Advanced Oxidation Pretreatment for Biological Treatment of Reclaimer Wastewater Containing High Concentration N-methyldiethanolamine

Abstract

A wastewater treatment configuration consisting of advanced oxidation pretreatment and biological wastewater treatment process (BWTP) was investigated to treat a reclaimer wastewater generated in a steel-making industry, which contained high concentration MDEA (N-methyldiethanolamine) of up to 20,548 mg/L and other pollutants such as formate, phenol, and thiocyanate. The Fenton, ozone, and peroxone methods were tested as candidates, and the peroxone method was chosen because it could selectively decompose MDEA resulting in the final MDEA and chemical oxygen demand (COD) removal efficiencies of 92.87% and 27.16%, respectively. Through the respirometer tests using the sludge of the BWTP, it was identified that the microbial toxicity of the peroxone-pretreated wastewater was negligible and the short-term biochemical oxygen demand (BOD) to COD ratio, indicating that the biodegradability of wastewater significantly increased from 0.103 to 0.147 by the peroxone pretreatment. Analysis of the oxygen uptake rate profiles also revealed that the microbial degradation rate of the pollutants present in the reclaimer wastewater was in the order of phenol > formate > thiocyanate > MDEA, which could be changed depending on the microbial community structure of the BWTP.

1. Introduction

Alkanolamines encompassing monoethanolamine (MEA), diethanolamine (DEA), diisopropanolamine, triethanolamine (TEA), N-methyldiethanolamine (MDEA), etc. are effective absorbents to capture acid gas components, such as CO₂ and H₂S from natural or mixed gases [1,2,3,4,5,6,7]. Particularly, MDEA, a tertiary alkanolamine, has been most widely used in the field because it selectively reacts with CO₂ and H₂S [1,3,8]. However, all alkanolamines including MDEA should be properly treated prior to being discharged into the environment since they are classified as hazardous substances [8].

Petrochemical and coal pyrolysis plants are major point sources generating a large volume of acid gases, and many of them adopt an acid gas capture process using alkanolamines [2,3,5]. In the acid gas capture process, varied organic and inorganic anions, such as glycolate, acetate, formate, oxalate, chloride, nitrite, nitrate, sulfate, and so on, are produced through the reactions between alkanolamines and acid gases. These anions lead to the accumulation of heat stable salts (HSS) within the process, which causes amine loss as well as decrease of process efficiency during the operation. To remove HSS, reclaimer unit processes, such as distillation, ion exchange, and electrodialysis, can be incorporated into the acid gas capture process [6,9]. When using a reclaimer unit process, however, so-called reclaimer wastewater containing high concentrations of alkanolamines and other organic matters is generated, so its adequate treatment should be considered together.

In general, wastewater containing alkanolamines is very difficult to treat by biological method because of low biodegradability of alkanolamines [1,2,3,4,7,8]. As alternatives, advanced oxidation processes (AOPs), such as Fenton, ozone (O₃), and photocatalytic oxidation, have been explored and reported to be effective not only in removing alkanolamines, but also in improving the biodegradability of the wastewater [2,3,4,7,10,11,12,13,14]. Fenton oxidation is a method that can decompose refractory pollutants present in wastewater through a chain reaction of hydroxyl radical (∙OH), which is a powerful oxidizer produced by the reaction of ferrous ion (Fe²⁺) and hydrogen peroxide (H₂O₂). This method has been known to be able to decompose various alkanolamines including MEA, DEA, and MDEA [2,3,7]. Ozone oxidation, meanwhile, use a chain reaction of hydroxyl radicals generated by ozone decomposition. This method is known to have an advantage of selective degradation for some target pollutants compared to the Fenton method [15,16]. Although ozone has been widely used in the field for a long time to treat various kinds of recalcitrant pollutants, its application cases for alkanolamine degradation are not widely studied in the literature [3,11,12,13]. The photocatalytic oxidation method has also been successfully used for the treatment of alkanolamines [4,10,14]. This method, however, is not suitable for treating a large amount of wastewater due to the maintenance problem of UV lamps [17]. Some combinatorial innovative AOP schemes, such as photo-Fenton, electro-Fenton, and peroxone using H₂O₂ together with ozone, have also been developed for the treatment of refractory pollutants [18,19,20,21]. However, more case studies are needed for their practical application in the field.

AOPs have usually been used to treat the wastewater with low concentrations of refractory pollutants or to resolve turbidity issue at the final stage of wastewater treatment [20,22]. In the case of high concentration wastewater, however, it is more efficient to use AOPs as pretreatment method for subsequent biological treatment, considering their high facility investment and reagent costs [2,3,4,7,15,23]. Duran-Moreno et al. [3] investigated Fenton and ozone pretreatment of oil refinery wastewater containing DEA with the concept of sequential AOP and biological wastewater treatment process (BWTP). Jagadevan et al. [13] also reported ozone pretreatment of metalworking fluids’ wastewater containing MEA and TEA showing the biodegradation results of ozone-pretreated effluents. According to our literature survey, however, there is no report to deal with AOP pretreatment of MDEA with the scheme of AOP-combined BWTP. In addition, the effect of AOP-pretreated effluent on the performance of BWTP has been rarely explored despite its significance.

In this study, an AOP-combined BWTP scheme is investigated for the treatment of reclaimer wastewater generated in a steel-making industry. In this scheme, the reclaimer wastewater containing extraordinarily high concentration MDEA is first pretreated by AOP and then further treated in a BWTP currently treating cokes wastewater, which is the largest BWTP in the field of the steel-making industry [20,24]. Three AOPs, i.e., Fenton, ozone alone, and peroxone, were tested as candidate pretreatment methods, and the most proper one was chosen based on their performances and chemical reagent costs required. To assess the effect of AOP-pretreated wastewater on the BWTP, microbial toxicity, biodegradability, and pollutant biodegradation kinetics were also investigated using a respirometer system.

2. Materials and Methods

2.1. Wastewater Characteristics

Table 1. Characteristics of wastewaters.

Parameter	Reclaimer Wastewater	Cokes Wastewater
COD (mg/L)	44,721	3967
TN (mg-N/L)	3064	263
NH3 (mg-N/L)	227	47
MDEA (mg/L)	20,548	n.d. 1
Formate (mg/L)	20,420	n.d. 1
Phenol (mg/L)	841	1024
SCN− (mg/L)	2479	455
pH	10.0	9.0

¹ not detected.

shows the characteristics of the reclaimer and cokes wastewater used in this study, which were obtained from POSCO (Pohang Iron and Steel Company, Republic of Korea). In the given AOP-combined BWTP scheme, it is possible to use any BWTP that is operated in the field without serious modification if its capacity is enough to handle the AOP-pretreated effluent. Since the BWTP we intend to use is currently treating cokes wastewater, its characteristics are also important in understanding the nature of microorganisms present in the BWTP. The reclaimer wastewater contained high levels of chemical oxygen demand (COD) above 44 g/L, mainly originating from MDEA and formate. Considerable amounts of phenol and thiocyanate, both of which are the major pollutants of the cokes wastewater, were also present in the reclaimer wastewater.

2.2. Advanced Oxidation Pretreatment Tests

Fenton oxidation was conducted after adjusting the initial pH of the reclaimer wastewater to 3.0 using 1 N H₂SO₄ referring to the optimal initial pH range of 2.5~4.5 [2]. After completely dissolving 7.69 g of FeSO₄∙7H₂O in 700 mL of pH-adjusted reclaimer wastewater, 34.5 wt% H₂O₂ was continuously added at a flow rate of 2.34 mL/min using a peristaltic pump. The continuous H₂O₂ feeding scheme was adopted to avoid the side reaction of hydroxyl radical with H₂O₂, which could scavenge the hydroxyl radical available for MDEA degradation [25]. The Fenton reaction started with the addition of H₂O₂ and proceeded for 60 min under room temperature and 300 rpm mixing condition. To check the reaction extent, 10 mL samples were collected every 5 or 10 min and all samples were kept refrigerated after the following procedures: 4 mL of 1 N NaOH solution was added per 1 mL of the sample immediately after sampling to terminate the Fenton reaction [2]; the samples were heated at 70 °C for more than 30 min because residual H₂O₂ present in the sample could affect COD analysis [3]; the samples were centrifuged at 3000 rpm for 3 min and filtered using 0.45 μm syringe filter (6746-2504, Whatman, Maidstone, UK) to remove the precipitate generated by the reaction of Fe²⁺ and NaOH.

In the ozone oxidation test, a specially designed experimental apparatus consisting of reactor (a closed system, 700 mL working volume), demister, ozone generator (PC57-10, Ozonetch, Daejeon, Republic of Korea), ozone analyzer (OM-1500B, Ozonetech, Daejeon, Republic of Korea), and gas flow meter was used as shown in Figure 1. No initial pH adjustment of the reclaimer wastewater was made since the optimum pH range for ozone oxidation was reported as over 8 [15,16]. The ozone generator supplied a O₃/O₂ mixed gas containing 56.0 mg-O₃/L of ozone into the reactor at a flow rate of 1 L/min. The ozone oxidation reaction began with the addition of the reclaimer wastewater into the reactor and lasted for 480 min under room temperature and 300 rpm. The ozone concentration in the effluent gas of the reactor was monitored in real time using the ozone analyzer. Ozone consumption in the reaction was calculated from the difference between the ozone concentrations before and after the wastewater addition. To investigate the efficiency of ozone oxidation pretreatment, 15 mL samples were collected every 60 min, and all the samples were preserved in a refrigerator after filtering using 0.45 μm syringe filter until pending analysis.

In the peroxone oxidation test, the ozone-supplying condition was slightly changed and H₂O₂ was added together with ozone. An O₃/O₂ mixed gas containing 32.5 mg/L of ozone was supplied at a flow rate of 1 L/min, and 5 wt% H₂O₂ solution was continuously fed at a flow rate of 0.1 mL/min using a peristaltic pump. Other reaction conditions and sampling methods for peroxone test are the same as the ozone oxidation test.

2.3. Respirometer Experiments

The respirometer system was setup as shown in Figure 2a, which consists of bioreactor containing microbial sludge, air supplier, and an online dissolved oxygen (DO) meter (HQ40d, HACH, Loveland, CO, USA). This respirometer system can assess the toxicity, the biodegradability based on short-term biochemical oxygen demand (BOD), and the microbial pollutant degradation kinetics of wastewater using the DO profile or its relevant exogenous oxygen uptake rate ($OU{R}_{exo}$) profile, which is obtained by injecting a wastewater sample into the bioreactor. The sludge used in the respirometer tests was obtained from the biological cokes wastewater treatment process of POSCO (Pohang Iron and Steel Company, Pohang, Republic of Korea).

To estimate the toxic effect of the sample wastewater on the microorganisms, two additional reference wastewater injections are required before and after the sample wastewater injection (see, Figure 2b). Then, the toxicity of the sample wastewater can be calculated by comparing the peak height (PH) or peak area (PA) of the References 1 and 2 injections as follows [26,27,28]:

$$\% Toxicity=\frac{PH{\left(or PA\right)}_{reference 1}-PH{\left(or PA\right)}_{reference 2}}{PH{\left(or PA\right)}_{reference 1}}\times 100 .$$

(1)

In this study, the raw and AOP-treated reclaimer wastewaters were used as the sample wastewater, and the cokes wastewater was used as the reference wastewater. As such, 3 mL of sample or reference wastewater was injected for each toxicity test.

The short-term BOD of the sample wastewater ($BO{D}_{st}$) can be calculated from the $OU{R}_{exo}$ profile as follows [27,29]:

$$BO{D}_{st}={{\displaystyle \int }}_{t\_pulse}^{t\_final}OU{R}_{exo}\left(t\right)dt\times \frac{V_{Sludge}+V_{sample}}{V_{sample}} ,$$

(2)

where $OU{R}_{exo}\left(t\right)$ is the oxygen uptake rate at time t by the exogenous respiration of the microorganisms, $t\_pulse$ is the time of the sample wastewater injection, $t\_final$ is the time when $OU{R}_{exo}$ returns to zero (endogenous respiration phase), $V_{Sludge}$ is the sludge volume before the sample wastewater injection, and $V_{sample}$ is the volume of the injected sample wastewater. The $OU{R}_{exo}$ profile can be obtained from the DO profile using the following equation:

$$OU{R}_{exo}=k_{L}a\left(O_{2,sat}-O_2\right)-OU{R}_{endo}-\frac{d{O}_2}{dt} ,$$

(3)

where $k_{L}a$ is the volumetric mass transfer coefficient for DO, $O_{2,sat}$ is the saturation concentration of DO, $O_2$ is the DO concentration, and $OU{R}_{endo}$ is the oxygen uptake rate by the endogenous respiration.

For all respirometer tests, the same following operating conditions were applied: initial bioreactor sludge volume of 700 mL, initial MLSS concentration of 5000 mg/L, air flowrate of 2 L/min, agitation of 300 rpm, and room temperature. For each respirometer test, at least two air off/on procedures were also conducted to determine $OU{R}_{endo}$, $k_{L}a$, and $O_{2,sat}$, which are required for the $OU{R}_{exo}$ calculation [27,28].

2.4. Analysis

MDEA concentration was analyzed using high performance liquid chromatography (e2695, Waters Corporation, Milford, MA, USA) with a UV detector at 215 nm. YMC-pack polymer C18 (250 × 4.6 mm, S-6 μm) column (YMC Co., Ltd., Kyoto, Japan)was used, and the mixed solution of 100 mM Na₂HPO₄ and 100 mM NaOH (6:4 as volume ratio, pH = 12) was used as eluent at a flow rate of 0.6 mL/min. For the analysis of formate, ion chromatography (883 Basic IC Plus, Metrohm, Herisau, Switzerland) was used with conductivity detector. Metrosep A Supp 5-250/4.0 column (Metrohm, Herisau, Switzerland) was used with the mixed solution of 2 mM Na₂CO₃/2 mM NaHCO₃ as eluent with a flow rate of 0.7 mL/min. COD, total nitrogen (TN), NH₃, and phenol concentrations of the samples were determined using colorimetric reagent kits (C-MAC [30], Republic of Korea) with a spectrophotometer (Qvis 3000H+, C-MAC) at a wavelength of 620, 410, 425, and 510 nm, respectively. MLSS and thiocyanate concentrations were determined according to the standard methods for the examination of water and wastewater [31]. All reagents used in this study were ACS Grade Sigma-Aldrich (St. Louis, MO, USA) products.

3. Results and Discussion

3.1. Performance of AOP Pretreatment

Figure 3 represents the MDEA and COD concentration profiles during the Fenton, ozone, and peroxone oxidation tests. Among various AOPs, we selected these three methods as candidates for our AOP-combined BWTP considering their ease of use in full-scale application. In Figure 3, one can easily notice that the Fenton method decomposed MDEA most rapidly compared to the others, but its final MDEA removal efficiency was relatively low compared to the other methods. In the Fenton oxidation, the hydroxyl radical generated from the reaction of Fenton reagents (Fe²⁺ and H₂O₂) is the key oxidant responsible for the pollutant degradation, so the rapid MDEA degradation rate implies the strong reactivity of hydroxyl radical with MDEA. The low MDEA removal efficiency was the result of using rather mild conditions for the Fenton reaction (H₂O₂/MDEA = 13.8 mol/mol and H₂O₂/Fe²⁺ = 58.33 mol/mol). We adopted this condition referring to the report of Dutta et al. [7] with the intention of improving the biodegradability of the reclaimer wastewater via partial oxidation of MDEA. In this condition, however, it seems that the added amount of Fenton reagents was not sufficient for the complete removal of MDEA.

The final MDEA and COD removal efficiency of the Fenton method were 70.3% (from 20,548 mg/L to 6111 mg/L) and 47.0% (from 44,721 mg/L to 23,710 mg/L), respectively. The different extent of MDEA and COD reduction means that MDEA was not completely oxidized but just partially decomposed into other organic intermediates as we intended. Harimurti et al. [4] reported the MDEA degradation mechanism by hydroxyl radical, where some biodegradable intermediates, such as formic acid (CH₂O₂), oxalic acid (C₂H₂O₄), and acetic acid (CH₃COOH), are generated during the MDEA degradation. Other researchers who investigated the alkanolamine degradation using Fenton method also reported similar results [2,3,7]. As can be seen in

Table 2. Characteristics of AOP-pretreated reclaimer wastewaters.

Parameter	Fenton Pretreated	Ozone Pretreated	Peroxone Pretreated
COD (mg/L)	23,710	34,487	32,575
MDEA (mg/L)	6111	1179	1382
Formate (mg/L)	5479	20,228	17,787
Phenol (mg/L)	- 1	10.57	10.71
SCN− (mg/L)	- 1	- 1	50.4

¹ Not determined.

, a significant amount of formate was still present in the Fenton-pretreated wastewater. Other intermediates presumed to be DEA, MEA, and oxalate were also observed in the HPLC peaks obtained for the MDEA analysis, although their concentrations were not determined quantitatively (see, Figure S1).

In the ozone oxidation test, the MDEA concentration was gradually decreased to 1179 mg/L, resulting in the final removal efficiency of 94.3%. COD was also reduced accordingly, but its removal efficiency was only 22.9%. It should be noted that the difference between the MDEA and COD reduction extent was much more remarkable in the ozone oxidation compared to the Fenton oxidation, which implies that the ozone method is more suitable for the selective degradation of MDEA. This can be also confirmed from the fact that formate, one of the major intermediates of MDEA degradation, remained at a level comparable to its initial concentration even after the ozone pretreatment (see Table 2). In the ozone oxidation, not only hydroxyl radical formed from ozone decomposition, but also ozone itself can act as a strong oxidant. According to von Gunten [15], ozone can react selectively with non-protonated amine. Kishore et al. [32] and Zahardis et al. [33] also reported that ozone had a high selectivity for the degradation of varied amine compounds. Hydroxyl radical, on the other hand, is known to be less selective than ozone despite its higher reactivity [15]. In our ozone oxidation test, it seems that most MDEA was degraded by ozone, while its degradation by hydroxyl radical was restricted, considering the report that hydroxyl radical generated from ozone could be rapidly scavenged by other organic and inorganic constituents of wastewater resulting in a nearly diffusion-controlled reaction scheme [34]. Reminding that the aim of introducing the AOP is not the complete degradation of MDEA but its partial oxidation, the ozone method seems to be more suitable for the given AOP-combined BWTP scheme than the Fenton method.

Although the ozone oxidation method was proved to have the advantage of selective MDEA degradation, it still had a problem of requiring a rather large amount of ozone supply. The peroxone method using H₂O₂ together with ozone could be a solution to reduce the ozone requirement, so its performance was also investigated. As can be seen from Figure 3, the peroxone method exhibited similar MDEA and COD removal trends comparable to the ozone method. The final MDEA and COD removal efficiencies were also compatible, showing 93.3% (from 20,548 mg/L to 1382 mg/L) and 27.2% (from 44,721 mg/L to 32,575 mg/L). These final MDEA and COD removal efficiencies were achieved with a greatly reduced ozone supply amount. In the ozone method, the amount of ozone consumed for the MDEA degradation was estimated to be 0.499 g-O₃/g-MDEA (see, Figure 4). In the peroxone method, on the other hand, it was reduced to 0.323 g-O₃/g-MDEA, indicating that the peroxone method could significantly reduce the ozone requirement for the MDEA degradation. However, it should be noted that the use of H₂O₂ in the peroxone method could negatively affect the selective degradation of MDEA because the diffusion-controlled reaction scheme described above could be deteriorated by the excessive generation of hydroxyl radical form H₂O₂. Indeed, the formate concentration in the peroxone-pretreated wastewater was slightly lowered than ozone oxidation case (see Table 2). Hence, the amount of H₂O₂ addition would be a critical operating condition for the peroxone method, which is indeed a trade-off problem between the benefit of saving the ozone requirement and the loss of the selectivity in the MDEA degradation. With the operating condition given in this study, the benefit was more significant and the selective oxidation of MDEA was still possible.

In this study, the operating conditions for each pretreatment method were not optimized in detail since the main purpose was just confined to the screening of an adequate AOP method applicable to the AOP-combined BWTP scheme adopted for the treatment of the reclaimer wastewater. However, the performance results of the AOP pretreatment tests clearly supported that the peroxone method would be the most proper one. Moreover, the benefit of using peroxone method was more obvious if comparing the chemical reagent costs required for the MDEA degradation (see Table S1). Therefore, we finally selected it as the AOP method for our AOP-combined BWTP scheme.

3.2. Effect of Peroxone-Pretreated Wastewater on BWTP

When using AOP as pretreatment method for BWTP, the effect of AOP-pretreated wastewater on the microorganisms in BWTP should be carefully examined. The respirometer system adopted in this study is very useful for this purpose because it can rapidly determine the toxicity, biodegradability, and even pollutant biodegradation kinetics of the concerned wastewater [26,27,28,29,35,36].

First, we investigated the toxicity of the raw and peroxone-pretreated reclaimer wastewater on the microbial sludge of the BWTP treating the cokes wastewater. The PH and PA toxicity of the raw reclaimer wastewater were determined to be 10.61% and 20.04%, respectively. These rather high toxicities imply that direct feeding of the raw reclaimer wastewater could adversely affect the performance of the BWTP. On the other hand, it was revealed that the PH and PA toxicities of the peroxone-pretreated wastewater were only 1.25% and −1.06%, respectively, which means that the peroxone method could be effectively combined with the BWTP without the concerns of inhibition effect. One can find the toxicity results for the Fenton and ozone pretreatments in Figure S2.

Figure 5a shows the $OU{R}_{exo}$ profiles for the cokes, raw reclaimer, and peroxone-pretreated wastewaters, from which the information about the biodegradability and the pollutant biodegradation kinetics of the wastewaters can be obtained. The $OU{R}_{exo}$ profiles for other AOP methods are also given in Figure S3 for comparison. The short-term BOD ($BO{D}_{st})$ of the cokes wastewater was calculated to be 1200 mg/L, and its consequent $BO{D}_{st}$/COD ratio, which could represent its biodegradability, was determined to be 0.302. This $BO{D}_{st}$/COD value should be the reference when comparing the biodegradability of other wastewaters, since the microbial sludge used in the respirometer was obtained from the BWTP treating the cokes wastewater. The $BO{D}_{st}$/COD ratio of the raw reclaimer wastewater was 0.103, which implies that its biodegradability was much lower than the cokes wastewater. This might be due to the presence of high concentration of MDEA in the wastewater. In a separate respirometer test using pure MDEA solution, it was proved that the microbial sludge of the BWTP treating the cokes wastewater can hardly degrade MDEA (see Figure 5b). The estimated $BO{D}_{st}$/COD ratio of pure MDEA was only 0.004. The low biodegradability of the reclaimer wastewater could be remarkably improved by peroxone pretreatment up to the level corresponding to the $BO{D}_{st}$/COD ratio of 0.147. The partial oxidation of MDEA by peroxone pretreatment seems to be responsible for this improvement of biodegradability.

The $OU{R}_{exo}$ profiles shown in Figure 5a can also be used to investigate the microbial degradation kinetics of the pollutants present in the wastewaters. With an appropriate kinetic model like the Monod equation, the involved model parameters can be estimated by numerical regression of $OU{R}_{exo}$ profile [35,36]. In this study, however, we just focused on the qualitative analysis of the $OU{R}_{exo}$ profile because there were so many pollutants in the wastewater, and thus, the consequent mathematical model system was extremely complicated. However, it should be noted that in the respirometer test it was basically assumed that the observed $OU{R}_{exo}$ profile is the sum of the individual $OU{R}_{exo}$ profile, which results from the biodegradation of the distinct pollutant in the wastewater [27,35]. Therefore, we conducted additional respirometer tests for the major pollutants using pure chemicals, which would provide some clues to be able to interpret the observed $OU{R}_{exo}$ profile (see Figure 5b).

As can be seen in Figure 5a, the $OU{R}_{exo}$ of the raw reclaimer wastewater rapidly increased immediately after the sample injection. This rapid increase observed in the initial phase is believed to be most deeply related with the phenol degradation, which can be confirmed from the $OU{R}_{exo}$ profile of pure phenol shown in Figure 5b. A similar rapid $OU{R}_{exo}$ increase can also be observed in the test using the cokes wastewater, of which the major pollutant was phenol itself. Although an excessive amount of MDEA (20,548 mg/L) also existed in the reclaimer wastewater, its contribution to the $OU{R}_{exo}$ would be trivial considering its extremely low biodegradability. The $OU{R}_{exo}$ of the peroxone-pretreated wastewater also increased soon after the sample injection. However, its increasing rate was clearly lower than the other cases, implying that the initial $OU{R}_{exo}$ increase of the peroxone-pretreated wastewater was more strongly related with the microbial degradation of other pollutant rather than phenol. The most probable pollutant was formate as can be seen in Figure 5b. In fact, formate became the most abundant pollutant (17,787 mg/L) in the wastewater after the peroxone pretreatment due to the result of partial oxidation of MDEA. In addition, the phenol and thiocyanate concentrations of the peroxone-pretreated wastewater were only 10.7 and 50.4 mg/L, respectively (see Table 2), which means that their contribution to the initial $OU{R}_{exo}$ increase would be minor.

The respirometer tests using pure chemicals revealed that the biodegradation rates of the involved pollutants were in the order of phenol > formate > thiocyanate > MDEA. It is interesting that formate, a well-known readily biodegradable compound, was degraded more slowly than phenol. This might be because the microbial sludge used in the respirometer tests had been acclimated to the cokes wastewater for a long time so that phenol-degrading microorganisms should be dominant in the sludge. Though the slow biodegradation rate of formate can be a temporary problem during the startup period of the proposed AOP-combined BWTP, we believe that it will be naturally resolved along with continued operation since the microbial community distribution within the sludge will be eventually optimized to the peroxone-pretreated wastewater through a new sludge acclimation process.

4. Conclusions

Treatment of an industrial wastewater containing high concentrations of refractory pollutants has always been a challenging problem in the field of wastewater treatment. In this study, an AOP-combined BWTP scheme was investigated for the treatment of the reclaimer wastewater containing a high concentration of MDEA and other organic pollutants such as formate, phenol, and thiocyanate. Although there were several choices of AOP applicable to the scheme, we concluded that the peroxone pretreatment using ozone and H₂O₂ would be the best, because it could selectively decompose MDEA into biodegradable compounds with a significantly reduced ozone requirement. When adopting AOP as a pretreatment step for the subsequent BWTP, its effect on the microorganisms in the BWTP should also be carefully examined. Some criteria such as microbial toxicity, biodegradability, and pollutant biodegradation kinetics could be imperative, and the respirometer test was very effective to evaluate these criteria. Although the detailed operating conditions should be further optimized in the future, we believe that the given AOP-combined BWTP scheme could be the feasible solution for the successful treatment of the reclaimer wastewater.