membranes

Article
Seawater Reverse Osmosis Performance Decline Caused by
Short-Term Elevated Feed Water Temperature
Thomas Altmann 1 , Paulus J. Buijs 2,3 , Andreia S. F. Farinha 2 , Vitor R. Proença Borges 2 , Nadia M. Farhat 2, *,
Johannes S. Vrouwenvelder 2 and Ratul Das 1,3, *

1 Innovation and New Technology, ACWA Power, 41st Floor, The One Tower, Sheikh Zayed Road,
Dubai P.O. Box 30582, United Arab Emirates
2 Water Desalination and Reuse Center (WDRC), Biological and Environmental Science and Engineering (BESE)
Division, King Abdullah University of Science and Technology (KAUST), Thuwal 23955-6900, Saudi Arabia
3 KAUST ACWA Power Center of Excellence (KAPCOE), Thuwal 23955-6900, Saudi Arabia
* Correspondence: [email protected] (N.M.F.); [email protected] (R.D.)

Abstract: The shortage of fresh water resources has made the desalination of seawater a widely
adopted technology. Seawater reverse osmosis (SWRO) is the most commonly used method for
desalination. The SWRO process is energy-intensive, and most of the energy in SWRO is spent on
pressurizing the seawater to overcome the osmotic barrier for producing fresh water. The pressure
needed depends on the salinity of the seawater, its temperature, and the membrane surface proper-
ties. Membrane compaction occurs in SWRO due to hydraulic pressure application for long-term
operations and operating temperature fluctuations due to seasonal seawater changes. This study
investigates the effects of short-term feed water temperature increase on the SWRO process in a
full-scale pilot with pretreatment and a SWRO installation consisting of a pressure vessel which
Citation: Altmann, T.; Buijs, P.J.;
contains seven industrial-scale 8” diameter spiral wound membrane elements. A SWRO feed water
Farinha, A.S.F.; Borges, V.R.P.; temperature of 40 ◦ C, even for a short period of 7 days, caused a permanent performance decline
Farhat, N.M.; Vrouwenvelder, J.S.; illustrated by a strong specific energy consumption increase of 7.5%. This study highlights the
Das, R. Seawater Reverse Osmosis need for membrane manufacturer data that account for the water temperature effect on membrane
Performance Decline Caused by performance over a broad temperature range. There is a need to develop new membranes that are
Short-Term Elevated Feed Water more tolerant to temperature fluctuations.
Temperature. Membranes 2022, 12,
792. https://doi.org/10.3390/ Keywords: reverse osmosis; SWRO; specific energy consumption (SEC); membrane compaction;
membranes12080792 membrane permeability
Academic Editors: Fu Liu and
Kaisong Zhang

Received: 23 July 2022

1. Introduction
Accepted: 16 August 2022
Published: 18 August 2022
The importance of desalination with respect to humanity’s ability to produce high-
quality fresh water sustainably and at a low cost cannot be overstated [1]. Two-thirds of
Publisher’s Note: MDPI stays neutral
the global population live under severe water scarcity for at least one month a year [2].
with regard to jurisdictional claims in
With its arid and semi-arid climate and limited freshwater resources, the Middle East relies
published maps and institutional affil-
mainly on desalination for sustenance [3–9]. Desalination is a proven technology that helps
iations.
alleviate the water stress in these regions. Seawater reverse osmosis (SWRO) has gained
the limelight as a promising desalination technology to source fresh water. SWRO holds
roughly a 69% market share among all desalination technologies [10,11]. SWRO requires
Copyright: © 2022 by the authors.
a high energy input to extract fresh water from seawater compared with the production
Licensee MDPI, Basel, Switzerland. from conventional sources such as rivers and lakes [12]. The specific energy consumption
This article is an open access article (SEC) of SWRO is the key parameter characterizing its performance [13,14]. SEC is the
distributed under the terms and energy required in kWh to produce a m3 of product water. It comprises contributions from
conditions of the Creative Commons all sections of the SWRO plant, i.e., the seawater intake, pretreatment, RO desalination
Attribution (CC BY) license (https:// composed of high-pressure pumps, membranes, energy recovery devices, and the product
creativecommons.org/licenses/by/ post-treatment section. The RO section of the SWRO plant contributes between 60 and 80%
4.0/). of the total energy demand [12–14].

Membranes 2022, 12, 792. https://doi.org/10.3390/membranes12080792 https://www.mdpi.com/journal/membranes

Membranes 2022, 12, 792 2 of 14

The energy consumed in the RO section depends primarily on the osmotic barrier
of feedwater, which is determined by its salinity and the hydraulic water permeability of
the membrane. The SEC generally benefits from membranes having higher permeability,
selectivity, and tolerance to fouling [15]. Several studies focus on how SEC can be decreased
by using membranes with a higher permeability [16–21]; a higher permeability allows for a
lower pressure required to achieve the same permeate flux [22]. Membrane deformation or
compaction due to increased hydraulic pressure and changes in feed water temperatures is
observed in SWRO [23–25], which reduces the permeability of the membrane and requires
the need for higher pressures to maintain a constant permeate flux [26]. Compaction is
primarily irreversible [27], and membrane deformation has been verified through mem-
brane autopsies after their use [25]. The fundamental behavior of compaction and its effect
on SWRO performance remains poorly understood. Studies focusing on experimental
compaction verification primarily address membrane deformation due to high hydraulic
pressures [28–31]. Few studies report on the variations of SWRO performance due to
expansion and shrinkage of membrane materials resulting from variations in feed water
temperatures, as is the case for SWRO in the Middle East. The seawater temperatures in
the Middle East vary seasonally due to the arid and semi-arid climate. Temperatures can
reach as high as 38 ◦ C in peak summers and as low as 19 ◦ C in winters. Figure S1 shows
the typical annual variations of seawater temperatures for this region. Global warming
and climate change have exposed the region to evident warming of seawater tempera-
tures. It is estimated that seawater temperatures in these regions can increase anywhere
between 0.02 and 0.07 ◦ C/year [32–34]. The seasonal extremities in temperatures affect
the performance of SWRO. Even pressurization of seawater to higher than 60 bar pressure
increases its temperature due to the inefficiencies of the pumps [35]. It is known that high
temperatures tend to impact membrane salt rejection and scaling negatively. In contrast,
low temperatures require higher pressures (more energy) to achieve the same permeate
water flux [36]. Due to the seasonal highs and lows in seawater temperatures in the region,
the membranes are subjected to thermal stress, which causes their deformation, reducing
permeability and hence affecting the SEC in the process [26]. Hence, a critical question
remains to be answered, “How do variations in seawater temperature affect the performance of
SWRO membranes in the long term?”.
In light of the aforementioned, in the present study, we analyze the performance
of commercially available 8” Polyamide thin-film composite (TFC) spiral wound SWRO
membranes operating in a full-scale SWRO pilot. The pilot test unit consisted of ultrafiltra-
tion and reverse osmosis fed by Red Sea seawater. To simulate the seasonal variations in
seawater temperatures in the Middle East and their effect on the SEC of the SWRO process,
we ran the pilot in recirculation mode and controlled the SWRO inlet seawater temperature.

