Adv Syst Sci Appl 2024; 02:203–218 Published online at https://ijassa.ipu.ru. Solving the Energy Consumption Barrier in Brackish Water Reverse Osmosis Desalination Plants: A Genetic Algorithm and Energy Recovery Approach Moumni Mohammed*, Massour El Aoud Mohamed, Moumni Fatima Zahra Sultan Moulay Slimane University, Khouribga, Morocco Abstract: Reverse osmosis desalination is an effective technology for supplying potable water to regions facing water stress. However, this process consumes a significant amount of energy, limiting its widespread adoption globally. This study aims to analyze the variations in specific energy consumption (SEC) in the reverse osmosis desalination process for brackish water at a plant in Morocco, considering the feed water parameters. Additionally, the study examines energy consumption with and without the implementation of energy recovery devices at this plant, which produces 10 million cubic meters of water annually. A genetic algorithm is utilized to identify the optimal combination of design and operational parameters to achieve the lowest SEC. The findings indicate that incorporating energy recovery devices in the future design of the plant could reduce the SEC by up to 30%. Keywords: desalination, brackish water, reverse osmosis, energy recovery 1. INTRODUCTION At present, increases in the global population, coupled with drought and desertifcation due to climate change, will undoubtedly aggravate water security [1]. Desalination technologies have been emerged to supply water from unconventional water sources, and reverse osmosis systems account for a large share of desalination facilities [2]. Reverse osmosis (RO) process is an important filtration process that is used extensively for the desalination of sea and brackish water all over the world [3]. Reverse osmosis is generating growing interest due to its energy efficiency and versatility compared to other desalination technologies. The need to supply drinking water to populations in regions suffering from water shortages has made the development of this technology essential. [4]. Reducing the Specific Energy Consumption (SEC) poses a significant challenge, driving technological innovation and research in the desalination industry. The energy cost of the desalination process (RO) can account for up to half of the total cost of producing one cubic metre of drinking water. [5]. Nevertheless, compared with other desalination technologies, such as multi-stage flash (MSF) [7], [8], multiple-effect distillation (MED) [9], membrane distillation (MD) [ [10], [11], [12], [13],and electrodialysis (ED) [15], the RO process consumes relatively little energy. Consequently, most large-scale seawater and brackish water desalination plants have been designed to use the RO method. According to several research studies, the SWRO process consumes between 2 and 5 kWh/m³ depending on the feed characteristics [24]. On the other hand, the total energy consumption of the BWRO process, including electrical energy, is between 1.5 and 2.5 kWh/m³ [20]. Many researchers and engineers have dedicated significant effort to finding innovative ways ∗Corresponding author: moumni.mohammed@gmail.com 204 M. MOUMNI, M. EL AOUD MOHAMED, M. FATIMA ZAHRA to reduce the energy consumption of the reverse osmosis process. They have focused on several key areas and methods to achieve this objective, including the following: • Improvement of RO membranes: Developing and utilizing high-performance membranes to enhance filtration efficiency and reduce energy requirements. • Reduction of membrane fouling: Implementing prevention and cleaning techniques to decrease membrane fouling, maintaining optimal efficiency and lowering energy consumption. • Efficiency of energy-consuming equipment: Enhancing the efficiency of pumps and other energy-intensive equipment to minimize energy use. • Energy recovery devices (ERDs): Using energy recovery devices to capture and reuse energy, thereby reducing the overall energy demand. • Innovative membrane module designs: Introducing new membrane module configurations to reduce pressure drops in the membrane channel, thus improving energy efficiency. • Process optimization: Applying optimization algorithms to dynamically adjust operational parameters and maximize energy efficiency in real-time. • Integration of renewable energy: Utilizing renewable energy sources, such as solar or wind power, to supply desalination plants, reducing reliance on traditional energy sources. First, advances in RO membranes have contributed to energy savings. High-performance reverse osmosis membranes, characterized by increased water flow and salt rejection rates, mitigate the excess pressure required above the osmotic pressure of seawater. Researchers have successfully developed an efficient RO system for low SEC water desalination by adjusting membrane type and size [25] and [26], respectively. Furthermore, low fouling propensity is critical for lowering energy consumption in the SWRO process because membrane fouling reduces water permeability, which requires higher operating pressure. Surface modification has been used to develop