membranes

Article
Reverse Osmosis Treatment of Wastewater for Reuse as Process
Water—A Case Study
Marjana Simonič

Faculty of Chemistry and Chemical Engineering, University of Maribor, 2000 Maribor, Slovenia;
[email protected]

Abstract: The aim of this work was to purify mixed wastewater from three different production
processes in such a manner that they could be reused as process water. The maximum allowed
concentrations (MAC) from the Environmental Standards for emissions of substances released
into surface water were set as target concentrations. Wastewaters contained solid particles, sodium,
aluminium, chloride, and nitrogen in high amounts. Quantitatively, most wastewaters were generated
in the production line of alumina washing. The second type of wastewater was generated from the
production line of boehmite. The third type of wastewater was from regeneration of ion exchangers,
which are applied for feed boiler water treatment. The initial treatment step of wastewater mixture
was neutralisation, using 35% HCl. The precoat filtration followed, and the level of suspended
solids was reduced from 320 mg/L to only 9 mg/L. The concentrations of ions, such as aluminium,
sodium and chlorides remained above the MAC. Therefore, laboratory reverse osmosis was applied
to remove the listed pollutants from the water. We succeeded in removal of all the pollutants. The
concentration of aluminium decreased below 3 mg/L, the sodium to 145 mg/L and chlorides to



193 mg/L. The concentration of nitrate nitrogen decreased below 20 mg/L.



Citation: Simonič, M. Reverse Keywords: alumina production; reverse osmosis; fouling; wastewater reuse
Osmosis Treatment of Wastewater for
Reuse as Process Water—A Case
Study. Membranes 2021, 11, 976.
https://doi.org/10.3390/ 1. Introduction
membranes11120976
Alumina (Al2 O3 ) is considered a basic catalytic material support due to its good
mechanical properties, such as high strength, chemical and physical stability, additionally,
Academic Editor: Asuncion
its high thermal resistance, and thermal conductivity [1]. Boehmite (AlOOH) is used as the
Maria Hidalgo
raw material for the preparation of alpha and gamma-alumina phases whose properties
such as morphology, specific surface area, and porosity, depend strongly on the boehmite
Received: 18 November 2021
Accepted: 9 December 2021
structure [2].
Published: 10 December 2021
The disposal of washing water into the environment is not allowed without treat-
ment, because some parameters, such as Al, Na and Cl, exceed the maximum allowed
Publisher’s Note: MDPI stays neutral
values [1]. Water impurities concentrate during boiler operation. Boiler water must be
with regard to jurisdictional claims in
softened properly before use. However, without periodic water removal (blowdown),
published maps and institutional affil- problems such as scale deposits, corrosion and embrittlement may occur [3]. A study was
iations. reported, conducted on the utilisation of a simulated boiler blowdown for incorporation
into cement-based materials. The results indicate that the use of waste brine in cemented
backfill applications is feasible. The phosphate addition could result in the formation of
deposits such as iron phosphate. Phosphate corrosion is assumed as a significant concern
Copyright: © 2021 by the author.
in phosphate treated steam boilers [4]. Basic feedwater treatment involves ion exchange.
Licensee MDPI, Basel, Switzerland.
For existing industries operating at a low scale, resin-assisted separation continues to offer
This article is an open access article
an attractive option [5]. After regeneration of an ion exchanger, acidic and alkaline effluents
distributed under the terms and are generated, containing chlorides, sulphates, nitrates, silicates, etc.
conditions of the Creative Commons Reuse of wastewater is recognised in most water-scarce countries [6]. Instead of
Attribution (CC BY) license (https:// discharging industrial effluents into rivers and streams, reverse osmosis (RO) membranes
creativecommons.org/licenses/by/ can be used to treat the wastewater and reuse it as process water in companies. Spiral
4.0/). wound modules (SWMs) are most widely adopted among the commercially available RO

Membranes 2021, 11, 976. https://doi.org/10.3390/membranes11120976 https://www.mdpi.com/journal/membranes

Membranes 2021, 11, 976 2 of 12

membrane modules [7]. The analysis of the design of spiral wound modules and correlation
of experimental and model values can help to identify the module geometry and spacer
design for specific applications [8]. The permeability coefficients with the feed spacer design
have a relevant impact on the performance of RO membranes in terms of production,
permeate quality and specific energy consumption [9]. This is due to the relationship
between the feed spacer geometry and pressure drop and concentration polarisation
phenomena. The operation of an RO system deals with the impact of fouling, [10] which
is the main concern of this technology and along with optimal operational conditions the
main thing responsible for losing performance and efficiency in a long term-operation [11].
Spiral wound modules offer ease of operation, fouling control, and a high permeation rate
and packing density. Applications of spiral wound modules include desalination, water
treatment, water reclamation and treatment of industrial wastewater. A water recycling
system has been documented using RO technology for the treatment of oleochemical
wastewater [12]. The fouled RO membrane required a high operating pressure in order
to obtain a consistent permeate flow. In general, RO membranes are subjected to surface
fouling and scaling, which can pose a significant problem when reverse osmosis is used [13].
Scaling is one of the limiting factors for increasing the flux recovery rate and improving
the efficiency of the process [14]. It is well known that filtration prior to RO is essential for
maintaining efficiency and protecting the membranes‘ functions. The use of RO membranes
largely reduces the chemical consumption when membrane separation is coupled with
chemical treatment, due to the capability of the RO membranes of removing dissolved
solids [12]. The application of the RO process compared to other conventional thermal
technologies for desalination of brackish water has increased remarkably, due to its high
purification efficiency at low cost, and low energy consumption [15].
The main objective of the present study was to investigate the possibilities of recycling
mixed wastewater from three different processes: The first after alumina washing, the
second after boehmite production and a third from the regeneration of ion-exchangers
for feed boiler water treatment. The chemical analyses of individual wastewaters did
not differ much, therefore, wastewater from all three lines was collected in a feed tank,
neutralised, and filtered using precoat filters. This was followed by RO treatment using
two selected membranes, FILMEC XLE (Dow) and ESPA (Nitto), chosen from among the
major membrane manufacturers. Treatment was tested in two membrane modes: With
and without concentrate recycling. The quality of the RO permeate after both modes was
compared with the Slovenian Regulations for water discharge into the environment. The
organic content was determined, expressed as chemical oxygen demand (COD).

