Background Art During recent decades, nitrate contamination in raw water sources have been increasing due to the intensive use of nitrogenous fertilizers, changes in land-use patterns (from pasture to arable), and the contamination of sewage and industrial effluents. It has been found that nitrate levels have been increasing in drinking water supplies in the European Economic Community, the United States, Canada, Africa, the Middle East, Australia, and New Zealand (Kappor and Viraraghavan, 1997). Because an increased nitrate uptake can link to several health hazards causing methaemolobinaemia or cancer risks due to nitrosamines or nitrosamides, limits have been set up to regulate the maximum nitrate levels in drinking water. In Europe, an EEC Directive on the quality of drinking water for human consumption specifies a maximum admissible concentration of 50 mg NO3-/l, but a guide level of 25 mg NO3-/l is recommended (European Community, 1980). The U. S. EPA establishes a maximum contaminant level (MCL) of 10 mg NO3--N/l corresponding to 44 mg N03-/l (Pontius, 1993). The Health and Welfare Canada has established a maximum acceptable concentration (MAC) of 10 mg N03--N/l and the nitrite of 3.2 mg/1 when nitrates and nitrites are measured separately in drinking water (Health and Welfare Canada, 1993).

Several technologies can be used for nitrate removal from waters. The chemical technologies include ion exchange and catalytic denitrification. The membrane technologies involve reverse osmosis, nanofiltration, electrodialysis and electrodeionization. Biological denitrification has also been applied. When comparing the various technologies in terms of effectiveness, ease of operation, reliability, cost and suitability for environmental regulation, ion exchange is found to be a highly competitive technology and is used in commercial large-scale applications, especially for drinking water treatment.

Conventional ion-exchange technology involves a process, in which the nitrate contained in water is exchanged with another anion that is already adsorbed on the strong basic ion-exchange resin. Also competing anions present in the water take part in this exchange and are adsorbed to the resin. This condition lowers the efficiency of nitrate removal from waters contaminated with e. g. sulphate.

After a period of operation, the exchange capacity of the resin is exhausted. The exhausted resin is then regenerated with a regeneration chemical. During the regeneration cycle, the nitrate is released from the resin and a waste stream is produced together with the regeneration chemical. The nitrate can be reduced to nitrogen, but the waste stream containing the regeneration chemical is still remaining. Disposal of the waste stream may lead to a significant waste disposal cost or regulatory complications as well as to loss of the regeneration chemical. The disposal of the waste solution is a big challenge for the application of the ion exchange technology. Attempts to recycle the regenerating solution have been made, but have failed because competitive anions are accumulated in the spent regenerant and the regeneration efficiency is not maintained.

The European patent application 291,330 discloses a process for treatment of ground-water containing nitrate, comprising the steps of contacting the water with an ion- exchange resin, regenerating the exhausted ion-exchange resin with a regenerant, and reconstituting the spent regenerant by subjecting it to electrolysis. In this process, however, chlorine gas may undesirably be evolved. In addition, sodium hypochlorite, which will destroy the ion-exchange resin, will be formed.

The British patent 1,432, 020 discloses a process for recovery of spent regenerating solutions for re-use in an ion-exchange system. This is an extensive process aiming at recovering the different components of the spent regenerating solutions by precipitation and electrodialysis.

Summary of the Invention One object of the present invention is to provide an integrated process allowing efficient removal and destruction of nitrate from waters contaminated with other anions.

A second object of the present invention is to provide such a process where spent regenerant solution can be continuously re-used.

Another object of the present invention is to provide such a process which does not give rise to any waste streams.

Still another object of the present invention is to provide such a process where competing anions are not accumulated in the re-used regenerant.

A further object of the present invention is to provide such a process where the regeneration efficiency is maintained.

Other objects of the invention should be apparent to a person skilled in the art when reading the more detailed description of the invention as presented below.

