Mass Transfer Coefficient as an Indicator of ... - ACS Publications

c School of Computing and Engineering, University of Missouri Kansas City, Kansas ... Keywords: Condensate polisher resin, Film forming amine, Ion exc...
0 downloads 0 Views 2MB Size
Download PDF
Article Cite This: Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

pubs.acs.org/IECR

Mass-Transfer Coefficient as an Indicator of Resin Performance: Impacts of Film-Forming Amines and Storage Time on Condensate Polishing Ion-Exchange Resins Maruful Hasan,† M. Montashirur Rahman,*,† Ahmedul Kabir,‡ Allen Apblett,‡ and Gary L. Foutch§ School of Chemical Engineering and ‡Department of Chemistry, Oklahoma State University, Stillwater, Oklahoma 74078, United States § School of Computing and Engineering, University of Missouri Kansas City, Kansas City, Missouri 64110, United States Downloaded via TUFTS UNIV on July 25, 2018 at 15:43:43 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

†

S Supporting Information *

ABSTRACT: Ion-exchange resins used in ultrapure water processing are known to lose effectiveness with fouling. Masstransfer coefficient (MTC) is one property that defines acceptable resin performance. MTC is measured by a simple column test with known feed concentration of ions at sufficient flow to achieve kinetic leakage out of bed. MTC deviation due to exposure to film-forming amines (FFAs), as well as impacts of resin storage time, are investigated. A filming amine (FA), octadecylamine (ODA), was evaluated experimentally for potential impacts on condensate polisher ion-exchange bead resins (Dowex MS 650C (H) and Dowex MS 550A (OH)). Results indicate significant reductions in MTC on both cationic and anionic resin. In addition, unused resin stored for excessive times, up to 16 years, showed MTC reduction.

■

through the bed.6 The concentration of ions in the feedwater that leaves the bed without the opportunity for exchange is referred to as kinetic leakage. Beds that exhibit elevated kinetic leakage are not effective for trace contaminant removal even if they have high remaining exchange capacity. For equilibrium leakage, the effluent concentration depends on the regeneration efficiency of the mixed bed. Equilibrium leakage will occur at a specific flow rate until the bed saturates with ions in the feed and begins to break through into the outlet water stream. At this high flow condition, any bed will exhibit kinetic leakage, and the entire bed depth may be considered the exchange zone. As the degree of fouling increases, kinetic leakage will occur at lower flow rates. When the feed concentration is higher, equilibrium leakage concentration will not be affected, but breakthrough will occur sooner, whereas kinetic leakage will be a function of both the feed concentration and the degree of fouling. The MTC test method uses a high flow rate and sufficient concentration to ensure that the bed operates with kinetic leakage. The overall MTC experiment (ASTM D 6302-98) has two primary functions: a qualitative indication that a problem may exist and the ability to quantify the MTC of the resin currently.711 A numerical value of MTC can be used to calculate the expected

INTRODUCTION Many processing industries, such as power, pharmaceutical, and chemical, use ultrapure water to minimize corrosion.1 Ionexchange resins can purify large volumes of water economically. Mixed-bed ion-exchange resin, composed of anionic and cationic beads, can purify water to parts per trillion levels. Unfouled ion-exchange resin exhibits very low resistance to ionic mass transfer and can be modeled by water-phase film diffusion. However, the resin can foul by repeated use or deviation from the recommended operation. When the exchange is slowed by fouling, factors like particle diffusion and exchange-site reaction need to be considered.2 An overall mass-transfer coefficient includes all factors and represents the quality of the ion-exchange resin.3 Resin fouling results in inefficient purification and reduced time until breakthrough.4 The two fouling mechanisms of ion-exchange resins are (1) loss of exchange-site capacity and (2) non-exchange site or bead fouling that creates a physical barrier to ionic mass transport.4 In addition to influent water impurities, chemicals used to minimize corrosion may also foul the resin.5 Even unused resin sitting in storage can degrade over time, resulting in a reduction in mass-transfer properties. Mass-transfer coefficient (MTC) is defined as the rate at which ions move from the bulk solution to the exchange site and has the same units as velocity. As a lumped parameter, numerous factors contribute to the MTC value. In practice, a low MTC indicates that the rate at which ions are removed by the exchange is slow compared to the rate at which water flows © XXXX American Chemical Society

Received: Revised: Accepted: Published: A

April 18, 2018 July 12, 2018 July 18, 2018 July 18, 2018 DOI: 10.1021/acs.iecr.8b01681 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

