Evaluation of the benefit of practically operating reverse osmosis system in the factory: taking the recycling of KI solution and water of the screen polarizing plate as an example

With the advancement of industrial production technology, the industrial process is also rapidly improved, and the characteristics of pollutants in the wastewater system are constantly changing. The scarcity of water resources is increasingly affecting the global economic development and ecological environment. In order to improve treatment efficiency, reducing industrial water consumption cost, exploring the potential advantages and feasibility of this wastewater recycling system, reducing the concentration of pollutants, and reducing the impact on the ecological environment will be an important topic of this research.

In the related studies of RO membrane technology (Chen et al. 2000; Benito & Ruíz 2002; Ozaki et al. 2002; Hsu 2004; Shenvi et al. 2015; Lin 2005; Hsu et al. 2018; Zheng et al. 2018), scholars also pointed out that RO membrane was suitable for the treatment of metal processing wastewater (Benito & Ruíz 2002; Ozaki et al. 2002). It could be used to purify the rinsing water of the electroplating layer and concentrate the plating metal ion so that the water can be used repeatedly in an electroplating bath, which brings significant economic benefits (Hsu et al. 2018). Therefore, this study is to evaluate the benefit of recycling the KI solution and water from a screen polarizing plate factory with the RO system.

With the increasing shortage of water resources in Taiwan, the recycling cost of RO water is about US$ 0.83/t, which can reduce the water pollution control fee and wastewater treatment fee as required by the administrative bureau by US$ 1/t. For example: assuming that a polarizing plate factory recycles 250 cubic meters of wastewater per day (CMD), it can save annually US$ 168,630; if the 2 CMD KI solution recycled per day (about US$ 5/L) is added, the annual saving will be US$ 3,650,000; after deducting the operating cost of US$ 109,000, the annual saving will be US$ 3,709,579, which indicates the benefit of recycling is remarkable.

Therefore, in this study, the benefits of operating an RO system for the purpose of KI solution and water recycling in a screen polarizing plate factory are evaluated, and the key points and directions are discussed as follows:

• I.

  analyzing water quality of waste liquid from the polarizing plate manufacturing process;

• II.

  discussing the mechanism of blocking and scaling of RO membrane;

• III.

  comparing the operations between RO membrane for water purification and anti-fouling RO membrane for water recycling; and

• IV.

  overall benefits of recycling both KI solution and water from the polarizing plate manufacturing process.

This research first conducts the water quality analysis test of the polarizing plate factory to understand the analysis of the components contained in the water, input the water quality and various influencing factors, and design the number of membranes and the arrangement and combination of the RO system by software simulation. The results of changes in pressure, membrane pressure (ΔP), changes in water quality, frequency of chemical washing, changes in water collection time, etc., to infer the optimal water collection, water collection time, frequency of chemical washing, and optimal chemical washing (CIP) for RO film Concentration, time, temperature and other parameters are used to derive operating costs and recovery years, test the best parameters, and integrate the above conditions to evaluate the benefits of this research. The research method flow diagram is shown in Figure 1 for details.

In this test, the membrane developed by H Company (Table 1) and the RO membrane produced by H Company were used. The membrane was made of polyamides and had a daily permeating water flow rate per membrane of 41.6 CMD; a desalination rate of 99.7% (minimum of 99.5%); a structure of low pollution spiral winding; a membrane area of 400 ft² (37.1 m²); a water feeding pipe of 34 Mil (0.864 mm); a maximum resistance to residual chlorine concentration of <0.1 PPM; a maximum pressure resistance of 600 psig (4.16 MPa); a maximum operating temperature of 113 °F (45 °C); a continuous pH range (chemical cleaning) of 2–10 (1–12); a maximum feeding water turbidity of 1.0 NTU; a maximum feeding water supply silt density index (SDI) (15 min) of 5 and a maximum inflow 75 GPM (17.0 m³/h).

