The method of controlling the process of ion exchange water softening
Ivan Tikhonov
The article deals with the problem of operational control of the process of ion water softening. The dependence of the electrical conductivity of water on its hardness value for the water softening process is presented. The article presents a control system for the water softening installation based on programmable relay. The results of this system of operational control of the softening process with analysis and conclusions are given.
For some reason, the first thing that comes to mind when I look at this graph is that it is an electrocardiogram of the ion exchange water softening unit. It’s the rhythm of her heart. As long as the heart beats, the installation is alive. As soon as the fluctuations in values of conductivity fade, the installation “dies”, ceases to soften water. But let’s talk about it in due course.
It is known that the electrical conductivity of water depends on the amount of substance (ions) dissolved in water. The electrical conductivity of the vast majority of natural waters is provided by six types of ions: Cations-Calcium (Ca), Magnesium (Mg), Sodium + Potassium (hereinafter Sodium Na); Anions – Bicarbonate (HCO3), Sulfates (SO4); Chlorides (Cl). Charge transfer in water is carried out by ions. The process of transferring an electric charge in water is discussed in detail in the article [Ivan Tikhonov «The influence of ion composition of water on its electrical conductivity» www.tiwater.info]. If we consider the electrical conductivity of water with hardness ions (Ca, Mg) and the electrical conductivity of softened water, i.e. water only with sodium cations, then the electrical conductivity of softened water will almost always be higher than the electrical conductivity of the original hard water. This is not due to the fact that the molar mass of sodium is slightly larger than calcium and magnesium with equivalent substitution (Na=23*2=46 vs. Ca=40, Mg=24) and, accordingly, the salt content of softened water increases, but because monovalent sodium loses less mobility in the environment of other ions than divalent calcium and magnesium ions and can transfer more electric charge than divalent calcium and magnesium.
The value of the electrical conductivity of a separate ion does not have a physical meaning, because it holds an ion shell of oppositely charged ions around it in water and, accordingly, experiences a retarding effect from the ions of the opposite charge. We can fairly accurately determine the value of the electrical conductivity of water only for a specific salt, and also determine the decrease in mobility (equivalent to electrical conductivity) when the concentration of this salt increases.
The ions that determine the electrical conductivity of natural water are shown in table 1.
Table 1
| Ions (mg-eq/l) | HCO3 | SO4 | Cl |
| Ca | Ca(HCO3)2 | CaSO4 | CaCl2 |
| Mg | Mg(HCO3)2 | MgSO4 | MgCl2 |
| Na | NaHCO3 | Na2SO4 | NaCl |
As can be seen from table 1, nine salts can be formed in water by six types of ions. (six salts of strong acids-sulphate and chloride; three salts of weak acid-bicarbonate-carbonic).
From the point of view of the mutual influence of ions, we can say that water contains 3 types of salts of the following valences:
– 1:1 (NaHCO3, NaCl)
– 2:1 (Ca(HCO3)2, Mg(HCO3)2, CaCl2, MgCl2, Na2SO4)
– 2:2 (CaSO4, MgSO4).
If water of this composition is softened by passing through a cationite in Na form, then the softened water will contain only Na ions as cations, i.e. only 1:1 (NaHCO3, NaCl) and 2:1 (Na2SO4).
The loss of the mobility of ions of 1:1 and 2:1 salts in the area of low concentrations (up to 10 mmol / l) is almost the same. That is, if the source water contains only 1:1 and 2:1 salts, then after replacing all the cations in such water with Na cations, the value of the electrical conductivity of such water will become less or will remain the same. In this case, the source water does not contain sulfate and the softened water will contain only 1:1 salts.
If the source water contains sulfate, it means that the water contains 2:2 salt and after softening, this salt becomes sodium sulfate, i.e. 2:1 salt. Accordingly, in this case, the drop in mobility of 2:1 salt is significantly less than the drop in mobility of 2:2 salt. Therefore, the electrical conductivity of such water will be greater than that of hard water. At the same time, the greater the concentration of salts in the source water is, the higher the electrical conductivity of the softened water relative to the source will be.
Conclusion: The more sulfate ion is found in the source water, the greater the difference between the value of the electrical conductivity of the softened water in relation to the electrical conductivity of the source water is, and the better it is possible to perform conductometric control over the operation of ion exchange water softening plants.
However, if the hardness of the source water is more than 8 mg-eq/l, even if there is no sulfate in the water, the electrical conductivity of the softened water will be greater than the electrical conductivity of the source water. Although the difference in this case will be insignificant, and this greatly complicates the quality control of the ion exchange softening process of such water using electrical conductivity.
It is possible to determine with sufficient accuracy the value of the electrical conductivity of the solution depending on its concentration for each specific salt. But as you can see from table 1, there are at least 9 salts in water that determine its electrical conductivity.
