Vol.25, No.4 (2014)

(2)

Article

Optimizing Cation to Anion Resin Ratio in Mixed-Bed Ion ExchangeM. M. RAHMAN1, Taekyung YOON2,Gary L. FOUTCH3*

1School of Chemical Engineering, Oklahoma State University, USA 740782Department of Environmental Engineering, Dong-eui University, Busan, Korea 614-714

3Department of Civil and Mechanical Engineering, University of Missouri Kansas City, USA 64110

(Manuscript received June 27, 2014; Accepted August 8, 2014)

Abstract

A mixed bed ion exchange (MBIE) column producing ultrapure water will be removed from service well before its capacity is exhausted. There is usually an operational criterion for replacing or regenerating the bed; for example, defining 50 parts per trillion (ppt) Na as an endpoint in Pressurized Water Reactors (PWRs). These plant-specific criteria depend on the water treatment objectives and include low day-to-day concentrations or the ability to handle an upset event such as a concentration spike. In mixing the bed initially the cation to anion resin ratio is frequently set to where the equivalence of cation capacity equals the equivalence of anion capacity. An alternative approach is to define the resin ratio based on the feedwater pH so that the initial breakthrough time of cations and anions are the same. This ratio calculation procedure using feedwater analysis and a rate limited ion exchange model is described and the impact of this approach is discussed.Keywords: MBIE, BWR, PWR, MBIE-SIM

1. Introduction

1.1 BackgroundThe concept of an “equivalent bed” to define the mixture of

anionic to cationic ion exchange resin used in ultrapure water generation is worthy of additional discussion. The purpose for mixing ion exchange resin types is to allow the neutralization reaction of the released hydrogen and hydroxide ions to form water. The overall result of a mixed-bed ion exchange column is to produce the purest water possible. By setting the resin ratio to an equivalent bed, the equivalence capacity of the cations and anions exchanged would be equal; an equivalent bed should breakthrough at the same time for both cations and anions if the water entering the bed is neutral. However, this is a reasonable assumption for only a limited number of applications; for example, polishing bottles used in microelectronics rinse water. Goldman (2002)1 discusses the composition and use of mixed-bed ion exchange hydrogel-forming polymers as absorbents. He states that deviating toward either enriched cationic or anionic ratios may be needed to compensate for differences in pK, neutralization, or required effluent pH.

* Corresponding authorE-mail: gary.l.foutch@okstate.edu

For the power industry, there are two situations where an equivalent bed should be reconsidered: where chemical additives are used to adjust the pH of water traveling through the steam cycle and to plan for a condenser tube leak event that would result in a spike of river, lake or ocean water into the system. There may be additional considerations for specificpower plant operating cases as well; such as, regulatory requirements, potential for radioactive exposure, or seasonal cooling water variability.

For cases where plant operating criteria specify a particulareffluent concentration for resin regeneration or bed removal, and that ion exchange resin saturation is not desired, then an equivalent bed cannot maximize the time until the terminationcriteria are reached. The reason is due primarily to the large difference in cation and anion exchange site selectivities. Anionic selectivities in ratio to that of hydroxide are typically 20 and above, depending on the anion. Cationic selectivities to hydrogen are 1.5 and slightly above depending on the cation. This gives a very sharp exchange front for anionic breakthrough and a more graduate pattern for cations. The times to achieve specific cationic and anionic breakthrough criteria approach each other at full saturation of the respective resins. However,

191

J. ION EXCHANGE

(xxx)

for power plant operation, condensate polishers are never intended to operate near the saturation point.

Given the feedwater analysis, a multicomponent ion exchange process can be modeled to generate individual ionic breakthrough curves. Plant specific operating criteria can then be used to define bed breakthrough actionable events.Examples of bed termination conditions include: achieving a specific ionic effluent concentration, loading the beds to a specified percentage of the original capacity, exceeding a bed pressure drop, approaching the transition from hydrogen cycle to amine cycle operation, or because regeneration occurs on a defined time schedule.An example of PWR (pressurized water reactor) bed-

