- Email: [email protected]
Two stage electrostatic separator for the recycling of plastics from waste electrical and electronic equipment
Journal of Electrostatics 71 (2013) 681e688
Contents lists available at SciVerse ScienceDirect
Journal of Electrostatics journal homepage: www.elsevier.com/locate/elstat
Two stage electrostatic separator for the recycling of plastics from waste electrical and electronic equipment Wessim Aksa a, Karim Medles a, Mohamed Rezoug a, Mohamed Fodil Boukhoulda a, Mihai Bilici b, Lucian Dascalescu b, * a b
Electrostatics and High Voltage Research Unit, IRECOM, University Djillali Liabes, 22000 Sidi-Bel-Abbes, Algeria Institut P’, CNRS-University of Poitiers-ENSMA, IUT, 4, avenue de Varsovie, Angoulême 16021, France
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 August 2012 Received in revised form 3 November 2012 Accepted 8 March 2013 Available online 1 April 2013
The aim of study was to evaluate the effectiveness of a new facility for recycling of plastics from granular waste electrical and electronic equipment. The installation consists of two sections, the products of a first tribo-aero-electrostatic separator being subsequently treated in two free-fall electrostatic separators. The tests were performed on a mixture of polycarbonate (PC) and polyamide (PA). Analysis of the purity of the products obtained was performed using a program of image processing in MATLAB. Products of very high purity (roughly 95% for both PC and PA) were obtained at a recovery rate higher than 70%. Ó 2013 Elsevier B.V. All rights reserved.
Keywords: Electrostatic processes Image processing Triboelectricity Recycling Waste recovery
1. Introduction Computers, printers, mobile phones, and other such appliances have shorter and shorter lifetimes [1], due to the very fast progress in electronics and information technology (IT). They represent an increasingly larger part of the waste electrical and electronic equipments (WEEE), the volume of which has reached an alarming level, especially during the last twenty years, when the markets have been saturated with huge quantities of new products of this type [2,3]. According to the recent statistics, the quantity of WEEE increased by 25% in five years, with the proportion of plastics amplified by 30% in the same period [1,4e6]. This situation has drawn the attention of both governmental and non-governmental on the necessity of developing effective methods for the recycling of WEEE. The electrostatic separation methods [7e9] have already proved to be a very effective solution for the recycling of insulating materials contained in this kind of waste. This non-pollutant
* Corresponding author. Tel.: þ33 545673245. E-mail addresses: [email protected], [email protected] (L. Dascalescu). 0304-3886/$ e see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.elstat.2013.03.009
technology is characterized by low energy consummation, as well as by reduced costs of operation and maintenance [10]. The tribo-electricity, a physical phenomenon involving the charge transfer between two bodies in contact [11e16] is the main charging mechanism employed for the separation of granular insulating materials in an intense electric field. The aim of the present work is to validate a new tribo-electrostatic separation process that has been designed for increasing the purity of the plastics recovered from WEEE. 2. Experimental set-up The installation is composed of two superposed, detachable electrostatic separators, attached to a same vertical support (Fig. 1). The upper section of the set-up is a tribo-aero-electrostatic separator that consists in a parallelepiped enclosure (height: 500 mm; width: 130 mm; depth: 110 mm), having two transparent walls in order to permit the visualization of phenomena, and two opaque lateral walls that have aluminum plate electrodes glued on their internal surfaces. These electrodes are typically energized from two adjustable DC high voltage supplies of opposite polarities 50 kV, to create an electrical field sufficiently strong to control the trajectories of charged granules.
682
W. Aksa et al. / Journal of Electrostatics 71 (2013) 681e688
Fig. 1. Schematic representation of the new double-stage electrostatic separator for mixed granular plastics.
Granule charging is produced by tribo-electric effect in the fluidized bed generated in the interior of this enclosure (Fig. 1). The fluidization air is furnished by a variable-speed blower. The uniformity of the fluidization bed is ensured by a custom-designed air diffuser, which is a finely-perforated plate situated at the bottom of the upper section of the installation. The granules are introduced in the separation enclosure by a funnel supplied by a fully-adjustable vibratory feeder. Under the combined action of the gravitational, aerodynamic, and electrical forces, they separate essentially in function of the polarity of their charge and exit the first separation stage through two slots that direct them to the lower section of the experimental set-up (Fig. 2). This section is composed of two free-fall electrostatic separators, the electrodes of which are aluminum plates (520 mm 100 mm) glued to four insulating PMMA boards (650 mm 110 mm). The upper edges of these four boards are positioned at the exit of the tribo-aero-electrostatic separator. The two vertical central plaques are fixed and connected to the earth, while the two exterior electrodes are connected to high voltage supplies of opposite polarities and can rotate to form angles ranging from 0 to 45 with respect to the vertical. The separated products are recovered in two identical