fouling-resistant RO membranes by improving hydrophilicity, reducing surface roughness, and decreasing concentration polarization at the membrane’s surface [16] and [17] . Moreover, to achieve an optimal combination for a reverse osmosis system with the lowest SEC, some researchers are focusing on the design of single-stage or two-stage RO systems. They analyze the performance of these systems with varying feed parameters, membrane permeabilities, and recovery rates. ( [1, 27–30]). Two-stage reverse osmosis can achieve a recovery rate of over 50% ( [31]).Mingheng [32]. It has been demonstrated that, without inter-stage booster pumps, single-stage is more energy efficient than two-stage in BWRO due to lower retentate pressure drop. Li et al. [33] investigated the validation of a model-based optimization for BWRO operations. In addition, energy-intensive equipment such as high-pressure pumps (HPPs), booster pumps (BPs), and ERDs have been developed and improved to reduce the overall energy consumption of the SWRO process. The efficiency of HPP has increased by about 90%, which is a practical limit for centrifugal pump efficiency. Large pumps are recommended for improving HPP efficiency, with pump efficiency in a large-scale desalination plant reaching up to 85% [18], [22]. In other areas, the integration of photovoltaic (PV) and wind power in desalination processes has been evaluated, with promising results in regions with high solar and wind potential. For example, Yahiaoui et al. optimized a PV-diesel generator-battery hybrid system for the city of Djanet, Algeria, using the Grey Wolf optimizer [23]. In the last 20 years, advancements in ERD have played a significant role in reducing the energy consumption of the SWRO process. The development of ERD equipment, including Francis turbines (FTs) and pressure exchangers (PXs), has significantly reduced energy consumption [18], [19]. The efficiency of PX can exceed 95% [18]. As a result, advancements in energy-consuming units have nearly Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) SOLVING THE ENERGY CONSUMPTION BARRIER... 205 reached their practical limits. Many research investigations have been conducted to compare ERDs for saline waters. Alexander et al. analyzed the choice between two types of ERD (turbocharger and isobaric ERD) for brackish water in his study [35], taking into account the conversion rate parameter and the Life Cycle Cost analysis. In addition to its efficiency of between 94 and 98 percent for a discharge rate of up to 110 m3 per hour, which is close to the discharge rate of the plant in this case study, (Alexander et al. concluded that the isobaric ERD is more energy efficient than the turbocharger, particularly at a conversion rate of 75 to 81 percent, which is also the case for the plant in this study. As a result, the isobaric ERD will be considered an energy recovery system for the RO process used in this plant, which lacks ERD. Furthermore, because the organization intends to expand this plant by adding more RO trains, an analysis of the SEC with and without ERD will be conducted to justify the feasibility of implementing this ERD at this plant. This analysis will determine whether implementing an energy recovery device (ERD) in the expanded plant is a viable option. To achieve this, a genetic algorithm will be applied to identify the optimal ERD configuration, maximizing energy efficiency while considering the specific parameters of the reverse osmosis process. The genetic algorithm, known for its ability to solve complex optimization problems, will help explore a wide range of possible solutions and converge on the best configuration. 2. DESCRIPTION OF BWRO DESALINATION PLANT The BWRO desalination plant consists of three production lines, as depicted in Fig 2.1. Each production line is structured in two stages. In this setup, the concentrate from the first stage serves as the feed water for the subsequent stage, allowing for additional permeate production. The permeate collected from the first stage is combined with the permeate from the second stage. The HPP increases the pressure of the pre-treated brackish water to a suitable value for Fig. 2.1. Schematic Diagram the membrane. The pressure required depends on the concentration and temperature of the feed water. Osmotic pressure increases with concentration, so the operating pressure must be higher than the osmotic pressure corresponding to the concentration of the rejected brine flow rate at the membrane rack outlet, as well as membrane fouling. This plant’s feed water concentration varies from 1g/l to 3g/l throughout the year, as does its temperature, which ranges from Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) 206 M. MOUMNI, M. EL AOUD MOHAMED, M. FATIMA ZAHRA Table 2.1. Equipment characteristics Number of production line 3 Number of membrane by line 600 Membrane Sectional area 35m2 Nominal Flow of HP pump 550m3/h Nominal Pressure of HP pump 20 bar Nominal Speed of HP pump 3000 rpm 7C° to 32C°. To manage the changing characteristics of brackish water and keep within the operating range of this plant, the recommended conversion rate for each production line with its two stages is 75%, with a production flow rate of about 390 m3/h. To ensure that the above instructions are followed, each HP feed pump is equipped with a VFD that allows the pump speed to be varied, which changes both the flow rate and the pressure of the membrane feed water. The article [36] used fuzzy logic to effectively control these parameters in the station object of this study. Table 2.1 shows the equipment used in this station’s RO process. 