desalination process engineering part ii · 2018. 3. 22. · mark wilf ph. d. phone: +1 858 444...
TRANSCRIPT
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
165
(NOM) decline. passage decline
Biological matter Some increase of
pressure drop. Some
permeability and salt
passage decline.
Initially in the lead
element(s).
Significant increase of
pressure drop. Some
permeability and salt
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).
Significant increase of
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
permeability and salt
passage
Significant increase of
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
195
Figure 5.7.4. Computer projection program – membrane elements look up table
Mark Wilf Ph. D.
Phone: +1 858 444 7334
RO Technology
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
198
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
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
E-mail: [email protected]
Webpage:www.rotechnology.net
199
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
********************************************************************* **** 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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
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
1-1 541.7 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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
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
1-1 541.7 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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
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
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
Ca 1.2 0.1 1.2 0.1 0.008 0.0 8.3 0.4 Mg 4.3 0.4 4.3 0.4 0.026 0.0 28.3 2.3 Na 144.4 6.3 162.8 7.1 4.732 0.2 1058.2 46.0 K 7.9 0.2 7.9 0.2 0.285 0.0 50.8 1.3 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 0.029 0.0 0.029 0.0 0.000 0.0 0.194 0.0 CO3 0.0 0.0 6.0 0.2 0.018 0.0 39.7 1.3 HCO3 4.1 0.1 3.4 0.1 0.182 0.0 21.7 0.4 SO4 8.5 0.2 8.5 0.2 0.064 0.0 56.1 1.2 Cl 248.8 7.0 248.8 7.0 7.456 0.2 1616.2 45.6 F 0.0 0.0 0.0 0.0 0.002 0.0 0.2 0.0 NO3 0.1 0.0 0.1 0.0 0.014 0.0 0.4 0.0 B 1.28 1.28 0.12 7.81 SiO2 0.0 0.0 0.000 0.1 CO2 3.87 0.00 0.00 0.00 TDS 420.5 444.2 12.91 2887.9 pH 6.2 10.5 9.3 11.3
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
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
1-1 541.7 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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
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
1-1 541.7 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
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 2.28 Stiff & Davis Saturation Index 0.78 0.15 1.26 Ionic strength 0.91 0.92 1.67 Osmotic pressure 34.0 bar 34.5 bar 62.5 bar
Mark Wilf Ph. D.
Phone: +1 858 444 7334
RO Technology
E-mail: [email protected]
Webpage:www.rotechnology.net
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
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
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 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
Ca 2.6 0.1 2.6 0.1 0.044 0.0 17.1 0.9 Mg 8.9 0.7 8.9 0.7 0.152 0.0 58.4 4.8 Na 301.0 13.1 342.7 14.9 27.159 1.2 2130.8 92.6 K 16.4 0.4 16.4 0.4 1.605 0.0 100.1 2.6 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 0.061 0.0 0.061 0.0 0.001 0.0 0.401 0.0 CO3 0.0 0.0 8.7 0.3 0.076 0.0 57.6 1.9 HCO3 8.2 0.1 3.0 0.0 0.441 0.0 17.4 0.3 SO4 17.3 0.4 17.3 0.4 0.377 0.0 113.5 2.4 Cl 510.7 14.4 510.7 14.4 43.154 1.2 3160.1 89.1 F 0.1 0.0 0.1 0.0 0.013 0.0 0.4 0.0 NO3 0.1 0.0 0.1 0.0 0.069 0.0 0.5 0.0 B 2.95 2.95 0.54 16.56 SiO2 0.0 0.0 0.000 0.1 CO2 2.83 0.00 0.00 0.00 TDS 868.3 913.6 73.63 5673.1 pH 6.6 10.5 9.8 11.3
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
NaOH, hypochlorite, ammonia
Brackish RO, NF, surface
water feed
Ferric coagulant, filtration
polymer, hypochlorite, sodium
bisulfite, acid (H2SO4 or
HCl), scale inhibitor
NaOH, hypochlorite, ammonia
Mark Wilf Ph. D.
Phone: +1 858 444 7334
RO Technology
E-mail: [email protected]
Webpage:www.rotechnology.net
230
Seawater RO, beach well feed Acid (H2SO4 or HCl), scale
inhibitor
CO2, CaCO3, Ca(OH)2,
NaOH, hypochlorite, ammonia
Seawater RO, surface water
feed
Ferric coagulant, filtration
polymer, hypochlorite, sodium
bisulfite, acid (H2SO4 or
HCl), scale inhibitor
CO2, CaCO3, Ca(OH)2,
NaOH, hypochlorite, ammonia
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
I/O I/O I/O I/O I/O I/O
Hot Backup
Control
Figure 8.4. Control system configuration – ‘hot” backup (PLC only)
PLC Processor PLC Processor
I/O I/O I/O I/O I/O I/O
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
253
Item description Units Value
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
Solution specific gravity
nominal
range
--
--
1.10
1.00 – 1.15
Dosing rate ppm 1.0
Number of Installed Tanks -- 1
Operating configuration -- Single duty
Materials -- FRP
Capacity m3 5
Number of Installed Pumps -- 3
Operating configuration -- 2duty, 1 standby
Pump type -- Hydraulic diaphragm
Pump materials
head
diaphragm
--
--
316 stainless steel
PTFE
Capacity of Each Pump l/hour 10
Pump discharge set pressure bar 2
Motor Size of Each Pump horsepower 0.3
Drive System of Each Pump -- SCR with tachometer feed back
Pretreatment – cartridge filters
Number of Units -- 11
Operating configuration -- 10 duty, 1 standby
Mark Wilf Ph. D.
