https://orcid.org/0000-0002-6211-2142
Abstract
A batch monopolar electrocoagulation system was developed and studied for the removal of phosphorus from anaerobic bioreactor effluent using iron as an electrode material. The study focused on the optimization of the independent variables, such as initial pH, retention time (RT), current density (CD) and inter-electrode distance (IED) using the response surface methodology (RSM) to maximize the removal of total phosphorus (TP). A quadratic model was fitted to the experimental data for TP removal. The optimal parameters were found to be pH of 6.75, RT of 11.06 min, CD of 300 A/m2, and inter-electrode distance of 1.5 cm resulting in 98.05% TP removal and energy consumption of 1.28 kWh/m3. A kinetic study for TP removal revealed that at optimal conditions, removal followed first-order kinetics (K = 0.185 m/min). Phosphorus was recovered from the post-precipitated sludge through combustion at 900 °C followed by acid leaching with sulfuric acid. Acid leaching tests were carried out with sulfuric acid for the post-precipitated sludge obtained at the optimum conditions. It resulted in around 91% of phosphorus recovery at a liquid-to-solid ratio of 100 mL/g.
HIGHLIGHTS
Effect of calcium in iron-electrocoagulation system.
Statistical optimization of parameters to maximize total phosphorus removal from anaerobic effluent.
Consideration of interactive effects of independent variables in process optimization.
INTRODUCTION
Global phosphorus consumption has gradually increased from 43.7 million tonnes in 2015 to 48.2 million tonnes in 2019 due to the growing demand for food production (Hermann et al. 2018). The agricultural sector accounts for 85% of phosphorous consumption (Johnston et al. 2014). As phosphate is a limited and nonrenewable resource, a rapid increase in phosphate consumption can put enormous strain on global phosphate rock reserves. It is estimated that the worldwide phosphate reserves will be depleted in 10 decades if the current rate of phosphate mining continues and no alternative measures are adopted (Gilbert 2009). To deal with the scarcity of phosphorus, it is crucial to identify alternate phosphorus sources. Phosphorus levels in typical anaerobically treated wastewater effluents and digestates are generally high (>40 mg/L) (Moerman et al. 2009; Huang et al. 2017). The recovery of phosphate from these waste streams is an appealing option for achieving long-term phosphorus supplies which otherwise would be discharged into the water environment causing adverse impacts. It has the potential to supply up to one-fifth of worldwide phosphorus demand (Yuan et al. 2012).
Agricultural runoff is the major contributor of phosphorus in water bodies. The presence of phosphorus above 0.02 mg/L in the water bodies causes eutrophication, which interferes with the reproduction of algae and microorganisms (Wan et al. 2020). This leads to the depletion of dissolved oxygen and generates toxins in the water, resulting in the death of aquatic organisms and causing damage to wildlife (Conley et al. 2009; Li et al. 2016) and can become a public health concern. The presence of phosphorus in the waste stream encourages the fouling of water pipes, resulting in costly maintenance (Attour et al. 2014). Therefore, removal or possible recovery of phosphate from wastewater discharges is crucial to maintain a balance of phosphorus in the aquatic environment, keeping in mind the aspects of sustainable development, and resource recovery.
Conventional methods such as chemical and biological phosphorus recovery methods are widely used industrially and can achieve a low phosphorus concentration (<1 mg/L) in the treated effluent (Yin et al. 2015; Wang et al. 2018; Li et al. 2019). However, chemical precipitation needs large amounts of chemical additives and is associated with high sludge production, whereas biological phosphorus removal is energy-intensive and complex in nature (Sengupta et al. 2015). Furthermore, the phosphorus content of the sludge produced by biological phosphorus removal is extremely low (2–3%) (Jupp et al. 2021). These shortcomings sparked the interest of researchers in exploring other alternatives for sustainable phosphorus recovery from anaerobic effluent. Recent studies on phosphorus removal and recovery using an electrochemical membrane (Kekre et al. 2021), electrochemical precipitation (Bakshi et al. 2020; Govindan et al. 2021; Li et al. 2021), and microbial fuel cell (Ye et al. 2019) are found to be promising. Among these methods, electrochemical precipitation (electrocoagulation) has gained considerable research interest due to its ease of use and high removal efficiency (up to 99%) (Lacasa et al. 2011).
Several studies on phosphorus removal using an electrocoagulation system have been published, in the majority of the studies, synthetic wastewater was used, which avoids the effect of organics and other constituent ions on phosphorus removal (Gharibi et al. 2010; Zhang et al. 2018; Bakshi et al. 2020; Zeng et al. 2021). To the best of our knowledge, the interaction of parameters such as pH and current density (CD), CD and retention time (RT) were not taken into account in studies using anaerobic effluents from industries (Tran et al. 2012; Huang et al. 2017). These parameters and their interactions might affect the mass of metal ion generation from the anode. Furthermore, consideration of the interactive effects in process optimization could decrease energy consumption, thereby lowering operational costs. Therefore, further research is warranted for addressing these gaps in the literature. This will help to achieve the desired phosphorus concentration in the effluent to meet discharge guidelines and can serve as a sustainable phosphorus source to meet global phosphorus demand while protecting the environment.
