.5 662.0 532.3

TDS, PPM 45034 5673 45034 44455 868 868 82813 172

Pressure,

bar2.0 2.0 2.0 72.0 1.0 11.5 1.0 1.0

9

428.1

196

2.0

1185.4

Temp 20C

Temp 36C

Figure 5.7.7. Split partial configuration of a 12,500 m3/day SWRO train

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



202

SPLIT PARTIAL TWO PASS & Permeate THROTTLING(1ST STAGE) WITH Pressure/Work Exchanger

RO program licensed to: Calculation created by: Blended flow: 12558.8 m3/d Project name: Kuwait Permeate flow: 13000.00 2500.00 m3/d HP Pump flow: 1203.7 122.5 m3/hr Raw water flow: 28447.7 m3/d Feed pressure: 76.2 11.6 bar Permeate recovery: 45.0 85.0 % Feedwater Temperature: 20.0 C(68F) Total system recovery: 44.1 % Feed water pH: 7.5 10.5 Element age: 3.0 years Chem dose, ppm, ppm 28.2 15.6 Flux decline % per year: 7.0 5.0 Fouling factor: 0.80 1.00 Salt passage increase, %/yr: 10.0 5.0 Average flux rate: 13.8 29.0 lm2hr Feed type: Seawater - open intake



1-1 541.7 10.0 5.5 13.8 1.01 74.3 1.0 SWC4B MAX 960 120x8 2-1 83.8 15.3 4.8 32.0 1.20 8.7 1.0 ESPA2 MAX 64 8x8 2-2 20.4 12.9 6.1 20.8 1.10 5.9 0.0 ESPA2 MAX 24 3x8

Raw water Adjusted Water Feed water Permeate Concentrate ERD Reject Ion mg/l mg/l mg/l mg/l mg/l mg/l

Ca 554.0 545.7 561.1 0.342 1019.7 992.0 Mg 1893.0 1864.5 1917.3 1.168 3484.2 3389.7 Na 13360.0 13171.7 13543.4 40.327 24566.6 23900.9 K 582.0 573.9 590.0 2.201 1069.7 1040.7 NH4 0.0 0.0 0.0 0.000 0.0 0.0 Ba 0.000 0.000 0.000 0.000 0.000 0.0 Sr 13.000 12.804 13.167 0.008 23.928 23.3 CO3 27.1 8.2 8.5 0.006 15.4 16.9 HCO3 200.0 224.6 230.9 1.043 418.2 405.3 SO4 3373.0 3350.4 3445.2 2.102 6260.8 6089.2 Cl 25011.0 24652.7 25349.1 62.924 45994.6 44748.2 F 1.9 1.9 1.9 0.010 3.5 3.4 NO3 0.9 0.9 0.9 0.019 1.6 1.6 B 5.50 5.53 5.67 0.375 9.80 9.5 SiO2 1.1 1.1 1.1 0.00 2.0 2.0 CO2 0.85 3.87 3.87 3.10 0.00 3.87 TDS 45034.6 44413.8 45668.3 110.53 82870.1 80622.7 pH 8.1 7.5 7.5 9.3 8.0

Raw water Feed water Concentrate CaSO4 / Ksp * 100: 30% 31% 65% SrSO4 / Ksp * 100: 44% 45% 94% BaSO4 / Ksp * 100: 0% 0% 0% SiO2 saturation: 1% 1% 2% Langelier Saturation Index 1.45 0.80 1.87 Stiff & Davis Saturation Index 0.48 -0.17 0.87 Ionic strength 0.91 0.92 1.67 Osmotic pressure 32.3 bar 32.7 bar 59.3 bar

H.P. Differential of Pressure/Work Exchanger: 1.0 ba

r Pressure/Work Exchanger Leakage: 1 %

Pressure/Work Exchanger Pump Boost Pressure: 2.9 bar

Volumetric Mixing: 6 %

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



203

SPLIT PARTIAL TWO PASS & Permeate THROTTLING(1ST STAGE) PASS 1 WITH Pressure/Work Exchanger

RO program licensed to: Calculation created by: Project name: Kuwait Permeate flow: 13000.00 m3/d HP Pump flow: 1203.7 m3/hr Raw water flow: 28447.7 m3/d Feed pressure: 76.2 bar Permeate recovery ratio: 45.0 % Feedwater Temperature: 20.0 C(68F) Feed water pH: 7.5 Element age: 3.0 years Chem dose,ppm (100%) 28.2 H2SO4 Flux decline % per year: 7.0 % Fouling factor: 0.80 Salt passage increase, %/yr: 10.0 Average flux rate: 13.8 lm2hr Feed type: Seawater - open intake



1-1 541.7 10.0 5.5 13.8 1.01 74.3 1.0 SWC4B MAX 960 120x8

Raw water 1 Feed water 1 Permeate 1 Concentrate 1 Ion mg/l meq/l mg/l meq/l Back Front mg/l meq/l

Ca 554.0 27.6 561.1 28.0 1.25 0.42 1019.7 50.9 Mg 1893.0 155.8 1917.3 157.8 4.27 1.45 3484.2 286.8 Na 13360.0 580.9 13543.4 588.8 144.44 49.17 24566.6 1068.1 K 582.0 14.9 590.0 15.1 7.86 2.68 1069.7 27.4 NH4 0.0 0.0 0.0 0.0 0.00 0.00 0.0 0.0 Ba 0.000 0.0 0.000 0.0 0.000 0.000 0.000 0.0 Sr 13.000 0.3 13.167 0.3 0.029 0.010 23.928 0.5 CO3 27.1 813.9 8.5 0.3 0.01 0.00 15.4 0.5 HCO3 200.0 3.3 230.9 3.8 4.07 1.26 418.2 6.9 SO4 3373.0 70.3 3445.2 71.8 8.46 2.61 6260.8 130.4 Cl 25011.0 705.5 25349.1 715.1 248.77 76.71 45994.6 1297.5 F 1.9 0.1 1.9 0.1 0.04 0.01 3.5 0.2 NO3 0.9 0.0 0.9 0.0 0.07 0.02 1.6 0.0 B 5.50 5.67 1.28 0.44 9.80 SiO2 1.1 1.1 0.01 0.00 2.0 CO2 0.85 3.87 3.87 3.87 3.87 TDS 45034.6 45668.3 420.5 134.79 82870.1 pH 8.1 7.5 6.2 5.77 8.0

Raw water Feed water Concentrate CaSO4 / Ksp * 100: 30% 31% 65% SrSO4 / Ksp * 100: 44% 45% 94% BaSO4 / Ksp * 100: 0% 0% 0% SiO2 saturation: 1% 1% 2% Langelier Saturation Index 1.45 0.80 1.80 Stiff & Davis Saturation Index 0.48 -0.17 0.80 Ionic strength 0.91 0.92 1.67 Osmotic pressure 32.3 bar 32.7 bar 59.3 bar

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



204

SPLIT PARTIAL TWO PASS WITH Pressure/Work Exchanger & Permeate THROTTLING(1ST STAGE) PASS 2

RO program licensed to: Calculation created by: Project name: Kuwait Permeate flow: 2500.00 m3/d Feed pressure: 11.6 bar Permeate recovery ratio: 85.0 % Feedwater Temperature: 20.0 C(68F) Feed water pH: 10.5 Element age: 3.0 years Chem dose, ppm (100%) 15.6 NaOH Flux decline % per year: 5.0 % Fouling factor: 1.00 Salt passage increase, %/yr: 5.0 Average flux rate: 29.0 lm2hr Feed type: Seawater - open intake



2-1 83.8 15.3 4.8 32.0 1.20 8.7 1.0 ESPA2 MAX 64 8x8 2-2 20.4 12.9 6.1 20.8 1.10 5.9 0.0 ESPA2 MAX 24 3x8

Raw water 2 Feed water 2 Permeate 2 Concentrate 2 Ion mg/l meq/l mg/l meq/l mg/l meq/l mg/l meq/l


Raw water Feed water Concentrate CaSO4 / Ksp * 100: 0% 0% 0% SrSO4 / Ksp * 100: 0% 0% 0% BaSO4 / Ksp * 100: 0% 0% 0% SiO2 saturation: 0% 0% 0% Langelier Saturation Index -4.75 0.10 2.45 Stiff & Davis Saturation Index -4.73 0.12 2.37 Ionic strength 0.01 0.01 0.05 Osmotic pressure 0.3 bar 0.3 bar 2.2 bar

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



205

SPLIT PARTIAL TWO PASS & Permeate THROTTLING(1ST STAGE) WITH Pressure/Work Exchanger

RO program licensed to: Calculation created by: Blended flow: 12558.8 m3/d Project name: Kuwait Permeate flow: 13000.00 2500.00 m3/d HP Pump flow: 1203.7 122.5 m3/hr Raw water flow: 28447.7 m3/d Feed pressure: 72.0 9.8 bar Permeate recovery: 45.0 85.0 % Feedwater Temperature: 36.0 C(97F) Total system recovery: 44.1 % Feed water pH: 7.5 10.5 Element age: 3.0 years Chem dose, ppm, ppm 22.6 35.8 Flux decline % per year: 7.0 5.0 Fouling factor: 0.80 1.00 Salt passage increase, %/yr: 10.0 5.0 Average flux rate: 13.8 29.0 lm2hr Feed type: Seawater - open intake



1-1 541.7 10.0 5.5 13.8 1.01 70.2 1.0 SWC4B MAX 960 120x8 2-1 92.1 15.3 3.8 35.2 1.22 7.2 1.0 ESPA2 MAX 64 8x8 2-2 12.1 10.2 6.1 12.3 1.02 5.0 0.0 ESPA2 MAX 24 3x8

Raw water Adjusted Water Feed water Permeate Concentrate ERD Reject Ion mg/l mg/l mg/l mg/l mg/l mg/l

Ca 554.0 545.8 561.2 0.504 1019.6 991.9 Mg 1893.0 1865.0 1917.7 1.722 3483.8 3389.3 Na 13360.0 13187.4 13559.1 62.748 24551.9 23887.1 K 582.0 574.6 590.8 3.441 1068.6 1039.7 NH4 0.0 0.0 0.0 0.000 0.0 0.0 Ba 0.000 0.000 0.000 0.000 0.000 0.0 Sr 13.000 12.807 13.170 0.012 23.924 23.3 CO3 48.5 14.2 14.7 0.020 26.7 27.4 HCO3 200.0 219.8 225.9 1.520 408.2 395.8 SO4 3373.0 3345.8 3440.4 3.110 6249.7 6078.8 Cl 25011.0 24675.0 25371.4 98.027 45964.5 44719.9 F 1.9 1.9 1.9 0.016 3.5 3.4 NO3 0.9 0.9 0.9 0.038 1.6 1.6 B 5.50 5.65 5.79 0.692 9.52 9.3 SiO2 1.1 1.1 1.1 0.00 2.0 2.0 CO2 0.63 2.83 2.83 2.26 0.00 2.83 TDS 45034.6 44449.8 45704.1 171.85 82813.4 80569.3 pH 8.1 7.5 7.5 9.8 8.1

