Evaluation of Pilot Scale Domestic Wastewater Reuse System in Terms of Irrigation and Industrial Process Waters in Turkey

ABSTRACT It is now inevitable to reuse the wastewater in urban areas in terms of climate change and circular economy. The aim of this study is to evaluate the reuse of secondary treatment as irrigation water and/or industrial process water by pilot-scale ultrafiltration (UF) and reverse osmosis (RO) units. The conductivity, pH, organic matter, heavy metals, micropollutants, coliform bacteria and sodium hazard assessments were analysedduring operation. An average flux of 60.02 ± 16 l/m2.h was obtained in UF system and 12.30 ± 2.93 l/m2.h was obtained in RO system. Both UF and RO permeates were compared with the national and the global irrigational water quality standards. It has been determined that UF effluent is suitable for irrigation, while RO permeate water can be used for industrial processes with UF pre-treatment. The total operation and maintenance (O&M) costs of two pilot plants were calculated at 0.252 US $/m3.


Introduction
Reuse of secondary wastewaters from wastewater treatment plants gaining attention for applications like irrigation, industrial and in some cases, direct potable use (Gikas and Tchobanoglous 2009;Paranychianakis, Salgot, and Angelakis 2010;Asano et al. 2007).Advanced biological treatment processes are generally used in the treatment of domestic wastewaters.However, these processes are often inadequate in removing non-biodegradable compounds and micropollutants and are not suitable for direct reuse.Conventional WWTPs, usually based on biological processes, are unable to fully remove these contaminants or any of their intermediate degradation products (Bunani et al. 2013).In other terms, the majority of wastewater treatment plants are designed primarily to remove organic nutrients, such as carbonaceous, nitrogenous and phosphorous organic substances.Use of advanced filtration techniques required for desired effluent quality for reuse.
Researchers have tested the efficiency of different advanced treatment technologies in the reuse of domestic and industrial wastewaters (Bunani et al. 2013;Cinperi et al. 2019).Membrane systems are used in combination with physical, chemical and biological processes to make the most efficient configuration in terms of water recovery/ reuse application.Membrane bioreactor (MBR) technology using ultrafiltration (UF) membranes can be combined with activated sludge processes or anaerobic biodegradation processes.Arévalo et al. tested water reuse after treatment by MBR and studied two different types of membranes, which are microfiltration and ultrafiltration.Permeates of both systems were shown to be suitable for unrestricted reuse based on Spanish reuse guidelines.The removal of the organic matter depends on the biological activity and is not affected by the membrane type.However, UF and microfiltration (MF) membranes were having different coliphage removal rate.The coliphage concentration was lower in the UF membrane due to its smaller pore size than MF membrane (Arévalo et al. 2012).
On the other hand, reverse osmosis (RO) systems with the ultrafiltration pre-treatment is the widely accepted treatment method for desalination and reuse.Both filtration technologies are effectively used for the removal of different organic compounds, and RO systems are effective in the separation of divalent and monovalent ions (Lopera, Agata, and Alonso 2019).
Nanofiltration (NF) and RO membranes are very effective after advanced biological treatment processes for direct or indirect use of treated wastewaters (Angelakis and Bontoux 2001).Cinperi et al. (2019) studied MBR, NF and RO pilot applications on wastewater coming from a woolen textile mill.Composite wastewater mixed from different processes was used as a feed stream.MBR + NF and MBR + RO pilot tests were conducted.Effluents were tested for the dyeing process in the lab scale and it was found that MBR + NF and MBR + RO did not negatively affect the dyeing when compared to softened groundwater.It was also found that color depth was increased with pilot plant effluent water (Cinperi et al. 2019).When the removal of dissolved organics and inorganics is desired, membrane technologies such as NF and RO should be applied.NF is capable of treating dissolved substances with a molecular weight of 300-1000 g/mol, for less than 300 g/mol, RO membranes should be used.RO can purify monovalent ions such as sodium and chlorine in the removal rate of 98-99.9%,while NF membranes can treat these ions by 50-90%.The applications of these membranes in water reuse also differ.NF is generally used to soften water and reduce TDS concentration in water, while RO is used to remove salts and ions (Sadr and Saroj 2015).Bunani et al. (2013) investigated different kinds of NF membranes (Ge Osmonics CK, Dow Filmtec NF270, and NF90) for the reuse of municipal wastewater.The use of NF90 between the three membranes has achieved the highest water quality.The membrane permeates were compared to irrigation water standards (Ayers and Wescot 1994).NF90 permeate had significantly lower concentrations in all parameters except for potassium in the irrigation water standards (Bunani et al. 2013).Bunani et al. also studied RO for the reuse of secondary treated urban wastewater.Brackish water RO (BWRO) and seawater RO (SWRO) membranes were tested at 10 and 20 bars, respectively for this purpose.According to the results, permeate water blended with the 20-30% secondary treatment was suitable for agricultural irrigation (Bunani et al. 2015).
In this study, the performances of two different pilot-scale treatment units, namely Ultrafiltration (UF), and Reverse Osmosis (RO), were evaluated in the removal of COD, TOC, SS, conductivity, anions, cations, microbiology and micropollutants from advanced biological treatment plant.The objective of this study was to assess the UF effluent and RO permeate separately with the irrigation and/or industrial water standards in terms of high sodium ions content.Because high sodium content in irrigation water affect the permeability of the soil and cause infiltration problems.Sodium hazard has therefore been assessed based on SAR (sodium adsorption ratio), ECw, SSP (soluble sodium percentage) and ESP (exchangeable sodium percentage).Morphological investigation of used membranes has been carried out for membrane fouling analysis.Energy and chemical consumptions have been recorded for operation and maintenance cost estimations.These cost calculations for the pilot-scale studies are necessary for obtaining representative results that can be applied to large-scale plants.Therefore, this pilot study provides extensive information about the reuse and feasibility of the treated domestic wastewaters.