These objects are achieved by a process for removal and destruction of dissolved nitrate from water containing the same, which comprises a removal step in the form of an anion-exchange operation, where nitrate is eliminated from said water in the form of a more concentrated solution thereof, said anion-exchange operation being nitrate selective; a destruction step, where said more concentrated nitrate solution is subjected to an electrolysis operation for nitrate destruction by electrochemical reduction thereof; a conditioning step, where solution resulting from the destruction step is adjusted by the addition of chloride ions so as to enable use of said solution as a regenerant in the anion-exchange operation while maintaining an operating capacity of said anion-exchange operation which is substantially unchanged from one anion-exchange operation to another; and a recycling step, where solution resulting from the conditioning step is used as a regenerant in the anion-exchange operation.

The use of a removal step with a nitrate selective ion-exchange operation makes it possible to remove nitrate ions also from waters contaminated with competing anions, without substantially lowering the efficiency of the removal.

The destruction step with an electrolysis operation where nitrate is reduced to nitrogen assures effective destruction of nitrate.

The conditioning step makes it possible to maintain the regenerating efficiency of the recycled regenerant by a sufficiently high chloride concentration to displace nitrate ions from the ion-exchange resin and thereby to maintain the operating capacity of the ion-exchange operation.

The fact that a minimum of competing anions are adsorbed to the ion-exchange resin, combined with efficient regeneration of the resin, means that less competing anions are accumulated in the spent regenerant.

The recycling step completes the process and enables an integrated process with complete re-use of spent regenerant and no waste streams.

For the sustainable operation of an ion-exchange system with continuous recycling of spent regenerant, three aspects must simultaneously be considered: the nitrate contained in the spent regenerant must be destructed as much as possible the accumulation of competitive anions in the spent regenerant must be discontinued in some way, and high regeneration efficiency for the ion-exchange bed must be maintained as the concentration of regeneration solution is changed with continued regeneration cycles.

Brief Description of the Drawings Figure 1 shows a flow diagram for a typical system configuration to perform the process of the present invention.

Figure 2 and 3 show results from Example 2 below.

Detailed Description of Preferred Embodiments of the Invention The integrated process comprises four steps: nitrate selective removal from water by ion exchange, electrochemical destruction of the nitrate in spent regenerant, conditioning of the spent regenerant, and recycling of the conditioned regenerant. According to the invention, preferably all, but at least a major part, of the spent regenerant is recycled.

The key points of the process are to keep mass balances for accumulated anions in a succession of repeated process cycles, and to maintain the regenerating

and operating efficiencies throughout repeated cycles.

When sulphate is present the mass balance for sulphate is achieved by a steady-state accumulation in spent regenerant.

The steady state is a dynamic phenomenon related to the re-use of spent regenerant in multiple operation and regeneration cycles. It depends on the properties of the ion-exchange resin, the concentration of anions in the raw water, and the conditions of operation and regeneration. If nitrate removal can be performed under such steady state conditions, a continuous process can be set up for multiple operating and regenerating cycles without the discharge of spent regenerant.

The efficiencies of operation and regeneration are maintained in two ways. In regeneration, the nitrate adsorbed in an operation cycle must be replaced by the regenerant as much as possible. Although both chloride and sulphate may theoretically be used as regenerants for an anion resin to displace nitrate, high concentration of sulphate in a regeneration solution gives a negative influence on regeneration efficiency. In operation, the removal of nitrate is achieved mainly by selective displacement using nitrate selective anion-exchange resins.

Selective nitrate removal by ion-exchange In the first step of the process, nitrate is removed from water by a selective anion-exchange resin. A nitrate selective resin has higher affinity for nitrate than for other major anions present in the water. The purposes of using a nitrate selective resin are to increase the operating capacity of the ion-exchange operation when the ratio of competing anions, e. g. sulphate, to nitrate is high in the water; to reduce the adsorption of other anions to the ion-exchange resin; and to decrease the accumulation of other anions in the regeneration solution.

