Sie sind auf Seite 1von 13

DESALINATION

ELSEVIER Desalination 135 (2001) 141-153


www.elsevier.conglocate/desal

Fifteen years of R&D program in seawater desalination at KISR


Part I. Pretreatment technologies for RO systems

S. Ebrahim*, M. Abdel-Jawad, S. Bou-Hamad, M. Safar


Kuwait Institute for Scientific Research, Water Desalination Department, P.O. Box 24885, Safat 13109, Kuwait
Tel. + 965 4878122/4; Fax + 965 4879238; e-mail: sebrahim@safat, kisr.edu, kw

Received 20 November 2000; accepted 29 November, 2000

Abstract

Reverse osmosis (RO) has been established as a reliable process for seawater desalination. The reliability of the
RO technique depends upon the extent of the pretreatment that must be designed to reduce the suspended and colloidal
materials in the feedwater. During the last 15 years of research and development work at the Doha Research Plant
(DRP), the Kuwait Institute for Scientific Research (KISR) has evaluated various types of pretw~atmenttechnologies
for RO systems. These included: conventional pretreatrnent (CP), microfiltmtion (MF) and the beachwell (BW) intake
systems. This paper gives brief description of these pretreatment technologies which were implemented at DRP and
discusses their performances

Keywords:Reverse osmosis; Seawater desalination; Prea'eatment technologies

1. Introduction performance of an RO system will only be as


Pretreatment o f the seawater feeding the good as the feedwater is properly pretreated
membrane in the desalination process is the most before feeding it to the system. Most RO systems
critical step in designing the plant. Currently, include some level of feedwater pretreatment
available RO membranes are generally not designed to
robust enough to operate directly on surface feed • Prevent fouling of the membranes;
seawater. Feedwater usually contains particles • Maintain performance (i.e., salt rejection and
that can adversely affect the performance and recovery) of the system; and
lifetime o f the RO membrane. Therefore, the • Extend the lifetime of the membranes.

Different pretreatment systems are used for


*Corresponding author.

Presented at the International Conference on Seawater Desalination Technologies on the Threshold of the New
Millennium, Kuwait, 4-7 November 2000.

0011-9164/01/$- See front matter © 2001 Elsevier ScienceB.V. All fights reserved
PII: S0011-9164(01)00146-1
142 S. Ebrahim et al. / Desalination 135 (2001) 141-153

RO plants. Generally, they are divided into con- (SWRO) desalination process to prevent memb-
ventional (i.e., disinfection, flocculation/coagu- rane fouling. Feed seawater is pretreated by the
lation and filtration processes), and non-conven- conventional method to lower the silt density
tional (i.e., microfiltration, ultrafiltration and index (SDI), remove excessive turbidity or
beachwell). suspended solids, and adjust and control the pH.
Conventional pretreatment is widely used for Additional pretreatment is carried out to inhibit
seawater RO plants, but is costly and space- or control the formation of scale and disinfect or
consuming. Dual-media filtration is a complex prevent biological and slime growth and
process and depends on the physico-chemical contamination of equipment. Type and size of
characteristics of the particles to be removed, the the conventional pretreatment system vary
filter grains and the filtration system as a whole. depending on the nature of the feedwater to be
Most important, the quantity and quality of the treated and type of membrane used in the RO
filtrate are not steady. system. Cost of the conventional pretreatment
The beachwell intake system has been system also depends on the type and size of the
identified as a reliable technique to replace the pretreatment system.
conventional surface seawater pretreatment. This At DRP, a conventional pretreatment system
system can deliver a continuous flow of naturally (common pretreatment) was used to treat surface
filtered seawater and has almost steady physical seawater needed to feed to three different RO
characteristics (i.e., temperature and water lines. The average values for the main con-
quality). In addition, elimination of the coagu- stituents of the surface seawater feed to the CP
lation, filtration and backwash systems reduce are presented in Table 1.
the amount of chemicals needed and power Each of the three lines was divided into two
consumed. Nevertheless, it has the drawback of stages as follows:
site dependency and limitation of water quantity.
Membrane separation (i.e., MF, UF) is also RO Line 1
used for pretreatment of feedwater. MF is a First stage Spiral wound, UOP-PA 1501
successful filtration technique applied to surface Second stage Spiral wound, UOP-PA 8600
seawater, brackish water, and industrial and
urban seawater. It relies on the membrane RO Line 2
separation technique that requires low pressure First stage Hollow fine fiber, DuPont B-10
to force good quality filtered water through the Second stage Hollow fine fiber, DuPont B-9
membrane, thus rejecting harmful contaminates,
including micro-organisms. RO Line 3
Since 1984, KISR has carried out an intensive First stage Plate and frame, Enro +Schleicher
R&D program to assess the technical viability & Schuell + FilmTec
and economic feasibility of desalting seawater Second stage Spiral wound, Hydronautics 8040B
using RO technology. The program has included
evaluation of conventional and non-conventional Each line had its own additional pretreatment
(MF and beachwell) pretreatment techniques. system, in addition to common pretreatment~ and
was designed to produce 1,000mVd of fresh
2. Conventional pretreatment water. The common pretreatment and RO
Lines 1 and 2 were in operation from 1984 until
The conventional pretreatment method is August 1990. Membranes of the first stage of
widely applied for seawater reverse osmosis RO Line 1 were replaced in July 1989 by a new
S. Ebrahim et al. / Desalination 135 (2001) 141-153 143