Phone: +1 858 444 7334
RO Technology
E-mail: [email protected]
Webpage:www.rotechnology.net
254
Item description Units Value
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
E-mail: [email protected]
Webpage:www.rotechnology.net
255
Item Description Units Value
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
Number of Installed Pumps -- 11
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
E-mail: [email protected]
Webpage:www.rotechnology.net
256
Item Description Units Value
Valve Type -- Ball Valve
Pump Materials Body
Stem Seals
--
-- --
Type 316 Stainless Steel
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
E-mail: [email protected]
Webpage:www.rotechnology.net
257
Table 10.5. CIP unit design data
Item Description Units Value
Number of Installed Tanks -- 1
Operating Configuration -- Duty
Materials -- FRP
Capacity m3 15.0
Number of Installed Pumps -- 1
Operating Configuration -- Duty
Pump Type -- Horizontal Centrifugal
Pump Materials
Casing
Impeller
--
--
Type 316L Stainless Steel
Type 316 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
E-mail: [email protected]
Webpage:www.rotechnology.net
258
Table 10.6. Post treatment unit design data
Item Description Units Value
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
Solution specific gravity
nominal
range
--
--
1.52
1.48 – 1.55
Dosing rate ppm 17.5
Number of Installed Tanks -- 1
Operating configuration -- Single duty
Materials -- FRP
Capacity m3 10
Number of Installed Pumps -- 3
Operating configuration -- 2 duty, 1 standby
Pump type -- Hydraulic diaphragm
Pump materials
head
diaphragm
--
--
316 stainless steel
PTFE
Capacity of Each Pump l/hour 10
Pump discharge set pressure bar 2
Motor Size of Each Pump horsepower 0.3
Drive System of Each Pump -- SCR with tachometer feed back
10.7. Equipment list
Mark Wilf Ph. D.
Phone: +1 858 444 7334
RO Technology
E-mail: [email protected]
Webpage:www.rotechnology.net
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
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
1-1 293.2 12.3 3.6 30.1 1.10 11.6 3.0 ESPA2 MAX 238 34x7 1-2 61.0 8.8 4.5 15.2 1.04 10.1 0.0 ESPA2 MAX 98 14x7
Raw water Feed water Concentrate CaSO4 / Ksp * 100: 3% 4% 40% SrSO4 / Ksp * 100: 3% 4% 40% BaSO4 / Ksp * 100: 115% 127% 1271% SiO2 saturation: 12% 12% 75% Langelier Saturation Index 0.62 0.07 2.28 Stiff & Davis Saturation Index 0.61 0.07 1.72 Ionic strength 0.04 0.04 0.24 Osmotic pressure 1.4 bar 1.4 bar 9.1 bar
Raw water Feed water Permeate Concentrate Ion mg/l meq/l mg/l meq/l mg/l meq/l mg/l meq/l
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
E-mail: [email protected]
Webpage:www.rotechnology.net
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
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
1-1 293.2 12.3 3.6 30.1 1.10 11.6 3.0 ESPA2 MAX 238 34x7 1-2 61.0 8.8 4.5 15.2 1.04 10.1 0.0 ESPA2 MAX 98 14x7
Raw water Feed water Permeate Concentrate Ion mg/l meq/l mg/l meq/l mg/l meq/l mg/l meq/l
Ca 120.0 6.0 120.0 6.0 19.086 0.0 792.8 39.5 Mg 4.6 0.4 4.6 0.4 0.732 0.0 30.4 2.5 Na 600.0 26.1 600.0 26.1 115.553 0.0 3829.6 166.5 K 5.5 0.1 5.5 0.1 1.116 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 0.008 0.0 0.330 0.0 Sr 1.900 0.0 1.900 0.0 0.302 0.0 12.552 0.3 CO3 1.2 0.0 0.2 0.0 0.185 0.0 1.2 0.0 HCO3 250.0 4.1 228.4 3.7 51.910 0.0 1426.7 23.4 SO4 170.0 3.5 188.7 3.9 27.208 0.0 1246.6 26.0 Cl 879.0 24.8 879.0 24.8 163.177 0.0 5651.2 159.4 F 3.8 0.2 3.8 0.2 0.835 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 3.04 110.1 CO2 7.02 24.20 10.28 24.20 TDS 2053.1 2049.1 383.1 13159.6 pH 7.7 7.2 6.5 7.8
Raw water Feed water Concentrate CaSO4 / Ksp * 100: 3% 4% 40% SrSO4 / Ksp * 100: 3% 4% 40% BaSO4 / Ksp * 100: 115% 127% 1271% SiO2 saturation: 12% 12% 75% Langelier Saturation Index 0.62 0.07 2.28 Stiff & Davis Saturation Index 0.61 0.07 1.72 Ionic strength 0.04 0.04 0.24 Osmotic pressure 1.4 bar 1.4 bar 9.1 bar
Mark Wilf Ph. D.
Phone: +1 858 444 7334
RO Technology
E-mail: [email protected]
Webpage:www.rotechnology.net
262
11.0. Example of design of 100,000 m3/day seawater RO desalination system