PROCESS OF ELECTRODIALYSIS FOR STABILIZING WINES WITH LOW WATER CONSUMPTION

Abstract

The present disclosure relates to a method for reducing water consumption in tartaric stabilization of wine by electrodialysis, comprising the following steps: passing an aqueous stream comprising a weak organic acid between a tank and an electrodialysis module; feeding the electrodialysis module with a stream of wine to be treated so that potassium bitartrate or calcium tartrate pass from the wine to be treated to the aqueous stream which flows between the tank and the electrodialysis module, so that during the passage of the aqueous stream through the electrodialysis module, the potassium bitartrate or calcium tartrate initially present in the wine stream are transferred to the aqueous stream and discarding the aqueous stream when it reaches a certain potassium bitartrate or calcium tartrate saturation point.

Claims

1. A method for reducing water consumption in tartaric stabilization of wine by electrodialysis, comprising: passing an aqueous stream comprising a weak organic acid between a tank and an electrodialysis module; supplying the electrodialysis module with a stream of a wine to be treated so that potassium bitartrate or calcium tartrate transfers from the wine stream to be treated to the aqueous stream that passes between the tank and the electrodialysis module, so that during a passage of the aqueous stream through the electrodialysis module, the potassium bitartrate or calcium tartrate initially present in the wine stream is transferred to the aqueous stream; and discarding the aqueous stream when it reaches a certain saturation point of potassium bitartrate or calcium tartrate.

2. The method according to claim 1 wherein the saturation point of potassium bitartrate or calcium tartrate in the aqueous stream occurs by the appearance of precipitate in the aqueous stream.

3. The method according to claim 1, further comprising filling the tank with water and adding the weak organic acid in order to obtain the aqueous stream.

4. The method according to claim 1, further comprising adding citric acid or malic acid to the aqueous stream.

5. The method according to claim 1, wherein discarding the aqueous stream comprises emptying the tank when the aqueous stream becomes saturated with potassium bitartrate or calcium tartrate.

6. The method according to claim 1, further comprising filling the tank with water by means of a first valve.

7. The method according to claim 1 wherein the weak organic acid is citric acid, malic acid or combinations thereof.

8. The method according to claim 6, wherein discarding the aqueous stream comprises emptying the tank by means of a second valve when the aqueous stream is saturated with potassium bitartrate or calcium tartrate.

9. The method according to claim 1, wherein discarding the aqueous stream comprises emptying the tank and wherein filling the tank and emptying the tank are carried out at a time interval less than or equal to 3 hours.

10. The method according to claim 1, further comprising adding citric acid to the aqueous stream such that a pH of the aqueous stream is at least 3.0.

11. The method according to claim 10, further comprising adding the citric acid to the aqueous stream such that the aqueous stream pH is 3.5-4.5.

12. (canceled)

13. (canceled)

14. The method according to claim 1, wherein the aqueous stream that passes between the tank and the electrodialysis module is a brine stream.

15. The method of claim 1, further comprising repeating passing the aqueous stream, supplying the electrodialysis with the wine stream, and discarding the aqueous stream until the wine stream is treated.

16. The method according to claim 8, wherein filling the tank with water by means of the first valve and emptying the tank by means of the second valve are carried out simultaneously in order to maintain a constant liquid level in the tank.

17. The method according to claim 8, wherein emptying the water tank by means of the second valve is carried out in laminar flow and at low speed by means of a distributor element.

18. The method according to claim 14, comprising a step of collecting the treated wine stream by means of a suction pump such that the pressure difference between the wine stream and the brine stream is at least 0.05 bar.

19. (canceled)

20. (canceled)

21. The method according to claim 1, wherein supplying the electrodialysis module with the wine stream to be treated is preceded by passing the wine stream to be treated through a particulate filter.

22. The method according to claim 1, further comprising adding a strong inorganic acid dissolved in water to the aqueous stream in order to dissolve precipitated potassium bitartrate or precipitated calcium tartrate in the electrodialysis module.