Raw water Feed water Concentrate CaSO4 / Ksp * 100: 27% 28% 59% SrSO4 / Ksp * 100: 40% 41% 86% BaSO4 / Ksp * 100: 0% 0% 0% SiO2 saturation: 1% 1% 1% Langelier Saturation Index 1.80 1.16 2.30 Stiff & Davis Saturation Index 0.78 0.15 1.27 Ionic strength 0.91 0.92 1.67 Osmotic pressure 34.0 bar 34.5 bar 62.5 bar

H.P. Differential of Pressure/Work Exchanger: 1.0 ba

r Pressure/Work Exchanger Leakage: 1 %

Pressure/Work Exchanger Pump Boost Pressure: 2.8 bar

Volumetric Mixing: 6 %

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



206

SPLIT PARTIAL TWO PASS & Permeate THROTTLING(1ST STAGE) PASS 1 WITH Pressure/Work Exchanger

RO program licensed to: Calculation created by: Project name: Kuwait Permeate flow: 13000.00 m3/d HP Pump flow: 1203.7 m3/hr Raw water flow: 28447.7 m3/d Feed pressure: 72.0 bar Permeate recovery ratio: 45.0 % Feedwater Temperature: 36.0 C(97F) Feed water pH: 7.5 Element age: 3.0 years Chem dose,ppm (100%) 22.6 H2SO4 Flux decline % per year: 7.0 % Fouling factor: 0.80 Salt passage increase, %/yr: 10.0 Average flux rate: 13.8 lm2hr Feed type: Seawater - open intake



1-1 541.7 10.0 5.5 13.8 1.01 70.2 1.0 SWC4B MAX 960 120x8

Raw water 1 Feed water 1 Permeate 1 Concentrate 1 Ion mg/l meq/l mg/l meq/l Back Front mg/l meq/l

Ca 554.0 27.6 561.2 28.0 2.60 0.62 1019.6 50.9 Mg 1893.0 155.8 1917.7 157.8 8.89 2.11 3483.8 286.7 Na 13360.0 580.9 13559.1 589.5 300.97 71.59 24551.9 1067.5 K 582.0 14.9 590.8 15.1 16.38 3.90 1068.6 27.4 NH4 0.0 0.0 0.0 0.0 0.00 0.00 0.0 0.0 Ba 0.000 0.0 0.000 0.0 0.000 0.000 0.000 0.0 Sr 13.000 0.3 13.170 0.3 0.061 0.015 23.924 0.5 CO3 48.5 1453.7 14.7 0.5 0.03 0.01 26.7 0.9 HCO3 200.0 3.3 225.9 3.7 8.17 1.79 408.2 6.7 SO4 3373.0 70.3 3440.4 71.7 17.35 3.79 6249.7 130.2 Cl 25011.0 705.5 25371.4 715.7 510.69 111.66 45964.5 1296.6 F 1.9 0.1 1.9 0.1 0.08 0.02 3.5 0.2 NO3 0.9 0.0 0.9 0.0 0.14 0.03 1.6 0.0 B 5.50 5.79 2.95 0.73 9.52 SiO2 1.1 1.1 0.02 0.00 2.0 CO2 0.63 2.83 2.83 2.83 2.83 TDS 45034.6 45704.1 868.3 196.26 82813.4 pH 8.1 7.5 6.6 5.96 8.1


Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



207

SPLIT PARTIAL TWO PASS WITH Pressure/Work Exchanger & Permeate THROTTLING(1ST STAGE) PASS 2

RO program licensed to: Calculation created by: Project name: Kuwait Permeate flow: 2500.00 m3/d Feed pressure: 9.8 bar Permeate recovery ratio: 85.0 % Feedwater Temperature: 36.0 C(97F) Feed water pH: 10.5 Element age: 3.0 years Chem dose, ppm (100%) 35.8 NaOH Flux decline % per year: 5.0 % Fouling factor: 1.00 Salt passage increase, %/yr: 5.0 Average flux rate: 29.0 lm2hr Feed type: Seawater - open intake




Raw water 2 Feed water 2 Permeate 2 Concentrate 2 Ion mg/l meq/l mg/l meq/l mg/l meq/l mg/l meq/l


Raw water Feed water Concentrate CaSO4 / Ksp * 100: 0% 0% 0% SrSO4 / Ksp * 100: 0% 0% 0% BaSO4 / Ksp * 100: 0% 0% 0% SiO2 saturation: 0% 0% 0% Langelier Saturation Index -3.46 0.87 3.20 Stiff & Davis Saturation Index -3.37 0.95 2.97 Ionic strength 0.02 0.02 0.10 Osmotic pressure 0.7 bar 0.7 bar 4.5 bar

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



208

6.0.Design for RO high pressure pump and ERD

The energy usage in RO system is an aggregate value of pumping energy for delivery of raw

water, head loses in pretreatment, energy to drive the high pressure pump, energy of treatment of

product water, energy of pumping product water to the distribution system and electric energy

required for the operation of auxiliary equipment (Figure 6.1).

Sizing of high pressure pumps and energy recovery devices is based design parameters of the RO

membrane units and configuration of the desalination plant.

The selection of type of pumping equipment, including selection of material of construction is

based on the type of application, specific site conditions and project specifications.

The general approach is to minimize number of pumping stapes in the plant through

configuration of proper hydraulic profile.

The hydraulic profile of the RO plant will vary with application and site conditions. In most

cases the pumping requirement will consist pumping to deliver raw water to the plant site with

sufficient head to pass the pretreatment step.

6.1.Raw water supply and transfer pumps

In brackish systems treating well water this pressure bust will be provided by well pumps that

would generate sufficient head for the raw water to pass the cartridge filters with effluent

pressure within the suction head required by the high pressure pump.

In RO systems in which the raw water is stored initially in the raw water reservoir, there will be a

transfer pump to generate head for passing raw water through the pretreatment system.

Similar consideration of pumping system configuration will apply to brackish system operating

on surface water.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



209

Etot = Erw + Eprt + Ehp + Eprod + Eserv + Eaux

NF BRO SWRO

35%

21% 30%

52%

15% 22%

80%

Feed supply Prod. pumping

E ~ 0.9 kwhr/m3

(3.4 kwhr/kgallon)

E ~ 1.2 kwhr/m3

(4.5 kwhr/kgallon)

E ~ 3.5 kwhr/m3

(13.2 kwhr/kgallon)

8%6%

HP Pump

Figure 6.1. Energy usage in RO desalination systems

In this case the pretreatment will be more extensive and could include media filtration followed

by cartridge filters. In case of pressurized media filtration only one pumping step after raw water

storage reservoir will be required to pass raw water through the pretreatment.

In case of gravity filters, the approach is to use gravity filters and the filtrate effluent clear well

as the buffer reservoir. This way the number of pumping steps in the pretreatment can be limited

to pumping from the filtrate clear well through cartridge filters to the suction of high pressure

pumps.

The pretreatment in seawater RO systems, treating seawater from beach wells, is configured in a

similar way as it is in the brackish plants. Accordingly the pumping requirements in the

pretreatment section of the plant are similar as it is in the brackish plants.

However, majority of seawater RO plants treat seawater delivered from the open sea intakes. The

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



210

hydraulic profile and pumping requirement will depend on type of pretreatment utilized.

The possible alternatives could include:

- Gravity media filtration followed by cartridge filters

- Submersed (vacuum driven) membrane filtration

- Pressurized medial filtration followed by cartridge filters

- Pressurized membrane filtration

In the above first two pretreatment alternatives the gravity media filtration and submersed

membrane filtration units provide the required buffer storage capacity and the pressure bust is

provided by the transfer pump after the filtrate storage clear well.

In case of pressurized media filtration or pressurized membrane filtration pretreatment system,

the usual configuration is to have a raw water storage reservoir prior to pretreatment system.

Transfer pumps at the outlet of storage reservoir provide pressure bust required to pass the

pretreatment system. Additional pumping units will be require after the pretreatment to pump

feed water to the suction of high pressure pumps.

Selection of type of transfer pumps will depend on site conditions and preference of the system

designer. The current tendency is to utilize split case one – two impeller horizontal pumps over

the vertical can turbine pumps, whenever possible. The horizontal pumps, having smaller

volume, are built utilizing smaller weight of metal alloy, therefore, are usually less expensive

than vertical can pumps.

Vertical pumps can be installed on the top of clear well, where horizontal pumps require side

access, which is not always available at sites with limited area.

The selection of materials of construction of transfer pumps will depend on the salinity of raw

water. In brackish applications stainless steel grade, equivalent to 316 L will be sufficient. In

seawater applications the pumps should be made of high alloy steels, equivalent to the super

duplex alloy.

If variability of feed pressure in RO unit is expected, due to fluctuation of feed water temperature

and/or salinity, the system designer should have decision which pumping stage in the system will

provide the variable head. Accordingly, the pump designated to provide the variable head will be

driven by electric motor equipped with a variable frequency drive (VFD). As the VFD introduces

additional energy transfer inefficiency in the range of 2%, the general approach is to install a

variable speed drive on the lower energy demand motors in the system to reduce efficiency

losses.

Except for small systems the transfer pumps are configured as a group of parallel unit pumping

to a common manifold rather than operate as dedicated pump to individual membrane trains. In

this arrangement the transfer pumps could be bigger and it is usually sufficient to have one spare

transfer pump for 4 – 5 transfer pumps in operation.

6.2.High pressure pumps

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



211

The decision on selection and configuration of high pressure pump will be driven by similar

considerations as discussed in the above section regarding the transfer pumps. As the high

pressure pumps are main contributor to the plant total energy use, it is important to select pumps

of high hydraulic efficiency.

In brackish and nanofiltrtaion applications the common configuration is to have high pressure

pump dedicated to individual trains. In seawater applications high pressure pumps can be

configured as train dedicated pumps or group in a “pressure center” configuration, as shown in

Figure 6.2.

In pressure center configuration the high pressure pumps form a group of pumps pumping to the

common high pressure feed manifold. The number of high pressure pumps is significantly

smaller than number of RO trains. The pumps are of high capacity and usually of high efficiency.

Reduction of number of high pressure pumps and high efficiency results in reduced capital cost

and reduced energy consumption.

Figure 6.2. Pressure centers configuration of a large capacity SWRO plant.

Due to high efficiency of the pumps and high efficiency of modern energy recovery devices,

there is little energetic penalty for reducing recovery rate and the recovery rate can vary in

relatively wide range with very small effect on the energy usage of the system.

Accordingly when 1 – 2 trains are taken out of operation (if for example a cleaning of membrane

elements is required), the system will continue to operate with the same number of pumps and

energy recovery devices but at lower recovery rate. The reduced recovery rate will result in

lower average osmotic pressure of the feed – concentrate stream. Therefore at a constant feed

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



212

pressure the remaining trains in operation will produce additional quantity of permeate, as

compared to the regular operating conditions.