Table 1
Average seawater quality at Doha

Parameter Concentration*,mg/1
TDS at 180°C 47,000 -~ 2,000
Total alkalinity, as CaCO3 175 ± 15
Carbonate 14 ± 8
Bicarbonate 185 ± 18
Carbon dioxide 14 ± 4
Sulfate 3,400 ± 300
Chloride 24,000 ± 700
Calcium 570 ± 45
Magnesium 1,700 ± 150
Sodium 12,300 ± 20
Potassium 470 ± 20
Total iron 0.08 ± 0.08
PH 8.2 ± 0.1
SDI value, % >6.5 ±

* Average concentration of 12 samples.

8" spiral wound configuration system from destabilization (coagulation and flocculation)
FilmTec (FT-30). Modules of RO Line 2 were and agglomeration.
also replaced in December 1988 by the new Destabilization means charge compensation
version o f hollow fiber DuPont (B-10T). With of the usually negative colloidal particles, thus
regard to RO Line 3, the plant was shut down in leading to the formation of microflakes. This is
March 1988 due to technical and economic achieved by mixing the untreated water with the
reasons related to the plate and frame configu- flocculant using a substantial amount of energy.
ration system. The flocculant, in this case FeC1SO4 (ferric
chloride sulfate), is added to the untreated water
at the inlet to the destablization tank. The energy
required for mixing is applied by the agitator.
2.1. Description o f common pretreatment system
Dwell time for the water is approximately 15 s at
The chlorinated surface seawater is delivered maximum flow. The ideal pH value for destabi-
to the common pretreatment system at the Doha lization and flocculation is set by adding H2SO4
East Power Station via a reinforced fiberglass (sulfuric acid), also at the inlet to the destabili-
pipe at a pressure o f 2-3 bar (Fig. 1). If the zation tank.
chlorine content is below the required level, In the agglomeration stage, energy is used to
chlorine gas is added. The removal of suspended combine the destabilized particles or micro-
and colloidal particles from the untreated water flakes. At this stage, mieroflakes are formed by a
is carded out by flocculation and filtration. The series of processes such as chemical trans-
process o f flocculation may be divided into two formations, electrokinetic effects and adsorption
separate stages that merge into one another: phenomena. Excess turbulence leads to a
144 S. Ebrahim et al. / Desalination 135 (2001) 141-153

Polyelectrolyte
dosing
Ferric chloride
sulphate dosing

Acid dosing

the Destabilization FlocculaUon stage


tank
Filtration stage

J
Fig. 1. Schematic diagram for the common pretreatment system at DRP.