termination is to regenerate when the effluent concentrationreaches 50 parts-per-trillion (ppt) sodium. The sodium concentration (mass fraction) breakthrough profile is dictated by a combination of the regeneration efficiency (equilibrium leakage) and sodium-throw wave that precedes the amine concentration down the bed. However, a key feature, and the basis for this discussion is that the anions show no signs of breaking yet. By enriching the cation resin fraction the time until 50 ppt effluent is increased. At some point, as the cation resin fraction increases an anion – typically chloride or sulfate –will increase in effluent concentration at an earlier time due toreducing anionic capacity. By performing a series of simulations over a broad range of cation to anion resin ratios acrossover from the initial cation to the initial anion will occur at a specific concentration and the corresponding time is defined. The lower the breakthrough concentration criterion, as defined

by the plant operator, the greater the impact of this ratio adjustment. If the breakthrough concentration criterion is higher,there is less gained by adjusting the cation to anion resin ratio because both resins will have a higher loading fraction and display near saturation equilibrium leakage.When a bed is operated to ensure against a concentration spike

from a condenser tube leak the strategy must be adjusted. In this case the cation-to-anion resin ratio should be proportional to the cooling water concentration ratio - the leaking water composition. For a small leak the reduction in anticipated run time for the bed can be estimated and included in operational strategy. Some plants operate with small leaks through high demand periods. For a large condenser tube leak the bed will flood with ions and saturate quickly. For a guillotine tube rupture (complete tube failure) saturation may occur within just a few minutes. Having an adjusted resin ratio may buy you some, but very little, additional time. For those cases the resin ratio will not make much of a difference either way.

For a BWR (boiling water reactor) case there is no amine

added to the water for pH control. Condensate polisher run times may be years. The criterion for taking a bed out of service may be the residual ion exchange capacity; for example, 50% of the original exchange sites in the hydrogen and hydroxide form.Sulfate is a common concern from leaching sulfonates originating from the cationic resin functionalities, but the bed can be tiered with an enriched anionic layer at the bottom of the bed to collect any sulfonates or sulfate generated passing through or originating from the cationic resin. Most BWRs use anionic heels for such a purpose.

1.2 Literature ReviewMixed bed ion exchange is a single column containing both

cationic and anionic resin and can achieve the highest deionization efficiency2. Both cationic and anionic exchange occurs simultaneously on adjacent resins beads. The H+ and OH-

1.2.1 MBIE-SIM Background

ions from the exchangers diffuse out of the beads and into the water where they react with one another to form water by the pH equilibrium reaction. This neutralization reaction reduces the number of ions in the water and increases the flux rates of ions into the resin beads; thus more efficiently generating an ultrapure product. This literature review consists of two major parts. The first part provides a brief history on the development of the current MBIE simulator (MBIE-SIM)program; which is used to perform all the simulations for this article. The second part of the literature review focuses on plant strategies for optimizing operating conditions of condensate polishers in PWRs and BWRs.

Kunin and MacGarvey (1951)3

The first theoretical study on MBIE was performed by Caddell and Moison (1954)

were the first to use a mixed bed ion exchange column. “Monobed” deionization eliminates the requirement of conventional double-bed (separated cation exchange resin bed and anion resin exchange bed) deionization technique. They conducted a thorough study from a technical and economic point of view and showed how monobed deionization can overcome many difficulties of the, then, conventional double-bed deionization technique.

4. They inspected the variables that influence the breakthrough of mixed-bed, and developed an empirical relationship between leakage and capacity. Kunin (1960)5

Many researchers have performed numerous studies focused on liquid-side mass transfer in binary ion exchange systems; Copeland et al. (1967)

further conducted an experimental study on the kinetics of mixed-bed deionization. He examined the effects of influent concentration, flow rate, bed depth, and temperature on bed performance and concluded that the ion exchange rate was controlled by a liquid-film mass transfer mechanism at low concentrations. This fact is one of the most important assumptions of MBIE modeling.

6; Copeland and Marchellow (1969)7;Glaski and Dranoff (1963)8; Kataoka et al. (1968, 1972)9, 10;

192

Vol.25, No.4 (2014)

(3)

Schlögl and Helfferich (1957)11; Smith and Dranoff (1964)12;Turner and Snowdon (1968)13; Franzreb et al. (1993)14;Chowdiah et al. (1996)15. Nernst-Planck (N-P) equation to describe the fluxes of ionic species were first applied by Schlögl and Helfferich (1957)11

Kataoka et al. (1987)

. They showed that the electric field, caused by the difference of their respective diffusivities, has a significant effect on the ion exchange rate of a binary system.

