In this study, emulsified oil removals have been studied electrochemically by using stainless steel sponge electrode beds. A first-order electroflotation model was developed and the model estimation were consistent with the experimental results. It was found out that the mean electroflotation rate constant was mainly a function of the voltage applied to the electrode beds. In addition, the properties of intermediate materials (electrical conductivity) placed between the anode and cathode electrode beds strongly affected removal yields. For the initial concentration of 57,150 mg/L, the chemical oxygen demand (COD) removal was obtained as 85% under the conditions of voltage gradient 15 V. The experiments were also performed by varying the electrode bed lengths. Even though higher oil yields were obtained at 27 cm bed length, similar oil yields were also obtained at 18 cm bed length, especially after 60 min, with less energy consumption. Therefore, the optimum electrode bed length was concluded to be 18 cm.
Oily wastewaters are generated from various industries, such as food, petrochemical, steel, leather, textile, transportation, machining and metals. Numerous techniques are available for the treatment of oil/water mixture which include gravity settlers, dissolved air flotation, adsorption, coagulation-flocculation and membrane filtration. However, these conventional separation techniques are often ineffective for the separation of emulsified oils, particularly for oil droplet sizes smaller than 10 μm (Hosny 1996; Bande et al. 2008). Electrochemical treatment processes, especially electrocoagulation, have been successfully applied in the literature for the emulsified oil removals by many studies (Chen 2004; Yang 2007; Kobya et al. 2011; Sangal et al. 2013; An et al. 2017). Electrocoagulation is related to chemical coagulation, and involves the supply of coagulant ions (Al3+, Fe3+) by the application of an electrical current to a sacrificial anode (usually aluminum or iron plates) placed into an electrochemical reactor. Hydrogen gas also evolves at cathode that is involved in a flotation process during electrocoagulation. In the case of electroflotation or electrolytic flotation, only oxygen and hydrogen gases form at the electrodes which are non-sacrificial electrodes such as platinum, gold, stainless steel, graphite, carbon felt and titanium as a result of water electrolysis reactions. Oil droplets adsorb onto gas bubbles and they rise up to the free surface, where oil can be removed by skimming. The dispersed gas bubbles formed from electrolysis are extremely fine and uniform and bubble diameters range in size from 20 to 40 μm (Bande et al. 2008; Montes-Atenas et al. 2010). Gas bubbles concentration can easily be controlled by varying current, thereby increasing the probabilities of bubble and oil droplet collision.
In order to obtain high yields at electrochemical treatment processes, working electrode surfaces should be high. Although the traditional vertical parallel plate electrodes are still conducted in the majority of electrochemical processes, significant researches have been conducted to investigate different shapes of electrodes (rod, mesh, tubes) including horizontally oriented (Genc & Eryilmaz 2017) and rotating electrodes (Rivera & Nava 2007). Three-dimensional electrodes, such as fluidized steel ball electrodes (Jung et al. 2015), provide a large specific electrode area and large numbers of the mass transport coefficient (Zhang et al. 2013; Kyzas et al. 2016). Hence, the studies on electrochemical treatment processes have been directed towards optimizing high specific electrode area by using three-dimensional electrode arrangements.
In the present study, three-dimensional fixed bed electrodes, which were formed by stainless steel sponges, were utilized for the treatment of oily wastewater. The effects of electrical potential, electrode bed lengths, pH and properties of intermediate materials on treatment efficiency were investigated experimentally. In addition, a first-order electroflotation model has been developed and shown that the electroflotation rate constant is mainly depended on the voltage applied to the electrodes. The electroflotation model predictions were consistent with the experimental observations. In addition, a preliminary test performed with real wastewater has shown that the steel sponge electrodes can be successfully utilized for the treatment of oily wastewater, however, some more work is needed to improve oil yields.