Two commercial nitrate selective anion-exchange resins (IMAC HP555 from Rohm & Haas and A520E from Purolite) have been tested in their chloride forms for ion-exchange. The relative affinities of these types of resins for anions present in water are: N03-> S042-> Cl-> HC03- (1) The relative affinities of a regular strong base anion- exchange resin for these anions are usually as follows: S042-> N03-> C1-> HC03- (2) Because the selective resins have a higher affinity for nitrate compared to other major anions present in nature waters, the adsorption of other anions becomes much less than to regular anion-exchange resins. The higher affinity of a resin for nitrate, the larger nitrate removal capacity the resin has.

When the nitrate removal capacity of the resin bed is exhausted in an operation cycle, the anions adsorbed together with nitrate in the operation cycle are replaced by chloride. During a regeneration cycle, co-current regeneration is carried out to recover the capacity.

Although it is impossible to totally separate nitrate from other anions in the ion-exchange operation cycle, the selective displacement using the nitrate selective resin makes the adsorption of other anions limited.

Carbonate, chloride and sulphate are major anions in most of nature waters. As a nitrate selective resin is used for the ion-exchange system and sodium chloride is used as the regenerant, sulphate is mainly concerned for the anion accumulation in spent regenerant according to the relative affinities of the resin for these anions (equation 1).

Electrochemical destruction of nitrate The nitrate reduction can be performed by any electrochemical method suitable for that purpose, but is preferably carried out in a two-compartment electrochemical cell equipped with cation-exchange membranes. A wide range of commercially available cation exchange membranes are suitable to incorporate in the cell and be used in the process. As cathode, graphite, platinum, platinised titanium, lead, nickel, nickel mesh, copper, or copper mesh electrodes have proved to be especially efficient.

The nitrate reduction on the cathode can be described as follows: M + N03-aq (M) NO3 adg (3) Nomads + Hz0 + e- # NO2- ads M + H20 + e M (H) ads + OH- (5) 2NO2~ads + 8M (H) ads N2g + 4H20 (6) Catalytic hydrogenation could be a part of the reduction of the nitrite and/or nitrate adsorbed on nickel graphite or platinum cathodes. On lead and copper cathodes electronation/protonation are the main reactions of the nitrate reduction NO3- ag 2NO2-aq + 8H+ + 2e-N2 + H20 (8) The main side reactions are hydrogen evolution H20 + 2e- # H2 + OH- (9) and nitrite reduction to ammonia

NO2-aq + 8H+ + 6e-j NH3 (10) The anolyte contains sulphuric acid. In the anodic compartment the main reaction is water electrolysis, oxygen evolution and proton generation: 2H20-4e'-Oz + 4H+ (11) Protons generated during the water oxidation are transferred to the cathodic compartment.

The main roles of the cation exchange membrane are to prevent re-oxidation of the intermediates and the products of the nitrate reduction process, to transfer protons generated on the anode from the anodic compartment to the cathodic compartment, and to avoid chlorine evolution on the anode, which would lead to the presence of active chlorine or to the production of hypochlorite (ClO-) or chlorate (Cl03-). These oxidative species can cause serious damage to the anion exchange resin.

Conditioning of spent regenerant In addition to discontinued accumulation of competing anions in spent regenerant, a high regeneration efficiency at such steady-state concentrations is required in order to have a low residual nitrate in treated water during succesive operation cycles. A high regeneration efficiency is maintained as for instance the sulphate concentration in the spent regenerant increases during the re-using cycles until a steady state is reached. It has been found that the regeneration efficiency is decreased as the sulphate concentration is relatively higher and the chloride concentration is relatively lower. However, the negative influence of sulphate on the regeneration can be reduced when a relatively higher concentration of chloride and/or a

relatively higher ratio of chloride to sulphate is kept in the spent regeneration solution. A NaCl concentration higher than about 9 % wt (corresponding to a chloride concentration higher than about 5 % wt) and/or a ratio in equivalents of chloride to sulphate higher than about 6 in the spent regeneration solution is an especially preferred value to keep a high efficiency of regeneration.