Mechanism of electrocoagulation.
The dominant reactions for iron as an electrode material are as follows (Linares-Hernández et al. 2010; Bernal-Martínez et al. 2013):
Phosphorus present in electrocoagulation sludge is bound to metal ions and requires additional processing for phosphorus extraction (Kyle & McClintock 1995). Available methods of phosphorus extraction from phosphorus-rich sludge include biological leaching (Mehta et al. 2015), hydroxyapatite precipitation (Lee et al. 2018), and leaching (Petzet et al. 2012; Atienza-Martínez et al. 2014; Damaraju et al. 2019; Monea et al. 2020). Among the various available methods of phosphorus recovery, leaching is considered economical. The combustion of sludge enhances the availability of phosphorus for recovery. The melting temperature of iron-rich sewage sludge ash ranges between 900 and 1,160 °C (Shao et al. 2007; Wang et al. 2012). Combustion of post-precipitated electrocoagulation sludge at a melting temperature of 900 °C results in the formation of acid-soluble phosphorus and acid-insoluble iron compounds like iron oxides (Fe2O3) from iron phosphate (Damaraju et al. 2019).
In this study, a batch monopolar iron-electrocoagulation reactor was developed to investigate the feasibility of phosphorus removal from anaerobic bioreactor effluent. The objectives of this study were to evaluate the effect of process variables (pH, RT, CD, inter-electrode distance (IED)) and their interaction on TP removal, to optimize the parameters for maximizing the TP removal efficiency, to develop a quadratic model for the response prediction using response surface methodology (RSM), and to evaluate the effect of liquid-to-solid (L/S) ratios on phosphorus extraction by acid leaching. Experimental runs were conducted at the optimized run condition and the observed phosphorus removal efficiency was compared with the predicted removal efficiency. The kinetics of TP removal was investigated to determine the rate of phosphorus removal. Acid leaching was used for the sludge obtained from the electrocoagulation system at optimized parameters to recover phosphorus.
MATERIALS AND METHODS
Sampling of the wastewater
Anaerobic bioreactor effluent was collected from an industrial wastewater treatment plant (McCain Foods, Florenceville-Bristol, Canada). The WWTP uses a large covered anaerobic lagoon for the treatment of its wastewater and to generate biogas. The collected sample was stored under 2 °C to avoid degradation of the organic matter present in it. The collected anaerobic bioreactor effluent was homogenized for initial characterization.
Experimental setup and procedure
Schematic of the EC batch reactor setup.
The physico-chemical characteristics of the anaerobic bioreactor effluent are shown in Table 1. The analysis to characterize the effluent was performed in triplicate. Determination of pH, conductivity, total phosphorus (TP), total dissolved solids (TDS), total suspended solids (TSS), calcium, magnesium, iron, and filtered chemical oxygen demand (fCOD) was conducted according to Standard Methods (APHA AWWA & WEF 2005). TP, calcium (Ca2+), magnesium (Mg2+), iron (Fe), and sodium (Na+) were measured using ICP-OES (VISTA-MPX Axial, Varian, Australia). The orthophosphate concentration was measured using UV spectrophotometer (DR 2700, HACH, USA). The orthophosphate (
) concentration agrees with the TP (PO4-P) concentration divided by the value 0.3261, as suggested by the Hach methods. The pH was adjusted using NaOH (50%) or HCl (37%). A digital pH meter (AB15, Fisher Scientific, USA) was used to measure the pH. Constant stirring was done using a magnetic stirrer. Anaerobic bioreactor effluent was filtered through glass fiber filters (Whatman, 47 mm) with a pore size of 1.2 μm to eliminate the effects of suspended solids in the experiment. During each run, samples of 10 mL were collected from the surface at an interval of 2.5 min and filtered through cellulose acetate membrane filters (Whatman, 47 mm) with a pore size of 0.45 μm. The filtrate was acidified using ACS-grade nitric acid for elemental analysis. Electrocoagulation effluent was centrifuged to separate flocs and precipitate from the treated effluent. This precipitate was oven dried at 105 °C for X-ray diffraction (XRD), energy-dispersive spectroscopy (EDS), and scanning electron microscopy (SEM).
