Assessing RO and NF Desalination Technologies for Irrigation-Grade Water

Abstract

In this work, the performance of a Reverse Osmosis (RO) process using different types of reverse osmosis (RO) and nanofiltration (NF) membranes is evaluated for brackish water desalination for producing irrigation-grade water. The proposed desalination system is a single-stage system, where three types of RO and two NF membranes were examined. The different desalination systems were simulated using ROSA72 software. In order to validate the theoretical model, the results obtained from the simulation were compared to those obtained from the experiment conducted in this work. The El-Moghra aquifer of Egypt is considered the test bed due to a considerable amount of data being available for this aquifer. The El-Moghra aquifer has 79 wells, and the available water data, when checked against several quality parameters, show that none of the investigated wells are suitable for direct irrigation without treatment due to problems of salinity, the sodium adsorption ratio, and low water quality according to the irrigation water quality index values. The obtained results show that nanofiltration membranes exhibited superior energy efficiency compared to reverse osmosis membranes. However, what sets the nanofiltration membranes apart is their ability to elevate water quality in 89.9% of the total investigated wells to an acceptable level for agricultural purposes. This underscores the nanofiltration membranes as a highly effective alternative to reverse osmosis membranes, demonstrating the capability to produce water suitable for irrigation while concurrently reducing operational costs due to the lower energy consumption in nanofiltration-based systems.

1. Introduction

The lack of freshwater resources and concerns about the safety of irrigation and drinking water have turned into significant barriers to long-term socioeconomic development in many countries. For example, due to the incremental demand and the depletion of traditional water resources over several years, Egypt experiences water supply issues, just like many other countries around the world. Since 1959, the Nile water pact between Egypt and Sudan allotted Egypt 55.5 billion cubic meters of water yearly while Egypt’s requirement is around 93.7% of its water supply.

Many countries, including Egypt, depend on desalination technologies to desalinate the water resources available in a country, such as brackish water or seawater, to meet the irrigation and drinking water demand. Brackish water has a salt concentration of less than 10,000 mg/L, which makes it more saline than freshwater (salt content of 1000 mg/L) but less saline than seawater (salt content between 25,000 and 45,000 mg/L) [1,2]. Desalination of brackish water requires less energy than seawater desalination for making freshwater. However, brackish water desalination still has environmental impacts, such as brine disposal and the potential for groundwater depletion. Therefore, it is crucial to carefully consider the environmental effects of brackish water desalination projects and implement appropriate mitigation measures [3].

The two most often used desalination techniques worldwide are reverse osmosis (RO) and multi-stage flash (MSF) [4]. In order to counteract the osmotic pressure, pressure is applied across a partially permeable membrane to remove ions, unwanted chemicals, and larger particles from saltwater using the RO water purification technique [5]. RO is the most prevalent type of desalination since it accounts for around 65% of all desalination facilities worldwide, placing it at the top in terms of ubiquity [6]. RO and NF are similar to other membrane filtration techniques like microfiltration (MF) and ultrafiltration (UF), where their tiny pores remove even smaller molecules. The pore widths of nanofiltration membranes, which range from 1 to 10 nanometers, are smaller than those of MF (100 to 10,000 nanometers) and UF (10 to 100 nanometers), but these widths are larger than those of reverse osmosis membranes (less than 1 nanometers). Most membranes are constructed from polymer thin films [7]. RO and NF are the primary techniques used to treat brackish water for drinking and irrigation [8,9,10]. The above sentences describe the relative pore size of different types of filters, including the two types of filters (RO and NF) used in this research. The reason for mentioning the sizes of MF and UF filters is to demonstrate that these filters will not be suitable for treating brackish water for the purpose of irrigation.

Several studies have been conducted comparing the effectiveness of NF and RO membranes, as well as hybrid NF–RO systems, for brackish water desalination. One such study by Chang et al. [11] investigated the impact of desalination on the Sodium Adsorption Ratio (SAR), a key factor in determining water quality for agriculture. The study found that while RO rejects both monovalent and divalent ions, resulting in a reduction in SAR value, NF only rejects divalent ions, and therefore, has a lesser effect on SAR. However, the study also noted that NF can negatively affect soil permeability, highlighting the importance of checking SAR values when using this method. Other research has compared the retention of nutrients in irrigation water by RO and NF. Ghermandi and Messalem [12] found that NF is more effective than RO at desalinating water with higher nutrient concentrations, leading to an 18% boost in biomass. Conversely, El Azhar et al. [13] compared the retention of sodium, chlorine, and total dissolved solids by RO and NF membranes in Monasra City, Morocco, and found that RO membranes retained these ions at a higher rate than NF membranes. One of the main issues with using RO for irrigation is the low concentration of magnesium ions in the resulting water, which is crucial for agriculture. To address this, Birnhack et al. [14] developed an integrated RO and NF system that increased the concentration of magnesium ions.

Several RO and NF membranes were investigated by Ghaffour et al. [15]. From the economic perspective, the BWRO-30 membrane has a lower operating cost than the SWRO-30 membrane. Bhojwani et al. [16] also concluded that the NF90 membrane costs less than the SWRO-30 membrane. Elazhar et al. [17] examined the prices and technical performance of RO and NF technologies. Their results show that NF had lower operating costs, higher productivity, and higher capital costs compared to reverse osmosis technology. As a result, NF technology is less expensive than RO technology for desalinating low-salinity groundwater because it uses fewer membranes, uses less energy, and is more productive. As NF requires less hydraulic pressure than RO and uses less energy as a result, its running costs are cheaper [12]. The membrane’s pores become clogged with age, which raises energy consumption and operational costs. In this regard, Ruiz-García and Ruiz-Saavedra [18] compared chemical cleaning and membrane change over a lengthy period of time (10 years), concluding that the former is the more cost-effective option.