Moreover, a suitable regeneration level should also be considered. The regeneration level is defined as the weight of regenerant used to regenerate a certain volume of ion-exchange resin. The absolute regeneration level includes regenerant added in the conditioning step and regenerant re-used from the last cycle. The absolute regeneration level does not strongly affect the economy of the process because the actual regeneration level (only accounting for regenerant added in conditioning) is much lower than the absolute regeneration level. However, the regeneration level affects the operating capacity and the quality of treated water in the following cycle. An absolute regeneration level higher than about 350 g NaCl/1 resin is recommended to keep a high efficiency of regeneration.

In addition, deionized or softened water should be used for conditioning of the regenerant and rinsing of regenerated resin to prevent from the precipitation of calcium sulphate and/or magnesium sulphate.

Recycling of conditioned regenerant After nitrate destruction and conditioning, the spent regenerant is re-used for regeneration of the an- ion exchange operation during another process cycle.

Thus, a process which does not give rise to any waste streams is provided.

Figure 1 schematically shows one embodiment of an apparatus for performing the process according to the

invention. The apparatus shown in the figure includes the following elements: an ion-exchange column 1, a regenerant tank 2, a spent regenerant tank 3, an electrochemical cell 4, and an anolyte tank 5.

Said apparatus works in the following way: Nitrate contaminated water passes through the ion-exchange column 1, loaded with chloride ions, where nitrate ions present in said water are adsorbed on the nitrate selective anion-exchange resin and are replaced by chloride ions.

When the ion-exchange column 1 is exhausted, it is regenerated with a sodium chloride solution from the regenerant tank 2. Nitrate is removed and replaced by chloride ions.

The spent regenerant solution, containing nitrate, is collected in the spent regenerant tank 3. Spent regenerant from the spent regenerant tank 3 and anolyte from the anolyte tank 5 are circulated through the electrochemical cell 4, where nitrate is reduced to nitrogen gas.

The spent regenerant is then transferred to the regenerant tank 2, where it is conditioned by addition of sodium chloride.

The conditioned regenerant in the regenerant tank 2 is re-used for regeneration of the ion-exchange column 1 in another process cycle.

In another embodiment, the apparatus comprises multiple ion-exchange columns to allow use of one column for nitrate removal when another is being regenerated.

Examples A nitrate removal process is demonstrated in the following examples. The process is an integrated process with nitrate removal by nitrate selective ion-exchange, destruction of nitrate in the spent regenerant in an electrochemical cell, and the full re-use of the regenerant by hindering the accumulation of sulphate and

conditioning the concentration of sodium chloride in the spent regeneration solution.

A synthetic waste water containing 100 mg/1 of nitrate, 500 mg/1 of sulphate and 160 mg/1 of chloride was used as raw water. The concentration of the synthetic water represented a relatively harsh situation compared to natural waters, especially on the ratio of nitrate to sulphate. During operation, the synthetic water was treated by passing it through an ion-exchange column (0 29 mm) with 20 g of nitrate selective resin (IMAC HP555 or A520E). The operation flow rate was usually 20-25 bed volumes/h (BV/h).

The nitrate concentration in the effluent was controlled to be less than 40 mg NO3-/1, which was taken as the end point of an operation cycle. As the nitrate concentration in effluent reached 40 mg N03-/l, the operation cycle was stopped, a regenerating cycle was then started using NaCl solution. Meanwhile, the spent regenerant was collected for electrolysis, conditioning and re-use in the next regeneration cycle.

All experiments were carried out with commercially available ElectroMP and ElectroSyn cells equipped with Nafion 324 or Nafion 350 type cation exchange membranes.

A wide range of commercially available cation exchange membranes are suitable to incorporate in the cell and be used in the process. The ElectroMP cell has a minimum geometrical surface area of 100 cm2. The minimum surface area in the EletroSyn cell is 400 cm2 for each electrode.