6.3.Optimized control methods for high pressure pump discharge head and

capacity

6.3.1. Application of energy recovery devices (ERD) in RO systems

Energy recovery devices (ERD’s) are applied in RO systems for the purpose to reduce energy

consumption. Addition of ERD increases system cost. Therefore, energy savings should be

sufficient to provided net economic benefits as calculate through life cycle cost.

Operation of ERD in RO membrane unit results in energy saving according to efficiency of

ERD, available concentrate flow and concentrate pressure.

Moving from seawater application to brackish application, both the flow rate of concentrate and

concentrate pressure are being reduced. In parallel to reduced availability of discharged energy

the economic incentive of utilizing ERD decrease as well.

Example of potential energetic benefits in RO applications is illustrated through results listed in

Table 6.1.

Table 6.1. Calculation of energy total energy usage in RO systems of permeate capacity of

40,000 m3/day

Application type NF RO low

salinity

RO high

salinity

Seawater RO

System recovery rate, % 85% 80% 65% 50%

Raw water pressure, bar 5 5 5 3

Feed pressure, bar 7 15 25 65

Concentrate pressure, bar 2 10 20 62

Permeate pumping pressure, bar 5 5 5 5

Energy, raw water, kwhr 364.2 387.0 476.3 379.0

Energy, feed pump, kwhr 437.1 1083.6 2286.3 8084.3

Energy recovered from the

concentrate, kwhr (%)

-9.9 (0.9) -88.4 (5.2) -390.6 (14.5) -2289.3

(35.3)

Energy product, kwhr 309.6 309.6 309.6 309.6

Energy total, kwhr 1101.1 1691.8 2681.3 6483.4

Specific energy, kwhr/m3 (kWhr/kgallon)

0.66 (2.5) 1.02 (3.9) 1.61 (6.1) 3.89 (14.7)

It is evident from the results listed in Table 6.1 that application of ERD has noticeable effect on

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



213

overall power consumption only in high pressure, low recovery RO applications.

6.3.1.1.Selection of ERD

All modern ERD’s have a proven record of providing reliable operation in field conditions.

Therefore, the selection of type ERD for given application is mainly based on economic benefits.

The important parameter in evaluation is the local electricity rate. At location with low electricity

rate utilization of low cost and lower efficiency energy recovery equipment will more beneficial.

At locations of high electricity rates, energy recovery equipment with premium efficiency will be

more beneficial.

Additional considerations are maintenance cost, footprint, simplicity of operation and hydraulic

considerations.

6.3.1.2.Pelton wheel

The Pelton Wheel ERD is shown on Figure 6.3.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



214

Figure 6.3. Pelton Wheel.

The device consist of drum with buckets mounted on a shaft. The concentrate exits the

concentrate nozzles under pressure and impinges the buckets. The kinetic energy of the

concentrate stream creates rotation of the drum and the shaft that is connected through the

electric motor to the high pressure pump (Figure 6.4) provides rotation energy to the electric

motor, reducing electric load required for operation of the high pressure feed pump.

The Pelton Wheel chamber, where the rotor is located, operates under atmospheric pressure.

Therefore, the Pelton Wheel equipment has to be installed at elevation that will provide sufficient

hydrostatic head for concentrate to flow under gravity at sufficient velocity to the concentrate

outfall.

It is important that the outlet channel, located below the Pelton Wheel device, will be designed

properly to prevent foaming of the concentrate as it flows out of the Pelton Wheel rotor. At some

installations an excessive foaming has been experienced as shown on Figure 6.5. .

Pelton

wheel

Concentrate

pipe

Feed water

pipe

High pressure

pump

Electric

motor

Pump

discharge pipe

Pumping system at the Larnaca plant

Figure 6.4. Pelton Wheel – electric motor – high pressure pump unit.

The Pelton Wheel can receive concentrate flow from number of RO trains, however, the usual

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



215

operation mode is according to the configuration shown in Figure 6.4, an integrated unit

connected to a single train.

In the past Pelton Wheel ERD has been used both in brackish and seawater applications. Today

their application is limited to SWRO almost exclusively.

Hydraulic efficiency of Pelton Wheel is the range of 84% and up to 88% in very large units.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



216

Figure 6.5. Concentrate foaming at the Pelton Wheel outlet

Energy requirement (E) of a RO membrane unit equipped with Pelton Wheel is calculated

according to the following equation. The units are kWhr/m3 of permeate.

E = 0.02777* (Pf/(p * m * R) * f - Pc * (1 – R) * t/m * c ) (6.1)

Where: Pf is feed pressure

DPc is the differential pressure of the concentrate available for the ERD

p, m and t are efficiencies of high pressure pump, electric motor and ERD respectively

f and c are densities of the feed and concentrate streams

6.3.1.3.Turbocharger

Hydraulic Turbocharger is configured as two impellers with the blades in opposite directions

connected through the common shaft (Figure 6.6).

Figure 6.6. Configuration of Hydraulic Turbocharger

The turbine impeller receives kinetic energy of the concentrate that induces rotation of the

turbine impeller that provides torque to the pump impeller which creates differential pressure

boost of the feed stream.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



217

Brackish RO train equipped with Hydraulic turbocharger in interstage position is shown on

Figure 6.7. The pictures also shows other components of turbocharger assembly that includes

two motorized valves: the bypass valve and backpressure control valve.

Hydraulic Turbocharger operating in seawater unit is shown in Figure 6.8. In this case the

turbocharger unit is placed after the high pressure pump providing pressure boost to feed stream.

Relation for calculation of pressure boost available from Hydraulic Turbochrger is listed on

Figure 6.9. The pressure boost is function of turbocharger efficiency, available concentrate

pressure and ration of flow of concentrate to the feed stream.

The turbocharger consist of two turbine impellers, so the combined efficiency is multiplier of

efficiency of individual impellers. Assuming maximum efficiency of turbine device of 0.9, the

ConcentrateInterstage

Figure 6.7. Brackish RO train with Hydraulic Turbocharger in the interstage position

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



218

Figure 6.8. Hydraulic Turbocharger positioned after high pressure pump in seawater RO unit

combined efficiency of the ERD will be 0.81 (or 81%). Such high efficiency is seldom

achievable. In large seawater units efficiency could reach 78% - 80%. In smaller brackish units

efficiency of 65% - 75% is more common.

P = Tef Rcf (Pc – P e)

P – pressure boost

Tef – turbocharger efficiency

Rcf – ratio of concentrate to feed flow (or interstage flow)

Pc – concentrate pressure at the RO unit exit

Pe – concentrate pressure at the turbocharger exit

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



219

Figure 6.9. Examples of configurations of seawater (left) and brackish (right) RO units with

Hydraulic Turbocharger

P = Tef Rcf (Pc – P e)

Example - seawater RO

P = 0.78 * (50/100) * (67.0 – 0.3) = 26 bar (377 psi)

Example - brackish RO

P = 0.78(16/40)(13.0 – 0.3) = 3.9 bar (56 psi)

Figure 6.10. Example of calculations of pressure boost provided by Hydraulic Turbocharger in

seawater and brackish RO membrane unit

The latest development is Hydraulic Turbocharger unit equipped with electric motor (figure

6.11). Such unit provides additional flexibility of increasing pressure boost beyond what is

available through recovery of energy of the concentrate stream.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



220

Figure 6.11. Hydraulic Turbocharger equipped with electric motor.

6.3.1.4.Pressure exchangers (isobaric devices)

The isobaric energy recovery devices are positive displacement energy recovery – pumping units

or very high hydraulic efficiency, frequently reaching 94 % - 96% even in small size units.

The schematic configuration of RO membrane unit with isobaric device is shown on Figure 6.12.

In this configuration the freed water stream is split into two streams. One stream (F1) is directed

to the high pressure pump (P1). The high pressure pump increases pressure of the F1 stream to

the required feed pressure. The flow rate of the F1 stream is approximately equal to the flow rate

of the permeate stream from the RO membrane unit.

The second stream (F2) is directed to the isobaric ERD. Flow rate of stream F2 is approximately

equal to flow rate of the concentrate stream from the RO membrane unit. In the isobaric ERD,

the F2 stream exchanges energy with the concentrate stream. At the exit from the isobaric ERD

the pressure of the F2 stream is somewhat lower than the required feed pressure. The additional

pressure boost is provided by the circulation pump P2. Prior to the entrance to the membrane

unit, both streams are combined together.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



221

F1 – 50 m3/hr,

220 gpm

P - 50 m3/hr,

220 gpm

P1

P2

Power

Recovery

deviceF1 = Permeate

F2 = Concentrate

F2- 50 m3/hr,

220 gpm

High

pressure

pump

Circulation

pump

2bar, 30 psi

100 m3/hr,

440 gpm

60 bar, 870 psi 100

m3/hr, 440 gpm

58 bar, 840 psi, 50

m3/hr, 220 gpm1 bar,

15 psi

55 bar,

797 psi

F2 – 60 bar,

870 psi

Energy consumption of RO

process: 2.14 kWhr/m3 (8.01

kWhr/kgallon) 46% reduction

F1

F2

Figure 6.12. Schematic configuration of RO membrane unit with isobaric energy recovery

device.

Two commercial types of isobaric ERD dominate at present the seawater RO applications

market. They are DWEER and PX. Both are positive displacement devices but of different

configuration.

The configuration of DWEER is shown on Figure 6.13. It consists of two parallel cylinders with

floating pistons. The cylinders are connected together at one end through the “link valve”

(concentrate end) and on the other end through manifold with check valves (seawater feed end).

The two cylinders operate in opposite cycles of filling in by RO concentrate and seawater. In the

energy exchange cycle, high pressure concentrate enters on cylinder through the link valve and

replaces under high pressure seawater that filled the cylinder before at low pressure. When the

floating piton reaches the end of the cylinder, the link valve, positioned at the other end, changes

its position and allows the concentrate to be replaced by low pressure seawater. The second

parallel cylinder operates in the same way but its cycle sequence timing is shifted to perform

opposite filling and discharge to the operation of the first cylinder.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



222

Figure 6.13. Configuration of DWEER energy recovery device

The isobaric devices make by DWEER could have flow capacity of up to 300 m3/hr. Number of

units can be connected together of large flow capacity is require. An assembly of large size

DWEER units at the 300,000 m3/day SWRO Ashkelon, Israel, desalination plant is shown on

Figure 6.14.

The configuration of PX isobaric ERD is shown on Figure 6.15. The PX configuration consists

of ceramic rotor in a vessel made of composite material. The ceramic rotor has number of radial

spaced parallel channels passing through the rotor. The PX vessel has four connections, two on

each end of the vessel.

