The effect of internal impellers on mixing in an electrochemical reactor with rotating rings electrodes

Chemical Engineering and Processing 88 (2015) 37–46

Contents lists available at ScienceDirect

Chemical Engineering and Processing: Process Intensification journal homepage: www.elsevier.com/locate/cep

The effect of internal impellers on mixing in an electrochemical reactor with rotating rings electrodes Huerta-Chavez Oliver a , Rodriguez-Arias Yonatan b , Victor X. Mendoza-Escamilla c , Mollinedo Helvio d , Miguel A. Morales-Mora e, Sergio A. Martínez-Delgadillo f, * a CÁTEDRAS CONACYT Posgrado de Mecátronica, Tecnológico de Estudios Superiores de Ecatepec TESE, Av. Tecnológico S/N, Col. Valle de Anáhuac, CP 55210, Ecatepec, Edo. de México b SEPI. ESIME Zacatenco, Instituto Politécnico Nacional, G.A. Madero, CP 07738, México D.F. Mexico c Depto. Electrónica, Universidad Autónoma Metropolitana Azcapotzalco, Av. San Pablo 180, Azcapotzalco, CP 07740, México D.F. México d UPIITA, Instituto Politécnico Nacional, Av. IPN 2580, G.A. Madero, CP 07340, México D.F. Mexico e PEMEX, Petroquimica, Subgerencia de Protección Ambiental, PEMEX-Petroquímica, Jacarandas 100, Colonia Rancho Alegre, Coatzacoalcos, Veracruz 96558, Mexico f Depto. Ciencias Básicas, Universidad Autónoma Metropolitana Azcapotzalco, Av. San Pablo 180, Azcapotzalco, CP 07740, México D.F. Mexico

A R T I C L E I N F O

A B S T R A C T

Article history: Received 25 May 2014 Received in revised form 14 July 2014 Accepted 6 December 2014 Available online 10 December 2014

An electrochemical reactor with rotating electrodes has been used to to remove pollutants from aqueous media. Poor mixing and passivation of electrodes surface have been identified as the major drawbacks for the operation of this type of reactors because they adversely affect the critical reactions that take place in the liquid bulk. In this work, three different reactor configurations are proposed and their performance on reactor mixing time and process costs is evaluated. CFD simulations, based on previously validated models, were used to observe mixing inside the electrochemical reactors. Three different arrays were used for the rotating rings electrodes: (a) without impellers, (b) with four internal vertical fins and (c) with a pitched blade central impeller. Power consumption, torque, and parameters such as turbulent intensity, mixing time, among others, were evaluated for all configurations. The reactor with no impellers showed two separated zones of recirculation, reducing the reactor mixing and performance. The reactor with pitched blade impeller, showed no significant improvement due to its low central impeller pumping capacity at low rotational speeds (150 rpm). The array with 4 vertical fins operated at 130 rpm presented the highest flow/power ratio, and the lowest mixing time. ã 2014 Elsevier B.V. All rights reserved.

Keywords: Electrochemical Reactor CFD simulation Mixing Impeller Hexavalent chromium

1. Introduction The hexavalent chromium (Cr(VI)) contained in the electroplating and galvanizing wastewaters is a major source of environmental contamination [1,2]. If not properly treated, wastewaters from these industrial processes can contaminate groundwater and cause serious health problems due to Cr(VI) high toxicity. Several methods have been used to remove Cr(VI) from wastewaters, namely, the use of bisulfite, evaporation, ion exchange, and ferrous sulfate, among others. As an alternative to these treatments, the electrochemical treatment has proved to be an effective method for Cr(VI) reduction from wastewaters. By using this technique, very low levels of concentration can be reached [3]. During the electrochemical treatment, Fe(II) is

* Corresponding author. Tel.: +525 553189007; fax: +525 53189011. E-mail address: [email protected] (S.A. Martínez-Delgadillo). http://dx.doi.org/10.1016/j.cep.2014.12.003 0255-2701/ ã 2014 Elsevier B.V. All rights reserved.

released from the anode reducing Cr(VI) into Cr(III) in the bulk liquid as shown in Eqs. (A)–(C) [4]. (a) Cr(VI) reduction through formation of ferrous ions Fe(II) as a result of oxidation at the steel anodes (solution): 6Fe2+(aq) + Cr2O72
(aq) + 14H+(aq) $ 6Fe3+(aq) + 2Cr3+(aq) + 7H2O(l) (A (l) (b) Cr(VI) reduction at the cathode: Cr2O7