Initial characteristics of the filtered anaerobic bioreactor effluent
| Parameter | Unit | Value |
|---|---|---|
| pH | – | 7.5 ± 0.25 |
| Conductivity | mS/cm | 2.7 ± 0.09 |
| fCOD | mg/L | 119.1 ± 5 |
| TP | mg/L | 45.6 ± 2 |
| NH3-N | mg/L | 141.5 ± 2.5 |
| Ca | mg/L | 70.9 ± 2 |
| Mg | mg/L | 12 ± 0.5 |
| Na | mg/L | 96.4 ± 3 |
| Parameter | Unit | Value |
|---|---|---|
| pH | – | 7.5 ± 0.25 |
| Conductivity | mS/cm | 2.7 ± 0.09 |
| fCOD | mg/L | 119.1 ± 5 |
| TP | mg/L | 45.6 ± 2 |
| NH3-N | mg/L | 141.5 ± 2.5 |
| Ca | mg/L | 70.9 ± 2 |
| Mg | mg/L | 12 ± 0.5 |
| Na | mg/L | 96.4 ± 3 |
Here Ci and Cf are the initial and final concentrations of TP.
Here, m represents the mass of the metal ions generated in grams, I represents the current in amperes, t represents the treatment time in seconds, M represents the material's molar mass in g/mol, z represents the valency of the produced ion (for iron, z = 2), F represents the Faraday's constant (96,500 C/mol). However, this equation does not consider electrode surface conditions.
Here E denotes the voltage in volts, I denotes the current applied in amperes, t denotes the RT in hours, and V denotes the volume of the effluent treated in m3.
Recovery of phosphorus from the post-precipitated electrocoagulation sludge was performed by combustion at 900 °C followed by acid leaching. Acid leaching tests were conducted with sulfuric acid with an acid load of 100 kg/kg P and various liquid-to-solid ratios (50, 100, 150, and 200 mL/g) (Atienza-Martínez et al. 2014). Experiments were conducted in duplicate. The samples were collected after 24 h for analysis.
Experimental design and model development
Independent variables and levels
| Variables | − 2(α) | − 1 | 0 | + 1 | + 2(α) |
|---|---|---|---|---|---|
| pH | 3 | 4.75 | 6.5 | 8.25 | 10 |
| RT (min) | 5 | 7.5 | 10 | 12.5 | 15 |
| Current density (A/m2) | 100 | 150 | 200 | 250 | 300 |
| IED (cm) | 0.5 | 0.75 | 1 | 1.25 | 1.5 |
| Variables | − 2(α) | − 1 | 0 | + 1 | + 2(α) |
|---|---|---|---|---|---|
| pH | 3 | 4.75 | 6.5 | 8.25 | 10 |
| RT (min) | 5 | 7.5 | 10 | 12.5 | 15 |
| Current density (A/m2) | 100 | 150 | 200 | 250 | 300 |
| IED (cm) | 0.5 | 0.75 | 1 | 1.25 | 1.5 |
Design matrix including experimental and predicted values
| Run order | Variables | TP removal efficiency (%) | ||||
|---|---|---|---|---|---|---|
| pH | RT | CD | IED | Experimental | Predicted | |
| 1 | −1 (4.75) | −1 (7.5) | −1 (150) | −1 (0.75) | 7.47 | 8.76 |
| 2 | 1 (8.25) | −1 (7.5) | −1 (150) | −1 (0.75) | 72.83 | 72.64 |
| 3 | −1 (4.75) | 1 (12.5) | −1 (150) | −1 (0.75) | 36.08 | 37.91 |
| 4 | 1 (8.25) | 1 (12.5) | −1 (150) | −1 (0.75) | 91.28 | 93.04 |
| 5 | −1 (4.75) | −1 (7.5) | 1 (250) | −1 (0.75) | 34.58 | 37.47 |
| 6 | 1 (8.25) | −1 (7.5) | 1 (250) | −1 (0.75) | 84.07 | 80.15 |
| 7 | −1 (4.75) | 1 (12.5) | 1 (250) | −1 (0.75) | 67.43 | 66.61 |
| 8 | 1 (8.25) | 1 (12.5) | 1 (250) | −1 (0.75) | 96.94 | 100.55 |
| 9 | −1 (4.75) | −1 (7.5) | −1 (150) | 1 (1.25) | 27.09 | 22.97 |
| 10 | 1 (8.25) | −1 (7.5) | −1 (150) | 1 (1.25) | 71.50 | 73.31 |
| 11 | −1 (4.75) | 1 (12.5) | −1 (150) | 1 (1.25) | 36.08 | 40.58 |
| 12 | 1 (8.25) | 1 (12.5) | −1 (150) | 1 (1.25) | 86.00 | 82.19 |
| 13 | −1 (4.75) | −1 (7.5) | 1 (250) | 1 (1.25) | 61.69 | 60.51 |
| 14 | 1 (8.25) | −1 (7.5) | 1 (250) | 1 (1.25) | 92.41 | 89.66 |
| 15 | −1 (4.75) | 1 (12.5) | 1 (250) | 1 (1.25) | 78.45 | 78.13 |