In this work, the performance of different types of RO and NF membrane-based desalination technologies is evaluated to produce irrigation-grade water from brackish water based on real data from the El Moghra aquifer, Egypt. A single-stage process is considered in this work, where three types of RO and two types of NF membranes are examined. The different desalination systems are simulated using ROSA software. In order to validate the theoretical model, the results obtained from the simulation are compared to those obtained from the experiment conducted in this work. The performance of these membranes is evaluated and compared in terms of the quality of permeated water for agricultural uses and specific energy consumption (SEC). As we have access to field data for the El Moghra aquifer of Egypt, we have selected this aquifer for our experimental work in this study. The methodology applied in this study will be applicable to any brackish water system (of course, the field data will be different).

2. Methodology

In this section, the data used in the study and the indicators used in evaluating water quality for agricultural purposes will be clarified. Three indicators are considered: salinity hazard, sodium hazard, and irrigation water quality index (IWQI). The specifications of the membranes used in the proposed desalination system and the operating conditions of the system will also be mentioned. The method used to simulate the proposed membranes will also be clarified to determine the quality of the produced water. The operating conditions such as the required pressure and the consumed energy are also investigated.

2.1. Irrigation Water Quality

Irrigation water quality refers to the chemical, physical, and biological characteristics of water that is used for irrigation. The quality of irrigation water can have a significant impact on the growth and development of crops, as well as the health of soil and productivity. The type and quantity of salts dissolved in irrigation water have a significant impact on its quality [19]. It is well acknowledged that the nature and severity of irrigation water quality issues varies depending on several variables, such as the type of soil and crop, the local climate, and the farmer using the water. However, it is now widely acknowledged that these issues can be divided into the following four main categories: (a) salinity hazard, (b) infiltration and permeability hazard, (c) specific ion toxicity hazard, and (d) other issues [20]. Among the most widely used important indicators to evaluate irrigation water are salinity (EC), sodium adsorption ratio (SAR), and irrigation water quality index (IWQI). The following will explain these indicators.

2.2. Salinity and Sodium Hazards (EC and SAR)

When brackish water is used for irrigation, salinization frequently results, which can have detrimental effects on the ecosystem. The degree of salinity threat could be determined by how well water conducts an electric current, which is the water quality standard that has the greatest impact on agricultural output. Both an electrical conductivity (EC) test and a total dissolved solids (TDS) analysis could be used to determine the salinity of water because conductance is strongly dependent on the total dissolved ionic solids [21]. High salt concentrations in irrigation water can restrict the kinds of crops that can be produced, have a detrimental impact on crop germination and yields, and deteriorate the soil. Additionally, when the osmotic pressure of the soil solution is high due to increased salt concentration from irrigation water, the water potential of the soil decreases, making it more difficult for plant roots to absorb water. This can lead to water stress in plants, which can affect plant growth, reduce crop yields, and even lead to plant death. According to electrical conductivity (EC), irrigation water may generally be divided into four groups, as indicated in Figure 1 and

Table 1. Salinity and sodium hazards classification based on the USSL staff diagram.

	Water Class	Hazard Class
Salinity hazard	C1 (excellent)	Low salinity
	C2 (good)	Medium salinity
	C3 (bad)	High salinity
	C4 (very bad)	Very high salinity
Sodium hazard	S1 (excellent)	Low sodium
	S2 (good)	Medium sodium
	S3 (bad)	High sodium
	S4 (very bad)	Very high sodium

.

High levels of sodium in irrigation water can have negative effects on soil, including soil compaction and reduced water infiltration. Sodium can also interfere with the uptake of other nutrients by plants, leading to reduced crop growth and productivity. Utilizing the Sodium Adsorption Ratio (SAR) is the most popular method for estimating the sodium hazard of irrigation water (SAR). Equation (1) is used to calculate the SAR [22]:

$$\mathrm{SAR}={\displaystyle \frac{{\mathrm{Na}}^{+}}{\sqrt{\frac{{\mathrm{Ca}}^{+2}+{\mathrm{Mg}}^{+2}}{2}}}}$$

(1)

where SAR is represented as (milli moles per liter) 0.5 or (mmoles/L) 0.5, and Na⁺, Ca⁺², and Mg⁺² are the concentrations of sodium, calcium, and magnesium ions in irrigation water, respectively, expressed in milli equivalents per liter (meq/L). In general, irrigation water with a SAR value above 9.00 is considered to be of poor quality for irrigation, as it can have negative impacts on soil and plant health [23]. Figure 1 can be used to evaluate the salinity and sodium hazards of irrigation water. The sodium hazard (SAR) is plotted on the y-axis versus the salinity hazard (EC) on the x-axis in this graph (Logarithmic scale). This figure was first created by USSL staff [24] and modified by Shahid and Mahmoudi (2014) [25]. There are 16 categories on the graph that can be used to categorize the water quality based on EC and SAR values, each category is shown in Figure 1 and Table 1.

Figure 1. USSL staff diagram for classification of irrigation water (modified by Shahid and Mahmoudi) [25].

Figure 1. USSL staff diagram for classification of irrigation water (modified by Shahid and Mahmoudi) [25].