21%. For higher concentrations of humic acid, the capacity loss was as high as 30%. Chernyshev et al.16 studied the standby corrosion prevention by ODA in a cogeneration plant. By analyzing eight hot water boiler stations, Chernyshev et al. showed that an ODA layer preserved most surfaces after a long-term outage. ODA lowered the intensity of the formation of scale in seasonal autumn and winter maximum load operation. ODA also had an inhibiting effect on the pitting corrosion of stainless steel. Bäßler et al.17 studied the behavior of pitting corrosion at temperatures between 150 and 250 °C. They found that the inhibiting effect of ODA is more marked at lower temperatures. Although many studies have been done on the corrosionresistant properties of the FFAs, there is very little information on the impacts of FFA on ion-exchange resins. Previous studies on ODA focused on the inhibiting properties of the ODA on the system materials. Ion-exchange resin performance is subjective to the exposure of chemical compounds in the circulating water. Therefore, presence of ODA in the circulating system is expected to influence ion-exchange resin operating capacity. The inefficiency of exchange can be identified using the MTC as a parameter of evaluation. In this article, MTC as an indicator of resin performance is investigated for two types of fouling mechanisms. The first section identifies the reduction in MTC due to a filming amine, ODA. The second aspect is the loss of performance capacity due to storage of resin for a long time.

performance of the resin in full-scale operation. Detailed theory of the MTC experiment is provided in the Supporting Information. Lee et al.2 used MTC as an indicator of resin fouling. They analyzed samples of different ages of resin from two coal-fired power plants in Oklahoma. The older resin samples showed consistently lower MTC than newer resins. In addition, one of the plants showed greater fouling with the result traced to organics in river water downstream of oil refineries used in the condensers. They also investigated the MTC changes with various experimental conditions, like influent flow rate and concentration and found no strong correlation between them as expected. Therefore, MTC represents the resin quality well with the difference between samples indicating the degree of fouling. FFAs, or film-forming amines, often referred to as polyamines or fatty amines, are typically long-chain cationic surfactants that reduce corrosion of metal surfaces in aqueous systems. The amine has one hydrophilic end and one hydrophobic end. The hydrophobic end makes the surfaces unwettable by creating a physical barrier that prevents water, oxygen, and other corrosive agents from reaching the metal surface. A more detailed explanation of the use of FFAs is provided in the Electric Power Research Institute’s (EPRI’s) 2010 report12 on the assessment of amines in fossil plant applications. FFAs are intended to deposit and protect surfaces and have a long history. One patent from 1949 covers their use in corrosion prevention.13 The patent describes the deposition of primary amines with straight carbon chains from C10 to C16 by adding them in small quantities to gases, vapors, and liquids. The FFAs form a monomolecular layer on materials they contact. Examples of FFAs include octadecylamine (ODA), hexadecylamine, and tetradecylamine, along with their related salts. Upgrades to the formulations were made over the next few decades. Walker and Cornelius14 described emulsion formulas containing tallow amines, ethylene oxide, and propylene oxide in water. There are numerous commercial FFAs available currently with proprietary formulations focused on specific applications. Their usage is an active research area, but with their chemistry unknown, only a few publications are cited within this paper. FFAs have a strong affinity to surfaces. Distribution ratios and average surface coverage of FFAs are covered by Voges and Hater.15 An autoclave was filled with deionized (DI) water containing a particular concentration of FFA to determine the distribution ratio. After achieving a certain pressure, water samples from the autoclave and condensate collector were analyzed with the concentration of filming amine measured by a Bengal Rose method. The distribution ratio is the quotient of concentration in the condensate divided by the concentration in the autoclave. Experiments determined which amines have the highest distribution ratios and average surface coverage. Voges and Hater found that the distribution ratio decreased as the operating pressure increased and vice versa. The average surface coverage was not a strong function of operating pressure but did increase linearly as the initial concentration of the FFA increased. Gonder et al.1 covered the fouling of anion exchanger by organic matter, such as humic and fulvic acids. They argued that the degradation products of cation exchangers caused fouling of anion-exchange resins. They also compared the breakthrough curve and capacity for fresh and fouled resin. For resin fouled by 0.13 mg/L humic acid, the capacity loss was