collectors, each subdivided in twenty compartments (length: 100 mm; width: 30 mm; depth: 85 mm). The 20 boxes of the two collectors are numbered as shown in Fig. 3. 3. Materials and method The experiments have been carried out with granules of polycarbonate (PC) and polyamide (PA) (Table 1). The mechanical separation of these granules is impossible, because they have similar shapes and mass densities. The analysis of the purity of the separated products is facilitated by the fact that the granules have different colors (Figs. 4e6). A set of 100 granules were weighted and the results were divided by 100 to give the weight of one granule. The high-voltage U1 applied to the electrodes of the upper section was adjusted at various values ranging between 24 and 48 kV. The electrodes of the two free-fall electrostatic separators of the lower section were energized at voltages U2 and U3 of opposite polarities and similar absolute values in the range 24e32 kV. These values were established after several preliminary experiments. Thus, it was observed that separation is very poor at voltages lower than 24 kV (the electric field strength is less than 2 kV/cm, not
W. Aksa et al. / Journal of Electrostatics 71 (2013) 681e688
enough to drive the granules to the electrodes). On the other hand, corona and spark discharges may occur from the edges of the plate electrodes at voltages higher than 48 kV (for the upper section) and 32 kV (for the lower section). The other variables of systems were maintained constant: the speed of the air blower: 13,000 rev/min; the relative humidity of the air RH ¼ 41 1%; the room temperature T ¼ 18 1 C; the mass and composition of the granular product samples m ¼ 100 g (PC) þ 100 g (PA) ¼ 200 g; the feed-rate: D ¼ 1 g/s. The movable electrodes of the inferior floor have been inclined at a ¼ 32 , their inferior edges correspond to the seventh compartment of each collector. The experimental modeling of the separation process has been performed using the response surface methodology [17,18], which recommends the use of a composite factorial experimental design and the adoption of a quadratic model. For the factors considered hereafter, namely the high-voltages U1, U2 and U3 applied respectively to the upper section electrodes, and to the two pairs of electrodes in the lower section of the experimental setup, the model y of the response (i.e. the mass m and in some cases the purity p of the material collected in the third to the seven compartment), will take the following form:
y ¼ a0 þ a1U1 þ a2U2 þ a3U3 þ a12U1U2 þ a13U1U3 þ a23U2U3 þ a11U21 þ a22U22 þ a33U23
683
(1)
The experimental data were analyzed with MODDE 5.0 software (Umetrics, Sweden) [19], which calculates the coefficients of the mathematical model, draws the response contours and identifies the best adjustments of the parameters for optimizing the process. Moreover, the program calculates two statistical criteria: the goodness of fit: R2, and the goodness of prediction: Q2. The latter is a measure of how well the model will predict the responses for new experimental conditions. In the case of a good mathematical the criteria R2 and Q2 have numerical values close to the unit.
4. Results and discussion The results of the seventeen experiments carried out according to the composite factorial design are reported in Tables 2 and 3. For each experiment, the overall purity of the granular materials collected in boxes #3 to 7 was analyzed with the image processing program previously-described and the results are given in Table 4.
Fig. 2. Photograph of the experimental set-up.
684
W. Aksa et al. / Journal of Electrostatics 71 (2013) 681e688
Fig. 5. Processed image of a PCePA granular mixture, showing only the PA granules.
Fig. 3. Schematic representation of the electrode system, energized from three highvoltage supplies U1 (positive polarity), U2 (positive), U3 (negative), and of the two 20-box collectors. The negatively-charged PC granules and positively-charged PA granules are mainly collected in the left side and right side collectors, respectively.
Table 1 Characteristics of the granules of plastics employed in the experiments. Granule Colour Form Height [mm] Mass [mg] Density [kg/m3]
PA Blue Cylindrical Ø 2.5 3.4 20 z1100
Fig. 4. Aspect of the PCePA granular material.
PC Orange Cylindrical Ø 3 3.6 25 z1200
The mathematical models of the responses PPC (goodness of fit R2 ¼ 98.3%, goodness of prediction Q2 ¼ 92.6%), PPA (R2 ¼ 97.1%, Q2 ¼ 88.0%), RPC (R2 ¼ 98.6%, Q2 ¼ 84.7%) and RPA (R2 ¼ 99.5%, Q2 ¼ 95.4%), computed with MODDE 5.0, were:
PPC ¼ 89.63 þ 2.3U1 þ 2.19U2 5.17U1 U2 2.75U1 U3 5.62U2 U3 13.43U22 þ 5.67U23
(2)
PPA ¼ 91.37 2.1U1 þ 2.1U2 þ 5.48U1 U2 8.35U1 U3 4.32U2 U3 7.96U21 7.28U22 þ 0.68U23
(3)
Fig. 6. Processed image of a PCePA granular mixture, showing only the PC granules.
W. Aksa et al. / Journal of Electrostatics 71 (2013) 681e688
685
Table 2 Mass of PC granules collected in boxes #1e7 (composite experimental design). U1 [kV]
U2 [kV]
U3 [kV]
24 48 24 48 24 48 24 48 24 48 36 36 36 36 36 36 36
24 24 32 32 24 24 32 32 28 28 24 32 28 28 28 28 28
32 32 32 32 24 24 24 24 28 28 28 28 32 24 28 28 28
Box # 1
2
3
4
5
6
7
8.19 6.17 8.66 9.53 5.92 7.18 7.46 11.18 6.39 7.35 5.51 9.09 8.93 9.07 6.61 11.08 8.03
10.16 9.54 7.8 8.63 7.4 7.89 9.17 12.5 6.85 9.62 8.42 7.34 10.01 10.85 9.48 11.62 9.12
19.45 16.75 12.28 14.65 12.3 12.93 13.36 16.55 10.32 13.6 14.85 10.1 13.28 14.53 14.53 16.93 13.93
15.81 12.44 12.00 11.5 13.94 11.55 13.54 13.25 12.34 15.05 14.33 10.51 12.51 14.37 15.15 15.6 15.4
22.4 21.38 20.3 18.72 20.33 18.24 18.58 17.58 18.74 18.23 20.47 23.97 17.58 19.02 18.7 17.77 16.94
15.56 19.59 16.94 18.66 19.33 20.16 16.63 16.58 18.73 15.04 18.26 17.71 16.83 15.23 17.43 13.14 14.84
12.61 13.00 20.18 15.67 20.19 20.1 21.23 10.52 25.66 18.46 16.59 20.00 19.1 16.00 14.8 11.81 17.12
RPC ¼ 84.58 1.06U1 1.74U2 1.14U1 U2 1.16U1 U3 1.25U2 U3 þ 1.62U21 4.17U23
(4)
RPA ¼ 86.21 4.85U1 0.81U2 þ 0.59U3 0.65U1 U2 þ 0.57U2 U3 þ 1.75U21 þ 1.19U23
(5)
According to these models, the purity of the separated materials may be as high as 96% for PC and 93% for PA (Figs. 7 and 8). The maximal value for the purity of PC is obtained at high U1 and moderate U2. At higher U2, the purity of the recovered product would be reduced as a consequence of the fact that the positivelycharged PA granules are deviated by the stronger impacts with the electrode of opposite polarity in the lower section of the installation. In the case of PA the best results are obtained at low U1 and high U3. The increase of U1 would intensify the impacts between the negatively-charged PC granules and the electrode of positive polarity in the upper section of the installation.