23. The method according to claim 14, further comprising identifying an increase in a degree of brine turbidity in the brine stream, identifying a decrease of flow rate of the brine stream, or identifying both the increase in the degree of brine turbidity in the brine stream and the decrease of the flow rate of the brine stream.

24. (canceled)

25. An electrodialysis device for carrying out the method according to claim 1.

26. (canceled)

Description

BRIEF DESCRIPTION OF THE FIGURES

[0062] These and other objects, features and advantages of the disclosure will be more apparent from the following detailed description when read in conjunction with the accompanying drawings.

[0063] FIG. 1 depicts a detailed scheme of one embodiment of the electrodialysis process for tartaric stabilization of wines.

[0064] FIG. 2 is a front view of the wine and brine streams exchange device in the electrodialysis module.

[0065] FIG. 3 depicts a scheme of an alternative process for implementing the brine renewal and discharging cycles.

DETAILED DESCRIPTION

[0066] An embodiment of he tartaric wine stabilization process is shown in FIG. 1. The raw wine to be stabilized, which is in a raw wine tank, continuously enters the tartaric stabilization process through the stream 10 and gets out partially deionized through the stream 44. This stream 44 has a non-return valve 14 to prevent treated wine, when under pressure in a treated wine tank, from returning to the membrane module after the process has been stopped.

[0067] During the continuous wine stabilization phase, valve 10A opens, and valve 11 and valve 41 are manipulated so that the wine passes through pump B1 and the treated wine flows out through stream 44. Raw wine passes through particulate filter 12 and enters the electrodialysis module via stream 17. Wine and brine flow rates are measured using rotameters 15 and 19, respectively. Wine treated by electrodialysis passes through the suction pump B3. Suction pump B3 is preferably a centrifugal pump so that there are no strong interactions between this pump B3 and pump B1 during start-up and during continuous process control.

[0068] During the continuous operation of treatment of raw wine, the specific conductivity of the raw wine and the treated wine is measured using the specific conductivity meters C1 and C2, respectively. The electric potential applied to the electrodialysis module is adjusted so that the reduction of the specific conductivity of the wine is equal to the desired value for the wine to be stable. In the electrodialysis module 100, potassium bitartrate or calcium tartrate pass from wine to the brine stream. This brine stream flows between tank 26 and electrodialysis module 100 through the streams 18 and 37.

[0069] In washing mode of the electrodialysis module 100, the wine is removed from the machine and valves 11 and 41 are manipulated in order to activate the streams 43 and 42. Thus, the washing solution, which is added to tank 26, circulates in all compartments of the electrodialysis module and returns to the tank 26. At the end, the washing solution is then discharged by opening the valve 36 or the valve 67.

[0070] One aspect of the disclosure consists in feeding water to the tank 26 discontinuously, during the wine treatment. Instead of operating with a given make-up water flow rate to the tank 26, the latter is periodically emptied and again refilled with acidified clean water at time intervals of preferably less than 2 hours. The main advantage of this mode of operation is that surprisingly it turns out possible to reduce direct specific water consumption to values below 5 L hL of wine, for deionization degrees of wine up to 35%. This means that electrodialysis can operate with a specific direct water consumption very similar to that of the cold process of tartaric stabilization,

[0071] In one embodiment, thus and under typical conditions of tartaric stabilization of wines, no precipitation of potassium bitartrate or calcium tartrate is observed and the process proceeds normally, even at relatively low temperatures of about 10° C. Experimentally, in the examples given, it was observed that the minimum direct consumption of water is obtained when brine is acidified with citric acid instead of sulfuric acid or nitric acid. For this reason, under the preferred operating conditions, the system should operate with brine acidified with citric acid.