In a similar way as in operation of DWEER, high pressure concentrate fills one of the rotor

channels, pressurized seawater that filled the channel before at low pressure. The rotor is in

continuous rotation and the channel filled with concentrate moves to position that connects

(through the rotor channel) the inlet of low pressure seawater and discharge of concentrate. At

this position low pressure seawater fills the channel and concentrate volume that has filled the

channel previously is discharged from the device. In a continuous process rotor channels are

filled with low pressure seawater, pressurized with high pressure concentrate and as the rotor

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



223

rotates, the concentrate is replaced with low pressure seawater again.

Figure 6.14. DWEER isobaric EDR assembly operating in 330,000 m3/day SWRO plant,

Ashkelon, Israel.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



224

Figure 6.15. Configuration of PX energy recovery device (ERI)

The flow capacity of PX is lower than DWEER, in the range of 50 m3/hr. In a similar way as

DWEER units, the PX units can be arranged in parallel assemblies as shown on Figure 6.16.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



225

Figure 6.16. Large assembly of PX ERD’s.

In the DWEER ERD the concentrate and seawater feed are separated by floating piston. In the

PX ERD there is direct contact between concentrate and seawater. For this reason mixing

between concentrate and seawater is somewhat higher in PX device. On the average in DWEER

the mixing of two streams is usually less than 3%. In the PX device the mixing can reach up to

6% of the seawater feed flow rate. Because, in each case only about 50% of the feed flow is

treated with ERD, the effect of salinity increase of the feed is only half on the volumetric mixing

in each case.

Introduction of isobaric ERD’s brought a significant decrease of energy requirement in SWRO

applications.

Current Isobaric ERD’s are not very effective in brackish application due to significantly

different ratios of concentrate to feed flows rates in BWRO.

Incentives for application of ERD in brackish RO units are much lower than in SWRO (see Table

6.1). In brackish applications Hydraulic Turbocharger is at present a more cost effective solution

than isobaric ERD’s.

Among new ERD devices one more interesting is ISave introduce recently by Danfoss. ISave is

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



226

an integrated unit consisting of pump, energy recovery unit and a motor, shown on Figure 6.17.

Figure 6.17. ISave ERD introduced by Danfoss.

The flow capacity of ISave units is limited to about 40 m3/hr. However, the advantage of ISave

is elimination of the circulation pump, required in system utilizing DWEER or PX.

6.3.1.5.Cost and economic benefits of ERD

Configuration of pumping unit depends on application, method of delivery of raw water to the

plant, selection of pumping equipment and plant size.

The membrane trains pumping units usually consist of transfer pumps, cartridge filters

(depending on filtration pretreatment type) high pressure pumps and energy recovery devices.

Listing of components of pumping system for a 100,000 m3/day SWRO unit, treating high

salinity gulf seawater at recovery rate of 45%, is provided in Table 6.2.

Table 6.2. Comparison of operating parameters of pumping unit in SWRO 100,000 m3/day plant.

Feed flow m3/day 231,200

Concentrate flow m3/day 129,200

Permeate flow M3/day 102,000

Energy recovery device type Isobaric Pelton Turbocharger

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



227

Cartridge filters pump (feed pump) # units 2+1 4 +1 4 +1

Cartridge filters pump flow/unit m3/hr 2,125 2,410 2,410

Total head bar 10 10 10

Name plate power kw 900 1,000 1,000

Cartridge filters (feed pump) # units 4+1 8+1 8+1

Cartridge filters flow/unit m3/hr 1,065 1,200 1,200

Energy recovery device feed flow rate m3/d 129,200 231,200

Cartridge filters pump (energy recovery

pump) #

units 3+1

Cartridge filters pump flow/unit m3/hr 1,800

Total head bar 3.5

Name plate power kw 280

Cartridge filters (energy recovery pump)

#

units 5 + 1

Cartridge filters flow/unit m3/hr 1,080

SWRO high pressure feed pump # units 4 + 1 8+1 8+1

High pressure feed pump flow/unit m3/hr 1,065 1,200 1,200

Pump differential head bar 65.0 65 35

Name plate power kw 2,800 1,900 1,750

Energy device flow rate m3/d 129,200 129,200 129,200

High pressure brine flow rate m3/d 132,200 129,200 129,200

High pressure brine pressure bar 70.0 70.0 70.0

Low pressure brine flow rate m3/d 132,200 129,200 129,200

Low pressure brine pressure bar 2.0 2.0 2.0

Circulation pump high pressure feed flow

rate

m3/d 129,200

Circulation pump high pressure feed

pressure

bar 65.0

Energy recovery units units 4+1 8 8

Feed flow per energy recovery unit m3/hr 1,350 1,200

Concentrate flow per energy recovery

unit

m3/hr 1,380 670 670

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



228

Number of devices per unit units 28

Circulation pump flow rate m3/hr 5,380

Total differential head bar 2.9

Number of units units 4 +1

Individual pump flow rate m3/hr 1,345

Name plate power Kw 170

Table 6.3. Summarized estimation of cost of pumping and energy recovery equipment in three

basic configuration alternatives: pumps + isobaric ERD’s, pumps + Pelton Wheel and pump +

Hydraulic Turbocharger.

As expected, the premium efficiency isobaric device equipment is the most expensive option.

However, it provides the lowest energy requirement and at locations of moderate to high

electricity rates this alternative is the most cost effective (based on a plant life cycle cost).

Table 6.3. Comparison of cost of pumping – energy recovery equipment alternatives for SWRO

100,000 m3/day plant.

Equipment Capacity,

m3/hr

Number of

units

$/unit Subtotal $

Pumps with isobaric devices

LP pumps (for HP pumps) 2,125 3 238,000 714,000

LP pumps for isobaric devices 1800 4 215,000 860,000

Cartridge filters housings 1,065 11 70,000 770,000

High pressure pumps 1,065 5 615,000 3,075,000

Circulation pumps 1,350 5 155,000 775,000

Isobaric devices 50 140 29,500 4,130,000

Total isobaric equipment 10,324,000

Pumps with Pelton ERD devices

LP pumps (for HP pumps) 2,410 5 275,000 1,375,000

Cartridge filters 1,200 10 75,000 750,000

High pressure pumps 1,200 9 Included

with PW

Pelton Wheel ERD units 670 9 630,000 5,670,000

Total Pelton Wheel equipment 7,795,000

Pumps with Hydraulic

Turbocharger devices

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



229

LP pumps (for HP pumps) 2,410 5 275,000 1,375,000

Cartridge filters 1,200 10 75,000 750,000

High pressure pumps 1,200 9 230,000 2,070,000

Hydraulic Turbocharger units 1,360/750 9 230,000 2,070,000

Total Pelton Wheel equipment 6,265,000

7.0.Chemical dosing equipment design

The role of chemicals used in RO applications is to improve effectiveness of pretreatment

process in reducing scaling tendency of the feed water and improve effectiveness of the filtration

step. In the post treatment process, chemicals are added to reduce corrosivity of the product

water and to create disinfection residual, according to local regulations.

Another group of chemicals are used during the membrane cleaning step. Objective of the

cleaning chemicals is to dissolve and disperse foulant deposits from the membrane surfaces.

Water produced by RO and NF systems should be of potable quality. Therefore, toxic chemicals

cannot be used during regular operation of the unit.

The composite membrane material is susceptible to oxidation damage. Therefore, strong oxidant

cannot be used at condition that would result in exposure of membrane to this type of chemicals.

7.1.Selection criteria for chemicals used in the RO process

Type of chemicals used in RO process depends on application and type and configuration of

pretreatment and post treatment process applied. Listing of chemicals potentially used according

to application is included in Table 7.1.

Table 7.1. Listing of chemicals used in RO and NF applications.

Application Pretreatment Post treatment

Brackish RO, NF, well water

feed

Acid (H2SO4 or HCl), scale

inhibitor

NaOH, hypochlorite, ammonia

Brackish RO, NF, well water

feed containing iron or

manganese

Hypochlorite, sodium

bisulfite, acid (H2SO4 or

HCl), scale inhibitor


Brackish RO, NF, surface

water feed

Ferric coagulant, filtration

polymer, hypochlorite, sodium




Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



230

Seawater RO, beach well feed Acid (H2SO4 or HCl), scale

inhibitor

CO2, CaCO3, Ca(OH)2,


Seawater RO, surface water

feed

Ferric coagulant, filtration

polymer, hypochlorite, sodium



CO2, CaCO3, Ca(OH)2,


Membrane cleaning Mineral acids, citric acid, NaOH, sodium trpoliphosphate,

sodium bisulfate, EDTA, ammonium bifloride, detergent

(SDBS), specialty cleaning chemicals

7.2.Procedures for determination of chemicals dosing rate

Procedure for determination of dosing rate of chemicals could be based on the following

considerations as listed in table 7.2.

Table 7.2. Procedures for determination of chemicals dosing rates

Chemical type Point of use Method for estimation of dosing rate

Mineral acid,

NaOH

Pretreatment, post treatment,

adjustment of water pH

According to carbonate system relations.

Target pH according to target LSI (section

5.5.2.5).

Scale inhibitor Pretreatment According to recommendations of supplier,

usually to target concentration of 100 ppm in

the concentrate stream

Ferric

coagulant

Pretreatment Based on results of pilot operation. Usually 1

– 20 ppm

Filtration

polymer

Pretreatment Based on results of pilot operation. Usually

0.1 – 0.5 ppm

Hypochlorite Pretreatment According to chlorine demand by the

oxidation process of metals and organics.

Intermittent or continuous dosage of 0.5 – 2.0

ppm. Shock chlorination 5 – 20 ppm.

Sodium

bisulfite

Pretreatment According to chlorine residual. Bisulfite to

chlorine ratio 3:1 ppm/ppm

Hypochlorite Post treatment According to demand and required chlorine

contact time and disinfection residual

Ammonia Post treatment According to required level of chloramines in

the product water. Usual ratio of ammonia to

free chlorine ~ 1:3

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



231

NaOH Post treatment According to carbonate system relations and

required pH of product water

CO2, CaCO3,

Ca(OH)2

Post treatment According to carbonate system relations.

Target pH according to target LSI (section

5.5.2.5).

Cleaning

chemicals

Cleaning process According to recommendation of membrane

manufacturer and operational experience.

Calculation of usage of chemicals is based on calculation of dosing rate, concentration of active

ingredient and specific density chemical, flow rate of the stream and plant on line factor. The

starting point is flow and mass balance included in process flow diagram (PDF) and chemical

composition of stream being treated.

7.3.Criteria for sizing of chemicals storage equipment

The determination of sizing of chemical storage equipment is driven by the dosing rate,

concentration of chemical, logistic of operation and expected supply lead time. The other

consideration is the size of storage for a given chemical and its shelf life. Marinating on the plant

site quantity of high rate usage chemical for a long period of operation could be prohibitively

expensive. Usually, a quantity of stable chemicals (for example scale inhibitor) used at low usage

rate is stored for a longer period. For chemicals with short shelf life (for example solution of

sodium bisulfate) a storage volume for a shorter operating period is being used.