2. Materials and Methods

Alumina washing waters are collected in a feed reservoir. The alumina is firstly
washed in an HCl acid solution and then filtered. The solid particles and the solution are
separated. The solution contains large amounts of inorganic compounds, and must be
treated before it is released into the environment. The second wastewater is generated after
boehmite washing. Boehmite is aluminium oxide hydroxide that forms from Al(OH)3 . The
lattice spacing in the material is 0.117 nm. The structure [2,16] as well as hydrophobicity,
light transmittance, thermal stability, and mechanical durability [17] has been characterised
extensively in the literature. Wastewater flows into the decanter, and it is then mixed
with the washing water from the alumina. The third type of wastewater is generated after
regeneration of an ion exchanger for boiler water treatment.
The amount of water from alumina washing is 72,000 m3 /y, from boehmite 14,500 m3 /y
and from boiler water 14,400 m3 /y. In such ratios (5:1:1) wastewater was mixed in the
100 m3 feed tank in the laboratory.
The feed solution after precoat perlite filtration flowed from the tank through the
valve and pump to a Culligan Aqua-Cleer RO system. The system was equipped with
volume counters (2) and pressure gauges (3), as seen from Figure 1. The system allows
transmembrane pressure (TMP) up to 15 bar. The sample flows through the 80 µm filter
 The feed solution after precoat perlite filtration flowed from the tank through the
valve and pump to a Culligan Aqua-Cleer RO system. The system was equipped with
volume counters (2) and pressure gauges (3), as seen from Figure 1. The system allows
Membranes 2021, 11, 976
transmembrane pressure (TMP) up to 15 bar. The sample flows through the 80 µm 3filter of 12
and 5 µm filter for removal of the bigger particles. Such pre-treated solution is gathered
in a feed tank (4). The permeate and retentate were withdrawn continuously from the
system in the initial four experiments. In the final experiment the permeate and
and 5 µm filter
concentrate forrecycled
were removalinto
of the
thebigger particles.
feed tank. The Such pre-treated
capacity was 1.5solution
L/h for is
thegathered
permeate in
a feed
and 1.9tank
L/h (4). Theretentate
for the permeateatand
roomretentate were withdrawn continuously from the system
temperature.
in theThe
initial
feedfour experiments.
samples In the final
and permeate wereexperiment
taken at thethe
endpermeate andand
of the trial concentrate were
subsequently
recycled into the feed tank. The capacity was 1.5 L/h for the permeate
analysed. The wastewater flux was measured after every repeated trial with mixed and 1.9 L/h for the
retentate
wastewater. at room temperature.

Figure
Figure 1.
1. Schematic
Schematic RO
RO process.
process.

The feed samplesXLE

Dow-FILMTEC and(DuPont-Filmtec,
permeate were taken at the endGermany)
Wiesbaden, of the trialand
andESPA-2521
subsequently anal-
(Nitto,
ysed. The wastewater
Nottinghamshire, UK)flux
ROwas measuredwere
membranes after used
everywith
repeated trial with mixed
the properties listed wastewater.
in Table 1.
Dow-FILMTEC
The membrane XLE
module (DuPont-Filmtec,
dimensions were 0.287Wiesbaden, Germany)
m in length and 0.117andmESPA-2521 (Nitto,
in diameter.
Nottinghamshire, UK) RO membranes were used with the properties listed in Table 1. The
membrane
Table 1. The module dimensions
RO membranes’ were 0.287 m in length and 0.117 m in diameter.
characteristics.

Parameter FILMTEC XLE-2521

Table 1. The RO membranes’ characteristics.
ESPA-2521
Producer DuPont-Filmtec (Germany) Hydranautics, Nitto (USA)
Parameter
Pmax FILMTEC XLE-2521
41 bar ESPA-2521
22 bar
Tmax
Producer 45 °C
DuPont-Filmtec (Germany) 45 Nitto
Hydranautics, °C (UK)
Pmax pH 41 bar 2–11 22 2–10
bar
Tmax 45 ◦wound
Spiral C composite 45 ◦ C composite
Spiral wound
Material/Type
pH 2–11 2–10
Material/Type polyamide
Spiral wound composite polyamide Spiral wound polyamide
composite polyamide

The analyses of the feed wastewater and permeate were performed according to ISO
The analyses
Standards in threeofreplicates.
the feed wastewater and permeate
The Standard weresummarised
methods are performed according
in Table 2.to The
ISO
Standards in three replicates. The Standard methods are summarised in Table 2.
analyses were chosen in accordance with the Slovene legislation for wastewater emission The
analyses were chosen in accordance with the Slovene legislation for wastewater emission
into the environment [18]. Important parameters among the general parameters are
into the environment [18]. Important parameters among the general parameters are sus-
suspended solids mass concentration and pH. Absorbance at 436 nm was measured as an
pended solids mass concentration and pH. Absorbance at 436 nm was measured as an
indication of inorganic contamination. The sum of all organic compounds in the water
indication of inorganic contamination. The sum of all organic compounds in the water
samples was determined as COD. Inorganic species are measured as chloride, sulphate and
nitrate ions using Ion chromatography (IC). Metals (Na, Cu, Cr, Ti, Zn, Ni, Mg), including
Al, were determined by inductively coupled plasma and mass detector (ICP-MS). SiO2
was determined spectrophotometrically. Suspended solids (SS) were analysed with an
Imhoff funnel.
Membranes 2021, 11, 976 4 of 12

Table 2. The methods used for wastewater chemical analyses.