[0072] In one embodiment, valve 67 periodically opens in order to discharge the brine into sewage 31, Once the tank is discharged, valve 28 opens and clean water enters into the tank through stream 33 until the tank is again refilled. This cycle repeats until all the wine is processed. Preferably, to prevent brine pump 62 from de-priming, valves 67 and 33 can be opened simultaneously in order to keep the liquid level constant in tank 26. In this case, the liquid level in the tank is measured with the level transmitter LT and valve 33 opens or closes in order to keep this level constant. Distributor 50 can also be used, which is designed so that clean water discharge is effected in laminar flow and at low speed, so that there is no significant mixing of clean water with concentrated brine, which is the denser and remains at the bottom of the tank.

[0073] Another aspect of the disclosure is the periodic addition of citric acid and a strong acid, such as nitric or sulfuric acid, to the brine. At the end of the brine discharge, water acidified with a food-grade strong acid may be passed for a short time, with a pH of less than 2, in order to dissolve potassium bitartrate or calcium tartrate, which may have precipitated in the electrodialysis module, The used volume of acidified water with strong acid should not exceed more than 1 L/hL of treated wine in a given cycle, one cycle being defined by filling and emptying tank 26 once. For this purpose, a strong acid metering pump 23 is used which injects strong acid from tank 21 to stream 93. The injection is made prior to the protective filter 20 of pump B2 to re-dissolve potassium bitartrate or calcium tartrate which may also have precipitated therein. After injection of strong acid, food grade citric acid is added to the brine at the beginning or during the cycle in order to keep the pH always below 4.0. The citric acid in the reservoir 22 is metered through metering pump 45.

[0074] Another aspect of the disclosure is a rigorous method of identifying whether there are wine losses or brine leaks to the wine in the electrodialysis modules based on monitoring the brine level in the tank, using the LT level meter. As valves 33 and 67 are closed, the water level in the tank should rise slowly due to the passage of water from the wine to the brine by electro-osmotic effect. The passage of current through the ion exchange membranes causes a residual water passage which is carried along with the ions. For this reason, the brine level tends to rise during wine electrodialysis. Under the conditions tested in the examples described below, this electroosmotic water flow was comprised between 0.3 and 0.5 L/(m.sup.2.h), depending on the degree of deionization of the wine. Once an experimental correlation between the electro-osmotic flow and the degree of deionization of the wine is determined, it becomes possible to estimate what is the typical increase in the brine level in a given time interval. If the change in level with time deviates from this correlation, there is a sign that either there is abnormal passage of wine to brine, or abnormal passage of brine to wine.

[0075] In a further embodiment, any change in brine level in a given time interval, which deviates from the normal system value, for a given degree of wine deionization, allows immediate recognition that either there are abnormal losses of wine or there is passage of brine to the wine. With the electrodialysis module tested in the examples described below, the electroosmotic water flow rate was around 10 to 20 L/h. Any deviation relative to this predicted value can only originate in abnormal wine losses or brine leaks into the wine. In practical terms, the system is able to detect this type of leak with an accuracy that is equal to half of the electro-osmotic water flow, i.e. between 5 and 10 L/h. Given that the nominal capacity of the electrodialysis machine used was 3000 L wine per hour, the system can detect wine or brine leaks of less than 0.15 to 0.3 L per hL of treated wine, which is extremely important for real-time machine diagnostics.

[0076] Another aspect of the disclosure comprises the precise control of the pressure difference between the wine stream and the brine stream within the electrodialysis module 100. In order to ensure that the risk of brine leakage to the wine is minimal, the pressure of the wine streams 17 and 38 should be at least 0.05 bar above the pressure of brine streams 18 and 37, respectively. This is done by measuring pressures using pressure gauges P1, P2 and P3 and regulating the frequency of pump motors B2 and B3 using a programmable logic controller (PLC). In continuous operation, the pressure of stream 17 is set at a given value by regulating the motor frequency of pump B1. The PLC then receives pressure's value from gauge P2 and regulates the frequency of pump motor B2 so that the pressure of stream 18 is at least 0.05 bar less than the pressure of stream 17. Preferably, this pressure difference should be comprised between 0.05 bar and 0.1 bar, it order to minimize wine losses. The pressure of stream 38 is maintained slightly above the pressure of stream 37 by regulating the motor frequency of the suction pump 83. The PLC receives the pressure recorded by the analog meter P3 and manipulates the motor frequency of suction pump 83 so that this pressure is 0.05 bar greater than pressure of stream 37. A particularity of this aspect of the disclosure is that the process is capable of reacting to increases in discharge pressure 44 of treated wine in order to ensure that there is never a dangerous pressure imbalance between streams 38 and 37. Increase of the pressure of stream 44, which has repercussions until stream 38, may occur when the wine level in the treated wine tank begins to rise.