disintegration of flakes that have already formed. type. The mixture of water and flakes flows
These processes take place in the cascade downstream through the filter material, where
configuration of tanks. To assist with agglome- various types of filter materials are placed on top
ration, the discharge of tanks is provided with a of one another. The size of grains is finer in the
dose of polyelectrolytic solution that has an direction of flow so that the particles to be
interlacing effect on the flakes that have formed. filtered out will be deposited uniformly over the
The energy necessary for coagulation is applied entire height of the filter so it is possible to avoid
by three agitators. To optimize the application of a large increase in pressure loss due to surface
energy, the speed of the agitators in the agglome- filtration. The layer distribution in the filters is as
ration basins is adjustable. Furthermore, the last follows:
agitator, in the direction of flow, is fitted with a
torque-measuring instrument. The dwell time of Grain size, m m Filter height, m
the water in the agglomeration tanks is approxi- Supporting layer various 0.3
mately 10min at maximum flow. The pretreated Silica sand 0.7-1.2 1
water leaves the last tank via a pipe and is taken Hydroanthracite 1.4-2.5 0.7
to the filter station.
The purpose of filtration is to remove the The mixture of water and flakes is distributed
flakes from the water. This is achieved by passing evenly over the filter layer and leaves the filters
the pretreated water through four parallel multi- via the pure-water discharge valves and the pure-
layer filters. The filters are of the open gravity water pipe. Flow through the filters is maintained
S. Ebrahim et al. / Desalination 135 (2001) 141-153 145

constant by the two filter outlet controls. The • Polyelectrolyte dosing system and three inline
filtering velocity is 10 m/h at maximum flow. coagulation filters to further reduce the silt
The filters can be backwashed using the density index (SDI) of the feed to less than
filtered water and air supplied by a compressor. 3.0.
The time and sequence of water and air can be • Sodium hydrogen sulphite dosing system to
adjusted at will. The pretreated water emerging remove residual chlorine in the feed.
from the multi-layer filters is collected in the • Three cartridge chlorine in the feed larger
pure-water pipe and fed to the intermediate tank than 5 lain. Two filters are always in operation
before it is delivered to the storage tank. Before and the third is on standby basis.
entering the storage tank, chlorine gas can be
added, if chlorine content is less than the RO Line 3
minimum level. The volume of filtrate being
pumped is controlled in the intermediate tank. • Acid dosing system to prevent carbonate
The filtrate in the storage tank is fed to the three scaling.
RO lines. • Antiscalant dosing system to prevent sulfate
scaling.
• Sodium hydrogen dosing system to remove
2.2. Chemical dosing systems for the RO Lines
residual chloride.
Before the filtered water coming from the CP • Two cartridge filters to filter out particles
enters the first stage of each RO line, further larger than 25 gm.
treatment is carded out separately to obtain the
quality feedwater specified by the manufacturers
of the RO modules. The pretreatment systems
2.3. Performance evaluation
for RO Lines 1, 2 and 3 consist of:
During the period from August 1984 until
RO Line 1 December 1989, CP operated with availability of
• Sodium hydrogen sulfite dosing system to more than 97%. The mean values for monthly
remove residual chlorine in the feed. parameters are presented in Table 2.
• Two parallel activated carbon filters to ensure Most of the time, the system was successfully
the complete removal of residual chlorine and controlled to give the designed quantity of
to act as a holding tank for the sodium filtrate with SDI value of 3.6; but, in some cases,
hydrogen sulphite reaction with chlorine. it failed to produce acceptable quality and the
• Antiscalant dosing system to prevent sulfate required quantity of filtered seawater.
scaling. The causes for these failures can be attributed
• Acid dosing system to prevent carbonate to several factors, mainly: clogging of dual
scaling. media filters, effect of pH, dosing rate of
• Two cartridge filters (micron filters) to filter FeCISO4, dosing rate of polyelectrolyte, energy
out particles larger than 5 lxrn. These two filters input, and climatic conditions (i.e., temperature,
are located just before the high-pressure dust storm, wind).
pumps. During the reported period (August 1984-
December 1989), a total of 21,828,201m 3 of
RO Line 2 filtrate was produced by the CP. Only 70% was
• Acid dosing system to prevent carbonate used as feed for RO lines, whereas the remaining
scaling. 30% was used to backwash the multi-layer filters
146 S. Ebrahim et al. / Desalination 135 (2001) 141-153