In order to determine the concentration of each salt in water, it is necessary to assume that the ion shell of each ion is evenly distributed, i.e. the ratio of different ions in any part of the water volume is the same ions at any time. In fact, there should be a uniform distribution of ions in the volume of water taken in mg-eq/l.
Accordingly, you can determine the share of each salt, and knowing the concentration of each ion (from chemical analyses), you can determine the concentration of each salt by multiplying the concentration of each ion by the share of each (corresponding) salt.
Knowing the concentration of each salt, you can determine the value of the electrical conductivity of each salt.
Then, by analogy, using the approach of the transference number in determining the electrical conductivity of an individual ion, we can determine the value of the electrical conductivity of water containing all the salts shown in table 1.
Based on the described approach, I developed a program for calculating the value of the electrical conductivity of water from its ionic composition. As a result of the work on this program, it became clear that it is possible to control the process of ion exchange water softening by the value of electrical conductivity. Moreover, using this program, you can determine the ionic composition of water. To do this, you need to know the electrical conductivity of the source and softened water, as well as the water hardness and alkalinity.
For example, let’s calculate the electrical conductivity of the source water based on known ionic composition:
Cations:
Source water
Сa – 0,001 mol/l (0,002 mg-eq/l)
Mg – 0,0006 mol/l (0,0012 mg-eq/l)
Na – 0,0009 mol/l (0,0009 mg-eq/l)
Softened water
Na – 0,0041 mol/l (0,0041 mg-eq/l)
The sum K – 0,0041 mg-eq/l
Anions:
HCO3 – 0,002 mol/l (0,002 mg-eq/l)
SO4 – 0,0007 mol/l (0,0014 mg-eq/l)
Cl – 0,0007 mol/l (0,0007 mg-eq/l)
The sum A – 0,0041 mg-eq/l
The sum K = The sum A
The calculation results are shown in table 2
Table 2 The calculation of the electrical conductivity of the source water
| Salts | mmol/l | mg/l | n | Cond. µS/cm |
| Ca(HCO3)2 | 0,487804878 | 79,02439 | 0,868324 | 91,00798 |
| СaSO4 | 0,341463415 | 46,43902 | 0,646076 | 71,87858 |
| CaCl2 | 0,170731707 | 18,95122 | 0,446775 | 42,41778 |
| Mg(HCO3)2 | 0,292682927 | 42,43902 | 0,811072 | 52,3246 |
| MgSO4 | 0,204878049 | 24,58537 | 0,586084 | 41,94853 |
| MgCl2 | 0,102439024 | 9,731707 | 0,401087 | 24,26331 |
| NaHCO3 | 0,43902439 | 36,87805 | 0,915811 | 40,2682 |
| Na2SO4 | 0,153658537 | 21,81951 | 0,580393 | 37,59439 |
| NaCl | 0,153658537 | 8,989024 | 0,477637 | 18,81977 |
| The sum | 2,346341463 | 288,8573 | 0,6869 | 420,5231 |
Table 3 shows the result of the calculation of the electrical conductivity value for fully softened water. Only sodium cations are used as cations in such water, so the calculation is made for only three salts.
Table 3 The calculation of the electrical conductivity of softened water
| Salts | mmol/l | mg/l | n | Cond. µS/cm |
| NaHCO3 | 2 | 168 | 0,92425 | 181,7691 |
| Na2SO4 | 0,7 | 99,4 | 0,586697 | 169,4232 |
| NaCl | 0,7 | 40,95 | 0,480596 | 85,20665 |
| Сумма | 3,4 | 308,35 | 0,706578 | 436,3989 |
The difference between the electrical conductivity of softened and source water is:
436-420=16 µS/cm.
Then we will proportionally increase the water hardness value by equivalently reducing the sodium value. Let’s take 6 values,
Са = 0,0001; Mg=0,00005; Na=0,004 mol/l (Total hardness – 0,3 mg-eq/l);
Са = 0,0002; Mg=0,0001; Na=0,0037 mol/l (Total hardness – 0,6 mg-eq/l);
Са = 0,0004; Mg=0,0002; Na=0,0031 mol/l (Total hardness – 1,2 mg-eq/l);
Са = 0,0006; Mg=0,0003; Na=0,0025 mol/l (Total hardness – 1,8 mg-eq/l);
Са = 0,0008; Mg=0,0004; Na=0,0019 mol/l (Total hardness – 2,4 mg-eq/l);
Са = 0,0001; Mg=0,0005; Na=0,0013 mol/l (Total hardness – 3,0 mg-eq/l).
For each value of calcium, magnesium and sodium, we use the program to determine the electrical conductivity. The calculation results are shown in figure 1.
Figure 1
As you can see, the dependence is linear. This is a well-known condition, for example, used in a patent (DOPSLAFF HARTMUT; DOPSLAFF CARSTEN H + «Blending control method with determination of untreated water hardness via the conductivity of the soft water and blended water», Application number NZ20130720562 20131217).