2



3+
(aq) + 7H2O(l) + 6e $ 2Cr (aq) + 14OH (aq)

(B

(l)

(c) Fe(III) reduction to Fe(II) at the cathode: Fe3+(aq) + e
$ Fe2+(aq)

(C)

Due to the Cr(VI) electrochemical reduction to Cr(III) is mainly carried out in the solution, mixing is a very important factor in the process. At the same time oxide film can be formed on the surface of the electrodes (passivation effect). Therefore, to increase the reaction

38

H.-C. Oliver et al. / Chemical Engineering and Processing 88 (2015) 37–46

rate in the bulk liquid and mitigate the electrode passivation, efficient mixing must be maintained in the reactor. To this purpose, an electrochemical reactor with rotating ring-shaped electrodes, as shown in Fig. 1(a), was designed and studied in previous works [5]. The Cr(VI) was removed from 500 mg/L to concentrations less than 0.5 mg/L, this means that there is a the reduction Cr(VI) in to Cr(III) efficiency higher than 99.9%. The operation conditions were: voltage of 2.75 V, current density of 190 A/m2 and pH of 1.5. The agitation generated by the rotational movement of the electrodes in the reactor, enhanced mixing and reduced the passivation oxide film on the electrode, thus decreasing the processing time. The effect of the rotating speed was evaluated, and it was found that as the rotational speed increased, the treatment time was reduced. However, at higher rotational speeds (i.e., 150 rpm and 230 rpm) there were no important differences in the processing times because the fluid behavior in the central region of the electrodes rings was barely affected by the rpm increment [6,7]. In order to enhance the agitation of the liquid and ensure optimal conditions for the electrochemical process, the electrochemical reactor presented in this work was modified to include impellers inside the central region of the rotating rings electrode. The intention was to generate greater and more homogeneous mixing and turbulence throughout the reactor specifically intended to create a mixing flow field circulating downwards inside the rotating rings and

upwards in the outer vicinity of the electrodes. For that purpose, two types of impellers were added to the electrodes array; one with a 45 pitched blade impeller mounted on the electrodes shaft (Fig. 1(c)) and a set of four internal fins symmetrically disposed directly on the inner side of the rings (Fig. 1(b)). However, both solutions demand an additional input of energy and therefore, their performances were compared against the original arrangement of rotating rings, based on a constant power consumption constraint. Power consumption, axial velocity, vorticity magnitude, turbulent intensity, mixing time, power number, pumping and circulation number were evaluated for each case. 2. Materials and methods Fig. 1 shows the three types of arrays tested: with no impellers (a), with four fins (b) and with a central 45 pitched blade impeller (c). The electrochemical reactor capacity was 18 L; the reactor consists of a cylindrical tank with four baffles arranged symmetrically and a set of 14 iron steel rings allocated in a sequence of one cathode followed by one anode. The electromechanical connection of the anode rings was made with a pair of external bar allocated 180 apart from each other. A similar array of bars was disposed 90 apart from them to connect the cathode components. Both arrays produce mixing outside the ring electrode. Such bars produce

Fig. 1. Electrochemical reactor (a) with no impellers, (b) with four internal fins, (c) with a central 45 pitched blade impeller and (d) detailed electrodes connection.