| 16 | 1 (8.25) | 1 (12.5) | 1 (250) | 1 (1.25) | 98.83 | 98.54 |
| 17 | −2 (3) | 0 (10) | 0 (200) | 0 (1) | 5.40 | 3.41 |
| 18 | 2 (10) | 0 (10). | 0 (200) | 0 (1) | 85.77 | 87.69 |
| 19 | 0 (6.5) | −2 (5) | 0 (200) | 0 (1) | 42.06 | 45.18 |
| 20 | 0 (6.5) | 2 (15) | 0 (200) | 0 (1) | 86.40 | 83.21 |
| 21 | 0 (6.5) | 0 (10) | −2 (100) | 0 (1) | 49.23 | 47.73 |
| 22 | 0 (6.5) | 0 (10) | 2 (300) | 0 (1) | 91.35 | 92.78 |
| 23 | 0 (6.5) | 0 (10) | 0 (200) | −2 (0.5) | 76.51 | 74.66 |
| 24 | 0 (6.5) | 0 (10) | 0 (200) | 2 (1.5) | 82.39 | 86.85 |
| 25 | 0 (6.5) | 0 (10) | 0 (200) | 0 (1) | 77.38 | 80.76 |
| 26 | 0 (6.5) | 0 (10) | 0 (200) | 0 (1) | 80.13 | 80.76 |
| 27 | 0 (6.5) | 0 (10) | 0 (200) | 0 (1) | 81.71 | 80.76 |
| 28 | 0 (6.5) | 0 (10) | 0 (200) | 0 (1) | 79.96 | 80.76 |
| 29 | 0 (6.5) | 0 (10) | 0 (200) | 0 (1) | 85.69 | 80.76 |
| 30 | 0 (6.5) | 0 (10) | 0 (200) | 0 (1) | 80.65 | 80.76 |
| 31 | 0 (6.5) | 0 (10) | 0 (200) | 0 (1) | 82.47 | 80.76 |
| Run order | Variables | TP removal efficiency (%) | ||||
|---|---|---|---|---|---|---|
| pH | RT | CD | IED | Experimental | Predicted | |
| 1 | −1 (4.75) | −1 (7.5) | −1 (150) | −1 (0.75) | 7.47 | 8.76 |
| 2 | 1 (8.25) | −1 (7.5) | −1 (150) | −1 (0.75) | 72.83 | 72.64 |
| 3 | −1 (4.75) | 1 (12.5) | −1 (150) | −1 (0.75) | 36.08 | 37.91 |
| 4 | 1 (8.25) | 1 (12.5) | −1 (150) | −1 (0.75) | 91.28 | 93.04 |
| 5 | −1 (4.75) | −1 (7.5) | 1 (250) | −1 (0.75) | 34.58 | 37.47 |
| 6 | 1 (8.25) | −1 (7.5) | 1 (250) | −1 (0.75) | 84.07 | 80.15 |
| 7 | −1 (4.75) | 1 (12.5) | 1 (250) | −1 (0.75) | 67.43 | 66.61 |
| 8 | 1 (8.25) | 1 (12.5) | 1 (250) | −1 (0.75) | 96.94 | 100.55 |
| 9 | −1 (4.75) | −1 (7.5) | −1 (150) | 1 (1.25) | 27.09 | 22.97 |
| 10 | 1 (8.25) | −1 (7.5) | −1 (150) | 1 (1.25) | 71.50 | 73.31 |
| 11 | −1 (4.75) | 1 (12.5) | −1 (150) | 1 (1.25) | 36.08 | 40.58 |
| 12 | 1 (8.25) | 1 (12.5) | −1 (150) | 1 (1.25) | 86.00 | 82.19 |
| 13 | −1 (4.75) | −1 (7.5) | 1 (250) | 1 (1.25) | 61.69 | 60.51 |
| 14 | 1 (8.25) | −1 (7.5) | 1 (250) | 1 (1.25) | 92.41 | 89.66 |
| 15 | −1 (4.75) | 1 (12.5) | 1 (250) | 1 (1.25) | 78.45 | 78.13 |
| 16 | 1 (8.25) | 1 (12.5) | 1 (250) | 1 (1.25) | 98.83 | 98.54 |
| 17 | −2 (3) | 0 (10) | 0 (200) | 0 (1) | 5.40 | 3.41 |
| 18 | 2 (10) | 0 (10). | 0 (200) | 0 (1) | 85.77 | 87.69 |
| 19 | 0 (6.5) | −2 (5) | 0 (200) | 0 (1) | 42.06 | 45.18 |
| 20 | 0 (6.5) | 2 (15) | 0 (200) | 0 (1) | 86.40 | 83.21 |
| 21 | 0 (6.5) | 0 (10) | −2 (100) | 0 (1) | 49.23 | 47.73 |
| 22 | 0 (6.5) | 0 (10) | 2 (300) | 0 (1) | 91.35 | 92.78 |
| 23 | 0 (6.5) | 0 (10) | 0 (200) | −2 (0.5) | 76.51 | 74.66 |
| 24 | 0 (6.5) | 0 (10) | 0 (200) | 2 (1.5) | 82.39 | 86.85 |
| 25 | 0 (6.5) | 0 (10) | 0 (200) | 0 (1) | 77.38 | 80.76 |
| 26 | 0 (6.5) | 0 (10) | 0 (200) | 0 (1) | 80.13 | 80.76 |
| 27 | 0 (6.5) | 0 (10) | 0 (200) | 0 (1) | 81.71 | 80.76 |
| 28 | 0 (6.5) | 0 (10) | 0 (200) | 0 (1) | 79.96 | 80.76 |
| 29 | 0 (6.5) | 0 (10) | 0 (200) | 0 (1) | 85.69 | 80.76 |
| 30 | 0 (6.5) | 0 (10) | 0 (200) | 0 (1) | 80.65 | 80.76 |
| 31 | 0 (6.5) | 0 (10) | 0 (200) | 0 (1) | 82.47 | 80.76 |
RESULTS AND DISCUSSION
Statistical analysis
value indicates the ability of the model to predict the response for a given set of operating conditions. After the removal of statistically insignificant terms, the modified quadratic model for TP removal efficiency was obtained as the following equation.