2. Materials and Methods

2.1. Pilot Plant Details
The pilot plant was designed for seawater desalination research; it was equipped
with state-of-the-art instrumentation and was fully automated through programmable
logic controllers (PLC), which recorded all the signals into a data logger. All sensors and
instruments used in the pilot plant are complied with industrial standards with a maximum
error of ±0.2%. The details of different sections of the pilot are shown in Figure 1 and
listed below:
Seawater intake—It was an open intake in the Red Sea, located 1.2 km from the coast
and at a depth of 10 m. There was a 2 mm mesh at the mouth of the intake pipe to prevent
the suction of larger objects and mussels. There was no chlorination at the intake.
Microstrainer—The seawater from the intake entered the pilot through a micro\strainer; it
consisted of a 250 µm mesh that prevented large objects from interfering with the pilot’s operations.
Seawater intake tank—This was an intermediate tank to collect the seawater before
ultrafiltration (UF).
 the suction of larger objects and mussels. There was no chlorination at the intake.
Microstrainer—The seawater from the intake entered the pilot through a mi-
cro\strainer; it consisted of a 250 µm mesh that prevented large objects from interfering
with the pilot’s operations.
Membranes 2022, 12, 792 Seawater intake tank—This was an intermediate tank to collect the seawater before 3 of 14
ultrafiltration (UF).
UF—Dow Ultrafiltration™ SFP-2880 membranes were used in the UF modules. It
was operated
UF—DowatUltrafiltration™
a flux of 75 LMH for 60 min
SFP-2880 and thenwere
membranes backwashed
used in thewith
UFthe UF permeate
modules. It was
at 100 LMH for 3 min. Chemically enhanced backwash was
operated at a flux of 75 LMH for 60 min and then backwashed with the UF permeate performed once daily. The at
filtrate
100 LMH from
for the UF Chemically
3 min. was stored enhanced
in the UF filtrate
backwash tank before
was being pumped
performed to The
once daily. the SWRO
filtrate
system.
from the UF was stored in the UF filtrate tank before being pumped to the SWRO system.
RO
RO section—Water
section—Water from from the
the UF
UF filtrate
filtrate tanks
tanks went
went through
through aa SWRO
SWRO feed
feed pump
pump and and
then
then a a high-pressure
high-pressure pump pump (HP(HP pump)
pump) to to the
the SWRO
SWRO pressure
pressure vessels.
vessels. It
It was
was aa two-pass
two-pass
system;
system; each
each pass
pass had two 8-
had two 8- and
and one
one 16-inch
16-inch pressure
pressure vessel.
vessel. Permeate
Permeate and brine from
and brine from
each pass could go to a tank for
each pass could go to a tank for storage. storage.
Recirculation tank tankand andcooler—There
cooler—Therewas was a provision
a provision for putting
for putting permeate
permeate and
and brine
brine
from thefrom thepass
first first into
passainto a recirculation
recirculation tank. tank.
This This
couldcould
also also be used
be used as feed
as feed to HP
to the the
HP pump
pump through
through the recirculation
the recirculation tanktank
pump.pump.TheTheflowflow
fromfrom the recirculation
the recirculation tanktank
to theto
the
HP HPpump pumpwent went through
through a cooler
a cooler to control
to control its temperature;
its temperature; it was
it was cooled
cooled using
using a cool-
a coolant,
ant,
and and a constant
a constant temperature
temperature was was maintained
maintained through
through a PLC.
a PLC. FigureFigure 1 shows
1 shows a process
a process flow
flow diagram for
diagram for the pilot. the pilot.

Figure 1. Schematic
Figure 1. Schematic of
of the
the seawater
seawater reverse
reverse osmosis
osmosis pilot
pilot used
used for
for the
the study, operated with
study, operated with seven
seven
8” diameter spiral wound SWRO membrane modules in
8” diameter spiral wound SWRO membrane modules in series.series.

2.2. Operating Conditions

To study the effect ofof seawater
seawater temperature
temperature on the performance of spiral wound
membranes, we we loaded
loaded one
one pressure
pressure vessel with22××400
vesselwith 400ftft 2 and 5 × 4402 ft2 area
SWRO membranes, 2 and 5 × 440 ft area 8”
8” diameter elements. Chlorine-free seawater was fed into the pilot plant. Chlorination has
diameter elements. Chlorine-free seawater was fed into the pilot plant. Chlorination
been shown to cause higher microorganism growth [37]; we avoided chlorination
been shown to cause higher microorganism growth [37]; we avoided chlorination for for that
that
purpose. No other chemicals were dosed after the test started. The pilot plant was operated
in two phases. The first phase was for membrane performance stabilization. The plant was
operated in a once-through mode during this phase at normal seawater conditions without
any chemical dosing. The second phase was aimed at testing the membrane performance
under varying temperatures. The plant was operated in recirculation mode during this
phase. The seawater was heated through the HP pump, and its temperature was controlled
using a water-cooled exchanger. The entire pilot test was conducted at 40% system recovery.
The seawater conditions and the standard operating parameters are listed below.
Membranes 2022, 12, 792 4 of 14

2.3. Seawater Conditions

• Temperature: unregulated, 24–35 ◦ C
• pH—natural, 8.3
• Total dissolved solids (TDS)—natural, approx. 41,500 ppm
Table 1 below lists the elemental composition of the seawater used for the pilot deter-
mined analytically using an Agilent 7500cx Inductively coupled plasma mass spectrometry
(ICP-MS) [38].