It is known that the filtration cycle of softening installations is calculated based on the hardness of softened water of 0.1 mg-eq/l. A Softening installation starts to regenerate with sodium chloride, when the value of the hardness of softened water is not more than 0.5 mg-eq/l. Accordingly, if the conductivity of softened water is a value of 435 µs/cm, we can assume that the water ionic composition contains residual hardness not more than 0.1 mg-eq/l.
To verify this assumption, it is convenient to use a circuit with remote control of the measured conductivity value. To do this, it was decided to use the equipment of the company OWEN. The company has a cloud service where data on the measured electrical conductivity of softened water is transmitted and it is possible to track the values of electrical conductivity in the form of a time graph, which is extremely convenient for such tasks.
An electrical conductivity sensor with a transmitter with an output signal of 4-20 mA was purchased to measure the electrical conductivity. OWEN provided a set of equipment consisting of a programmable relay, a network gateway whit GPRS to transmit the values of electrical conductivity measured by the sensor to cloud.
This programmable relay is used to obtain the electrical conductivity of softened water through the 4-20 mA analog input from the electrical conductivity sensor. Then, using the network gateway connected via the RS – 485 interface, the value is passed to “OwenCloud”. On the OwenCloud service, the values of the electrical conductivity of softened water are visualized as a graph depending on the measurement time.
Using this control scheme, the following experiment was performed.
Water is softened in an ion exchange column. The initial water hardness is 2.8 mg-eq/l. The volume of ion exchange resin in the column is 0.55 l. The water flow through the column is 50 l/h. The linear filtering speed is 32 m/h. The calculated filter cycle is 220-240 liters.
Figures 2 and 3 show data for monitoring the operation of the water softening column from the moment of regeneration to the moment of complete depletion of the resin by sodium ions using the “OwenCloud” service.
Figure 2 shows that the column began to soften the water after regeneration at 11:15. It can be seen that the electrical conductivity of the softened water is slightly increased due to residual salts after cleaning the resin of the regeneration solution. Then the electrical conductivity of the water is kept almost constant. It is more convenient to track the level of electrical conductivity in figure 3, since you can clearly see the change in electrical conductivity due to the increased scale along the electrical conductivity axis.
At 12:00, in fact, at the beginning of the filtration cycle, the value of electrical conductivity was 380 µs/cm. The water analysis on the hardness showed a value of 0.04 mg-eq/l. For 2.5 hours, the conductivity of water remained at the level of 380 µs/cm and hardness not more than 0.05 mg-eq/l. After 3.5 hours of the operation of softening the hardness of the softened water was 0.12 mg-eq/l and the conductivity was at 377 µs/cm. The volume of softened water by this time amounted to 175 liters. Then for an hour there was a decrease in the electrical conductivity values up to 373 µs/cm, the hardness was 0.3 mg-eq/l, and then for another half an hour the conductivity fell to 372 µs/cm and hardness increased up to 0.5 mg-eq/l. After 6.5 hours of work, the conductivity began to fall rapidly, and after 7 hours reached a value of conductivity of the original hard water.
Using the obtained data, it is possible to build a dependence that will clearly show how much of the exchange capacity of the resin provides the filter cycle, and how much of the exchange capacity ensures the quality of the softened water obtained in the volume of this filter cycle. Figure 4 (4.1) shows the dependence of the softened water hardness on the volume of water that has been softened. As can be seen from the figure, the Hardness of the softened water no more than 0.1 mg-eq/l was provided in the volume of the filter cycle of 180 liters. At the same time, the hardness of the softened water equal to 0.2 mg-eq/l was provided in the volume of 250 liters. Only after 300 liters of softened water, the filtrate hardness was slightly more than 0.5 mg-eq/l. The recommended filter cycle for this amount of resin was 240 liters. The water hardness was no more than 0.2 mg-eq/l.
As we can see, a sharp increase in the filtrate hardness occurs after the filtrate hardness of 0.5 mg-eq/l. Therefore, it is recommended to begin the regeneration of the unit at the filtrate hardness of 0.5 mg-eq/l, and to calculate the filter cycle for the filtrate hardness of 0.1 mg-eq/l.
Figure 3 shows that a sharp decrease in conductivity occurred after the hardness of the filtrate was 0.3 mg-eq/l, respectively, knowing the difference in conductivity values between the treated and source water, and monitoring the level of degradation of the conductivity of the softened water we can clearly track the level of hardness of the filtrate installation softening.
Figure 3.1 shows the results for the conditions in figure 3 with a different method of installing the conductivity sensor, which allowed smoother measurements of the conductivity value. In this case, the nature of changes in the electrical conductivity of softened water is exactly similar to the experiment in figure 3.