16

Haub and Foutch (1984)

studied the film-diffusion controlled liquid-side mass transfer in a ternary system. Flux expressions for the ions with equal valences and ions with different valences were developed separately. The numerical solution of Kataoka’s model matched reasonably with their experimental results. However, the Kataoka model was limited to systems with three species only.

17, 18 were the first to model rate-limited mixed-bed ion exchange at ultrapure water concentrations. Since concentrations of interest are frequently lower than those of dissociated H+ and OH- from water, they considered the dissociation of water equation to develop a model for hydrogen cycle MBIE. The major advantage of their model was the consideration of separate material balances for each resin. The model was limited to binary monovalent systems such as Na+ - Cl-

Divekar et al. (1987)at 25

19 extended Haub and Foutch’s model to incorporate temperature effects. They studied the correlations of physical properties such as diffusivity, viscosity, and dissociation constant as a function of temperature for certain species and implemented them into the model. Zecchini et al. (1990, 1991)20, 21 extended the Divekar et al. model to simulate a ternary system of monovalent ions with amines. Pondugula et al. (1994)22

Other contributors of the current MBIE-SIM model are Bulusu (1994)

extended the model to divalent ternary systems. Also, the effect of desulfonation of the cation resin on column performance was included in the current Mixed Bed Ion Exchange Simulator (MBIE-SIM) model.

23; Sunkavalli (1996)24; Hussey (1996)25, 26; Liu (1999)27;Yi (2004)28. Of particular note is the work of Franzreb et al. (1993)14 who developed a method of handling an arbitrary number of ionic species. Yi was the first to include radioactive isotope species in the MBIE-SIM model. To summarize the above work, a detailed description of the multicomponent MBIE model is presented by Yi and Foutch (2004)29

1.2.2 PWR and BWR polisher operating strategies

.

Plant control programs have specific goals for operating conditions. Minimization of ionic and particulate impurity transport to the steam generators in PWR systems and to the reactor circulation cycle in BWR systems are the major goals. In particular, rates of corrosion accelerates in partially occluded regions of PWR steam generators due to formation of acidic or basic solutions, reduces plant availability and in some casescauses steam generator replacement. In BWRs, ingress of ionic

impurities, sulfate and chloride, is known to accelerate stress corrosion cracking of reactor system materials31

The performance requirements vary markedly depending on the criteria used and the type of plant. For example, polisher effluent chloride requirements vary from 6 to 50 ppt at BWR units to 1.25 to 5 ppb (several hundred times greater) at PWR once-through steam generator units

. Since of leakage of cooling water is a major source of ionic impurity ingress in both PWRs and BWRs, techniques for removing such impurities from the condensate are of significant interest.

31

A variety of possible techniques are available to increase hydrogen cycle operating time: 1) increase of fraction of cation in mixed bed, 2) increase bed depth, 3) install a cation bed upstream of the mixed bed

. Thus approaches to polisher design and operation vary markedly.

30

In PWR systems, the operating time to amine breakthrough decreases with increasing pH and increases with cation resin operating capacity and cation to anion resin ratio. In this case, a 2 to 1 cation to anion volume ratio (equivalent bed) is most often recommended as a reasonable compromise between run length and anion exchange efficiency

. This article focuses on optimizing existing systems based on feedwater analysis, thereby options 2and 3 are not considered.

31

2. Experimental

. In BWR systems, only condensate impurity ions need to be removed since no pH control additives are used, i.e., the condensate is at near neutral pH. In this case, an equivalent resin ratio is usually recommended for ionic removal, but other resin ratios may be selected to improve corrosion product removal. This article addresses alternative methods to optimize operating criteria of condensate polishers based on feedwater analysis and plant specifications.