■

MATERIALS AND METHODS Resins were exposed to defined concentrations of amines at known operating conditions in laboratory test columns. Variables included amine concentration, operating temperature, water flow rate, and contact time. The test columns were removed from service, and the resins were analyzed for properties that define their performance. Laboratory and Experimental Setup. The experimental system was based on ASTM International Standard 6302 (98).7 A detailed description of the experimental setup, test equipment, analytical capability, and procedures are given in a 2002 EPRI report.18 The system consists mainly of carboys, pumps, and a test column. Figure 1 represents the MTC testing apparatus. The dilute stock solution (feed carboy 1) was pumped into the main line while a branch line was used to deliver the FFA (feed carboy 2) into the mainline upstream of the static mixer before the ion-exchange resin column. In addition to the MTC apparatus, auxiliary units were used for regeneration of resin and FFA exposure. An ion exchange chromatography (IEC) and personal computer (PC) were used to analyze the sample concentration. The test equipment is provided in Table S1, Supporting Information. Information on the chemicals used in the resin testing is provided in Table S2. The experimental column used in this study was made of Pyrex glass. These columns had been used in a previous investigation to study the incomplete regeneration of cationic resin.19 Fritted disks were attached to both ends of the column to contain resin within the column. The column inside diameter was 1 in. [25.4 mm], with length of 24 in. [609.6 mm]. Strong anionic and cationic resins utilized in whole bed condensate polishing systems were used. Three-year-old unused resins were also tested. These were Dowex MS 650C (H) cationic resin and Dowex MS 550A (OH) anionic resin B

DOI: 10.1021/acs.iecr.8b01681 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

to the column as a slurry. The transfer was done slowly and carefully to minimize air bubbles in the column. Then the resin was rinsed by DI water for 30 min at 50 mL/min, and 6.5% of sulfuric acid (standard regeneration) or 8.5% of sulfuric acid (strong regeneration) was used as the regenerant solution. Up to four-bed volumes of regenerant solution were passed through the resin at 25 mL/min. A volume of 100 mL of anionic resin was measured by coring and transferred as a slurry to another column. The same procedures were followed for anionic resin as those followed for cationic resin. Five percent sodium hydroxide was used as standard regenerant, and 8% sodium hydroxide was used as strong regenerant. Rinsing. Rinsing was done by gravity. The column was connected with an overhead carboy containing DI water. The effluent was collected in a waste carboy. The flow rate of DI water was controlled by the carboy valve. At first, the flow rate was approximately 5 mm of water above the resin. After 15 min, the flow rate of DI water increased to 100 mL/min. The effluent sample conductivity was measured by a conductivity meter. The rinsing procedure continued until the effluent conductivity was less than 5 μS/cm. Resin Mixing and Column Preparation. A 150 mL of cationic resin and 75 mL of anionic resin aliquot was transferred by coring into the plastic storage container. Extra water was decanted, and the resins were mixed thoroughly with a glass rod. The mixed resin was transferred to the test column as a slurry slowly to minimize air pockets. A small amount of DI water was used to rinse the resin adhering to the side of the column. The water level was kept at not more than 5 mm above the resin surface to minimize separation. After the resin loading, the column was installed into the MTC test unit. MTC Experimentation. The FFA (ODA) solution was prepared and stored in carboy 2 (Figure 1). Because of the strong affinity of ODA for surfaces, the ODA solution was injected right before the column. A metering pump was used to pump the ODA solution into the main feed line. The branch line could be isolated using valve 2. The speed of the metering pump was adjusted to ensure the correct concentration of ODA would be injected into the column. A static mixer was used to mix the DI water from the main feed line and the ODA solution from the branch line. The effluent was collected in a waste tank. There was a sample collection provision before the column. This sample collection branch also helped to vent and air during startup and allowed continuous solution flow. A high flow rate (∼1 L/min) was used to create kinetic leakage in the column (defined as effluent concentration above the equilibrium leakageas required by the ASTM method. Concentration of specific ions were measured for both influent and effluent samples. Ion Exchange Chromatography Procedure. A solution of 3.2 mmol of Na2CO3 and 10 mmol of NaHCO3 was used for the anion eluent solution. The cation eluent solution was 2 mmol of nitric acid and 0.7 mmol of PDCA (2,6-pyridine dicarboxylic acid). A solution of 100 mmol of H2SO4 was used for regeneration, and DI water was used for rinsing. The baseline was set for 70 min, and then the chromatograph was calibrated using standard solutions ranging from 0.01 to 10 ppm. The sample from the experiment was injected, and the chromatograph detected the concentration automatically. Exposing with ODA. One of the exposure procedures was using gravity. ODA solution was made in a low-density polyethylene carboy that was then connected to the test column containing the resin. The amine solution was slowly

Figure 1. Kinetic test apparatus.