ANOVA results
| Source | DF | Adj SS | Adj MS | F-value | p-value | Remarks |
|---|---|---|---|---|---|---|
| Model | 14 | 19,639.7 | 1,402.8 | 111.40 | 0.000 | Significant |
| Linear | 4 | 16,092.6 | 4,023.1 | 319.47 | 0.000 | Significant |
| A | 1 | 10,656.6 | 10,656.6 | 846.22 | 0.000 | Significant |
| B | 1 | 2,168.6 | 2,168.6 | 172.20 | 0.000 | Significant |
| C | 1 | 3,044.6 | 3,044.6 | 241.77 | 0.000 | Significant |
| D | 1 | 222.9 | 222.9 | 17.70 | 0.001 | Significant |
| Square | 4 | 2,627.2 | 656.8 | 52.16 | 0.000 | Significant |
| A*A | 1 | 2,235.7 | 2,235.7 | 177.54 | 0.000 | Significant |
| B*B | 1 | 499.9 | 499.9 | 39.69 | 0.000 | Significant |
| C*C | 1 | 203.2 | 203.2 | 16.14 | 0.001 | Significant |
| D*D | 1 | 4.0 | 4.0 | 0.32 | 0.579 | Not significant |
| 2-Way interaction | 6 | 919.9 | 153.3 | 12.18 | 0.000 | Significant |
| A*B | 1 | 76.4 | 76.4 | 6.07 | 0.025 | Significant |
| A*C | 1 | 449.3 | 449.3 | 35.68 | 0.000 | Significant |
| A*D | 1 | 183.1 | 183.1 | 14.54 | 0.002 | Significant |
| B*C | 1 | 0.2 | 0.2 | 0.01 | 0.909 | Not significant |
| B*D | 1 | 132.9 | 132.9 | 10.55 | 0.005 | Significant |
| C*D | 1 | 78.1 | 78.1 | 6.20 | 0.024 | Significant |
| Error | 16 | 201.5 | 12.6 | – | – | – |
| Lack-of-fit | 10 | 161.9 | 16.2 | 2.45 | 0.142 | Not significant |
| Pure error | 6 | 39.6 | 6.6 | – | – | – |
| Total | 30 | 19,841.2 | – | – | – | – |
| Source | DF | Adj SS | Adj MS | F-value | p-value | Remarks |
|---|---|---|---|---|---|---|
| Model | 14 | 19,639.7 | 1,402.8 | 111.40 | 0.000 | Significant |
| Linear | 4 | 16,092.6 | 4,023.1 | 319.47 | 0.000 | Significant |
| A | 1 | 10,656.6 | 10,656.6 | 846.22 | 0.000 | Significant |
| B | 1 | 2,168.6 | 2,168.6 | 172.20 | 0.000 | Significant |
| C | 1 | 3,044.6 | 3,044.6 | 241.77 | 0.000 | Significant |
| D | 1 | 222.9 | 222.9 | 17.70 | 0.001 | Significant |
| Square | 4 | 2,627.2 | 656.8 | 52.16 | 0.000 | Significant |
| A*A | 1 | 2,235.7 | 2,235.7 | 177.54 | 0.000 | Significant |
| B*B | 1 | 499.9 | 499.9 | 39.69 | 0.000 | Significant |
| C*C | 1 | 203.2 | 203.2 | 16.14 | 0.001 | Significant |
| D*D | 1 | 4.0 | 4.0 | 0.32 | 0.579 | Not significant |
| 2-Way interaction | 6 | 919.9 | 153.3 | 12.18 | 0.000 | Significant |
| A*B | 1 | 76.4 | 76.4 | 6.07 | 0.025 | Significant |
| A*C | 1 | 449.3 | 449.3 | 35.68 | 0.000 | Significant |
| A*D | 1 | 183.1 | 183.1 | 14.54 | 0.002 | Significant |
| B*C | 1 | 0.2 | 0.2 | 0.01 | 0.909 | Not significant |
| B*D | 1 | 132.9 | 132.9 | 10.55 | 0.005 | Significant |
| C*D | 1 | 78.1 | 78.1 | 6.20 | 0.024 | Significant |
| Error | 16 | 201.5 | 12.6 | – | – | – |
| Lack-of-fit | 10 | 161.9 | 16.2 | 2.45 | 0.142 | Not significant |
| Pure error | 6 | 39.6 | 6.6 | – | – | – |
| Total | 30 | 19,841.2 | – | – | – | – |
Model summary
| Parameter | R2 | Adjusted R2 | Predicted R2 | Plof (lack of fit) |
|---|---|---|---|---|
| Total phosphorus | 98.96 | 98.27 | 96.27 | 0.142 |
| Parameter | R2 | Adjusted R2 | Predicted R2 | Plof (lack of fit) |
|---|---|---|---|---|
| Total phosphorus | 98.96 | 98.27 | 96.27 | 0.142 |
Predicted vs actual TP removal efficiency.