The modeling program used to simulate performance of an ion exchange bed was MBIE-SIM described by Yi and Foutch (2004)29. The model assumes ionic mass-transfer through a film surrounding each bead, flux rates linked through Nernst-Planck relationships and bulk phase neutralization. Bed operating conditions were typical of BWRs and PWRs, some with condenser tube leaks. For these example cases a mixed bed operating at 200 gal/min with a bed depth of 4.0 ft and diameter of 3.66 ft was simulated assuming a condenser tube leak bed inlet concentration resulting in 1 PPM (parts per million) of Na and an equivalent amount (1.458 PPM) of Cl. This is a relatively small bed operating at a low flow rate. To extend this work to other plants or applications then details must be input accurately into the model. Numerous breakthrough curves were generated over a range of cation to anion resin ratios keeping all other input variables constant. Breakthrough curves for Na and Cl (for the BWR conditions), and NH3, Na and Cl (for the PWR conditions) were generated. For discussion of the significance of cation to anion resin ratio, the resin amounts were adjusted

193

J. ION EXCHANGE

(xxx)

until the times to a specific effluent concentration were achieved. Crossover was defined as the point where breakpoints for both the leading cation and anion matched.

Table 1 Input

Parameter Value

data for BWR

Bed Diameter (feet)Bed Depth (feet)Flowrate (gpm)Void FractionTemperature

3.6642000.3525

Cationic Resin Dowex Monosphere 650C

Bead Diameter (mm)Capacity (mequiv. /ml.)Form

0.6251.9Hydrogen

Anionic Resin Dowex Monosphere 550A

Bead Diameter (mm)Capacity (mequiv. /ml.)Form

0.591.1Hydroxide

Feed Water Chemistry

Ion Feed Concentration (ppb)

NaCl

+ 1000- 1458

Table 2 Input

Parameter Value

data for PWR

Bed Diameter (feet)Bed Depth (feet)Flowrate (gpm)Void FractionTemperature

103.5 31700.3535

Cationic Resin Dowex Monosphere 650C

Bead Diameter (mm)Capacity (mequiv. /ml.)Form

0.6251.9Hydrogen

Anionic Resin Dowex Monosphere 550A

Bead Diameter (mm)Capacity (mequiv. /ml.)Form

0.591.1Hydroxide

Feed Water Chemistry

Ion Feed Concentration (ppb)

NaCl

+

NH

-

4

1000

+

10002000

3. Results and DiscussionThere are numerous criteria for terminating the service cycle

of an ion exchange bed and these must be plant specific. A first consideration is to define the primary purpose for condensate polishing: whether the beds are in use for general water cleanup

or operate for the potential event of a condenser tube leak. Other factors that will not be discussed here include whether the beds have a role in particle adsorption or sulfate leaching. The concentration targets that we will discuss are arbitrary and simplistic compared to full-scale, plant-specific criteria.

An example of the crossover phenomenon can be seen in the laboratory experimental data presented in Fig. 1 with NaCl solution in a mixed bed. These plots are representative of a large amount of data that were generated by Yoon (1990)32 and present initial to final concentration ratios of Na and Cl as a function of the total water processed. The total amount of resin was fixed and only the cationic to anionic resin ratio was changed. For conditions where the flowrate is constant, the x-axis could be expressed as operating time.

Fig. 1 Experimental data showing crossover points fordifferent percentage of cation resin32

As can be seen in Fig. 1. there is a slight decrease in the crossover for the amount of water that can be processed in moving from 67% to 40% cationic resin. The same change in resin percentage yields a significant decrease in crossover location. These experimental observations are reinforced with the following simulation results.

3.1 Boiling Water Reactor In BWRs steam is generated by direct contact with the nuclear

reactor. As a result, water is maintained at near neutral conditions to minimize the risk of precipitation or alkylation reactions on materials in the core that might affect heat transfer. For a BWR condensate polisher case as shown in Fig. 2 initialconcentrations remain generally flat due to equilibrium leakage from the bed. This equilibrium value is based on the amount of the respective ion on the resin prior to the start of the service cycle from either regeneration efficiency, if the resins were previously operated through a bed cycle, or from residual ions on the beds as a result of manufacturing or initial rinsing if they are new. In Fig. 2 Cl shows a significant equilibrium base with the initial ionic loading fraction on the anionic resin sites of 0.0001. For an example case, Na effluent reaches 50 ppt at 3.0

194

Vol.25, No.4 (2014)

(5)

days with Cl at 9.0 days for an equivalent bed (37% cationic by volume). If the cation fraction is increased to 40%, the 50 ppt Na effluent is 4.0 days while Cl is 8.1 days. This shows nearly a full day of extra operating time before a 50 ppt action level is reached. By increasing the percentage cation further, eventually Na and Cl will reach 50 ppt at the same time yielding the optimum ratio to give the longest time of operation.