2.3. Specific Ion Toxicity

2.3.1. Sodium Toxicity

Detecting sodium toxicity in plants can be challenging, as its symptoms are similar to those caused by other factors such as drought or nutrient deficiencies. Symptoms of sodium toxicity, such as leaf burn, scorch, and dead tissue along the leaf edges, can also be the result of other factors. Therefore, it is important to confirm the presence of sodium toxicity using other methods. One effective way to confirm sodium toxicity is by testing the irrigation water for its sodium content and comparing it to recommended levels. High levels of sodium in irrigation water can serve as an indicator of sodium toxicity in plants. However, it is important to note that the severity of the toxic effects of sodium on plants depends on various factors, such as the plant species and soil type being used. In addition to testing the irrigation water, visual symptoms of sodium toxicity can also be used to detect potential problems. Plants that are showing signs of sodium toxicity may have stunted growth, reduced yields, and a general decline in health. Soil testing can also be used to determine the level of sodium in the soil, which can help to identify potential sources of sodium toxicity. To prevent sodium toxicity in plants, it is important to manage the sodium content of irrigation water and soil. This can include selecting irrigation water sources with lower sodium content, using irrigation methods that reduce the buildup of sodium in the soil, and implementing soil amendments that can help to leach out excess sodium from the soil. By managing irrigation water’s sodium content and monitoring for symptoms of sodium toxicity, farmers can ensure healthy soil and optimal crop yields [20].

2.3.2. Chloride Toxicity

Chloride toxicity is a significant concern in irrigation water quality. Chloride is a naturally occurring ion that is found in many water sources, but high levels of chloride in irrigation water can be detrimental to plant growth and health. When chloride levels in irrigation water exceed the tolerance threshold of plants, it can cause leaf burn, stunted growth, and reduced yields. Additionally, high chloride levels can lead to soil salinization, which can further exacerbate plant stress and reduce crop productivity. Chloride toxicity is particularly problematic in arid and semi-arid regions where water resources are limited, as irrigation water sources often have higher chloride concentrations due to evaporation and concentration. Effective management of chloride levels in irrigation water is crucial for crop production, and it involves careful monitoring of water quality, selection of appropriate irrigation methods, and implementation of appropriate soil and water management practices [21].

2.3.3. Bicarbonate Toxicity

Bicarbonate toxicity is another important factor to consider in irrigation water quality. Bicarbonate is a common constituent of irrigation water, and its concentration can vary depending on the source of the water. High bicarbonate levels in irrigation water can lead to several problems, including soil alkalinity, reduced nutrient availability, and reduced crop growth and yield. Bicarbonate ions can react with calcium ions in the soil, forming calcium carbonate, which can reduce soil permeability and cause waterlogging. This can further lead to increased soil salinity, which can exacerbate plant stress and reduce crop productivity. Additionally, high bicarbonate levels can interfere with the uptake of essential nutrients by plants, such as iron and zinc, leading to nutrient deficiencies and reduced crop quality. Therefore, it is important to monitor bicarbonate levels in irrigation water and take appropriate measures to manage its concentration. This may involve using acidification treatments to reduce bicarbonate levels or selecting crops that are more tolerant to high bicarbonate concentrations. Proper management of bicarbonate levels in irrigation water is crucial for maintaining healthy soil and crop growth, and it can ultimately lead to improved crop yields and quality [26].

2.4. The Model of Irrigation Water Quality Index (IWQI)

Due to the multiplicity of irrigation water quality indicators (EC, SAR, and toxicity of sodium, chloride, and bicarbonate), it was necessary to use one indicator that combines the most important indicators into a single number that is easy to use to determine water suitability for irrigation purposes. This model was created by Meireles et al. [27]. In their study, Meireles et al. developed a model for assessing irrigation water quality. The first step involved identifying the parameters that were considered most relevant for irrigation use. In the second step, the researchers established a definition of quality measurement values (qi) and aggregation weights (wi). The quality measurement values (qi) were estimated based on the value of each parameter, using both the irrigation water quality parameters proposed by the University of California Committee of Consultants (UCCC) and the criteria established by Ayres and Westcot [19]. The water quality parameters were represented by a non-dimensional number, with higher values indicating better water quality. The higher the value, the better the water quality, according to a non-dimensional number used to represent water quality standards. The irrigation water quality index (IWQI) approach is commonly used in evaluation since it yields a single value utilizing electrical conductivity (EC), sodium adsorption ratio (SAR), and concentrations of sodium (Na), bicarbonate (HCO₃), and chloride (Cl). The equation below is used to calculate the indicator qi [27].

$$\mathrm{IWQI}={{\displaystyle \sum }}_{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{q}}_{\mathrm{i}}{\mathrm{w}}_{\mathrm{i}}$$

(2)

$${\mathrm{q}}_{\mathrm{i}}={\mathrm{q}}_{\mathrm{imax}}-\left\{{\displaystyle \frac{\left[\left({\mathrm{x}}_{\mathrm{ij}}-{\mathrm{x}}_{\inf }\right){\ast \mathrm{q}}_{\mathrm{iamp}}\right]}{{\mathrm{x}}_{\mathrm{amp}}}}\right\}$$