H.-C. Oliver et al. / Chemical Engineering and Processing 88 (2015) 37–46

mixing outside the ring electrode, see Fig. 1(d). The electrode assembly is mounted on a shaft which is connected to a variable speed motor. Fig. 1 shows the three arrays tested in this work; the original design, lacked of 4 fins or central impeller, (Fig. 1(a)), the array with four internal fins attached to the electrodes array (Fig. 1(b)) and the one with central 45 pitched blade impeller (Fig. 1(c)). 2.1. Numerical simulations CFD software Fluentã was used to perform the simulations of the hydrodynamic behavior of the reactor. A full 3D model was used owing to the asymmetries in the upper part of the geometry of the electrodes array. Due to the complex geometry of the models, tetrahedral cells were used for all the meshes. Simulations were carried out with various mesh densities to check for mesh independency of the solution. Grid independence was checked from 800,000 to 1,500,000 cells. Mesh refinements were made in the regions of high gradient around the reactor geometry (Fig. 2). In this model, the steady-state multiple reference frame (MRF) approach was used, the solution domain was divided into an inner region containing the rotational parts (impellers and the ring electrodes body) and an outer region containing the stationary tank with baffles, as shown in Fig. 2. This approach is recommended when the interaction between the impeller and the baffles is not strong, that is, where the baffle-impeller gap is large enough. The rotational system consisting of the ring-shaped electrodes, pitched blade impeller and internal fins were assumed as a moving wall (with no slip condition) attached to the moving reference frame; the surfaces of the tank including the baffles were considered to be a stationary wall (with no slip condition). The realizable k–e turbulent model was applied with the following values for the constants: C2e = 1.9, TKE Prandtl number = 1 (turbulent kinetic energy), TDE Prandtl number = 1.2 (specific dissipation rate). The realizable k–e model, developed by Shih et al. [8], has proved to be a significant improvement over the standard k–e model, at predicting flows involving rotation, recirculation and flows with complex secondary flow features. The realizable k–e turbulent model and MRF approach was used by Lie et al. [9] and Santos-Moreau et al. [10] with good results. The governing equations were discretized using a pressure-based segregated algorithm along with the standard scheme for pressure, the semi-implicit pressure-linked equation (SIMPLE) algorithm for the pressure–velocity coupling, and the second order upwind scheme for the momentum. The convergence criteria were verified for the residuals of all the variables <5  10
4 and further checks

39

for the convergence were made confirming that the values of the torque and dynamic pressure were constant. 2.2. Power, pumping and circulation number The power consumed by the impeller is an important design parameter and is a function of the characteristics of the system such as impeller speed, impeller diameter, and impeller design, physical properties of the fluid, tank size and geometry, location of the impeller and baffle design. It can be obtained from the torque applied to the stirring system [11–13]. Power applied to the impeller is transferred to the fluid causing fluid motion. The power is dissipated into heat within the fluid over time by means of viscous dissipation. The power consumed by the impeller during stirring should be equal to the power dissipated by the impeller in the liquid (Eq. (1)). Z R Z H Z 2p redrdzdu (1) P¼ 0

0

0

Hence, the impeller power number can be calculated from the volume integration of turbulent energy dissipation rate predicated by the CFD model [14], where r is the density and e is the turbulent energy dissipation rate. The power consumed is often represented in a dimensionless form through the power number NP, which is defined in Eq. (2). Np ¼

P

rN3 D5

(2)

where D is the outer diameter of the impeller. The pumping capacity Q, of an impeller is the volumetric flow rate passing through the planes established by the impeller rotation. Pumping number is defined in Eq. (3) [15–17]. R z2 2 z1 pRU r dz Q NQ ¼ ¼ (3) rND3 ND3 where Q is the pumping capacity, Ur is the mean radial velocity (m s
1), and R is the radius of the ring electrodes. And the circulation number NC is defined by Eq. (4). RR Q up 2 0 r prU z dr NC ¼ ¼ (4) rND3 ND3 where Uz is the axial velocity (m s
1), Rr represents the radial distance to vorticity center and Qup is the upward mass flow rate (upflow) evaluated at the up plane surface (Fig. 3(b)). It was also

Fig. 2. Computational model and grid regions.

40

H.-C. Oliver et al. / Chemical Engineering and Processing 88 (2015) 37–46

Fig. 3. (a) Cylindrical surface used to evaluate Q and NQ, (b) Plane surfaces used to evaluate the upward mass flow rate at the up (Qup) and down(Qupd) plane surfaces of the reactor.

evaluated the upward mass flow rate (Qupd) at the down plane surface shown in Fig. 3(b). 2.3. Mixing time The mixing time (t99) was evaluated for the different arrangements using a tracer pulse injection at the top of the reactor. The tracer had the same properties as the liquid (water) and its concentration was measured as a function of time in seven points in the reactor (shown in Fig. 4). The uniformity U was evaluated based on the coefficient of variation (COV), which is the ratio between standard deviation s Y and the average, as can be seen in Eqs. (5) and (6), respectively. C OV ¼

sY Y

U ¼ 1
C OV

(5)