3. MATERIELS AND METHODS Before evaluating the specific energy consumption (SEC) of the (RO) process implemented in this station, it is essential to first model the RO process and develop a simulation model. This initial step was thoroughly addressed in the work presented in article [37].Using Matlab/Simulink software tools, a numerical model was developed and validated against real plant values, as well as against the model used by Arun-Joseph and Vasanthi Damodaran in Table 3.1. RO Process Equations Definition Equation Osmotic Perssure (bar) π = i.Cf .R.T (3.1) Permeate flux (m.s−1) Jw = Aw(∆P −∆π) (3.2) Solvent permeability (m.s−1.Pa−1) Aw = A0exp(6.433− 1885 T ) (3.3) Permeat flow (m3/h) Qp = S.JW (3.4) Drop Pressure Across Membrane (bar) ∆P = ( Pf + Pc 2 − Pp) (3.5) ∆P = ∆π + ( Qp S.Aw ) (3.6) Conversion rate Y = Qp Qf (3.7) their study [38]. The equations used for this modeling are summarized in Table 3.1. With this model in place, the current SEC of the RO process can be evaluated both with and without an energy recovery device (ERD). This comparison will demonstrate the potential benefits of installing an ERD in the plant. Following the conclusion to install the ERD, it is necessary to determine the optimal parameters for the RO process, taking into account the parameters of Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) SOLVING THE ENERGY CONSUMPTION BARRIER... 207 the ERD, which will also be modeled using equations. This optimization will be carried out using a genetic algorithm to ensure the most efficient operation of the expanded plant. 4. EVALUATION OF SEC AT BWRO Electricity consumption is a major obstacle in the development of the RO process, with High- Pressure Pumps (HPPs) being the primary energy consumers. These pumps need substantial energy to overcome the osmotic pressure of saline water. Optimizing energy use involves lowering the feed’s osmotic pressure and selecting the most efficient pump. HPPs account for about 75% of total specific energy consumption, with the remainder coming from membrane operations [26]. This study, therefore, focuses on the energy consumption of HPPs in the RO process for freshwater production. The equation below shows a mathematical method for estimating the SEC of the RO membrane without the need for an ERD system: Es = Pf .Qf Qp.η (4.1) The model generated in Matlab/Simulink to evaluate the fluctuation of SEC in the station under study can be used to get an understanding of the predicted energy consumption and losses at the discharge. Three additional scenarios are added to the three cases given in the article [37] to study the energy fluctuation, as shown in Table 4.1. Furthermore, the energy assumed to be lost during discharge can be calculated by adding the flow rate and pressure of the concentrates. Given that the organization intends to expand this station with more trains, and based on the results, which show that the energy lost to create a cubic meter of permeate can average 0.3 kwh, a ERD for the new entity must be installed. 5. MODELING ERD As previously mentioned, an isobaric Energy Recovery Device (ERD) will be utilized for the RO process implemented in this plant. Rotary isobaric devices use a small rotor to recover hydraulic energy from the concentrate stream. The rotor contains ducts that alternately fill with high-pressure brine and low-pressure feed water. As the rotor turns, these ducts are exposed to high- and low-pressure zones alternately, effectively replacing the high-pressure brine with saltwater on a 1-to-1 basis. The timing of the water exchange ensures that the chamber is never completely empty, creating a static water piston that prevents the two streams from mixing. [39]. The equations presented below are used to determine the best design for the RO process with an ERD. The installation of an ERD in an existing RO process necessitates knowledge of its operating parameters. Figure 5.1 shows the ERD system’s inputs and outputs. Fig. 5.1. The ERD system’s inputs and outputs Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) 208 M. MOUMNI, M. EL AOUD MOHAMED, M. FATIMA ZAHRA 5.1. Flow Calculation The ERD requires for its adequate choice the preknowledge of the inlet flows(Qic, Qif ) and outlet flows (Qoc,Qof ) of this equipment. Water flows from the concentrate of RO process and the low pressure pumps are the two types of water flows that enter the ERD. Qic can be determined as is defined, the concentrate flow rate, Qc which is a function of the RO process feed flow Qf and its conversion rate Y : Qic = Qf .