Table 1. Elemental composition of seawater used for the pilot test.

Cations Unit Value Anions Unit Value

Calcium Ca++ mg/L 485 Bicarbonate HCO3 − mg/L 153
Magnesium Mg++ mg/L 1649 Chloride Cl− mg/L 22,918
Sodium Na+ mg/L 12,549 Bromide Br− mg/L 78.29
Potassium K+ mg/L 464 Sulfate SO4 −− mg/L 3220
Strontium Sr++ mg/L 5.32 Fluoride F− mg/L 1.34
Barium Ba++ mg/L 0.01 Carbonate CO3 −− mg/L 6.0
Nitrate NO3 − mg/L 1.3
Other parameters
TDS mg/L 41,540 Boron B mg/L 4.9
Conductivity µS/cm 59,800 Silica Si mg/L 1.4
pH - 8.24 TOC C mg/L 3.0
Temperature T ◦C 25 Density ρ g/L 1028

2.4. Operating Parameters

• Feed temperature ~25–40 ◦ C (after HP pump).
• Feed conductivity ~60.0 mS/cm.
• Feed flow ~7.5 m3 /h.
• Recovery—40% constant.
• Permeate flow—3.0 m3 /h constant and no permeate split operation.
• Brine flow—4.5 m3 /h.
• Membrane flux calculated 2 × 400 + 5 × 440 ft2 and 3.0 m3 /h = 10.75 LMH.
• Operation conditions were maintained by brine and feed control valves.
• SWRO inlet temperature during recirculation operation was maintained by circulating
the seawater through a water-cooled exchanger, which was operated automatically
through a PLC.
Seawater temperature, conductivity, oxidation reduction potential (ORP), and RO inlet
temperature after the HP pump were continuously monitored during the test. Parameters
required for membrane performance normalization were recorded in the distributed control
system (DCS) at an interval of 30 s.

2.5. Experiment Conditions

This pilot test was primarily aimed at studying the effects of the seasonal extremities
of seawater temperature on the performance of SWRO. The pilot was operated in a once-
through mode for the first two weeks for membrane performance stabilization. Afterward,
seawater recirculation was started to maintain a constant SWRO inlet temperature. The
temperature was kept at 25 ◦ C for two days for stabilization. Afterward, the SWRO inlet
temperature gradually rose to 40 ◦ C and operated at 40 ◦ C for one week. The extreme
temperatures in the summer season usually last for a week in the Gulf; therefore, the
duration was selected to replicate these conditions. After a week of operation at 40 ◦ C, the
 through mode for the first two weeks for membrane performance stabilization. After-
ward, seawater recirculation was started to maintain a constant SWRO inlet temperature.
The temperature was kept at 25 °C for two days for stabilization. Afterward, the SWRO
inlet temperature gradually rose to 40 °C and operated at 40 °C for one week. The extreme
temperatures in the summer season usually last for a week in the Gulf; therefore, the du-
Membranes 2022, 12, 792 5 of 14
ration was selected to replicate these conditions. After a week of operation at 40 °C, the
temperature gradually decreased to 25 °C, and the steady-state pilot operation at 25 °C
continued under recirculation mode. A summary is shown in Table 2.
temperature gradually decreased to 25 ◦ C, and the steady-state pilot operation at 25 ◦ C
continued
Table 2. Pilotunder recirculation
operating conditionsmode.
during A
thesummary is shown in Table 2.
different phases.

Operation
Table 2. Pilot operating Phase during the different
conditions Water phases. Time
Phase Operation Mode
Description Temperature (°C) (day)
Operation Phase Water
Phase Time (day) Once Mode
Operation Recircula-
Description Temperature (◦ C)
Through tion
Once Through Recirculation
1a Not regulated * 0 to 16 X
1a Stabilization of of
Stabilization membranes
Not regulated * 0 to 16 X
1b membranes
25 17–18 X
1b 25 17–18 X
2 Temperature increase From 25 to 40 19–20 X
Temperature
2 Operation From 25 to 40 19–20 X
increase at high
3 40 20–27 X
temperature
Operation at
3 40 20–27 X
Temperature
high temperature decrease
4 35, 30, and 25 27–30 X
reaching 25 °C
Temperature
4
* natural seawaterdecrease
temperature.
◦
35, 30, and 25 27–30 X
reaching 25 C
* natural seawater temperature.
3. Results
This study looked at the effects of short-term feed water temperature increase on the
3. Results
SWROThis process
study performance in a full-scale
looked at the effects pilot
of short-term containing
feed pretreatment
water temperature andonan
increase theSWRO
SWRO
installation.
process performance in a full-scale pilot containing pretreatment and an SWRO installation.
Figure
Figure22shows
showsthethe pH
pH and
and conductivity
conductivity trends
trends of
of the
the feed,
feed, permeate,
permeate, andandbrine
brine
streams
streams during the pilot run. Figure S2 shows the oxidation reduction potential(ORP)
during the pilot run. Figure S2 shows the oxidation reduction potential (ORP)ofof
seawater.
seawater.These
Thesemeasurements
measurementswere weredonedonewith
withhandheld
handheldinstruments
instrumentstotovalidate
validatedata
data
from
fromonline
onlinesensors.
sensors. The
ThepHpH and
and conductivity
conductivity trends seen in
trends seen in Figure
Figure 22 were
wereasasexpected
expectedinina
afull-scale
full-scaleplant.
plant.

Figure
Figure2.2.pH
pHand
andconductivity
conductivityprofile
profileofofthe
the(A)
(A)seawater
seawaterfeed,
feed,(B)
(B)SWRO
SWROpermeate,
permeate,and
and(C)
(C)SWRO
SWRO
brine
brineduring
duringthe
the30-day
30-daypilot
pilottest
teststudy.
study.