Parameter Standard Method Apparatus

T (◦ C) ISO 10523 Thermometer
pH ISO 10523 pH-meter, MA 5740
A (436 nm) SIST EN ISO 7887 Spectrophotometer Carry 100
κ (mS/cm) EN 27888 Conductivity-meter
COD (g/L O2 ) ISO 6060 Digestion, Titration
SS ISO 38409-H9-2 Imhoff funnel
Metals ISO 17294-2 ICP-MS Agilent 7700x
Ions (Cl, SO4 , NO3 ) ISO 10304-1 IC Metrohm IC 761
Phosphorus ISO 6878 Spectrophotometer Carry 100
Silicium dioxide SM4500-SiO2C Spectrophotometer Carry 100

2.1. Calculations
Rejection R was determined according to Equation (1):

R = (cf − cp )/cf (1)

Concentration factor f c was determined according to Equation (2):

f c = 1/(1 − q) (2)

where
cf = concentration in the feed solution (mg/L/)
cp = concentration in the permeate solution (mg/L/)
q = water flow (L/h)
One of the most common methods of determination of fouling is the membrane
filtration index (MFI) [19]. Cake layer formation is proposed as the dominant mechanism.
The MFI test is performed by filtration of water through a 0.45 um filter with constant
pressure in dead-end mode. The MFI can be calculated as seen from Equation (3) [19]:

t µ Rm µ I
= + V (3)
V TMP.A 2 TMP A2
By plotting t/V versus V (permeate volume) the MFI is defined as the slope of a
straight line after the initial linear section. TMP represents transmembrane pressure (bar),
A is the membrane area (m2 ), Rm is the membrane resistance (m−1 ), µ is viscosity (Pa. s)
and t is the filtration time (s). The higher the fouling potential for a given solution, the
higher will be the MFI value.

2.2. SEM Imaging

The washing solutions after membrane cleaning were examined with an FEI, SIRION-
400 field emission scanning electron microscope (FESEM).

3. Results and Discussion

3.1. Physico-Chemical Analyses
After filtration of all samples, the parameters listed in Table 3 were measured in the
wastewater samples. Wastewater was taken from the washing process line 10 times and
mixed. Such average sample was analysed for all parameters according to the Slovenian
Regulations. Washing water in alumina production is denoted in Table 3 as WW1, washing
water in boehmite production is denoted in Table 3 as WW2, and water after regenera-
tion during feed boiling water condition as WW3. The results are gathered in Table 3.
Al exceeded 3 mg/L in WW1. The suspended solids’ mass concentration was too high
in the washing water in boehmite production (WW2), sodium was above the MAC and
aluminium was very high, measured at 33 mg/L. In WW3 phosphorus was just above the
MAC and the pH value was too alkaline. Absorbance at 436 nm exceeded the MAC for all
Membranes 2021, 11, 976 5 of 12

samples. MAC represents the maximum allowed concentration according to the Slovenian
Regulations [18]

Table 3. Measured chemical parameters after alumina washing.

MAC c (WW1) c (WW2) c (WW3)

Parameter
(mg/L) (mg/L) (mg/L) (mg/L)
Al 3 8.41 33.03 1.13
SiO2 250 0.74 2.86 40.1
Na 200 152 245 166
Cu 0.5 0.01 0.01 0.01
Cr 0.5 0.08 0.006 0.003
Ti 1 0.001 0.001 0.001
Fe 2 <0.1 <0.1 0.17
Zn 2 <0.01 0.02 0.14
Ni 0.5 <0.001 <0.001 0.003
Mg 10 1.7 0.1 0.3
N-NH3 10 21.1 0.06 0.2
N-NO3 20 11 9.2 9.3
SO4 2000 22.7 15.6 15.7
PO4 1 <0.05 0.08 1.03
Cl 250 21 16 203
TOC 30 8 9 7
SS 3 9 11 0.1

3.2. Pretreatment with Precoat Filtration

The mixture of wastewater streams was neutralised to pH = 7 using HCl 35% (weight
percents). An aliquot of 45 L of the mixture was vacuum filtered through 2 cm of perlite
layer (200 g). The emphasis was on suspended solids‘ removal. The removal was 96%,
which means that the measured concentration decreased down to 11 mg/L. The results
are in good agreement with the reported study showing a 95% reduction in turbidity [20].
The concentration was acceptable for further treatment. Settleable solids decreased below
0.1 mg/L and were not problematic. Among metals Al was measured, and the removal was
64%, which was still above 3 mg/L. Other parameters from Table 3 were below the MAC.
During the second trial the same aliquot of mixture was filtered through 350 g of a
2.8 cm perlite layer. The results were similar compared with the first trial, but the filtration
time was prolonged by some 15 min. The suspended solids’ removal was 97%, which
means that the measured concentration decreased down to 9 mg/L. It was acceptable for
further treatment. Settleable solids decreased below 0.1 mg/L, and were not problematic.
Among metals Al was measured, and the removal was 66%, which was still above 3 mg/L.
Other parameters were below the MAC.
According to the results and the literature [21] it could be concluded that increasing
the amount of precoat did not improve the regeneration result.