Synthetic oily wastewaters were produced from bor oil (Petrol Ofisi), which is a commonly used cutting oil in the metal cutting industry in Turkey. It is paraffin-based mineral oil and contains surfactants and other chemicals such as biocides, lubricating agents, pressure additives, anti-foam agents and corrosion inhibitors. The synthetic wastewater samples were prepared by adding bor oil to tap water (2%). Then the mixture was stirred mechanically (Heidolph R2R 2020) at a stirring speed of 2,000 rpm for 30 min. The properties of the synthetic wastewater are presented in Table 1.
|Parameters .||Value .|
|Parameters .||Value .|
The electroflotation cell was made of plexiglass and its dimension were 37 cm in length, 17 cm in width and 10 cm in height (Figure 1). The lengths of inlet and outlet sections were 5 cm. The anode and cathode beds were formed by using stainless steel sponges (Figure 2). The heights of anode and cathode beds were 2 cm and 7 cm, respectively. The porosities of both electrode beds were around 97% and the corresponding numbers of sponges were 6 and 30, respectively. Polyurethane materials were placed in between electrode beds and the electrical potentials were supplied by using a DC power supply (18 V, 10 A). Stainless steel sponge electrodes were utilized only once and were submerged.
The electroflotation experiments were first performed at batch operating mode by recirculating the effluent to the inlet. Then the experiments were operated at continuous scale by adjusting water flowrate (50, 100 and 150 mL/min) using a peristaltic pump (Masterflex).
Turbidity measurements have been used as an indication of oil removals in this study since a very strong linear correlation was observed between turbidity and oil percentage of wastewater. The evaluated regression coefficient was very close to 1 (R2 = 0.9976). The turbidities of samples were measured by using Aqualytic AL450T-IR and three readings were taken for each operating conditions. In addition, at the optimized operating conditions, chemical oxygen demand (COD) removal percentages were also investigated. The COD of the water samples was determined depending on the procedure presented in the closed reflux titrimetric method.
DISCUSSION AND RESULTS
Electroflotation mechanism in the stainless steel sponge electrode cell
The electroflotation experiments were first performed at batch mode by recycling (150 mL/min). In the study, 5, 10 and 15 V potential differences were applied to the stainless steel sponge electrode beds. The observed turbidity removal percentages are shown in Figure 3. It can be seen that there is no turbidity removal, i.e. oil removal, when there is no potential gradient in the electrode beds. This result clearly indicates that the oil droplets are not adsorbed onto the stainless steel sponges and the intermediate polyurethane materials used. The highest oil removal efficiency was around 85% and observed at the highest applied voltage (15 V). On the other hand, the attained removal efficiency was only around 30% when the applied potential difference was 5 V.
According to the result presented in Figure 4, it can be concluded that the steady states were reached around 40 min for the three studied flowrates. The steady state turbidities were around 5,467, 11,986 and 16,645 NTU, respectively, for 50, 100 and 150 mL/min. Since the hydraulic residence times (60, 30 and 20 min) can be evaluated for the studied flowrates, one can obtain the mean electroflotation rate constant from Equation (5). The electroflotation rate constant ‘k’ was evaluated as 0.04 min−1 for the three studied flowrates. Therefore, it can be concluded that ‘k’ is mainly a function of applied potential difference to the electrode beds and the applied voltage is the main parameter for controlling removal efficiency.
Oil removal has been also tested by performing a test where COD variations with time was measured. Similar to the turbidity removal efficiencies, an exponential behavior was observed for the COD removals and 85% COD removal was obtained under the 15 V potential difference.
Influence of stainless steel sponge electrode bed lengths on removal efficiency
Gas bubbles concentration increases as a result of increases in electrode surface area with the improvements in bed length. In this study, three electrode bed lengths (L = 9, 18 and 27 cm) have been studied at constant bed porosity (0.97) and an optimum electrode bed length is investigated. The variations of turbidity removals with time are shown in Figure 5. The currents passing through the electroflotation cell in these experiments were also depicted. The highest turbidity removal efficiencies were attained at L = 27 cm up to 50 min while the removal efficiencies at L = 18 cm and L = 27 cm were almost coincide after 60 min.