The lower purities for PA can be explained by the fact that the size and mass of each PC granule is smaller that of a PA granule and thus easier for the airflow to move them in the collecting zone of PA granules. In spite of the increased voltage applied to the electrodes, gravity and aerodynamic forces are stronger that the electrical ones and the granules do not have sufficient time to move out of that zone. They are collected and counted as impurity even if charged with the opposite polarity that the rest of the granules. The best recovery results 86.4% for PC and as high as 93.7% for PA are obtained at the lower values of U1, for reasons similar to those exposed for the purity. The voltages U2 and U3 applied to the electrodes in the lower section of the installation have lesser impact on the recovery of the two products. The predictions in Figs. 7 and 8 can be used for adjusting the values of the applied voltages such as to obtain the optimum performances in terms of purity and recovery. For instance, according to software MODDE 5.0, the higher purity and recovery of PC and PA are respectively: PPC ¼ 94.96%, PPA ¼ 94.57%, RPC ¼ 84.78% and RPA ¼ 94.9%, which correspond to
Table 3 Mass of PA granules collected in boxes #1e7 (composite experimental design). U1 [kV]
U2 [kV]
U3 [kV]
Box # 1
2
3
4
5
6
7
24 48 24 48 24 48 24 48 24 48 36 36 36 36 36 36 36
24 24 32 32 24 24 32 32 28 28 24 32 28 28 28 28 28
32 32 32 32 24 24 24 24 28 28 28 28 32 24 28 28 28
20.41 4.11 14.14 5.01 23.69 5.34 13.81 3.8 12.39 11.15 11.39 12.15 10.45 14.14 9.08 16.35 20.41
13.84 6.8 18.9 5.9 14.46 5.61 15.14 7.07 12.49 11.89 14.25 13.54 11.29 14.03 15.73 13.67 13.84
14.25 8.03 15.38 9.51 12.5 8.74 14.23 7.83 15.45 10.71 13.76 12.77 11.55 12.73 12.65 12.34 14.25
8.96 7.38 10.08 9.65 11.21 9.96 9.48 8.54 10.11 11.2 12.05 10.00 7.61 9.66 9.03 9.16 8.96
15.88 17.21 18.05 19.27 17.05 24.71 19.88 21.09 20.95 24.42 24.22 20.81 20.00 16.78 17.37 18.02 15.88
14.69 22.6 13.38 24.16 16.09 27.43 15.53 24.65 18.77 10.22 15.19 18.61 20.81 15.18 15.79 14.04 14.69
7.77 31.69 9.23 25.72 6.91 18.00 7.33 23.7 10.86 10.05 10.55 14.96 15.48 16.71 17.49 15.42 7.77
686
W. Aksa et al. / Journal of Electrostatics 71 (2013) 681e688
Table 4 Purity and recovery of PA and PC products collected in boxes #1e7 (composite experimental design). Voltage U1 [kV]
Voltage U2 [kV]
Voltage U3 [kV]
Purity PPA [%]
Recovery RPA [%]
Purity PPC [%]
Recovery RPC [%]
24 48 24 48 24 48 24 48 24 48 36 36 36 36 36 36 36
24 24 32 32 24 24 32 32 28 28 24 32 28 28 28 28 28
32 32 32 32 24 24 24 24 28 28 28 28 32 24 28 28 28
58.33 81.47 81.34 83.73 85.3 76.21 92.22 59.97 83.81 78.94 79.49 84.63 89.01 91.04 94.42 93.73 94.13
94.41 85.55 93.05 80.94 94.05 85.81 94.38 84.13 92.69 83.64 86.74 85.95 86.83 88.38 85.86 85.81 86.19
62.72 83.29 89.14 89.24 80.95 90.71 85.1 73.95 86.36 90.17 74.02 76.16 93.24 95.16 90.65 90.25 92.45
82.39 84.11 83.23 81.35 86.6 84.63 83.37 75.88 86.63 85.57 86.85 83.36 80.72 79.89 84.36 84.82 85.02
the following optimal values of the voltages applied to the electrodes: U1 ¼ 24 kV, U2 ¼ 27.96 kV and U3 ¼ 24.36 kV. An experiment performed with these values of the high-voltages conducted
to the following results: PPC ¼ 95.57%, PPA ¼ 94.88%, RPC ¼ 86.53% and RPA ¼ 94.3%, which are very close to the prediction made by MODDE 5.0.