(Dow Chemical Company). Physical properties of these resins are listed in Table S3. All the resins were regenerated and rinsed with ultrapure water before experimentation. A Metrohm 790 personal IEC analyzed the influent and effluent samples. A personal computer (PC) was used to record and evaluate the chromatograph data and make operational changes. An injection valve was used for an individual sample or DI water injection. A low-pulsation double-piston highpressure pump delivered a flow range of 0.2−2.5 mL/min at a maximum pressure of 25 MPa. The column chamber was insulated to ensure a thermally stable condition for the separation column and to shield the system against the electromagnetic interface. A cationic column and an anionic column with a suppressor were used in the chromatograph. The suppressor module was pressure-resistant and regenerated automatically. An integrated two-channel peristaltic pump with a flow rate of 0.5 mL/min was used to regenerate and rinse the suppressor module. A conductivity detector operated at 40 °C varied by less than 0.01 °C. Solution conductivity was measured with a Milwaukee MW301 m, and pH was measured with a Hach H138 Elite miniLab ISFET pH Meter. The conductivity meter was factory-calibrated but checked with a National Institute of Standards and Technology-traceable conductivity standard. The pH meter was calibrated at two points: pH 7.0 and either 4.0 or 10.0, depending on the measurement range. Experimental Procedure. Several resin bed heights in the test column were used. Also, single-bed and mixed-bed experimentations were performed. The experimentation procedure for 15-in. [381 mm] mixed-bed column is explained in the following sections. Separation of Anion and Cation Resin. Mixed resin was backwashed with DI water to separate the phases. The backwash flow rate was sufficient to give a 50% bed expansion. Next, the anion resin on the top was removed using an aspiration assembly while cation resin was taken from the bottom. The middle layer of mixed resin was not taken to minimize cross-contamination. The procedure was repeated to give satisfactory separation. Regeneration. Regeneration was done on all separated resin samples before MTC testing. A volume of 200 mL of cationic resin was measured by coring. Then the resin was transferred C

DOI: 10.1021/acs.iecr.8b01681 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

experimental errors associated with the MTC, the reduction in MTC values is attributed to the FFA exposure. A reduced MTC means fewer ions will be transferred to the exchange site and, consequently, a lower exchange efficiency. Although the actual nature of reduction mechanism is unclear, based on the kinetic experiments a reasonable assumption is that resins exposed to ODA will cause a reduction in ion exchange capacity. After regeneration by 6.5% sulfuric acid, the MTC was still about 12% lower than the original unexposed MTC. Therefore, even after regeneration, ion-exchange efficiency could not be restored to the original value. This means that the impact of ODA on ion-exchange resin is shown to have lasting effects. On the operations level, this reduction in MTC implies a shorter operating time for a plant and frequent regeneration of the bed will be required. Figure 3 shows the MTC variation for anion-exchange resins when exposed to ODA. The trend in Figure 3 is similar to

passed through the column. Another procedure was to make a small amount (∼50 mL) of FFA solution and inject it directly into the resin in the test column. Subsequently, DI water was passed slowly to the column by gravity for 3 h. Experimental Procedure for Stored Resins. The experimental setup follows ASTM procedure D-6302-98 (reapproved 2004) used to evaluate changes in kinetic performance of ion-exchange resins. The theory behind the experiment is based on the derivation of Harries and Ray.3 Test conditions were limited to kinetic rather than equilibrium leakage by ensuring the test column flow rate was sufficiently high. This effluent concentration was influenced by a combination of flow velocity, feed concentration, and column bed depth. The feed solution was pumped at 1000 mL/min through the column. Both influent and effluent samples were collected and analyzed by IEC. Old unused resins were first regenerated by the acid or base recommended by the manufacturer. Resins were rinsed with deionized water before loading the test column. Na+ and Cl− of different concentrations ranging from 50 to 1000 ppb were fed to the column. An effluent sample was collected and Na+ and Cl− concentrations were measured by an IEC.

■

RESULTS AND DISCUSSION Figure 2 depicts the impact of ODA on MTC for cationic resin. An ODA solution of 0.5 ppm was made in isopropanol.

