The DI Process
When minerals dissolve in water, they dissociate into positively charged cations and negatively charged anions. The demineralization process exchanges these ions into hydrogen (H+) and OH– producing "pure" water in the effluent. Equation 1 uses sodium chloride (NaCl) as an example.
These reactions are most complete when the pH of the effluent from either resin is nearer the neutral mark. Since the cation reaction produces an acidic effluent and the anion is more of a neutralization reaction, the anion reaction is the more complete.
A logarithmic measurement, pH measures the acidity and alkalinity of a solution. For example, a pH of 3 is 10 times more acidic than a pH of 4; a pH of 1 is 100 times more acidic than a pH of 3 (pH 7 is neutral).
The effluent of a two-bed DI system is generally alkaline or high pH (8.5 to 10). This is because the low pH of the cation reaction (pH 2.5 to 3.0) causes some of the residual sodium left on the resin from the previous regeneration to "leak" off during the run. Sodium is the least tightly held of the cations. Depending on the level of regeneration and water composition, this can amount to 1 percent to 2 percent of the total cations leaking as sodium (Na+). In the anion bed, this sodium is converted to sodium hydroxide (NaOH), which causes the high pH. One ppm of Na+ in the effluent will give a pH of about 9.5 and 4 ppm will result in a pH of around 10.0. This pH will also give rise to an increase of silica in the final effluent (leakage from the anion). It is, therefore, apparent that improving the leakage performance of the cation resin is key to overall good product quality from the system.
Factors Affecting Cation and Anion Leakage
As influent water makes its way down through the cation vessel, more and more cations are exchanged for H+, thus reducing the pH. The higher the influent total dissolved solids (TDS), the lower the pH and the higher the leakage of sodium. Higher percentages of influent sodium also will increase the sodium leakage because more sodium will be left on the cation exchanger after regeneration. This is why it is not advised to pre-soften feed water going to a DI system. Since the "strength" of the acid produced by the cation exchanger affects leakage, waters that are high in the strong acids such as chlorides and sulfates will cause a higher Na+ leakage. Conversely, waters that are high in bicarbonate or alkalinity will give more favorable cation leakage characteristics.
One way to reduce overall cation leakage is to increase the acid dosage during regeneration. In fact, the regenerant level is about the only controllable variable. However, there is a rapid drop off in the economics for increasing the acid dose compared to the increase in capacity and reduction of leakage. Typical acid levels are approximately 8 pounds per cubic feet of resin.
In anion resin, the circumstances are different. Anion resins love strong acids such as chlorides and sulfates but have a tougher time removing the weak ionized acids such as bicarbonates and silica. This especially is true if there is a considerable Na+ leakage from the cation, which converts to NaOH in the anion exchanger. Efficient operation of the anion exchanger is not as concerned with the overall TDS level because the removal of the acidic anions results in a neutral pH (see Reaction 2). However, the anion is affected adversely by high levels of weak acids (which we already have deemed beneficial to the operation of the cation.) Again, increasing the regenerant level will improve the operation of an anion but the gains are limited by economics. Typical caustic levels are approximately 8 to 9 pounds per cubic feet of anion. Typical two-bed effluent with 400 to 500 ppm feed water would be about 250 to 400K ohms (2 to 4 µmhos), a pH of 9.5 and silica leakage of 20 ppb.
Counter-Current vs. Co-Current Regeneration
Another option that can be considered to lower leakages is the use of counter-current regeneration as opposed to co-current regeneration. This technique greatly improves the efficiency of an ion exchange bed as the least regenerated resins remain at the top of the bed at the end of the regeneration cycle. When the next service cycle begins, the highly regenerated resin at the bottom of the bed serves to ensure good effluent quality at the start of the run. In a co-current system some resin also is left unregenerated or partially exhausted. However, this resin is situated at the bottom of the bed and upon start-up of the service cycle a higher leakage will be seen. Typically, counter-current systems can produce water with 80 percent to 90 percent lower leakage than their co-current counterparts. An added benefit of counter-current systems is that operating capacity increases by about 5 percent to 10 percent over co-currently regenerated resins because the run can be extended due to the lower average leakages.1
For proper counter-current regeneration the resin bed must be held in place with a mechanical screen or by completely filling the bed with resin (packed bed). Other methods include the use of blocking flows of air or water that involve multiple distributors. The regenerant flow and blocking flow are in opposite directions, holding the bed in place. Although some of the benefits are retained, efficiency drops off appreciably if the bed is allowed to "float" during up-flow regeneration (typical of standard household automatic valves). This also precludes realizing any real benefit from counter-current regeneration in the portable exchange industry.