Frequently the size of storage of treatment chemicals is defined in specification as number of

days of desalination plant operation between deliveries. The usual period of operation between

deliveries ranges from 7 – 30 days.

7.4.Selection of chemical dosing pumps capacity and materials of construction

Dosing units represent relatively small fraction of the overall system cost. However, their proper

operation is critical to operation of the desalination system. For this reason is recommended to

build in a sufficient spare capacity into the configuration of the dosing unit.

The required spare dosing capacity is addressed by the number and capacity of the dosing

pumps. The prudent approach is to have each dosing skid equipped with three pumps, each of

dosing capacity sufficient for delivery of full required quantity of chemical at designed operation

condition of the desalination plant. At the regular operating conditions two pumps are in

operation at 50% capacity and one pump is idle. This configuration provides a two level

protection against malfunction of the dosing equipment.

Dosing systems usually handle aggressive solutions therefore require suitable material of

construction of wetted parts.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



232

Representative selection of materials of construction of chemical dosing systems components are

listed in Table 7.3

Table 7.3. Representative materials of construction for chemical dosing systems

Metallic components 316 stainless steel (alloy 20 for sulfuric acid)

Thermoplastic components Polypropylene

Elastomers PTFE/Viton

Dosing pump head, valves and seats PVC, ceramic, EPDM

Dosing pump diaphragm PTFE

Storage tank FRP (steel for sulfuric acid)

Piping, tank fill Sch 80 PVC

Piping, vent Sch 80 PVC

Piping, pump suction Sch 80 PVC

Piping, pump discharge Sch 80 PVC

Injection quill Hastelloy C276

Valves PVC

7.5.Example of sizing of chemical dosing system for SWRO plant of permeate

capacity 100,000 m3/day operating at recovery rate of 45%.

Table 7.4. System process information for 100,000 m3/day SWRO system

Permeate capacity, m3/day 100,000

Raw water flow, m3/hr 10,466

Feed water flow, pass1, m3/hr 9,630

Feed water flow, pass 2, m3/hr 980

Permeate water flow, pass 1, m3/hr 4,333

Permeate water flow, pass 2, m3/hr 833

Permeate water flow, blended, m3/hr 4,186

Permeate flow through post treatment, m3/hr 628

Table 7.5. Usage of treatment chemicals in 100,000 m3/day SWRO system

Chemical Stream Flow rate Dosing

rate

Concent

ration

Quantity

per day

Quantity

per year

(**)

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



233

Units m3/hr ppm % ton/d ton/y

Sulfuric acid Raw water 10,466 32.0 96 8.3 3,166

Ferric sulfate Raw water 10,466 10.0 40 6.3 2,402

Filtration

polymer

Raw water 10,466 0.5 100 0.11 42

Sodium

hypochlorite

Raw water 10,466 5 (*) 12.5 0.42 160

Sodium bisulfite Feed water 9,630 3 (*) 40 0.07 27

Sodium

hydroxide

Second pass

feed

980 30 48 1.5 573

Scale inhibitor Second pass

feed

980 1.5 100 0.04 14

Carbon dioxide Permeate post

treatment

628 293 100 4.4 1,677

CaCO3 Permeate post

treatment

628 667 96 10.5 4,005

Sodium

Hydroxide

Permeate post

treatment

628 2 40 0.07 25

Sodium

Hypochlorite

Final product 4,186 1.5 12.5 1.2 416

(*) Intermittent chlorination – 1hr/day

(**) Quantity increased by 10% of safety margin

Size of the storage tanks should be based on daily usage multiplied by required length of

operation between deliveries, specified in plant scope book.

Sizing of dosing pump is based on daily usage and reagent density. Example of sixing of dosing

pumps is provided in Table 7.6. The sizing is for a single pump delivering 100% of chemical

quantity. The number of pumps for a dosing skid is usually 3. Two pumps in operation, each

providing 50% of the dosing flow, and one spare. The pumps are sized that each pump is capable

to deliver 100% of the required flow.

Table 7.6. Sizing of chemical dosing pumps

Chemical Stream Flow rate Dosing

rate

Concentr

ation

Reagent

density,

kg/l

Dosing pump

capacity

(100% flow)

Units m3/hr ppm % ton/d l/hr

Sulfuric acid Raw water 10,466 32.0 96 1.84 190

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



234

Ferric sulfate Raw water 10,466 10.0 40 1.40 190

Filtration

polymer

Raw water 10,466 0.5 100 1.00 5

Sodium

hypochlorite

Raw water 10,466 5 (*) 12.5 1.20 350

Sodium bisulfite Feed water 9,630 3 (*) 40 1.30 54

Sodium

hydroxide

Second pass

feed

980 30 48 1.52 42

Scale inhibitor Second pass

feed

980 1.5 100 1.00 2

Carbon dioxide Permeate

post

treatment

628 293 100 4.4 185

CaCO3 Permeate

post

treatment

628 667 96 1.6 275

Sodium

Hydroxide

Permeate

post

treatment

628 2 48 1.52 2

Sodium

Hypochlorite

Final

product

4,186 1.5 12.5 1.2 42

(*) Intermittent chlorination – 1hr/day

8.0. Instrumentation and control system

The RO desalination system consists of large number of process equipment and subunits that

operate in parallel, interacting with each other in a specific sequence of operation and within the

values of designed operating parameters. Operation of these complex units requires real time

operation control, only possible through the electronic instrumentation and control system.

The instrumentation and control system is designed to achieve the following objectives:

• Protect system from operation at conditions that may result in personnel injury or

equipment damage

• Maintain required sequence and timing of equipment operation

• Maintain operation of equipment within the design process limits

• Maintain production of the design quantity and quality of water

• Store and processes operational data, generates reports, displays information

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



235

• Enable controlled intervention in system operation

In addition to protect system operators, equipment and controlling operation to produce required

quantity and quality of product water, the control system collects and manages the operating

data.

Operational data storage and management is required to satisfy regulatory agencies, the end user

and to satisfy terms of equipment warranty. Properly managed operational data are also required

for adequate scheduling of maintenance operation, maintaining adequate quantity on hand of

process materials and parts, follow up on process economics, etc..

8.1.Process control strategy

The basic operative objective of the control system is to control operation of the membrane unit

to produce required quantity of permeate at the design recovery rate.

The control loop to achieve these two objectives is depicted in Figure 8.1.

RO process control

FIT

FIT

Electric

motor

Motor

controller

Valve

controller

Product line

Concentrate lineHigh pressure

feed pump

Figure 8.1. Basic process control of RO membrane unit

As shown in Figure 8.1, two loops control operation of the RO membrane unit:

- Permeate flow control loop

- Measurement of permeate flow and control of motor speed or position of the feed

valve

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



236

- Recovery rate control loop

- Measurement of permeate and concentrate flows. Control of position of

concentrate valve according to ratio of permeate-concentrate flows

The above two loops control production rate of permeate while maintaining membrane unit

operation at the design recovery rate.

8.2.Control loops in RO system

In large RO desalination plants the control system is divided into loops that control operation of

process segments.

Instrumentation and operation control loops could include:

• Intake control loop

• Pretreatment control loop

• Main RO control loop

• RO trains control loop

• Motor control center loop

• Electrical circuits and VFD loop

• Permeate post treatment and storage control loop

• Wastewater neutralization and discharge control loop

The configuration of control system is shown schematically in Figure 8.2.

The overall process is controlled by a master programmable logic controller (PLC). Usually the

process control is distributed to local PLC’s controlling logic segment of the process, as shown

on Figure 8.2. The local PLC’s are connected to and controlled by the master PLC.

The operators control the process through the human – machine interface (HMI) display stations.

The control system also includes data server for storage of reference process information and

operating data collected by sensors connected to the PLC’s.

In addition to maintain operational availability of the control system, also during the power

outage, supply of emergency power to the PLC is provided through the uninterrupted power

supply (UPS) units.

Additional level of availability is provided through spare control equipment that is on line in

addition to the main system.

Figure 8.3. shows schematic configurations of a control system without redundancy. In this

configuration malfunction of PLC or critical input/output (IO) units will result in shut down of

the desalination plant.

Figure 8.4. shows schematic configuration of control system with “hot” backup of second PLC.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



237

Human

Machine

Interface (HMI)

Human

Machine

Interface (HMI)

Human

Machine

Interface (HMI)

Master

Programmable

Logic Controller

(PLC)

Raw water supplyPretreatment

system

Filtrate transfer

pumps

HP pumps and RO

trains

Product water

treatment

Product water

storage and

pumping

Auxiliary systems

Control room

Field located control loops

Database

server

Figure 8.2. Schematic configuration of control system in RO desalination plant

I/O I/O I/O I/O I/O I/O

PLC Processor

Figure 8.3. Control system configuration – no backup control equipment

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



238

The backup PLC is continuously of line with the base PLC. In malfunction of one PLC the

control is transferred to the second PLC.

PLC Processor PLC Processor


Hot Backup

Control

Figure 8.4. Control system configuration – ‘hot” backup (PLC only)

PLC Processor PLC Processor


Synchronized link

Figure 8.5. Control system configuration – complete backup (PLC and IO’s)

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



239

Figure 8.5 shows configuration of a control system with full backup. The two PLC operate in

parallel, being synchronized in real time. In addition the more critical IO’s also have on line

backup.

The increased redundancy of control system comes at increasing cost, which is measured against

potential loss of water production from the desalination plant.

8.3.Process control and performance normalization software

The specialized process control software for small RO units is offered by number of vendors. For

larger desalination plants the software control software package is developed as a part of design

process of the project.

The software is developed based on system process and instrumentation diagram (P&ID),

instrumentation and equipment specification, desalination process description and control

process narrative, including decision tree (decision diagram), developed by process engineer(s).

Based on the above information software engineer develops software package that will be

utilized by PLC to control the desalination plant.

Adequacy of the software control package to control operation of the desalination plant is tested

during the plant commissioning step.

Another important software package, usually operated through PLC, is the membrane

performance normalization software. The normalization software is a computer program that is

used to compare performance of the membrane unit with set of reference operational data. Such a

comparison enables operator to evaluate stability of membrane performance, determine if

performance trend indicates that the membrane fouling process is occurring in the membrane

elements and decide if membrane elements maintenance procedure (membrane elements

cleaning, membrane elements replacement, etc..) is required.

The normalization programs in a spreadsheet format are available from all major membrane

manufacturers. These programs correspond to generic configuration of RO membrane unit.

Normalization programs in different formats and reflecting configuration of membrane unit in a

given desalination plant can be developed based on the ASTM membrane normalization

procedure: D4516-00 Standard practice for standardizing reverse osmosis performance data.