Fig. 7. MODDE 5.0 predicted equal-purity [%] and equal-recovery [%] contours for the two granular products: PC (at U3 ¼ 24 kV) and PA (at U2 ¼ 28 kV).
W. Aksa et al. / Journal of Electrostatics 71 (2013) 681e688
687
Fig. 8. MODDE-predicted purity and recovery of PC and PA as function of the three applied voltages. The upper and lower curve on each graph indicates the limits of the 95% confidence level interval.
5. Conclusions The experiments described in this paper prove the effectiveness of a novel electrostatic separation process for the mixed granular plastics: 1) In order to ameliorate the purity and recovery of plastic materials recuperated from waste electrical and electronic equipments, the products of the first tribo-aero-electrostatic separation can be submitted to a second sorting, in the electrical field of two free-fall electrostatic separators. The tribo-aero-electrostatic separator and the two electrostatic separators of freefall represent respectively the upper and lower sections of the same installation. 2) The techniques of image processing and analysis facilitate the determination of the purity and recovery of the separated products.
3) The experimental design methodology enables the modeling and the optimization of the installation. 4) Higher voltages (32 kV) applied to the upper section of the separator may increase the purity of the PC product, but decrease that of the PA product. Lower voltages (24 kV) are recommended when the goal is to improve the recovery of both products. 5) Increasing the voltages applied to the lower section of the separator do not necessarily improve the performances, because the stronger impacts between the particles and the electrodes may drive them into a wrong direction. However, both the purity and the recovery of the products are quite robust to the changes in voltages applied to this section. References [1] F.O. Ongondo, I.D. Williams, T.J. Cherrett, How are WEEE doing? A global review of the management of electrical and electronic wastes, Waste Manag. 31 (2011) 714e730.
688
W. Aksa et al. / Journal of Electrostatics 71 (2013) 681e688
[2] H.I. Kang, J.M. Schoenung, Electronic waste recycling: a review of U.S. infrastructure and technology options, Resour. Conserv. Recycl. 45 (2005) 368e400. [3] UNEP, The Great E-Waste Recycling Debate, UN Environmental Program, 2007. [4] M. Schluep, Sustainable Innovation and Technology Transfer Industrial Sector Studies. Recycling e from E-waste to resources, UN Environmental Programme and UN University, Bonn, 2009. [5] H. Aizawa, H. Yoshida, S. Sakai, Current results and future perspectives for Japanese recycling of home electrical appliances, Resour. Conserv. Recycl. 52 (2008) 1399e1410. [6] S.M. Al-Salem, P. Lettieri, J. Baeyens, Recycling and recovery routes of plastic solid waste (PSW): a review, Waste Manag. 29 (2005) 2625e2643. [7] O.C. Ralston, Electrostatic Separation of Mixed Granular Solids, Elsevier, Amsterdam, 1961. [8] A. Tilmatine, S. Bendimered, F. Boukhoulda, K. Medles, L. Dascalescu, Electrostatic separators of particles. Application to plastic/metal, metal/metal plastic/plastic mixtures, Waste Manage. 29 (2009) 228e232. [9] I.I. Inculet, G.S.P. Castle, J.D. Brown, Electrostatic separation of plastics for recycling, Part. Sci. Technol. 16 (1998) 91e100. [10] L. Dascalescu, A. Samuila, A. Iuga, R. Köhnlechner, Electrostatic separation technologies for waste electric and electronic equipment (WEEE) recycling. Entrepreneurial opportunities, Environ. Eng. Sustain. Dev. Entrep. 1 (2012) 5e11.
[11] G.S.P. Castle, Contact charging between insulators, J. Electrostat. 40&41 (1997) 13e18. [12] A. Iuga, L. Calin, V. Neamtu, A. Mihalcioiu, L. Dascalescu, Tribocharging of plastics granulates in a fluidized bed device, J. Electrostat. 63 (2005) 937e942. [13] L. Calin, A. Mihalcioiu, A. Iuga, L. Dascalescu, Fluidized bed device for plastic granules triboelectrification, Part. Sci. Technol. 25 (2007) 205e211. [14] L. Calin, L. Caliap, V. Neamtu, R. Morar, A. Iuga, A. Samuila, L. Dascalescu, Tribocharging of granular plastic mixtures in view of electrostatic separation, IEEE Trans. Ind. Appl. 44 (2008) 1045e1051. [15] M. Bilici, L. Dascalescu, C. Dragan, O. Fati, A. Iuga, A. Samuila, Tribocharging and electrostatic separation of mixed granular solids in fluidized bed devices, IEEE Trans. Ind. Appl. 18 (2011) 1476e1483. [16] M. Bilici, L. Dascalescu, T. György, V. Barna, F. Rahou, A. Samuila, Experimental modeling of the tribo-aero-electrostatic separation of mixed granular plastics, in: Conf. Rec. IEEE-IAS Ann. Meet., 2011. [17] N.L. Frigon, D. Mathews, Practical Guide to Experimental Design, Wiley, New York, 1996. [18] L. Eriksson, E. Johansson, N. Kettaneh-Wold, C. Wikström, S. Wold, Design of Experiments. Principles and Applications, Learnways AB, Stockholm, 2000. [19] Umetrics AB, MODDE 5.0. User Guide and Tutorial, Umetrics, Umea, Sweden, 1999.