The total energy consumptions SEC and SECR were calculated with Equations (1) and (2) by using the current values presented in Figure 5. The results are shown in Table 2. The evaluated SEC and SECR values were consistent with the results presented in the literature (Karhu et al. 2012; Ochando-Pulido et al. 2017). When the electrode bed length increases, the total energy consumption also increases. This is expected theoretically because the potential gradient across the electrode bed is constant at 15 V and the current increases with bed length. Even though turbidity removal efficiency stays constant at 90%, the energy consumption is much higher at L = 27 cm. Therefore, when the turbidity removal efficiencies and energy consumptions are taken into consideration, the optimum electrode bed length is concluded to be 18 cm.
|Bed length (cm) .||Turbidity removal (%) .||Energy consumption (kWh) .||SEC kWhm−3 .||SECR kWhkg−1 .|
|Bed length (cm) .||Turbidity removal (%) .||Energy consumption (kWh) .||SEC kWhm−3 .||SECR kWhkg−1 .|
Influence of intermediate electrode bed materials on removal efficiency
Three polyurethane materials (A, B, and C) were placed between the anode and cathode beds. The intermediate materials A and B were in the form of fabric, while the material C was a sponge. The height of each material was 1 cm when they were placed between the electrode beds. The hydraulic conductivities of materials A, B and C were measured as 0.0101, 0.0062 and 0081 m3.m−2.s−1, respectively. The turbidity removal efficiencies and corresponding currents passing through the electroflotation cell are shown in Figure 6. The use of material A results in higher removal efficiencies and higher currents, while only 8% removal efficiency was observed for the utilization of material C. The porosities of materials A and B were similar but the electrical conductivities were quite different (Figure 6). According to Faraday law, the gas formation rates at the electrode surfaces are a strong function of electrical current passing through an electrochemical cell. Since the electrical conductivity of material C is low, the current passing through the electroflotation cell was low resulting in less bubble generation. This can be the cause of almost no oil removal in the utilization of material C.
Influence of pH on removal efficiency
Experiments were performed at pH 6, 8 and 10, applying 15 V potential difference to the electrode beds (L = 18 cm). The pH adjustments were made with hydrochloric acid and sodium hydroxide additions. The corresponding turbidity removal efficiencies were 94, 90 and 92%. The cause of slight variations in turbidity removal efficiency with pH can be explained depending on the stability of bor oil water dispersion. The pH of bor oil in water dispersion is in the range of 7–8 (Table 1). The zeta potentials of bor oil droplets were measured −40, −120 and −70 mV at pH 6, 8 and 10, respectively, by Malvern Zetasizer-4. These results show that oil droplets have a negative surface charge by adsorbing hydroxyl ions (Rios et al. 1998). The variations effect of pH on the zeta potentials of bor oil droplets can be explained by considering that droplets contain pH-dependent ionizable functional groups, both acidic and basic, that can undergo dissociation and protonation (Knecht et al. 2010). According to the zeta potential measurements, bor oil dispersion is most stable at pH 8 where oil droplets have maximum negative surface charges.
Electroflotation of real wastewater samples
Bilge water collected from ships was treated with the sponge electrode beds under 15 V gradient. The wastewater was very viscous and contained high amount of grease. Therefore, the experiment was carried out after diluting the wastewater with 5% by volume of tap water and skimming of grease over the top. The treatment efficiency was around 65%, which was much lower than the treatment efficiency of synthetic wastewater. During the treatment, oils in bilge water adhered to the steel sponge electrodes and it was not possible to separate it from the electrode surfaces. In the case of bor oil, no adsorption of oil onto the electrode surfaces were observed. There was almost no change in the mass of steel sponge electrodes during the experiment and only some color changes were observed.
Based on the electroflotation experiments performed by using oily wastewaters, the following conclusions are drawn:
Turbidity variations with time were successfully estimated by using a first-order electroflotation model.
85% oil removal efficiency was obtained by applying 15 V potential difference to the stainless steel sponge bed electrodes.
The optimum bed length was found to be 18 cm depending on both oil removal efficiency and power consumption.
The removal efficiency was a strong function of intermediate electrode bed material properties as well, such as electrical conductivity.