Ambarlı advanced biological wastewater treatment plant (WWTP)
The WWTP is owned by the Istanbul Water and Sewerage Administration (ISKI) located in Ambarlı, Istanbul which had a capacity of 400,000 m 3 /day flow rate.The WWTP has a wastewater treatment line and a sludge treatment line.The wastewater treatment line comprises primary treatment (coarse and fine screening, grit and oil removal system) and secondary treatment (an extended aeration biological reactor, bio-phosphorous reactors and final clarifiers).The sludge is first thickened in the line by using gravity thickener and then dewatered.It also has a cogeneration technology that helps the use of the plant's energy demand.Pilot containers were mounted at the end of the plant's treatment line.

Pilot plant
Pilot plant fed with effluent coming from a secondary clarifier of the activated sludge WWTP.It consists of two cargo containers.One container has membrane systems consisting of an ultrafiltration module of 60 m 2 and the other container has six reverse osmosis modules of 37.2 m 2 and the related equipment.The UF and RO membranes used in the pilot plant was Inge Dizzer XL 0.9 MB 60 W and Oltremare BR3-8040, respectively.The cartridge filters used in the system was made of polypropylene with a 1-micron filtering.Both systems had an average design capacity of 100 m 3 /day.UF plant had an automatic coagulation dosing system for the feed line and a caustic, acid, and chlorine dosing system for the backwash line.The reverse osmosis system had sodium metabisulfite (SMBS) and an antiscalant dosing system in the feed line.Automatic Cleanin-Place and flushing system setup was used for the chemical cleaning of the RO membranes.Control panel schematic of the UF and RO pilot plants was given in Figure 1.

