Environ. Sci. Technol. 2009, 43, 4119–4123
Fully Automated Measuring Equipment for Aqueous Boron and Its Application to Online Monitoring of Industrial Process Effluents S E I I C H I O H Y A M A , * ,† K E I K O A B E , † HITOSHI OHSUMI,† HIROKAZU KOBAYASHI,† NAOTSUGU MIYAZAKI,‡ KOJI MIYADERA,‡ AND KIN-ICHI AKASAKA‡ Central Research Institute of Electric Power Industry, 1646 Abiko, Abiko-shi, Chiba, Japan, and DKK-TOA Corporation, 2-214 Sakuragaoka, Higashiyamato-shi, Tokyo, Japan
Received January 12, 2009. Revised manuscript received April 1, 2009. Accepted April 6, 2009.
Fully automated measuring equipment for aqueous boron (referred to as the online boron monitor) was developed on the basis of a rapid potentiometric determination method using a commercial BF4- ion-selective electrode (ISE). The equipment can measure boron compounds with concentration ranging from a few to several hundred mg/L, and the measurement is completed in less than 20 min without any pretreatment of the sample. In the monitor, a series of operations for the measurement, i.e., sampling and dispensing of the sample, addition of the chemicals, acquisition and processing of potentiometric data, rinsing of the measurement cell, and calibration of the BF4- ISE, is automated. To demonstrate the performance, we installed the monitor in full-scale coal-fired power plants and measured the effluent from a flue gas desulfurization unit. The boron concentration in the wastewater varied significantly depending on the type of coal and the load of power generation. An excellent correlation (R 2 ) 0.987) was obtained in the measurements between the online boron monitor and inductively coupled plasma atomic emission spectrometry, which proved that the developed monitor can serve as a useful tool for managing boron emission in industrial process effluent.
Introduction Boron naturally occurs in an oxygenated form such as boric acid or borate salt, which are abundant in seawater, volcanic spring water, and mineral ore (1). Industrially, boron compounds are utilized widely in the production of, for example, heat-resistant glass, semiconductors, soaps, detergents, and insecticides, which are anthropologic sources of boron emission. Additionally, boron concentrations much higher than the natural background are found in municipal wastewaters in industrialized countries. Boron is an essential micronutrient for plants and animals, but also is toxic in large doses. For plants, different species require different levels of boron for optimum growth and its allowable range is extremely narrow. For humans, excessive * Corresponding author phone: +81 4 7182 1181; fax: +81 4 7183 2966; e-mail:
[email protected]. † Central Research Institute of Electric Power Industry. ‡ DKK-TOA Corporation. 10.1021/es900062f CCC: $40.75
Published on Web 04/23/2009
2009 American Chemical Society
ingestion over a long period may result in eczema, stomachache, and nausea (2). Thus, the World Health Organization (WHO) reports the drinking-water guideline value of boron to be 0.5 mg/L (3) and the drinking water standards are set at 1 mg/L in the U.S., the EU (4), and Japan. In Japan, the Environmental Quality Standard for water pollution, which is a target level to be achieved in public water to protect human health, is set at 1 mg/L for boron (5). Reflecting this establishment, the National Effluent Standard for boron is set at 10 mg/L and 230 mg/L for effluents discharged to terrestrial water bodies and to coastal water bodies, respectively (6). Boron is one of the trace elements contained in coal (7, 8). Since boron is relatively volatile, the boron in coal is released to the gas phase (flue gas) owing to its combustion. Then the gaseous boron species cool and condense on solid surfaces or in aqueous solutions by adsorption or absorption. In pulverized coal-fired power plants with a flue gas desulfurization (FGD) unit, the boron in coal is released into the flue gas during combustion and then transferred downstream to the flue gas cleaning system. The boron in coal is mainly partitioned into clinker (bottom ash of the boiler), fly ash, and FGD wastewaters (9-12). The main source of aqueous boron in power plants is effluent from the FGD unit. Thus, we must achieve better control of boron emission to meet the regulations. In order to manage and control boron emission effectively, we must monitor the concentration of boron in the effluent. This study is aimed at developing fully automated measuring equipment for monitoring process effluent and at demonstrating its performance and effectiveness for the management of boron emission.