3.3. Filtration with a Reverse Osmosis Membrane

The linear plot for permeability with Millipore water was determined using both
RO membranes, FILMTEC XLE-2521 (XLE) and ESPA-2521 (ESPA), as shown in Figure 2.
Similar values were also obtained in another study [22]. The permeability of both mem-
branes was followed in the range from 6 bar to 10 bar. A small deviation was observed
at 7 bars with both membranes, indicating that the highest flux was reached at 7 bar and
10 bar TMP. The permeate flux remained very similar to that of the trial at 10 bars with
concentrate withdrawal. The wastewater fluxes were determined at 111 L/(m2 .h) for XLE
and 112 L/(m2 .h) for ESPA at 10 bar. At other TMP, the flux was slightly lower compared
to that at TMP = 7 bar and TMP = 10 bar. Therefore, 7 bar and 10 bar TMP were chosen for
further trials with both membranes. The optimal TMP between 7 and 8 was reported in the
literature [22,23].
 observed at 7 bars with both membranes, indicating that the highest flux was reached at
7 bar and 10 bar TMP. The permeate flux remained very similar to that of the trial at 10
bars with concentrate withdrawal. The wastewater fluxes were determined at 111 L/(m2.h)
for XLE and 112 L/(m2.h) for ESPA at 10 bar. At other TMP, the flux was slightly lower
Membranes 2021, 11, 976 compared to that at TMP = 7 bar and TMP = 10 bar. Therefore, 7 bar and 10 bar TMP6were
of 12
chosen for further trials with both membranes. The optimal TMP between 7 and 8 was
reported in the literature [22,23].

Figure 2. Flux in dependence of TMP for both membranes.

The fluxes of mixed wastewater are shown in Figure 3 for for both
both membranes
membranes at at two
two
different TMP; 7 bar and 10 bar were chosen according to the millipore water
different TMP; 7 bar and 10 bar were chosen according to the millipore water flux and the flux and
the literature
literature [22].[22].
TheThe fluxes
fluxes of of both
both membranes
membranes were
were thesame
the sameatat77bar,
bar, and
and the
the curves
curves
overlapped each other. Slightly higher fluxes were achieved at 10
overlapped each other. Slightly higher fluxes were achieved at 10 bar with bar with ESPA compared
ESPA
with the XLE
compared with membrane. The same trends
the XLE membrane. were trends
The same observedwerewith the timewith
observed dependence:
the time
During
Membranes 2021, 11, x FOR PEER REVIEW 30 min
dependence: the flux
During decreased,
30 min the fluxand then it stabilised
decreased, 25 L/(m2at.h)25for
and then itatstabilised L/(mXLE 7and
2.h) of 13
for
27 L/(m 2 .h) for ESPA at 10 bar.
XLE and 27 L/(m .h) for ESPA at 10 bar.
2

In Figure 3, a relatively uniform flux was observed from 50 min to 200 min of RO
operation.
40 Also, it was very important that, after 200 min, there was no need for cleaning
for both membranes. The membranes with a higher negative surface charge and greater
35
hydrophilicity are less prone to fouling [12]. The ESPA membrane is more hydrophilic, as
seen from
30 Figure 2, and it can be expected that fouling mechanisms would affect the ESPA
membrane25 to a lesser extent than the XLE. This claim was confirmed, as the flux of mixed
J (L/m2h)

process water was higher when using ESPA compared with XLE at 10 bars (Figure 3). The
20 in agreement with Koo [12].
results are
15
10
5
0
0 50 100 150 200 250 300 350 400
t (min)

XLE 7 bar XLE 10 bar ESPA 10 bar ESPA 7 bar

Figure3.3.Flux
Figure Fluxdependent
dependenton
ontime
timefor
forboth
bothmembranes.
membranes.

3.3.1.
InChemical
Figure 3, Analyses after
a relatively RO
uniform flux was observed from 50 min to 200 min of RO
operation.
AfterAlso, it was
perlite very important
filtration, wastewater that,
from after 200streams
three min, there
waswas
mixedno in
need
the for cleaning
volume ratio
for both membranes. The membranes with a higher negative surface charge
5:1:1 from alumina washing, boehmite washing and wastewater after regeneration of and greater
hydrophilicity are less
boiler feed water, prone to fouling
respectively. The ratio[12].
is The
basedESPA membrane
on actual is moregeneration
wastewater hydrophilic, in as
the
seen from Figure
company. Such a2,mixture
and it can be expected
represented thethat
feedfouling
for an mechanisms wouldanalysis
RO. The chemical affect the ESPA
is shown
membrane
in Table 4.to a lesser extent than the XLE. This claim was confirmed, as the flux of mixed
process water was higher when using ESPA compared with XLE at 10 bars (Figure 3). The
results are
Table 4. in agreement
Measured with
chemical Koo [12].in the wastewater mixture (mixWW).
parameters

MAC c (mixWW)
Parameter
(mg/L) (mg/L)
Al 3 9.02
SiO2 250 15
Membranes 2021, 11, 976 7 of 12

3.3.1. Chemical Analyses after RO

After perlite filtration, wastewater from three streams was mixed in the volume ratio
5:1:1 from alumina washing, boehmite washing and wastewater after regeneration of boiler
feed water, respectively. The ratio is based on actual wastewater generation in the company.
Such a mixture represented the feed for an RO. The chemical analysis is shown in Table 4.