8.4.Instrument selection criteria and their location in the RO system

The instruments for RO membrane desalination units are selected according to specifications

included in desalination plant scope book. In lack of specifications provided by project engineer

the EPC usually follows the requirements of the client or EPC’s past experience with the

instrumentation equipment.

In addition to the instrumentation equipment required for the monitoring of the RO process, there

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



240

is additional instrumentation equipment required to satisfy warranty terms of equipment

suppliers and equipment necessary to enable proper operation of the auxiliary equipment.

The desalination plant environment includes exposure to high salinity water and exposure to

stream at extreme pH range. These environmental conditions should be also considered during

selection of the instrumentation equipment.

For the control and proper monitoring of the operation of RO membrane trains the following

minimum set of instruments is required:

- Measurement of feed water temperature

- Measurement of feed water pH

- Measurement of permeate pH

- Measurement of concentrate pH

- Measurement of feed water redox potential (if chlorination – dechlorination is being

applied in the pretreatment)

- Measurement of feed water turbidity

- Measurement of feed water SDI (if automatic SDI instrument is available)

- Measurement of feed water conductivity

- Measurement of permeate water conductivity

- Measurement of concentrate conductivity

- Measurement of feed pressure

- Measurement of concentrate pressure

- Measurement of permeate pressure

- Measurement of permeate flow

- Measurement of concentrate flow

In a multistage RO membrane unit it is a good practice to provide also pressure

indicators/transmitters to monitor the interstage pressures and flow meters to monitor permeate

flow rates produced by the individual stages.

8.5.Frequency of data collection and representative range of operating

parameters

In PLC controlled system vales of operating parameters, translated to 4 – 20 mv output are

continuously transmitted to PLC. The programmer of the control system should configure the

control system to collect and store operational data in a sensible frequency. Additional

requirement is to include in data collection software filters that would ignore erroneous values

collected in transition periods of system shut down, startup or other periods when unit is not in

normal operation.

Process engineer that has a task of writing narrative of the computer software operation for the

programmer has to specify range of acceptable values of operating parameters at various

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



241

operating conditions of the membrane unit. Eventually the program will have to differentiate

properly between erroneous data that should be rejected and values of operating parameters that

are outside the designed operational limits that should trigger an alarm and require intervention

of the operator.

The usual operating period of set of membrane elements is 5 – 7 years. The designer of the

control and data collection system has to make decision what volume of system operational data

will be available on line and how the data will be presented for viewing and for calculation of the

normalized performance of the membrane system. Regardless of frequency of collection of the

operational parameters the data should be presented for evaluation that represents data collection

frequency of one point per shift or day.

During the plant commissioning and acceptance test period this frequency could be higher but

eventually should be reduced.

However, plant operator should have the option to see operational data at higher frequency for

selected operational interval, as required to evaluate some operational events of the system.

8.6.Methods of control of operation of chemicals dosing systems

Chemical dosing system serve critical role in the operation of RO desalination plant. Operation

of the chemical dosing system affects quality of the feed water, scaling tendency of the

concentrate and compliance of permeate water quality with requirements of potable water

supply.

For some of the chemicals their affect of the water stream can be detected by measurements of

water quality parameters. This is the case of addition of acid, base or reducing agents (sodium

bisulfate) by measuring pH and REDOX potential. This is not the case when dosing scale

inhibitor, which is a neutral organic compound, being added at a very low dosing rate, usually in

the range of 1 – 3 ppm .

Even in case of acid or caustic addition the operation of dosing system cannot be driven only by

measurement of pH of the stream, as the inertia of the pH adjustment and the distance between

dosing and sampling points, would make the accurate and consistent pH adjustment difficult.

The dosing system should be driven by target concentration of dosed chemical in the stream and

flow rate of the stream, i.e. volumetric quantity of chemical transfer to the stream by the dosing

pump.

Majority of dosing system utilize positive displacement, diaphragm pumps. These pumps create

pulsation flow of the chemical solution and therefore the regular flow meters would not be

effective in accurate measurement of the flow rate.

One of the common methods of monitoring flow rate of chemical solution is to measure level

change of chemical solution in a calibrated tank. The calibrated tank is connected to the chemical

storage tank and to the dosing pump. The tank is filled from the storage tank at the defined low

level in the calibrated tank. The dosing pump suction is connected to the calibrated tank.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



242

The control is based on volumetric measurements of the decreased volume (and consequently the

fluid level) of a calibrated tank. This tank is sized for 1 hour of operation between fillings with

such diameter that at nominal production and chemical dosing, level will drop at a ratio of 1 cm

per minute. A level transmitter on each tank continuously reads the level. This data is analyzed

and compared to the required dosage according to the water flows measured by flow meters in

the plant.

Dosing adjustment is achieved by varying the pumps motor speed utilizing the variable

frequency driver on the pump motor.

If a feed back signal is available (pH, ORP, or Cl2), it is overriding the control if the value of the

signal reached or exceeded the designed limits.

Other method of monitoring flow of dosing chemicals is utilization of heat sensor and optical

tracers.

The heart sensor method utilizes a flow through cell equipped with heating element and

temperature sensor. The temperature inside the flow through cell, measured by temperature

sensor, is proportional to the flow rate (temperature will increase with decreased flow rate). The

temperature readings are compared with the calibration curve of flow vs. temperature and the

motor speed of the dosing pump is adjusted accordingly.

The optical tracer method is based on monitoring with optical sensor the presence of fluorescent

material added in small quantity to dosed chemical (usually the scale inhibitor). Readings of the

optical sensor can be related to quantity of chemical dosed and utilized to control the dosing

pump and trigger alarm if necessary.

8.7.Pumps process control in brackish and nanofiltration applications

In low pressure applications; brackish and nanofiiltration systems, water pressure required for

passage through the cartridge filters and to develop sufficient suction pressure for the high

pressure pumps is usually provided by the well pumps.

On the startup, these pumps do not pump water to the desalination system. The initial flow is

diverted to drain for number of minutes. This arrangement is to assure that the initial flow, that

could contain high concentration of particulate matter, will not reach the membrane elements.

After the predetermined time period the drain valve is being closed and in parallel the valve,

connecting the raw water delivery pump to the RO system, opens gradually. The rate of valve

opening should provide feed pressure increase that will not exceed 0.7 bar/sec.

During this step the high pressure pump discharge valve and the concentrate valves are

completely or partially open. The RO membrane unit (RO train) is being flushed with the feed

water with the objectives to remove air that could be trapped in piping and pressure vessels.

After period of about 3 – 10 min, the concentrate valve is closing to the position corresponding

to the normal operating conditions of the membrane unit. The motor of high pressure pump starts

and the pressure in RO membrane unit gradually rises. The discharge valve of the high pressure

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



243

pump opens gradually and the pressure in the membrane unit is allowed to increase at the rate

not exceeding 0.7 bar/sec till the permeate flow reaches the design flow rate and/or the feed

pressure reaches the design value.

8.8.Pumps process control in seawater RO applications

Control of operation of high pressure pumps and energy recovery devices is depended on

configuration of the pumping system in relation to membrane train configuration.

In train dedicated configuration high pressure pump and energy recovery device form a pumping

unit. Each unit is dedicated to operation of a specific membrane train (Figure 8.6).

In a pressure centers configuration the high pressure pumps and ERD’s are group to operate in

parallel as two pressure centers, each group operating with all trains.(Figure 8.7).

8.8.1. “Train dedicated” configuration

The pumping units utilizing the Pelton Wheel and Hydraulic Turbocharger ERD’s are more

suitable for operation as train dedicated units. While pumping unit utilizing pressure exchangers

can operate as a membrane train dedicated unit and also in systems configures as pressure

centers.

The instrumentation and location of flow control valves for pumping unit with Pelton Wheel is

shown on Figure 8.8

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



244

ERD

ERD

ERD

RO train

RO train

RO train

Figure 8.6. Pumping units and RO membrane trains in “train dedicated” configuration

ERD

RO train

RO train

RO train

ERD

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



245

Figure 8.7. Pumping units and RO membrane trains in “pressure centers” configuration

V - 2V-3

V - 1

FIT

V - 4

M

V - 5

PITCIT PIT PIT

FIT PIT CIT TIT

PIT CIT

Figure 8.8. Pumping unit with Pelton Wheel EFD

The sequence of startup of system with Pelton Wheel is listed below.

1. Set up regulating valves of the Pelton Wheel nozzles to the design position

2. Open completely discharge valves of high pressure pump

3. Start booster pump, increase pressure to 6 bar

4. Operate booster for 15 min till all the air dissolves

5. Pelton Wheel turns and rotates high pressure pump at ~600 RPM

6. Partially close discharge valves of high pressure feed pump and start the motor

7. Slowly opens discharge valves of high pressure feed pump for pressure increase of ~ 0.7

bar/sec

8. Valves on the PV set to position for designed concentrate flow

9. Adjust valves again after 1 – 2 hr

The instrumentation and location of flow control valves for pumping unit with Hydraulic

turbocharger is shown on Figure 8.9 and 8.10.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



246

1

2

3

4

PIT

PIT

PIT

PIT

PIT

VF

D

PIT

PIT

Figure 8.9. High pressure pump and hydraulic turbocharger in feed entry position

1

2

3

4

Stage 1Stage 2

PIT

PIT

PIT

PIT

PIT

PIT

VF

D

PIT

PIT

Figure 8.10 Hydraulic turbocharger in the interstage position

The startup of the RO trains equipped with hydraulic turbocharger is similar to startup sequence s

described above for the unit with Pelton Wheel ERD. The feed water flow is initiated by the

booster pump(s). Operation of the booster pump(s) creates flow of the concentrate that flows

through the ERD’s providing initial pressure boost. High pressure pumps should start operation

with discharge valve partially closed to maintain slow rise of the feed pressure (< 0.7 bar/sec).

Gradually opening the discharge valve and increasing speed of the motor by regulating VFD,

brings the feed pressure to the design value.

Operational sequence of pumping system including isobaric ERD

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



247

The schematics of pumping unit that includes isobaric ERD is shown on Figure 8.11.

Starting pumping system with isobaric ERD (PX) – dedicated pumping unit

1. Start booster that delivers water to HPP and to ERD

2. Operate at pressure 2.5 – 6 bar till all the air dissolves

3. Maintain flow within the flow limit of ERD

4. Start the circulation pump

5. Close discharge valve on HPP and start the pump motor

6. Slowly open the HPP pump valves for pressure increase ~0.7 bar/sec

V-3

FIT

VFD

V - 5

PITCIT

PIT

PIT

FIT PIT CIT TIT

V-8

V-9

V-10

V-14

V-15

V-16

FIT

V-18

V-20

V-21

V-22 V-23V-25

V-26

FIT

PIT

CIT

V-11

V-19CIT

ACF

DG

BH

Figure 8.11. High pressure pumping unit utilizing isobaric ERD.