Figure 3 summarizes the entire pilot operation for 30 days. It shows the variation in
normalized flux, feed pressure, and temperature during the pilot operation (a detailed nor-
malization procedure is described in Section S1 of Supplemental Materials). As mentioned
earlier, the pilot was operated in two phases at a constant recovery of 40% (Figure S3).
The first phase was to stabilize membrane performance and test the pilot equipment and
instrumentation. The system stabilization was confirmed by the feed pressure and the
permeate conductivity readings. During the second phase, the pilot was operated in a
recirculation mode under controlled temperatures, which was maintained by the heat of
the HP pump and a water-cooled heat exchanger. The permeate conductivity increased
at a higher temperature (Figure 4); however, when the temperature was lowered to 25 ◦ C,
the conductivity was lower than during the conditioning phases, signaling membrane
structural changes. Figure 5 below shows the operational data (feed pressure and salt
 and instrumentation. The system stabilization was confirmed by the feed pressure and the
permeate conductivity readings. During the second phase, the pilot was operated in a
recirculation mode under controlled temperatures, which was maintained by the heat of
the HP pump and a water-cooled heat exchanger. The permeate conductivity increased at
Membranes 2022, 12, 792 a higher temperature (Figure 4); however, when the temperature was lowered to 25 6 of°C,
14
the conductivity was lower than during the conditioning phases, signaling membrane
structural changes. Figure 5 below shows the operational data (feed pressure and salt re-
jection) during
rejection) duringthe
therecirculation
recirculationoperation;
operation; seawater
seawater feed
feed temperature to the
temperature to the RO
RO varied
varied
between 25◦°C to 40
between 25 C to 40 C. ◦°C.

Membranes 2022, 12, x FOR PEER REVIEW 7 of 15

Figure 3.
Figure 3. Operating
Operating data
data of
of the
the pilot
pilot test,
test, data
data captured
captured from
from the
the distributed
distributed control
control system
system (DCS).
(DCS).
The plots (A) show the variation in normalized flux and (B) feed pressures with changes
The plots (A) show the variation in normalized flux and (B) feed pressures with changes in feed in feed
temperature. The dotted vertical lines indicate the period where the feed water temperature was
temperature. The dotted vertical lines indicate the period where the feed water temperature was
maintained at 40 °C.
maintained at 40 ◦ C.
 Figure 3. Operating data of the pilot test, data captured from the distributed control system (DCS).
Membranes 2022, 12, 792 The plots (A) show the variation in normalized flux and (B) feed pressures with changes in feed 7 of 14
temperature. The dotted vertical lines indicate the period where the feed water temperature was
maintained at 40 °C.

Membranes 2022, 12, x FOR PEER REVIEW 8 of 15

Figure4.
Figure 4. Variations
Variations ininthe permeate
the conductivity
permeate and flux
conductivity andduring the pilot
flux during operation.
the pilot operation.

Figure 5. Pilot operating data during recirculation operation. (A) Variation in feed pressure with
Figure 5. Pilot operating data during recirculation operation. (A) Variation in feed pressure with
temperature
temperature(B) (B)Variation
Variationin
insalt
salt rejection
rejection with
with temperature under constant
temperature under constantrecovery
recoveryofof40%
40%and
and feed
flow 3 /h. 3
feedrate
flowofrate
7.5ofm7.5 m /h.

It can be seen from Figure 5A that a lower pressure was required to maintain a con-
stant recovery at an elevated temperature. Permeate flux increased as the feed tempera-
ture increased. The viscosity of seawater decreased as temperature increased, and the wa-
ter permeation rate through the membrane increased. Permeate flux typically increased
 Membranes 2022, 12, 792 8 of 14

It can be seen from Figure 5A that a lower pressure was required to maintain a con-
stant recovery at an elevated temperature. Permeate flux increased as the feed temperature
increased. The viscosity of seawater decreased as temperature increased, and the water
permeation rate through the membrane increased. Permeate flux typically increased with
temperature linearly with viscosity [39–41]. When the RO feed temperature increased from
25 ◦ C to 40 ◦ C, the feed pressure reduced from 56.5 bar to 54.8 bar. On the contrary, when
the RO feed temperature decreased from 40 ◦ C to 25 ◦ C, the feed pressure increased from
56.7 bar to 60.25 bar. The additional pressure buildup during the reduction in operating
temperature resulted from operation at higher water temperature, possibly causing com-
paction of the membrane surface. Higher temperatures also increased the solubility of the
solute, and a higher diffusion rate of the solute through the membrane was possible [41],
causing a reduction in salt rejection by the membrane, as observed in Figure 5B. Figure 5B
shows that when the RO feed temperature increased from 25 ◦ C to 40 ◦ C, salt rejection
decreased. However, when the feed water temperature returned to 25 ◦ C, we saw an
Membranes 2022, 12, x FOR PEER REVIEW 9 of
increase in salt rejection by 0.2%. This improvement in salt rejection possibly results from
membrane surface compaction, but it comes at the cost of additional energy expenditure,
as discussed earlier. Salt rejection is calculated as follows:
 𝜎 
𝑆𝑅(%%) = 11− σpermeate ∗ 100
SR 𝜎 ∗ 100 (1)
σseawater
where σ denotes electrical conductivity.
where σ denotes electrical conductivity.
Figure6 below
Figure 6 below presents
presents the normalized
the normalized operation
operation data fordata for the
the pilot pilot (a nor-
(a detailed detailed no
malization
malization procedure
procedure is described
is described in Section
in Section S1 of Supplementary
S1 of Supplementary Materials).
Materials).

Figure6.6.Normalized
Figure Normalized operation
operation datadata forpilot.
for the the pilot. NPF denotes
NPF denotes the normalized
the normalized permeatepermeate
flow; NDP,flow; ND
thenormalized
the normalized differential
differential pressure;
pressure; andthe
and NSP, NSP, the normalized
normalized saltEach
salt passage. passage. Each line to
line corresponds correspon
to the
the axisaxis
withwith similar
similar color. color.