Another very important point needs to be clarified. Let’s build the dependence between the electrical conductivity of the softening unit filtrate and its hardness obtained in the above experiment. The dependency graph is shown in figure 5. This is a dependency graph similar to the graph in figure 1. Only the graph in figure 1 uses calculated data. The difference is immediately obvious. The graph in figure 5 is linear only in the hardness range from 0.5 to 3.0 mg-eq/l. In the range from 0 to 0.5, there is a sharp nonlinear decrease in the value of water conductivity. To clarify this fact, several experiments were conducted and all of them confirmed this dependence. Probably this “special” behavior of electrical conductivity in the area of small concentrations of filtrate hardness ions is determined by the fact that calcium and magnesium ions at concentrations less than 0.1 mg-eq/l form ion pairs with sulfate and, accordingly, these ions do not carry an electric charge. When the concentration of hardness ions in the filtrate is 0.1 mg-eq/l, their uniform distribution starts in the ion shells and the decrease in the electrical conductivity values with increasing hardness of the filtrate is linear, because calcium and magnesium are embedded in the ion shells of chlorides, bicarbonates and sulfates.
This circumstance must be taken into account when the process of monitoring the operation of the softening unit is organized using the value of the electrical conductivity. Otherwise, the filter cycle of the softening unit can be defined as significantly smaller than it actually is, with all the ensuing consequences (overspending of salt, increased amount of waste water, etc.).
This control technology was introduced at the Saratov dairy plant. Screenshots of the recorded parameters of the Na – cation water unit in the steam boiler room of the Saratov dairy plant are shown in figures 6, 7, and 8.
The water conductivity value is measured using one and the same sensor. In this connection, the electrical conductivity of the source water is measured every 3.5 hours for 30 minutes. The rest of the time, the electrical conductivity of the softened water is measured. Due to this there are characteristic differences in the measurement of electrical conductivity values that are similar to the heart rate.
Using a single conductivity sensor allows you to avoid measurement errors and clearly control the difference between the electrical conductivity of softened and source water. A patent application has been filed for this solution.
In figure 6, you can see that the electrical conductivity of water is constantly monitored. When a peak is equal to 100, the electrical conductivity of the source water is measured for 30 minutes. The rest of the time (3.5 hours) the electrical conductivity is measured in the softened water. In figure 7, a peak-200 shows that one of the softening plant filters (TWIN) is being regenerated. You can see that the regeneration lasts for 2 hours.
Figure 8 shows a large scale on the ordinate axis. You can track how the electrical conductivity value changes between the source and softened water. In this case, the value of the source water is 353 µS/cm, softened 368 µS/cm. A difference of 15 µS/cm indicates that the softening unit produces softened water, i.e. it operates within the established filter cycle.
After installing this control system at the factory, the optimal filter cycle of the softening unit was determined within three days. It was – 35 m3. At the same time, the calculated filter cycle installed earlier in the automatic control valve of the softening unit was 25 m3. The filter cycle was increased by more than 30 %. An internal leak of the control valve of a single softening filter running on the TWIN system was also detected. Because of this, one filter constantly gave a small hardness increase, which was quite problematic to detect when I only performed one analysis of water hardness per month.
Currently, the softening system is constantly monitored remotely, which in the context of the coronavirus pandemic has proved to be an extremely effective and safe way to remotely control the water softening plant.
Summing up, I want to say that the control of the water softening process by electrical conductivity is quite possible and it significantly increases the efficiency of the softening plant without conducting constant chemical analyses of water for hardness. Moreover, the control of electrical conductivity, with the appropriately competent organization of the process, is more effective than conducting periodic chemical analyses. Since the electrical conductivity is measured constantly and immediately, it reacts to changes in the filtrate hardness. At the same time, this control makes it extremely easy to determine the optimal filter cycle of the softening plant and it significantly simplifies and improves the quality of commissioning at the softening plants. The hardness of the initial water of surface sources varies depending on the time of year. For example, the water hardness in the Volga river varies from 2.8 to 4.8 mg-eq/l (usually from 3.0 to 4.2). Accordingly, the filter cycle of softening plants operating on such water will change up to 30 %. Knowing this fact, modern automatic softening systems are set to the highest hardness value. At the same time, the filter cycle could be increased by 20-30% in the summer-autumn period (with minimal hardness). The conductivity control makes it easy to track this and even a non-specialist can set up the optimal filter cycle depending on the current value of the initial water hardness.
It is necessary to say that the way of measuring the electrical conductivity of water for monitoring water softening plants is very important and, in many ways, it is the main one. I developed and filed a patent for a method of measuring the electrical conductivity of water. Currently, the technology for controlling water softening systems is undergoing industrial testing.
Figure 2
Figure 3
Figure 3.1
Figure 4
Figure 4.1
Figure 5
Figure 6
Figure 7
Figure 8