Table 2 3. Microfiltration technique


Mean parameters for seawater feed and filtrate for the
Microfiltration (MF) and ultrafiltration (UF)
period from August 1984 to December 1989
are two attractive pretreatment techniques that
Parameter Average are increasingly being used in drinking water,
seawater and wastewater applications. Several
Seawater feed flow rate, m3/h 493.11 projects have shown that MF and UF are capable
Pressure of seawater feed, bar 2.06 of consistently reducing turbidities to <0.1 NTU,
Temperature of seawater feed, °C 26.24 regardless of the influent turbidity level. UF and
SDI of seawater feed, % min 6.47 MF have also been shown to be very effective
pH of seawater feed 8.14
for the removal of total coliform (TC) bacteria,
SDI of seawater filtrate, % min 3.6
Giardia and Cryptosporidium. UF is also very
pH of seawater filtrate 6.21
effective in removing viruses.
Acid dosing (H2SO4)rate, mg/1 82.24
It is expected that MF will compete strongly
FeCISO4 dosing rate (Fe3+),mg/1 3.04
with conventional pretreatment in terms of relia-
Polyelectrolyte dosing rate, mg/1 0.62
bility and total operating costs. It is consequently
C12content in seawater feed, mg/1 0.15
believed that MF will be more widely applied to
C12content in filter inlet, mg/1 1.7
CI2 content after storage tank, rag/1 0.16 RO feedwater pretreatment. The main aspects of
cost reduction are expected to include:
1. Reduction in capital cost:
• No standby capacity
and to reorganize and resettle the filter media • Reduction in the size of chemical dosing
(infiltration), and very small quantity was wasted systems
due to overflow. More details of the performance • Elimination off'me filters in the RO system
of this system is presented in [1]. 2. Reduction in operating costs:
• Less membrane replacement cost (due to
lengthened membrane useful life)
• Less chemical consumption cost (no
2.4. Conclusions chemicals are needed for disinfection,
coagulation and dechlorination)
The conventional pretreatment system used at
• Elimination of cartridge filters cost (no
DRP succeeded in providing feedwater for the
cartridge filters are needed before the RO
RO lines with SDI less than 4.0. The availability
step)
o f this system was more than 97% during the
• Less RO system down-time
operated period; however, this system was very
• Less maintenance cost for the high pressure
difficult to control. Several factors obstructed the
pump and the measuring instruments
continuous full utilization of this system. Among
• Less labor cost (less manpower is needed
these factors were: instability of SDI value, high
to operate the conventional pretreatment
rate of chemical consumption (i.e. acid, flocculant
system and to clean the membranes and
and flocculant-acid), frequent backwashing and
maintain the system)
high rate of water consumption for this purpose,
dangerous hazard resulting from acid storing and At DRP, a study was carded out to investigate
handling, and relative high operating cost the viability of MF as a pretreatment technique
resulting from excess consumption of chemical, for SWRO plants. Details of this study are
filtrate and energy. presented in [2].
S. Ebrahim et al. / Desalination 135 (2001) 141-153 147