Figure 3. Changes of the MTC for exposure to ODA of anion (8-in. [203.2 mm] bed height).

cation-exchange resins with minor difference in numerical values. The unexposed anion resin provided the best MTC values for kinetic leakage condition. The average MTC was lowered by approximately 18% when exposed to ODA compared to a 20% reduction for cation-exchange resins. However, since anion-exchange resins have nearly 50% lower exchange capacity compared to cation-exchange resin (anionexchange resin capacity is 1.1 equiv/L for Dowex 550A whereas the cation-exchange resin, Dowex 650C, has a capacity of 2.0 equiv/L), the impact of ODA is relatively significant for anion-exchange resin. The lower exchange capacity coupled with the reduction in MTC values will cause early breakthrough for anionic contaminants. The regenerated resin MTC values are not significantly better than those of exposed resin, yielding a 2% MTC recovery after treatment with 5% NaOH. The relatively lower percentage recovery of MTC after regeneration of anionic resin compared to cationic resin implies a stronger fouling mechanism for anion-exchange resins. One possible explanation could be that the film formed due to exposure to ODA prevents the quaternary amine functional group in the anionic resin from further exchange. For power plant applications, the amine formulation, entrained with water in the steam cycle, has a high affinity for metal surfaces and deposits as a protective film. This film is a barrier for corrosive chemicals and dissolved gases and, because of its

Figure 2. Changes of the MTC for exposure to ODA of cation (8-in. [203.2 mm] bed height).

Although solubility is not limited in isopropanol, the concentration chosen was based on solubility of ODA in water. To compensate for the low concentration, additional solution was passed through the resin sufficient to saturate with ODA stoichiometrically. The resin was exposed to three bed volumes (0.3 L). MTC variation is nearly linear with respect to influent concentration for all resin types. Increase in Na+ concentration showed little influence on MTC. This is reasonable as the experimental conditions were kept at kinetic leakage condition; thus, equilibrium relationships, which are dependent on the influent concentrations, were not affecting MTC. However, resin exposed to ODA solution showed MTC variation, indicating resin fouling. Unexposed resin provides the highest MTC for ion exchange over the concentration range. ODA-exposed resin showed the lowest MTC among the three resin conditions. The average MTC for ODA-exposed resin was about 20% lower than unexposed resin. Excluding the D

DOI: 10.1021/acs.iecr.8b01681 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research

several times using a stopwatch and cylindrical test tube. Influent and effluent samples were diluted to a different ratio and tested in an ion chromatograph to obtain the standard errors of measurement. Surface of the resin in the test column was not perfectly flat and bed height data were taken from a different side of the column to measure the variability. Then propagation of error analysis was done. From the column material balance (assuming negligible accumulation in the bulk fluid), we find MTC

hydrophobic characteristics, minimizes wettability. Because filming occurs on all contacted surfaces, the impact of these amines on ion-exchange resin performance is seen on both types of ion-exchange resins. For optimal ion-exchange resin performance, no limitations to ionic mass transfer from the bulk water to the anionic or cationic exchange site should exist. FFAs reduce the efficiency of ion-exchange resins by numerous mechanisms. These include physical surface blockage of a pore or gel structure, permanent filling of space within the resin beads, the establishment of a surface charge or zeta potential that repulses diffusing ions, and permanent bonding of exchange sites within the beads that reduces effective exchange capacity. Even if bonding is nonpermanent and the amine is removablefor a typical resin regeneration acid or caustic treatmentthere could still be reduced operating time between regenerations if external filming blocks internal exchange sites. This blocking affect would reduce the effective capacity for the resin to remove contaminants. In addition to MTC measurements, the effect of ODA sorption on ion-exchange resin was also measured. One of the first steps in this investigation was to determine the maximum uptake of the FFA by the resin. The resins were exposed to an excess of 0.01 M FFA for 2 weeks, and the uptake was determined by titration of the unreacted amine and comparison with the starting concentration. The results are provided in Table S4. Notably, sorption by the sodium and proton forms was similar to each other. Sorption of ODA to the anion resins was not detectable, but because the experiment was performed in isopropanol, this reflects only that the ODA does not partition onto the anion resin from this solvent in which it is much more soluble. This is very different from aqueous ODA, which coats or fouls the resin easily. However, it can be concluded that the affinity of ODA for cation resin is significantly higher than it is for anion resin. The ion-exchange capacity, apparent density, and moisture content of resins before and after exposure to ODA are shown in Table S5. The exchange capacity for the Dowex MS 650C (H) cation resin decreased by 20.7% when exposed to ODA. The cation resin exposed to ODA becomes more fouled during regenerating, falling to 26.3% from the original. Presumably, this is due to redistribution of ODA in the resin from the resin backbone to the ion-exchange sites. This movement of the amine would be facilitated by the enhanced solubility of the protonated amine over the neutral amine. Anion resin, Dowex MS 550A (OH) exposed to ODA, showed a reduction in exchange capacity of approximately 37−38%. The anion capacity was not fully recovered following regeneration with a 17.7% loss for ODA-exposed resin. Regardless of the resin types, only partial recovery of the ion-exchange capacity indicates that the impact of the FFA, ODA, has an irreversible fouling mechanism. Error Analysis. Mass-transfer coefficient (k) depends on a number of factors, which can contribute to the variability. Resin bead diameter (dp) was supplied by the manufacturer and treated as constant, irrespective of the known swelling that occurs for cationic resin in the hydrogen form. Packed-bed resin was used for the experiment and voidage (ε) was considered constant. Error analysis was done for a single bed resin experiment, so resin volume fraction (R) was unity. In addition, column diameter (A) was constant as the single column was used. Other input measurements were considered subject to error and standard errors of measurement were calculated for these data. Volumetric flow rate was measured