Experimental Section Rapid Determination in Potentiometric Measurement. To build a process monitor for aqueous boron, a simple and rapid determination method that can be applied to on-site measurement in a compact size is required, even if its analytical accuracy is sacrificed to some extent. The typical measurement methods of boron, i.e., inductively coupled plasma atomic emission spectrometry (ICP-AES) and colorimetric analysis (spectrophotometry), require bulky and costly analytical equipment or a time-consuming complicated pretreatment, both of which are difficult to apply to on-site analysis. We focus on potentiometric measurement using a commercial tetrafluoroborate (BF4-) ion-selective electrode (ISE). In the conventional potentiometric measurement (13-19), aqueous boron compounds react with fluoride ions (F-) in acidic solution to form BF4-, which is measured with a BF4ISE. In the case of H3BO3 with NaF, the reaction is H3BO3 + 4NaF f 4Na++ BF4 + 3OH
(1)
To achieve complete conversion and accelerate the reaction rate, an excess amount of F- is added to boron in acidic solution, wherein the full conversion requires roughly 1 h at ambient temperature. Figure 1(a) shows an example of the reaction profile, in which 10 mg/L (as boron) of H3BO3 is converted to BF4- at pH 1. Thus, higher temperatures (333-353 K) are usually employed to accelerate the reaction (13, 18, 19). Under the conditions that 100 mmol/L of NaF and 0.1-1 mmol/L of H3BO3 are present in solution, thermodynamic calculation indicates that the equilibrium conversion of H3BO3 to BF4- is 99% at pH lower than 3 (20). Thermodynamically, pH 3 is sufficient for complete conversion, but kinetically, pH 1 is required for the rapid conversion of H3BO3. VOL. 43, NO. 11, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4119
FIGURE 1. BF4- profiles in the conversion of H3BO3 with NaF. An excess amount of NaF was added to a 10 mg/L (as boron) H3BO3 solution which was acidified with H2SO4 at pH 1. (a) The complete conversion requires 0.5-1 h. (b) The final concentration of BF4can be kinetically estimated on the basis of the profile in the initial 10 min.
FIGURE 2. Online boron monitor (dimensions: 500 mm width × 500 mm depth × 1500 mm height). (a) Inside view. (b) Close-up of the measurement cell. We devised a rapid determination method without the need to heat the sample (21, 22). By analyzing the reaction kinetics in the case of a large excess of F- over H3BO3 in acidic solution, we found that the reaction obeys the firstorder kinetics with regard to H3BO3, where the BF4- concentration CB is given by CB ) CA0(1 - exp(-kt)) + CB0
(2)
where CA0: initial concentration of H3BO3, CB0: initial concentration of BF4-, CB: concentration of BF4- at arbitrary time, and k: rate constant. By applying eq (2) to a BF4- profile, we can obtain three unknown parameters, CA0, CB0, and k, from the curve fitting, where the sum of CA0 and CB0 is the final BF4- concentration that we must obtain. If we can precisely estimate the parameters from a BF4- profile in a short period, a reduction in the measurement time can be achieved without the need to heat the solution. When F- was added to the sample to 50 times the stoichiometric ratio (reaction (1)) and the solution was acidified with H2SO4 at roughly pH 1, the measurement of BF4- in the initial 10-15 min is sufficient for estimation ((10%). Thus, the final BF4- concentration can be kinetically obtained from the initial BF4- profile (Figure 1 (b)), leading to a reduction in the measurement time, which is beneficial in the case of on-site manual measurement. The details of the determination method are described elsewhere (21, 22). Instrumentation. On the basis of the above method, we constructed the fully automated measuring equipment for aqueous boron, referred to as the online boron monitor. Its internal views are shown in Figure 2. Figure 3 shows a 4120
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 11, 2009
schematic of the system, which automates a series of operations required for the measurement: the sampling and dispensing of the sample, addition of chemicals, acquisition and processing of potentiometric data, rinsing out of the measurement cell, and calibration of the BF4- ISE. The monitor was equipped with a measurement cell made of acrylic resin with an internal volume of 50 mL, which is maintained at 313 ( 2 K by circulating heated water in a water jacket around the cell. The cell housed a BF4- ISE and a reference electrode, both of which were obtained from DKK-TOA Corporation (Figure 2(b)). The BF4- ISE (EL7464L) is a liquid membrane electrode, which disperses tetrafluoroborate trioctylammonium as an ionophore, in the polyvinyl chloride (PVC) membrane. The ISE can measure the concentration of BF4anion in the range of 0.1-1000 mg/L as boron, but in practice, 1-100 mg/L in the case of high matrix samples containing ions interfering with the electrode. The ions severely interfering with the BF4- ISE are perchloride (ClO4-), iodide (I-) and cyanide (CN-) ions, while nitrate (NO3-), sulfide (S2-) and bromide (Br-) ions interfere weakly with the BF4measurement (23). The sample is collected through a sampling loop, the volume of which is adjustable between 1 and 10 mL by changing its length. After a certain amount of sample is injected into the cell, 5 mL of 9 M H2SO4 solution is added, and then the solution is diluted with deionized water by a factor of 5-50 depending on the loop volume, which is controlled by a liquid-level gauge in the cell. The monitor waits for 3 min for the solution temperature to become stable, and then 10 mL of 0.5 M NaF solution is