3.1. Description o f the microfiltration unit After passing through a coarse strainer, raw
surface seawater is fed to a break tank and
The MF 0VIemcor 20M10 CMF) unit is a
injected with 2.5ppm of sodium bisulfite
skid-mounted MF machine designed to remove
(NaI-ISO3) to remove any traces of chlorine.
impurities larger than 0.2 lain from feedwater. It
From the break tank, the feed stream is fed to the
is designed to operate in the direct filtration mode.
shell side of the two MF modules where
The machine consists of filtration modules,
filtration takes place.
circulation pump, associated valving, pipework,
The microfiltration system uses an air
instrumentation and a control system, all mounted
backwash stream to clean the hollow fiber
in a stainless steel frame. Fig. 2 shows a
membranes. Backwash is automatically controlled
schematic diagram of the unit.
by a PLC. Air at high pressure is injected into
The unit has two microfiltration modules made
the center of the hollow fibers and bursts through
from polypropylene, and each has a nominal
the membranes, removing the foulants that have
membrane filtration area of 10m2. A feed pump
accumulated on the membrane.
drives the raw seawater into the filtration modules
Direct flow microfiltration, incorporating the
form a buffer tank. The process valves are fitted
air backwashing technique, is an energy-efficient
with pneumatic actuators. A programmable logic
system. Unlike UF, this system requires rela-
controller (PLC) mounted in the control cabinet,
tively low shear rates as the membrane is
controls the pilot solenoid valves and pump
primarily kept clean by the air backwash. The
operation. The PLC also monitors various control
high porosity of the membrane (70% porosity)
switches and other inputs, and illuminates the
also allows operation at very low pressures of
appropriate indicator lamps during machine
typically less than 10m head. The unit capacity
operation.
is 6m3/h.

Feed transfer
pump

Feed pump I~ Backwash drain


Drain ~ ....

Fig. 2. Schematicdiagram ofmicrofiltration system.


148 S. Ebrahim et al. / Desalination 135 (2001) 141-153

3.2. Performance evaluation adjustment and coagulant dosing systems,


destabilization and agglomeration tanks and
During the course of operation, several tests
stirrers, sand and dual-media filters, filtered-
and modifications were made to optimize the
water transfer pumps, a filter backwash system, a
performance of the unit. These included chlorine
treated-water storage tank, a dechlorination
dosage, backwash interval, cleaning cycles and
system, scale inhibitor dosing systems (i.e. acid
feed temperature [2].
and threshold scale inhibitors), and cartridge
It was found that the SDI value varied
filters. The usage of a beachwell seawater intake
between 0.24 and 3%, with an average value of
system reduces not only the capital costs (i.e., the
2.42, knowing that the SDI of the seawater feed
surface seawater intake and pretreatment sys-
was over 6.5%. The average filtrate flowrate was
tems), but also costs for membrane degradation,
3.2mVh. The optimal backwash interval was
operation, and manpower. In addition, elimination
10 min at a variable feed flowrate. The MF unit
of the coagulation, filtration and backwash
was capable of producing good quality water
systems reduces the amount of chemicals needed
suitable as a feed for RO systems.
and the power consumed.
At DRP, a study was carried out to assess the
3.3. Conclusions viability of using the beachwell intake system as
With regard to the MF system, the overall a feed for SWRO plants. The detailed outcome
performance of the MF system revealed that it is of this study is presented in [3].
capable of treating surface seawater to produce
good quality water on a continuous basis that is 4.1. Description o f the beachwell system
suitable to feed an RO system without further
Several test holes were drilled near the beach
treatment (average SDI value of 2.42% and
at the Doha site to find out the most suitable
optimal backwash frequency every 10 min).
strata and depth of the well with the right quality
as seawater intake. Enough quantity and seawater
quality was found at 30m depth and 30m from
4. Beachwell intake system
the beach side. This water was free from
Beachwell seawater intake has been identified hydrogen sulfide and had low carbonate and
as a promising technique to replace conventional bicarbonate content, which is suitable as feed for
surface seawater pretreatment. Beachwell water SWRO plants.
intake systems have been successfully used with The beachwell consists of a well pump, well
minimum pretreatment to feed RO plants in casing, screens, submersible pumps, risers, pump
several countries. The beachwell system, com- starters, a complete dosing system and inter-
prised of wells, well casings, a screen, sub- connecting pipes. A block diagram of the system
mersible pumps, riser, pump starters, chemical is shown in Fig. 3. The feedwater is filtered
dosing systems and interconnecting pipes, can naturally through the beachwell filtration sand
replace a conventional surface seawater pretreat- system. The system capacity is 90 mVh.
ment system that usually includes a pipeline or
channel to convey surface seawater to the pump
4.2. Performance evaluation
pit, a coarse filtration system to prevent the
intrusion of fish and large objects, seawater During the course of operation, the beachwell
transfer pumps, a chlorine dosing system, p H water temperature remained relatively steady
S. Ebrahim et al. / Desalination 135 (2001) 141-153 149