(1− Y ) Where : Qf = Qp Y = Qc +Qp (5.1) Each ERD is is characterised by Brine flow loss or Leak (L) and Overflush (Of) [34] where : Qif , Qof and Qoc can be defined using the following equations: Flow balance: Qic +Qif = Qof +Qoc (5.2) Overflush (Of): Overflush range is provided by isobaric ERDs manufactures: Of = Qif −Qof Qof (5.3) Table 4.1. Simulation values and actual values comparaison Cnd Unit case1 case2 case3 T C° 20 13,5 24 Cf g/l 2 1.6 1.26 Pa ST* Mb** Er % ST Mb Er % ST Mb Er % Pf bar 15,9 15,2 4,4 18,1 17,8 1,65 12,9 14,1 9,3 Pc1 bar 13,9 13,2 5,03 15,2 15,7 3,28 10,5 12,1 15,2 Pc2 bar 11,9 11,2 5,88 12,9 13,7 6,2 8,95 10,1 12,84 Qp1 m3/h 305,4 299,04 2,1 296,9 323 8,7 298,9 287,3 3,88 Qp2 m3/h 76,05 74,4 2,16 91,3 80,4 11,9 80 71,5 10,6 Qc1 m3/h 230,45 225,6 2,1 215 243 13 226,6 216,6 4,4 Qc2 m3/h 154,4 151,15 2,1 140,2 163,1 16,33 146,6 145.1 1 Cp1 g/l 0,04 0,02 50 0,01 0,01 0 0,02 0,02 0 Cp2 g/l 0,09 0,06 33,3 0,02 0,03 50 0,06 0,04 33,3 Y % 71,2 72,7 2,1 73,5 71,1 3,2 72,1 71,2 1,2 Es kwh/m3 0,8 0,76 5 0,85 0,89 4,4 0.68 0.71 4,4 Ec kwh/m3 0,17 0,15 7,5 0,16 0,2 23,5 0,15 0,16 6 Cnd Unit case4 case5 case6 T C° 24 10 12 Cf g/l 1 2 1.2 Pa ST* Mb** Er % ST Mb Er % ST Mb Er % Pf bar 13,4 13,95 4,1 19,1 19,55 2,35 17,8 18,09 1,6 Pc1 bar 11,6 11,95 3 17,7 17,55 1,1 15,9 16,09 1,1 Pc2 bar 10,2 9,95 2,4 15,9 15,5 2,5 14,3 14,09 1,4 Qp1 m3/h 291,04 285,62 1,8 345,2 338,1 2 330,2 325,28 1,4 Qp2 m3/h 72,05 71,11 1,3 96,3 84,17 12,5 82,1 80,98 1,3 Qc1 m3/h 220,45 215,47 2,2 257,2 255,07 0,8 243,6 245,39 0,7 Qc2 m3/h 149,4 144,36 3,3 173,2 170,9 1,3 167,6 164,41 1,9 Cp1 g/l 0,04 0,02 50 0,05 0,04 20 0,02 0,02 0 Cp2 g/l 0,07 0,05 28,5 0,11 0,09 18,1 0,07 0,06 14,2 Y % 71,19 72,1 1,2 71,1 71,1 0 72,1 71,2 1,2 Es kwh/m3 0,66 0,69 4,5 0,92 0,98 6,5 0,88 0,91 3,4 Ec kwh/m3 0.15 0,13 13,3 0,25 0,24 4 0,22 0,21 3 Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) SOLVING THE ENERGY CONSUMPTION BARRIER... 209 Brine flow loss or Leak (L): is provided by isobaric ERDs manufactures as a function of temperature and brine input flow per isobaric ERDs unit [40]: L = Qic −Qof Qic (5.4) Therefore the output flow rate from the side of feed membranes Qof can be expressed as : Qof = Qic.(1− L) (5.5) And Qif : Qif = (Of + 1).Qof (5.6) Finally by using Flow balance: Qic +Qif = Qoc +Qof (5.7) The output flow rate from ERD to the rejection can be expressed as : Qoc = Qic +Qif −Qof (5.8) 5.2. Salinity Calculation After defining the inlet and outlet flows of the ERD system, the salinity of the flows in the ERD sides can be determined. Cicis the salinity of the concentrate of the RO process Cc and Cif is the salinity of feed water with low-pressure flow which is equal to the salinity Cf . It remains to determine the concentration of outflows from the ERD system Coc and Cof which are respectively the salinity of reject flow and the salinity of pressured flow by the ERD. Each isobaric ERD has its own mixing Mx ratio [41] which is the ratio of the volume of brine that transfers into a volume feed water and can be calculated with the following equation independent of the pressure exchanger high and low pressure flow balance: Mx = Cof − Cif Cic − Cif (5.9) Using the equation (5.9) Cof can be expressed as a function of the previously known parameters: Cof = Mx(Cic − Cif ) + Cif (5.10) According to the salinity balance : Qic.Cic +Qif .Cif = Qoc.Coc +Qof .Cof (5.11) Coc can be expressed as: Cic = (Qic.Cic +Qif .Cif )−Qof Qoc (5.12) 5.3. Pressure Calculation The ERD’s input pressures Pic and Pif are respectively, the one of the from the concentrate flow rate of the RO process Pc, and the second is the pressure of low pressure flow rate coming from the low pressure pump, which is generally the circulation water pressure taken at approximately 2 bar. Moreover, the ERD outlet pressure may be estimated as follows: The output pressure recovered by ERD Pof : Pof = Pic −∆p1 (5.13) The pressure of the discharge from the ERD Poc: Poc = Pif −∆p2 (5.14) Where : ∆p1 and ∆p2 are the losses across the isobaric ERD. Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) 210 M. MOUMNI, M. EL AOUD MOHAMED, M. FATIMA ZAHRA 6. BOOSTER PUMP A multi-stage brackish system without an interstage boost can be designed in a similar manner to a single-stage system. In this scenario, the previous stage’s concentrate is used to pressurize a feedwater stream for the first stage. The circulation pump compensates for pressure losses in the membrane stages, pipework, and ERD. Because the brackish and seawater RO processes are not the same, the implementation of isobaric ERDs in brackish water RO systems must be handled differently. The small amount of pressure loss caused by membranes, friction in the ERD, and the piping circuit necessitates the use of a booster pump in isobaric ERD. This pump is used as the ERD’s output in single-stage seawater systems. The ERD booster pump, on the other hand, can play two important roles in a two-stage brackish water system by being installed between stages one and two. The ERD booster pump acts as an inter-stage booster pump in this configuration, lowering the required pressure from the main high-pressure feed pump and balancing flux between stages 1 and 2 [35]. In this study, the booster pump will function as a circulation pump to recover pressure lost in the circuit (membranes, ERD, and pipes). 