FIGURE 3. Schematic diagram of online boron monitor. CV: check valve, ISE: BF4- ion-selective electrode; MC: measurement cell; P: pump; RE: reference electrode; SL: sample loop; ST: stirrer; STD H: standard solution (high); STD L: standard solution (low); SP: syringe pump; TC: thermocouple; WB: water bath; WJ: water jacket. added to the solution, and the measurement starts. The absolute calibration is automatically carried out between two BF4- concentrations of one order difference, e.g., 50 and 500 mg/L in the case of a 50-fold dilution of the sample, in the same manner as the measurement. Thus, the online boron monitor can measure the concentration of boron compounds in the range of 1-500 mg/L, which covers the Japanese effluent standards, in less than 20 min without the need for any pretreatment. All chemicals employed here are of reagent grade and obtained from Wako Pure Chemical Industries. The consumable supplies in the measurement are the electrodes and low-cost chemicals, thereby enabling a large reduction (to 1/9) in the running cost as compared with the existing equipment based on the conventional spectrophotometry. Field Demonstration Test. The monitor was installed in two full-scale coal-fired power plants (plant A with a 250 MW unit and plant B with a 700 MW unit) and the boron concentration in the FGD wastewater was monitored for 1 and 4 months, respectively. The monitor automatically measured the effluent at 3 h intervals and was calibrated daily. For the verification of the measurements, the FGD wastewater samples were periodically collected and analyzed with ICP-AES. In both power plants, the boron compounds in the effluent was treated in a downstream wastewater treatment facility and discharged to public water bodies with the boron concentration below the effluent standard. Plant A was operated with nonblended coal combustion (one kind of coal at a time), whereas plant B was operated with blended combustion, in which a mixture of different kinds of coal is burned at different compositions.
Results and Discussion Plant A. The online boron monitor was successfully operated at plant A for one month. Figure 4 shows the boron behavior in the FGD effluent, as revealed by the monitor. The values measured by ICP-AES are also depicted as blue triangles in Figure 4(a). The boron concentration varied significantly in the range of 40-210 mg/L. Probable factors governing the boron behavior in the FGD effluent, i.e., the boron content in coal and the load of power generation, are also shown in
FIGURE 4. Boron concentration in FGD effluent at plant A. (a) Boron concentration in FGD effluent. (b) Boron content in coal. (c) Load of power generation. (d) Response slope and standard potential E0 for calibration of ISE. Figure 4(b) and (c). Three types of coal with different boron contents were used during the test period. It is apparent that the type of coal is a governing factor of the boron level in the effluent, but no proportional relationship was recognized between the boron concentration in the effluent and the VOL. 43, NO. 11, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4121
FIGURE 5. Relationship between values measured with the ICP-AES and the online boron monitor (at plant A). boron content in coal. Two reasons are envisaged for this. The total amount of boron released during the combustion is governed by the boron content, but its partitioning would be governed by other factors, i.e., the properties of coal, such as calcium content or alkalinity (9, 10). If the partitioning of boron largely differs owing to the chemical properties of coal, the boron concentration would not be proportional to the boron content in coal. In addition, if the FGD operation differs greatly depending on the properties of coal, i.e., sulfur content, this also should affect the boron behavior. The power generation was constant at 250 MW in full load operation during the test period, except from the 13-15th days, in which the plant was shut down and then restarted (Figure 4(c)). In response to this, a V-shaped profile was recognized in the boron concentration at around the 16th day. This indicates the constant operation of the FGD unit even during the shutdown period. Thus, the boron concentration in the effluent reflects changes in the type of coal and the load of power generation with a delay of 2-3 days. Such details of the boron behavior in the FGD effluent have never been revealed before because the measurement of boron concentration at a power plant is usually