Table 4. Measured chemical parameters in the wastewater mixture (mixWW).

MAC c (mixWW)
Parameter
(mg/L) (mg/L)
Al 3 9.02
SiO2 250 15
Na 200 229
Cu 0.5 0.01
Cr 0.5 0.08
Ti 1 0.001
Fe 2 <0.1
Zn 2 <0.01
Ni 0.5 <0.001
Mg 10 1.6
N-NH3 10 1.1
N-NO3 20 10
SO4 2000 24
PO4 1 <0.05
Cl 250 255
TOC 30 8
SS 3 2

The concentrations of heavy metals were not problematic, because the measured
values of all heavy metals were below the MAC, as seen from Table 4. Suspended solids
decreased below 3 mg/L. Exceeded were concentrations of Al, Na and Cl. Organic com-
pounds were below 30 mg/L C. Thus, in further experiments, the focus was directed
towards measurement of the mentioned parameters. The XLE membrane was applied at
TMP 7 bar. All metal concentrations remained below the MAC. Suspended solids decreased
to 1.2 mg/L. Na decreased from 245 to 11 mg/L which is above 95%. In the production
process Al2 O3 decreases with the increase of alkali concentration [24], therefore, the con-
centration of Na should be low in the permeate. The measured values were acceptable.
Similarly, Cl decreased to 13 mg/L, which also means 95% removal. Al decreased below
MAC (3 mg/L) to 1 mg/L, which means 88% efficiency. Although the ammonia concentra-
tion was not problematic, it could be further reduced by lowering pH value below 6 [25].
The study shows ammonnia removal at 99.8%.
In the next experiment ESPA was applied at TMP = 7. The results were very similar
to those of the XLE membrane: Na and Cl decreased by 96% and Al by 91%. Suspended
solids decreased to the same value of 1.2 mg/L. The results of NaCl rejection correlated
well with the reported values, around 98% for the XLE, and a little higher for the ESPA
membrane at 8 bar [23].
In the third experiment the XLE membrane was tested at TMP 10 bar. Very similar
results were obtained as at 7 bar. Only the efficiencies were a little lower compared with
7 bar for Al, Na and Cl, up to 90%.
In the fourth experiment the ESPA membrane was tested at TMP 10 bar. Very similar
results were obtained as at 7 bar. Only the efficiencies were a little lower compared with
7 bar, for Al, Na and Cl, up to 91%.
The next experiments were performed with reverse osmosis in concentration mode
with both membranes. The concentrate from the Aqua Cleer system was returned to the
feed solution. The concentration factor was calculated at 5. The main advantage of such
treatment is decreasing of retentate quantity production and waste streams‘ minimisation.
 In the fourth experiment the ESPA membrane was tested at TMP 10 bar. Very similar
results were obtained as at 7 bar. Only the efficiencies were a little lower compared with
7 bar, for Al, Na and Cl, up to 91%.
The next experiments were performed with reverse osmosis in concentration mode
Membranes 2021, 11, 976 with both membranes. The concentrate from the Aqua Cleer system was returned to the
8 of 12
feed solution. The concentration factor was calculated at 5. The main advantage of such
treatment is decreasing of retentate quantity production and waste streams` minimisation.
The results using ESPA are presented in Figure 4. The mass concentration of Na and Cl
The results using ESPA are presented in Figure 4. The mass concentration of Na and Cl
ions increased due to the concentration of the feed. After the treatment, the concentrations
ions increased due to the concentration of the feed. After the treatment, the concentrations
of both ions decreased below the MAC. The concentration of Na was decreased from 1678
of both ions decreased below the MAC. The concentration of Na was decreased from
mg/l in the feed to 145 mg/l below the MAC of 200 mg/l. Due to neutralisation using HCl
1678 mg/l in the feed to 145 mg/l below the MAC of 200 mg/l. Due to neutralisation
prior
usingtoHCl
precoat
priorfiltration
to precoatand concentration
filtration of the RO feed,
and concentration theRO
of the concentration of chlorides
feed, the concentration
in
of chlorides in the RO feed tank was measured at 2376 mg/l. The concentrationtoof193
the RO feed tank was measured at 2376 mg/l. The concentration of Cl decreased Cl
mg/l, well below
decreased to 193 the MAC.
mg/l, well below the MAC.

Figure 4.
Figure Massconcentration
4. Mass concentration in
in feed
feed (denoted
(denoted 1)
1) and
and permeate
permeate (denoted2).
(denoted2).

Suspended solids
Suspended solids were
were removed
removed below
below 22 mg/L.
mg/L.Al
Alalso
alsoremained
remainedbelow
below33mg/L.
mg/L.
The chemical analysis of water treated with the XLE membrane was a little worse. The
The chemical analysis of water treated with the XLE membrane was a little worse.
membrane started to foul after concentration factor 3, therefore, the initial concentrations
The membrane started to foul after concentration factor 3, therefore, the initial
were lower compared with the experiment using the ESPA membrane. The concentrations
concentrations were lower compared with the experiment using the ESPA membrane. The
of Na, and Cl decreased below the MAC, but Al remained above 3 mg/L. From the
concentrations of Na, and Cl decreased below the MAC, but Al remained above 3 mg/L.
experiments we can conclude that the ESPA membrane is more appropriate for wastewater
treatment and water reuse.

