Sodium chloride is generally preferred for regeneration for reasons of cost and efficiency. When available sea water can be used quite effectively. The use of softened water for make up of regenerant and rinse is often recommended to avoid the precipitation of calcium carbonate in and around the CQ-N250 (or any other resin used in this application). Although the precipitation is not particularly detrimental in the short term, the long term effects may include increased resin attrition and leakage of nitrates.

CQ-N250 is processed to insure that it meets the requirements for use in the treatment of potable water. On installation it is recommended that the resin be regenerated with two bed volumes of 6% NaCl followed by a rinse of four bed volumes of potable water, prior to use.

The pressure drop or headloss across a properly classified bed of ion exchange resin depends on the particle size distribution, bed depth, and void volume of the exchange material as well as on the viscosity (and hence on the temperature) of the influent solution. Factors affecting any of these parameters, for example the presence of particulate matter filtered out by the bed, abnormal compressability of the resin, or the incomplete classification of the bed will have an adverse effect and result in an increased headloss. Depending on the quality of the influent water, the application and the design of the plant, service flow rates may vary from 10 - 40 bed volumes/hour (1 - 5 gpm/ft3). Typical pressure drop data is given in Fig. 1.

During upflow backwash, the resin bed should be expanded in volume by between 50 and 70%. This operation will free it from any particulate matter, clear the bed of bubbles and voids, and reclassify the resin particles, ensuring minimum resistance to flow. Bed expansion increases with flow rate and decreases with temperature, as shown in Fig. 2. Care should be taken to avoid over expansion of the bed.

The high selectivity of   CQ-N250  for nitrate over sulphate ensures that any necessary reduction in nitrate levels can be achieved even in the presence of high influent sulphate concentration. Hence it so offers the advantage over standard strong base resins that its exchange capacity for nitrates is less affected by a high influent concentration of sulphate. For this reason, although CQ-N250  has a lower total exchange capacity than standard strong base anion resin, its use can produce advantageously higher throughputs for the following reasons.

Both standard gel type or macroporous strong base resins are quite capable of effective nitrate removal where sulphate to total anion ratios are low. However, on account of the high selectivity for sulphate in dilute solutions which follows the order,

HCO3- < Cl- < NO3- < SO4=

selective displacement of nitrate by sulphate results in the effective nitrate removal capacity being reduced by sulphate loading. Apart from the obvious disadvantage of the reduction of treated water obtained on cycling, the exchange of both nitrate and sulphate by chloride will result in a less palatable and sometimes less acceptable water than the influent supply, in that the treated water may be more corrosive and the limits for chloride concentration may be exceeded.

CO-CURRENT REGENERATION

Fig. 3 and Fig. 4 give the operating capacity and nitrate leakage respectively which may be obtained using cocurrent regeneration at the given regeneration levels. Values obtained from Fig. 3 are expressed in terms of nitrate throughput, corrected for nitrate leakage, and hence may not be used directly to determine the throughput of water. All ion concentration values are either on a ppm or a meq/l basis for ratio determination.

Similarly Fig. 5 and Fig. 6 give the values for countercurrent regeneration. It should be noted that in this case the nitrate leakage is lower for a given regeneration level. Hence the possibility to blend treated with untreated water on a 50% basis is a useful option which can make counter-current regeneration attractive. On the other hand the choice of co-current regeneration can result in the production of higher volumes of treated water of satisfactory quality for direct use. The higher leakage (ln, lnp, in the equations above) so reduces the load on the ion exchange bed that for a given operating capacity greater throughputs per cycle are obtained. This latter effect can influence the throughput more than differences in basic operating capacity. It therefore follows that both capacity and leakage for alternative modes of regeneration should be evaluated before recommending specific design conditions.

CO-CURRENT REGENERATION

Presupposing that the objective of the nitrate removal treatment is to obtain potable water of a quality which meets the World Health Organization (WHO) limit, where the nitrate/(nitrate + sulphate) ratio is higher than 0.6, a nitrate selective resin is not necessary. A standard strong base resin can give higher throughputs as a result of its higher total capacity. It will be seen that up to the ratio of 0.6 the curves in Figs. 3 - 6 are continuous to show where CQ-N250 is the recommended resin. The discontinuous curves are given so that comparisons may be made with alternative resins. Where lower leakages than the WHO limit are required, for example in the processing of certain foods, CQ-N250  will often give a superior performance to the standard resins even where nitrate/(nitrate + sulphate) ratios are higher than 0.6. One particular advantage here is that there is no slug of highly concentrated nitrate at breakthrough as is found with standard resins, hence the possibility to excessively contaminate the food product by overrunning the bed is avoided.

Depending upon the throughput requirement the resin volume is chosen so as to operate within the flow rate stipulations given in the standard operating conditions above. A design factor of 0.9 is also recommended as is customary. Hence throughput/liter of resin for design purposes will be 313 x 0.9 = 281.7 liters (2124 U.S. gal/ft3). In this example the leakage is 17.3 ppm as CaCO3 (21.4ppm as NO3), hence the useful option to blend treated water with raw water on a 50% basis could be applied. It would be of no advantage to move to counter-current regeneration in this case. Reference to Fig. 5 will show that the basic capacity curve is very similar. However the throughput will be lower, because the reduced leakage increases the ion exchange load for a given throughput. When on the other hand nitrate concentrations or ratios are higher, it may be advantageous to operate counter-current rather than increase the regeneration level while operating co-current. In this way a suitable blend may be obtained with lower regenerant costs (and costs of disposal).