In order to avoid the interference of other ions, the ultra-pure water was used as the reagent water during the experiment process, and the reagent types used are as follows:

• 1.

  Sodium hydroxide (NaOH): Merck, 99.6%

• 2.

  Sodium carbonate (Na₂CO₃): Ferak, 99.5%

• 3.

  Hydrochloric acid (HCl): Merck, 30%.

• 4.

  EDTA tetrasodium salt: Daojiu, 99.5%

• 5.

  Sodium hypochlorite (NaOCL): MERCK, 50%

• 6.

  Sodium bicarbonate (NaHCO₃): Ferak, 99.5%

• 7.

  Acid (H₂SO₄): Merck, 95–97%

• 8.

  Phosphoric acid (H₃PO₄): Merck, 85%

The complete RO system setup was mainly composed of four parts, including pretreatment, RO host (membrane filtration), posttreatment and system cleaning.

The flowchart of the RO system of this study is shown in Figure 2 and Table 2 shows the RO system units.

The design of the RO membrane filtration system in the RO main machine (membrane filtration) included the membrane module, pressure vessel arrangement, high-pressure pump, pipeline, instruments and meters, etc. The general principles of its design were to reduce the operating pressure of the system as much as possible, save the cost of the membrane module, maintain the long-term stability of the system, reduce its cleaning and maintenance costs and achieve the increase of permeating water flow rate and KI recycling rate. The most significant factor in the design of the RO membrane system was the fouling tendency of feeding water. The fouling of the membrane module was due to the existence of particulate matters, colloidal matters and organic matters in feeding water and their deposition on the membrane surface. The deposition rate of fouling materials increased with the increase of the average treated water (permeation) rate (permeating water load per unit membrane area) and the module recycle rate (which affected the concentration polarization). Therefore, the overly high system average permeation rate and the system recycle rate can easily lead to a higher fouling rate and more frequent chemical cleaning. Design guidelines were empirical values obtained by a comprehensive study based on the design and operation data of a large number of engineering projects involving different types of water sources, and the system was designed according to these guidelines so as to reduce the cleaning frequency in operation and prolong its life cycle.

In the posttreatment and system cleaning steps, a reasonable RO system was designed for different feeding water sources and permeating water quality in the hope of reducing the cleaning frequency, prolonging the cleaning cycle, improving the long-term stability of the system and reducing its operating cost.

The data of the parameter test results of actual factory units shown in Figure 3 were recorded and analyzed in the hope of finding the optimal parameters suitable for this water source and smooth operation of the RO membrane, which are as follows:

• 1.

  After the results indicated that the feeding water met the requirements by testing and analyzing the feeding water of the device, only the adjustment of the feeding water could be conducted.

• 2.

  The pressure control of water supply pump and automatic water quality monitoring system were adjusted.

• 3.

  It should carry out the inspection to see if all pipelines of the device were perfectly connected, if the pressure gauges were complete, if the joints of low-pressure pipelines were tight and sufficient.

• 4.

  The discharge valve in front of the pump was turned on, the pretreatment equipment was started and the water supply rate was adjusted to make it greater than the feeding water rate of the device.

• 5.

  All pressure gauge switches, water inlet valves, concentrated water outlet valves and permeating water outlet valves were fully turned on.

• 6.

  The outlet valve in front of the pump was turned off, and the water feeding valve of the device was turned off after the module was fully filled with water.

• 7.

  When the inlet pressure of the high-pressure pump was greater than 0.2 MPa, the high-pressure pump was started, and the water feeding valve of the device was turned on slowly while the concentrated water discharge valve was turned off. The total pressure difference of the device was controlled to be less than 0.3 MPa; after the permeating water was discharged for 2 min, the permeating water outlet valve was turned off, the device was running for 15 min and all the high- and low-pressure pipelines and instruments should be inspected to see if they work normally.

• 8.

  The water feeding valve and concentrated water outlet valve were adjusted so that the ratio of permeating water to concentrated water was 3:1.