The Geometric Approach
Several years ago, researchers looked at methods of improving upon the typical leakage performance of standard DI resins, particularly the cation exchanger. The quest was to improve upon the ability to regenerate the cation resin with standard chemical levels and standard equipment. The basic thesis was to improve the ease with which the regenerant would penetrate the resin matrix. Smaller resins (fine mesh) and uniform particle sized (UPS) resins showed promise but both still left moderate quantities of contaminant behind after regeneration, and there was a limit to just how small a resin could be made and still be contained within the existing equipment. The quest finally was answered when they discovered a method of making a cation resin with only a thin shell of surface functionality and, therefore, no need to achieve a deep regeneration.Nearly 30 percent of a resin’s reactive sites lie just below the surface in the first 10 percent of its outer radius. Nearly 60 percent of its volume is only 25 percent of its shell and 95 percent is contained at approximately 60 percent of the radius depth. (See Table 1.)
Shallow Shell Technology
Over the past 25 years, a dozen or more scientific papers have promoted the idea that if a resin had a shorter diffusion path or shallow shell, it could be more readily regenerated. This, in turn, would yield higher regenerant efficiency, produce lower leakages and conserve water. Partial functionality was the approach. Early versions of these resins somewhat resembled a boiled egg, only round. It soon was discovered that the shallow shell also was a very uniform shell. Not all beads were fuctionalized to the same percentage depth, but they were functioned to the same uniform depth and that depth was controllable. A resin now could be manufactured that would duplicate the positive attributes of both uniform beads and fine mesh beads at the same time without cost or pressure drop limitations.
Recent improvements in this basic technique have led to the development of a family of shallow shell resins that have been confirmed through field trials to have capacities that are 15 percent to 25 percent higher than conventional resins (Table 2) while providing 50 percent to 70 percent lower leakage. In short, shallow shell resins are able to demonstrate counter-current leakage properties in a conventional co-current regenerated system. (See Table 3.) Commercial field installations have confirmed that reductions of 15 percent to 25 percent in regenerant levels are possible while maintaining normal capacity and reducing leakages.
The explanation for the superior performance of the shallow shell resin lies in its ability to be more completely regenerated than its conventional cousins at any given regenerant level. A laboratory study was carried out on a softening version of this resin to examine a conventional resin and a shallow shell resin for residual hardness immediately after regeneration and immediately at breakthrough exhaustion (leakage of 17.1 ppm as CaCO3). The results are shown in Table 2.
Benefits Translate to Hydrogen (DI) Cycle
A two-year study of a 100 cu. ft. industrial DI installation utilizing the hydrogen form of the shallow shell resin vs. a side-by-side conventional DI cation resin showed similar results compared to the softening studies. Reduced sodium leakage from the cation vessel translated to a lower TDS effluent from the anion vessel with a more neutral pH and longer runs due to the higher capacity, lower leakage and reduced rinse requirements. Examples of lower leakage provided by shallow shell resin vs. a conventional resin, using either hydrochloric or sulfuric acid, are summarized in Table 3.
This improvement in sodium leakage translates directly to higher quality and capacity with lower silica slippage and lower rinse requirements. This is a simple solution to improving the quality of two-bed and PE deionizers.
Conclusions
The easier a resin is to regenerate, the more efficiently it will utilize the regenerant chemical and the more complete the regeneration will be at any chemical dosage. Shallow shell resins facilitate that ease and provide lower rinse requirements and lower the leakage in service.
The shallow shell technology concept represents the first truly significant improvement in ion exchange resin design in more than 50 years.
Editor’s Note: The resin described is a patent-pending product developed by the Purolite Co. and was introduced in 1997. It is referred to commercially as Purolite SST-60 (representing a 60 percent shell ratio). The above data were collected by a combination of bench study, pilot studies, field study and commercial scale installations.About the Authors
Jim Sabzali is the North American sales and marketing manager of The Purolite Co., Bala Cynwyd, Pa. Sabzali has 20 years of experience with synthetic resins and has been with Purolite since 1990. He can be reached at 800-343-1500; fax 610-668-3962; e-mail [email protected].
C.F. "Chubb" Michaud, CWS-VI, is technical director of the Systematix Co. of Placentia, Calif., which he founded in 1982. With more than 30 years of experience in water and fluid treatment, he is a member of the Water Quality Association's Ion Exchange Task Force and chairman of its subcommittee for component providers in the manufacturer/supplier section. Michaud can be reached at 714-993-2482; fax 714-528-2482; e-mail [email protected].
References
- Helfferich F., Ion Exchange (McGraw-Hill, New York, 1958).
- Kunin R., Ion Exchange Resins 2nd ed. (Wiley & Sons Publishers, New York, 1958).
- Irving J., "The Optimization of the Separation of Component Ion Exchange Resins in Mixed Bed Exchangers," Technical Paper given at the International Water Conference, Pittsburgh, Pa., October 1991.
Equation 1
(1) Cation Reaction NaCl + R*H Æ RNa+ + HCl
(2) Anion Reaction HCl + R*OH Æ RCl– + H+OH– (H2O)
where R* represents the resin