Contents lists available at SciVerse ScienceDirect
Journal of Electrostatics journal homepage: www.elsevier.com/locate/elstat
Two stage electrostatic separator for the recycling of plastics from waste electrical and electronic equipment Wessim Aksa a, Karim Medles a, Mohamed Rezoug a, Mohamed Fodil Boukhoulda a, Mihai Bilici b, Lucian Dascalescu b, * a b
Electrostatics and High Voltage Research Unit, IRECOM, University Djillali Liabes, 22000 Sidi-Bel-Abbes, Algeria Institut P’, CNRS-University of Poitiers-ENSMA, IUT, 4, avenue de Varsovie, Angoulême 16021, France
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 August 2012 Received in revised form 3 November 2012 Accepted 8 March 2013 Available online 1 April 2013
The aim of study was to evaluate the effectiveness of a new facility for recycling of plastics from granular waste electrical and electronic equipment. The installation consists of two sections, the products of a first tribo-aero-electrostatic separator being subsequently treated in two free-fall electrostatic separators. The tests were performed on a mixture of polycarbonate (PC) and polyamide (PA). Analysis of the purity of the products obtained was performed using a program of image processing in MATLAB. Products of very high purity (roughly 95% for both PC and PA) were obtained at a recovery rate higher than 70%. Ó 2013 Elsevier B.V. All rights reserved.
Keywords: Electrostatic processes Image processing Triboelectricity Recycling Waste recovery
1. Introduction Computers, printers, mobile phones, and other such appliances have shorter and shorter lifetimes [1], due to the very fast progress in electronics and information technology (IT). They represent an increasingly larger part of the waste electrical and electronic equipments (WEEE), the volume of which has reached an alarming level, especially during the last twenty years, when the markets have been saturated with huge quantities of new products of this type [2,3]. According to the recent statistics, the quantity of WEEE increased by 25% in five years, with the proportion of plastics amplified by 30% in the same period [1,4e6]. This situation has drawn the attention of both governmental and non-governmental on the necessity of developing effective methods for the recycling of WEEE. The electrostatic separation methods [7e9] have already proved to be a very effective solution for the recycling of insulating materials contained in this kind of waste. This non-pollutant
* Corresponding author. Tel.: þ33 545673245. E-mail addresses: [email protected], [email protected] (L. Dascalescu). 0304-3886/$ e see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.elstat.2013.03.009
technology is characterized by low energy consummation, as well as by reduced costs of operation and maintenance [10]. The tribo-electricity, a physical phenomenon involving the charge transfer between two bodies in contact [11e16] is the main charging mechanism employed for the separation of granular insulating materials in an intense electric field. The aim of the present work is to validate a new tribo-electrostatic separation process that has been designed for increasing the purity of the plastics recovered from WEEE. 2. Experimental set-up The installation is composed of two superposed, detachable electrostatic separators, attached to a same vertical support (Fig. 1). The upper section of the set-up is a tribo-aero-electrostatic separator that consists in a parallelepiped enclosure (height: 500 mm; width: 130 mm; depth: 110 mm), having two transparent walls in order to permit the visualization of phenomena, and two opaque lateral walls that have aluminum plate electrodes glued on their internal surfaces. These electrodes are typically energized from two adjustable DC high voltage supplies of opposite polarities 50 kV, to create an electrical field sufficiently strong to control the trajectories of charged granules.
682
W. Aksa et al. / Journal of Electrostatics 71 (2013) 681e688
Fig. 1. Schematic representation of the new double-stage electrostatic separator for mixed granular plastics.
Granule charging is produced by tribo-electric effect in the fluidized bed generated in the interior of this enclosure (Fig. 1). The fluidization air is furnished by a variable-speed blower. The uniformity of the fluidization bed is ensured by a custom-designed air diffuser, which is a finely-perforated plate situated at the bottom of the upper section of the installation. The granules are introduced in the separation enclosure by a funnel supplied by a fully-adjustable vibratory feeder. Under the combined action of the gravitational, aerodynamic, and electrical forces, they separate essentially in function of the polarity of their charge and exit the first separation stage through two slots that direct them to the lower section of the experimental set-up (Fig. 2). This section is composed of two free-fall electrostatic separators, the electrodes of which are aluminum plates (520 mm 100 mm) glued to four insulating PMMA boards (650 mm 110 mm). The upper edges of these four boards are positioned at the exit of the tribo-aero-electrostatic separator. The two vertical central plaques are fixed and connected to the earth, while the two exterior electrodes are connected to high voltage supplies of opposite polarities and can rotate to form angles ranging from 0 to 45 with respect to the vertical. The separated products are recovered in two identical
collectors, each subdivided in twenty compartments (length: 100 mm; width: 30 mm; depth: 85 mm). The 20 boxes of the two collectors are numbered as shown in Fig. 3. 3. Materials and method The experiments have been carried out with granules of polycarbonate (PC) and polyamide (PA) (Table 1). The mechanical separation of these granules is impossible, because they have similar shapes and mass densities. The analysis of the purity of the separated products is facilitated by the fact that the granules have different colors (Figs. 4e6). A set of 100 granules were weighted and the results were divided by 100 to give the weight of one granule. The high-voltage U1 applied to the electrodes of the upper section was adjusted at various values ranging between 24 and 48 kV. The electrodes of the two free-fall electrostatic separators of the lower section were energized at voltages U2 and U3 of opposite polarities and similar absolute values in the range 24e32 kV. These values were established after several preliminary experiments. Thus, it was observed that separation is very poor at voltages lower than 24 kV (the electric field strength is less than 2 kV/cm, not
