3.1 Turbidity
To assess the efficacy of the first filter in turbidity removal, two methods (photographing the filter and a turbidity test) were employed. Typically, the first filter in each system is housed in a transparent container, allowing visual observation of its color change from white to brown, indicating the need for replacement. Over a two-month period, the clogging depth of this filter was monitored weekly, comparing the results of turbidity tests and filter clogging depth. Most membrane manufacturers recommend that water entering the membrane should have a turbidity of less than 1 NTU. Hence, surpassing this value in the output turbidity was established as the criterion for replacing the first filter. Fig. 3 illustrates changes in output turbidity for the first filter of System 1, while Fig. 6 depicts the weekly color change procedure over three months. Replacements were carried out in September, January, and June based on the appearance color of the filter. Fig. 6 highlights significant color changes on the filter surface after two months of installation. Fig. 4 showcases variations in the turbidity of the inlet and outlet water of the first filter of System 2, and Fig. 7 displays its color change procedure. Replacements occurred in October, January, and April based on appearance color, with the output turbidity exceeding 1 NTU about two months after the first filter replacement, as indicated in Fig. 4. Fig. 7 demonstrates color changes starting one month after filter installation, progressing to the inner layers by the end of the second month. Fig. 5 and Fig.8 depict changes in the color of the first filter in System 3 and variations in raw water turbidity and the first filter's output turbidity. Replacements took place in October, January, April, and July, guided by appearance color. An increase in output turbidity was observed approximately one month after filter installation, while raw water turbidity averaged 1.1 NTU during the sampling period, often exceeding 1 NTU. Fig.7 indicates that the color change on the filter surface began one week after installation. Thus, in these three systems, signs of first filter clogging manifested about one month after installation, reaching a potential maximum within second months.
3.2 TDS
Fig. (9-a) illustrates the fluctuations in input and output Total Dissolved Solids (TDS) for System 1 over the course of one year. The input TDS ranged from 700 to 1100 mg/l, while the output TDS remained below 100 mg/l. This consistently met the admissible limits recommended by ISIRI (1000 mg/l) and WHO (600 mg/l). However, a TDS level below 100 mg/l may indicate excessive removal of essential water solutes, including calcium, magnesium, and fluoride, which are beneficial to the human body but are removed at a high percentage by the membrane (Warsinger, D. et al., 2016; Bellona, et al., 2004). Fig. (9-b) displays the variations in input and output TDS for System 2 throughout the sampling period, one year after membrane installation. Input TDS fluctuated between 700 and 1100 mg/l, and the output TDS ranged from 100 to 200 mg/l. Similar to System 1, the output TDS remained below the recommended standards. However, there was an increase of approximately 100 mg/l compared to System 1, possibly attributed to the longer lifespan of its membrane. Fig. (9-c) demonstrates the changes in input and output TDS for System 3 over one year. The membrane in this system was installed two years prior to the start of sampling. Input TDS varied between 800 and 1300 mg/l, and the output TDS ranged from 300 to 500 mg/l. In the last month of the sampling period, the device's membrane was replaced, resulting in a decrease in output TDS to 50 mg/l. Despite being within standard limits, consumers found the taste of the water unacceptable, leading to the replacement of the filter approximately 3 years after its installation. Comparing the performance of the three systems reveals that in the first year of membrane operation, the output TDS consistently remained below 100 ppm. In the second year, this parameter increased to around 200 ppm, and in the third year, it sharply rose to the permissible limit recommended by WHO. It is essential to exercise caution with low TDS concentrations, as the low mineral content in purified water can pose health risks, including potential cardiovascular problems, reduced thyroid function, and increased risk of diseases such as osteoporosis and cancer. Therefore, it is advisable to consider minimum values for parameters related to human health, in addition to adhering to standards such as those recommended by WHO, which specify maximum amounts for receiving essential ions during the operation of these devices (Table 3) (Rosborg & Kozisek, 2019; WHO, 2017). The research conducted by Sediq et al. (2013) and Normoradi et al. (2016) revealed that the output TDS was below 100 mg/l, aligning with the TDS output of System 1 in the present study. Additionally, Bonnélye et al. devised a treatment approach encompassing seven stages, incorporating a Reverse Osmosis (RO) treatment system to eliminate TDS from Curacao drinking water. The final TDS concentration was determined to be below 150 ppm, consistent with the TDS output of System 2 in this investigation.Dinarello et al. (2016) observed an output TDS of approximately 500 mg/l, a value that correlates with the output TDS from System 3 during the last month of the sampling period in the current study. It's worth noting that these studies did not provide information about the lifespan of the membranes used in the experiments.