Experimental procedure
The effluent originating from the final clarifiers was first introduced into the UF container.To protect the UF membranes, influent water was passed through automatic filter and cartridge filters.UF permeate was collected in a buffer tank, which was having a volume of 20 m 3 .The filtration cycle for the UF membranes was a 2-minute backwash every 40 minutes of filtration.Chemically enhanced backwash was also applied every 12 hours of operation, using caustic, acid and chlorine.Automatic coagulation dosing system was not activated during the experiments.UF membrane was tested under 3 bars feed pressure.
UF permeate coming from the buffer tank was fed to the RO system.So there are two effluent streams that was evaluated; first UF effluent which was fed with the secondary clarifiers of the WWTP and the second stream was RO effluent which was fed with the effluent coming from the UF pilot system.RO membranes were operated with crossflow mode applying an approximately 50% recovery rate.During the RO operation, constant feed pressure of 10 bar was applied.

Online measurements and water analysis
Online flow rate, pressure, pH, turbidity and conductivity were measured using integrated probes in the pilot plants.The total organic carbon (TOC) were measured on a Shimadzu TOC-VWP analyser with an ASI-V autosampler.Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES, Perkin Elmer, Optima 7000) was used to measure the concentrations of Ni, Cu, Al, B, Si, Pb, Fe and Mn in the feed, UF permeate, RO permeate and concentrate streams.Anion-cation concentrations were measured by using Ion Chromatography (IC, Dionex, ICS-3000) for whole streams.Determination of the other parameters (such as COD, SS, Total Coliforms and Escherichia coli) was carried out with the Standard Methods for the Examination of Water and Wastewater (APHA, AWWA, and WEF 2005).These parameters were analysed during the operation period.Min, max, average and standard deviation values were reported from the 30

Calculations
Rejection (R) and flux results of UF and RO membranes were obtained from equation (Equations 1 and 2).C p , C 0 , J v , V, A and t indicate permeate concentration, feed concentration, permeate flux (L/m 2 .h), the volume of permeate, active area of the membrane and time, respectively.Sodium adsorption ratio (SAR), soluble sodium percentage (SSP) and exchangeable sodium percentage (ESP) were calculated according to cation results by using equation (Equation 3), (Equation 4) and (Equation 5), respectively.
Effluents from the pilot plant were evaluated for water quality analysis and compared with irrigation water standards and also Spanish Regulations for Water Reclamation and Reuse -Legal Framework Royal Decree 1620/2007.RO concentrate water quality was also compared with the Turkish Water Pollution and Control Regulation (SKKY) (SKKY 2004; Spanish Royal Decree 1620/2007, Spanish Regulations for Water Reuse.Real Decreto 1620/2007  ).One sample t-test was applied for both UF effluent and RO effluent quality parameters.Hypothesized mean (μ) for UF effluent and RO effluent, was accepted as a limitation of the related parameter in 'Spanish Regulations for Water Reclamation and Reuse -Legal Framework Royal Decree 1620/2007' and 'American Boiler Manufacturers Association (ABMA)', respectively.The significance level of probability (p-value) was 0.05.

Results and discussion
Results obtained from the UF and RO pilot systems were evaluated into two perspectives for 130 days of operation; performance evaluation and permeate water quality analysis.In the end, the cost of operation and maintenance were calculated.

Performance evaluation of the UF and RO
UF and RO permeate flow rate were recorded during the operation.An average flux of 60.02 ± 16 L/m 2 .hwas obtained in the UF system.During the operation, flux and pressure changes were shown, respectively in Figures 2 and 3.
Feed pressure was measured in the influent line of the UF module, just after cartridge filters.The average feed pressure of 2.41 ± 0.29 bar was measured during the operation.The average permeate pressure was 1.98 ± 0.41 bar.Differential pressure reflects the transmembrane pressure (TMP) of the UF module.An average TMP value of 0.45 ± 0.41 bar was obtained.When considering the increasing flow rate of the system, TMP starts to increase in the second month of the operation.
Feed pressure fluctuates based on the status of the automatic filter and the cartridge filters at the time of the data recording.Flux was also fluctuated based on the UF membrane fouling, the status of the automatic filter and cartridge filters.Also, the UF permeation cycle affects the automatically collected data based on the recording time of the filtration period.Therefore, highly fluctuated flux data were obtained.However, average flux reflects the UF membrane performance during the whole operation.
For the RO system, an average permeate flux of 12.30 ± 2.93 L/m 2 .hwas obtained.Flux changes during the operation were shown in Figure 4.The pressure of feed for RO was constant at 10 bar.Average influent and effluent conductivities were 1592 ± 375 µS/cm and 105 ± 43 µS/cm, respectively.Rejection performance based on the conductivity measurements for the RO membrane was 93.30 ± 3.65% for 130 days.
In the last 45 days of operation, Ambarlı Advanced WWTP capacity overload has caused several problems in the final clarifiers and the effluent quality of the WWTP was significantly reduced.That's why, in this period, UF TMP has increased and RO membrane flux has decreased.