The normalized data presented clearly shows a significant decrease in NPF, a redu
tion in NSP, and a constant increase in NDP. The continuous increase in normalized d
ferential pressure can be attributed to the deposits on the membrane surface during t
pilot run. Inductively coupled plasma mass spectrometry (ICP-MS, Agilent 7500cx) an
ysis was performed to check the major elemental composition of deposits on the mem
brane and feed spacer at the end of the 30-day study when the membrane element w
autopsied. The deposits contained primarily Iron, Nickel, Magnesium, and Mangane
(Table 3). A detailed elemental composition of the deposits on the feed spacer and mem
brane surface can be found in Supplementary Materials (Table S3 and Section S2).
Membranes 2022, 12, 792 9 of 14

The normalized data presented clearly shows a significant decrease in NPF, a reduction
in NSP, and a constant increase in NDP. The continuous increase in normalized differential
pressure can be attributed to the deposits on the membrane surface during the pilot run.
Inductively coupled plasma mass spectrometry (ICP-MS, Agilent 7500cx) analysis was
performed to check the major elemental composition of deposits on the membrane and
feed spacer at the end of the 30-day study when the membrane element was autopsied.
The deposits contained primarily Iron, Nickel, Magnesium, and Manganese (Table 3). A
detailed elemental composition of the deposits on the feed spacer and membrane surface
can be found in Supplementary Materials (Table S3 and Section S2).

Table 3. The feed spacer and membrane surface coupon major elemental composition from the
membrane autopsy.

Element Value (mg/m2 )

Fe 191.1
Mg 17.6
Ni 13.7
Mn 1.4

From Figure 6, it can be seen that there was a 0.43 bar increase in normalized differential
pressure (NDP), corresponding to a 2.5% decrease in normalized permeate flow (NPF)
during the once-through operation mode. There was a subsequent increase of 0.3 bar in
NDP and a corresponding 13.9% decrease in NPF during the recirculation operation mode,
which raised the temperatures to 40 ◦ C. Hence, the increase in NDP was certainly not the
reason behind the steep decrease in NPF, which increased the pressure required to maintain
the recovery rate. There was also a reduction of 0.2% in normalized salt passage (NSP) post
recirculation operation, which points toward a possible membrane surface compaction.
The changes in NPF and NSP are with reference to the initial conditions at the startup
of the experiments; hence, they represent the actual change in the parameter rather than
measurement errors.
The changes that occurred to the membrane surface due to high-temperature operation
directly impacted the specific energy consumption (SEC) of the RO process (Figure 7).
Additional energy (higher pressure) was required to maintain recovery ratios (constant
permeate flux), as shown in Figure 7. When the RO inlet temperature was increased from
25 ◦ C to 40 ◦ C, a decrease of 2.5% in the SEC of the process was seen (Figure 7). The
SEC decrease was due to an increase in the water permeation rate through the membrane
resulting from a reduction in viscosity of the RO feed at higher inlet temperatures. However,
when the feed temperature returned to 25 ◦ C, we saw an increase in SEC of the process
by 7.5%. It was expected that as the viscosity of the fluid returned to its original value
of 25 ◦ C, the SEC would increase back by the same amount of 2.5%, which was not
the case. The additional energy required to maintain production was attributed to the
possible compaction of the membrane surface, which resulted from surface deformations
of the porous support due to thermal stress [42–44]. Figure 8 shows a scanning electron
microscopy image of the used membrane after operation, where the thickness of the
polyamide layer was observed to be 28 µm. This same virgin membrane had a polyamide
layer thickness of 40 µm, as reported by [45], which was also confirmed by the membrane
OEM. Compaction is primarily irreversible, and these structural changes permanently
impact the energy consumption of the RO process [27].
 support due to thermal stress [42-44]. Figure 8 shows a scanning electron microscopy im-
age of the used membrane after operation, where the thickness of the polyamide layer was
observed to be 28 µm. This same virgin membrane had a polyamide layer thickness of 40
µm, as reported by [45], which was also confirmed by the membrane OEM. Compaction
Membranes 2022, 12, 792
is primarily irreversible, and these structural changes permanently impact the energy10con-
of 14
sumption of the RO process [27].

Membranes 2022, 12, x FOR PEER REVIEW 11 of 15

Figure 7.
Figure Variations in
7. Variations in specific
specific energy
energy consumption
consumption (SEC)
(SEC) during
during the
the pilot
pilot test.
test.

Figure 8. Scanning electron microscopy (SEM) image of the used membrane after operation.
Figure 8. Scanning electron microscopy (SEM) image of the used membrane after operation.

4.4.Discussion
Discussion
Structuralchanges
Structural changes
in in thin-film
thin-film composite
composite (TFC)(TFC) polyamide
polyamide RO membranes
RO membranes com-
combined
bined
with with changes
changes in solvent
in solvent viscosityviscosity anddiffusivity
and solute solute diffusivity
governgovern the relationship
the relationship betweenbe-
ROtween
inletRO inlet temperature,
temperature, mass transfer,
mass transfer, membranemembrane transport,
transport, and SECandof SEC
theofRO
theprocess.
RO pro-
cess. Considering
Considering only changes
only changes in and
in solute solute and solvent
solvent propertiesproperties will misconstrue
will misconstrue the im-
the impact of
ROpact of RO
inlet inlet temperature
temperature on the performance
on the performance of RO membranes,
of RO membranes, which has which
beenhas been dis-
discussed
cussed theoretically
theoretically by researchers
by researchers [41,46].
[41,46]. The dataThe
fromdata
thefrom
pilot the
testpilot test suggested
suggested that the
that the perfor-
performance
mance of TFC polyamide
of TFC polyamide RO membranes
RO membranes is sensitive
is sensitive to changes,
to changes, even temporarily,
even temporarily, in the
ROin inlet
the RO inlet temperature.
temperature. Membrane Membrane manufacturers
manufacturers need tofor
need to account account for the structural
the structural changes
changes in the membranes due to thermal stress and provide flux data for the modules
over a wide temperature range up to 40 °C.
The results presented in this study underscore the importance of seasonal variations
in the temperature of seawater and their potential impact on the SEC of a RO process
(Figure 9). In a water-stressed region such as the Middle East, SWRO is widely adopted
Membranes 2022, 12, 792 11 of 14