Adding of ERD to the operation

1. Circulate low pressure feed water to ERD to remove air

2. Slowly open valve of high pressure brine to ERD

3. Start the circulation pump

Stop ERD unit

1. Stop the circulation pump

2. Close the high pressure brine valve

3. Flush the unit from brine

4. Close high pressure feed valve

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



248

8.8.2. “Pressure centers” configuration

In the pressure center configuration the high pressure pumps and ERD’s operate in

the same way as in dedicated pumping unit – membrane train configuration. The

important considerations are during operations when membrane trains are added or

shut down, while rest of the system is in operation.

Adding membrane train to the operating system

1. Once feed water pH and turbidity is within range, Modulate all valves to pre-flush

positions (concentrate and ERD bypass valves full open, however feed valve could either

be open or at a preset position to limit permeate production)

2. Start the RO feed pump at minimum speed and begin pre-flush timer (pre-flush timer is

operator adjustable within a specified range)

3. Following completion of pre-flush, modulate the concentrate control valve and ERD

bypass valve to preset positions (also open feed valve to full open if preset position is

used)

4. Place RO feed pump into automatic control and allow adjustment of the pump speed

(system feed pressure) to maintain the design permeate capacity

5. Place concentrate control valve into automatic control and allow valve to modulate to

maintain the design recovery rate

6. Place ERD bypass valve into automatic control and allow valve to modulate to maintain

the design 2nd stage permeate capacity

7. All control loops are now in automatic and RO unit is operating at nominal conditions

Membrane train shutdown sequence

1. Upon initiation of normal skid shutdown by operator, control loops are removed from

automatic control mode and post flush sequence is initiated

2. Reduce the RO feed pump to minimum speed and modulate all valves to post-flush

positions (typically the same as pre-flush positions)

3. Begin post-flush timer (post-flush timer is operator adjustable within a specified range) 4. Following completion of post-flush, stop the RO feed pump and associated raw water

well (supply) pumps and close the feed valve

5. As an option, the concentrate control valve and ERD bypass valves can be closed once

feed valve is confirmed closed

6. RO unit is now in standby mode

Emergency shutdown sequence of membrane train

1. Emergency stop RO Feed Pump

2. Stop associated raw water well (supply) pumps and close the feed valve

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



249

3. Upon confirmation of closed feed valve, close all other control valves

9.0.Selection of materials of construction of equipment components

Desalination environment is highly corrosive due to presence of saline water with high

concentration of chloride ions. High chloride ion concentration breaks the protective oxide layer

on the metal surface and enables corrosion process to proceed.

The nature of corrosion process depends on environmental conditions. The most common in

saline environment are: galvanic corrosion, pitting corrosion and crevice corrosion.

The galvanic corrosion process requires presence of two metals of dissimilar activity and water

containing dissolved oxygen according to reaction:

2Fe + O2 + 2H2O = 2Fe(OH)2 (9.1)

The dissimilar metals could be two sections of the metal component of different activity. The

electrons travel from the more active metal site (anode) to the less active site (cathode), where

they react with water forming hydroxyl ions. The hydroxyl ions react with metal on the anodic

site, forming corrosion products.

The pitting corrosion occurs under condition of very low dissolved oxygen concentration and

high concentration of chloride ions. This is localized corrosion due to breaks in the protective

oxide layer that cannot be recreated due to deficiency of oxygen.

The crevice corrosion takes place at small areas of stagnant water with variable concentration of

dissolved oxygen. This type of corrosion frequently occurs under gaskets connecting metal parts.

Corrosion is mitigated by proper selection of materials of construction, maintaining proper flow

velocity, passivation of metal surfaces and in some cases by applying cathodic protection.

The cathodic protection addresses galvanic corrosion by use of sacrificial anode(s), which supply

electrons to the metal part being protected. The metal part becomes cathode and has negative

potential, usually in the rage of -1.0 V for sufficient corrosion protection.

The anode, made usually of aluminum, is slowly dissolving in the process in place of metal part

that otherwise would corrode.

The passivation process of metal parts (usually piping parts), is applied as a last fabrication step.

It consist of immersing of metal parts in a mixture of hydrochloric and nitric acid. In this step all

the impurities are dissolved and flushed out from the metal surfaces, which allows proper

formation of corrosion protective layer of metal oxides.

The construction materials are selected according to operating pressure, process segment and

feed water salinity.

For operation involving low pressure conditions: raw water supply, pretreatment and post

treatment, the tendency is to utilize components made of polymeric materials. For the high

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



250

pressure operation, low salinity water systems utilize type 316 steel and SMO and Duplex alloys

are used for seawater applications.

Example of materials selection is provided in Table 9.1.

Table 9.1. Selection of piping material according to application

Application Piping material

Low pressure, up to 500 mm (

<=20”)

Polypropylene

Low pressure, over 500 mm (>20”) FRP (GRP), HDPE

RO permeate 316L SS, PVC (< 12”), PP

High pressure brackish 316L SS (PREN = 24 – 31)

High pressure seawater, up to 350 mm

( <=14”)

SMO (PREN = 44 – 48)

High pressure seawater, over 350 mm

( >14”)

Duplex (PREN = 43)

Alloy steels are categorized by the pitting resistant equivalent (PREN)

PREN = %Cr+3*%Mo+16*%N (9.2)

The higher the PREN number is the more resistant steel is against corrosion.

Table 9.2. lists PREN number for common alloy steels used in desalination applications. For

seawater applications, usually project specifications includes PREN value of over 42.

Table 9.2. Relevant composition and PREN values of alloy steels

Alloy type Cr Mo N PREN

304L 17.0 – 19.5 - 0.12 – 0.22 18.9 – 23.0

316 16.5 – 18.5 2.0 – 2.5 0.11 max 23.1 – 28.5

316L 17.0 – 19.0 2.5 – 3.0 0.11 max 25.3 – 30.7

904 19.0 – 21.0 4.0 – 5.0 0.15 max 32.2 – 39.9

2250 22.1 3.1 0.18 34.3

254 SMO 19.5 – 20.5 6.0 – 7.0 0.18 – 0.25 42.2 – 47.6

654 SMO 24.0 7.3 0.5 53.9

Zeron 100 24.0 – 26.0 3.0 – 4.0 0.20 – 0.30 37.9 – 45.7

To assure stability of piping components the flow velocity should be in proper range, as listed in

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



251

Table 9.3.

Table 9.3. Recommended flow velocity range in RO applications

Material type Flow velocity, m/sec (ft/sec)

Stainless steel 2.5 – 3.5 (8 – 12)

PVC, schedule 80 1.5 – 2.0 (5 – 7)

PVC, schedule 40 1.0 – 1.5 (3 – 5)

FRP (GRP) 1.5 – 2.0 (5 – 7)

10.0. Example of design of 100,000 m3/day brackish RO desalination system

The following example provides preliminary design specifications of brackish water system of

product water capacity of 100,000 m3/day.

The assumptions are that the system will operate with good quality well water. Therefore, the

pretreatment will be limited to acidification, addition of scale inhibitor and cartridge filtration.

10.1. Raw water source The raw water is provided from brackish well of salinity of about 2,000 ppm TDS. The well water contains very low concentration of suspended particles. Turbidity is below 0.2 NTU and SDI about 1. Table 10.1. Well Water Quality

Constituent Units Concentration, mg/l

TDS mg/l 2053

Temperature C 28

pH 7.7

Turbidity NTU 0.2

Calcium mg/l 120

Magnesium mg/l 4.6

Sodium mg/l 600

Potassium mg/l 5.5

Barium mg/l 0.05

Strontium mg/l 1.9

Carbonate mg/l 1.2

Bicarbonate mg/l 250

Sulfate mg/l 170

Chloride mg/l 879

Fluoride mg/l 3.8

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



252

Constituent Units Concentration, mg/l

SiO2 mg/l 17.0

10.2 Product Water Quality

The required product water quality is < 500 ppm TDS. The RO system produces permeate of

salinity of about 90 ppm TDS. The permeate water will be blended with well water at the ratio of

8.5:1.5 respectively. For blending well water prior to acidification and addition of scale inhibitor

will be used.

Table 10.2. Permeate and product water quality

Constituent Units Concentration RO permeate, mg/l

Concentration after blending, mg/l

TDS mg/l 88.5 383

Temperature C 28 28

pH 6.0 7.2

Calcium mg/l 1.3 19.0

Magnesium mg/l 0.05 0.7

Sodium mg/l 30.0 115.5

Potassium mg/l 0.3 1.1

Barium mg/l 0.0 0.0

Strontium mg/l 0.0 0.3

Carbonate mg/l 0.0 0.0

Bicarbonate mg/l 17.0 55.0

Sulfate mg/l 2.0 27.2

Chloride mg/l 36.8 163.2

Fluoride mg/l 0.3 0.8

SiO2 mg/l 0.6 3.0

10.3. Pretreatment System

The pretreatment system will include three process steps: acidification, addition of scale inhibitor

and cartridge filtration. Acid dosing system, scale inhibitor dosing system and cartridge filtration

units will be common to all ten trains.

Design data of the pretreatment systems are listed in Table 10.3.

Table 10.3. Pretreatment System Design Data

Item description Units Value

Pretreatment – acid dosing system

Solution strength

nominal

percent

96

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



253


Solution specific gravity

nominal

range

--

--

1.83

1.80 – 1.85

Dosing rate ppm 20.0

Number of Installed Tanks -- 1

Operating configuration -- Single duty

Materials -- Steel

Capacity m3 10.0

Number of Installed Pumps -- 3

Operating configuration -- 2 duty, 1 standby

Pump type -- Hydraulic diaphragm

Pump materials

head

diaphragm

--

--

Alloy 20

PTFE

Capacity of Each Pump l/hour 50.0

Pump discharge set pressure bar 2

Motor Size of Each Pump horsepower 0.5

Drive System of Each Pump -- SCR with tachometer feed back

Pretreatment – scale inhibitor dosing system

Solution strength

nominal

percent

100


nominal

range

--

--

1.10

1.00 – 1.15

Dosing rate ppm 1.0



Materials -- FRP

Capacity m3 5


Operating configuration -- 2duty, 1 standby


Pump materials

head

diaphragm

--

--

316 stainless steel

PTFE

Capacity of Each Pump l/hour 10




Pretreatment – cartridge filters

Number of Units -- 11


Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



254


Type -- Wound Horizontal or Vertical

Elements

Materials

Housing

Filters

--

--

316L Stainless Steel

Polypropylene

Capacity per Filter m3/hr 420

Cartridges per filter (100 cm length) cartridges 100

Cartridge filter Rating microns 10

Design operating pressure bar 10

Differential pressure

clean

dirty

bar

bar

0.1

1.0

10.4. Equipment Description

The RO system consists of ten trains, each arranged with two stages of treatment with vessels in

a 34:14 array, seven elements per vessel. The trains are supplied as individual skids. Each train

also includes a feed pump, instrumentation, valves, flow meters, piping and sample tray.