7. THE SEC OF BWRO WITH ERD In order to evaluate the SEC by a BWRO equiped with a ERD, firstly it is necessary to express the energy recovered in the concentrate part of the process, which is function of the RO brine pressure Pc , flow rate of the concentrate Qc, and the ERD efficiency ηerd as shown : Eerd = Pc.Qc.ηerd (7.1) by introducing a booster pump (BP) with efficiency ηBP , the energy consumed by this pump must be considered in order to evaluate the energy saving envisaged in the new design. In fact, this pump takes the water flow Qof coming from the ERD with the pressure Pof in order to achieve the required pressure at the RO process inletPf , therfore, the energy consumed by(BoP) can be expressed as follows: EBP = (Pf − Pof )Qof ηBP (7.2) The pressure provided by the booster pump PBP : PBP = Pf − Pof (7.3) The SEC by the BWRO with a ERD will be the combined energy consumed by the HP pump and the Booster pump [?]: SEC = (Qf −Qof ).Pf ηHP .Qp + PBP .Qof ηBP .Qp (7.4) 8. RESULTS AND DISCUSSIONS The Matlab/Simulink model developed in the article [37] and validated by the real values of the studied station makes it possible to evaluate the SEC by the process RO and especially the HP pumps. Furthermore, the values considered in the above study are taken at the start of exploitation of the said plant. However, after two years of operation, it was discovered that feed pressure values had increased for the same feedwater properties (salinity and temperature). This is due to membrane fouling and its life span. Following the development Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) SOLVING THE ENERGY CONSUMPTION BARRIER... 211 of the operating parameters of the RO process in this plant, it was observed that feed pressure and concentrate pressure rise by 2 to 3 bars. As a result, the SEC increases, proving the significance of the current study. Using the ERD modeling equations and the RO model developed in Matlab/Simulink, the ERD has been implemented into this model while taking into account the inclusion of a booster pump, which is essential for an RO process with an ERD. The proposed design is illustrated in Figure 8.3. Then, the Specific Energy Consumption (SEC) of the three reverse osmosis (RO) trains implemented in the plant under study is evaluated by considering various states of the saltwater feed, specifically by varying its salinity and temperature. These two parameters have a direct impact on the SEC of an RO process, as indicated in several studies ( [6,20,21]), regardless of the system’s design and structure. The graphs below illustrate the evolution of SEC for the RO process in this plant with and without Energy Recovery Device (ERD). When the feed water salinity is set at 1.85 g/L and its temperature is varied within the recorded range at the station, the SEC of the RO process with ERD decreases from 2.1 kWh/m³ to 1.45 kWh/m³ as the feed water temperature increases (Figure 8.1). In contrast, the SEC for the current design without ERD decreases from 3.05 kWh/m³ to 2.15 kWh/m³ under the same conditions (Figure 8.2). In the same way, the feed water temperature was set to 22°C, and the salinity was adjusted within the station’s operating range. The SEC with an ERD ranges from 1.55 kwh/m3 to 1.72 kwh/m3 (Figure 8.5), while the SEC without an ERD ranges from 2.27 kwh/m3 to 2.53 kwh/m3. Implementing the ERD at this station can reduce the SEC by an average of 30%, even with the installation of a booster pump, according to the study’s findings. In addition, the SEC formula (7.4) takes into account the energy used by this pump. To evaluate the SEC Fig. 8.1. SEC with T(with ERD) Fig. 8.2. SEC with T(without ERD) Table 8.1. Algorithm Genetic parameters Parameter Number of Variables Population size Mutation Migration Crossover Iteration Value 6 50 for five 0,71 0,41 0,7 155 Definition of variables Xi X1 X2 X3 X4 X5 X6 Variable Y L Qif Of Qf Qp Lower value Lb 0.7 0.003 136 0 490 370 Upper value Ub 0.75 0.013 162 0.05 550 400 Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) 212 M. MOUMNI, M. EL AOUD MOHAMED, M. FATIMA ZAHRA Fig. 8.3. The proposed Design provided by the RO process in the three production lines, several salt water scenarios that closely resemble reality were captured and then simulated on the Matlab platform. It is clear that the inclusion of an ERD has an important effect on energy saving, and according to the 3D multivariate representation (Figure 8.6 and Figure 8.7), it can be observed that the temperature variation has a significant effect on the SEC more than the salinity variation Fig. 8.4. SEC with Cf (without ERD) Fig. 8.5. SEC with Cf (with ERD) Table 8.2. Results of Genetic Algorithm Pressure (bar) 10 14 17 20 22 25 SEC(Kwh/m3) 1,07 1,45 1,73 2,01 2,29 2,48 Y 0,7 0,7 0,7 0,7 0,7 0,7 L 0,03 0,03 0,03 0,03 0,03 0,03 Qif (m3/h) 136,64 136,92 136,16 136,66 136,99 136,65 of (m3/h) 0,04 0,04 0,04 0,04 0,04 0,04 Qf (m3/h) 549,54 549,88 549,85 549,45 549,99 549,58 Qp(m3/h) 412,99 412,95 412,68 412,78 412,99 412,93 Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) SOLVING THE ENERGY CONSUMPTION BARRIER... 