Effluent water quality.
Water quality was analyzed through the pilot plant.Pilot plant COD and TOC rejections during the operation were given in Figure 5.The average COD value of feed water which was WWTP effluent water and UF effluent were 40 ± 17 and 19 ± 7 mg/L, respectively.The average COD value of 5 ± 4 mg/ L was achieved with RO permeate.RO concentrate was having an average value of 63 ± 27 mg/L COD.When compared to SKKY standards even the RO concentrate was under the limit value of 100 mg/L of COD.TOC measurements for the influent, UF effluent, RO permeate and RO concentrate were 18 ± 4, 16 ± 4, 0.32 ± 0.18, and 36 ± 15 mg/L, respectively.The average TOC rejection for the whole pilot plant was 98 ± 3%.
COD, TOC, SS, SS, heavy metals, anions and cations concentration during the operation has been given in Table S1.As heavy metals measured in water are in the form of monovalent and multivalent ions, the removal of these metals by UF could not occur, as UF membranes could not remove monovalent and divalent ions from water.While NF can remove divalent or larger ions, RO must be preferred to remove monovalent ions, simultaneously.Thus, only RO membrane removal efficiencies were taken into consideration.Averagely, more than 80% Si, Fe and Mn removal efficiencies were obtained from RO, on the other hand only 22% B removal was observed as expected.The low rejection on some of the heavy metal elements like boron, lead, copper and aluminium observed.This may be related to membrane type used in the system, which was brackish water model.One of the most problematic parameters in treated wastewater is nitrate.Kim et al. (2019), focusing on the recovery of wastewater by removing undesirable ions from treated wastewater, applied the membrane capacitive deionization process on a pilot scale.While the nitrate removal efficiency achieved is equal to the result of Kim et al. (2019) (>75%), other anions and cations removals were obtained higher in our study.Detailed COD, SS TOC, anion, cation removal efficiencies were listed in Table S1.UF effluent and BW RO membrane permeate salinity measured as electrical conductivity (ECw) and its specific ion toxicities compared to irrigation water guidelines in Table 3 (Bunani et al. 2015;US EPA 2012).The conductivity measurements and rejection of the BW RO membrane were given in Figure 6.The rejection for the conductivity was 93.3 ± 3.65%.RO Permeate conductivity was 105 ± 43 µS/cm.The results of the microbiological analysis were listed in Table 1.For the RO permeate, Fecal Coliforms and Total Coliforms in 10 of the total samples taken during the operation were measured as 0. However, 4 of the total samples introduce a small number of coliform bacteria because of the possible contamination of RO permeate, while taking samples from the plant.Only 4 of the samples were contaminated among the 14 samples.Average Fecal Coliform and Total Coliform in the RO permeate were obtained as 0 CFU/100 mL.UF effluent and RO concentrate were also measured and results are presented in Table 1.UF effluent microbial removal efficiency calculated as 99.995%, which means above Log 4 bacterial removal efficiency is possible with just adding UF after secondary wastewater.
Fifteen different types of micropollutants were chosen as target micropollutants and their removal rates were studied, as summarized in Table 2. DDT (dichlorodiphenyltrichloroethane) is a common type of pesticide used to control insects in agriculture and insects carrying diseases like malaria.It does not have any odor or taste.DDE (dichlorodiphenyldichloroethylene) and DDD (dichlorodiphenyldichloroethane) are similar chemicals to DDT.Aldrin, Dieldrin, Endrin, Isodrin, Heptachlor, Benzo(a)pyrene, Pyrene, and DDT-like chemicals are organochlorine insecticides on the list of persistent organic pollutants (POPs).Even some developing countries still use these compounds due to their low cost and versatility in industry, agriculture and public health (Zhang et al. 2004).Fluoranthene, Benzo(b)Fluoranthene, Benzo(k)Fluoranthene are polycyclic aromatic hydrocarbons (PAH) substances.PAHs are prevalent in urban watersheds, and their concentrations in urban waterways are associated with the degree of urbanization and population density (Meyer, Lei, and Wania 2011).These compounds are toxic to mammals, act like hormones, disrupting the reproductive systems of humans and wildlife.It also allows to monitor the effects of industry, agriculture and surface flow simultaneously.The micropollutant analysis was carried out for WWTP effluent, UF effluent, RO permeate and RO concentrate streams which were taken at different times during operation.The results were listed in Table 2 in the ppt (ng/L) unit.Among the measured micropollutant parameters, they were not detected in the WWTP effluent, UF effluent and RO permeate streams.Very low micropollutant concentrations were detected only in RO concentrate samples.Cd and Hg concentrations were also measured for three of WWTP effluent, UF effluent, RO permeate and RO concentrate samples.Cd and Hg results were under detection limit (<0.001 ppm and <0.005 ppm, respectively) of AAS, apart from RO concentrate of one sample which has 0.013 ppm Hg.