Fig. 2 BWR effluent breakthrough curve at two different cation resin fractions (37 and 40%)

Note however, that the impact on full saturation is hardly significant with all curves approaching 100 ppb in about 11 days. If the primary operating objective is routine condensate polishing, or for even small condenser tube leaks then this resin ratio change is worth considering as long as no other plant operating criteria are violated. For the case of a major tube leak there is little benefit, but also likely little harm, because breakthrough will differ by very little as the bed saturates quickly.

From the slopes of the Na and Cl curves we can see that for lower target concentrations the cation to anion ratio can be increased, but for higher concentrations the ratio must be decreased. In practice, the resin ratio can be optimized only after specific plant requirements are identified and actual feedwater chemistry analysis is known.

Increasing the cationic fraction further will show a longer time until 50 ppt Na effluent occurs, but will result in earlier breakthrough of Cl. At some ratio there is a crossover point at which the Na and Cl concentrations are equal. Plots of this computation are presented in Fig. 3 and 4. These plots continue lower until the equilibrium leakage from the bed occurs and the effluent concentrations level off. As time decreases there is no point where Cl concentration exceeds Na concentration, so these curves do not have a minimum through the initial bed operation.

Fig. 3 BWR crossover time related to the percentage of cation resin. Crossover is defined as when the effluent of the first cation and first anion match.

Figures 3 and 4 can be used to define how this particular bed can operate at different resin ratios. First, the curves are generated for specific plant operating conditions, feed water concentrations; and, in the case of a tube leak, the anticipated leak rate and leak concentrations. For this example the conditions are presented in Table 1. Then the plant specific effluent concentration is defined. For 50 ppt Na breakthrough the crossover point will be 47% cationic resin fraction (from Fig.4) and the longest time achievable will be 6 days (from Fig. 3). If the breakthrough goal were 1 ppb then 44% cationic resin would allow operation for 8 days. Using the crossover point for these calculations also means that the Cl concentration would be

Fig. 4 BWR crossover concentration versus the percentage of cation resin. Once the required end-point concentration is defined by plant operator the percentage can be used in Fig 3 to estimate the run time for the bed.

0.001

0.01

0.1

1

10

100

1000

30 40 50

Cro

ssov

er c

once

ntra

tion

(ppb

)

Percentage of cation of resin

195

J. ION EXCHANGE

(xxx)

breaking to the same concentration at the same time. If you focus specifically on the total concentration of Na and Cl, say 50 ppt, then an optimization curve can be developed, as in Fig. 5. At the peak, the maximum number of days this bed can operate to is defined. This is also where the concentrations of Na and Cl are equal. On the rising (left) side of the curve, Na breaks to 50 ppt while on the falling (right) side of the curve, Cl breaks to 50 ppt. To repeat this computation for other breakthrough concentrations requires reviewing the complete concentration profiles that were used to generate Figs. 3 and 4.

Fig. 5 BWR maximum operating time obtainable as a function of percentage cation resin.

3.2 Pressurized Water ReactorIn PWRs water within the steam cycle does not contact the

nuclear core, but rather a steam generator. As a result, water in the steam cycle does not contact the core, should not be radioactive (except for steam generator leakage) and materials of construction can be protected by pH control chemicals to minimize erosion and corrosion. Water pH is controlled by addition of ammonia, and low molecular weight amines; such as, ethanolamine, morpholine and a few others. Ammonia is the amine we will consider for the simplified analysis we are discussing here. So the three ions of interest become NH4

+, Na+

and Cl-

An interesting characteristic of amine-cycle operation is Na throw. This phenomenon occurs because there are some Na ions left on the resin after regeneration and an amine breakthrough front “scrubs” or “pushes” these ions off the resin and adds these ions to the Na exchange front. They are the result of the localized relative selectivities of Na

.

+

For the PWR cases we performed a series of simulations under conditions outlined in Table 2 with varying cation to anion resin ratios. The sequence of simulations was similar to those performed for the BWR case described previously. First, simulations with an equivalent bed of resin gave the baseline case for comparison.

and amonium. With amine selectivity higher than Na the amine is preferred. If there were no residual Na on the resin its exchange front would look like a normal breakthrough curve and not have the spike in

concentration associated with Na throw. The more efficientregeneration (fewer Na ions on the resin) the less significant the Na throw spike. For highly inefficient regeneration the Na spike off the bed can be quite significant; even to the point of exceeding an initial bed effluent criterion. In these test cases we selected a new or pristine bed with essentially no initial Na loading.