213 Fig. 8.6. SEC without ERD Fig. 8.7. SEC with ERD Fig. 8.8. SEC compared to Pf Fig. 8.9. Energy saving and Qp because its range of variation in this station is limited. As concluded in various studies, including the article [37], changes in saltwater parameters (temperature and salinity) cause variations in the feed pressure. The relationship between SEC and feed pressure is illustrated in Figure 8.8, which shows that SEC increases with rising feed pressure. Figure 8.9 depicts the energy savings, defined as the difference in SEC between the RO process with ERD and the process without ERD. The curve indicates that energy savings increase with the produced flow (permeate flow). The reduction rate of SEC with the ERD remains approximately 30%, regardless of changes in permeate flow. Therefore, deploying the ERD is beneficial for RO stations with high energy consumption, especially for saline water within the same variable range of salinity and temperature. This advantage can be further supported by a techno-economic analysis. By incorporating the genetic algorithm-based optimization approach, significant improve- ments in energy efficiency can be achieved for the RO process in water desalination plants. The observed 30% reduction in the SEC by adding an ERD in the RO process highlights the potential benefits of using this approach. Additionally, the genetic algorithm- based optimization can identify the most suitable ERD configuration based on the specific parameters of the RO process, leading to further improvement in energy efficiency. The Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) 214 M. MOUMNI, M. EL AOUD MOHAMED, M. FATIMA ZAHRA Fig. 8.10. Best individual combination of ERD deployment and genetic algorithm-based optimization could provide a more sustainable and cost-effective approach to water desalination. In order to optimize the specific energy consumption (SEC) of the RO process using a genetic algorithm, the formula (7.4)for SEC it’s developed as a function of the parameters that contribute to the choice of the ERD(overflush: OF and leak: L). The equation can be expressed as follows: SEC = (1− (1− Y )(1− L)).Pf .Qf ηHP .Qp + PBP .Qif (Of + 1).ηBP .Qp (8.1) The optimal values of the SEC were obtained by using the Genetic Algorithm on the given formula, considering the feed pressure values ranging from 7 to 25 bar. It is important to note that the feed pressure values were calculated based on the salinity and temperature of the water. The parameters of the applied genetic algorithm are detailed in Table 8.1 and the obtained optimization results are presented in Table 8.2. Upon analyzing the values obtained through the genetic algorithm, it can be concluded that the specific energy consumption (SEC) of the implemented RO process in the studied station ranges from 0,73 to 2,3. This means that, at the maximum feed pressure values considered in this station, energy consumption can be reduced by over 40%. Table 8.2 summarizes the optimal values of the variables xi required for the operation of the RO process in this station, taking into account the parameters that characterize the ERD. For the process to function optimally, an appropriate ERD should have an overflush value of 0,04 and a leak value of 0,03. The Figure 8.10 shows that variables X3,X5 and X6 are the parameters that have the most influence on energy optimization in this RO process. 9. CONCLUSIONS The analysis of specific energy consumption (SEC) in the Moroccan plant’s reverse osmosis desalination process for brackish water reveals significant potential for energy savings. The study shows that by using a genetic algorithm to optimize design and operational parameters, the incorporation of energy recovery devices can reduce SEC by up to 30%. These findings highlight the significance of energy optimization in increasing the feasibility and sustainability of reverse osmosis desalination. Implementing these technologies can make the process more cost-effective and environmentally friendly, encouraging widespread Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) SOLVING THE ENERGY CONSUMPTION BARRIER... 215 adoption in water-scarce regions. Future research will focus on implementing intelligent control systems for the operating parameters to ensure consistent production under optimal conditions. This approach aims to maintain the desired output quality while adapting to varying feed water characteristics, thus ensuring the process remains efficient and reliable. 