3.3.2. Membrane Fouling Studies

The MFI was determined according to the results of Equation (3). The MFI was
determined for untreated wastewater and pre-treated wastewater. The formation of a cake
layer was confirmed, due to the much lower fouling index after the pre-coat filtration prior
to RO treatment. The MFI of untreated water was determined at 5.7 s/L2, and that of
pre-treated water with a precoat filter decreased to 1.9 s/L2 , as seen from Figure 5. The
results are in agreement with suspended solids‘ removal as discussed in Section 3.2.

3.3.3. Membrane Cleaning

In wastewater after alumina washing, different concentrations of Al, SIO2 , Cl, and Na
ions are still present, which contribute to the inorganic fouling on the membrane‘s surface.
Sodium phosphate species react with the deposited scales [4]. Fouling was more alleviated
using the ESPA membrane, due to intensification of the inorganic fouling. The results of
flux in Figure 4 show that the flux of XLE was lower than the ESPA membrane, which
was probably due to stronger interactions of the inorganic compounds with the membrane
compared to the ESPA.
 The MFI was determined according to the results of Equation (3). The MFI was
determined for untreated wastewater and pre-treated wastewater. The formation of a cake
layer was confirmed, due to the much lower fouling index after the pre-coat filtration prior
to RO treatment. The MFI of untreated water was determined at 5.7 s/L2, and that of pre-
Membranes 2021, 11, 976 9 of 12
treated water with a precoat filter decreased to 1.9 s/L2, as seen from Figure 5. The results
are in agreement with suspended solids` removal as discussed in Section 3.2.

MFIfor
Figure5.5.MFI
Figure for1.1.Untreated
Untreatedmixed
mixedwastewater,
wastewater,2.2.Pre-treated
Pre-treatedmixed
mixedwastewater.
wastewater.

After 200 min

3.3.3. Membrane of working the membrane fouled (not shown in paper), because there
Cleaning
was no flux and cleaning was necessary. Biofouling was controlled by UV lamp in accor-
In wastewater after alumina washing, different concentrations of Al, SIO2, Cl, and Na
dance with the literature [26]. The cleaning agents were among a wide variety of chemicals,
ions are still present, which contribute to the inorganic fouling on the membrane`s surface.
including acids (HCl, HNO3 , and H2 SO4 ), a base (NaOH), a complexing agent (EDTA), a
Sodium
surfactantphosphate
(SDS), and species
theirreact with the deposited
combination. To achieve scales
high[4]. Fouling
cleaning was more
efficiency, thealleviated
effects of
using
Membranes 2021, 11, x FOR PEER REVIEW the ESPA membrane, due to intensification of the inorganic
physical factors were studied (velocity, temperature, and time). The result showed fouling. The results
10 of 13of
that
flux
the intwoFigure
stages,4 show thatand
caustic the flux of XLEcleaning
detergent was lower than theNaOH-SDS
including ESPA membrane,followedwhichby was
acid,
probably
provideddue to stronger
an effective interactions
recovery. Since weof could
the inorganic compounds
not disassemble with the(we
the module membrane
had only
compared
thisCleaning to
one), afterthe
with ESPA.
HCl probably
performance of removed the Ca
the cleaning, theand Mg salts,
effluent which
solution are deposits
with soluble infromacids.the
This After 200
finding
membrane inmin
issurface
goodof
wasworking
agreement thewith
observed membrane
whatSEM
through hadfouled
been (not shown
observed
imaging. From in paper),
from
Figure the because
6 itwater
can there
recycling
be seen that
was
plant
small noparticles
in flux and
another cleaning
study,
were whereby
washed was necessary.
NaOH
out Biofouling
was found
in a range from to was
be more
below 1 µmcontrolled
effective
up to 20inµm.by UV lamp
recovering
Based onthe in
the
accordance
flux than HCl
chemical with the literature [26]. The cleaning agents were among a
[12]. it was assumed that some aggregates of Al and SiO2 were formed, as
analyses wide variety of
chemicals,
well as Naincluding
or Ca saltsacids
such(HCl, HNO3, and H2SO4), a base (NaOH), a complexing agent
as sulphates.
(EDTA), a surfactant (SDS), and their combination. To achieve high cleaning efficiency,
the effects of physical factors were studied (velocity, temperature, and time). The result
showed that the two stages, caustic and detergent cleaning including NaOH-SDS
followed by acid, provided an effective recovery. Since we could not disassemble the
module (we had only this one), after performance of the cleaning, the effluent solution
with deposits from the membrane surface was observed through SEM imaging. From
Figure 6 it can be seen that small particles were washed out in a range from below 1 µm
up to 20 µm. Based on the chemical analyses it was assumed that some aggregates of Al
and SiO2 were formed, as well as Na or Ca salts such as sulphates.

Figure 6. SEM image of deposit from the membrane after cleaning (line: 10 µm).
Figure 6. SEM image of deposit from the membrane after cleaning (line: 10 µm).

The cake layer forms by deposition of material on the membrane`s surface rather than
by penetration, in accordance with the SEM image.
Membranes 2021, 11, 976 10 of 12

Cleaning with HCl probably removed the Ca and Mg salts, which are soluble in acids.
This finding is in good agreement with what had been observed from the water recycling
plant in another study, whereby NaOH was found to be more effective in recovering the
flux than HCl [12].
The cake layer forms by deposition of material on the membrane‘s surface rather than
by penetration, in accordance with the SEM image.