Evalutation of pilot effluents in terms of wastewater reuse
The results obtained from average UF effluents and RO permeate were compared with 'Guidelines for interpretation of water quality for irrigation in Turkey' in Table 3 to discuss the potential irrigation problems.(US EPA 2012; Bunani et al. 2015) The degree of restriction of UF effluent on use is graded as 'slight to moderate' according to SAR, EC w , TDS, Na and Cl results.On the other hand, infiltration, B, NO 3 -N, SSP and pH values have no restriction on agricultural irrigation.p-value for EC w , TDS, Na, Cl, B and NO 3 -N are found lower than 0.05 for UF effluent, thus it can be said that the difference between UF effluent and limitations are significant.In other words, the effluent UF quality guarantee to meet the irrigation limitations.Looking at the RO permeate results, most of the values are graded as unrestricted, but when the SAR and EC w results are evaluated together, it is not suitable for agricultural irrigation.Because excessively low salinity causes the soluble salt and minerals to be washed away from the soil which is not a desired agricultural situation.To avoid possible irrigation problems, it is recommended to use RO permeate by blending with UF effluent or even with the WWTP effluent (Bunani et al. 2015).According to 'Legal Framework Royal Decree 1620/2007 establishing the legal requirements for reuse water' E. coli, and suspended solids (SS) parameters were compared with the water quality results of UF effluent and RO permeate.The quality of each water was suitable for consumable and non-consumable agricultural products according to this guideline.(Spanish Royal Decree 1620/2007, Spanish Regulations for Water Reuse.Real Decreto 1620/2007 2007) The use of reclaimed water for boiler make-up water is an alternative reuse of RO permeate stream.Required water quality for boiling water differs according to the operating pressure of the boiler.Typically, higher operating pressure needs higher-quality water rather than the lower one.The characterization of RO permeate water and the maximum recommended boiler water limitations are shown in Table 4, comparatively.It can be seen that RO permeate water can be used as boiler water until 1500 psi pressure.The difference between RO effluent and boiling water limitations in terms of TDS, conductivity, and silica were found significant (p < 0.05).Hence, this high-quality RO permeate stream can benefit nearby industrial zones and production facilities.A similar pilot-scale treatment process was applied to a treated industrial wastewater stream by Ozbey-Unal et al. (2020).The effluent of the MF/RO process combination was found efficient in using for industrial cooling or boiling water, which upholds our results (Ozbey-Unal et al. 2020).Aziz and Kasongo (2021) applied a three-stage membrane process, consist of UF/NF/RO, to the secondary effluent of the MBR plant for water reuse.RO effluent of their pilot-scale plant was found appropriate for cooling water for following standards: COD < 30 mg/L, NH 3 < 1 mg/L, PO 4 < 7 mg/L, EC < 1445 µS/cm and turbidity < 36 NTU (Aziz and Kasongo, 2021).According to these standards and the results (shown in Table S1 and Table 3), pilot RO effluent stream is suitable as cooling water, again.Although there was no NF process in our study, the effluent qualities were quite similar to Aziz and Kasongo's study.This argument shows that the results obtained from the study correspond to other studies in the literature.
Ambarlı WWTP was in the coastal region so that the RO concentrate can be discharged to the Marmara Sea.According to SKKY standards (SKKY Table 23), RO concentrate meets the requirements for all the parameters listed.(SKKY 2004)