PVDF or PP frames and EPDM gaskets (O-rings) were incorporated in the cell for proper sealing. A PE turbulence promoter provided uniform fluid distribution and good mass transfer conditions. The operation mode for all experiments was batch mode.

As cathode, a graphite electrode was used in such a way that solid graphite acted as current feeder which was pressed/glued to graphite felt. The catholyte compartment therefore was packed; the distance between the membrane

and the cathode was reduced to 2 mm from the original 9 mm gap.

The electrolytes of the anodic and cathodic chambers had a volume of 2 litres in the ElectroMP cell. The anolyte was 0.9 M of sulphuric acid. The acid concentration remained stable after several cycles of electrolysis. The anolyte level was maintained the same; deionized water was added to the anolyte, as a part of the water had been electrolysed.

The pressure on the cell was maintained at 0.2 bar, which provided a constant flow in each compartment. The temperature of the electrolytes was below 40 °C. The process parameters are summarised in Table 1.

Table 1 The process parameters of the electrochemical nitrate reduction Active electrode surface 200-400 cm2 Current density 1 kA/m2 Electrode gap 2 mm cathode side; 9 mm anode side Temperature 25-34 °C Superficial velocity 0.5 m/s Membrane Nation 324,350 The operation-regeneration-operation cycles were continuously run without discharge of spent regenerant. A steady-state concentration of sulphate was reached in the spent regeneration solution after the initial cycles.

Because of the complete re-use of spent regenerant, a high regeneration level and co-current regeneration were used for this process.

The efficiency of the continuous process for nitrate removal in such a succession of repeated cycles is illustrated in the following examples.

Example 1 The effluent quality of the ion-exchange system was good in continuous operation cycles. As shown in Table 2 and Table 3, the average concentrations of residual nitrate in the effluent were below 15 mg N03-/1 when the control limit was set to less than 40 mg NO3-/l. The concentrations of other anions in the effluent were close to those in the raw water. Nitrate was selectively removed in the operation cycles.

A comparison between an early cycle and a later cycle is also given in Table 2 and Table 3. Although the early operation cycle (cycle 2) was regenerated by a regenerant containing a relatively lower concentration of sulphate and the later operation cycle (cycle 9) was regenerated under a high (steady-state) concentration of sulphate, the effluent quality did not show any significant difference between these operation cycles.

The efficiency of nitrate removal in operation was maintained throughout the operation cycles.

Table 2 The effluent quality of the ion-exchange column using the nitrate selective resin IMAC HP555

Cycle 2 Cycle 9 Effluent N03-S042-Cl-pH Effluent N03-SOg2-Cl-pH (BV) (mg/1) (mg/1) (mg/1) (BV) (mg/1) (mg/1) (mg/1) 12 10 4 584 6.27 11 6 0 537 6.73 30 8 26 563 6.19 29 7 16.2 558 6.74 48 11 158 470 6.51 48 7 474 129 6.75 66 6 377 311 6.19 67 7 358 307 6.24 84 6 501 213 6.14 86 6 441 192 6.42 104 5 547 181 6.21 104 7 526 184 6.34 122 6 557 171 6.21 125 7 538 177 6.40 142 7 555 175 6.25 144 8 567 183 6.41 163 11 557 173 6.32 163 10 536 172 6.48 172 14 558 174 6.27 173 11 534 173 7.19 181 17 551 172 6.24 182 14 561 179 6.44 190 21 545 174 6.18 191 18 567 180 6.38 199 28 542 170 6.23 200 19 534 169 6.35 209 34 538 170 6.18 209 25 527 167 6.36 218 39 537 174 6.19 218 31 529 167 6.37 Average 15 437 258 6.24 Average 12 447 232 6.51 Notes: Cycle 2 is an early operation cycle that has been regenerated by re-use of spent regenerant containing a relatively lower concentration of sulphate. Cycle 9 is a later operation cycle that has been regenerated using spent regenerant containing the steady- state concentration of sulphate.