3.3.4. Process Scheme

The process scheme was designed based on the results. Figure 7 shows the scheme of
the process water treatment. The water from alumina scrubbing, wastewater generated after
boehmite washing and the effluents after regeneration of an ion exchanger for boiler water
treatment, were gathered in a collection tank with the volume 100 m3 . After neutralisation
with 35% HCl the pre-coat filtration was performed using a pressure leaf (Kelly) filter.
A powerful centrifugal pump is required for dosing the perlite precoat. The pre-treated
Membranes 2021, 11, x FOR PEER REVIEW 11 of 13
water then flows on to the RO plant. The permeate is collected in a 60 m3 tank. After the
treatment, the water can be reused as process water in the production process of alumina
and boehmite.

Figure 7. Schematic process scheme for wastewater treatment.

In7.Table
Figure 5 theprocess
Schematic costs are evaluated,
scheme based treatment.
for wastewater on the company’s data (a) and the litera-
ture [27]. According to the results 85% of water could be reused. The cost would decrease
due toInwater
Tablesavings andare
5 the costs reduced wastewater
evaluated, based discharge by 82,000data
on the company’s Eur per year.
(a) and The
the annual
literature
[27]. According to the results 85% of water could be reused. The cost would decrease The
cost was calculated to be 46,630 Eur (Table 5), based on 10 years of depreciation. due
payback
to water period
savingsofand
less reduced
than one wastewater
year indicates a high return
discharge on investment
by 82,000 for the
Eur per year. proposed
The annual
RO wastewater
cost treatment
was calculated to be plant.
46,630 Eur (Table 5), based on 10 years of depreciation. The
payback period of less than one year indicates a high return on investment for the
Table 5. Cost estimation.
proposed RO wastewater treatment plant.
Total Cost
Table 5. Cost estimation. Annual Costs/10 y Depreciation (Eur)
(Eur)
Cost of plant [27] 180,000 Total Cost 18,000 Costs/10 y
Annual
Energy a 7730 (Eur) 7730
Depreciation (Eur)
Chemicals [27] 3900 3900
Costa of plant [27]
Labour 11,000 180,000 11,000 18,000
Energya a
Wastewater discharge 6000 7730 6000 7730
Chemicals [27]
Total 208,630 3900 46,630 3900
a Compay’s data.
Labour a 11,000 11,000
Wastewater discharge a 6000 6000
Total 208,630 46,630
a Compay’s data.

4. Conclusions
The study shows that the wastewaters could be treated in order to get quality for
water reuse. The results confirmed that three flows could be collected, neutralised and
Membranes 2021, 11, 976 11 of 12

4. Conclusions
The study shows that the wastewaters could be treated in order to get quality for water
reuse. The results confirmed that three flows could be collected, neutralised and treated
further. After the mixture of differet wastewater types the water could be reused. The
sodium content reduced by 95%, chloride from 255 down to 193 mg/L, and nitrogen below
10 mg/L. Metals such as iron and aluminium decreased below detection limits. Also, the
TOC was measured below 10 mg/L. For RO the most suitable membrane was Hydranautics
ESPA. It was calculated that 7 membrane modules with an area of 37.1 m2 each would
satisfy the needs for treatment of 250 m3 /d. The process scheme showed that we could
reduce the water consumption effectively at 85% by the proposed wastewater treatment.

Author Contributions: Conceptualization, M.S.; methodology, M.S.; investigation, M.S.; writing—

original draft preparation, M.S.; writing—review and editing, M.S. The author has read and agreed
to the published version of the manuscript.
Funding: This research was funded by the Slovenian Research Agency and Support of Research,
Process Systems Engineering and Sustainable Development, Programme P2-0032.
Data Availability Statement: Data supporting the reported results can be found here: https://dk.
um.si/Dokument.php?id=145897 (accessed on 2 December 2021).
Acknowledgments: The author would like to thank the funders of the Slovenian Research Agency,
Process Systems Engineering and Sustainable Development, Programme P2-0032.
Conflicts of Interest: The author declares no conflict of interest.