(6)

3. Results and discussion In Fig. 5, the results of axial velocity (m/s), vorticity magnitude (1/s) and turbulent intensity (%) for the three types of arrays tested in the electrochemical reactor are presented for 150 rpm. For all of the cases, there is an upward flow (axial velocity, colored in yellow and red), surrounding the outer part of the rings arrays, which has a natural circulation structure due to the electrode assembly rotation, which produces a radial velocity component in the fluid mass, expelling part of the fluid towards the walls of the tank, increasing the dynamic pressure in the periphery and generating a low pressure zone around the electrodes rotation axis. For the case of the reactor without impellers (Fig. 1(a)), the overall effect produces a downward axial flow (axial velocity, colored in blue) inside the central region of the rotating rings, and upward flow outside the electrodes, generating two recirculation flow zones. The primary zone is located at the upper part of the reactor (Fig. 6(a) and Fig. 7(a)) taking up about 65% of the total flow. The secondary zone of recirculation, surrounds the lower outer part of the electrodes assembly and is trapped in the bottom of the reactor, generating recirculation flow zones. The reactor without impellers has the lowest vorticity magnitude and turbulent intensity inside the rings electrode, as illustrated in Fig. 5(e) and (i), respectively. In addition, both zones of recirculation are not symmetric and they seem to be

disconnected. The dynamics of the flow in the primary one promotes better mixing and reaction in the bulk liquid, between the reactants dissolved in the liquid and those ones released from the anode. On the other hand, the reactants released from the anode in the primary zone, do not reach the secondary zone. Therefore, the mixing is poor in the reactor and the electrochemical process becomes less efficient. In order to improve the mixing and to reduce the recirculation zones, the rotational speed was increased from 150 rpm to 230 rpm with a high increment in the power from 4.9 W to 16.0 W. However, flow pattern was barely affected by the rpm increment according with other studies [6,7]. Although the axial velocity, vorticity and turbulent intensity were increased as shown in Fig. 5(d), (h) and (l), respectively, the downward flow (colored in blue in Fig. 5(d)) grew slightly deeper compared to the 150 rpm case (Fig. 5(a)). Vorticity was clearly increased, but a similar zone with low vorticity magnitude inside the rotating electrodes (Fig. 5(h)) remains as for the 150 rpm case (Fig. 5(e)). The turbulent intensity (Fig. 5(l)) increased inside the electrodes. Despite the power consumption increment (three times higher than for 150 rpm), the 230 rpm flow pattern shows the same zones of recirculation found in the 150 rpm case. The secondary recirculation zone was not

Fig. 4. Injection and sampling points in the reactor to evaluate the mixing time (t99).

H.-C. Oliver et al. / Chemical Engineering and Processing 88 (2015) 37–46

41

Fig. 5. Contours of axial velocity (m/s), vorticity magnitude (1/s) and turbulent intensity (%) for the electrochemical reactor: (a, e and i) with no impellers, (b, f and j) with four internal fins and (c, g and k) with a central 45 pitched blade impeller at 150 rpm and (d, h and l) with no impellers at 230 rpm. (For interpretation of the references to color in the text, the reader is referred to the web version of this article.)

Fig. 6. Velocity vectors for the reactor without impellers at: (a) 150 rpm and (b) 230 rpm.

42

H.-C. Oliver et al. / Chemical Engineering and Processing 88 (2015) 37–46

Fig. 7. Axial velocity streamlines for the reactor without impellers at: (a) 150 rpm, (b) 230 rpm and (c) with 4 fins at 150 rpm.

significantly reduced and continued taking out energy from the central flow stream (Fig. 6(b) and Fig.7(b)), deflecting the flow from the electrodes (main reaction zone), causing poor mixing and large dead zones in the lower part of the reactor. These zones of recirculation increase in intensity, producing a constriction (or venturi) that accelerates the flow upwards into the primary zone and it is the cause for which the maximum axial velocity vectors occur in the outer periphery of the electrodes array, indicated by the arrows in Fig. 5(d) and Fig. 6(b). In Fig. 6(a), another dead zone with low turbulence and low velocity (into the blue circle) is formed at the bottom of the reactor under the electrodes rotating axis. Although the rotational speed was increased to 230 rpm, this zone did not disappear as shown in Fig. 6(b). Considering all the points set out above, it is confirmed that the rotational speed increase did not have an important effect on the performance of the electrochemical reactor, in spite the rpm increased [18]. The observed drawbacks in the reactor with no impellers, led to evaluate the two other arrays to improve the flow circulation throughout the reactor and to increase the overall efficiency of the reactor by reducing the secondary recirculation zone. One