Table 3 The effluent quality of the ion-exchange column using the nitrate selective resin A520E

Cycle 2 Cycle 9 Effluent NO3-so42-Cl-pH Effluent N03-SOg2-Cl-pH (BV) (mg/1) (mg/1) (mg/1) (BV) (mg/1) (mg/1) (mg/1) 11 9 0 537 6.80 10 10 6 549 6.81 28 9 12 566 6.49 28 10 14 565 6.54 45 8 89 513 6.36 46 10 118 488 6.22 63 8 325 338 6.02 64 15 609 598 6.12 80 8 470 221 6.12 81 8 477 230 6.00 98 7 521 189 5.98 98 7 534 192 6.06 117 7 530 174 5.71 115 7 554 185 5.66 134 7 529 172 6.07 132 7 554 180 5.57 152 10 538 178 5.54 150 8 561 180 5.55 162 8 539 175 5.60 158 8 550 178 5.54 170 9 533 177 5.61 168 10 564 179 5.57 180 10 538 175 5.61 177 11 550 173 5.56 188 12 531 174 5.55 186 8 552 177 5.66 197 14 529 174 5.58 195 10 557 178 5.67 206 17 532 173 5.70 203 13 545 175 5.50 214 20 530 171 5.78 212 18 554 177 5.53 223 26 567 181 5.79 221 22 547 175 5.50 232 28 523 168 5.90 230 25 545 172 5.81 240 34 503 163 5.61 238 31 539 172 5.58 249 40 519 165 5.68 247 46 533 168 5.66 Average 14 443 239 5.88 Average 14 473 255 5.81 Notes: Cycle 2 is an early operation cycle that has been regenerated by re-use of spent regenerant containing a relatively lower concentration of sulphate. Cycle 9 is a later operation cycle that has been regenerated using spent regenerant containing a steady- state concentration of sulphate.

Example 2 Spent regenerant was continuously re-used in a regeneration-operation-regeneration chain of repeated cycles. Two nitrate selective resins, IMAC HP555 (Rohm & Haas) and A520E (Purolite), were tested for their regeneration and operation efficiencies, and for the accumulation of sulphate in spent regenerant. As shown in Figure 2, the accumulation of sulphate in the spent regeneration solutions was terminated after certain regeneration-operation-regeneration cycles. The concentration of sulphate in the spent regeneration solutions reached a steady state without discharge of spent regenerant. This means that a nitrate removal process can be performed with a succession of repeated cycles as the regeneration efficiency can be maintained under steady-state conditions. Figure 3 shows that the operation capacities of the ion-exchange columns were very stable during the regeneration-operation- regeneration cycles, also when the sulphate accumulation in spent regenerant had reached steady state. More detailed information on the regeneration parameters and operation features is given in Table 4 (for IMAC HP555) and Table 5 (for Purolite A520E). The results indicate that the quality of the effluents was not changed during the repeated process cycles.

Table 4 The results of the column tests using the nitrate selective resin IMAC HP555 Cycle Regeneration Operation (average in effluent) NaCl Level NO3- SO42- Cl- pH (%) (g NaCl/l resin) (mg/1) (mg/1) (mg/1) 1 6.4 459 16 400 144 6.88 2 8.9 482 12 415 276 6.24 3 7.4 476 13 412 282 6.56 4 9.0 486 13 383 287 6.50 5 9.0 486 9 422 253 6.49 6 9.0 486 13 427 250 6.57 7 9.0 486 13 425 245 6.48 8 9.0 486 11 432 248 6.52 9 9. 0 486 14 452 228 6.57

Table 5 The results of the column tests using the nitrate selective resin A520E Cycle Regeneration Operation (average in effluent) NaCl Level N03-SO4z-Cl-pH (%) (g NaCl/1 resin) (mg/1) (mg/1) (mg/1) 1 8.9 458 16 447 248 5.78 2 9.0 462 14 443 239 5.96 3 9.0 462 14 439 238 6.05 4 9.0 462 12 441 244 5.95 5 9.0 462 12 448 248 6.06 6 9.0 462 13 459 239 5.99 7 9.0 462 11 430 230 6.03 8 9.0 462 11 470 240 5.85 9 9.0 462 14 473 255 5.86 10 9.0 462 15 440 253 5.88 11 9. 0 462 16 432 232 6.11