[0077] Another aspect of the disclosure is a device for exchanging the streams 17, 18, 37 and 38 when the polarity of the electrodialysis module 100 is intended to be reversed. In this case, streams 17 with 18, and 38 with 37 need to be switched simultaneously. The daily change of polarity helps to increase the longevity of the electrodialysis module membranes and electrodes. This exchange can be done manually, but there is always the risk that the exchange is wrongly done and there will be streams crossing, passing the wine to the brine and the brine to the wine. An obvious way to overcome this problem is to use two synchronized four-way valves. These valves are, however, very expensive, increasing the cost of the process. In this disclosure, a hand-held device is used to exchange streams. This device is shown in FIG. 2.

[0078] In a first form of polarity of the electrodialysis module, wine enters by stream 17 and exits by stream 38, through flexible tubes, while brine enters by stream 18 and exits by stream 37, also through flexible tubes. The flexible tubes fit into the metal tubes 108, 106, 120 and 121 through tri-clamps 105. The metal tubes are immobilized by mechanical connections to the top and bottom of the electrodialysis module. The flexible tubes, at the same time, are attached to the metal plates 103 and 104. These metal plates are centrally connected to a metal axis 101 and can rotate about this axis. The metal axis 101 is supported by the support bars 110 and 111 which are welded to the metal tubes 106, 121, 108 and 120 as shown in FIG. 2. The handle 102 is attached to the metal plates 103 and 104 in order to ensure that the plates rotate together.

[0079] To switch the streams, the tri-clamps 105 are opened and the metal plates 103 and 104 rotate 180 degrees with the aid of the handle 102. After the metal plates 103 and 104 rotate 180 degrees, the stream 17 passes through the metal tube 108, the stream 18 passes through the metal tube 120, the stream 37 passes through the metal tube 121, and the stream 38 passes through the metal tube 106. Thus, the module becomes able to operate in the second polarity mode, i.e. with inverted polarity, compared to the initial polarity.

EXAMPLE 1

Citric Acid

[0080] Industrial scale trials of tartaric stabilization of wines were performed to determine specific water consumptions, using a pilot plant with a scheme similar to that of FIG. 1. Valves 33 and 67 were manually operated during brine discharge and fresh water addition to the tank, This facility had a nominal capacity of 3000 L of wine/hour, a tank 23 with a nominal volume of 200 L. The electrodialysis module was supplied by General Electric (USA) using CR67 homogeneous cation exchange membranes and AR204 homogeneous anion exchange membranes, with an effective area of 30 m.sup.2 of membrane pairs. The wine feed pressure was 2.0 bar and the brine feed pressure was 1.95 bar. The pressure before non-return valve 75 was 1.20 bar and the pressure set point P3 before pump was adjusted to 1.25 bar. Brine pH was maintained between 3.5 and 4.0 throughout all trials, with only a concentrated citric acid solution added to tank 26. It was not necessary, in any trial, to use a strong acid to lower the brine pH to clean the electrodialysis module. The load loss in the module was always constant throughout the tests, which showed no precipitation of potassium bitartrate or calcium tartrate.