alternative for improving the performance of the reactor was to add four vertical fins mounted inside the rings assembly and arranged symmetrically about the rotation axis, as shown in Fig. 1(b). This new configuration of the reactor, shows considerable improvement of the flow field and reduction of the secondary flow loops formed in the lower region near the lower five rings as shown in Fig. 7(c). They are pushed down to the bottom zone of the reactor thus reducing the secondary recirculation zone, increasing the primary one. In this case, the downward fluid reaches deeper inside the rotating rings electrode than in the other two cases, which allows greater mass flow rate into the flow of the primary zone (Fig. 5(b)), increasing it to 85% of the total reactor volume (Fig. 7(c)), improving mixing in the reactor. In addition, the homogeneity of the central flow (axial flow) and the venturi throat were increased. Also, the vorticity fields and turbulent intensity were improved, as shown in Fig. 5(f) and (j), respectively, reaching the highest values of all three reactor configurations tested at 150 rpm. Even though, the secondary recirculation zone was reduced and the vorticity and turbulent intensity were increased at higher rpm for this configuration, from 150 rpm to 230 rpm and to

Fig. 8. Axial velocity pathlines for the reactor with 4 fins at: (a) 230 rpm and (b) 500 rpm.

H.-C. Oliver et al. / Chemical Engineering and Processing 88 (2015) 37–46

0.45

Turbulent intensity vorcity magnitude velocity magnitude

20

0.4 0.35 0.3

15

0.25 0.2

10 0.15

Velocity magnitude (m/s)

Vorticity (1/s) or Turbulent intensity (%)

25

0.1

5

0.05 0

no impellers 150 rpm

4 fins 130 rpm

4fins 150 rpm

central impeller 150 rpm

0

Fig. 9. Turbulent intensity, velocity magnitude, and vorticity magnitude for each case.

500 rpm, it was no possible to eliminate the zone of recirculation in the lower portion of the reactor as shown in Fig. 8(a) and (b). Such flow loops prevent fluid from the secondary (bottom) zone reaches the electrodes, so there is a possibility to enhance reactor performance by increasing the flow in the primary zone above the 85% achieved with this configuration. Consequently, another modification was tested: a kind of pitched blade impeller at mid-height of the electrodes array was added (see Fig. 1(c)). Both, impeller and electrode rings were rotated at the same rotational speed, 150 rpm. However, it had a slight effect on the downward axial flow (Fig. 5(c)), which reaches a little deeper than in the reactor without impeller (Fig. 5(a)), the vorticity and turbulent intensity, showed little improvement near the impeller, as shown in Fig. 5(g) and (k), respectively. In spite of the poor results obtained with the latter reactor configuration, several reasons were recognized not to rule out this alternative. The first reason is that the down flow generated by the electrode rotation is too high for the pumping impeller capacity which also rotates at 150 rpm then, it cannot improve the down-pumping action, on the contrary, it blocks the flow inside the electrodes array. Therefore, tests should be performed at rotational impeller speeds higher than 150 rpm. The second reason is that the impeller was placed at the middle height of the rings electrode, where the main hydrodynamic field generated by rotation of the electrode assembly dominates the flow fields. Therefore, results suggest that lowering the impeller position could reduce the central flow blockage. In this case, it seems that the blade angle exceeds the natural angle of impact, it induces a