The two stage membrane units will operate at recovery rate of 85%. Permeate throttling will be applied to the first stage to equilibrate fluxes between stages. The permeate line from the first stage in each train will be equipped with permeate throttling valve. Permeate throttling will be in the range of 0 – 3 bar. The trains are fed from a single header from the cartridge filters, which splits to each train and its associated feed pump. The feed pumps are equipped with variable frequency drives (VFDs) that modulate the pump speed to maintain constant flow. Each train has an automated isolation valve to isolate the train in response to an alarm situation or when required for maintenance. Pressure is monitored throughout the system to provide warning to changes in operating conditions and potential problems with the system. Feed pressure is monitored at the feed pump inlet and discharge. Pressure of permeate and reject are monitored for the first stage and second stage. Conductivity is monitored for both permeate and reject from each train. In addition the pH, conductivity, and flow for the combined reject from the trains are monitored. Sample ports are located on the permeate line from each pressure vessel, the feed line (pre-cartridge filter), the permeate line, and the reject line. The outline of the design data for the RO system is listed in Table 10.4

Table 10.4. System Design Data

Item Description Units Value

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



255


General RO Membrane System

System Capacity Product m3/day 100,000

RO System Feed Flow m3/hr 4,170

RO system Permeate Flow m3/hr 3,540

Blending flow 630

Operating Configuration -- Ten two stage membrane units

System Recovery Rate percent 85

RO Membrane Trains

Number of Trains -- 10

Pressure Vessel Array -- 34:14

System Permeate Capacity, per Train m3/day 8,500

Feed Flow per Train m3/hr 417

Concentrate Flow per Train m3/hr 63

Train Recovery Rate percent 85

RO feed Pumps


Operating Configuration -- 10 Duty, 1 standby

Pump Type -- Vertical Turbine, In-Line

Pump Materials

Discharge Head Shaft Impellers

-- -- --

Type 316L Stainless Steel Nitronic 50

Type 316 Stainless Steel

Capacity of Pump m3/hr 450

Total Dynamic Head of Pump m 150

Motor Size of Pump horsepower 380

Drive System for Pump -- Variable Speed

RO Unit Pressure Vessel Racks

Number of Trains -- 10

Number of Pressure Vessels per Train -- 48

Number of First Stage Pressure Vessels -- 34

Number of Second Stage Pressure Vessels -- 14

Number of Membrane Elements per Vessel

-- 7

Design Pressure bar 13.4

Membrane Elements

Number of Elements per Train -- 336

Area per Element m2 40

Average Flux Rate l/m2/hr 25.8

Element Type -- ESPA2MAX

Concentrate Control Valve

Number Installed Per Train -- 1

Operating Configuration -- Duty

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



256


Valve Type -- Ball Valve

Pump Materials Body

Stem Seals

--

-- --


Type 316 Stainless Steel PTFE

Size mm 150

Design Flow m3/hr 60 - 105

Rate Pressure Drop bar 12.0

Actuator -- Electric Modulating

10.5. Cleaning in Place (CIP) unit

The Cleaning in Place system is used to circulate chemicals through the RO membrane elements

when they have become scaled or fouled. The major components of a system are a tank with a

heater, a pump, and a cartridge filter.

Chemical solutions selected to effectively remove the observed fouling are mixed in the CIP

tank. The CIP solution is heated with the heater to increase the effectiveness of the cleaning. The

tank is equipped with a temperature element to monitor the temperature. The tank also has a level

indicator and high and low level switches to prevent the level from going too high or too low

during the recirculation of the cleaning solution.

The CIP pump is used to circulate the cleaning fluid through the RO membranes. One train is

cleaned at a time while the other trains remain in operation. The cartridge filter removes particles that are entrained in the solution during cleaning.

The pressure, temperature, and pH of the CIP solution are measured at the pump discharge. A

pressure indicator after the cartridge filter allows the differential pressure to be monitored. Flow

of CIP solution is measured on the CIP skid before the solution is fed to the RO skid.

Piping in the CIP system allows the solution to be re-circulated within the system while it is

being heated and mixed. Disposal of the used solution is through a line to the sanitary sewer.

Cleaning is typically performed every 6 to 12 months. Chemicals used in CIP include:

High pH – 0.1% NaOH

0.25% Na-DDS

Low pH – 1% citric acid

0.2% HCL

Equipment and instruments in the CIP system are identified in Table 10.5.

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



257

Table 10.5. CIP unit design data




Materials -- FRP

Capacity m3 15.0



Pump Type -- Horizontal Centrifugal

Pump Materials

Casing

Impeller

--

--

Type 316L Stainless Steel


Capacity m3/hr 280

Total Dynamic Head m 50

Motor Size horsepower 85

Drive System -- Constant Speed

Number of Installed Heaters -- 2

Operating Configuration -- Lead, Lag

Capacity of Each Heater kW 120

Service Voltage V 480

Cartridge Filters -- 1

Materials -- FRP

Cartridge Size Micron 5

10.6. Post treatment

The post-treatment process consists of five counter-current forced-draft degasifiers for the removal of CO from RO permeate. The permeate flow from the RO trains is manifolded together into a common discharge pipeline that conveys the total NF process permeate flow to the degasifiers. The permeate is discharged into the top of the degasifiers, where it flows into a distribution tray that distributes the flow evenly over the internal plastic packing material. The permeate flows from the top of the decarbonator down through the plastic packing material to the base of the unit, where the permeate then passes through a pipe into the clearwell below. Air is forced upward through the packing material by a blower connected to the side of each of the decarbonators. The downward flow of water is thereby contacted by the counter-current flow of air. As the water flows downward, the packing promotes fluid distribution and greatly increases the liquid phase to gaseous phase mass transfer area. Carbon dioxide present in the permeate then diffuses from the liquid to the gaseous phase and exists with the exhaust air at the top of the decarbonator. The well water for blending will be connected to the permeate line at the outlet from the degasifier. Design data for the degasifiers are contained in Table 10.6

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



258

Table 10.6. Post treatment unit design data


Post-treatment - Degasifiers

Number of Units -- 1

Operating Configuration -- 1 Duty

Type -- Forced Draft

Materials Tower Packing

-- --

FRP

Polypropylene

Capacity per Unit m3/hr 850

Tower Diameter m 4

Tower Height (straight shell) m 7.5

Minimum Packing Depth m 2

Packing Volume per Unit cf 18

Maximum Hydraulic Loading m/hr 70

Air to Water Ratio m/m 2.5

Blower Capacity scm/sec 4.7

Static Pressure cm wc 5.0

Motor Size hp 15

Post-treatment – sodium hydroxide dosing system

Solution strength

nominal

percent

40


nominal

range

--

--

1.52

1.48 – 1.55

Dosing rate ppm 17.5



Materials -- FRP

Capacity m3 10




Pump materials

head

diaphragm

--

--

316 stainless steel

PTFE

Capacity of Each Pump l/hour 10




10.7. Equipment list

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



259

A detailed equipment list of all major equipment was prepared for the project. This list is based

on the preliminary design data laid out in this report.

Table 10.7. Major equipment list

Equipment Number of units Capacity, m3/hr Power req., hp Comments

RO feed pump 11 450 380 VFD

CIP pump 1 280 85 Constant

Cartridge filters

Cartridge filters 11 420

Cartridge filter,

CIP

1 280

RO trains 10 354

Tank, sulfuric

acid

1 10 m3

Tank scale

inhibitor

1 2 m3

Tank, CIP 1 15 m3

Degasifier 5 850

Blower 5 4.7 scm/sec

Chemical feed

pump, sulfuric

acid

3 50 l/hr 0.5

Chemical feed

pump, scale

inhibitor

3 10 l/hr 0.3

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



260

10.8 Computer projections for RO membrane unit without blending

Permeate THROTTLING(1ST STAGE)

RO program licensed to: Calculation created by: Project name: Brackish RO Permeate flow: 8500.00 m3/d HP Pump flow: 416.7 m3/hr Raw water flow: 10000.0 m3/d Permeate throttling(1st st.) 3.0 bar Feed pressure: 13.4 bar Permeate recovery: 85.0 % Feedwater Temperature: 28.0 C(82F) Feed water pH: 7.2 Element age: 3.0 years Chem dose, ppm (100%): 19.1 H2SO4 Flux decline % per year: 7.0 Fouling factor: 0.80 Salt passage increase, %/yr: 10.0 Average flux rate: 25.8 lm2hr Feed type: Well Water






Ca 120.0 6.0 120.0 6.0 1.278 0.1 792.8 39.5 Mg 4.6 0.4 4.6 0.4 0.049 0.0 30.4 2.5 Na 600.0 26.1 600.0 26.1 30.062 1.3 3829.6 166.5 K 5.5 0.1 5.5 0.1 0.342 0.0 34.7 0.9 NH4 0.0 0.0 0.0 0.0 0.000 0.0 0.0 0.0 Ba 0.050 0.0 0.050 0.0 0.001 0.0 0.330 0.0 Sr 1.900 0.0 1.900 0.0 0.020 0.0 12.552 0.3 CO3 1.2 0.0 0.2 0.0 0.000 0.0 1.2 0.0 HCO3 250.0 4.1 228.4 3.7 16.953 0.3 1426.7 23.4 SO4 170.0 3.5 188.7 3.9 2.010 0.0 1246.6 26.0 Cl 879.0 24.8 879.0 24.8 36.855 1.0 5651.2 159.4 F 3.8 0.2 3.8 0.2 0.312 0.0 23.6 1.2 NO3 0.0 0.0 0.0 0.0 0.000 0.0 0.0 0.0 B 0.00 0.00 0.000 0.00 SiO2 17.0 17.0 0.57 110.1 CO2 7.02 24.20 24.20 24.20

TDS 2053.1 2049.1 88.5 13159.6

pH 7.7 7.2 6.0 7.8

Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



261

10.9 Computer projections for RO membrane unit with blending

PERMEATE BLENDING AND PERMEATE THROTTLING(1ST STAGE)

RO program licensed to: Calculation created by: Project name: Brackish RO Permeate flow: 8500.00 m3/d HP Pump flow: 416.7 m3/hr Raw water flow: 11500.0 m3/d Blended flow: 10000.0 m3/d Feed pressure: 13.4 bar Permeate recovery: 85.0 % Feedwater Temperature: 28.0 C(82F) Blending ratio: 15.0 % Permeate throttling(1st st.) 3.0 bar Feed water pH: 7.2 Element age: 3.0 years Chem dose, ppm (100%): 19.1 H2SO4 Flux decline % per year: 7.0 Fouling factor: 0.80 Salt passage increase, %/yr: 10.0 Average flux rate: 25.8 lm2hr Feed type: Well Water







Mark Wilf Ph. D.

Phone: +1 858 444 7334

RO Technology



262

11.0. Example of design of 100,000 m3/day seawater RO desalination system

desalination process engineering part ii · 2018. 3. 22. · mark wilf ph. d. phone: +1 858 444...

Documents