The model equation includes linear terms, quadratic terms, and interaction terms explaining the relationship of factors with the response variable. Initial pH was found to be the most important factor among linear and squared terms affecting TP removal. The interaction of initial pH and CD was found to be the most significant interaction term affecting the response. For the main effect, a positive coefficient indicates that TP removal increases as the main effect increases, whereas a negative value would indicate that TP removal decreases as the main effect decreases.
Effect of independent process variables
Initial pH
The Initial pH of the solution is a critical factor for the removal and recovery of TP from wastewater. It affects the conductivity of the anaerobic effluent and anodic dissolution. TP removal efficiency was low under highly acidic conditions and gradually increased with an increase in pH. It denotes that the phosphorus removal mechanism is strongly influenced by the initial pH of the solution. Phosphate tends to precipitate as ferric phosphate (FePO4) at acidic pH due to a lack of hydroxide ions (OH−). During the electrocoagulation process, the solution pH increases due to the formation of OH− ions. Whereas at a higher pH range (>7), the formation of ferric hydroxide (Fe(OH)3) occurs. Ferric hydroxide acts as an adsorbent that provides active sites for the adsorption of phosphate (Sincero & Sincero 2003). Huang et al. (2017) discovered that phosphorus recovery from anaerobic bioreactor effluent using an iron-electrocoagulation system decreases as the initial pH increases, whereas Damaraju et al. (2020) discovered that removal efficiency increases as the pH approaches neutral. However, interference of calcium ions in the iron-electrocoagulation system for phosphorus removal has not been observed so far. During the experiments, the formation of hydroxyapatite was observed at alkaline conditions (pH > 7.5). In this study, TP removal was found to be increasing with an increase in pH up to 9. Further increase in the initial pH of the solution leads to electrode passivation thereby causing a decrease in removal efficiency. Electrode passivation can be reduced with electrode polarity reversal. Higher removal efficiency at alkaline pH can be justified in terms of phosphorus adsorption by ferric hydroxide and the formation of calcium phosphate.
Current density
The CD has a significant impact on the rate of electrocoagulation process (Bektaş et al. 2004). According to Faraday's law, an increase in CD causes an increase in the production of metal ions (coagulants), resulting in higher phosphorus recovery (Dolati et al. 2017). Furthermore, as CD increases, the rate of bubble formation increases while the size of the bubble decreases. Both effects are advantageous for high pollutant removal via H2 flotation (Holt et al. 2002; Kobya et al. 2006). Within a CD range of 100–300 A/m2, the phosphorus removal efficiency is found to increase with increasing CD.
Retention time
TP removal efficiency is found to be increasing with an increase in RT up to 12.5 min. For a given CD, the mass of the coagulant added from the anode into the solution is directly proportional to the RT as shown in Equation (10). Furthermore, a longer RT allows enough time for the reaction to occur under the continuous addition of metal ions (Ramcharan & Bissessur 2017). Further increase in RT beyond 12.5 min had a negligible effect on the removal efficiency because enough metal ions were already present in the solution. This observation is in agreement with other findings (Lacasa et al. 2011; Attour et al. 2014; Bakshi et al. 2020; Damaraju et al. 2020)
Inter-electrode distance
IED is an important factor affecting the performance of electrocoagulation systems. Within a range of 0.5–1.5 cm, the efficiency of phosphorus removal was found to be increasing with an increase in IED. The maximum phosphorus removal was observed at an IED of 1.5 cm. At a low IED (1 cm), movement of colloids and fluid through the electrode gap is obstructed resulting in the accumulation of precipitates on the electrode surfaces. As a result, electrical resistance increases (Phalakornkule et al. 2009). Furthermore, colloidal particle interaction taking place at a low IED affects flotation and settling of the precipitates which leads to increased resistance and affects TP removal (Sridhar et al. 2011; Shankar et al. 2014).