• 9.

  The conductivity of permeating water was inspected, and when it met requirements, the permeating water outlet valve was turned on first and then the permeating water outlet valve was turned off.

• 10.

  Notes on the adjusting process:

• • (1)

    The water feeding pressure should not be greater than 1.5 MPa during adjustment.

  • (2)

    If the feeding water temperature was higher or lower than 25 °C, it should be corrected according to the water temperature-permeating water flow rate curve, and the recycle rate should be controlled to be 75–90%.

  • (3)

    After the device runs continuously for 4 h, if the desalination rate could not reach the designed removal rate, the desalination rate of each component of the device should be inspected, and faulty components (if any) should be replaced.

  • (4)

    When it was found that there was water leakage in the high-pressure pipeline, the pressure of the device should be relieved, and it was strictly forbidden to loosen the high-pressure joint under high-pressure condition.

The test of and comparison between traditional pure water RO membrane CPA3-LD and anti-fouling recycle membrane LFC3-LD are shown in Table 3.

In this study, the process of polarizing plate factory was run by using the RO membrane system equipment for more than 3 months, where KI waste liquid was concentrated by anti-fouling RO membrane while RO permeating water was recycled so that it enters the water purification system for reuse. The system stability was proven by raw water quality analysis, RO membrane selection, RO membrane microanalysis and system test in an actual factory; the recycling effect of chemical cleaning was analyzed and the optimization of parameters was made according to the data change. Finally, the economic benefits of the recycling period of the actual factory were evaluated.

In Table 4, it can be observed from Day 1 to Day 5 that the total organic carbon (TOC) value is between 0.58 and 0.79, which is less than 1 ppm and better than the TOC requirement of 3 ppm for tap water quality. During the 3 months of operation, the water quality meets the standard of recycling TOC of less than 1 ppm. The conductivity value of the water sample is 2,899–3,298 μs/cm, and the pH value of wastewater is between 7.5 and 7.9. It can be inferred that this portion of water originates from ultra-pure cleaning without adding any chemical agents. Somewhat different from the pH of 7.5–7.9 of wastewater selected for this experiment, the water quality of this experiment was far superior to that of chemical mechanical polishing (CMP), because too much oxidants and acid and alkali-containing substances were added to the CMP water. Roche tube water was not recovered as the water quality of the wastewater source was extremely complex. Too many variables would increase the treatment cost, thus reducing the treatment recycling rate and benefits.

RO membrane test results are plotted as shown in Figure 4. The test time was about 4 months. The pressure difference rising rate can be observed from the figure, which is far higher in the first stage than in the second stage. Therefore, the key factor is in the first stage. The conventional RO pressure difference at the beginning of the first stage was 1.5 kg/cm² and that of the LD anti-fouling membrane was 0.5 kg/cm². After continuous operation for 1.5 months, the blocking pressure difference was 5.6 and 2.9 kg/cm², respectively, and chemical CIP cleaning was carried out at the time to restore the pressure difference to 4.1 and 0.9 kg/cm². After continuous operation for 3 months, the blocking pressure difference was 7.6 and 7.1 kg/cm², respectively, and chemical CIP cleaning was carried out to restore the pressure difference to 6.0 and 0.9 kg/cm², respectively. Therefore, it could be clearly seen that the anti-fouling membrane could prevent blockage and had a cleaning restoration that was much better than that of the traditional membrane, so the LFC3-LD membrane was chosen for actual factory testing. It showed a graph of traditional membrane and RO membrane in delta-pressure first stage during 270 days of operation period. The red line (LD technology, pressure loss from 0.8 kg/cm² to 1.2 kg/cm²) was more stable than the blue line (conventional technology, pressure loss from 0.8 kg/cm² to 4.4 kg/cm²) and the difference for total pressure loss from these two technologies was 3.2 kg/cm². The traditional membrane needed continuous cleaning, had unstable water flow and decreased recycling water rate which could increase employees' workloads (Figure 5).