in the membranes due to thermal stress and provide flux data for the modules over a wide
temperature range up to 40 ◦ C.
The results presented in this study underscore the importance of seasonal variations
in the temperature of seawater and their potential impact on the SEC of a RO process
(Figure 9). In a water-stressed region such as the Middle East, SWRO is widely adopted
as a low-cost and sustainable clean, fresh water source for human use [3,4,8,9]. There-
fore, a rigorous methodology considering all parameters that affect the SWRO process
efficiency and its SEC is essential to optimize the final water tariff. The energy consumed
by the RO section of an SWRO plant can account for ~40% of the total water cost of a
desalination plant [12,47]. Hence, considering changes in the membrane structure is of
utmost importance in accounting for its effect on the pressure required to achieve a specific
permeate flux. Most desalination project contracts in the Middle East are offered under a
BOO (Build, Own, Operate) structure, with a long concession period (~20–25 years) [8,48].
The contracts have technical guidelines to be met to achieve final water quality, and they
also define the water tariff for the entire concession period. To avoid unforeseen changes
in water cost during the concession period due to changes in energy consumption by the
Membranes 2022, 12,RO process,
x FOR it is of paramount importance that all parameters that account for changes in
PEER REVIEW 13 of
feed pressure required to achieve a specific permeate flux are considered during project
design, including the changes in membrane structure due to seasonal variations of seawater
temperature. Concerted computational investigations are needed to develop models to
ing their operation. However, such studies are beyond the scope of this work. The deve
account for increased energy requirements in RO due to membrane structural changes
opment of newer RO membranes with materials that can resist the effects of such therm
during their operation. However, such studies are beyond the scope of this work. The
compaction at elevated temperatures up to 40 °C is also needed [42,47,48].
development of newer RO membranes with materials that can resist the effects of such
thermal compaction at elevated temperatures up to 40 ◦ C is also needed [42,47,48].

Figure 9. Average specific

Figure energyspecific
9. Average consumption (SEC) before,(SEC)
energy consumption during, and after
before, short-term
during, and after(7-day)
short-term (7-da
RO-feed
RO-feed seawater seawater
temperature temperature
increase. increase.
The change The change
in SEC can alsoinbeSEC can alsowith
correlated be correlated
the changewith
in the chang
in NPF,
NPF, as discussed as discussed
in Figure 6. in Figure 6.

5. Conclusions
To summarize, this study provided an experimental verification for the impact
temperature variations on the energy consumption of an SWRO process. It provided in
sights into rationally designing new SWRO projects. The study was conducted at a ful
Membranes 2022, 12, 792 12 of 14

5. Conclusions
To summarize, this study provided an experimental verification for the impact of
temperature variations on the energy consumption of an SWRO process. It provided
insights into rationally designing new SWRO projects. The study was conducted at a
full-scale pilot plant consisting of a seawater intake point, microstrainer, ultrafiltration,
and an SWRO installation consisting of a pressure vessel containing seven industrial-scale
800 diameter spiral wound membrane elements.
Results showed that:
• A SWRO feed water temperature of 40 ◦ C, even during a short period of 7 days,
caused a permanent performance decline, as illustrated by a strong specific energy
consumption (SEC) increase of 7.5%.
• A 7.5% increase in SEC, depending on the plant size, translates into an additional
operating cost of USD 250,000 a year for a 60,000 m3 /day production capacity plant to
USD 2.5 M a year for a 600,000 m3 /day capacity plant [47].
• There are financial consequences, in addition to contractual implications, for the use of
additional energy. Since the energy required by the plant is defined during project de-
velopment, it may require payment of an inflated tariff for the additional consumption.
The authors conclude with the hope that this study heralds coordinated efforts from
the membrane manufacturers to provide insights into the changes in membrane structures
due to temperature variation during RO operation, and that their subsequent effect on
energy consumption is accounted for in membrane projection programs.

Supplementary Materials: The following supporting information can be downloaded at: https://
www.mdpi.com/article/10.3390/membranes12080792/s1, Figure S1: Average daily temperature for
the year 2018–2019 for seawater at a full-scale desalination plant in the Arabian Gulf; Figure S2: The
oxidation reduction potential (ORP) profile of the seawater feed used for the pilot test. ORP was
constant, indicating no membrane damage due to chlorination; Figure S3: Pilot operation at a constant
recovery of 40%. Table S1: EC-TDS conversion coefficients; Table S2: Correction Factors used for
normalization; Table S3: Elemental composition of the sticky brown deposit on the feed spacer and
membrane surface. Refs. [41,49–54] are cited on supplementary materials
Author Contributions: Conceptualization, R.D.; Data curation, A.S.F.F. and R.D.; Formal analysis,
A.S.F.F., N.M.F. and R.D.; Investigation, T.A. and R.D.; Methodology, T.A., P.J.B., N.M.F. and R.D.;
Project administration, P.J.B.; Resources, V.R.P.B.; Supervision, T.A.; Validation, N.M.F., J.S.V. and
R.D.; Writing—original draft, R.D.; Writing—review & editing, T.A., P.J.B., A.S.F.F., V.R.P.B., N.M.F.,
J.S.V. and R.D. All authors have read and agreed to the published version of the manuscript.
Funding: The authors thank ACWA Power and KAUST Research Translation Funding (REI/1/4455-01-0).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Acknowledgments: We thank Muhammad Altaf, Abdulaali T. Alharbi, and Achraf Boussira for their
assistance in the pilot operation. Figure 1 was produced by Ana Bigio, Scientific Illustrator at KAUST.
Conflicts of Interest: The authors declare that they have no known competing financial interests or
personal relationships that could have appeared to influence the work reported in this paper.