VALVE seawater ranged between 0.1 and 7.8 ppm. Most


measurements, however, were between 3.5 and
4.5ppm. The dissolved oxygen content in the
beachwell water was generally less than 1 ppm.
Highly dissolved oxygen content was noticed
when there were leaking pipe joints or after
shutdowns of the submersible pump.
The SDI of the beaehwell water was high
immediately after the well was drilled, but
decreased to less than 1% after nine months of
continuous pumping. This delay in obtaining low
CABLE SDI values is believed to be due to inadequate
well development. The recorded SDI values of
the beachwell water for the first 15 months
I VALVE
ranged from 0.5 to 5.35%, with a mean value of
EN
1.89%. The SDI values of the pretreated surface
LE IK)MP seawater for the same period ranged from 1.3 to
6.31%, with a mean value of 3.43%. Eliminating
the first four months of operation, the SDI of the
beachwell water ranged from 0.51 to 4.27% with
a mean value of 1.45%. The corresponding pre-
treated surface seawater SDI values for the same
period were 2.9-3.46%, with a mean value of
3.08%. It was noticed that the SDI of the beach-
Fig. 3. Cross-sectionof beachwellseawaterintake system. well water consistently remained about 1% after
reaching this value. This is considered to be the
most valuable merit of the beachwell water as
(ranging from 26 to 28.5°C), whereas the surface RO feedwater.
seawater temperature at DRP fluctuated between
11.7 and 33°C. The average beachwell water
temperature was 27.29°C, which is higher than
4.3. Conclusions
the average temperature of the pretreated surface
seawater for the same period (i.e. 24.27°C). Beachwell seawater intake provided high-
At the start, the beachwell water pH was quality feedwater free from contaminants and at
found to be one unit less than that of the surface a reasonably steady temperature (i.e. 27.3°C
seawater, and ranged from 7.2 to 7.7. Sulfuric average) throughout the study duration. The SDI
acid was added, therefore, to the surface seawater values were always lower than those of the
at the rate of 23 ppm to reduce the pH to appro- conventionally pretreated surface seawater and
ximately 6.8 to achieve optimum operating con- remained consistent at about 1% after it reached
ditions. The average conductivity of the pretreated this value. Moreover, feedwater derived from the
surface seawater was 58,000 lxS/cm, whereas that beachwell did not require pH adjustment with
of beachwell water was 53,6401aS/cm for the recovery of less than 40%, whereas the
same period. conventionally pretreated surface seawater
The dissolved oxygen content of the pretreated required 23 ppm of H2SO4 dosing.
150 S. Ebrahim et aL / Desalination 135 (200]) 141-153

5. Technical and economic comparison H2SO4 and FeCISO4 were dosed into the
between three pretreatment systems seawater stream before entering the buffer tank.
The feed pump delivered seawater to two sand
At DRP, an experimental study was carried
filters connected in parallel where the water was
out to evaluate the performance of three pretreat-
filtered and fed to the RO unit.
ment systems (i.e. conventional, MF, beachwell)
NaHSO3 was dosed to insure the complete
on a comparative basis. Details of this study are
removal of residual chlorine. The SDI is conti-
outlined in [4]. The study used in-line coagulation
nuously monitored. The system has a capacity of
and flocculation followed by pressurized filtration
8mVh.
systems. A schematic diagram of this system is
shown in Fig. 4.
5.1. Performance evaluation
The system consisted of four l m ~ glass-
reinforced plastic (GRP) filter tanks, feed and The in-line pretreatment system was operated
backwash pumps, filtrate tank (pipes and valves), for one year along with the MF and beachwell
three chemical dosing tanks equipped with systems. Performance parameters obtained from
pumps and agitators for H2SO4, FeC1SO4 and the three systems were compared against each
NaHSO3 and four sand filters. other. The temperature of the beachwell water
The buffer tank was fed with seawater from varied between 24.5 and 25°C during winter and
the intake of the Doha East Power Station. summer, respectively, giving almost a steady