(3)

where q_imax is the class’s maximum value of qi, x_ij is the parameter’s observed value, x_inf is the value that corresponds to the class’s lower limit to which the parameter belongs, q_iamp is the class’s amplitude, and x_amp is the class’s amplitude to which the parameter belongs. The maximum value found in the physical–chemical and chemical analyses of the water samples was taken into consideration while evaluating x_amp, the last class of each parameter. The Irrigation Water Quality Index (IWQI) developed by Meireles et al. [27] assigns specific weights to various parameters to assess water suitability for irrigation. Electrical Conductivity (EC) and Sodium (Na), which hold a weight of 0.211 and 0.204, respectively, reflect their critical roles in salinity and soil permeability. Sodium Adsorption Ratio (SAR) and Chloride (Cl⁻) are weighted at 0.189 and 0.194, respectively, while Bicarbonate (HCO₃⁻) is weighted at 0.204. These weights are used to calculate the IWQI, offering a comprehensive measure of water quality for irrigation. The normalized wi values were set up so that their sum was one. The Irrigation Water Quality Index (IWQI), as proposed by Meireles et al. [27], categorizes water suitability for irrigation based on its chemical characteristics, considering soil properties and plant tolerance. Water with an IWQI above 85 is deemed excellent (no restriction) and suitable for all soil types and crops, including sensitive ones, with no significant adverse effects. Good quality water (IWQI 70–85) is appropriate for most soils and moderately sensitive crops, though monitoring is recommended (low restriction). Moderate quality water (IWQI 55–70) may cause some salinity or sodicity issues, particularly in poorly drained soils, and is best used for moderately tolerant crops (moderate restriction). Poor quality water (IWQI 40–55) poses significant risks, suitable only for tolerant crops and requiring careful soil and water management (high restriction). Finally, water with an IWQI below 40 is considered unsuitable for irrigation, potentially leading to severe soil degradation, and is not recommended for most crops (severe restriction).

3. Experimental Data

To meet the water demand of Egypt, about 2.1 billion cubic meters of groundwater are added to Egypt’s water resources each year [28]. About 0.35 billion cubic meters per year are contributed through the desalination of salty and brackish water [28]. El-Moghra is one of the six primary groundwater aquifers in Egypt: Nile, Nubian Sandstone, El-Moghra, Fissured, and Hardrock, Coastline [29]. As mentioned earlier, the El-Moghra aquifer has been selected for our experimental part of the work due to the fact that a number of earlier studies can be found on this aquifer, which are briefly discussed below.

Abdel Mogith et al. [30] examined roughly 200 wells in the El-Moghra aquifer region to identify their physical, chemical, and hydraulic properties. The study also aimed to assess the aquifer potential and investigate any lateral and vertical changes. To ensure the best agricultural investment in the El-Moghra aquifer, it is recommended to focus on areas east of longitude 30°00′00″ E, where the water level is above −27 m and the water salinity is less than 2000 ppm [30]. The possibility of using groundwater from the El Moghra aquifer for irrigation was evaluated by Eltrabaly and Moghazy [31]. They found that the water is not suitable for irrigation in general, and that water with high levels of salinity (above 9000 micrograms/cm) are treated with desalination devices before they are suitable for irrigation. Ismail et al. [32] investigated the impact of groundwater irrigation on onion yield in Wadi El Natrun. Their results reveal that irrigation productivity is increased by lowering the salinity of the water prior to irrigation. They used RO technology to reduce the salinity of irrigation water.

The El-Moghra region is located in the northern part of the Western Desert in Egypt, covering an area of approximately 966 square kilometers. It is situated 40 km south of the city of El-Hamam in Matrouh Governorate, and 40 km west of the city of El-Alamein [31]. The region experiences hot summers and warm winters, with an average annual temperature ranging from 14.20 °C to 24.40 °C degrees Celsius. Rainfall is scarce, with an average annual precipitation of 9.9 mm. The rate of evaporation varies throughout the year, reaching a maximum of 14 mm per day in the summer and a minimum of 5.1 mm per day in the winter (https://www.weather-atlas.com/en/egypt/el-dabaa-climate#temperature accessed on 25 June 2023). The Moghra aquifer is divided into two parts based on recharge: a small, renewable portion and a larger, fossil portion (non-renewable) [33]. Water salinity near the EL-Qattara Depression is 31,000 mg/L and near Wadi El-Farigh is 290 mg/L [31]. The Egyptian government has taken a keen interest in the region due to its inclusion in a significant agricultural reclamation project. This project, initiated by the government, has an ambitious goal of reclaiming one and a half million acres of land. The project is aimed at expanding the agricultural sector in Egypt by implementing advanced farming techniques and irrigation systems. This initiative is expected to stimulate economic growth and generate employment opportunities in the region [31].

Data

In this study, the data used for the water quality assessment in the El Moghra region of Egypt were collected by Eltarabily and Moghazy [31]. Their research mainly concentrated on evaluating the groundwater’s purity, determining its chemical properties, and determining whether or not it was safe to utilize for irrigation. They collected data on the chemical properties of groundwater in 79 wells but studied in 46 wells only (Figure 2). The data collected by them includes electrical conductivity (EC) values measured in micro-Siemens per centimeter (μs/cm), as well as the concentration values of six different salts in milligrams per liter (mg/L or ppm): Sodium (Na), Calcium (Ca), Magnesium (Mg), Bicarbonate (HCO₃), and Chlorine (Cl). These data will be used to conduct a comprehensive water quality assessment for agricultural purposes, as shown in Table S1 [31]. Also, the table contains the maximum, minimum and average values for each variable.

4. Desalination System and Simulation

Five different membranes were used in this study, three of which are RO membranes and the other two are NF membranes. The five membranes used are similar in some properties, each with a maximum pressure of 41 bar and maximum temperature of 45 °C (https://www.dupont.com/products, accessed on 25 June 2023).