Variable interaction and response
Response surface plots for TP removal: (a) pH and RT, (b) pH and CD, (c) pH and IED, (d) RT and IED, (e) CD and IED.
Response surface plots for TP removal: (a) pH and RT, (b) pH and CD, (c) pH and IED, (d) RT and IED, (e) CD and IED.
Contour plots for TP removal: (a) pH and RT, (b) pH and CD, (c) pH and inter-electrode distance, (d) RT and IED, (e) CD and IED.
Contour plots for TP removal: (a) pH and RT, (b) pH and CD, (c) pH and inter-electrode distance, (d) RT and IED, (e) CD and IED.
According to the interaction plot of RT and IED described in Figures 4(d) and 5(d), TP removal increases with an increase in RT. The higher removal efficiency was observed in Figure 5(d) at retention times greater than 11 min, and this was independent of IED. The interaction plots for CD and IED are depicted in Figures 4(e) and 5(e). The higher removal efficiency was observed when the CD was greater than 200 A/m2 and the IED was greater than 0.87 cm.
Effect on COD removal
The electrocoagulation process can remove a variety of contaminants from wastewater, including nutrients, suspended solids, and organic matter and improves the water quality of the treated effluent (Emamjomeh & Sivakumar 2009). The concentration of organic matter in an aqueous solution is typically expressed by COD. According to the study carried out by Li & Sheng (2021), an increase in organic matter resulted in a decrease in soluble Fe2+ ions in the aqueous solution. The primary reason for this claim was that organic matter removal during electrocoagulation occurs as a result of coagulation by Fe2+ or adsorption on the surface of ferric hydroxide. This resulted in coagulant (Fe2+ ions) consumption by organic matter in the wastewater, making it a limiting factor for phosphorus removal.
The effect of the iron-electrocoagulation system on COD removal was investigated using factorial plots generated with Minitab. Factorial plots describe the relationship between the process variables and the response. The initial pH of the solution influences the stability of the metal hydroxide compounds formed during the reaction as well as the efficiency of COD removal. The COD removal efficiency was found to increase with increasing pH from 3 to 10. The factorial plot for the initial pH shown in Figure S1(a) shows that high COD removal can be achieved for the iron electrode under alkaline conditions. Similar findings were observed by Ebba et al. (2021) for a pH range of 3–7.5. During this study, an increase in RT resulted in an increase in COD removal efficiency (Figure S1(b)). COD removal was found to increase with increasing CD, as illustrated in Figure S1(c). However, increasing the CD above 250 A/m2 resulted in a decrease in COD removal efficiency. This could be because electrode passivation occurs at high CD, resulting in a decrease in metal ion addition. Increased COD removal with an increase in RT and CD is due to an increase in the addition of metal ions and hydroxyl ions into the solution. An increase in IED between 0.5 and 1.5 cm has no effect on COD removal efficiency. The factorial plot for IED is shown in Figure S1(d). However, when compared to 0.5 and 1.5 cm inter-electrode distances, 1 cm IED resulted in low COD removal.
Optimization
Optimized conditions for the TP removal efficiency were obtained after evaluating the significance of independent process variables. The factors affecting TP removal were optimized using MiniTab software (version 20). This study found that optimal conditions are a pH of 6.75, an RT of 11.06 min, a CD of 300 A/m2 and an IED of 1.5 cm. At the optimum conditions, the model predicts that all the phosphorus present in the anaerobic bioreactor effluent will be removed. Confirmation runs were performed using the model's predicted optimal conditions. Table 6 displays the experimental results under optimal conditions.
Verification of experimental results at optimal conditions
| Optimum conditions | TP removal efficiency (%) |
|---|---|
| Model response | 100 |
| Experimental results | 98.05 |
| Error | 1.95 |
| Standard deviation | ±0.75 |
| Optimum conditions | TP removal efficiency (%) |
|---|---|
| Model response | 100 |
| Experimental results | 98.05 |
| Error | 1.95 |
| Standard deviation | ±0.75 |
pH = 6.75, RT = 11.06 min, CD = 300 A/m2, IED = 1.5 cm, V = 11.57 V.
Electrocoagulation sludge of 276 ± 11 mg was obtained from a liter of anaerobic effluent during the optimal electrocoagulation runs. XRD analysis of the sludge was carried out to find out the presence of crystalline compounds. The results as shown in Figure S2 show the amorphous nature of the sludge. EDS analysis of the sludge shown in Figure S3 revealed the presence of C, O, P, Cl, Ca, and Fe as primary elements. The presence of phosphorus indicates that it has been successfully adsorbed onto iron hydroxides. The formation of amorphous structures with large grooves (binding sites) was revealed by SEM image (Figure S4) of the sludge.