I
Acid SBS Anti-Scalant

to RO System

Del. Pump ..............

_.21 12 i

Fig. 4. Schematicdiagram of in-line conventionalsystem.


S. Ebrahim et al. / Desalination 135 (2001) 141-153 151

feed temperature; whereas variation of surface remained almost consistently at about 1.2% after
seawater temperature, fed to the MF and in-line reaching this value.
system, was comparatively large, between 18 The SDI values of the filtrate from the MF
and 32°C for winter and summer, respectively. system were acceptable throughout the operation
The values of the SDI o f the filtrate from the period. The SDI varied between 0.24 and 3%
conventional pretreatment were not steady and and the average SDI value was 2.02%, knowing
varied between 2.5 and 4%. The SDI of the that the SDI of the seawater feed was over 6.5%.
beachwell water remained steadily below 2% Table 3 shows average physical and chemical
after one month of continuous pumping. This parameters values of the three systems.
delay in getting low SDI values within a No significant reduction in the COD and
reasonable time is believed to be due to pumping BOD values were noticed in the filtrate o f the
some sand particles into the system. Recorded conventional pretreatment system, whereas these
SDI values o f beachwell water ranged between values were reduced substantially in the filtrate
0.3 and 2.5%, with an average value of 1.2%. It of the MF system and the beachwell water.
was also noticed that the SDI o f this system

Table 3
Average physical and chemical analysis of the seawater feed and filtrate of the pretreatment systems

Filtrate
Analysis, ppm Seawater feed Conventional Beachwell intake Microfiltration
pH 8.34 7.2 77 8.3
Total alkalinity 145 99 144 144
p-value 193 1.93 0 1.14
TDS 42,555 40,706 42,860 40,928
Na+ 13,493 13,098 14,015 13,224
K÷ 443 440 346 468
Ca+ 515 507 597 468
Mg+ 1,570 1,553 1,451 1,582
A1÷ 1.16 1 0.9 1
F- 1.53 1.5 1.9 1.6
CI- 22,185 22,108 22,294 22,250
SO42- 3,154 3,180 3,154 3,239
Fe2÷ 0.026 0.025 0.025 0.025
Sr2+ 7.41 7 11 7.5
Ba2+ 0.025 0.025 0.025 0.025
Cu2÷ 0.025 0.025 0.025 0.025
HCO3 143 100 143 142
CO2 1.91 13 7.6 2
PO4 0.025 0.025 0.284 0.025
NH3 0.43 0.39 1 0.034
SDI >6.5 3.3 1.2 2.02
Total count, 861 1,861 56.91 x 105 964
colonies/100 ml
COD 400 524 228 221
BOD 6 10 1.7 3
152 S. Ebrahim et al. / Desalination 135 (2001) 141-153