Table 2. The characteristics of the membranes.

System No.	Membrane Name	Membrane Type	Area (m2)	Stabilized Salt Rejection
M1	BW30-4040 (RO)	Polyamide Thin-Film Composite	7.2	99.5
M2	LP-4040 (RO)	Polyamide Thin-Film Composite	7.2	99.2
M3	TW30-4040 (RO)	Polyamide Thin-Film Composite	7.2	99.5
M4	NF270-4040 (NF)	Poly piperazine Thin-Film Composite	7.6	97.0
M5	NF90-4040 (NF)	Poly piperazine Thin-Film Composite	7.6	98.7

shows some of the important properties of the membranes used in the study, which affect the performance of the membranes, which are the effective area, stabilized salt rejection, and the membrane material. However, they differ in productivity and salt rejection ability, thus, the produced water quality varies from membrane to membrane. For all membranes, a constant value of productivity of one cubic meter per hour (m³/h) was imposed. The proposed system is a single-stage system, either based on RO membranes (M1, M2, and M3) or NF membranes (M4 and M5). To make the system more efficient, numerous recovery values have been tried for it to achieve the highest productivity. Recovery (R) is the ratio between the permeated outflow and the feed water flow. After several attempts, the recovery was set to 18% for the system for all membranes. Figure 3 shows schematic diagrams of the proposed single-stage system.

ROSA is a software tool used for designing reverse osmosis (RO) and nanofiltration (NF) systems. It was developed by DuPont Water Solutions (formerly Dow Water Solutions) and is used in the water treatment industry to design RO plants that meet specific water treatment specifications. The software is specifically designed to help users make the right choice for their water treatment needs and to monitor the performance of FilmTec and DesaliTec RO systems after they have been installed. It is a widely used industry-standard software for RO system design (https://www.dupont.com/water/resources/rosa-software.html, accessed on 25 June 2023).

5. Experiments and Validation of the Model

A set of laboratory experiments were conducted to ascertain the program’s ability to model a single-stage desalination system. A tap water membrane (TW30-18-12) was used, which is the miniature size of the membrane (M3, TW30-4040). A high-pressure pump with a maximum pressure and discharge of 9 bar and 3 L/min was used, respectively. Three tanks were used, one of them for feeding water with a capacity of 30 L. The other two tanks are for brine and permeated water, with a capacity of 25 and 10 L, respectively. No membranes or filters were used before the main RO membrane to protect it so as not to affect the results. A controller valve was fixed at the outlet of the brine to control the pressure. Experiments were carried out at a room temperature of 25 °C. Two pressure gauges are fixed at the inlet of the pressure chamber and the outlet of the brine solution.

In order to confirm the ability of the software used to accurately simulate the desalination systems based on reverse osmosis, three sets of experiments were performed in the laboratory. Laboratory measurements were compared with the results of the program (ROSA software) in all groups to assess the ability of the program to simulate the desalination systems based on RO. The first set of experiments used different values of feed water salinity and feed hydraulic pressure. Six values of salinity were used in the first group, which were 250, 500, 750, 1000, 1250, and 1500 ppm of Sodium chloride (NaCl) solution at two values of feed pressure (2 and 4 bar). The second group used a constant value of feed water salinity of 1500 ppm of sodium chloride solution but with six values of feed pressure (1, 2, 3, 4, 5 and 6 bar). The third group used feed water formed in the laboratory to match water samples taken from three wells in the study area. The three wells with the lowest saline (W47, W54, W65) were chosen to perform the experiments. In all experiment groups, the test was carried out as follows. The pressure is adjusted first using the brine outlet valve, then the experiment begins after ensuring that the tanks of both permeated and brine water are completely emptied. Pressure and temperature are monitored during operation until 25 L of water is pumped from the feed tank, which has a total volume of 30 L. During operation, the feed tank is not completely emptied so as not to change the operating conditions. The time elapsed for infusion of 25 L is recorded, and the volume of both permeated and brine water is measured using a sensitive weighing scale. The change in water density resulting from the change in salinity values is ignored and the weight in kilograms is considered to be numerically equal to the volume of water (unit weight of water is 1 kg/L). The salinity of both permeated and brine water was measured in the first and second sets of experiments, while all concentrations of sodium ions, chloride, bicarbonate, calcium, and magnesium were measured in the third group of experiments.

In the following, the results of modeling using the ROSA software are compared with the experimental results of three sets of experiments. Table S2 shows the values of productivity per square meter of membrane area (flux) and salt rejection for the first and second sets of experiments. Flux refers to the rate at which water passes through the desalination membrane per unit area. It is a critical parameter that indicates the efficiency and performance of the desalination process. The flux is typically measured in units of volume per unit area per unit time, such as liters per square meter per hour (LMH) or cubic meters per square meter per day (m³/m²/day). The same values inferred by the modeling and the relative error for each experiment are also shown. The table demonstrates the high capability of ROSA software in modeling reverse osmosis systems and confirms its reliability in predicting the performance of a specific reverse osmosis system. Additionally, it indicates that the program’s ability to infer salt rejection was very high, with a maximum relative error of 2.05%. However, its ability to infer system productivity was moderate but with an acceptable relative error, and the maximum relative error was 15.79% when inferring system productivity.