TP removal under experimental conditions was found to be 98.05%, which agrees with the predicted response. At optimal conditions, the energy consumption was determined to be 1.28 kWh/m3. The efficacy of the electrocoagulation system for phosphorus removal in this study was compared to the optimization conditions of previously conducted studies treating various types of wastewaters (Table 7).
Efficacy of EC on various types of wastewater
| Serial No | Type of wastewater | Optimum conditions | Electrode material | Removal efficiency (%) | References | |||
|---|---|---|---|---|---|---|---|---|
| pH | RT (mins) | Current density/Current/Voltage | IED (cm) | |||||
| 1 | Sludge anaerobic supernatant | 3 | 80 | 37.5 A/m2 | 2 | Fe | 99 | Huang et al. (2017) |
| 2 | Municipal wastewater | 7 | 20 | 382 A/m2 | 1 | Mild steel | 97 | Tran et al. (2012) |
| 3 | Palm oil mill effluent | 6.4 | 7.69 | 77.8 A/m2 | – | Fe | 73 | Damaraju et al. (2019) |
| 4 | Synthetic | 7 | 14 | 11.5 V | 3 | Scrap Al | 90 | Bakshi et al. (2020) |
| 5 | Dairy manure | 7.4 | 100 | 0.6 A | 4 | Low carbon steel | 96.7 | Zhang et al. (2016) |
| 6 | Synthetic | 5.5 | 40 | 40 V | 1.5 | Fe | >99 | Gharibi et al. (2010) |
| 7 | Synthetic | 7.4 | 34 | 21 A/m2 | 1.8 | Fe | 90.24 | Zeng et al. (2021) |
| 8 | Anaerobic bioreactor effluent | 6.75 | 11.06 | 300 A/m2 | 1.5 | Fe | 98.05 | Present Study |
| Serial No | Type of wastewater | Optimum conditions | Electrode material | Removal efficiency (%) | References | |||
|---|---|---|---|---|---|---|---|---|
| pH | RT (mins) | Current density/Current/Voltage | IED (cm) | |||||
| 1 | Sludge anaerobic supernatant | 3 | 80 | 37.5 A/m2 | 2 | Fe | 99 | Huang et al. (2017) |
| 2 | Municipal wastewater | 7 | 20 | 382 A/m2 | 1 | Mild steel | 97 | Tran et al. (2012) |
| 3 | Palm oil mill effluent | 6.4 | 7.69 | 77.8 A/m2 | – | Fe | 73 | Damaraju et al. (2019) |
| 4 | Synthetic | 7 | 14 | 11.5 V | 3 | Scrap Al | 90 | Bakshi et al. (2020) |
| 5 | Dairy manure | 7.4 | 100 | 0.6 A | 4 | Low carbon steel | 96.7 | Zhang et al. (2016) |
| 6 | Synthetic | 5.5 | 40 | 40 V | 1.5 | Fe | >99 | Gharibi et al. (2010) |
| 7 | Synthetic | 7.4 | 34 | 21 A/m2 | 1.8 | Fe | 90.24 | Zeng et al. (2021) |
| 8 | Anaerobic bioreactor effluent | 6.75 | 11.06 | 300 A/m2 | 1.5 | Fe | 98.05 | Present Study |
Kinetics of TP removal
Ln(Ct) vs time.
Acid leaching
Effect of L/S ratio on phosphorus yield using sulfuric acid.
CONCLUSIONS
In this study, a monopolar electrocoagulation reactor was developed to investigate the feasibility of the removal and recovery of phosphorus from the effluent of an anaerobic bioreactor treating industrial wastewaters. The response surface method was used to investigate and model the effects of independent process variables. The statistical optimization predicted complete (100%) removal of phosphorus at the optimum condition of pH of 6.75, RT of 11.06 min, CD of 300 A/m2, and IED of 1.5 cm, whereas the removal of 98.05% was observed during experimental conditions. Energy consumption at optimized conditions was found to be 1.28 kWh/m3. The batch monopolar system is efficient in removing phosphorus from the anaerobic bioreactor effluent. A kinetic study for TP removal at optimal conditions revealed that the TP removal rate was well fitted to the first-order rate model. Acid leaching data show that the maximum phosphorus recovered from the post-precipitated sludge ash is 91% at a liquid-to-solid ratio of 100 mL/g. More research should be carried out to study the electrode passivation and the effect of electrode polarity reversal on the performance of electrocoagulation reactors.
ACKNOWLEDGEMENTS
The authors are thankful for the funding support provided by the Natural Sciences and Engineering Research Council (NSERC) of Canada, New Brunswick Innovation Foundation, and the University of New Brunswick. The authors are grateful to Steve Cogswell at the University of New Brunswick's Microscopic and Microanalysis Facility for SEM and Ven Reddy for XRD at the University of New Brunswick's Geochemical and Spectrographic Facility.
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.
%20cropped.png?versionId=5955)