Fouling analysis
After completing the pilot study, specimens were taken from the RO spiral wound membranes for autopsy analysis.At the end of the pilot operation, this autopsy aimed to investigate RO membrane sheets and determine the fouling on the membrane surface.The latest module in the row of the pressure vessel which encountered more concentrated feed water was chosen for the analysis.RO membrane sheets were observed by cutting the spiral wound membrane.Three membrane sheets were chosen for the SEM-EDS and Confocal Microscope analysis.Biofouling and inorganic fouling of the RO membranes were evaluated using the analysis results. Figure 7 presents the actual photos of the RO membrane sheets and several SEM images showing the most significant foulants on the membrane surface which were observed in the autopsy analysis.
No substantial amount of fouling was observed on the membranes when inspected physically by the eye, dirty areas were very rare.Random three sheets of the spiral wound module were chosen for the SEM.A closer inspection by SEM images reveals amorphous and soft layers on the membrane surface (Figure 7b).Hard-edged and crystalline substances were not observed on any of the SEM samples.The amorphous substances were sludge-like and can be cleaned by rubbing with flushing water.
EDS analysis was conducted to understand the chemical content of the fouling layer observed on the SEM.The analysis showed that the chemical content of the fouling layer is composed primarily of Carbon and Sulphur.High carbon content could result from the organic content.However, the source of the carbon was mostly the membrane polyamide surface itself.Sulphur was also high on EDS measurements.It was also because of the polysulfone support layer of the membrane.No Calcium element was detected in EDS spectrum (Figure S1).Therefore, the Sulphur reading was not coming from the CaSO 4 scaling.It was not observed on the SEM.There was no significant fouling on RO membrane surfaces based on the EDS.
Red dots on the pictures taken from the confocal microscope represents (Figure 7) dead cells on the membrane surface.There are no living microorganisms observed.Dead ones were only a few.After a successful operation of 130 days, fouling analysis has shown that the Ambarlı Advanced WWTP effluent is suitable for RO filtration, and UF pretreatment of the RO system worked as expected, protected the RO membranes successfully.

Operation and maintenance costs
This plant has two following treatment units that can be listed as UF unit (UF module with mechanical and cartridge filters and chemical dosage), and RO unit (brackish water RO membrane, and chemical dosage line).The pilot plant can treat WWTP effluent effectively despite the unstable wastewater characteristics.The plant capacity was designed for 100 m 3 /day, however, the average capacity of 66 m 3 /day was reached during the operation.Operation and maintenance costs include electricity, chemical and cartridge filter replacement cost.The electricity consumptions were 0.38 kWh/m 3 for UF and 1.49 kWh/m 3 for the RO system with the total consumption ratio of 2 kWh/m 3 and the total energy cost was calculated as 0.24 US $/m 3 .Chlorine, caustic, acid, antiscalant and SMBS consumption of the UF and RO system were recorded and calculated per cubic meter of treated water.Based on these records, chlorine, caustic, acid, antiscalant, SMBS consumption per cubic meter were 0.07 mL, 88 mg, 0.38 mL, 32.2 mL and 500 mg, respectively.Cartridge filters changed one time only and five 20" cartridges were used in the system.Maintenance cost excluding energy consumption of the system was calculated as 0.012 US $/m 3 for the UF and RO pilot plants in total.
The operation and maintenance cost for UF was found to be 23% of the total calculation and was 0,058 US $/m 3 .The RO system has the higher energy and chemical consumption, which resulted in the cost of 0,194 US $/m3.For both UF and RO pilot units with an average capacity of 66 m 3 /day, total operation and maintenance cost were calculated as 0.252 US $/m 3 .