W. Aksa et al. / Journal of Electrostatics 71 (2013) 681e688
enough to drive the granules to the electrodes). On the other hand, corona and spark discharges may occur from the edges of the plate electrodes at voltages higher than 48 kV (for the upper section) and 32 kV (for the lower section). The other variables of systems were maintained constant: the speed of the air blower: 13,000 rev/min; the relative humidity of the air RH ¼ 41 1%; the room temperature T ¼ 18 1 C; the mass and composition of the granular product samples m ¼ 100 g (PC) þ 100 g (PA) ¼ 200 g; the feed-rate: D ¼ 1 g/s. The movable electrodes of the inferior floor have been inclined at a ¼ 32 , their inferior edges correspond to the seventh compartment of each collector. The experimental modeling of the separation process has been performed using the response surface methodology [17,18], which recommends the use of a composite factorial experimental design and the adoption of a quadratic model. For the factors considered hereafter, namely the high-voltages U1, U2 and U3 applied respectively to the upper section electrodes, and to the two pairs of electrodes in the lower section of the experimental setup, the model y of the response (i.e. the mass m and in some cases the purity p of the material collected in the third to the seven compartment), will take the following form:
y ¼ a0 þ a1U1 þ a2U2 þ a3U3 þ a12U1U2 þ a13U1U3 þ a23U2U3 þ a11U21 þ a22U22 þ a33U23
683
(1)
The experimental data were analyzed with MODDE 5.0 software (Umetrics, Sweden) [19], which calculates the coefficients of the mathematical model, draws the response contours and identifies the best adjustments of the parameters for optimizing the process. Moreover, the program calculates two statistical criteria: the goodness of fit: R2, and the goodness of prediction: Q2. The latter is a measure of how well the model will predict the responses for new experimental conditions. In the case of a good mathematical the criteria R2 and Q2 have numerical values close to the unit.
4. Results and discussion The results of the seventeen experiments carried out according to the composite factorial design are reported in Tables 2 and 3. For each experiment, the overall purity of the granular materials collected in boxes #3 to 7 was analyzed with the image processing program previously-described and the results are given in Table 4.
Fig. 2. Photograph of the experimental set-up.
684
W. Aksa et al. / Journal of Electrostatics 71 (2013) 681e688
Fig. 5. Processed image of a PCePA granular mixture, showing only the PA granules.
Fig. 3. Schematic representation of the electrode system, energized from three highvoltage supplies U1 (positive polarity), U2 (positive), U3 (negative), and of the two 20-box collectors. The negatively-charged PC granules and positively-charged PA granules are mainly collected in the left side and right side collectors, respectively.
Table 1 Characteristics of the granules of plastics employed in the experiments. Granule Colour Form Height [mm] Mass [mg] Density [kg/m3]
PA Blue Cylindrical Ø 2.5 3.4 20 z1100
Fig. 4. Aspect of the PCePA granular material.
PC Orange Cylindrical Ø 3 3.6 25 z1200
The mathematical models of the responses PPC (goodness of fit R2 ¼ 98.3%, goodness of prediction Q2 ¼ 92.6%), PPA (R2 ¼ 97.1%, Q2 ¼ 88.0%), RPC (R2 ¼ 98.6%, Q2 ¼ 84.7%) and RPA (R2 ¼ 99.5%, Q2 ¼ 95.4%), computed with MODDE 5.0, were:
PPC ¼ 89.63 þ 2.3U1 þ 2.19U2 5.17U1 U2 2.75U1 U3 5.62U2 U3 13.43U22 þ 5.67U23
(2)
PPA ¼ 91.37 2.1U1 þ 2.1U2 þ 5.48U1 U2 8.35U1 U3 4.32U2 U3 7.96U21 7.28U22 þ 0.68U23
(3)
Fig. 6. Processed image of a PCePA granular mixture, showing only the PC granules.
W. Aksa et al. / Journal of Electrostatics 71 (2013) 681e688
685
Table 2 Mass of PC granules collected in boxes #1e7 (composite experimental design). U1 [kV]
U2 [kV]
U3 [kV]
24 48 24 48 24 48 24 48 24 48 36 36 36 36 36 36 36
24 24 32 32 24 24 32 32 28 28 24 32 28 28 28 28 28
32 32 32 32 24 24 24 24 28 28 28 28 32 24 28 28 28
Box # 1
2
3
4
5
6
7
8.19 6.17 8.66 9.53 5.92 7.18 7.46 11.18 6.39 7.35 5.51 9.09 8.93 9.07 6.61 11.08 8.03
10.16 9.54 7.8 8.63 7.4 7.89 9.17 12.5 6.85 9.62 8.42 7.34 10.01 10.85 9.48 11.62 9.12
19.45 16.75 12.28 14.65 12.3 12.93 13.36 16.55 10.32 13.6 14.85 10.1 13.28 14.53 14.53 16.93 13.93
15.81 12.44 12.00 11.5 13.94 11.55 13.54 13.25 12.34 15.05 14.33 10.51 12.51 14.37 15.15 15.6 15.4
22.4 21.38 20.3 18.72 20.33 18.24 18.58 17.58 18.74 18.23 20.47 23.97 17.58 19.02 18.7 17.77 16.94
15.56 19.59 16.94 18.66 19.33 20.16 16.63 16.58 18.73 15.04 18.26 17.71 16.83 15.23 17.43 13.14 14.84
12.61 13.00 20.18 15.67 20.19 20.1 21.23 10.52 25.66 18.46 16.59 20.00 19.1 16.00 14.8 11.81 17.12
RPC ¼ 84.58 1.06U1 1.74U2 1.14U1 U2 1.16U1 U3 1.25U2 U3 þ 1.62U21 4.17U23
(4)
RPA ¼ 86.21 4.85U1 0.81U2 þ 0.59U3 0.65U1 U2 þ 0.57U2 U3 þ 1.75U21 þ 1.19U23
(5)
According to these models, the purity of the separated materials may be as high as 96% for PC and 93% for PA (Figs. 7 and 8). The maximal value for the purity of PC is obtained at high U1 and moderate U2. At higher U2, the purity of the recovered product would be reduced as a consequence of the fact that the positivelycharged PA granules are deviated by the stronger impacts with the electrode of opposite polarity in the lower section of the installation. In the case of PA the best results are obtained at low U1 and high U3. The increase of U1 would intensify the impacts between the negatively-charged PC granules and the electrode of positive polarity in the upper section of the installation.