43

impeller stall, thereby contributing to produce the recirculation vortices behind the impeller (yellow color) as shown in Fig. 5(c). Based on these results, simulations were performed with the four fins configuration, at lower rotation speeds (130 rpm) so as to reduce power and obtain a value similar to those obtained for the other configurations. After power adjustment, values of turbulent intensity, velocity magnitude, vorticity magnitude and axial velocity in the reactor were still the highest of all cases. Figs. 9 and 10 show comparisons among the different cases tested. As can be seen, the highest values of turbulent intensity, velocity magnitude, vorticity magnitude and axial velocity were obtained with the four fins configuration driven at 150 rpm, which agrees with the previous discussion. Moreover, those parameters are the highest inside the rotating rings electrode and best distributed into that zone. As can be seen, the reactor with 4 fins operated at 130 rpm has the best hydrodynamic behavior (e.g., highest, vorticity, turbulent intensity and pumping mass flow rate (Q) for the same power consumption or even lower than the other arrays. Although the rotating rings electrode is not considered as a conventional impeller the NP,NQ and NC were evaluated, and the results are presented in Fig. 11. For the four fins reactor an NP = 4.4 was obtained, for all the rpm tested from 75 rpm to 500 rpm. Values of NQ = 0.72 and NC = 0.57 were calculated for the four fins reactor at 130 rpm. These values are comparable in magnitude with the values of other kind of impellers [19]. Fig. 12, shows the upward mass flow rate evaluated in the up (Qup) and down (Qupd) transversal section of reactor as defined in Fig. 3(b). As seen, the upward mass flow rate at the upper zone is similar in all the cases, but the up flow evaluated at the bottom zone of the reactor is lower in the case of the rotating ring electrode with no impellers and central impeller, which agrees with the discussed above. The upward mass flow rate in the down zone for the cases of 4 fins (130 rpm and 150 rpm) is about 55% higher than for the cases with central impeller and no impeller. The results support that the reactor with 4 fins running at 130 rpm shows a better performance than the reactors with no impellers and central impeller at 150 rpm, even though the rpm were reduced. To corroborate the results obtained before, the mixing times (t99) for the different arrangements were evaluated. Fig. 13 shows the tracer concentration as a function of time at seven points in the reactor (shown in Fig. 4), for the three cases at 150 rpm. Based on these results and Eqs. (5) and (6), the mixing times (t99) were obtained (Table 1). As can be seen, the lowest mixing time (12 s) was obtained when the reactor with 4 fins was operated at 150 rpm, which

0.8

5

0.15

Nq

Nc

Np

0.7

4.5 4

0.1

0.6

Nq or Nc

Axila velocity (m/s)

0.05 0 -0.05

0.5

3

0.4

2.5

Np

3.5

2

0.3

1.5 0.2 1

-0.1

0.1 -0.15

0.5 0

0 No impellers 150 rpm

-0.2

no impellers 150 rpm

4 fins 130 rpm

4fins 150 rpm

Fig. 10. Axial velocity (m/s).

central impeller 150 rpm

4 fins 130 rpm

4 fins 150 rpm central impeller 150 rpm

Fig. 11. Power number (NP), flow number (NQ) and secondary circulation flow number (NC) ratio for the different configurations.

44

H.-C. Oliver et al. / Chemical Engineering and Processing 88 (2015) 37–46

Fig. 12. Upward mass flow rate evaluated at the up (Qup) and down (Qupd) transversal section of the reactor.

Fig. 13. Variation in the tracer concentration as a function of time at 7 points in the reactor at 150 rpm: (a) without impellers, (b) with 4 fins and (c) with central impeller 4 PBT.

agrees with the results obtained before. As shown in Fig. 14, for the reactor operated with 4 fins, the tracer is distributed faster than the others arrays. For the reactor without impellers, 3 s after injection, the tracer reaches only the upper section of the reactor. A similar behavior was obtained for the reactor with central impeller (Fig. 14(a) and (c), respectively). On the other hand, for the reactor with 4 fins, the tracer was distributed better in the upper and inferior section of the vessel (Fig. 14(b)). In addition, 7 s after the tracer injection, the concentration in the reactor with four fins (Fig. 14(b)) reached higher uniformity than in the other cases, followed by the reactor with the central impeller (Fig. 14(c)), being that without impellers (Fig. 14(a)) the one that showed the lowest uniformity

Table 1 Comparisons of the power consumption, the pumping mass flow/power ratio and torque of the electrochemical reactor with no impellers versus the other three arrays. Electrochemical reactor array

Power

Flow/ power

Torque Mixing time (s)