[0081] The results regarding these pilot trials of tartaric stabilization of wines are shown in Table 1. Three red wines and two white wines with wine volumes up to 32 000 L were tested. The facility worked without precipitation problems of potassium bitartrate or calcium tartrate, and without severe membrane clogging, during 22 hours of operation over 3 days. The tank was cyclically filled with 150 L of fresh water. At the end of the cycle, water was only added after the brine began to reach the bottom of the tank. Due to the manual operation of valves 67 and 33, there was always a fraction of the brine that was never fully discharged in order to prevent pump B2 from de-priming.

[0082] The results presented in Table 1 show that under typical conditions of tartaric stabilization of wines, it is possible to achieve specific consumptions that are less than 5 L/hl of wine, in one of the cases being it possible to be up to 2.6 L/hL. The residence time of the brine in the tank was always comprised between about 1.5 and 2.0 hours.

[0083] The brine tank was quite oversized in the pilot plant, its volume being possible easily reduced to 50 L in a new plant. This 50 L volume still contains enough water to easily wash the membrane module. For this reason, the average recommended times for brine replacement with fresh water can be reduced by up to half an hour. Reducing residence time lowers the risk of precipitation of potassium bitartrate or calcium tartrate or other poorly soluble salts. In the limit, the brine residence time in the tank can still be further reduced, but in this case it is already necessary to use, in addition to the brine tank, a second larger tank to wash the electrodialysis module.

[0084] The water consumption results in Table 1 can still be normalized by dividing the water consumption by the degree of specific conductivity reduction in %. This calculation for the tested wines gives a water consumption of the electrodialysis process which is approximately comprised between 0.1 and 0.2 (L/hL of wine)/(% specific conductivity reduction). Given that the amount of potassium bitartrate or calcium tartrate passing into brine increases linearly with the degree of deionization, the previous range of water consumption per % of specific conductivity reduction allows an estimate of the average water consumption for a given wine intended to be stabilized. In these trials, the electro-osmotic water flow rate from wine to brine ranged from 10 to 20 L/h, depending on the degree of deionization of the wine.

TABLE-US-00001 TABLE 1 Results of operation of tartaric stabilization of various wines by electrodialysis. Time Total interval Average Initial duration between water spec. % time disch. consumption Wine cond. reduction of the Number of brine (L H.sub.2O/ Wine vol. (miliS/ of spec. T process of brine tank hL type (L) cm) cond. (° C.) (min) discharges.sup.a) (min) wine) red 32000 2.75 15% 10.6 690 6 115 2.8 red 10500 2.28 21% 10.8 235 2 117.5 2.9 red 8000 2.62 16% 10.9 180 2 90 3.8 white 5000 1.52 25% 10.5 130 1 130 2.6 white 4000 1.82 26% 13.5 95 1 95 3.8 .sup.a)The volume of fresh water added to the tank in each cycle was 150 L

EXAMPLE 2

Citric Acid vs. Sulfuric Acid

[0085] To study the impact of using citric acid on reducing the water consumption of electrodialysis, the same batch of wine was stabilized in two different trials using sulfuric acid in one case and citric acid in another one to acidify the brine. The electrodialysis prototype used was the same as in example 1, operating with discontinuous water addition and brine discharge.

[0086] In each test the red wine flow rate was 3000 L/h and 6000 L of wine was processed. Red wine had an initial specific conductivity of 2380 micros/cm, the degree of deionization imposed was 25% and the brine pH was maintained between 3.5-4.0 by periodic addition of acid. The applied electric potential was 150 Volt and the electric current was comprised between 14.5-17 Ampere.

[0087] In the citric acid trial, about 1 kg of acid/hour was used to keep the pH within the previous range. In the sulfuric acid trial about 0.5 kg h was used.

[0088] In the citric acid trial, the brine tank was initially filled with 200 L of water. With this volume it was possible to process 6000 L of red wine, without occurring precipitate formation in the brine. Under these conditions, water consumption was 3.3 L of water/hL of wine. A 750 mL sample of brine was also taken at the end of the cycle which was left at room temperature for 24 hours. At the end of this period, a precipitate formed at the bottom of the bottle was observed, indicating that the brine was supersaturated. After filtration and drying of the precipitate, 5.5 g/L was found to precipitate.