k=

f V ijj Ci yzz lnjjj eff zzz 6(1 − ε)R Az k Ci {

dp

dp 1 6(1 − ε)R A ij Cif yz zzz lnjjjj eff z k Ci {

Considering k=M

V z

(1)

= M(constant), which leads to

(2)

We can differentiate the above equation to get the propagated error for MTC,

f f 1 ijj Ci yzz V ij 1 yz V ij Ci yzz zzez + M jjj f zzzeCif lnjjj eff zzze v + M 2 lnjjjj eff z z k Ci { z jk Ci z{ z k Ci { V ij 1 yz + M jjjj eff zzzzeCieff z k Ci { (3)

ek = M

e v , ez , eCif , and eCieff are the standard error of measurement for volumetric flow rate, bed height, influent concentration, and effluent concentration, respectively. Figure 4 shows the results of the error analysis for new Dowex 650 C (H) cation resin MTC experiment.

Figure 4. Error analysis for MTC experiment.

Results of error in MTC came low (about ten power minus seven ranges). A new ion chromatograph measured the concentration. Therefore, standard error of concentration was low (0.001−0.004 range). Variability was significant in volumetric flow rate measurement, but the input unit in the equation was in cubic meter per second. In addition, the numerical value of volumetric flow was also low. Therefore, the main error contributor was the variability of bed height. However, considering the entire factor, the final error in the MTC was minimal and can be ignored for MTC comparison. E

DOI: 10.1021/acs.iecr.8b01681 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research Impact of Resin Storage Time. MTC for used ionexchange resins was discussed by Lee et al.2 who found that fouled resin has 30−50% lower coefficients than that predicted or measured for unfouled resin. In this study, MTC of unused old resin was measured to determine the degradation or fouling due to storage time only. Dowex brand anionic and cationic resin of different ages were used. Figures 5 and 6 show Na+ and Cl− MTC data for various influent concentrations and ages of resin. MTCs are essentially

resins, MTCs were more than 50% lower than those for new resins. The deviation of MTC with age even when unused indicates that the resins will be unusable after a certain amount of time. This loss of functionality is independent of the process operating conditions (influent concentration of contaminants, bed geometry, flow rates, and exposure to FFAs). The reduction is more severe for anionic resins−as expected for the more unstable amine functionality. For instance, MTC for 3-year-old dry resin was 45% lower than new resin while 5year-old wet anion resin showed 35% reduction. Similar to cation-exchange resins, better performance was observed for the resins stored in wet condition. For 10-year-old dry samples of anionic resins, MTCs were more than 50% lower than those for new resins. The relatively quicker loss of efficiency for anion resins compared to cation resins is attributable to the quaternary amine functional group in anion-exchange resins. We are aware that there is slight variability in flow rates, influent concentration, and bed depth from batch to batch that may account for some of the differences on MTC but these effects were not considered. Nearly horizontal lines on Figures 5 and 6 reinforce the hypothesis that MTC is independent of ionic influent concentration, as expected. As such, MTC is a property of the resin and not a function of the experimental method used to obtain its numerical value. Figure 7 shows the change of average MTC of dry cationic resins of different age. The steady decrease of MTC continues

Figure 5. Sodium MTC for old cationic resins at various influent concentrations.