References
1. Gude, V.G.; Nirmalakhandan, N.; Deng, S. Desalination using solar energy: Towards sustainability. Energy 2011, 36, 78–85.
[CrossRef]
2. Mekonnen, M.M.; Hoekstra, A.Y. Four billion people facing severe water scarcity. Sci. Adv. 2016, 2, e1500323. [CrossRef] [PubMed]
3. Berktay, A. Environmental approach and influence of red tide to desalination process in the Middle East region. Int. J. Chem.
Environ. Eng. 2011, 2.
4. Ghaffour, N.; Missimer, T.M.; Amy, G.L. Combined desalination, water reuse, and aquifer storage and recovery to meet water
supply demands in the GCC/MENA region. Desalination Water Treat. 2013, 51, 38–43. [CrossRef]
Membranes 2022, 12, 792 13 of 14

5. Tahir, F.; Baloch, A.A.; Ali, H. Resilience of Desalination Plants for Sustainable Water Supply in Middle East. In Sustainability
Perspectives: Science, Policy and Practice; Springer: Berlin/Heidelberg, Germany, 2020; pp. 303–329.
6. Das, R.; Arunachalam, S.; Ahmad, Z.; Manalastas, E.; Mishra, H. Bio-inspired gas-entrapping membranes (GEMs) derived from
common water-wet materials for green desalination. J. Membr. Sci. 2019, 117185. [CrossRef]
7. Das, R.; Arunachalam, S.; Ahmad, Z.; Manalastas, E.; Syed, A.; Buttner, U.; Mishra, H. Proof-of-Concept for Gas-Entrapping
Membranes Derived from Water-Loving SiO2 /Si/SiO2 Wafers for Green Desalination. JoVE 2020, e60583. [CrossRef]
8. Ghaffour, N.; Missimer, T.M.; Amy, G.L. Technical review and evaluation of the economics of water desalination: Current and
future challenges for better water supply sustainability. Desalination 2013, 309, 197–207. [CrossRef]
9. Ghaffour, N. The challenge of capacity-building strategies and perspectives for desalination for sustainable water use in MENA.
Desalination Water Treat. 2009, 5, 48–53. [CrossRef]
10. Altmann, T.; Das, R. Process improvement of sea water reverse osmosis (SWRO) and subsequent decarbonization. Desalination
2021, 499, 114791. [CrossRef]
11. Jones, E.; Qadir, M.; van Vliet, M.T.H.; Smakhtin, V.; Kang, S.M. The state of desalination and brine production: A global outlook.
Sci. Total Environ. 2019, 657, 1343–1356. [CrossRef]
12. Voutchkov, N. Energy use for membrane seawater desalination – current status and trends. Desalination 2018, 431, 2–14. [CrossRef]
13. Yousry, A.; Ridwan, M.G.; Altmann, T.; Rousseva, A.; Azab, K.; Das, R. Performance Model for Reverse Osmosis. Chem. Eng. Res.
Des. 2022, in press. [CrossRef]
14. Karabelas, A.; Koutsou, C.; Kostoglou, M.; Sioutopoulos, D. Analysis of specific energy consumption in reverse osmosis
desalination processes. Desalination 2018, 431, 15–21. [CrossRef]
15. Pillai, S.; Santana, A.; Das, R.; Shrestha, B.R.; Manalastas, E.; Mishra, H. A molecular to macro level assessment of direct contact
membrane distillation for separating organics from water. J. Membr. Sci. 2020, 608, 118140. [CrossRef]
16. Busch, M.; Mickols, W.E. Reducing energy consumption in seawater desalination. Desalination 2004, 165, 299–312. [CrossRef]
17. Zhu, A.; Rahardianto, A.; Christofides, P.D.; Cohen, Y. Reverse osmosis desalination with high permeability membranes—cost
optimization and research needs. Desalination Water Treat. 2010, 15, 256–266. [CrossRef]
18. Wilf, M. Design consequences of recent improvements in membrane performance. Desalination 1997, 113, 157–163. [CrossRef]
19. Bartels, C.R.; Andes, K. Consideration of energy savings in SWRO. Desalination Water Treat. 2013, 51, 717–725. [CrossRef]
20. Zhu, A.; Christofides, P.D.; Cohen, Y. Minimization of energy consumption for a two-pass membrane desalination: Effect of
energy recovery, membrane rejection and retentate recycling. J. Membr. Sci. 2009, 339, 126–137. [CrossRef]
21. Mishra, H.; Arunachalam, S.N.M.; Domingues, E.M.; Das, R. Perfluorocarbon-free membranes for membrane distillation. U.S.
Patent Application No. 17/056,809, 2021.
22. Okamoto, Y.; Lienhard, J.H. How RO membrane permeability and other performance factors affect process cost and energy use:
A review. Desalination 2019, 470, 114064. [CrossRef]
23. Mehdizadeh, H.; Dickson, J.M.; Eriksson, P.K. Temperature effects on the performance of thin-film composite, aromatic polyamide
membranes. Ind. Eng. Chem. Res. 1989, 28, 814–824. [CrossRef]
24. Kimura, S. Analysis of reverse osmosis membrane behaviors in a long-term verification test. Desalination 1995, 100, 77–84.
[CrossRef]
25. Gonzalez-Gil, G.; Behzad, A.R.; Farinha, A.S.; Zhao, C.; Bucs, S.S.; Buijs, P.; Vrouwenvelder, J. Clinical Autopsy of a Reverse
Osmosis Membrane Module. Front. Chem. Eng. 2021, 3, 23. [CrossRef]
26. Davenport, D.M.; Ritt, C.L.; Verbeke, R.; Dickmann, M.; Egger, W.; Vankelecom, I.F.J.; Elimelech, M. Thin film composite
membrane compaction in high-pressure reverse osmosis. J. Membr. Sci. 2020, 610, 118268. [CrossRef]
27. Peterson, R.; Greenberg, A.; Bond, L.; Krantz, W. Use of ultrasonic TDR for real-time noninvasive measurement of compressive
strain during membrane compaction. Desalination 1998, 116, 115–122. [CrossRef]
28. McGovern, R.K.; McConnon, D.; Lienhard, J.H. The effect of very high hydraulic pressure on the permeability and salt rejection
of reverse osmosis membranes. In Proceedings of the 2015 IDA World Congress on Desalination and Water Reuse, San Diego, CA,
USA, 4 September 2015.
29. Davenport, D.M.; Deshmukh, A.; Werber, J.R.; Elimelech, M. High-Pressure Reverse Osmosis for Energy-Efficient Hypersaline
Brine Desalination: Current Status, Design Considerations, and Research Needs. Environ. Sci. Technol. Lett. 2018, 5, 467–475.
[CrossRef]
30. Alexiadis, A.; Wiley, D.E.; Vishnoi, A.; Lee, R.H.K.; Fletcher, D.F.; Bao, J. CFD modelling of reverse osmosis membrane flow and
validation with experimental results. Desalination 2007, 217, 242–250. [CrossRef]
31. Ladner, D.A.; Subramani, A.; Kumar, M.; Adham, S.S.; Clark, M.M. Bench-scale evaluation of seawater desalination by reverse
osmosis. Desalination 2010, 250, 490–499. [CrossRef]
32. Hereher, M.E. Assessment of Climate Change Impacts on Sea Surface Temperatures and Sea Level Rise—The Arabian Gulf.
Climate 2020, 8, 50. [CrossRef]
33. Nesterov, O.; Temimi, M.; Fonseca, R.; Nelli, N.R.; Addad, Y.; Bosc, E.; Abida, R. Validation and statistical analysis of the Group
for High Resolution Sea Surface Temperature data in the Arabian Gulf. Oceanologia 2021, 63, 497–515. [CrossRef]
34. Alawad, K.A.; Al-Subhi, A.M.; Alsaafani, M.A.; Alraddadi, T.M. Decadal variability and recent summer warming amplification of
the sea surface temperature in the Red Sea. PLOS ONE 2020, 15, e0237436. [CrossRef]
Membranes 2022, 12, 792 14 of 14