Table S3 presents the rejection values for each ion calculated from simulations and experiments, along with the mean error between the two values. The table confirms that the program has a high degree of accuracy in predicting the rejection of each ion, with a maximum relative error of 3.86%. This finding supports the possibility of relying on the program to simulate RO and NF systems and accurately predict the rejection of each ion. The relative error in Tables S2 and S3 was calculated using the following equation.

$$\mathrm{relative}\mathrm{error}={\displaystyle \frac{\mathrm{ROSA}-\mathrm{EXP}}{\mathrm{ROSA}}}\times 100$$

where ROSA is the result of simulation using ROSA software, and EXP is the result of the experimental work.

6. Results and Discussion

In the study area, an evaluation of groundwater quality was conducted in 79 wells using previously discussed indicators, with the aim of determining their suitability for agricultural purposes. Subsequently, the water produced by five different desalination membranes was evaluated to determine the best membrane in terms of water quality for agricultural use. The next step included studying the effect of blending source water with desalinated water on water quality for agricultural purposes. The goal was to determine the optimal blending ratio that would produce water suitable for agricultural use. This was followed by a comparison between the proposed membranes in terms of energy consumption, which is one of the factors that effectively affects the cost of operating desalination systems.

6.1. Assessment of Source Water Quality

The groundwater of all well classifications based on the SAR and EC values of their water and their suitability for agricultural use is shown in Figure 4 using the USSL staff diagram. It can be observed from this figure that most wells are situated in zones C4-S3 and C4-S4, which are characterized by poor water quality and are not suitable for agricultural purposes. Figure 5 presents the percentage of wells for each water quality classification for agricultural uses based on all indicators (EC, SAR, and IWQI). According to Figure 5, all wells are categorized as having very high salinity (C4-very poor water quality) and are therefore unsuitable for direct irrigation prior to treatment. In terms of the sodium adsorption ratio (SAR), around 97.5% of the wells are unsuitable for direct irrigation, as 74.7% of the wells had a very high SAR value, while it was high in 22.8% of the wells. Only 2.5% of the wells were classified as suitable for direct irrigation without treatment based on SAR value, as shown in

Table 3. The number of wells and their percentage for each classification for agricultural uses based on EC and SAR values.

Parameter	Hazard Class	Water Class	Number of Wells	Percentage of Wells Number (%)
Salinity (EC)	Low salinity	(C1) Excellent	0	0.0
	Medium salinity	(C2) Good	0	0.0
	High salinity	(C3) Bad	0	0
	Very high salinity	(C4) Very bad	79	100
Sodium Adsorption Ratio (SAR)	Low sodium	(S1) Excellent	0	0
	Medium sodium	(S2) Good	2	2.5
	High sodium	(S3) Bad	21	26.5
	Very high sodium	(S4) Very bad	56	71.0

.

In the study area, the quality of irrigation water was assessed using the irrigation water quality index (IWQI), which considers several parameters, including electrical conductivity (EC), sodium adsorption ratio (SAR), and the concentration of sodium, chlorine, and bicarbonate ions. The results of the IWQI calculations are presented in Table S4, which shows the IWQI values and classifications for each well individually. The table indicates that out of the total of wells evaluated, 52 have very poor water quality while 27 have poor water quality. This suggests that not all wells in the study area are suitable for direct irrigation without proper treatment. The three indicators used in the evaluation process confirmed that the groundwater in the study area is not appropriate for irrigation without appropriate treatment. Therefore, this study emphasizes the need to treat the groundwater before using it for agricultural purposes. It is essential to ensure that the water used for irrigation does not negatively impact the soil, crops, or human health.

6.2. Permeated Water Assessment

The groundwater in the study area, as determined from an assessment of 79 wells, is not appropriate for direct agricultural use due to high levels of salinity, the sodium adsorption ratio, and a poor irrigation water quality index. To address this issue, the study proposes five desalination membranes that are used in a single-stage desalination system, including reverse osmosis and nanofiltration membranes. The quality of the desalinated water will be assessed using the three indicators previously mentioned. The aim is to determine the effectiveness of each membrane in addressing the issues with the source water and its suitability for agricultural purposes. Overall, the study seeks to provide a viable solution to the issue of unsuitable groundwater in the study area, enabling its use for agricultural purposes. Finally, the membranes will be compared based on energy consumption, which is the most influential factor in operating cost and environmental impact.

Table S5 provides information on the classification of water in wells before and after desalination using the five different membranes. The first and third membranes (M1 and M3), which are RO membranes, demonstrated a high ability to remove salts. Since most of the wells had a salinity level of less than 100 parts per million (this was referred to in Table S5 of the classification C0), the water produced from these two membranes (M1 and M3) needed to have some salts added back in to make it suitable for plant consumption. For most wells, the water produced by these two membranes may not be suitable for agricultural purposes. Conversely, the fourth NF membrane (M4) was the least effective in removing salts, resulting in high levels of salt in the water produced, making it unsuitable for agricultural use. The table demonstrates that the M2 (RO) and M5 (NF) membranes were effective in producing water suitable for irrigation, with most wells classified as being of good or very good quality. RO membranes are recognized for their impressive ability to effectively eliminate monovalent and divalent ions, as well as salts [34,35]. RO membranes consist of a composite structure that includes a thin selective layer, an underlying support layer, and a nonwoven fabric. The thin selective layer (100–300 nm) is solely responsible for the membrane’s selectivity, while the other two layers provide mechanical strength. The rejection of ions occurs through the solution–diffusion mechanism within the selective layer [36]. Desalination applications have extensively employed RO membranes, attaining a rejection rate of salts and ions of up to 99% [37]. Conversely, NF membranes operate based on a combination of hydrodynamic filtration, the solution–diffusion mechanism, and the Donnan exclusion principle. In contrast, mass transport in RO membranes is governed solely by the solution–diffusion mechanism [36]. The pore size of NF membranes is small enough to reject divalent ions, but they also reject 20–30% of monovalent ions due to the aforementioned mechanisms.