Two bed ratios are presented in Fig. 6. Several points can be made from these groups of curves. For the equivalent bed, again, the key breakthrough curves are cationic. The NH3

concentration was set by pH and was significantly larger than the Na concentration. As a result, despite a higher selectivity coefficient for NH3, without initial Na on the bed the NH3

concentration is proportionally higher through the entire breakthrough curve. The Na and Cl concentrations fed to the bed are those of a large condenser tube leak and the resulting breakthrough time to 1 ppb is less than 1.0 days for Na and about 5 days for Cl. Note that the Cl curve has an initial low value that rises to 0.01 ppb between 0.5 and 1.5 days. This concentration change is directly related to the aminebreakthrough curve and is the result of the pH change. At the amine breaks the pH is moving from near neutral to 9.6 over this time and the change in OH concentration drives this rise in observed Cl by localized equilibrium near the bottom of the bed. Once the effluent concentration is stable then the Cl curve establishes a relatively stable level as well.

Fig. 6 PWR effluent breakthrough curves for three ions at two different cation resin fractions.

Percentage of cation resin inside parenthesis

196

Vol.25, No.4 (2014)

(7)

With the large difference in break time until 1.0 ppb for Na and Cl then an adjustment in the resin ratio can extend the time until any ion achieves 1.0 ppb. After a series of simulations with adjusted resin ratios a crossover point was observed near 1.0 ppb. The second set of curves on Fig. 6 is for this cationic ratio of 60%; with the crossover occurring near 2.0 days – as seen in the boxed area. As with the BWR case an additional day of operation at this condition is possible. However, note that for this case the amine break would have occurred a few hours before the crossover point so the 1.0 ppb point for any ion would be about 1.7 days, still an extension in time over the equivalent bed ratio.

By numerous simulations over a wider range of cation to anion resin ratios, then plotting the time until the first ion reaches a specific concentration, an optimum can be defined. Figure 7presents the optimum cation ratio that can achieve a 0.05 ppb effluent concentration. This plot indicates that a 0.05 ppbconcentration can be delayed as long as 23 hours if the cation fraction is 0.58.

Fig. 7 PWR maximum operating time possible as a function of the percentage of cation resin

For normal PWR operation bed times are most commonly associated with the amine breakthrough curve. This is because PWRs regenerate the beds and Na throw is likely to occur in subsequent service cycles. To avoid Na throw beds are removed from service before the amine concentration starts to rise. In point of fact, if the regeneration efficiency was near perfect, or the bed was initially pristine, then operation through the amine break should not be an issue. Even when regeneration is not complete, known regeneration efficiency can be used to define the height of a Na throw peak. Since plants do not sample the cationic resin for a residual sodium measurement then selecting

to operate solely in the hydrogen cycle avoids an Na peak whether it would be significant or not.

Figure 8 presents amine breakthrough curves for the case in Table 2, but with 10 ppt Na and Cl. The NH3

Figure 9 shows the time expected over the range of bed resin ratios to a 50 ppt end point. Since the anion concentration (Cl) is particularly low and does not breakthrough within this range of conditions, then there is no optimum with these simulations.

concentration remains to control the steam cycle pH. For the small-bed, low-flow example simulations of this case there is very low Na and Cl effluent concentrations and the breakthrough criteria is dominated by the amine. Even with the case of 75% cationic resin there is still sufficient capacity for Cl removal.

Fig. 8 PWR breakthrough curve for normal operation

Fig. 9 PWR operating time curve for normal operation

197

J. ION EXCHANGE

(xxx)

There are several practical considerations with these example simulations. Many plants try to take an overall best approach to the use of their ion exchange resins. They want the best daily ultrapure water with the protection for a small tube leak. In the event of a moderate tube leak the plant will be challenged regardless. In the event of a major leak water chemistry will be severely challenged. However, unless detailed calculations are performed using specific plant operating conditions and condenser water analysis there is little way to know whether ion exchange bed run times can be extended while still meeting the power plant objectives.