5.2. Economic comparison 5.3. Conclusions


An economic assessment study was carded • Conventional pretreatment of surface seawater
out to assess the three systems. Table 4 shows produces unsteady feeAwater quality and
the unit cost o f the three systems for a capacity quantity, and, therefore, it is considered not
o f 27,276 m3/d [2]. reliable for a RO pretreatment system.
The study revealed that the total unit water • Results indicate that the MF and beachwell
costs produced by beachwell, MF and conven- systems are suitable for pretreating surface
tional system were 11.082, 12.264 and 28.153 seawater feed for RO plants.
fils/m3, respectively (1000 ills = US$3.33). The • Results indicate that the beaehwell system is
beachwell system gives substantial cost savings capable of producing excellent quality water
and is the most cost-effective among the three (SDI less than 2%) suitable as RO feed. The
available techniques for seawater pretreatment. If microfiltration process is also comparable
for some reason the beachwell system is with the beachwell system.
technically not feasible, such as appropriate • Based on actual operational data for one year,
locations in sufficient number are not available, it is concluded that the beachwell system is
the MF system, which is found to be almost the most reliable source of feedwater for RO
equally cost-effective as the beachwell system, is plants, with no standby capacity and no
the next cost-effective system for seawater chemical dosing being required, followed by
pretreatment. the microfiltration system.

Table 4
Comparison of unit cost of seawater pretreatmentby alternative systemsfor 1995 (plant capacity27,276 mVd)

Cost component Unit cost, fils/ms


Conventionalsurface Beachwell Mierofiltration
A Capital cost (depreciation) 10.365 2.917 3.537
Feed pumps 0.119 -- 0.119
Chlorine dosing plant 0.015 -- --
Structure - intake 0.387 0.387 0.387
Structure - outfall 0.387 -- 0.387
Pretreatment plant 9.457 2.530
Beachwell -- -- --
Mierofiltrationplant __ m 2.644

B Operatingcost 17.788 8.165 8.727


Electricity 9.648 6.426 7.090
Chemicals 4.662 -- 1.637
Filters 3.478 1.739 --

Total unit cost (A+B), fils/m3. 28.153 11.082 12.264

*1000 ills = US$3.33


S. Ebrahim et al. / Desalination 135 (2001) 141-153 153

• The techno-economic evaluation revealed that • A conventional surface pretreatment system


the total unit costs produced by beachwell, is not recommended due to its economics and
MF and conventional surface systems were operating drawbacks. It can be considered as
11.082, 12.264 and 28.153 fils/m3, respec- a pretreatment for a seawater RO plant only if
tively. beachwell and MF systems are not feasible.
• The conventional surface seawater treatment • In selecting a seawater pretreatment system
is the highest in cost. In addition, it is the for RO, MF or beachwell should be evaluated
most complicated method for seawater pre- for biological content prior to the final
treatment. Hence, it should not be considered. selection of the system.
• MF and beachwell systems are comparable in
cost (12.264 and 11.082 fils/m3) for small to
medium size RO plants. For large RO plants,
Acknowledgment
MF process is the favorable treatment.
The authors acknowledge the financial
contribution of the Ministry of Electricity and
Water (MEW) and the Kuwait Foundation for
6. Conclusions
the Advancement of Sciences (KFAS). The
• Proper pretreatment is the most critical factor authors wish, also, to acknowledge the assistance
for successful long-term performance of a of their colleagues in the Water Desalination
SWRO. Department (WDD) and the Techno-Economics
• Since 1984, KISR has carried out an intensive Division (TED) at the Kuwait Institute for
R&D program to assess the technical viability Scientific Research (KISR).
and economic feasibility of desalting SWRO
technology with special emphasis on pretreat-
ment of the seawater feed.
References
• The program included evaluation of three
pretreatment techniques: conventional, micro- [1] S.H. Ebrahim, M.M. Abdel-Jawad and M. Safar,
filtration and beachwell. Desalination, 102 (1995) 179.
• In designing new commercial seawater RO [2] S. Ebrahim, S. Bou-Hamed, M. Abdel-Jawad and
desalination plants, a beachwell intake system N. Burne, Desalination, 109 (1997) 165.
should be considered as a first priority for [3] M. Abdel-Jawad and S. Ebrahim, Desalination, 99
small to medium size plants. If this system is (1994) 57.
technically not feasible, a MF system should [4] S. Bou-Hamad, M. Abdd-Jawad, S. Ebrahim, M.
be considered as an alternative pretreatment AI-Mansour and A. AI-Hijji, Desalination, 110
method. (1997) 85.

Das könnte Ihnen auch gefallen