No impellers 4 fins at 130 rpm 4 fins at 150 rpm Central impeller

1.0 0.7836 0.8956 0.9051

1.0 1.2761 1.1164 1.1047

1.0 25.5 0.952 13.5 1.0367 12.0 0.9363 21.0

H.-C. Oliver et al. / Chemical Engineering and Processing 88 (2015) 37–46

45

Fig. 14. Tracer distribution after 3 s and 7 s after injection, in the reactor at 150 rpm: (a) without impellers, (b) with 4 fins and (c) with central impeller 4 PBT.

of the three cases, which agrees with the results obtained before (Table 1). As it was previously mentioned, reducing the rotational electrode speed at 130 rpm, the power consumption is reduced and mixing time (13.5 s) is maintained lower than the other cases making this array the more adequate of all, which is also in agreement with the results obtained. Finally, it is important to mention that experimental measurements of P were in accordance with those ones calculated by CFD simulation within an error less than 5%. Table 1 shows the comparisons of the power consumption, the pumping mass flow/power ratio and the torque of the electrochemical reactor with no impellers versus the other three arrays. It is evident that the four fins case at 130 rpm requires less than 21% of power and less the torque (4.5%) than the reactor with no impellers and it has the best performance of all the arrays tested. It must be mentioned that there were not important variations in the voltage in the electrodes, and then the power consumption by the electrochemical process is almost constant. The changes in the power consumptions were mainly due to the different configurations as it is summarized in Table 1. It is important to point out that as it was described in Eqs. (A)–(C), during the electrochemical process there is an homogeneous chemical reaction (Cr(VI) reduction by the Fe(II), both soluble at pH 1.5), which is carried out in the bulk liquid, therefore it is not an electroflotation process. Finally, in order to reuse the treated water, the Cr(III) and Fe(III) formed must be separated after the electrochemical process. This can be done, forming the insoluble hydroxides of chromium and iron, increasing the pH to 7.5–8.0. The sludge (hydroxides) are separated from the treated water in a settler, where the conditions are totally different (very low or no mixing) and it can be an interesting subject of other study, because the hydroxides formed could be recovered and reused, as a byproduct with added value. 4. Conclusions The reactor without impellers presented the highest mixing time (t99) and the lowest flow/power ratio of all arrays tested. Therefore, hydraulic performance of the electrochemical reactor can be improved by the addition of impellers or 4 fins, inside the

rotating rings electrode. It was found that the array with 4 fins impellers inside the rotating rings electrode at 150 rpm, presented higher values of turbulent intensity, velocity magnitude, vorticity magnitude and axial velocity than the other arrays. In addition, its mixing time was the lowest of all arrays. However, the torque is the highest in comparison with the other cases. By means of reducing the rotating speed of the electrode with 4 fins, from 150 rpm to 130 rpm, it was possible to reduce the power consumption and torque, maintaining lower mixing times than the reactors without impellers or central impeller at 150 rpm. In addition, the highest flow/power ratio was reached with this array. In the case of the reactor with central impeller, the central flow is blocked due to low rotation speed (150 rpm) of the central impeller, which reduced the flow/power ratio and then, the reactor mixing efficiency. The values of NP, NQ and NC obtained in this work are comparable in magnitude with the values of other kind of impellers. Further work needs to be done to study impeller geometry, position and rotational speed (relative to the electrodes assembly). Acknowledgments Financial supports of this work by the Consejo Nacional de Ciencia y Tecnología (Proyecto No. CB-2011/169786) are gratefully acknowledged. References [1] L. Liu, X. Ma, Technology-based industrial environmental management: a case study of electroplating in Shenzhen, China, J. Clean. Prod. 18 (2010) 1731–1739. [2] G. Kong, R. White, Toward cleaner production of hot dip galvanizing industry in China, J. Clean. Prod. 18 (2010) 1092–1099. [3] S.A. Martínez, M.G. Rodríguez, Removal of chromium hexavalent from rinsing chromating waters electrochemical reduction in a laboratory pilot plant, Water Sci. Technol. 49 (2004) 115–122. [4] M.G. Rodríguez, S.A. Martínez, Removal of Cr(VI) from wastewaters in a tabular electrochemical reactor, J. Environ. Sci. Health, Part A. Toxic/Hazard. Subst. Environ. Eng. A 40 (12) (2005) 2215–2225. [5] M.G. Rodríguez, V. Mendoza, H. Puebla, S.A. Martínez, Removal of Cr(VI) from wastewaters at semi-industrial electrochemical reactors with rotating ring electrodes, J. Hazard. Mater. 163 (2009) 1221–1229. [6] H.R. Mollinedo, S.A. Martínez-Delgadillo, V. Mendoza-Escamilla, C. GutierrezTorres, J.A. Jimenez-Bernal, Evaluation of the effect of the rotational electrode