Figure 7. Average MTC for old cationic dry resins of different ages. Figure 6. Chloride MTC for old anionic resins at various influent concentrations.

as the age of the resin increases. At this rate, resin MTC can fall below 0.0001 m/s in approximately 5 years and become unusable. Figure 8 shows the average MTC of dry anionic resins of different ages. MTC loss was higher in the first 3 years for anionic resin than cationic resin. Therefore, the anionic resin may become unusable more quickly than cationic resin. The trend of MTC loss in both cases is close to linear. So, for both cases, linear equations were obtained to relate age and MTC for dry resin.

constant over the feedwater concentration range within experimental error. These results are similar to the data from Harries and Ray5 that showed slightly lower MTCs at lower chloride feed concentrations in experiments with fouled resins. Resin stored dry were available for ages of 10 and 16 years. Regardless of the storage condition (dry or wet), resins showed MTC decrease with age for both cation- and anion-exchange resins. In Figure 5, cationic resin with sulfonate functionality, both 3-year-old dry and 5-year-old wet resin show 22% MTC decrease compared to unexposed fresh cation resin. Although the wet resins were 2 years older than the dry resins, the similar loss of MTC efficiency indicates better protection from fouling for the wet condition. For both 10- and 16-year-old dry sample

■

CONCLUSIONS A series of experiments to evaluate ion-exchange resin properties before and after exposure to FFA was performed. Experimentation was performed with octadecylamine (ODA), a nonproprietary, commercially available FFA. Results indicate F

DOI: 10.1021/acs.iecr.8b01681 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Industrial & Engineering Chemistry Research

■

Article

AUTHOR INFORMATION

Corresponding Author

*Tel.: 405-385-2572. Fax: 405-744-6338. E-mail: montashirur. [email protected]. ORCID

M. Montashirur Rahman: 0000-0002-8399-2206 Allen Apblett: 0000-0002-1251-9497 Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS The constructive and detailed comments of three anonymous reviewers are greatly acknowledged. NOMENCLATURE as = Specific surface area of resin bead (m2/m3) dp = Harmonic mean diameter of the spherical particle (m) Ci = Fluid-phase concentration (mg/L) Cfi = Fluid-phase concentration of feed (mg/L) Ceff i = Fluid-phase concentration of effluent (mg/L) Ci* = Concentration at liquid−solid interface (mg/L) F = Volumetric flow rate (m3/s) FR = Volumetric fraction of resin (m3/m3 bed) k0 = Overall mass-transfer coefficient (m/s) kf = Film mass-transfer coefficient (m/s) kp = Particle mass-transfer coefficient (m/s) K = Distribution factor Kf,i = Average film mass-transfer coefficient (m/s) qi = Resin-phase concentration (mg/L) T = Time (s) ur = Fluid velocity (m/s) Z = Bed depth (m) Ε = Porosity

Figure 8. Average MTC for old anionic dry resins of different ages.

that ion-exchange resins are impacted by exposure to ODA. Reductions in MTC were permanent. Contact between ODA and ion-exchange bead resin should be avoided at the higher dosages, which suggested achieving sufficient layup protection when added over a short period (hours). ODA exposure reduced the ion-exchange capacity by 20% for cationic resin while anionic resins lost 37% of the original capacity. Regeneration was more effective for anionic resins. If exposure does occur, using maximum regenerant concentrations will be required to use the resins further. Operators should anticipate that (a) MTC will likely stabilize at 10−20% lower than that of new resins after the first regeneration and (b) the use of regenerant, both in concentration and frequency, is likely to be significantly greater. At lower MTC, the initial cationic breakthrough will occur earlier. The earlier breakthrough may result in more frequent regenerations and reduced bed effectiveness in the case of concentration ingress entering the polishers, such as condenser tube leak. Lower MTC for both cationic and anionic resins is shown to yield a reduced ion-exchange capacity; thus, the acceptability of ion-exchange resin can be evaluated through MTC measurement. The results show the steady decrease of ionexchange resin MTCs, indicating deteriorating resin quality, even when unused. MTC decreased linearly with age. For maximum effectiveness, the resin should be used within a reasonable time. Indefinite storage cannot be assumed.