4. ConclusionOptimization of cation to anion resin ratio based on feedwater

chemistry can provide additional time of operation for a mixed bed ion exchange system. The approach should be considered in addition to existing plant specific performance criteria giving operators an expanded list of options. The technique has the greatest potential for cases where low continuous concentration is the primary objective for the condensate polishers. If the objective is protection against a condenser tube leak, then adjusting the resin ratio to match the charge ratio of cations to anions in the cooling water can add some response time that is a function of the size of the leak.

References1. S. A. Goldman, United States Patent 6,380,456, Apr. 30

(2002).2. F. G. Helfferich, Ion Exchange, McGraw-Hill, New York,

(1962).3. R. Kunin and F. X. McGarvey, Industrial & Engineering

Chemistry, 43, 734-740, (1951).4. J. R. Caddell and R. L. Moison, Performance of ion

exchange resins at high flow rates, U.S. Atomic Energy Commission, Technical Information Service, Oak Ridge, Tenn., (1955).

5. R. Kunin, Elements of ion exchange, Reinhold, New York,(1960).

6. J. P. Copeland, C. L. Henderson and J. M. Marchello, AIChEJournal, 13, 449-452, (1967).

7. J. P. Copeland and J. M. Marchello, Chemical EngineeringScience, 24, 1471-1474, (1969).

8. F. A. Glaski and J. S. Dranoff, AIChE Journal, 9, 426-431,(1963).

9. T. Kataoka, N. Sato and K. Ueyama, Journal of ChemicalEngineering of Japan, 1, 38-42, (1968).

10. T. Kataoka, H. Yoshida and K. Ueyama, Journal ofChemical Engineering of Japan, 5, 132-136, (1972).

11. R. Schlögl and F. Helfferich, The Journal of ChemicalPhysics, 26, 5-7, (1957).

12. T. G. Smith and J. S. Dranoff, Industrial & EngineeringChemistry Fundamentals, 3, 195-200, (1964).

13. J. C. R. Turner and C. B. Snowdon, Chemical EngineeringScience, 23, 221-230, (1968).

14. M. Franzreb, W. H. Höll and H. Sontheimer, ReactivePolymers, 21, 117-133, (1993).

15. V. N. Chowdiah, G. L. Foutch and G.-C. Lee, Industrial &Engineering Chemistry Research, 42, 1485-1494, (2003).

16. T. Kataoka, H. Yoshida and T. Uemura, AIChE Journal, 33,202-210, (1987).

17. C. E. Haub and G. L. Foutch, Industrial & EngineeringChemistry Fundamentals, 25, 373-381, (1986).

18. C. E. Haub and G. L. Foutch, Industrial & Engineering,Chemistry Fundamentals, 25, 381-385, (1986).

19. S. V. Divekar, G. L. Foutch and C. E. Haub, Industrial &Engineering Chemistry Research, 26, 1906-1909, (1987).

20. E. J. Zecchini, 9106894 Ph.D., Oklahoma State University,(1990).

21. E. J. Zecchini and G. L. Foutch, Industrial & EngineeringChemistry Research, 30, 1886-1892, (1991).

22. S. K. Pondugula, M.Sc. Thesis, Oklahoma State University, (1995).

23. R. Bulusu, M.Sc. Thesis, Oklahoma State University, (1994).

24. S. K. V. Sunkavalli, M.Sc. Thesis, Oklahoma State University, (1996).

25. D. F. Hussey, M.Sc. Thesis, Oklahoma State University(1997).

26. D. F. Hussey, 9979168 Ph.D., Oklahoma State University,(2000).

27. Y. Liu, 9918802 Ph.D., Oklahoma State University, (1998).28. Y. Jia, 3164266 Ph.D., Oklahoma State University, (2004).29. Y. Jia and G. L. Foutch, Reactive and Functional Polymers,

60, 121-135, (2004).30. Electric Power Research Institute (EPRI), Technical Report,

Design and operating guidelines for condensate polishers in nuclear power plants, (1991).

31. Electric Power Research Institute (EPRI), Technical Report,Condensate polishing guidelines for PWR and BWR plants,(1997).

32. T. Yoon, 9035222. Ph.D., Oklahoma State University, AnnArbor, (1990).

198