3.3 pH
Based on Fig.10, the pH levels of all three systems declined over the sampling period, reaching 5.8, 6.7, and 6.5 for systems 1, 2, and 3, respectively, indicating an acidic shift in the water. This acidification is attributed to the removal of alkalinity factors, specifically carbonate and bicarbonate ions, by the membranes. The elimination of these ions disrupts the balance between (H+) and (OH-) ions, leading to water acidity, as highlighted by WHO (2017). WHO recommends a minimum pH of 6.5 for drinking water, and Figure 14 illustrates that, during the sampling period, the pH of effluent from all three systems dropped to the minimum limit, occasionally turning acidic. Hence, close attention is necessary to address the acidification in the output of these systems due to potential long-term health effects associated with consuming acidic water. Water acidification can result in undesirable taste and increased corrosiveness, especially in proximity to metal pipes too (Rezaeinia et al., 2018).Studies conducted by Rezaeinia et al. (2018) and Rummi Devi Saini (2017) in the medical field indicate that the consumption of acidic water and other acidic beverages can lead to mineral imbalance in the body. Cancer patients often experience electrolyte imbalance, including conditions like hyponatremia, hypokalemia, hypocalcemia, and hypomagnesemia, underscoring the significance of minerals in drinking water (Rosborg & Kozisek, 2019). While it is possible to adjust the pH of membrane outlet water using a mixer tap to increase Total Dissolved Solids (TDS), this practice is not recommended for a long duration in Ahvaz city. Because sediment accumulation in the mixer tap can lead to blockages in this system over time. Figure 10-c in system 3 shows that replacing the fourth filter in July resulted in a decrease in TDS from 400 to 50 mg/l and a pH drop from 7 to 6.5. Despite this, ongoing monitoring of the pH in the output water from these devices is advisable. Confirmation of water acidification in the outlet has been observed in other studies as well, such as those by Yari et al. (2011), Miranzadeh et al. (2016), Jafaripour et al. (2016), and Dehghani et al. However, these studies focused on investigating the effects of water acidification rather than the pH changes patterns. In contrast, some studies consistently reported that the output pH remained within the standard range, potentially attributed to the effective performance of forth, and sixth filters during sampling period (Babaei et al., 2013; Twanger et al.2014; Dinarello et al.2015; Naqi-pour et al. 2018; Rezaei-nia et al.2018; Al-Jayyousl et al.2019; Zejin Zhang et al.2020).
Fig. 11 depicts the variations in pH versus TDS for the inlet and outlet water of all three systems. In these illustrations, water TDS is arranged in ascending order, and the corresponding pH values are indicated. Figures 11-a, 11-c, and 11-e demonstrate that the TDS of the water entering the devices consistently surpassed 690 mg/l, while the pH levels remained within the standard range of 6.5 to 8.5 for all instances depicted. Based on the data depicted in figures 11-b, 11-d and 11-f, there were instances where the pH levels fell below the recommended minimum limit, and in these cases, the TDS values were also below 50 mg/l. This suggests a potential for water acidification with decreasing TDS levels. According to Rosborg & Kozisek (2019), they proposed a minimum TDS level of 100 mg/liter for drinking water.
3.4 Microbial quality
Throughout the sampling period, the presence of total coliforms was assessed in both the input and output water. The results indicated the presence of total coliforms in only three samples of input water, with no total coliforms detected in any of the output water samples. This outcome aligns with the findings of Dehghani et al. (2013) and Twanger et al. (2010). The absence of coliforms in the outlet can be attributed to the effective performance of the membranes, which successfully remove these organisms, leaving them in the form of biofilm behind the filter (Morris et al., 2008). In contrast, studies by Nasrallahi et al. (2009), Yari et al. (2011), and Ebrahimi et al. reported the presence of coliforms in the output. This occurrence might be linked to a delayed replacement of the membrane, allowing coliforms to pass through due to the leakage phenomenon (Morris et al., 2008). Lin et al. (2020) demonstrated that coliforms were present in the outlet, and to eliminate them, either membrane replacement or the installation of an ultraviolet (UV) unit as the last stage in the outlet is necessary (Lin et al., 2020).