46

[7]

[8]

[9]

[10]

[11]

[12]

H.-C. Oliver et al. / Chemical Engineering and Processing 88 (2015) 37–46 speed in an electrochemical reactor using computational fluid dynamics (CFD) analysis, Ind. Eng. Chem. Res. 51 (2012) 5947–5952. S.A. Martínez, J. Ramírez, H.R. Mollinedo, V. Mendoza, C. Gutiérrez, J.A. Jiménez, Determination of the spatial distribution of the turbulent intensity and velocity field in an electrochemical reactor by CFD, Int. J. Electrochem. Sci. 8 (2013) 274–289. T. Shih, W.W. Liou, A. Shabbir, Z. Yang, J. Zhu, A new k–e eddy viscosity model for high Reynolds number turbulent flows, Comput. Fluids 24 (3) (1995) 227–238. M. Lie, C. Jing-Cai, J. Hong, Y. Chao, X. Wei-Hong, J. Wan-Qin, Design and analysis of a submerged membrane reactor by CFD simulation, Chem. Eng. Technol. 36 (2013) 1874–1882. V. Santos-Moreau, L. Brunet-Errard, M. Rolland, Numerical CFD simulation of a batch stirred tank reactor with stationary catalytic basket, Chem. Eng. J. 207–208 (2012) 596–606, doi:http://dx.doi.org/10.1016/j.cej.2012.07.020. K.K. Singh, S.M. Mahajani, K.T. Shenoy, A.W. Patwardhan, S.K. Ghosh, CFD modeling of pilot-scale pump-mixer: single-phase head and power characteristics, Chem. Eng. Sci. 62 (5) (2007) 1308–1322, doi:http://dx.doi. org/10.1016/j.ces.2006.10.028. R. Zadghaffari, J.S. Moghaddas, J. Revstedt, Large-eddy simulation of turbulent flow in a stirred tank driven by a Rushton turbine, Comput. Fluids 39 (7) (2010) 1183–1190, doi:http://dx.doi.org/10.1016/j.compfluid.2010.03.001.

[13] A. Ryma, I. Mokni, H. Mhiri, P. Bournot, A 3D CFD simulation of a self inducing pitched blade turbine downflow, Energy Convers. Manage. 64 (2012) 633–641, doi:http://dx.doi.org/10.1016/j.enconman.2012.06.005. [14] T. Kumaresan, J.B. Joshi, Effect of impeller design on the flow pattern and mixing in stirred tanks, Chem. Eng. J. 115 (3) (2006) 173–193, doi:http://dx.doi. org/10.1016/j.cej.2005.10.002. [15] I. Fort, On hydraulic efficiency of pitched blade impellers, Chem. Eng. Res. Des. 89 (6) (2011) 611–615, doi:http://dx.doi.org/10.1016/j.cherd.2010.10.005. [16] G. Chun-Yan, W. Jia-Jun, G. Xue-Ping, F. Lian-Fang, CFD simulation and PIV measurement of the flow field generated by modified pitched blade turbine impellers, Chem. Eng. Res. Des. (2013) . http://dx.doi.org/10.1016/j. cherd.2013.08.024. [17] V.V. Ranade, An efficient computational model for simulating flow in stirred vessels: a case of Rushton turbine, Chem. Eng. Sci. 52 (24) (1997) 4473–4484, doi:http://dx.doi.org/10.1016/S0009-2509(97)00292-3. [18] S. Martinez-Delgadillo, H. Mollinedo, V. Mendoza-Escamilla, C. GutiérrezTorres, J. Jiménez-Bernal, C. Barrera-Diaz, Performance evaluation of an electrochemical reactor used to reduce Cr(VI) from aqueous media applying CFD simulations, J. Clean. Prod. 34 (2012) 120–124. [19] L. Mingzhong, G. White, D. Wilkinson, K.J. Robert, Scale up study of retreat curve impeller stirred tanks using LDA measurements and CFD simulation, Chem. Eng. J. 108 (2005) 81–90.