35. Cattaert, A. High pressure pump efficiency determination from temperature and pressure measurements. In Proceedings of the
2007 IEEE Power Engineering Society Conference and Exposition in Africa-PowerAfrica, Johannesburg, South Africa, 16–20 July
2007; pp. 1–8.
36. Koutsou, C.P.; Kritikos, E.; Karabelas, A.J.; Kostoglou, M. Analysis of temperature effects on the specific energy consumption in
reverse osmosis desalination processes. Desalination 2020, 476, 114213. [CrossRef]
37. Saeed, M.O.; Jamaluddin, A.; Tisan, I.; Lawrence, D.; Al-Amri, M.; Chida, K. Biofouling in a seawater reverse osmosis plant on the
Red Sea coast, Saudi Arabia. Desalination 2000, 128, 177–190. [CrossRef]
38. Strickland, J.D.H.; Parsons, T.R. A Practical Handbook of Seawater Analysis; Fisheries Research Board of Canada: Ottawa, ON,
Canada, 1972.
39. Goosen, M.F.A.; Sablani, S.S.; Al-Maskari, S.S.; Al-Belushi, R.H.; Wilf, M. Effect of feed temperature on permeate flux and mass
transfer coefficient in spiral-wound reverse osmosis systems. Desalination 2002, 144, 367–372. [CrossRef]
40. Water, D.; Solutions, P. FILMTEC™ Reverse Osmosis Membranes Technical Manual. 2013. Available online: http://msdssearch.
dow.com/PublishedLiteratureDOWCOM/dh_08db/0901b803808db77d.pdf (accessed on 22 July 2022).
41. Jin, X.; Jawor, A.; Kim, S.; Hoek, E.M.V. Effects of feed water temperature on separation performance and organic fouling of
brackish water RO membranes. Desalination 2009, 239, 346–359. [CrossRef]
42. Aghajani, M.; Wang, M.; Cox, L.M.; Killgore, J.P.; Greenberg, A.R.; Ding, Y. Influence of support-layer deformation on the intrinsic
resistance of thin film composite membranes. J. Membr. Sci. 2018, 567, 49–57. [CrossRef]
43. Handge, U.A. Analysis of compaction and life-time prediction of porous polymer membranes: Influence of morphology, diffusion
and creep behaviour. Polym. Int. 2017, 66, 521–531. [CrossRef]
44. Kitano, H.; Takeuchi, K.; Ortiz-Medina, J.; Ito, I.; Morelos-Gomez, A.; Cruz-Silva, R.; Yokokawa, T.; Terrones, M.; Yamaguchi, A.;
Hayashi, T.; et al. Enhanced desalination performance in compacted carbon-based reverse osmosis membranes. Nanoscale Adv.
2020, 2, 3444–3451. [CrossRef]
45. Sharabati, J.-A.-D.; Guclu, S.; Erkoc-Ilter, S.; Koseoglu-Imer, D.Y.; Unal, S.; Menceloglu, Y.Z.; Ozturk, I.; Koyuncu, I. Interfacially
polymerized thin-film composite membranes: Impact of support layer pore size on active layer polymerization and seawater
desalination performance. Sep. Purif. Technol. 2019, 212, 438–448. [CrossRef]
46. Lakner, G.; Lakner, J.; Racz, G.; Kłos, M. Temperature dependence modelling of reverse osmosis. Desalin. Water Treat 2020,
192, 431–436. [CrossRef]
47. Desaldata, I. The IDA Water Security Handbook 2020–2021; Media Analytics Ltd.: Oxford, UK, 2020; pp. 1–41.
48. Wolfs, M.; Woodroffe, S. Structuring and financing international BOO/BOT desalination projects. J. Struct. Financ. 2002, 7, 19–24.
[CrossRef]
49. Safar, M.; Jafar, M.; Abdel-Jawad, M.; Bou-Hamad, S. Standardization of RO membrane performance. Desalination 1998, 118, 13–21.
[CrossRef]
50. Zhao, Y.; Taylor, J.S. Assessment of ASTM D 4516 for evaluation of reverse osmosis membrane performance. Desalination 2005,
180, 231–244. [CrossRef]
51. Toray Industries, I. TorayTrak. Available online: https://www.water.toray/knowledge/tool/trak/ (accessed on 7 July 2022).
52. Ruiz-García, A.; Nuez, I. Long-term performance decline in a brackish water reverse osmosis desalination plant. Predictive model
for the water permeability coefficient. Desalination 2016, 397, 101–107. [CrossRef]
53. Farhat, N.M.; Vrouwenvelder, J.S.; Van Loosdrecht, M.C.M.; Bucs, S.S.; Staal, M. Effect of water temperature on biofouling
development in reverse osmosis membrane systems. Water Res. 2016, 103, 149–159. [CrossRef]
54. Jawor, A.; Hoek, E.M.V. Effects of feed water temperature on inorganic fouling of brackish water RO membranes. Desalination
2009, 235, 44–57. [CrossRef]