[0089] In the sulfuric acid trial, the brine tank was initially filled with 200 L of water. With this volume it was possible to process 3750 L of red wine. In this sulfuric acid trial, white and fine precipitates were formed in the brine tank after 75 minutes. At that time the tank brine was immediately discharged and water was introduced again to continue the process. In this case, the water consumption, referred to a brine charging and discharging cycle, was 5.3 L of water/hL of wine.

[0090] Thus, in these trials there was a saving of 2 L of water/hL of wine when citric acid was used versus sulfuric acid.

[0091] These two tests show that the use of citric acid makes it possible to operate under conditions of maximum brine supersaturation. On the other hand, this example also shows that it is possible to simply recover the potassium bitartrate or calcium tartrate contained in the brine under operating conditions with citric acid, as they precipitate naturally and at room temperature.

EXAMPLE 3

Citric Acid vs. Nitric Acid

[0092] To study the impact of using citric acid on reducing the water consumption of electrodialysis, the same batch of wine was stabilized in two different trials, using nitric acid in one case and citric acid in another case to acidify the brine. The electrodialysis prototype used was the same as in examples 1 and 2, operating with discontinuous water addition and brine discharge.

[0093] In each test the red wine flow rate was 3200 L/h and a total of 17000 L of wine was processed. The red wine had an initial specific conductivity of 2510 microS/cm, the degree of deionization imposed was 30%, and the brine pH was maintained between 3.5-4.0 by periodic addition of acid. The applied electric potential was 150 Volt and the electric current was comprised between 15-21 Ampere.

[0094] In the nitric acid trial, the brine tank was initially filled with 200 L of water, and nitric acid was added to ensure an initial pH of 3.0. Throughout the process, doses of nitric acid were added to maintain the brine pH between 3.5-4.0. After 73 minutes the brine was cloudy, its flow rate decreased and precipitate formation began. Under these conditions, water consumption was 5.1 L of water/hL of wine.

[0095] In the citric acid trial, the process was started with the brine tank equally filled with 200 L of water. In this case the amount of citric acid required to obtain an initial pH of 3.0 was added. Throughout the process, doses of citric acid were added to maintain the brine pH also between 3.5-4.0. The process proceeded normally with a decrease in flow rate only at 109 minutes. The brine was cloudy but with lower turbidity degree than in the nitric acid test. Under these conditions, water consumption was 3.4 L of water/hL of wine.

[0096] Thus, in these trials there was a saving of 1.7 L of water hL of wine when citric acid was used versus nitric acid.

[0097] These two trials show that with the use of citric acid it is possible to operate under maximum brine supersaturation conditions and to treat more wine with the same volume of water.

EXAMPLE 4

Malic Acid

[0098] An additional industrial scale tartaric wine stabilization test was performed to determine specific water consumptions using the same pilot plant as examples 1 to 3 but with the changes shown in FIG. 3 and using malic acid to acidify the brine. In order to reduce water consumption, a new feed tank 200 with two compartments 201 and 202 was used. The volume of each tank 201 and 202 was 100 L. The valves 28, 210, 220, 36 and 67 were manually operated during brine discharge and fresh water addition to the tank. This new configuration makes it possible to minimize the amount of concentrated brine residual left in the system after the brine charging and discharging cycle.

[0099] In this example 40000 L of red wine with an initial conductivity 2390 micros cm and a pH 3.53 was used, with the desired deionization degree of 20%. The wine was fed continuously to the machine at a flow rate of 3000 L/h and the brine was renewed periodically in cycles with a duration of about 75 minutes. The electric potential applied to the membrane module was adjusted to reduce the specific electrical conductivity of the wine by 20%. The wine feed pressure was 2.0 bar and the brine feed pressure was 2.5 bar. The pressure before non-return valve 75 was 1.20 bar and pressure set point P3 before pump was adjusted to 1.25 bar.