This attribute renders NF membranes well-suited for situations where retaining some salt is advantageous, particularly in specific industrial processes [38].

The first, second, and fifth membranes (M1, M2, and M5) had the most significant impact on the sodium adsorption ratio, with the first two membranes (M1 and M2) achieving a classification of S1 in most wells. The fifth membrane had a relatively high effect on the sodium adsorption rate, raising the classification to S1 and S2 in 51 and 27 wells, respectively. The high effectiveness of the M1, M2 and M5 membranes in reducing the SAR value is attributed to their ability to remove sodium ions, which is the primary factor influencing SAR value. Conversely, the M3 and M4 membranes had little effect on the sodium adsorption rate, and most wells were classified as unsuitable for irrigation due to the high sodium hazard.

RO membranes are exceptionally proficient at decreasing the SAR (sodium adsorption ratio) of water by effectively removing both monovalent and divalent ions, such as sodium, calcium, and magnesium [39,40]. The compact pore structure of RO membranes hinders the movement of ions, leading to a noteworthy decrease in SAR values [41]. Numerous studies have documented SAR reductions exceeding 90% when employing RO membranes for water treatment [42,43]. In contrast, NF membranes exhibit a moderate effect on SAR values. As a result, the reduction in SAR values achieved by NF membranes is not as substantial as that of RO membranes. Studies have indicated SAR reductions ranging from 50% to 70% when using NF membranes [38].

Table 4. Number of wells for each classification based on EC and SAR after desalination by all membranes.

	Raw Water	M1	M2	M3	M4	M5
C1S1	0	10	58	0	0	5
C1S2	0	0	0	1	0	0
C2S1	0	0	18	0	0	46
C2S2	0	0	0	0	0	25

shows the number of wells for each membrane in which the membrane succeeded in raising the water classification to good or very good (C1S1, C2S1, C1S2, and C2S2) according to the salinity and sodium hazard index. Also, water with a salinity of less than 100 ppm was considered unsuitable for irrigation because it did not contain the nutrients necessary for plant growth. The table clearly shows the success of both the second and fifth membranes (M2 and M5) in improving the water quality for irrigation purposes in 76 wells to become good or very good. Despite the success of both membranes, there is a preference for the reverse osmosis membrane, which is the second membrane, due to the higher quality of the water produced by it compared to the fifth membrane. On the contrary, the first membrane succeeded in raising the water quality in only 10 wells, while the third and fourth membranes failed in almost all wells.

6.3. IWQI

The selection of an effective technique for evaluating water quality for irrigation purposes is a challenging task, especially when there are multiple indicators to consider, as was the case in our current case study. While membranes can efficiently reduce salts, they can negatively impact the SAR value. However, the irrigation water quality index (IWQI) simplifies this process by incorporating five indicators into a single value, making evaluating the quality of desalinated water easier. In our study, we assessed the quality of desalinated water using all proposed membranes. Table S6 displays the IWQI values for all membranes after desalination for all wells. To determine the best membrane for each well, we also included the maximum value of IWQI for each well in the table. Our findings show that RO and NF membranes are generally effective methods for improving water quality. The M2 and M5 membranes consistently had higher IWQI values than the other membranes, demonstrating their superior performance. Specifically, M2 and M5 membranes were the best fit for practically all wells, while the M4 membrane was effective in only two wells. The M2 membrane was the most effective of 27 wells, while the M5 membrane was the best of 50 wells. The number of wells for each membrane and for each classification is shown in

Table 5. Number of wells for each classification based on IWQI values after desalination by all membranes.

Classification	M1	M2	M3	M4	M5
SR	76	20	74	73	1
HR	3	27	5	6	7
MR	0	20	0	0	58
LR	0	12	0	0	13
NR	0	0	0	0	0

. The table shows that the water produced by the M1, M3, and M4 membranes had a very low quality for agricultural use and was unsuitable for irrigation in practically all wells. In contrast, the M2 and M5 membranes were superior to the other membranes, with the M5 membrane being relatively superior due to the majority of wells rating the water as good when this membrane was used for desalination.

6.4. Rejection

Figure 6 shows the average rejection value for all membranes for EC, Na, Ca, Mg, HCO₃, and Cl. The figure shows that the RO membrane systems (M1, M2, and M3) have the highest ion rejection values. The findings align with earlier studies, which also demonstrated that RO membranes exhibit a minimum ion rejection rate of 99 percent [44,45]. However, the second membrane (M2), despite being an RO membrane, exhibited a decline in salt rejection performance. This decline can be attributed to the larger void volume of this particular membrane compared to its counterparts. Consequently, the larger voids necessitate lower operating pressures, leading to lower energy consumption. Unfortunately, this characteristic also allows for a higher passage of ions through the membrane compared to the M1 and M3 membranes. NF-based systems had a low rejection power for monovalent ions, such as sodium and chloride, and a high rejection for bivalent ions, like calcium and magnesium, which is consistent with the findings of Lew et al. (2020) [46]. The systems based on RO membranes (M1, M2, and M3), on the other hand, had a high ability to reject all ions. The rejection ability of membranes depends on several factors, such as the membrane material and the properties of pores (e.g., pore size and surface charge of the membrane). RO membranes are characterized by their small pore size, which gives them a high ability to reject more ions compared to nanofiltration membranes.