10. ABBREVIATIONS [ERD] Energy Recoverey Device [RO] Reverse Osmosis [HP] High Pressure [BW] Brackish Water [BWRO] Brackish Water Reverse Osmosis [SWRO] Sea Water Reverse Osmosis [HPP] High Pressure Pump [BP] Booster Pump [LPP] Low Pressure Pump [VFD] Variable Frequency Drive [SEC] Specific Energy Consumed (kwh/m3) [Pf ] Feed Pressure (bar) [Pc] Concentrate Pressure (bar) [Pp] Permeat Pressure (bar) [Qf ] Feed flow rate (m3/h) [Qc] Concentrate flow rate (m3/h) [Qp] Permeat flow rate (m3/h) [Cf ] Feed water salinity (g/l) [Cc] Concentrate salinity (g/l) [Cp] Permeat salinity (g/l) [T ] Temperature (C°) [Es] Specific energy consumed without ERD (kwh/m3) [Ec] Energy lost on the rejection (kwh) [Eerd] Energy recovered by ERD (kwh) [EBP ] Energy consumed by BP (kwh) [ηHP ] HPP efficiency [ηBP ] BP efficiency [ηerd] ERD efficiency [Pif ] ERD input Pressure from LPP (bar) [Pic ] ERD input Pressure from the concentrate (bar) [Pof ] ERD output Pressure from ERD to BP (bar) [Poc] ERD output Pressure from ERD to the rejection (bar) [Qif ] ERD input flow rate from LPP (m3/h) [Qic] ERD input flow rate from the concentrate (m3/h) [Qof ] ERD output flow rate from ERD to BP (m3/h) [Qoc] ERD output flow rate from ERD to the rejection (m3/h) [Cif ] ERD input salinity of flow rate from LPP (g/l) [Cic] ERD input salinity of flow rate from the concentrate (g/l) [Cof ] ERD output salinity of flow rate from ERD to BP(g/l) [Coc] ERD output salinity of flow rate from ERD to the rejection (g/l) [L] ERD Flow Loss [Of ] ERD over flush range Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) 216 M. MOUMNI, M. EL AOUD MOHAMED, M. FATIMA ZAHRA [Mx] Mixing Ratio [∗ST ] Station Values [∗ ∗Mb] Matlab/Simulink Values [Er] Margin of error between ST and Mb [i] Number of ions dissociated in the case of an electrolyte [R] Ideal gas constant R = 8.314 (J .mol−1.k−1) [S] Surface of the membrane (m2) REFERENCES 1. Shah Abedi, M., Hashemi, S. H. & Fazeli, M. (2022). Feasibility of Increasing Water Recovery of Inland Reverse Osmosis Systems and the Use of Reject Brine, Arab J Sci Eng, 47, 6525–6534. doi: 10.1007/s13369-021-06451-4 2. Eltamaly, A. M., Ali, E., Bumazza, M., et al. (2021). Optimal Design of Hybrid Renewable Energy System for a Reverse Osmosis Desalination System in Arar, Saudi Arabia, Arab J Sci Eng, 46, 9879–9897. doi: 10.1007/s13369-021-05645-0 3. Emad, A., Ajbar, A. & Almutaz, I. (2011). Periodic control of a reverse osmosis desalination process, Journal of Process Control, 22, 218–227. doi: 10.1016/j.jprocont.2011.09.001 4. Bartman, A. R., Zhu, A., Christofides, P. D. & Cohen, Y. (2011). Minimizing energy consumption in reverse osmosis membrane desalination using optimization-based control, Journal of Process Control, 20, 1261–1269. doi: 10.1016/j.jprocont.2010.09.004 5. Adda, A., Naceur, W.M. & Abbas, M. (2016). Modelisation et optimisation de la consommation d’energie d’une station de dessalement par procede d’osmose inverse en Algerie, Revue des Energies Renouvelables, 19(2), 157–64. 6. Al-Obaidi, M. A., Alsarayreh, A. A., Al-Hroub, A. M., Alsadaie, S. & Mujtaba, I. M. (2018). Performance analysis of a medium-sized industrial reverse osmosis brackish water desalination plant, Desalination, 443, 272–284. doi: 10.1016/j.desal.2018.06.010. 7. Al-Karaghouli, A. & Kazmerski, L. L. (2013). Energy consumption and water production cost of conventional and renewable-energy-powered desalination processes, Renew. Sust. Energ. Rev., 24, 343–356. 8. Semiat, R. (2008). Energy issues in desalination processes, Environ. Sci. Technol., 42, 8193–8201. 9. Miller, S., Shemer, H. & Semiat, R. (2015). Energy and environmental issues in desalination, Desalination, 366, 2–8. 10. Al-Obaidani, S., Curcio, E., Macedonio, F., di Profio, G., Al-Hinai, H., et al. (2008). Potential of membrane distillation in seawater desalination: thermal efficiency, sensitivity study and cost estimation, J. Membr. Sci., 323, 85–98. 11. Jeong, K., Park, M., Ki, S. J. & Kim, J. H. (2017). A systematic optimization of Internally Staged Design (ISD) for a full-scale reverse osmosis process, J. Membr. Sci., 540, 285–296. 12. Khayet, M. (2013). Solar desalination by membrane distillation: Dispersion in energy consumption analysis and water production costs (a review), Desalination, 308, 89–101. 13. Kim, J., Kim, J. & Hong, S. (2018). Recovery of water and minerals from shale gas produced water by membrane distillation crystallization, Water Res., 129, 447–459. 14. Park, K., Kim, D. Y. & Yang, D. R. (2017). Theoretical analysis of pressure retarded membrane distillation (PRMD) process for simultaneous production of water and electricity, Ind. Eng. Chem. Res., 56, 14888–14901. 15. Al-Karaghouli, A. & Kazmerski, L. L. (2013). Energy consumption and water production cost of conventional and renewable-energy-powered desalination processes, Renew. Sust. Energ. Rev., 24, 343–356. Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) SOLVING THE ENERGY CONSUMPTION BARRIER... 217 16. Miller, D. J., Dreyer, D. R., Bielawski, C. W., Paul, D. R. & Freeman, B. D (2017). Surface modification of water purification membranes, Angew. Chem. Int. Ed., 56, 4662– 4711. 17. Guha, R., Xiong, B., Geitner, M., Moore, T., Wood T.K., et al (2017). Reactive micromixing eliminates fouling and concentration polarization in reverse osmosis membranes, J. Membr. Sci., 542, 8–17. 18. Voutchkov, N. (2018). Energy use for membrane seawater desalination–current status and trends, Desalination, 431, 2–14. 19. Qureshi, B. A. & Zubair, S. M. (2016). Energy-exergy analysis of seawater reverse osmosis plants, Desalination, 385, 138–147. 