[0100] The trial began with the compartments 201 and 202 filled with water. 4.5 L of a solution with 150 of malic acid was added to compartment 201. The valve 220 was then opened and the valve 210 was closed in order to circulate the brine through the stream 93. At the end of each cycle, the valves 67, 210, 28 and 36 were opened and the valve 220 was closed. Immediately fresh water entered into the brine stream 93 in order to completely remove the concentrated brine flowing in the membrane module. At the same time, the concentrated brine was discharged to the sewage 31.

[0101] The time required for fresh water to flow out through stream 37 was about 30 seconds. At the end of this time interval, valves 67 and 36 were again closed, and valve 220 was opened, and 4.5 L of the malic acid solution was added manually. The valves 28 and 210 had to be kept open for about 5 minutes to allow compartments 201 and 202 to refill again with fresh water. After this step, the valves 28 and 210 were closed.

[0102] During the test, the pH, specific conductivity and turbidity of the brine were recorded using manual meters. Although turbidity measurement was done manually, FIG. 4 proposes the placement of a TT turbidity meter in order to automate the process. The brine flow rate, which was measured with a rotameter, was also recorded. The results regarding this red wine tartaric stabilization test are presented in Table 2 for the second brine renewal cycle. These results show that it is possible to operate with 75 min brine discharge cycles. In the final phase of the cycle, a reduction in brine flow rate was observed and an increase in brine turbidity, originated by the presence of small crystals that were visible in the rotameter. After removal of brine and introduction of fresh water with malic acid, the brine flow rate returned to the normal value, which showed that there was no irreversible blockage of the brine channels of the membrane module. This test shows that, with the process herein described wherein malic acid is used, it is possible to make tartaric stabilization of a red wine by electrodialysis, with 20% deionization, using only 2.7 L of fresh water/hL of wine. The test further shows that brine discharge can be done immediately when the brine flow rate begins to decrease, or when turbidity begins to increase significantly.

TABLE-US-00002 TABLE 2 Operating results of tartaric stabilization of a red wine during the second cycle of brine water renewal, using malic acid to acidify the brine. The results presented refer to the brine..sup.a Specific electric conductivity Time (min) pH (microS/cm) Turbidity (NTU) Flow rate 0 3.7 5.48 5 3000 3 3.9 9.10 7.8 3000 36 4.4 15.30 12 3000 45 4.6 17.50 17 3000 57 4.7 19.50 15 3000 72 4.9 >20 24 2980 75 4.9 >20 50 2900 .sup.aThe processed red wine had a specific electrical conductivity of 2390 microS/cm and a pH 3.53. The degree of deionization of the wine by electrodialysis was 20% and the temperature of the wine was 14° C. The brine was initially acidified with 4.5 L of aqueous malic acid solution with a concentration of 150 g/L.

[0103] Although only particular embodiments of the present disclosure have been represented and described, the skilled person will be able to introduce modifications and replace some technical characteristics for equivalent ones, depending on the requirements of each situation, without departing from the scope of protection defined by the appended claims.

[0104] The embodiments presented are combinable with each other. The following claims further define embodiments.

REFERENCES

[0105] [1] Bories, A., V. Sire, D. Bouissou, S. Goulesque, M. Moutounet, D. Bonneaud, and F. Lutin. “Environmental impacts of tartaric stabilisation processes for wines using electrodialysis and cold treatment.” South African Journal of Enology and Viticulture 32, no. 2 (2011): 174. [0106] [2] Forsyth, Karl. “Comparison between electrodialysis and cold treatment as a method to produce potassium tartrate stable wine.” AWRI Project number: PCS 10004 (2010): 53-58. [0107] [3] Allison, Robert P. “High water recovery with electrodialysis reversal.” In Proceedings American Water Works Assoc. Membrane Conference, pp. 1-4. 1993.