The lower purities for PA can be explained by the fact that the size and mass of each PC granule is smaller that of a PA granule and thus easier for the airflow to move them in the collecting zone of PA granules. In spite of the increased voltage applied to the electrodes, gravity and aerodynamic forces are stronger that the electrical ones and the granules do not have sufficient time to move out of that zone. They are collected and counted as impurity even if charged with the opposite polarity that the rest of the granules. The best recovery results 86.4% for PC and as high as 93.7% for PA are obtained at the lower values of U1, for reasons similar to those exposed for the purity. The voltages U2 and U3 applied to the electrodes in the lower section of the installation have lesser impact on the recovery of the two products. The predictions in Figs. 7 and 8 can be used for adjusting the values of the applied voltages such as to obtain the optimum performances in terms of purity and recovery. For instance, according to software MODDE 5.0, the higher purity and recovery of PC and PA are respectively: PPC ¼ 94.96%, PPA ¼ 94.57%, RPC ¼ 84.78% and RPA ¼ 94.9%, which correspond to
Table 3 Mass of PA granules collected in boxes #1e7 (composite experimental design). U1 [kV]
U2 [kV]
U3 [kV]
Box # 1
2
3
4
5
6
7
24 48 24 48 24 48 24 48 24 48 36 36 36 36 36 36 36
24 24 32 32 24 24 32 32 28 28 24 32 28 28 28 28 28
32 32 32 32 24 24 24 24 28 28 28 28 32 24 28 28 28
20.41 4.11 14.14 5.01 23.69 5.34 13.81 3.8 12.39 11.15 11.39 12.15 10.45 14.14 9.08 16.35 20.41
13.84 6.8 18.9 5.9 14.46 5.61 15.14 7.07 12.49 11.89 14.25 13.54 11.29 14.03 15.73 13.67 13.84
14.25 8.03 15.38 9.51 12.5 8.74 14.23 7.83 15.45 10.71 13.76 12.77 11.55 12.73 12.65 12.34 14.25
8.96 7.38 10.08 9.65 11.21 9.96 9.48 8.54 10.11 11.2 12.05 10.00 7.61 9.66 9.03 9.16 8.96
15.88 17.21 18.05 19.27 17.05 24.71 19.88 21.09 20.95 24.42 24.22 20.81 20.00 16.78 17.37 18.02 15.88
14.69 22.6 13.38 24.16 16.09 27.43 15.53 24.65 18.77 10.22 15.19 18.61 20.81 15.18 15.79 14.04 14.69
7.77 31.69 9.23 25.72 6.91 18.00 7.33 23.7 10.86 10.05 10.55 14.96 15.48 16.71 17.49 15.42 7.77
686
W. Aksa et al. / Journal of Electrostatics 71 (2013) 681e688
Table 4 Purity and recovery of PA and PC products collected in boxes #1e7 (composite experimental design). Voltage U1 [kV]
Voltage U2 [kV]
Voltage U3 [kV]
Purity PPA [%]
Recovery RPA [%]
Purity PPC [%]
Recovery RPC [%]
24 48 24 48 24 48 24 48 24 48 36 36 36 36 36 36 36
24 24 32 32 24 24 32 32 28 28 24 32 28 28 28 28 28
32 32 32 32 24 24 24 24 28 28 28 28 32 24 28 28 28
58.33 81.47 81.34 83.73 85.3 76.21 92.22 59.97 83.81 78.94 79.49 84.63 89.01 91.04 94.42 93.73 94.13
94.41 85.55 93.05 80.94 94.05 85.81 94.38 84.13 92.69 83.64 86.74 85.95 86.83 88.38 85.86 85.81 86.19
62.72 83.29 89.14 89.24 80.95 90.71 85.1 73.95 86.36 90.17 74.02 76.16 93.24 95.16 90.65 90.25 92.45
82.39 84.11 83.23 81.35 86.6 84.63 83.37 75.88 86.63 85.57 86.85 83.36 80.72 79.89 84.36 84.82 85.02
the following optimal values of the voltages applied to the electrodes: U1 ¼ 24 kV, U2 ¼ 27.96 kV and U3 ¼ 24.36 kV. An experiment performed with these values of the high-voltages conducted
to the following results: PPC ¼ 95.57%, PPA ¼ 94.88%, RPC ¼ 86.53% and RPA ¼ 94.3%, which are very close to the prediction made by MODDE 5.0.