There are four main causes of membrane fouling: adsorption and growth of microorganisms, adsorption of organic matters, aggregation of colloids and particulate matters, and precipitation of inorganic matters. In the process of practical application, the combined action of these four reasons leads to membrane fouling (Hsu et al. 2018). Zheng et al. (2018) found that bio-fouling was the main contributor in the RO process and Ca, Mg, Si, Fe and bio-derived OMs played significant roles in organic fouling. Biological pollution and microbial community are considered as key contributors together in the results of SEM analysis. The fouling at the end cover and membrane entry also indicates significant biological fouling, and the vertical distribution of microbial community is found in the cross section of fouling. This study is helpful in clarifying the components and main contributors of RO membrane fouling and improving our understanding of membrane fouling mechanism and control strategy. In the comparison, Figure 6(a3), 6(a4)/(b3), 6(b4) shows the pure water and recycle spacer blockage and structural phenomena. As the water source of pure water membrane was tap water and the recovered water was potassium iodide wastewater, it was found that the spacer of recycling membrane was obviously polluted and blocked after operation, and in particular there was very obvious iodine deposit on the b4 grid, while it was clean on the a4 grid. From the microscope observation in Figure 7, it can be seen that the pure water membranes a1–a8 are slightly polluted and blocked, while the recycle membranes b1–b8 have white crystals formed on the grid as shown in Figure 7(b8) (400×), which are pure potassium iodide crystals. In Figure 8(a)–8(h), there are not only white deposits on the net column but also gray-white granular deposits at the intersection of the net columns. It is speculated that it is because there were dead angles at the corners, and the deposition phenomenon is the most remarkable where the flow rate was the lowest.

In Figure 9(a)–9(d), when the crystals are observed at high magnifications of 700–3,500 × , they are found to be potassium iodide crystals, and the width of the trace crystal substance is about 35–45 μm, which is gray-white at the highest point on the surface of the net column. On the surface of pure water RO membrane in Figure 10, the intercepted particles observed in Figure 10(a) at 350× magnification are brown in color and analyzed as trace iron deposits in tap water, while obvious gray-black deposited particles observed in Figure 10(b) and 10(c) 700–1,400× are analyzed as SiO₂ particles and trace iron deposits. On the surface of the recycling RO membrane, as shown in Figure 11(a)–11(d), the crystal substance on the surface, under the magnification of 350–2,450 × , is found to be completely piled up at the highest place. Upon analysis, the gray iron deposits are mixed with the gray potassium iodide crystals.

In Figure 12(a)–12(d), translucent nematodes are surprisingly found in the surface substance in addition to colloidal dispersion at the magnification of 350–1,450 × , with a quite large number and a length of about 180–200 μm. As the trace TOC in water is the food for nematodes to grow and absorb nutrients, it is also one of the main reasons for RO membrane bio-fouling (Yu et al. 2018). According to the study, different bacterial species and their extracellular polymeric substances (EPSs) significantly affected the fouling potential of the RO membrane in the wastewater recycle, and EPSs components with molecular weight (MW) exceeding 10 kDa were separated by the ultrafiltration membrane and proved to have higher membrane fouling potential.

Besides, through 3D micrographs in Figure 13(a)–13(c) (original color), Figure 13(d) and 13(e) (pseudo-color) and Figure 13(g)–13(I) (wire frame), it can be seen that the recovering RO membrane surface in brown is a stack of KI crystals, while Figure 13(j) and 13(k) show that its height is about 0.907 μm (1.812 − 0.905 μm).

From Figure 14(a)–14(c), it can be further observed that the uneven accumulation of white potassium iodide crystal deposit on the grid support layer of the cross structure of KI_support anti-fouling membrane, the highest position of which is the red block in the upper left corner, with a height of 94.721 μm, and the stacking height of white potassium iodide crystal is 47.37 μm (green part). The red block at the highest position can reach a height of 94.721 μm.