performed a few times a week at most. The boron concentration showed a relatively sluggish response with regard to time probably because of the large buffer tank for the effluent. In contrast, the process effluents with a small amount of or no buffer would show a rapid change in the boron concentration. In such cases, the proposed rapid determination method will offer more merits compared with the conventional ISE determination. The daily calibration values suggest a change in the ISE performance. The performance is estimated from the response slope and the standard potential E0, which are shown in Figure 4(d). The response slope is the difference in the electrode potential when measuring two concentrations of one order difference, and E0 is the potential of 1 mg/L BF4solution. The response slope increased linearly from -58 to -55 mV, whereas E0 decreased from 313 to 293 mV, both of which brought about a slight shift during the test period. Thus, no significant deterioration of the ISE membrane was observed. Figure 5 shows the values measured with the monitor and the ICP-AES. An excellent correlation (R2 ) 0.987) was obtained between the two methods, which demonstrates the effectiveness of the online boron monitor as a process monitoring tool. Plant B. The boron profile over a period of 4 months at plant B is shown in Figure 6. The boron concentration varied drastically in the range of 100-500 mg/L, which roughly reflects the boron content in coal. For the initial 2.5 months, a good correlation was obtained between the monitor and ICP-AES measurements. Thereafter, large discrepancies were 4122
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 11, 2009
FIGURE 6. Boron concentration in FGD effluent at plant B. (a) Boron concentration in FGD effluent. (b) Boron content in coal. (c) Load of power generation. (d) Response slope and standard potential E0 for calibration of ISE. observed for 2 weeks, which is depicted as A in Figure 6(a). The values measured with the monitor were 2-fold those measured with the ICP-AES. During this period, we observed no significant change in the sample composition, i.e., no ionic species such as ClO4-, I- and CN-, which are seriously interfering with the BF4- electrode (23). However, some foaming phenomena in the FGD unit were reported during this period. Thus we suspect that the discrepancies might be caused by bubbles becoming attached onto the ISE membrane. The air bubbles may come from the sample (effluent) or from the utility water in sample dilution. Since we did not see any perturbation in the calibration (Figure 6(d)), we infer that the air bubbles come from the effluent. We have experimentally confirmed that air bubbles on the membrane reduce the electrode potential, leading to higher boron concentrations. As a measure against the foaming, we modified the cell configuration by tilting the BF4- ISE that had been installed perpendicularly in the cell, so that air bubbles will have difficulty in attaching to the ISE membrane and also can be easily removed from there. This configuration has already been put into practice in the monitor that is installed at a different plant and no problem has arisen so far. Additionally, the use of antifoaming agents would be another solution for dissipating the foam. In this case, we must examine prior to its use that the agent results in no or little interference with the ISE. For the initial 3 months, the response slope of BF4- ISE increased linearly, but remained below -50 mV, as shown in Figure 6(d), which is a criterion for the exchange of the ISE membrane. Then, the slope fluctuated greatly and exceeded -50 mV. A similar behavior was also observed in E0. This suggests that the life of the ISE membrane is 3 months under the test conditions. The fluctuation of values measured
with the monitor during period B, shown in Figure 6(a), is affected by the life of the ISE membrane. The online boron monitor, a fully automated measurement device, was developed on the basis of the rapid potentiometric determination using a BF4- ISE, which measures boron compounds in the range of 1-500 mg/L in less than 20 min without the need for any pretreatment of the sample. In the demonstration tests at two power plants, the online boron monitor was successfully operated, and the detailed behavior of boron in the FGD effluent, which reflects the type of coal and the load of power generation, was revealed. The life of the ISE membrane was estimated to be 3 months from the test results. The online boron monitor was proved to be a useful tool in managing the boron emission in process effluent. In this study, we diluted an effluent and measured a higher boron concentration in the range of several hundred mg/L. With and without the sample dilution, the online boron monitor measures the boron concentration in the range of 1-10 mg/L. Thus, if interference by coexisting ions in the sample is negligible, it can also be applied to a municipal wastewater treatment process and a desalination process, in both of which boron should be controlled at around 1 mg/L.