Fig. 7. MODDE 5.0 predicted equal-purity [%] and equal-recovery [%] contours for the two granular products: PC (at U3 ¼ 24 kV) and PA (at U2 ¼ 28 kV).
W. Aksa et al. / Journal of Electrostatics 71 (2013) 681e688
687
Fig. 8. MODDE-predicted purity and recovery of PC and PA as function of the three applied voltages. The upper and lower curve on each graph indicates the limits of the 95% confidence level interval.
5. Conclusions The experiments described in this paper prove the effectiveness of a novel electrostatic separation process for the mixed granular plastics: 1) In order to ameliorate the purity and recovery of plastic materials recuperated from waste electrical and electronic equipments, the products of the first tribo-aero-electrostatic separation can be submitted to a second sorting, in the electrical field of two free-fall electrostatic separators. The tribo-aero-electrostatic separator and the two electrostatic separators of freefall represent respectively the upper and lower sections of the same installation. 2) The techniques of image processing and analysis facilitate the determination of the purity and recovery of the separated products.
3) The experimental design methodology enables the modeling and the optimization of the installation. 4) Higher voltages (32 kV) applied to the upper section of the separator may increase the purity of the PC product, but decrease that of the PA product. Lower voltages (24 kV) are recommended when the goal is to improve the recovery of both products. 5) Increasing the voltages applied to the lower section of the separator do not necessarily improve the performances, because the stronger impacts between the particles and the electrodes may drive them into a wrong direction. However, both the purity and the recovery of the products are quite robust to the changes in voltages applied to this section. References [1] F.O. Ongondo, I.D. Williams, T.J. Cherrett, How are WEEE doing? A global review of the management of electrical and electronic wastes, Waste Manag. 31 (2011) 714e730.
688
W. Aksa et al. / Journal of Electrostatics 71 (2013) 681e688
[2] H.I. Kang, J.M. Schoenung, Electronic waste recycling: a review of U.S. infrastructure and technology options, Resour. Conserv. Recycl. 45 (2005) 368e400. [3] UNEP, The Great E-Waste Recycling Debate, UN Environmental Program, 2007. [4] M. Schluep, Sustainable Innovation and Technology Transfer Industrial Sector Studies. Recycling e from E-waste to resources, UN Environmental Programme and UN University, Bonn, 2009. [5] H. Aizawa, H. Yoshida, S. Sakai, Current results and future perspectives for Japanese recycling of home electrical appliances, Resour. Conserv. Recycl. 52 (2008) 1399e1410. [6] S.M. Al-Salem, P. Lettieri, J. Baeyens, Recycling and recovery routes of plastic solid waste (PSW): a review, Waste Manag. 29 (2005) 2625e2643. [7] O.C. Ralston, Electrostatic Separation of Mixed Granular Solids, Elsevier, Amsterdam, 1961. [8] A. Tilmatine, S. Bendimered, F. Boukhoulda, K. Medles, L. Dascalescu, Electrostatic separators of particles. Application to plastic/metal, metal/metal plastic/plastic mixtures, Waste Manage. 29 (2009) 228e232. [9] I.I. Inculet, G.S.P. Castle, J.D. Brown, Electrostatic separation of plastics for recycling, Part. Sci. Technol. 16 (1998) 91e100. [10] L. Dascalescu, A. Samuila, A. Iuga, R. Köhnlechner, Electrostatic separation technologies for waste electric and electronic equipment (WEEE) recycling. Entrepreneurial opportunities, Environ. Eng. Sustain. Dev. Entrep. 1 (2012) 5e11.
[11] G.S.P. Castle, Contact charging between insulators, J. Electrostat. 40&41 (1997) 13e18. [12] A. Iuga, L. Calin, V. Neamtu, A. Mihalcioiu, L. Dascalescu, Tribocharging of plastics granulates in a fluidized bed device, J. Electrostat. 63 (2005) 937e942. [13] L. Calin, A. Mihalcioiu, A. Iuga, L. Dascalescu, Fluidized bed device for plastic granules triboelectrification, Part. Sci. Technol. 25 (2007) 205e211. [14] L. Calin, L. Caliap, V. Neamtu, R. Morar, A. Iuga, A. Samuila, L. Dascalescu, Tribocharging of granular plastic mixtures in view of electrostatic separation, IEEE Trans. Ind. Appl. 44 (2008) 1045e1051. [15] M. Bilici, L. Dascalescu, C. Dragan, O. Fati, A. Iuga, A. Samuila, Tribocharging and electrostatic separation of mixed granular solids in fluidized bed devices, IEEE Trans. Ind. Appl. 18 (2011) 1476e1483. [16] M. Bilici, L. Dascalescu, T. György, V. Barna, F. Rahou, A. Samuila, Experimental modeling of the tribo-aero-electrostatic separation of mixed granular plastics, in: Conf. Rec. IEEE-IAS Ann. Meet., 2011. [17] N.L. Frigon, D. Mathews, Practical Guide to Experimental Design, Wiley, New York, 1996. [18] L. Eriksson, E. Johansson, N. Kettaneh-Wold, C. Wikström, S. Wold, Design of Experiments. Principles and Applications, Learnways AB, Stockholm, 2000. [19] Umetrics AB, MODDE 5.0. User Guide and Tutorial, Umetrics, Umea, Sweden, 1999.