We can discuss the stability of the RO membrane system from two aspects. The first one is to make a curve analysis of ΔP of the system, i.e., pressure difference vs time (Figure 15). First, observe Table 1, when ΔP was 3.3 kg/cm² on the 14th day, chemical washing was conducted, in which NaOH at pH = 12 was first used at normal temperature and then HCl at pH = 2 was used at normal temperature. As a result, ΔP was returned to 1.6 kg/cm². The operation continued until the 38th day in the second step, when ΔP was 4.7 kg/cm² and chemical cleaning was conducted. After cleaning with NaOH at pH = 12 at 45 °C, and then with HCl at pH = 2 at normal temperature, ΔP was returned to below 0.4 kg/cm², indicating that NaOH heating was very effective.

Another factor to be discussed is the curve of time vs permeating water flow rate. If the curves are consistent with our design values, for example, this system is designed to produce a permeating water flow rate of 10 m³/h within 3 months. If after the blockage, the water flow rate can be restored to 10 m³/h after chemical cleaning, the system should be stable at this time. Otherwise, chemical cleaning should be done as just described to achieve recuperability. In Table 1 and Figure 16, after 14 days of operation, the initial flow rate was decreased from 16 to 8 m³/h. As described above, it was restored to 13.5 m³/h after chemical cleaning. And when the operation continued until the 38th day, the permeating water flow rate was once again reduced to 8 m³/h, but was restored to the initial 16 m³/h after chemical cleaning with NaOH at pH = 12 at 45 °C, indicating that heating has an obvious relation to the restoration of water flow rate.

From the combined comparison of pressure loss vs permeating water flow rate vs time of RO system in Figure 17, it can be observed that the pressure loss, permeating water flow rate and time are negatively correlated. On the 14th day, when the pressure loss was increased to ΔP 3.3 kg/cm², the permeating water flow rate was decreased to 8 m³/h, and on the 38th day, when ΔP was increased to 4.7 kg/cm², the permeating water flow rate was decreased to 8 m³/h. The RO system pressure loss value will decrease with the number of days, and the RO system water flow value will increase with the number of days, using time as the benchmark. It is obvious that the pressure loss is in reverse relation to the permeating water flow rate.

Another (approach) in Figure 18 was the comparison between the feeding water KI concentrations of 390–466 ppm with the concentrated water KI concentration of 2,999–3,366 ppm. If KI concentration had no deviation but kept stable around 9–10 times the design value in 3 months, it would mean that the recycling membrane operated normally without damage.

• 1.

  In the polarization process, a large amount of water is used, and the discharge of wastewater generated from etching and pickling is continuous. In order to recycle the process water, the RO membrane system is often used for wastewater recycling. There are several main control points in the RO system operation, which mainly depend on the liquid level of RO permeating the water tank. The RO system permeating water recycle will be started when the liquid level is low, or when the feeding water tank of the RO system is at an ultra-high level.

• 2.

  To judge whether the RO membrane system operates normally and whether the equipment is damaged or not, we can observe the concentration of KI (potassium iodide) after the RO membrane system is used for 3 months. If the KI concentration is 9 to 10 times the design value, and the KI concentration of feeding water and that of concentrated water are always stable without deviation, both the RO membrane system and equipment are normal.

The benefits are estimated for the recycling of cutting and grinding wastewater of an actual IC packaging and testing factory, which recovers 250 CMD wastewater at an operation water flow rate of 10.5 m³/h and a recycling rate of 85%. The consumables, electricity charges, labor cost, interest costs, etc. for the 5 years of operation, and the economic benefits of saving costs in the average annual recycle are estimated. The relevant consumables data for benefit evaluation are shown in Table 5.