Conclusions
The main results obtained from the pilot plant operation of UF and RO showed that Ambarlı Advanced WWTP effluent requires post-treatment for suitable reuse as irrigation or industrial process water.
UF and RO membrane performances were stable for 130 days except for the days of the WWTP overload.
Extended water quality analyses were done to compare the UF effluent and RO permeates for the reuse applications such as irrigation and industrial process water.Moreover, none of the target micropollutants and Cd and Hg were detected in the WWTP effluent, UF effluent, and RO permeate.UF effluent was suitable for urban and irrigational water reuse while RO permeate was suitable for the industrial process or boiler make-up water.Alternative usage scenarios such as blending RO permeate with UF or WWTP effluent is also possible for high-quality irrigation water.Besides, RO concentrate meets the requirements to discharge to Marmara Sea.
SEM-EDS and confocal microscope analysis of used RO membrane samples showed that no significant scaling and fouling occurred on the membrane surface.
Total O&M cost was 0.252 US $/m 3 and will be lower on fullscale application.Based on the SAR, EC w and other parameters measured, effluent from the pilot UF can be used for irrigational use and effluent from the pilot RO can be used as process water in the nearby industrial facilities.
measurements.While FEI Quanta Feg 250 scanning electron microscope (SEM) was used to visualize the fouling on the RO membranes, microanalysis of the precipitated elements was carried out in an energy-dispersive spectrometer (EDS).RO membranes were stained with Live/Dead Backlight Viability Kit and then analysed by Nikon C4 confocal microscope to visualize the bacterial fouling on the membrane surface.15 different target micropollutants were monitored during operation.Cd and Hg analysis completed with Thermo Scientific ICE 3500 Atomic Absorption Spectrophotometer and other micropollutant parameters (dichlorodiphenyltrichloroethane (DDT); dichlorodiphenyldichloroethylene (DDE); dichlorodiphenyldichloroethane (DDD); Aldrin; Dieldrin; Endrin; Isodrin; Heptachlor; Benzo(a)pyrene; Pyrene; Fluoranthene; Benzo(b) Fluoranthene; Benzo(k)Fluoranthene) were measured by using Thermo Scientific TSQ 8000 Evo Triple Quadrupole electron ionization mode Gas Chromatograph-Mass/Mass Spectrometer (GC-MS/MS).

Figure 1 .
Figure 1.The SCADA control panel of the two systems; ultrafiltration and reverse osmosis. 2007

Figure 2 .
Figure 2. Pilot UF flux during the pilot operation.

Figure 3 .
Figure 3. Pilot UF feed, permeate line pressure, and TMP changes during the operation.

Figure 4 .
Figure 4. Pilot RO permeate flux changes during the operation.

Figure 6 .
Figure 6.Pilot RO membrane feed and permeate conductivities and rejection during the operation.

Figure 7 .
Figure 7. Actual photos of the 4"x40" RO membrane sheets after pilot use (a) SEM images (b) and confocal microscope images (c).

Table 1 .
analysis results of the UF effluent RO permeate and RO concentrate.

Table 2 .
Micropollutant analysis results of the WWTP effluent, RO permeate, and RO concentrate (ng/L).
N/D: Not detected

Table 3 .
Irrigation water standards compared with the UF effluent and RO permeate of the pilot plant.

Table 4 .
Comparison of pilot-scale RO permeate and American Boiler Manufacturers Association (ABMA) maximum recommended boiler water limit.