6.5. Specific Energy Consumption

In light of the crucial role of energy consumption in determining the cost of desalination systems, the study focused on highlighting the energy usage associated with different proposed membranes. This emphasis stems from the fact that energy consumption significantly contributes to the overall cost of desalinated water production and also the environmental impact. Figure 7 graphically illustrates the electricity consumption required by each membrane to produce one cubic meter of desalinated water for four wells in the study area. The results depicted in Figure 7 clearly demonstrate a positive correlation between the salinity of the water source and the energy consumed by all membranes. This relationship can be attributed to the increase in osmotic pressure of the water as its salinity rises. The higher the salinity, the more energy is needed to overcome the osmotic pressure and facilitate the desalination process. Among the membranes evaluated in the study, M3 exhibited the highest energy consumption, closely followed by M1. Both M3 and M1 are based on RO technology, which requires substantial energy input for the desalination process. M2, another reverse osmosis membrane, ranked third in terms of energy consumption. Notably, the nanofiltration membranes showed relatively lower energy consumption compared to reverse osmosis-based membranes. Specifically, M4 proved to be the least energy-consuming membrane, followed by M5. Considering both water quality and energy consumption data, the fifth membrane (M5), based on nanofiltration technology, emerged as a highly promising alternative to reverse osmosis membranes. This membrane not only produces water of satisfactory quality but also stands out for its cost-effectiveness and lower energy demands. Such findings are of great scientific importance as they provide valuable insights into selecting appropriate desalination technologies for specific contexts.

7. Conclusions

This work assesses the performance of a single-stage desalination process using 5 different types of membranes. Three of the membranes are RO and two of the membranes are NF. The brackish water from 79 wells of El-Moghra (Egypts) has been used as the feed water for the desalination process. Three indicators were used to evaluate water quality: electrical conductivity (EC), the sodium adsorption ratio (SAR), and the irrigation water quality index (IWQI). It is noticed that water in most of the wells does not meet the required qualities for agricultural purposes in terms of EC, SAR, and IWQI. The groundwater is therefore unsuitable for irrigation as it was classified as low quality in all wells based on salinity hazard classification. Additionally, it was rated as bad or very bad in 97.50% of the wells based on the SAR rating. According to the irrigation water quality index (IWQI), none of these 79 wells are suitable for direct irrigation. The water was classified in 52 wells as having severe restrictions (SR) and in 27 wells as having high restrictions (HR).

Simulation software (ROSA) was used to predict the operating conditions of the system and the water quality produced from the 79 wells using different types of membranes in the desalination process. The quality of the water produced from the proposed desalination membranes was evaluated to determine which membrane succeeded in raising the water quality to an acceptable level for irrigation purposes. A comparison of the membranes was made in terms of the ability to reject salts and energy consumption. Finally, based on the evaluation, the ideal membrane was recommended for each well, i.e., the one capable of producing water of suitable quality for irrigation at the lowest energy cost.

According to the classification of freshwater based on electrical conductivity and the sodium absorption ratio, the results show the superiority of reverse osmosis membranes over nanofiltration membranes. The second and fifth membranes were successful in producing irrigation water in 76 wells, while the first membrane was successful in only 10 wells. Neither the third nor the fourth membrane succeeded in any well. Due to the high salt rejection capacity of both the first and third membranes, the water produced by them contains less than 100 ppm of salt, which makes it unsuitable for agricultural purposes. And due to the weak salt rejection capacity of the fourth membrane, the water produced by it contains high levels of salts, making it unsuitable for agricultural purposes as well. These waters can be used for other purposes, such as industry or domestic use, but care must be taken not to use them for irrigation or agriculture.

Evaluating water quality for irrigation is challenging when considering multiple indicators. Membranes can reduce salts, but negatively impact SAR. The Irrigation Water Quality Index (IWQI) simplifies this process by combining five indicators. In the current study, all proposed membranes significantly improve the water quality; M2 and M5 membranes consistently had higher IWQI values, making them the best fit for most wells. While M4 was only effective in two wells, M1, M3, and M4 had very low-quality water for agricultural use, while M2 and M5 were superior, with M5 being the most effective.

Reverse osmosis membranes (M1, M2, M3) showed the highest rejection values, while NF-based systems had lower rejection for monovalent ions (Na, Cl) and higher rejection for bivalent ions (Ca, Mg). RO’s small pore size contributes to its superior ion rejection compared to nanofiltration membranes with larger pores.

Nanofiltration membranes exhibited superior energy efficiency, with the fourth membrane being the most energy-efficient, followed by the fifth membrane. However, what sets the fifth membrane apart is its ability to elevate water quality in 71 wells to an acceptable level for agricultural purposes. This underscores the nanofiltration membranes as a highly effective alternative to reverse osmosis membranes, demonstrating the capability to produce water suitable for irrigation while concurrently reducing operational costs due to the lower energy consumption in nanofiltration-based systems.

The research provides valuable insights for assessing water quality for agricultural purposes in El Moghra region and suggests suitable desalination systems, addressing the challenges posed by energy requirements in remote areas. The study recommends examining alternative membranes from different manufacturers to assess their effectiveness.

Finally, the approach adopted in this work can be extended to any other aquifers either in Egypt or elsewhere to provide safe water for use in irrigation.