■

■

REFERENCES

(1) Gönder, Z. B.; Kaya, Y.; Vergili, I.; Barlas, H. Capacity loss in an organically fouled anion exchanger. Desalination 2006, 189 (1−3), 303−307. (2) Lee, G. C.; Foutch, G. L.; Arunachalam, A. An evaluation of mass-transfer coefficients for new and used ion-exchange resins. React. Funct. Polym. 1997, 35 (1−2), 55−73. (3) Harries, R. R.; Ray, N. J. Anion exchange in high flow rate mixed beds. Effluent Water Treat. J. 1984, 24 (4), 131−139. (4) Chowdiah, V. N.; Foutch, G. L.; Lee, G. C. Binary Liquid-Phase Mass Transport in Mixed-Bed Ion Exchange at Low Solute Concentration. Ind. Eng. Chem. Res. 2003, 42 (7), 1485−1494. (5) Raught, D. P.; Foutch, G. L.; Apblett, A. Ion Exchange Resin Fouling by Organic Amines in Secondary Systems at U.S. Nuclear Power Plants. Power Plant Chem. 2005, 7 (12), 741. (6) Hussey, D. F.; Foutch, G. L. Ion-Exchange Kinetics for Ultrapure Water. In Ion Exchange and Solvent Extraction; Sengupta, A. K., Marcus, Y., Eds.; Marcel Dekker: New York, 2004; pp 339−373. (7) ASTM Standard D6302-98. Standard Practice for Evaluating the Kinetic Behavior of Ion Exchange Resins; ASTM International: West Conshohocken, PA, 2017; www.astm.org. (8) Jia, Y.; Foutch, G. L. True multi-component mixed-bed ionexchange modeling. React. Funct. Polym. 2004, 60 (0), 121−135. (9) Frisch, N. W.; Kunin, R. Kinetics of Mixed-Bed Deionization: I. AIChE J. 1960, 6 (4), 640−647. (10) Pandey, A. K. Kinetic Study of Ion Exchange Column Operation For Ultrapure Water Application. MS Thesis, Oklahoma State University, 2008. (11) Cussler, E. L. Diffusion: mass transfer in fluid systems; Cambridge University Press, 2009.

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.iecr.8b01681. Theory of MTC experiments, determination of influence of ODA on chemical and physical properties of resins, batch tests, and ion-exchange capacity measurement procedure, list of tests of equipment, list of chemicals, detailed physical properties of the resin, uptake of FFAs by ion-exchange resins, and resin properties from batch tests (PDF) G

DOI: 10.1021/acs.iecr.8b01681 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX

Article

Industrial & Engineering Chemistry Research (12) EPRI Assessment of Amines for Fossil Plant Applications; EPRI Technical Report No. 1017475, 2010; Electric Power Research Institute: Palo Alto, CA, 2010. (13) Lewis, K. H. Method of protecting systems for transporting media corrosive to metal. U.S. Patent 7,198,706, April 25,1997. (14) Walker, J. L.; Cornelius, I. T. E. Filming amine emulsions. U.S. Patent 3,860,430A, November 5,1973. (15) Voges, N.; Hater, W. Distribution ratio and average surface coverage of film forming amines. VGB Powertech 2010, 90 (3), 72− 76. (16) Chernyshev, E. V.; Veprov, E. N.; Petrov, V. A.; Bogdanov, S. L.; Levina, T. Y.; Petrova, T. I.; Kashinskii, V. I.; Zonov, A. A.; Verkhovskii, A. E. Increasing the corrosion resistance of equipment due to the use of film-forming amines. Power Technol. Eng. 2006, 40 (1), 34−37. (17) Bäßler, R.; Uhlemann, M.; Mummert, K. Inhibiting effect of octadecylamine on pitting corrosion behaviour of stainless steel type 1.4541 up to 250 C. Mater. Corros. 1999, 50 (3), 146−153. (18) EPRI Investigation of ETA Interactions in Mixed Bed Ion Exchange Systems − Phase 1; EPRI Technical Report No. 1003599, 2003; Electric Power Research Institute: Palo Alto, CA, 2003. (19) Lee, J.; Foutch, G. L. Investigation and Findings on the Thermal Stability of Neutralizing Amines. Presented at the 10th International Conference on Cycle Chemistry in Fossil and Combined Cycle Plants, Seattle, WA, 2012.

H

DOI: 10.1021/acs.iecr.8b01681 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX