Analytical Work

The chemical oxygen demand (COD), pH, biochemical oxygen demand (BOD5), electrical conductivity (EC), total suspended solids (TSS), ammoniacal nitrogen (NH3-N), and heavy metals (magnesium (Mg<sup>2</sup>+), calcium (Ca<sup>2</sup>+), iron (Fe<sup>2</sup>+), zinc (Zn<sup>2</sup>+), copper (Cu<sup>2</sup>+), chromium (Cr<sup>2</sup>+), cadmium (Cd<sup>2</sup>+), lead (Pb<sup>2</sup>+), arsenic (As<sup>3</sup>+), cobalt (Co<sup>2</sup>+), and manganese (Mn<sup>2</sup>+)) were tested before and after each aeration run. The level of BOD5 was estimated utilizing Method 5210B. The DO was tested utilizing a DO meter (model 1000, YSI Inc., Greene County, OH, USA). COD concentration was tested utilizing the closed reflux colorimetric method (5220B—DR2500 HACH, Loveland, CO, USA). Color was determined using the DR 2800 HACH spectrophotometer at 455 nm wavelength.

A portable digital pH/mV meter (model inoLab pH 720, WTW, Weilheim, Germany) was used to measure the pH and EC. TSS was determined utilizing method 2540D, dried at 103–105 ◦C, which included the following procedure: preparation of filter disc, selection of filter type and sample volume, analysis of samples, and calculation of Equation (4) (APHA 2012) [31].

$$\text{TSS mg/L} = \text{I}(\text{A} - \text{B}) \text{ mg/V} \,\text{mL},\tag{4}$$

where A is the weight of filter-dried residue (mg), B is the weight of the filter, and V is the sample volume.

NH3-N level was determined using the phenate method (4500-NH3 F) utilizing a DR2500 spectrophotometer at 640 nm. Heavy metals were measured using atomic absorption spectroscopy (Unicam 929 AA Spectrophotometer, UNICO, Franksville, WI, USA). All physico-chemical parameters and heavy metals were measured according to standard methods for examining water and wastewater [32]. Di fferent leachate/POME ratios (1:0, 0.9:0.1, 0.7:0.3, and 0.5:0.50) were prepared in 1000-mL bottles to investigate the removal e fficiency of targeted parameters.

### **3. Results and Discussion**

### *3.1. E*ff*ects of Aeration Time Variation on the Removal E*ffi*ciency during the Aeration Process of Leachate Treatment*

The maximum removal e fficiencies in only leachate aeration for COD, TSS, color, and NH3-N reached 44.12%, 43.5%, 55%, and 97%. Figure 1 shows the removal e fficiency influenced by the time of reaction for the targeted parameters. The removal e fficiencies of both COD and TSS increased slightly in the same way. However, the increased color removal e fficiency was characterized by fluctuation, as shown on Figure 1, while the removal e fficiency of NH3-N increased sharply in the first week of aeration, reaching 93% on the seventh day; it did not demonstrate a further marked increase in removal efficiency, reaching 97% by the 24th day. This is in line with many studies reporting that NH3-N can be removed under the e ffect of the gas stripping process during aeration [21], while the removal is significantly improved with aeration time increasing up to a specific point [33] due to the reaction of NH3 with water, as shown in Equation (5).

$$\mathrm{NH\_3 + H\_2O \leftrightarrow NH\_4 \,^+ + OH^-}.\tag{5}$$

*Processes* **2020**, *8*, 601

From Equation (5), increasing the pH due to the formation of OH− will increase the concentration of NH3. The increase in pH enhances the ammonia stripping during aeration [21] and enhances the removal of ammonia, after which, owing to the recarbonation of lime in leachate, the pH begins to decline by absorbing CO2 from the ambient air [34].

**Figure 1.** Effects of reaction time variation on COD, TSS, color, and NH3-N removal efficiency (with natural pH, leachate only, aeration 20 L/min).

*3.2. E*ff*ects of Reaction Time Variation on the Removal E*ffi*ciency during the Aeration Time of Leachate*/*POME Treatment (Ratio 900 mL Leachate*/*100 mL POME)*

In this stage, the effect of the reaction time on leachate with a ratio of 900 mL leachate/100 mL POME was investigated during the aeration time. The optimum reaction time was reached on the 24th day of aeration, and the maximum removal efficiencies for COD, TSS, color, and NH3-N reached 91%, 54%, 50%, and 98%, respectively. All targeted parameters increased as the aeration time increased (Figure 2).

**Figure 2.** Effects of reaction time variation on COD, TSS, color, and NH3-N removal efficiency (with natural pH, leachate/POME (900 mL leachate/100 mL POME), aeration power 20 L/min).

### *3.3. E*ff*ects of Reaction Time Variation on the Removal E*ffi*ciency during the Aeration Time of Leachate*/*POME Treatment (Ratio 700 mL Leachate*/*300 mL POME)*

In this stage, the effects of reaction time variation on the removal efficiency during the aeration time of leachate/POME treatment with a constant ratio (700 mL leachate/300 mL POME) were investigated for the targeted parameters under the same condition of natural pH and an aeration power of 20 mL/min. The results showed an increase in the removal efficiency of all targeted parameters. The maximum removal of NH3-N, COD, TSS, and color was 96%, 89%, 53%, and 41%, respectively. As shown in Figure 3, there was a higher removal efficiency for NH3-N than COD, but color showed the lowest removal efficiency.

**Figure 3.** Effects of reaction time variation on COD, TSS, color, and NH3-N removal efficiency (with natural pH, leachate/POME (700 mL leachate/300 mL POME), aeration power 20L/min).

### *3.4. E*ff*ects of Reaction Time Variation on the Removal E*ffi*ciency during the Aeration Time of Leachate*/*POME Treatment (Ratio 500 mL Leachate*/*500 mL POME)*

As shown in Figure 4, the ratio between leachate and POME was 500 mL/500 mL. The effect of reaction time variation on the removal efficiency during the aeration time for targeted parameters was investigated with the same conditions for the other ratios (natural pH, aeration power 20 L/min). The maximum removal efficiencies for COD, TSS, color, and NH3-N were 89%, 21%, 42%, and 94%. The removal efficiency of NH3-N increased sharply in the first 10 days of aeration, reaching 90%, and it did not demonstrate a further marked increase in removal efficiency, reaching 94% by the 24th day.

**Figure 4.** Effects of reaction time variation on COD, TSS, color, and NH3-N removal e fficiency (with natural pH, leachate/POME (500 mL leachate/500 mL POME), aeration power 20 L/min).

### *3.5. E*ff*ects of the Leachate*/*POME Mixing Ratio on the Leachate Aeration Process*

The treatment of leachate was implemented using POME in four ratios (leachate only, 900 mL leachate/100 mL POME, 700 mL leachate/300 mL POME, and 500 mL leachate/500 mL POME) with a natural pH and an aeration power of 20 L/min. Accordingly, a limited removal e fficiency of COD (43%) was found when treating leachate without any POME dosages. As shown in Figure 5, adding 100 mL, 300 mL, and 500 mL of POME improved the removal e fficiency of COD (COD removal of 85%, 88%, and 88%, respectively).

The results of the removal e fficiencies for the targeted parameters using several dosages (leachate only, 900 mL leachate/100 mL POME, 700 mL leachate/300 mL POME, and 500 mL leachate/500 mL POME) are illustrated in Figure 5. The maximum removal e fficiency was NH3-N, reaching 97% during the aeration process for leachate only and 900 mL leachate/100 mL POME, while the maximum removal for COD reached 88% (700 mL leachate/300 mL POME and 500 mL leachate/500 mL POME ratios). The lowest removal was 51% for TSS, followed by 52% for color for the 700 mL leachate/300 mL POME ratio.

**Figure 5.** Effect of the leachate/POME ratios on COD, TSS, color, and NH3-N removal e fficiency after 24 days of mixing aeration (with natural pH, aeration power 20 L/min).

### *3.6. Analysis of Variance*

One of the key objectives of the RSM is to calculate the best value for the control variables that can maximize or minimize a response over a particular area of concern. A good-fitting model is defined to provide a proper representation of the mean response to achieve the optimum value [34]. A total of 13 experiments with di fferent dosages (leachate/POME) and reaction times at room temperature (28 ◦C) were performed using central composite design (CCD); the outcomes were analyzed using analysis of variance (ANOVA), as shown in Table 3. The POME dosage used to treat stabilized leachate was evaluated in terms of its e ffectiveness in the removal of COD, TSS, color, and NH3-N.

The coe fficient of determination ( *R<sup>2</sup>*) for COD, TSS, color, and NH3-N was 0.9927, 0.8218, 0.9854, and 0.991, respectively. This mean that the models are good in Equations (6)–(9) due to the coe fficient of determination being high and close to 1 (Table 4). Prior to the data analysis, the assumption of normality should be tested. The assumption of normality showed that the data roughly fit a bell-shaped curve for all responses, as presented in Figure 6.

$$\text{COD removal} = +80.18 + 25.81 \times \text{A} + 8.60 \times \text{B} - 19.97 \times \text{A}^2 - 4.00 \times \text{B}^2 - 0.98 \times \text{A} \times \text{B};\tag{6}$$

$$\text{TSS removal} = +49.93 + 4.32 \times \text{A} + 15.76 \times \text{B} - 6.39 \times \text{A}^2 - 0.089 \times \text{B}^2 + 1.02 \times \text{A} \times \text{B};\tag{7}$$

$$\text{Color removal} = +45.74 + 1.35 \times \text{A} + 16.14 \times \text{B} - 17.50 \times \text{A}^2 - 7.62 \times \text{B}^2 + 1.20 \times \text{A} \times \text{B};\tag{8}$$

$$\text{NHI}\_3\text{-N removal} = +97.24 - 3.05 \times \text{A} + 6.72 \times \text{B} - 1.06 \times \text{A}^2 - 5.17 \times \text{B}^2 + 1.92 \times \text{A} \times \text{B}.\tag{9}$$



### *Processes* **2020**, *8*, 601

**Figure 6.** Normal probability plots for (**I**) COD, (**II**) TSS, (**III**) color, and (**IV**) NH3-N removal.

The two variables, POME dosage and aeration time, displayed an important impact (*p*-value < 0.05) for the linear and quadratic models on the targeted factors of COD, color, and NH3-N, as shown in Table 4 for the ANOVA results. On the other hand, a higher-order model like a third-order polynomial model or more complicated model can be used to improve the model for TSS removal.




**Table 4.** *Cont*.

Abbreviation. DF: degrees of freedom; Cor: corrected; CV: coefficient of variation; Pred: predicted; Adeq: adequate.

In addition, the *p*-values for the interaction effect were 0.4607, 0.7681, and 0.1294 for the removal efficiency of COD, TSS, and color. In other words, the interaction effect between POME dosage and aeration time was insignificant (*p*-value > 0.05) for the removal efficiency of COD, TSS, and color. That implies that the two factors function independently. Conversely, the *p*-value (0.0013) for the interaction effect was significant for NH3-N removal efficiency. Figure 7 shows the interaction between the two (variables) factors (POME dosage and aeration time) and their behaviors in terms of removal of the targeted parameters (COD, TSS, color, and NH3-N). The *p*-values for lack of fit were 0.8314, 0.0523, and 0.0647, which indicates that the lack of fit was insignificant (*p*-value > 0.05), which means that the model is appropriate for the removal efficiency of COD, color, and NH3-N, while the *p*-value for lack of fit was significant (*p*-value < 0.05) for TSS removal efficiency, indicating the model is not appropriate.

The impact of POME dosage and aeration time on the selected responses is illustrated in Figure 8. The descriptions for the behavior of each response for POME dosage and aeration time are shown as the surface of a three-dimensional plot for the maximization of the four targeted responses (COD, TSS, color, and NH3-N) (Figure 8). All response plots demonstrate clear peaks, suggesting that the maximum area of impact is well known with the selected boundaries of the POME dosage and aeration time.

**Figure 7.** The impact of the POME dose combination and the aeration time for the removal of (**I**) COD, (**II**) TSS, (**III**) color, and (**IV**) NH3-N.

**Figure 8.** Response surface plot for (**I**) COD, (**II**) TSS, (**III**) color, and (**IV**) NH3-N removal.

### *3.7. Optimization of Leachate Treatment Using POME*

Design-Expert 6.0.7 software offers strong tools for setting up an optimal experiment of the treatment process to identify the optimum value of removal efficiency for COD, TSS, color, and NH3-N. In accordance with the approach of software optimization, the required target was within the range for each experimental condition (POME dosage and aeration time). To obtain the highest output, the responses (COD, TSS, color, and NH3-N) were described as a maximum value. The program incorporates individual desirability into a single number and then searches on the basis of the response target to optimize this feature. The optimum conditions and respective percentage removal efficiencies were established, and the COD (89.83%), TSS (66.7%), color (91.7%), and NH3-N (94%) removal results are illustrated in Table 5 The desirability function for these optimum conditions was recorded as 0.935. Additional experiments under optimal conditions were performed to verify agreemen<sup>t</sup> with the outcome experiments from models and the experiments. The results from the laboratory experiment were 87.51%, 65.62%, 53.10%, and 91.8% for the removal of COD, TSS, color, and NH3-N, respectively (Table 5). There was close agreemen<sup>t</sup> between the removal efficiencies for all response parameters gained from the experiments and those estimated by models. These results are more efficient than the result of Banch et al. [8], who conducted tests for the same parameters except for color. Tatsi et al. [35] achieved a color removal of about 100% for partially stabilized leachate. However, the reported residuals for COD, TSS, color, and NH3-N were recorded as 430 mg/L, 1620 mg/L, 1780 Pt-Co, and 66 mg/L, respectively, which are still higher than the effluent discharge limits.


**Table 5.** Optimal response results from the model prediction and laboratory.

### *3.8. Heavy Metal Analysis*

The removal efficiency of targeted heavy metals from stabilized leachate was evaluated under the achieved optimum experimental conditions, and the results are given in Table 6. The removal efficiency for the targeted heavy metals ranged between 99.00% (Cd<sup>2</sup>+) and 6.85% (As<sup>3</sup>+). The order of residuals of heavy metals in leachate was Fe2<sup>+</sup> > Mn2<sup>+</sup> > As3+ > Zn2<sup>+</sup> > Cr2+ > Cu2+ > Co2+ > Pb+ > Cd2+ from highest to lowest concentration. The heavy metal residuals were 1856, 31.72, 16.15, 10.76, 2.80, 1.28, 0.53, 0.28, and 0.23 μg/<sup>L</sup> for Fe2+, Mn2+, As3+, Zn2+, Cr2+, Cu2+, Co2+, Pb<sup>+</sup>, and Cd2+, respectively, while the removal efficiency was 90.73%, 34.75%, 6.85%, 96.16%, 93.78%, 96.95%, 95.24%, 93.30%, and 99.00% respectively. Only the concentration of Mn2<sup>+</sup> existed out of the limits, while the other targeted heavy metals were within the limits. The removal of heavy metals is attributed to the high level of suspended solids and metal complexes in POME which may act as a natural coagulant. The suspended solids and metal complexes improve the charged exchange and accelerated the adsorption and deposition of dissolved heavy metals in wastewater [36–40]. This study showed more efficient removal than Banch et al. [8] for Fe2+, As3+, Zn2+, Cr2+, Cu2+, Co2+, and Cd2+.

**Table 6.** Effect of the POME dosage and aeration time on heavy metal removal (POME dosage 188.32, aeration time 21 days).