Example 3 The nitrate contained in spent regeneration solution was destructed by an electrochemical reduction process before conditioning. The catholyte was composed of sodium chloride, sodium nitrate and sodium sulphate. The initial ratios of the three main components of the catholyte were varied since the concentration of the anions changes after several cycles of regeneration of the nitrate selective anion exchange resin. The sulphate concentration was varied between 32 and 134 mmol/1.

Selective nitrate reduction and water electrolysis were the two main reactions. The conditions of the electrolysis were set to reduce the nitrate selectively.

Over 99 % of the nitrate in the catholyte was reduced.

Sulphuric acid (0.8-0. 9 M) was used as anolyte and water electrolysis with oxygen evolution was the main reaction in the anodic compartment. The protons generated from the electroysis were transferred into the cathodic compartment.

The pH of the catholyte and the anolyte remained stable after several cycles of nitrate removal and reduction. Volumetric changes in the two compartments resulted in an increased level in the catholyte and a decreased in the anolyte. The anolyte level was adjusted by water addition after each cycle of the reduction.

In the anodic compartment the electrolyte was not changed. In a total of over the 150 hours of operation there was no need for acid addition to the sulphuric acid anolyte. In the cathodic compartment the same sodium chloride solution was used.

The parameters and results are shown in Table 6-9.

Table 6 Results obtained from the electrochemical reduction of nitrate in an effluent in which the initial sulphate concentration was lower than the nitrate concentration.

The conversion rate for nitrate was over 91 % and the current efficiency was 46 %. Catholyte Anolyte Time Current Tension Volume S042-N03-N03-Volume H+ (h) (A) (V) (1) (mmol/1) (mmol/1) (mg/1) (1) (mol/1) 0 20 4.40 2.00 32.03 86.44 5360 2.00 0.80 1.0 20 4.35 2.00 31.89 68.11 4223 1.95 2.0 20 4.40 2.10 31.84 31.76 1969 1.85 3.0 20 4.50 2.10 31.41 7.70 478 1.80 4.0 20 4. 60 2.20 30. 66 1. 43 88 1.75 0.91

Table 7 Results obtained from the electrochemical reduction of nitrate in an effluent in which the initial sulphate and nitrate concentrations were in the same range. The conversion rate for nitrate was over 92 % and the current efficiency was 84 %.

Catholyte Anolyte Time Current Tension Volume 5042-NO3-N03 Volume H' (h) (A) (V) (1) (mmol/1) (mmol/1) (mg/1) (1) (mol/1) 0.0 20 4.20 2.20 50.2 53.1 3296 2.00 0.80 1.0 20 4.00 2.20 49.1 5.2 310 1.95 2.0 20 4.00 2.20 49.2 0.2 10 1.85 2.5 4.00 2.20 48.8 0.1 7 3.0 20 4.10 2.25 47.9 0.3 19 1.85 3.5 4.20 2.25 48.1 0.0 0 1.80 4.0 20 4. 20 2.30 48. 0 0. 0 0 1.78 0.90

Table 8 Results obtained from the electrochemical reduction of nitrate in an effluent in which the initial sulphate concentration was higher than the nitrate concentration.