References
1. El-Shafie, M.; Kambara, S.; Hayakawa, Y. Alumina particle size effect on H2 production from ammonia decomposition by DBD
plasma. Energy Rep. 2020, 6, 25–30. [CrossRef]
2. Mohammadi, M.; Khodamorady, M.; Tahmasbi, B.; Bahrami, K.; Ghorbani-Choghamarani, A. Boehmite nanoparticles as versatile
support for organic–inorganic hybrid materials: Synthesis, functionalization, and applications in eco-friendly catalysis. J. Ind.
Eng. Chem. 2021, 97, 1–78. [CrossRef]
3. Afroughsabet, V.; Zhao, P.; Peterson, K.; Shvarzman, A.; Gitarts, A. Evaluation of mortar produced with boiler blowdown brine.
Constr. Build. Mater. 2021, 278, 122459. [CrossRef]
4. Pourabdollah, K. Fouling formation and under deposit corrosion of boiler firetubes. J. Environ. Chem. Eng. 2021, 9, 104552.
[CrossRef]
5. Taqvi, S.A.A.; Sohail, M.; Uddin, F. Utilization of Ion-Exchange Technology for Boiler Feed Water Production-Design and Testing.
Chem. Eng. 2016, 1, 26–35.
6. Manikandan, S.; Subbaiya, R.; Saravanan, M.; Ponraj, M.; Selvam, M.; Pugazhendhi, A. A critical review of advanced nanotech-
nology and hybrid membrane based water recycling, reuse, and wastewater treatment processes. Chemosphere 2021, 132867.
[CrossRef]
7. Gu, B.; Xu, X.X.; Adjiman, C.S. A predictive model for spiral wound reverse oomosis modules: The effect of winding geometry
and accurate geometric details. Comput. Chem. Eng. 2017, 96, 248–265. [CrossRef]
8. Schwinge, J.; Neal, P.R.; Wiley, D.E.; Fletcher, D.F.; Fane, A.G. Spiral wound modules and spacers: Review and analysis. J. Membr.
Sci. 2004, 242, 129–153. [CrossRef]
9. Ruiz-Garcia, A.; Pestana, I.N. Feed spacer Geometries and Permeability Coefficients. Effect on the Performance in BWRO
Spiral-Wound Membrane Modules. Water 2019, 11, 152. [CrossRef]
10. AlSawaftah, N.; Abuwatfa, W.; Darwish, N.; Husseini, G. A Comprehensive Review on Membrane Fouling: Mathematical
Modelling, Prediction, Diagnosis, and Mitigation. Water 2021, 13, 1327. [CrossRef]
11. Cai, Y.H.; Galili, N.; Gelman, M.; Gilron, J. Evaluating the impact of pretreatment process on fouling of reverse osmosis membrane
by secondary wastewater. J. Membr. Sci. 2021, 623, 119054. [CrossRef]
12. Koo, C.H.; Mohammad, A.W.; Suja, F. Recycling of oleochemical wastewater for boiler feed water using rverse osmosis
membranes—A case study. Desalination 2011, 271, 178–186. [CrossRef]
13. Amma, L.V.; Ashraf, F. Brine Management in Reverse osmosis Desalination: A UAE Perspective. In Proceedings of the Advances
in Science and Engineering Technology, Dubai, United Arab Emirates, 4 February–9 April 2020. [CrossRef]
14. Anis, S.F.; Hashaikeh, R.; Hilal, N. Reverse osmosis pretreatment technologies and future trends: A comprehensive review.
Desalination 2019, 452, 159–195. [CrossRef]
15. Al-Karaghouli, A.; Kazmerski, L.L. Energy consumption and water production of conventional and renewable-energy-powered
desalination processes. Renew. Sustain. Energy Rev. 2013, 24, 343–356. [CrossRef]
Membranes 2021, 11, 976 12 of 12

16. Wefers, K.; Misra, C. Oxides and Hydroxides of Aluminum. V: Alcoa Technical Paper; Alcoa Laboratories: Pittsburgh, PA, USA, 1987.
17. Zhang, C.; Cheng, L.; Zhang, M.; Long, Z.; Meng, F.; Lin, H. Robust and durable transparent superhydrophobic boehmite
(γ-AlOOH) film by a simple hydrothermal method. Ceram. Int. 2021, 47, 11694–11701. [CrossRef]
18. Regulation, Regulation on the Emission of Substances and Heat from the Discharge of Wastewater from Municipal Wa-
tertreatment Plant, Official Gazette of Republic of Slovenia, No 45/07,63/09, 105/2010. 2007. Available online: https:
//leap.unep.org/countries/si/national-legislation/regulation-emission-substances-and-heat-discharge-waste-water (accessed
on 2 December 2021).
19. Kim, D.H.; Lee, C.; Nguyen, T.T.; Adha, R.S.; Kim, C.; Ahn, S.J.; Son, H.; Kim, S. Insight into fouling potential analysis of a
pilot-scale pressure-assisted forward osmosis plan for diluted seawater reverse osmosis desalination. J. Ind. Eng. Chem. 2021, 98,
237–246. [CrossRef]
20. Rani, S.R.S.; Kumar, V. Insight on appplications of low-cost ceramic membranes in wastewater treatment: A mini-review. Case
Stud. Chem. Environ. Eng. 2021, 4, 100149. [CrossRef]
21. Maxim, L.D.; Niebo, R.; McConnell, E.E. Perlite toxicology and epidemiology—A review. Inhal. Toxicol. 2014, 26, 259–270.
[CrossRef]
22. Madaeni, S.S.; Koocheki, S. Application of Taguchi method in the optimization of wastewater treatment using spiral-wound
reverse osmosis element. Chem. Eng. J. 2006, 119, 37–44. [CrossRef]
23. Kosutic, K.; Dolar, D.; Kunst, B. On experimental parameters characterizing the reverse osmosis and nanofiltration membranes’
active layer. J. Membr. Sci. 2006, 282, 109–114. [CrossRef]
24. Wang, Y.; Zhang, T.; Lyu, G.; Ma, L.; Zhang, W. Multi-material circulation optimization of the calcification-carbonation process
based on material balance and phase transformation for cleaner production of alumina. J. Clean. Prod. 2021, 290, 125828.
[CrossRef]
25. Shin, C.; Szcuka, A.; Jiang, R.; Mitch, W.A.; Criddle, C.S. Optimization of reverse osmosis operational conditions too maximize
ammonia removal from the effluent of an anaerobic membrane bioreactor. Environ. Sci. Water Res. Technol. 2021, 7, 739–747.
[CrossRef]
26. Harif, T.; Elifantz, H.; Margalit, E.; Herzberg, M.; Lichi, T.; Minz, D. The effect of pre-treatment on biofouling of BWRO membranes:
A filed study. Desalination Water Treat. 2011, 31, 151–163. [CrossRef]
27. Papapetrou, M.; Cipollina, A.; La Commare, U.; Micale, G.; Zaragoza, G.; Kosmadakis, G. Assessment of methodologies and data
used to calculate desalination costs. Desalination 2017, 419, 8–19. [CrossRef]