20. Alsarayreh, A. A., Al-Obaidi, M. A., Al-Hroub, A. M., Patel, R. & Mujtaba, I. M. (2019). Evaluation and minimisation of energy consumption in a medium-scale reverse osmosis brackish water desalination plant, Journal of Cleaner Production, 248, 119220. doi: 10.1016/j.jclepro.2019.119220 21. Anqi, A. E., Alkhamis, N. & Oztekin, A. (2015). Numerical simulation of brackish water desalination by a reverse osmosis membrane, Desalination, 369, 156–164. doi: 10.1016/j.desal.2015.05.007 22. Ruiz-Garcia, A., Nuez, I., Carrascosa-Chisvert, M. D. & Santana, J. J. (2020).Sim- ulations of BWRO systems under different feedwater characteristics. Analysis of operation windows and optimal operating points, Desalination, 491, 114582. doi: 10.1016/j.desal.2020.114582 23. Eltamaly, A. M., Ali, E., Bumazza, M., et al. (2021). Optimal Design of Hybrid Renewable Energy System for a Reverse Osmosis Desalination System in Arar, Saudi Arabia, Arab J Sci Eng, 46, 9879–9897. doi: 10.1007/s13369-021-05645-0 24. Xevgenos, D., Moustakas, K., Malamis, D. & Loizidou, M. (2016). An overview on desalination and sustainability: renewable energy-driven desalination and brine management, Desalination and Water Treatment, 57(5), 2304–2314. doi: 10.1080/19443994.2014.984927 25. Fethi, K. (2003). Optimization of energy consumption in the 3300 m3/d RO Kerkennah plant, Desalination, 157(1–3), 145–149. doi: 10.1016/S0011-9164(03)00394-1. 26. Al-Obaidi, M. A., Kara-Zaitri, C. & Mujtaba, I. M. (2018). Significant energy savings by optimising membrane design in the multi-stage reverse osmosis wastewater treatment process, Environmental Science: Water Research and Technology, 4(3), 449–460. doi: 10.1039/C7EW00455A 27. Alanood, A. Alsarayreh, A., Mudhar, A., Al-Obaidi, B., Shekhah K., et al. (2021). Performance evaluation of a medium-scale industrial reverse osmosis brackish water desalination plant with different brands of membranes. A simulation study, Desalination, 503, 114927. doi: 10.1016/j.desal.2020.114927. 28. Qiu, T. & Davies, P.A. (2012).Comparison of Configurations for High-Recovery Inland Desalination Systems, Water, 4(3), 690–706. doi: 10.3390/w4030690 29. Wei, Q. J., McGovern, R. K. & Lienhard, J. H. V. (2017). Saving energy with an optimized two-stage reverse osmosis system, Environmental Science: Water Research & Technology, 4. doi: 10.1039/c7ew00069c 30. Al-Huwaidi, J. S., Al-Obaidi, M., Jarullah, A. T., Kara-Zaı̈tri, C. & Mujtaba, I. M. (2021). Modeling and simulation of a hybrid system of trickle bed reactor and multistage reverse osmosis process for the removal of phenol from wastewater, Computers and Chemical Engineering, 153, 107452. doi: 10.1016/j.compchemeng.2021.107452 31. Kim, J., Park, K. (2020). Optimization of two-stage seawater reverse osmosis membrane processes with practical design aspects for improving energy efficiency, Journal of Membrane Science, 601, 117889. doi: 10.1016/j.memsci.2020.117889 32. Li., M. (2020). Optimization and Plant Validation of BWRO Operation, in Analysis and Design of Membrane Processes: A Systems Approach. Melville, NY: AIP Publishing. doi: 10.1063/9780735421790-005 Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) 218 M. MOUMNI, M. EL AOUD MOHAMED, M. FATIMA ZAHRA 33. Li, M. & Noh, B. (2012). Validation of model-based optimization of brackish water reverse osmosis (BWRO) plant operation, Desalination, 304, 20–24. doi: 10.1016/j.desal.2012.07.029 34. Sanchez, J. M. S., Castillo, N. S. & Castillo, R. S. (2007). Mathematical Model for Isobaric Energy Recovery Devices, Proc. of IDA World Congress-Maspalomas (Gran Canaria, Spain), 21–26. 35. Drak, A., Adato, M. (2014). Energy recovery consideration in brackish water desalination, Desalination, 339, 34–39. doi: 10.1016/j.desal.2014.02.008 36. Moumni, M. & Massour, M. (2022). Fuzzy logic control of a brackish water reverse osmosis desalination process, Computers and Chemical Engineering, 167, 108026. doi: 10.1016/j.compchemeng.2022.108026. 37. Moumni, M. & Massour, M. (2021). Modeling of reverse osmosis process at a brackish water desalination station, Proc. of 7th International Conference on Optimization and Applications (ICOA) (Wolfenbüttel, Germany). doi: 10.1109/ICOA51614.2021.9442632. 38. Arun, J., Vasanthi, D. (2019).Dynamic simulation of the reverse osmosis process for seawater using LabVIEW and an analysis of the process performance, Computers and Chemical Engineering, 121, 294–305. doi: 10.1016/j.compchemeng.2018.11.001 39. Lance, R., Pinto, L. & Pinto, J. M. (2015). Energy Recovery in Desalination: Returning Alternative Water Supplies to Consideration, Florida Water Resources Journal. 40. Heinz, L. (2022). Reverse Osmosis Seawater Desalination Volume 2: Planning, Process Design and Engineering. A Manual for Study and Practice. Berlin, Germany: Springer Nature. 41. Huang, B., Pu, K., Wu, P., Wu, D. & Leng, J. (2020). Design, Selection and Application of Energy Recovery Device in Seawater Desalination: A Review, Energies, 13, 4150. doi:10.3390/en13164150 Copyright © 2024 ASSA. Adv Syst Sci Appl (2024) Introduction Description of BWRO desalination plant Materiels and Methods Evaluation of SEC at BWRO Modeling ERD Flow Calculation Salinity Calculation Pressure Calculation Booster pump The SEC of BWRO with ERD RESULTS AND DISCUSSIONS CONCLUSIONS Abbreviations