The conversion rate for nitrate was over 80 % and the current efficiency was over 61 %. Catholyte Anolyte Time Current Tension Volume S042-N03-N03-Volume H+ (h) (A) (V) (1) (mmol/1) (mmol/1) (mg/1) (1) (mol/1) 0.0 20 4.34 2.20 80.5 42.4 2627 2.00 0.79 1.0 20 4.40 2.20 76.1 9.1 567 1.95 2.0 20 4.50 2.30 75.4 0.5 32 1.85 0.83 0.0 20 4.40 2.20 96.5 42.0 2601 1.95 0.80 1.0 20 4.40 2.20 94.0 11.2 696 1.90 2.0 20 4. 60 2.30 93. 5 0. 9 53 1.85 0.84

Table 9 Results obtained from the electrochemical reduction of nitrate in an effluent in which the initial sulphate concentration was close to the highest concentration that could be obtained with the nitrate selective ion exchange resin after several cycles of regeneration. The conversion rate for nitrate was close to 90% and the current efficiency was over 52 %. Catholyte Anolyte Time Current Tension Volume SO42-N03-N03-Volume H+ (h) (A) (V) (1) (mmol/1) (mmol/1) (mg/1) (1) (mol/1) 0.0 20 4.30 2.00 134.4 31.0 1924 2. 00 0.82 1.0 20 2.00 129.5 3.3 206 1.95 2.0 20 2.10 129.4 0.1 8 1.85 3.0 20 2.10 129.3 0.1 6 1.80 4.0 20 4. 57 2.20 129. 0 0. 0 <5 1.75 0.88

As shown by the results, the nitrate reduction was successfully carried out in an electrolyte containing sulphate and chloride in high concentrations. The current efficiency of the nitrate reduction depends on the initial nitrate concentration and on the conversion rate.

The reduction process is more efficient in solutions which are more concentrated on nitrate. The selectivity of the nitrate reduction was not effected by the increasing sulphate concentration. Increasing initial nitrate/sulphate ratio resulted in an increase in the nitrate reduction efficiency as indicated in Table 10.

Table 10 The current efficiency of the nitrate reduction depends on the ratio of nitrate to sulphate Ratio of N03-/S042-Current efficiency (%) > 1-46 < 1 > 52 Example 4 As shown in Table 11, the ratio of chloride to sulphate in the regeneration solution played an important role for the regeneration efficiency. A high ratio of chloride to sulphate should be kept in a spent regeneration solution in order to maintain a good regeneration efficiency and to decrease residual nitrate in effluent in the following operation cycles.

Table 11 Comparison of the regeneration efficiencies for IMAC HP555 with different ratios of chloride to sulphate in the regenerant Regeneration Operation NaCl S04z-Ratio Capacity N03-SOq2-Cl-pH in effluent (meq/l) (meq/l) (Cl-/SO42-) (eq/l resin) (mg/1) (mg/1) (mg/1) 769 770 1.00 0.267 20 396 264 6.37 1149 390 2.95 0.304 18 438 247 6.07 1538 390 3.95 0.319 15 417 244 6.21 1538 245 6. 29 0.332 12 442 238 6.55 The choice of a suitable end point for controlling the ion-exchange operation can improve the separation of nitrate from sulphate, by which the adsorption of sulphate should be displaced as much as possible in an operation cycle. As shown in Table 12, the release of sulphate and the operation capacities were obviously increased as the end point (EP) of operation was changed from 20 mg N03-/l to 40 mg NO3-/l for the residual nitrate in effluent. Meanwhile, there were little changes in the average nitrate concentrations in effluent. Because relatively larger amounts of sulphate were released into the treated water near the end of the operation cycle, choosing a relatively higher end point as controlling limit for residual nitrate could reduce the steady-state concentration of sulphate in spent regenerant.

Table 12 The inflence of the end point (EP) of operation on mass balance (IMAC HP555) Cycle Capacity Average N03-in Average so42-in (eq/L resin) effluent (mg/1) effluent (mg/1) EP=20 ppm EP=40 ppm EP=20 ppm EP=40 ppm EP=20 ppm EP=40 ppm 1 0.289 0.331 12 16 369 400 2 0.271 0.300 9 12 396 415 3 0.297 0.332 10 13 394 412 4 0.260 0.278 10 13 368 383 5 0.331 0.353 6 9 410 422 6 0.282 0.323 8 13 398 427 7 0.305 0.336 9 13 406 425 8 0.305 0.337 9 11 418 432 9 0. 296 0. 327 10 14 435 452