Literature Cited (1) Cotton F. A.; Wilkinson, G.; Murillo C. A.; Bochmann, M. Advanced Inorganic Chemistry, 6th ed.; Wiley: New York, 1999. (2) World Health Organization, Boron, Environmental health criteria monograph 204; IPCS; Geneva1998 (http://www.inchem.org/ documents/ehc/ehc/ehc204.htm). (3) Guidelines for Drinking-Water Quality, Recomendations, 3rd ed.; World Health Organization: Geneva, 2004; Vol. 1. (4) EU Drinking Water Directive, 1998 (98/93/EC). (5) http://www.env.go.jp/en/water/wq/wp.pdf (accessed date: 2009/ 04/19). (6) National Effluent Standards. http://www.env.go.jp/en/water/ wq/nes.html (accessed date: 2009/04/19). (7) Clarke, L. B.; Sloss, L. L. Trace Elements - Emissions from Coal Combustion and Gasification, IEACR/49; IEA Coal Research: London, 1992. (8) Ito, S.; Yokoyama, T.; Asakura, K. Emissions of mercury and other trace elements from coal-fired power plants in Japan. Sci. Total Environ. 2006, 368, 397–402.
(9) Clemens, A. H.; Damiano, L. F.; Gong, D.; Matheson, T. W. Partitioning behaviour of some toxic volatile elements during stoker and fluidized bed combustion of alkaline sub-bituminous coal. Fuel 1999, 78, 1379–1385. (10) Clemens, A. H.; Deely, J. M.; Gong, D.; Moore, T. A.; Shearer, J. C. Partitioning behaviour of some toxic trace elements during coal combustionsthe influence of events occurring during the deposition stage. Fuel 2000, 79, 1781–1784. (11) Iwashita, A.; Sakaguchi, Y.; Nakajima, T.; Takanashi, H.; Ohki, A.; Kambara, S. Leaching characteristics of boron and selenium for various coal fly ashes. Fuel 2005, 84, 479–485. (12) Noda, N.; Ito, S. The release and behavior of mercury, selenium, and boron in coal combustion. Powder Technol. 2008, 180, 227– 231. (13) Carlson, R. M.; Paul, J. L. Potentiometric determination of boron as tetrafluoroborate. Anal. Chem. 1968, 40, 1292–1295. (14) Gulens, J.; Leeson, P. K. Direct ion-selective electrode determination of micromolar boron as tetrafluoroborate. Anal. Chem. 1980, 52, 2235–2237. (15) Lanza, P.; Mortera, G. Potentiometric determination of boron in silicon with an ion-selective electrode. Anal. Chim. Acta 1975, 75, 149–384. (16) Lanza, P.; Mortera, G. The determination of boron in waste waters with an ion-selective electrode. Ann. Chim. 1983, 73, 371–384. (17) Johnson, M. L.; Ward, W. G. Analysis of boric acid by tetrafluoroborate specific ion electrode. Met. Finish. 1988, 86, 49–51. (18) Imato, T.; Yoshizuka, T.; Ishibashi, N. Potentiometric flowinjection determination of boron by using a flow-through tetrafluoroborate ion-selective poly(vinyl chloride) membrane electrode. Anal. Chim. Acta 1990, 233, 139–141. (19) Imato, T.; Yoshizuka, T.; Ishibashi, N. Flow-injection determination of boron by using poly(vinyl chloride) membrane-based tetrafluoroborate ion-selective electrode. Bunseki Kagaku 1993, 42, 91–98, in Japanese. (20) Katagiri, J.; Yoshioka, T.; Mizoguchi, T. Basic study on the determination of total boron by conversion to tetrafluoroborate ion (BF4-) followed by ion chromatography. Anal. Chim. Acta 2006, 570, 65–72. (21) Ohsumi, H.; Ohyama, S.; Kudo, S.; Sakata, M. A simple and rapid determination method of boron in wastewater using ionselective electrode. J. Environ. Chem 2004, 14, 81–89, in Japanese. (22) Ohsumi, H.; Ohyama, S.; Kudo, S.; Sakata, M. Japanese Patent 4210146, 2008. (23) Tetrafluoroborate Ion Electrode, ION-LB08102; DKK-TOA Corp: Tokyo, 2007; Instruction Manual (in Japanese).
ES900062F
VOL. 43, NO. 11, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4123