If the owner directly discharges the wastewater without a recycling system, the costs for wastewater treatment and water pollution will be about US$ 168,630, and the cost for KI recycling will be about US$ 3,650,000. If the recycling system is used, US$ 3,709,579 can be saved per year, and the device will be recovered after about 0.08 year, as shown in Table 6.

In this study, the KI waste liquid from the polarizing plate factory is concentrated with anti-fouling RO membrane, and the permeating water of the RO system is recycled so that it enters the water purification system for reuse, thus not only achieving the benefits of double recycles but also realizing the circular economy advocated by the government due to the increasing shortage of water resources. The recycle cost of RO water is about US$ 0.83/t, eliminating the water pollution fee and wastewater treatment fee of about US$ 1/t to be collected by the Bureau of Science and Technology. Assuming that the polarizing plate factory recycles 250 CMD wastewater per year, it can save annually US$ 168,630 and is shown in Table 7. If the 2 CMD KI solution is recovered per day, about US$ 5/L is added, the annual saving will be US$ 3,650,000 after deducting the operating cost of US$ 109,000.

Figure 19 shows the annual expenses to be paid by the company before and after the construction of the water recycling system. Before the construction, the KI recycle fee, wastewater treatment fee, water pollution fee and water purification fee should be paid a total of US$ 3,818,630, but after the construction, only the operation fee and labor cost totaling US$ 109,015 should be paid. That is, the expense of about US$ 3,709,579 can be saved each year. According to this study, the recycling of the cutting and grinding wastewater from an actual IC packaging and testing factory can achieve the benefits of double recycles, which can not only achieve the purpose of water recycling but also realize the double benefits of the circular economy advocated by the government. In the future, it is hoped that it can be popularized in all packaging and testing factories, which will not only bring business opportunities to manufacturers, save considerable expenses, but also reduce the pollution of the earth, save a lot of water resources and water treatment energy. It can really serve multiple purposes, and it is a recycling technology worthy of vigorous promotion.

• I.

  There are two mechanisms that affect the permeating water flow rate of the RO system: one is the blockage caused by the ion crystallization, and the other is the blockage caused by the biomembrane. KI (potassium iodide) chemical is generally added in the process to improve the recycle rate, but it can make the down tank (reaction tank) reach the peak period of reaction. When KI chemical is cleaned, the conductivity will become high, and when the RO conductivity removal rate is poor, it will make the recycled water quality unable to meet requirements and also indirectly affects the improvement of recycled water volume. Therefore, if the RO membrane system runs stably, the good recycling efficiency can be achieved, the recycling system can produce permeating water stably, the concentration of KI will be relatively stable, and the recovered water quality can be purer.

• II.

  The stability of the RO membrane system can be discussed from two aspects. The first step is to analyze the curve of the pressure difference of the system vs time. The second step is that when the ΔP reaches 4.7 kg/cm² on the 38th day and cleaning with NaOH and HCl is conducted, if the ΔP returns to below 0.4 kg/cm², it indicates that the RO membrane system is very stable.

• III.

  With the increase of time, the relationship between pressure loss and the RO permeating water flow rate will become negatively correlated. On the 14th day, when the pressure loss was increased to ΔP 3.3 kg/cm², the permeating water flow rate was decreased to 8 m³/h, and on the 38th day, when ΔP was increased to 4.7 kg/cm³, the permeating water flow rate was decreased to 8 m³/h. It is obvious that the pressure loss is in reverse relation to the permeating water flow rate.

• IV.

  To achieve the vision of effective utilization of water resources and meet the requirements of the management strategies and regulations for effective utilization of industrial water issued by relevant government departments, the economic benefits of water recycling are important, and depend on proper pipeline division, water flow diversion and the economic scale of water recycling. About US$ 670,000 is needed for wastewater treatment and water pollution fees, while US$ 560,000 can be saved per year from the benefits of the recycling technology according to this study, for which the investment can be recovered in about 1.14 years.

All relevant data are included in the paper or its Supplementary Information.